The Cheshire Basin: basin evolution, fluid movement and mineral resources in a Permo-Triassic rift setting

By J A Plant D G Jones G Warrington R A Chadwick A E Milodowski R Metcalfe

Bibliographical reference: Plant, J A, Jones, D G, and Haslam, H W (editors). 1999. The Cheshire Basin: Basin evolution, fluid movement and mineral resources in a Permo-Triassic rift setting. (Keyworth, Nottingham: the British Geological Survey.)

British Geological Survey

The Cheshire Basin: basin evolution, fluid movement and mineral resources in a Permo-Triassic rift setting

Edited by J A Plant, D G Jones and H W Haslam

Keyworth, Nottingham: British Geological Survey 1999. © NERC copyright 1999 First published 1999. ISBN 0 85272 333 4

• Editors
• J A Plant D G Jones H W Haslam British Geological Survey, Keyworth
• Authors
• J A Plant D G Jones G Warrington R A Chadwick A E Milodowski R Metcalfe British Geological Survey
• G Wadge G Ferrier Environmental Systems Science Centre (ESSC), Reading
• D F Ball J Bulat D J Evans I N Gale H W Haslam P J Hooker E K Hyslop N S Jones S J Kemp G A Kirby M J Leng A C Morton R A Nicholson D J Noy C A Rochelle W J Rowley T J Shepherd N J P Smith B Spiro G E Strong A S D Walker S R Young British Geological Survey
• M J Arthur A D Bradley B Birch J J W Higgo C McDermott D Savage A A Wilson Formerly British Geological Survey

Printed in the UK for the British Geological Survey by Nuffield Press, Abingdon C3 12/99

Maps in this book use topography based on Ordnance Survey mapping.

Datasets relating to this study are available under licence and may be obtained from the Manager, Analytical and Regional Geochemistry Group, British Geological Survey, Keyworth, Nottingham NG12 5GG, Tel. 0115 936 3100.

(Front cover) Cover photograph: Depth contours to the base of the Permo-Triassic.

(Rear cover)

Foreword

The work embodied in this book, a resource analysis of the Cheshire Basin, is based on a multidisciplinary study of basin processes. It is built on earlier multidisciplinary studies by BGS, such as those of the Lake District and North Wales and, in particular, studies to develop metallogenic models and exploration criteria for carbonate-hosted ore deposits in the Pennines and for gold.

Since 1835, the British Geological Survey has undertaken systematic geological mapping and curation of data for the UK. More recently, the development of geophysical, geochemical, hydrogeological and many other types of surveys, and the acquisition of data from companies, has meant that the activities and responsibilities of the Survey have expanded greatly, particularly since it established the National Geoscience Information Service (NGIS). It is increasingly possible to take a multidisciplinary approach to regional studies, and the potential is further increased by developments in computer manipulation of information. Data are now captured and stored in a series of relational databases, structured so that they can be analysed interactively, for example by use of Image Analysis, Geographical Information Systems or modelling packages. The BGS is at the forefront of such developments in the preparation and manipulation of systematic geoscience datasets for a wide range of economic and environmental applications.

The present study is based on analysis and modelling of systematic data covering the Cheshire Basin of north-west England, and includes work to develop an expert system for resource analysis for Cu-Pb-Zn-Ba, halite, hydrocarbons and water in this and other sedimentary basins. The study incorporates advanced basin analysis and hydrogeological and hydrochemical modelling with the results of geological, sedimentological, geophysical (seismic and potential field), geochemical (major and trace elements and istopes), mineralogical and petrological studies, including heavy-mineral and fluid-inclusion studies.

The Cheshire Basin was selected for study as one of a series of sedimentary basins formed in the complex Permo-Triassic rift systems which cut north–south across Britain and its continental shelf. These systems are important reservoirs of oil and gas, water and, in some places, halite, potash and geothermal resources. The Cheshire Basin contains large resources of halite which, in the nineteenth century, were important in establishing the chemical industry in Britain. It is also an important source of potable water, and is increasingly a target for gas exploration. It contains base-metal mineralisation of sedimentary-copper type, formerly worked at the Alderley Edge and other mines.

The study provides an analysis not only of the resource potential of the Cheshire Basin but also of the processes involved in deposit formations, which are relevant to the evaluation of comparable basins onshore and offshore in the UK and elsewhere. It has particular relevance to the Irish Sea, a major hydrocarbons province. It also provides new information on the structural and stratigraphical evolution of north-west England during the Permo-Triassic and hence makes a contribution to an understanding of the geological history of the British Isles.

The multidisciplinary study of the Cheshire Basin took place when my predecessor, Dr Peter Cook, was Director. He took a close interest in the project, and I am pleased to see the results brought to publication.

David A Falvey, PhD Director

Aknowledgements

(Figure 13)) for lithological key. From Wilson (1993)." data-name="images/P1000266.jpg">(Figure 12), (Figure 14), (Figure 15), (Figure 17), (Figure 18), (Figure 19) and (Figure 21) and (Plate 3) are reproduced by permission of the Yorkshire Geological Society. Thanks are due to the following for permission to publish borehole logs: British Salt Limited for (Figure 14), (Figure 16) and (Figure 18); British Gypsum plc for (Figure 20) and (Figure 21), and North West Water for (Figure 13. H C Ivimey-Cook contributed to (Figure 22) and (Figure 24) and to the associated text.

Permission to use commercial seismic reflection data and to display interpretations thereof was obtained from the following: AmBrit Resources Limited, Amerada Hess Limited, Amoco (UK) Exploration Company, BP Exploration Operating Company Limited, British Gas Exploration and Production, Cairn Energy Onshore Limited, CONCORD Gas Limited, Edinburgh Oil & Gas plc, Enterprise Oil plc, Fina Exploration Limited, Hamilton Oil Company Limited, Kelt UK Limited, Mobil North Sea Limited, Premier Consolidated Oilfields plc, Shell UK Exploration and Production, Sovereign Oil and Gas plc, Trafalgar House, Triton Resources UK, Tullow Exploration Limited and Union Texas Petroleum Limited. This is gratefully acknowledged.

The authors thank the NRA (now part of the Environment Agency) for donating core samples to the BGS and for providing access to, and allowing sampling of, cores held by the NRA; particular thanks are due to M Thewsey (NRA Northwest) and P R Stewart (NRA Severn-Trent) for their assistance. Cores from the A556 (M56–M6) road improvement scheme were provided by Exploration Associates

Limited with the approval of Allott & Lomax acting on behalf of the Department of Transport. Samples were collected from borehole core from the A49 (Weaverham to Lower Whitley) road improvement scheme by permission of Soil Mechanics Limited. The extra-basinal pebbles were mostly selected from a collection held at the Lapworth Museum, Birmingham. The authors thank the many landowners in Shropshire and Cheshire who permitted access for the collection of hydrocarbon samples, including Mrs Colthurst of Pitchford Hall, Tarmac Roadstone (for Haughmond Hill), and ECC Grinshill Stone Quarries Limited; John Kunka supplied the oil sample from Hem Heath Colliery; D B Thompson supplied a hydrocarbon-bearing sample from Alderley Edge. The authors thank P V R Sorensen for permitting access to West Mine, Alderley Edge; Derbyshire Caving Club for facilitating access to other mines at Alderley Edge; Shropshire Caving and Mining Club for facilitating access to Clive Mine; and ICI Chemicals and Polymers Limited for a visit to Winsford salt mine. D E Highley contributed to the section on industrial minerals, and R C Jones carried out the factor analysis.

The authors thank M D Eggboro and M Thewsey of Northwest NRA, and S W Fletcher of the Severn-Trent NRA, for facilitating access to records held by the NRA.

The authors thank Professor R Gurney for his support of the work at NESSC. P Pan contributed at different stages in this work.

Abstract

The Cheshire Basin is part of a complex north–south Permo-Triassic rift system, bounded by major, syndepositional normal faults. It marks a rapidly subsiding segment of the rift, forming an asymmetrical half-graben with a preserved fill of over 4500 m of Permo-Triassic red beds and, locally, Jurassic marine strata.

The lower part of the Triassic basin fill includes the Sherwood Sandstone Group (SSG), deposited in aeolian and fluvial environments, which, offshore, is an important hydrocarbon reservoir. This is overlain by the Mercia Mudstone Group (MMG), which accumulated in playa and tidal-flat environments and which contains two major halite formations and units rich in anhydrite and gypsum. The SSG is host to widespread baryte mineralisation and a more localised polymetallic assemblage, containing Cu with lesser amounts of Pb, Co, V, Zn, Mn, Ni and As and traces of Au.

The basin formed during east–west extension by reactivation of Caledonian-Variscan basement structures. It was initiated in Permian times, with the main phase of extension and subsidence during the early Triassic. This was followed by thermal relaxation subsidence and further episodic extension from the late Triassic to the early Cretaceous. Thermal relaxation subsidence probably continued into late Cretaceous times, the basin reaching its maximum depth in end-Cretaceous or early Cenozoic times, with more than 6000 m of sediment in the basin depocentre. Subsequent uplift and basin inversion in the Cenozoic led to the removal of more than 2000 m of strata from the depocentre, including most of the Jurassic marine sequence.

Temperatures at the base of the Permo-Triassic reached about 70°C in late Triassic times, increased further during the Jurassic and Cretaceous and exceeded 100°C in earliest Cenozoic times. Because of subsequent uplift and erosion, present-day temperatures exceed 60°C only in the deepest parts of the basin.

The SSG was laid down within a long northward-draining fluvial system. Whole-rock petrography and geochemistry, along with heavy-mineral and isotopic studies, indicate the interplay of distant and local sources of sediment, which evolved through time. Significant differences between the SSG of the Cheshire Basin and other UK Permo-Triassic basins suggest dilution of southerly sourced sediment by more locally derived material along the palaeo-drainage system.

The source terrain included igneous, sedimentary and metamorphic rocks, with probable inputs from Cornubia and Armorica mixed with detritus from Wales, the Welsh Borders, Anglo-Brabant massif and the Pennines.

The differences in heavy-mineral contents of fluvial and aeolian sandstones indicate that the aeolian sediment was not reworked from southerly derived fluvial material but was probably introduced from the Pennine landmass to the east.

Eodiagenesis in the SSG was dominated by the formation of micro-nodular non-ferroan dolomite, iron-oxide grain coatings, and hematite and anatase replacement of detrital ferromagnesian minerals and Ti-Fe oxides. Authigenic smectite or corrensite also formed as grain coatings and replacements of unstable detrital phases. Sandstone porosity was reduced during early mesodiagenesis by compaction and pressure-solution, and the precipitation of quartz and K-feldspar (± albite) overgrowths and cements and non-ferroan dolomite cements. Well-sorted aeolian sands with low contents of detrital chert and lithics were least affected by compaction.

Major anhydrite or other evaporite mineral cements probably affected much of the SSG, but have been largely removed near the surface by meteoric groundwaters. They formed by remobilisation from the overlying MMG during early burial. The complex calcite-baryte-Cu-Ag-Pb-As-Co-Hg-Ni-Se-V-Zn-Mn-sulphide mineralisation replaced the evaporite cements. It formed from low-temperature (<80°C) brines of generally low salinity (<5 wt% NaCl equivalents) during baryte precipitation, but higher during sulphide deposition (up to 18 wt% NaCl).

Hydrochemical modelling helped constrain diagenetic and ore-forming processes. Oxidised red-bed diagenetic fluids were simulated and mixed with more reducing fluids, for example derived from Carboniferous basement. The predicted diagenetic mineral assemblages are reasonably consistent with observed assemblages. The model basement fluid contained much less Cu and Pb, and slightly more Zn, than the red-bed fluid. Small degrees of mixing between the two fluids are critical in controlling heavy-metal mobility: less than 5% basement fluid strips all the Cu from the red-bed fluid. Sorption may have been an important process controlling the distribution of metals.

Modelling of fluid flow demonstrates that the distribution of evaporites is critical. A model for fresh groundwater predicts topographically-driven flow, with recharge in the south-east and discharge to the west and north. However, if the dissolution of evaporites is simulated, the high-density brines flow downwards in the basin centre and outwards towards the margins.

A four-stage metallogenic model is proposed for the Cheshire Basin ore deposits involving:

1. Remobilisation of metals by breakdown of primary minerals and incorporation into or sorption onto hematite, clays and carbonates during eodiagenesis.
2. Scavenging of metals, mainly from mudstones in the upper part of the basin fill, associated with the flow of density-driven brines from the MMG through the SSG towards basin-bounding faults.
3. Faulting near the basin margin, enabling the influx of small quantities of reducing fluid from the basement, causing precipitation of the ore assemblage.
4. Alteration of the ore assemblage due to meteoric groundwater influx following basin inversion in the Cenozoic.

It seems most likely that the known mineralisation in the Cheshire Basin formed during late Triassic (post MMG) to early Jurassic extension.

The strongly faulted eastern and north-eastern parts of the basin appear to have provided fluid-flow conditions favourable to ore formation. The major north-trending normal faults may have created ore-fluid migration pathways by structurally juxtaposing the SSG and the MMG, but it is possible that sub-vertical east–west transfer faults formed more effective conduits to fluids rising from the underlying Carboniferous rocks.

Prospectivity for mineralisation is considered to be lower in the extreme south-west of the basin, where faults trending north-east may have been less suitable for fluid flow. In the north-west of the basin, isotopic evidence suggests a greater influence of reducing fluids derived from underlying organic-rich Carboniferous basement. Carbon isotopic signatures of diagenetic carbonates appear to distinguish areas of potential Cu mineralisation from zones of dominantly Fe sulphides.

The Cheshire Basin holds significant resources of industrial minerals. In addition to glacial sand, boulder clay and building stone, there are vast resources of halite in the MMG, notably in the area of the Prees Syncline.

The sandstones of the SSG and the Permian Collyhurst Sandstone Formation are potential hydrocarbon reservoirs. They have considerable secondary porosity due to dissolution of early diagenetic cements. The MMG halites and mudstones are excellent seals. Potential source rocks include Dinantian and early Namurian shales and Westphalian oil-shales, cannel coals and bituminous coals. These rocks were buried during Carboniferous basin formation and also during Permo-Triassic and later events; thermal modelling suggests that both phases could have generated hydrocarbons. Consideration of thermal gradients, thermal modelling, burial history and possible migration pathways enables several structural leads to be identified in the Helsby Sandstone and Collyhurst Sandstone formations.

The main aquifer units in the Cheshire Basin belong to the SSG and the underlying Permian sandstones. The MMG forms a confining layer above the aquifers.

An expert system, linked to a GIS, provided the capability of analysing several types of spatial data. The design included five modules: mineralogy-petrology-diagenesis; provenance; structure; fluid flow; and mineralisation. The mineralogy-petrology-diagenesis module and the provenance module were taken to an advanced stage of development.

Summary

The results of a multi-disciplinary study of the Cheshire Basin of north-west England are presented. The aim of the project was an integrated study of the basin history with particular reference to the formation of resources. Data were captured in digital form and analysed using software developed by BGS as well as proprietary software. Work to develop an expert system integrated with a GIS was also carried out.

The Cheshire Basin, in common with other Permo-Triassic basins in Britain and its continental shelf, developed during a period of east–west regional crustal extension associated with development of the Arctic–North Atlantic rift system to the north and the Tethys–Central Atlantic–Gulf of Mexico rift-wrench system to the south. The basin lies within a complex north–south Permo-Triassic rift system which extends for more than 400 km from the English Channel Basin in the south, to the East Irish Sea Basin and beyond in the north. The basins of the rift system are bounded by major, syndepositional normal faults, which controlled basin development. In Permo-Triassic times the Cheshire Basin marked a particularly rapidly subsiding segment of the rift system, forming an asymmetrical half-graben with a preserved fill of over 4500 m of Permo-Triassic red beds and, locally, Jurassic marine strata. Offshore parts of the rift system contain important hydrocarbon reservoirs, and the Cheshire Basin itself has become a focus for hydrocarbon exploration (especially for gas). It also contains important halite and water resources, as well as polymetallic mineralisation of sediment-hosted copper deposit type (SCD). Such ore deposits presently supply approximately 10% of the world's copper and 30% of its cobalt production.

Palaeogeographical reconstruction suggests that, in common with other Permo-Triassic basins containing economic deposits of halite and SCDs, the Cheshire Basin lay in the equatorial zone, to the north of the Variscan mountain chain, during sedimentation, early diagenesis and ore formation. The lower part of the Triassic basin fill includes an arenaceous red-bed sequence, the Sherwood Sandstone Group (SSG), deposited in aeolian and fluvial environments, which, offshore, is an important hydrocarbon reservoir. This is overlain by the Mercia Mudstone Group (MMG), an argillaceous red-bed sequence which accumulated in playa and tidal-flat environments and which contains two major halite formations and units rich in anhydrite and gypsum. The SSG is host to widespread baryte mineralisation and a more localised polymetallic assemblage, containing Cu with lesser amounts of Pb, Co, V, Zn, Mn, Ni and As; traces of Au have been found at Alderley Edge. The polymetallic mineralisation occurs in traps controlled by faulting and an overlying impermeable cover of Mercia Mudstone, and has several of the characteristics of SCDs of continental red-bed type, as exemplified by the Dorchester deposit in New Brunswick, Dzhezkazgan in Kazakhstan and Nacimiento in New Mexico. It currently lacks an extensive overlying cover of reduced rocks, which limits the potential for the discovery of large-tonnage, high-grade Kupferschiefer-type SCDs.

The Cheshire Basin is flanked to the east and west by Carboniferous and Lower Palaeozoic rocks and its Permo-Triassic fill is unconformable on Upper Carboniferous strata, except in the extreme south-west where the basement comprises Lower Carboniferous and, lorglly, Lower Palaeozoic rocks. The basin is markedly asymmetrical, being deepest in the south-east, close to the basin-bounding Wem–Bridgemere–Red Rock Fault System (WBRRFS). In this area the basal Permo-Triassic is at depths greater than 4500 m, the basal SSG at depths greater than 3500 m, and the basal MMG at depths greater 1750 m. The cumulative normal throw on the WBRRFS presently exceeds 4000 m. In contrast, the western margin of the basin is relatively unfaulted, and characterised by depositional onlap

Depth-of-burial and backstripping studies indicate that the basin was initiated in Permian times, with the main phase of extension and subsidence during the early Triassic. This was followed by thermal relaxation subsidence and further episodic extension in the late Triassic, early Jurassic, late Jurassic and early Cretaceous. Thermal relaxation subsidence probably continued into late Cretaceous times, the basin reaching its maximum depth at about the end of the Cretaceous or early in Cenozoic times, with more than 6000 m of sediment indicated in the basin depocentre. Subsequent uplift and basin inversion in the Cenozoic led to the removal of more than 2000 m of strata from the depocentre, including most of the Jurassic marine sequence which could have provided a reducing seal for migrating ore fluids. It is suggested that basin inversion was accompanied by several hundred metres of oblique-reverse displacement along the WBRRFS.

Analysis of the major basin-controlling faults indicates that the basin formed by transtensional reactivation of Caledonian–Variscan basement structures. The smaller intra-basin faults show a lesser degree of basement control, the predominance of north–south-trending faults indicating an extension vector oriented 083°–263°, approximately east–west, consistent with regional fault geometries elsewhere in the graben system. Extension factors from fault restorations, based on fractal analysis of fault-size distribution, are typically in the range 1.10 to 1.15, broadly consistent with subsidence analysis.

Modelling of the basin's thermal history indicates temperatures of about 70°C at the base of the Permo-Triassic during the principal phase of basin subsidence, which culminated in late Triassic times. Temperatures in the basin increased further during Jurassic and Cretaceous times, the maximum subsurface temperatures being attained in earliest Cenozoic times when the deepest parts of the Permo-Triassic basin fill were well in excess of 100°C. Because of subsequent uplift and erosion, present-day temperatures exceed 60°C only in the deepest parts of the basin.

Geochemical, petrographic, mineralogical, fluid inclusion and stable isotope (δ¹³C_, δ¹⁸O_, δ³⁴S) investigations were used to assess the sedimentological, diagenetic and geochemical evolution, and the movement of palaeofluids in the basin with particular reference to the formation of SCDs and hydrocarbon accumulations.

The SSG was laid down within a long northward-draining fluvial basin system, originating in the Armorican Highlands to the south of Britain and developed along the Permo-Triassic rift system. It comprises quartz-arenite, quartz-sublitharenite and minor subarkosic sandstones and siltstones with thin beds of ferruginous silty mudstone. The MMG, transitional from the SSG at its base, passes upwards into a complex sequence of laminated gypsiferous, anhydritic and dolomitic mudstones and siltstones with fine-scale cycles of siltstone-dolomicrite-anhydrite, or anhydritic dolomicrite; it contains two thick halite formations.

The lithic clasts and heavy minerals seen in whole-rock petrographic studies of the early formations of the SSG indicate derivation from a complex source terrain, which included high-grade metamorphic rocks, syenite, acid volcanics, quartzite, sandstones and siltstones. Reworked monazite nodules resemble those described from Lower Palaeozoic sedimentary rocks in Wales, south-west England, Brittany and Belgium. Certain igneous clasts, and fossils contained in some pebbles, indicate Cornubian and Armorican sources.

The evidence suggests that the source terrain evolved through time. Albite became more important, and volcanic fragments less important, in the later formations of the SSG. The feldspar may reflect the increased importance of granitic inputs from Armorica or Cornubia, whilst the sedimentary clasts suggest more local derivation from the Anglo-Brabant Massif, Pennines or Wales.

In general, there is no gross variation in detrital mineralogy in the SSG across the Cheshire Basin. However, significant differences between the SSG of the Cheshire Basin and other UK Permo-Triassic basins suggest dilution of southerly sourced sediment by more locally-derived material along the palaeo-drainage system.

Detrital chert is prominent in the SSG of the Cheshire Basin, but it appears to be less significant in adjacent Permo-Triassic basins. It is probable, therefore, that the source was the abundant cherry horizons in nearby Dinantian limestones of North Wales, the Welsh Borders, the south Pennines and the Midlands (Anglo-Brabant Massif). This suggests that Namurian and Westphalian strata had already been at least partly eroded from these areas.

Variations in heavy-mineral assemblages and the chemistry of detrital tourmaline in the SSG and in earlier Permian sandstones also suggest an interplay between the major Armorican or Cornubian sourceland to the south of the basin and more local sources. The greatest influence of the southerly source took place during deposition of the Chester Pebble Beds Formation, coinciding with the time of greatest runoff from this source and the ensuing higher-energy depositional conditions. The heavy-mineral assemblages associated with these sediments are relatively rich in detrital monazite, and the tourmaline suites are dominated by Fe-rich varieties.

The differences in heavy-mineral contents of fluvial and aeolian sandstones indicate that the aeolian sediment was not reworked from southerly derived fluvial material during arid periods but was introduced from another region. In view of the easterly palaeowind direction, the most likely source was the Pennine landmass.

The Wilmslow Sandstone Formation is characterised by an apatite-poor heavy-mineral assemblage which overlies apatite-rich material. The most likely explanation for this is apatite dissolution during the interval represented by the Hardegsen Unconformity, at the boundary between the Wilmslow Sandstone and the Helsby Sandstone formations. Similar apatite depletion has been observed in the German Buntsandstein, deposited at a similar time and in similar facies. Such events are widespread in the Buntsandstein, and one of them may correlate with the zone of apatite depletion in the Cheshire Basin.

Sm-Nd isotope data are also generally consistent with the interplay of more distant southerly sources, from Cornubia or Armorica, and more local inputs from the Midlands, Welsh Borders or Wales. Many of the whole-rock samples have signatures characteristic of recycled material derived from Gondwana, like much of the Lower Palaeozoic sedimentary succession of Britain and, therefore, not very diagnostic of source.

The whole-rock geochemistry of the SSG indicates that the aeolian sediments (Kinnerton Sandstone) deposited prior to the first and most significant fluvial influx into the basin from the south (Chester Pebble Beds) differed in composition from later sediments, suggesting that the formations above the Chester Pebble Beds contained a certain amount of fluvially transported sediment derived from the south; evidence of this effect was observed in the heavy-mineral investigations.

Comparison of the aeolian facies below and above the Chester Pebble Beds (the Kinnerton and Wilmslow sandstones, respectively) is consistent with the addition of granitic material, presumably from a Cornubian or Armorican source, in the later formation. There are indications that this may decline in importance towards the north of the basin.

The middle part of the MMG is geochemically distinctive, with a composition suggesting a greater input from a basic igneous source relative to acid igneous or sedimentary components. The nearest source of basic material was probably the Carboniferous volcanics and intrusives of the south Pennines.

Eodiagenesis in the SSG was dominated by the formation of micro-nodular non-ferroan dolomite, iron-oxide grain coatings, and hematite and anatase replacement of detrital ferromagnesian minerals and Ti-Fe oxides. Authigenic smectite or corrensite also formed as grain coatings and replacements of unstable detrital phases. Sandstone porosity was reduced during early mesodiagenesis by compaction and pressure-solution, and the precipitation of quartz and K-feldspar (± albite) overgrowths and cements and non-ferroan dolomite cements. Well-sorted aeolian sands with low contents of detrital chert and lithics were least affected by compaction.

Major anhydrite or other evaporite mineral cements, which probably affected much of the SSG, now remain only in the deeper parts of the basin, having been largely removed nearer the surface by meteoric groundwater percolation.

Anhydrite and subsequent halite cements, formed by remobilisation from the overlying MMG during the early stages of burial, were introduced into the SSG along fractures. The complex calcite-baryte-Cu-Ag-Pb-As-Co-Hg-Ni-Se-V-Zn-Mn-sulphide mineralisation occurs adjacent to major fault structures and replaced earlier anhydrite or halite cements.

Isotope and fluid-inclusion studies provide constraints on the genesis of the ore deposits. Depositional temperatures for the gangue minerals associated with the copper ores support a low-temperature (<80°C) brine model for the red-bed Cu-Pb mineralisation. Fluid salinities were generally low (<5 wt% NaCl equivalents) during baryte precipitation but were higher and more variable during sulphide deposition (up to 18 wt% NaCl). Laser ablation ICP-MS analyses for individual fluid inclusions in the upper halite units of the MMG indicate the development of Na-Ca-Mg-Cl brines highly enriched in K, Ba and other ore-forming elements during the diagenetic evolution of the mudstone evaporitic sequence. Such fluids would have migrated down sequence and displaced earlier fluids within the SSG and thereby contributed to the overall metalliferous potential of the basin fill. Baryte δ³⁴S values of 15–20‰ CDT are similar to those of anhydrite and gypsum in the SSG and are consistent with Permo-Triassic sea water. There is a systematic variation in the δ¹³C signature of late-diagenetic carbonate veins regionally across the basin. This is indicative of an increased contribution of a carbon component from increasingly mature, organic-rich sediments towards the north-west. The δ¹⁸O of vein calcite (–13 to – 4‰) covaries with δ¹³C, also indicating an increase in the contribution of a warmer, deeper-derived fluid source, consistent with increasing organic maturation indicated by the δ¹³C signature towards the north-west.

The SSG has higher median Ba and Pb contents than the MMG, but lower concentrations of Co, Cr, Cu, Ni, V and Zn. Median Cu contents of the SSG are only 8% of average upper continental crust (ucc) values, whereas Zn contents are slightly higher at 11% of average ucc and Pb concentrations much greater at 45%. For the MMG, Cu values are 50%, Zn 75% and Pb 35% of average ucc contents.

It is not known whether the lower values in the SSG have a primary genetic cause, perhaps reflecting its predominantly aeolian derivation, or result from preferential removal of Cu and Zn by fluids passing through the basin.

The total contents of Cu, Pb and Zn in the basin, estimated from present-day concentrations and estimates of deposited rock thicknesses, are shown in the table below. An estimate for the total Cu extracted from the Alderley Edge mines is 3500 t, approximately 0.005% of this total.

If the basin fill originally contained average upper crustal levels of metals, then 173 Mt of Cu, 84 Mt of Pb and 481 Mt of Zn would have been removed from the SSG and 60 Mt of Cu, 63 Mt of Pb and 84 Mt of Zn from the MMG. On the other hand, if original contents were at average sandstone levels for the SSG, then 30–150 Mt of Cu, 8 Mt of Pb and 132–326 Mt of Zn would have been leached. For the MMG, original contents equivalent to average shale would have yielded 168 Mt of Cu, 78 Mt of Pb and 223 Mt of Zn. It is unlikely, however, that MMG metal contents would have been as high as average shale, so these figures can be regarded as upper limits. Hence, depending on the composition of the original basin fill, there could have been the potential to form world-class copper deposits associated with the basin, where the chemistry and transport conditions of ore-forming solutions and the depositional environment were favourable.

Geochemical and hydrochemical computer packages were used to model and constrain the processes controlling diagenesis and ore formation during basin evolution. SUPCRT92 was used to calculate equilibrium constants for pH and redox-buffering equilibria; EQ3/6 to model aqueous speciation and fluid mixing; HYDRAQL to model sorption; and SUTRA to simulate coupling between the evolution of fluid salinity and fluid flow. Oxidised red-bed diagenetic fluids (log fO_2bars = 0.7 to –40) were simulated for temperatures from 25°C to 125°C and mixed with more reducing fluids (log f0_2bars = –50), such as those that might be derived from Carboniferous basement rocks, at a temperature of 125°C. Red-bed diagenesis was modelled by interaction between a sediment with a composition approximating to bulk continental crust and Na-Ca-Cl solutions of up to 2 molal. The predicted eo- and mesodiagenetic mineral assemblages (mainly quartz and calcite, with smaller quantities of hematite, mica and clays) are reasonably consistent with observed assemblages. The composition of the basement fluid was constrained by equilibria involving typical mudrock minerals, and the pH and redox conditions were fixed by silicate and organic equilibria respectively. The model basement fluid contained orders of magnitude less Cu, and Pb and slightly larger amounts of Zn than the red-bed formation fluid. Small degrees of mixing between the red-bed and basement fluids are shown to be critical in controlling heavy-metal mobility: a mixture with less than 5% basement fluid is sufficient to strip all the Cu from the model red-bed formation fluid.

Modelling of the fluid flow in the basin, assuming Darcian flow in a two-dimensional porous medium, demonstrated that the distribution of evaporites is critical. A model based on fresh groundwater predicts that the fluid flow is topographically driven, with recharge zones in the south-east and discharge to the west and north; the MMG forms a relatively impermeable layer, and the underlying aquifers are under confined conditions. However, if the dissolution of evaporites, such as those of the MMG, is simulated, the flow of the high-density brines created is downwards in the centre of the basin (reversing the freshwater flow directions) and outwards towards the basin margins.

Sorption may have been an important process controlling the distribution of metals. Sorption, particularly to oxide surfaces, depends on pH and redox conditions, since both metal speciation and the nature of the surface depend on these variables. Copper can occur as Cu⁺ or Cu²⁺ whereas Pb and Zn occur only as divalent ions. Under reducing conditions Cu⁺, like Ag⁺, may be only weakly sorbed at neutral-to-acid pH. This could explain the preferential removal of Cu relative to Zn into ore fluids and, consequently, its much greater importance in the ores.

The results of the Cheshire Basin study provide new insights into the formation of its copper-lead-baryte deposits and probably SCDs generally, especially those of continental red-bed type. Geochemical, isotopic, petrological and modelling studies all suggest that one component of the ore fluid was a density-driven brine derived from bitterns in the MMG. Flow may have been initiated by compaction-driven dewatering of the MMG via the SSG during mesodiagenesis, and was associated with the development of an anhydrite or halite cement in the SSG throughout the basin. The metalliferous brines derived from the MMG may also have scavenged metals from the underlying SSG, as they migrated both downwards and laterally towards the basin-margin faults. The SSG is predominantly aeolian, however, and may have contained lower quantities of Cu and other metals on authigenic hematite, clays or calcite. These authigenic phases were, in part, formed by the alteration of detrital ferromagnesian and other primary minerals during eodiagenesis. Basins containing higher proportions of argillites or sediments rich in diagenetically altered ferromagnesian minerals may have higher metal contents. Supporting thermodynamic calculations demonstrate that the complex chemistry of the ores (Cu-Pb+Ag-As-Co-Hg-Mn-Ni-Se-V-Zn) and local variations in mineral parageneses can be explained by interaction of mesodiagenetic brines with poorly crystalline iron-oxide phases in red-bed mudstones under variable Eh-pH conditions.

It is suggested that the mineralisation in the north-east of the basin, including that at Alderley Edge, formed by the mixing of a small proportion (<5%) of a deep reducing fluid from the Carboniferous sequence beneath the Cheshire Basin with oxidising density-driven brines from evaporitic bitterns in the MMG, which remobilised and transported metals from the upper part of the basin fill (MMG and SSG). This is consistent with the overall low temperatures (<80^°C) indicated by fluid inclusions and the agreement between computed flow-path models for gravity-driven brine circulation below the MMG and the distribution of ore occurrences in the Cheshire Basin. Such a mixing model may also explain the spatial zonation of ore minerals in the known deposits, in which Cu mineralisation tends to extend further from faults than the Pb-Zn mineralisation. Additional features of the deposits which can be explained by the mixing model are: local bleaching of sandstones; quartz overgrowths; and baryte concentrations near faults. The complex metal-sulphide parageneses of the deposits possibly reflect chemical evolution of the red-bed formation fluids, owing to variations in redox conditions and sorption characteristics of iron oxides during diagenesis. The temporal and spatial distribution of sulphide minerals in the Cheshire Basin deposits is thus consistent with precipitation during an interplay between a chemically evolving red-bed formation fluid and mixing-controlled redox processes near to basin-margin faults. Best estimates for the timing of mineralisation suggest a late-Triassic to early-Jurassic event coincident with a major phase of extension. The present-day predominance of secondary (supergene) Cu-mineral assemblages is probably linked to Cenozoic uplift and the influx of low salinity, oxygenated groundwaters into the SSG.

The strongly faulted eastern and north-eastern parts of the basin appear to have provided fluid-flow conditions favourable for the development of ore bodies. The major north-trending normal faults may have created ore-fluid migration pathways by structurally juxtaposing the SSG and the MMG, but it is possible that the subvertical transfer faults were of greater importance. The latter structures trend roughly east–west, cutting and locally offsetting the larger north–south normal faults, and may have exerted an important influence on the development of ore bodies. Steeply dipping fractures probably formed more effective conduits to fluids rising from the underlying Carboniferous basement rocks than the more gently inclined normal faults, as the latter would tend to be closed by the weight of their hangingwall blocks.

Prospectivity for mineralisation of Alderley Edge type is considered to be lower in the extreme south-west of the basin, which is underlain by Lower Palaeozoic and Lower Carboniferous rocks and where faults trending north-east (basement Caledonide trend) may have been less suitable for fluid flow. In the north-west of the basin, isotopic evidence, indicative of an increased proportion of reducing fluids from organic-rich rocks, suggests a greater influence of fluids derived from underlying organic-rich Carboniferous basement.

Comparison of the distribution of authigenic Cu sulphides with the variation in δ¹³C of late mesodiagenetic carbonates contemporaneous with ore deposition indicates that areas with carbonates with δ¹³C PDB < –8‰ were dominated by the deeper-sourced reducing fluid which inhibited Cu mobilisation and mineralisation. The –8‰ δ¹³C PDB contour for the carbonates appears to provide a useful tool for delineating areas of potential Cu mineralisation, where values exceed –8‰, from zones of dominantly Fe sulphides, where δ¹³C PDB is less than −8‰.

A four-stage metallogenic model is proposed for the Cheshire Basin ore deposits involving:

1. Remobilisation of metals by breakdown of primary minerals and incorporation into or sorption onto hematite, clays and carbonates during eodiagenesis. Mudstones of the MMG are shown to provide an important source of Cu, Pb and Zn, but they form a relatively small proportion of the basin fill of the Cheshire Basin, which is predominantly arenaceous, aeolian-dominated SSG. Mass-balance calculations suggest the availability of some tens of millions of tonnes of Cu and Pb, and in excess of 100 million tonnes of Zn for mineralisation during this phase of basin development.
2. Scavenging of metals, mainly from mudstones in the upper part of the basin fill, associated with the flow of density-driven brines from the MMG through the SSG towards basin-bounding faults.
3. Faulting near the basin margin, which provided an efficient plumbing system for the influx of small quantities of a reducing fluid from the basement, causing precipitation of the ore assemblage.
4. Alteration of the ore assemblage as a result of the influx of meteoric groundwater following basin inversion in the Cenozoic.

One of the factors affecting the prospectivity of the basin is the timing of ore-forming events. The evidence presented here suggests that mineralisation cannot be simply related to the flow of diagenetic fluids in the context of progressive basin evolution but was related instead to later tectonism associated with reactivation of basin-bounding fault systems after cementation and lithification had occurred. Regional considerations, particularly evidence from more fully preserved sedimentary sequences to the south of the basin, indicate that extensional basin subsidence continued through early Jurassic times, with renewed extension in the late Jurassic and early Cretaceous associated with sea-floor spreading in the southern part of the North Atlantic region. It seems most likely that the known mineralisation in the Cheshire Basin occurred sometime during the late Triassic (post MMG times) to early Jurassic period of extension. Such a model has considerable implications for Cu exploration, especially for Kupferschiefer-type SCDs, in Permo-Triassic basins elsewhere in the UK.

The Cheshire Basin holds significant resources of industrial minerals. In addition to minor quantities of glacial sand, boulder clay and building stone (in the SSG), there are important resources of halite in the Northwich and Wilkesley Halite Formations of the MMG. The quantity of halite in Cheshire has been estimated at 117 km³ and forms a vast future resource, notably in the area of the Prees Syncline.

Potential hydrocarbon reservoirs in the Cheshire Basin exist in the Helsby Sandstone, Wilmslow Sandstone and Kinnerton Sandstone formations of the SSG, the Collyhurst Sandstone Formation in the underlying Permian and Westphalian and Namurian sandstones. Many of these Permo-Triassic sandstones have considerable secondary porosity due to dissolution of early diagenetic cements. The MMG halites and mudstones are excellent seals, where sufficiently thick. Potential hydrocarbon source rocks include Dinantian (Brigantian) basinal shales, early Namurian shales (Holywell Shales), Westphalian oil-shales and cannel coals (oil-prone) and bituminous coals (gas-prone), all of which outcrop on the periphery of the Cheshire Basin. These are well known source rocks in other oil provinces and some of them are likely to be present at depth beneath the Cheshire

Basin. These source rocks have undergone two phases of burial, firstly during Carboniferous basin formation and secondly during Permo-Triassic and later burial; thermal modelling suggests that both phases were sufficient to have resulted in hydrocarbon generation. Consideration of thermal gradients, thermal modelling, burial history and possible migration pathways enables several structural leads to be identified in the Helsby Sandstone, Collyhurst Sandstone and Carboniferous sandstone formations.

Hydrocarbons generated from Carboniferous times to the present day are likely to be trapped beneath the base Permian unconformity, by Ruabon (Etruria) Marl seals. Drilling to test late Namurian and early Westphalian reservoirs can be justified. However, wells sited on early Namurian and older rocks are not recommended because these rocks have unfavourable reservoir characteristics.

Although not a primary target, testing of the Helsby Sandstone should not be ignored. The mineralisation model requires a contribution from Carboniferous fluids and suggests that sealed traps may occur down-dip of mineralised structures in several areas:

1. On the block between the Wem and Hodnet faults.
2. Between the Helsby Sandstone outcrop at Bickerton and the Wem–Audlem Sub-basin.
3. Between the King Street Fault and Alderley Edge.
4. South-west of the Brook House Fault.

The main prospective areas for Carboniferous reservoirs are: near Wigan and in the subsurface around Runcorn; where basal Westphalian Sandstone unconformably overlies Dinantian rocks; on the possible extension of the Derbyshire High north-west from Buxton to the Alderley–Stockport area; and downdip of the Milton Green inlier and on the block between the Wem and Hodnet faults.

The main aquifer units in the Cheshire Basin belong to the SSG and the underlying Permian sandstones, principally the Collyhurst Sandstone Formation. The MMG forms a confining layer above the Triassic and Permian aquifers.

An expert system linked to the ARC/INFO GIS was designed for the Cheshire Basin project, to provide the capability of analysing several types of spatial data relating to sedimentary basins. The design includes five modules: mineralogy-petrology-diagenesis; provenance; structure; fluid flow; and mineralisation. The user is guided by data-selection menus based on criteria such as lithostratigraphy or borehole depth. The GIS performs spatial manipulation and display of information, whilst the knowledge base of the expert system, which encapsulates the expertise of the specialists involved, analyses the data.

The mineralogy-petrology-diagenesis module and the provenance module were taken to an advanced stage of development. They provide capabilities such as defining the 3-D spatial extent based on criteria such as the occurrence of a particular mineral or assemblage of minerals and mapping the uncertainty levels of provenance sources within specific horizons.

Chapter 1 Introduction

J A Plant and D G Jones

Background

Several types of mineral deposit, including red-bed copper, Mississipi Valley-type (MVT) lead–zinc, sedimentary exhalative (sedex) lead-zinc, and some uranium deposits, along with petroleum and natural gas, can be related to the tectonic and thermal evolution of sedimentary basins. The deposits form as a result of the flow of brines or hydrocarbons in response to hydrostatic gradients related to topographic relief, compaction, osmotic pumping, thermal gradients, and deformation and tectonism. Sedimentary basins are also host to resources of coal and evaporitic minerals (such as gypsum, halite and potash), and contain important groundwater resources and, in some basins, low-enthalpy geothermal resources. Moreover, in Britain and much of western Europe, the principal centres of population and industrial and agricultural activity are in areas underlain by major sedimentary basins.

Considerable progress has been made in recent years in basin analysis (e.g. McKenzie, 1978; Royden, 1986; Kuznir and Ziegler, 1992), modelling of the geochemical and fluid flow in basins (e.g. Bethke, 1985, 1986, 1989; Bethke et al., 1988; Harrison and Summa, 1991; Cathles and Smith, 1983), understanding the geochemistry and diagenesis of basin fill (e.g. Walker, 1967, 1989) and the formation of associated mineral deposits, especially base metals (Roberts and Sheahan, 1988; Boyle et al., 1989; Brown, 1993; Sangster, 1993) and hydrocarbons (Tissot and Welte, 1978; Demaison and Murris, 1984). The present study is aimed at applying these findings to understanding the formation of resources (Cu-Pb-Ba, halite and hydrocarbons) associated with the Cheshire Basin of north-west England, which contains mainly red-bed sandstones and, higher in the sequence, mudstones and evaporites. The Cheshire Basin was selected as being representative of the basins formed in the Permo-Triassic rift systems which cut north–south across the British Isles and the continental shelf and which host important resources (hydrocarbons, industrial minerals and water) onshore and offshore. The aim  was to develop an integrated approach to the resource analysis of red-bed sedimentary basins generally, based on analysis of the large range of datasets held by the BGS.

The study

The study used advanced basin analysis, and geochemical and fluid-flow modelling. The results were integrated with information on the geology, geophysics, stratigraphy and palaeogeography of the basin and on the geochemistry, diagenesis and heavy-mineral suites of the basin fill. The data were compiled in digital form and plotted and presented using software developed in BGS as well as proprietary software. Work to develop an expert system for interactive resource analysis of sedimentary basins using multidataset analysis by GIS, based on the Cheshire Basin study, was initiated.

It is hoped that this report will not only provide new methods for assessing the resources of red-bed sedimentary basins, but also stimulate fresh discussion on all aspects of the geology of Permo-Triassic rift system basins in Britain and elsewhere.

The Cheshire Basin

The Cheshire Basin is a major Permo-Triassic extensional basin within a complex north–south-trending rift system (Figure 1) which stretches for about 400 km from the English Channel Basin in the south to the East Irish Sea Basin and beyond in the north. Offshore, similar Permo-Triassic rift systems underlie the North Sea. The basins of the rift system are bounded by major, syndepositional normal faults, which controlled basin development. In Permo-Triassic times the Cheshire Basin marked a particularly rapidly subsiding segment of the rift system, and has a preserved Permo-Triassic sediment thickness of almost 4000 m. The Permo-Triassic rocks of the basin, particularly the Sherwood Sandstone Group, have been the focus of detailed investigation because of their importance as reservoirs for hydrocarbons, water and geothermal resources. The Cheshire Basin was selected for detailed study because it is one of the few onshore basins with some unexplored potential for hydrocarbons (especially for gas), it contains important water and halite resources, and, unlike many British Permo-Triassic basins, it contains Cu-Pb-Zn mineralisation. Modelling the formation of the ore deposits can provide considerable information on palaeo- and present-day fluid flow of importance for resource analysis of British Permo-Triassic basins generally.

The resources of the basin

The copper mineralisation and its global context

The mineralisation which occurs largely around the margins of the basin is of a sedimentary continental red-bed copper association. The most important occurrence is at Alderley Edge, where Cu, Pb, and Zn, with minor amounts of Ag, Co, V, Ni and Mn, occur as disseminations or in fault breccias, mainly in three conglomerate and sandstone units near the top of the Lower Triassic Sherwood Sandstone Group (Warrington, 1980). The mineral association is characteristic of the continental red-bed copper association (Ixer and Vaughan, 1982; Naylor et al., 1989), a sub-type of sediment-hosted stratiform copper deposits (SCDs) (Gustafson and Williams, 1981; Brown, 1993). Although few in number, SCDs such as White Pine, Michigan, the Kupferschiefer of Poland, and the Zambian Copper Belt, are among the world's most important sources of copper, recently accounting for 10% of the world's copper production, as well as 30% of its cobalt and increasing quantities of silver.

Most models for such ore deposits depend on a diagenetic brine-expulsion model (e.g. Gustafson and Williams, 1981; Brown, 1993; Naylor et al., 1989) whereby brines equilibrated with red-bed sequences, and carrying Cu, Pb, Zn and Ba as chloride complexes, deposit Cu and other metals on crossing a redox boundary, such as that provided by sulphide- bearing grey beds. Two sub-types of sedimentary copper deposits are distinguished by Eckstrand (1984), paralic marine (Kupferschiefer type) and continental red-bed type. In the former, anoxic marine rocks are associated with red beds, while in the latter the anoxic rocks are of fluvial or lacustrine origin. The Cheshire Basin deposits have the closest similarity to the latter group, examples of which are the Dorchester deposit in New Brunswick, Dzhezkazgan in Kazakhstan and Nacimiento in New Mexico. In both sub-types, copper is precipitated by reduction in anoxic sediments, diagenetic pyrite being a common reductant, whereas in the copper deposits of the Cheshire Basin it has been suggested that hydrocarbons, including methane, acted as the reducing agent responsible for precipitation of the ore (Warrington, 1980).

Such models derive from the classic work of Walker (1967, 1974, 1976, 1989) in the south-west USA and Mexico, showing that the continental red-beds are first-cycle, immature sediments deposited in oxidising conditions. They owe their red colour to the early diagenetic alteration of ferromagnesian minerals to hematite and other iron oxyhydroxides. The alteration is associated with the release of Cu and other elements held in primary minerals onto hematite and clays from which they can be mobilised to form ore deposits. The present study examines the exploration criteria and metallogenic models previously proposed for SCDs and suggests important modifications for the development of prospectivity and resource analysis criteria for such mineralisation in sedimentary basins generally.

Halite

The Permo-Triassic red beds of the UK are host to commercially valuable evaporites, including sylvite (Cleveland), anhydrite and gypsum (Cheshire, Cumbria and Nottinghamshire) and halite (Cleveland and Cheshire). In the Cheshire Basin, the resources of halite from the two major salt-bearing formations (the Northwich Halite and Wilkesley Halite formations) have been estimated to be of the order of 28 cubic miles (Pugh, 1960). Brine extraction, presently by controlled pumping in the Holford and Warmingham Brinefields, is from the Northwich Halite Formation; salt is also mined from the 'Bottom Bed' of this formation. A total of 5.5 Mt per year of salt is produced from the Cheshire Basin.

Future resources are vast, particularly in the virtually untapped area of the Prees Syncline, where both halite-bearing formations are present over large areas. There is also potential for formation of deep cavities for storage of gas or liquids in the deeper parts of the syncline.

The presence of halite in the Cheshire Basin is important not only as a resource, but also as a control on fluid flow and for metallogenic modelling.

Hydrocarbons

Red beds are important hydrocarbon reservoirs in many parts of the world. In the UK, Permo-Triassic rocks are important hydrocarbon reservoirs in the North Sea and the Wessex and East Irish Sea basins. Producing fields from Triassic reservoirs in basins adjacent to Cheshire, at Formby and in the East Irish Sea, together with oil shows in the Needwood Basin in the Midlands, indicate that both oil and gas have been generated over a wide area. In the East Midlands oilfields the source and the reservoirs are of Carboniferous age, and in this study we consider not only the prospectivity of the Permo-Triassic but also that of the Upper Carboniferous rocks beneath the Cheshire Basin.

Groundwater

Permo-Triassic red beds are major sources of potable water in Cheshire, Lancashire and Nottinghamshire, and are exploited as geothermal reservoirs in the Wessex Basin. Utilisation of groundwater resources has played an important role in the urban and industrial development of the Cheshire Basin. The Permo-Triassic sandstones are prolific aquifers, exploited largely where they are not confined by the overlying Mercia Mudstone Group and younger formations. Recharge to the aquifer is, in places, restricted by thick drift cover, and groundwater generally flows towards the major rivers draining the basin.

Beneath the large conurbations, groundwater has been over-exploited historically, leading to depression of the water table below sea level. This has resulted in saline intrusion, particularly along the Mersey Estuary and the Manchester Ship Canal. Reduction or cessation of pumping in the last few decades has led to rising groundwater levels, particularly beneath Liverpool, in part in response to the contribution made by leaking sewers and water mains.

The present-day shallow groundwater flow system overlies a deeper saline groundwater system, which appears to be driven by the density of brines derived from the overlying halite deposits in the central and south-eastern part of the basin.

Structure of the book

Chapter 2 of the volume gives an account of the stratigraphy of the Permian to Lower Jurassic basin fill, including sections on the regional setting, biostratigraphy and sedimentology. The basin structure and its evolution are discussed in Chapter 3, based on the interpretation of seismic reflection data calibrated by borehole information and on geophysical modelling of gravity and aeromagnetic data. This is followed in Chapter 4 by an account of the provenance of the basin fill, incorporating geochemical (including Nd/Sm isotope data) and mineralogical studies, especially of the heavy-mineral suite. In Chapter 5 the diagenesis of the basin is discussed, based on new petrographic, clay mineralogical, geochemical (organic and inorganic), isotope (sulphur, carbon and oxygen) and fluid-inclusion data. Hydrogeological and hydrogeochemical models consistent with the evolution of the Cheshire Basin and its diagenesis and geochemistry are then presented in Chapter 6. All these data are used to model the metalliferous mineral occurrences in the basin in Chapter 7, which also presents a resource analysis (for halite, hydrocarbons and water, as well as Cu-Pb-Zn-Ba) and includes a review of known resources, mining and extraction history, resource models, exploration criteria and prospectivity of the basin. The report concludes (Chapter 8) with an account of development work on an Expert System for resource analysis of red-bed sedimentary basins generally.

Sample collection

The solid geology of the Cheshire Basin is poorly exposed because of an extensive cover of glacial drift. 'Where outcrops do occur they are generally affected by surface weathering.. Because of these two factors nearly all of the samples studied for the project were obtained from boreholes or from underground mines.

There was little borehole core from the basin in the BGS collections at the start of the project. A search was therefore made for borehole core from external sources. Most of the core acquired for the project came from the National Rivers Authority (NRA), now part of the Environment Agency (EA) and comprised two sets of boreholes. One set, in the north of the basin, extending into the Wirral and the Lancashire coastal strip (in the area between the Cheshire Basin and the East Irish Sea Basin) came from the EA (North West Region), while the other, in the south of the basin, formed part of Shropshire groundwater investigations and was held by EA (Midlands Region). Further cores were obtained from site investigations for the A556 road improvement link to Manchester Airport (from Allott and Lomax) and the Weaverham bypass (Soil Mechanics) and from mineral exploration for evaporites. All of this material is currently held in the core storage facility of the BGS and it is intended that a representative set of borehole cores will be retained. Full use was made of existing core and samples held by BGS from stratigraphical, geothermal, water and site-investigation boreholes, as well as from exploration for coal, oil and salt resources.

Most of the boreholes sampled (see (Figure 2) and (Table 1)) were relatively shallow (<200 m) and many of them were subject to the effects of Cenozoic and modern weathering and shallow groundwater flow. Because groundwater abstraction or monitoring boreholes are heavily represented, there is a bias towards the permeable Permo-Triassic sandstone aquifer rocks (principally those of the Sherwood Sandstone Group (SSG)) around the margins of the Cheshire Basin.

Less borehole core was available from the Mercia Mudstone Group (MMG) and other formations in the centre of the basin, but MMG strata were sampled from BGS boreholes, such as Wilkesley and Crewe, and from exploration for evaporites and site-investigation boreholes. In addition to borehole material, the MMG (Northwich Halite) was sampled underground at the Meadowbank Rock Salt Mine, Winsford. Due to the lack of available material only a limited study was made of pre-SSG strata. The Collyhurst Sandstone, Manchester Marl and earlier strata were sampled from a small number of boreholes.

To supplement the borehole material a few surface outcrop samples were collected, principally from the south of the basin, and samples were taken from the old mine workings at Alderley Edge and Clive. The locations sampled are illustrated in (Figure 2) and details are given in (Table 2).

Samples from the cores, or outcrops, were selected to give as wide a geographical and stratigraphical coverage as possible. A standard sample-handling scheme was devised (Figure 3) to avoid conflict between the requirements of different parts of the project; wherever possible, reference material was retained prior to any destructive treatment, such as preparation of the rocks for chemical analysis, thin sections, physical properties or palynology. In total more than 850 samples were collected during the project, of which more than 450 were from the SSG and over 300 from the MMG.

Particulars of each sample were recorded in a computer database. The information included location (grid reference, borehole name, borehole reference number, depth etc.), stratigraphy, collector and a description of the sample.

Chapter 2 Stratigraphy and sedimentology

G Warrington, A A Wilson, N S Jones, S R Young and H W Haslam

Tectonic setting

Overviews of the tectonic development of Britain and neighbouring regions before and during the deposition of the Permo-Triassic sediments of the Cheshire Basin are given by, among others, Anderton et al. (1979), Cope et al. (1992), Coward (1990, 1993), Dewey (1982), Leeder (1982, 1988) and Soper et al. (1987). This section (on Tectonic setting) has been compiled from these accounts and from Chapter 3 of this volume.

Pre-Variscan and Variscan

During Precambrian and Palaeozoic times, the tectonics of north-west Europe were dominated by the sequential accretion of magmatic arcs and older continental fragments against and onto the stable North American craton.

In early to middle Devonian times, oblique collision between the Laurentian and Avalonian land masses produced regional compressional deformation over much of Britain, with widespread uplift, folding and basin inversion. These important Acadian tectonic events effectively marked the end of the Caledonian orogeny in Britain, by which time closure of the mid-European (Rheic) and Iapetus oceans had resulted in the Armorican, Laurentian, Avalonian and Baltican terranes being fused into a single land mass, Laurasia, with Britain close to its southern margin (Figure 4). Major Acadian structures were susceptible to later reactivation and played an important part in the subsequent structural development of the region.

Throughout much of north-west Europe and the Appalachians, the Variscan orogeny was a further episode in this long history of progressive accretion. Laurasia was separated from the continent of Gondwana to the south by a proto-Tethys Ocean. Closure of this ocean, combined with the buffering effects of various small terranes, had, by end-Carboniferous times, created the Variscides as a collision orogenic belt.

In Germany, the internal, or Moldanubian, zone of the Variscides (Figure 5) shows structures dominated by a gently dipping schistosity and gneissic banding, with overthrust sheets of eclogitic and granulitic material. The tectonic transport direction was towards the north-west. To the north-west of this zone, the Saxo-Thuringian and Rheno-Hercynian zones show lower-grade metamorphism but similarly north-west-directed shears and thrusts. Together they indicate several hundred kilometres of crustal shortening. In north-west France crustal shortening may have been considerably less, and in south-west England the shortening across Devon and Cornwall has been estimated at 150 km.

The Variscides of southern England form a continuation of the outer, or Rheno-Hercynian, zone of the European Variscan belt and can, in part, be interpreted in terms of thin-skinned thrust tectonics, with a thrust transport direction dominantly towards the north-west or north-north-west. There is, however, evidence of thick-skinned thrust tectonics beneath north Devon, South Wales and the Mendips, where the structures trend approximately east–west, oblique to the general north-west-directed movements in the thin-skinned zones.

The Devonian and Lower Carboniferous sedimentary sequence in south-west England reflects a marine transgression from the south. Terrestrial deposits were succeeded first by shallow-marine sediments and subsequently by deep-water shales, cherts and volcanic rocks. During Namurian and Westphalian times the Variscan thrust front advanced from the south while the flysch-like sediments of the Culm were deposited in north Devon and Cornwall.

To the north of the developing Variscan fold belt in southern Britain lay a foreland on which extensional basins developed in Carboniferous times on the northern side of a persistent Wales–London–Brabant high. Extension in latest Devonian and early Carboniferous times produced a series of grabens and half-grabens in the Caledonian basement, which exerted the dominant control on early Carboniferous sedimentation to the north of the high. In Namurian times, crustal extension progressively gave way to regional post-extensional subsidence, the early Carboniferous deposition of carbonate and detrital sediments on a block-and-basin topography being followed by the accumulation of thick, more uniform sequences of deltaic strata. Later in the Carboniferous, subsidence was replaced by uplift, as Variscan compressive forces became progressively more dominant. The main effects of this regional Variscan compression were the reversal of Caledonian lines of weakness and structural inversion of the Carboniferous basins. Minor inversion occurred sporadically in latest Dinantian and Namurian times, but the main period of basin inversion, roughly coeval with emplacement of the Variscan thrust front, took place in Westphalian and Stephanian times. The Variscan front now effectively forms the northern limit of the Variscan fold belt.

During Devonian times, a hot semi-arid climate prevailed over much of the North Atlantic area, Britain lying close to the equator. Though Britain remained near the equator throughout the Carboniferous and early Permian, the climate changed from arid in the earliest Dinantian to semi-arid and then monsoonal in later Dinantian times. Later in the Carboniferous, an increasingly humid tropical climate prevailed over much of north-west Europe, which ended in latest Westphalian to Stephanian times when the evidence of red beds and oxidised coal seams indicates the onset of the lengthy Permo-Triassic arid period.

Post-Variscan

During Permo-Triassic times, northern Europe lay within the arid hinterland of the Pangaean supercontinent, newly formed by the collision of Laurasia and Gondwana. A tectonic regime of regional crustal extension became established, many of the earlier Caledonian and Variscan structures being reactivated as basin-controlling normal faults. The Cheshire Basin formed part of a north–south rift system (see (Figure 1) and Chapter 3 of this volume). The climate was continental, with low rainfall, and the basin was filled with a sequence of continental sediments, mostly of aeolian, fluvial and playa-lake origin. Erosional areas lay to the east (Pennines) and the south-west and west (Wales). Marine incursions entered from the north-west in Late Permian and Mid and Late Triassic times, and from the south and west at the end of the Triassic.

The Cheshire Basin

The Cheshire Basin (Figure 6) and (Figure 7); Enclosure 1 is one of the largest British onshore post-Variscan rift basins. It is about 100 km long, with a maximum width of about 55 km, and covers an area of at least 3500 km². The Permo-Triassic fill rests unconformably on, or is in faulted contact with, folded Carboniferous and older rocks and is contiguous with the fill of the East Irish Sea Basin (EISB) to the north-west, and the Stafford, Needwood and Worcester basins to the south-east and south.

The basin fill comprises formations assigned to the Appleby, Cumbrian Coast, Sherwood Sandstone, Mercia Mudstone, Penarth and Lias groups (Figure 8). It consists of a dominantly arenaceous sequence (to the top of the Sherwood Sandstone Group; SSG) overlain by a largely argillaceous sequence with evaporites in the lower part (Mercia Mudstone Group; MMG) and limestones in the upper part (Penarth and Lias groups). Though the nature of the basin fill was recognised in the last century, its thickness was not proved by drilling until after 1970. A gravity survey (White, 1949) broadly defined the structure of the area as a half-graben, bounded to the east by a major fault. Kent (1949) proposed depths to the basin floor of more than 1500 m in the south, near Prees, and about 2000 m further north near Northwich. More recent geophysical studies provided a better definition of the structure of the basin and the form of its floor (Gale et al., 1984); these interpretations have been further refined by studies of seismic reflection data and the records of deep boreholes at Knutsford and Prees (Evans et al., 1993; and see Chapter 3 of this volume).

Glacial deposits are widespread in the Cheshire Basin and exposures of the basin fill are thus limited (Figure 9). The dominantly arenaceous lower part of the succession is better exposed than the overlying, largely argillaceous sequence. Surface exposures have been augmented by borehole sections in parts of the basin. The arenaceous lower part of the basin fill has been extensively drilled for water-supply purposes, largely in the peripheral areas in the north, west and south of the basin. Parts of the overlying largely argillaceous sequence have been drilled in connection with salt extraction in the Northwich–Winsford area and elsewhere. Additionally, information is available from boreholes drilled, largely in connection with the exploration of concealed coalfield prospects, around the northern and northwestern margins of the basin, and from boreholes drilled by the

Geological Survey, principally in the Stockport, Chester and Nantwich districts (geological sheets 98, 109, 122; (Figure 10). Several deep boreholes penetrating the full basin fill have been drilled in search of hydrocarbons (see below).

The basin floor

The form of the basin floor has been interpreted from geophysical data (Gale et al., 1984; and see (Figure 36)). Around the northern, western and southern margins of the basin the fill rests on folded and faulted pre-Permian rocks (Figure 7), which are mostly Carboniferous (up to Westphalian D) in age though Ordovician rocks occur at the south-western margin, south of Oswestry. The eastern and south-eastern margins are delineated by the Wem–Bridgemere–Red Rock Fault System.

Few boreholes have proved the entire succession of the basin fill or the nature of the basin floor beneath the deeper parts of the structure. In 1973, the Prees Borehole (Figure 7) and (Figure 11) (Trend Petroleum, [SJ 558 344]: Colter and Barr, 1975; Colter, 1978) proved reddened Carboniferous (Upper Silesian) shales at c.3610 m KB (metres below Kelly Bushing) with Lower Palaeozoic (Ordovician to Silurian) rocks at c.3850 m (Evans et al., 1993); the latter may be in faulted contact with the Carboniferous rocks (Wilson in James, 1983; Evans et al., 1993). Nearly 50 km to the north-north-east, Carboniferous shales, sandstones and coals of Westphalian age were proved from c.2820 m KB to TD (3045.7 m KB) in the Knutsford No. 1 Borehole (Figure 7) and (Figure 11) (Gas Council Exploration, [SJ 7027 7786]: Colter and Barr, 1975). A small inlier of Carboniferous rocks has been proved near Chester (Poole and Whiteman, 1966; Earp and Taylor, 1986); a borehole in this area (Milton Green: Esso Petroleum, [SJ 4374 5692]) (Figure 7) proved some 1500 m of Carboniferous (Dinantian and Silesian) rocks resting on Ordovician sediments and intrusive igneous rocks (Whittaker et al., 1985; Earp and Taylor, 1986; Penn, 1987).

A summary account of the distribution and structures of the pre-Permian rocks of the basin floor and in the surrounding area is given on pp.55–67.

The basin fill: overview of stratigraphical evolution

At the beginning of the Permian, Britain lay deep within the Pangaean supercontinent, in a belt of easterly winds and situated within a few degrees of the equator (Smith and Taylor, 1992). The area was dominated by desert erosion and deposition. Crustal tension, possibly associated with the onset of rifling in the North Atlantic area, led to local subsidence and the development of a series of fault-bounded basins, including the Cheshire Basin and the adjacent East Irish Sea and Stafford basins (Figure 6). Aeolian sands (Collyhurst Sandstone Formation and its correlatives) were deposited in all three basins, while at the centre of the East Irish Sea Basin (EISB) a saline lake developed. To the north-east and south-west, the Cheshire Basin was surrounded by bare, rugged hills. The sands were deposited on an uneven erosional surface of mainly Carboniferous rocks, and this surface was further modified by syndepositional faulting.

Early in the Late Permian, a marine incursion from the north-west into the northern part of the Cheshire Basin led to the deposition of the dolomitic and gypsiferous mudstones of the Manchester Marls Formation. In the southern part of the basin at this time, the Bold Formation represents a continuation of the conditions that obtained during the deposition of the Collyhurst Sandstone. Towards the end of the Permian, the sea retreated and the Bold and Manchester Marls formations were succeeded by the Kinnerton Sandstone. Marine conditions persisted for longer in the area of the East Irish Sea, with the deposition of the St Bees Shales and Evaporites. The Kinnerton Sandstone Formation (the basal formation of the SSG) consists dominantly of aeolian sands, though evidence of fluvial action suggests the presence of rivers in the interdune areas.

Similar conditions continued into the Triassic. Britain now lay about 15 to 20° north of the equator, in a monsoonal climate (Warrington and Ivimey-Cook, 1992). The Kinnerton Sandstone was succeeded by conglomerates and sandstones of the Chester Pebble Beds Formation. These sediments were deposited from a northward-flowing river system which entered the Cheshire Basin via the Stafford Basin. The abundance of pebbles decreases northwards; pebble beds extend into the south-eastern extremity of the EISB, beyond which lie fluvial sandstones (the St Bees Sandstone Formation) similar to the Wilmslow Sandstones which, in the Cheshire Basin, overlie the Chester Pebble Beds. The Wilmslow Sandstone Formation consists of sandstones, with some siltstones and mudstones, deposited from the north-west-flowing river system, and also features dunes formed by easterly winds.

The succeeding beds of the Bulkeley Hill Sandstone Formation are only locally present, and their deposition was followed by a period of faulting and erosion, now marked by the Hardegsen disconformity.

The Helsby Sandstone Formation (at the top of the SSG) and its correlative in the EISB, the Ormskirk Sandstone Formation, were deposited on the Hardegsen erosion surface, under the influence of easterly winds and rivers from the south-east.

The Tarporley Siltstone Formation, the lowest formation of the MMG, was laid down in intertidal and playa environments and marks a transition to less sandy deposits. It is followed by a sequence of formations dominated by dolomitic mudstone (Boffin, Byley, Wych, and Brooks Mill Mudstone formations) or halite (Northwich and Wilkesley Halite formations). The Cheshire Basin was periodically connected to the sea during this time, via the EISB and then south-westward through the Celtic Sea Basin. The land surface varied between dry, desert conditions, shallow temporary lakes, and seas or lakes of a more lasting nature in which thick halite units developed. Apart from infrequent fluvial sandstones, the detrital sediment was entirely fine grained, deposited as aeolian sediment (structureless facies; oxidised) or in shallow water (laminated facies; oxidised and reduced laminae). Groundwater throughout this time contained relatively high concentrations of magnesium, leading to the precipitation of dolomite and magnesium-bearing clays in the mudstones during early diagenesis. Nodules of gypsum also grew in the mudstones, close to the surface. Subsidence accompanied by movement along contemporary faults continued until Carnian times (Wilkesley Halite), when conditions became more uniform over a wider area (Warrington and Ivimey-Cook, 1992). The MMG in the EISB is broadly similar to that in the Cheshire Basin.

The Blue Anchor Formation, at the top of the MMG, closely resembles the laminated facies of the Brooks Mill Mudstone, being composed of wind-blown detritus deposited in shallow water, except that it consists almost entirely of reduced facies; terrestrial and/or freshwater life may have been more prolific, perhaps because of a more reliable rainfall.

A major marine incursion ensued. The Penarth Group consists of mudstones, fine sandstones and limestones with marine fossils, and this fauna became more diverse in the overlying Lias Group.

The basin fill: stratigraphical details

Stratigraphical nomenclature

The stratigraphical nomenclature used is that of Hull (1869), and Pugh (1960), as revised by Warrington et al. (1980) and further amended, in part, by the British Geological Survey (BGS Lexicon of Named Rock Units) and by Wilson (1993) (Figure 8); and see the 1:250 000 scale geological map, Enclosure 1).

Succession

In its most complete form, in the northern part of the basin, the fill comprises the Appleby, Cumbrian Coast, Sherwood Sandstone and Mercia Mudstone groups; to the south, younger deposits of the Penarth Group and the lower part of the Lias Group are also present (Figure 7) and (Figure 8).

Appleby Group

The Collyhurst Sandstone Formation (BGS Lexicon of Named Rock Units) outcrops around the northern and north-eastern margins of the basin, in the Wigan, Manchester, and Stockport districts (geological sheets 84, 85, 98; (Figure 10) (Tonks et al., 1931; Jones et al., 1938; Magraw, 1961; Taylor et al., 1963) where it rests unconformably upon Carboniferous rocks and is overlain conformably by the Manchester Marls Formation. Further south in the basin the formation becomes almost indistinguishable from the overlying Manchester Marls where the latter pass laterally into a more silty or arenaceous facies, as recognised in the Prees Borehole (Evans et al., 1993,. fig. 5) (Figure 7) and (Figure 11). Where these formations become indistinguishable, their correlatives are regarded as forming the lower part of the Kinnerton Sandstone Formation of the Sherwood Sandstone Group (Warrington et al., 1980; Meadows and Beach, 1993a, fig. 2) which outcrops on the western side of the basin in the Flint, Chester, Wrexham, Nantwich and Oswestry districts (geological sheets 108, 109, 121, 122, 137; (Figure 10).

The Collyhurst Sandstone is of continental, aeolian, origin. Palaeocurrent vectors indicate deposition under the influence of winds from an easterly direction, with the possible formation and migration of transverse dunes with barchanoid crests or, locally, of longitudinal seif dunes (Thompson, 1985, fig. 7). Further south, in the Stafford and Worcester basins (Figure 6), the lower part of the aeolian Bridgnorth Sandstone Formation is equivalent to the Collyhurst Sandstone. The Bridgnorth Sandstone is interpreted as having accumulated in a combination of transverse and barchanoid draas, with superimposed oblique crescentic and linear dunes, under the influence of fluctuating winds from a predominantly easterly direction (Karpeta, 1990). To the north-west, in the EISB, the Collyhurst Sandstone comprises mainly aeolian deposits (Jackson et al., 1995).

Changes in the thickness of the formation in the Manchester area have been interpreted as evidence of syndepositional faulting (Tonks et al., 1931). However, Poole and Whiteman (1955) considered that these variations were related largely to deposition on an uneven land surface, and that the apparent relationship to fault lines was fortuitous in all but one instance. Seismic reflection data (Figure 42) and (Figure 43) show, however, that the Collyhurst Sandstone does thicken across faults, supporting the views of Tonks et al. (1931) (see also Evans et al., 1994). In the Manchester area the formation is locally absent but ranges in thickness up to 261.8 m (Tonks et al., 1931; Poole and Whiteman, 1955); in the Stockport district the formation is up to 281 m thick (Taylor et al., 1963). Within the basin the formation is 557 m thick in the Knutsford Borehole and 515 m in the Prees Borehole (Evans et al., 1993) (Figure 11).

There is no direct evidence for the age of the Collyhurst Sandstone, but the youngest underlying beds are Late Carboniferous (Westphalian D) in age, and the overlying Manchester Marls Formation is Late Permian (Kazanian to Tatarian). The Collyhurst Sandstone is thus constrained only within the post-Westphalian to pre-Kazanian interval (Figure 8), though it is generally regarded as Early Permian (e.g. Smith et al., 1974).

Cumbrian Coast Group

This newly introduced group (BGS Lexicon of Named Rock Units) incorporates the Manchester Marls Formation, which is recognised around the northern and north-eastern margins of the basin (Figure 7), in the Wigan, Manchester and Stockport districts (geological sheets 84, 85, 98; (Figure 10) (Tonks et al., 1931; Jones et al., 1938; Taylor et al., 1963).

The Manchester Marls rest on the Collyhurst Sandstone or locally overlap that formation and rest unconformably on Carboniferous rocks (e.g. Tonks et al., 1931). Around Manchester the formation varies in thickness from 40.4 to 75.06 m (Tonks et al., 1931) and in the Stockport area it is up to 100 m thick (Taylor et al., 1963).

The Manchester Marls comprise red, rarely green, dolomitic and gypsiferous mudstones. In the lower part of the formation there are lenses and thin beds of calcareous mudstone, dolomitic limestone and iron2stained limestones, containing calcareous microfaunas and shelly macrofossils indicative of a marine depositional environment (Logan, 1967; Pattison, 1970; Pattison et al., 1973). Upwards, the formation becomes more arenaceous and passes into the Kinnerton Sandstone Formation of the SSG. A concealed development of the Manchester Marls extends south for at least 40 km from the outcrops at the northern end of the basin (Gale et al., 1984, fig. 7). This development is 148 in thick in the Knutsford Borehole (Evans et al., 1993) but it passes laterally into more arenaceous beds, and the concealed limits of the formation are, therefore, poorly defined.

A silty and arenaceous correlative 70 m thick is recognised nearly 50 km to the south-south-west in the Prees Borehole (Evans et al., 1993) (Figure 11); this arenaceous unit has been formalised as the Bold Formation (BGS Lexicon of Named Rock Units). Elsewhere the formation is indistinguishable from the underlying Collyhurst Sandstone and merges with the equivalent of that formation to form the lower part of an arenaceous sequence assigned to the Kinnerton Sandstone Formation (Warrington et al., 1980; Meadows and Beach, 1993a, fig. 2). An interbedded succession of aeolian sandsheet deposits and fluvial sheetflood facies in the Speke Reservoir Borehole (p.38) is attributed to the Bold Formation.

Regional studies show that the Manchester Marls accumulated marginally to a depocentre sited to the north-west, in the EISB. Towards that depocentre the formation passes laterally into concentric belts of dolomitic and anhydritic mudstones disposed around a body of halite (Colter and Barr, 1975; Jackson et al., 1987, fig. 7a, b) and is in continuity with the St Bees Evaporites and St Bees Shales further north (Jackson and Mulholland, 1993, fig. 3).

The age of the Manchester Marls is, on biostratigraphic evidence, Late Permian (Kazanian to Tatarian) (c.260–250 Ma, Forster and Warrington, 1985) (Figure 8).

Sherwood Sandstone Group

The formations of this group have an extensive outcrop in the Cheshire Basin, almost surrounding that of the Mercia Mudstone Group and younger deposits (Figure 7). The principal formations (Warrington et al., 1980) are the Kinnerton Sandstone Formation, Chester Pebble Beds Formation, Wilmslow Sandstone Formation and Helsby Sandstone Formation (Figure 8); in the western part of the basin, the Bulkeley Hill Sandstone Formation is present locally between the Wilmslow and the Helsby sandstones (Figure 8).

The upward increase in sand content in the Manchester Marls and the transition into the sandstones of the SSG reflect palaeogeographical changes that resulted in the exclusion of marine influences from the northern part of the Cheshire Basin and the neighbouring EISB. The Chester Pebble Beds, Wilmslow Sandstone and Bulkeley Hill Sandstone were deposited under continental conditions as parts of a dominantly fluvial facies association that spread northwards from southern England, through the Worcester, Stafford and Cheshire basins, into the EISB in Early Triassic times (Warrington and Ivimey-Cook, 1992, map Trla). The Bulkeley Hill Sandstone and Wilmslow Sandstone are progressively cut out at an unconformity at the base of the Helsby Sandstone (Warrington et al., 1980; Evans et al., 1993). This has been correlated with the Hardegsen disconformity (Trusheim, 1963) in the Middle Bunter of Germany (Warrington et al., 1980). Evidence of the regional nature of this unconformity, and of its presence within the concealed SSG in the Cheshire Basin, has been compiled from geophysical studies (Evans et al., 1993). Above the unconformity, the Helsby Sandstone reflects the re-establishment, in the early Mid-Triassic (Warrington and Ivimey-Cook, 1992, map Trlb), of a continental fluvial system which drained northwards into the EISB. Continental sedimentation in the EISB (Bushell, 1986; Stuart and Cowan, 1991; Cowan, 1993; Meadows and Beach, 1993a, b) was superseded by the deposition of fine-grained sediments with evaporites, which form the lowest part of the MMG and were deposited partly in water of marine origin. During Mid-Triassic times this environment, and the area of MMG deposition, expanded southwards to occupy the Cheshire Basin and other areas further south in England where the SSG had accumulated previously (Warrington, 1970a, b, 1974; Warrington and Ivimey-Cook, 1992, maps Trlb, Tr2).

Kinnerton Sandstone Formation

This formation is predominantly of continental, aeolian facies. Detailed petrographical examination (p.90) suggests that, in some instances, fluvial reworking of the aeolian sandstones occurred in an interdune environment. The Kinnerton Sandstone lacks fossils, but, because of its lateral equivalence to the aeolian Collyhurst Sandstone and the partly marine Manchester Marls, it is regarded as largely Permian in age (Figure 8). The thickness of the Kinnerton Sandstone increases southwards from 73 m in the Knutsford Borehole to 110 m in the Prees Borehole (Figure 7) and (Figure 11) (Evans et al., 1993); this trend parallels a thinning of the underlying Manchester Marls and may reflect the southward passage of the upper part of that formation into more arenaceous beds. Palaeowind vectors (Thompson, 1985, fig. 7) indicate that the Kinnerton Sandstone, like the Collyhurst Sandstone, was deposited under the influence of winds from an easterly direction. In the Chester district (geological sheet 109; (Figure 10) the upper part of the formation shows large-scale deformed cross-bedding which Thompson (1989) compared with structures described by Glennie and Buller (1983) from Permian aeolian deposits in the North Sea Basin. In the Chester district, 14 m of very micaceous sandstones, described by Thompson (1989) as at the 'predicted horizon' of the Manchester Marls, forms a transition upwards into the Chester Pebble Beds. These sandstones may correlate with the Bold Formation, a unit recognised by geologists of the North West Water Authority in the Runcorn district (geological sheet 97; (Figure 10) (Campbell et al., 1981; Thompson, 1989).

Chester Pebble Beds Formation

This formation outcrops principally on the western and north-western sides of the basin but is also recognised in smaller areas in the north-east; minor outcrops in the southeast are contiguous with more extensive outcrops in the adjoining northern part of the Stafford Basin. The Chester Pebble Beds Formation is recognised at depth in the Prees and Knutsford boreholes (Figure 7) and (Figure 11) (Colter and Barr, 1975; Colter, 1978; Penn, 1987; Evans et al., 1993), where it rests on the Kinnerton Sandstone and is overlain by the Wilmslow Sandstone (Figure 8) and (Figure 11). Elsewhere, it locally rests unconformably on the Bold Formation, Manchester Marls, Collyhurst Sandstone and Carboniferous rocks and, in the south-eastern part of the basin, is overlain unconformably by the Helsby Sandstone.

The Chester Pebble Beds comprise conglomerates, pebbly sandstones and sandstones (the Red Pebbly Sandstone Lithofacies of Thompson, 1970a) and are of continental origin. Horizontally bedded gravels are interpreted as representing longitudinal sheet bars, and flat- and cross-bedded gravel associations were deposited as mid-channel bars in confined braided channels with considerable bar–channel relief. These facies types are recognised principally in the south-east. Thick cross-bedded pebbly sandstones also occur there, and persist northwards through the basin; they represent deposition in transverse bars and dunes in channels of a braided river system. Interbedded finer sediments (argillaceous cross-bedded sandstones, siltstones and mudstones) reflect less turbulent braided river systems dominated by transverse bars and dunes (Steel and Thompson, 1983; Thompson, 1970a, 1985). Palaeocurrent vectors (Thompson, 1970a, 1985, fig. 8) indicate deposition from rivers that flowed north-westwards from the Stafford Basin, through the Cheshire Basin, into the EISB. The abundance of pebbles decreases in the downstream direction, and the formation is traceable only to the south-eastern periphery of the EISB (Colter and Barr, 1975; Colter, 1978; Jackson et al., 1987; Jackson and Mulholland, 1993). The formation is 175 m thick in the Prees Borehole and 166 m at Knutsford (Evans et al., 1993) (Figure 11). North-westwards, into the EISB, the formation shows a decrease in pebble content and becomes indistinguishable from sandstones of the overlying Wilmslow Sandstone Formation. Beyond the limit of the occurrence of pebbles in the Chester Pebble Beds, the lateral equivalents of that formation and the Wilmslow Sandstone occur in the St Bees Sandstone Formation (Jackson and Mulholland, 1993, fig. 3). Thompson (1989) considers that some 14 m of non-pebbly, argillaceous, very micaceous, cross- and planar-bedded sandstones that were mapped as the basal beds of the Chester Pebble Beds Formation in the Chester district (geological sheet 109; (Figure 10) are equivalent to the Bold Formation.

No stratigraphically useful fossils are known from the Chester Pebble Beds which, from their position above the Late Permian Manchester Marls are assessed as Early Triassic in age (Figure 8).

Wilmslow Sandstone Formation

This formation outcrops principally in the western and northern parts of the basin. It is recognised at depth in the Prees and Knutsford boreholes (Figure 7) and (Figure 11) (Colter and Barr, 1975; Colter, 1978; Penn, 1987; Evans et al., 1993). It rests on the Chester Pebble Beds; at outcrop in west Cheshire and within the basin it is succeeded by the Bulkeley Hill Sandstone Formation (Poole and Whiteman, 1966; Warrington et al., 1980; Evans et al., 1993). Elsewhere the Bulkeley Hill Sandstone is overlapped by the Helsby Sandstone which comes to rest unconformably on, or completely cuts out, the Wilmslow Sandstone.

The Wilmslow Sandstone comprises predominantly red, fine-grained, argillaceous, cross-bedded sandstones (the Soft Sandstone Lithofacies of Thompson, 1970b), with some interbedded siltstones and mudstones. It was deposited under continental fluvial and aeolian conditions in a braided river system dominated by transverse bars and dunes. Aeolian dunes developed in interdistributary areas under the influence of easterly winds; palaeocurrent vectors from water-laid sediments indicate transport to the north-west (Thompson, 1985, fig. 9).

The formation is 920 m thick in the Knutsford Borehole and 540 m thick at Prees (Figure 7) and (Figure 11); the thinner sequence encountered in the Prees Borehole is attributed to stratigraphical cut-out by normal faulting (Evans et al., 1993).

The Wilmslow Sandstone lacks stratigraphically useful fossils. Trace fossils, including vertebrate tracks, from the formation in its type area (see Sarjeant, 1974) are significant for environmental interpretation only. From its position beneath an inferred representative of the Hardegsen disconformity, which affects beds high in the Middle Bunter of Germany (Trusheim, 1961, 1963), and below the Mid-Triassic (Anisian) Helsby Sandstone, the Wilmslow Sandstone is assessed, with the Chester Pebble Beds, as Early Triassic in age.

Bulkeley Hill Sandstone Formation

This formation, the 'Keuper Sandstone Passage Beds' of Poole and 'Whiteman (1966), is recognised at outcrop only in west Cheshire, where it is up to 21 m thick (Poole and Whiteman, 1966; Earp and Taylor, 1986; Evans et al., 1993). It succeeds the Wilmslow Sandstone conformably but is overlapped by the Helsby Sandstone above an unconformity that was correlated (Warrington et al., 1980) with the Hardegsen disconformity (Trusheim, 1961, 1963). Evidence of the presence of this unconformity within the concealed SSG in the Cheshire Basin, and of the concealed extent of the Bulkeley Hill Sandstone, has been compiled from geophysical studies (Evans et al., 1993, figs 8, 10) which suggest a maximum thickness for the formation of about 220 m towards the basin centre.

The formation comprises massive, well-bedded, fairly coarse brown sandstones, with interbedded red flaggy sandstones, soft millet-seed sandstones and red-brown mudstones. It lacks fossils, but its stratigraphical position is analogous to that of the Wilmslow Sandstone and it is similarly assessed as Early Triassic in age.

Helsby Sandstone Formation

This formation has an almost continuous, though faulted, outcrop which fringes that of the MMG in the northern, western and southern parts of the basin. Small, isolated occurrences are found on the eastern side of the basin (Figure 7), adjacent to the Red Rock Fault. The Helsby Sandstone was proved at depth in the Prees and Knutsford boreholes (Figure 7) and (Figure 11) (Colter and Barr, 1975; Colter, 1978; Penn, 1987; Evans et al., 1993). It succeeds the Wilmslow Sandstone and, Bulkeley Hill Sandstone unconformably (Warrington et al., 1980; Evans et al., 1993) and is overlain, conformably, by the Tarporley Siltstone Formation, the lowest formation of the MMG.

The stratigraphy and sedimentology of the Helsby Sandstone in the Cheshire Basin have been studied by Thompson (1969, 1970a, b, 1985) who recognised a Soft Sandstone Lithofacies and a Red Pebbly Sandstone Lithofacies in the formations of the SSG (Thompson, 1970b). The Soft Sandstones, of fluvial, fluvio-lacustrine and aeolian origin, occur principally in the Wilmslow Sandstone but are also present in the Helsby Sandstone. The Red Pebbly Sandstones, representing fluvial-channel and associated overbank deposits of low- to moderate-sinuosity stream systems, occur in the Chester Pebble Beds and the Helsby Sandstone. Thompson (19706) proposed division of the Helsby Sandstone into members distinguished by the predominance of fluvial or of aeolian sediments. The principal members are the Thurstaston Soft Sandstone Member, the Delamere Pebbly Sandstone Member, and the Frodsham Soft Sandstone Member which occur broadly in that stratigraphical order. In view of the amount of possible interdigitation and lateral variation in thickness illustrated by Thompson (1970b, figs 6, 7) it is more appropriate to consider these members as facies associations whose stratigraphical order may vary laterally; a similar interpretation is placed upon the constituents of the correlative Ormskirk Sandstone Formation in the EISB (Meadows and Beach, 1993a, b; see below). The Helsby Sandstone is 230 m thick in the Prees Borehole and 205 m at Knutsford (Figure 7) and (Figure 11) (Evans et al., 1993).

BGS mapping in the Chester and Nantwich districts (geological sheets 109 and 122; (Figure 10) places the base of the Helsby Sandstone at the unconformable base of the Delamere Member, assigning the Thurstaston Member to the underlying Wilmslow Sandstone (Poole and Whiteman, 1966; Earp and Taylor, 1986). In the southern part of the basin, in the Oswestry and Wem districts (geological sheets 137 and 138; (Figure 10), the Ruyton and Grinshill Sandstones have historically been assigned, largely or entirely (Pocock and Wray, 1925; Wedd et al., 1929), to the sequence now termed Helsby Sandstone, and are shown as such on the 1:250 000 geological map (Enclosure 1).

In the early stage of Helsby Sandstone sedimentation, braided river channels were the sites of deposition of pebbly transverse sand bars and dunes and some pebbly channel deposits during high-discharge phases. Thompson (1970b, 1985) has recognised parts of three distributary courses, represented by coarse pebbly members in the northern part of the basin. Sediments of the Thurstaston Member facies accumulated between these distributaries and are more widely developed across the basin; they comprise low-energy fluvial sands and some possible aeolian dunes that formed on dry interdistributary tracts. The succeeding stage of Helsby Sandstone sedimentation witnessed the more widespread deposition of pebbly fluvial sands, of the Delamere Member facies, in distributaries of low to moderate sinuosity. Subsequently, fluvial deposition waned substantially and aeolian deposits of the Frodsham Member facies dominated the basin; localised fluvial deposits formed in discrete distributary channels, mainly in the eastern part of the basin. Palaeocurrent studies (Thompson, 1969, 1985, figs 9–11) indicate that aeolian facies in the formation accumulated under the influence of easterly winds and that fluvial transport was generally north-westwards. The fluvial systems were sourced in catchment areas in central and southern England and drained into the EISB (Warrington, 1970b; Warrington and Ivimey-Cook, 1992, maps Trlb, Tr2).

The extension of the three members of the formation north-westwards into the EISB (Colter and Barr, 1975; Colter, 1978; Ebbern, 1981; Colter and Ebbern, 1978, 1979; Bushell, 1986; Wilson and Evans, 1990; Stuart and Cowan, 1991) has been questioned by Meadows and Beach (1993b). In the EISB the equivalent beds are termed the Ormskirk Sandstone Formation (Jackson et al., 1987; Jackson and Mulholland, 1993) in which Stuart and Cowan (1991) recognised five major facies associations. Meadows and Beach (1993a, b) have recognised seven facies associations comparable with those of Stuart and Cowan (1991), comprising major and minor fluvial-channel sandstones, aeolian-dune and sandsheet sandstones, sheetflood deposits, and playa-lake and playa-margin deposits. Palaeogeographical interpretations (Meadows and Beach, 1993a, fig. 8; 1993b, figs 10, 11) suggest that the stratigraphical order and lateral relationships of these facies associations do not necessarily correspond to those proposed by Thompson (1970b) for similar associations in the Helsby Sandstone in the Cheshire Basin.

The age of the Helsby Sandstone is, on biostratigraphic evidence, early Mid-Triassic (Anisian) (Figure 8). In the absence of stratigraphically significant fossils from formations lower in the SSG, the position of the Permian-Triassic boundary is poorly constrained. In the northern part of the basin it is within some 1200 m of strata between the highest fossiliferous beds in the Manchester Marls and the base of the Helsby Sandstone; it may, therefore, occur within the topmost Manchester Marls or the lower part of the SSG. Further south in the basin it is considered to occur within the Kinnerton Sandstone (Figure 8) (Warrington et al., 1980).

Mercia Mudstone Group

This group has an extensive outcrop, enclosing the outliers of the Penarth and Lias groups in the southern part of the basin and almost surrounded by the outcrop of the SSG.

The MMG comprises six predominantly fine-grained elastic formations, the Tarporley Siltstone, Bollin Mudstone, Byley Mudstone, Wych Mudstone, Brooks Mill Mudstone and Blue Anchor formations, and two halite-bearing units. The lower halite-bearing unit, the Northwich Halite Formation occurs between the Bollin and Byley mudstones, and the upper unit, the Wilkesley Halite Formation, between the Wych and Brooks Mill mudstones (Figure 8) and (Figure 13)) for lithological key. From Wilson (1993)." data-name="images/P1000266.jpg">(Figure 12). The MMG succession in the Cheshire Basin was first demonstrated in the Geological Survey Wilkesley Borehole (Pugh, 1960).

The mudstone formations in the MMG include structure-less and laminated units. The former constitute the bulk of the Wych and Brooks Mill mudstones and the lower part of the Bollin Mudstone, and also occur within the Byley Mudstone. Laminated beds are dominant in the upper part of the Bollin Mudstone and also occur in the Byley Mudstone. The structureless units are probably largely aeolian in origin; Arthurton (1980) suggested that relict structures in these mudstones originated as 'ploughed ground' or zardeh, an irregular surface caused by growth of halite or gypsum crystals in sediments, which acted to trap wind-blown dust. Gypsum nodules grew in the sediment, probably in slightly upraised areas where gypsiferous solutions pulsed close to ground level (Shearman, 1970). The laminated units contain sedimentary structures indicative of deposition in shallow water.

Tarporley Siltstone Formation

This formation, formerly called the 'Keuper Waterstones', was defined by Warrington et al. (1980) from exposures around Tarporley. The lowest formation in the MMG, it is at least 280 m thick in the Ashley Borehole [SJ 7738 8355], south of Altrincham (Figure 13) and between 200 and 230 m thick in several boreholes elsewhere in the Cheshire Basin. The multiple seismic reflector associated with the Helsby Sandstone–Tarporley Siltstone succession (Figure 32) is identifiable widely in the Cheshire Basin. Towards the basin edge, however, 100 m is a more common thickness for the Tarporley Siltstone, with a likelihood that the unit thins out locally on the eastern fringe near Alsager. Exposures [SJ 900 656] to [SJ 892 652] near North Rode viaduct on the River Dane, which are seen near the basin edge, were first noted by Ormerod (1848); the formation here is rich in siltstone, but contains only scattered sandstones. The well exposed upper half of the section exposed here has been measured (Figure 13).

The Tarporley Siltstone comprises a distinctive facies that consists of alternating siltstones, reddish brown and greenish grey mudstones and thin fine- to medium-grained sandstones; desiccation surfaces and pseudomorphs after halite point to arid conditions. Cross-lamination, burrows and reptilian footprints are also recorded; Ireland et al. (1978) considered the formation to be intertidal in origin, on the basis of sedimentological and ichnofossil evidence from north-west Cheshire. Examination of the Saughall Massie Borehole, however, revealed abundant sedimentary structures (Figure 13) indicative of deposition in an ephemeral lake (playa) environment (see p.40).

In the Malpas area, a large part of the formation comprises a reddish brown cross-bedded sandstone, some 180 m thick. This unit, the Malpas Sandstone, contains abundant windblown grains (Poole and Whiteman, 1966) but includes water-laid sandstone, with current-ripple lamination and clasts of mudstone.

In the south-eastern part of the basin, near Norton in Hales [SJ 720 380], beds of Tarporley Siltstone facies alternate with up to six sandstone units some 20 to 30 m thick, of both aeolian and subaqueous origin, with rip-up clasts and cross-bedding. These strata have been assigned to the Helsby Sandstone in recent BGS mapping (geological sheet 123).

Bollin Mudstone Formation

This formation was defined by Wilson (1993); it replaces the former 'Lower Keuper Marl' and the 'lower mudstone division of the Mercia Mudstone Group' of Earp and Taylor (1986). The formation is defined as the beds between the highest sandstone interbed at the top of the Tarporley Siltstone and the base of the first bed of halite over 2 m in thickness, marking the base of the Northwich Halite. The type section is in the banks of the River Bollin [SJ 7953 8453] to [SJ 8281 8350], close to Manchester Airport (Figure 14).

In the northern half of the basin there is commonly a repetition of the Tarporley Siltstone facies in the overlying Bollin Mudstone. Formerly known as the 'Upper Keuper Sandstone', this was first recognised in the Marston borehole (De Rance, 1895). It has not been separately mapped at surface but has been identified in boreholes sunk for a road scheme north-west of Knutsford (Wilson, 1993).

The formation probably exceeds 500 m in thickness at the depocentre, which extends from the Northwich area, southwestwards to Wych Brook (Figure 15). The less well exposed lower half consists of structureless reddish brown mudstones with some laminated reddish brown and greenish grey mudstone and siltstone. The upper half is largely interlaminated mudstone and dolomitic siltstone, dominantly reddish brown in colour but with a number of greenish grey beds up to 4 m in thickness. Beds of halite up to 2 m in thickness occur locally in the topmost 20 m.

The upper part of the Bollin Mudstone is seen in a number of scattered sections along the river Bollin and in Wych Brook, near Malpas (Figure 14). Loose blocks in the road cutting to the underpass for the Manchester Airport runway [SJ 812 830] display sedimentary structures, which include desiccation cracks (Plate 1), pseudomorphs after halite (Plate 2), wrinkle marks, current ripples, cross-lamination, groove marks and convolute bedding (Wilson, 1993); similar structures occur in the stream sections and in borehole cores. Trace fossils occur very rarely. An insect wing and the branchiopod Euestheria minuta (Zieten) have been recorded by Thompson (1966) from Giant's Castle Rocks, the best exposure on the Bollin. Acritarchs of marine origin occur in this formation (Warrington in Earp and Taylor, 1986; Warrington in Wilson, 1993) suggesting periodic marine incursions across wide coastal plains or sabkhas.

Northwich Halite Formation

This formation, formerly the 'Lower Keuper Saliferous Beds', was named by Warrington et al. (1980) from Meadowbank Mine near Northwich. The thickest known development, of 283.2 m, is in Byley Borehole [SJ 7207 6942] (Figure 16). There is thinning towards the east, but the correlation in the marginal sequences is uncertain (Figure 16). Similarly, to the south-west, in Wych Brook, the narrow width of the collapsed zone (wet-rockhead) overlying the halite suggests attenuation of the formation.

Earp and Taylor (1986) estimated that some 25% of the Northwich Halite consists of mudstone. Several of the mudstone beds appear to correlate for many km across the basin (Figure 16). The best known of these is the 'Thirty-Foot Marl', which separates the 'Top' and 'Bottom' beds of the Northwich area. The halite occurs in beds which are virtually pure halite and in others in which there are varying amounts of mudstone and siltstone with crystals of halite, termed haselgebirge (Wilson, 1993) (Plate 3).

Meadowbank Mine [SJ 6525 6825] is currently working the 'Bottom Bed', some 30 m thick and consisting of 95% pure sodium chloride. This contains the giant polygons figured by Ward (1898) and subsequently described by Tucker (1981) and Tucker and Tucker (1981). The polygons, best seen in the mine roof, are 4 to 14 m across. In cross section the edges of the polygons are defined by salt-filled fissures some 1.5 m deep, with other, deeper, fissures to some 5 m; Tucker (1981) considered that the polygons, which resemble those formed at the sediment surface in contemporary salt flats in Iran, resulted from diurnal variations in temperature.

Further evidence that deposition of the Northwich Halite was in shallow water comes from laboratory experiments and studies of core by Arthurton (1973). He concluded that halite formed in shallow brine pools, at the water surface and on the bed of the pool. The enormous thickness of saliferous beds was probably related to persistent subsidence during a dominantly thermal sag phase of basin formation. The fact that subsidence was concentrated close to the eastern margin of the basin, however, suggests continued down-west displacements on the Wem–Bridgemere–Red Rock Fault System.

The bromine content of the Northwich Halite (Haslam et al., 1950) and the geochemistry of Sr, K and Mg (Tucker, cited in Thompson, 1989) suggest a marine origin for the brines. Small amounts of K and Mg salts occur, suggesting that 98% of the original sea water had evaporated.

Byley Mudstone Formation

The Byley Mudstone, the lower part of the former 'Middle Keuper Marl' and the 'middle mudstone division of the MMG' in the Cheshire Basin (Wilson, 1993) comprises a distinctive alternation of laminated and structureless mudstones and varies from 150 to 182 m in thickness. The contact with the overlying Wych Mudstone is taken at the top of the highest substantial (about 1 m thick) laminated greenish grey unit. The type section is in the Byley Borehole [SJ 7207 6942], which also shows the thickest development and appears to lie near the depocentre for a combination of the two formations (Figure 17). The strata are best known from boreholes, but the lowermost beds are seen in the River Dane [SJ 8808 6536] and the uppermost in Wych Brook (Figure 18).

The laminated facies consists of interlaminated siltstones and mudstones, each layer usually 1 to 5 mm in thickness. Siltstones are frequently dolomitic (Arthurton, 1980). Sedimentary structures figured by Wilson (1993) are desiccation cracks, pseudomorphs in siltstone after halite, and convolute and cross-lamination. The mudstones are either greenish grey or reddish brown, with most greenish grey beds towards the base.

The structureless mudstones are chiefly reddish brown and silty, consisting of clay minerals and detrital quartz silt. Anhydrite and dolomite are minor components (Arthurton, 1980). Gypsum veins are common; bands of nodules of anhydrite or gypsum occur sporadically.

Wych Mudstone Formation

This formation, the upper half of the former 'Middle Keuper Marl' (Wilson, 1993), consists almost entirely of reddish brown structureless mudstone with anhydrite or gypsum nodules (Plate 4) and numerous veins of gypsum. Nodules vary from a few mm to 10 cm in thickness. At surface, these beds weather to a sticky red clay and are seen in scattered cliffs in the valleys of Wych Brook [SJ 4887 4469] to [SJ 4931 4530] and on the River Dane west of Congleton [SJ 7870 6766], [SJ 8469 6407]. In the Wilkesley Borehole the formation is 186.44 m thick (Figure 19). Of the section proved in the Crewe Heat-Flow Borehole [SJ 6827 5452] (Figure 19), 106 m of core is held at BGS, Keyworth, for reference purposes.

A few laminated beds occur, either greenish grey or reddish brown in colour. In the vicinity of Sandbach 7 km north-east of Crewe, thin beds of halite occur towards the top of the formation (Figure 19). Two thin beds of halite were logged at lower levels in the sequence in the Crewe Heat-Flow Borehole, and these match gaps in the core at the Wilkesley Borehole, where halite may also have been present.

The red mudstones were probably largely aeolian in origin, akin to loess (Taylor et al., 1963; Wills, 1970).

Wilkesley Halite Formation

This formation, comprising the former 'Upper Keuper Saliferous Beds', was named by Warrington et al. (1980) after the only fully cored sequence of this formation, in the Wilkesley Borehole (Figure 20). The Wilkesley Halite is the thicker of the two saliferous formations in the Cheshire Basin. The only complete penetrations of the Wilkesley Halite are in the Prees Syncline, where the unit was 404.50 m thick in the Wilkesley Borehole and about 350 m in the Prees Borehole.

The upper half of the formation is somewhat purer than the average analysis for the Northwich Halite (Poole and Whiteman, 1966), and mudstone partings are under 3 m in thickness (Figure 20). In contrast, individual beds of the halite in the lower half of the formation tend to be less pure and there are numerous red and grey mudstone partings, up to 12 m in thickness.

The Wilkesley Halite contains four significant sandstone beds 0.30 to 1.65 m in thickness, as well as a few thinner sandstone layers, all within the mudstone partings. In contrast, there are no sandstone beds in the Northwich Halite.

Besides the main outcrop in the Prees Syncline, further occurrences of the Wilkesley Halite in and west of Nantwich were proved in the Burland Borehole [SJ 6018 5333] and probably in the area 5 km south of Crewe, on the evidence of subsidence hollows and seismic reflection data. An outlier around Sandbach, restricted to the lower beds of the formation, marks the site of the only exploitation of brine from the Wilkesley Halite in the present century.

Since the Wilkesley Halite is not seen in mines and there are no specimens of any of the clastic interbeds, less is known about the likely origin of this formation than that of the Northwich Halite. Much of the lower part of the Wilkesley Halite does, however, consist of haselgebirge, a lithology known to have been deposited in shallow water during the deposition of the Northwich Halite.

Brooks Mill Mudstone Formation

This formation, the former 'Upper Keuper Marl', is generally drift-mantled and is named from small exposures [SJ 6296 4370], [SJ 6315 4384] near Brooks Mill on the south bank of the River Weaver (Wilson, 1993). The base of the formation is at the top of the highest halite bed in the Wilkesley Halite and the top is placed at the incoming of the greenish grey mudstones of the Blue Anchor Formation. The only complete cored section, 161 m thick, is in the Wilkesley Borehole; in the Prees Borehole 205 m were recorded from cuttings and borehole geophysical logs. Other, partial, intersections of the upper half of the formation were in boreholes AU15 and 17 (Figure 21) (sunk south of Audlem and curated for reference purposes at BGS, Keyworth) and in the BGS Plattlane Borehole.

The Brooks Mill Mudstone consists dominantly of structureless reddish brown mudstones with anhydrite or gypsum nodules (Plate 5). The sequence in the Wilkesley Borehole is divisible into three parts, as follows:

1. Lower strata, below the anhydrite beds. These are structureless reddish brown mudstones 52 m thick, with anhydrite or gypsum nodules and veins of halite. Sandstone beds of greenish grey and reddish brown colour, up to 1.05 m in thickness, occur at four main levels near the middle of the sequence.
2. Anhydrite beds. Beds of anhydrite up to 0.9 m thick, with mudstone bands up to 0.3 m thick containing anhydrite nodules, make up a member 6.4 m thick.
3. Strata above the anhydrite beds. Structureless reddish brown and chocolate-coloured mudstones 103 m thick contain many nodules of anhydrite and gypsum, ranging in size from a few mm to 0.2 m; many gypsum veins are seen. Greenish grey beds occur in the topmost 18 m of the formation and are best known from boreholes near Audlem.

The structureless mudstones which dominate this formation are, like those in the Wych Mudstone, probably largely of aeolian origin.

Blue Anchor Formation

This formation, formerly known as the 'Tea Green Marl', was renamed by Warrington et al. (1980). Exposures of the formation occur in the headwaters of the River Weaver [SJ 633 436], [SJ 617 429], [SJ 653 428] and it has been cored in the Wilkesley and Plattlane boreholes; a partial core from the AU17 Borehole is preserved at BGS Keyworth.

The Blue Anchor Formation comprises poorly laminated, greenish grey mudstones, with some chocolate-coloured mottling and a transitional colour change to reddish brown at the base. Desiccation cracks, minor erosion surfaces, and horizons rich in polished sand grains of aeolian origin have been noted. Fish remains and the crustacean Euestheria minuta (Zieten) have been recorded (Figure 22).

Correlation and comparison with the sequence in the East Irish Sea Basin

The MMG successions in the Cheshire Basin and the adjacent EISB show an increasing degree of similarity at progressively higher levels in the sequence.

The lowest formation, the Tarporley Siltstone of Cheshire and its approximate equivalent on palynological grounds, the Hambleton Mudstone Formation on the eastern margin of the EISB, show little similarity of facies. Intertidal sediments near Daresbury in Cheshire (Ireland et al., 1978) contrast with probable near-coastal sabkha sediments near Blackpool (Wilson, 1990).

The lower half of the Bollin Mudstone of Cheshire, like the Singleton Mudstone Formation of Blackpool, contains much structureless reddish brown mudstone, but there may be rather more laminated strata in Cheshire. Beds of halite are not found in Cheshire, but do occur in the EISB, the Rossall salts lying low in the Singleton Mudstone Formation and the multiple Mythop salts higher in the sequence (Figure 13)) for lithological key. From Wilson (1993)." data-name="images/P1000266.jpg">(Figure 12). The Mythop salts, in particular, thicken markedly towards the depocentre of the EISB (Wilson, 1990, fig. 3).

The upper half of the Bollin Mudstone, like the Thornton Mudstone Member of Blackpool in the EISB, consists very largely of interlaminated mudstones and dolomitic siltstones, with abundant desiccation cracks and pseudomorphs after halite. Alternations of mudstone beds of greenish grey and reddish brown colour are in roughly equal proportions at Blackpool and Walney Island (Wilson, 1990, fig. 9), but by contrast in the Cheshire Basin the greenish grey beds are less numerous than the reddish brown ones.

The Northwich and Preesall halites were, on palynological evidence, laid down at about the same time and it is possible that the two basins had become a single depositional area by then. The Northwich Halite does not persist eastwards into the Stafford Basin, however.

The Byley Mudstone, which overlies the Northwich Halite, is closely similar to the Coat Walls Mudstone Member of Blackpool. In each there is a close alternation of structureless and laminated mudstone, with an upwards increase in the amount of structureless mudstone. Greenish grey beds occur in laminated strata at many levels, with the greatest concentration towards the base of the formation. It would seem likely that the Cheshire and East Irish Sea basins lay within the same depositional tract. However, palynological evidence suggests that, despite the close lithological similarities, the Anisian–Ladinian boundary lies closely above the base of the Coat Walls Mudstone Member in the Blackpool area (Wilson and Evans, 1990, fig. 14) whereas in Cheshire it is apparently located near the top of the Byley Mudstone (Wilson, 1993).

The Wych Mudstone of Cheshire closely resembles the Breckells Mudstone Formation of the EISB in the abundance of gypsum and anhydrite nodules in structureless reddish brown mudstone.

The Wilkesley Halite of Cheshire is the thicker of the two great salt deposits and is represented to the south-east, in the Stafford Basin, by the Stafford Halite. Equivalent strata in the EISB are probably indicated by recent salt-solution collapse breccias at Preesall, and further salts may occur offshore in the Keys Basin (Jackson and Mulholland, 1993; Jackson et al., 1995).

The equivalent of the Brooks Mill Mudstone of Cheshire is not preserved on land at the eastern margin of the EISB, but is almost certainly present in the same facies offshore. Contemporaneous beds as far away as Port More in Northern Ireland and at Keyworth in Nottinghamshire exhibit similarity of facies, with numerous nodules and some beds of anhydrite (Figure 21). There is similar widespread similarity of facies at the level of the Blue Anchor Formation, though this has not yet been verified in the EISB.

Penarth Group

The presence of the Penarth Group, formerly the 'Rhaetic' (Warrington et al., 1980), in the southern part of the basin was first recognised by Maw (1870). The outliers of those beds and the succeeding Lias Group in the Prees and Frith Farm areas of the Nantwich and Wem districts (geological sheets 122, 138; (Figure 7) and (Figure 10) have been delineated by the Geological Survey (Pocock and Wray, 1925; Poole and Whiteman, 1966; Wilson in James, 1983; Wilson, 1993). The succession has been proved in the Plattlane [SJ 5140 3645] and Wilkesley boreholes (Figure 7) and (Figure 22) (Poole and Whiteman, 1966) and in the Prees borehole (Figure 7) and (Figure 23) (Colter and Barr, 1975; Colter, 1978; Penn, 1987; Evans et al., 1993).

The Penarth Group comprises the Westbury Formation and the succeeding Lilstock Formation. The Westbury Formation comprises dark grey and black, fissile, micaceous silty mudstones, with thin lenses and beds of siltstone, fine sandstone, and limestone. Shelly fossils are abundant and fish remains are common; the fauna is indicative of a marine depositional environment. The formation is 7.79 and 7.81 m thick in the Wilkesley and Plattlane boreholes, respectively. In the Lilstock Formation, pale grey-green calcareous mudstones and siltstones, which are commonly micaceous, laminated and finely cross-bedded, comprise the lower part of the Cotham Member. They contain marine bivalves and include slumped beds (Figure 22). Slumped beds occur in the Gotham Member at many other localities in England; Mayall (1983) suggested that these widespread disturbed beds reflect earthquake activity. Fossils are scarce in the upper part of the Cotham Member and comprise crustaceans, ostracods and fish remains indicative of fresh to brackish water environments. The Cotham Member is overlain by a thin (<0.15 m), fine-grained, argillaceous limestone, which may represent the Langport Member and is here included in the Lilstock Formation. The Lilstock Formation is 5.77 and 5.74 m thick in the Wilkesley and Plattlane boreholes respectively.

The Penarth Group rests on a minor unconformity, represented by an erosion surface at the top of the MMG, and is succeeded conformably by the Lias. It is 13.55 m thick in the Wilkesley and Plattlane boreholes (Figure 22).

The age of the group is, on biostratigraphical evidence, late Late Triassic (Rhaetian) (c.210–205 Ma, Forster and Warrington, 1985) (Figure 8).

Lias Group

Rocks of the Lias Group were first noted in the Cheshire Basin by Murchison (1835). Outliers of these beds, the youngest Mesozoic rocks preserved in the basin, have been delineated by the Geological Survey in the Frith Farm and Prees areas of the Nantwich and Wem districts (geological sheets 122, 138; (Figure 7) and (Figure 10) (Pocock and Wray, 1925; Poole and Whiteman, 1966; Wilson in James, 1983; Wilson, 1993).

The succession was partly proved in the Wilkesley and Plattlane boreholes (Figure 7) and (Figure 24) (Poole and Whiteman, 1966). The Prees Borehole proved 597 m of beds here assigned to the Lower Lias (Figure 7) and (Figure 23) (Colter and Barr, 1975; Colter, 1978; Penn, 1987; Evans et al., 1993) which are succeeded, at outcrop, by 30 m of beds assigned to the Middle Lias; the youngest bed recognised is the Marlstone Rock Bed (Pocock and Wray, 1925; Poole and Whiteman, 1966; Cope et al., 1980).

The Lias comprises dominantly grey fissile mudstones, shales and thin limestones, and was deposited in a marine environment.

The succession in the Prees outlier is, on biostratigraphic evidence, latest Triassic (late Rhaetian) and Early Jurassic (Hettangian to late Pliensbachian) in age. The boundary between the Triassic and Jurassic systems is placed at the level of the appearance of ammonites of the genus Psiloceras (Cope et al., 1980; Warrington et al., 1980). In the Wilkesley Borehole (Figure 24) this level is 9.37 m above the base of the Lias (Poole and Whiteman, 1966), the lowest beds of which are therefore assigned a Late Triassic (Rhaetian) age.

Estimates for depths of burial over the basin are given on pp.67–70. The calculated depth of burial of 1580 m for Prees (Table 4), (Figure 54) provides a local estimate for the thickness of post-Pliensbachian sediments deposited and subsequently eroded.

Biostratigraphy

Biostratigraphical information is available for the Manchester Marls Formation, the Helsby Sandstone Formation, the MMG (below the Wilkesley Halite), and the Penarth and Lias groups.

Manchester Marls Formation

The Manchester Marls Formation has yielded plant remains (including spores and pollen), calcareous microfossils (foraminifera and ostracods), and macrofossils, including gastropods, scaphopods, bivalves and fish remains (Binney, 1841, 1855, 1862; Geinitz, 1889, 1890; Roeder, 1890a, b, c, 1892; Tonics et al., 1931; Jones et al., 1938; Stoneley, 1958; Taylor et al., 1963; Logan, 1967; Pattison, 1969, 1970; Pattison et al., 1973). Plant remains are recorded throughout the formation, but the other fossils occur only in the lower beds which include thin limestones (Pattison, 1970). The fossil associations are comparable to those known from the Zechstein (Upper Permian) sequence in the North Sea Basin and from correlative beds in the EISB (Pattison et al., 1973; Jackson et al., 1987) and are indicative of a Late Permian age. The occurrence of the pollen Lueckisporites virkkiae in the Manchester Marls at Manchester (Pattison et al., 1973) implies, by comparison with the palynomorph succession in the type area of the Permian in Russia (Warrington in Smith et al., 1974), a Kazanian to Tatarian age (c.260–250 Ma, Forster and Warrington, 1985) (Figure 8).

Helsby Sandstone Formation

This formation has yielded plant remains (including spores and pollen), crustaceans, remains of reptiles, and ichnofossils. Of this material, only the spores and pollen and the reptilian remains are of biostratigraphical value; the crustaceans (Euestheria: Brockbank, 1891; Warrington, 1963) and the trace fossils, though significant for the interpretation of the environment of deposition, are of little value for dating purposes. The ichnofossils include traces and trails of invertebrates (Pollard, 1981, 1985) and the celebrated vertebrate ichnofauna recovered mainly from sites in north Cheshire but also known from other localities in the basin. The vertebrate tracks are the subject of an extensive literature reviewed by Swinton (1960), Sarjeant (1974, 1984, 1985), Delair and Sarjeant (1985) and Tresise (1989). The basal member of the Helsby Sandstone at Alderley Edge [SJ 859 774] (Figure 7) has yielded miospores (spores and pollen). The assemblage (Warrington, 1970b) includes Angustisulcites klausii and is now considered indicative of an Anisian (early Mid-Triassic) age (Benton et al., 1994). Reptilian remains (Rhynchosaurus articeps) from the topmost beds of the SSG and the basal beds of the overlying Tarporley Siltstone (MMG) at Grinshill [SJ 518 328] (Figure 7) in the southern part of the basin (Benton, 1990) are also interpreted as Anisian in age (Benton et al., 1994).

Mercia Mudstone Group

Fossils are commoner in parts of the MMG than in the SSG in the Cheshire Basin. They include: ichnofossils; spores, pollen and other palynomorphs (acritarchs); and crustaceans (Euestheria). The trace fossils, acritarchs and crustaceans are significant for the interpretation of depositional environments but not for dating purposes.

Records of spores and pollen from independently dated Triassic successions elsewhere in Europe (Visscher and Brugman, 1981; Van der Eem, 1983; Brugman, 1986) provide a basis for the dating of part of the MMG succession in the Cheshire Basin. The base of the Anisian Stage is marked by the lowest occurrences of Stellapollenites thiergartii and Angustisulcites spp. The highest occurrence of S. thiergartii and the lowest occurrence of Ovalipollis pseudoalatus are at or just above the Anisian–Ladinian boundary. Slightly above the base of the Ladinian, Camerosporites secatus and Duplicisporites spp. appear, followed by Echinitosporites iliacoides, signifying an early Ladinian (Fassan substage) age. Miospore assemblages from the MMG in the Cheshire Basin (Warrington, 1970a; Warrington in Earp and Taylor, 1986; Fisher, 1972a, b) indicate, by reference to the above criteria, that the Anisian succession extends upwards from the Helsby Sandstone to a level above the Northwich Halite. In addition to occurring in the basal Helsby Sandstone at Alderley Edge (Warrington, 1970b), Angustisulcites spp. are recorded, together with Stellapollenites thiergartii and other taxa, such as Perotilites minor, that are indicative of an Anisian age, from the overlying Tarporley Siltstone at Liverpool (Fisher, 1972a, b). Palynological evidence of an Anisian age has also been obtained from the Tarporley Siltstone in the Chester district (geological sheet 109, (Figure 10; Warrington in Earp and Taylor, 1986). Assemblages containing Angustisulcites spp., P. minor and S. thiergartii have also been recovered from the Boffin Mudstone, and Angustisulcites spp. occurs with Tsugaepollenites oriens in the upper part of the Byley Mudstone, which is now assessed as late Anisian in age. The palynological evidence suggests that the Anisian–Ladinian boundary lies at the top of the Byley Mudstone (Figure 8) or in the lower part of the Wych Mudstone (Wilson, 1993; Benton et al., 1994). Few miospore assemblages have been recovered from higher levels in the MMG. The Wilkesley Halite may, on very sparse palynological evidence, span the Mid to Late Triassic (Ladinian–Carnian) boundary or be entirely Carnian (early Late Triassic) in age. There is no satisfactory biostratigraphical evidence for the age of the succeeding Brooks Mill Mudstone and Blue Anchor formations. The latter, which contains sporadic crustaceans (Euestheria) and fish remains, is, however, overlain by the Penarth Group, of Rhaetian (late Late Triassic) age (see below). The succession between the Wilkesley Halite and the Penarth Group is therefore considered to be Late Triassic and to comprise beds of Norian and probably early Rhaetian age (Figure 8).

Penarth Group

Anderson (1964) recorded ostracods from the Penarth Group of the Plattlane Borehole. The macrofaunas from the group in the Plattlane and Wilkesley boreholes were documented by Ivimey-Cook (pp.48–50 and Appendix 2 in Poole and Whiteman, 1966). The presence of taxa such as Protocardia rhaetica and Rhaetavicula contorta in the bivalve fauna (Figure 22), and of the dinoflagellate cyst Rhaetogonyaulax rhaetica in palynomorph assemblages, indicates that the Penarth Group is of late Late Triassic (Rhaetian) age (c.210–205 Ma, Forster and Warrington, 1985).

Lias Group

Fossils (other than ammonites) from the Lias in the Wilkesley and Plattlane boreholes were recorded by Ivimey-Cook (pp.48–50 and Appendix 2 in Poole and Whiteman, 1966), and the foraminifer biostratigraphy has been reported by Copestake (1989) and Copestake and Johnson (1989). The ammonite biostratigraphy of these boreholes was determined by Donovan (pp.50–51 and Appendix 2 in Poole and Whiteman, 1966) and has been revised by Ivimey-Cook (Figure 24). The base of the Jurassic is placed at the level of the appearance of ammonites of the genus Psiloceras (Cope et al., 1980; Warrington et al., 1980); in the Wilkesley Borehole (Figure 24) this occurs 9.37 m above the base of the Lias. Above that level in the Wilkesley Borehole the ammonite faunas indicate the presence of the Psiloceras planorbis, Alsatites liasicus and Schlotheimia angulata biozones, of Hettangian age, and the Arietites bucklandi biozone and part of the Arnioceras semicostatum biozone of early Sinemurian age (Figure 24). In the Plattlane Borehole only the lower part of the planorbis biozone is present (Figure 24). The youngest unit in the Lias succession preserved in the Prees outlier is of late Pliensbachian (Pleuroceras spinatum biozone) age (Pocock and Wray, 1925; Poole and Whiteman, 1966; Cope et al., 1980).

Application of biostratigraphic results

The biostratigraphic dating of units in the basin fill provides a basis for the assessment of the rates of sediment accumulation in the Cheshire Basin. The preserved fill ranges in age from Early Permian (inferred) to Early Jurassic (late Pliensbachian) (Figure 8), from a maximum age of about 300 Ma at the Carboniferous–Permian boundary (Hess and Lippolt, 1986) to c.187 ± 5 Ma at the end of the Pliensbachian (Hallam et al., 1985). The poorly constrained lower part of the fill, comprising the Collyhurst Sandstone, Manchester Marls, Chester Pebble Beds and Wilmslow Sandstone formations, comprises nearly 1900 m of mostly coarse clastic sediments that accumulated during a maximum time of 58 ± 5 Ma (Forster and Warrington, 1985; Hess and Lippolt, 1986). The part of this sequence comprising the Late Permian (Kazanian–Tatarian) Manchester Marls and the succeeding units below the Mid-Triassic Helsby Sandstone accumulated in about 18 ± 5 Ma and is represented by some 1300 m of sediments. The succession comprising the Manchester Marls and beds up to and including those of Anisian age above the Northwich Halite spans approximately 25 ± 5 Ma and is represented by more than 2300 m of sediments. The Anisian succession comprises more than 1000 m of sandstones, mudstones and evaporites that accumulated in as little as 7 Ma (Forster and Warrington, 1985). The remainder of the MMG, together with the Penarth Group and basal (Triassic) beds of the Lias, comprises nearly 900 m of mostly fine-grained elastic sediments and evaporites which accumulated over a period of some 30 Ma (Ladinian–Rhaetian; Forster and Warrington, 1985). The 600 m of Jurassic rocks preserved in the basin accumulated over a period of some 18 Ma (Forster and Warrington, 1985; Hallam et al., 1985). Burial history (subsidence) curves constructed by Evans et al. (1993, fig. 11) show that the most rapid subsidence occurred during Early and Mid Triassic times (see also Chapter 3). During this phase the SSG and part of the MMG (below the Wilkesley Halite Formation) accumulated under conditions of fault-controlled subsidence.

Sedimentology

A sedimentological analysis was made of selected boreholes with continuous, or near continuous, core (Table 3), (Figure 2). The boreholes contained strata ranging from the Collyhurst Sandstone to the Tarporley Siltstone. A full account of the study is given in Jones (1994). Detailed logs of boreholes through the MMG are summarised by Wilson (1993) and in (Figure 13)) for lithological key. From Wilson (1993)." data-name="images/P1000266.jpg">(Figure 12), (Figure 13), (Figure 14), (Figure 16), (Figure 18), (Figure 19), (Figure 20) and (Figure 21).

Appleby Group

Collyhurst Sandstone Formation

The upper 20 m of this formation is present in the Speke Reservoir Borehole. It consists of moderate to dark reddish brown, fine- to coarse-grained sandstones, interpreted, in common with the formation as a whole (pp.16–17), as aeolian-dune deposits. This interpretation is based on the recognition of aeolian grainfall and grainflow laminations. Alternating fine and coarse-grained foresets are interpreted as grainfall laminations, produced by the deposition of saltating grains blown over dune crests (Hunter, 1977). Sharp-based, coarse-grained foresets are interpreted as grainflow laminations, produced by the avalanching of sediment over dune crests (Hunter, 1977). Dewatering structures indicate that the sediment was wet, probably as a result of a high water table.

Cumbrian Coast Group

Manchester Marls And Bold Formations

The Manchester Marls Formation was not present in any of the boreholes studied, but a 10 m sandstone unit underlying the Chester Pebble Beds in the Speke Reservoir Borehole is attributed to the Bold Formation. If this attribution is correct, the overlying Kinnerton Sandstone must have been eroded in this area. The core is distinct from the surrounding units in that it is finer grained and contains an interbedded succession of wind-blown and water-lain sands. The wind-blown sandstones are interpreted as aeolian sandsheet deposits, formed in an area of predominantly aeolian sand where dunes are absent (Pye and Tsoar, 1990). Such deposits typically occur marginal to dune fields and develop in environments where conditions do not favour the development of dunes. Poorly defined laminations are interpreted as aeolian grainfall deposits. Sets of cross-bedding represent the deposits of small aeolian dunes that migrated across the sandsheet. Massive sandstones are probably secondary in origin, as a result of the destruction of lamination. Siltstone laminae with desiccation cracks result from deposition of fines in ephemeral pools on the sandsheet. Wavy and convoluted silty laminae represent accretion of wind-blown sediment onto a damp sediment surface.

The water-lain sandstones are interpreted as fluvial sheet-flood facies, the product of weak tractional sheetfloods deposited by unconfined flows across the sandsheet. The presence of current-ripple cross-lamination and cross-bedding indicates that unidirectional currents were dominant. Also present are low-relief ripple form sets with straight crests and symmetrical ripple forms; these are wave formed and indicate reworking within a shallow subaqueous setting. Presumably, following sheet-flooding, ponding of water created small lakes in which waves could be generated.

Sherwood Sandstone Group

Kinnerton Sandstone Formation

Aeolian dune deposits were identified in the Stanlow Borehole (250 m of reddish brown friable sandstones with a weak calcareous cement) and the Halewood Borehole (7.65 to 68.80 m and 100.50 to 130.00 m, orange to pinkish brown, fine- to coarse-grained sandstones, poorly cemented and friable). There are foreset stratification types which are distinctly aeolian, with grainfall lamination as the dominant type, formed by sand blown over the dune crest and deposited on the lee slope. Lenses of well-sorted, coarse-grained sandstone are interpreted as grainflow horizons, formed by sediment avalanching. Overturned folds result from grainflow in non-cohesive dry sands (McKee et al., 1971). Pinstripe lamination is interpreted as subcritically climbing translatent stratification (Hunter, 1977), produced by the migration of ripples under conditions of net deposition (Pye and Tsoar, 1990).

Aeolian sandsheet deposits were recorded in the Halewood Borehole (68.8 to 100.5 m), formed by the accretion of aeolian-supplied fine sand and silt. The crudely laminated nature of the deposit indicates irregular accretion of sediment onto a damp surface with convolutions formed as a result of loading into water-saturated sediments. Aeolian dunes migrated across the sandsheet when the sediment was dry.

As noted earlier the formation in general is of continental, aeolian facies (p.18).

Chester Pebble Beds Formation

This formation is generally represented by continental, fluvial deposits (see p.19). Erosion surfaces were produced by episodes of channel scouring, with pebbles overlying these surfaces representing lags. Moderate to high-energy, unidirectional currents generated sets of cross-bedding, with trough cross-bedding produced by sinuous-crested, subaqueous dunes and planar cross-bedding produced by straight-crested, subaqueous dunes.

A fine- to medium-grained sandstone, 2 m thick, in the ICI Widnes Borehole shows low-angle cross-bedding, low-relief rippleform laminae that lack foresets, and wavy and convoluted silty laminae. These features suggest an aeolian origin, probably by reworking of channel deposits.

Wilmslow Sandstone Formation

In the boreholes examined (Table 3), this formation consists of sandstones of aeolian dune and interdune facies. In general it includes both fluvial and aeolian facies (see pp.18–19).

The aeolian dune facies comprises dominantly fine- to medium-grained sandstones. Foresets comprise alternating coarser- and finer-grained laminae. There are rare claystone laminae, discontinuous and broken, which were formed by the accretion of mud-grade sediment, probably in localised pools along dune bases; drying of these pools led to the desiccation and breakage of the laminae.

The aeolian interdune facies occurs interbedded with the dune facies, and comprises mainly fine-grained reddish brown to pale buff sandstones. Interdunes are low-lying areas between dunes, characterised either by deflation or deposition (Ahlbrandt and Fryberger, 1982). The deposition of sandstone and the lack of features such as deflation lags and ventifacts indicates that this is a depositional interdune sequence. Such sequences can be classified as wet, damp or dry, depending on the nature of the sediment surface (Kocurek, 1981). Siltstone laminae result from deposition of fines in ephemeral pools when the interdune was wet. Wavy and convoluted silty laminae (the dominant type in these sections) develop by accretion of wind-blown sediment onto a damp interdune surface. Drying of the interdune area is indicated by the occurrence of desiccation cracks and from the presence of the low-angle lamination, a typical feature of dry interdunes (McKee and Bigarella, 1979). These interdune successions were characterised by fluctuations in the nature of the interdune surface, with damp conditions dominating.

Helsby Sandstone Formation

The Thurstaston Soft Sandstone Member was examined in the Gallantry Bank boreholes (GB 80-1 and GB 80-4), where it typically comprises medium- to coarse-grained, reddish brown sandstones, interpreted as aeolian dune deposits on the basis of aeolian foreset stratification and the absence of fluvial features (such as erosion surfaces and mudstone pebble horizons). A 0.3 m claystone horizon containing sandstone lenses, convolute lamination and broken and upturned claystone laminae is interpreted as a wet interdune deposit, formed by the deposition of mud within a pool of water; drying of this pool led to the desiccation and breakage of the claystone laminae.

The Delamere Pebbly Sandstone Member was examined in the Gallantry Bank (GB 80-1 and GB 80-4) and Saughall Massie boreholes, where it consists of interbedded fine-, medium- and medium- to coarse-grained sandstones with subordinate claystones and siltstones, interpreted as deposits of low-sinuosity fluvial channels. Extraformational quartz pebbles were carried by high-energy currents. Intraformational pebbles are locally derived, by erosion of overbank fines and in-channel drapes. Moderate- to high-energy, unidirectional currents are responsible for generation of sets of cross-bedding. Weakly defined lamination probably formed by deposition of sediment from suspension. Mudstones were also deposited from suspension, during periods of low flow within the channel. Discontinuous and broken claystone laminae are the product of sediment exposure and desiccation.

The Frodsham Soft Sandstone Member in the Saughall Massie Borehole is an orange-brown, fine- to medium-grained, poorly cemented sandstone. It is interpreted as an aeolian dune succession, because of the presence of aeolian foreset stratification and the absence of fluvial features. A medium-grained basal sequence containing thin wavy, slightly convoluted silt laminae, micaceous in parts and faintly cross-laminated, is interpreted as a damp to wet interdune deposit, with sediment supplied by sheetflood events.

Mercia Mudstone Group

Tarporley Siltstone Formation

A 95 m sequence of the Tarporley Siltstone was examined from the Saughall Massie Borehole. An intertidal mudflat origin has previously been suggested for this formation, mainly on the basis of a marine trace-fossil assemblage from exposures in north-west Cheshire (Ireland et al., 1978), but no evidence for these trace fossils was seen in the Saughall Massie Borehole.

Two main facies were recognised in the borehole: a saline mudflat and an ephemeral playa lake. The saline-mudflat facies consists dominantly of reddish brown structureless siltstones and minor very fine to fine grained sandstones. This facies is considered to have formed from a combination of water-lain and wind-blown fine-grained sediment, deposited on an irregular salt-encrusted subaerial surface. The dominance of mudrock indicates a mudflat environment. Irregular sandstone lenses are thought to reflect entrapment on a disrupted, salt-encrusted surface.

The ephemeral-playa-lake facies is characterised by thinly to thickly interlaminated claystones, siltstones and sandstones. The heterolithic nature of this facies reflects alternating deposition of coarser- and finer-grained sediment, characteristic of ephemeral playa lakes. The well-laminated nature of the facies suggests that the sediments were predominantly water lain. Sandstones contain sedimentary structures indicative of tractional processes and hence are interpreted as the product of weak or distal sheetflood events. The thinly laminated nature of many of the sandstones is indicative of subaqueous flows. The occurrence of wave ripples, and evidence for wave reworking of current ripples, suggests deposition within shallow bodies of water of sufficient fetch to experience small waves, probably wind generated. Interlaminated siltstones and claystones were deposited from suspension, following sheet-flooding. The common occurrence of desiccation cracks in the mudstones indicates that lake drying was common, further evidence that these were ephemeral playas. Pseudomorphs after halite show that conditions were saline, possibly as a result of lake drying, with evaporation leading to the formation of brines saturated with halite.

Chapter 3 Structure and evolution of the basin

R A Chadwick, D J Evans, W J Rowley, N J P Smith, A S D Walker, B Birch and J Bulat

The Cheshire Basin is markedly asymmetrical in cross-section, having, in general terms, the form of a faulted half Oben, deepest in the south-east near the Wem–Bridgemere–Red Rock Fault System which forms its south-east margin. The present-day cumulative throw on this faulted margin in places exceeds 4000 m. In contrast, the western margin of the basin is relatively unfaulted, forming a feather-edge characterised by depositional onlap. A colour-shaded gravity image of the basin (Figure 49), (Figure 50), (Figure 51)." data-name="images/P1000279.jpg">(Figure 25) illustrates its gross morphology, the Permo-Triassic basin fill corresponding closely to a large negative gravity anomaly.

This account of the structure and evolution of the basin is based mainly on the interpretation of seismic reflection data, complemented by geological information from boreholes (Figure 2) and outcrop (Enclosure 1). Additionally, potential-field interpretations (of gravity and aeromagnetic data) have been used to provide further information on the deep structure. Details of the seismic and potential-field interpretative methodologies are given in Chadwick et al. (1994a).

Overview of structural evolution

Early Palaeozoic evolution

South-west of the Cheshire Basin, outcrops of Lower Palaeozoic strata of the Welsh Caledonides are widely developed. These sediments were deposited close to the eastern margin of the Lower Palaeozoic Welsh Basin and are cut by major north-east-trending Caledonian faults (Figure 26).

During early Palaeozoic times, present-day Precambrian inliers such as those of the Longmynd probably formed elevated areas around the margin of the Welsh Basin. Thin sequences were deposited here, Tremadocian and Caradocian strata being the only Ordovician rocks preserved.

A thicker, more complete Cambrian–Silurian sequence was deposited west of the Pontesford–Linley Fault, which is interpreted as a down-west syn-sedimentary fault (Smith, 1987). Ordovician rocks in the Welsh Basin are associated with arc volcanics, whereas Silurian rocks are turbidites derived from the rising Caledonides in the west, and probably shed into a foreland basin. Both sequences contrast with much thinner shallow-water shales and reefs deposited contemporaneously on the Midland Massif to the east. The Pontesford–Linley Fault probably developed in Cambrian times, ceasing normal movement in the late Ordovician. From then to late Llandovery times, basin inversion occurred, with reversal of the Pontesford–Linley Fault and development of the Shelve Anticline. Preservation of Caradoc rocks in a depression east of the Pontesford–Linley Fault, unconformably overlying both Uriconian and Longmyndian rocks, suggests that basin inversion occurred soon after Caradoc times.

Cambrian and Tremadoc rocks probably subcrop beneath the Carboniferous rocks west of the Pontesford and Hodnet faults as far as the Prees Borehole (Smith, 1987), consistent with late Ordovician basin inversion beneath the southern part of the Cheshire Basin. A borehole in the Hanwood Coalfield penetrated Coal Measures, ,which rest on a ?Cambrian sandstone.

Upper Llandovery strata are unconformable on rocks of Longmyndian to Ordovician age, and thicken eastwards towards the Pontesford–Linley Fault, indicating some reactivated downthrow to the west. It is likely that a Silurian basin lay to the north of the Shelve inlier, as well as to the south, where littoral deposits were described by Whittard (1952).

Much of the Welsh Basin and the subsurface to the north-east are devoid of rocks of Devonian age. In contrast, thick successions are preserved on the Midland Massif at outcrop and in the subsurface, suggesting that Acadian inversion of the Welsh Basin formed a sediment source for the Old Red Sandstone rivers draining south-eastwards.

The Caledonian Orogeny comprised two main events in this region. The first was associated with the late Ordovician basin inversion. The second, Acadian, event (Soper et al., 1987), in early to mid-Devonian times, uplifted the whole Welsh Basin. It is likely that Acadian movements included large transpressive displacements (commonly down to the east) on many of the major structures such as the Bala Fault, the Clun Forest Disturbance, the Church Stretton Fault and the Pontesford–Linley Fault (Figure 26). The Caledonian orogenic front may correspond to the Church Stretton Fault, imaged as a west-dipping thrust on seismic reflection data (Smith, 1987) and separating, at the surface, rocks with steep dips in its hangingwall block from a relatively undeformed footwall block to the east. The Pontesford–Linley structure is a prominent feature on images of gravity and magnetic fields (Figure 27) and has a prolonged history of reactivation, possibly from Proterozoic times onwards (Wills, 1978; Woodcock, 1984; Soper et al., 1987). Reactivation of this structure appears to have strongly influenced both Carboniferous and Permo-Triassic structural development of the Cheshire Basin (see below).

Carboniferous evolution

The plate-tectonic process which most strongly influenced Carboniferous structural development of the Cheshire Basin region was the formation, several hundred km to the south, of a collision-type orogenic belt in the Iberian–Armorican–Massif Central region of the Hercynides (Leeder, 1982). Northwards subduction of the Rheic Ocean resulted in north–south (back-arc) extension to the north of this orogenic belt.

Broadly speaking, Carboniferous basin development in the Cheshire Basin region can be split into two main phases. In early Carboniferous times, rapidly subsiding fault-controlled extensional basins developed between structurally elevated, largely emergent blocks. This was followed, in late Carboniferous times, by more regional subsidence, which was characterised by a lack of major fault control and led to submergence and depositional onlap of the earlier structural highs.

Details of Carboniferous basin development are sketchy because the seismic reflection data do not image the Carboniferous sequence effectively. Comparison with the better described early Carboniferous extensional basins of northern England (e.g. Fraser and Gawthorpe, 1990; Kirby et al., in press) suggests that an extensional Carboniferous basin underlies the Permo-Triassic Cheshire Basin. In general terms, because of its position close to St George's Land, the basin is likely to have a thinner sequence than the basins farther north. It is, moreover, likely to have a geometry markedly different to the overlying Permo-Triassic basin, given the roughly orthogonal extension directions (approximately north–south in the Carboniferous, approximately east–west in the Permo-Triassic). Structural trends of the principal early Carboniferous faults were strongly influenced by the structural grain of the underlying basement. Major Caledonian basement features such as the Bala and Pontesford–Linley faults were reactivated as down-to-the-north-west normal faults which controlled extensional basin development.

In latest Carboniferous times, final closure of the Rheic Ocean culminated in the Variscan Orogeny, with large-scale thrust and nappe emplacement and regional crustal shortening in northern France, Belgium, southern England, South Wales and southern Ireland. A zone of northerly directed thrusts, the Variscan Front, marks the northern limit of regional Variscan compressional deformation in southern England; central and northern England lie to the north, on the Variscan Foreland. Variscan deformation was much less pervasive here, being largely restricted to the reversal of pre-existing Dinantian normal faults and associated basin inversion.

The main Variscan movements in the Cheshire Basin region probably postdated the preserved Westphalian rocks and predate deposition of the Permian strata which rest unconformably on the Carboniferous beds. Dating the precise onset of the main Variscan movements is difficult, but evidence from central England (Glover et al., 1993b) suggests that inversion may have commenced as early as Westphalian C times.

The main effect of Variscan compression in the region appears to have been reversal of important Caledonian lines of weakness and associated early Carboniferous faults, which led to structural inversion of the Carboniferous basins. Evidence of this comes from the south-east margin of the Cheshire Basin, where seismic reflection data show west-dipping reflections between 0.1 and 0.65 s two-way travel time, beneath the Permo-Triassic cover (Figure 46). Seismic character and regional correlation suggest that these reflections correspond to concealed Westphalian Coal Measures. Their relationship to the Carboniferous succession imaged at the extreme eastern end of the line, where the Keele Formation (Barren Measures) outcrops, suggests that the formations are separated by a down-east reverse fault. Precise displacement of the Westphalian strata is difficult to assess, because of their complex structure and uncertain seismic character matches. In the footwall of this fault, other, smaller, down-east reverse faults displace the Coal Measures and Barren Measures. These faults are related to the many Variscan compressive structures seen at outcrop in the area, which range in size from the Potteries Syncline to the numerous small thrusts seen in coal workings in the North Staffordshire Coalfield (Taylor et al., 1963; Evans et al., 1968).

The reverse faults lie roughly along the north-easterly continuation of the Clun Forest Disturbance and the Pontesford–Linley Fault (Figure 26) and are interpreted as being associated with Variscan transpressive reactivation of these or related basement fractures.

A few km to the south, seismic reflection and gravity data (Figure 39) and (Figure 51) show the sub-Permian basement structure along Section 4 of (Figure 38). Carboniferous strata pinch out eastwards beneath the Wem–Audlem Sub-basin, with Lower Palaeozoic rocks forming the pre-Permian subcrop on the hangingwall block of the Wem Fault. In contrast, east of the Wem Fault, on its footwall block, the gravity interpretation (Figure 51) and outcrops to the south (British Geological Survey, 1990), indicate that thick Carboniferous strata occur beneath the Permo-Triassic cover. As elsewhere in the basin, this structural configuration appears to be the consequence of basement-fault reactivation (Figure 28). Variscan reversal of the Wem Fault (or its structural precursor) led to uplift of its hangingwall block and inversion of the Carboniferous basin.

The Edgerley Fault, in the western part of the basin (Figure 40) may also be associated with reactivation of the Bala Fault, another important Caledonian, Carboniferous and Variscan basement structure.

Permo-Triassic and post-Triassic evolution

By earliest Permian times, Variscan continental collision had led to final suturing and consolidation of the Pangaean supercontinent. The Cheshire Basin region lay deep within this continental mass. Variscan basin inversion and regional uplift had resulted in considerable elevation of the land surface, which, during Permian times, underwent progressive peneplanation. Contemporaneously, regional sag basins developed in the North Sea region. By late Permian, and, particularly, early Triassic times, the north-west European region formed an isthmus between the rapidly developing Arctic–North Atlantic rift system to the north and the Tethys–Central Atlantic–Gulf of Mexico rift-wrench system to the south (Figure 29). Regional crustal extension became established as the dominant tectonic process.

a)                  Carboniferous basin development

b)                  end-Carboniferous (Variscan) basin inversion

c)                  Permo-Triassic extension and basin development, with reactivation of earlier fault.

The Cheshire Basin formed part of a Permo-Triassic rift system which extended from the English Channel Basin in the south to the East Irish Sea Basin in the north. The geometry of the basins of this rift system was strongly influenced by the extensional reactivation of older (Caledonian and Variscan) structural features in the underlying basement rocks.

The post-Triassic evolution of the Cheshire Basin is poorly understood because of the paucity of preserved strata. It is likely, however, that regional crustal extension continued episodically, prior to the onset of North Atlantic sea-floor spreading in mid-Cretaceous times. Regional considerations, and, particularly, evidence from more fully preserved sedimentary sequences to the south, indicate that extensional basin subsidence continued into early Jurassic times (Figure 30)a; the preserved Lias succession near Prees is about 600 m thick (see p.33). Renewed extension in the late Jurassic and early Cretaceous was associated with sea-floor spreading in the southern part of the North Atlantic region (Figure 30)b as Iberia and America separated. Erosion associated with development of the widespread late Cimmerian unconformity (Rawson and Riley, 1982) is likely to have affected much of the Cheshire Basin region in the early Cretaceous. Most of any lower Cretaceous deposits, together with a substantial thickness of Jurassic and possibly Triassic strata, are likely to have been removed at this time, particularly from the basin edges. In mid-Cretaceous (Cenomanian) times, the onset of sea-floor spreading in the North Atlantic effectively terminated extension in north-west Europe. Deposition recommenced in the late Cretaceous, as regional, post-extensional shelf subsidence became established, with deposition of the laterally uniform Chalk. Maximum basin development was probably attained in earliest Cenozoic (Palaeocene) times (see p.70). Subsequent Cenozoic uplift comprised two distinct components. Regional uplift took place, possibly due to crustal underplating associated with development of the Icelandic Plume to the north-west (Brodie and White, 1994). Superimposed on this, a more localised uplift due to basin inversion, associated with Alpine continental collisions to the south, led to severe erosion and removal of a large part of the basin fill. Basin inversion, not necessarily contemporaneous with the regional uplift, corresponded to one or more of three inversion episodes in southern Britain: the late Cretaceous inversion in the southern North Sea (Glennie and Boegner, 1981; Badley et al., 1989), the Palaeocene inversion of the Wessex Basin (Lake and Karner, 1987) or the main Oligo-Miocene inversion of southern Britain and the southern North Sea (Chadwick, 1993; Badley et al., 1989). Well-documented cases of basin inversion reasonably close to the region include the Sole Pit Trough (Van Hoorn, 1987) and the Wessex Basin (Lake and Karner, 1987), where the main phase of inversion occurred in Oligo-Miocene times, corresponding to major Alpine nappe development. It is considered that the principal inversion of the Cheshire Basin occurred at this time, though earlier, minor, phases cannot be ruled out.

The detailed structure of the Cheshire Basin and aspects of the Permo-Triassic and subsequent phases of basin evolution, are examined in more detail below.

Present-day structure

Permo-Triassic and Jurassic rocks

Seismic stratigraphy and mapping

Evans et al. (1993) identified the principal Permo-Triassic seismic reflectors, and established a characteristic seismic stratigraphy for the basin fill, based on the Knutsford and Prees boreholes which bottom in Coal Measures and Lower Palaeozoic strata respectively (Figure 31). Seismic-reflection data near the Knutsford Borehole illustrate the distinctive seismic character of the sedimentary sequence in the central part of the basin (Figure 32).

The MMG is imaged as a seismically banded sequence. Laterally continuous zones of high-amplitude reflections alternate with seismically less-reflective zones, corresponding to thick saliferous and mudstone units respectively. The Bollin Mudstone and Tarporley Siltstone formations and the Helsby Sandstone Formation of the underlying SSG can usually be identified as groups of moderate- to high-amplitude continuous reflections (reflectors a, b and c respectively on (Figure 32).

The Permo-Triassic succession beneath the Helsby Sandstone Formation gives rise to a seismically layered interval, within which three poorly reflective zones are developed, corresponding to the Wilmslow Sandstone, Chester Pebble Beds and Collyhurst Sandstone formations. These three transparent zones are separated by two layers of high-amplitude, coherent reflections, corresponding to the top of the Silicified Zone within the Wilmslow Sandstone Formation (Colter and Barr, 1975), and the Manchester Marls (and lateral equivalents) and Kinnerton Sandstone formations. Reflections from these strata are of high amplitude with good continuity and can be traced across much of the Cheshire Basin. The base of the Permo-Triassic succession is in places marked by a high-amplitude, continuous event, commonly associated with truncation of underlying reflections from Westphalian strata.

Farther south, seismic data near the Prees Borehole have a character closely similar to that near Knutsford (Evans et al., 1993). This characteristic seismic stratigraphy, together with data from many shallow boreholes and the surface geology, enables Permo-Triassic reflectors to be identified across much of the basin with a fair degree of precision. In the north-west of the basin however, seismic data are of generally poorer quality, with the base of the Permo-Triassic succession and, in particular, the base of the SSG being somewhat uncertain. Also, in the vicinity of the faulted eastern margin of the basin, complex structure and large thickness variations, together with poor data quality in some places, make interpretation within individual fault blocks difficult.

Seven principal reflectors were picked during interpretation of the seismic data, of which four, Base Jurassic, Base MMG, Base SSG and Base Permo-Triassic (Figure 31), were depth-converted (Chadwick et al., 1994a). Initial structure-contour maps were produced at a scale of 1:50 000. They were subsequently reduced, simplified to produce 1:100 000 compilation maps, and digitised. Major faults were digitised as polygons, but most of the faults were digitised as simple line traces, fault-width (heave) information thereby being lost. The digital data were then gridded using the Dynamic Graphics ISM and EarthVision software to produce colour-shaded depth maps (Figure 33), (Figure 34), (Figure 35) and (Figure 36).

Principal structural features

The principal structural features of the Cheshire Basin are illustrated in (Figure 38)." data-name="images/P1000291.jpg">(Figure 37) and (Figure 38). The basin is quite heavily faulted; more than 600 individual faults or fault segments have been seismically resolved and, in view of the sparse data coverage in parts of the basin, it is likely that considerably more faults are present. Of the seismically resolvable faults, the large majority are small, subplanar, normal faults. Fault-dips are moderate to steep (55–80°), being generally somewhat steeper in the south of the basin. Most of these smaller faults penetrate to the surface; throws are typically in the range 100–500 m and tend to decrease upwards, indicating both Permo-Triassic and probable post-Triassic syndepositional movement. A few significantly larger normal faults are present, which are subplanar and of considerable length (several tens of km), with dips in the range 45–75°. Throws are typically in the range 500–2500 m, decreasing upwards, again indicating syn-depositional basin-controlling displacements. The large normal faults divide the basin into several smaller structural provinces, each comprising a set of tilted fault blocks.

Curved listric fault geometry is developed only locally (e.g. the King Street Fault), and evidence of faults detaching onto the salt units of the MMG is sparse. Similarly, no evidence was found of large-scale salt movement and diapirism.

The structural complexity of the Cheshire Basin means that its form departs in places from that of a simple southeast-deepening half graben. In the north, the basin is roughly symmetrical (Figure 38), Section 1, the principal structural feature being the Sandbach–Knutsford Sub-basin which here forms a westerly deepening asymmetrical graben bounded by the Brook House Fault. In the north, the Red Rock Fault appears to have a rather small throw, so the eastern margin of the basin is not marked by a major structure. The central part of the basin (Figure 38), Section 2 is also roughly symmetrical, with a slight eastward deepening asymmetry. The principal structure in this part of the basin is the Sandbach–Knutsford Sub-basin, forming a symmetrical graben bounded to the west and east by the Overton and Alderley faults respectively. The Red Rock Fault marks the eastern margin of the basin, with a throw in this area of at least 1800 m down to the west. True half-graben asymmetry is developed in the southern part of the basin (Figure 38), Sections 3 and 4, where the south-east-deepening Wem–Audlem Sub-basin forms the dominant structural feature. The Wem and Bridgemere faults, which typically have present-day normal throws (at the base Permo-Triassic level) of some 2500 m, down to the west, form the south-east margin of the basin in this area.

The principal structural features described below have been identified from the detailed seismic reflection interpretation. The major faults correspond to previously recognised surface structures but names for the newly recognised sub-basins and blocks are, as yet, provisional.

The deepest parts of the Cheshire Basin are presently within the Wem–Audlem and Sandbach–Knutsford sub-basins, where the pre-Permian basement lies at depths greater than 4000 m (Figure 36).

The north-north-east-trending Wem–Audlem Sub-basin, in the south of the Cheshire Basin has the form of a classic half graben (Figure 38), Sections 3 and 4; (Figure 39), deepening to the south-east, with the Wem and Bridgemere faults forming its south-eastern margin. The sub-basin gradually shallows northwards, passing into the Lymm High.

Pre-Permian basement lies at depths greater than 4500 m in the hangingwall block of the Wem Fault, west of the Prees Borehole, and greater than 4250 m in the hangingwall block of the Bridgemere Fault, south-east of the Burford Borehole (Figure 38)." data-name="images/P1000291.jpg">(Figure 37). The Permian, SSG and MMG all thicken south-east into the sub-basin (Figure 38), Sections 3 and 4, indicating that it developed as a structural feature throughout the extensional phase of basin evolution, by contemporaneous normal displacements on the Wem and Bridgemere faults.

The two areas in the Wem–Audlem Sub-basin where the pre-Permian basement is at the greatest depth appear to have had significantly different structural histories. Whereas the base of the SSG and MMG and the base of the Jurassic succession also reach their greatest depth in the area west of Prees, the more northerly area of deep pre-Permian basement, south-east of Burford, is not well defined at higher stratigraphical levels. Indeed, at the base of the MMG it marks a local structural high (Figure 34). Depth-of-burial and back-stripping studies (see pp.70 and 74) indicate that the latter area marked the Permo-Triassic depocentre of the Cheshire Basin, but suffered significantly more uplift than surrounding areas during Cenozoic basin inversion.

The north-trending Sandbach–Knutsford Sub-basin lies in the eastern part of the Cheshire Basin. In the south it forms a roughly symmetrical graben bounded to the west by the King Street Fault and to the east by the Alderley Fault (Figure 38), Section 2. Farther north it becomes markedly asymmetrical, deepening towards its western boundary, the Brook House Fault (Figure 38), Section 1. Pre-Permian basement lies at depths in excess of 4000 m in the hangingwall block of the King Street Fault. Permian strata, the SSG and the MMG all thicken across the bounding faults of the sub-basin, indicating that it developed as a structural feature throughout the extensional phase of basin evolution by contemporaneous normal displacement of these faults.

West of these two important sub-basins, the Permo-Triassic basin fill is much thinner. The western updip flank of the Wem–Audlem Sub-basin – the Western Slope – is characterised by a Permo-Triassic sequence which has a regional east-south-east dip (towards the Wem–Audlem Sub-basin) and which thins gradually towards the more or less unfaulted western basin margin. Westward thinning occurs partly as a result of stratigraphical attenuation of the sequence but principally by the present-day erosion surface, which cuts across progressively older strata to the west. Faults in this area trend north-east to north-north-east and are mostly small, down-west normal faults, forming minor, south-east-dipping tilt-blocks (Figure 38), Sections 3 and 4. A more prominent tilt-block, the Milton Green Inlier is demarcated to the west by the important Edgerley Fault, of which it forms the footwall block (Figure 38), Section 3; (Figure 40). The westerly edge of this east-dipping tilt-block suffered sufficient relative uplift during the extensional phase of basin development for pre-Permian basement rocks (Carboniferous strata) to be exposed at surface at the present day.

North of the Western Slope, an area of shallow pre-Permian basement, generally less than 1000 m deep, has many small normal faults, throwing both down-east and down-west to create a complex system of tilt-blocks (Figure 38) Section 2. Scattered larger normal faults, such as the down-east Dungeon Banks Fault (Figure 38), Section 1 are present locally. In the north-west corner of the basin the Ellesmere Saddle (Figure 38), Section 1 has outcropping pre-Permian basement to the north-east and south-west, and connects the Cheshire Basin to the West Lancashire–East Irish Sea Basin.

East of the Ellesmere Saddle, the northern part of the Cheshire Basin is characterised by regional southerly dips, with Permo-Triassic strata dipping and thickening southwards into the Wem–Audlem and Sandbach–Knutsford sub-basins. Stratigraphical thickening is particularly pronounced in the Permian succession, which is extremely condensed close to the northern edge of the basin (Figure 41). A prominent feature of the northern part of the basin is the Lymm High (Figure 38), Section 1, which forms a broad, faulted structural block, with an axis plunging southwards into the Wem–Audlem Sub-basin. The eastern margin of the Lymm High is marked by the Brook House and King Street faults, forming the uplifted footwall blocks of these faults. The sedimentary sequence is thin, especially compared with the Sandbach–Knutsford Sub-basin to the east (Figure 38), Sections 1 and 2. The Permian succession is particularly condensed on the footwall block of the Brook House Fault (Figure 42), suggesting that the Lymm High was a prominent feature throughout (but particularly early in) the extensional phase of basin development.

The Alderley High, which forms the north-eastern corner of the Cheshire Basin (Figure 43), is bounded to the west by the Mobberley and Alderley faults, and to the east by the Red Rock Fault at the basin margin (Figure 38), Section 1. In its northern part the Alderley High forms the updip part of the hangingwall block of the down-east Mobberley Fault (Figure 44), cut by many smaller north- or north-north-west-trending, mostly down-east, normal faults. Farther south it forms the footwall block of the down-west Alderley Fault, and, except in the south, is heavily faulted. In addition to the dominant north-trending normal faults, east- to north-east- trending cross-faults occur locally, which are thought to be steep structures which acted as transfer faults during extensional basin development. Some of these structures occur close to the Alderley Edge mineral deposits and may have played an important part in the development of the ore bodies (see Chapter 7 and Summary). The Alderley High is characterised by relatively thin Permian and SSG sequences, the former thinning eastwards towards the basin margin (Figure 45). The MMG is absent or partially preserved, having been removed by later erosion.

The eastern margin of the Wem–Audlem Sub-basin is marked by the large faults of the complex Wem–Bridgemere–Red Rock Fault System (WBRRFS), the Blakenhall and Ternhill terraces forming relatively undisturbed tracts within the fault system.

The Blakenhall Terrace, between the Bridgemere and Wem faults (Figure 46), forms a small structural block, transitional between the deep Wem–Audlem Sub-basin to the west and the margin of the Cheshire Basin to the east. The Permian sequence is very thin or absent, suggesting that displacement on the WBRRFS was largely restricted to the Bridgemere Fault in Permian times, initiating the Wem–Audlem Sub-basin in its hangingwall block, but leaving its footwall block (the Blakenhall Terrace) largely subaerially exposed. By early Triassic times, displacement along the Wem Fault led to deposition of SSG on the Terrace, which nonetheless continued as a structural feature during subsequent extensional development, with SSG and MMG sequences markedly thinner than in the Wem–Audlem Sub-basin to the west.

The Ternhill Terrace lies to the east of the WBRRFS (Figure 46), and is bounded to the east by the Hodnet Fault. It is similar to the Blakenhall Terrace, though somewhat larger, and is also characterised by a thin, locally absent, Permian sequence. It probably existed as an active structural feature throughout the extensional phase of basin develop ment, with SSG and MMG sequences markedly thinner than in the Wem–Audlem Sub-basin to the west (Figure 39).

The Wem–Bridgemere–Red Rock Fault System (WBRRFS), a major structure over 110 km long, forms the eastern margin of the Cheshire Basin and was the dominant influence on basin development (Figure 38), (Figure 39), (Figure 45) and (Figure 46). The fault system comprises an anastomosing network of large down-west normal faults. The trend is north-east in the south, changing to a northerly trend in the north. The reactivation of older basement structures played an important part in determining the location and geometry of the fault system (see p.76). The cumulative throw on the WBRRFS is greatest in its southern-central section where it locally exceeds 4000 m at the base of the Permo-Triassic succession (Figure 36). Throws on individual structures (the Wem and Bridge-mere faults) locally exceed 2500 m. The throw on the fault system gradually diminishes northwards, as the Bridgemere Fault dies out and the Wem Fault passes into the Red Rock Fault (Figure 38), Sections 1 and 2. Where it marks the eastern margin of the basin (Figure 45), the observable throw on the Red Rock Fault is typically 1000–2000 m, but this decreases northwards to a few hundred metres, or less, in the north-east of the basin. However, because Carboniferous rocks outcrop on the footwall block of the Red Rock Fault, along much of its length, only minimum estimates of its true throw can be made; the real throw may be much greater. The constituent faults of the WBRRFS have planar or sub-planar fault-surfaces, which show a systematic variation in dip along the length of the fault system. The Red Rock Fault has moderate westerly dips, typically in the range 45–60° (Figure 38), Section 2. Further south the Wem and Bridgemere faults are significantly steeper, with north-westerly dips in the range 60–70°. Locally, close to its junction with the Wem Fault, the Bridgemere Fault is very steep (up to 80°), and has the appearance on seismic data (Figure 46) of a negative flower-structure (sensu Harding, 1985), indicative of strongly oblique extension (see also p.82). Small transfer faults (sensu Gibbs, 1984) are associated with the WBRRFS. They appear to trend approximately east to north-east and cut into the hangingwall blocks of the fault system, such as the Alderley High. Particularly well-developed transfer faults are associated with offsets in the Bridgemere Fault, one such feature marking the northern margin of the Blakenhall Terrace.

All preserved stratigraphical units thicken markedly westwards across the WBRRFS, e.g. (Figure 39),(Figure 45) and (Figure 46), resulting in an upward decrease of displacement and indicating that the system controlled the development of the basin throughout the extensional phase of evolution.

Depth-of-burial studies (p.70) suggest that during Cenozoic basin inversion the hangingwall blocks of the WBRRFS suffered differential uplift of up to 1000 m relative to the footwall blocks. Associated reversal of the WBRRFS may also, therefore, have been up to 1000 m locally. Further evidence of this reversal is provided by local reverse displacements and the development of tight hangingwall-anticlines (Figure 45) in parts of the WBRRFS. Prior to inversion, therefore, the cumulative down-west normal throw on the WBRRFS may in places have approached 5000 m.

Two important down-west normal faults associated with the WBRRFS are the Hodnet and Alderley faults. The Hodnet Fault (Figure 38), Section 3 lies to the east of the WBRRFS, in its footwall. It appears to splay southwards from the Wem Fault, forming the eastern boundary of the Ternhill Terrace and demarcating the Cheshire Basin from the Stafford Basin to the south-east. The fault is poorly understood, because of sparse seismic coverage, but gravity data (Figure 51) suggest that it has a considerable westerly downthrow. The Alderley Fault (Figure 38), Section 2, lies in the hangingwall block of the WBRRFS, splaying north from the Wem Fault, and is approximately 30 km long. It separates the Alderley High from the much deeper Sandbach-Knutsford Sub-basin to the west, and makes an important contribution to the cumulative downthrow at the basin margin. The sub-surface nature of this fault is poorly understood (Figure 45) because of the quality of the seismic data. Its throw, at the pre-Permian basement, is believed to be of the order of several hundred metres (locally in excess of 1000 m), decreasing markedly upwards.

Three large down-east normal faults, the Brook House, King Street and Mobberley faults, lie in the hangingwall of the WBRRFS, and link with it at depth in an antithetic relationship. The Brook House Fault (Figure 38), Section 1, which constitutes the major structure in the northern part of the Cheshire Basin, is about 15 km long, with a northwesterly trend, and separates the Lymm High from the northern part of the Sandbach–Knutsford Sub-basin; preserved strata thicken eastwards across the fault. Thickening of the Permian sequence is particularly marked (Figure 42), indicating that the fault was an important structure early in basin evolution. The throw on the fault locally exceeds 1000 m, decreasing upwards and northwards. The King Street Fault (Figure 38), Section 2 is 30 km long, trends north–south, and links to the Brook House Fault via subvertical, possible transfer, faults. It marks the western margin of the Sandbach–Knutsford Sub-basin (Figure 47). The fault is typically planar, and quite steeply dipping (c.70°), but in parts a less steep, somewhat listric geometry is developed (Figure 38), Section 2. Throws on the King Street Fault approach 1000 m at pre-Permian basement level, but diminish north and south. Eastwards stratigraphical thickening occurs across the fault, particularly in the SSG, but in contrast to the Brook House Fault, thickening of the Permian sequence is minor. The Mobberley Fault (Figure 38), Section 1 lies to the east of the King Street and Brook House faults and is 30 km long, with a north–south trend. Like these faults, it has an easterly downthrow locally approaching 1000 m and is largely syn-depositional, with commensurate eastward thickening of stratigraphical units (Figure 44).

Further west, the Overton and East Delamere faults (Figure 38), Sections 2 and 3 form a major north–south structure 70 km long. The faults have planar geometry and moderate (c.50–60°) easterly dips, steepening somewhat in the south. Easterly downthrows locally exceed 1000 m, at the top of pre-Permian basement. There is only minor eastward thickening of preserved strata across the faults, suggesting that most of the displacement took place after deposition of the SSG.

The principal fault in the west of the basin is the Edgerley Fault (Figure 38), Sections 2 and 3, 60 km long and with a north–south trend. It is unusual for the Cheshire Basin in that it displays a markedly curved (listric) profile, with a rollover developed in its hangingwall block (Figure 40). The fault downthrows to the west, locally almost 1000 m where pre-Permian basement (Carboniferous) outcrops in the footwall block at the western edge of the Milton Green inlier. Seismic data suggest that the sub-cropping Carboniferous sequence is different on either side of the fault, and thus displacements pre-dated deposition of the Permian. The fault may reflect reactivation of the Bala Fault, an important basement fracture (p.44).

In addition to the many normal faults, there are a small number of reverse faults (Figure 42) and (Figure 48). They are generally steep, with small reverse throws which appear to have formed by strongly oblique-reverse displacements in lithified (brittle) basin fill. A larger reverse fault is visible near the eastern margin of the basin, which, together with the folding described above, suggests widespread reversal of the basin-margin faults, probably associated with Cenozoic basin inversion (see pp.45 and 70).

Carboniferous rocks

Carboniferous rocks, comprising a Dinantian sequence overlain by Silesian (Namurian and Westphalian) rocks, underlie much of the basin except for a small area in the south where lower Palaeozoic rocks come to subcrop (Figure 140). The Carboniferous succession is clearly imaged on seismic profiles beneath the northern and western parts of the basin (Figure 40), (Figure 41), (Figure 42) and (Figure 43), where it is locally more than 4000 m thick. Elsewhere, the thickness of Carboniferous strata is constrained by the gravity interpretation. Gravity stripping (Chadwick et al., 1994a) indicates that the observed negative Bouguer anomaly (Figure 49), (Figure 50), (Figure 51)." data-name="images/P1000279.jpg">(Figure 25) cannot be wholly explained by the Permo-Triassic basin fill: considerable thicknesses of upper Carboniferous strata are also required. 2-D gravity modelling along basinwide traverses indicates upper Carboniferous (Namurian and Westphalian) thicknesses in excess of 2000 m beneath the northern part of the basin, gradually decreasing southwards (Figure 49), (Figure 50) and (Figure 51).

Much of the lithostratigraphical information relating to the concealed Carboniferous sequence beneath the Cheshire Basin comes from boreholes and outcrops outside the basin.

Dinantian rocks

Dinantian rocks in the Cheshire Basin are proved only in the Milton Green, Blacon East, Ternhill, Stoke-on-Tern and Bowsey Wood boreholes, and the top Carboniferous Limestone reflector cannot be traced very widely from Milton Green, so deductions about Dinantian rocks and structures are based largely on the surrounding outcrops.

North-east Wales

In north-east Wales a north-north-west-striking outcrop of Dinantian rocks (Smith and George, 1961) reveals synsedimentary fault control by the Llanelidan (Bala) and Aqueduct faults (Figure 52). Both faults have thicker Dinantian successions on their northern (hangingwall) blocks. Subsequent dating of the sequence (Somerville and Strank, 1984) indicates that although late Asbian and Brigantian rocks thicken on the downthrown sides of the faults most of the displacement occurred early in the Dinantian. Arundian to early Asbian rocks and older, undated Basement Beds are thick to the north of the faults and absent on the footwall blocks to the south. Information from deep boreholes to the east indicates that the Milton Green Carboniferous inlier corresponds to a Dinantian structural high (Dinantian rocks are c.320 m thick), forming an eastern extension of the Cyrn-y-Brain Anticline (Figure 52). On the outcrop to the west at Minera, Dinantian rocks are 260 in thick. The Blacon East Borehole (Figure 2), drilled to the north of Milton Green, confirms an eastern extension of the hangingwall basin formed above the Llanelidan Fault, here termed the Blacon Sub-basin. The borehole proved thick Pendleian to Brigantian rocks, with a thick Lower Bowland Shale equivalent. A seismic reflector, deeper than the borehole, dips towards the subsurface continuation of the Llanelidan Fault, suggesting thickening of the early Dinantian rocks in this sub-basin. Dinantian rocks in the Blacon East borehole total more than 400 m, with the possibility of a further 600 m beneath terminal depth.

Eastern margin of the Cheshire Basin

In the south of the basin the WBRRFS throws Triassic strata against Westphalian rocks. Further north, near Astbury (Figure 139), Namurian and Dinantian rocks outcrop on the footwall block. The section at Astbury is thinner than that drilled in the Gun Hill Borehole (Figure 139) to the east (Evans et al., 1968). A south-west-deepening sub-basin, the north-west continuation of the Widmerpool Gulf (or Staffordshire Basin; Trewin and Holdsworth, 1973), lies between the Woo Dale Borehole (Smith et al., 1985) and Gun Hill. Small-scale tilt blocks and basins, with the same polarity, are seen on seismic profiles beneath the Cheadle Coalfield. A syn-sedimentary fault may form the eastern margin of the North Staffordshire Coalfield, between Astbury and Gun Hill, as envisaged for Namurian times (Trewin and Holdsworth, 1973). If, as seems likely, the Red Rock Fault acted as a Dinantian syn-sedimentary fault (see below) then Astbury stood on a structural high which was bounded by two differently trending Dinantian syn-sedimentary faults. At Apedale (Figure 139) the borehole may have drilled down the limb of a fold (840 m of ash) or intersected a volcano of possible Dinantian age (Rees and Wilson, 1998).

Southern margin of the Cheshire Basin

Beneath the southern part of the Cheshire Basin, where Carboniferous reflectors are few, the Carboniferous structure cannot be resolved directly. Along the south-eastern margin of the basin large Permo-Triassic faults connect southwestwards to major Caledonian structures such as the Pontesford–Linley structure. These faults are comparable to the Llanelidan (Bala) Fault in their longevity. Outcrops suggest that, in common with many pre-existing faults, the Wem–Red Rock and Pontesford–Linley–Hodnet faults acted as down-west syn-sedimentary normal faults in Dinantian times (see p.43). The Church Stretton Fault probably passes between the Edgmond Borehole (Figure 2) and Stoke-on-Tern. No Dinantian rocks were encountered in the Edgmond Borehole, whereas more than 200 m are present at Stoke-on-Tern. Over 100 m of Dinantian was proved in the Ternhill Borehole, close to the Hodnet Fault.

The cross-section (Figure 52) shows the Dinantian sequence thickening southwards to the limit of outcrop at Llanymynech, which may be related to syn-depositional movement on the Wem Fault.

In the subsurface, seismic profiles show thickening of the Dinantian sequence southwards from Milton Green. Termination of deep reflectors may occur against a fault, which may be the continuation of the Aqueduct Fault.

Dinantian rocks are overlapped by unconformable Westphalian D rocks beneath the south-western part of the basin.

North-eastern margin of the Cheshire Basin

Deep seismic reflectors in this area, not proved by drilling, probably correspond to rocks in Dinantian basins, controlled by north-north-west-trending faults. One of these faults may be the East Delamere–Overton–Runcorn Fault. To the west, a thin sequence is present, similar to that at Milton Green; there are deeper reflectors to the east. The basin may link northwards to the Rainford Sub-Basin (underlying the western part of the Lancashire Coalfield), where eastwards-thinning Dinantian may reach the Brook House Fault.

Namurian rocks

Regional considerations and relationships elsewhere in the Pennine Basin (Fraser and Gawthorpe, 1990; Kirby et al., in press) suggest that most of the Namurian and later Carboniferous rocks were deposited during the post-extensional (thermal relaxation) phase of basin formation. Since rocks of Kinderscoutian age commonly overlie the earlier extensional structural highs, it is possible that the base of the Kinderscoutian marks the syn- to post-extensional transition.

North-western part of the Cheshire Basin

Approximately 170 m of Namurian rocks were intersected in the Milton Green Borehole on the Cyrn-y-Brain High, and more than 800 m in the Blacon East Borehole in the Blacon Sub-basin; Namurian strata were also intersected in a few other boreholes. Geophysical log correlation with wells drilled in Lancashire suggests that a full basin sequence is present at Blacon East, including thick Kinderscoutian, Arnsbergian and Pendleian strata. The intersection in the Milton Green Borehole could represent either a very condensed, complete sequence or a sequence lacking the early stages. The critical marine band at about 1137 m is not definitive (N J Riley, personal commmunication). The Blacon West borehole terminates in probable Kinderscoutian rocks. These sandstones probably represent southerly derived protoquartzite deltas, in contrast to the feldspar-rich grits of the northern part of the Pennine Basin (Jones, 1980). The Aqueduct Grit is a distinctive low-gamma unit within the Namurian Series in the Milton Green Borehole.

South-western part of the Cheshire Basin

In the Talwrn Borehole, Namurian rocks may reach 230 m in thickness; they probably thin northwards, in the Wrexham district, on to the Cyrn-y-Brain High. Namurian rocks have a narrower outcrop width (and may, therefore, be thinner) south of the Aqueduct Fault, compared to its northern hangingwall block. In the Oswestry district, the outcrop width suggests that the Namurian rocks thin to the south. Sandstones are poorly developed and thin in comparison with the main Pennine Basin; they are quartzitic in composition, having been derived from St George's Land (Wales). Towards the top of the Namurian succession the Aqueduct Grit has a feldspathic composition, consistent with the incursion of southward spreading Pennine deltas. In the extreme southwest of the basin, Namurian rocks are overlapped by unconformable Westphalian D sttata, beneath Permian rocks.

Northern and eastern parts of the Cheshire Basin

East of the Red Rock Fault a basinal Namurian succession shows the relationship between the southern and northern elastic sources more clearly. The Pendleian, Arnsbergian–Alportian turbidite sandstones were derived from the south

(Trewin and Holdsworth, 1973). The Kinderscoutian delta did not reach this basin, possibly because of the intervening Woo Dale tilt-block high (Smith et al., 1985; Collinson, 1988). Marsdenian delta sandstones, derived from the north, swung round to flow axially into the basin from the southeast (Jones, 1980). This area was called the Staffordshire Basin by Trewin and Holdsworth (1973), but appears to be contiguous with the Widmerpool Gulf. Collinson et al. (1977) have speculated that this basin extended north-west to join the thick Marsdenian deltaic sandstones derived from the north-east.

Northward thinning of the southerly derived turbidites in the Macclesfield district (Evans et al., 1968) may have caused starved sedimentation of Edale Shale type beneath this part of the Cheshire Basin, resulting in the deposition of potential hydrocarbon source rocks.

This part of the basin is therefore probably underlain by the north-west-trending continuation of the Widmerpool Gulf. The aeromagnetic data (Chadwick et al., 1994a) support a structural continuation of magnetic basement from the Derbyshire Dome beneath the north-eastern part of the basin.

In the Bowsey Wood Borehole (Figure 139) only Yeadonian rocks are present. They are unconformable on Dinantian rocks (Ramsbottom et al., 1978) but in faulted contact with Westphalian rocks, consistent with overlap of thick Westphalian rocks (Productive Coal Measures) across the feather edge of Namurian strata. Namurian rocks are also absent from the Market Drayton High.

Westphalian rocks

Westphalian strata are present beneath much of the Cheshire Basin. Within the Westphalian succession the Symon Unconformity separates sequences thickening in opposite directions. Thus, beneath the unconformity, Productive Coal Measures, encompassing rocks of Westphalian A, B and part of C, thicken northwards to a depocentre near Manchester (Calver, 1968), with splitting of seams in a northerly direction. In contrast, red beds above the unconformity appear to thicken southwards, successively down-cutting Productive Coal Measures, Namurian and Dinantian strata to rest unconformably on Silurian and older rocks.

North-western part of the Cheshire Basin

This area includes the Milton Green Inlier and the Flintshire and Denbighshire Coalfields. The Productive Coal Measures are about 450 m thick in the boreholes and about 500 m thick in the Denbighshire Coalfield (Glover et al., 1993b). Strata above the Symon Unconformity (the Ruabon Marl, Coed-yr-Allt Formation and Erbistock Beds) are about 530 m thick at Milton Green and thicken southwards on seismic profiles.

South-western part of the Cheshire Basin

The Carboniferous outcrop terminates in the Oswestry district, west of Eardiston copper mine. The Symon Unconformity is represented here by overlap of the Productive Coal Measures, Namurian and Dinantian rocks by the Ruabon Marl and Coed-yr-Allt Formation. The thinning and failing of the formations beneath the unconformity is due to the proximity of St George's Land to the south. The possible thickening of the red beds above the unconformity is discussed below. No Productive Coal Measures are known from the Shrewsbury Coalfield, where the red beds rest on Cambrian rocks (Pocock et al., 1938). The Leebotwood Coalfield also features red beds unconformable on late Precambrian Longmyndian rocks. The basal beds here contain bitumen, and oil' was collected from the Shrewsbury Coalfield.

In the north of the Shrewsbury Coalfield the Lower Braggington Fault (Pocock et al., 1938), forming the southerly continuation of the Wem Fault, downthrows to the south-east, providing evidence for Variscan reversal of this fault (see p.44).

South-eastern part of the Cheshire Basin

In the Coalbrookdale Coalfield, Productive Coal Measures (c.100 m thick) rest unconformably on Silurian rocks; the Dinantian rocks of the Little Wenlock area have been overlapped here. In the Madeley Coalfield, Productive Coal Measures are known in the north, but have been removed in the south by the Symon Unconformity.

The North Staffordshire Coalfield has thick Productive Coal Measures (1500 m; Glover et al., 1993b), overlain by red beds (780 m; Glover et al., 1993b). The northward thickening of the Productive Coal Measures reflects the Pennine depocentre near Manchester (Calver, 1968). The more complete sequence of Productive Coal Measures, compared with that of Madeley, suggests that the Symon Unconformity was formerly a south-dipping palaeoslope.

Northern part of the Cheshire Basin

The Knutsford Borehole reached the base of the Permo-Triassic sequence at 2821 m, and penetrated a further 224 m of coals and limestones of the Upper Coal Measures which are probably part of the Newcastle Formation, equivalent to the Coed-yr-Allt Formation to the west of the basin. If this correlation is correct, the Keele Formation, equivalent to the Erbistock Beds, must have been eroded prior to Permian deposition. Rocks of the Etruria Formation, equivalent to the Ruabon Marl, would be expected below terminal depth in the borehole, with Productive Coal Measures at greater depth.

In the Lancashire Coalfield to the north and north-east of the Cheshire Basin, Productive Coal Measures are 1600 m thick, thinning westwards (Kirby et al., in press). The overlying red beds are about 320 m thick (Ramsbottom et al., 1978). Seismic reflection data (Figure 41) and (Figure 43) suggest that similar thicknesses extend southwards beneath much of the northern part of the Cheshire Basin.

Pre-Carboniferous rocks

The basement rocks which underlie the Permo-Triassic and Carboniferous strata of the Cheshire Basin are thought to be principally of early Palaeozoic age; Lower Palaeozoic strata were intersected in the Milton Green Borehole (where they underlie Carboniferous rocks) and in the Prees Borehole, as well as beneath the Dinantian outcrop to the west in several shallow provings in the Halkyn–Minera mining district of north-east Wales. Reflections from rocks of this age are visible on seismic profiles beneath terminal depth in the Milton Green Borehole. No Devonian rocks are known, but a borehole at Edgmond discovered Precambrian volcanic rocks beneath Coal Measures (Falcon and Kent, 1960). These rocks, probably of Uriconian affinity, continue in the subsurface to Lilleshall, the Wrekin and south-east to Caer Caradoc (Smith, 1987), and are probably responsible for some of the observed aeromagnetic anomalies (Figure 27).

South-west of the Cheshire Basin, Lower Palaeozoic strata outcrop widely and are cut by major north-east-trending Caledonian faults such as the Bala, Clun Forest, Pontesford–Linley and Church Stretton structures (Figure 26). Many of these features are associated with marked aeromagnetic anomalies (Figure 27), indicative of structural penetration to considerable depth. Reactivation of these structures played an important part in the subsequent structural evolution of the region.

Basin analysis

Basin subsidence history

Depth-of-burial study

The maximum depth of burial of the Mesozoic succession preserved in the Cheshire Basin has been estimated using a method that relies on sediment compaction characteristics. As sedimentary deposits are progressively buried, their porosity decreases and, consequently, their density increases. This relationship has been demonstrated by laboratory and theoretical work (Sclater and Christie, 1980; Baldwin and Butler, 1985) and observed in borehole logs (e.g. Lang, 1980). Compaction also depends on sediment type: for example, mudstones compact more quickly and to a greater degree than sandstones. Most importantly, the compaction process is irreversible: sediments do not decompact significantly upon uplift. Thus, once a set of standard compaction curves has been established, the maximum depth of burial of a sample of known lithology and density can be determined. An estimate of the thickness of overburden eroded from the sample site is then obtained by subtracting the present depth of the sample from its maximum burial depth.

The velocity of acoustic waves in rocks is also related to rock porosity (e.g. Lang, 1980), so it is possible to use sonic velocity measurements to estimate the depth of burial.

The reliability of these methods depends on the measured physical properties being due solely to sediment compaction. However, there are other factors that may alter the density or sonic velocity of sedimentary rocks, such as recrystallisation of limestones or changes in the cement of sandstones. Previous studies (e.g. Marie, 1975; Magara, 1976) indicate that the best estimates of depth of burial are obtained from argillaceous lithologies.

Compaction curves

The stratigraphical succession in the Cheshire Basin is not now at its maximum depth of burial, having undergone uplift and erosion during the Cenozoic (see p.45). In order to estimate the maximum depths of burial, standard compaction curves were plotted (Figure 53) for arenaceous and argillaceous lithologies, based on data from the Wessex Basin (Chadwick, 1985) and theoretical studies (Sclater and Christie, 1980; Baldwin and Butler, 1985). Constant densities were assumed for halite and anhydrite (Cermak and Rybach, 1982).

Borehole logs

Density logs from the Burford, Crewe Heat Flow, ICI RM 72A, ICI RM 73 and Prees boreholes (Figure 54) were used to estimate the maximum depths of burial and the thicknesses of eroded overburden for each formation intersected (Table 4). Many of the mudstone units in the Cheshire Basin contain significant thicknesses of evaporites, dominantly halite. In the Wych and Byley mudstone formations and the upper part of the Bollin Mudstone Formation, evaporite beds can be identified on density logs and can thus be eliminated. Elsewhere in the MMG however, evaporites are mixed with mudstones as lamellae and nodules, and histograms of density values show bimodal distributions. One peak corresponds to halite; the other, corresponding to the argillaceous matrix, was used for burial estimation.

Results obtained from the sandstone units were less consistent than those from the mudstones, possibly reflecting postdiagenetic alteration of cement. At the Burford Borehole, where a porosity log is available for the SSG, the sandstones have porosities between 6 and 24%, which may produce an underestimate of the depth of burial. Other units are conglomeratic (e.g. the Chester Pebble Beds) and, despite the use of histograms to distinguish between pebble and matrix densities, are less likely to yield reliable results.

Cross-plots of sonic and density logs were made for each formation intersected by the Prees, Burford and Crewe boreholes, the only boreholes where both types of log were available. The cross-plots for the Lower Lias, Brooks Mill Mudstone, Bollin Mudstone, Malpas Sandstone, Helsby Sandstone, Wilmslow Sandstone and Kinnerton Sandstone formations at Prees and Burford showed clear linear trends, and best-fit lines were constructed to convert sonic-log values from other boreholes to density values, enabling the depth of burial to be estimated. The cross-plot for the Bollin Mudstone at Crewe did not show a linear trend, and density values estimated from it were much higher than those obtained from the Prees and Burford cross-plots. This may indicate a calibration error in the Crewe sonic-log data (the density log gives a sensible value for halite and is therefore thought to be correctly calibrated). Cross-plots for the Wilkesley Halite, Northwich Halite, Wych and Byley Mudstone, Tarporley Siltstone and Collyhurst Sandstone formations showed too much scatter to enable reliable trend lines to be plotted.

Sonic logs from the Blacon East, Codsall, Elworth, Knutsford, North Stafford and Ranton boreholes (Figure 54) were processed using a method similar to that used for density logs, to calculate mean and modal sonic transit times for each of the formations for which reliable cross-plots were available. These were then converted to equivalent density values using the sonic-density cross-plots and averaged (Table 5).

Both observed and sonic-derived density values were used to estimate maximum depths of burial from the standard compaction curves (Figure 53). The thicknesses of eroded overburden were then calculated by subtracting the present-day depths. Results are summarised in (Table 4) and (Table 5). (Figure 54)a shows the adopted eroded overburden thicknesses, based mainly on figures from the (probably more reliable) mudstone units.

The reliability of the estimated values for overburden thickness was tested by adding them to the observed depth to the base of the Bollin Mudstone Formation in each borehole and contouring the resulting data. This yielded a smooth surface indicating that the estimates are, relative to each other, consistent. There may nevertheless be systematic error in the uplift values because of the difficulty in obtaining appropriate standard burial curves.

Palaeotemperatures calculated from apatite fission-track data have been interpreted (Lewis et al., 1992; Green et al., 1993) to indicate overburden thicknesses in the range 2700–3300 m over the basin. Holliday (1993) and Chadwick et al. (1994b) have argued that, for various technical reasons, those authors overestimated the depths of burial. If these factors are taken into account the fission-track palaeotemperatures are in broad agreement with the results of the present study.

A gridded isopach map of the eroded overburden was generated using ISM, and displayed in EarthVision (Figure 54)b. The estimated overburden had a maximum thickness of more than 2000 m around the Burford and Crewe boreholes – significantly greater than elsewhere in the basin. The structure contour map for the base of the MMG (Figure 34) also shows a pronounced structural culmination in this area, suggesting that basin inversion has resulted in preferential uplift of the basin depocentre in the hangingwall block of the WBRRFS. Basin inversion was probably accomplished, at least in part, by reversal, locally in excess of 500 m and possibly up to 1000 m, of the WBRRFS.

It is possible to restore the basin to its configuration at the time of maximum burial by adding the eroded-overburden grid to the present-day depth grids (Figure 55). At this time, the basin had a sedimentary fill locally in excess of 6000 m thick. The base of the SSG lay at depths locally greater than 5000 m, the base of the MMG at depths locally greater than 3000 m and the base of Jurassic strata at depths locally greater than 2000 m. A single depocentre in the hangingwall block of the WBRRFS is apparent (Figure 55)d. The presence of two apparent depocentres in the present basin (Figure 36) appears to be an effect of later basin inversion. Regional tectonic considerations and apatite fission-track data (Lewis et al., 1992) suggest that the maximum depth of burial probably occurred about 60 Ma ago in Palaeocene times, immediately prior to Cenozoic uplift, inversion and erosion.

Backstripping results

The BGS basin-modelling program HOTPOT (Rowley et al., 1993) reconstructs the depositional history of a basin from the present-day basin fill by successively stripping away one layer at a time to restore, by decompaction, the thicknesses of underlying layers. Layers representing eroded material can be incorporated into the backstripping process. The layer information used for backstripping is summarised in (Table 6). The lithological compositions of the preserved layers were estimated from the logs of the Prees and Wilkesley boreholes (Figure 2), while the lithologies of eroded layers were based on those of Chadwick et al. (1994b). The ISM structure-contour grids for Base Jurassic, Base MMG, Base SSG and Base Permo-Triassic (p.47) were thinned to a grid-node spacing of 1 km to reduce the data volume to the limits required by the HOTPOT program. The present-day topography was assigned a single basin-wide value of 120 m above OD. Isopach maps for the present-day preserved thicknesses of the layers are given in Chadwick et al. (1994a); results from the maps are discussed below.

In order to provide a full backstripped history it was necessary firstly to divide the eroded overburden into its component stratigraphical units. This is a poorly constrained exercise, based principally on regional subsidence patterns. The total thickness of eroded overburden strata (Figure 54)b, assumed to have been deposited prior to 60 Ma ago (mid- Palaeocene times), comprises three components: Upper Cretaceous and Palaeocene (principally the Chalk), Jurassic (with possible lower Cretaceous), and Permo-Triassic. The Upper Cretaceous and Palaeocene component was assumed to have formed a uniform blanket 400 m thick over the whole area (consistent with it having been deposited in the post-extensional phase of basin subsidence). The combined thickness of the Triassic and Jurassic components was thus calculated by subtracting 400 m from the total-overburden grid. The fact that no Triassic strata were eroded where Jurassic rocks are still preserved and the assumption that preserved Jurassic strata (after late Cimmerian erosion) would have been of negligible thickness towards the northern and western limits of the basin, were used to constrain the Permo-Triassic and Jurassic components of the eroded succession. It is not possible to subdivide the eroded succession further.

Selected maps of the backstripped sequences are shown in (Figure 56). Preserved thicknesses of Permian strata are locally in excess of 1300 m (Chadwick et al., 1994a). Initial thicknesses (decompacted) were greater than this (Figure 56)a, locally in excess of 1800 m, and define a north-east-trending depocentre lying in the hangingwall block of the WBRRFS. Thicknesses decrease gradually north-west from this depocentre, so, at the end of Permian times (Figure 56)a, the Cheshire Basin had the overall form of a half graben. Minor local depocentres lay in the southern and northern parts of the Wem–Audlem Sub-basin, and the Sandbach–Knutsford Sub-basin, with evidence of syn-depositional fault control.

Preserved thicknesses of the SSG (Chadwick et al., 1994a) appear to define two depocentres in the southern and northern parts of the Wem–Audlem Sub-basin, which locally have more than 2200 m of strata. A third depocentre lies in the Sandbach–Knutsford Sub-basin with over 1800 m of preserved strata. Strong syn-depositional fault control is clearly evident. At the end of SSG deposition the basin had a maximum depth in excess of 3500 m (Figure 56)b.

The MMG is presently thickest in the Wem–Audlem and Sandbach–Knutsford Sub-basins (Chadwick et al., 1994a). In the former a complete sequence is preserved, locally in excess of 1500 m thick. In the latter, although the sequence is less complete, it is locally more than 1700 m thick, suggesting that the area marked the original MMG depocentre. Lesser preserved thicknesses (<1300 m) between the two sub-basins reflect the effects of later erosion related to inversion (see p.70) rather than depositional thinning. Separate depocentres, characteristic of the earlier basin evolution, were no longer so clearly defined, and it seems likely that there was a single depocentre for the MMG. End-Triassic depths (Figure 56)c cannot be tightly constrained, because they incorporate a putative amount of eroded Permo-Triassic strata. Nevertheless, it is estimated that at the end of Triassic times the basin contained more than 5000 m of strata. It is likely that Triassic strata (particularly the MMG) spread considerably beyond the present-day boundaries of the basin, onto adjacent block areas.

Only small outliers of Jurassic strata are preserved. They lie in the central part of the Wem–Audlem Sub-basin, to the south of the MMG depocentre, and it is thought unlikely that they correspond precisely to the Jurassic depocentre. Their preservation probably reflects the fact that there was less uplift and erosion in this area than further north in the basin.

Sedimentary thicknesses (Figure 56)d are likely to have been locally greater than 6000 m immediately prior to deposition of the Chalk (97 Ma ago), but estimates are poorly constrained, incorporating a putative amount of Permo-Triassic, Jurassic and lower Cretaceous overburden. The tendency towards a single basin depocentre clearly continued in Jurassic times, as fault-controlled subsidence gradually diminished and post-extensional regional subsidence prevailed.

Sedimentary thicknesses in early Palaeocene times, incorporating the total estimated thickness of eroded strata, locally approached 6500 m (Figure 56)e, and marked the maximum development of the basin, immediately prior to uplift and erosion.

A single-node burial-history plot was prepared using the HOTPOT model for a location in the Wem–Audlem Sub-basin where the most complete sequence is preserved (Figure 57). The onset of measurable basin subsidence corresponds to the deposition of the oldest preserved strata in the basin (the Collyhurst Sandstone). The age of these beds is poorly constrained, estimates ranging from 300 to 260 Ma, but an age of 265 Ma is here assigned to them following Chadwick and Evans (1995). Late Cimmerian (early Cretaceous) uplift is almost wholly unconstrained and is not modelled here (the true subsidence is likely to have been somewhat greater at the end of the Jurassic (144 Ma) than shown, followed by uplift to 97 Ma, before renewed subsidence in the late Cretaceous). Nevertheless, the main aspects of basin subsidence can be identified. Very rapid (fault-controlled) early Triassic subsidence marked the main phase of basin development. This was followed by more gradual subsidence in Jurassic and Cretaceous times, followed in turn by Cenozoic uplift. Burial-history plots for other locations in the basin are given in Chadwick et al. (1994a).

Estimation of the super-regional extension factor

Subsidence analysis is a powerful means of estimating crustal extension factors, because crustal thinning and subsidence are related by isostasy. For a locally (Airy) isostatically balanced basin:

(Equation 1)

Average basin subsidence S = T_c(1 – 1/γ)(ρ_c-ρ_m)/(ρ_{s –} ρ_m)

where

ρ_c = density of crust (2.80 gcm⁻³)

ρ_s = density of basin fill (l.03 gcm⁻³ for a sediment-starved basin filled only with sea water)

ρ_m = density of mantle (3.33 gcm⁻³)

T_c = Thickness of crust prior to extension (31.2 km)

γ = crustal extension factor

Deep-seismic-reflection data suggest that on the UK continental shelf small sedimentary basins are not fully isostatically compensated by crustal thinning, being, to some extent, supported by the flexural strength of the lithosphere. Thus, an accurate estimate of crustal thinning cannot be obtained by analysing the subsidence of an area as small as the Cheshire Basin. Isostatic balance is, however, achieved on a regional scale, typically over distances of about 100 km (e.g. Warner, 1987). A super-regional subsidence model was therefore developed, comprising a 250 X 250 km square with the Cheshire Basin at its centre, over which the total Permian to Palaeocene sedimentary thickness (corresponding to the maximum burial) was estimated (Chadwick et al., 1994a). The total sediment thickness outside the Cheshire Basin was estimated from the thicknesses of preserved sediment (from British Geological Survey, 1985b) and estimated thicknesses of eroded strata (c.2000 m over the East Irish Sea Basin, decreasing to c.1000 m over the adjacent massifs). HOTPOT was used to correct for sediment loading by normalising sediment thicknesses to equivalent thicknesses of sea water, giving the sediment-starved (tectonic) subsidence over the super-region (Figure 58); this has a mean value of 888 m. Using this value in the Airy isostatic equation (above) gives a super-regional crustal extension factor g = 1.14. Thus, the amount of post-Variscan crustal extension in the super-region was about 14%. This is consistent with regional 2-D gravity modelling (Chadwick et al., 1994a) which suggests that the crystalline crust may have thinned from a nominal initial thickness of 32.5 km, to about 28 km beneath the basinal areas.

Fault analysis

Basement-fault reactivation and regional extension direction

As discussed above, the structures of the Cheshire Basin were, at least in part, controlled by the reactivation of older, basement faults. This is typified by the WBRRFS, at the south-east margin of the basin, details of which are revealed by seismic data (Figure 46). The large down-west normal faults of the WBRRFS lie in the hangingwall blocks of Variscan reverse faults, the base of the Permo-Triassic succession being progressively faulted down to the west. This was a consequence of extensional reactivation of the basement reverse faults in Permo-Triassic times which led to collapse of their hangingwall blocks by upward propagation of steeper normal faults, to form the basin margin. This type of thrust reactivation, involving hangingwall collapse, is a well-documented feature of the Mesozoic basins of the UK (e.g. Chadwick et al., 1983; Stein, 1988). A few km to the south (Figure 38), Section 4, a similar situation is apparent, the Wem Fault probably forming by extensional reactivation of an older fault (Figure 28). These two examples, though differing in detail, suggest that development of the whole south-east margin of the basin was controlled by reactivation of underlying basement structures, which were either directly or indirectly related to the Clun Forest Disturbance and the Pontesford– Linley Fault (Figure 26).

On a more regional scale, the Permo-Triassic rift system of southern Britain can be divided structurally into three segments which reflect the structural grain of the underlying basement. Thus, in the English Channel and Wessex basins, basin-controlling faults have a dominant east–west (Variscoid) trend; in the Midlands (the Worcester Basin) the faults have a dominant north–south (Malvernoid) trend and in north-west England (the Cheshire Basin) north-east (Caledonoid) basin-fault trends are prominent (Chadwick and Evans, 1995). Because the orientations of the basin-margin faults were controlled by basement structures they were not necessarily perpendicular to the regional extension direction and consequently suffered varying degrees of oblique-normal displacements. An analytical model developed by Chadwick and Evans (1995) relates the displacement components of a set of variably oriented faults to a single regional extension direction. Application of this model to the marginal faults of the rift basin indicated that faults trending north–south had displacements which closely approximated to dip-slip, while faults trending east–west suffered markedly oblique-slip displacements. This is interpreted as indicative of a cumulative Permo-Triassic extension vector oriented roughly west-southwest.

Extension direction from fault orientations

The strikes of all faults cutting the Base Permo-Triassic surface were measured using the 1:100 000 scale compilation maps (Chadwick et al., 1994a). Long faults were divided into individual segments of uniform strike. In total, 633 faults or fault segments were measured and the results plotted on a rose diagram (Chadwick et al., 1994a). Fault strikes are heavily biased towards an approximately north–south direction. Some 55% of fault strikes lie between 330° and 010°, and 17% of faults have the commonest (modal) strike of 350–360°. In contrast, east–west fault strikes are notably rare.

There is a systematic geographical variation of fault orientation (e.g. (Figure 38)." data-name="images/P1000291.jpg">(Figure 37). In the southern half of the basin, faults are dominantly north-east-trending, whereas over much of the central and northern parts of the basin they strike mainly north–south. In the far north-east a north-north-west trend is prominent. In order to analyse the differing trends more fully, separate rose-diagrams were plotted for the north-eastern, central and southern areas of the basin (Figure 59).

In the southern area 206 fault-strikes were measured (Figure 59). Trends are dominantly north-north-east to north-east, with 71% of fault-strikes between 0° and 060°. These trends are roughly parallel to the Wem and Bridgemere faults. A degree of Caledonian basement control is indicated, consistent with the reactivation of basement faults at the basin margin. Statistically, however, the observed trends are somewhat unusual, in that they do not have a normal distribution about a mean direction. The commonest (modal) fault strike (16%) is 0–10°. This lies at the extreme limit of the dominant range, giving a strongly skewed overall distribution. This may be due to a lack of adequate constraints on fault strikes in this area, resulting from the sparse seismic cover. Trends could not be accurately determined for the many small faults which intersect only one seismic line. For these faults either the surface geology or the gravity shaded-relief maps (Chadwick et al., 1994a) were used to indicate the dominant fault trend.

In the north-eastern area 95 fault strikes were measured (Figure 59). The dominant north-north-westerly trend is clearly evident, 60% of fault strikes lying between 310° and 350°, with a modal direction (19%) of 330–340°. A subsidiary peak at 290–300° relates principally to poorly constrained small faults near the northern edge of the basin and may be an artefact. The dominant fault strike is not parallel to the main Caledonoid trend, and may reflect a reactivated Carboniferous structural trend: a strong north-north-westerly trend is evident in the Carboniferous rocks of the Rossendale area to the north, visible on the aeromagnetic data (Figure 27), and also manifest as large north-north-westerly trending faults (Kirby et al., in press).

The majority of the faults in the basin are in the central area. Faults are present on all scales ranging from very large structures at the basin margin through major intra-basin faults to small structures of limited lateral extent. In all, 332 fault strikes were determined, giving a well-defined pattern (Figure 59). The main range of fault strikes is very narrow, 76% of faults falling between 330° and 010°, with a modal direction (26%) of 350–360°. The single dominant direction is estimated at 353°. The roughly uniform distribution of fault strikes either side of this direction indicates that the dataset is well constrained. The trends of the major basin-controlling faults may reflect underlying basement structure, but the multitude of intermediate and small intra-basin faults are more likely to have formed as dip-slip normal faults, perpendicular to the dominant extension direction which is therefore estimated at 083–263°. A small but significant peak occurs in the fault distribution at a trend of 265–270° (Figure 59), perpendicular to the dominant fault trend and corresponding to several small faults cross-cutting the dominant north–south faults, sometimes offsetting them and commonly demarcating adjacent areas of differing structural style. These faults are interpreted as transfer faults, parallel to the extension direction. This trend, albeit poorly constrained, is consistent with the extension direction derived from the dip-slip faults.

The observed fault trends in the Cheshire Basin thus show evidence of basement control, particularly at the south-east margin and, more generally, in the south of the basin. Most of the faults, however, particularly the small intra-basin structures, appear to have formed perpendicular to the extension direction, which is estimated at 083–263° (though the north-westerly fault trends seen in the north-east of the basin suggest the possibility of north-easterly directed extension). Results are generally in good agreement with the regional Permo-Triassic extension direction in southern Britain, estimated by Chadwick and Evans (1995) using different structural criteria at 075–255°. The extension direction may not have remained constant throughout the evolution of the basin. Fault-strike analysis probably reveals the early extension direction which initiated basin development and determined the overall structure of the basin. Subsequent, different, extension directions may have simply reactivated existing faults, in oblique-slip modes. Regional tectonic considerations (see p.44) indicate that roughly east–west extension was probably dominant throughout the Permo-Triassic and early Jurassic, the period which encompassed most of the extensional faulting.

Estimation of extension factor by fault-displacement analysis

The heave of a fault is defined as the horizontal displacement of a reference horizon across that fault, measured perpendicular to the fault. For a dip-slip normal fault, therefore, the heave is equal to the horizontal displacement across the fault in the extension direction, and provides a measure of the extension across the fault. In the more general case of an oblique-slip fault, the apparent heave can be defined as the horizontal displacement of a measured horizon across the fault in the extension direction. Thus by summing the apparent heaves of faults which cut a cross-section parallel to the extension direction, a measure of the total extension can be obtained.

Fifteen equally spaced transects AA to OO were chosen, parallel to the assumed extension direction (083–263°) and crossing the basin so as to encompass the Permo-Triassic outcrop and the basin-margin faults (Figure 60). The apparent heaves on all faults intersected by each transect were measured from the 1:100 000 photoreduced structure-contour maps (Chadwick et al., 1994a), and summed to give a cumulative apparent heave (Table 7). The transects vary in length from 32.0 km to 64.2 km, with cumulative heave values in the range 0.8 to 5.9 km. The cumulative heaves on transects AA to JJ are marked with an asterisk signifying that they are minimum values, because on these transects the eastern basin margin includes faults (principally the Red Rock Fault) with pre-Permian basement rocks at outcrop in the footwall blocks. The true heaves of these faults at the base of the Permo-Triassic cannot be seen; observed heaves are therefore minimum estimates. The cumulative apparent heave on each transect gives a first-order estimate of the extension along that transect.

However, estimates of extension based on the sum of observed fault heaves tend to be too low. A possible explanation for this lies in the presence of small faults, beneath the seismic-resolution threshold, which together can account for substantial amounts of hidden extension (e.g. Walsh et al., 1991). In the basins of the North Sea, for example, extension estimates based on restoration of (observed) faults are significantly less than estimates derived from other evidence such as subsidence analysis (Sclater and Shorey, 1989; Ziegler and Van Hoorn, 1989).

Several recent studies suggest that fault populations in a particular basin are consistent across a wide range of scales. In other words, the relative number of large and small faults remains the same at all scales. This scale-invariance is known as a fractal, or power-law, size distribution (e.g. Kakimi, 1980; Marret and Allmendinger, 1992; Walsh and Watterson, 1992). Thus, the hidden extension due to small faults can be addressed statistically by applying fractal scaling laws which quantify scale-invariant phenomena.

The distribution of measured fault heaves can thus be used to estimate hidden extension. This approach has the advantage that heaves and extension are directly related:

ΔΕ_TOTAL = ΣΔh_resolved + ΣΔh_hidden

where ΔΕ_TOTAL is the total extension on a set of faults

Δh is the heave on each fault

A major disadvantage of analysing measured heave data, however, is that the heaves of small faults, whose distribution critically affects the prediction of hidden extension, are extremely difficult to measure accurately from seismic reflection data. Seismic data are much more amenable to the accurate measurement of small fault throws. In the analysis described below, therefore, the throw distribution of faults is the principal analytical tool. Thus:

TOTAL = Σt_resolved +Σt_hidden

where;

T_TOTAI. is thecumulative throw on a set of faults

t is the throw on each fault

Following Marrett and Allmendinger (1992), if a fault population follows a fractal size distribution, the cumulative number of faults (N) with throw greater than or equal to t may be written:

N = t^−c i.e. log₁₀N = –c.log₁₀t

where c is the power-law exponent which characterises the relative number of large and small faults.

Fault-displacement analysis was carried out by examining the extension-parallel transects (Figure 60), which give multiple 1-D sampling of the fault population. For each transect, the throw of each intersecting fault was measured and assigned to a specific category. (Table 8) shows the number of faults in each throw category seen on each transect, the total number of faults in each category, and the cumulative number of faults with throws greater than each category boundary. It is evident that small faults are much more abundant than large faults. Thus, only 2 faults have throws greater than or equal to 2500 m, 19 have throws of 1000 m or more, 148 have throws of 100 m or more and 355 (all of the observed faults intersecting the transects) have throws greater than the lower limit of seismic resolution (typically about 35 m). By plotting this cumulative number data on a log-log plot, fractal size distribution can be identified by points which lie on a straight line whose gradient is equal to –c (Figure 61).

Thus, roughly, fractal behaviour is shown by faults which have throws in the range 100–1250 m, such faults defining a straight line of gradient –0.90. Faults outside this range however, appear to deviate somewhat from the ideal fractal distribution.

There are two main reasons why large faults (throws 1250 m) may depart significantly from the linear fractal trend. Firstly, observed throws on some of these faults, for example the Red Rock Fault, represent minimum estimates (due to the lack of preserved strata in the footwall block and also to the effects of subsequent Cenozoic reversal), which are significantly less than their true original throws. If it were possible to correct present-day observed throws to true original throws, the result would be to move the plotted locations of large faults to the right, closer to the fractal trend. Secondly, very large faults may depart from ideal fractal behaviour, because of mechanical difficulties connected with their formation. The fractal trend predicts the existence of a small number of very large faults, but faults of this size may develop less often than predicted because displacement tends to transfer to new faults when the throw has reached a certain critical value (dependent on the angle of the fault and the physical properties of the faulted medium). The large faults that form the eastern margin of the Cheshire Basin may be examples of this process.

Measured small faults (throws 5100 m) also depart somewhat from the linear fractal trend, though not to the same extent as the very large faults. This is probably the result of inadequate sampling. Small faults are likely to be under-sampled, either because they are of limited length (e.g. Walsh and Watterson, 1987) or because they are close to the limit of seismic resolution and went unnoticed by the interpreter.

It is concluded therefore, that the measured fault throws (Figure 61) exhibit an approximate fractal size distribution for a range of throws from c.100 to c.1250 m, with a likely fractal distribution down to the effective lower limit of seismic resolution (c.35 m).

The cumulative throw due to displacements on all faults can therefore be estimated from the N largest faults (i.e. the resolved faults):

T_TOTAL = t₁ + t₂ + t₃ + …….. + tN + Σt_hidden

For 1-D sampling, the hidden cumulative throw for small faults can be determined by extrapolating the fractal size distribution (Marrett and Allmendinger, 1992):

(Equation 1)

For this expression to be useful, the fault population must be such that c is less than unity. This allows Σt_hiddentoconverge, with the implication that the largest faults in the population account for most of the cumulative throw. Thus, we can solve for Σt_hidden beneath a chosen throw threshold, by substituting the following values into the above equation:

t_N = 100 m (minimum throw of fully sampled faults exhibiting fractal behaviour)

N = 134 (number of faults exhibiting fractal behaviour with throws ≥100 m)

c = 0.90

Thus:

Σt_<100 = 120 500 m

where Σt_<100 is the cumulative throw of all faults with less than a 100 m throw, likely to have intersected all of the analysed transects.

To estimate the hidden extension, the cumulative heave of the small faults (rather than their cumulative throw) must be calculated. The dips of the small, but seismically resolved, faults (throws typically <500 m) on transects AA–OO fall in the range 60–80°, with a mean of 70°; the smaller faults tend to have dips at the upper end of the range. A value of 75° was therefore taken as a representative dip of all faults with throws <100 m. Thus the cumulative heave, ΣΔh_<l00 of these faults can be calculated:

ΣΔh_<l00 = t_<100/tan75

from which ΣΔh_<l00 = 32.3 km.

This 32.3 km of heave attributable to faults with throw <100 m was apportioned between the 15 transects AA-OO as follows. First, the total heave attributable to faults with throws __100 m was measured for each transect (Table 9) and then summed for all transects. The resulting cumulative heave attributable to all faults with throws .1.00 m was 44.6 km. Thus the cumulative heave on faults of throw <100 m accounts for 32.3/44.6, or 72%, of the cumulative heave on faults of throw 100 m, and for 32.3/(32.3 + 44.6), or 42%, of the cumulative heave on all faults (i.e. the assumed total extension). The latter figure falls within the range of 25–60% quoted by Marret and Allmendinger (1992) and is very similar to the 40% estimated by Walsh et al. (1991) in two independent studies of faults in the Viking Graben.

The essential condition for this approach is that the fractal relationship displayed by the whole fault set is appropriate throughout the basin. In detail, this may not be the case. Work done by Chadwick (1997) on individual transects suggests that two fractal fault populations may be present. However, because of the limited number of faults intersected on each transect, this is subject to significant statistical uncertainty. For simplicity, therefore, a single basinwide fractal relationship (c = 0.90) was applied to each transect individually. Thus, for each transect the total heave on faults with throws <100 m was computed from the measured total heave on faults of throw .100 m (Table 9), to give the total heave attributable to all faults and hence the total extension.

When corrected for the contribution of small faults, total heaves along transects AA-OO are in general considerably greater than the heaves derived from only seismically resolved faults (Table 7). They are typically more than 4 km, and more than 8 km in the central part of the basin (transects GG and HH). The corrected total heave on transect AA of 0.7 km, is slightly less than the observed total heave (0.8 km) because of over-estimation, in the latter figure, of the heaves of very small faults. The corrected heave, of course, does not incorporate heave measurements on faults with throws less than 100 m, these being accounted for in the fractal extrapolation.

The crustal extension factor g can be computed for each transect, where

g = extended section length / pre-extension section length

Extension factors vary from 1.02 in the far north of the basin to about 1.12 in the south, with a peak of 1.15 in the central part (Table 9). These values are comparable to the super-regional extension factor derived from the subsidence analysis (p.75). It would be expected that extension factors derived from analysing the Cheshire Basin itself (presumably a site of enhanced extension), would be significantly larger than the super-regional extension factor, since the latter includes interbasin massifs which are presumably only weakly extended. However, as discussed earlier, the extension factors derived from the fault analysis are minimum estimates, because the observed heaves on the major basin-bounding faults are usually significantly less than the true original heaves (due to lack of preserved correlative strata on the footwall blocks and the effect of subsequent reversal during basin inversion; see p.70). These effects are difficult to quantify, but may result in under-estimates of several hundred metres of heave along each transect. If these effects could be accounted for, the fault-derived extension factors from the basin would probably significantly exceed the subsidence-derived super-regional extension factor.

Development of extension with time

Thus far, the fault analysis has concentrated on displacements at the base of the Permo-Triassic, which indicate the total extension undergone by the Cheshire Basin. Fault heaves at different horizons can be used to determine the incremental development of the basin as it extended. Because of the progressively smaller outcrop areas of horizons above the base of the Permo-Triassic (Figure 33), (Figure 34), (Figure 35), (Figure 36) it is not possible to use all of the 15 extension-parallel transects. Neglecting the Base Jurassic level, which is of limited extent, the Base MMG has the smallest area of outcrop. Segments of transects DD-LL (termed DD* -LL* ) were chosen to encompass the Base MMG outcrop and its associated faults (Figure 60). Heaves were measured for all the faults on DD*-LL* at Base Permo-Triassic, Base SSG and Base MMG (it was not necessary to apply the fractal extrapolation techniques described above, since relative rather than absolute amounts of extension were being determined). Total heaves at each stratigraphical level are given in (Table 10). Despite considerable variation between transects, a systematic pattern of incremental extension is apparent. On each transect, observed heaves for Base Permo-Triassic are greater than those for Base SSG, which in turn are greater than those for Base MMG. Differences in heave proportions from one transect to another in part reflect displacements on different fault segments at different times, but may also be due to problems in measuring and sampling faults. The cumulative profile indicates that measured heaves at Base MMG and Base SSG are typically 56% and 80% of the total observed heaves. (Table 11) shows these percentage heaves expressed as cumulative relative displacements. As sediments are deposited across a fault, the effect of differential compaction is to amplify fault displacements at higher stratigraphical levels relative to deeper displacements, giving a false measure of true displacements. This effect was analysed by examining present-day and backstripped (decompacted) displacements across the King Street Fault, a typical intrabasin structure, at a point where its displacement history approximately matches the cumulative relative displacements of the faults analysed. Using the King Street Fault as an analogue for the cumulative fault behaviour, correction for compaction was applied to the observed cumulative relative displacements (Table 11).

Using these cumulative relative displacements it is possible to reconstruct the incremental extension along each transect (Table 10) and thereby the extension history of the basin. Basin extension in relation to displacements on the eastern margin faults is shown in (Figure 62). The steeply dipping north-east-trending Wem and Bridgemere faults suffered oblique-normal displacements (sinistral transtension), forming a partial transform south-east margin to the basin (characterised by 'flower-structure' fault geometries). The moderately dipping Red Rock and Alderley faults trend more north-south and experienced near dip-slip normal displacements. Maximum extension occurred close to the basin depocentre cf. (Figure 55)d. Extension at the northern end of the basin (Table 9) diminishes markedly, suggesting that the present northern margin may have been close to the margin of the basin throughout its evolution. It is possible that north of the basin, regional extension was transferred westwards into the West Lancashire and East Irish Sea basins. In contrast, extension factors at the southern end of the Cheshire Basin remain roughly uniform at about 1.12, suggesting that, before uplift and erosion, the basin extended some considerable distance to the south, its present southern limit being a much later exhumed feature of no particular depositional significance.

Thermal modelling

The BGS HOTPOT basin-modelling software (Rowley et al., 1993) integrates the depositional history of a basin with thermal parameters (heat flow, surface temperature and thermal conductivity) to calculate the thermal history of the basin. The method assumes 1-D vertical conductive heat transfer through the basin, with heat flow and surface temperature allowed to vary with time.

The thermal conductivity of sedimentary rocks depends mainly on lithology but also increases systematically with depth of burial, in a manner similar to density. For this study, a set of thermal conductivity v. depth v. lithology curves (Chadwick et al., 1994a), derived from the almost complete Mesozoic to Neogene sequence in Denmark (Balling et al., 1981) were used to calculate thermal-conductivity values for each backstripped layer. Thermal conductivity also varies with temperature (e.g. Cermak and Bodri, 1986), and the HOTPOT program iteratively adjusts the interpolated thermal conductivities as it computes temperatures.

Present-day subsurface temperatures

The present-day average annual surface temperature of the Cheshire Basin area is approximately 10°C.

The heat-flow measurements used to compile the BGS Heatflow Map of the UK (Gebski et al., 1987) were supplemented for this study by data supplied by K E Rollin (personal communication). Observations are too sparsely distributed and laterally variable to produce a reliable large-scale heat-flow map of the area. Instead, they were analysed statistically and plotted as a histogram (Chadwick et al., 1994a). This showed a normal distribution of heat-flow values with a mean value of 52 mWm⁻² and a standard deviation of 12 mWm⁻², indicating a true heat flow in the range 40–64 mWm⁻² with a 70% confidence limit. Three present-day models were computed, using heat-flow values of 40, 52 and 64 mWm⁻².

Present-day temperature maps for the bases of the Jurassic, MMG, SSG and Permian, computed using the mean present-day heat flow of 52 mWm⁻² (Figure 63), indicate that the basin is quite cool, with mean temperatures ranging from 20°C at the base of the MMG to 37°C at the base of the Permo-Triassic. The highest temperatures at the base of the Permo-Triassic in the deeper parts of the basin locally exceed 70°C in parts of the Wem–Audlem Sub-basin and 65°C in the Sandbach-Knutsford Sub-basin. Calculated temperatures near the Alderley Edge mineralisation range from 25°C at the base of the SSG to 33°C at the base of the Permo-Triassic.

Decreasing the heat flow to 40 mWm⁻² depressed mean basin subsurface temperatures by about 3°C at the base of the MMG, and 6°C at the base of the Permo-Triassic. Increasing the heat flow to 64 mWm⁻² raised the mean temperatures by about 2°C at the base of the MMG and 7°C at the base of the Permo-Triassic. The corresponding temperature changes for the deepest parts of the Permo-Triassic sequence were respectively −15°C (for a heat flow of 40 mWm⁻²) and +15°C (for a heat flow of 64 mWm⁻²).

Palaeo-subsurface temperatures

In order to compute palaeo-subsurface temperatures it is necessary to take into account possible heat-flow variations with time. A simple model for predicting heat flow during basin evolution is McKenzie's (1978) hypothesis of instantaneous uniform lithospheric extension, which predicts that instantaneous extension of the lithosphere by a factor b causes simultaneous thinning of the lithosphere by a factor 1/b. Extension does not vary with depth, the crustal extension factor being the same as the extension factor in the lithospheric mantle. Lithospheric extensional thinning is accomplished by elevation of the lithospheric isotherms and the development of a positive thermal anomaly, heat flow increasing instantaneously by a factor b. The subsequent decay of this thermal anomaly can be calculated from the one-dimensional heat-flow equation (McKenzie, 1978).

Evidence from deep seismic reflection data (e.g. Klemperer and White, 1989; White, 1990) broadly supports McKenzie's hypothesis, indicating that, on a regional scale, the lithosphere stretches by pure-shear extension. Bearing in mind the regional nature of isostasy, it is possible to estimate the lithospheric extension factor b, and thereby the heat flow, from a regional assessment of basin subsidence. The mean tectonic subsidence in the Cheshire Basin super-region (Figure 58) is 888 m (corresponding to the maximum depth of burial, at c.60 Ma), consistent with a regional crustal extension factor (γ) of 1.14 (p.75). The extension that led to this cumulative extension factor was distributed over a long period of time, from latest Permian times to the early Cretaceous. For the purposes of this study a single, instantaneous pulse of extension occurring at the beginning of the Triassic (251 Ma) is assumed. (Figure 64) illustrates the subsidence that would result from this pulse of extension, with an extension factor of β =1.14. The predicted tectonic subsidence at c.60 Ma, corresponding to the maximum depth of burial in the super-region, very closely matches the observed value of 888 m.

Heat-flow variation with time, constrained by the present-day observed mean value of 52 mWm⁻², can be predicted from the extensional model (Figure 64). Prior to extension, the heat flow is estimated to have been 51.9 mWm⁻², comprising a sub-lithosphere component of c.33 mWm⁻² (McKenzie, 1978) and a lithospheric/crustal component of c.18.9 mWm⁻² caused by radioactive heat-producing elements which occur preferentially in the upper crust. The fact that these heat-producing elements are concentrated strongly in the uppermost few kilometres of crust means that deep lithospheric temperatures are not significantly different from those assumed by the McKenzie (1978) model. As a result of extension, heat flow increased instantaneously to 56.5 mWm⁻² and decreased gradually thereafter to 52 mWm⁻² at the present day.

There are few estimates of average surface temperatures for Britain in the past. Curry (1992) gives a value of 20°C for the early Palaeocene, corresponding to the maximum depth of burial in the region. Prior to this, surface temperatures are poorly constrained. Since the climate was either tropical or desert throughout the Mesozoic, temperatures were clearly much higher than today. For this study a value of 20°C is assumed.

Palaeotemperature maps

HOTPOT was used to compute palaeotemperatures for the full basin history, based on the palaeo-heat-flow variation illustrated in (Figure 64) (note that this is contrained by the mean value of present-day heat flow - the above-mentioned uncertainty limits still apply).

Computed temperatures at the base of each layer 60 Ma ago (early Palaeocene), when the Cheshire Basin was at its maximum depth of burial are shown in (Figure 65). The temperature of the basin fill was much higher than at present, with mean values ranging from c.80°C at the base of the MMG to c.95°C at the base of the Permo-Triassic. In the Wem–Audlem Sub-basin, peak temperatures reached c.90°C locally at the base of the Jurassic sequence, c.105°C at the base of the MMG, c.130°C at the base of the SSG and c.140°C at the base of the Permo-Triassic. Corresponding values in the Sandbach-Knutsford Sub-basin were Base MMG c.95°C, Base SSG c.120°C and Base Permo-Triassic c.135°C. The Alderley Edge area was, at this time, assumed to be buried under nearly 1500 m of Chalk, Jurassic and upper Triassic strata (subsequently eroded). The computed temperatures at the base of this overburden were c.65°C, at the base of the SSG c.80°C, and at the base of the Permo-Triassic 90°C.

Mapped subsurface temperatures 97 Ma ago, corresponding to the end of late Cimmerian erosion, immediately prior to the onset of post-extension subsidence in the late Cretaceous, are given in Chadwick et al. (1994a). Mean temperatures ranged from c.75°C at the base of the MMG to c.90°C at the base of the Permo-Triassic. By this time the basin had a single depocentre, to the north-east of the present deepest part of the Wem–Audlem Sub-basin, in which temperatures attained almost c.100°C at the base of the MMG, c.125°C at the base of the SSG and c.140°C at the base of the Permo-Triassic. The Alderley Edge area was at this time buried beneath about 1100 m of upper Triassic and Jurassic strata. Temperatures at the top and base of the preserved SSG at were c.55°C and c.75°C, and 85°C at the base of the Permo-Triassic.

Mapped subsurface temperatures 205 Ma ago, at the end of the Triassic, are given in Chadwick et al. (1994a). The mean temperatures in the basin were c.40°C at the base of the MMG, c.60°C at the base of the SSG and c.70°C at the base of the Permo-Triassic. Peak temperatures of c.60°C at the base of the MMG; c.90°C at the base of the SSG and >100°C at the base of the Permo-Triassic were reached locally in the Sandbach-Knutsford Sub-basin. In the Alderley Edge area, temperatures were c.60°C at the base of the SSG and c.70°C at the base of the Permo-Triassic.

A thermal history plot for the Wem–Audlem Sub-basin (Figure 66), corresponding to the burial history plot in (Figure 57), illustrates the thermal development of the basin fill, in particular the thermal maximum at c.60 Ma and the rapid cooling thereafter. (The apparent fall in temperature during the period 242 to 224 Ma, when the SSG was rapidly buried and compacted by the deposition of the MMG, is probably an artefact of the thermal-conductivity and depth data.) Thermal history plots for other locations are given in Chadwick et al. (1994a).

Pseudo-maturity estimation

HOTPOT does not perform organic maturity modelling, but maturity estimates may be obtained by displaying the temperature data with a pseudo-maturity scale. This divides the temperature data into four colour bands, representing under-mature (below 100°C), oil maturation (100 to 150°C), gas maturation (150 to 220°C) and over-mature (above 220°C).

Computed pseudo-maturity maps of the bases of selected Permo-Triassic layers from the end of the Triassic to the early Palaeogene are given in Chadwick et al. (1994a). In summary, the deepest parts of the Permian sequence entered the oil window towards the end of the Triassic and left it during Cenozoic uplift. A large proportion of the Permian deposits were in the oil window at the time of maximum burial (Figure 67) and some parts were at temperatures almost high enough for gas generation. The deepest parts of the SSG were in the oil window from early Jurassic to mid or late Palaeogene times. The deepest parts of the MMG were perhaps only briefly in the oil window from late Cretaceous to early Palaeogene times e.g. (Figure 67)b. Upper Carboniferous strata underlying the basin would have entered the oil window earlier than the Permo-Triassic strata, probably in mid to late Triassic times, and remained in the oil generation zone for longer, probably until the early Neogene. They would also have attained higher levels of maturity, well into the gas-generation zone at the time of maximum burial in the early Palaeocene.

Summary and conclusions

Caledonian basement rocks form the oldest tectono-stratigraphical unit in the Cheshire Basin region. They were deposited in Lower Palaeozoic times and have subsequently suffered several phases of deformation, resulting in the development of major, predominantly north-east-trending fault zones. The last major deformation episode, in early Devonian times, corresponded to the late Caledonian (Acadian) orogeny and resulted in major transpressive faulting, uplift and regional erosion.

Carboniferous strata were deposited unconformably on the Caledonian erosion surface. Initially, Dinantian rocks were deposited in fault-bounded extensional basins (formed in a north–south extensional regime) which, at least in part, were controlled by the reactivation of older, basement, structures. Later Carboniferous deposition appears to have been of a more regional nature. Carboniferous rocks underlie most of the Cheshire Basin, being thickest in the north where more than 4000 m of strata are preserved, with about 2000 m of Upper Carboniferous beds. At the end of Carboniferous times Variscan compressive stresses led to major transpressional fault reactivation, basin inversion, and regional erosion.

The Cheshire Basin is filled largely with Permo-Triassic strata, which rest unconformably on Carboniferous rocks and, locally, in the south, on Caledonian basement. The basin is roughly elliptical in plan, with a long axis trending north-east. It is markedly asymmetrical in cross-section, having, in general terms, the form of a faulted half graben, deepest in the southeast, where the base of the Permo-Triassic succession lies at depths locally in excess of 4500 m. The Wem–Bridgemere–Red Rock Fault System forms the south-east margin of the basin, with a present-day cumulative throw in places greater than 4000 m. In contrast, the western margin of the basin is relatively unfaulted, forming a feather edge characterised by depositional onlap. The internal structure of the basin is complex, with numerous faults. Intra-basinal normal faults range from structures with displacements in excess of 1000 m to features below the limit of seismic resolution, the larger faults subdividing the basin into a system of tilt-blocks.

The Permo-Triassic and subsequent evolution of the Cheshire Basin can be summarised as follows:

265–251 Ma (late Permian)

The initial measurable phase of basin subsidence corresponded to deposition of the Permian sequence. Crustal extension, approximately east–west in direction, reactivated basement faults to produce a south-east-deepening half graben. Intrabasin syn-depositional normal faulting produced local depocentres with sediment thicknesses locally in excess of 1800 m. Many of the principal basin-controlling normal faults were initiated at this time. Extension during this period accounted for approximately 24% of the total post-Variscan extension.

251–242 Ma (early Triassic)

Early Triassic subsidence was associated with the deposition of much of the SSG. Strong crustal extension, generally directed east to east-north-east, produced rapid fault-controlled subsidence and the development of separate, well-defined basin depocentres. The principal basin normal faults probably underwent their main phase of development at this time, with sinistral transtensional displacements on the WBRRFS the main basin-controlling factor. This phase of basin evolution accounted for approximately 28% of total basin extension, with more than 3500 m of strata deposited locally by the end of the period.

242–205 Ma (mid and late Triassic)

The upper SSG, the MMG and the Penarth Group were deposited during Middle and Upper Triassic subsidence. Crustal extension and syn-depositional normal faulting continued, but to a lesser extent than previously. Rates of subsidence decreased gradually with time, and a single depocentre developed, where, by the end of the Triassic, sedimentary thicknesses probably exceeded 5000 m.

205–144 Ma (Jurassic)

Preserved Jurassic strata are restricted to outliers of limited extent, so the Jurassic phase of basin evolution is poorly constrained. Regional considerations suggest that east–west extension probably continued into early Jurassic times, with continued fault-controlled basin subsidence. It is likely that the preserved Jurassic strata lie somewhat to the south of the Jurassic depocentre, which probably overlay the late Triassic depocentre. Extension probably recommenced in late Jurassic times, possibly with a markedly different extension direction (approximately north–south?). Total sedimentary thicknesses at the end of this period may have exceeded 6000 m locally, with the deepest parts of the Permian succession probably entering the oil window.

144–97 Ma (early Cretaceous)

No strata of this age are preserved, so accounts of this phase of basin evolution are speculative. Regional considerations suggest that crustal extension continued into early Cretaceous times, but contemporaneous regional uplift caused severe erosion of the basin margins, with lesser erosion of the basin depocentre. The effects of this erosion cannot be quantified, but it led to development of the important late Cimmerian unconformity, a regional feature of the UK continental shelf. It is likely that crustal extension ceased in early Cretaceous times, extension in the period from 242 Ma to 97 Ma constituting perhaps 48% of the total post-Variscan extension. Cumulative extension factors attained at the end of this period were significantly in excess of the presently measured values of 1.10 to 1.15. The deepest parts of the SSG and the Permian succession probably entered the oil window during this period.

97–60 Ma (late Cretaceous and early Palaeocene)

This phase of basin evolution was probably characterised by regional, post-extensional, thermal-relaxation subsidence, broadly corresponding to deposition of the Chalk. Maximum basin development was probably attained in early Palaeocene times, with the accumulation of nearly 6500 m of strata in the basin depocentre. Peak subsurface temperatures were attained at this time, with the deepest parts of the MMG and all underlying strata within the oil window.

60–0 Ma (Late Palaeocene to present)

Regional uplift commenced in Palaeocene times, perhaps associated with development of the Iceland Plume. Superimposed on this, basin inversion caused uplift of the basin depocentre and probable oblique reversal of the major basin-controlling faults. Inversion probably culminated in the Oligo-Miocene, associated with Alpine compressional events to the south. Uplift varied from about 1500 m outside the basin (the regional component), to well over 2000 m in the basin depocentre where inversion was most pronounced. Considerable cooling accompanied uplift, with the entire basin fill leaving the oil window.

Chapter 4 Provenance of the basin fill

D G Jones, A C Morton, M J Leng, H W Haslam, A E Milodowski, G E Strong and S J Kemp

Introduction

Sediment provenance studies provide valuable information on basin development and the relationships between basins and their hinterland. An understanding of provenance can help to establish the location and character of source areas, identify sediment entry points, determine transport routes and delineate intrabasinal sediment dispersal patterns. Identification of changes in provenance enables local correlations to be established, which are of importance in resource studies, particularly in delineating and modelling hydrocarbon reservoirs or host rocks for mineral deposits. Provenance information can also constrain the unroofing history of the source terrain (including the timing of fault movements), tectonic setting, crustal evolution, extent of sediment recycling, and contemporary climatic conditions. Finally, the composition of the basin fill has important implications for the fertility of the basin as a source of metals for ore fluids; for example, a high content of unstable minerals could lead to the release of large amounts of metals into pore-fluids as they break down, whereas an infill of mature sediments, consisting of more stable mineral species, would provide much lower metal concentrations. This factor could thus affect the sizes of ore bodies generated during basin evolution.

Many of the processes that affect the basin fill can obscure information on source: climatic factors can modify the mineralogy and geochemistry, sediment recycling produces a bias towards mature grains, and diagenesis can modify the mineral assemblage (Haughton et al., 1991). The chemical and mineralogical composition of a sedimentary rock is a product of provenance, weathering, transport, deposition and diagenesis.

A full understanding of the provenance of sedimentary basin fills requires a multidisciplinary approach. An integrated study maximises the provenance information from a given basin and provides the best assessment of factors such as diagenesis and regional variations in lithology and grain size (Humphreys et al., 1991).

Several approaches have been used here to shed light on the provenance of the Permo-Triassic fill of the Cheshire Basin, including studies of heavy-mineral assemblages, heavy-mineral chemistry, clay mineralogy, whole-rock petrography, whole-rock geochemistry and radiogenic isotopes (of individual clasts in conglomeratic facies, and of whole rocks). An assessment of the sedimentology of key boreholes in the basin was made, to provide a framework for the study, and this was placed in the context of previously published accounts (see pp.38–40) to give background information on the depositional environments of the sediments and the flow directions in which fluvial and aeolian processes operated.

In order to minimise the effects of surface weathering, most of the samples studied were obtained from boreholes. However, most of the boreholes sampled were relatively shallow (<200 m) and may, therefore, have been affected by

Tertiary and modern weathering and shallow groundwater flow. A few surface outcrop samples were also included in the study, as well as samples from the old mine workings at Alderley Edge and Clive. Sample locations are shown in (Figure 2) and summarised in (Table 1) and (Table 2).

General petrographic characteristics

Detailed petrographic studies were carried out to elucidate the provenance and diagenetic history of the basin. An account of diagenesis is given in Chapter 5, but a general petrographic description of the rocks is relevant to provenance studies and is, therefore, reported in this section. In total, 286 samples were examined petrographically from 50 cored boreholes and 5 outcrop or mine locations (Table 1) and (Table 2); (Figure 2) by a range of methods including, optical, scanning electron, backscattered scanning electron (BSEM) and cathodoluminescence microscopy. BSEM identification of minerals was aided by qualitative energy-dispersive X-ray microanalysis (EDXA). In all, 92 samples were analysed by X-ray diffraction analysis (XRD) to determine their clay mineralogy: 54 from the MMG, 35 from the SSG and 3 from the Manchester Marls Formation. Details of the techniques used are given in Milodowski et al. (1994) and Kemp (1994), and in Chapter 5 of this volume.

Sherwood Sandstone Group

Kinnerton Sandstone Formation

The Kinnerton Sandstone Formation was examined in boreholes from both the northern and southern parts of the Cheshire Basin. The SSG examined in the Perry Farm and Little Ness boreholes may include this formation, but the stratigraphy of these boreholes is very poorly constrained.

The Kinnerton Sandstone Formation is composed of pebble-free sandstones. It is normally considered to be mainly aeolian, with occasional beds of fluvial sandstone (cf. Evans et al., 1993, and see pp.18 and 39). In the borehole sequences examined in this study, however, fluvial sedimentation appears to have been more significant. Thin-section petrography reveals that many sediments consist of a mixture of angular grains and well-rounded millet-seed grains, suggesting that they may be fluvial, interdune sediments containing reworked aeolian sand. Massive and cross-bedded aeolian sandstones in the formation are typically moderately to well sorted (using the definitions of Longiaru, 1987) and have abundant well-rounded millet-seed grains. Finer-grained flaser- and ripple-laminated interdune sandstones are much less well sorted. Parallel-laminated sandstones often show good sorting characteristics within individual laminae.

The sandstones are largely cherry sublitharenites with subordinate quartz arenites (using the definitions of Pettijohn, 1975). The detrital mineralogy comprises major quartz with minor (1–10%) K-feldspar, ferruginous mudclasts, chert grains and volcanic rock fragments (Plate 6). Albite is usually present in trace (<1%) amounts but is more abundant in some sandstones. Sandstone rock fragments were found in the Kinnerton Sandstone Formation from the Rainhill Borehole. The proportion of polycrystalline quartz to simple single-crystal quartz increases in the south of the basin, where lithic fragments are also more abundant. These include metaquartzite and altered garnetiferous metamorphic rocks (Child's Ercall Borehole) and sphene-bearing syenite (Hodnet Station Borehole), in addition to cherts and altered acid/felsic volcanic rocks. Lithic fragments reach major proportions (up to 40%) in sandstones in the Hodnet Station Borehole. In contrast, in the northern boreholes (Halewood, Mickle Trafford, Rainhill, Seedley Print Works and Stanlow) the lithic fragments comprise largely chert with less abundant fine-grained volcanic rocks. However, lithic and albite grains would originally have been more abundant, since these rocks contain significant secondary porosity as a result of dissolution of these types of grains.

Minor to trace amounts of detrital biotite and muscovite, and trace amounts (<1%) of magnetite, ilmenite, chromite, altered Ti-Fe oxide, rutile, zircon, apatite, epidote and garnet are also present. Detrital garnets in the Child's Ercall Borehole display snow-ball inclusion fabric. Rare detrital grains of monazite of diagenetic origin (containing inclusions of chlorite and white mica), similar to those described from Lower Palaeozoic rocks of central Wales (Milodowski and Zalasiewicz, 1991a), were identified in sandstones from the Halewood Borehole. This type of monazite has been found in Palaeozoic rocks from the Welsh Massif, south-west England, Belgium and Brittany, as well as in rocks ranging from Precambrian to Permian in age further afield in Europe and other parts of the world (Donnot et al., 1973; Burnotte et al., 1989; Cooper and Read, 1984; Read et al., 1987; Milodowski and Zalasiewicz, 1991a; Smith et al., 1994), although it has not been reported in rocks from northern England or Scotland (Read et al., 1987).

Detrital clay is present in some sandstones, in silty laminae or as intergranular matrix in less well sorted sandstones. The clay mineralogy of samples from the north of the basin (Halewood, Mickle Trafford and Stanlow boreholes; (Table 1); (Figure 2) consists of varying proportions of illite, smectite and chlorite.

Chester Pebble Beds Formation

Observations on the Chester Pebble Beds Formation are biased towards the north-west and south-east of the basin, and there is little apparent difference between these areas.

The rocks are typically light brown to red-brown in colour, and are often strongly cemented by carbonate (calcite and/or dolomite) cements. Lithologically the formation is very variable, even in a single locality, consistent with complex fluvial facies associations (Chapter 2). The rocks comprise quartz arenites, lithic arenites, quartz sublitharenites, cherty sublitharenites, arkoses, subarkoses, quartz wackes and lithic wackes (definitions after Pettijohn, 1975), with occasional beds of finely laminated ferruginous silty mudstone and siltstone. Sand grains in sandstones and the sandy matrix of conglomerates are typically angular to subangular, although some sandstones contain laminae of coarser well-rounded aeolian sand.

The mineralogy of the Chester Pebble Beds is very similar to that of the underlying Kinnerton Sandstone, although the sandstones generally contain a greater proportion of lithic clasts, chert and K-feldspar. Micaceous sandstones, with minor amounts of detrital muscovite and biotite, are also common. Lithic fragments are dominated by chert and quartzite, with subordinate silicified volcanic clasts and silicified bioclastic rock fragments, the latter probably derived from the Carboniferous. Ferruginous mudstone and siltstone intraclasts are also very common, particularly in pebbly sandstones and conglomerates. Trace detrital components include zircon, rutile, magnetite, ilmenite, altered Ti-Fe oxides and rare igneous monazite and green amphibole.

Detrital clay is significant in many sandstones, forming thin laminae or disseminated within the sandstone matrix. Samples from the north of the basin (Mickle Trafford, Littleton and ICI Widnes boreholes) were found to contain an assemblage of illite, smectite and chlorite, similar to the underlying Kinnerton Sandstone Formation, although locally it appears that the Chester Pebble Beds are more chloritic.

Wilmslow Sandstone Formation

The Wilmslow Sandstone Formation was examined in a wide coverage of boreholes from the north and south of the basin, and also at outcrop at Grinshill (Deakins Quarry, (Table 2); (Figure 2). The formation is generally more argillaceous than the underlying rocks, although mineralogically very similar. The sandstones are dominated by massive, cross-bedded, parallel-laminated, fine- to medium-grained, cherty sublitharenites and subarkoses, but lithic wackes, quartz wackes, litharenites, quartz arenites, lithic greywackes, silty mudstones and siltstones are also present. Sorting is generally poor to moderate, but coarser rocks tend to be better sorted. Many sandstones contain a ferruginous matrix clay. Detrital grains are predominantly angular to subrounded but disseminated coarser, well-rounded, fluvially reworked aeolian, or aeolian-derived, grains are commonly present.

Detrital mineralogy comprises major quartz (simple and polycrystalline grains), major to minor chert and K-feldspar, and minor muscovite, biotite and albite. Albite appears, qualitatively, to be more abundant in the Wilmslow Sandstone Formation than in the other SSG formations. Fragments of volcanic rock, including silicified rhyolitic and highly altered ferruginous varieties, are probably less significant than in the Kinnerton Sandstone and Chester Pebble Beds. Other lithic clasts include metaquartzite, mudstone intraclasts and clay pellets, and silicified foraminiferal limestone fragments (probably Carboniferous). Mudstone intraclasts are common in the bases of some fining-upwards fluvial cycles. Small well-rounded quartz or quartzite pebbles are also locally present. Rare collophanic bone fragments were identified in the Clotton Borehole. Trace detrital minerals are dominated by altered Ti-Fe oxides, magnetite, zircon, apatite and tourmaline.

The clay mineralogy of the formation was examined in adjacent red and grey-green sandstones from the Mickle Trafford Borehole. Both sandstone types contain a similar clay-mineral assemblage of smectite, illite and minor chlorite. Outcrop material taken from Deakins Quarry is uniquely dominated by smectite with only minor illite. XRD scans of the 060 diffraction spacing indicates that the smectite species is dioctahedral (1.50 Å spacing) and the illite trioctahedral (1.54 Å). The commonest dioctahedral smectite-group mineral, containing divalent interlayer cations, is montmorillonite.

Helsby Sandstone Formation And Grinshill Sandstone

The Helsby Sandstone Formation was studied largely in the north and north-west of the basin. In the south of the basin, samples were examined from the locally defined Grinshill Sandstone, which may be stratigraphically equivalent (Chapter 2).

The Bulkeley Hill Sandstone Formation, which lies between the Wilmslow Sandstone Formation and the Helsby Sandstone Formation, was not examined in this study.

In the Malpas area (west-central part of the basin) and in the Wilkesley Borehole (Figure 2), a thick sequence of cross-bedded aeolian and water-lain sandstones (Ma1pas Sandstone) is considered to be part of the Tarporley Siltstone Formation (Wilson, 1993; Chapter 2, this volume). This sandstone was not, however, fully penetrated in the Wilkesley Borehole and may here represent the top of the Helsby Sandstone Formation (Warrington, personal communication). The petrological and sedimentological characteristics of the Malpas Sandstone are so similar to those of the Helsby Sandstone Formation that they are considered together in this section.

The Helsby Sandstone Formation is dominated by quartz arenites interbedded with subordinate quartz and cherry sublitharenites and subarkosic sandstones, and occasional quartz wackes, quartz-rich ferruginous siltstones and silty mudstories. The sandstones vary from coarse to fine grained, and from poorly sorted fluvial sands (with angular to sub-angular grains) to moderate to well-sorted aeolian sandstones (with abundant millet-seed grains). The rocks vary from red to brown with local bleached (or green, reduced) horizons, particularly siltstones. Interlaminated siltstones and mudflake breccias are common in the upper parts of the sequence, for example in the Saughall Massie Borehole, as it passes into the Tarporley Siltstone Formation.

The detrital mineralogy of the formation is similar to that of the underlying rocks. It consists of major quartz (single crystal and polycrystalline), major to minor K-feldspar, and lithic fragments (dominated by chert). Mudstone intraclasts are also abundant in some sandstones, particularly at the base of fluvial sandstone beds. In general, however, quartz arenites dominate the Helsby Sandstone Formation and the sandstones are more feldspathic. In contrast with underlying formations, volcanic rock fragments and quartzite clasts are less common. However, in the Malpas Sandstone of the Wilkesley Borehole, well-rounded, low-grade (weakly cleaved), metamorphic, silty mudstone grains, composed of illite, quartz and chlorite, are present with abundant chlorite-mica stack intergrowths identical to those observed in the Lower Palaeozoic rocks of the Welsh Basin (Evans and Adams, 1975; Craig et al., 1982; Woodland, 1985; Milodowski and Zalasiewicz, 1991b). This suggests that sediment was derived from Lower Palaeozoic basement rocks in the Midlands or Wales. Other detrital components include minor to trace biotite, muscovite, albite and calcic plagioclase, and traces of zircon, apatite, rutile, ilmenite, magnetite and altered Ti-Fe oxides.

The clay mineralogy of the formation is generally dominated by illite, with only trace quantities of smectite and chlorite. In the west (Gallantry Bank) the Helsby Sandstone Formation lacks chlorite. The upper part of the Bootle Borehole (above c.76 m) and one sample from the Thornton Borehole have significantly more smectite.

Mercia Mudstone Group

Tarporley Siltstone Formation

The Tarporley Siltstone Formation has a distinctive facies of alternating siltstone, reddish brown and greenish grey mudstone and thin, fine- to medium-grained sandstones. BSEM petrography clearly shows that many of these sediments actually comprise small complex, but well-defined, fining-upwards cycles rather than discrete siltstone and mudstone laminae (Plate 7). The Tarporley Siltstone Formation in the Marston Salt Union Borehole locally contains occasional thin (<1 mm) laminae of aphanitic lath-like synsedimentary anhydrite.

Coarser sandstones and siltstones in the Tarporley Siltstone Formation are similar to those of the underlying Helsby Sandstone Formation. Mineralogically, the rocks comprise largely detrital angular quartz silt and clay minerals, with minor amounts of detrital K-feldspar, albite, mica (muscovite and biotite), chert fragments, and Ti-Fe oxides. Minor amounts of fine-grained (<1 μm) iron oxide (presumably hematite) are present in the red-brown lithologies but largely absent in the green or grey varieties. Concentrations of detrital heavy minerals (ilmenite, zircon > rutile >> magnetite, tourmaline > apatite, monazite, xenotime, chromite) are common at the base of siltstone laminae and interbeds (Plate 8). In one sample from the Saughall Massie Borehole, unusual well-rounded medium sand-grade detrital thorite grains, showing possible metamict shrinkage cracking/crazing, are relatively common, together with unusual detrital monazites which resemble authigenic monazites described from the Lower Palaeozoic of Wales (Milodowski and Zalasiewicz, 1991a).

XRD studies of the Tarporley Siltstone Formation from the Wood Lane Borehole indicate that the <2 μm fraction clay minerals are dominated by illite with a variable proportion of smectite and chlorite. A further sample from Bridge Quarry has a <2 pm fraction which is dominated by ?corrensite with minor illite.

Bollin Mudstone Formation

Detailed BSEM petrography of the mudrocks reveals a complex sedimentary fabric. Many of what appear to be siltstones are in fact silty dolomicrites and anhydritic dolomicrites developed in complex rhythmic cycles, 0.2–4 mm thick. In the Wilkesley and Mobberley Town 5 boreholes (and probably equivalent rocks to the Bollin Mudstone Formation in the Coton Fields Borehole) the laminated strata typically comprise 0.2–2.0 mm laminae of a basal dolomicrite or dolomicritic silt, interlayered with, and grading upwards into, 0.1–3 mm of anhydritic dolomicrite and sometimes finally anhydritic mudstone (Plate 9) or a thin layer of virtually pure aphanitic felted-lath anhydrite (Plate 10). Lenticles of felted-lath anhydrite (up to 10 mm long and resembling starved ripples) occur in low-angle ripple and cross-laminated dolomicrite and dolomicritic siltstone within laminated dolomicritic siltstone matrix. Anhydrite and dolomicrite laminae are also locally disrupted by soft-sediment deformation with slumping and dewatering structures (e.g. flame structures). These features indicate a very early synsedimentary (pre-burial) origin for most of the anhydrite. Nodular and micro-nodular aphanitic anhydrite locally disrupts the sedimentary fabrics (Plate 11), and clearly developed within the sediment shortly after deposition. These fabrics are similar to those of sabkha sediments described by Shearman (1966), Kinsman (1969) and Dean et al. (1975).

Siltier facies contain mainly fine angular detrital quartz with minor albite, K-feldspar in a clay matrix, and traces of apatite, anatase, rutile, Ti-Fe oxides, zircon, monazite and xenotime. Silt-grade and fine sand-grade detrital muscovite and biotite are locally abundant within micaceous laminae, and elsewhere are generally common minor components. Minor to trace amounts of fine-grained disseminated iron oxide are present in red-brown rocks but absent in green varieties.

XRD analysis of 14 representative samples from four boreholes (Table 1), indicated a clay-mineral assemblage of illite and minor chlorite. Corrensite was also identified in some samples. It is commonly found in many modern-day hypersaline evaporitic lacustrine and associated syndepositional diagenetic environments (Velde, 1977; Khoury, et al., 1982; Jones and Weir, 1983) where conditions are alkaline (pH 8–10), and its occurrence is consistent with the inferred depositional environment of the Bollin Mudstone Formation.

Northwich Halite Formation

The whole-rock petrography of the Northwich Halite Formation was examined only in samples from the Meadowbank Salt Mine (Winsford). The formation comprises up to 75% halite with beds of mudstone making up the remaining 25% (Wilson, 1993; Chapter 2, this volume).

The halite horizons are nearly monominerallic but in places contain scattered inclusions and thin laminae (<1 mm) of minor, perhaps synsedimentary, anhydrite and non-ferroan dolomicrite. Petrographically the halite has suffered extensive, and probably multiple, recrystallisation. Interbedded mudstones and siltstones are very similar to the laminated anhydritic-dolomicritic siltstones found in the underlying Bollin Mudstone Formation and the overlying Byley Mudstone Formation. Euhedral, often large (centimetre-scale) synsedimentary crystals of halite have grown displacively within the mudstones and siltstones (haselgebirge facies).

XRD analysis of the clay mineralogy of the formation, in borehole samples as well as those from the Meadowbank Salt Mine, show some variation. The Mobberley Town samples, and the younger sample from A556 Borehole 446, exhibit an ate, corrensite and chlorite assemblage. The samples from the Meadowbank Salt Mine display an illite and chlorite assemblage with only a trace of corrensite, while the deeper sample from A556 Borehole 446 has an assemblage of illite and chlorite only.

Byley Mudstone Formation

The Byley Mudstone Formation comprises red and green interlaminated siltstones and mudstones, which are often very dolomitic and anhydritic, together with a structureless mudrock facies. BSEM petrography shows that the anhydritic and dolomitic rocks are typically very finely laminated (0.2–5 mm laminae), similar to those of the Bollin Mudstone Formation, and comprise cycles of siltstone and dolomitic siltstone which pass gradationally into dolomicrite, anhydritic dolomicrite and/or felted-aphanitic anhydrite laminae (Plate 12). These are well developed in the Crewe Heat Flow Borehole and in the Winsford W6 Borehole. As in the Bollin Mudstone Formation, the BSEM reveals very fine-scale cross-lamination in these anhydritic-dolomitic sediments (Plate 13), indicating deposition under a gentle current regime. The sediments also show soft-sediment deformation, including slumping, convoluted bedding and flame structures resulting from early synsedimentary dewatering (Plate 13). Ptygmatically folded (compactionally deformed) small syneresis cracks are common in some samples. They are interpreted as subaqueous shrinkage cracks developed under hypersaline conditions (similar to those found in other hypersaline environments: cf. Donovan and Foster, 1972), filled by felted masses of fine anhydrite sediment or anhydritic–dolomicritic silt.

Nodular displacive anhydrite is also sometimes present (Plate 13) and appears to be eodiagenetic in origin. In general, the Byley Mudstone Formation is similar to the Bollin Mudstone Formation and also represents deposition in a shallow hypersaline environment, periodically drying out with localised marginal sabkha development. XRD analysis of eight samples from the Crewe Heat Flow Borehole shows a uniform clay-mineral assemblage of illite, corrensite and minor chlorite.

Detrital components comprise mainly silt-grade angular quartz and clay minerals, with minor K-feldspar, albite, muscovite and traces of green chloritised biotite, fine titanium oxides, zircon, monazite and ilmenite. Minor to trace amounts of fine-grained disseminated iron oxide are present within red-brown rocks but are absent in green variants.

Wych Mudstone Formation

The Wych Mudstone Formation, as typified by the sequence in the Crewe Heat Flow Borehole, is represented by dominantly structureless reddish brown siltstones, silty mudstones and mudstones, with occasional thin very fine quartz arenites. The rocks often contain abundant disseminated dolomite and anhydrite and may display localised greenish reduction spots. Nodules of gypsum and anhydrite ranging in size from <1 mm to 10 cm may be present. These expansively disrupt the sediment and in some cases may produce a pseudo-breccia. Scattered halite crystals may also be present.

The detrital components comprise largely angular quartz silt and very fine sand floating within a fine clay or dolomicritic clay matrix, with minor amounts of K-feldspar and coarse detrital muscovite. As in the underlying Byley Mudstone Formation, the clay mineralogy of the Wych Mudstone Formation consists of a uniform assemblage of illite, corrensite and minor chlorite.

Wilkesley Halite Formation

No petrographic analysis was undertaken for the Wilkesley Halite Formation. However, Wilson (1993) indicates that it has broadly similar characteristics to the underlying Northwich Halite Formation (see also Chapter 2). XRD analysis of two samples from the Arclid Bridge 2 Borehole identify a clay-mineral assemblage of illite, corrensite and minor chlorite similar to that in the underlying MMG formations (except for the Tarporley Siltstone Formation, which contains appreciable smectite).

Brooks Mill Mudstone Formation

Petrographic analysis of the Brooks Mill Mudstone is limited to the British Gypsum Audlem AU17 Borehole. The rocks are described by Wilson (1993) as dominantly structureless reddish brown mudstones with subordinate sandstones and anhydrite nodules, but detailed petrographic examination revealed an intercalation of reddish brown silty pelloidal mudstones with thin laminae of ferruginous clay-pellet sandstone, structureless to weakly laminated dolomitic mudstones, and reddish brown or green silty dolomicrites and ripple-laminated silty dolomitic mudstones and siltstones.

Mineralogy is dominated by angular quartz silt and sand in a clay matrix, with minor to trace amounts of K-feldspar, muscovite, biotite and chlorite, and trace amounts of apatite, altered Ti-Fe oxides, zircon, xenotime and monazite. XRD analysis of 13 samples of Brooks Mill Mudstone Formation from the Audlem AU17 and Lotus Ltd, Stafford boreholes show either an assemblage of illite and chlorite, or illite, corrensite and chlorite.

Blue Anchor Formation

No petrographic analysis of this formation was undertaken, although XRD analysis of a single sample from the Audlem AU17 Borehole indicates a clay-mineral assemblage of illite and chlorite only.

Heavy-mineral studies

Heavy-mineral suites in sandstone samples from eighteen borehole sections (Table 1); (Figure 2) and (Figure 72) were examined. The boreholes were chosen to optimise regional and strati-graphical cover. Seventeen of them were from the Cheshire Basin or from the coastal strip between the basin and the East Irish Sea Basin. The other section was from the Little Hay Borehole in the Needwood Basin, south-east of the Cheshire Basin (Figure 72). As well as using the mineralogy to constrain the provenance of the sandstones, the study of heavy minerals aimed to:

1. characterise individual stratigraphical units on the basis of their detrital mineralogy,
2. assess whether these variations were useful for correlation purposes,
3. determine if regional variations in mineralogy exist across the basin, and
4. assess the degree and nature of diagenesis by an analysis of the dissolution patterns of detrital heavy minerals.

Over 100 samples were selected for conventional heavy-mineral analysis, and tourmaline compositions were determined on a subset of 22. The tourmaline analysis was carried out to confirm whether differences in overall mineral assemblages were associated with changes in detrital tourmaline compositions, and to establish whether further variations could be detected.

Methods

Heavy-mineral analysis was carried out on the 63–125 μm fraction using standard preparation procedures (see Morton and Johnsson, 1993; Morton, 1993). A basic characterisation of the non-opaque detrital heavy-mineral suites was made by counting 200 grains using the ribbon method of Galehouse (1971). Identification was made on the basis of optical properties, as described for grain mounts by Mange and Maurer (1992). Provenance-sensitive heavy-mineral indices (Morton and Hallsworth, 1994) were determined by making additional counts, where possible, of at least 100–200 grains of each pair of minerals, again using the ribbon method. The index values used in this study are defined in (Table 12). The relative abundances of the three TiO₂ group minerals (rutile, anatase and brookite) were determined by counting a minimum of 50 grains.

The compositional ranges of individual tourmaline populations were obtained from electron microprobe analyses. Tourmaline grains in the heavy-mineral residues were picked with a needle during optical examination using a polarising microscope. All tourmalines observed during one or more traverses across the residue were extracted, approximately 60 being chosen per sample. The grains were placed on double-sided adhesive tape, coated with carbon, and analysed using a Link Systems AN 10/55S energy-dispersive X-ray analyser attached to a Cambridge Instruments Microscan V electron microprobe. The count time was 50 seconds for each grain. Because of the presence of B, Li, H and F, none of which can be detected by energy-dispersive X-ray analysis, analytical totals do not exceed about 88%. Analyses with totals significantly below 88% because of grain surface roughness, imperfect orientation or chemical contamination, were rejected. The data were then normalised to 100% to ensure data comparability.

Most of the compositional differences between tourmaline grains recorded in this study relate to variations in the major components FeO and MgO and in the trace components CaO and TiO₂.

General observations

The heavy-mineral suites examined in this study are relatively restricted in diversity, the only consistently abundant minerals being apatite, TiO₂-group minerals, tourmaline and zircon. These species form an average of 95.6% of the heavy-mineral assemblages. Garnet is generally minor, although it forms over 20% of the assemblages in about 3% of samples and between 10% and 20% in a further 7%. Of the other species identified, monazite is present in minor amounts in many samples, but chrome spinel, clinopyroxene and epidote are very rare. It should be noted that the monazites recorded here are detrital monazites of igneous or metamorphic origin and do not include the derived diagenetic monazite nodules noted on p.91. Staurolite is generally very rare, but is an important constituent of the assemblages in the Little Hay Borehole. Tourmaline compositions are diverse, with all samples showing a wide range in FeO and MgO contents. Nevertheless, there are indications of differences between formations, in terms of both major and trace elements (Figure 68). Full results are reported in Morton (1993).

Most of the garnet grains are moderately to strongly etched, indicating that this mineral has undergone dissolution in the subsurface and that its abundance may have been reduced post-depositionally. As garnet is more stable than staurolite, kyanite, titanite, epidote, amphibole and pyroxene (Morton, 1984), the scarcity of these minerals may not be provenance-related, but rather the result of post-depositional dissolution. The indication from the Little Hay Borehole is that staurolite, at least, was originally present in some sandstones. By comparison with the experimental work of Hansley (1987), the extent of garnet dissolution suggests that pore-fluid temperatures exceeded 80°C. This accords with the results of apatite fission-track analysis (Lewis et al., 1992), which suggests that palaeotemperatures may have exceeded 90°C. In the North Sea, similar degrees of garnet dissolution occur when burial depths reach about 3000 m (Morton, 1984, 1986). However, maximum temperatures and burial depths for the SSG cannot be inferred directly from the experimental work of Hansley (1987) or the North Sea analogy, because dissolution rates also depend on pore-water chemistry, fluid flux and heat flow. The relatively advanced garnet dissolution observed in the SSG of the Cheshire Basin therefore implies either (i) that the sequence has been deeply buried and then uplifted; (ii) that the pore waters were considerably more aggressive than in the North Sea; (iii) that high pore-fluid temperatures resulted from fluid movement updip or along fault planes from the deeper parts of the basin; or (iv) a combination of the above. Garnet is slightly more abundant in the south of the area (4.9%) than in the north (3.6%), but there is a very large scatter in both areas and the difference is not considered significant in terms of indicating differences in burial history between the northern and southern parts of the basin.

Many of the samples contain secondary diagenetic phases, the most frequent being carbonate, pyrite and anatase (both neoformed and as overgrowths on detrital anatase). Other species foundin the residues, but not included in the count, are micas (notably biotite), phyllitic rock fragments and opaque minerals.

Sandstones below the Chester Pebble Beds Formation

Thirty sandstone samples from formations underlying the Chester Pebble Beds Formation were included in the study (Table 1). Most are from sandstones assigned to the Kinnerton Sandstone Formation. One sample of Bold Formation, two of Collyhurst Sandstone Formation and two of arenaceous units near the top of the Manchester Marls Formation were also included. The most complete coverage is from the Stanlow Borehole, where samples cover approximately 250 m of vertical section. As data from the Bold, Collyhurst Sandstone and Manchester Marls formations fall within the ranges shown by the Kinnerton Sandstone (Table 12)." data-name="images/P1000323.jpg">(Figure 69), they are all considered together (Table 13). They have a relatively low mean ATi value, but with a large scatter. Monazite is scarce, and thus MZi is generally low. RZi values are relatively low, and rutile forms a high proportion of the TiO₂ group.

The Kinnerton Sandstone in the Stanlow Borehole shows stratigraphical variations that may provide a basis for correlation, at least on a local level. Some of the high degrees of scatter (notably in the ATi value) may be a result of these stratigraphical changes in provenance. The Stanlow Borehole sequence may be divided into three units (Figure 70). The lowermost unit has relatively low ATi and RZi values (ATi: m=26.5, s=13.5, RZi: m=23.1, s=6.9), and monazite and garnet are scarce. The base of the overlying unit is marked by a sharp increase in the abundance of garnet between 131.5 m and 122.1 m, but there is little difference in other parameters. For example, ATi remains very similar (m=23.8, s=7.7), and monazite is absent; RZi is slightly higher (m=30.5, s=7.2), but this difference is not statistically significant. The increased abundance of garnet is therefore not necessarily related to a change in provenance, although this remains a possibility. The uppermost unit is defined by a decline in garnet abundance, an increase in ATi (m=49.5, s=11.2), the appearance of consistent levels of monazite (mean MZi of 1.2), and a further increase in RZi (m=33.6, s=3.2). There is therefore a clear change in sand provenance in the upper part of the Stanlow section. The low ATi value in the uppermost sample is attributed to apatite dissolution due to present-day weathering, since apatite grains in this sample show marked dissolution features.

Tourmaline compositions were determined for eight samples from below the Chester Pebble Beds (Table 14). Comparison of the mean tourmaline compositions of each sample suggests differences in provenance. Six of the eight samples are relatively low in FeO and high in MgO. Mean CaO and TiO₂ contents for each sample range from 0.63 to 0.75% and 0.85 to 0.97% respectively. In contrast, the two remaining samples have higher FeO, lower MgO and lower CaO. TiO₂ contents are similar. These two tourmaline populations are termed Kinnerton Type 1 and Kinnerton Type 2 respectively ((Table 14) and (Figure 71). Type 2 occurs in Kinnerton sandstones of the Newport UDC and Childs Ercall boreholes, whereas Type 1 occurs in Collyhurst, Bold and Kinnerton sandstones of the Stanlow, Speke Reservoir and Childs Ercall boreholes. The tourmaline data therefore appear to indicate that more than one type of source rock was involved. Type 2 is confined to the south, whereas Type 1 is prevalent in the north, although it is also present in one of the two samples from Childs Ercall in the south. This may imply a palaeogeographical control on the detrital tourmaline assemblages, with more FeO-rich and MgO-poor tourmalines supplied from the southern part of the source terrain than the north. There are no clear differences in the associated heavy-mineral assemblages, and the downhole variations seen in the Stanlow Borehole are not mirrored by significant changes in tourmaline composition.

Chester Pebble Beds Formation

Thirty-two sandstone samples from the Chester Pebble Beds were analysed from eight boreholes in the Cheshire Basin and from the equivalent Cannock Chase Formation (Warrington et al., 1980) in Little Hay Borehole in the Needwood Basin (Table 1); (Figure 72). There is some doubt as to whether the succession in the Perry Farm Borehole (5 samples) is Chester Pebble Beds or Kinnerton Sandstone; the heavy-mineral ratios are more characteristic of the former.

The heavy-mineral ratios and tourmaline compositions differ significantly from those in the underlying sandstones (Table 13) and (Table 14). The formation is characterised by higher monazite abundances, reflected by generally high MZi values (Table 12)." data-name="images/P1000323.jpg">(Figure 69). Despite this, some samples have low MZi values similar to those of the underlying sandstones, indicating a continuation of input from low-monazite sources as well as input from a new, high-monazite source. Mean MZi is 4.0 (s=2.0), noticeably higher than in any other formation (Table 13). Excluding two samples from the base of the Chester Pebble Beds Formation in the Speke Reservoir Borehole, ATi values are high with relatively little scatter. RZi is higher than in underlying sandstones, although the proportion of rutile within the detrital TiO₂ component is similar (47.6%). In the thickest cored section (ICI Sports Ground Borehole, Widnes), ATi values increase towards the top. This pattern is also seen in the Speke Reservoir section, where the base of the formation has very low ATi values.

Tourmalines were analysed from six Chester Pebble Beds sandstones (Table 14), (Figure 71). Apart from one sample (Holcroft Lane Borehole, 143.4 m), the assemblages are uniformly rich in FeO and poor in MgO. The anomalous sample has lower FeO and higher MgO, and is more comparable to samples from the overlying Wilmslow Sandstone Formation. It also has low MZi (0.5). The boundary between the Wilmslow Sandstone Formation and the Chester Pebble Beds Formation has been placed at 135 m in the borehole, some 8 m above the anomalous sample. This indicates either that the position of the boundary requires revision or that Wilmslow-type sources had become of major importance during late Chester Pebble Beds times.

Wilmslow Sandstone Formation

Twenty-nine samples from the Wilmslow Sandstone Formation were studied from five boreholes (Table 1). Included in this total are samples from the Bootle Golf Course Borehole, although the lithostratigraphy of these is in doubt. The geological map suggests that the borehole should start in Helsby Sandstone, but the geochemistry and the borehole log suggest that the Wilmslow Sandstone is represented, at least in the deeper parts of the hole. The heavy-mineral characteristics are very clearly of Wilmslow Sandstone affinity.

The sample set contains two strikingly different heavy-mineral assemblages, one with abundant apatite and one virtually devoid of apatite (Table 12)." data-name="images/P1000323.jpg">(Figure 69), (Table 13). The apatite-rich assemblage characterises 15 of the 29 samples. It has a much higher mean ATi than the apatite-poor assemblage but mean MZi and RZi are not significantly different. Mean Ru/TiO₂ values are 31.5% and 20.8% respectively (Table 13). The apatite-poor assemblage occurs throughout the Sansaw Heath Borehole, and overlies the apatite-rich assemblage in the Bootle and Holcroft Lane boreholes. The boundary appears to be sharp and without any transition, although there may be a gradation on a smaller scale than the sample interval.

The tourmaline compositions of apatite-poor and apatite-rich sandstones from the Wilmslow Sandstone Formation are very similar (Table 14). FeO values are slightly different, but this is not considered significant in view of the relatively large standard deviation and the similarity in the other components. The tourmaline data show a close affinity with the overlying Helsby Sandstone (Figure 71).

One possible explanation for the observed differences in mineralogy is that there was a change in sandstone provenance. The lack of any difference in tourmaline composition argues against this, and the apparent abruptness of the variation is also incompatible with this interpretation, as unroofing tends to cause gradational rather than abrupt changes. The associated change in RZi value from 43.0 to 48.2 (Table 13) is more likely to be due to the development of secondary anatase than to a change of provenance, as the ratio of zircon to rutile shows very little difference between the two assemblages. Secondary anatase commonly develops during acidic groundwater flushing events, and, although wholly secondary anatase grains were not counted during the RZi determinations, many of the primary anatase grains that were included in the count were present in the 63–125 um fraction only as a result of secondary overgrowth (see also Morton and Hallsworth, 1994).

It is more likely that the apatite-poor assemblage was formed from the apatite-rich assemblage through apatite dissolution, as a result of an intra-Triassic, or a much more recent, event. Apatite dissolution takes place under the influence of acidic groundwaters (Morton, 1986) related to periods of subaerial exposure. Although almost all the borehole sequences studied are very shallow, there is only one sample (in the Stanlow Borehole) where apatite depletion can be attributed to surface weathering, and here apatite has not been entirely removed, so it is unlikely that the apatite-poor assemblages are due to recent dissolution. On the other hand, the stratigraphical position of the apatite-poor assemblage, which appears to be in the upper part of the Wilmslow Sandstone beneath the Hardegsen disconformity, suggests that the action of acidic groundwater during an intra-Triassic event is the most likely cause.

A number of regionally widespread deep weathering profiles have been observed in the Buntsandstein of Germany and Switzerland (Ortlam, 1980), deposited in a similar environment and climatic setting. These weathering profiles are associated with major disconformities, such as the Hardegsen disconformity, and with zones of apatite depletion. It is possible that the apatite dissolution event noted in the Cheshire Basin has regional significance and correlates with one of the events in the German Buntsandstein.

The provenance of the Wilmslow Sandstone differs significantly from that of the Chester Pebble Beds. Monazite is less abundant, with a mean MZi close to 1.0 compared with 4.0 for the Chester Pebble Beds, and there are fewer FeO-rich and more MgO-rich tourmaline grains (Figure 68) and (Figure 71), (Table 14). Therefore, although the source of the Chester Pebble Beds is likely to have continued to contribute to the basin fill, generating the relatively high MZi values seen in some Wilmslow Sandstone samples, the source of much of the material appears to have greater affinities with that supplying the Kinnerton, Collyhurst and Bold sandstones. However, there are some differences, notably an increase in ATi and RZi values and a decrease in the rutile/anatase ratio, which suggest that the source area evolved to some extent, possibly by unroofing through continued erosion.

Helsby Sandstone Formation

Eleven samples from the Helsby Sandstone were examined (nine from the Thornton Borehole and one each from the Wood Lane and Hondslough Farm boreholes), together with one sample from the overlying Tarporley Siltstone in the Coton Fields Borehole (Figure 2), (Table 1).As noted in the previous section, the Helsby Sandstone may also be represented at the top of the Bootle Borehole. Apart from the greater abundance of garnet in Coton Fields (probably because of less advanced dissolution), the Helsby Sandstone Formation has generally uniform characteristics. ATi is high, MZi is low, RZi is moderate and 35.6% of the TiO₂ group comprises rutile (Table 13). Overall, these parameters are very similar to those associated with the apatite-rich Wilmslow Sandstone.

In the Thornton Borehole, there is a slight difference between the mineralogy of aeolian and fluvial sandstones. The five aeolian sandstone samples have mean MZi values of 0.0, whereas the four fluvial sandstone samples have mean MZi values of 2.4. This suggests that the provenance of the fluvial sandstones has some affinity with that of the Chester Pebble Beds Formation, and that this influence is absent in the aeolian facies.

Three tourmaline populations were studied from the Helsby Sandstone, two from the Thornton Borehole and one from the Wood Lane Borehole. Their characteristics are similar to those from the Wilmslow Sandstone Formation (Figure 68) and (Figure 71), (Table 14), although they tend towards higher FeO and MgO values. This supports the evidence from the conventional heavy-mineral data, that the provenance of the Helsby Sandstone Formation is similar to that of the Wilmslow, but with some influence from Chester Pebble Beds sources.

The samples from the Helsby Sandstone Formation are characterised by a greater abundance of secondary minerals, notable carbonate and pyrite, than in the underlying sandstones. This may be because the Helsby Sandstone Formation underlies the impermeable MMG, which probably acted as a source and caprock to mineralising fluids (Warrington, 1980; Chapters 5 and 7, this volume).

Sm-Nd isotope study

The Nd-isotope signatures of whole rocks and pebbles from the basin fill were examined in order to obtain information about the rocks which were weathered and eroded to produce the detritus deposited in the Cheshire Basin during Permian and Triassic times. The isotope signatures of the sedimentary rocks are thought to represent the average isotope composition of their sources, as sedimentary processes tend to cause homogenisation. In contrast, the isotope signature of an individual clast from the Chester Pebble Beds Formation will reflect that of a specific source. By comparing the Nd-isotope values of the whole-rocks with those of the pebbles it was hoped to gain information concerning the provenance of the sediments, specifically the relative importance of the different sources in contributing to the basin fill. Comparison was made between the data obtained in this study and available data from possible source rocks.

Sample collection

Sedimentary rocks

Samples from the SSG were selected on the basis of the results of the first stage of heavy-mineral investigations. The samples (Table 1) came from boreholes in the north-west part of the basin and the area where the basin connects to the East Irish Sea Basin. A typical sample was selected from each of the four sandstone formations and, in addition, samples with unusual heavy-mineral contents from the Kinnerton Sandstone, Wilmslow Sandstone, and Helsby Sandstone formations were analysed, as these were considered more likely to have a secondary source component.

Five samples were collected from the MMG, covering the range of colour and grain size; they included examples of both structureless and laminated facies, ranging from red-brown to grey-green to light grey in colour. One sample of fine sandstone and siltstone was taken from the basal Tarporley Siltstone Formation and four comprising siltstones and mudstones from lithologies higher in the sequence (Table 1).

In addition, one sample of dark grey mudstone was collected from the Carboniferous (Kinderscoutian (R₁) stage of the Namurian) basement beneath the basin to provide a comparison with the Permo-Triassic basin fill.

Pebbles

Clasts from the Chester Pebble Beds Formation in the Cheshire Basin were analysed, and comparative data were obtained for similar conglomeratic units deposited in the more southerly (proximal) Needwood, Knowle and Worcester basins (Cannock Chase and Kidderminster formations).

Pebbles were selected from the Chester Pebble Beds at two outcrop localities in the north-west of the basin, within the same area as the whole-rock samples (Figure 72): road cuttings in the village of Dunham-on-the-Hill and a small quarry on the east side of the A56 (Table 2) and (Table 15). Several hundred pebbles were collected from each locality. A large amount of the vein-quartz material, obvious clasts of sandstones and mudstones (as these represent multicycle deposits), and pebbles too small for analysis (less than 1 cm in diameter) were discarded. After cleaning, cutting and thin sectioning, eight clasts were deemed suitable for analysis, in terms of size, composition and degree of weathering. The eight samples included a basalt, an acidic tuff, a schist and five quartzites.

Pebbles were also selected from a collection (held at the Lapworth Museum, Birmingham) of clasts from the Kidderminster Formation in the Knowle Basin and the north of the Worcester Basin (Table 2) and (Table 16); (Figure 72). Large pebbles, up to 4 cm in diameter, were selected, covering the range of igneous and metamorphic rock types: tourmalinite, rhyolite, quartzite, granite, gneiss and a variety of schists.

A large sample of syenite, from the Almington Gravel Works in the Staffordshire Basin, just outside the south-east margin of the Cheshire Basin, which was held in the BGS sample collections, was also examined. This pebble was of sufficient size (20 cm) to enable it to be dated by K-Ar and Rb-Sr methods.

Analytical methods

200 mg of sample was decomposed in a bomb at 120°C for three days using concentrated HF and HNO₃. The residue was converted to chloride using 6M HCl, and Sm and Nd were separated from the bulk REE fraction using columns filled with Biobeads coated with bis Di-ethyl hexyl hydrogen phosphate. Sm and Nd concentrations were obtained by isotope dilution using a mixed ¹⁴⁹Sm-¹⁵⁰Nd spike solution on a VG 354 multicollector mass spectrometer. Errors are quoted as two standard deviations from measured or calculated values. Analytical uncertainties are estimated to be 0.005% for ⁸⁷Sr/⁸⁶Sr, 0.003% for ¹⁴³Nd/¹⁴⁴Nd ratios, 0.5% for ⁸⁷Rb/⁸⁶Sr and 0.3% for ¹⁴⁷Sm^/144Nd ratios.

The syenite pebble from Almington was crushed, milled in agate and divided into two. One portion was sieved at 125–250 μm to yield individual grains and the second portion tema milled in agate to a powder. Amphibole for K-Ar analysis was separated from the 125–250 ìm fraction using a Franz electromagnetic separator. Potassium was analysed in duplicate by conventional mixed-acid digestion followed by flame photometry using a lithium internal standard. Argon was extracted by fusion under vacuum using external RF induction heating and analysed by the isotope dilution method in an MM1200 mass spectrometer. Ages were calculated using the decay and other constants recommended by Steiger and Jäger (1977) and errors are quoted at the 95% confidence level. The amphiboles from the K-Ar separation were analysed for Rb-Sr dating, as were alkali feldspars and the whole rock powder. The Sr and Rb were separated by a conventional ion-exchange technique, and ratios and concentrations by isotope dilution were measured on the Finnegan-MAT 262. Analytical uncertainties are estimated to be 0.005% for ⁸⁷Sr/⁸⁶Sr ratio, and 0.5% for ⁸⁷Rb/⁸⁶Sr ratio.

Results

The Sm-Nd results are given in (Table 17), with ε_Nd values calculated at approximately the time of deposition (240 Ma = Anisian, on the timescale of Forster and Warrington, 1985), and show a scatter of values between +0.7 and −12.6. The age chosen lies approximately at the SSG–MMG boundary (see Chapter 2 of this volume). In the fine-grained sedimentary rocks there are two fields of data, –3.0 to –3.7 and –7.6 to –11.2, whereas the pebbles scatter between –3.8 and –12.6, except for the more radiogenic syenite (CHB 343) at +0.7. There is no significant difference between the intraand extra-basinal pebble suites. Most pebbles have ε_Nd values in the range –3.8 to –8.2, with one schist at –10.8 and two quartzites and an acidic tuff lying between –12.1 and –12.6.

Discussion

There are many bordering and distal massifs which were available to contribute detritus into the basin during the Triassic, including: the Welsh Massif; the Lake District, the Southern Uplands and Highlands Massifs; the Anglo-Brabant Massif; the Cornubian Massif; and more distal sources under or beyond the English Channel, such as the Armorican Massif. Some of these source areas comprise rocks of similar age and composition, and consequently have similar isotope ratios. For example Lower Palaeozoic fine-grained sedimentary rocks with similar ε_Nd values are found in Wales, the Lake District, the Southern Uplands and south-west England. The palaeocurrent evidence for fluvial inputs from the south-west and aeolian transport from the east, rules out some of these sources, particularly those to the far north. Sources to the west, such as those of Wales and the Welsh Borders, could have provided lateral input to a northerly flowing river system, but could have contributed to aeolian sediments only indirectly through reworking of fluvial material. Each landmass comprises rock types which may have been added during different phases of crustal accretion; for example the Avalonian basement exposed in North Wales and along the Welsh borders formed part of the Welsh Massif, along with intrusive and extrusive Lower Palaeozoic igneous and sedimentary rocks. Extensive Nd-isotope data are available for each of these potential contributors to the Cheshire Basin (Evans, 1989; Leng and Evans, 1994).

In other areas data are less abundant, though some information is available from the igneous Midlands basement, which may have been partly exposed or formed part of the London-Brabant Massif (Henney et al., 1993), and from the Dalradian and Lewisian of the Grampian highlands (O'Nions et al., 1983). The Cornubian Massif, another potential source, would have been shedding Permian mudstone as well as material from the Hercynian granites (Darbyshire and Shepherd, 1994). Some data are also available from the Armorican Massif of north-west France and from Icartian and Biscay granulites (Bernard-Griffiths et al., 1985; Guerrot et al., 1989; D'Lemos and Brown, 1993; Darbyshire, unpublished data). Information on provenance was obtained by comparing the Nd-isotope ratios (depicted as ε_.Nd(240), values) for the whole rocks and pebbles with data from possible source areas, all calculated for 240 Ma and shown as (Figure 73). The results can be summarised as follows:

1. Samples from two of the sedimentary rocks (a non-pebbly whole rock from the Chester Pebble Beds (CHB 164) and the Carboniferous basement sample (DJD 577)) and four pebbles (CHB 269, 275, 461, 466), give unradiogenic ε_Nd(240) values of between –10 and –13. This range is common in the UK in ancient rocks, derived from multiply recycled Palaeoproterozoic terrains such as Gondwana (Lower Palaeozoic sedimentary rocks from the Midlands, south-west England, Wales, the Welsh Borders and the Lake District all have this isotopic signature; Thorogood, 1990) and thus the isotopic data are not diagnostic. However, the palaeocurrent evidence would particularly favour south-west England as a source, whilst the Midlands, Wales and the Welsh Borders could have supplied lateral inputs to the major fluvial system.

2. Most of the sedimentary rocks have a slightly more radiogenic signature (−7 to −9), between values for Gondwana-like terrains and those for other source areas such as Cornubia, Armorica, Avalonia and the Midlands, suggesting that a mixture of these two types of source may be represented. The second source may be similar to most of the pebbles, as they have  ε_Nd(240) values between –8 and –4. Of these sources, palaeocurrent data support a Cornubian or Armorican derivation. The sphericity of the pebbles and the degree of sorting suggest that the pebbles were transported great distances or reworked for long periods prior to deposition. Armorican faunas have been reported from the Budleigh Salterton Pebble Bed exposed on the south coast of England, which may occur at a similar stratigraphical level to the Chester Pebble Beds (Warrington and Ivimey-Cook, 1992). The tourmaline-quartz pebble (CHB 274) suggests input from the Cornubian granites. A mix of Midlands, Armorican and Cornubian sources in the sedimentary rocks would fit with the aeolian and fluvial palaeoflow information.

3. Two sedimentary samples CHB 84 and CHB 119 (from the Tarporley Siltstone and Wilmslow Sandstone formations) have ε_Nd(240) values of between –3 to –4. These values are similar to those of Armorica or the North Wales igneous suite. Palaeocurrent evidence would favour the Armorican source, but there could have been a lateral supply of sediment from the west into the major northward-flowing fluvial system.

4. The syenite boulder (CHB 343) is likely to have had a very local source, because of its size; other larger boulders were also present in the Almington Gravel Works. The source may be now buried, but the ε_Nd(240) values and the radiometric ages suggest that it was probably intruded during the same phase of activity as the Lower Palaeozoic igneous complex of North Wales.

The average K-Ar age for the syenite is 356 ± 15 Ma (i.e. older than Lower Carboniferous). This age represents the minimum age of the pebble and cannot be treated as a definitive age because of the likelihood of Ar loss from the amphiboles during chemical alteration. Rb-Sr isochrons were calculated using different combinations of mineral and whole rock. The results are inconclusive, but they clearly indicate a Lower Palaeozoic source:

5. Most of the sediments give crustal residence ages (T_DM) between 1.4 and 1.7 Ga, with most of the data for the pebbles lying between 1.3 and 1.9 Ga. These are typical of average upper-crustal values throughout the Phanerozoic, and are thus not diagnostic of source. They may reflect sediment reworking or sampling of rocks with a broad spectrum of ages (Mearns et at, 1989; Thorogood, 1990). Three of the igneous rock pebbles have slightly younger model ages: the syenite and a rhyolite from outside the Cheshire Basin, with values of 0.9 and 1.1 Ga respectively, and a rhyolite from the basin 1.2 Ga. These younger ages suggest incorporation of material from a second source.

The Nd-Sm isotope data thus tend to support other lines of evidence in favour of significant fluvial inputs from the Cornubian or Armorican areas to the south, mixed with more local influxes from the Midlands, the Welsh Borders or Wales.

Whole-rock geochemistry

The elements that give the best information for provenance studies are those contained in the more resistant detrital minerals — the phases which survive the processes of weathering, erosion, transport, deposition and diagenesis relatively unscathed. These elements include the Al group, REE, Y, Th, Sc, Nb, Co and Zr, all of which have very low sea-water to upper-crust partition coefficients and low residence times in sea water, and are strongly excluded from natural waters (Whitfield, 1979; Turner et al., 1980; Taylor and McLennan, 1985; McLennan, 1989). Elements of the Al group tend to have limited value because they have low dispersal in common igneous rocks and fairly constant ratios to other elements. REE may suffer diagenetic migration and fractionation (Burnotte et al., 1989; Milodowski and Zalasiewicz, 1991a), although this may not affect their bulk geochemistry. A further potential problem is the recycling of sedimentary detritus, which tends to reduce geochemical contrast and leads to complex mixes of sources. It also tends to introduce a bias towards mature grains which are less useful for discerning source (Haughton et al., 1991).

The mobile elements Ca, Mg, Ba, Mn and Na largely reflect diagenetic carbonates, sulphates and halides. In red beds, labile minerals are readily broken down under oxidative weathering to produce iron-oxides and clays. More mobile elements may be removed into groundwaters. Resistate minerals may survive this process and still provide information on provenance, and immobile elements in less stable phases may be retained in the bulk volume of rock after the breakdown of their original host minerals, by adsorption onto clays or oxide coatings or incorporation into the lattice of clays.

Sample collection, preparation and analysis

More than 580 samples were collected for chemical analysis, including more than 300 from the SSG and over 200 from the MMG (Table 1) and (Table 2); (Figure 2). The stratigraphical position of some samples is uncertain, and these are categorised in the text that follows as ?Helsby, etc.

A standard scheme of preparation and chemical analysis was followed. After cleaning, the cores were jaw crushed twice, riffle split and then ground in an agate tema mill. The resulting powder was coned and quartered and a portion reduced to a <50 μm particle size by processing for 20 minutes in a P5 agate planetary ball mill 12 g of this powder was mixed with elvacite binder and processed for a further 10 minutes in the ball mill, reducing the particles to <30 μm, before pressing the mixture into a pellet for trace-element XRF analysis. Major elements were also determined by XRF, on lithium tetraborate fusion beads. A few of the samples from the Wilkesley Halite, which consisted largely of halite with minor mudstone, were placed in water to remove the salt, and the residue was processed in the same way as the whole-rock samples.

Approximately 50 samples from the SSG and the Tarporley Siltstone Formation were analysed for REE by ICP-MS.

Presentation and interpretation of geochemical data

Several approaches were taken in displaying and interpreting the geochemical data. Factor analysis was carried out to examine the inter-relationships between different elements and to provide information on the mineral phases with which the elements were associated. The aim was also to assess whether elements were likely to provide information relevant to provenance, diagenesis or other aspects of the study. The analysis was undertaken on the SSG and MMG as a whole, and on the individual component formations.

Overall relationships were examined at the stratigraphical group and formation levels, and to some extent within formations, by the use of Exploratory Data Analysis, in particular the use of box-and-whisker plots (e.g. Tukey, 1977; Chambers et al., 1983; Kûrzl, 1988).

In .these plots the median is used as the measure of the centre of the population, with a box defined by upper and lower hinges which are esssentially the upper and lower quartiles (Figure 74). The difference between the hinges is termed the h-spread. Fences are defined at both ends of the box at 1.5 times the h-spread (inner fences) and three times the h-spread (outer fences). Whiskers extend from the box to the extreme values within the inner fences. Sample values between the inner and outer fences are shown as possible outliers, whilst probable outliers are depicted outside the outer fence. Notches are used to differentiate between populations: they represent confidence intervals relative to the median such that, if the notches do not overlap, the two populations are significantly different at the approximately 95% confidence level. The technique is robust and resistant to outliers: up to 25% of the data can be wild without affecting the median or the box.

Individual samples were compared by the use of normalised multi-element diagrams (spidergrams). These diagrams enable all the data to be seen at a glance, and comparison to be made between one suite and another. The normalising values used are those for average upper continental crust (Taylor and McLennan, 1985, p.46; except P₂O₅ from Weaver and Tarney, 1984). Thus a detrital sediment derived, unmodified, from a source area of average crustal composition would plot as a horizontal line of normalised value equal to 1.0, and the manner in which the pattern departs from this horizontal line reflects the degree to which the source rocks differ from average upper continental crust and the ways in which they were modified by the chemical and physical processes of weathering, transport, deposition and diagenesis.

Box-and-whisker plots and spidergrams are complementary approaches, the former often suggesting possible relationships which can be tested by the use of the latter.

Factor analysis

R-mode factor analysis was used, with varimax factor rotation. SiO₂ was excluded from the analysis, to avoid the problem of data closure, and U and Mo were omitted because there were few determinations above the detection limits. In some datasets one or two other elements were also excluded, where they produced spurious factors, and the factor analysis was rerun (Ba and Sr in the Kinnerton Sandstone, Na and Pb in the Wilmslow Sandstone, Pb in the Helsby Sandstone, Na in the Bollin Mudstones). The Helsby Sandstone data were run once with all samples included and again with mineralised samples from West Mine and Clive Mine omitted.

Sherwood Sandstone Group

Factor analysis of the SSG as a whole (Table 18) reveals aluminosilicate, mineralisation and carbonate factors. Factor analysis of individual formations shows many common features between formations, but also some differences. The first factor is always an aluminosilicate factor (clay minerals, micas, feldspars) with significant factor loadings in every dataset for Al₂O₃, TiO₂, K₂O, Ce, Cr, La, Nb, Rb, Th and Y. Also included in the first factor are Fe₂O₃ and V (though not for the Kinnerton Sandstone), P₂O₅ (not Kinnerton and Wilmslow sandstones), Sr (not Wilmslow Sandstone) Na₂O and Zr (not Chester Pebble Beds or Wilmslow Sandstone), MgO (not Chester Pebble Beds or Helsby Sandstone), Co, Ni and Zn (Chester Pebble Beds and Wilmslow Sandstone only), Ba (Kinnerton Sandstone only). This factor could also include secondary iron oxides, which are more abundant in the clay-rich rocks of the SSG. These oxides can adsorb significant levels of metals, such as V, Cr, Co, Ni and Zn, which may explain the presence of Fe₂O₃ and these metals in this factor.

The association of many elements with aluminosilicate minerals is also clear from correlation matrices of the elements; generally, many elements correlate closely with Al₂O₃including TiO₂ Fe₂O_3, K₂O_, Ce, La, Nb, Rb and Y. In addition, Ni and Cr correlate with Al₂O₃ in all formations except the Helsby Sandstone. Ba, Sr, Th, V and Zn also correlate with Al₂O₃ in the Chester Pebble Beds and Sr, Th, V and Zn in the Wilmslow Sandstone.

A common factor in all units (either second or third in terms of the percentage of variance it covers) is a carbonate factor including CaO, LOI and MnO (only CaO and MnO for the Kinnerton Sandstone). The absence of MgO from this factor probably reflects the fact that MgO resides in clays as well as carbonate, and should not be interpreted as suggesting that calcite is a more important carbonate mineral than dolomite. Both dolomite and calcite are present as diagenetic phases and both are manganoan (Chapter 5; Milodowski et al., 1994; Jones and Haslam, 1994). MgO is present in this factor for the Helsby Sandstone, suggesting that in this formation dolomite is the most important MgO mineral. This is supported by petrographical observations, particularly of eodiagenetic nodular dolomite, in the Helsby sandstone (p.127) and by geochemical assessment of preserved carbonate contents (p.152).

The second factor for the Kinnerton Sandstone includes Fe₂O₃, Co, Ni and V, suggesting an iron-oxide or sulphide association. The Chester Pebble Beds and Wilmslow Sandstone have a probable heavy-mineral factor, dominated by Zr, with a less significant score for Th in the Chester Pebble Beds and P₂O₅ in the Wilmslow Sandstone. Baryte (with or without other mineralisation) emerges as a probable minor factor in all but the Kinnerton Sandstone.

Taken as a whole, the MMG does not show a very coherent result from factor analysis, probably because many elements are present in more than one mineral. Slightly clearer results are obtained if elements are normalised against the aluminosilicate content (e.g. as ratios to TiO₂; (Table 19)). This produces (1) a clay (aluminosilicate) factor with K₂O, Rb, Al₂O₃, P₂O₅, (2) a factor with positive Ce, Zr and Y scores and negative Cr, Ni, Fe₂O₃ and V scores (heavy mineral versus iron-oxide or acid versus basic), (3) a carbonate (probably dolomite) factor (LOI, MgO, MnO, CaO) and (4) a second possible heavy-mineral factor (La, Ce, Y).

When individual formations of the MMG are considered (Table 19) the dominant factor is, as in the SSG, an aluminosilicate one which includes Al₂O₃, TiO₂, K₂O, Ce, Cr, La, Nb and Rb in all formations. However, unlike the SSG, Fe₂O₃, P₂O₅ and Ni are always present, but Th and Y do not always appear. Dolomite is important, as MgO usually appears with LOI and MnO, but not always CaO. This is confirmed by petrographic work, which showed that dolomite cements with late Mn-rich overgrowths are dominant rather than calcite. CaO is absent from the carbonate factor, probably because of the complication introduced by the presence of gypsum and anhydrite. Anhydrite is a major primary precipitate in many of the mudrocks and is associated with later replacive gypsum (Chapter 5; Milodowski et al., 1994).

Minor factors include:

• Tarporley Siltstones: As, Co, Pb
• Bollin Mudstones: As, Co, Ni (? sulphides)
• Northwich Halite: Zr
• Byley Mudstones: Ti, P, Ba, Ce, La, Nb, Y, Zr
• Wych Mudstones: La, Y, Zr (heavy minerals)

As-Co-Ni sulphides, plus galena, pyrite, Zn and Cu sulphides have been seen in petrographic examination (Chapter 5; Milodowski et al., 1994). Other minor factors are not geologically reasonable and probably reflect the variable mineralogy, in which evaporites (including halite) and exotic clay minerals (such as corrensite) are commonly, but not ubiquitously, present.

It is therefore possible to separate elements related to heavy minerals and aluminosilicates, which may provide information on provenance, from others associated with cement phases or mineralisation which give no such information.

The conclusions reached here are similar to those of Haslam and Sandon (1991) which were based on a similar statistical approach.

Sherwood Sandstone Group

Comparison between formations

The Kinnerton and Wilmslow sandstones display less variability than the Chester Pebble Beds and the Helsby Sandstone, and the interquartile range for most elements is usually narrower (Figure 75). Samples taken from the Kinnerton and Wilmslow sandstones are dominantly of aeolian facies, whereas the Pebble Beds are mainly fluvial and the Helsby samples include a significant number from fluvial facies. The range of lithologies in the fluvial sediments, from coarse-grained channel deposits to fine-grained overbank material, is generally much greater than that of the aeolian deposits, which would explain the wider range of geochemistry.

The Chester Pebble Beds and the Kinnerton Sandstone have higher Fe₂O₃/Al₂O₃ but lower Zr/Al₂O₃ than the other formations (Figure 76).

The Helsby and Kinnerton sandstones have lower K₂O levels than the other SSG formations. It is noteworthy that K₂O/Al₂O₃ decreases as values of both these elements increase (Figure 77). This suggests that K-feldspar (with higher K₂O/Al₂O₃) is a more important influence at lower levels and micas and clay minerals (with lower K₂O/Al₂O₃) have more influence at higher levels. A similar effect is seen in the relationships of some other elements with Al₂O₃ (e.g. Rb and, in some datasets, Ba and Sr).

The Kinnerton Sandstone differs in many respects from most or all of the other units (Figure 75) and (Figure 76). It has higher SiO₂ than all other units except the ?Helsby and Grinshill sandstones, higher Fe₂O₃ than the Wilmslow Sandstone and later units and low Al₂O₃, TiO₂, P₂O₅, Th, Y except compared to Helsby, ?Helsby and Grinshill sandstones. Concentrations of Na₂O, Ba, Sr, Rb, As, Pb are low except compared to the Grinshill Sandstone, K₂O levels low except compared to the Helsby and Grinshill sandstones and Zr low except compared to the Chester Pebble Beds and Helsby Sandstone.

The Fe₂O₃/Al₂O₃ ratio is higher than in all the other units, Ce/Al₂O₃ is higher than in all formations except the Grinshill Sandstone, and Ni/Al₂O₃ is high except compared to the Helsby and Grinshill sandstones (Figure 76). The Kinnerton Sandstone has low Rb/Al₂O₃ except in comparison to the Grinshill Sandstone.

The Kinnerton Sandstone is predominantly an aeolian unit, deposited early in the basin's history, prior to the first major influx of fluvial sediments represented by the Chester Pebble Beds. Its distinctive geochemistry probably reflects a lack of fluvially transported detritus from the south. Palaeoflow indicators show that it was deposited from easterly winds, suggesting inputs from the southern Pennines or Midlands. All units deposited subsequent to the Chester Pebble Beds, as well as that unit itself, have some geochemical signature related to the fluvial material from the south.

Comparison between Kinnerton And Wilmslow Sandstones

It is interesting to compare the Kinnerton and Wilmslow sandstones, as both are predominantly aeolian, particularly in the boreholes sampled. Such a comparison should provide some insight into the nature of the wind-blown sediment and the importance of fluvially transported sediment in the younger unit, either as a direct fluvial input or reworked by aeolian processes.

Both sandstones generally have higher SiO₂ than other units, but the Kinnerton Sandstone has higher values than the Wilmslow Sandstone. The Wilmslow Sandstone has higher Al₂O₃, Na₂O, TiO₂, K₂O, MgO, P₂O₅, Sr, Ba, Rb, Th, Zr, Y, Ce, La, Cu, As, and Pb (Figure 75). As was noted on p.90, the Kinnerton Sandstone is dominated by cherry sublitharenites, thus explaining the high SiO₂ content, whilst the Wilmslow Sandstone is more argillaceous and, therefore, richer in Al₂O₃ and related elements.

The concentrations of most elements are strongly influenced by the effects of variable dilution by quartz. To allow for this, the results were normalised to alumina. The Wilmslow Sandstone has higher Na₂O/Al₂O₃, K₂O/Al₂O₃, Rb/Al₂O_3, Ba/Al₂O_3, Sr/Al₂O₃  Zr/Al₂O_3, and Th/Al₂O₃ (Figure 76). These differences suggest higher feldspar, zircon and monazite contents, consistent with inputs from a granitic source area such as Cornubia or Armorica. The Wilmslow Sandstone Formation includes arkosic and subarkosic rocks, which were not seen in the Kinnerton Sandstone (pp.90–91). However, as was noted earlier, there are diagenetic monazites present in the SSG, as well as those from igneous sources; Cornubia or Armorica could be the source of these, but they could equally be supplied from other areas of Palaeozoic rocks in Wales or the Midlands.

In contrast, the Kinnerton Sandstone has higher Fe₂O₃, CaO, LOI, Zn, Fe₂O₃ /Al₂O₃, Ce/Al₂O_3, Ni/Al₂O₃, and V/Al₂O₃. These reflect greater carbonate content (Jones and Haslam, 1994) and the more highly developed iron-oxide cement (Chapter 5; Milodowski et al., 1994). The higher Ni/Al₂O₃ could indicate a more basic igneous input, possibly from Carboniferous volcanics, although volcanic clasts observed in the formation (p.91) appear to be of acid or felsic composition.

There is no significant difference in MnO content between the two sandstones, but the levels are low in both compared with those of the Chester Pebble Beds and the Helsby Sandstone.

Geographical variations

It is difficult to test geographical variations within formations rigorously because of the uneven distribution of samples, but some differences can be seen for the Kinnerton and Wilmslow sandstones. For example, K₂O/Al₂O₃values for the Kinnerton and Wilmslow sandstone are higher in the south of the basin than in the northern half (Figure 78). This suggests that K-feldspars are more important (relative to micas and clay minerals) in the south than in the north.

The other notable variation is in Ni/Al₂O₃ (Figure 78). In the Kinnerton Sandstone this ratio is significantly higher in the south than the north, which may be related to the greater variety of lithic clasts seen in the south (p.91). The Wilmslow Sandstone, in contrast, has lower Ni/Al₂O₃ in the south. This may be related to the high value of the K₂O/Al₂O₃ ratio in the south: if a large proportion of the Al₂O₃ is contained in feldspar and a correspondingly lower proportion in clays, and if the Ni is predominantly in the clay fraction, the Ni/Al₂O₃ ratio is likely to be low.

The Wilmslow Sandstone data may thus be interpreted in terms of a decline in coarser-grained fluvial detritus northward, as evidenced by the decrease in pebble content of the Chester Pebble Beds in this direction, with an associated decrease in feldspar and increase in clay-mineral contents. This is consistent with the decrease in grain size from proximal to distal facies, along the general northerly palaeoslope, noted by Burley (1984). The differences in the Kinnerton Sandstone suggest variations in source area.

Facies variations

The geochemistry of samples examined sedimentologically was assessed by comparing samples from aeolian dune, interdune, fluvial channel, mudflat and playa environments (Figure 79). The most obvious difference is that the finer-grained playa and mudflat facies are poorer in SiO₂ and richer in most other elements.

Compared with the fluvial facies, the aeolian facies is richer in SiO₂, P₂O₅, Co and Cu and poorer in CaO, MgO, MnO, LOI, Pb and Th. The aeolian facies are thus richer in quartz, have comparable alumina contents to fluvial deposits, and have less carbonate cement.

The aeolian sandstones have lower K₂O/Al₂O₃ and Fe₂O₃/Al₂O₃ but higher TiO₂/Al₂O₃, Zr/Al₂O₃ and Cr/Al₂O₃. This suggests a lower ratio of K-feldspar to mica plus clay minerals and higher levels of heavy minerals such as zircon and possibly chromite.

Rare-earth elements

REE analysis was carried out on selected samples from the SSG and the Tarporley Siltstone Formation to assess whether the patterns would provide any useful information on provenance. Chondrite-normalised plots show a general decline in values from LREE to HREE, with a small negative Eu anomaly (Figure 80). The main difference observed is related to grain-size. The finer-grained samples have higher contents of REE and different normalised patterns. On a chondrite diagram they show a steeper slope from LREE to HREE and a slightly more pronounced Eu anomaly. The difference is more marked when the data are normalised to Post-Archaean average shale (Taylor and McLennan, 1985). On such diagrams (Figure 80) the finer-grained samples show flat profiles close to 1; i.e. they are very similar to average shales. The coarser samples are depleted in total REE compared to average shale, an effect which is most marked for the LREE. The total depletion is consistent with observations (e.g. Cullers et al., 1979) that REE are concentrated in the fine fraction.

When plotted against Al₂O₃ (Figure 81), the LREE show a better correlation than the HREE, indicating that they are preferentially associated with aluminosilicates. Plots of La against Al₂O₃, TiO₂ and K₂O show a strong correlation, but with a change of slope at higher values. This suggests a change of host mineral with changing grain size. There is much more scatter on plots of these elements against Lu, revealing a more variable mineral association. Plots of La against other REE (Figure 82) show a very close correlation, as would be expected, with the adjacent REE, Ce, Pr and Nd. For Sm and heavier REE the correlation is slightly less strong and there is a change of slope, with lesser concentrations of the heavier REE at higher levels, compared to La. This is very apparent on the La v Lu plot, on which the samples divide into two groups, the finer-grained samples having lower Lu/La ratios.

In conclusion, there appears to be a mineralogical control on the REE in the SSG and the Tarporley Siltstone. This suggests that the variation is the result of changes in relative mineral concentrations, associated primarily with transport and deposition rather than original source area.

In order to remove grain-size effects, the La/Lu ratio was normalised to Al₂O₃, and box-and-whisker plots were drawn (Figure 83). These show a marked difference between the Kinnerton Sandstone and other units, consistent with its rather different general geochemistry. This may be related to the fact that the Kinnerton Sandstone is mainly of aeolian facies, deposited before the first major fluvial influx of sediment into the basin represented by the Chester Pebble Beds. Later aeolian facies were probably, at least in part, reworking exotic material introduced from the south by rivers, while fluvial facies of the Wilmslow and Helsby sandstones are known to have this southerly derivation (Thompson, 1985).

Mercia Mudstone Group

Comparison between formations

An overall comparison of the different lithostratigraphical units was made using box-and-whisker plots (Figure 84) and (Figure 85). The Tarporley Siltstones have higher SiO₂ than other units, and higher K₂O (except for the Brooks Mill Mudstones). The unit has lower levels of Na₂O (except compared to the ?Bollin Mudstones), LOI, CaO, MgO and Sr. The higher SiO₂ reflects the coarser grain size and higher quartz content. Al₂O₃ contents are, however, not significantly different, although the Tarporley Siltstones have the highest median content; this suggests overall that the detrital component, including K-feldspar and mica, is greatest in this unit. This is borne out by petrographic study which shows a high siliciclastic component, very similar to that of the underlying SSG. The carbonate content, on the other hand, is lower.

The Byley Mudstones are generally distinct from the other units, but most similar to the Wych Mudstones and ?Bollin Mudstones. They have lower SiO₂ (except in comparison with the Stafford Halite and ?Bollin Mudstones), TiO₂ (except for the ?Bollin Mudstones and the Wilkesley Halite) . and Zr. They are generally lower in Nb, Y, Co, Ce, Cu, Ba, La, Rb and V, but higher in LOI, CaO and MgO. Anhydritic and dolomitic rocks are well developed in the Byley Mudstones and, to a lesser extent, in the upper Bollin and Wych mudstones.

Mudstones above the Tarporley Siltstones near Stafford (Stafford Basin), which were assigned as ?Bollin Mudstones, seem to be geochemically most similar to the Byley Mudstones of the Cheshire Basin.

For some elements the lower part of the MMG (Tarporley Siltstones to Northwich Halite) is similar to the Brooks Mill Mudstones, but somewhat different from the middle part of the succession (Byley Mudstones to Wilkesley Halite). Thus the middle part of the succession has generally lower SiO₂, TiO₂, Nb, Zr, Y, Ce, Co and La. The common pattern of these elements is for similar, and relatively high levels in the Tarporley Siltstones, Bollin Mudstones and Northwich Halite, a marked fall in the Byley Mudstones, and then a consistent rise in median concentrations up to the Brooks Mill Mudstones. The petrography of the Tarporley Siltstone and Brooks Mill Mudstone is similar, with both formations containing a significant arenaceous siliciclastic component.

In order to remove the possible effects of dilution by quartz, and hence changes in aluminosilicate mineral contents, elements were plotted as ratios to Al₂O₃ (Figure 85). The Tarporley Siltstones have higher K₂O/Al₂O₃ than other units except the ?Bollin Mudstones. Values are also generally high in the Bollin and Brooks Mill Mudstones, and in the Stafford Halite. This ratio appears to be related to grain size. There is a good correlation between K₂O and Al₂O₃, but the ratio is generally higher in sandstones than in mudstones. This probably reflects a higher proportion of K-feldspar to phyllosilicates in the sandstones than the mudstones, and/or differences in phyllosilicate composition.

The Bollin Mudstones, ?Bollin Mudstones and Northwich Halite have higher TiO₂/Al₂O₃. The Tarporley Siltstones and Bollin Mudstones have lower Sr/Al₂O₃.

The middle (Wych and Byley) mudstones have higher Na₂O/Al₂O₃ than the other non-halitic units. The ?Bollin Mudstones and the Stafford Halite have low Fe₂O₃/Al₂O₃ compared with all except the Tarporley Siltstones. The Wych and Byley mudstones have low Y/Al₂O₃, Ce/Al₂O₃, Zr/Al₂O₃, La/Al₂O₃ and Rb/Al₂O₃ (except compared with the Wilkesley Halite) and high Cr/Al₂O₃, Ni/Al₂O₃ compared with most other units.

The gross differences between formations were also examined by comparing geochemical profiles in the Wilkesley Borehole: samples show higher Cr/Al₂O₃ and Ni/Al₂O₃ and lower Zr/Al₂O₃ in the Wych and Byley mudstones than in any of the other formations (Figure 86).

Geographical variation is also apparent within the Byley and Wych mudstones, as the Cr/Al₂O₃ and Ni/Al₂O₃ ratios are higher and the Zr/Al₂O₃ ratio lower, in the Crewe, Wilkesley and Lower Wych boreholes than in other boreholes (Figure 87).

These variations in Cr/Al₂O₃, Ni/Al₂O₃ and Zr/Al₂O₃ suggest the presence of two compositional types in the source material, a basic igneous composition (represented by high Cr/Al₂O₃ and Ni/Al₂O₃) and an acid igneous or sedimentary composition (represented by high Zr/Al₂O₃).

As a contrast to the above relationships, there is a positive correlation between Zr/Al₂O₃ and Cr/Al₂O₃ in the lower part of the Wilkesley Halite of Wilkesley Borehole (Figure 86). This implies that here the similar behaviour of the heavy minerals zircon and chromite during transport and deposition is the dominant influence, rather than any difference in provenance.

Comparison between facies

Two broad facies types can be recognised in the MMG: (1) laminated mudstones and siltstones and (2) structureless mudstones (Arthurton, 1980; Wilson, 1993; and see Chapter 2 of this volume). Both facies are present in the Byley Mudstones of the Crewe Borehole. Samples of each type from similar depths were compared (Figure 88). The samples from the structureless facies were all oxidised; those from the laminated facies were dominantly oxidised, though with some reduced patches. There is no significant geochemical difference between the laminated and structureless facies. The detrital material, whether water-laid or deposited on dry land, is of wind-blown origin (Arthurton, 1980), and thus is of similar provenance in the two facies types. The influence on chemical composition exerted by depositional conditions (as represented by the different facies types) thus appears to be subordinate to the influence of provenance and diagenesis.

The Saughall Massie Borehole (Figure 89) shows a contrast between the Tarporley Siltstones and the underlying Helsby Sandstone. There is little variation in Cr/Al₂O₃, but the Ni/Al₂O₃ profile is almost a mirror image of that of the Zr/Al₂O₃ ratio throughout much of the Tarporley Siltstones, possibly reflecting input from different sources. In the Helsby Sandstone and basal Tarporley Siltstone, in contrast, the Ni/Al₂O₃ and Zr/Al₂O₃ profiles run almost parallel, with Cr/Al₂O₃ showing a similar trend in the Delamere Member, suggesting that sedimentological factors were the dominant influence.

Comparison of the Manchester Marls Formation with the Mercia Mudstone Group

There are strong differences between the Manchester Marls and the MMG. Although SiO₂, CaO and LOI in the former lie between the levels of the Tarporley Siltstones and other MMG units, Al₂O₃, TiO₂, Fe₂O₃, K₂O, Cr, Ce, Co, Pb, La, Rb, Nb, V and Y are higher than in the MMG and Na₂O and MgO lower (Figure 84). This suggests generally that the Manchester Marls have higher quartz and aluminosilicate contents than the MMG (i.e. a greater detrital component) and lower levels of carbonates, sulphates and, perhaps, Mg-rich clays.

When the data are normalised to Al₂O₃ (e.g. (Figure 85) Na₂O/Al₂O₃, K₂O/Al₂O₃ and P₂O /Al₂O₃ are lower than all the MMG (except for the Wilkesley Halite in the case of K₂O and Stafford Halite in the case of P₂O₅). Many elements are similar to certain groups of the MMG but differ from others. The Bollin Mudstones are most similar overall (TiO₂, Cr, La, Ni, Ce, Zr, Rb, Sr, V) but differ for Nb and Th, as well as the major elements referred to above.

As a contrast to the above relationships, there is a positive correlation between Zr/Al₂O₃ and Cr/Al₂O₃ in the lower part of the Wilkesley Halite of Wilkesley Borehole (Figure 86). This implies that here the similar behaviour of the heavy minerals zircon and chromite during transport and deposition is the dominant influence, rather than any difference in provenance.

Comparison between facies

Two broad facies types can be recognised in the MMG: (1) laminated mudstones and siltstones and (2) structureless mudstones (Arthurton, 1980; Wilson, 1993; and see Chapter 2 of this volume). Both facies are present in the Byley Mudstones of the Crewe Borehole. Samples of each type from similar depths were compared (Figure 88). The samples from the structureless facies were all oxidised; those from the laminated facies were dominantly oxidised, though with some reduced patches. There is no significant geochemical difference between the laminated and structureless

Discussion and conclusions

The different approaches described in this chapter provide complementary information on provenance. Some of the studies relate to the whole rock (geochemistry, petrography and isotope studies), while others relate to specific constituents (heavy-mineral studies, and isotopic studies of individual pebbles).

The SSG was laid down within a long northward-draining fluvial basin system, originating in the Armorican Highlands to the south of Britain and developed along a series of linked Permo-Triassic extensional basins along the west side of the country (e.g. Audley-Charles, 1970; Warrington, 1970b; Warrington and Ivimey-Cook, 1992; Chapters 2 and 3). The fluvial system extended from the Wessex Basin in the south, through the Worcester Basin, Stafford Basin and Cheshire Basin, into the East Irish Sea Basin and the basins of the Solway, Northern Ireland and the Firth of Clyde in the north.

Whole-rock petrography

The lithic clasts and heavy minerals seen in whole-rock petrographic studies of the Kinnerton Sandstone Formation indicate derivation from a complex source terrain, which included high-grade metamorphic rocks, syenite, acid volcanics, quartzite, sandstones and siltstones. The reworked monazite nodules, noted in this formation in the Halewood Borehole, resemble those described from Lower Palaeozoic sedimentary rocks in Wales, south-west England, Brittany and Belgium (though none have been reported from northern Britain), suggesting derivation from one of these areas.

The whole-rock petrography of the Chester Pebble Beds suggests a similarly mixed source terrain. The increase in lithic clasts (chert, sandstones, quartzite, silicified volcanics and bioclastics) and feldspar compared to the underlying formation indicates changes in the relative importance of different sediment sources. The greater importance of albite and reduction of volcanic fragments in the Wilmslow Sandstone Formation suggests a continued evolution of the source terrain. This trend continued into the Helsby Sandstone, which also, locally, shows clear evidence for inputs from the Lower Palaeozoic of the Midlands or Wales. The feldspar may reflect the increased importance of granitic inputs from Armorica or Cornubia, whilst the sedimentary clasts suggest more local sources in the Carboniferous of the Anglo-Brabant Massif, Pennines or Wales and perhaps the Devonian. Clasts in the Chester Pebble Beds in the Cheshire Basin, as in other basins, are dominated by quartzite and vein quartz. The quartzites rarely contain derived Lower Palaeozoic fossils, and chert and limestone pebbles rarely include Carboniferous fossils (Thompson, 1970a). Previous studies of the Kidderminster Formation of the West Midlands and Budleigh Salterton Pebble Beds of south-west England equivalent formations to the Chester Pebble Beds – have identified components of Cornubian and Armorican affinity (e.g. Wills, 1950, 1956, 1976; Fitch et al., 1966). Certain igneous clasts have been compared with rocks in south-west England (Campbell Smith, 1963), whilst faunas in quartzite pebbles are most similar to those of the Armorican Peninsula of Brittany and Normandy (Cocks, 1993).

With the exception of the Kinnerton Sandstone Formation, which displays an apparent decrease in the proportion of polycrystalline quartz to single-crystal quartz and a decline in the content and change in the composition of lithic fragments from south to north, the SSG shows no gross variations in detrital mineralogy across the Cheshire Basin. However, there are significant differences between the SSG of the Cheshire Basin and other UK Permo-Triassic basins. Changes in sediment composition may be related to dilution of the southerly-sourced sediment by more locally derived material along the palaeo-drainage system. This change is accompanied by a gradual decrease in grain size, and a change from proximal to distal fluvial facies from south to north (Burley, 1984).

In the Cheshire Basin, sandstones are characterised by moderately low feldspar contents (typically between 5 and 10% feldspar, which is dominantly K-feldspar), and are represented by cherry litharenites and sublitharenites, with quartz arenites becoming more significant in the Helsby Sandstone Formation. The SSG of the Wessex Basin is significantly different in character and is represented by very feldspathic sandstones: lithic arkoses, arkoses and subarkoses with up to 55% K-feldspar (Burley, 1984; Knox et al., 1984; Milodowski et al., 1986; Strong and Milodowski, 1987; Bath et al., 1987). The SSG of the Lincolnshire–East Yorkshire Basin is similarly feldspathic, although the proportion of K-feldspar (10–35%) is lower than for the Wessex Basin (Milodowski et al., 1987; Bath et al., 1987). The St Bees Sandstone of the Cumbrian coast and the East Irish Sea Basin (cf. Strong et al., 1994) also appears to be significantly more feldspathic and chert-free than equivalent Cheshire Basin strata (Kinnerton Sandstone Formation, Chester Pebble Beds Formation and Wilmslow Sandstone Formation). Furthermore, the SSG of the Wessex Basin (and to some extent in the Lincolnshire–East Yorkshire Basin) includes a high proportion of detrital Ba-rich K-feldspar. This is not a characteristic of the SSG of the Cheshire Basin.

The Ormskirk Sandstone of the East Irish Sea Basin, and the Bromsgrove Sandstone Formation of the Stafford, Knowle and Worcester basins (stratigraphical equivalents of the Helsby Sandstone Formation), represented mainly by quartz arenites, sub-litharenites and subarkoses, are very similar to the Helsby Sandstone Formation of the Cheshire Basin (Ali and Turner, 1982; Burley, 1984; Strong, 1993). The Bromsgrove Sandstone Formation, in particular, contains similar igneous and metamorphic fragments, dominated by altered acid volcanics. However, whereas detrital chert is prominent in the SSG of the Cheshire Basin it appears to be less significant in adjacent Permo-Triassic basins. Since chert is not as important in other basins, the most likely source is from the common cherry horizons in Dinantian limestones of North Wales, the Welsh Borders, the south Pennines and the Midlands (Anglo-Brabant Massif). This suggests that Namurian and Westphalian strata had already been at least partly eroded from these areas. There is evidence for this in the South Pennines, where neptunian dykes in Dinantian limestones in the Ashbourne area are filled with probable Triassic sediment, and some pocket deposits in the limestones may be of Triassic age (Aitkenhead et al., 1985; Chisholm et al., 1988). Areas where Dinantian strata are known or inferred to subcrop beneath Permo-Triassic rocks are outlined south of the Pennines, in the Welsh Borders and in the area of the London–Brabant Massif (British Geological Survey, 1985a; Whittaker, 1985).

Heavy-mineral studies

The heavy-mineral study has provided constraints on provenance and diagenesis. Variations in heavy-mineral assemblages and the chemistry of detrital tourmaline indicate that a number of distinct sources supplied sediment to the Cheshire Basin during the Permo-Triassic, and that these varied in relative importance with time. In broad terms, two types of assemblage were recognised, one with abundant detrital monazite (i.e. not including diagenetic monazite nodules) and Fe-rich tourmalines, and one with lower detrital monazite abundances and Mg-rich tourmalines. The former characterises the Chester Pebble Beds Formation, whereas the latter is typical of both the earlier (Collyhurst Sandstone Formation, Manchester Marls Formation, Bold Formation and Kinnerton Sandstone Formation) and later (Wilmslow Sandstone Formation and Helsby Sandstone Formation) sandstones (Table 12)." data-name="images/P1000323.jpg">(Figure 69) and (Figure 71).

The provenance of the sandstones is best viewed, on the basis of the heavy-mineral data, as resulting from the interplay between the major Armorican or Cornubian sourceland to the south of the UK proposed in previous palaeogeographical reconstructions (Wills, 1950; Warrington and Ivimey-Cook, 1992) and more local sources adjacent to the Cheshire Basin. The maximum influence of the southerly source took place during deposition of the Chester Pebble Beds Formation, coinciding with the time of greatest runoff from this source and ensuing higher-energy depositional conditions. The heavy-mineral assemblages associated with these sediments are relatively rich in detrital monazite, and tourmaline suites are dominated by Fe-rich varieties. The available regional information on Triassic sandstones of the UK shows an overall decrease in the abundance of detrital monazite from south to north (Figure 90), indicating that the southerly source area was characterised by high amounts of detrital monazite: the progressive northward decrease in MZi values is interpreted as the result of dilution by more local material low in detrital monazite.

More local sources were of greater importance in the preceding Collyhurst, Kinnerton and associated sandstones and the overlying Wilmslow and Helsby sandstones, in which detrital monazite is markedly less abundant and tourmaline suites are generally more Mg-rich. There are minor regional and stratigraphical variations in the nature of this sandstone type, reflecting local differences in the nature of the source lithology, but it is difficult to specify the precise sources. Aeolian sandstones in formations below and above the Chester Pebble Beds were dominantly transported from the east (Thompson, 1985). The differences in mineralogy compared with the fluvial sandstones indicate that the aeolian sediment was not reworked from southerly-derived fluvial material during arid periods, but was introduced from another region. Given the palaeowind direction, the most likely sediment source was the Pennine landmass. During the Triassic, the Pennine area was probably mostly blanketed by Late Carboniferous elastics, but the scarcity of modern heavy-mineral data on Late Carboniferous sandstones from the region prevents a more detailed assessment of their suitability as source material. The only comparable data presently available are from Namurian sandstones of the Lancaster area. These have broadly similar heavy-mineral assemblages but differ in having generally higher detrital monazite abundances (Hallsworth, 1991), and cannot therefore have been involved to a significant degree. It is possible that younger Carboniferous sandstones lacked this component, and it is equally possible that there are regional variations in detrital monazite distribution. This question can only be resolved by undertaking studies of potential source lithologies on the Pennine landmass.

The involvement of Lower Palaeozoic sediments of the Welsh Massif, or Lower Palaeozoic from the Midlands, cannot be entirely ruled out. However, the Lower Palaeozoic sediments of the Welsh landmass to the west and south-west are not likely to have been a potential source, given the prevailing easterly winds, unless sediment was introduced by eastward-flowing rivers and then redistributed by aeolian activity. In many respects, the heavy-mineral assemblages in the Lower Palaeozoic are similar to those of the low-monazite Cheshire Basin sandstones: the assemblages are generally low in diversity, lack detrital monazite and chrome spinet, and have a mean ATi value of 59, similar to the mean Cheshire Basin value (Morton et al., 1992). However, RZi values are distinctly lower, and it is considered on this basis that the Lower Palaeozoic rocks of the Welsh Massif are unlikely to have been significant contributors.

The presence of appreciable amounts of detrital monazite in some samples from the Wilmslow and Helsby sandstones indicates that sediment supply from the south continued throughout the deposition of the later formations of the SSG (after the Chester Pebble Beds) but that its importance in the Cheshire Basin was markedly reduced. The evidence from the Helsby Sandstone of the Thornton Borehole, in which aeolian sandstones have very low MZi values and fluvial sandstones have significantly higher detrital monazite abundances, provides good evidence that the northward-flowing fluvial system was still supplying monazite-rich detritus. Thus the fluvial input continued to have a similar character to that of the Chester Pebble Beds, whereas the aeolian sediments appear to have a strong affinity with those of the earlier Collyhurst and Kinnerton sandstones.

There are stratigraphical and regional variations in the nature of the low-monazite assemblages: for example, the Wilmslow and Helsby sandstones have higher ATi and RZi values and lower rutile/anatase ratios than the Kinnerton and associated sandstones. Similarly, there is evidence for fluctuations in garnet abundance, as in the Stanlow Borehole. These variations are considered to relate to minor variations in the source-rock lithology, probably resulting from a combination of unroofing during erosion and geographical variations. Detailed study of potential source materials may throw further light on the cause of these variations and may help to provide further constraints on patterns of sediment supply.

The Wilmslow Sandstone Formation is characterised by two distinct heavy-mineral assemblages, one being apatite-rich and the other apatite-poor. The apatite-poor sandstones overlie the apatite-rich material in all cases where their strati-graphical relationships can be determined. There is little evidence that this difference resulted from changes in sediment provenance, and the most reasonable explanation for the variation is that it resulted from apatite dissolution during the interval represented by the Hardegsen disconformiry, at the boundary between the Wilmslow Sandstone and the Helsby Sandstone formations. Similar apatite depletion has been observed in the German Buntsandstein, deposited at a similar time and in similar facies. These events are widespread in the Buntsandstein, and it is possible that one of them may correlate with the zone of apatite depletion in the Cheshire Basin. Ideally, the variation in apatite abundance should be examined by reference to a cored sequence unequivocally passing through the boundary between the Wilmslow Sandstone and Helsby Sandstone, but no such material was available for study.

Garnet dissolution is at an advanced stage in most of the samples studied. Comparison with previous studies suggests that pore-fluid palaeotemperatures reached at least 80°C during burial. By analogy with the North Sea, burial depths may have reached 3 km, but it is possible that dissolution took place at shallower levels, either through higher heat flow or through expulsion of hot fluids from deeper levels. The Little Hay Borehole, outside the confines of the Cheshire Basin, shows preservation of garnet and the less stable species staurolite, suggesting that the fluid-flow and thermal history of the Cheshire Basin may differ from the area to the south and south-east.

Isotopic studies

Sm-Nd isotopic studies generally support the interplay of more distant southerly sources and more local inputs, as suggested by the heavy-mineral results. The data are consistent with a mixture of Cornubian or Armorican material and detritus from the Midlands, Welsh Borders or Wales. Many of the whole-rock samples have signatures characteristic of recycled material derived from Gondwana, like much of the Lower Palaeozoic sedimentary succession of Britain and, therefore, not very diagnostic of source.

Whole-rock geochemistry

The whole-rock geochemistry of the SSG indicates that the predominantly aeolian sediments (Kinnerton Sandstone) deposited prior to the first and most significant fluvial influx into the basin (Chester Pebble Beds) differed in composition from later sediments. The whole-rock geochemistry suggests that the formations above the Chester Pebble Beds contained a certain amount of fluvially transported sediment derived from the south, and elements of this effect were observed in the heavy-mineral investigations.

Comparison of mainly aeolian facies above and below the Chester Pebble Beds (the Kinnerton and Wilmslow sandstones) is consistent with the addition of granitic material, presumably from a Cornubian or Amorican source, in the later formation. There are indications that this may decline in importance towards the north of the basin.

The middle part of the MMG (Byley and Wych mudstones) is geochemically distinctive, with a composition suggesting a greater input from a basic igneous source relative to acid igneous or sedimentary components. The nearest source of basic material was probably the Carboniferous volcanics and intrusives of the South Pennines.

Chapter 5 Diagenesis of the Permo-Triassic rocks

A E Milodowski, G E Strong, T J Shepherd, B Spiro, S J Kemp, E K Hyslop, D G Jones, M J Leng, H W Haslam, A D Bradley, R A Nicholson and G Warrington

Introduction

Apart from a few studies (principally those of Ixer and Vaughan (1982), Holmes et al. (1983), Burley (1984, 1987) and Naylor et al. (1989)), very little detailed modern information has been published on the diagenetic history of the Cheshire Basin.

As part of the present investigation, petrological studies were carried out to address the diagenesis of the Permo-Triassic rocks of the basin and its genetic relationship to the sandstone-hosted red-bed Cu-Pb-Ba mineralisation and hydrocarbons. The main objectives were:

1. to identify the principal effects of diagenesis on the Permo-Triassic rocks;
2. to establish the relative chronology of diagenetic mineralisation (diagenetic paragenesis) and other modifications (e.g. compaction, porosity generation) which are relevant to the mobility of ore fluids and hydrocarbons, and their relationships to the burial history of the Cheshire Basin;
3. to assess any relationships between diagenetic phenomena in rocks at the basin margins – which contain hydrocarbon shows and Cu-Ba-Pb metal deposits – and the pattern and history of basinal fluid movement and associated mineralisation; and
4. to identify the source and nature (thermal and chemical characteristics) of the fluids responsible for diagenesis and mineralisation. Implicit in this objective was the need to test various models proposed for the mineralisation, in particular the mixing of hydrocarbon and metalliferous fluids, as suggested by Warrington (1980).

To this end, a programme of analysis was undertaken, using detailed petrographic studies by optical microscopy, scanning electron microscopy, backscattered scanning electron microscopy and cathodoluminescence microscopy, allied to X-ray diffraction analysis of clay minerals, fluid-inclusion microthermometry and chemistry, whole-rock lithogeochemistry, stable-isotope geochemistry of diagenetic cements and mineralisation, and analysis of residual organic matter and oil seeps from in and around the basin.

General petrographic characteristics of the basin infill are given in Chapter 4.

Sampling and analytical methods

Mineralogy, petrography and whole-rock geochemistry

As outlined in Chapter 1, most samples were obtained from borehole material, or from underground mine exposures, with only a small number of outcrops being sampled. Sampling procedures are described in Chapter 1, sample locations are shown in (Figure 2), and sample details are summarised in (Table 1) and (Table 2). Due to the lack of available material, only a very limited study was made of pre-SSG strata.

In total, 286 samples were examined petrographically in thin section by optical petrographic microscope, cathodoluminescence (CL) and backscattered scanning electron microscopy (BSEM). BSEM analysis was carried out using a Cambridge Instruments Stereoscan 5250 scanning electron microscope (SEM) fitted with a KE-Developments Ltd solid-state (diode-type) backscattered electron detector. A 20 kV electron beam was used. The identification of minerals was aided by qualitative energy-dispersive X-ray microanalysis (EDXA) using a Link Systems 860 X-ray microanalyser fitted to the SEM. A limited number of mineralised samples were also analysed by electron microprobe analysis (EMPA) using a Cameca SX50 electron microprobe.

A total of 92 samples were analysed by X-ray diffraction analysis (XRD) to determine their clay mineralogy; 54 samples from the MMG; 35 samples from the SSG, and 3 samples from the Manchester Marls. XRD analyses were performed with a Philips PW1700 series X-ray diffractometer controlled by Philips APD1700 software running on a Digital MicroVax 2000 computer. The X-ray generator was fitted with a Co-target tube and was operated at 45 kV and 40 mA. Analyses were made on oriented mounts of nominal <2 μm fraction separated material, scanned over the range 1.5–32° 2θ at a speed of 0.8° 2θ/minute in the air-dried and glycol-solvated states and after heating at 375°C for 2 hours and 550°C for 2 hours. A detailed account of the methodology and the clay mineralogy is given by Kemp (1994).

A limited number of SSG samples were studied by fission-track registration (FTR) analysis to examine the distribution and mobilisation of uranium during diagenesis. The method employed was based on that of Kleeman and Lovering (1967). Lexan polycarbonate plastic was used as a detector and placed in contact with polished thin sections of SSG. These were then irradiated with thermal neutrons in a reactor to fission ²³⁵U. After irradiation, the Lexan plastic was etched for 5–10 minutes at 60°C with NaOH solution to reveal fission-track damage. Quantitative estimation of the uranium content of different host mineral phases was achieved by irradiation of uranium-doped standard glasses along with the sample batches and consequent track counting on an equal-area basis using an optical microscope. A neutron fluence of c.2 X 10¹⁵ thermal neutrons per cm² was chosen to yield discrete tracks with minimum overlap to ensure reasonable counting accuracy.

Whole-rock geochemical analyses were carried out on 580 samples, including more than 300 from the SSG and more than 200 from the MMG. The preparation and analytical procedures are outlined on p.107 and reported in more detail in Jones and Haslam (1994).

Stable-isotope studies

δ¹³C and δ¹⁸O stable-isotope determinations were made on 52 well-characterised whole-rock samples containing carbonate cements (dolomite, ferroan dolomite, ankerite and calcite), and on hand-separated fracture-filling carbonate minerals. Samples were selected after careful petrographic observation to identify the type, and estimate the amount, of carbonate present, and to establish the number of generations of mineralisation. In most cases, samples were chosen of single generations of single carbonate minerals. Preparation of samples for stable-isotope analysis followed McCrea (1950) as modified by Rosenbaum and Sheppard (1986). Small amounts of material were ground to a fine powder by hand in an agate pestle and mortar, and 10–100 mg samples (depending on the estimated carbonate content) were taken for ¹³C/^{1 2}C and ¹⁸O/¹⁶O analysis. Carbon dioxide was extracted from samples by reacting with excess 100% orthophosphoric acid in vacuo at a constant temperature of 25 ± 0.05°C. δ¹⁸O was calculated using the fractionation constants (α¹⁸ (CO₂-calcite) = 1.01025 (Friedman and O'Neil, 1977) and α¹⁸ (CO₂-dolomite) = 1.01109 (Sharma and Clayton, 1965). Isotopic analysis was performed at the NERC Isotope Geoscience Laboratory, Keyworth using a VG SIRA 10 triple collector mass spectrometer. Results are quoted on the conventional 8 scale in per mil (‰) deviation from the PDB standard. The overall analytical precision is ±0.1‰ for carbon and oxygen.

δ³⁴S stable isotope analyses were made on 15 hand-picked monominerallic sulphide concentrates (pyrite, chalcopyrite, galena) from diagenetic cements and vein mineralisation. Thirty-three δ³⁴S stable isotope determinations were also made on whole-rock samples containing sulphate cements (anhydrite, gypsum and baryte) and on pure separates of sulphate' minerals from vein mineralisation. Sulphides were analysed using the method of Robinson and Kusakabe (1975) and sulphates using that of Coleman and Moore (1978). Sulphide samples were mixed with cuprous oxide and heated at 1075–135°C for 45 minutes. Sulphate samples were also mixed with cuprous oxide, but with the addition of quartz powder, and heated at 1125–1135°C for 65 minutes. Both reactions yield SO₂ gas which was extracted and collected by condensation in a n-pentane/liquid nitrogen trap and analysed on a VG SIRA 10 Series 2 mass spectrometer after separation of CO₂. Results are quoted on the conventional del (δ) scale in per mil (‰) deviation from the CDT standard. The overall analytical precision is ±0.1‰.

Fluid-inclusion studies

Fluid-inclusion analyses were undertaken on selected materials, and included both conventional microthermometric analysis and geochemical analysis of single inclusions using the recently developed Laser Ablation MicroProbe Inductively Coupled Plasma Mass Spectrometer (LAMPICP-MS). Thermometric data were obtained for inclusions in baryte and calcite cements and fracture mineralisation in the SSG from Alderley Edge, Clive Mine, Bickerton and the Thornton Borehole. LAMP-ICP-MS analysis was carried out on halite samples (Table 20). Halite samples were prepared as 3 mm polished slices; rock and mineral samples as ultra-thin (30 μm) doubly polished wafers.

A new method for producing these ultra-thin sections, involving the use of low-expansion, thermally stable, cold-setting adhesives, was developed for the fluid-inclusion studies. This was necessary because the baryte and calcite contain very small, poorly developed, low-temperature inclusions. A calibration protocol for the analysis of single inclusions by LAMP-ICP-MS was developed using fluid-inclusion analogues, and the ablation response of aqueous inclusions in halite (Shepherd and Chenery, 1995). The fluid-inclusion investigations are described in detail by Shepherd (1994).

Organic geochemistry

Rock samples and surface oil and tar seeps were subjected to solvent leaching in Soxhlet apparatus to isolate the organic matter, and were subsequently analysed by gas chromatography. The leachates were separated by column chromatography using hexane to elute the straight-chain alkanes. Full details of the methods used are given in Bradley and Nicholson (1995).

Several of the oil and tar samples were analysed semi-quantitatively for their organic-bound trace-metal content by LAMP-ICP-MS. This has been shown to be a potentially promising method (Bradley et al., 1992; Hughes and Bradley, 1993; Bradley et al., 1994), and further work is in progress to develop the technique so that quantitative results can be obtained.

Sherwood Sandstone Group

General comments

Much of the SSG was accessible only at the margins of the Cheshire Basin, where samples were obtained from surface exposures, shallow mine-workings (e.g. Alderley Edge, Clive) and relatively shallow (<310 m) borehole cores. Deep-basin borehole material was limited to small amounts of core from the Wilkesley and Prees boreholes. Recognition of diagenetic processes and the construction of a paragenetic sequence can be particularly difficult and complicated for rocks deposited in a basin-margin setting. Such environments may experience a more complex history of fluid-rock interaction than the deeper parts of a basin. During burial and basin dewatering, considerable volumes of fluids may be expelled outwards towards the basin margin. In contrast, in the basin centre, only small pore-volumes of water may pass through a given volume of rock, unless another source of pore fluid is available. The marginal areas of a basin are also very susceptible to erosion and to invasion by meteoric groundwater during basin inversion. Consequently, basin-margin areas may experience extreme changes in the type of pore fluids and hence display a complex diagenetic history. Waters expelled laterally from the basin centre during burial will tend to be warm, saline and reducing. In contrast, meteoric groundwaters invading the basin margins during uplift will tend to be cool, of low ionic strength, wealdy acidic and oxidising.

The Cheshire Basin, like the adjacent East Irish Sea Basin (Bushell, 1986; Jackson et al., 1987; Stuart and Cowan, 1991), has had a long and complex history, with episodes of burial followed by inversion (see Chapters 3 and 6), and a full diagenetic record may not be preserved. Evaporite minerals such as halite and anhydrite, precipitated from saline brines, are readily dissolved by dilute meteoric ground-waters and may be only partially preserved or lost. Also, minerals precipitated from oxidising and/or low-ionicstrength porewaters (e.g. iron oxides) can be transformed, replaced, reduced and redissolved when these meteoric fluids are displaced by warm, reducing, basinal brines, or their chemistry is changed by increasing P-T conditions (e.g. illitisation of smectite). Where evaporites are involved, whether within, above or below a sequence, the situation can be very complex. Because these minerals are extremely soluble, multiple phases of precipitation, dissolution, recrystallisation and redistribution can be expected, and it is in this context that interpretation of the diagenesis of the Permo-Triassic rocks in the Cheshire Basin must be considered.

Since the diagenetic history of the sandstone and siltstone facies of the Tarporley Siltstone Formation is broadly similar to that of the SSG, it is included in the following discussion. The degree of diagenetic modification, however, varies significantly depending on the nature of the sandstone facies. The diagenetic regimes encountered can be classified, following Schmidt and McDonald (1979a), according to whether diagenetic reactions are controlled by near-surface environments prior to burial (eodiagenesis), by conditions prevailing during burial (mesodiagenesis) or by near-surface processes following uplift or basin inversion (telodiagenesis). In general, eodiagenetic and early mesodiagenetic processes in the SSG sandstones tended to effect a reduction of the primary intergranular pore space through compaction, early pressure solution, grain overgrowth and cementation. Later mesodiagenetic processes removed some of the earlier cements and dissolved unstable framework grains, resulting in the formation of secondary porosity.

The diagenetic history of the SSG in the Cheshire Basin can be further subdivided within a relative temporal framework of eight broad Diagenetic Episodes, referred to as DE1 to DE8, along similar lines to that described by Strong et al. (1994) for the Permo-Triassic rocks of west Cumbria. Each Diagenetic Episode consists of a sequence of events or mineralogical features. Not all events or features comprising a Diagenetic Episode may be present in all samples, nor indeed in all boreholes or locations. Nor, in some cases, can the relative paragenesis of all component events within a given Diagenetic Episode be determined unequivocally. However, each Diagenetic Episode has a number of general characteristics which identifies it, and which can be recognised (or inferred) in most of the samples and locations studied. This summary scheme provides a useful framework within which the diagenetic evolution of the SSG of the Cheshire Basin can be described, interpreted and related to other features such as regional mineralisation and structural and geological events. The main aspects of each Diagenetic Episode are discussed below and summarised in (Table 21).

DE1: Red-bed and pedogenic eodiagenesis

Early red-bed diagenetic and pedogenic modification can be recognised in nearly all sandstones from all SSG formations. The most obvious feature is the coating of many detrital grain surfaces by thin (1–2 μm) veneers of very fine red-brown to black iron oxide. (DE1a). The iron oxide appears to have precipitated directly on grain surfaces and is responsible for the reddish brown coloration of the rocks. In some rocks, however, the coatings are due to iron-oxide impregnation of infiltrated detrital clay cutans around grains. The iron oxide is apparent in most sandstones, but it is absent in green or grey-white lithologies and reduction spots, from which the iron-oxide coatings have been leached by reducing mesodiagenetic fluids. Whole-rock geochemistry of the SSG shows a clear difference in Fe content (total Fe as Fe₂O₃) between reddish brown oxidised facies and grey or grey-green reduced facies, with significantly higher levels in the oxidised rocks. There is, however, no significant difference in Cu, Pb, Zn, Cr or MnO concentrations (Figure 91). The iron-oxide staining is closely associated with oxidative alteration and hematite replacement of detrital mafic minerals (principally biotite, chlorite, magnetite and ilmenite) and volcanic lithic fragments. Fine-grained Ti oxides (probably largely anatase) often accompany iron-oxide alteration products. Many biotite (and some muscovite) flakes have developed fine-grained hematite and anatase along partly exfoliated cleavage planes (Plate 14). These authigenic oxides are volumetrically small but provide a weak binding between detrital grains. They are characteristic of eodiagenesis in present-day sediments in hot desert or semi-arid environments (e.g. Walker et al., 1978). Similar red-bed eodiagenetic features have been described previously from the Cheshire Basin (Burley, 1984, 1987; Thompson, 1985) and are common characteristics of both aeolian and fluvial sediments of the SSG in other UK Permo-Triassic basins (Ixer et al., 1979; Ali and Turner, 1982; Burley, 1984, 1987; Bushell, 1986; Milodowski et al., 1986, 1987; Strong and Milodowski, 1987; Strong, 1993; Strong et al., 1994).

A common feature of most sandstones is the development of micronodular to pisolitic non-ferroan dolomite cements (DE1b). In most sandstones the dolomite typically comprises microgranular aggregates of interlocking xenotopic microspar, commonly with later euhedral overgrowths developed into intergranular areas. Radial-spherulitic crystal fabrics are also found but are less abundant. The dolomite was probably precipitated originally as micritic dolomite microconcretions but in most cases (especially in the north of the basin) the original fabric has been lost through recrystallisation. Original micritic fabrics are still evident in some of the boreholes, and are particularly well preserved in the Mickle Trafford, Thornton and Perry Farm boreholes (Plate 15). Dolomite micronodules (even when recrystallised) and dolomite-cemented patches within the sandstones (where adjacent nodules have coalesced to form a more continuous intergranular cement) usually display locally expansive, uncompacted grain fabrics indicating a near-surface origin. They also contain dispersed iron-oxide staining, commonly as complex concentric banding (spheroids) and radial-fibrous fabrics. Similar carbonate spheroids have been described by Strong and Pearce (1995) from Permo-Triassic sandstones of Cumbria. Thus, it is evident that DE1b carbonate authigenesis is coeval with the development of red-bed DE1a iron oxide ('desert varnish'). Early iron oxides and dolomite nodules predate diagenetic quartz, K-feldspar, anhydrite and calcite cements. These cements often occlude dolomite and iron oxides, and early dolomite cements locally inhibit the development of quartz and feldspar overgrowths.

While eodiagenetic DE 1b dolomite occurs widely across the Cheshire Basin in all SSG formations, it appears to be most extensively developed in the Chester Pebble Beds and Helsby Sandstone formations in the north of the basin. It may originally have been much more abundant but has been extensively corroded and removed by late telodiagenetic processes. In sandstones at outcrop, and in the upper levels of many boreholes, delicate skeletal radial-concentric structures preserved in hematite, and spherical mouldic cavities in matrix clays and iron oxyhydroxide (with euhedral rhombic cavities around their margins e.g. (Plate 16) corresponding to former dolomite overgrowths), are often all that testify to the original presence of eodiagenetic dolomite cements, which have since been dissolved by recent groundwaters. Nodular non-ferroan calcite cement with similar paragenetic and morphological characteristics is also found in the SSG. However, in contrast to the dolomite, nodular calcite has a very restricted stratigraphical and geographical distribution, occurring only in the the south-east corner of the Cheshire Basin, to the east of the Hodnet and Wem faults (in the Kinnerton Sandstone Formation of the Hodnet Station and Childs Ercall boreholes and the Chester Pebble Beds of the Bearstone Mill Pumping Station Borehole) and in the western margins of the basin (in the Helsby Sandstone Formation of the Gallantry Bank boreholes). There is some evidence that eodiagenetic calcite is also present in the Chester Pebble Beds of the Newport UDC Borehole, but it is difficult to differentiate it unambiguously from abundant late mesodiagenetic calcite. Eodiagenetic calcite and dolomite are mutually exclusive within any given sandstone horizon, although in the Hodnet Station Borehole, dolomite-bearing and calcite-bearing sandstones interdigitate.

The dolomite and calcite cements are similar to eodiagenetic cements described from the SSG in other Permo-Triassic basins (Burley, 1984, 1987; Milodowski, et al., 1986, 1987; Strong and Milodowski, 1987; Strong et al., 1994) and are interpreted as pedogenic or caliche-type dolocretes and calcretes. Stable-isotope analyses of nodular dolomite show compositions lying between c.0 and –5‰ δ¹³C (most <–2‰) and between c.0 and –5‰ δ¹⁸O (Table 22). Most analysed samples had either recrystallised during DE3 mesodiagenesis or contained inseparable DE3 mesodiagenetic carbonate mineralisation (dolomite overgrowths etc.), and the most reliable analyses for eodiagenetic dolomites indicate compositions of 0 to –1‰ δ¹³C and c.0 to –4.5‰ δ¹⁸O (Field I,(Figure 92). This is similar to eodiagenetic dolomite from the East Yorkshire-Lincolnshire Basin (Bath et al., 1987) and Cumbria (Burley, 1984). The high δ¹³C of the dolomite indicates that there was no significant influence from soil organic-derived CO₂ and that groundwater-dissolved carbonate had a provenance probably inherited from detrital carbonate of marine origin, with the possibility of exchange only with atmospheric CO₂ (δ¹³C = c.−7‰). Petrographic evidence identified the presence of detrital silicified marine limestone (?Carboniferous Limestone), and it is plausible that HCO₃⁻ derived from the dissolution of detrital marine (?Carboniferous) limestone (δ¹³C −0‰) dominated the carbon isotopic signature of the eodiagenetic SSG porewaters. The δ¹⁸O values are heavier than for calcrete calcites from the SSG of the Wessex Basin, although similar to dolocrete from East Yorkshire and Lincolnshire (Bath et al., 1987). To some extent this difference may be due to the variation in oxygen isotope fractionations between calcite or dolomite and solution (c.2‰; Fritz and Smith, 1970). Palaeogeographical effects or a greater degree of evaporative enrichment of δ¹⁸OH₂O in near-surface water recharging the SSG in the Cheshire Basin and the East Yorkshire- Lincolnshire Basin may also account for the difference (cf. Bath et al., 1987). Work by Salomons et al. (1978) shows that with increasing evaporation soil carbonates become increasingly isotopically heavy. Surface conditions in the Cheshire Basin during SSG times may have been similar to those observed in modern arid continental environments: dolomite precipitated in these environments has typically high δ¹³C and δ¹⁸O values (McKenzie, 1981). In this respect, the isotopic characteristics of the nodular eodiagenetic dolomite in the SSG are similar to those of the primary dolomicrite in the overlying MMG (Table 22). MMG dolomicrites are closely associated with, and interlaminated with, primary anhydrite (p.153), in sequences thought to represent evaporative lacustrine or sabkha deposition (Wilson, 1990, 1993).

The restriction of eodiagenetic calcite to the south-east margin of the Cheshire Basin, with dolomite eodiagenesis in the west and north-west, resembles the eodiagenetic lithofacies zonation observed in the Wessex Basin. Here, the proximal or basin-marginal areas are characterised by calcrete development which is superseded laterally, towards the basin depocentre, by dolocrete development, at the periphery of a distal inland sabkha or playa environment, in turn characterised by gyperete and/or anhydrite precipitation (Strong and Milodowski, 1987; Holloway et al., 1989). Similar lithofacies zonation in recent Australian arid palaeodrainage systems has been described by Arakel and McConchie (1982), who consider that dolomitisation occurs in zones where groundwater levels are subject to fluctuation. In the Cheshire Basin (at least with respect to the Kinnerton Sandstone and Chester Pebble Beds formations), the diagenetic zonation parallels the drainage direction of the Lower Triassic fluvial system. Calcrete-dominated wetter (or higher water-table) facies in the south-east pass rapidly, in a north-westerly direction, across the basin margin growth faults (Wem–Bridgemere–Red Rock Fault System), into a more distal evaporitic or arid diagenetic facies. Similarly, a basin-margin calcrete zone further west, in the Bickerton area, is rapidly superseded by dolocrete to the north and east.

DE2: Early silicate and smectite (±early pyrite) diagenesis

Minor, weakly developed, tiny overgrowths of authigenic quartz, and to a lesser extent adularia-like K-feldspar (DE2) are sometimes developed in close association with DE1b grain coatings of hematitic iron oxide in many of the sandstones. They probably developed early, before significant burial compaction, since they are seen impacted into less competent grains of mudstone and siltstone in rocks which have suffered burial compaction. DE3 mesodiagenetic dolomite cement encloses, and therefore postdates, DE2 quartz and feldspar. DE2 quartz and feldspar authigenesis may be widespread in development but is indistinguishable from later authigenic quartz and feldspar in many of the sandstones since complete cement paragenetic fabrics in the rocks are seldom preserved, as a result of late telodiagenetic dissolution of carbonate (and ?evaporite) cements.

Sandstones in the north-west of the basin sometimes display evidence of early diagenetic clay-mineral authigenesis. Typically, the authigenic clay forms delicate pore-lining 'box-work' fabrics. It is largely restricted to sandstones with well-developed infiltrated clay cutans from which it appears to develop. Authigenic clay is also seen replacing expanded and exfoliated altered detrital biotites and muscovites (Plate 17). The clay predates major compaction: clay replacements of micas are distorted, and pore linings are commonly spatted from grain surfaces by compactional deformation (Plate 17). DE2 clay coatings are now preserved largely as K-Al-rich illite or mixed-layer illite-smectite. However, in the Helsby Sandstone Formation (and in sandstones of the overlying Tarporley Siltstone Formation) in the Saughall Massie Borehole, SEM-EDXA indicates that these coatings are Mg-rich saponitic, smectitic or corrensitic clay. These clay coatings probably account for much of the smectite identified by XRD in the clay fraction separated from bulk sandstone. The boxwork or cauliflower morphology of the authigenic clay is typical of smectite, and although the clay is now preserved as illite, mixed-layer illite-smectite or corrensite, it seems likely that it was originally smectite.

Precise relationships with other diagenetic phases are difficult to ascertain. DE2 clays tend to be associated with eodiagenetic clay cutans which inhibit the formation of DE4 quartz and feldspar overgrowths. Consequently, the paragenetic relationship between DE2 clays and DE4 quartz and feldspar are seldom seen, although they are quite clearly superseded by DE6 carbonate and sulphide mineralisation. However, these clays are not observed coating DE4 precipitates, and may rarely be seen trapped beneath DE3 carbonate cements which suggests that the clay authigenesis predates these cements.

Framboidal pyrite was seen very rarely in the SSG and Tarporley Siltstone Formation. It occurs in trace amounts in the Helsby Sandstone Formation of the Hondslough Farm Borehole but is more abundantly developed in the muddier sediments of the Tarporley Siltstone Formation, as in the Saughall Massie and Marston Salt Union boreholes. Its precise paragenetic relationships are not clear. In the Saughall Massie Borehole framboidal pyrite is clearly affected by compaction and predates the main DE4 quartz cements, which locally occlude it. It only occurs in locally reduced green siltstone horizons which have been leached of early DE1b iron oxide grain coatings. It also replaces exfoliated altered detrital biotite grains. Framboidal pyrite is typically associated with bacterial sulphate reduction during diagenesis (Hudson, 1982). It would therefore appear to have formed during early diagenesis but postdating DE1b. However, the sediments must have been buried sufficiently for the porewaters to attain sufficiently reducing conditions within these organic-poor strata to stabilise ferrous sulphide precipitation.

DE3: Early carbonate mesodiagenesis

DE3 is recognised in all formations in all boreholes hosting eodiagenetic dolomicrite nodules. It is characterised by the development of idiotopic rhombic overgrowths of nonferroan dolomite on cores of DE1b nodular dolomite (Plate 18) and (Plate 19). In many samples, the eodiagenetic dolomite has recrystallised to microspar aggregates or single crystals in optical continuity with the DE3 overgrowths. The dolomite overgrowths are often strongly zoned and become increasingly ferroan in later growth stages. Similar features are reported by Burley (1984) and Strong et al. (1994) from the margins of the East Irish Sea Basin. DE3 dolomite and DE3recrystallised early dolomite nodules are the dominant carbonate cement in much of the SSG around the northern rim of the Cheshire Basin. In many samples, particularly in the upper levels of the boreholes, this dolomite is inferred to have been present previously but removed during late telodiagenesis. This inference is based (as for DE2b) on the observation of rhombic and euhedral mouldic cavities (characteristic of dolomite morphology), preserved beneath an encrustation of late telodiagenetic secondary iron and manganese oxyhydroxides, or within patchy mesodiagenetic calcite cements e.g. (Plate 16).

DE3 dolomite is a relatively early mesodiagenetic cement. It predates the main development of authigenic (DE4) quartz and feldspar overgrowths and cements, since detrital grains enclosed in dolomite rarely show evidence of quartz and feldspar authigenesis. Dolomite also predates and is enclosed within major calcite (DE6) and anhydrite (DE5 and DE6) cements. DE3 fabrics appear to have developed during the early stages of burial compaction, cementing detrital grains which show slight grain compaction. However, the DE3 dolomite itself is sometimes seen to have suffered some compactional deformation (e.g. stress cracks filled by later calcite, anhydrite and pyrite cements) and less-competent grains such as detrital micas and mudclasts may be compactionally deformed around adjacent DE3 dolomite crystals e.g. (Plate 18). The compositional characteristics of DE3 dolomite can be interpreted to represent precipitation during early mesodiagenesis (early burial) from initially oxidising shallow groundwaters (i.e. precipitation of non-ferroan dolomite) which progressively became more reducing with increasing burial (i.e. resulting in the precipitation of ferroan dolomite as the porewaters became enriched in reduced Fe²⁺, and to a lesser extent Mn²⁺ species).

Although pure DE3 dolomite overgrowths could not be separated for stable-isotope analysis, composite mineral concentrates of DE3-recrystallised dolomites with DE3 overgrowths were analysed (Table 22). DE3 dolomites are isotopically lighter than eodiagenetic dolomite (Field II, (Figure 92) with δ¹³C between +1 and –4‰ and δ¹⁸O values between –6.5 and –4.0‰. The lighter isotopic composition is consistent with precipitation from a warmer pore fluid as burial progressed. In this respect DE3 dolomite is similar to early mesodiagenetic non-ferroan dolomite described by Burley (1984) from the SSG in Lancashire. Both Burley's results and those of the present study show that DE3 dolomites have isotopic compositions between those of eodiagenetic DE1 b dolomites and deep burial (DE6) carbonates.

DE4: mesodiagenetic quartz–feldspar authigenesis

Many of the sandstones show the development of a major episode of syntaxial overgrowth and cementation by authigenic quartz, K-feldspar and to a lesser extent albite. The paragenetic relationships between authigenic quartz, K-feldspar and albite overgrowths are often complex and impossible to resolve unambiguously. Burley (1987, p.430) identified up to 12 cathodoluminescence colour zones in the quartz cements of the SSG of the Kirkham Borehole, near Preston. It seems likely that all three minerals were closely syngenetic, although more than one generation of mineral overgrowth may be represented within the overall DE4 framework. Authigenic quartz is generally more significant than authigenic feldspar, and it almost fills the intergranular porosity in some sandstones (Plate 20).

Detrital grain boundaries within clean quartz-cemented sandstones generally show only simple point, long and rare sutured grain contacts, indicating only a moderate degree of compaction prior to burial. The incidence of more complex grain boundaries (triple point and sutured contacts and pressure welding) is rare except in poorly sorted sandstones and litharenites with a very high proportion of lithic fragments, especially chert-rich sandstones. Some quartz overgrowths may be compacted into, and deform, less competent mudstone clasts, or they may display the development of compactional stress cracks which have been mineralised by later DE5 and DE6 cements e.g. (Plate 21). These petrographic relationships suggest that the SSG was actively undergoing the early stages of burial during DE4 quartz–feldspar mesodiagenesis but had not yet achieved maximum burial compaction.

The degree of quartz (± feldspar) development is very variable – even within the same formation and in the same borehole sequence. For example, it is well developed in the Helsby Sandstone Formation but absent in the Wilmslow Sandstone Formation in the Marston Salt Union. Borehole. Generally, however, authigenic DE4 quartz (± K-feldspar) tends to form only minor or patchy cements in most of the SSG from around the north-western and western margins of the basin. It is particularly well developed in certain sandstone beds in the Helsby Sandstone Formation (Prees, Hondslough Farm and Marston Salt Union boreholes), Wilmslow Sandstone Formation (Clotton Borehole) and Chester Pebble Beds (ICI Widnes and Clotton boreholes). In contrast, except for the Bearstone Mill Borehole (Chester Pebble Beds) where it forms a minor cement, DE4 is absent, or only rarely present, around the southern margin of the basin or to the east of the Wem–Bridgemere–Red Rock Fault System.

Overgrowths tend to be inhibited in muddier sandstones with well-developed eodiagenetic grain-coating clay cutans, and by the presence of extensive DE1b and DE3 dolomite cementation. However, even in the north, a number of cleaner sandstone beds lack DE4 quartz–feldspar development. Despite this, the sandstones preserve relatively open and uncompacted fabrics, suggesting that the porosity is secondary and must originally have been filled by a cement (which prevented or restricted quartz–feldspar authigenesis) prior to any major burial compaction. Since both late post-quartz DE6 calcite, and DE1 and DE2 dolomite are preserved in many of these rocks, it is evident that the removed cement was not a carbonate, but may have been an evaporite mineral (e.g. halite, gypsum or anhydrite). This could imply a locally very complex (?diachronous) relationship between DE4 and DE5, and that early DE5 cementation may locally have interrupted or inhibited DE4 development. The paragenetic relationships between these two Diagenetic Episodes are not always unequivocal.

Minor fine authigenic anatase is often seen included in DE4 quartz cements. The anatase is associated with altered detrital Ti-Fe oxides. Minor anatase is also sometimes present in late DE3 dolomite cements, and its paragenesis may span DE3 to DE4. Trace amounts of fine-grained authigenic pyrite are rarely present, as inclusions trapped beneath DE4 quartz overgrowths, in some sandstones. However, significant sulphide diagenesis and mineralisation is only associated with DE6.

DE5: mesodiagenetic evaporite cementation

Much of the intergranular porosity in the SSG is secondary (rejuvenated) porosity, resulting from the dissolution of an earlier diagenetic pore-filling cement. This is indicated by the lack of compaction in many of the porous sandstones, and the presence of oversized pores (cf. Schmidt and McDonald, 1979b). Burley (1984, 1987) suggested that much of the secondary porosity in the SSG resulted from late mesodiagenetic carbonate cement dissolution. However, although extensive carbonate dissolution (mainly of DE1b and DE3 dolomites) is evident in the SSG around the basin margins, it is related largely to telodiagenesis (DE8). Important in this respect is the lack of carbonate-cement dissolution in porous sandstones from deep boreholes (Prees and Wilkesley) near the basin depocentre. Strong (1993) suggested that anhydrite, gypsum and possibly halite were precursor cements that were later removed to yield secondary porosity in SSG rocks of the Preston area of Lancashire.

Significant characteristics of these rocks are: (1) the lack of penetration of vein calcite (and other late diagenetic vein minerals, such as ferroan dolomite, ankerite, baryte, galena and quartz) into friable, uncemented and very porous sandstone wall-rock; and (2) the ability of these often very soft and friable sandstones, with little quartz overgrowth or cementation, to sustain very brittle fracturing. Both of these features imply that the sandstones were well lithified and cemented by an earlier cement ('DE5') prior to fracturing and consequent mineralisation. Petrographic relationships indicate that DE5 must have postdated DE4 quartz and feldspar authigenesis, although in some rocks it appears to have arrested or prevented DE4 development. This cement also limited sandstone compaction to some extent. Since calcite, ferroan dolomite and ankerite filling these fractures often show little evidence of dissolution in contact with open pores, it must be concluded that the pre-existing cement could not have been a carbonate mineral. This is observed in many localities (e.g. in West Mine, Alderley Edge and boreholes at Childs Ercall, Bewsey, ICI Widnes, Perry Farm, Saughall Massie and Wilkesley), indicating a widespread distribution of the former DE5 cement throughout the Cheshire Basin.

Anhydrite is a very likely candidate cement: poikilotopic anhydrite is observed as a major cement in the Malpas Sandstone of the Wilkesley Borehole (and other sandstones of the Tarporley Siltstone Formation elsewhere in the basin). In most cases, petrographic relationships show that the anhydrite encloses (and therefore postdates) DE4 quartz and feldspar authigenesis. However, in the Wilkesley Borehole, the relationships are not straightforward. Poikilotopic anhydrite can inhibit (therefore predate) or enclose (therefore postdate) DE4 quartz overgrowths. Furthermore, it also encloses DE6 ferroan dolomite and ankerite in late fault-related veins. Anhydrite cements both uncompacted and compacted sandstone fabrics. Multiple generations of anhydrite cementation and remobilisation of anhydrite persist from the early stages of burial through to deep burial mesodiagenesis. DE6-associated anhydrite veining in the Wilkesley Borehole also cuts porous open-textured sandstones, again with no evidence of penetration into the adjacent porous wall-rock. This implies that a different evaporite mineral, probably halite, must have been present in addition to anhydrite in some sandstones, but has subsequently been leached by circulating groundwater. Halite and anhydrite are both seen cementing siltstones and fine sandstones in parts of the overlying MMG, but halite was not directly observed in the SSG.

Anhydrite cements are a common feature of the SSG in many UK Permo-Triassic basins. It is observed as a major cement in the Wessex Basin, the East Yorkshire–Lincolnshire Basin and the East Irish Sea Basin (Milodowski et al., 1986, 1987; Strong and Milodowski, 1987; Strong et al., 1994). However, it tends only to be preserved in the deeper parts of the basins. Its removal (rather than carbonate dissolution) during late mesodiagenesis and/or during telodiagenesis appears to account for much of the secondary porosity in the SSG in these basins. This study indicates that anhydrite (and probably halite) cement (and their subsequent dissolution) have been similarly important in the Cheshire Basin.

DE6: Major mesodiagenetic silica-carbonate-sulphidesulphate cementation and fracture mineralisation

A major complex episode (DE6) of carbonate, sulphide and sulphate mineralisation, often interspersed with discrete mineral dissolution events followed DE5. DE6 is very closely associated with faulting and fault-related fractures. In many of the sandstones around the basin margin it is only poorly preserved because of late telodiagenetic removal of DE6 carbonate and sulphide cements, and can only be inferred from relict fabric information. The effects of this episode of diagenesis are widespread across the Cheshire Basin, and petrographic relationships show that DE6 is closely related to tectonic fracturing and syngenetic with major regional carbonate-baryte (± sulphide) fracture-hosted mineralisation. DE6 can be subdivided into a number of different stages of mineralisation and authigenesis. The relationships between some of these subdivisions are not always clear, and fine details of the mineral paragenesis vary from site to site – particularly with respect to the main areas of polymetallic mineralisation such as Alderley Edge, Clive, Bickerton – and may be episodic. However, all DE6 features are mesodiagenetic.

DE6A

The early stages of this episode are evident in numerous small faults in the SSG. They typically manifest themselves as 1–5 cm-wide zones consisting of a closely spaced anastomosing network of incipient or diffuse fine white veins. Similar features can be seen offshore in the underlying Collyhurst Sandstone (e.g. Irish Sea Well 110/11-2). Sometimes white bleaching or green reduction of the red-brown sandstone wall-rock can be seen locally for up to 1–2 cm either side of the vein, indicating the passage of reducing fluids along these structures. In most cases, the faults show normal displacements. These veins represent fine fault gouges or microbreccias comprising highly sheared and microgranulated wall-rock material cemented by fine-grained chalcedonic silica (Plate 22). They are particularly well developed in the Clive and Alderley Edge mines, where the major fault planes are heavily silicified and more crystalline xenomorphic heterogranular granoblastic quartz cements are also developed in association with the chalcedonic silicified fault rocks. These features clearly indicate brittle fracturing of the sandstones. Therefore, although many of the rocks are now soft, friable and uncemented, they must have been very well cemented and lithified at the time of DE6a fracturing. This cement — inferred to be of DE5 — must have been removed subsequently to DE6.

DE6a silica mineralisation predates DE6 carbonate mineralisation, since there is no evidence of inclusion of fragmented DE6 carbonate mineralisation within the gouge material, and it is cross-cut by DE6e calcite and baryte veining (Plate 23). The relationship between DE6a silica and sulphide mineralisation is complex. With the exception of the Hondslough Farm, Mickle Trafford and Thornton boreholes (which carry some traces of pyrite inclusions within DE6a silica cement or quartz overgrowths), sulphide is not generally associated with DE6a in boreholes around the northern or southern margins of the basin. However, in the Alderley Edge (West Mine) and Clive areas these early silicified fault gouges carry very complex assemblages of minor baryte, sulphide, selenide and arsenic minerals inter-grown with quartz and chalcedonic cements.

At Clive, quartz cements in the fault rocks may locally develop stretched cross-fibre crystal habit which is characteristic of the multi-stage crack-seal mechanism of fracture mineralisation, rather than simple dilational fracturing and infill (Ramsey, 1980; Barker, 1990). This process is considered by some authors to provide strong evidence for hydrofracturing, possibly related to overpressuring (Shearman et al., 1972; Ramsey, 1980; Stoneley, 1983; Barker, 1990). It points to a possible genetic link between faulting, DE6 mineralisation, overpressuring and hydrofracturing. Within the silicified fault rock, euhedral overgrowths on entrained detrital quartz grains have trapped tiny (<5 um) inclusions of baryte, pyrite, galena, covellite, Cu₂S (?chalcocite or digenite), chalcopyrite, molybdenite, and rare Ag₂S (?argentite or acanthite) associated with Cu₂S. It is clear that the fault rock has been reactivated several times, and later-stage chalcedonic silica cements introduced: at least two generations appear to be present. The chalcedony contains abundant tiny inclusions of baryte, Ag₂Se (?naumannite), Cu-Ag-selenide (?eucairite; CuAgSe), Cu₂S (?chalcocite or digenite), chalcopyrite, pyrite, galena and cinnabar. The precise mineralogical identity of some of these tiny inclusions is uncertain since they could only be observed under BSEM or by EMPA, with identification based on elemental ratios determined by EDXA. Ag₂Se (?naumannite) is the most abundant inclusion in the chalcedonic silica. It characteristically occurs in globular or framboidal aggregates enclosed in microcolloform silica (Plate 24). In places the silica has been replaced (following telodiagenetic alteration) by blue amorphous gel-like hydrous copper silicate (?chrysocolla). Framboidal sulphosalt morphology (e.g. pyrite) is typically associated with bacterial sulphate reduction during diagenesis (Hudson, 1982). Microbial mediation is not necessarily a prerequisite, as framboidal sulphide formation also occurs inorganically (Rickard, 1970, 1975; Sweeny and Kaplan, 1973; Greer, 1978). However, it does imply that mineralisation occurred at relatively low temperatures. There is a discrete zoning of sulphosalt inclusion minerals within the quartz overgrowths and chalcedonic cements, which indicates the following complex paragenetic sequence within DE6a at Clive:

1. pyrite + Cu₂S + Ag₂S + baryte
2. quartz
3. galena + Cu₂S
4. quartz
5. baryte + Ag₂Se+ Cu₂S + chalcopyrite + pyrite + galena + cinnabar + molybdenite
6. chalcedony
7. Cu-Ag-selenide + ?chalcopyrite +?pyrite
8. chalcedony

Similar inclusion fabrics are seen in quartz overgrowths at Alderley Edge (Plate 25) and (Plate 26), although the chalcedonic cement in silicified fault rock is free of sulphides. There are, however, differences in DE6a inclusion mineralogy between Clive and Alderley. DE6a inclusions at Alderley Edge comprise largely octahedral CoAsS (probably cobaltite), complex octahedral Co-Ni-As-sulphides (possibly As-substituted siegenite or cobaltite-gersdorffite solid-solutions), rare NiAs₂ (probably corresponding to pararammelsbergite) similar to inclusions identified by Ixer and Vaughan (1982), and galena, with minor earlier sphalerite (e.g. (Plate 27)). Inclusion crystals range from <5 to 50 pm in size. The cobaltite and Co-Ni-As sulphides sometimes contain minor amounts of Zn and Fe. As in Clive, the inclusions seen in West Mine, Alderley Edge are arranged in mineralogically distinct zones within the growth bands of the authigenic quartz but have the following paragenetic sequence:

1. quartz + chalcedony
2. sphalerite
3. galena
4. cobaltite (CoAsS), Co-Ni-As-sulphide, pararammelsbergite (NiS₂)
5. galena
6. quartz

This paragenetic assemblage, in turn, differs significantly from that reported by her and Vaughan (1982) and Naylor et al. (1989) from the mineralisation at nearby Stormy Point, Engine Vein and Wood Mine at Alderley Edge, and from Bickerton in the west. Bravoite, pyrite and chalcopyrite, found by these authors as inclusions in early quartz overgrowths, were not seen in West Mine. Although material from Stormy Point, Engine Vein and Wood Mine were not studied as part of the present project, their DE6a quartz-hosted Ni-As-Cu-Fe sulpharsenide assemblage would appear to be geochemically distinct from the Ni, Co, As-free and Se, Ag, Cu, Fe, Pb, Hg, Ba-rich DE6a sulphide assemblage found at Clive Mine in the south of the basin, the Se, Cu, Ag, Hg, Fe, Ba-free, and Co, Ni, As, Pb-rich sulphide assemblage observed in West Mine, Alderley Edge, and the pyrite-rich assemblage which characterises the Hondslough Farm, Mickle Trafford and Thornton boreholes in the north of the basin. Comparison of these studies with the present investigation demonstrates that the geochemical characteristics of diagenetic fluids during the DE6a stage of metal mobilisation and mineralisation were quite variable both locally and across the Cheshire Basin and, furthermore, the paragenesis of red-bed mineralisation is more complex than has been reported by her and Vaughan (1982) and Naylor et al. (1989).

DE6a quartz and silica authigenesis in all localities is restricted to fractures and fault rocks, and it appears that DE6a silicate (± sulphosalt) -cements do not penetrate for more than 1–5 cm (in most cases «1 cm) into the adjacent (now porous) sandstone wall-rocks (implying that they were tightly cemented by an earlier diagenetic cement). The close relationship between faulting and quartz-silica mineralisation, overgrowths and quartz grains is very clear in West Mine and at Clive. In these localities, well-developed quartz overgrowths and silica cements are seen in sandstone in the fault rock and sandstone immediately adjacent to the faults e.g. (Plate 28), but even a few tens of centimetres away from faults the same sandstone shows no evidence of DE6a quartz overgrowths e.g. (Plate 29).

DE6B

Further fracturing followed DE6a and was accompanied by dolomite mineralisation and authigenesis (DE6b). Early precipitates were non-ferroan dolomite, but the dolomite progressively became ferroan then ankeritic in successive growth stages. DE6b cements of dolomite, ferroan dolomite and ankerite are normally seen as the earliest minerals in later DE6 mineralised veins (e.g. Hondslough Farm Borehole, Thornton Borehole, West Mine Alderley Edge, Wilkesley Borehole), or inferred to have been present from rhombic mouldic cavities in late DE6 veins (e.g. Clive Mine). DE6b relationships are best seen in the Thornton Borehole where DE6b dolomite occurs enclosed in DE6e ferroan calcite, both in veins and locally penetrating the adjacent wall-rock. In both situations the dolomite–ankerite is encrusted or corroded by DE6d pyrite which is also entrapped beneath DE6e calcite (Plate 30). Dolomite mineralising fluids appear to have locally corroded the DE5 pore-filling evaporite cement soon after fracturing, resulting in patchy ankerite–dolomite cementation of the sandstone matrix. In detail, DE6b displays numerous corrosion surfaces internally between crystal grown zones (Plate 30) and (Plate 31) which represent hiatuses in mineralisation, when the minerals were redissolved to some extent in response to changes in fluid geochemistry. In the Wilkesley Borehole, DE6b dolomite–ankerite vein and matrix cements are enclosed in later DE6c anhydrite.

Only the Wilkesley Borehole provided sufficient isolated DE6b dolomite–ankerite for stable-isotope geochemistry. However, additional data were also obtained from Marston Salt Union Borehole for early nodular dolomite which had been largely recrystallised and replaced by DE6b dolomite. Similar late fracture-mineralising dolomite enclosed in anhydrite occurs in the Byley Mudstone Formation of the Crewe Heat Flow borehole. δ¹⁸O and δ¹³C results are presented in (Table 22) and (Figure 92) (Fields III and IV). The early-stage DE6b non-ferroan to weakly ferroan vein and matrix dolomite has lighter δ¹⁸O (–5.5 to –7.5‰) than earlier DE1b and DE3 carbonates. This is consistent with deeper burial and precipitation from warmer mineralising fluids (Friedman and O'Neil, 1977; Irwin et al., 1977; Anderson and Arthur, 1983; Moore, 1989). In contrast, δ¹³C is similar to that of the earlier carbonates, which suggests that the carbon isotopic signature of the pore fluids during early DE6b was still largely influenced by re-working or dissolution of detrital carbonate (or early diagenetic carbonate) components within the SSG. As mineralisation and diagenesis progressed, the later, more ferroan dolomite-ankerite shifted to lighter δ¹⁸O and δ¹³C (Figure 92). Late diagenetic ankerite and ferroan dolomite with similar isotopic characteristics have been observed in the East Irish Sea Basin, where it also occurs as a pore-filling and fracture-mineralising cement (Burley, 1984, 1987). The observed trend from isotopically heavy early DE1b carbonate cements through progressively lighter DE3 to DE6b dolomiteankerite is consistent with progressive burial diagenesis cement sequences seen in the East Irish Sea (Burley, 1984, 1987).

DE6c

Within the SSG and the Tarporley Siltstone Formation, DE6c is only recognised in the Malpas Sandstone and Tarporley Siltstone Formation of the Wilkesley Borehole. It occurs as coarsely crystalline, platey-to-fibrous, white vein-filling anhydrite, cutting both anhydrite-cemented and porous sandstones. The vein anhydrite encloses earlier wall-lining euhedral DE6b dolomite and ankerite. In the porous sandstones the vein anhydrite fills only the fracture, showing little or no penetration into the adjacent porous wall-rock (Plate 32). This clearly indicates that an earlier cement must have been present in the host rock, preventing intergranular precipitation. Since neither vein-filling anhydrite, nor vein and matrix carbonate cements show any evidence of dissolution, the earlier (DE5) host-rock cement must have been another soluble mineral – most probably halite cement (especially considering the major development of bedded halite in the overlying MMG). Anhydrite cements are present in other sandstones, and it would appear that the pattern of cements is variable in the Wilkesley Borehole: with DE4 anhydrite in some sandstones; DE4 halite (since removed) in other sandstones; and later cross-cutting DE6c anhydrite veining. It is not possible to distinguish between possible DE6b and earlier DE5 in sandstone where both anhydrite cements may be present.

The extent of DE6c effects across the basin are now impossible to evaluate. Anhydrite is very soluble, and DE6c features are not preserved in sandstones around the basin margin because they have suffered extensive telodiagenetic leaching. Similar anhydrite veining is seen in the overlying Boffin Mudstone Formation (MMG) in Wilkesley, and is generally preserved more widely in the MMG (e.g. Meadowbank Mine, Winsford, Coton Fields, Crewe Heat Flow, British Salt BH4 Stafford boreholes). This suggests that DE6c was widespread in its influence, affecting the MMG in addition to the SSG.

DE6D

A major episode of sulphide mineralisation (DE6d) followed DE6b dolomite–ankerite mineralisation in many parts of the Cheshire Basin. Sulphides are not observed or preserved in all localities (especially at outcrop or at shallow depths) and therefore the former extent of DE6d is difficult to evaluate, although the distribution of occurrences indicates that it was probably widespread in effect, if not in scale. Copper sulphides and pyrite ± a complex assemblage of Pb-Zn-AsNi-Co-Ag-Hg-Bi-Mo-Se sulphides are found in the SSG to the south-east of the Mersey and in the Wirral Peninsula. However, to the north of the Mersey and in the Mickle Trafford Borehole, pyrite and sometimes sphalerite are the only sulphide minerals developed. Minor galena, sphalerite, and Ni-Co-As-Cd-sulphide mineralisation, associated with bituminous hydrocarbon residues containing concentrations of uraninite and uranium silicate, are found locally within the underlying Manchester Marls and appear to belong to the same episode of diagenesis. The hydrocarbon residues are similar to bitumens described elsewhere from around the margins of the East Irish Sea Basin (Eakin, 1989; Milodowski, et al., 1990; Parnell and Swainbank, 1990). DE6d generally resulted in only small disseminations of authigenic sulphide which, in common with other DE6 mineralisation, is usually closely associated with faulting and fracturing. However, in the Alderley Edge, Clive–Grinshill, Eardiston, Pim Hill, Yorton, Hawkstone–Wixhill and Bickerton areas, sulphide mineralisation was more extensively developed and is associated with major fault structures e.g.(Figure 93).

DE6c anhydrite veining (now preserved only in the deep basin) and DE6d sulphide mineralisation are nowhere observed together and their relationship remains unresolved, although both diagenetic features postdate DE6b. Relationships between DE6d mineralisation and other diagenetic fabrics are particularly clearly demonstrated in the Thornton Borehole. Here, fault-related veins have been mineralised by a complex sequence of cements; early DE6a chalcedonic silica is cross-cut by fractures mineralised initially by DE6b dolomite then ankerite, subsequently by DE6d pyrite and, finally, enclosed within and locally cross-cut by DE6e ferroan calcite (e.g. (Plate 30) and (Plate 33). Pyrite is found either encrusting, or corroding and replacing, earlier DE6b dolomiteankerite cements (Plate 30) and (Plate 31). Despite the sandstone wall-rocks being very open and porous, the pyrite mineralisation typically only penetrates a few millimetres into the wall-rock and abuts against open intergranular porosity (Plate 33). However, locally pyrite may extend a few tens of centimetres away from major mineralised fractures as patchy areas of cement. Again, it is clear from evidence for brittle fracturing of these now-friable rocks that the sandstones were originally cemented prior to fracturing and pyrite mineralisation. The presence of pristine earlier DE6b dolomite and later DE6e calcite cements and vein-fills clearly demonstrates that the secondary porosity results from the dissolution of a major soluble non-carbonate pore-filling cement. This is inferred to have been DE5 anhydrite (or halite). The pyrite appears to have corrosively replaced the wall rock cement and to some extent the detrital minerals (Plate 33).

South of the Mersey (with the exception of the Mickle Trafford Borehole) pyrite commonly occurs as inclusions in DE6e calcite cement in green or bleached rocks. It is accompanied by disseminated traces of chalcopyrite, Cu₂S (?chalcocite or digenite) and CuS (?covellite), with or without sphalerite, galena and Ni-Co-As sulphides. In most samples the sulphide paragenesis is unclear, but in siltstones and sandstones in the Tarporley Siltstone and Helsby Sandstone formations of the Saughall Massie, Hondslough Farm and Coton Fields boreholes, minor early DE1 framboidal pyrite is seen to be overgrown and included within coarser euhedral DE6d pyrite which, in turn, is partially replaced and overgrown by minor chalcopyrite. The chalcopyrite contains tiny inclusions of galena and sphalerite (either exsolved from chalcopyrite or precipitated prior to chalcopyrite). Similar sulphide mineralisation relationships are observed in the SSG in the Stowell Park Borehole [SJ SP 084 118] in the Worcester Basin (Green and Melville, 1956). A copper-mineralised sandstone (BGS reference collection E24678) examined from this borehole by BSEM for comparison with the Cheshire Basin samples showed a similar paragenesis: early framboidal pyrite, overgrown and partly replaced by coarse pyrite, followed by chalcopyrite then chalcocite, and finally enclosed in late (?DE6e-equivalent) calcite. The chalcopyrite in the Stowell Park material also contained inclusions of Ag-Cu sulphide.

The main areas of copper mineralisation (Clive, Alderley Edge etc.) have a very complex DE6d paragenesis. Paragenetic relationships identified in Clive Mine and West Mine are summarised in (Figure 94). Published mineral parageneses (Ixer and Vaughan, 1982; Naylor et al., 1989) for the other mines in the Alderley Edge and Bickerton mining areas are also shown in (Figure 94) for comparison with the diagenetic framework established here.

Only relicts of DE6d are preserved in Clive Mine, and evaluation of the detailed paragenesis was not possible. Relicts of altered Cu-Fe sulphide (probably chalcopyrite), largely replaced by secondary copper hydroxycarbonates and hydroxysilicates are all that remain of the primary DE6d sulphide ore. Traces of Bi-rich secondary oxide or oxyhydroxide, and ill-defined Pb-As secondaries indicate that minor Bi, Pb and As sulphides were also originally present in the DE6d assemblage.

Pyrite, chalcopyrite, and possibly chalcocite, appear to have been the main ore minerals in West Mine. These minerals can be found as relicts partially replaced by secondary iron-oxyhydroxides preserved in later DE6e calcite-baryte vein mineralisation developed along major fault planes (West Mine Boundary Fault) and remnants of the principal mineralised sandstone ore bodies (Figure 93). Calcite veinlets with chalcopyrite in the West Mine Boundary Fault clearly cross-cut earlier DE6a silicified fault gouge and contain brecciated fragments of DE6b dolomite-ankerite mineralisation, indicating repeated fault activity. Different faults within West Mine display different mineralisation characteristics. Late galena-baryte vein mineralisation is developed within the Chain Shaft Fault. Baryte was the initial vein mineralisation, locally penetrating a few centimetres into the adjacent wall-rock. The baryte-mineralised fault was then reactivated in places brecciating the earlier fill, and subsequently mineralised with galena. The galena is also largely limited to the fault, despite the porous nature of the present sandstone. Both baryte and galena cross-cut and include relict copper-sulphides, which here may have been chalcocite rather than chalcopyrite, indicating that they postdate the main copper-mineralising episode. The relationship between vein calcite (associated largely with the West Mine Boundary Fault) and galena (only observed in the Chain Shaft Fault) was not seen in West Mine, but by analogy with observations from the other Alderley Edge mines (Ixer and Vaughan, 1982; Naylor et al., 1989) it would appear that calcite veining postdates galena mineralisation. Traces of Ag₂S (?argentite or acanthite), HgS (?cinnabar) and associated, closely intergrown fine-grained gold grains (<2 urn), occur as inclusions within the relict chalcopyrite adjacent to the West Mine Boundary Fault.

Overall, observations in the Cheshire Basin suggest that Fe-sulphide (pyrite) mineralisation is followed by Cu+Fe sulphide mineralisation. However, in the majqr mineralised areas DE6d mineralisation varies in detail between different sites (Figure 94). In West Mine, trace Ag-Hg-Au-sulphide mineralisation preceded major Cu-Fe-sulphide mineralisation, which was subsequently followed by baryte then galena mineralisation. In the other Alderley Edge mines, and at Bickerton, minor Co-Ni-Fe and Cu-Fe sulphides preceded the major Cu-Fe-sulphide mineralisation which appears to have occurred in two main stages, the first accompanied by minor amounts of As-Ag-Sb. No Pb mineralisation was recorded by Naylor et al. (1989) from Bickerton but galena-Zn-sulphide mineralisation separated the two main episodes of Cu mineralisation at Stormy Point, Wood Mine and Engine Vein (Alderley Edge). No evidence of two episodes of Cu-sulphide mineralisation was apparent in West Mine, and galena is the last stage of sulphide mineralisation; this may indicate that only the first stage of Cu mineralisation is represented here.

DE6E

Major calcite and baryte cementation and veining followed the main stage of sulphide diagenesis and mineralisation. The calcite is weakly ferroan to non-ferroan and typically forms poikilotopic pore-filling cements (lustre-mottle textured in hand specimen) which enclose and often replace earlier carbonate cements. Up to two generations of calcite cement are recognised in many places, but mineralisation is more complex in the Alderley Edge area where four generations of calcite veining (associated with repeated fracturing and brecciation of earlier-mineralised vein fabrics) can be distinguished in West Mine (Figure 94). Major baryte cementation and vein mineralisation are closely associated with DE6e calcite. In the Gallantry Bank boreholes, major poikilotopic baryte cement is seen to post-date, or is coeval with, major poikilotopic weakly ferroan calcite cement. This is almost certainly an oversimplification since Naylor et al. (1989) identified three possible generations of baryte in this area, syngenetic with calcite mineralisation which continued after baryte mineralisation ceased (Figure 94). At least three generations of baryte can be distinguished in the Clive Mine, but whether this belongs to DE6e or DE6d is unclear since DE6e calcite is absent or not preserved. Ixer and Vaughan (1982) and Naylor et al. (1989) also recognised three possible generations of baryte in the Alderley Edge area, although only one generation has been recognised with any certainty in West Mine in the present study. In common with other DE6 mineralisation, the calcite and baryte vein-fills often show little or no penetration into now-porous sandstone wall-rocks (cf. Plate 33).

6180 and δ¹³C data for DE6e calcites are presented in (Table 22) and(Figure 92). DE6e calcite is isotopically lighter than early (DE1b + DE3) diagenetic carbonates (although δ¹⁸O is similar to DE6b dolomite), consistent with precipitation during late diagenesis (deep burial) from a warm fluid. Overall, there is a slight shift towards lower δ¹⁸O values from south to north across the basin, but the trend is not marked. In contrast, the minimum δ¹³C values at each locality, which represent samples with the least contamination by earlier diagenetic carbonates, show a very strong systematic variation across the basin (Figure 95): from moderately heavy carbon signatures in the south (−2.4‰ in the Childs Erc211 Borehole) to very light signatures in the north and north-west (–26.9‰ in the Thornton Borehole). The heavy DE6e calcite δ¹³C signatures in the south of the basin closely resemble early DE1b-DE3 carbonate signatures, suggesting that DE6 carbonate was derived largely by re-working of early diagenetic and detrital carbonate components within the SSG. The δ¹³C characteristics of DE6e from the north of the basin are more typical of carbonates precipitated from fluids influenced by CO₂ derived from deep-burial diagenetic thermal maturation and decarboxylation of organic matter (Irwin et al., 1977; Boles, 1978; Moore, 1989; Franks and Forester, 1984). This implies that deeper-sourced fluids, probably from basement Carboniferous rocks undergoing thermal maturation of organic matter, significantly influenced DE6e carbonate diagenesis and mineralisation in the northern and north-western part of the basin.

(Figure 95) also illustrates the distribution of DE6 δ¹³C in comparison with the occurrence of DE6 sulphide species. Discrete copper-mineral authigenesis is only present in areas where DE6 carbonates are isotopically heavier than –8‰ δ¹³C _PDB (i.e. in the Wirral and to the south and east of the basin). Pyrite, rather than copper-sulphide mineralisation and authigenesis, dominates in areas where the associated DE6 carbonates are isotopically lighter than –8‰ δ¹³C _PDB (i.e. to the north of the Mersey and around the Mickle Trafford Borehole). Geochemical modelling studies (Chapter 6) indicate that Cu is only mobile in relatively oxidising groundwaters in red-bed sediments. The –8‰ δ¹³C_PDB contour for DE6 carbonate may therefore represent the redox boundary north of which (except locally in the Wirral) fluids were too reducing for copper mobilisation during DE6 because of a significant contribution from reducing mineralising fluids derived from deeper-sourced organic-rich (?Carboniferous) rocks undergoing thermal maturation. South of this area, there is little evidence of deeper-sourced fluids having influenced the SSG. Fluids here were more oxidising and the carbon-isotope composition was influenced largely by fluid–rock interactions within the red-bed SSG.

Greater significance of intraformational water–rock interaction in the south of the basin may also account for the heavier δ¹⁸O characteristics of DE6 carbonates. Although δ¹⁸O fractionation between water and precipitated carbonate is predominantly temperature-controlled (e.g. Friedman and O'Neil, 1977; Irwin et al., 1977; Anderson and Arthur, 1983; Moore, 1989) and isotopically lighter carbonate is precipitated with increasing temperature, by the same process subsurface water-rock interaction results in progressively heavier δ¹⁸O values in subsurface fluids (Land and Prezbindowski, 1981). Thus δ¹⁸O fractionation in carbonate cements may reflect the degree of water-rock interaction as well as the influence of temperature. In this respect, the slightly heavier δ¹⁸O found in DE6 carbonates in the south of the basin is also consistent with greater influence of water–rock interaction within the Permo-Triassic pile (including reworking of intraformational marine carbonates), in contrast to significant influence from deeper-sourced fluids at higher temperature.

Geochemical profiles at West Mine, Alderley Edge

A series of spot samples of wall-rock were taken at intervals on traverses on each of the three levels worked in the western part of West Mine, extending north-eastwards from the south-west boundary fault to the furthest extent of the stoped-out ground. These represent sampling away from the main ore-fluid conduit, at the margins of the mined ore bodies. Many of the samples are enriched in Cu, Pb, Zn, MnO, Ba, Sr, Co, Ni, As and Cr, with concentrations significantly above the median values for the Helsby Sandstone (Figure 96) and (Figure 97). These values fit well with the observed ore-mineral assemblage at Alderley Edge as described above and by Ixer and Vaughan (1982) and Naylor et al. (1989). Cu contents are the most enhanced relative to median values, attaining a peak of approximately 1000 X median. Ba is the next most enriched, followed by, in order of relative abundance, Co, Ni, Zn, As, MnO, Pb, Sr and Cr; the latter being only 5 X median, at most.

SiO₂ values (Figure 96) are generally greater than the Helsby Sandstone median of 84.3% on the lower and upper levels, but at or below the median for the middle level. With one exception, the Al₂O₃ contents are below the median for the formation. This suggests that the host rocks for mineralisation in West Mine are relatively clean, quartz-rich sandstones with low contents of clay, mica and feldspar. They are predominantly aeolian and contain few mudstone intercalations (Warrington, 1980).

Although concentrations of ore elements are significantly higher than average, the patterns of individual elements often differ from one level to another and those of different elements are rarely similar, even on the same level (Figure 96) and (Figure 97). The main exception is the close agreement of the Ba and Sr profiles, which indicates that most of the Sr occurs in baryte. Co and Ni patterns are generally similar, and MnO and As are comparable to Fe₂O₃. On the lower level there is a coincidence in the peak values of Cu, Co, Ni, Fe₂O₃, As and MnO. CaO contents indicate that significant calcite was only found 5–20 m from the fault on the middle level.

The overall complexity of the profiles and the lack of inter-element correlations suggest that, generally, each element occurs in a different mineral phase. Thus not only do Cu, Pb and Zn occur as separate sulphides (or oxidative derivatives), but also Co-Ni, As-Fe-Mn and Cr appear to be in distinct mineral species. The complex patterns also fit with the complex paragenesis at West Mine, as described earlier in this section, and with the irregular pattern of stoped ground seen in the workings, with variable distance of mined sandstone relative to the bounding fault on each of the levels. The patterns have almost certainly been severely affected by later oxidation of the primary sulphides – for example, very recent telodiagenetic mobilisation of Cu is strikingly apparent in the mine (see DE8: Late telodiagenesis).

DE7: Late mesodiagenetic cement dissolution, hydrocarbon migration and illite authigenesis

Based on criteria discussed above, it is clear that much of the present porosity in the SSG and Tarporley Siltstone Formation is secondary. Most of this is probably of telodiagenetic origin and results from the dissolution of carbonate as well as evaporite cements within the relatively recent (Tertiary–Recent) meteoric groundwater regime following basin uplift and inversion. Minor hydrocarbon shows are found within the SSG around the basin margin (Thornton, Mickle Trafford, Alderley Edge, Grinshill, Little Ness; pp.169 and 222): This hydrocarbon occurs in locally porous horizons, yet largely predates any carbonate cement dissolution. Some sandstones also display traces of pore-lining fibrous illite developed within pores not cemented by carbonates. It was not possible to evaluate the extent and significance of illite development in the samples obtained due to its poor preservation: most illite fibres had collapsed onto pore walls as a 'result of drying during core/sample storage (cf. discussion by McHardy et al., 1982). Using the criteria defined by Schmidt and McDonald (1979b), many illite-lined pores in the sandstones are considered to be secondary. Yet some carbonate cements are well preserved, and therefore it seems likely that the secondary porosity results from the dissolution of evaporite cements prior to illite authigenesis. Fibrous illite is best developed in the Bewsey Borehole, where it is seen to develop largely from earlier-diagenetic smectite-illite boxwork coatings on grain surfaces but locally it rests on well-developed DE4 quartz-feldspar overgrowths and patchy DE6 calcite cements.

DE8: Late telodiagenesis

Late telodiagenetic effects are evident in much of the unconcealed SSG and Tarporley Siltstone Formation from around the basin margins. They are characterised by oxidative dissolution of carbonate and sulphide cements and formation of secondary iron and manganese oxyhydroxides as pseudo-morphs and pore linings. Many of these secondary iron-manganese oxides contain enrichments of Ba, Cu, As, Co, Ni, Pb, Zn detectable by EDXA at levels probably of the order of up to 1%. U fission-track analysis also demonstrates that U is mobilised during telodiagenesis and locally concentrated in these secondary iron and manganese oxyhydroxides which are developing from dissolving ferroan calcite and ferroan dolomite-ankerite cements, and altering detrital mafic grains and volcanic rock fragments. Caution is therefore required when interpreting the trace-element characteristics of the SSG rocks, especially with respect to the transition metals and other redox-sensitive elements, which have quite clearly been remobilised and redistributed in most of the samples examined.

Detrital igneous rock fragments, feldspars and micas are also affected by telodiagenesis and are commonly seen undergoing replacement by coarse vermicular or book-like crystals of kaolinite. Detrital chert grains also appear to be unstable and are commonly partially dissolved.

In the Alderley Edge and Clive copper mining areas, DE8 telodiagenesis has resulted in a spectacular and very complex assemblage of secondary Cu-Pb-Co-Ni-As-P carbonate, hydrated silicate and hydroxide/oxide minerals. The assemblages are dominated by chrysocolla, dioptase, malachite, azurite, mimetite, pyromorphite, cerrusite, hydrocerrusite and smithsonite. More detailed accounts of the complex secondary supergene mineralisation are given in Chapter 7 of this volume and by Warrington (1994a, b).

Preserved carbonate contents from whole-rock geochemistry

Whole-rock geochemistry of the SSG shows a good correlation between CaO, MgO and LOI (and MnO, except in the Kinnerton Sandstone). There is a clear association of these elements with carbonate minerals, which are present only as diagenetic cement or vein-filling phases; there is no significant component of detrital carbonate. It is possible to identify many points on CaO and MgO v LOI plots which fall on dolomite and calcite trends (Figure 98). From such plots, for individual formations, an estimate can be made of the preserved calcite and dolomite contents of the rocks.

The Kinnerton Sandstone shows strong calcite and dolomite trends. The maximum dolomite content is about 9% (Halewood Borehole), but most samples contain less than 2%. Calcite contents range up to up to about 28% (Newport UDC Borehole), but are mainly less than 3%. Dolomite predominates in the north-west of the basin and calcite in the southeast. This geographical division probably reflects the distribution of nodular eodiagenetic carbonates as described on p.127.

The Chester Pebble Beds also show both calcite and dolomite trends. Dolomite and calcite contents are each mostly less than 9% but, although dolomite appears to predominate, its maximum level is only about 16% (ICI Widnes Borehole), whereas the maximum for calcite is almost 40% (Bearstone Mill Borehole). As in the Kinnerton Sandstone, higher dolomite values are restricted to the north-west of the basin, but high calcite contents occur in both the south-east and the north-west.

The Wilmslow Sandstone generally has low levels of carbonates. Dolomite is more prevalent than calcite, but contents of each are mainly below 2%. One sample from the Marston Salt Union Borehole has about 8% dolomite and another, from the same borehole, around 7% calcite. Three samples from Holcroft Lane Borehole have higher calcite values, ranging up to 14%.

Dolomite also seems to be the more important carbonate in the Helsby Sandstone, with maximum levels up to about 25% (Mobberley Town 5 Borehole), although most samples have less than 6%. Calcite ranges up to about 17% (Hondslough Farm Borehole), but most samples contain less than 7%.

There appears to be some facies control of the preserved carbonate content: greater quantities are found in the Chester Pebble Beds and Helsby Sandstone, which contain significant fluvial facies, than in the Wilmslow or Kinnerton sandstones which, in the samples collected, were dominantly of aeolian origin.

Mercia Mudstone Group

General comments

The MMG shares many sequences of diagenetic cements and vein-fills in common with those found in the underlying SSG. However, because of the much finer grain size, the paragenetic fabrics are often difficult to define. The Tarporley Siltstone Formation (in particular the more arenaceous facies) shows very similar characteristics to the SSG and the description of its diagenesis has been largely included with that of the SSG strata. As in the SSG, diagenetic modification can be ascribed to eodiagenetic (pre-burial), mesodiagenetic (burial) and telodiagenetic (near-surface meteoric alteration) processes. The diagenesis is summarised in (Table 23).

Eodiagenesis: DE1, DE2 and the development of oxidised and reduced facies

The most visually distinctive diagenetic feature is the green and red colouration commonly observed in the MMG. In the laminated facies, which occurs chiefly in the Bollin Mudstone Formation and Byley Mudstone Formation, there are alternations of oxidised and reduced laminae. Colour lamination is generally concordant with lithological lamination. The boundaries between laminae of different colours may be sharply defined, though diffuse boundaries are common where grey laminae overlie red laminae. The rocks of the structureless facies are basically brownish red, with local pale grey-green spots and patches; at the top of structureless units, where they are overlain by grey sediments of the laminated facies, the colour is commonly grey. Arthurton (1980) considered that the differentiation into oxidised and reduced sediments was 'partly a primary sedimentary feature and partly the result of early diagenetic reduction of ferric iron compounds in pre-existing oxidised sediment'.

Previous geochemical studies in the MMG have shown that the reduced laminae generally contain less ferric iron than the interbedded oxidised laminae (Dunham in Stevenson and Mitchell, 1955, p.62, quoting Maw, 1868 and Moody, 1905; Aljubouri, 1972; Haslam and Sandon, 1991; H W Haslam, unpublished data from the Kirkham borehole to the north of the Cheshire Basin, near Preston).

The same feature has been found in the present study, with total iron (as Fe₂O₃) significantly lower in reduced samples of the MMG (Figure 99); the relationship also holds when the data are normalised to Al₂O₃ (Figure 100). Examination of spidergrams of oxidised and reduced laminae in the MMG confirms that in most sets of neighbouring samples Fe₂O₃ contents are lower in the reduced laminae e.g.(Figure 101) and (Figure 102).

In general, the oxidised facies have a higher detrital component (reflected by SiO₂ and Al₂O₃ contents) and a lower diagenetic component (CaO and MgO) than the reduced facies. An example, from the Tarporley Siltstones of Saughall Massie Borehole, clearly shows higher Ca, Mg and Mn and lower contents of most other elements in the reduced samples, suggestive of a greater carbonate content and a lower detrital component (Figure 102). In other examples, such as the Byley Mudstones of Crewe Borehole, the reduced samples contain more sand or silt than the neighbouring oxidised samples and hence have higher SiO₂ and lower Al₂O₃ (Figure 101). The presence of a greater proportion of argillaceous material in the oxidised rocks explains the fact that values for Pb, Zn and Cr are also generally higher in the oxidised facies (Figure 99), relationships which do not persist when the data are normalised to Al₂O₃ (Figure 100).

Values for Cu (and, in some samples, Co) are significantly greater in the reduced facies (Figure 99)-(Figure 102) as a result of mesodiagenetic processes (DE6: see pp.158 and 174).

BSEM petrographic observations show clearly that whilst the red mudstones contain minor disseminations of fine-grained iron oxide, this is absent in the green and grey facies which instead contain major to minor amounts of framboidal pyrite or fine-grained pyrite (Plate 34). As discussed in connection with the SSG (see p.133) framboidal pyrite is commonly considered to be associated with early diagenetic bacterial sulphate reduction (typically within a few centimetres of the sediment–water interface). The pyrite in the MMG is clearly eodiagenetic and can be seen to predate compaction, as demonstrated by compactional deformation of sedimentary laminae, and less competent mica grains, around framboidal aggregates.

The detrital component of the sediment is considered to be wind-blown (Arthurton, 1980), derived from desert terrain, and contained ferric material, formed in the source area, as pellicles around quartz grains and as discrete, fine dust particles. Some organic matter was also deposited in the sediment (e.g. Warrington and Ivimey-Cook, 1992). Early diagenetic bacterial decomposition of the organic matter would have created local reducing pore fluids leading to reduction of the ferric compounds to ferrous iron associations. (This process is clearly demonstrated in the SSG of the Stowell Park Borehole, Gloucestershire, where abundant framboidal pyrite is concentrated around detrital woody or coaly fragments dispersed within the sandstones.) The final state of each bed would have depended on the relative proportions of ferric iron and organic matter in the detritus. If, when all the organic matter had been oxidised, there remained some ferric oxide, the rock would be red-brown in colour. If, on the other hand, the organic matter was present in sufficient abundance to reduce all the ferric iron, the resulting rock would be grey-green in colour, and the reducing environment would be favourable for the preservation of sulphide minerals and microfossils. This process explains the fine alternation of oxidised and reduced laminae in a sequence of sediments deposited under similar conditions.

It is likely that the ferrous iron produced by the reduction of ferric oxides was initially taken into solution in the pore water. Some was probably incorporated into pyrite and iron-bearing silicates, such as chlorite and corrensite; some may have been co-precipitated with Ca and Mg in carbonate cement; and the remainder was removed in solution, accounting for the low total iron content.

Another early diagenetic or synsedimentary feature is the development of nodular anhydrite in many mudstones. This caused expansive disruption of sedimentary fabrics and is therefore considered to have formed near-surface, probably in evaporitic or sabkha-type environments in the MMG. At shallower levels, the anhydrite has been replaced by gypsum as a result of telodiagenetic hydration (e.g. in the Wych Mudstone Formation of the Crewe Heat Flow Borehole). Small (up to 5 mm) displacive clusters of needles and euhedral poikiloblastic swallow-tailed twinned crystals of gypsum were precipitated near the surface in some porous siltstones in the Brooks Mill Mudstone Formation (Audlem AU17 Borehole) probably akin to the development of modern-day 'desert rose' gypsum in sabkha environments. However, these have been completely pseudomorphed by later mesodiagenetic pyrite and chalcopyrite. Patchy micritic non-ferroan calcite cement is also locally present as expansive nodules within pelloidal mudstones in the Brooks Mill Mudstone Formation (Audlem AU17 Borehole) and may represent calcrete development.

Early diagenesis can be summarised as follows, using the scheme adopted for the SSG:

DE1a: synsedimentary development of red-bed clay and iron-oxide grain coatings (Tarporley Siltstone Formation only).

DE1b: precipitation of near-surface disruptive nodular anhydrite and gypsum 'desert rose' cements in unconsolidated mudstones in sabkha environments (all formations); precipitation of nodular dolocrete (Tarporley Siltstone Formation only); precipitation of nodular calcrete (Brooks Mill Mudstone Formation only).

DE2a: early pore-filling and pore-lining authigenic corrensite and smectite; these were recognised in the Tarporley Siltstone Formation directly as authigenic minerals (Chapter 4), but corrensite identified by XRD in other formations is consistent with early synsedimentary evaporative lacustrine environments. Relationship with DE1 is unclear.

DE2b: reduction of disseminated detrital fine-grained ferric oxide in association with localised bacterial sulphate reduction and decomposition of organic matter, accompanied by the precipitation of framboidal pyrite.

This broadly resembles DE1–3 diagenesis in the SSG. However, since these features are synsedimentary in origin they are obviously not syngenetic with DE1 diagenesis in the SSG).

Mesodiagenesis: DE3

Burial diagenesis in the MMG shows many similarities with the sequences of cement and vein paragenesis in the underlying SSG and is summarised in (Table 23).

The dolomicritic–anhydritic mudstones usually display some degree of recrystallisation. Sometimes this is associated with infilling of cubic dissolution cavities after synsedimentary halite crystals (Plate 35). Dolomite-cemented fabrics are typically uncompacted but in the later growth stages the dolomite becomes ferroan. Early dolomite recrystallisation and cementation also clearly overprints (therefore postdates) the soft sediment deformation and dewatering structures. This suggests that dolomite authigenesis developed during the very early stages of burial, in porewaters which were initially oxidising, or possibly syngenetic with DE2 (since early non-ferroan dolomite crystals may be coated in fine pyrite). However, as burial progressed the porewaters became more reducing below the zone of bacterial sulphate reduction (cf. Irwin et al., 1977). The later stages of dolomite growth became ferroan, particularly in the reduced rocks (this may be related to the release of Fe²⁺ in solution when ferric oxides were reduced in DE2b). A parallel can be seen between this and the DE2 to DE3 stage of diagenesis defined in the SSG.

Mesodiagenesis: DE4

Silty laminae in the laminated dolomicritic–anhydritic mud-rocks commonly display well-developed euhedral quartz, K-feldspar and albite grain overgrowths. These may locally fuse together to form a rigid cement. Some quartz-cemented silt laminae and interbeds preserve relatively open uncompacted fabrics, while other horizons show some degree of compaction. Quartz cements clearly developed after the formation of dewatering fabrics observed in the MMG. However, quartz cements are inhibited by earlier non-ferroan dolomite cements, but are often included by later anhydrite cements. Within the main halite of the Northwich Halite Formation and in anhydrite beds in the Wych Mudstone Formation, authigenic euhedral porphyroblasts of quartz are developed, locally replacing the evaporites and also preserving, as corroded inclusions, early unrecrystallised and largely uncompacted primary aphanitic anhydrite fabrics e.g. (Plate 36). This clearly indicates that silica was mobilised early during mesodiagenesis, after dolomite mineralisation but before later anhydrite remobilisation and cementation. The paragenetic relationship of silicate authigenesis in the MMG mirrors DE4 in the SSG.

Mesodiagenesis: DE5

Poikilotopic anhydrite cements appear later during the diagenesis of the MMG. They post-date and enclose quartz and dolomite overgrowths in coarser sediments, and infill dissolution cavities after halite, lined by earlier dolomite cements. Halite has been clearly remobilised during mesodiagenesis and reprecipitated within veins and as poikilotopic cements in the MMG. Although some halite dissolution fabrics predate DE3 dolomite and DE5 anhydrite cements, in other cases cross-cutting halite veins and cements clearly postdate, corrode and enclose earlier DE5 anhydrite cements. These fabrics again mirror the DE5 stage evaporitic cements postulated in the SSG.

Mesodiagenesis: DE6

Late-stage ferroan dolomite and ankerite occur as rhombic overgrowths on corroded mesodiagenetic dolomite in the Brooks Mill Mudstone Formation (Audlem AU17 Borehole). The ferroan dolomite–ankerite is seen to postdate quartz overgrowths and be replaced by late diagenetic chalcopyrite, chalcocite and pyrite. Unfortunately, the relationship between this and 'DE5' anhydrite cement is not seen. However, in the Crewe Heat Flow Borehole veins of (DE6c equivalent) anhydrite lined by earlier ferroan dolomiteankerite (identical to the DE6b–DE6c paragenetic sequences seen in the SSG and therefore probably related) place ferroan dolomite-ankerite authigenesis in the MMG in a similar paragenetic position to DE6b features in the SSG.

The framboidal pyrite in the MMG is commonly overgrown and replaced by later diagenetic sulphides of Cu, Fe ± Zn, Pb, As, Co, Ni, Ag, Hg and, rarely, traces of Ag-Cu-Au (± Zn) alloy (e.g. sample CHB 777 from the Stafford Halite, British Salt BH4). Coarser late-stage minor chalcopyrite, chalcocite and pyrite also locally replace anhydrite cements in the same rocks. These authigenic sulphides also corroded and replaced ferroan dolomite and ankerite cements. In the Brooks Mill Mudstone Formation (Audlem AU17 Borehole) chalcopyrite and late relatively coarsely crystalline pyrite both replace and pseudomorph small clustered needles and swallow-tail crystals of eodiagenetic gypsum. This sequence: early framboidal pyrite, followed by coarser pyrite, overgrown and overprinted by chalcopyrite, corresponds very closely with the paragenetic relationships illustrated by DE6d in the SSG. The direct replacement (although very small scale) of anhydrite cements by later sulphides is common in many of the MMG rocks and lends support to the mechanism of sulphide mineralisation by replacement of anhydrite-cemented wall-rock, proposed in the SSG (p.175).

Samples from oxidised and reduced laminae of the Byley Mudstone Formation in the Crewe Borehole (Figure 101) are seen to be very similar in chemical composition, the only systematic differences being that Fe is lower and Cu higher in the reduced laminae than in neighbouring oxidised laminae. In a suite of samples from the Tarporley Siltstone Formation of the Saughall Massie borehole (Figure 102) the reduced samples contain more Ca than the oxidised samples, accompanied by higher Mg and Mn. This can be attributed to the presence of major DE6e calcite cements within the reduced rocks and none in the red laminae. This suggests that the reduction-affected strata here were more permeable at the time of diagenesis. Most other elements are slightly lower in the reduced samples, compensating for the higher Ca and Mg. The differences between the oxidised and reduced samples observed in the Byley Mudstone (lower Fe and higher Cu in the reduced laminae) are also apparent in the Tarporley Siltstone, and additionally the reduced laminae are here enriched in Co. This can be attributed to later minor to trace chalcopyrite, chalcocite and covellite authigenesis in the green facies of the Tarporley Siltstone Formation (which can be directly correlated with the DE6d stage of diagenesis seen in the underlying SSG). This tends to be focused around, and replace, earlier DE1 framboidal pyrite.

Similarly, rare late cross-cutting calcite veinlets carrying small amounts of baryte may be correlated with DE6e calcite veining in the SSG.

Mesodiagenesis: DE7

Apart from late fibrous pore-lining illite in the arenaceous facies of the Tarporley Siltstone Formation the higher strati-graphical units do not appear to demonstrate diagenetic features similar to DE7 in the SSG.

Telodiagenesis: DE8

The most significant telodiagenetic effect in the MMG is the development of collapse breccias and major subsidence associated with the large-scale dissolution of thick halites by contemporary groundwater, particularly around Nantwich in

Cheshire. This is described in more detail by Wilson (1993). However, other telodiagenetic effects include the hydration of anhydrite cements and their replacement by gypsum in near-surface MMG rocks. Gypsum has a larger molar volume than anhydrite. The hydration process causes expansive disruption of the rocks which is associated with the development of cross-fibre or 'satin-spar' gypsum veining formed by a progressive crack-seal mechanism of hydration possibly causing overpressuring during expansion, resulting in hydrofracturing and mineralisation (cf. Shearman et al., 1972; Ramsey, 1980; Stoneley, 1983; Barker, 1990). Secondary gypsum veins tend to be subhorizontal but are often non-systematic and irregular. During anhydrite hydration and replacement by gypsum, minor local porphyroblasts of authigenic celestite are produced as a result of the release of Sr from the altering anhydrite (Plate 37).

Origin of sulphates and sulphides

Sulphur-isotope data of sulphur-bearing mineral species may be used to place constraints on the source of sulphur in the red-bed mineralisation in the SSG. Sulphur stable-isotope (δ³⁴ScDT) data for DE6 baryte and sulphide minerals from the SSG and the Tarporley Siltstone Formation are presented in (Table 24) and 8³⁴S variations within the minerals are illustrated in (Figure 103) (which incorporates additional data for a wider range of minerals from Naylor et al., 1989).

Sulphide minerals

(Figure 103) shows a very wide range of δ³⁴S values (approximately –9 to +33‰) for sulphide minerals, even for the same mineral species. The range in values is wider than that reported previously (+1.3 to +15.7‰) by Naylor et al. (1989), and galenas in particular show a wide range in composition. The data clearly demonstrate that the sulphur was not derived from magmatic sources, which typically exhibit δ³⁴S of 0 ± 4‰ (Ohmoto, 1986).

Naylor et al. (1989) suggested that the different isotopic compositions were reflected in the textural variability of galenas from Engine Vein (Alderley Edge): they found that massive vein galenas had a mean δ³⁴S of 14.9‰, whereas disseminated galena cementing sandstones and conglomerates yielded δ³⁴S values of 5.9 ± 3‰ (variation 4.1–11.5‰). This relationship did not hold for galenas examined in the present study, which included analyses of galena from the Engine Vein, Wood Mine, West Mine and Stormy Point areas of the Alderley deposits. Indeed, even a massive vein galena from Engine Vein dump material (CHB 505) was found to have δ³⁴S as low as 8.1‰, i.e. within the range reported for disseminated ore impregnations by Naylor et al. (1989).

Sulphides from the north-western part of the basin (Saughall Massie, Thornton, and ICI Widnes boreholes) are also shown in (Figure 103). Although the data set is limited, the observed variation in δ³⁴S is even greater than that seen in the Alderley Edge area. Authigenic chalcopyrite replacing framboidal DE2 pyrite from Saughall Massie (CHB 307/1) has the lightest δ³⁴S (–8.7‰) recorded from the basin. This value is consistent with biogenic reduction of a sulphate source with an isotopic composition similar to that of the maximum observed in the Alderley Edge sulphide and sulphate mineralisation (Ohmoto, 1986). The range in values between these two points could be indicative of a system open to H₂S, or of a non-equilibrium system.

Fluid-inclusion studies show that mineralisation is likely to have occurred at low temperatures (<70–80°C) from moderately saline fluids (see p.167 and Shepherd, 1994). Therefore, it is kinetically unlikely that isotopic equilibrium was achieved within the sulphate(fluid)–sulphide(mineral) system (Ohmoto and Lasaga, 1982; Ohmoto, 1986). At these temperatures, the most probable mechanisms by which reduced sulphur may be generated at the depositional sites are: diagenesis of unstable sedimentary sulphides; decomposition of organically bound sulphur (unlikely in the organic-poor SSG), and biogenic sulphate reduction (requiring introduction of an organic nutrient source in the SSG). Abiogenic sulphate reduction is generally not considered to be viable at temperatures less than 175°C (Ohmoto, 1986).. It has been proposed earlier, that the SSG may have had significant anhydrite cement prior to DE6 mineralisation (p.139) and this could have provided an in situ source of sulphate. Under low-temperature conditions, aqueous sulphide and sulphate species have a very low rate of sulphur-isotope exchange and retain their original δ³⁴S signatures (Ohmoto, 1986). Thus, the range of δ³⁴S in the sulphides between –8.1‰ (in chalcopyrite associated with framboidal pyrite from Saughall Massie) and heavy sulphides from Alderley Edge with maxima (δ³⁴S = +16.9‰) close to that of DE6d–DE6e baryte (Figure 103) would be consistent with biogenic sulphate reduction under variable non-equilibrium conditions. The upper limit of δ³⁴S (c.15–18‰) in the Alderley Edge mines is probably close to the original δ³⁴S of the fluid-phase sulphate. Highly variable δ³⁴S and disequilibrium relationships are typical of red-bed-associated Cu deposits and mineralising systems with low fluid H₂S/metal ratios (Ohmoto, 1986).

DE6d sulphides from the Thornton and ICI Widnes boreholes are anomalously heavy (δ³⁴S = +19 to +34‰). These boreholes are situated to the north of the Mersey and fall outside the area of discrete Cu sulphide mineralisation, with associated DE6 carbonates being characterised by very light δ¹³C values (Figure 95). The isotopically light calcites indicate a significant role for fluids which have been influenced by organic matter (see p.148). Low δ¹³C

O₂ can be released during early diagenesis of organic matter and during thermal decarboxylation. The first stage largely coincides with the bacterial sulphate reduction zone, resulting in low δ³⁴S sulphides in open systems. δ³⁴S of organic matter can increase during diagenesis by sulphate reduction and fixation of organic sulphur or metal sulphides in closed systems (Ohmoto, 1986). Subsequent maturation of organic-sulphur compounds can produce very heavy H₂S and this may account for the heavy δ³⁴S in DE6 pyrite in the north-west. Fluid-inclusion analysis of associated DE6e calcite also indicated that the fluids in the north-west were geochemically different to those at Alderley Edge, containing a significant MgCl₂ component (Shepherd, 1994). All these data argue strongly for the influence of a deeper-sourced fluid derived from matured organic-rich sediments (?Carboniferous basement) to the north of the Mersey and in the Chester (Mickle Trafford) area.

Isotopic relationships between baryte mineralisation in the SSG and evaporites in the MMG

In contrast to the sulphides, δ³⁴S ratios of baryte from Alderley Edge have a limited range of values (c.+11‰ to +18‰), which do not vary significantly between the different mines (Figure 103). Baryte values are generally heavier than those for the sulphide minerals, with the mode value (c.17‰) representing the upper limit for sulphide. As discussed above, the baryte and sulphide probably derived sulphur from similar sources but were deposited under conditions of isotopic disequilibrium. The 8MS values of DE6 baryte from the Clive–Grinshill, and Gallantry Bank areas are similar to each other (Figure 103) but are generally isotopically heavier (c.+ 15‰ to +20‰) than at Alderley Edge.

Naylor et al. (1989) noted that the baryte from Clive–Grinshill and Bickerton was isotopically similar in δ³⁴S values to the evaporites within the MMG. They therefore concluded that much of the baryte sulphate was derived ultimately from Triassic evaporites. They invoked the involvement of a lighter sulphate component, derived by oxidation of reduced sulphide from the underlying Carboniferous, to account for the lighter baryte in the Alderley Edge area. Their comparisons of baryte with evaporite sources were based on only three samples of anhydrite and gypsum (from the Tarporley Siltstone, Bollin Mudstone and Northwich Halite formations) and limited published data (Taylor, 1983) for the Brooks Mill Mudstone Formation. A more systematic study of sulphur isotope variation within the MMG evaporites in the Cheshire Basin was undertaken in the present study, and the data are summarised in (Figure 104).

(Figure 104) compares δ³⁴S values for the primary/DE1 and hydrated (gypsified) Cheshire Basin evaporites and hydration products with a best estimate for the isotopic composition of sea water during the Middle and Upper Triassic (Curve A; based on calculations by Taylor (1983) and with data for other European Middle and Upper Triassic evaporite deposits.). A modified sea-water curve (Curve B, (Figure 104) is obtained for the Anisian if data from the Rot evaporites constitute (or reflect) a localised, rather than global, event (Taylor, 1983). δ³⁴S data for sulphates were obtained from all formations except the Byley Mudstone Formation and Northwich Halite Formation.

At 25°C, gypsum and anhydrite are enriched in ³⁴S by only 1.5‰ relative to aqueous sulphate (Ohmoto, 1986; Raab and Spiro, 1991). Consequently, hydration and replacement of primary anhydrite by gypsum will not produce any significant isotopic fractionation, and secondary gypsum should be similar to the original anhydrite, which in turn should closely approximate to the δ³⁴S value of the lacustromarine waters from which the MMG was deposited. Values of δ³⁴S obtained from late (DE8) vein gypsum need to be considered more cautiously, since they could introduce sulphate mobilised from a stratigraphically different level to the host rock. However, petrographic examination of these rocks indicates that most of the gypsum veining is closely associated with hydration of pre-existing synsedimentary anhydrite. This is supported by the isotopic data, which show a close correspondence between the composition of primary anhydrite and that of secondary gypsum within the same stratigraphical horizon (Figure 104). This suggests that the late telodiagenetic (DE8) gypsum veining represents only intraformational redistribution of calcium sulphate. The data show that below the Byley Mudstone Formation, δ³⁴S compositions of MMG evaporites closely follow the Triassic marine sulphate evolution (corrected for the Röt event) curve (Curve B), indicating a strong influence of Triassic marine-derived sulphate in the evaporitic and sabkha sediments of the Lower MMG. The S isotopic compositions are significantly depleted in ³⁴S, with respect to the Triassic sea-water curve, in the Wych Mudstone Formation and again in the Brooks Mill Mudstone Formation. This indicates a strong continental-derived brine influence, and greater isolation from Triassic marine circulation, in the Wych and Brooks Mill Mudstone Formations. The lighter δ³⁴S sulphate in these continental brines could feasibly have been dominated by re-working of earlier Permian sulphates (Permian sea water, particularly, was lighter than at other times, with δ³⁴S values of c.10‰). Although only a single analysis was obtained from the Wilklesey Halite Formation, the δ³⁴S value of primary anhydrite indicates a return to an evaporative environment influenced by a Triassic (Carnian) marine sulphate source, and is consistent with the development of a large thickness of halite in this formation.

(Figure 104) shows that the MMG sediments and DE1 evaporites generally became progressively depleted in ³⁴S with time; from values between 14.8‰ and 20.8‰ in the Tarporley Siltstone and Bollin Mudstone formations to values between 10.3‰ and 14.3‰ in the Brooks Mill Mudstone Formation. Comparison of these data with baryte from the SSG-hosted Cu-Pb-Ba mineralisation (Figure 103) clearly shows that the δ³⁴S values for baryte are all within the range of values of primary evaporites in the MMG. Since little ³⁴S fractionation occurs between solution and mineral on dissolution and re-precipitation of anhydrite or gypsum at low temperature (δ³⁴S = 1.5‰ at 25°C; Ohmoto, 1986), the baryte compositions in the SSG can all be accounted for by precipitation from sulphate ultimately derived from the MMG. There is no need to invoke the involvement of a second, deeper source of sulphur (as was proposed by Naylor et al., 1989) to account for the lighter S-isotopic signatures of baryte in the Alderley Horst. The differences in δ³⁴S between the Alderley Horst mineralisation and Cu-Pb-Ba mineralisation in the south (Clive, Grinshill), and west (Bickerton) of the basin may simply reflect tapping of brines or sulphate sources from different parts of the MMG. The δ³⁴S data suggest that sulphur in the mineralised areas in the south and west of the basin was originally derived from sulphate brines dominantly tapped from sources within the lower part of the MMG (Tarporley Siltstone Formation to Northwich Halite Formation). In contrast, the Alderley Horst mineralisation has δ³⁴S signatures that could reflect sulphate from brines derived throughout most of the MMG sequence (Tarporley Siltstone Formation to Brooks Mill Mudstone Formation) but with a possible bias towards sediments between the Bollin Mudstone Formation and the Wilkesley Halite Formation.

The subtle differences in metalliferous mineralisation between the different areas (e.g. Pb, Zn, As, Ni, Co are greater in the Alderley area than in the south and west of the basin; whereas Se is enriched in the Clive mineralisation but not in the Alderley deposits) may similarly be related to different parts of the MMG being brine-tapped.

Fluid inclusion studies

Fluid inclusions potentially provide an important source of information concerning the temperature and composition of diagenetic fluids and for reconstructing the thermal burial history of sediments (Guilhaumou, 1993). The most important inclusions for this purpose are those preserved in quartz overgrowths and intergranular diagenetic cements. A second approach, though less direct, is to utilise the fluid inclusions in fracture cements. Unequivocal interpretation is only possible, though, if the age relationship between fracturing and mineral diagenesis is clearly understood.

For the SSG, detailed examination of more than 30 sandstone samples failed to identify early diagenetic (DE3- D4) inclusions of suitable size (>2 μm) for micro-thermometric analysis in the quartz overgrowths or carbonate cements. For this stage of the burial history, the only data are those of Naylor et al. (1989). Consequently, analysis was restricted to the DE6 fracture cements and vein material associated with faulting and fault-related fractures. To address the mineralising role of gravity-driven evaporite brines from the MMG into the underlying SSG, a laser ablation ICP-MS analysis was carried out of liquid inclusions in halite from the Wilkesley Halite Formation.

In the mineralised areas at Clive and Alderley Edge, three possible fluid-inclusion host minerals were considered: DE6 quartz, baryte and calcite (see (Table 25)). At both localities the quartz is very finely crystalline or chalcedonic and free from microscopic inclusions. By contrast, the baryte is ubiquitously inclusion-rich but dominated by very small (<5 μm) monophase, aqueous inclusions. Calcite is more restricted in its distribution but, where present, contains rare two-phase aqueous inclusions with low vapour/liquid ratios. Overall, the inclusions, crystal morphology and mineral assemblages are compatible with deposition from low-temperature fluids. Only seven of the samples were ranked suitable for further study (see (Table 26)).

In the Wilkesley Halite, the halite is commonly inter-grown with reddish-brown mudstone and displays relic chevron or cornet textures suggesting primary crystallisation fabrics. The thicker layers, however, comprise massive, colourless halite with no internal structure, indicative of secondary recrystallisation. Since the two members are interbedded, the recrystallisation fabrics are considered to be early diagenetic in origin. Thus the fluid inclusions represent samples of the evaporitic bitterns trapped during or soon after halite crystallisation. Five samples were selected for detailed LAMP-ICP-MS analysis (see (Table 28)). The brine inclusions are of three main types: monophase aqueous, two-phase (liquid+vapour) aqueous, and complex liquid-solid ± vapour varieties. The two-phase inclusions display variable and anomalously high vapour/liquid ratios, some of which are located on cleavage fractures and can thus be attributed to liquid loss. Others, however, are slightly overpressured and forcefully expel liquid when ruptured. This is attributed to the degassing of CO₂ or air that was trapped during inclusion formation. The complex liquid-solid inclusions are restricted to halite that is characterised by dense clusters of micron-sized anhydrite needles. Compared with the simple liquid or liquid + vapour varieties, the inclusions are very irregular in shape and are considered to have formed along inter-crystal boundaries; a process that would favour the contained sylvite (KC1) or carnallite (KMgCl₃.6H₂O), daughter salts indicative of potassium salt precipitation. The raw thermometric data for aqueous inclusions in baryte and calcite are given in (Table 26). To acquire equilibrium melting temperatures for the monophase liquid inclusions, it was necessary to overstretch them by fast cooling to random occlusion of solid material. None of the inclusions contain ed sylvite (KCl) or carnallite (KMgCl₃.6H₂O ), daughter salts indicative of potassium salt precipitation.

Microthermometric analysis of inclusions in vein minerals and fracture cements

The raw thermometric data for aqueous inclusions in baryte and calcite are given in (Table 26) . To acquire equilibrium melting temperatures for the monophase liquid inclusions, it was necessary to overstretch them by fast cooling to −100°C. This caused inelastic deformation of the inclusion walls and the development of anomalous vapour bubbles. Having achieved an artificial two-phase (liquid + vapour) system at room temperature, melting temperatures could then be measured under stable, equilibrium three-phase conditions (i.e. solid + liquid + vapour) at sub-ambient temperatures.

Clive Mine

Baryte from Clive Mine displays two discrete fluid populations: a primary, low-temperature (<80°C), low-salinity fluid (0.5 to 9.0 wt% NaCl equivalents) and an overprint of higher-temperature (homogenisation temperature, T_h = 87–102°), high-salinity fluids (14–21 wt% NaCl equivalents). Both fluids display low first-melting temperatures (<40°C), consistent with mixed Na-divalent-cation chloride fluids as shown in (Table 27).

In theory, first melting temperatures (T_fm) are equivalent to eutectic temperatures (T_e), but for very small inclusions it is not always possible to recognise the initiation of melting. However, inclusions with T_fm values <−50°C indicate that the fluids contain significant amounts of CaCl₂ and can be modelled using the NaCl–CaCl₂H₂O salt-water system. For those inclusions with very high ice-melting (T_ice)temperatures (–0.3 to –1.7°C; 0.5 to 3.0 wt% NaCl equivalents) but no observable first melting, interpretation is more equivocal. Two explanations are possible: (1) the ice-melting temperatures indicate very dilute NaCl fluids, or (2) the fluids are sulphatic. Reference to the system H₂O-Na₂SO₄ predicts that significant quantities of Na₂SO₄ may be held in solution (up to 6 wt%) with little depression of the freezing point of pure water. Thus fluids associated with certain generations of baryte may have higher values of TDS (total dissolved salts) than estimated by reference to the system H₂O-NaCl.

West Mine, Alderley Edge

Material from West Mine proved very poor in inclusions. The calcite and baryte samples comprise clear crystals and display few micro- or macroscopic defects - characteristics typical of slow growth. Only six inclusions were of sufficient size to obtain thermometric data. T_ice melting temperatures for baryte (–0.3 to –0.8°C; 0.5 to 1.8 wt% NaCl equivalents) are similar to the high T_ice group recognised at Clive Mine. Calcite overgrowths to the baryte (CHB 699) do not contain primary inclusions: the only data (T_ice –0.3°C; 0.5 wt% NaCl) refer to a single secondary inclusion.

Thornton Borehole

The sulphidic carbonate vein intersected in this borehole is characterised by monophase aqueous primary/pseudosecondary (P/PS) and secondary (S) inclusions. The P/PS inclusions have consistently higher salinities (11.5 to 18.5 wt% NaCl equivalents) than the corresponding S inclusions (0.5 to 2.9 wt% NaCl equivalents). On first examination, the P/PS salinities appear comparable to those of the S inclusions at Clive Mine. However, an important difference is noted. The grouping of T_fm, values around −35°C suggests the congruent melting of MgCl₂.12H₂O and hence an additional MgCl₂ component. Data for phase relations in the quaternary system NaCl-MgCl₂-CaCl₂-H₂O are not available, but as a first approximation the measured final hydrohalite (T_hh) and ice-melting (T_ice) temperatures indicate monovalent/divalent cation ratios of –0.80 to 0.95, assuming that MgCl₂ behaves similarly to CaCl₂ in lowering the final hydrohalite melting temperature (Oakes et al., 1990).

Discussion

Baryte mineralisation in the Cheshire Basin is ubiquitously associated with low-temperature, variable low- to moderate-salinity fluids (0.5 to 9.0 wt% NaCl equivalents). No evidence could be found for unequivocal primary, two-phase (liquid+vapour) inclusions, and hence the thermal conditions for baryte deposition cannot be positively defined. However, the presence of monophase aqueous inclusions suggests that the fluids were unlikely to have been hotter than 70–80°C. Metastability (i.e. the failure to develop a vapour bubble) is a common characteristic of low-temperature inclusions (Roedder, 1971; Shepherd et al., 1985) and typical of sediment-hosted baryte deposits. The preferred temperature regime is in good agreement with the estimated mean basin temperature of −60°C at the base of the SSG for the end of the Triassic (Chapter 3).

At Clive Mine the ore fluids responsible for baryte deposition were generally more saline than those at Alderley Edge. Furthermore, the secondary inclusions record an overprint of hotter, more saline fluids (87–102°C; 14 to –21 wt% NaCl equivalents) which are similar to the late-stage fluids (17 wt% NaCl equivalents) trapped in calcites from Bickerton Copper Mine (Naylor et al., 1989). Baryte at Bickerton contains only monophase aqueous inclusions, indicating lower fluid temperatures. A further distinguishing feature of the Clive brines is that the inclusions show sharp and conspicuous first melting at temperatures close to the NaCl-CaCl₂-H₂O eutectic. This indicates a significant component of Ca in the ore fluids and suggests a possible link with the Pb-Zn-Ba ores of the adjacent West Shropshire Orefield (Pattrick and Bowell, 1991). Here, veins west and east of the Pontesford-Linley Fault (a south-westerly extension of the Wem-Bridgemere-Red Rock Fault System) carry baryte and traces of secondary copper minerals and are associated with higher-temperature fluids (>100°C) carrying up to 28 wt% dissolved salts. It is possible, therefore, that the high-temperature, saline overprint at Clive is a distal expression of this hydrothermal activity.

Fluids associated with fracture-controlled calcite-pyrite mineralisation in the SSG of the Thornton Borehole on the fringe of the East Irish Sea Basin (Figure 95) and (Figure 2), are alsocool (<80°). Although their average bulk salinity is similar to fluids in the southern part of the Cheshire Basin, they appear to be MgCl₂-rich.

With regard to the primary Cu-Pb sulphide ores at Clive and Alderley, fluid-inclusion data for the sulphide mineral- ising fluids sensu stricto could not be obtained. However, on the basis of the intimate relationship between sulphides and gangue minerals, and the obvious geochemical and mineralogical similarities between Alderley, Clive and Bickerton, it is concluded that the fluids responsible for the deposition of the metalliferous ores were similarly low tem- perature. No evidence could be found for the introduction of liquid hydrocarbons before or during sulphide deposi- tion. Thus, excluding late-stage fluids at Clive Mine, the fluid-inclusion evidence supports a low-temperature brine model for the genesis of Cheshire Basin red-bed Cu-Pb mineralisation.

LAMP-ICP-MS analysis of inclusions in halite

Results

Analytical results for individual fluid inclusions in halite are given in (Table 28). The data have been corrected for back- ground, isotopic abundance and relative sensitivities, andhave been normalised to Sr, since the method, as developed to date, cannot provide absolute concentrations. Based on data for present-day subsurface brines beneath Permian evaporites in the Upper Palaeozoic Palo Duro Basin and Mesozoic–Cenozoic basins of the Gulf of Mexico (Fischer and Kreitler, 1987; Land and Macpherson, 1992), a Sr value of 1000 ppm was selected for normalisation. Reported Sr concentrations range from 500 to several 1000 ppm but a value of 1000 ppm is considered representative of fluids with Ca values similar to those of the Wilkesley brines. Ideally, normalisation to Na would have been adopted. However, in the case of halite-hosted inclusions, distinction between host Na and solution Na could not be assured. As a further check on the correct choice of Sr concentration, estimated Na/K wt ratios (4 to 14), assuming NaCl saturation at 20^°C and Sr-normalised K values, are in general agreement with average Na/K wt ratios (7 to 10) reported for inclusions in massive halite (Ayora et al., 1993) The acquisition of trace-element data (Rb, Sr, Ba, Pb, Cu, Mn, Cs) for single inclusions marks a breakthrough in palaeo-fluid analysis; slight uncertainties in their true abundances have no significant influence upon the overall interpretation. Working detection limits for the trace elements are cal- culated to be in the range 5–10 ppm. For the purposes of interpretation however, normalised values less than 5 ppm are not shown in (Table 28).

Interpretation

Data for the stratigraphically lowest samples (HAL 2 and 5) in the Wilkesley Halite Formation are conspicuously depleted in K, Mn, Cu, Rb, Ba and Pb compared with data from all higher samples. This is consistent with the early precipitation of halite during the initial stages of evaporitic concentration. As the evaporitic process advances there is a progressive enrichment in these elements. In samples HAL 1 and 3 from the stratigraphically higher halite bed, the inclusions represent the parent evaporitic brine present during the main stages of halite precipitation. Cl/Br ratios obtained for the salt matrix also support this interpretation. Values for Mg and K are in good agreement with the fluid data for massive Miocene halite units in the Lorca basin, south-east Spain (Ayora et al., 1993). However, HAL 1 and 4 also contain highly potassic inclusions which are thought to represent late-stage potassic bitterns that migrated down through the underlying halite units to be trapped as inclusions during early burial diagenesis of the evaporite-mudstone sequence. These fluids are appreciably enriched in Mg, Rb and Pb, the systematic positive correlations with K (Figure 105)a, b, c) indicating that the trace-element enrichments are a direct result of an evaporitic process. Ba and Mn are enriched during the main stage of halite precipitation but show no specific relationship to advanced K enrichment (Figure 105)d, e. Ca concentrations tend to remain generally constant throughout the salt profile and probably reflect chemical buffering by anhydrite or gypsum. Compared with typical marine evaporites, the Ca/Mg and Ca/K ratios, however, are unusually high. According to Hardie (1990) this is diagnostic of the CaCl₂-rich brines found in intra-cratonic basins and may thus indicate restricted connection between the Cheshire Basin and an open-ocean source. Mg varies considerably between individual inclusions but is highest in K-rich inclusions (Figure 105)a. This is in accordance with increasing enrichment in K and Mg during advanced evaporation and the trend towards carnallite (KMgCl₃.6H₂O), sylvite (KCl) and bischofite (MgCl₂.6H₂O) precipitation. A significant feature of the data is the relatively high Mg/Ca ratio for the evaporitic bitterns and, by inference, high Mg/Na ratios. This imparts a characteristic signature to evaporitic bitterns and allows distinction from Ca-rich formational and oilfield brines whose chemistry reflects greater influence of fluid-mineral interaction at higher temperatures (Moldovanyi and Walter, 1992).

Most importantly, the estimated concentrations of Ba, Cu and Pb are well within the predicted ranges for highly metalliferous fluids. Thus the evaporite bitterns of the MMG constitute a newly discovered and potentially major source of metals for sandstone-hosted mineralisation in the Cheshire Basin. The complex inter-element correlations between K, Mn and Cu (Figure 105)e, f, g indicate, however, that the observed metal enrichments are not solely a product of evaporitic concentration but involve other processes operating during the diagenesis of the mudstone–evaporite sequence. The principal of these was probably metal adsorption/desorption on iron oxides (Chapter 7).

Organic geochemistry

Late mesodiagenetic hydrocarbon shows were found in the basin fill (p.150), and hydrocarbon seeps from in and around the basin were examined to assess the nature of the organic component. Both red sandstones and paler grey-green sandstones (mostly collected from borehole cores) contained some extractable organic matter, but only the latter contained alkanes in sufficient quantities for analysis.

Nine surface oil and tar seeps were examined. Most of them were just outside the basin margins, collected from locations that suggested a possible link to the processes of hydrocarbon generation within the basin Other outcrop and borehole samples were also collected, where there was a likelihood of organic matter being present. The seeps were generally too biodegraded to have recognisable alkane patterns, but other outcrop samples (e.g. oil trapped within vein material) were sufficiently well preserved to allow useful data to be obtained. Because of the effects of weathering and biodegradation, it was only possible to get numerical data from ten of the samples.

One sample, from a seep in Westphalian strata at Hem Heath Colliery just outside the south-east margin of the basin (Figure 139), is a mature, light crude oil, showing few signs of oxidation or biodegradation. Peaks on the gas chromatography (GC) trace for this sample were readily identifiable and therefore the retention times were used to confirm the identity of the peaks in the other samples, in addition to the routine standard run with each batch.

Data for samples for which numerical data could be obtained are summarised in (Table 29), along with more limited data on the other samples for which an extract was obtained. The Maximum Carbon Number is the highest straight-chain alkane identifiable from the GC trace. These values, together with the Peak Carbon Number (the number of carbon atoms in the alkane present in the greatest concentration – not necessarily the greatest peak height), provide information on the origin of the sample. Very high carbon numbers (C>38) generally indicate derivation from higher plants, as the waterproofing waxes present in their leaves tend to be the highest molecular-weight organic matter (OM) preserved. The values also give insight into the degree of thermal maturity: as thermal maturity advances, highmolecular-weight compounds break down into compounds of lower molecular weight.

The Carbon Preference Index (CPI) is the concentration ratio of odd-numbered to even-numbered alkanes, and is also a thermal-maturity indicator. High CPI values (above 1.5) indicate relatively immature samples. Low CPI values may mean higher maturity, or a lack of higher n-alkanes stemming from a terrestrial source. The latter explanation is favoured for the samples in (Table 29).

Pristane and phytane are isoprenoid alkanes formed by breakdown of the side chains in the chlorophyll molecule. The remaining ring structures become porphyrins which can form organo-metallic complexes (e.g. nickel or vanado-porphyrins), which can be used to group hydrocarbons according to their source. The pristane/phytane ratio (Pri/Phy) and the ratios between pristane and phytane and their parent alkanes, (Pri/C17) and (Phy/C18), are useful provenance and maturity indicators. Pri/Phy ratios less than 1.5 are generally considered to indicate marine-shalecarbonate precursors, while ratios greater than 2.0 indicate terrestrial origins. High Pri/C17 and Phy/C18 ratios (>1.0) generally indicate immaturity.

With two exceptions, all the samples are thermally mature. The data are consistent with migration of hydrocarbons from the Westphalian sequence beneath or adjacent to the basin. The two exceptions, from Haughmond Hill and to a lesser extent Burton Point, yielded conflicting data which may be due partly to weathering and biodegradation effects.

Half the samples show Pri/Phy between 1 and 2 and, in view of the poor state of preservation of most of them, it would be unwise to classify them too rigidly. However, the extreme Pri/Phy values are interesting. The three highest values (>2.0) are from samples on the southern and eastern margins of the basin, while the three lowest (with ratios <0.4) are from the north-west of the basin and the East Irish Sea Basin. The data suggest an original marine OM source for hydrocarbons in the north-west, but a predominantly terrestrial OM source in the south and east. Study of the chromatograms for some of the samples (Figure 106) suggests the presence of two or more separate phases within the OM. The lighter and better-preserved fraction was originally thought to be a possible present-day contaminant, but the peak patterns show few similarities to a refined diesel-type oil, which would be the most likely modern source. Therefore this lighter oil is thought to belong to a second oil-generation event.

This hypothesis is supported by variations in the Ni/V data. The trace-metal contents of crude oils, in particular the Ni/V ratio in porphyrin complexes, have long been used to correlate different wells and producing fields (e.g. Hyden, 1961; Lewan, 1984; Witherspoon and Nagashima, 1957), although considerable disagreement exists over the factors influencing changes in this ratio. Preliminary examination of Ni/V ratios of samples from in and near the Cheshire Basin suggest two separate sources, or generation events, for the Cheshire Basin OM (Bradley et al., 1992; Hughes and Bradley, 1993; Bradley et al., 1994).

The differences in the organic geochemistry from north-west to south-east mirror patterns in the δ¹³C data for diagenetic carbonates (p.148), which are interpreted in terms of an increase in carbon of organic origin in the dissolved carbonate in the parent fluids. This component was derived from organic matter subject to a higher degree of thermal maturation probably within Carboniferous strata. The organic data from the basin suggest that the type of maturing organic matter in the basement may also have been important, with that from a marine source leading to a greater influx of reducing fluids in the north-west of the basin. It is known that organic carbon of marine phytoplanktonic origin is much less refractory than terrestrial organic matter (e.g. Waples, 1983) and would therefore be more prone to oxidation and hence the creation of reducing fluids.

Discussion and conclusions

Detailed petrographic and stable-isotopic investigation of the SSG and MMG in the Cheshire Basin have revealed a complex sequence of diagenetic modifications. Three stages of diagenesis are recognised: changes attributed to near-surface synsedimentary processes (eodiagenesis), processes associated with significant burial (mesodiagenesis), and late-stage alteration caused by relatively recent invasion of the sedimentary pile by meteoric groundwaters (telodiagenesis). The principal diagenetic features, and their timing within this relational framework are summarised in (Table 21) (SSG + Tarporley Siltstone Formation) and (Table 23) (MMG).

Early diagenesis of the SSG in the Cheshire Basin was similar to that observed in other UK Triassic basins (Burley, 1984, 1987; Milodowski et al., 1986, 1987; Strong and Milodowski, 1987; Strong, 1993; Strong et al., 1994). It involved the infiltration of detrital clay into sandstones, the oxidative breakdown of unstable detrital ferromagnesian and Fe-Ti oxides in the basin with the released Fe and Ti precipitated as ferric oxide (now hematite) and titanium oxides (probably anatase) as films and coatings around detrital grains (DE1a) similar to modern-day desert varnishes. This was closely associated with the evaporation of capillary porewaters in near-surface soils and sediment to form nodular or microsphaeroidal pedogenic carbonates; i.e. dolocretes and to a lesser extent calcretes (DE1b). There is some evidence of possible zonation in the distribution of calcrete and dolocrete within the lower part of the SSG (Kinnerton Sandstone and Chester Pebble Beds formations). Calcrete formation appears to have been restricted to the south-eastern and western margins of the basin, whereas dolocrete development dominated in the main basinal and more distal northern margins. A similar zonation is observed in the Wessex Basin (Strong and Milodowski, 1987; Holloway et al., 1989). This may be compared with the eodiagenetic facies zonation of modern large arid inland drainage basins in Australia (Arakel and McConchie, 1982), where calcretes are developed at basin margins and are replaced by dolocrete (then gyperete – not seen in the Cheshire Basin), playas and salinas towards the basin centres as conditions become more saline.

Similar early diagenetic processes affected sandy facies in the MMG (principally in the Tarporley Siltstone Formation). However, as a result of major development of sabkha environments in shallow marine-lacustrine or lagoonal environments, significant amounts of nodular anhydrite grew within the sediments, as well as crusts on subaerially exposed mudflats, as a result of near-surface evaporation of saline groundwaters. Thick deposits of halite were deposited by repeated evaporation of shallow ponded bodies of water. Sulphur-isotope data for the lower MMG (Tarporley Siltstone Formation to Northwich Halite Formation) and the later Wilkesley Halite Formation indicate that primary and eodiagenetic evaporites were derived largely from brines of Triassic marine origin. However, evaporites formed during the middle MMG (Wych Mudstone Formation) and towards the end of the MMG (Brooks Mill Mudstone Formation) were probably influenced significantly by isotopically lighter brines of continental origin, perhaps deriving a large part of their sulphate by re-working of pre-existing Permian sources (characteristically depleted in ³⁴S).

Reduction of Fe³+ and sulphate occurred in the early pore waters, associated with bacterial decomposition of localised organic matter. This led to the removal of detrital and earlier eodiagenetic ferric oxides, and to the precipitation of minor authigenic framboidal pyrite. The effect was to produce characteristic reduced green beds and reduction spots, common in the MMG and seen locally in the finer-grained parts of fluvial cycles in the SSG.

Authigenic smectitic and corrensitic pore-lining clays precipitated in some sandstones in the Helsby Sandstone and Tarporley Siltstone formations, and in part replaced detrital clays and micas. Corrensite was identified in most MMG formations but its diagenetic fabrics were not observed. However, corrensite formation is a common feature in modern alkaline-hypersaline lacustrine sediments (Velde, 1977; Khoury, et al., 1982; Jones and Weir, 1983). It seems likely, therefore, that the corrensite observed in the MMG is similarly of eodiagenetic origin, since the depositional environments of the MMG were likely to have been hypersaline. These diagenetic modifications are considered to have occurred during later stages of eodiagenesis (DE2).

During the early stages of burial (late eodiagenetic to early mesodiagenetic), many of the fine-grained eodiagenetic carbonate cements (calcite, dolomite) and evaporitic sediments (halite, anhydrite) in the MMG and the SSG were recrystallised and remobilised (DE3). Neoformation and overgrowth of idiomorphic dolomites and the formation of authigenic quartz and feldspar cements and overgrowths (DE3–DE4) are common features in both the SSG and the MMG. In the MMG, evidence can be found for early diagenetic dissolution of halite (DE3) prior to the development of early mesodiagenetic dolomite (DE3). Authigenic dolomite cements became increasingly ferroan as burial progressed and pore waters became more reducing.

Later mesodiagenetic fabrics show that anhydrite was remobilised and redistributed as a mesodiagenetic (syncompactional) pore-filling cement (DE4) in siltstones and dolomicrites in the MMG. Many of the porous sandstones of the SSG have fabrics indicating that the porosity is largely secondary, resulting from the dissolution of an early mesodiagenetic cement (i.e. oversized pores and preserved uncompacted fabrics). The well-preserved carbonate, baryte and sulphide infill of cross-cutting veins related to faulting do not penetrate a significant distance into the now-porous, and often very friable, wall-rocks. This indicates clearly that these rocks were lithified and well-cemented prior to major tectonic fracturing. The preservation of carbonate fabrics within the rocks indicates that the removed cement was most probably an evaporite mineral. Anhydrite-cemented sandstones (Malpas Sandstone) similar to the Helsby Sandstone Formation are preserved at the base of the MMG in the Wilkesley Borehole, lending support to the conclusion that much of the SSG was cemented by anhydrite during mesodiagenesis (DE5). There is circumstantial evidence to suggest that halite cements may also have been significant in some sandstones, at least in the Wilkesley Borehole, where now porous rock is cross-cut by later vein dolomite and anhydrite (DE6) without any penetration of the vein mineralisation into the adjacent open porosity. The evaporite cements have been removed after mineralisation as a result of late mesodiagenetic dissolution (DE7).

Many of the diagenetic features are common to both MMG and SSG rocks. However, eodiagenesis and early mesodiagenesis within the MMG and SSG were obviously diachronous since they are contemporary with sedimentation and shallow burial. However, it seems very likely that later mesodiagenesis in both formations was linked and related to brine expulsion along permeable paths (porous sandstones and tectonic fractures) during burial dewatering of the MMG.

During late mesodiagenesis (DE6) in the MMG, Cu, Pb, Co, As, Zn, Ag (± trace amounts of Au) formed minor amounts of authigenic sulphide minerals (e.g. chalcopyrite, chalcocite, galena, Co-Ni-As-sulphide and rare metallic Ag-Au-Cu). Copper is variably enriched in the reduced beds of the MMG (with values in excess of 3000 ppm in some analysed samples) and other metals are enriched to a lesser degree (e.g. up to 120 ppm Co). It is suggested that diagenetic brines in the MMG mobilised these metals in solution (by leaching them from the oxidised rocks where they may have been associated with the ferric oxides) and were thus a potential source of mineralising fluids – as is also seen from the geochemical compositions of fluid inclusions trapped within MMG halites. The quantities of metals involved are not great, possibly because of the limited supply of reduced sulphur, but the presence of metal sulphides in the reduced laminae of the MMG illustrates, in microcosm, the processes by which migrating Cu-bearing saline solutions deposit their load of Cu (and sometimes other metals) when they encounter reducing conditions.

Similar authigenic sulphide paragenetic assemblages are observed in the SSG. They developed as significant accumulations associated with major block faults around the basin margins – principally in the Alderley, Clive–Grinshill and Bickerton areas in the east, south and west of the basin. Fracturing clearly took place after the rocks were tightly cemented, probably by anhydrite ± halite which have since been lost, and mineralising solutions were introduced along faults and related fractures. Host-rock fabrics in West Mine (Alderley Edge) and Clive indicate that mineralisation of the adjacent wall-rock sandstones must have occurred largely by progressive corrosion, dissolution and replacement of the preexisting tight matrix cement, rather than by simple pore-filling of a porous sandstone. In the main Cu-Pb-Ba-mineralised areas DE6 paragenesis was complex and episodic, and mineralisation followed repeated episodes of faulting and fracture reactivation. Initial faulting was accompanied by silicification of fault gouges and localised quartz and chalcedony cementation in the immediately adjacent wall-rocks, accompanied by trace amounts of sulphide and baryte. Early microcolloform and framboidal textures indicate low temperatures consistent with fluid-inclusion data from associated baryte and calcite. This was followed by further fracturing and dolomite-ankerite mineralisation (DE6b). Major sulphide (primarily chalcopyrite, chalcocite, galena) ± baryte mineralisation (DE6d) came after another episode of faulting and fracturing. Later repeated fracturing (at least three generations) was associated with calcite and baryte (DE6e). Anhydrite veining (DE6c) post-dating DE6b dolomiteankerite is observed only in the deeper boreholes near the basin centre (e.g. Wilkesley and Crewe boreholes).

Fluid-inclusion data for baryte and calcite from the Alderley Edge and Clive Mines confirm the low-temperature (<70–80°C) nature of the mineralising fluids. Fluid salinities average 0.5 to 9.0 wt% NaCl equivalents, except in the Clive area where there is an overprint by hotter, more saline fluids which may be linked to hydrothermal activity in the West Shropshire orefield. No evidence could be found for the migration of liquid hydrocarbons into the SSG before or during base-metal mineralisation. Though the presence of gaseous hydrocarbons in the SSG prior to mineralisation cannot be discounted, a model for sulphide deposition involving light hydrocarbon gases remains unproven.

Chemical analysis of brine inclusions in halite from the Wilkesley Halite Formation demonstrates that the early diagenetic evaporite fluids were already significantly enriched in Cu, Pb and Ba prior to their migration into the SSG. Subsequent leaching of metals from the SSG would further enhance their mineralising potential.

Sulphur isotope data provide strong evidence that the sulphur in DE6 sulphide mineralisation and associated baryte mineralisation was derived from the overlying MMG. However, petrographic evidence alluded to above indicates that this sulphur was already in place, probably as anhydrite cementing the SSG sandstones, and therefore was not brought into the SSG at the same time as the metals. The δ³⁴S signatures of the baryte represent the signature of a precursor DE5 anhydrite. This implies that anhydrite-saturated brines must have penetrated from the MMG into the SSG during DE5 (either across major faults, or under gravity flow, or as a result of compactional dewatering of the MMG – displacing more dilute SSG formation waters). As temperatures increased during progressive burial, anhydrite would have become supersaturated and be precipitated as pore-filling cements. Later DE6 brines from the MMG may have been low-sulphate NaCl brines, in contrast to the earlier DE5 anhydrite-saturated fluids. The later brines would have been capable of transporting Ba and other metals in solution. Petrographic observations in the MMG show that anhydrite is corroded by NaCl brines during halite remobilisation (e.g. in Meadowbank Mine, Winsford). The late DE6 NaCl-type brines would have dissolved the anhydrite cements on entering the SSG, and immediately any Ba in solution would have been deposited as baryte. This is consistent with the fracture-controlled distribution of baryte in the SSG, and with the evidence for several different fluids having been produced in the MMG at different times.

Differences in the sulphur isotopic signature of baryte between different deposits across the basin can be accounted for by differences in the stratigraphical level of the MMG from which the DE5 anhydrite-mineralising brines were originally sourced. In the south and west of the basin, DE5 fluids have affinities suggesting derivation largely from the lower part of the MMG (i.e. below the Byley Mudstone Formation). In contrast, isotopic signatures in the Alderley Horst area suggest that DE5 brines were derived from a slightly higher stratigraphical level in the MMG (i.e. largely above the Tarporley Siltstone Formation and below the Brooks Mill Mudstone Formation).

The mechanism for metal-sulphide precipitation in the SSG is more problematic since it requires reduction of the sulphate. Biogenic sulphide reduction could have been partly responsible, but requires an organic carbon source to be introduced into these organic-poor rocks. It is unlikely that this component was introduced in the late DE6 metalliferous brines derived from the MMG because the reducing nature of the fluid would have severely limited the transport of Cu (Chapter 6). It is much more likely that mixing with a deeper-sourced sulphide-bearing fluid caused sulphides to precipitate. Carbon and oxygen stable-isotope data for DE6 carbonates provide strong evidence for the progressively greater influence of a deeper-sourced fluid derived from an organically-matured source (probably Carboniferous basement) towards the north-west of the basin. It is suggested that north of the Mersey, and in the Chester area, this fluid dominated, making conditions too reducing for Cu mobilisation and consequently any mineralisation was limited to pyrite (± trace galena and sphalerite) and minor baryte. This dominance may be related to geographical differences in the nature of the basement underlying the basin, and/or to fluids interacting with different parts of the Carboniferous sequence due to variations in the nature of the subsurface plumbing system. The organic geochemistry of hydrocarbons in and around the basin suggests a marine organic matter source in the north-west of the basin and, in contrast, a terrestrial organic matter source in the south-east. Comparison of the distribution of DE6 authigenic Cu sulphides with the variation in δ¹³C of DE6 carbonates indicates that areas with DE6 carbonates with δ¹³C_PDB < –8‰ were dominated by the deeper-sourced reducing fluid which inhibited Cu mobilisation and mineralisation. The –8‰ δ¹³C_PDB contour for DE6 carbonates appears to provide a useful tool for delineating areas of potential Cu mineralisation, where values exceed –8‰, from zones of dominantly Fe-sulphides, where δ¹³C_PDB is less than –8‰.

Chapter 6 Hydrogeology and hydrogeochemistry

R Metcalfe, I N Gale, D F Ball, J J W Higgo, P J Hooker, D J Noy, C A Rochelle and D Savage

Introduction

The sedimentary sequence in the Cheshire Basin is chemically and hydrogeologically heterogeneous, and this heterogeneity exerted a control on the metalliferous mineralisation of the basin. In order to understand the diagenetic evolution of the basin fill and the genesis of the ore deposits, it is necessary to evaluate the chemical variability and chemical evolution of the formation waters.

The first sections of this chapter give an account of the evolution of the palaeo-groundwater flow, deduced from palaeogeographical reconstructions of the Cheshire Basin area, at selected stages of its development up to the present day.

The rest of the chapter is devoted to a study of fluid–rock interactions and groundwater hydrochemistry. Literature reviews and generic models were produced in order to shed light on the various processes which control the modern and palaeo-hydrochemistry of the formation waters of the basin. The results of this work are reported by Higgo et al. (1993), Metcalfe and Rochelle (1994), Rochelle et al. (1993), Metcalfe et al. (1993) and Savage (1993), and a synthesis is given in Metcalfe et al. (1994b). The present account gives a summary of these works and assesses their implications for an understanding of the evolution of the Cheshire Basin. The approach taken was to evaluate the available literature on red-bed diagenesis and red-bed-hosted ore formation and to use the results as a pointer to uncertainties in current understanding. These uncertainties were then investigated further by means of theoretical modelling. The chemistry of modern groundwaters in the Cheshire Basin was evaluated in order to place constraints upon these models. Finally, new observations on the petrography of the sediments and their associated ore deposits, described elsewhere in this report, are discussed in the light of the theoretical modelling results.

Palaeo-groundwater flow

Introduction

Fluid flow plays a key role in the development of sedimentary basins, particularly in the transport and deposition of solutes, the emplacement of ore bodies and the accumulation of hydrocarbons. This role, and the processes controlling fluid flow during various stages of basin development, have been discussed by Downing and Penn (1991).

Sedimentary basins were classified by Cousteau et al. (1975) on the basis of hydrodynamic conditions. During sedimentation, compaction-induced lateral movement of water towards the margins of the basin is the dominant process. Following uplift, leading to erosion, gravity-induced flow of meteoric water becomes more important. The major driving force in the latter situation is the force of gravity, and the fluid potential field reflects the shape of the water table, which itself is a subdued impression of the land surface (Toth, 1963). In the gravity systems, the three-dimensional pattern of the fluid potential is controlled by upland divides (which are recharge areas) and lowland, coastal or submarine groundwater sinks (or discharge areas). In both compaction-induced and gravity-induced systems groundwater flow is influenced by permeability variations in the rock sequences.

In addition to compaction and gravity-induced flow, other processes may become significant at certain stages of basin development, including thermally induced and density-induced flow. This can result from either dissolved ion concentrations or thermal differences or a combination of the two. In regional flow systems where both shallow, fresh groundwater and deep, saline groundwater are found, density-induced flow can dominate the topographical, or gravity-induced flow (Senger, 1993).

There is commonly hydraulic continuity throughout the basin sediments, and water moves through deposits of widely differing hydraulic properties, including those normally regarded as aquicludes. Large volumes of water can discharge from extensive confined aquifers by cross-formational flow through overlying low-permeability beds.

Preferred flow routes are considered to be important avenues of groundwater flow, whether through faults, fractures, solution-fissures, or more permeable layers and vertical zones in sandstones and mudstones. Even compacted shales can contain microfractures, allowing the passage of fluids. After an initial decrease in permeability caused by compaction, the permeability of shales increases and, as they become harder, microfractures form in response to tectonic stresses. Normal faults, in particular, provide preferred flow routes, especially in compacted sedimentary rocks (Neglia, 1979).

Fluid-potential patterns are recognised from potentiometric surface maps based on hydraulic heads calculated from fluid pressures measured during a variety of tests, including drill-stem and production tests. However, information from well-tests relates only to present-day conditions, and movement of fluids in the geological past was related to physical conditions in the basin at the time. Therefore, to assess the probable hydrodynamic conditions at different times, it is necessary to interpret the palaeogeography and identify physical features that probably controlled groundwater flow.

Interpretation of paleo-groundwater flow from palaeogeographical reconstruction

The palaeogeography of the United Kingdom has been compiled into the Atlas of palaeogeography and lithofacies (Cope et al., 1992). This comprehensive publication was used to follow the development of the Cheshire Basin from Permian times to the present day. At each stage of development, the hydrogeological regime of the basin is deduced from the relative positions of land masses, seas, sabkhas and depositional or erosional environments. Although crude, this interpretation will give the likely history of groundwater quality and, in some instances, the likely direction of movement.

Permian

Early Permian deposition in the Cheshire Basin area was characterised by aeolian sands and sand dunes in a desert environment. The basin lay between higher ground to the north-east and south-west, while to the north-west lay an area of sabkhas and hypersaline brines. During the late Permian (c.255 Ma), the Zechstein Sea encroached from the east and the Bakevellia Sea extended into Lancashire as far as Manchester but did not cover the Cheshire Basin. These evaporative seas contained brines, and adjacent groundwater bodies would probably have been saline. In the Cheshire Basin alluvial deposits, largely arenaceous, continued to be laid down and probably contained a fresh groundwater body which drained to the north-east towards an area of hyper-saline lakes and sabkhas. Towards the end of the period, alluvial deposits dominated the basin and fresh groundwater systems probably prevailed.

Triassic

In the early Triassic (c.246 Ma) sandy conglomeratic deposits continued to be laid down, the sediments being derived from the south: By the mid-Triassic (c.242 Ma), the system drained into a shallow hypersaline sea to the north of the Wirral. This sea gradually encroached over the Cheshire Basin (c.235 Ma), surrounded by aeolian and alluvial deposition. The sediment beneath the shallow sea would have been penetrated by the dense brines which would have displaced any fresh groundwater they may have contained. Additionally, compaction of the sediment pile would have expelled water towards the periphery of the basin at sea level.

In late Triassic times, the Cheshire Basin area was again an area of alluvial deposition and a shallow fresh groundwater system was probably established, while saline water probably persisted at depth. The depth of the system would have depended on the position of sea level, but the density of the brines in the sediments may have prevented them being flushed out. During the latest Triassic (Rhaetian, c.206 Ma), shallow seas covered the area and saline water would probably have displaced any fresh groundwater as the sea encroached over the land.

Jurassic

Jurassic sediments were deposited and eroded sequentially in the Cheshire Basin throughout the period, though few are now preserved. Generally, when the area was inundated and deposition was occurring, the groundwater in the underlying permeable sediments would probably have been replaced by saline water unless the sediments already contained denser brines. During periods of uplift and erosion, fresh groundwater would percolate to depths related to the ground elevation above sea level and the density of the brines in the aquifers.

During the early Jurassic (c.204 to c.184 Ma) the Cheshire Basin area was inundated, and the surrounding area contained several land masses. In mid-Jurassic times (c.179 Ma) the Welsh and Pennine land masses coalesced around the basin, which emerged as a low-lying alluvial basin (c.177 Ma) which extended to the southern Irish Sea. The sea varied in extent throughout the mid-Jurassic but did not inundate the Cheshire Basin until c.158 Ma. During the remainder of the mid-Jurassic and into the late Jurassic the area remained submerged but the sediments deposited there have been subsequently eroded. The Cheshire Basin was again emergent as part of the largely low-relief Anglo-Welsh land mass in late Jurassic times (c.139 to c.137 Ma).

Cretaceous

During the early Cretaceous (c.129 to c.109 Ma) the Cheshire Basin was an area of low relief; the Pennine land mass lay to the north-east and the Welsh land mass to the south-west. This would have been a period of erosion and a fresh groundwater flow system would have been established. Subsequently (c.97 Ma) the land was flooded from the south and by the end of the late Cretaceous (c.75 Ma) the Cheshire Basin was inundated, with a Welsh land mass possibly located to the south-west. Fresh groundwater in the Triassic sandstones would probably have been replaced by saline water.

Tertiary and Quaternary

Most of the British Isles was land throughout the Tertiary, although the coastline moved considerably during the period. Erosion and development of river systems and coastlines would have determined groundwater flow systems. In mid to late Oligocene times (30 to 24 Ma) a south-west-flowing river in the southern Irish Sea is postulated. The drainage of the Cheshire Basin could have formed a tributary to this system.

Sea-level fluctuations during the Quaternary would have resulted in coastlines moving and base drainage levels changing. These levels would determine the shape and depth of fresh groundwater flow systems and the extent to which saline groundwater was flushed from aquifers. During periods of relatively low sea level, it is likely that saline water in deeper parts of the basin would be flushed out, but a subsequent rise in sea level would only replace saline water along a coastal strip by saline intrusion. Glaciation would impact on the groundwater flow system by restricting recharge to aquifers covered by the ice sheet and where permafrost dominated. Additionally, the ice front would impact on the surface drainage systems. In the Cheshire Basin, the north-westerly drainage pattern was diverted to the south into what is now the Severn drainage basin.

Summary

The history of sedimentation and erosion in the Cheshire Basin has contributed to the present distribution of groundwater quality and flow regime. Groundwater in the sediments will have changed in quality and flow direction, dependent upon the dominant mechanism at the time. During periods of marine deposition, especially in evaporative' seas or in sabkha environments, saline water or brine is likely to replace any fresh water in submerged aquifers. When these areas are subsequently emergent and erosional conditions prevail, fresh groundwater will flush the terrestrial system to a depth related to topography, geology, precipitation and the density of saline water or brine already in the aquifer.

The palaeo-hydrogeological reconstruction suggests that the Cheshire Basin would have been filled with highly saline brines during Permian and Triassic times except during the later Triassic when a shallow fresh groundwater system may have been established. With alternating inundation and emergence during the latest Triassic and in Jurassic and Cretaceous times, shallow fresh groundwater systems would be periodically established and then replaced by encroaching saline water. The density of this water would, however, be less than that of the Triassic brines, some of which could be present at depth to this day.

Since the last major inundation of the Cheshire Basin in the Cretaceous, uplift and erosion has established shallow, fresh groundwater systems superimposed on deeper saline and brine groundwater systems, particularly where the arenaceous aquifers are unconfined. With removal of Cretaceous and Jurassic sediments, these systems would have evolved into the present system.

Present-day groundwater flow system

Groundwater levels and flow directions

On a local scale, the SSG consists of several aquifer units with varying rest water levels over short distances. On a regional scale, however, water-level contours can be drawn to give general trends in groundwater flow. These contours are shown in (Figure 108), (Figure 109), (Figure 110), (Figure 111), (Figure 112), (Figure 113), (Figure 114), (Figure 115), (Figure 116), (Figure 117), (Figure 118) (Figure 119) is shown." data-name="images/P1000361.jpg">(Figure 107), together with the regional flow directions which are largely controlled by the rivers draining the basin and the major groundwater abstraction schemes. In general, high groundwater levels are found in areas of recharge such as the Mid-Cheshire Ridge, although levels have been falling here as a result of large-scale groundwater abstraction. Natural seasonal fluctuations in groundwater levels are low, about 1 m, because of the aquifer's high specific yield of 6 to 14%.

Along the south-east margins of the SSG outcrop the potentiometric surface is up to 100 m above OD, though 40 to 50 m above OD is more common. Between Whitchurch and Chirk in the west, the contours of the potentiometric surface run east–west and slope northwards from 70 to 20 m above OD. Groundwater flow in the western part of the basin is influenced by the drainage pattern of the River Dee which is also the principal discharge area in this part of the basin. In the vicinity of the Dee, the natural potentiometric surface is less than 10 m above OD and it has been lowered to below sea level where over-abstraction has occurred, such as at Stanlow and Ellesmere Port. Lucey (1987) measured groundwater pressures in boreholes using packers and found that freshwater heads vary by less than 0.4 m over a 100 m depth interval. Near-surface groundwater flow directions may differ considerably on a local scale from the regional flows encountered at depth.

North of the Mersey, several large groundwater mounds exist which have led, along with considerable faulting, to complex patterns of groundwater flow. The River Mersey provides the main area of groundwater discharge, causing northerly groundwater flows beneath the southern Manch ester conurbation. The potentiometric surface falls uniformly from 100 m above OD at Prestbury to 20 m above OD in central Manchester.

The mudstones of the Manchester Marls, which separate the Collyhurst Sandstone from the Triassic aquifers, form a significant barrier to groundwater flow in the north of the basin as they dip southward beneath the Mersey (Brassington et al., 1992).

Rising groundwater levels recorded in the Liverpool area in recent years are due to reduced rates of abstraction: a rise of about 2 m between 1975 and 1985 was seen in several observation boreholes in the Liverpool area (Brassington and Rushton, 1987).

These groundwater flow regimes are the present-day active systems where the SSG is not overlain by the MMG. Where the aquifer is confined, groundwater is not ekploited and so there is very little information on which to interpret flow directions. Some information can be gleaned from the few deep boreholes that have been drilled for hydrocarbon exploration, but no deep groundwater samples have been collected and no pressure measurements have been made. Without these data it is not possible to confirm even the most basic estimates of flow directions at depth.

Groundwater recharge and residence times

Most recharge reaches the SSG via the thick cover of drift. Thick layers of low-permeability till are widespread, and it is estimated that only 2% of the potential recharge can infiltrate in these areas. In the west of the basin, the SSG is in contact with Lower Carboniferous Limestone through a series of en-échelon faults with throws up to 300 m locally. The contact with the limestones probably only provides a small proportion of the total recharge to the aquifer.

In the Liverpool area, rainfall is the main source of recharge, but there is also a significant contribution from leaking sewers and water mains. It is estimated that 66 Mld leaks from the mains in Liverpool (Brassington and Rushton, 1987). Induced recharge from the River Mersey could also contribute 12 Mld, but this figure is steadily falling owing to reduced rates of groundwater abstraction and the consequent rise in groundwater level (Brassington and Rushton, 1987).

Residence times are generally short because of the high frequency of fractures in the upper parts of the aquifer. An average residence time of 1300 years has been estimated for the upper 500 m of aquifer (Lucey, 1987). Below this level, residence times increase markedly, and the highly saline waters present at depth provide evidence of very slow-moving groundwater. These saline waters result from the solution of halite and some have chloride concentrations in excess of 100 000 mg/l.

Saline intrusion and saline groundwater

A long history of over-pumping from boreholes on both sides of the Mersey estuary, particularly at Liverpool and along the Manchester Ship Canal, has resulted in saline intrusion into the Permo-Triassic aquifer. Chloride concentrations of 6000 mg/l in these areas, combined with high sulphate, have led to the abandonment of many deep boreholes.

Naturally saline groundwater is also present at depth. Geophysical logs from NCB boreholes indicate high-salinity groundwater in the Collyhurst Sandstone in the north of the basin. Fresh groundwater to a depth of 200 m is underlain by a transition zone in which the electrical conductivity (EC) rises to between 40 and 50 mS/cm at 450 m. This salinity is found to the base of the aquifer at 850 m where it overlies the Carboniferous. Similar profiles are found in several boreholes; in some the salinity rapidly increases beneath the Manchester Marls, which restricts groundwater flow and has isolated the Collyhurst Sandstone aquifer from overlying sandstone units (Brassington et al., 1992). The surface of this body of saline water shows considerable topographic variations, which can be attributed to flushing by regional groundwater flow and upconing in areas of over-abstraction (Tellam et al., 1986).

Further evidence of saline groundwater at depth was obtained from a sample collected near Chester from the Coal Measures underlying the Triassic at a depth of about 500 m. The conductivity of the groundwater was about 50 mS/cm and the Total Dissolved Solids (TDS) was measured to be approximately 30 g/l.

A saline spring at Aldersey [SJ 4565 5652] has a recorded chloride concentration of up to 11 000 mg/l, which is approximately equivalent to a TDS of 20 g/l. This water is likely to be a mixture, resulting from the dilution of deep brine by meteoric water. The spring issues from Drift underlain by the Erbistock Beds of the Coal Measures, which form an inlier to the south of Chester. The driving force required to cause the saline water to rise to the surface may originate in the outcrop of Coal Measures to the west or from the density-driven system beneath the MMG to the east. The latter mechanism is discussed in the following sections, and is thought to be the most likely source of the saline water since the regional flow is from the east.

Regional groundwater flow

An interpretation of the present-day regional flow regime was made from the available data. Where an aquifer has been exploited there are ample data to form conceptual as well as numerical models as an aid to understanding the active flow in the aquifers. Where the SSG underlies the MMG, however, the aquifer has not been exploited and the only information available is from a few deep wells drilled for research or hydrocarbon exploration.

In view of this limited level of information, the main focus of attention was on the potential effects of brine densities on groundwater flow patterns in the basin. In particular, a comparison was made between the flows predicted under the assumption of purely freshwater flow and those obtained by including the effects of brine density.

Cross-section model

Groundwater flows in the Cheshire Basin have been modelled as Darcian porous-medium flow in a two-dimensional vertical section using the computer code SUTRA (Voss, 1984). This code calculates both the groundwater pressures and the transport of a solute within the groundwater. When the concentration of the solute is large the two equations are coupled through the dependence of the fluid density on the solute concentration, leading to a non-linear system of equations. These equations are solved by the finite element method, using a grid of isoparametric quadrilateral elements. Values for pressure and solute concentration are calculated for each node of this grid.

The geological section

A geological section across the Cheshire Basin on a north-west to south-east line through its centre (Figure 108), (Figure 109), (Figure 110), (Figure 111), (Figure 112), (Figure 113), (Figure 114), (Figure 115), (Figure 116), (Figure 117), (Figure 118) (Figure 119) is shown." data-name="images/P1000361.jpg">(Figure 107) and (Figure 108) was chosen in order to minimise the error involved in assuming a two-dimensional flow field. The flow in the basin is clearly complex and three-dimensional, but it is not possible with current codes and available computing facilities to simulate the coupled processes on this scale. The geology on the chosen section was simplified into a small number of units as shown in (Figure 108). The hydraulic conductivities and porosities adopted for each of these units are given in (Table 30). Little direct evidence for these values is available from in-situ measurements, so they can only be regarded as reasonable estimates. No attempt was made to take account of any heterogeneity in the units. This section was divided into a grid of 735 elements and 802 nodes for input to SUTRA.

Temperature and salinity data

Salinity data were interpreted from geophysical logs from four boreholes (Prees, Burford, Elworth and Knutsford; see (Figure 2) for locations located within the region of MMG cover (D K Buckley, written communication). Only Elworth is close to the line of the section chosen for the two-dimensional model. The availability of geophysical logs dictates the method used to calculate formation-water resistivity. In this instance, resistivity and porosity logs were available, the porosity being obtained directly from the neutron log or calculated from the density or sonic travel time log. Archie's law was used to transform porosity data to formation-water salinity at formation temperature.

The temperature distribution throughout the basin was evaluated from a much larger dataset (Gale et al., 1984) and was extrapolated along the selected section. (Table 31) gives the data for temperatures and pore water conductivities (mS/cm), calculated at both formation temperature and normalised to 25°C, together with estimates of TDS. Temperature gradients in the MMG are around 23°C/km and in the SSG are about 14°C/km.

The salinity (TDS) estimates show a considerable degree of variability, but many of the shallower results indicate values in the range 20–30 g/1 TDS whilst several of the deeper results indicate larger values of the order of 60–80 g/1 TDS. The two results from Elworth are very high despite being obtained at relatively shallow depths. Although these estimates have not been confirmed by direct measurement of salinity and temperature they are considered to represent the salinity found at depth.

Boundary conditions

The boundary conditions adopted for the freshwater calculations were 'no-flow' along the bottom and eastern edges of the model grid, and 'fixed head' (equal to elevation of the topographic surface) along the top surface. For the brine-dependant calculations, all inflowing surface water was assumed to have zero solute concentrations. The source of the brine was assumed to be the MMG, where extensive halite beds are found (Evans et al., 1993). Nodes here were given a fixed concentration corresponding to the brine density under consideration. Three source-zone densities were considered: 1050, 1100, and 1200 kg/m³. The freshwater calculations were run as a steady-state simulation while the brine-driven flows were followed in transient mode from an initial condition in which all water was fresh, except in the source zone of the MMG, until an apparent steady state was achieved. This varied depending upon the source-zone density in use, requiring a simulated time of about 1.5 Ma with the densest brine and about 10 Ma with the least-dense brine. The variation of density between the brines and the freshwater is sufficient to make the equations strongly nonlinear and it was found necessary to adopt large values for the hydrodynamic dispersivity (2 km) in order to ensure convergence of the solution throughout the simulations.

Model results

The freshwater flow field is represented by a plot of logarithmically scaled flow vectors in (Figure 109) and by a selection of particle pathlines in (Figure 110). The flow is topographically driven through the basin from the high ground in the south-east towards the Mersey estuary in the north-west. In addition, the MMG forms a confining layer to the Triassic sandstone aquifers, which are under artesian conditions in this calculation. Flow rates in the sandstone layers are typically of the order of 5.0e⁻¹⁰ m/s from east to west, while those in the MMG are of the order of 3.0e⁻¹¹ m/s.

The second calculation that was performed assumed that saturated brines with a density of 1200 kg/m³ form a continuous band in the lower part of the MMG. (Figure 111) shows the contours of density obtained after the system has evolved to a steady state, while (Figure 112) and (Figure 113) display the fluid-flow pattern as flow vectors and pathlines respectively. It is seen that the flow is now generally downwards in the central part of the basin under the influence of the dense brines originating in the MMG, reversing the artesian conditions suggested by the freshwater calculation. The deep basinal brines now circulate upwards at the eastern and western margins, with strong mixing vortices associated with the boundary faults. Flow rates in the sandstone layers are typically of the order of 7.0e⁻¹¹ m/s, while those in the MMG are of the order of 5.0e⁻¹¹ m/s.

It will be noted from (Figure 111) that the implication of this calculation is that much of the basin should be filled with brines at or near to halite saturation. Despite the limitations on the values of salinity calculated on p.180, it seems unlikely that this is in fact the case. Of a number of possible sources of error, it was decided to consider the sensitivity of the model to the density of the brine source. Whilst it is clear that water in close contact with a halite layer should rapidly attain a saturation density, it may be that, averaged over the area of an element of the calculational grid used here, a lower effective source density would be more appropriate.

The brine source density was accordingly reduced to 1100 kg/m³ and 1050 kg/m³ for two further calculations, both running in transient mode as before until an apparent steady state was obtained. The results for the 1100 kg/m³ case are shown in (Figure 114), (Figure 115), and (Figure 116). These show very similar patterns to those found for the previous case but with the lower peak brine density filling much of the basin. Flow rates in the MMG are reduced to about 2.0e⁻¹¹ m/s.

The results for the 1050 kg/m³ source density case are given in (Figure 117), (Figure 118), and (Figure 119), and show some significant differences from the earlier calculations. The mixing vortex on the eastern margin has now expanded to cover about half the basin, and the upper part of it introduces lower-density fluid to much of the upper SSG. The densest fluid is now restricted to a descending plume whose location may be determined by the shape of the lower boundary of the MMG. The groundwater flow in the MMG is no longer uniformly up or down but may vary from place to place, and the flow rates are reduced to the order of 5.0e⁻¹² m/s.

Discussion

These calculations illustrate that the distribution of brine sources can have a marked effect on the circulation of groundwaters in a basinal environment. In the central part of the Cheshire Basin, where halite beds occur in the near-surface layers, the brines may effectively prevent the topographically driven through-flow of fresh water, giving rise to a pattern of mixing cells with regions on the basin margins where deep basinal fluids are upwelling and mixing with the fresh water. This provides a possible mechanism for the formation of mineral deposits near such marginal faults. Geochemical studies by Metcalfe et al. (1993), outlined in this chapter (pp.195–200), suggest that several features of the mineralisation found here could be explained by the mixing of the basinal water with deeper reducing fluids introduced up basin-margin faults. Diagenetic studies (Chapter 5) also indicate the importance of brines derived from the dissolution of evaporites in the formation of mineral deposits.

These calculations do not include the effects of temperature variation. The deeper regions of the basin are currently believed to be at about 70°C (p.83), which might give rise to a density reduction of some 4–5% – much less than the 20% difference in density between fresh water and the densest brines. However, the lower density and the lower viscosity at depth would be likely to increase the rate of circulation. With less dense brines, which appear to be more likely from the limited data available, it would be expected that thermal perturbations of the flow fields may be significant.

A second important factor that was not taken into account in the calculations is the three-dimensionality of the true flow field. The generally oval shape of the basin means that the ratio of the area of recharge around the rim to the area of the brine source is greater than the two-dimensional model allows. It seems reasonable to suggest that this would enhance the creation of large mixing cells such as that seen in the 1050 kg/m³ example and that these might therefore be found at higher source densities than can be obtained with the two-dimensional calculations.

Further factors which would affect the flow fields in the basin to a greater or lesser extent include heterogeneity of the formations, extensive faulting, and possibly osmotic effects in the mudstone formations. Also, it has been assumed that the flows are essentially in steady state but it has been noted that the transients in the brine-driven flow calculations may take of the order of 10 Ma to settle. This is a long duration in comparison with events such as climate change which might affect recharge conditions. Hence a steady state might never be attained. Transients of a similar duration may be found in very-low-permeability clay and shale formations following, for example, external loading events. The inclusion of these features is beyond the scope of the current study, which has been aimed primarily at understanding the potential effects of dense brines on flow.

Conceptual model of regional flow

The present-day groundwater flow regime in the Cheshire Basin can be divided into two parts. First, the basin margins where flow is dominated by topographic driving forces with recharge occurring in areas of high ground and discharge from lower areas. Second, the central basin, covered by the MMG, where flow is strongly influenced by the density variations arising from halite dissolution and mixing with fresh water near the margins. It seems likely that the flow field is broken into a number of large mixing cells which cause the salinity to vary considerably from place to place within the basin.

Downing et al. (1987) suggested that regional flow in the Cheshire Basin would tend to follow peripheral routes, where the permeability is enhanced by fractures. Lateral flow is also suggested by the low heat flow measured, which implies that rising groundwater flow is not a feature. This hypothesis is supported by the results of the current modelling, where density-induced, downward groundwater flow would result in low heat flow. However, the water must be rising in some areas. Failure to detect these areas may be because of the small number of data points or because they are masked by the shallow, freshwater groundwater flow systems.

Review of previous studies of red-bed diagenesis and mineralisation

Introduction

This section is a summary of a review by Metcalfe et al. (1994a) of previous studies of red-bed sequences, to provide a framework for interpreting the hydrogeochemistry of the Cheshire Basin. Red beds play a fundamental part in the genesis and spatial distribution of natural, resources, being important hosts of hydrocarbon accumulations and deposits of metalliferous minerals worldwide. Copper mineralisation is perhaps the most economically important type of ore deposit in red beds, but other heavy metals also occur, including important deposits of U, Pb, Zn and Ag.

The regional distribution, detailed geometry and chemical composition of these resources are functions of a complex interplay between fluid flow paths and the chemical evolution of the fluid phase and the host red bed. In order to understand how this interplay controls the distribution of economic resources, it is necessary to understand the relationship between fluid chemistry and red-bed diagenesis.

A great deal has been published on the mineralogical and petrological aspects of this problem (e.g. Walker, 1967, 1974, 1976, 1989; Waugh, 1978; Turner, 1980; Holmes et al., 1983; Burley, 1984; Zielinski et al., 1983; Bath et al., 1987; Flint, 1987) and the associated ore deposits have also received considerable attention (e.g. Turner, 1980; Nash et al., 1981; Ixer and Vaughan, 1982; Merino et al., 1986; Sverjensky, 1984, 1987, 1989; Kirkham, 1989; Walker, 1989; Bechtel and Puttmann, 1991; Sawlowicz, 1991). However, with some notable exceptions (Rose, 1976, 1989; Sverjensky, 1984, 1987, 1989) previous studies have tended to concentrate on the evolution of red-bed mineral assemblages during diagenesis, and to date relatively little attention has been paid to the chemical evolution of the fluid phase or the theoretical basis for modelling red-bed diagenesis. In the present study an attempt is made to fill this gap by providing an overview of fluid evolution during red-bed diagenesis.

Diagenesis

Early diagenesis (eodiagenesis)

By using modern red beds as an analogue for ancient red-bed sequences, a number of studies have shown that the early diagenetic waters are generally alkaline and oxidising and cause in-situ weathering and dissolution of unstable ferromagnesian detrital minerals (pyroxenes, amphiboles, magnetite, biotite and olivine) (Walker, 1967, 1976, 1989; Walker and Honea, 1969; Walker and Runnells, 1984; Walker et al., 1978; Zielinski et al., 1983, 1986; Flint, 1987). With progressive diagenesis, mineral dissolution releases K, Fe, Mg, Al and Si into the pore waters and eventually causes chlorite and smectite to replace ferromagnesian minerals, and illite-smectite to replace feldspar.

Features of early diagenetic conditions include the formation of intergranular and replacive carbonates, and often extensive calcretes (Bath et al., 1987; Strong and Milodowski, 1987). The oxidising groundwater conditions mean that carbonates are characteristically non-ferroan, since iron is in its ferric state and is not easily incorporated into the carbonate mineral structure.

Rarely, zeolites including clinoptilolite, erionite and analcime may occur when groundwaters are oxidising and at near-neutral pH, or when alkaline brines are evaporated in closed basins (Surdam and Sheppard, 1978; Hartley et al,, 1991). Zeolite formation appears to be controlled partly by the presence of suitable detritus, such as volcanic grains of suitable composition.

In modern red beds the interdigitation of alluvium with fanglomerates and sabkha evaporite deposits is common. This leads to a gross chemical heterogeneity of the deposit, which in turn leads to the development of a number of different formation-water types (Figure 120). For example, calcium-carbonate waters may develop where meteoric waters weather silicate and carbonate minerals, but calcium sulphate and sodium chloride waters may evolve where there is dissolution of evaporite minerals. In general, halite is the dominant constituent of both marine and non-marine evaporites, and gypsum also occurs in these environments. Evaporating sea water never becomes bicarbonate-rich because: (1) the dominant anions in sea water are Cl^– (c.19 400 mg/l) and SO₄²⁻(c.2710 mg/l); and (2) the concentrations of Ca (412 mg/l) and Mg (1290 mg/l) are much greater than the concentration of HCO₃ (c.120 mg/l). Carbonate minerals are thus only a minor feature of marine evaporites. Continental evaporites, in contrast, often precipitate from bicarbonate-dominated waters, leading to the precipitation of a wide range of carbonate minerals including trona (NaHCO₃.Na₂CO₃.2H₂O) and burkeite (Na₂CO₃.2Na₂SO₄) (Eugster, 1980).

The redox conditions of diagenetic waters are often expressed in terms of redox potentials (Eh) or oxygen fugacities (fO₂). However, natural waters rarely approach electrochemical redox equilibrium, so Eh values are difficult to interpret in terms of overall system redox states. Additionally, fO₂ values do not always have a clear physical meaning, and although in groundwaters Eh values can often be measured, fO₂ values must generally be calculated from analytical data. Furthermore, fO₂ and Eh values do not show a simple correspondence and, for a given aqueous system, both parameters will change with temperature, making comparisons between redox conditions at different temperatures imprecise. In spite of these problems, both Eh and fO₂ are of practical value if used as semi-quantitative indices of redox, and enable general comparisons to be made between redox conditions in different environments.

Eodiagenetic waters include some of the most oxidising red-bed formation waters; organic-free waters at 25°C have Eh values up to approximately +500 mV, due to buffering by the atmosphere, with a redox state represented by logfO₂,bars = –0.7. However, it is likely that eodiagenetic waters originating in density-stratified saline lakes may become locally very reducing, with Eh approximately –400 mV to –500 mV, due to the trapping of organic matter in bottom brines (Sonnenfeld, 1984). The pH conditions are likely to be quite variable, generally lying in the range 6–10, which is typical of near-surface waters (Stumm and Morgan, 1981), but possibly exceeding 10 in continental evaporating lakes (e.g. Sonnenfeld, 1984; Darragi and Tardy, 1987). In such continental settings, HCO₃⁻ is often the most abundant anion, and evaporation may lead to the precipitation of carbonates and smectites, thereby reducing the Ca and Mg content of waters to the point where they are dominated by Na, K and HCO₃⁻. When CO₂ is independently buffered, for example by the atmosphere, then the pH is governed by:

HCO3⁻+ H⁺ = CO₂(g) + H₂O     (1)

fCO₂ [H₂O] / [HCO₃⁻[H⁺]) = Constant

This means that pH may reach >10 because as [HCO₃⁻] increases during evaporation [H⁺] must decrease.

In waters of marine origin, and in the absence of microbial activity, pH conditions generally decrease as ionic strengths increase during evaporation, owing to changes in the activity coefficients of the carbonate species which buffer pH (Krumgalz, 1980). For example, during atmosphere-equilibrated sea-water evaporation at 25°C, the pH falls from 8.2 to 6.7 at the point when halite saturates. However, microbial processes such as sulphate reduction may lead to an increase in pH to 8–9 (Sonnenfeld, 1984).

Burial diagenesis (mesodiagenesis)

As a red bed undergoes burial diagenesis (mesodiagenesis), formation waters become progressively more reducing and of generally lower pH. Early-diagenetic smectite transforms to illite, leading to the liberation of aqueous silica, Fe and Mg, according to reactions such as:

1.84Ca_{0 025}Na_0.1K_{0 2}Mg1.15Fe²⁺_{0 5}Fe³⁺_0.2Al³⁺_1.25Si^3.5O₁₀(OH)₂ + 0.232K⁺ Smectite

+ 6.72H⁺ = K_0.6Mg_0.25Al_2.3Si_3.5O₁₀(OH)₂ + 2.94SiO₂(aq) + 0.92Fe²⁺ Ilite

+ 0.368Fe³⁺ + 1.866Mg²⁺ + 0.046Ca²⁺ + 0.184Na⁺ + 4.2H₂O    (3)

This leads to precipitation of authigenic quartz and feldspar overgrowths from the formation waters (Figure ^.121); Bath et al., 1987; Strong and Milodowski, 1987; McBride, 1989).

Mesodiagenetic formation waters become progressively more reducing, and frequently reduce ferric iron. This is particularly the case when the red bed acts as a hydrocarbon reservoir, when conditions may become sufficiently reducing to cause the formation of pyrite from early hematite (Burley, 1984). The crystal structures of dolomite and calcite accommodate Fe²⁺ more readily than Fe³⁺, and late mesodiagenetic carbonates are typically ferroan, in contrast to eodiagenetic carbonates (e.g. Burley, 1984; Strong and Milodowski, 1987).

Modern deep groundwaters from red beds place constraints on the chemistry of mesodiagenetic waters. Waters from Permo-Triassic red beds in the Wessex Basin reach temperatures of more than 75°C (Smith, 1986) and can be considered diagenetic waters. Most deep waters in Permo-Triassic red beds of the UK are of NaCl: type;salinities are highly variable, but may approach halite saturation (>300 g/1 TDS; Edmunds, 1986). Most of the salinity appears to come from halite dissolution in the Permo-Triassic (Edmunds, 1986), and it seems likely that this process has been a major control of water salinity throughout diagenesis.

Neglecting localised redox variations (e.g. within individual organic-rich beds), red-bed sequences as a whole are likely to encounter the most reducing conditions during late burial diagenesis, due to the introduction of hydrocarbon-bearing waters originating in hydrocarbon source-rocks outside the red-bed sequence (Figure 120). The redox state of such waters can be estimated by assuming that the pH is controlled by equilibria involving silicates and that the redox state is buffered by equilibria involving light organic acids. For example, if the activity of K+ is known, then the pH can be constrained from equilibria such as:

Using the calculated pH, the redox state of the water can be constrained with equilibria such as:

Appropriate concentrations of K+ can be estimated from published analyses of oilfield brines (e.g. Gulf Coast USA, data from Carpenter et al., 1974). The activity coefficients of K+ and light organic species can then be calculated using a suitable equation of state such as the HKF equation (Helgeson et al., 1981); the equilibrium constants for the relevant equilibria can be obtained using the computer code SUPCRT92 (Johnson et al., 1991) and thermodynamic data for organic species from Shock and Helgeson (1990). This suggests that at a temperature of 125°C, reducing late-diagenetic waters will be slightly acidic, with a pH of about 5 and a redox state represented by a logfO₂,_bars of about −50 (approximately equivalent to Eh = −200 mV).

Phase boundaries were calculated for 25°C and 100°C at a pressure of 1 bar using the SUPCRT92 code (Johnson et al., 1991). As K+ activities increase due to mineral dissolution during burial, red-bed formation waters will move generally from the kaolinite stability field (A) to the muscovite stability field (B), and then to the K-feldspar stability field (C). Subsequent influx of dilute meteoric, or acidic basinal waters may then move the fluid composition back towards the kaolinite stability field.

Movement and mixing of formation waters from different parts of a red-bed sequence during mesodiagenesis may exert an important effect on diagenetic mineral assemblages. For instance, influxes of dilute meteoric waters which have low K+/H+ activity ratios will tend to produce diagenetic kaolinite, whereas higher-temperature waters have higher K+ concentrations due to mineral dissolution and will produce diagenetic illite (Figure 121). Cooling of high-temperature muscovite-equilibrated waters (point C for 100°C on (Figure 121) may give rise to authigenic K-feldspar (e.g. at point C for 25°C on (Figure 121) during ascent.

Diagenesis during uplift (telodiagenesis)

Telodiagenesis is characterised by progressively decreasing temperatures and formation-water salinities and by progressively more oxidising conditions. This is due to the influx of dilute neutral to low-pH meteoric waters which displace basinal brines and lead to the dissolution of evaporites and the removal of sulphate and halite cements (e.g. Burley, 1984; Parnell, 1992a). This also causes the dissolution of carbonate cement and feldspar, and leads to the precipitation of hematite due to the oxidation of Fe²⁺ liberated during the dissolution of ferroan carbonates. With increasing dilution, SiO₂(aq) and K⁺(aq) activities are depressed, leading to the precipitation of kaolinite (Burley, 1984; Bath et al., 1987; Strong and Milodowski, 1987; Parnell, 1992a; (Figure 121).

Modern groundwater chemistry can be used as an indicator of late-diagenetic mineral transformations. Edmunds et al. (1984) and Bath et al. (1987) found that invasion of meteoric waters into the SSG had flushed any pre-existing Na-Cl-dominated saline waters from the aquifer, and that the groundwater chemistry was dominated by reactions with carbonate cement, detrital dolomite and sulphate minerals.

The pH of pure water in equilibrium with modern atmospheric levels of CO₂ is about 5.6 (Stumm and Morgan, 1981), and this can be used as an estimate of the minimum pH of the late-diagenetic waters. Away from outcrop, however, the pH will rise owing to the interaction between meteoric water and diagenetic carbonates in the red bed. In the SSG of the east Midlands, pH increased from about 7 to about 8 at greater distances from the outcrop (Edmunds et al., 1984). However, in spite of this flushing, there is still a considerable redox gradient down-dip. Edmunds et al. (1984) found that only the waters in the unconfined part of the aquifer they studied were sufficiently oxidising to contain dissolved oxygen, and down dip from the outcrop, Fe²⁺ and then Fe²⁺ and HS⁻ become important in controlling redox potentials as Eh falls from +300 mV to 0 mV.

The nature of metal-transporting formation waters

The characteristics of the waters required for the mobilisation of Cu and other heavy metals in red beds have been evaluated by Rose and co-workers (Rose, 1976, 1989; Rose and Bianchi-Mosquera, 1985), Zielinski et al. (1983, 1986), Eugster (1985), Sverjensky (1984, 1987, 1989), Nash et al. (1981), Durrance (1986), and Hofmann (1990, 1991, 1992). Geological evidence suggests that the mineralising fluids in the Kupferschiefer ore deposits were late-diagenetic basinal brines which migrated through the Rotliegendes sandstones, leached metals from Fe-oxide mineral coatings, and ascended the flanks of basement highs. To transport metals in sufficient concentration, these basinal brines would have been more oxidising than oilfield brines with a redox state buffered by organic matter. The brines deposited their heavy-metal content as a result of encountering reducing conditions in the organic-rich, pyritic Kupferschiefer shales and Zechstein limestone above. Jowett (1986) calculated that a groundwater velocity of 13 cm per year and a Cu solubility of 1000 mg/l in 20–30% Ca-Na-Cl brines could have formed the Cu deposit at Lubin, Poland, in less than 6 Ma, but that, with more realistic Cu solubilities of a few mg/l, convective recycling of water would be needed in order to produce the deposit.

The mineral zonation in the Kupferschiefer concurs with progressive increase of H₂S in an oxidising fluid containing Cu, Fe and possibly Pb and Zn. This may occur through a combination of the addition of H₂S (e.g. through fluid mixing) and the reduction of SO₄²⁻, which is already present in the solution. Textures indicate that organic matter and pyrite are common reductants (Gustafson and Williams, 1981; Sawlowicz, 1992), and water-rock interactions involving these might act both to reduce oxidised sulphur already in solution, and to add reduced sulphur to the solution.

Uranium deposits form where oxidising meteoric waters encounter more reducing groundwaters as they move down-dip through sandstone aquifers, for example where there is organic-rich detritus. In general, under the oxidising conditions which favour U mobility, Fe is locked in hematite and the rock is characteristically coloured red. In contrast, Fe is usually mobilised as Fe²⁺ under reducing conditions which favour U ore precipitation, leading to U deposits being hosted by drab grey- to white-coloured sediments.

The redox boundaries which are associated with sediment colour changes in U ore deposits may extend for some hundreds of metres. However, more localised variations in redox conditions are responsible for commonly observed, hematite-depleted, high-U bleached spheroids, generally from a few mm to a few tens of cm in diameter. The origin of such spheroids has been a subject of some debate, and localised reduction has been variously ascribed to organic detritus (e.g. Durrance, 1986), and to bacterially mediated redox reactions involving dissolved reductants such as organic acids and methane (e.g. Hofmann, 1990, 1991, 1992).

The temperatures of ore-forming fluids in red beds are not as well constrained as for other types of base-metal ore deposits, such as MVT Pb-Zn deposits, but a number of lines of evidence point to low-moderate temperatures below about 150°C. The exsolution of chalcopyrite from bornite heated above 75°C was used by Rose (1976) to infer low temperatures (<100°C) for Cu-ore-forming fluids in red beds. Similar temperatures are indicated by the replacement of early diagenetic framboidal Fe sulphides by Cu sulphides in the Kupferschiefer (Sawlowicz, 1992), and by limited fluid-inclusion data, which suggest temperatures as low as 60–70°C (Naylor et al., 1989) (see also Chapter 5). Slightly higher temperatures of about 130°C are indicated by oxygen isotopic data for illites associated with Kupferschiefer mineralisation (Bechtel and Hoernes, 1993). Indirect evidence for the likely temperatures of mineralising fluids comes from fluid inclusions in other types of sediment-hosted ore deposits, such as MVT deposits which are thought to have a similar origin in basinal brines (e.g. Sverjensky, 1989). These data on MVT deposits suggest temperatures in the range 100–150°C, and occasionally up to 200°C (Roedder, 1984). Burial depths during ore deposition in many sedimentary basins such as the Permo-Triassic basins of the UK often suggest maximum formation-water temperatures between 100°C and 200°C, which are higher than temperatures estimated for Alderley Edge (see pp.83–85 for estimates of palaeo-subsurface temperatures in the Cheshire Basin). In contrast, U-deposits associated with red beds are often formed at near surface temperatures.

The theoretical constraints on the likely compositions of mineralising fluids have been assessed by Rose (1976, 1989). In summary, he concluded that the aqueous speciation of Cu in the system Cu-O-H-S under oxidising (Eh>300 mV) conditions was dominated by Cu²⁺, but the solubility was predicted to be >1 ppm only at pH<6.2. He considered that the carbonate complexes CuCO₃(aq) (pH 7.4–9.3), or Cu(CO₃)₂²⁻ (at pH > 9.3) may account for appreciable Cu solubility in the presence of CO₂. For P_CO2 ranges of 10⁻¹.⁵–10^−2.5 bars, malachite was predicted to be the stable mineral and at higher levels of P_CO2 azurite (2CuCO₃.Cu(OH)₂) was predicted to be stable.

Due to the abundance of halite-bearing evaporites in many red-bed sequences, the dominant Cu-complexing ligand is Cl⁻. However, the Cl⁻ complexation of Cu²⁺ is very weak, and the solubility of Cu in oxidising (Eh > 300 mV) , Cl-rich fluids may be reduced due to the precipitation of atacamite (Cu₄Cl₂(OH)₆). In contrast, Cl⁻ forms very strong complexes (Cu₂Cl⁻, CuCl₃²⁻) with Cu⁺ (the cuprous ion). These complexes are stable under more reducing conditions than Cu(II) minerals (malachite CuCO₃.Cu(OH)₂, tenorite CuO, brochantite Cu₄(SO₄)(OH)₆, atacamite), and more oxidising than native copper or Cu sulphides. Therefore, these cuprous Cl complexes are probably the most important Cu-transporting species over most of the range of redox conditions encountered during red-bed diagenesis. Complexes with CN⁻, F⁻, NH₃(aq) are unimportant, whereas SO₃²⁻ and S₂O₃²⁻ complexes may be important in weathering conditions. The complex Cu(HS)₃²⁻ is important only at very high HS⁻ concentrations, since Cu sulphides have low solubilities. Complexes with oxalate, formate, acetate and fulvate may be important in organic-rich pore fluids.

The Eh-pH conditions of most Cu-bearing red-bed formation waters are compared with the fields of stability of common Cu-ore minerals in (Figure 122). The figure is constructed for a total pressure of 1 bar and a temperature of 25°C, since reliable data for higher pressures and temperatures are not available for the aqueous Cu species which are shown. The shaded field on the diagram approximates the commonest ranges of pH and Eh over which Cu is likely to be mobilised as Cl complexes and is not intended to represent the total possible field of Cu mobility. The lower pH limit is taken as 5, which is near the value predicted from silicate equilibria such as equation (4) above, and the upper pH limit is taken as 9, which is near the maximum value found in most groundwater systems (e.g. Hem, 1989). The exact boundaries of the field are dependent on water composition and temperature, but nevertheless the figure serves as an aid in comparing the conditions which favour Cu transport with mineral stabilities in different compositional systems. Therefore, the field of Cu mobility as cuprous chloride has been superimposed on Eh-pH plots which illustrate Fe, Pb, Zn and U speciation (Figure 123).

Evaluation of the complexation behaviour of other heavy metals reveals that Fe is insoluble as hematite under most conditions required for Cu mobility (Figure 123)a, whereas Pb is soluble as PbCl₂(aq) (at pH < 7; (Figure 123)b, and Zn is soluble as ZnSO₄(aq) (at pH < 8.1; (Figure 123)c. Silver behaves differently, being soluble under only the most oxidising conditions as AgCl₂⁻ (at Eh > 200 mV). Therefore, Ag-rich Cu deposits may only form under more oxidising conditions.

Uranium mobility is similarly favoured by oxidising conditions, which allow formation of uranyl (U0₂²⁺) complexes rather than uranous (U⁴⁺) complexes (Langmuir, 1979; Nash et al. 1981; Durrance, 1986). However, in contrast to Cu, Pb, Zn etc., the most common ligand involved in U transport is generally carbonate (Figure 123d), particularly in the presence of meteoric waters, although phosphate, vanadate, fluoride and silicate complexes may also be important. An important feature is that although common U ore minerals are oxides, such as uraninite (UO₂; and the variety pitchblende, U₃O₈), they form under pH-redox conditions which are similar to those under which Cu, Pb and Zn sulphides form (cf. (Figure 122) and (Figure 123)d.

Sverjensky (1984, 1987, 1989) has considered the chemical evolution of brines during diagenesis and the possible effects on metal mobilisation and transport. His thesis is that a single basinal brine may evolve chemically to become an ore-forming fluid for a range of different types of heavy-metal sulphide deposits (Figure 124). Central to this idea are the chemical buffering properties of the aquifer along which the basinal brine migrates. Sverjensky considered a basinal brine initially saturated with galena, chalcopyrite, pyrite, muscovite, kaolinite and quartz but undersaturated with respect to sphalerite, with up to 1 ppm Pb, 0.1 ppm Cu and 5 ppm Zn. Passage of this fluid through a carbonate-cemented aquifer would produce a mineral deposit characterised by high Zn/Pb and (Zn + Pb)/Cu ratios. Transport of the same fluid through a quartz-cemented sandstone could exhaust the buffering capacity of the aquifer relatively quickly such that the resulting mineral deposit would be galena-rich. Maintenance of the oxidation state near the hematite-magnetite buffer would prevent mobilisation of Cu (solubility of Cu in a basinal brine is c.0.07 mg/l; (Figure 125). However, if the basinal brine were to migrate through a red bed containing hematite and anhydrite, the oxidation state and SO₄²⁻/H₂S ratio of the fluid would be increased by reactions such as:

CaSO₄ = Ca²⁺ + SO₄²⁻(8)

4Fe₂O₃ + H₂S + 14H⁺ = 8Fe²⁺ + SO₄²⁻ + 8H₂O (9)

Migration of this more oxidising fluid through a red bed could scavenge the necessary Cu, Zn and Pb to form a mineral deposit with much greater Cu contents [Cu > (Pb + Zn)]. In this model, redox conditions of logfO₂,_bars greater than −46 at 125°C are necessary to mobilise Cu as cuprous chloride complexes (Figure 125).

Bleaching of red-bed sandstones is a feature of some redbed-hosted ore deposits (e.g. Ixer and Vaughan, 1982) and is a result of the destabilisation and dissolution of hematite. Dissolution of hematite requires fluids more reducing and/or of lower pH than most of those waters capable of mobilising significant cuprous chloride (compare (Figure 122) and (Figure 123)a. Similarly, bleaching due to the reduction of hematite to magnetite would require conditions more reducing than the field of significant Cu mobility (logfO₂,_bars less than about –50.3; (Figure 125).

Relative to mean continental crustal concentrations, some red beds are depleted in Cu, but are undepleted in Zn, Pb and Co (Rose and Bianchi-Mosquera, 1985; Haslam and Sandon, 1991), while ore deposits associated with red beds contain variable proportions of these metals (Gustafson and Williams, 1981). These observations contrast with thermodynamic predictions which suggest that Cu, Pb and Zn should all be soluble at similar pH and in brines with similar overall chemistry (see pp.195–197), and imply that additional factors must have influenced metal mobility. One possibility is that sorption exerts a control on metal behaviour, and this possibility is considered further below.

Since red beds are important hydrocarbon reservoirs in many parts of the world, organic species are likely to be important components of red-bed diagenetic fluids. The organic fluid component is likely to exert a major control on red-bed ore deposits by directly controlling mineral stability, and by controlling the speciation of metal cations in solution. Increasingly, attention is being drawn to controls on cation mobility and secondary mineral stability by these organic components (e.g. Surdam et al., 1984; Manning, 1986; Hennet et al., 1988; Thornton and Seyfried, 1987; Seewald et al., 1990). Liquid hydrocarbons may transport cations such as V, Ni, Cu and Zn (Manning, 1986), while Ca, Mg, Fe, Al, Sr, Mn, U, Th, Pb, Cu and Zn may be transported as complexes with light mono- and di-carboxylic acids (e.g. Giordano and Drummond, 1991; Holm and Curtis, 1990; Fein, 1991; Harrison and Thyne, 1992). Furthermore, it appears that the inorganic chemistry of aqueous solutions exerts a control on the way in which the organic component behaves: for example, high Ca concentrations lead to the formation of Ca oxalate and the prevention of oxalate complexes transporting metal cations (Hennet et al., 1988; Huang and Longo, 1992).

The role of organic acids in controlling mineral stability has been a subject of some debate, and it is uncertain, for example, whether the primary role of the acids is to increase reaction rates (e.g. Stoessell and Pittman, 1990), or whether metal-organic complexing enhances mineral solubility (e.g. Surdam et al. 1984; Surdam et al., 1989). However, it seems that organic acids may buffer the pH and redox states of fluids (e.g. Lundegard and Land, 1986; Shock, 1988; Huang and Longo, 1992), and therefore these acids may have the capacity to influence complexing between metals and inorganic ligands even where complexing between metals and organic species is unimportant.

Theoretical models for diagenesis and ore formation

Approach to modelling

Theoretical models of diagenesis and ore formation can be used to place important constraints on interpretations of field and analytical data. Such models should be viewed as valuable interpretative tools that limit the range of plausible models for actual observations; the models do not prove that a particular interpretation is correct.

In the present study, the approach adopted was to construct some generic theoretical models for red-bed diagenesis and ore formation. These were used to suggest a plausible interpretation of data from the Cheshire Basin. The generic character of the models means there are differences between their input parameters and actual data for samples from the Cheshire Basin. However, these differences do not invalidate the application of the generic model results in constructing a conceptual model for diagenesis and mineralisation in the Cheshire Basin.

The importance of pH and redox state in red-bed diagenesis and mineralisation

In general, the diagenesis of silicate and carbonate minerals is influenced greatly by pH and only to a small degree by redox state, whereas the distribution of heavy metals is controlled by both pH and redox state. This is particularly important in red beds because during diagenesis the formation waters of these sediments acquire a wide range of pH and redox states. However, unlike the solute content of formation waters (which can be constrained from groundwater analyses from deep boreholes) and diagenetic mineral assemblages (which can be examined directly), pH and redox states are usually inaccessible to direct measurement. This is because of problems associated with making measurements in highly saline waters, and the inherent difficulty of collecting groundwaters without perturbing their in-situ pH and redox states, due to factors such as degassing and interaction with steel drilling equipment. The pH and redox states of red-bed formation waters must therefore usually be estimated theoretically and/or by using experimental data. In order to assess the significance of fluid mixing as a control of pH and redox state during red bed diagenesis, three important processes are considered: mineral-fluid equilibria; fluid mixing; and sorption.

Mineral–fluid equilibria

An equilibrium thermodynamic model was used to simulate mineral–fluid interactions during red-bed diagenesis, using the code EQ3/6 (Wolery, 1992a, b; Wolery and Daveler, 1992) according to the scheme in (Figure 126).

The diagenetic modelling used an approach similar to that of Bruton (1989), and the initial, pre-diagenetic sediment composition was estimated by assuming that a typical red-bed sequence is a representative sample of the bulk upper continental crust. Accordingly, estimates of the upper-crustal mineral composition, with mean abundances of 25 ppm of Cu, 20 ppm of Pb and 70 ppm of Zn (from Taylor and McLennan, 1985) were used. These values differ from the concentrations measured for red beds in the Cheshire Basin (where mean values, excluding samples from mines, were 33 ppm Cu, 10 ppm Pb and 33 ppm Zn; Chapter 4), but this difference does not invalidate the application of the model results to the Cheshire Basin. Tenorite (CuO), cerussite (PbCO₃) and zincite (ZnO) were taken as the hosts of these metals because there are only limited thermodynamic data for other ore-forming minerals.

The initial compositions of the diagenetic waters were considered to be buffered by the atmosphere, which constrained their initial CO₂ content at a value represented by logfCO₂,_bars = –3.5. The initial pH was taken to be 8, near the middle of the range observed for near-surface waters in modern arid environments. In common with most natural saline waters, the initial waters were taken to be chloride dominated. In such waters Na⁺ and Ca²⁺ are the commonest cations, and fluid-inclusion studies of a wide range of MVT ore deposits and red-bed evaporites suggest that ore fluids also typically have Na and Ca as the dominant cations (e.g. Roedder, 1984; see also Chapter 5 of this volume). Therefore, the initial computer simulations used a 1:1 (molar ratio) Na:Ca chloride solution in order to approximate the initial diagenetic water.

The proportions of rock and water play a major role in controlling diagenesis. The simplest model assumes that the system is closed, and that the porewater-rock system reaches equilibrium. In the example, the water/rock volume ratio was taken to be the same as the percentage porosity of the rock (a value of 25% was used). In nature, open-system conditions generally prevail, but models of these require significantly more complex computer codes than are required to model closed systems. A major uncertainty is the degree of equilibration between any moving fluid, the red-bed aquifer through which it passes, and any minerals which it precipitates. This is a function of reaction kinetics and the rate of fluid movement through the rock, which are coupled since mineral precipitation or dissolution will be an important control on rock permeability.

It can be concluded that diagenetic temperatures during ore deposition are usually less than about 150°C (see above). Since maximum burial depths can often be inferred from stratigraphical evidence and diagenetic mineral assemblages to be less than 4 km, the maximum burial pressures are often less than about 1 kbar. Although many codes, such as EQ3/6, cannot use pressure as an independent variable, mineral equilibria during red-bed diagenesis will be relatively insensitive to pressure. In the example, early diagenesis was simulated for a temperature of 25°C and 1 bar, whereas later diagenesis was simulated at a temperature of 125°C and a pressure of 2.32 bars (the pressure being fixed by the equilibrium curve for liquid water and water vapour at 125°C).

Equilibrium modelling which neglects reaction kinetics and which allows all thermodynamically stable phases to precipitate during a simulation, tends to produce unrealistic mineral assemblages. For example, in the absence of user-defined constraints on precipitation, the EQ3/6 model predicted numerous geologically unreasonable zeolites to form during early diagenesis. A knowledge of the actual mineral phases which occur in red beds, must therefore be used to decide what minerals should be allowed to precipitate during a simulation. The only minerals which were allowed to appear in the early diagenetic mineral assemblage in this case were those listed in (Table 32), plus tenorite (CuO), cerussite (PbCO₃), zincite (ZnO), cuprite (Cu₂O), chalcocite (Cu₂S), covellite (CuS), galena (PbS) and sphalerite (ZnS).

In order to simulate later, deeper, higher-temperature diagenesis in the presence of red-bed evaporites, the modelled early diagenetic waters were equilibrated with anhydrite. The early diagenetic mineral assemblages were reacted at this higher temperature with the resulting model pore water to yield a model late-diagenetic pore water (Figure 126); (Table 33).

This approach gave a reasonable representation of eodiagenetic and early mesodiagenetic mineral assemblages, characterised predominantly by quartz and calcite, with smaller quantities of hematite, mica and clay phases (Table 32). However, certain features of the model are inconsistent with observed mineral parageneses (see Chapter 5). Notably, late burial diagenetic feldspar overgrowths were not predicted. This is because the waters never attained realistic salinities, which meant that feldspars never reached saturation.

The model becomes increasingly inaccurate for waters with ionic strengths greater than about 1 molal, although for the range of water compositions considered here, reasonable results were given for ionic strengths up to about 2 molal. Varying the salinity of the starting water up to this value had little effect on the diagenetic mineral assemblage. However, in late mesodiagenesis, evaporite dissolution may result in ionic strengths significantly greater than 5 molal (at halite saturation at 25°C ionic strengths are c.7 molal) and therefore late mesodiagenesis cannot be modelled accurately. Although the specific-ion interaction model (Pitzer, 1973) provides a theoretical framework for modelling such concentrated waters, the necessary data are extremely limited. A lack of Pitzer coefficients for silica means that silicate equilibria cannot be simulated. Furthermore, the Pitzer data are applicable only over limited temperature ranges, with data for carbonates, for example, being limited to 25°C.

When pH and redox states were allowed to drift from their initial values, it was found that logfO₂,_bars remained almost constant during burial diagenesis, whereas pH fell by about 1.5 units. This is because the model calculated that all Fe²+ in the initial sediment was oxidised during early diagenesis under atmosphere-buffered conditions, so that during later diagenesis oxygen was not consumed by Fe oxidation. In nature, it is more likely that a significant proportion of the Fe will remain in the ferrous state until after burial has isolated the sediment from the atmosphere, and more reducing conditions will evolve as Fe oxidation depletes oxygen in the formation waters.

A further major problem is the prediction of geologically reasonable concentrations of metals in solution. In the example, only very small proportions of the available Cu in the red bed were predicted to be released to solution (c.1% for 1 M Cl solution), whereas all the available Pb and Zn were liberated from the model rock. This contrasts with the experimental simulations of diagenesis performed by Zielinski et al. (1986) which indicated that up to c.45% of Cu, but only c.20% of Pb and Zn should be available to the solution. These experimental data suggest that 1 kg of sandstone could liberate up to c.70 mg Cu, c.20 mg Pb and c.40 mg Zn (taking values for metal concentrations in sandstones from Gustafson and Williams, 1981). One possible explanation is that because the model formation waters remained highly oxidising, as discussed above, Cu is effectively locked in tenorite (Figure 122). For this reason, additional models were run specifying more realistic reducing redox states (to logfO_2,bars = –40) for the later diagenetic waters. However, these gave similar results whatever redox states were selected, with Cu concentrations remaining at c.0.01 ppm, and Pb and Zn concentrations remaining at c.20 ppm and c.70 ppm respectively. Alternatively, the inconsistency with experimental results may be due to a poor knowledge of the actual phases that control the solubility of Pb, Zn and Cu during diagenesis and a lack of thermodynamic data, especially high-temperature data, for potential phases that might control the solubility of Pb, Zn and Cu.

Fluid mixing

In order to assess the importance of fluid mixing it is necessary to estimate the proportions of fluid of different redox states which are necessary to produce significant changes in the aqueous chemical speciation and precipitated mineral assemblages. It is instructive to consider the amount of reducing fluid which is required to mix with an oxidised, metal-bearing red-bed formation water, in order to strip it of heavy metals owing to sulphide formation. A limiting case is where the most oxidising water that can be envisaged mixes with the most reducing water that is realistic. When the formation water is at its most oxidised, the maximum amount of a reducing water will be required to remove heavy metals completely.

The most oxidising water that can be envisaged is one which has been buffered by the atmosphere throughout the oxidation of all ferrous iron in the rock. This water was represented by the model-atmosphere-equilibrated, late meso-diagenetic pore water described above. The most reducing burial diagenetic water which can be envisaged is an oilfield water with a redox state imposed by organic matter. Such a redox state is appropriate for many reduced sediments which are actually associated with red-bed Cu mineralisation, such as the Kupferschiefer. These conditions were modelled as illustrated in (Figure 127), by assuming that silicate equilibria controlled pH, that organic-acid equilibria controlled redox states (see equations 4–7 above), and that the water was equilibrated with a typical mudrock mineral assemblage such as might occur in a hydrocarbon source rock: plagioclase, illite, chlorite, calcite, quartz, galena, sphalerite, pyrite and chalcopyrite. Salinity was specified by making the simple assumption that the solution was charge-balanced with Cl⁻. Other parameters were fixed by using oilfield-brine analyses as a guide to reasonable values. For instance, alkalinity was fixed by assuming 4000 mg/l acetate (near the maximum concentration reported for oilfield brines by Carothers and Kharaka, 1978). This gave a model reducing oilfield water of the composition shown in (Table 33) and with a metal content of 1.3 X 10⁻³ ppm Pb, 1.8 X 10⁻² ppm Zn, and 6.3 X 10⁻⁹ ppm Cu.

The limiting case was investigated by using the EQ6 code to model progressive addition of the reducing oilfield water to the model-atmosphere-equilibrated, late meso-diagenetic pore water. The model was terminated when sphalerite appeared as a product, since the common mineral zonation in red-bed ore deposits suggests that sphalerite generally occurs late in the sulphide paragenesis (e.g. Gustafson and Williams, 1981). The results of this model are illustrated in (Figure 128) and (Table 33). There is a sharp redox boundary when the reducing oilfield water constituted about 5% of the mixture, with logfO₂,bars decreasing from c. –5 to c.–45 (Figure 128)a. At this point, all of the Cu was removed from the solution, reflecting the low solubility of Cu sulphides compared with Pb and Zn sulphides (Figure 128)b. The concentrations of Pb and Zn decreased only after all the Cu had been stripped from the solution, and the onset of Pb removal from solution commenced before the onset of Zn removal. (Figure 128)b also shows that metal concentrations fell initially, during the addition of the first 3% of reducing oilfield water, as a consequence of simple dilution.

In the model, dolomite and baryte were the dominant predicted gangue minerals, while quartz, muscovite, albite and hematite formed a minor proportion of the mineral assemblage (Figure 128c). Trace amounts of albite and muscovite were present throughout the paragenesis, while trace quantities of hematite were predicted by the model initially but disappeared when about 30% of the mixture was reducing oilfield water. Thus, the mixing model can reproduce the major features of many red-bed hosted copper deposits.

It is significant that for this limiting case only 5% of reducing oilfield water is needed to strip all the Cu from the oxidised late mesodiagenetic pore water. This implies that only a small amount of mixing between different fluids could be a major control of heavy-metal mobility and ore-deposit formation. The unrealistically small Cu concentrations in the initial late burial diagenetic pore water do not invalidate this conclusion, because Cu sulphides are relatively insoluble and in sulphur-rich fluids, such as those which are typical of evaporite-bearhig red-bed sequences, sulphur speciation is a major control on metal mobility. It is likely that formation-water redox states more reducing than the atmosphere, and outside the stability field of Cu oxides (Figure 122), would mobilise more Cu than in the model. However, such waters will also lie closer to the field of sulphide stability, so that smaller quantities of reducing water will be required to cause sufficient reduction of the system to result in Cu-sulphide precipitation.

The role of sorption

Equilibrium thermodynamic models are of limited applicability because they generally assume that mineral dissolution-precipitation equilibria govern the composition of the aqueous phase. However, equilibria between dissolved and sorbed species may be a major control on the composition of diagenetic waters, particularly for trace metals. The dominant sorbents in nature are the common hydrous oxides of Fe, Al, Mn and Si, and, since Fe oxides are common constituents of red beds, sorption is likely to be particularly important in controlling metal mobility in red beds.

Oxide and hydroxide surfaces are capable of adsorbing or dissociating H+ from a surface OH⁻ group to acquire a net surface charge, according to:

S – OH + H⁺ = SOH₂ +  (10)

S – OH = SO⁻ + H⁺   (11)

where S-OH indicates surface sites occupied by OH. These charged sites affect the electrostatic attraction for other ions on or near the surface of the solid and create sites for further reaction, for example by:

S – OH + Me⁺ = SO⁻Me⁺ + H⁺ (12)

where Me⁺ is a metal cation. Ions bound only by the electrostatic forces are said to be non-specifically sorbed whereas those bound at surface sites are specifically sorbed. The abundance of these SOH₂+ and SO⁻ charged sites is sensitive to solution pH, and thus the proportion of a metal cation which is specifically sorbed is a function of pH (Figure 129), which in turn is a function of other equilibria, such as that in equation (4). Specific adsorption of cations onto oxide surfaces increases from low values to complete adsorption over a relatively small pH range to form an adsorption edge. At low pH where the surface sites are generally positively charged, sorption of metal cations is low, whereas at higher pH sorption increases and can result in the 100% retention of metals. At intermediate pH values sorption may cause retardation, but not complete retention of a metal. Factors other than pH which affect sorption are:

1. Concentration of the metal, which affects sorption as the number of occupied sites approaches a significant proportion of the total sites.
2. Competing cations, which reduce the number of sites available to the cation of interest.
3. Redox state, which controls aqueous speciation and the solubility of sorbing phases.
4. Temperature, since sorption generally decreases as temperatures rise.
5. The sorbent/sorbate ratio.

The importance of sorption during red-bed diagenesis has been evaluated by calculating the position of the sorption edge for Cu, Pb, Zn, Co and Ag when these are sorbed by hydrous ferric oxide (FeOOH). The HYDRAQL code (Papelis et al., 1988) was used in conjunction with a diffuse layer model, surface complexation data from Dzombak and Morel (1990) and the HYDRAQL database for solution species (Papelis et al., 1988). An oxide surface area of 600 m²/g was assumed, and the surface densities of strong and weak sites were taken to be 0.005 and 0.2 mol per mol Fe respectively (Dzombak and Morel, 1990). The effects of varying trace-metal concentration, ionic strength and surface-site concentrations on the proportion of metal which is sorbed were investigated. Redox conditions were within the stability field of Cu +, owing to a lack of data for the cuprous ion. Under these conditions, increasing the ionic strength from 0.1 M to 1 M and varying the metal concentration by four orders of magnitude had no significant effect on the sorption curves. However, increasing the surface area, or the density of sites available for sorption, lowered the pH at which 100% of the metal was sorbed.

(Figure 129) shows the calculated graphs for sorption of Pb, Cu, Zn, Co and Ag (5 X 10⁻⁷ M) in 0.1 M NaCl onto goethite with surface areas of 53 400 and 53.4 m² per litre of solution. The position of the pH edge is partly dependent on the sorbate/sorbent ratio. Similar graphs are obtained in each case, but the adsorption edge is shifted up by about 1.5 pH units for the lower sorbate/sorbent ratio (Figure 1296 compared to (Figure 129c).

The 'Approximate Copper Transport Field as Cu⁺-Cl⁻ in (a) is the same as the shaded field in (Figure 122). Sorption was modelled for Eh conditions consistent with Cu²⁺ stability; Cu' is considered likely to sorb in a similar fashion to Ag⁺. It is evident that under the range of pH required for Cu- and Ag- chloride stability (a), significant proportions of Cu and Ag may, at certain pH values, be rendered immobile (b and c). In the presence of iron oxides, the fields of Cu and Ag mobility may be significantly smaller than the approximate fields of Cu and Ag chloride transportation in (a).

Variable redox conditions during red-bed diagenesis do not greatly affect sorption of Pb, Zn, Co and Ag, because the oxidation states of these metals will be mostly constant. In contrast, Cu can form Cu²⁺ and Cu⁺ complexes, and redox may control the sorption of this metal, although a lack of data for Cu⁺ prevented modelling this. Sorption of Cu+ may be weaker than for Cu²⁺, because Cu⁺ probably behaves like Ag⁺ which forms strong chloride complexes (AgCl₂⁻ and AgCl₃²⁻, AgCl₄³⁻) and which is weakly sorbed at acid and neutral pH (Figure 129)b. However, it should be noted that Cu is mobile as Cu complexes under more reducing conditions than those under which Ag is mobile as Ag+ complexes.

Recently Rose and Bianchi-Mosquera (1993) carried out a series of experimental simulations of red-bed conditions, and investigated sorption of Pb, Zn, Co, Ni, Cu and Ag by goethite under a range of redox conditions. Their experimental adsorption curves were similar to the modelled curves, but compared to Pb, the Cu adsorption edge was at slightly lower, rather than slightly higher, pH value, possibly because of differences in Eh, and hence speciation.

From (Figure 129 it is clear that under conditions of pH similar to those bounding the field of Cu transport as cuprous chloride, goethite is likely to sorb all Cu²⁺, and, if Cu⁺ behaves like Ag⁺ and sorbent/sorbate ratios are suitable, may sorb significant amounts of Cu⁺. It is likely that Cu⁺ and Ag⁺ are more mobile if the adsorbing phase is hematite rather than goethite (Rose and Bianchi-Mosquera, 1993). This is consistent with the release of heavy metals to solution due to the transformation of goethite to hematite during burial diagenesis.

Modern groundwaters in the Cheshire Basin

The application of generic models of diagenesis and ore formation to the Cheshire Basin can be constrained by investigating the hydrochemistry of modern formation waters from the basin. Such an investigation also allows an evaluation of the relationships between modern groundwaters and the fluids responsible for the diagenetic and ore-mineral assemblages found in the basin. Information was obtained from relevant published literature, unpublished theses and the databases maintained by the National Rivers Authority (NRA; now part of the Environment Agency (EA)).

Previous studies of the hydrogeology of the Cheshire Basin

Campbell (1982) described the permeability of the Permo-Triassic sandstones of the lower Mersey basin, using data from boreholes at High Croft Farm, Padgate, Halewood and Kenyon Junction. He gave determinations of intrinsic permeability, porosity, density and grain-size distribution, but presented no groundwater chemical data. Similarly, Fletcher (1977) described the hydrogeology of the 'Bunter Sandstone' of the basin, north of Shrewsbury, and although data he determined included permeability and transmissivity, derived from borehole pumping tests, no chemical data for the water samples are presented.

Haslam et al. (1950) gave some analyses of Br in brines produced artificially during the exploitation of the Cheshire salt deposits. The Br contents ranged from 97.9 to 152.7 mg/l, and decreases of up to 30 mg/l were observed for brines from individual boreholes during pumping over an interval of three years (Figure 132, arrow A).

The most comprehensive study of the hydrochemistry of the groundwaters in the West Cheshire aquifer have been given in Lucey (1987). In addition, two recent studies describe the hydrochemistry of the Permo-Triassic aquifer in the northern part of the Cheshire Basin, near the Mersey (Tellam, 1995) and in the Liverpool area (Tellam, 1996).

Lucey (1987) reported groundwater chemical data for 150 sites throughout the Cheshire Basin. These sites included agricultural wells and boreholes, some springs and a subset of the boreholes in the NRA's monitoring borehole network. For all the sites, analyses for Ca²⁺, Mg²⁺, Na⁺, K⁺, alkalinity, SO₄²⁻, Cl⁻, NO₃⁻, Fe²⁺, SiO₂, dissolved oxygen and pH were undertaken. In addition, analyses are given for ²H, ¹⁸O and ³H for 32, 47 and 67 samples respectively. In recharge areas, carbonate dissolution owing to the action of dissolved CO₂ in meteoric waters was concluded to be the major process supplying Ca²⁺;, Mg²⁺, and HCO₃. Groundwater salinities were reported to increase along flow paths, and this was ascribed to the slow flushing of old saline waters. All the saline waters were found to be of Na – Cl type and the maximum salinities at the base of the basin fill were estimated from electrical conductivity data, from an 850 m coal-exploration borehole at Collinge [SJ 41429 71112], to reach approximately 1.3 times sea-water concentration. Lucey used these observations to conclude that, inland, the dominant source of salinity in the saline waters is most probably halite dissolution. However, in the vicinity of the Mersey Estuary saline intrusion of marine waters was inferred.

Availability of data from unpublished sources

Apart from Lucey (1987) and Tellam (1995, 1996), the two main sources of hydrochemical data for the Cheshire Basin are the north-west and Severn-Trent regional offices of the NRA (now the North West and Midlands regions of the EA). These offices hold records for several thousand groundwater monitoring or abstraction sites, including boreholes, springs and surface-water sampling localities. For most of these sites there are no chemical data. The most complete datasets are those for the networks of observation wells maintained by the EA. The locations of these wells are shown in (Figure 130). The distribution of the wells reflects their purpose, namely the monitoring of potable groundwater resources. Accordingly, the observation wells are generally in areas underlain by sediments of the SSG, which form the major groundwater aquifers. There are very few wells located on areas underlain by the MMG, since this generally acts as an acquiclude (Lucey, 1987; Tellam, 1995; (Figure 130).

The EA North West regional office retains records for 249 groundwater observation wells in the general vicinity of the northern part of the Cheshire Basin, while the Midlands regional office holds data for 160 localities (Figure 130). The records for these wells include chemical and pumping data, and there are analyses of packer, foam flush, water flush and core pore-water samples.

Data for 11 boreholes from the EA North West region, which were considered to have the most complete datasets, were obtained for further interpretation: Littleton [SJ 4350 6699]; Stanlow [SJ 4309 7620]; Mickle Trafford [SJ 4409 7073]; Halewood [SJ 4639 8381]; Rainhill [SJ 4928 8972]; Organsdale 1 [SJ 5507 6830] Clotton Trial [SJ 5280 6360]; Wood Lane [SJ 5444 6486]; Thornton [SJ 3426 9384]; Southworth Hall 1 [SJ 6233 9309]; and Southworth Hall 2 [SJ 6244 9384]. Core pore-water analyses were available for the boreholes at Littleton, Stanlow, Mickle Trafford, Wood Lane, Southworth Hall 1, Thornton, and core material was available for the boreholes at Littleton, Stanlow, Mickle Trafford, Halewood, Rainhill, and Thornton.

The deepest EA borehole in the basin is the Clotton Trial Observation Borehole [SJ 4309 7620] (EA record no. SJ/56/61) and a water analysis is available for a sample from the SSG at 305 m depth in this borehole. Additionally, a groundwater sample from a borehole near Chester, from a depth of approximately 500 m in the Coal Measures, was analysed for major, minor and trace elements, and for ¹⁸O/¹⁶O and ²H/¹H isotopes (Table 34).

Interpretation of groundwater analyses

Assessment of data quality

Groundwater data must be interpreted with caution, since there are a number of factors that tend to cause water samples to differ chemically from in-situ formation waters –particularly drilling-mud contamination, reaction with well casing, degassing during sampling, and atmospheric contamination during sampling. Determinands such as Eh, pH, O₂, CO₂ and Fe are most likely to be affected by these processes, but determinations of all trace constituents are likely to have greater uncertainties than those of major and minor constituents. Groundwater samples may also be products of mixing in the borehole, either between groundwaters from different horizons or between waters occupying different void spaces in the same horizon (e.g. water from fractures and water from the rock matrix, mixed in proportions disturbed by drilling). In spite of these problems, the groundwater analyses from the Cheshire Basin do allow some useful interpretations to be made.

Summary of the main groundwater types

The main groundwater types in the Cheshire Basin are summarised on a Piper diagram in (Figure 131), and representative groundwater analyses are given in (Table 34). The major-element concentrations enable four main types of groundwater to be discerned: saline (TDS > 10000 mg/l) Na-Cl dominated waters; brackish (TDS 1000–10000 mg/l) Na-Cldominated waters; fresh (TDS 5 1000 mg/l) Na-Cl type waters; and fresh (TDS 51000 mg/l) waters with a Ca-Na-

These boreholes are circled on (Figure 130). All the water analyses are averages for a particular formation or for an entire borehole. The sizes of the circles in the central, lozenge-shaped area are proportional to the salinities of the waters. The deep water from the Coal Measures in a borehole near Chester (CMC) is considerably more saline than any of the shallower groundwaters. The chemistries of the fresh shallow groundwaters show a trend from Na-Cl to Ca-Na-HCO₃ dominated.

HCO₃ chemistry. The approximately linear trend which is seen in the lower left triangular portion of (Figure 131) is consistent with calcite and dolomite dissolution during progressive water-rock interaction.

These chemical variations are consistent with the conclusions of Lucey (1987) and Tellam (1995, 1996), and can be interpreted in terms of the dissolution of halite and carbonate by fresh meteoric recharge waters (giving rise, respectively, to NaCl and mixed Ca-Na-HCO₃ chemistries) and marine saline incursion near the Mersey Estuary (giving rise to Na-Cl chemistry). The salinity levels which are reached will depend on a variety of factors, including the availability of carbonate and evaporite minerals in the aquifer, the fluxes of recharge waters, and the degree of saline intrusion (which is in turn linked to the quantities of groundwater abstracted from the aquifer).

The origins of the groundwaters

Although Lucey (1987) presented a considerable quantity of 180/160_, 2H/1H and³H data, he did not give a detailed interpretation. The limited stable-isotope data are consistent with most of the waters being initially of meteoric origin, with δ¹⁸O between –1.3 and –9.4‰_smow, δ²H between –12 and –63‰_smow and all waters lying close to the meteoric water line (for the UK δ²H‰_smow = 6.6 X δ¹⁸O‰_smow +1; Bath et al., 1979). However, Lucey noted that in general samples from boreholes close to the Mersey Estuary showed a trend of increasing heavy-isotope abundances with increasing salinities, while saline waters from inland showed an opposite trend. The trend in the near-coast samples can be explained by mixing of meteoric and marine waters. In contrast, because salinity generally increases with depth, the inland trend also implies that deeper inland samples are depleted in heavier isotopes, although this cannot be confirmed since most of Lucey's data cannot be related to a specific depth. A similar decrease in heavy-isotope abundances with increasing depths has been described from the East Midlands Triassic aquifer and has been interpreted as a palaeoclimatic effect, with the most depleted samples reflecting recharge during Pleistocene glacial episodes (Bath et al., 1979). Thus, the stable-isotope data for groundwaters from the Cheshire Basin are consistent with an increasing groundwater age with depth, and with the deeper waters possibly being recharged during colder climatic conditions than at present.

Tellam (1995) arrived at a similar conclusion regarding the age and origin of the waters. He pointed out that the saline waters did not necessarily arise through evaporite dissolution during the Pleistocene. Instead, he suggested that the depleted isotopic compositions of the present-day saline waters could plausibly have been produced by mixing between a parent brine (with δ¹⁸O_smow around –9‰) and fresh Pleistocene recharge waters (with δ¹⁸O^smow between –9.3‰ and –13.5‰).

The tritium (³H) data given by Lucey also potentially allow constraints to be placed upon groundwater ages. Tritium has a half-life of 12.43 years (IAEA, 1983), and groundwaters with ³H concentrations above backround levels (4–25 Tritium Units (TU)) are considered to have a component of water which was recharged after the beginning of the atmospheric testing of nuclear bombs in the early 1960s: 1964 is usually taken as 'time zero' (IAEA, 1983). Lucey reported ³H levels in the range 0.6–430 TU, with 30% of the samples having ³H concentrations > 25 TU.

Thus approximately one third of all the groundwaters sampled by Lucey contain a significant component which was recharged within the last 30 years. However, unlike the stable-isotope data, the ³H values show no correlation with salinities. Because only a limited number of samples came from known depths the reasons for this are uncertain, but it is probably because variations in the ³H and stable-isotope data reflect recharge over different time scales: the variations in ³H relate to time scales of the order of 10¹ years whereas differences in stable-isotope compositions may correspond to time scales of the order 10³ –10⁴ years. Thus, although there is a large-scale trend towards lighter stable-isotope compositions with increasing depth, small-scale fluctuations in salinity at shallow level may be the reason why no shallow depth-related trend in ³H can be discerned.

The water sample from the Coal Measures in the borehole near Chester gave δ¹⁸O = –6.9‰ and 8²H = –43‰, suggesting that the basement waters in this borehole are meteoric in origin, like the inland waters sampled by Lucey (1987) from the Permo-Triassic.

The use of Br and Cl data as indices of the origins of salinity

During the crystallisation of halite from a brine, any Br present will partition preferentially into the aqueous phase and the halite will possess a much lower Br/Cl ratio than the brine. This often enables Br/Cl ratios to be used to distinguish groundwater salinity of sea-water origin (Br/Clmolar = 1.54 X 10⁻³) from groundwater salinity originating from halite dissolution (Br/Clmolar < 1.54 X 10⁻³; e.g. Rittenhouse, 1967; Carpenter et al., 1974; Fontes and Matray, 1993).

Mean Br concentrations for a number of groundwater samples from the Cheshire Basin are plotted against corresponding Cl concentrations in (Figure 132), together with similar data from other UK Permo-Triassic aquifers. The fresh waters (from the Rainhill, Halewood and Mickle Trafford boreholes), brackish waters (Stanlow borehole) and saline waters (from the Coal Measures in the borehole near Chester) all have Br/Cl ratios close to those of diluted or evaporated sea water. This requires some explanation since, at face value, these ratios appear to suggest a sea-water origin for much of the groundwater salinity in the basin.

The waters from the Rainhill, Halewood and Mickle Trafford boreholes are all quite shallow (maximum depth of sample 125 m) and a possible explanation for the Br/Cl ratios is the presence of sea spray in the meteoric recharge waters; this is consistent with the fact that these boreholes are located near to the Mersey Estuary. An alternative view is that the Br/Cl ratios reflect a small degree of mixing with marine waters which form a saline incursion from the estuary in this area.

The salinities of the waters from the Stanlow borehole are too high (c.0.1 times that of sea water) for sea spray to explain the measured Br/Cl ratios, and it is more likely that the salinity is due to intrusion of sea water from the Mersey Estuary. The waters from the Stanlow borehole give mean Br/Clmolar = 1.48 X 10⁻³ ± 4.1 X 10⁻⁴ (with the uncertainty quoted at the 1 s level), which is almost identical to that of sea water.

The Br/Cl ratio in the water from the Coal Measures in the borehole near to Chester requires some other explanation. The salinity is too high (about 0.85 times that of sea water) to be explained by sea spray, and saline intrusion from the Mersey Estuary can also be ruled out because the borehole was sited some distance inland and because the oxygen and hydrogen isotopic data suggest a meteoric rather than marine origin. The fact that the salinity of the Coal Measures water is close to that of sea water precludes the possibility that the Coal Measures water acquired its salinity from sea water and its isotopic signature through mixing with dilute meteoric water. Instead, the most likely source of the salinity is halite dissolution in the Permo-Triassic basin fill. This hypothesis is supported by the fact that the water has Na/Cl_molar = 0.97 (sea water has Na/Cl_molar = 0.86).

However, as explained above, a water which had acquired its salinity through evaporite dissolution would be expected to have a Br/Cl ratio significantly less than the sea-water ratio, whereas the Coal Measures water has Br/Cl_molar = 2.02 X 10⁻³, some 30% higher than the sea-water value. A possible explanation is that the water, having acquired its Na-Cl-dominated salinity through halite dissolution in the Permo-Triassic sequence, subsequently gained Br by reaction with organic matter in the Coal Measures. Reactions involving organic matter have been postulated as a possible cause of high Br/Cl_molar ratios of up to 4.08 X 10⁻³ in the Coal Measures of north-east England (Edmunds, 1975).

The Cheshire Basin groundwaters are generally similar to those from other Permo-Triassic aquifers (Figure 132), which plot either near to the sea-water-related trend (consistent with salinity derived from sea water) or significantly to the left of this trend (consistent with salinity originating through halite dissolution). Analyses of waters from the Cheshire Basin produced during brine pumping (Haslam et al., 1950) all plot to the left of the sea-water trend on the Br versus Cl plot (Figure 132), as is to be expected since their salinity undoubtedly originates through halite dissolution. However, these data also show a trend towards lower Br contents with increasing degrees of brine pumping, consistent with the preferential liberation of Br from the halite structure during the first stages of halite dissolution (small quantities of Br can be accommodated in halite).

Relationships between modern groundwaters and fluids responsible for observed diagenetic and ore-mineral assemblages

The present-day groundwaters in the Cheshire Basin are generally fresh. A consequence of this is that diagenetic minerals are being dissolved rather than precipitated. Most of the groundwater samples from the Permo-Triassic sequence contain less than 1000 mg/l TDS, with a maximum salinity of about 20 000 mg/l TDS. This sample set is biased towards fresh waters, since most samples were taken from wells designed to monitor water supplies for domestic consumption, located in areas where MMG is absent, but there is other evidence to suggest that the groundwaters throughout the basin generally have low salinities and never become brines (TDS > 100 000 mg/l): (1) wireline electrical conductivity data suggest that maximum salinities at the base of the Permo-Triassic reach only c.1.6 X sea-water concentrations (pp.180–181); (2) the maximum salinity in inland water samples from the Permo-Triassic aquifers is only c.0.6 X seawater concentration (Lucey, 1987; Tellam, 1995, 1996); and (3) the salinity of the single water sample from the Coal Measures in the borehole near Chester was only 0.85 X seawater concentration. These values are lower than the salinities which might be anticipated in view of the abundance of halite throughout the Cheshire Basin. A halite-equilibrated water at 25°C would have a salinity of c.7.5 X sea water, and other Permo-Triassic basins in the UK with halite-bearing evaporites do in fact contain brines (Figure 132). Indeed, in the Wessex Basin, formation waters in Permo-Triassic aquifers approach halite saturation (e.g. Allen and Holloway, 1984). Thus, the limited salinity data for waters at the base of the Cheshire Basin suggest that the current salinities of formation waters in the basin are much lower than the maximum levels that could plausibly have been reached during diagenesis and are lower than elsewhere. A possible explanation is that the meteoric water has continued to flush the SSG via recharge in the south-east and discharge in the west and north. Periods of lower sea level would assist this process, thus limiting the continued increase in solutes found in other basins.

The available hydrochemical data for the modern ground-waters allow some constraints to be placed on the relationships between formation waters in the basin fill and those in the basement. It is likely that the water from the Coal Measures in the borehole near Chester acquired its salinity through halite dissolution, suggesting that groundwater has passed into the basement from the basin fill: there are no known evaporites in the region except for those in the Permo-Triassic sediments of the Cheshire Basin.

The measured redox potentials (Eh) and trace-element contents of the modern groundwaters allow the chemical characteristics of modern groundwaters and potential ore-forming fluids to be computed. Although, as discussed previously (pp.189–190), Eh measurements must be interpreted with caution, they can give semi-quantitative indications of redox conditions which are useful when comparing different groundwaters. It can be seen from (Table 34) that the modern groundwaters from the Permo-Triassic rocks of the Cheshire Basin are generally oxidising, and the measured Eh values are reasonably consistent with the waters having being influenced by the atmosphere (c.+300 mV). By comparison with (Figure 122) it can be appreciated that these waters lie near to the boundary between the field of tenorite stability and the field in which red-bed brines are capable of transporting Cu⁺-Cl⁻ complexes. Thus the modern groundwaters are not in redox equilibrium with Cu sulphides but would instead tend to oxidise any sulphides in the rock. Although the redox-pH conditions of the modern formation waters from the Permo-Triassic sequence do seem consistent with some possible mobilisation of Cu, the actual measured levels of dissolved Cu are low (<1 mg/l; (Table 34)) and, as expected, these dilute waters are unlikely to give rise to Cu ore deposits. However, the predicted oxidation of primary sulphides and the possibility of limited Cu mobilisation (depending upon the precise redox conditions) is broadly consistent with the oxidised Cu mineral assemblage that is actually found in the deposits Chapter 5). It is also noteworthy that malachite is predicted to be a stable mineral for Pco₂ levels in the range 10^−1.5 – 10^−2.5 bars (Rose, 1976, 1989; and see pp.192–195 above). Although levels of CO₂ in atmosphere-equilibrated waters are lower than this (P_CO2 in the atmosphere is 10^−3.5 bars), such elevated CO₂ pressures can be generated in zones of groundwater recharge due to biogenic activity in soils (e.g. Appelo and Postma, 1994). Consequently, the redistribution of Cu from primary sulphide mineralisation, and its reprecipitation as malachite, can be accommodated within the overall conceptual framework as a relatively recent, telodiagenetic phenomenon.

Similarly, the modern groundwaters will not be in redox equilibrium with either galena or sphalerite and would oxidise these minerals if brought into contact with them (Figure 123). As in the case of Cu, the concentrations of Pb and Zn in the shallow groundwaters from the Permo-Triassic sequence are low (<1 mg/l; (Table 34)) and hence these waters would not be likely to form deposits of Pb or Zn.

In contrast to the shallower waters from the Permo-Triassic basin fill, the groundwater from the Coal Measures in the borehole near Chester gave Eh = −38 mV, consistent with basement waters being more reducing than waters from the cover sequence. This result is to be expected because the Coal Measures have a much larger organic content than the Permo-Triassic strata, and the sample from the Coal Measures was taken from a relatively great depth so that it was more likely to be isolated from the atmosphere than shallower samples. These observations suggest that although the water in the Carboniferous basement probably originated in the Permo-Triassic cover, it has been resident in the Coal Measures for long enough to be reduced through water–rock interactions.

There are no analyses of heavy metals for the groundwater from the Coal Measures in the borehole near Chester. However, the Eh values suggest that the water could be in redox equilibrium with Cu, Pb, Zn and Fe sulphides. The observations suggest that the water is unlikely to be able to transport Cu, Fe or Pb although, given the uncertainties in the measured Eh and any calculated aqueous speciation, it is more likely that the water could mobilise Zn (cf. (Figure 122) and (Figure 123).

Synthesis and discussion

The general characteristics of the Permo-Triassic sediments which fill the Cheshire Basin are typical of red beds (e.g. Coker and Barr, 1975; Burley, 1984). Therefore, from a comparison with generic models of red-bed diagenesis and red beds elsewhere, it seems that diagenetic fluids in the Cheshire Basin probably had widely varying pH and redox conditions (possibly as great as logfO_2,bars from −0.7 to about −50; Eh from about +500 mV to about –500 mV; and pH from about 10 to about 5). Also, these comparative and theoretical investigations suggest a likelihood that the metal-bearing fluids which gave rise to the ore deposits in the basin originated within the Permo-Triassic fill as saline formation waters with TDS up to c.10 X sea-water concentration. By analogy with other red-bed basins, peak temperatures in the Cheshire Basin probably lay within the range 60°–150°C. However, if a late Triassic age is accepted for the Alderley Edge mineralisation, temperatures are likely to have below 100°C over the greater part of the basin. It is clear that reduction of such fluids must occur in order to generate ore deposits. Suitable reducing conditions could occur in organic-rich sediments which are encountered by a red-bed formation water migrating out of the basin, or by mixing with reduced aqueous fluids or hydrocarbon gases. In the Cheshire Basin, the mineralisation is generally related to structural traps where faulting juxtaposes red-bed aquifers against relatively impermeable mudrocks, particularly those in the MMG (Warrington, 1980; Carlon, 1981a, c; Ixer and Vaughan, 1982; Naylor et al., 1989). Possible reductants are: (1) reducing aqueous fluids (e.g. reduced by passing through organic-rich mudrocks), (2) liquid hydrocarbons, and (3) hydrocarbon gases.

The first of these is unlikely to be of major importance within the basin, since most of the mudrocks of the basin are red and lacking in organic matter. Where reduced mudrocks are present, they provided sites for the precipitation of Cu from mobile oxidising groundwaters (see Chapter 5); the evidence does not suggest that significant quantities of mobile, reducing fluids were produced. The possibility that ore formation might be due to migration of hydrocarbon fluids was first suggested by Warrington (1980), while Naylor et al. (1989) postulated that methane derived from the underlying Coal Measures was the reductant. No hydrocarbons have yet been found in fluid inclusions from minerals in the ore deposits, which may suggest that this model is incorrect. However, the theoretical mixing model described on p.198 suggests that any reduced fluid would only need to comprise c.5% of the fluids mixing to form the ore deposits. Therefore, it is possible that hydrocarbons have simply not been detected. Alternatively, the redox conditions may have been imposed by organic compounds, such as light organic acids. These may not have been resolved by the analytical techniques employed in this study or may have been effectively removed by processes such as oxidation during fluid mixing.

Possible pathways for any fluid reductant are either: (1) within the red-bed aquifers themselves; or (2) along the faults which are associated with the ore deposits. If the first option is correct, then clearly the reducing fluid and metal-bearing formation water cannot have migrated together and must have migrated into the structural traps from different directions or at different times. Since baryte, which is an important gangue mineral in the ore deposits, appears to be concentrated near to faults, the second hypothesis is the more likely.

The previously described theoretical models of diagenesis, fluid mixing and sorption allow useful constraints to be placed upon diagenesis and ore genesis in the Cheshire Basin. However, as discussed on pp.195–197 and 198–200, various compromises have to be made when fixing the conditions used for the modelling, in order to make the models as representative as possible of likely natural conditions while taking into account the limitations of the experimental data used in the models, the chemical models themselves, and the computer codes. The models could not simulate precisely the conditions under which diagenesis and ore genesis are thought to have occurred in the Cheshire Basin. In particular, the lack of data for likely eodiagenetic metal-solubility-controlling phases and the inability to model solutions more concentrated than about 1 molal limits the accuracy of the EQ3/6 models, while the HYDRAQL models of sorption cannot simulate realistic temperatures. However, the theoretical modelling has considerable value as a tool for constraining possible explanations for actual field and petrographical observations. For example, the EQ3/6 prediction that ore deposition requires only small amounts of reducing fluid (less than about 5% of reducing fluid mixed with red-bed formation water) supports the postulated origin of mineralisation by the ascent of reducing fluids along faults.

The EQ3/6 model could reproduce some of the major features of the mineral parageneses observed in the Cheshire Basin ore deposits (Chapter 5). Both the model and the observations indicate: (1) hematite formation before the main sulphide-forming event; (2) destabilisation of hematite adjacent to sulphide mineralisation, leading to bleached sandstones; (3) quartz overgrowths; (4) concentration of Pb and Zn ores close to faults, while Cu ore is, in addition, more widely disseminated throughout the sandstone adjacent to the faults; and (5) concentration of baryte close to the faults.

In terms of the model, the spatial distribution of Cu, Pb and Zn sulphides can be explained as a consequence of the redox controls on the relative stabilities of these minerals. Thus, the theoretical prediction that with increasingly reducing conditions Cu sulphides precipitate before Pb sulphides, which in turn precipitate before Zn sulphides (Figure 128), is reflected in the actual mineral distributions relative to the inferred source of reducing fluid, the faults.

Similarly, the concentration of baryte close to the faults is consistent with the model prediction. During diagenesis, the red-bed formation waters in the Cheshire Basin are likely to have had large amounts of dissolved sulphate, owing to the presence of anhydrite in the basin. In such sulphate-bearing waters, Ba is unlikely to have been a significant proportion of the total load, owing to the very low solubility of baryte (e.g. from the EQ3/6 database, at 100°C the reaction BaSO₄ = Ba²⁺ + SO₄²⁻ has log K = –9.51). In contrast, any reducing water from the Carboniferous is unlikely to have contained proportionately greater amounts of sulphate and may therefore have mobilised relatively large quantities of Ba. This can be appreciated in (Table 33), which tabulates the compositions of waters used in the mixing model. The model anhydrite-equilibrated late mesodiagenetic pore water contained 738 mg/l sulphur and only 0.3 mg/l Ba, whereas the model reducing-oilfield water contained only 0.25 mg/l sulphur, but 55 mg/l Ba. Therefore, the model predicted baryte to form only when these two fluids mixed.

The presence of baryte widely distributed throughout the SSG (in addition to its presence concentrated close to faults) is not accounted for by the theoretical model, and may in part be due to the release of Ba into formational fluids during feldspar diagenesis and the precipitation of baryte when these fluids mixed with waters having a high sulphate content derived from evaporite dissolution. Analysis of fluid inclusions in the halite show that the evaporite brines were capable of transporting more than 10 mg/l of Ba. This would provide a supply of Ba throughout the basin (pp.164–169; (Table 28)). Furthermore, comparison with red-bed sequences elsewhere (see pp.189–195) suggests that a wide range of water types may have formed at different places and at different times throughout the Cheshire Basin. Some of these waters may have contained low sulphate concentrations and thus been capable of mobilising significant Ba. This is supported by diagenetic studies (Chapter 5) which suggest that sulphate and chloride brines were generated at different times during the basin's diagenetic history.

The general conclusions of the theoretical EQ3/6 model for ore formation described previously are independent of the nature of the reductant: a similar ore mineral paragenesis would be predicted whatever reductant is used in the model since the red-bed formation waters are the most important source for metals. Similarly, the amounts of reductant which are necessary in order to deplete metals in these formation waters are independent of the reductants which are present. However, although broadly consistent with actual spatial distribution of the ore minerals in the Cheshire Basin deposits, the EQ3/6 prediction that Cu sulphides preceded Pb sulphides and then Zn sulphides, is inconsistent with the apparent temporal sequence of ore minerals in the Cheshire Basin. Petrographic observations reveal a much more complex paragenesis than that predicted by the model, with different mineral parageneses being seen in different ore deposits in the basin (see Chapter 5).

This suggests that simple reduction of a red-bed diagenetic fluid cannot explain the observed sequence of ore mineralisation, and that, instead, mineralisation may have involved several different fluids. However, EQ3/6 and HYDRAQL modelling together suggest a mechanism by which a complex ore-mineral paragenesis can be derived within the framework of a mixing model involving an oxidised fluid and a reducing fluid. The modelling suggests that during progressive diagenesis the chemical speciation of solutes in diagenetic waters, and the ability of solid phases in the rock to sorb dissolved species, would have evolved through time. This could have generated very complex ore-mineral parageneses by controlling the supply of metals to sites of reduction and ore formation.

This may be illustrated by the following example. An early diagenetic, oxidising (Eh = +500 mV) fluid with neutral to alkaline pH (7–8) would probably have been incapable of mobilising significant Cu since it would lie in the field of tenorite stability (Figure 122). Similarly, such a water may not have been capable of mobilising Pb since phosgenite or cerrusite might have been stable in its presence (Figure 123)b, but in contrast, significant aqueous Zn could have been mobilised (Figure 123c). Thus, such a diagenetic water could have acquired only Zn during migration through a red bed, and caused formation of early sphalerite on mixing with a reducing fluid. The lowering of pH from about 8–9 to 5, as is plausible in progressive diagenesis (see pp.189–191) might be expected to release first Pb to solution and then Cu (in the form of CuCl⁺) if the solution remains more oxidising than about +500 mV (cf. (Figure 122) and (Figure 123). However, although not directly applicable to red-bed ore formation, since they were calculated only for a temperature of 25°C, the sorption curves in (Figure 129) suggest that Cu²+ would be released to solution before Pb²+, since the latter would remain sorbed for longer than the former.

As discussed previously, these redox conditions are unrealistic, except for early diagenetic fluids, and it is more likely that diagenesis occurs mostly in the presence of a more reducing fluid, with an Eh in the range –200 mV to +500 mV. If, under these conditions, the pH of a red-bed formation water fell as before, then it is possible that Cu+ (which may behave like Ag+) would desorb before Zn²+, which would in turn desorb before Pb²+ (Figure 129). Thus, as in the original EQ3/6 prediction (p.198), mixing between this red-bed formation water and a reducing fluid might initially produce Cu sulphides, but these would be followed by Zn sulphides and then Pb sulphides, in contrast to the previous EQ3/6 model. Additionally, a fall in pH might not be necessary in order to produce this effect, since progressive diagenesis might be expected to liberate the metal cations progressively owing to the increasing crystallinity of sorbing Fe oxides. This can be appreciated from (Figure 129) which illustrates how, as the surface area of the sorbing phase decreases, the sorption edges shift towards higher pH (cf. (Figure 129)b and (Figure 129)c.

This assessment suggests that, although current models are not sufficiently refined to reproduce the observed ore parageneses exactly, mixing between only two basic fluids, a relatively oxidised red-bed formation water and a reducing fluid, could generate a complex ore-mineral paragenesis. Hence, it is not necessary to invoke the action of many different fluid types in order to explained the observed ore parageneses in the Cheshire Basin.

Although chemical data for modern groundwaters are limited, and none of these waters are apparently related to eodiagenetic and mesodiagenetic fluids, the general validity of modelling ore formation by mixing of two fluids with different redox states is supported by the measured Eh values. The waters from the Permo-Triassic aquifers are significantly more oxidising than the water from the underlying Coal Measures (Figure 122); (Table 34)). If the measured Eh values are indeed representative of in situ redox conditions, and if the water in the Coal Measures did indeed originate in the Permo-Triassic fill of the basin, then the implication is that this latter water has acquired its redox state during water-rock interactions within the Coal Measures. This supports the hypothesis that ancient fluids resident in these two sequences may have had a similar redox contrast.

If hydraulic connectivity between the cover and basement (as inferred from the limited data for Coal Measures water) occurred across the entire Cheshire Basin, then any reducing fluid which was present in the basement initially has probably been displaced. In this case, it can be hypothesised that such initial reducing fluid may have ascended along faults, giving rise to ore formation. The fluid-flow model on pp.180–189, in which dense brines descend in the middle of the basin before rising near the margins, suggests a mechanism by which such a circulation system might be established. However, this cannot be proven in the absence of more widely distributed data for deep basement waters beneath the basin. Since the deep borehole near Chester was located near the margin of the basin, it is possible that the waters sampled from the Coal Measures there were recharged from the basin margin, and that in fact fluid flow was the reverse of that in the ore genetic mixing model. Unfortunately, at present there is insufficient knowledge of the deep hydrogeology of the basin to decide which, if either, of these ideas is correct.

Conclusions

The main conclusions of the study of the current basinal groundwater flow system, the palaeohydrogeology and the role of fluids in the diagenesis and emplacement of minerals can be summarised as follows:

• The movement of groundwater and the hydrogeochemical interactions that occur during the deposition, diagenesis and development of a sedimentary basin are key to the movement, concentration and emplacement of mineral and hydrocarbon deposits. Because of the importance of Eh-pH conditions as controls on metal mobilisation, red-bed formation waters may evolve from metal-transporting fluids to metal-precipitating fluids and back again as they migrate through a red-bed sequence.
• Flow in the present-day groundwater system in the Cheshire Basin, in the areas not underlying the MMG, is controlled by topographic variations in the water table in the multiple-aquifer system of the Permo-Triassic sandstone. The aquifers have high porosity and permeability, especially in the near-surface zone where cement between sand grains has been removed by dissolution. Flow is, however, largely controlled by fracture flow, and fracture density has a large influence on local aquifer transmissivity, particularly at depth.
• The active freshwater system gains recharge at outcrop or via Quaternary drift cover. In places, this extensive drift cover is sandy in nature, stores significant quantities of groundwater, and is in hydraulic continuity with the underlying sandstones. Elsewhere, the drift comprises clays and till, which are relatively impermeable and, where they exceed 2 m in thickness they can restrict recharge to an estimated 2% of potential.
• Natural groundwater discharge is to the rivers that drain the area, namely the Mersey, the Dee, the Severn and their tributaries. Exploitation of groundwater has modified this system considerably around major conurbations such as Liverpool and Manchester and to the north of the Mersey, where large cones of depression have developed and induced saline intrusion and upconing of indigenous saline waters at depths of a few hundred metres.
• Where the SSG underlies the MMG, the aquifer has not been exploited and has only been investigated in a few deep boreholes. Available evidence was used to constrain a two-dimensional vertical cross-sectional model using the SUTRA computer code. Geological parameters and boundary conditions were either selected from data measured within the Cheshire Basin or inferred from similar geological environments.
• Initial runs of the model assumed that the fluid in the basin was fresh, and the driving force was demonstrated to be the topography of the water table. The effects of temperature and salinity variations were also considered. The results of modelling different scenarios illustrated that the distribution of brine sources has a marked effect on the circulation of brines in a basinal environment.
• In the central part of Cheshire Basin, the extensive halite beds near the base of the MMG, overlying the SSG, could act as a source of brines, the density of which will drive the groundwater flow system. Even at low relative densities (1.050) this density-driven flow seems to be able to dominate the fresh-water topographically driven flow. However, the density and viscosity reduction resulting from the temperature increase towards the base of the basin fill may be sufficient to cause thermal perturbations of the flow fields.
• The current conceptual model of groundwater flow in the Cheshire Basin is one dominated by density-driven cells moving downwards from several sources within the MMG cover and mixing with fresher water already in the SSG or being introduced by gravity-driven flow. Minor thermal perturbations may also play a part near the base of the aquifer, but the general downward movement is supported by the low heat flow measured.
• The outlet of this groundwater circulation system is probably around the margins of the basin into the active, freshwater circulation system, where it is diluted, or directly into the sea. One possible surface manifestation of discharge of this saline groundwater is the brine spring at Aldersey.
• Palaeo-hydrogeological reconstruction suggests that the Cheshire Basin would have been filled with highly concentrated brines during Permian and Triassic times. With alternating inundation and emergence during the Jurassic and Cretaceous periods, shallow fresh groundwater systems would have been periodically established and then replaced by encroaching saline water. The limited chemical and geophysical evidence for the salinities of deep waters in the basin suggests that any early diagenetic brines have been flushed away by less dense post-Triassic waters.
• The chemical dataset of groundwater analyses in the Cheshire Basin is biased towards the chemistries of dilute waters, because almost all samples have been taken from sites designed to monitor potable groundwater supplies. The maximum salinities shown by water samples from the basin are approximately equivalent to sea-water concentration. However, it is possible that some brine could still be present at depth.
• Since the last major inundation of the Cheshire Basin in the Cretaceous, uplift and erosion have established shallow, fresh groundwater systems superimposed on deeper saline and brine groundwater systems, particularly where the arenaceous aquifers are not confined by the MMG. These systems would have evolved to the present system with removal of Cretaceous and Jurassic sediments and, more recently, with the influences of glaciation, changing sea levels and changing coastlines.
• The groundwaters from the modern shallow groundwater system in the Permo-Triassic strata are oxidising and would oxidise any ore-forming sulphide minerals with
• which they came into contact. These waters can be considered to be telodiagenetic waters which are moving towards equilibrium with their host rocks mainly by mineral dissolution reactions.
• The chemistry of a single groundwater sample from the Coal Measures beneath the margin of the Cheshire Basin suggests that in the region of the Cheshire Basin there is hydraulic connectivity between the basement and the fill. This sample from the Coal Measures is relatively reducing and has a redox state consistent with sulphide stability.
• Diagenesis of the red-bed sediments within the Cheshire Basin is likely to have involved chemically heterogeneous waters. By comparison with evidence from red-bed sequences elsewhere, diagenesis of the Cheshire Basin red beds could have involved large ranges of redox and pH (broadly, logfO_2,bars from –0.7 to about –50; and Eh from mV . to about +500Mv to about –200 mV (although Eh may reach about –500 mV in certain low-temperature alkaline environments); and pH from about 10 to about 5).
• Theoretical models are able to simulate the major mineralogical transformations that occur during red-bed diagenesis and ore formation. These models can reproduce the main features of ore-mineral parageneses in the Cheshire Basin.
• Theoretical models suggest that extremely small amounts of fluid mixing can exert a critical control on both diagenesis and ore formation in red beds such as those of the Cheshire Basin. The addition of less that 5 parts of a reducing, hydrocarbon-bearing water to 95 parts of red-bed formation water will result in the precipitation of all the Cu in the latter.
• Such a mixing model for ore formation in the Cheshire Basin is consistent with models for palaeo-fluid flow in the basin, in which waters descend to considerable depths near the middle of the basin, migrate to the basin's margins, and then ascend marginal faults.
• Sorption, and chemical variations in waters during progressive diagenesis, may possibly explain the complexity of actual ore-mineral parageneses in the ore deposits of the Cheshire Basin. However, current theoretical models are not sufficiently advanced to reproduce this complexity.

Chapter 7 Resources of the basin: base metals, industrial mineral, hydrocarbons and groundwater

J A Plant, D F Ball, A D Bradley, R A Chadwick, D J Evans, D G Jones, G A Kirby, A E R A Nicholson, T J Shepherd, N J P Smith, G Warrington, and A A Wilson

Red beds play two important roles in the occurrence of sedimentary-basin resources: firstly, they act as a source of heavy metals, which are bound in the solid phases or dissolved in metalliferous formation waters; and, secondly, owing to their high mean porosities and permeabilities, they act as conduits for groundwater flow during basin evolution and may become reservoirs for mineralising fluids, hydrocarbons or groundwater. Red beds may also act as chemical conditioning agents for groundwaters passing through them.

Red beds thus play a fundamental part in the genesis and the spatial distribution of natural resources, and they are important hosts of hydrocarbon accumulations and deposits of metalliferous minerals worldwide. For example, sediment-hosted Cu deposits in red-bed sequences account for about 25% of the world's Cu production and reserves (Kirkham, 1989), exceeded in importance only by porphyry Cu deposits. Red-bed sandstones are also important sources of uranium and account for more than 75% of U production in the USA (OECD, 1986). Furthermore, single red-bed formations may control the distribution of a wide variety of economic resources. The Permo-Triassic red beds of the UK host: major oil and gas reservoirs (North Sea, Wessex and East Irish Sea basins); Cu-Pb-Zn mineralisation (Cheshire Basin); and commercially valuable evaporites, including sylvite (Cleveland Basin), gypsum and anhydrite (Cheshire, Cumbria and Nottinghamshire) and halite (Cleveland and Cheshire basins). Additionally, these red beds may act as major sources of potable water (Cheshire, Lancashire and Nottinghamshire), and as geothermal reservoirs (Wessex Basin).

The resources of the Cheshire Basin are considered under the headings Base-metal mineralisation, Industrial minerals, Hydrocarbon prospectivity, and Groundwater.

Base-metal mineralisation

In this section we first consider the principal characteristics and structural and stratigraphical setting of the known mineralisation in the Cheshire Basin and review previous models for its genesis. Ore deposit models associated with fluid flow in red-bed sedimentary basins generally are then considered, and a model is developed for the Cheshire Basin mineralisation based on the new evidence presented here. Finally, the prospectivity of the Cheshire Basin and comparable settings elsewhere in the UK for red-bed Cu deposits and other types of mineralisation potentially associated with Permo-Triassic rifted basins are discussed.

Structural and stratigraphical setting

A full account of the structural and stratigraphical setting of the base-metal mineralisation in the Cheshire Basin is given in Chapters 2 and 3 and in Warrington (1994b), and only a brief summary is given here.

The mineralisation comprises disseminated cement and fracture-hosted baryte and more localised polymetallic mineralisation, hosted in formations of the SSG. Several generations of baryte occur, mainly near to fault zones but also as pore-filling cements in fluvial or aeolian arenaceous horizons (Thompson, 1985, 1989). Baryte is particularly widespread near the base of the Helsby Sandstone Formation and in the upper part of the Wilmslow Sandstone Formation around the basin margins.

The polymetallic mineralisation comprises mainly Cu with some Pb sulphides, together with secondary Cu and Pb minerals and minor amounts of Zn, Ag, Co, V, Ni, As, Sb and Mn; traces of Au, Hg and Se also occur at Clive and Alderley Edge. The mineralisation occurs mainly in the upper part of the SSG where it is faulted against, or capped by, the now relatively impermeable argillaceous rocks of the overlying MMG. The size and form of the deposits reflect structural controls, and the sedimentary facies of the host rocks. The main locations of the mineralisation are shown in (Figure 133).

Alderley Edge

At Alderley Edge the mineralisation occurs in SSG rocks dipping south-west at 10 to 15°, in faulted contact with younger formations of the MMG across the Kirkleyditch Fault in the east, and the Alderley Fault in the west (Figure 134). The MMG formerly capped the sandstone, constraining fluid movement and forming a potential trap for hydrocarbons and ore fluids (Warrington, 1980).

The stratigraphical succession at Alderley Edge is shown in (Table 35). Widespread Cu mineralisation with Pb and Zn occurs adjacent to fault zones in both fluvial and aeolian sandstones of the uppermost Wilmslow Sandstone Formation and the Engine Vein Conglomerate, Wood Mine Conglomerate and West Mine Sandstone in the Helsby Sandstone Formation.

The characteristics of the ore bodies at successively higher levels in the host rocks can be examined in some 12 km of old mine workings, which form three geographical groupings (Figure 136) are identified by the appropriate letter." data-name="images/P1000389.jpg">(Figure 135).

The eastern group, comprising the Engine Vein Mine and Stormy Point workings, in the uppermost Wilmslow Sandstone Formation and basal Helsby Sandstone Formation (Engine Vein Conglomerate) (Warrington 1965, 1980) follow north-west-trending faults (Figure 134). They were probably worked mainly for Pb, although disseminated Cu was stoped in Engine Vein Mine (Figure 136)a. The remains of the ore bodies are indicated by bleached sandstones. The mineralised zone extends c.15 m into the footwall, consistent with fluid migration away from the fault, constrained up-dip by the fault and down-dip below south-west-dipping aquicludes.

The central group of workings in the Wood Mine Conglomerate is more extensive and irregular (Figure 136) are identified by the appropriate letter." data-name="images/P1000389.jpg">(Figure 135), with more stratiform ore bodies; it is bounded to the north and south by north-west-trending normal faults, and is also cut by a small fault with similar trend (Warrington, 1965, map 3; 1980). The fault at the northern margin of Wood Mine is the western extension of the fault which cuts the lower Helsby Sandstone Formation and the Wilmslow Sandstone Formation in Engine Vein Mine. The mineralisation, which is disseminated in fluvial conglomeratic sandstones, occurs up-dip of the southern boundary fault in the Wood Mine Conglomerate. The form of the ore bodies is complex, probably reflecting interaction of stratigraphical and structural controls and the presence of aquicludes, as well as the distribution of early diagenetic carbonate and sulphate cements in the host rock sequence. In the north of the mine (Figure 136)b an ore body about 20 m wide extended south, normal to the fault, for about 60 m into the footwall (Warrington, 1965, map 3) and smaller ore bodies occurred up-dip from the middle fault. The ore bodies in Wood Mine have convex fronts (Warrington, 1965, fig. 1 b), consistent with fluid flow away from faults. Some of the ore bodies are flat and confined to a single fluvial cycle whereas others extend through several cycles.

In West Mine, complex deposits hosted entirely in the West Mine Sandstone were worked opencast and in the subsurface over an area of 600 m by 200 m (Figure 136) are identified by the appropriate letter." data-name="images/P1000389.jpg">(Figure 135) The workings intersect several north-west-trending normal faults, which downthrow to the north-east. The workings are fault-bounded to the south-west; a complex ore body extended throughout almost the full thickness of the West Mine Sandstone in the hanging wall immediately up-dip of the fault (Figure 136)c–e. Mineral fabrics (Chapter 5) are consistent with migration of mineralising fluids away from the faults.

Clive

At Clive, in the south of the Cheshire Basin (Figure 133) and (Figure 137), the ore body occurs in the Grinshill Sandstone Formation adjacent to a normal north-north-east-trending fault (Figure 136)f and (Figure 137). The mine workings total about 0.9 km in length. The ore body is in the upper 10 m of the Grinshill Sandstone Formation immediately below the Tarporley Siltstone Formation and extends about 250 m along the fault and up to 8 m downdip into the hanging wall. Corresponding beds up-dip in the footwall were unproductive. The fault is highly silicified.

Other sites

The nature of the mineralisation at other sites in the Cheshire Basin (Figure 133) is inferred from documentary evidence, geological maps and comparisons with relationships at the Alderley Edge and Clive mines.

The Mottram St Andrew site immediately east of the Alderley block is adjacent to the Kirkleyditch Fault where the Helsby Sandstone Formation is thrown down east against the older Wilmslow Sandstone Formation (Figure 134). Mining was along a north-west-trending fault comparable to that at Engine Vein Mine, Alderley Edge.

At Bickerton in the west of the basin, mineralisation is associated with a north-east-trending normal fault which downthrows the MMG east against the SSG; the mineralisation is hosted by the Wilmslow Sandstone Formation in the footwall. The ore body was thin and narrow (Carlon, 1981a).

At Eardiston in the south-west of the Cheshire Basin, a north-north-east-trending normal fault downthrows east-dipping MMG beds west against the upper SSG. The fault is truncated to the north and south by sub-parallel east–west faults (Carlon, 1981b). MMG deposits overlying the SSG east of the main fault, and those in the hanging wall, created a trap for migrating fluids.

The Pim Hill site is on a north–south fault to the west of a small graben (Figure 137), where the Tarporley Siltstone Formation and underlying Grinshill Sandstone Formation are downthrown east against the Wilmslow Sandstone Formation. The geology of the Yorton site, which is also on a north–south fault which downthrows MMG beds east against the SSG, is similar (Figure 137).

At Wixhill mineralisation is associated with a north-east-trending normal fault, which downthrows the Bollin Mudstone Formation north-west against the Wilmslow Sandstone Formation (Figure 137). The mineralisation is likely to be in the fault and in the Wilmslow Sandstone Formation in the footwall.

Other mineralisation near Redcastle and Hawkstone (Figure 137) is also associated with faults.

Copper carbonate has been worked at Bearstone (Gibson, 1925, p.64) in beds assigned to the Helsby Sandstone Formation which dip north-west and interdigitate with the Tarporley Siltstone Formation; baryte and Cu mineralisation is visible at [SJ SJ 7220 3925] (Rees and Wilson, 1998). The mineralisation is in a small graben normal to the north-east-trending Wem and Hodnet faults.

Principal features of the known polymetallic mineralisation

Common features of the polymetallic mineralisation in the Cheshire Basin are as follows:

• The deposits are hosted in formations of the SSG, usually the Helsby Sandstone Formation but sometimes the Wilmslow Sandstone Formation, in stratigraphical or structural proximity to the MMG in either the footwall or hangingwall of faults.
• Deposits occur mostly in traps formed by the interaction of faults and aquicludes.
• The MMG appears to have formed the main aquiclude, but mudstone beds in the Helsby Sandstone Formation are also important locally.
• There is evidence that the faults associated with the poly-metallic mineralisation had been fluid pathways earlier in basin evolution. Deposition of phases such as quartz and chalcedony reduced the porosity around faults, providing a baffle against lateral fluid flow but probably enhancing fluid flow along faults during ore formation. More importantly, porosity was also reduced by the presence of evaporite cements (anhydrite and possibly halite) which petrographic evidence suggests were present at the time of mineralisation (Chapter 5).
• The most important known ore bodies, such as those of Alderley Edge and Clive, are in areas associated with fault structures normal to the direction of maximum (east–west) extension (see Chapter 3); whereas occurrences, such as that at Bickerton, associated with north-east-trending strike-slip faults, are relatively minor.
• The ore bodies display features consistent with fluid migration away from faults, with lateral flow constrained by cementation below aquicludes.
• Aeolian rocks host the largest, laterally extensive ore bodies; small, irregular discontinuous ore bodies occur in fluvial rocks adjacent to minor aquicludes.

Previous models of ore genesis for the known mineral deposits of the Cheshire Basin

The origin of the base-metal mineralisation in the Cheshire Basin, especially that at Alderley Edge, has been discussed extensively in the literature (Warrington, 1994a). Early workers, including Hull (1864) and Dewey and Eastwood (1925), favoured a syngenetic origin, with detrital Pb and Cu sulphides deposited contemporaneously with the Helsby Sandstone Formation host rocks. Subsequently Taylor et al. (1963) and Warrington (1965) proposed epigenetic models with late hydrothermal fluids (of Cenozoic and Jurassic age, respectively) moving up the faults; Warrington suggesting that a concealed Jurassic granite was the source of ore fluids. A model proposed by King (1968) stressed the importance of precipitation from downward-percolating metalliferous groundwater. Warrington (1980) later advocated deposition from intrastratal brines which migrated into structural traps; metallic ions were envisaged as partly leached from sediments through which the brines had passed, with precipitation of the base metals possibly being promoted by hydrocarbons. Additional, or alternative, sources of metallic ions may have been fluids of more deep-seated origin Similar models have been proposed by most workers subsequently. Holmes et al. (1983) proposed a diagenetic origin for the mineralising fluids and suggested that the ore solutions were oxidising basinal brines of moderate salinity with trace metals released during diagenesis of the Permo-Triassic sediments within the basin. A basin brine expulsion model was also favoured by Naylor et al. (1989), with precipitation as a result of interaction with methane from the underlying Coal Measures. In contrast, Thompson (1985) suggested that the metals may have originated in the black shales of the underlying Coal Measures.

The occurrence of the mineralisation beneath impermeable horizons at several locations in the basin has been noted by several authors (Warrington, 1965, 1980; Carlon, 1975; Thompson, 1985). Prior to erosion, the Tarporley Siltstone Formation provided a regional seal, indicating that the predominant direction of ore-forming fluids was up basin-margin fault systems. At present, the evaporite-bearing MMG lies topographically below the mineralised horizons in the SSG. Naylor et al. (1989) used this evidence, together with sulphur-isotope data for the baryte mineralisation which is consistent with an evaporitic source in the MMG, to suggest that mineralisation occurred during and after Tertiary inversion of the basin when sulphate-bearing fluids rose through the fault systems.

Most authors now propose ore deposition beneath the MMG, involving reductants such as (1) reducing aqueous fluids, (2) liquid hydrocarbons or (3) hydrocarbon gases. The possibility that ore deposition might be due to migration of hydrocarbons was first suggested by Warrington (1980), while Naylor et al. (1989) postulated that methane derived from the underlying Coal Measures was the reductant.

The nature and timing of the ore-forming processes is important not only in understanding the genesis of the relatively minor known occurrences of the Cheshire Basin, but also in developing exploration criteria for red-bed Cu and other ore-deposit types of possible economic significance in the Cheshire Basin region, elsewhere in the UK and in other parts of the world.

Models for ore deposits associated with fluid flow in red-bed sedimentary basins

General observations

In their classic paper of 1981, Gustafson and Williams proposed that most sediment-hosted stratiform deposits of both Cu and Pb-Zn are part of a continuum of ore deposits formed as a result of the migration of diagenetic brines from sedimentary basins to sites of deposition. They attributed the formation of Cu deposits to cool sulphate-rich brines migrating up dip to reducing sites of deposition at an early stage of basin evolution, whereas Pb-Zn deposits were thought to form from higher-temperature, more-evolved reducing fluids from greater depth in basins.

The diagenetic basin model for both Cu and Pb-Zn sediment-hosted deposits, which derives from the classic work of authors such as White (1958) and Beales and Jackson (1966), has generally replaced the syngenetic models of Schneiderhohn (1932) and Garlick (1961 and 1989) and the magmatic epigenetic models of, for example, Lindgren (1933) and Sales (1959).

Recently, Brown (1993) has described the features characteristic of sediment-hosted stratiform Cu deposits (SCDs) as follows:

1. the presence of an important Cu zone; exceptionally with abundant Pb and Zn, and sometimes with economically significant Ag and Co;
2. the common location in, or associated with, rift basins filled with continental red beds ± bimodal volcanic strata;
3. the peneconformable, stratiform nature of the Cu zone (including both ore-grade and subeconomic zones);
4. the uniform lateral continuity of mineralisation along bedding, suggesting (erroneously) a sedimentary origin;
5. the predominance of fine-grained, disseminated sulphides, typically in layers following host rock stratification;
6. the zoned distribution of metals and ore minerals;
7. the presence of host sediments that, in the ore zone, have been prepared syndiagenetically with reducing agents and abundant sulphur;
8. a major thickness of permeable, coarse-grained, red-bed clastic sediments in the immediate footwall of the Cu zone;
9. a temporal and spatial association of host rocks with strata formed in warm arid climates (evaporitic units, red beds);
10. the post-sedimentary, diagenetic introduction of Cu; and the deposition of Cu from aqueous, chloride-rich solutions across the redoxcline between footwall red beds and reduced, sulphide-bearing grey beds of the host strata.
11. the deposition of Cu from aqueous, chloride-rich solutions across the redoxcline between footwall red beds and reduced, sulphide-bearing grey beds of the host strata.

Some principal characteristics of SCDs are shown in (Table 36).

The common association of SCDs with rift basins has become widely recognised only in recent years. The spatial association of many SCDs with continental rifts and the early emplacement of Cu in rift-fill sediments have been used to suggest that the deposits are fundamentally related to the normal development and evolution of continental rift basins at low latitudes (Kirkham, 1989).

Examples and economic significance

Although few in number, economic occurrences of SCDs ((Table 36)) are among the most important (and most sought after) sources of Cu. Their consistent grades and lateral continuity make them highly attractive exploration targets. In addition to Cu, they provide significant amounts of Co (Central Africa), Pb (Poland) and Ag (the USA and Poland). Some deposits contain resources of Au, U, platinum-group elements (PGEs) and rare-earth elements as by-products. SCDs have recently accounted for approximately 10% of the western world's production of Cu, 30% of its Co, and increasing proportions of its Ag (British Geological Survey, 1995).

General model

The general (or diagenetic-overprint) model for sedimentary Cu deposits is considered to be a two-stage process (Lovering, 1963) involving (1) a chemically reducing, commonly carbonaceous and pyritic, grey sediment initially enriched in sulphur (iron sulphide and/or gypsum/anhydrite) by primary, syndiagenetic processes; with (2) Cu ± associated metals overprinted during a later influx of ore fluids from coarse-grained, highly porous and permeable, continental red-bed sediments (Figure 138). Metal deposition is considered to be a low-temperature chemical reaction between the abundant reduced sulphur and oxidised saline red-bed pore fluids carrying soluble metal-chloride complexes. The distinctive contact between the red and grey beds, or fossil redoxcline, is usually an important exploration target.

Copper-bearing sulphides (e.g. chalcocite, digenite, bornite and chalcopyrite) and other base-metal sulphides (e.g. galena and sphalerite) occur in well-defined zones away from the redoxcline. Chalcocite and digenite (the least soluble) are precipitated close to the redoxcline, whereas bornite and chalcopyrite are deposited progressively further away, toward the original pyritic zone. A hematitic zone immediately upstream of the Cu deposits reflects the influx of oxidising ore solutions.

The diagenetic-overprint model suggests that the timing and mode of mineralisation is related more closely to hydraulic solution flow along aquifers during normal basin diagenesis than to flow along major epigenetic structures, and in some deposits it is suggested that overprinted mineralisation occurred in poorly consolidated sediments immediately following sedimentation. In other deposits there is evidence that ore-forming solutions were focused along large, crosscutting structures which may post-date lithification. The most probable source of the ore-forming solutions is considered to be red-bed basin fill in continental rifts.

Red beds are important as a source of heavy metals, and their high mean porosity and permeability enable them to act as conduits for fluid flow during basin evolution and to become reservoirs for mineralising fluids, hydrocarbons or groundwater. They can also act as chemical conditioning agents for migrating groundwaters.

Continental red beds include fluvial, lacustrine, paralic and aeolian sediments, deposited in rapidly subsiding, synsedimentary, fault-bounded basins (e.g. Turner, 1980; Frostick and Reid, 1987). They may also form along continental margins or the edges of mountain belts (Walker, 1976; Turner, 1980). Their deposition in continental settings allows them to remain in oxidising, normally alkaline, diagenetic conditions for sufficient time for unstable ferromagnesian minerals, such as olivine, pyroxene and amphibole, to break down, producing iron oxyhydroxides and hematite which give red beds their characteristic colour. With progressive diagenesis, mineral dissolution releases K, Fe, Mg, Al and other metals, including ore-forming elements, into pore waters.

Metals (essentially Cu, with variable amounts of Ag, Pb, Zn, Co, PGE and Mo, depending on the composition of the red-bed basin fill and the chemistry of solid-fluid interaction) are thought to be carried as metal-chloride complexes which can attain elevated concentrations in warm, oxidised chloride-rich, sulphide-poor brines.

In the classic diagenetic-overprint model, basin-scale fluid circulation is considered to reflect high heat flow associated with rifting, possibly aided by magmatic heat. Gravity-driven convection is also considered to play a role in the formation of some ore deposits.

Sulphur is thought to be concentrated initially as fine-grained, disseminated eodiagenetic iron sulphides in organic-rich sediments and/or as evaporitic sulphates (gypsum/ anhydrite, in some cases in the ore forming brines) which are subsequently reduced to sulphide. Reduced sulphur may also be present in salinas and sabkas, or provided by migrating hydrocarbons.

There is a clear association of SCDs with continental red beds formed in hot arid climates in which evaporites commonly occur in the host section. Kirkham (1989) emphasised that SCD mineralisation is associated with evaporites formed within 30^° of the palaeoequator.

Classification of SCDs

Red-bed sedimentary Cu deposits of this general group have been subdivided by Eckstrand (1984) into paralic marine (Kupferschiefer type) and continental (Red-bed type). The former group includes the high-grade deposits (1–5% Cu) which range up to 500 million tonnes, with Co an important by-product in Zambia and Zaire; they are typically deposited at the base of major reducing marine transgressive units overlying red beds.

In the type localities in Poland and Germany, Kupferschiefer mineralisation is at the contact between Lower Permian volcanics and red beds and Upper Permian marine carbonates and evaporites. Recent studies indicate that the mineralisation is epigenetic and deposited from latediagenetic convecting fluids (Jowett, 1986, 1989).

Some Kuperschiefer-type SCDs are associated with PGE enrichment. At the Lubin and Polkawice mines in Poland, Pt-rich (>10 ppm) shales have been mapped along strike lengths in excess of 1.5 km (Kucha, 1982). Maximum values of >200 ppm Pt were found over 50 m strike lengths along the contact between white, oxidised sandstone, and black, reduced shale containing thucholite, ketones, kerogen, phenols, tertiary alcohols and aromatic hydrocarbons. The precious metals are thought to have been concentrated in the shale during late-diagenetic flow, by auto-oxidation and desulphurisation of the organic matter, with PGE perhaps acting as catalysts (Kucha, 1982). The Coronation Hill deposit in the Northern Territory of Australia also contains PGEs and U (Needham and Stuart-Smith, 1987).

Continental red-bed-type ores are smaller (c.1–10 million tonnes) and of lower grade (1–2% Cu). They are associated with anoxic fluvial or lacustrine rocks overlying or interbedded with red beds (Eckstrand, 1984).

Ore-deposit model for the Cheshire Basin

The main characteristics of the known ore deposits of the Cheshire Basin are most closely related to continental redbed-type sedimentary Cu deposits, according to the classification of mineral deposit types of Eckstrand (1984) and Plant and Tarney (1994).

A model for these ores is considered here on the basis of the detailed geochemical, thermodynamic and palaeohydrogeological data in the preceding chapters and other new evidence presented in this report.

The structural and stratigraphical evolution of the Cheshire Basin, its geochemistry, diagenetic history and palaeohydrology, and the results of fluid flow and geochemical modelling are generally consistent with a modified diagenetic-brine-expulsion model for the known ore deposits of the basin. The source of the Cu and associated metals is considered to have been the red-bed sedimentary fill of the basin; and mineralogical and petrological data are consistent with the alteration of ferromagnesian minerals to hematite and other secondary phases in situ. The basin fill comprises sedimentary rocks derived from a mixture of igneous, sedimentary and metamorphic sources (Chapter 4; Turner, 1980) originating from unradiogenic Gondwana basement to the south and from more local sources (i.e. North Wales, south Pennines). The contents of ferromagnesian minerals and their contained levels of Cu and other ore-forming elements (OFEs) were probably lower than in red-bed basins in the USA and Mexico, reflecting differences in source rocks (Turner, 1980; Walker 1967; Walker et al., 1978) – especially in comparison with the Michigan Basin which contains a high proportion of altered basalt (e.g. Seasor and Brown, 1989).

The contents of OFEs in the Cheshire Basin are low compared with average upper continental crust and average sandstone levels (Table 37). The extent to which the low levels are primary or reflect removal of OFEs in migrating fluids has a considerable bearing on models of ore genesis. Petrographical studies of the SSG (Chapter 5) have shown that these sandstones contained ferromagnesian minerals at the time of deposition, and it is not unreasonable to suppose that the levels of OFEs in the sediments were in the normal ranges for sandstones. Most of the detrital matter that formed the MMG was already oxidised at the time of deposition, with OFEs associated with ferric oxides and oxyhydroxides, and OFE levels may also have been in the normal ranges for organic-free argillaceous rocks.

If the present low levels of Cu and other OFEs in the Cheshire Basin reflect their removal in mineralising brines, then the total amount available for mineralisation is a function of the volume and density of the basin fill. The total volume of MMG and SSG strata preserved in the basin is estimated, from the interpreted seismic reflection data, as 1300 km³ and 3100 km³ respectively. Making an allowance for erosion, the original depositional volume of the SSG was probably of the order of 3300 km³. The ratio of preserved MMG/SSG thicknesses then gives an original depositional volume of about 2000 km³ for the MMG. Data for the preserved total content of OFEs in the basin and totals corrected for erosion are given in (Table 38), based on average densities of 2402 kgm⁻³ (SSG) and 2335 kgm⁻³ (MMG).

If the basin fill originally contained average upper-continental-crust levels of metals, then 173 Mt of Cu, 84 Mt of Pb and 481 Mt of Zn has been leached from the SSG and 60 Mt of Cu, 63 Mt of Pb and 84 Mt of Zn from the MMG. However, if the original contents of OFEs in the SSG were at average sandstone levels, then only 30–150 Mt of Cu, 8 Mt of Pb and 132–326 Mt of Zn would have been available. Whilst it is obviously not possible to establish the original concentrations of OFEs in the basin it is clear that the basin fill was potentially a significant source of OFEs.

The recorded ouput of ore at Alderley Edge mines after 1857 was about 250 000 tons (Warrington, 1981), from which 3500 tons of Cu metal may have been recovered, assuming an average Cu content of 1.4%. This represents <0.002% of the total Cu which would have been removed from the basin fill, assuming an original concentration close to that of average upper continental crust. The percentages for Pb and Zn are much greater than those of Cu, although much less Pb and Zn is present in the ores.

Metals in red beds are often sorbed from early syndepositional and eodiagenetic porewaters onto Fe oxides during early diagenesis and may be released during recrystallisation as a result of later burial diagenesis (Zielinski et al., 1983, 1986). However, the bulk rock may remain closed to metal migration. The lower values of OFEs in the SSG at present might reflect preferential removal of Cu and Zn by fluids passing through the basin, with greater volumes flowing through the more porous and permeable SSG than moving through the finer-grained MMG. The work of Zielinski et al. (1983) suggests that similar amounts of Pb and Zn would have been in leachable sites (associated with Fe oxides, Mn oxides and clays) but that approximately twice the amount of Cu would have been in such positions.

Sorption and desorption may be very important processes in controlling the release of metals to fluids (Chapter 6). Because Cu can occur as Cut and Cu²⁺, whereas Pb and Zn occur only as divalent ions, redox has a particularly important influence on Cu retention or release. Metcalfe et al. (1994a, and in Chapter 6) suggest that Cu⁺ may behave like Ag⁺ and be weakly sorbed only at neutral to acid pH. This could explain the preferential removal of Cu relative to Pb and Zn into ore fluids and hence its much greater importance in the known ores of the Cheshire Basin.

Early diagenesis in the SSG involved the breakdown of detrital ferromagnesian minerals and the reprecipitation of Fe in ferric oxyhydroxides, followed by infiltration of detrital clay into sandstones and the deposition of low-ferroan dolocretes and calcretes in conditions of high pH and Eh. Sulphur-isotope data suggest that sulphate was derived mainly from brines of Triassic marine origin, although higher in the sequence sulphate is isotopically lighter, possibly reflecting an increased continental component or derivation from Permian sources. During the late phases of eodiagenesis, reduction by bacteria of Fe³+ and SO₄²⁻ led to the production of reduced green beds and reduction spots with authigenic pyrite in fine-grained fluvial-cycle sediments and with corrensite deposition in hypersaline sections of the MMG.

The known ore deposits of the Cheshire Basin appear to have formed during the late stages of mesodiagenesis; the rocks had been lithified and well cemented prior to major fracturing associated with reactivation of synsedimentary fault systems. During earlier phases of SSG mesodiagenesis, early carbonate cements (calcite and dolomite) and evaporites (halite and anhydrite) together with silica were remobilised. Evidence in the MMG is consistent with early diagenetic dissolution of halite, prior to the development of early mesodiagenetic dolomite. It is suggested that much of the SSG was cemented by anhydrite (and/or possibly) halite prior to ore mineralisation (Chapter 5). This petrographical evidence, together with sulphur and carbon isotope data (which are consistent with major recycling of primary evaporitic phases from the MMG within the basin) and fluid-inclusion data (which indicate the high content of OFEs in evaporitic bitterns from the MMG) are all consistent with the results of fluid-flow modelling (Chapter 6).

All the evidence suggests that chloride and sulphate cements and vein mineralisation in the Cheshire Basin fill were derived by fluxes of fluid from the MMG into the SSG, initially as a result of compaction and gravity-driven brine circulation. Fluid-flow models indicate that a high-density brine from such a source would have descended through the basin under the influence of gravity and induced fluid flow towards the basin margins where the known ore deposits occur. Such a fluid could also have mobilised and transported large quantities of OFEs as chloride complexes in the near- neutral, oxidising, low-temperature (70–80°C), high-salinity (up to 9 wt % NaCl equivalent) conditions. Diagenetic evidence is consistent with early anhydrite-saturated fluids fluxing through the SSG, followed by later low-sulphate NaCl brines derived from the MMG. Baryte is thought to have been deposited when BaCl₂ carried in the later low-sulphate brine reacted with earlier-formed anhydrite cement, probably during several different phases of fluid flow in the basin. The MMG thus appears to have generated different fluids at different times.

The new evidence presented here indicates that mineralisation in the known sulphide ore deposits occurred when high-density metalliferous brines from the MMG fluxed through the basin under the influence of gravity and encountered a small volume of reducing fluid in fault systems around the margins of the basin, especially in the north (Chapters 5 and 6). Some of the OFE contents of the fluids could have been picked up during their passage through the SSG, and some may have been introduced with the deeper, reducing fluids. The extent to which the reducing fluid was derived from the East Irish Sea Basin to the north-west or the Coal Measures sequences beneath the northern Cheshire Basin is unclear, but structural considerations favour the latter source (Chapter 3). There is no direct evidence for the migration of hydrocarbons into the SSG before or during base-metal mineralisation, so models favouring the involvement of light hydrocarbon gases remain unproven. Host-rock fabrics in Alderley Edge and Clive mines indicate that mineralising solutions were introduced along faults and related fractures, with progressive corrosive dissolution and replacement of the pre-existing tight cement rather than by simple pore filling in a porous sandstone. The mineralisation appears to have been complex and episodic, probably reflecting repeated episodes of faulting and fracture reactivation in the basin. Complex interaction occurred between an ore fluid sourced in the MMG, transporting and mobilising OFEs as it flowed down and outwards towards the basin margin fault systems, and small components of a reducing phase, possibly derived by thermobaric dewatering and organic maturation in deep Carboniferous basins, migrating in relation to fault movement and seismic pumping. In terms of fluid sources, it is suggested that in the north-west of the basin the greater importance of the reducing fluid precluded Cu mineralisation. Comparison of the distribution of authigenic Cu sulphides with the variation in δ¹³C of late-mesodiagenetic carbonates contemporaneous with ore deposition indicates that in areas with carbonates with δ¹³C_PDB < –8‰ the deeper-sourced reducing fluid inhibited Cu mobilisation and mineralisation (Figure 95). The –8‰ δ¹³C_PDB contour for the carbonates appears to provide a useful tool for delineating areas of potential Cu mineralisation, where values exceed –8‰, from zones of dominantly Fe sulphides, where δ¹³C_PDB is less than –8‰.

This new model for the Cheshire Basin ore deposits, summarised in (Table 39), has considerable implications for understanding red-bed type SCDs and perhaps SCDs generally.

One of the factors affecting the prospectivity of the basin is the timing of ore-forming events. The evidence presented here suggests that mineralisation cannot be simply related to the flow of diagenetic fluids in relation to progressive basin evolution, as suggested, for example, by Holmes et al. (1983), but was related instead to late tectonism in the basin associated with reactivation of basin-bounding fault systems after cementation and lithification had occurred. Regional considerations, particularly evidence from more fully preserved sedimentary sequences to the south of the basin, indicate that extensional basin subsidence continued to early Jurassic times with renewed extension in the late Jurassic and early Cretaceous associated with sea-floor spreading in the southern part of the North Atlantic region. It seems most likely that the known mineralisation in the Cheshire Basin occurred at some time during the late Triassic (post-MMG times) to early Jurassic period of extension. Such a model has considerable implications for Cu exploration in the Cheshire Basin and in other Permo-Triassic basins in the UK.

Prospectivity

Evidence for the extraction of large quantities of metals from the Cheshire Basin fill by the migration of pervasive high-density brines, variously enriched in sulphate and chloride as a result of interaction with evaporitic phases especially in the MMG, .has considerable implications for exploration for SCDs in the Cheshire Basin and elsewhere. This is particularly the case if mineralisation occurred during the late Triassic to early Jurassic period of crustal extension or subsequently. The MMG appears to have acted as a major source of ligands and metals for the known ore deposits, as well as being the dominant control on ore-fluid migration, and there is thought to be the potential for the discovery of further small-tonnage high-grade Cu deposits beneath the MMG. The presence of aeolian sediments beneath an aquiclude (normally MMG shales), associated with intersections between faults normal to the direction of maximum extension and those subparallel to the extension direction and sealed by a secondary cement, provide the best combination of exploration criteria for the known red-bed Cu-ore deposits of the Cheshire Basin.

Marked Cu anomalies, sometimes with coincident Pb and Zn anomalies, are apparent on BGS regional geochemical data over the area. In addition to those associated with the known mineralisation at Alderley Edge and Bickerton, anomalies also occur near Frodsham, Tarporley and Baschurch (although other anomalies, such as those near Manchester and Ellesmere Port, almost certainly represent man-made contamination). Extensive Ba anomalies also occur over Alderley Edge and throughout the area, extending from the Peckforton Hills northwards to Helsby in the west of the basin, including the areas of the known mineralisation at Bickerton, and along the eastern margin of the basin south of Crewe.

The relationship between trace Cu phases and carbon-isotope data for associated mesodiagenetic carbonates (the distribution of the –8‰ contour) in the Cheshire Basin (Figure 95) suggests that there is limited potential for Cu mineralisation in the north-west of the basin. In the southwest of the basin, the lack of Westphalian rocks in the basement, which are thought to have been the source of the reducing phase involved in ore deposition, may also limit the potential for continental red-bed-type Cu deposits. The potential for PGE or Au concentrations in the basin has not been tested, although PGEs and Au would have been readily mobilised in the chloride-rich basinal brines and traces of Au have been found in this study at West Mine, Alderley Edge (Chapter 5). The concentrations of precious metals in red-bed ore deposits generally probably reflects the original contents and types of ferromagnesian minerals in the basin fill as well as the compositions of migrating fluids.

Some of the evidence presented here suggests that Cu mineralisation occurred in the Cheshire Basin during the late Triassic to early Jurassic phase of crustal extension. The Blue Anchor Formation, Penarth Group and Lias Group of the Prees outlier in the south-east of the basin, and rocks of similar age elsewhere in central and southern Britain, comprise dark mudstones and limestones which could have provided a reducing horizon capable of precipitating sulphide minerals from ore solutions originating in the red-beds beneath. Hence, if ore formation occurred following deposition of the Lower Jurassic, more extensive SCD deposits of Kupferschiefer, rather than continental-red-bed type, may have formed near the top of the red-bed sequence. In the Cheshire Basin the Blue Anchor Formation, Penarth Group and Lias Group rocks were largely eroded during the Tertiary. Elsewhere, however, they are present from north-east England to Devon in the south-west, overlying red beds of the MMG. Although no significant Cu mineralisation has been recorded at this stratigraphical level, the model suggests the potential for such mineralisation associated with the contact between red beds and the reducing cover rocks.

Numerous minor occurrences of epigenetic Cu occur in the Permo-Triassic areas of England. At Middleton Tyas, secondarily enriched Cu ores occur in Carboniferous host rocks (Wadge et at, 1982; Dunham and Wilson, 1985). The Lower Permian Marl Slate of north-east England (Hirst and Dunham, 1963) and offshore (Haslam, 1982) is the lateral equivalent of the Kupferschiefer. It is enriched in Cu as well as Pb, Zn, Mo, Ni, Co and V, and is probably consistent with new models for epigenetic ore deposition in the Kupferschiefer regions of eastern Europe (Kucha, 1990; Jowett, 1991; Cathles et al., 1993). Small Cu shows are also known from the Worcester Basin (N J P Smith, unpublished report) and similar diagenetic histories to those of the Cheshire Basin have been recorded also in the Wessex Basin (Milodowski et at, 1986; Strong and Milodowski, 1987) and around the margins of the East Irish Sea Basin (Strong et al., 1994). Mineralisation offers a possible explanation for a zone of anomalous values for Cu and other elements in stream sediments extending over Permo-Triassic rocks from Teesside to York (British Geological Survey, 1996); the diagenetic history of the Permo-Triassic rocks of east Yorkshire and Lincolnshire (Milodowski et al., 1987) is similar to that of the Cheshire Basin. Hence it is considered that there may be the potential for the discovery of significant SCDs in the Permo-Triassic rocks on and offshore of the UK where they are overlain by reducing marine sequences.

The other main types of ore deposit potentially associated with the Cheshire Basin include unconformity-type deposits. These are particularly important for OFEs such as Cu and U which are deposited at redox boundaries. The greatest potential for such deposits would be at sites where oxidising hydrothermal brines circulating in the basin have encountered reducing conditions, for example near to the contact between the red beds and the Coal Measures in the north of the basin. BGS regional geochemical data show Cu anomalies close to the contact between the Permo-Triassic and the underlying Westphalian immediately to the south-east of the Cheshire Basin, in the adjoining Stafford Basin. Dewey and Eastwood (1925) recorded Cu mineralisation from the Kidderminster Formation at Shore Hill, Huntingdon Quarry and in Fair Oak Colliery shafts near Cannock. In the geological record such deposits generally occur in sedimentary basins of mid-Proterozoic age, however.

There may also be some potential for roll-front U (V-MoSe) deposits, although the regions (such as the Pennines, Lake District and North Wales) that formed upland areas likely to have supplied the oxidising meteoric fluid during the telediagenetic phase in the evolution of the Cheshire Basin contain few evolved acid volcanics or granites to provide a source of U, in contrast to the environs of the classic U deposits of Wyoming and the Colorado Plateau in the western USA.

Industrial minerals

The Cheshire Basin holds significant resources of industrial minerals. Glacial sand, which is largely devoid of gravel, is an important source of construction sand, for use in concrete, mortar and asphalt, and is also one of the principal sources of silica sand in Britain. Deposits in the Chelford area are the sole source of sand for the manufacture of flat glass and those near Congleton are valued as foundry sands on account of their particle-size distribution and purity. Boulder clay is extracted for brickmaking near Warrington, and the SSG is worked for construction fill near Warrington and as a building stone at Grinshill. However, the principal industrial-mineral resource of the basin is halite, of which it is a major producer.

Halite

The quantity of halite in Cheshire has been estimated at some 28 cubic miles (117 km³) (Pugh, 1960), and forms a vast future resource. The industry is thought to have started in Roman times near Northwich and continued afterwards in Northwich and Nantwich. Salt was produced by evaporation in open pans in wych houses, and the early salt towns feature in the Domesday Book. In the Middle Ages, Middlewich also became a centre, but gradually Northwich became preeminent due to the proximity of the navigable River Weaver. Underground mining followed an early discovery of halite in an auger hole at Marbury, near Northwich, in 1670. A large group of shallow mines was sunk nearby in the 'Top' and 'Bottom' beds of the Northwich Halite. Some became flooded and the resultant brine was pumped, causing erosion of the pillars of support and some spectacular subsidences (Ward, 1898; Calvert, 1915). The last mine in the Northwich area, the Adelaide Mine, was abandoned in 1928.

Initially natural (wild) brine was pumped. Sixty-four small companies amalgamated in 1888 to form the Salt Union, which later became part of ICI, and most of the brine was sent by pipeline to Weston, near Runcorn, where large vacuum evaporators were installed in 1911. The wild brine pumping caused accelerated subsidence over wide areas, in linear and crater patterns. The main centres of wild brine extraction were at Winsford and Northwich, sited on the Northwich Halite. Wild brine was also formerly extracted from the Northwich Halite in the Heatley–Agden Saltfield near Lymm, and at Lawton [SJ 8050 5727]. Extensive brining was formerly undertaken in the lower part of the Wilkesley Halite, near Sandbach, Elworth and Middlewich. The only wild brine currently produced is at New Cheshire salt works at Wincham, near Northwich, from the lower part of the Northwich Halite.

Almost all brine extraction is now by controlled pumping in the Holford Brinefield [SJ 700 730] of ICI and the Warmingham Brinefield [SJ 696 613] of British Salt. All the brine comes from the Northwich Halite Formation and is from cavities some 100 m across and up to 170 m in height formed by a triple tube system involving water, compressed air and a brine outlet (Notholt and Highley, 1973; Pass, 1993). The shape of the cavity is monitored by sonar surveys. Both sites are under a cover of Byley Mudstones, overlying the halite.

Salt mining at Meadowbank Mine, Winsford began in 1844, but the mine was shut down from 1892 to 1928. Since 1928 Meadowbank has been expanded into the major producer of UK rock salt for winter road maintenance (the largest use) and for sugar-beet fertiliser. Mining is currently all in the 'Bottom Bed' of the Northwich Halite in galleries 7.5 m in height. Annual production varies with the winter climate and averages some 1.1 to 1.2 million tonnes. The Meadowbank Mine is operated by the newly formed Salt Union Ltd, which acquired the mine from ICI.

In recent years white salt production from brine, all of which comes from Cheshire, has averaged 1.3 to 1.4 Mt/y, whilst salt-in-brine, used by the chemical industry as a feedstock for the manufacture of soda ash, chlorine and caustic soda, has averaged 3 Mt/y. The sole extractor of chemical industry feedstock is ICI. When the above production figures are combined with those of the mine, there is an annual total of some 5.5 Mt/y salt production in Cheshire.

The future resources are vast, notably in the virtually untapped area of the Prees Syncline where both the Northwich and Wilkesley halites are present. This area includes a probable eastern extension of the Wilkesley Halite outcrop towards Betley, near Stoke-on-Trent, with the likelihood there of Northwich Halite at depth. There is also potential for formation of deep cavities for storage of gas or fluids, in the deeper parts of the syncline.

Hydrocarbon prospectivity

This prospectivity study is based on released information and an earlier account (Smith et al., 1995). Most of the analytical work on the project concentrated on the Permo-Triassic rocks, since the primary aim was to provide a metallogenic model for the Cu-Pb-Ba mineralisation. The prospectivity of the Carboniferous rocks is based on work done over a number of years on the Pennine Basin. Because there is a dearth of organic matter within the Permo-Triassic red beds, samples had to be collected outside the Cheshire Basin, from Carboniferous rocks (see pp.169–172). During the course of the work it has become apparent that there are important differences between the hydrocarbon prospectivity of the Cheshire Basin and its three adjacent hydrocarbon provinces, the East Irish Sea Basin, the (East Midlands) Pennine Basin and the southern North Sea. All share the fundamental property that hydrocarbons were generated from Carboniferous rocks.

As described earlier in this volume, the Cheshire Basin contains Permo-Triassic red beds (described in detail in Chapter 2) unconformably overlying Carboniferous Coal Measures (summarised on pp.64–66). Red beds are important hydrocarbon reservoirs in many parts of the world. In the UK, Permo-Triassic rocks are important hydrocarbon reservoirs in the North Sea, Wessex and East Irish Sea basins. Carboniferous rocks provide important oil and gas-prone source rocks in the East Irish Sea, southern North Sea, East Midlands and Europe generally, and provide reservoirs in the East Midlands. Carboniferous reservoirs are not proven in the Cheshire Basin, however, and there are no mature Permo-Triassic or younger source rocks in that area or in the East Midlands. Producing fields from Triassic reservoirs in basins adjacent to Cheshire, at Formby and in the East Irish Sea, together with oil shows in the Needwood Basin prove that oil and gas have been generated and have migrated, even into

Recent sedimentary deposits (Formby). In the East Irish Sea Basin the reservoir for all the hydrocarbons is the Ormskirk Sandstone Formation (equivalent to the Helsby Sandstone Formation), sealed by MMG halites and mudstones. The older sandstones of the SSG and the Collyhurst Sandstone are not productive, despite the latter being sealed by the Manchester Marls. The key to the hydrocarbon prospectivity lies in resolving whether migration from the Carboniferous source rocks into Permo-Triassic rocks has occurred, or the extent to which hydrocarbons are trapped within Carboniferous rocks.

Hydrocarbon occurrences

There are only about a dozen recorded hydrocarbon shows in Permo-Triassic rocks of the Cheshire Basin sensu stricto, and none of these contains live oil. There are, however, limited developments of oil sands in the Triassic of the nearby Needwood Basin and commercial oil and gas production from Permo-Triassic rocks of the East Irish Sea Basin. In addition, the Formby Oilfield produced from a Pleistocene sand in one well and from Triassic rocks in the other wells (Lees and Taitt, 1946). In contrast, hydrocarbon shows in Carboniferous rocks surrounding the Cheshire Basin and beneath the Permo-Triassic rocks are commonplace and include live oil (Figure 139).

Permo-Triassic shows

In the Cheshire Basin, the Little Ness Borehole (Figure 139) drilled on the Wem Fault, was found to contain hydrocarbons. Minor oil stains were reported in the Ashton and Ellesmere Port 6 boreholes. Redwood (1922) referred to a water well at Anderton having become contaminated with petroleum in 1888, and also mentioned that 'jets of inflammable gas which occasionally occur in the neighbouring salt mines tend to establish the genuineness and importance of the find', but these records could not be substantiated in the BGS archive. The Mere Brow Borehole encountered bitumen in the Collyhurst Sandstone, and minor hydrocarbon traces are reported in the Mickle Trafford Borehole, and at the Grinshill and Alderley Edge copper mines in the Cheshire Basin (Chapter 5).

The Thornton and Saughall Massie boreholes (in part of the West Lancashire Sub-basin, the onshore extension of the East Irish Sea Basin) also contain minor hydrocarbon shows. The Burton Point oil show is on the Irish Sea coast at the mouth of the Dee Estuary. These shows occur in locally porous horizons and predate carbonate cement dissolution. Their presence may be related to nearby large faults, in the case of the Mickle Trafford and Little Ness boreholes, and the close proximity to the main area of Carboniferous subsurface shows in the north of the basin for most of the others. An oil seep in sandstone just to the south of Liverpool city centre has recently been reported (Harriman and Miles, 1995); it has a carbon isotope composition similar to that of the Burton Point oil, 3‰ heavier than that from Formby.

On the south-east margin of the nearby Needwood Basin, bitumen-impregnated Bromsgrove Sandstone (equivalent to the Helsby Sandstone of the Cheshire Basin) was noted by Fox-Strangways (1905). Ickes (1923) reported obtaining 2.5% petroleum by the use of solvents, with Carboniferous oil characteristics. The bituminous horizon was traced in the area between Winshill and Bladon near the north-north-east-trending basin margin fault.

Good oil seeps were rediscovered in 1937 at Formby in the West Lancashire Sub-basin, in Recent peat overlying Tarporley Siltstones (Cope, 1939). The Formby Oilfield is very shallow, with the top reservoir at 53 m in the discovery well. A Carboniferous source and mother pool were suspected but not found, despite drilling five deep boreholes (Falcon and Kent, 1960). Formby lies above the subsurface part of the inverted Craven Basin. Interpretation of seismic reflection data shows that Dinantian rocks form the subcrop (Figure 140) stretching north-east from the seeps. Northwestward migration of oil, sourced from the Bowland Shales, could be envisaged towards this structural high. The basin depocentre, against the Pendle Fault, probably formed the oil kitchen, rather than an area to the north, as suggested by Falcon and Kent (1960), or migration from west of the Formby Point Fault (Trueblood et al., 1995). Several north–south Permo-Triassic faults have tapped the migrating oil and, with a very thin MMG seal, seepage to the surface has occurred. The Formby Oilfield produced 71 560 barrels of oil. This oilfield should be considered as the first discovery in the East Irish Sea Basin and highlights how the interplay between Carboniferous source-rock maturity, Triassic reservoirs, seals and faulting may be the key to further successes.

Carboniferous shows

Carboniferous rocks in all the coalfields surrounding the Cheshire Basin have fair to good shows of hydrocarbons, with live oil, and are comparable in this respect to the Nottinghamshire–Yorkshire Coalfield. Many of the coal seams are gassy, attracting two of the first three UK coal-bed methane exploration boreholes. In the Flintshire Coalfield, at Point of Ayr Colliery, a methane drainage scheme was required to remove gas from underground workings. Similarly, at Parkside Colliery in the Lancashire Coalfield, gas was drained and sold to a Warrington factory. At Wolstanton Colliery, in the North Staffordshire Coalfield, excess methane was sold to local potteries.

Live oil has been encountered in the Coalbrookdale, Lancashire and North Staffordshire coalfields (Strahan, 1920). In some instances this may be interpreted as migrating oil intercepted by mining, although in the Coalbrookdale area oil-bearing sandstones probably have a local cannel coal source. The oil has migrated to the extremity of the Carboniferous basin (Longmynd–Shelve area). If there are similar sources beneath the Cheshire Basin, migration north-westwards from Wem–Wilkesley and north-east from Crewe–Knutsford would be expected.

In Cheshire, weak hydrocarbon shows were recorded in the Blacon East Borehole, in Namurian sandstones. The Churton Borehole, in the Milton Green Inlier, had oil shows in the Erbistock and Coed-yr-Allt formations (Table 40), suggesting that the Ruabon Marl seal is breached in the Inlier. The saline waters (see pp.229–230) might also reflect this. More than a dozen boreholes drilled by the National Coal Board in the concealed part of the Lancashire Coalfield encountered shows of gas and oil. Some of the shows are in the red beds (Westphalian C-D), but most are in the Productive Coal Measures sandstones. Philips Bridge Borehole, south of Stockport, had an oil show in Westphalian A rocks.

In Clwyd, Redwood (1922) reported oil in water wells at Rosehill and Erbistock Lodge, about a kilometre apart, and the borehole at Overton Bridge had oil shows in the Coedyr-Allt and Ruabon Marl formations. Small amounts of gas and oil were encountered during shaft and tunnel building in the lead–zinc mines of north-east Wales (e.g. Halkyn Mine; Smith, 1921).

In Lancashire, many good oil shows were encountered during mining in the Lancashire Coalfield (Pinfold, 1958). In the Wigan area, oil seeps were common within the mines, for example at Sovereign Pit at Westleigh, Lea Green, Moss Hall, Pemberton, Bridgewater and Golborne Collieries (Strahan, 1920). These seeps may result from this area being a Westphalian palaeo-high, towards which migration could have occurred over a long period of time. Many boreholes to the south-west of the Lancashire Coalfield have also encountered oil shows (Figure 139). Cronton A3/5 Borehole penetrated a 3 m oil sand in the Upper Coal Measures (Pinfold, 1958).

Gas seeps have been known for centuries, from the celebrated burning (water) wells in the Wigan area. They are also common in the mines (especially Parkside Colliery) and caused a blow-out in a National Coal Board drilling rig at Tanhouse (Pinfold, 1958).

In Shropshire, the Coalport Tar Tunnel briefly produced 4500 gallons of oil per week shortly after intercepting an accumulation in the Coalport Formation in 1786. By 1842, however, only three barrels per year were sold (Brown and Trinder, 1979). Numerous other nearby localities also experienced seepages (e.g. Lincoln Hill, Broseley, Priorslee, Dawley, Coalport Bridge, Tarbatch Dingle), indicating that migration into the Westphalian D Coalport Formation sandstones had occurred over a considerable distance. D'Arcy Exploration Company drilled three unsuccessful boreholes at Coalport. Other shows (e.g. in the Pitchford Hall well, the West Shropshire Orefield mines and the Shrewsbury Coalfield oil in the west, and those south-west of the present area covered by Westphalian rocks) demonstrate that oil migration reached well beyond the limit of the Productive Coalfield source rocks. Selley (1992, fig. 6), in a cross-section, showed the relationship of some of the seeps to the basal Westphalian unconformity.

In Staffordshire, oil-stained Namurian rocks were found where the Red Rock Fault crosses the river Dane and at Dingle Brook near Bosley in Arnsbergian mudstones (Evans et al., 1968, pp.37 and 176). BP's borehole at Bosley was dry. Oil was also encountered in many of the mines in the North Staffordshire Coalfield (Meir Hay, Fair Lady, Hem Heath, Mossfield and Wolstanton).

Nearly all the early exploration wells drilled to appraise these shows were drilled north-east of the Coalfield, at Werrington, Gun Hill and Nooks Farm, on large surface anticlines, spudded into Namurian rocks. Nooks Farm Borehole encountered sub-economic gas in the latest Dinantian Onecote Sandstone. More recent drilling has been within or to the south-west of the Coalfield.

Source rocks and hydrocarbon generation

Source rocks

Potential hydrocarbon source rocks include Dinantian (Brigantian) basinal shales, early Namurian shales (Holywell Shales), Westphalian oil shales and cannel coals (oil-prone) and bituminous coals (gas-prone), all of which outcrop on the periphery of the Cheshire Basin. Some of these are likely to exist at depth under the Cheshire Basin. They are well-known source rocks in other oil provinces: early Namurian shales (Edale Shales) are considered to be the source rocks of the East Midlands oilfields and the Holywell Shales are the source rocks for the southern East Irish Sea oilfields (Trueblood et al., 1995). The Formby Oilfield overlies Bowland Shales, which subcrop beneath Permian rocks.

Oil was produced at Leeswood in the Flint Coalfield by heating the oil shales and cannel coals. In the Flint Coalfield, curly cannel yielded 80 gallons per ton, smooth cannel 35 gallons and oil shale 33 gallons (Wedd and King, 1924). Cannel at Clanway Colliery (North Staffordshire) yielded 60–70 gallons of oil per ton (Gibson, 1925). These Westphalian source rocks can reasonably be expected to underlie the Cheshire Basin between the North Staffordshire, Lancashire and North Wales coalfields. Such terrestrial coaly source rocks produce waxy oils and could conceivably be the source of the Hem Heath and Coalport oil (see below). The latter may have migrated south-west towards the margin of the Carboniferous basin.

Hardman et al. (1993) attribute the East Irish Sea and Formby oils to sources in the Sabden Shales, by analysis of material from a west Lancashire well. Although there are sour gases and oil in the East Irish Sea, in addition to the sweet gas of the Morecambe Gasfields, Hardman et al. (1993) favour derivation from the same source. Similarly, oil in the southern part of the basin is sourced from the Holywell Shales (Trueblood et al., 1995), which, beneath the Douglas Oilfield, is within the oil window (Trueblood et al., 1995). The oil is 44° API, slightly more mature than the Formby oil.

As part of the present investigations, hydrocarbon seeps were sampled from localities in and around the Cheshire Basin (pp.169–173). Precambrian samples were from the Shrewsbury area, where oil from the basal Westphalian has seeped into the underlying Longmyndian rocks. The Coalport Tar Tunnel sample is from Westphalian rocks a short distance above Silurian rocks. The Snailbeach Pb mine yielded bitumen from veins, thought to be Triassic in age, in Ordovician rocks.

The Row Brook bituminous breccia contains 4.5% organic carbon. The breccia and sandstone are very porous, despite overgrowths on quartz grains (Parnell, 1983). Five of the Triassic samples are from within the Cheshire Basin, Burton Point being adjacent to the East Irish Sea Basin. The sample from well 110/7-2 is a show in the Permian Collyhurst Sandstone. The Ecton sample is from Dinantian rocks at the Waterbank Mine dumps. The River Dane sample is from Triassic rocks close to the show recorded by Evans et al. (1968). The Hem Heath oil is from a sample collected in the coal mine. The five most biodegraded samples, without recognisable alkane patterns, are from the Westphalian and pre-Westphalian rocks in the Shrewsbury area (Tar Tunnel, Pitchford Hall, Row Brook, Snailbeach and Haughmond Hill).

The data (Table 29) suggest derivation from two different sources. The low Pri/Phy ratios of Burton Point, 110/7-2 (East Irish Sea Basin) and Mickle Trafford may reflect marine (Holywell Shale) source rocks. These data support the results of the carbon isotope composition of the late carbonate cement (pp.140–150), indicating an influx of fluids from mature Carboniferous rocks. The chromatograms of these samples (see (Figure 106) suggest the presence of two separate phases within the organic matter. The light oil may belong to a second oil-generation event.

A different source, within the Coal Measures, is indicated by the high Pri/Phy ratios of the Hem Heath, Little Ness and River Dane samples.

High Pri/C17 and Phy/C18 ratios (>1.0) generally indicate immaturity. Since most of the samples appear to be thermally mature, and as all are from the surface or shallow boreholes, it is likely that they represent migration from depth.

Geothermal gradients

Present-day geothermal gradients and temperatures

The present-day geothermal gradient in the basin is low, at about 20°C/km (Burley et al., 1984), and there is no evidence for significantly higher values in Permian to Tertiary times. Temperatures in the Lancashire coal mines gave an average of 20–22°C/km, although Rose Bridge Colliery was higher, at 38°C/km (Graham, 1922; Verma, 1981). The geothermal gradient in the North Staffordshire Coalfield is 37°C/km (Verma, 1981).

The average surface temperature is approximately 10°C and the average heat flow is calculated at 52 mW/m². Thermal modelling of the basin using the BGS HOTPOT software suggests that the basin is relatively cool, with mean temperatures ranging from 20°C at the base of the MMG to 37°C at the base of the Permo-Triassic (p.83). The highest temperatures at the base of the Permo-Triassic succession, in the deeper parts of the basin, locally exceed 70°C in parts of the Wem–Audlem Sub-basin and 65°C in the Sandbach–Knutsford Sub-basin.

Carboniferous geothermal gradients

There is some evidence from the East Irish Sea Basin that geothermal gradients were significantly higher in Carboniferous times. Roberts (1989, fig. 3B) showed that the assumed Mesozoic and Permian gradient (related to the present-day geothermal gradient) is lower than that in the Carboniferous rocks of well 112/30-1. In addition, there is a downhole variation in maturity from low gradients in post-Carboniferous rocks to high gradients in Carboniferous rocks in well 110/11-1. This may result from a geothermal gradient which was significantly higher during the Carboniferous. Hamilton Oil Company have, however, invoked 20 000 feet (6000 m) of Variscan uplift to explain this feature (Trueblood et al., 1995). The total preserved thickness of the Carboniferous in the Quadrant 109 Syncline in the East Irish Sea was estimated at 6250 m by Jackson et al. (1987).

Coal rank data from the Keele 1 Borehole in the North Staffordshire Coalfield (Millott et al., 1946) suggest that there may also have been variations in the geothermal gradient during the Carboniferous, with higher gradients in the early Westphalian than in late Westphalian times. A depth-maturity plot of Keele 1 (Figure 141) reveals by calculation that the geothermal gradient may have been as high as 81°C/km in early Westphalian times, declining to 29–45°C/km in late Westphalian times. The heat flow is therefore also thought to have been higher in early Westphalian times, as the thermal conductivities of the Westphalian rocks are relatively constant.

The coal rank maturity data (volatile matter percentages) in Keele 1 have been converted to equivalent vitrinite reflectance values to compare them with values from the Knutsford Borehole (Figure 142). Westphalian D rocks in Knutsford are significantly more mature than the highest Westphalian rocks in Keele 1, just leaving the oil window. If a similar amount of uplift has been experienced by both sites, this implies a geothermal gradient of approximately 34°C/km since Permian times, somewhat higher than at present. If, however, the geothermal gradient has been approximately constant at its present level since Permian times, differences in the amount of Variscan uplift would be suggested.

Evidence for lateral variations in geothermal gradients comes from coal rank maturity data from North Staffordshire, which suggest that palaeogeothermal gradients during Westphalian times were higher in this area than in the Lancashire and East Midlands coalfields, perhaps due to its closer proximity to Westphalian volcanism.

Burial history and hydrocarbon generation

Carboniferous source rocks in this region have undergone two phases of burial, firstly during Carboniferous basin formation and secondly during Permo-Triassic and later burial. Thermal modelling suggests that both phases led to sufficiently high temperatures in potential source rocks for hydrocarbon generation, although migration pathways would have been quite different.

Thermal modelling of the Permo-Triassic fill of the Cheshire Basin suggests that maximum temperatures were attained during the period of maximum burial in Palaeocene times (Chapter 3, this volume). Severe reductions in average porosities of Permo-Triassic sandstones, resulting from this burial, are likely towards the basin centre (see p.231). Average temperatures at the base of the Permo-Triassic were 95°C, with peak temperatures in the deepest part of the basin of 140°C (see (Figure 65)d. The destruction of garnets in the Cheshire Basin provides independent evidence of temperatures higher than 80°C, whereas their preservation in the Little Hay Borehole on the South Staffordshire High suggests that here these temperatures were not attained (Chapter 4, this volume). The lower part of the Permian sequence entered the oil-maturation window towards the end of the Triassic and left it during late Palaeogene uplift. Peak maturity was reached in early Palaeogene times, corresponding to the maximum depth of burial. Much of the Permian sequence was in the oil-maturation zone at this time, with some parts being at temperatures almost high enough for gas maturation. The base of the SSG was in the oil-generating window from early Jurassic to mid or late Palaeogene times, and the base of the MMG from late Jurassic to mid Palaeogene times (Chapter 3, this volume).

Modelling of underlying Carboniferous rocks is more problematical, as their maturity at the end of Carboniferous basin development is difficult to estimate. Carboniferous strata generally possess a maturity, due to Carboniferous burial and Variscan tectonics, which is markedly higher than post-Carboniferous rocks. Even if this observation is disregarded, the Carboniferous rocks would have become mature earlier than Permo-Triassic strata because of their greater depth. They probably reached the oil window in mid to late Triassic times, and they remained in the oil-generation zone for longer than post-Carboniferous rocks, probably until early Neogene times. They would also have attained higher levels of maturity, well into the gas-generating window, at maximum depth of burial in early Palaeogene times.

Hydrocarbon generation is therefore thought to have begun during Carboniferous times, from source rocks in the Pennine Basin depocentre which lay to the north of the area now occupied by the Cheshire Basin. It is likely that some of these hydrocarbons migrated south-westwards, away from the Pennine depocentre, towards the margins of the Carboniferous basin, into the area now underlying the Cheshire Basin. As subsidence of the Permo-Triassic Cheshire Basin proceeded it is likely that the underlying Carboniferous sediments, hitherto immature, also started to generate hydrocarbons. This is likely to have produced oil on the margins of the Cheshire Basin and gas towards the centre. The fate of these hydrocarbons, however, is difficult to elucidate.

Migration and migration pathways

Deep saline fluids, in the Collyhurst Sandstone and in the Coal Measures, have been recorded from geophysical log interpretation, from samples from a deep borehole near Chester, and from a 11 000 mg/l saline spring at Aldersey (Milton Green Inlier) (Chapter 6). Waters of the Aldersey type are present throughout the Erbistock Formation, along the entire southern and eastern margin of the Milton Green Inlier (Earp and Taylor, 1986, fig. 27). Fresh water was recovered by drill stem test from the Coed-yr-Allt Formation, in the Milton Green Borehole. Saline water was recovered from Dinantian rocks in the same borehole.

Two surface springs near the boundary of the Needwood and Widmerpool basins have elevated temperatures, and that at Kedleston contains hydrogen sulphide (Stephens, 1929). Springs with higher than average temperatures are geographi cally associated with hydrocarbon shows in the South Pennine Orefield, and the Ridgeway Borehole, near Sheffield, encountered hot salt water with hydrogen sulphide in Dinantian rocks.

Downing and Howitt (1969) showed that saline concentrations in Namurian and Westphalian groundwaters, in the East Midlands oilfields, increased northwards towards the basin depocentre. The highest values at Trumfleet (east of the Pennines) are matched at Formby 4, suggesting that the pattern west of the Pennines may also show a decrease southwards, in the Cheshire Basin. The East Midlands oilfields, which occur in a north-east-trending swathe, generally fall within values of less than 100 000 mg/l.

The low measured heatflow within the basin is consistent with a model of density-induced downward groundwater flow, as discussed in Chapter 6, upward flow taking place up the basin margin faults.

Migration within Carboniferous rocks

During Carboniferous times, it is likely that hydrocarbons migrated south-westwards away from the Pennine Basin depocentre towards its margin, to the area now underlying the Cheshire Basin. Late Carboniferous structural inversion of the Pennine Basin would have caused some hydrocarbons to migrate north-eastwards, away from the Cheshire Basin area, but erosion of potential reservoirs, in the Pennines, has destroyed any such hydrocarbons. These migration directions are analogous (as a mirror image) to those in the East Midlands (Kirby et al., 1987).

Migration of hydrocarbons since Carboniferous times may also have been largely within Carboniferous rocks, if the seals were not breached. Permo-Triassic syn-sedimentary faulting is, however, likely to have modified migration directions. The large faults produced sub-basins which will have had their own different migration directions. Migration was probably up-dip towards the Milton Green inlier. The most prospective area for this oil is undoubtedly to the south-east of the inlier, because there is probably no seal to the Carboniferous at outcrop. Hydrocarbon shows were present in the Carboniferous red beds in the Churton Borehole, drilled in the Milton Green Inlier.

The section of rocks drilled at Madeley, to the north of Coalbrookdale, illustrates how oil migrated south-westwards out of the basin (Figure 143). Madeley 1 and 3, to the north, have a more complete Westphalian succession, which thickens northwards. The Symon Unconformity, in contrast, cuts down into the earlier Westphalian rocks, removing most of the Westphalian B from Madeley 6 and all the Productive Coal Measures from Madeley 4, where the red beds come to rest on Wenlock Shale. This indicates that the Pennine Basin was full and already inverting, with the thickest post-Symon rocks displaced to the south onto St George's Land. Oil generated from the coals of the North Staffordshire Coalfield and its subsurface extension could have migrated toward the basin margin, within the thinning Productive Coal Measures. Near Madeley 6 and 4 the migrating oil transferred across the Symon Unconformity into sandstones of the Hadley and Coalport formations, reaching Coalbrookdale and beyond.

Migration into Permo-Triassic rocks

Evidence of migration of hydrocarbons into Permo-Triassic rocks, in addition to the few, poor hydrocarbon traces, comes mainly from isotope studies. The carbon-isotope characteristics of a late carbonate cement (late Triassic age) in the north and west of the Cheshire Basin indicate the presence of fluids derived from thermal maturation and decarboxylation of organic matter. This suggests fluid migration from Carboniferous to Permo-Triassic rocks at the southern margin of the Lancashire Coalfield and the West Lancashire Basin. There is no evidence of such a fluid component in the south and east of the basin (Chapter 5).

The migration pathways for fluids derived from the Carboniferous were probably not the main north–southtrending, inclined normal faults, with large displacements, which are probably sealing (Knott, 1994). Subvertical, east–west transfer faults may, however, have formed pathways for upward migration of fluids from Carboniferous rocks. Hydraulic connectivity between the Carboniferous and Permo-Triassic rocks is suggested by the chemistry of a single groundwater sample from the Coal Measures, but there are no other data to substantiate this interpretation (Chapter 6).

Most of the samples used for the present study were from shallow boreholes, which limits their usefulness in prediction of subsurface conditions (Chapter 5). Fluid-inclusion studies found no evidence of migration of hydrocarbons into the SSG of the Cheshire Basin, either before or during base-metal mineralisation. A model for sulphide deposition involving light hydrocarbon gases remains unproven (pp.218–20). Low-temperature fluids (70–80°C) were responsible for the mineralisation, and the model favoured requires a Carboniferous fluid component probably comprising no more than 5% of the total fluid budget. At Clive copper mine, south of the Wem Fault, there is evidence for a later, hotter, more saline fluid which deposited baryte. There is chemical affinity of these fluids with those of the West Shropshire Orefield, where saline fluids over 100°C were responsible for mineralisation (Chapter 5) and bitumen is recorded in Lower Palaeozoic rocks.

Migration of uraniferous bitumen into the Carbonifeous host rocks of the Ty Gwyn copper mine in North Wales is dated by ²⁰⁷Pb/²⁰⁶Pb as 248 ± 21 Ma (Early Triassic) (Parnell and Swainbank, 1990).

The small number of hydrocarbon shows in post-Carboniferous rocks invites comparison with the East Midlands. Although generation of hydrocarbons here also continued during Permian and Mesozoic times, there was no significant hydrocarbon migration into post-Carboniferous rocks. It is possible that, at the margins of the Cheshire Basin, hydrocarbons could have migrated into Permo-Triassic rocks. However, no such indications are known, although they clearly occur in the smaller Needwood Basin. The Needwood Basin, however, has a much thinner Permo-Triassic cover than Cheshire, and it would be easier for hydrocarbons to migrate vertically there.

The Formby Oilfield and surface seeps indicate that migration has continued, locally, into Recent times.

Reservoir rocks

Potential reservoirs are, from the top downwards: the Helsby Sandstone, Wilmslow Sandstone and Kinnerton Sandstone formations of the SSG; the Collyhurst Sandstone Formation; and Westphalian and Namurian sandstones. Many of the Silesian sandstones encountered in the Milton Green and Blacon East boreholes are thin, but thicker equivalents were found in the Erbistock Borehole.

Permo-Triassic reservoirs

Many of the Permo-Triassic sandstones have considerable secondary porosity due to dissolution of early diagenetic cements of non-ferroan dolomite, anhydrite and possibly halite and unstable framework grains (Chapter 5). This is a typical feature of other UK Permo-Triassic basins (Milodowski et al., 1986). Other authors (Burley, 1984; Bushell, 1986) have interpreted secondary porosity development as due to leaching of carbonate and feldspar phases, which then allowed entry of hydrocarbons in the East Irish Sea Basin.

Lovelock (1977) analysed porosities and permeabilities of four boreholes from south of the Wem Fault, two from the Vale of Clwyd, and an unspecified number of other boreholes. In the upper part of the Collyhurst Sandstone porosities were 10–15% and permeabilities up to 20 mD, but in the lower part they reached above 25% and up to 1800 mD respectively. Downing and Gray (1986) reported an average porosity of 13% from geophysical logs. In the Chester Pebble Beds the porosity in cemented parts was 14%, combined with low permeability, in contrast to the coarse clean sands with porosities of 22% and permeabilities of up to 2500 mD, which were probably not representative of subsurface properties. Geophysical logs indicate porosity of about 11%. In the Wilmslow Sandstone geophysical logs gave an average of 13% porosity (Downing and Gray, 1986). All these values are derived from samples taken towards the margins of the basin.

The geophysical logs in the Knutsford and Prees boreholes in the Cheshire Basin can be closely compared to the Formby wells. The same broad divisions are present in the northern East Irish Sea wells. British Gas laid the foundations of the correlation of the two areas and the relationships between the geophysical logs, the porosity-permeability and the diagenetic history of the sandstones (Colter and Barr, 1975; Colter, 1978; Colter and Ebbern, 1978; Ebbern, 1981; Bushell, 1986; Stuart and Cowan, 1991). The most distinctive log marker is the top Silicified Zone (Colter and Barr, 1975, fig. 8). This marker can be traced in nearly all the wells and separates porous sandstones above (known variously as the upper part of the St Bees Sandstone, the Calder Sandstone and, in the Cheshire Basin, the upper part of the Wilmslow Sandstone) from lower-porosity, cemented sandstones below (lower part of St Bees Sandstone, lower part of Wilmslow Sandstone, Chester Pebble Beds etc.). A decrease in porosity with depth is seen clearly in the sonic logs of the Knutsford and Prees boreholes, below the top Silicified Zone, reaching a minimum within the Chester Pebble Beds (D Buckley, personal communication). Below the Chester Pebble Beds and the base of the St Bees Sandstone (Wilmslow Sandstone) there are indications of more porous sandstones, comprising the Kinnerton Sandstone Formation and Collyhurst Sandstone Formation (D Buckley, personal communication). Changes in porosity in the upper part of the SSG, showing no systematic downhole increase, are attributed to facies control. Aeolian sandstones are more porous than fluvial sandstones. The fluvial Delamere Member has a higher sonic velocity than the Frodsham and Thurstaston members.

Modelling of the burial history of this basin suggests that compaction-induced porosity reduction is likely to have occurred towards the centre of the basin, reducing average porosities at the base of the SSG to between 5 and 7% (J Rowley, personal communication; (Figure 144). This is the result of Jurassic and Cretaceous burial. Porosity reduction has not occurred in the Permian sandstones of the Prees and Knutsford boreholes. The reason for this may be the early, secondary porosity noted in the shallow boreholes (Chapter 5), which was not destroyed by the later burial.

A downhole porosity reduction within the Permian sandstones is apparent in the Collyhurst Sandstone of the Prees Borehole, but the overlying Kinnerton Sandstone Formation has a lower porosity. The Collyhurst Sandstone, in the Knutsford Borehole, shows more variable and higher porosity. A number of factors may prevent porosity reduction occurring. Many sandstones with early cements can withstand compaction of grains (G E Strong, personal communication). Early migration of hydrocarbons into these rocks would have forestalled this porosity decrease with burial; there is, however, no evidence for such early migration. Low geothermal gradients result in low porosity reduction with depth.

Westphalian reservoirs

Laboratory tests on core from three British Coal boreholes demonstrated porosities less than 15% and permeabilities less than 1.6 mD (Downing and Gray, 1986, p.91). In the East Midlands oilfields, Coal Measures sandstones, which produce oil, have porosities of about 7–20%, with an average value of 12%. Upper Coal Measures sandstones, which are not productive, have higher porosities, 12–19%. Permeabilities range from 0.06 to 37 mD, locally 400 mD for the Productive Coal Measures and 2 to 160 mD for the Upper Coal Measures sandstones (Downing and Gray, 1986).

In the Bothamsall Oilfield Hawkins (1978) found porosities of 3–12% in coarser grained channel sandstones of Westphalian A age. Early oil migration prevented quartz diagenesis.

Reddening of Coal Measures sandstones, in contrast to mudstones, has occurred during Permian and Triassic times in the Potteries Coalfield, indicating some porosity and permeability of the sandstones (Rees and Wilson, 1998).

Namurian reservoirs

In the East Midlands porosities average 12% and permeabilities 14 mD. A northerly decrease in porosity and permeability has been noted in the East Midlands oilfields, which is interpreted as a Carboniferous burial effect (Downing and Gray, 1986). Similar northward loss of porosity and permeability is envisaged for the Cheshire Basin area. The south and west of the Basin may have the best reservoir properties.

Seal rocks

The MMG halites and mudstones are excellent seals, where sufficiently thick. At Formby, however, there is no salt and the MMG is thin, resulting in a very leaky oilfield.

There are no other significant mudstone beds within the Triassic succession. The Manchester Marls can be considered a reasonable seal where there is mudstone and, to the north-west, halite, but this formation becomes sandy in the south and west of the Cheshire Basin (Chapter 2).

Although gas is normally more difficult to retain, the small gasfields in the East Midlands (e.g. Hatfield Moors and Calow) have not required evaporites to maintain the seal. It may be deduced, therefore, that Carboniferous shales are adequate to seal the hydrocarbons beneath the Variscan Unconformity, particularly where Ruabon Marl (Etruria Marl) is present. At Hatfield Moors the productive Oaks Rock (Westphalian B) is overlain by 20 m of shales, beneath the Variscan Unconformity.

Knott (1994), in a study of fault displacements and fault zone widths, has suggested that sealing is likely to be greater on wide faults of large displacement, which have correspondingly thicker zones of reduced permeability and porosity. Pinfold (1958) pointed out that faults dipping at 40–50°, which are common in the Lancashire Coalfield area, are likely to be sealing because of the weight of the rock column on the hangingwall.

Comparison with neighbouring oil and gas provinces

The active tectonics of Permo-Triassic times in the Cheshire Basin contrast with the relatively steady accumulation of sediments on the gently shelving margin to the southern North Sea Basin in the East Midlands. This is likely to have led to differences in hydrocarbon distribution in these two areas. However, the sub-basins of the Cheshire Basin are relatively simple structures compared with the chaotic faulting in the East Irish Sea Basin (Jackson et al., 1987) and again differences in hydrocarbon distribution between the Cheshire and East Irish Sea basins are to be expected.

East Midlands

The East Midlands and the Cheshire Basin are alike in appearing to have few hydrocarbon accumulations in Permian and later rocks. However, there are numerous oilfields in Carboniferous rocks of the East Midlands whereas, to date, there have been no commercial discoveries in the Carboniferous rocks surrounding the Cheshire Basin.

The first oilfield was discovered at Hardstoft as a result of belated attempts to obtain home oil supplies during the First World War. Oil was discovered at Eakring on the eve of the Second World War and subsequently a steady stream of fields has been developed in the East Midlands. These oilfields have certain common characteristics: the reservoirs are mainly late Namurian and Westphalian A sandstones (although Hardstoft produced from uppermost Dinantian strata and Rempstone from early Namurian rocks), and are sealed below post-Permian rocks. By contrast, the western half of the Pennine Basin has no oilfields; there are as many oilfields in the East Midlands as there are dry holes to the west of the Pennines. This spectacular lack of success, west of the Pennines, may be because by far the largest percentage of wells drilled in the west of the Pennine Basin have spudded into large structures at stratigraphic levels below the reservoirs producing in the East Midlands (Table 41) and only a few wells have seals below post-Permian rocks. In other words, it is possible that inappropriate targets have been drilled; the lack of success to date, therefore, should not necessarily deter further exploration.

The API of East Midlands oils ranges from 26 to 43°, with a range of 31–43° at the Eakring Oilfield alone. This variation probably reflects the multi-source derivation, encompassing the complete oil window. The oil at Formby is 37° API. The East Irish Sea fields contain a more mature oil, as at Douglas (44°) and Lennox (45°) (Department of Trade and Industry, 1994).

The carbon isotope values of East Midlands oils and circum Cheshire Basin samples appear to be broadly comparable (Eakin, 1989).

Parnell (1992b) has tabulated the metal enrichment of bitumens in Carboniferous-hosted ore deposits, indicating that there are a number of close associations between ores and hydrocarbons in Carboniferous rocks. Examples surrounding the Cheshire Basin include Breedon, Ecton, Odin, Windy Knoll, Minera and the West Shropshire Orefield. The age of ore and bitumen migration into these host rocks is not known with certainty, and a link between ore fluid and hydrocarbon migration is not universally agreed upon (Ewbank et al., 1995; Quirk, 1996).

St George's Land formed the southern margin of the Carboniferous Pennine Basin; in the East Midlands, Westphalian rocks overlap regressive Namurian rocks along a north-east-trending line onto this former landmass. The Welton Oilfield was discovered where the basal Westphalian sandstones rest unconformably on Dinantian rocks. Near the southern margin of the Potteries Coalfield and beneath the south-east Cheshire Basin a similar stratigraphical situation may exist, with thinning of the Namurian rocks beneath the Potteries Coalfield (Rees and Wilson, 1998).

East Irish Sea Basin

The principal difference between the Cheshire and East Irish Sea basins is the presence of hydrocarbons in the Ormskirk Sandstone Formation of the East Irish Sea Basin and their apparent absence in its correlative, the Helsby Sandstone Formation, in the Cheshire Basin. This may relate to differences in structural history and stratigraphy, but it may also relate to the availability of suitable source rocks.

The main Carboniferous rift is orientated east–west, through the southern North Sea, beneath the Cleveland, Craven and East Irish Sea basins towards Ireland. The main rich oil-prone source rocks also occupy this zone (Sabden Shales, Bowland Shales and Edale Shales) and are now believed to have generated the hydrocarbons in the Irish Sea fields (Hardman et al., 1993; Trueblood et al., 1995). The Cheshire Basin is more peripheral to the Carboniferous Pennine Basin than the East Irish Sea Basin and is therefore thought less likely to be underlain by good oil-prone source rocks at depth. There are a few positive indications, however, such as oil shale in the Heswall Borehole, and also the unproved possibility that Dinantian basins (in addition to the Blacon area) may have locally rich oil-prone source rocks. Coals are abundant and widespread beneath the Cheshire Basin (Figure 140). Beneath the East Irish Sea Basin much of the subcrop consists of Namurian rocks.

Prior to the 11 th Round licence award to Hamilton in 1989, in the southern part of the East Irish Sea Basin, no significant discovery of oil had been made in the basin since the Morecambe Gasfield discovery in 1974. This long delay can be attributed to a lack of interest in the Carboniferous source rocks; even with the recent discoveries, there is a reluctance to extend the play, by exploration of Carboniferous rocks. It now appears that the classic model of oil on the margins and gas in the centre of the basin (Morecambe Gasfield) could be appropriate.

Significant oil was generated during Jurassic times, but gas generation required burial deeper than 4000 m according to Hardman et al. (1993). The amount of Tertiary uplift is estimated by various methods to be over 2000 m (Roberts, 1989; and see Chapter 3 of this volume). The apatite fission track studies were interpreted to indicate 3000 m of regional uplift (Lewis et al., 1992) and 3700 m at the Douglas Oilfield (Trueblood et al., 1995). However, Holliday (1993) suggested that these figures are an overestimate. Hamilton reached a similar conclusion in view of the lack of halokinesis in the vicinity of the Douglas Oilfield (Trueblood et al., 1995).

The SSG has many similar reservoir characteristics in the Cheshire Basin (Chapter 5) and the East Irish Sea Basin (Burley, 1984). Porosities of up to 21% and permeabilities of up to 400 mD are reported for the Morecambe Gasfield (Colter and Ebbern, 1978), in beds probably equivalent to the Frodsham Member of the Helsby Sandstone. Ebbern (1981) gave porosities of 5–20% and permeabilities of 0.1–1000 mD for the Ormskirk Sandstone, whereas the St Bees Sandstone had lower permeabilities (0.1–100), owing to the presence of detrital clays. The presence of platy illite reduces the permeability of the reservoir. At the Douglas Oilfield the Frodsham Member has a porosity of 17% and the Delamere Member 12% (Trueblood et al., 1995).

Coal-bed methane

Two coal-bed methane wells have been drilled near the western margin of the Cheshire Basin. Many of the coal seams of the Lancashire, North Staffordshire, Flint and Denbighshire coalfields are gassy, having required British Coal to set up methane drainage schemes in these coalfields at Point of Ayr, Parkside, Wolstanton and other collieries. Coal Measures are present beneath almost all of the Cheshire Basin. The coal-bed methane potential would appear to be very good (Glover et al., 1993a).

Conclusions and recommendations

General

The main conclusion is that hydrocarbons generated from Carboniferous times to the present day are likely to be trapped beneath the base Permian unconformity, by Ruabon Marl (Etruria Marl) seals, and that drilling to test late Namurian to early Westphalian reservoirs, although deeper than present exploration, can be justified. Wells sited on the crop of early Namurian and older rocks are not recommended because these rocks have unfavourable reservoir characteristics.

These conclusions are at variance with those of Fraser et al. (1990), who presented a pessimistic view of Carboniferous basins outside the East Midlands, which contributed partly to the evacuation of the area by the major oil companies.

Two key questions, relevant to the hydrocarbon prospectivity of the Cheshire Basin, are:

1. Why have no Carboniferous oilfields been discovered in north-west England comparable to those in the East Midlands?
2. Why is the Ormskirk Sandstone Formation of the East Irish Sea Basin productive of hydrocarbons whereas no discoveries have been made in its correlative Helsby Sandstone Formation in the Cheshire Basin?

Fraser et al. (1990) identified factors which are crucial to the prospectivity of Carboniferous basins in northern England. They do not recommend drilling deeper than 2500 m, because their core data indicate that permeability and porosity have declined at this depth to 1 mD and 10% respectively. However, there is sufficient variability in their data (about 5% difference in porosity at the same depth) to extend the cut-off a little deeper, and it can be argued that migration of oil into reservoirs at an early stage could have occurred before the detrimental effects of burial.

The critical burial depth for the oil window in the East Midlands of 1900–3600 m (Fraser et al., 1990) is probably not applicable to the Cheshire Basin. Carboniferous rocks in Knutsford lie near the oil floor at 3000 m, but have probably been uplifted from below 4500 m in Tertiary times (Figure 142). Coal-rank data indicate that present-day geothermal gradients, and possibly Carboniferous gradients, are lower in the Lancashire Coalfield than in the North Staffordshire and Nottingham–Yorkshire coalfields. This indicates important regional differences.

The distribution of good-quality basinal pro-delta source rocks is an important factor (Fraser et al., 1990). It is possible that Namurian pro-delta shales exist beneath the Cheshire Basin. Connections between the Widmerpool–Staffordshire Dinantian Basin and south Lancashire were suggested by Collinson et al. (1977; (Figure 145) and are tentatively in agreement with seismic reflection interpretation of the Carboniferous succession in parts of the northern Cheshire Basin. This gulf may not have been filled with delta turbidites and channel sands until Marsdenian times. Early Namurian sedimentation may, therefore, have led to the formation of rich source rocks.

The degree of post-Variscan trap modification (Fraser et al., 1990) is probably most severe within the Cheshire Basin. The Formby Oilfield is an example of the probable breaching of an accumulation in Carboniferous rocks, with migration of oil into Triassic and Recent sediments. PostVariscan modification of migration directions also occurred. Migration to the north (in the west) and north-east (in the east) is suggested by the main two tilt blocks forming the Cheshire Basin.

Structures can be mapped in Carboniferous rocks beneath the Permo-Triassic in some places, but additional seismic reflection data would be required in some areas prior to drilling. Only one large anticline (like Eakring, with Westphalian C–D rocks removed) is mapped, in the south-west of the basin (Figure 140). Westphalian reservoirs, beneath the level drilled at Knutsford, lie just below the oil floor. Oil-prone Namurian and Dinantian shales are probably present in the subsurface. Gas-prone Westphalian coals are present within a southerly thinning succession into which migration of hydrocarbons is thought to have occurred. Failure to find encouraging hydrocarbon shows in post-Carboniferous rocks and subsequent dry commercial wells within the basin are seen as evidence that the hydrocarbons may be securely trapped in the Carboniferous rocks. Modification of migration pathways by Permo-Triassic tectonism is obviously greater and more complex than in the East Midlands, but this should not be a deterrent to exploration.

If hydrocarbons are securely trapped within the Carboniferous rocks, it is possible that overpressuring has occurred. Hunt et al. (1994) believe that gas generation, not compaction, has caused the overpressuring in the USA Gulf Coast. Overpressuring is a characteristic of the deepest sedimentary basins such as the Anadarko Basin, USA, where the main overpressured zone, in several compartments, encompasses Devonian and Carboniferous rocks from 3 to 7 km deep (Al-Shaieb, 1991). The main result of overpressuring is high porosity and permeability, preserved by seals, across which the pressure can increase by 5.0 psi/ft compared with a normal gradient of 0.46 psi/ft.

In some North American basins (Western Canada, Eastern Ohio etc.) an abnormally pressured gas-saturated zone lies trapped below normally pressured water (Masters, 1984).

In the Cheshire Basin the model of density-induced downward groundwater flow (Chapter 6) may have prevented upward migration of hydrocarbons (I N Gale, personal communication).

Although it is not the primary target, testing of the Helsby Sandstone should not be ignored. The mineralisation model (pp.218–220) requires a contribution from Carboniferous fluids. The following structural leads are based on the idea that sealed traps may occur down dip of the mineralised localities.

Structural Leads: Helsby Sandstone Play

Structural leads can be identified by combining the contour maps of the top SSG with the outcrop of the Helsby Sandstone Formation. Four areas would be worth drilling (Figure 146), based on this criterion and the up-dip surface Cu mineralisation. These are the known occurrences of Cu, which, according to the mineralisation model, require a Carboniferous fluid component.

On the block between the Wem and Hodnet faults

Generation of hydrocarbons will have occurred to the north of the Wem Fault. Hydrocarbon migration up the Wem Fault may have occurred (Little Ness Borehole). A MMG seal is present, although it is not thick and does not contain halite. A nearby outcrop contains Cu mineralisation in the Helsby Sandstone. The top Helsby Sandstone is at very shallow depth hereabouts.

Between the Helsby Sandstone outcrop at Bickerton and the Wem–Audlem Sub-basin

Hydrocarbon migration up the tilt-block towards Milton Green may have occurred and sealing by a MMG succession with halite is assured. The outcrop to the north contains Cu mineralisation.

Between the King Street Fault and Alderley Edge

Generation of hydrocarbons is possible beneath the depocentre against the King Street Fault. Migration of hydrocarbons up the block toward Alderley Edge may have occurred. The seal by the MMG is good. Alderley Edge Cu mines are situated on the Helsby Sandstone outcrop to the north-east.

South-west (footwall block) of the Brook House Fault

Generation of hydrocarbons may have occurred beneath the Wem–Audlem Sub-basin, with migration northwards, complemented by generation east of the Brook House and King Street fault. A MMG seal is present and there are good shows to the north in Westphalian rocks in the subsurface and at outcrop.

Structural leads: Collyhurst Sandstone play

This play depends on the presence of the Manchester Marls seal. An eastward facies change in this formation is evident along the northern margin of the basin, from sands to marls and halite may be developed in the subsurface between Knutsford and Prees (but excluding the boreholes), but this area has not been explored at this level.

Halite is widely present in the Manchester Marls in the East Irish Sea Basin (Colter and Barr, 1975; Jackson et al., 1987). A thin bed (15 cm) was intersected at 625 m in the Heswall Borehole on the Wirral (Wade, 1910), on the high between the East Irish Sea and Cheshire basins. It would seem plausible to expect evaporites in a partly restricted basin (in the deepest part of the Cheshire Basin) prior to the establishment of the north-flowing Budleighensis River (Chester Pebble Beds), which amalgamated the individual rift basins for the first time.

Structural leads: Carboniferous plays

The main area of Carboniferous exploration targets is related to the south- and west-thinning Silesian rocks. The main shows are near Wigan and in the subsurface of the Runcorn sheet (number 97). The Westphalian rocks are thinner here than to the east, towards Manchester, and north, towards Burnley (Calver, 1968). Migration of hydrocarbons southwestwards from the Pennine Basin occurred during Silesian times. Permo-Triassic tectonism will have caused re-migration, although the hydrocarbons may still be retained within Carboniferous rocks. Hydrocarbons within Carboniferous rocks, west of the King Street Fault, probably migrated west or north; those within Carboniferous rocks east of the King Street Fault are likely to have migrated north-east.

Using an East Midlands analogue, there may also be a play where the basal Westphalian sandstone unconformably overlies Dinantian rocks. This would be located further south-west than the main prospective area indicated on (Figure 146).

A third play is the possible extension of the Derbyshire High from the Buxton area north-west to the Alderley–Stockport area. On the Chapel-en-le-Frith sheet (number 99) the northern limit of this high is recognised by thinning of the Kinderscoutian Shale Grit. This thinning coincides with the Dinantian synsedimentary Bakewell Fault, which was postulated between the Eyam and Woo Dale boreholes to account for the thickness changes in the Dinantian rocks (Smith et al., 1985). On the Buxton sheet (number 111) to the south, thickening of the Kinderscoutian Longnor Sandstones occurs near the point where the Asbian apron reef belt ceases, and marks the southern margin of this high (the Dinantian outcrop boundary also swings from south-east–north-west to north–south here). The area south of the reef belt can be defined as the Widmerpool–Staffordshire Basin.

The prolongation of the Derbyshire High into the Cheshire Basin is supported by the aeromagnetic evidence. A broad positive aeromagnetic anomaly trending north-west extends from the Derbyshire Dome into the north-east corner of the Cheshire Basin (Figure 147). Here the Red Rock Fault and associated faults are subordinate to the King Street Fault, which controls the dip from the Pennines.

Hydrocarbons could possibly have re-migrated north-east into this high. The attractiveness of this position is enhanced by the likelihood of Silesian reservoirs of better quality than farther east. It is likely, therefore, that the Derbyshire high continues north-west and becomes progressively concealed by Westphalian and Permo-Triassic rocks, on the north-east margin of the Cheshire Basin (Figure 146).

Other possible plays occur down-dip of the Milton Green inlier and on the block between the Wem and Hodnet faults, in areas also considered propective at the Helsby Sandstone level (Figure 146)

Groundwater

General hydrogeology

The main aquifer units in the Cheshire Basin belong to the SSG and the underlying Permian sandstones, principally the Collyhurst Sandstone Formation. The distribution of these aquifers over most of the area is shown on the Hydrogeological map of Clwyd and the Cheshire Basin (British Geological Survey, 1989). The sandstones occur around the perimeter of the basin and describe a broad arc from Market Drayton south to Shrewsbury and then north-west to Oswestry and Wrexham. They are also found around Chester and the Wirral and they continue through Merseyside to Manchester. The eastern boundary of the basin is mainly fault controlled, which has resulted in a much smaller area of sandstone outcrop, the overlying MMG sediments being faulted against Carboniferous rocks between Macclesfield and Newcastle-under-Lyme.

The MMG overlies the Triassic and Permian aquifers and attains a considerable thickness over much of the eastern two-thirds of Shropshire and the Cheshire Plain. It forms a confining layer, and its thickness restricts development of the underlying aquifers. Consequently, there are few data available in this area on which to make a hydrogeological interpretation.

The aquifers of the SSG represent a major groundwater resource in north-west England. Transmissivity values range from 10 m²/d to more than 10 000 m²/d, but for most parts of the upper aquifer they lie between 100 and 400 m²/d. Laboratory determinations of intergranular permeability of core samples rarely give values high enough to equate to the transmissivities derived from aquifer tests. Brassington and Walthall (1985) used packer tests and closed-circuit television logging to demonstrate that high permeability zones are associated with fracturing, which is commonly found to depths of up to 150 m.

Yields from boreholes vary considerably, but failures are almost unknown. Individual discharges of 501/s are common from large-diameter boreholes and some discharges exceed 1001/s. Near surface, the sandstone units are generally friable due to weathering, which reduces the degree of cementation. However, beneath the effects of weathering, fracturing provides the main permeability and hence matrix permeability is proportionally less important. The Wilmslow Sandstone Formation tends to produce the largest yields because of its higher density of fracturing and less-cemented matrix, though the more pebbly Helsby Sandstone and Chester Pebble Beds are also very productive aquifers.

A large number of borehole abstractions are centred around the margins of the Mersey, with several major well fields located to the north of Chester along the mid-Cheshire ridge, including over 60 boreholes of 200 m depth. There are several production boreholes across the Cheshire Plain within the area of the aquifer outcrop and, to the north of Shrewsbury, major groundwater abstraction projects, such as the Shropshire Groundwater Scheme, have been carried out over the past 20 years and are ongoing.

Total groundwater abstraction in the northern part of the basin peaked in 1960 at 93 Ml/d. By 1982, this had decreased to about 88 Mid. In Liverpool, groundwater abstraction increased until the mid-1950s to 70 Ml/d, since when there has been a steady decline to less than 20 Ml/d in 1982.

Groundwater was formerly of potable quality almost everywhere, with chloride ion concentrations of less than 60 mgCl/1. However, overpumping beneath large conurbations has resulted in ingress of saline water from the Mersey and the Manchester Ship Canal and a decrease in quality (6000 mgCl/1) over progressively larger areas. Where water levels are rising due to reduction in abstraction, there is generally not a corresponding improvement in quality.

Saline water (up to 100 000 mgCl/1) is found at depth over extensive areas beneath the active groundwater system and, in places, this is being drawn up by pumping. The saline water results from the dissolution of halite in the MMG. A large area of saline water of this origin is found to the east of the Carboniferous inlier at Milton Green (Earp and Taylor, 1986). This is thought to represent a rising limb of a deep basinal saline flow system (Chapter 6). Water quality can also be poor in Drift deposits and in the limited water resources that are exploited from the MMG. Nitrate concentrations in groundwater in the SSG have increased over the last three decades due to increased use of nitrogenous fertilisers in areas where recent recharge is occurring.

Hydraulic properties of aquifer units

In the following sections, the data relating to the hydraulic properties of the rocks are derived from both field pumping test data and laboratory measurements. Intrinsic permeability (mD) measurements have been converted to hydraulic conductivity units (m/d), assuming that the aquifer contains fresh water at 10°C. Although many of the sandstone aquifers in the Cheshire Basin have relatively high values of porosity and hydraulic conductivity, localised fracturing is the principal cause of high transmissivities.

Collyhurst and Kinnerton Sandstone Formations

These formations comprise fine- to medium-grained aeolian sandstones. In the north of the basin, the two units are separated by the Manchester Marls Formation, which is thought to act as an aquiclude. At outcrop, the sandstones are regarded as being very good aquifers, with intergranular hydraulic conductivities commonly greater than 1 m/d and yields of 20 to 301/s from large-diameter boreholes. Porosities are high, especially in sections beneath the Chester Pebble Beds and close to outcrop where they are in the range 20 to 24%. In the Shropshire area, field hydraulic conductivity varies between 1.8 and 12 m/d and storativity from 2.1 X 10⁻² to 5 X 10⁻³.

Chester Pebble Beds Formation

These conglomerates and pebbly sandstones, overlying the Collyhurst and Kinnerton Sandstone Formations, gradually thicken to the north-west to 225 m in the Manchester area and between 316 and 375 m in the Warrington area. The porosity of the sandstones at outcrop is 20 to 30% and the hydraulic conductivity ranges from 0.5 to 5 m/d. Hydraulic conductivities are, in general, higher in the south of the basin, but Fletcher (1977) states that around the River Tern they are only half that of the Collyhurst Sandstone Formation.

Wilmslow Sandstone Formation

This weakly cemented, medium to fine-grained sandstone crops out in the west and north of the basin, and on the basin margins it is up to 280 m thick. Physical properties cover a wide range. Near outcrop, porosity ranges from 13 to 30%, with associated hydraulic conductivities of around 0.006 to 6 m/d.

Helsby Sandstone Formation

This formation has an almost continuous outcrop in the north-west and south of the basin. It comprises 100 to 250 m of poorly cemented fine to coarse-grained sandstones and minor mudstone bands, with breccia and conglomerates at the base. Of the three members of the formation, the central Delamere Member is the least porous.

In mid-Cheshire, where the Delamere Member is absent, porosity and hydraulic conductivity values are highest, attaining values of 30% and 2.5 m/d respectively. Near outcrop in north Cheshire, hydraulic conductivity is as high as 10 m/d, but porosity is reduced by suture contact pressure solution to 25–30%.

Tarporley Siltstone Formation

The Tarporley Siltstone Formation, at the base of the MMG, is transitional between the Helsby Sandstone Formation and the overlying mudstone sequence. It ranges in thickness between 10 and 270 m. It generally has low values of hydraulic conductivity, although faulting can create hydraulic continuity with the Helsby Sandstone.

Drift

Extensive and thick deposits of glacial and fluvioglacial material are present in the Cheshire Basin and the upper Mersey valley around Manchester. Granular deposits commonly underlie sheets of till, but the sequences are extremely variable and complex. Large bodies of well-sorted sand to the east of Chester separate two tills of different ages, and this general succession extends from Wrexham to Shrewsbury.

The sand and gravel can exceed 70 m in thickness and form a significant groundwater resource on a local scale, with artesian conditions present in the floors of many valleys and areas of low ground. The sand and gravel deposits act as an important source of recharge to the SSG and also as a means of discharge.

Extensive sheets of till and clay Drift, however, act as effective barriers to recharge over much of the Cheshire Plain. More than half of the area of outcrop is covered by greater than 2 m of clay Drift, where it is estimated that only 2% of potential recharge can infiltrate.

Chapter 8 Expert-GIS

G Wadge, G Ferrier and C McDermott

Computing is a valuable tool for the analysis of sedimentary basins. Butler (1992) has argued that rule-based expert computer systems (knowledge-based systems) should play a major role in petroleum resource assessments in such basins. Increasingly, new exploration applications for knowledge-based systems are being proposed (e.g. Aminzadeh and Simaan, 1991), but there is a need for exemplary case studies of the applications of these techniques, for hydrocarbons and for metalliferous minerals (Wolf, 1994). In an earlier project concerned with the prospectivity of the Carboniferous basins of the Pennines for metalliferous minerals, Wadge et al. (1991, 1992) argued that not only were computers necessary for databasing and modelling but they were also needed to capture the knowledge, both operative and interpretive, of the project team. In that study they described a prototype PC-based computer system (BURMIN) that showed how expert-system and image-processing software could be used in combination to manage and model processes for mineralisation in basins. The Cheshire Basin project has built on the concept of providing a capability of interactive interrogation of the database and knowledge base, and has developed the Cheshire Basin Expert-GIS.

The main aims of developing the Cheshire Basin ExpertGIS were to create a workstation computer system within BGS that would:

1. create a permanent digital record across the breadth of sub-disciplinary knowledge brought to bear on the project,
2. allow retrieval of project results and enable results to be replicated and explored further, and would have value for training in basin analysis, and
3. provide a means of integrating cross-disciplinary knowledge applied to a broad problem – prospectivity for red-bed copper mineralisation.

This Chapter describes the progress towards these goals. It first presents an overview of the system design and implementation, with a summary of the knowledge bases held by the system. It then describes in more detail the way the ExpertGIS works in respect of three topics: diagenetic domains, handling vagueness and uncertainties in provenance determination, and spatial models of red-bed mineralisation.

System design and implementation

As the title implies, the system (Figure 148) combines a Geographical Information System (GIS) with Expert System software. The design was partly controlled by financial and institutional considerations and partly by the nature of the data themselves. The system was developed jointly by BGS and the NERC Unit for Thematic Information Systems (now the Environmental Systems Science Centre). It had to be quasi-operational in nature, and required a networkable machine of sufficient power. A Sun Sparcstation running under SunOS was chosen, a machine type which is used by both BGS and ESSC and is capable of running ORACLE database software, the BGS standard.

The structure of the basin is mainly known from seismic reflection and other geophysical modelling. Unlike some hydrocarbon reservoirs and ore bodies, where sufficient borehole data are available to interpolate in three dimensions with some confidence, there are few deep boreholes in the Cheshire Basin and the data do not justify the use of fully three-dimensional representation. To hold the spatial data, the ARC/INFO GIS (v.6) was used, which is capable of high-quality cartographic representation, basic grid processing and 2.5-D display of gridded surfaces. More importantly for this application, it has its own programming language, ARC Macro Language (AML), which allows subroutines to be constructed and provides a mechanism to communicate with external programs, the UNIX operating system and the TCP/IP level of message passing.

There is at present no consensus data model for representing 3-D geological data, though progress to that end is being made. Different sub-disciplines of the geosciences tend to use different, overlapping models to represent their physical parameters in the computer. It was therefore necessary to accommodate these different sub-disciplinary data models. NEXPERT OBJECT is a powerful expert-system software package that permits, indeed requires, that these data models be specified explicitly in an object-oriented way in conjunction with the rules of how to treat the data. It has a graphical interface that permits the developer to visualise the linkages between the data types and their relationships within the rules (Figure 149). Another very useful capability of NEXPERT OBJECT is the built-in linkage to various databases, including ORACLE. The proposed development of a more object- oriented version of ARC/INFO in future may make this system combination even more attractive.

NEXPERT OBJECT responds to a request for data, or an answer to a question, by testing a number of rules relevant to the problem. If these rules require additional data, these could be extracted from the database. Alternatively, if the rules require a spatial input, this would be taken from the GIS. When the immediate problem has been successfully addressed, the rules may require further action: drawing a new map, writing new records to a database, writing text to the screen, or the testing of a further ruleset.

Communications

One of the main implementation tasks of the ARC/INFO NEXPERT OBJECT system was the creation of a means of communication between the two packages. The development of this communication link has been discussed in Ferrier et al. (1993), McDermott et al. (1993) and Ferrier and Wadge (1994). Using ARC/INFO's AML it is possible to pass arguments, execute an external program, and receive a string in return using the function 'Task'. There are various ways to achieve the linkage, none of which are ideal for all possible network communications. In one version, custom software handles the message passing using the Transport Library Interface, with either the ARC/INFO or the NEXPERT OBJECT process acting as client or server. The system has been implemented in such a way that the user sees nothing of either the communications link or NEXPERT OBJECT itself, and does not need to know how ARC/INFO works.

User interface

The graphical user interface was designed using the widget and menu creating capabilities of ARC/INFO's AML (Figure 150). There are two basic ways to interrogate the system using the interface: in a supervised or unsupervised manner. Unsupervised operation allows the system to search for an answer to the question or hypothesis posed, unconstrained by any spatial considerations. In supervised use, the user defines the sampling space in three dimensions. First the map dimensions of the area of interest are chosen, either by coordinates, or by a cursor defining a region, or by choosing a specific point (e.g. representing a borehole). Then the user narrows the choice in the vertical or statigraphical dimension. This can be by borehole depth limits, by choosing an individual formation, by upper and lower bounding formations, or by lithology.

In the top right-hand corner is an overview of the object-oriented structure of a knowledge base with a magnified version of the properties and their values (e.g. depth=209.08) for one object. In the top left is the rule 'checklow' in the knowledge base 'sepfin21.kb'. If the statements in the If box are true then the actions in the 'Then Do' box are carried out. The syntax, although appearing obscure, is not really difficult and there is a facility to explain the rule's meaning linguistically in readily understandable form (not shown here). At the bottom left is an overview of the rule structure for the knowledge base, part of which is displayed, magnified, at bottom right, showing relationships between rules and status indicators. Question marks indicate untested statements; they change to ticks once the rules are fired.

The Cheshire Basin knowledge bases

The system operates as a series of modules, matching the main topics studied in the Cheshire Basin project and presented more fully in earlier chapters of this volume. The modules are summarised as follows:

Provenance

Determination of the source areas of the constituent sedimentary detritus of the basin. The data used (Chapter 4) include heavy minerals and their ratios, tourmaline compositions, Sm/Nd ratios, uranium in apatites, clay mineralogy, sedimentology, whole-rock geochemistry, and palaeoflow indicators.

Diagenesis

The mineralogical and chemical changes that the sediment undergoes during burial and uplift. The type and order of the development of authigenic minerals were determined from examination of thin sections, together with backscattered electron microscopy (Chapter 5). Inferences are made of the burial histories of parts of the basin.

Structure

The three-dimensional shape of the floor of the basin and the main stratigraphical horizons at the present day. Models of these structural features at key times during basin evolution were created using backstripping and decompaction techniques. The present-day structure is defined by four faulted surfaces (base Permo-Trias, base SSG, base MMG and base Jurassic), modelled from seismic-reflection sections constrained by knowledge of the surface geology, borehole cores and magnetic and gravity-field data. The palaeostructure and thermal states were modelled for 251, 242, 205, 97 and 60 Ma ago (see Chapter 3).

Hydrogeology

The present and past groundwater flow regimes, and the aquifer hydrochemistry during diagenesis and, particularly, the main sulphide mineralisation events (Chapter 6). The modern, shallow, groundwater flow regime is shown in map form, and the palaeoflow modelling in 2-D section across the basin axis. The geochemical and hydro-chemical modelling is essentially aspatial.

Mineralisation

Understanding the timing, ambient conditions and driving processes of the red-bed copper mineralisation in the Cheshire Basin, so that the prospectivity for the whole basin can be evaluated. This requires input from most of the other modules, together with data on the petrology, mineralogy and geochemistry of the ore bodies, stable-isotope geochemistry, and the temperature and composition of fluid inclusions. The geological setting of the known areas of mineralisation around the basin margins constrains the structural settings of the mineralisation event.

The knowledge bases need access to data in each module to run. Within each module there may be one or more knowledge bases containing rules that embody procedural knowledge (e.g. how to calculate a value or define a category) and heuristic and interpretive knowledge (e.g. z is likely if x and y are present). The information for the knowledge-based systems was acquired by talking to project staff and reading interim reports. Not all the modules were completed in the time available. This was because much of the data and knowledge only became available towards the end of the project, giving insufficent time to populate the system with data and knowledge. The Provenance and Diagenesis modules are the most complete, but the Mineralisation module, because it integrates the results from other modules, is not.

A system such as this should be able to address major integrative modelling tasks, such as identifying the volumes of basin fill that are most prospective for end-Triassic copper mineralisation. Wadge et al. (1992) argued that modelling of the mineralisation process itself offered the best approach to such tasks. To do this in a rigorous way would require a quantified process model solving for mass and energy flux using partial differential equations. The system would require access to data on: (1) the structure and thermal state of the basin at the time of mineralisation, (2) the geochemistry of the source rock, (3) the petrophysical properties of the rocks, and (4) the hydrochemistry of the pore fluid. In a knowledge-based system such as the Cheshire Basin ExpertGIS, the system would work backwards from the posed question or hypothesis, assessing the model boundary conditions and parameters. When it knew what data it required to run the model it would request them from the appropriate knowledge bases in the sub-disciplinary modules. Within these knowledge bases it is possible to generate ranked lists of methods to evaluate values and generate data products. For example, if the optimal method A cannot be applied try method B; otherwise use default values C.

Operation and results

Diagenesis domains

Authigenic mineral species were identified from thin-section observations of borehole (and surface) samples (Chapter 5). Textural observations allowed the relative ages and conditions of formation of those minerals to be inferred. The experience and interpretive skill of the petrologists were used to synthesise the global picture of diagenesis from the large number of thin-section observations.

The diagenesis knowledge base uses these basic observations coded into sequences of mineral/texture pairs. For instance, the paragenetic sequence, D3Q4K8, would indicate that the creation of a dolomite cement (D3) was followed by quartz cements and overgrowths (Q4) and later by replacement K-feldspar (K8). To create a global paragenesis for a set of samples a pattern-matching technique was used. A template sequence such as the above is chosen from the database. Each thin section is then queried in turn. If a mineral is not present in the template, a new mineral 'object' is created dynamically by the knowledge base (a useful facility of Nexpert Object, rendering it unnecesary to know the dimensions of the problem space before starting to solve it) with any range information attached (e.g. formed after the quartz cements/overgrowths but before the K-feldspar replacement). If the mineral species exists in the template its position is checked and if that does not match that of the template a new mineral phase is created. After testing all the samples requested, the rule base checks all the range limits of all the minerals and phases to determine the relative relationships and resolve any ambiguities if possible. The paragenesis of some minerals may be exactly defined in this way whilst others may remain poorly constrained.

The end result is a relative-time plot of the global paragenetic order of the formation of authigenic minerals.

(Figure 151) shows one such system result for borehole samples from the Kinnerton Sandstone Formation in the northern part of the basin. This paragenesis, although abbreviated, is equivalent in overall form to that deduced for all the SSG sandstones (Chapter 5).

Some mineral/textural and mineral combination/textural relationships are inferred to be indicative of general diagenetic conditions, while others have no such diagnostic value. It is standard practice to use such inferences to categorise parts of mineral paragenetic sequences into eo-, meso- and telodiagenetic periods, corresponding to shallow, pre-compaction changes, deep-burial changes and shallow, post-uplift changes. In the Cheshire Basin, for instance, nodular dolomite is a typical eodiagenesis indicator whilst quartz overgrowths and cements occur during mesodiagenesis and kaolinite pore-filling typifies telodiagenesis (Chapter 5). Using these three categories it is possible to map all seven possible combinations of diagenetic domains, and (Figure 152) shows an example using Thiessen polygons to show all borehole data for the SSG. In the central northern and central-southern parts of the basin there are boreholes with the full eo-, meso- and telodiagenesis combination, whilst more marginal areas tend to have less complete diagenetic domains.

Another capability of the system is to map (using the same Thiessen polygon approach) the occurrence of key authigenic mineral species and combinations. In (Figure 153) we see the distribution of all minerals thought to be associated with the end-Triassic mineralisation event. Four categories are shown: baryte mineralisation, simple sulphide mineralisation, simple sulphide and baryte mineralisation, and complex sulphide mineralisation. This latter category is only recognised at Alderley Edge mine and the Gallantry Bank prospect, which are represented by the blue triangles. Note the concentration of sulphide mineralisation in the north and east parts of the basin.

Handling vagueness and uncertainty in provenance

The Cheshire Basin lay in low latitudes during the Triassic and was part of a north–south system of connecting rifts. The SSG comprises aeolian-deposited arenites borne on easterly winds and fluvial sediments associated with rivers draining from the south or south-east (Chapters 2 and 4). The succeeding MMG comprises mainly the deposits of a continental playa.

The source rocks for the detritus of the SSG probably come from four main areas: (1) Armorica (including Cornubia), (2) London–Brabant massif, (3) Wales, and (4) Pennines (Chapter 4). The exact nature of these source rocks is not known. They are likely to contain mixtures derived from other sources, and none of them contain uniquely diagnostic components that allow the provenance to be determined unequivocally. The inferences that can be made from the mineralogical and geochemical character of the detritus are inherently vague. For example, the linguistically vague statement 'High monazite contents indicate an Armorican source' is typical of the kind of geological interpretation that is possible (and not only in provenance studies). Within the knowledge base for Provenance, therefore, there is a need to be able to express this degree of vagueness, and also to give some comparative measure of the degree of certainty that a hypothesis (e.g. that the sandstones of a particular formation are derived mainly from Armorican rocks) is true. Three models of approximate reasoning have been assessed for use in the system: the subjective Bayesian method, evidential reasoning (Dempster-Shafer theory) and Fuzzy Logic. Fuzzy Logic seems to be the most promising, as it supplies both a capability for expressing vagueness and a measure of certainty of truthfulness. It is being applied with increasing success, for example, in hydrocarbon migration modelling (Kacewicz, 1993). A sample can be a member of more than one fuzzy

class, say moderate and high, because the distributions of classes overlap. A rule using fuzzy logic in Nexpert Object has a hypothesis to be tested. In the example of the hypothesis 'This sample had an Armorican source', the rule is:

If Monazite/Zircon is very high

AND

If Apatite/Rutile is high

AND

If Tourmaline is very FeO rich

AND

If Palaeoflow Direction is northerly

AND

If Nd/Sm is moderate

THEN Hypothesis is True (Certainty = n)

The result of the application of this rule to all samples of the Chester Pebble Beds Formation is shown in (Figure 154). The Chester Pebble Beds represent one of the notable fluvial excursions into the dominantly aeolian sedimentation regime of SSG deposition. The results are couched in terms of four fuzzy classes, from low to very high possibility that the Chester Pebble Beds, in aggregate at a certain borehole, were derived from an Armorican source. It is evident that the samples with a higher level of possibility of derivation from an Armorican source lie in the south of the basin, the area closest to the source and the area where any northerly flowing river system would have debouched into the Cheshire Basin.

Qualitative spatial modelling of end-Triassic red-bed mineralisation

At the end of the Triassic, when the thermal phase of crustal extension in the Cheshire Basin was ending (Evans et al., 1993; Chapter 3), a hydrogeological regime is believed to have existed in which brines from the centre of the basin escaped upwards and outwards towards the basin margin and mixed with a small component of reducing fluid derived from the Carboniferous basement and moving along faults (Chapter 6 and 7). The metal content (particularly Cu) of the mixture was precipitated in faulted sandstones near the top of the SSG. In modelling the mineralisation event, the following factors need to be taken into account:

1. The dynamics of the process by which metals were extracted from the source rocks (the MMG and SSG) is not known, so no useful constraints can be placed on this element of the event.
2. The ambient structural model for the event should be that at 205 Ma. The principal axis of extension was almost east–west (083°), so regional permeability should have been enhanced in a north–south direction. North–south faults seem likely pathways for supplying the deep reducing fluid, whilst east–west transfer faults connecting to these may have supplied the basinal brines (Chapter 3). The transfer faults are small and not well mapped individually, unlike the north–south faults. There is no information on intra-formational zones of high permeability acting as fluid migration conduits.
3. The basin is floored by Upper Carboniferous rocks (capable of supplying the fault-conducted reducing fluids) except in the extreme south-west of the basin, and prospectivity may thus be lower in the south-west.
4. Cross-sectional hydrogeological flow models (Chapter 6) where brines dominate the pore waters show upward, density-driven, flow vectors onto the margins with mixing vortices at major basin-margin faults such as the Red Rock Fault and the Overton Fault. This pattern may have existed at the end of the Triassic (Chapter 7).
5. All the known examples of mineralisation occur in sandstones (chiefly the Helsby Sandstone Formation) near the top of the SSG and close to faulted juxtaposition with mudstones of the MMG (Chapter 7).

The useful spatial components that can be used from these are:

• Lower basin-wide prospectivity in the south-west of the basin.
• Flow vectors perpendicular to the end-Triassic basin margins.
• Increased prospectivity at the intersection of these with the top of the Helsby Sandstone Formation.
• Increased prospectivity where the Helsby Sandstone Formation is cut by north–south faults.
• Increased prospectivity where the Helsby Sandstone Formation and MMG mudstones are faulted together.

In any qualitative spatial model, subjective decisions have to be made concerning values such as the spatial dimensions at which planar intersections are represented, the relative weightings that are given to combinations of factors, and whether results are to be represented in binary form (prospective or non-prospective), by category (high, medium or low), or as a ratio map (on a scale of 0 to 1). Provided a record of these choices is kept and expert guidance is built in, an Expert-GIS can produce some valuable models of this kind. (Figure 155) shows the results of one such simple model: a binary map showing increased prospectivity where the top of the Helsby Sandstone Formation is cut by north-south faults, expressed as a minimum-width grid scale of 1 km. Validating such models for the Cheshire Basin is very difficult because there are so few known surface expressions of mineralisation and their areal extent is very small.

Conclusions

1. A workstation Expert-GIS was created for the Cheshire Basin project, comprising ARC/INFO, NEXPERT OBJECT and ORACLE, which allows the user to interrogate the data and knowledge acquired by earth scientists with different specialist skills. The user does not need to know anything of the workings of the individual components to the system.
2. The system produces both spatial (e.g. maps of values) and aspatial (e.g. lists of values) products. Some of the spatial products can give new insights into basin geology because they are not the standard ways of interrogating data and displaying results (e.g. diagenesis). This increased capability partly derives from a hierarchical menu system for choosing the samples/rock mass that allows the user to make a rapid synthesis of results from different subsets.
3. The successful application of Fuzzy Logic techniques within the provenance part of this geological Expert-GIS indicates a promising area for future development in this type of system. Geological data and analysis are often both vague and uncertain, and a systematic means of expressing vagueness and uncertainty is of great potential value.

References

Most of the references listed below are held in the Library of the British Geological Survey at Keyworth, Nottingham. Copies of the references can be purchased subject to current copyright legislation.

AHLBRANDT, T S, and FRYBERGER, S G. 1982. Introduction to eolian deposits. 11–47 in Sandstone depositional environments. SCHOLLE, P A, and SPEARING, D (editors). Memoir of the American Association of Petroleum Geologists, No. 31.

AITKENHEAD, N, CHISHOLM, J I, and STEVENSON, I P. 1985. Geology of the country around Buxton, Leek and Bakewell. Memoir of the British Geological Survey, Sheet 111, 168 pp.

AL-SHAIEB, Z. 1991. Compartmentation, fluid pressure important in Anadarko exploration. Oil and Gas JournaL July 8, 52–55.

ALI, A D, and TURNER, P. 1982. Authigenic K-feldspar in the Bromsgrove Sandstone Formation (Triassic) of central England. Journal of edimentary Petrology, Vol. 52, 187–197.

ALJUBOURI, Z A. 1972. Geochemistry, origin and diagenesis of some Triassic gypsum deposits and associated sediments in the East Midlands. Unpublished PhD thesis, University of Nottingham.

ALLEN, I L, and HOLLOWAY, S. 1984. The Wessex Basin. Investigation of the geothermal potential of the United Kingdom. (Keyworth, Nottingham: British Geological Survey.)

AMINZADEH, F, and SIMAAN, M. 1991. Expert systems in exploration. 244 pp. (Tulsa: Society of Exploration Geophysicists.)

ANDERSON, F W. 1964. Rhaetic Ostracoda. Bulletin of the Geological Survey of Great Britain, No. 21, 133–174.

ANDERSON, T F, and ARTHUR, M A. 1983. Stable isotopes of oxygen and carbon and their application to sedimentologic and paleoenvironmental problems. 1.1–1.151 in Stable isotopes in sedimentary geology, ARTHUR, M A (editor). SEPM Short Course, No. 10.

ANDERTON, R, BRIDGES, P H, LEEDER, M R, and SELLWOOD, B W. 1979. A dynamic stratigraphy of the British Isles. (London: Allen Unwin.)

APPELO, C A J, and POSTMA, D. 1994. Geochemistry, groundwater and pollution. (Balkema, Rotterdam.) 536 pp.

ARAKEL, A V, and MCCONCHIE, D. 1982. Classification and genesis of calcrete and gypsite lithofacies in paleodrainage systems of inland Australia and their relationship to carnotite mineralization. Journal of Sedimentary Petrology, Vol. 52, 1149–1170.

ARTHURTON, R S. 1973. Experimentally produced halite compared with Triassic layered halite-rock from Cheshire, England. Sedimentology, Vol. 20, 145–160.

ARTHURTON, R S. 1980. Rhythmic sedimentary sequences in the Triassic Keuper Marl (Mercia Mudstone Group) of Cheshire, north-west England. GeologicalJ ourna4 Vol. 15, 43–58.

AUDLEY-CHARLES, M G. 1970. Triassic palaeogeography of the British Isles. Quarterly Journal of the Geological Society of London, Vol. 126, 48–89.

AYORA, C, GARCIA-VEIGAS, J, and PUEYO, J-J. 1993. X-ray microanalysis of fluid inclusions and its application to the geochemical modelling of evaporite basins. Geochimica et Cosmochimica Acta, Vol. 58, 43–55.

BADLEY, M E, PRICE, J D, and BACKSHALL, L C. 1989. Inversion, reactivated faults and related structures: seismic examples from the southern North Sea. 201–209 in Inversion tectonics. COOPER, M A, and WILLIAMS, G D (editors). Special Publication of the Geological Society of London, No. 44.

BALDWIN, B, and BUTLER, C O. 1985. Compaction curves. Bulletin of the American Association of Petroleum Geologists, Vol. 69, 622–626.

BALLING, N, KRISTIANSEN, J I, BREINER, N, POULSEN, K D, RASMUSSEN, R, and SAXOV, S. 1981. Geothermal measurements and subsurface temperature modelling in Denmark. GeoSkrifier Aarhus Universitet, Vol. 16, 1–172.

BARKER, A J. 1990. Introduction to metamorphic textures and microstructures. (Glasgow and London: Blackie.)

BATH, A H, EDMUNDS, W M, and ANDREWS, J N. 1979. Palaeoclimatic trends deduced from the hydrochemistry of a Triassic sandstone aquifer, United Kingdom. Isotope Hydrology, International Atomic Energy Agency, No. 2, 545–566.

BATH, A H, MILODOWSKI, A E, and SPIRO, B. 1987. Diagenesis of carbonate cements in Permo-Triassic sandstones in the Wessex and East Yorkshire-Lincolnshire Basins, UK: a stable isotope study. 173–190 in Diagenesis of sedimentary sequences. MARSHALL, J D (editor). Special Publication of the Geological Society of London, No. 36.

BATH, A H, MILODOWSKI, A E, and STRONG, G E. 1987. Fluid flow and diagenesis in the East Midlands Triassic sandstone aquifer. 127–140 in Fluid flow in sedimentary basins and aquifers. GOFF, J C, and WILLIAMS, B P J (editors). Special Publication of the Geological Society of London, No. 34.

BEALES, F W, and JACKSON, S A. 1966. Precipitation of lead-zinc ores in carbonate reservoirs as illustrated by Pine Point orefield, Canada. Transactions of the Institution of Mining and Metallurgy (Section B), Vol. 75, B278–285.

BECHTEL, A, and HOERNES, S. 1993. Stable isotopic variations in clay minerals: a key to the understanding of Kupferschiefer-type mineralization, Germany. Geochimica et Cosmochimica Acta, Vol. 57, 213–306.

BECHTEL, A, and PUTTMANN, W. 1991. The origin of the Kupferschiefer-type mineralization in the Richelsdorf Hills, Germany, as deduced from stable isotope and organic geochemical studies. Chemical Geology, Vol. 91, 1–18.

BENTON, M J. 1990. The species of Rhynchosaurus, a rhyncosaur (Reptilia, Diapsida) from the Middle Triassic of England. Philosophical Transactions of the Royal Society of London, B328, 213–306.

BENTON, M J, WARRINGTON, G, NEWELL, A J, and SPENCER, P S. 1994. A review of British Middle Triassic tetrapod assemblages. 131–160 in In the shadow of the dinosaurs: Early Mesozoic tetrapods. FRASER, N, and SUES, H-D (editors). (Cambridge University Press.)

BERNARD-GRIFFITHS, J, PEUCAT, J J, SHEPPARD, S, and VIDAL, PH. 1985. Petrogenesis of Hercynian leucogranites from the southern Armorican Massif: contribution of REE and isotopic (Sr, Nd, Pd and O geochemical data to the study of source rock characteristics and ages. Earth and Planetary Science Letters, Vol. 74, 235–250.

BETHKE, C M. 1985. A numerical model of compaction-driven groundwater flow and heat transfer and its application to the palaeohydrology of intracratonic sedimentary basins. Journal of Geophysical Research, Vol. 90, 6817–6828.

BETHKE, C M. 1986. Hydrologic constraints on the genesis of the Upper Mississippi Valley mineral district from Illinois Basin brines. Economic Geology, Vol. 81, 23–249.

BETHKE, C M. 1989. Modelling subsurface flow in sedimentary basins. Geologische Rundschau, Vol. 78, 129–154.

BETHKE, C M, HARRISON, W J, UPSON, C, and ALTANER, S P. 1988. Supercomputer analysis of sedimentary basins. Science, Vol. 239, 261–267.

BINNEY, E W. 1841. Sketch of the geology of Manchester and its vicinity. Transactions of the Geological Society of Manchester, Vol. 1, 35–66.

BINNEY, E W. 1855. On the Permian beds of the north-west of England. Memoir of the Literary and Philosophical Society of Manchester, 2nd Series, Vol. 12, 209–269.

BINNEY, E W. 1862. Additional observations on the Permian beds of south Lancashire. Memoir of the Literary and Philosophical Society of Manchester, 3rd Series, Vol. 2, 29–47.

BOLES, J R. 1978. Active ankerite cementation in the subsurface Eocene of southwest Texas. Contributions to Mineralogy and Petrology, Vol. 68, 13–22.

BORISENKO, A S. 1977. Study of the salt composition of solutions in gas-liquid inclusions in minerals by the cryometric method. Soviet Geology and Geophysics, Vol. 18, 11–19.

BOYLE, R W, BROWN, A C, JEFFERSON, C W, JOWETT, E C, and KIRKHAM, R V. 1989. Sediment-hosted stratiform copper deposits. Special Paper of he Geological Association of Canada, No. 36, 3–38.

BRADLEY, A D, HUGHES, C D, and CHENERY, S R N. 1994. Laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS) of oils and tars - trace metal ratios made easy? Extended abstracts, 1994 British Organic Geochemical Society meeting, Aberdeen, July 1994.

BRADLEY, A D, HUGHES, C D, HOBBS, S F, and AULT, L. 1992. Determination of trace metals in oils and tars. Part 1: Initial comparison of data obtained by DRES and ICP-AES using samples from Staffordshire and Shropshire. British Geological Survey Technical Report, WI/92/9.

BRADLEY A D, and NICHOLSON, R A. 1995. Methods for geochemical fingerprinting of oils and tars. British Geological Survey Technical Report, WI/95/21.

BRASSINGTON, F C, LUCEY, P A, and PEACOCK, A J. 1992. The use of down-hole focused electric logs to investigate saline groundwaters. Quarterly Journal of Engineering Geology, Vol. 25. 343–349.

BRASSINGTON, F C, and RUSHTON, K R. 1987. A rising water table in central Liverpool. Quarterly Journal of ngineering Geology, Vol. 20. 151–158.

BRASSINGTON, F C, and WALTHALL, S. 1985. Field techniques using borehole packers in hydrogeological investigations. Quarterly Journal of Engineering Geology, Vol. 37, 1–72.

BRITISH GEOLOGICAL SURVEY. 1985a. Map 1: Pre-Permian geology of the United Kingdom (South). Scale 1: 1 000 000. (Keyworth, Nottingham: British Geological Survey on behalf of Petroleum Engineering Division, Department of Energy.)

BRITISH GEOLOGICAL SURVEY. 1985b. Map 2. Contours on the top of the pre-Permian surface of the United Kingdom (South). Scale 1:1000 000. (Keyworth, Nottingham: British Geological Survey on behalf of Petroleum Engineering Division, Department of Energy.)

BRITISH GEOLOGICAL SURVEY. 1989. Hydrogeological map of Clwyd and the Cheshire Basin. 1:200 000. (Keyworth, Nottingham: British Geological Survey.)

BRITISH GEOLOGICAL SURVEY. 1990. Mid-Wales and Marches. Solid Geology. 1:250 000. (Southampton: Ordnance Survey for the British Geological Survey.)

BRITISFI GEOLOGICAL SURVEY. 1995. World mineral statistics 1990–94: production, exports, imports. (Keyworth, Nottingham: British Geological Survey.)

BRITISH GEOLOGICAL SURVEY. 1996. Regional geochemistry of north-east England. (Keyworth, Nottingham: British Geological Survey.)

BROCKBANK, W. 1891. Notes on the discovery of Estheria minuta var. Brodieana, by Mr C E De Rance, FGS, at Alderley Edge. Memoir of the Proceedings of the Manchester Literary and Philosophical Society, Vol. 4, 12–13, 31–32.

BRODIE, J, and WHITE, N. 1994. Sedimentary basin inversion caused by igneous underplating: Northwest European continental shelf. Geology, Vol. 22, 147–150.

BROOKINS, D G. 1984. Geochemical aspects of radioactive waste disposal. (New York: Springer-Verlag.)

BROWN, A C. 1993. Sediment-hosted stratiform copper deposits. Geoscience Canada, Vol. 19, No. 3, 125–141.

BROWN, I, and TRINDER, B. 1979. The Tar Tunnel. Ironbridge Gorge Museum Trust Guide, No. 4.04. 11 pp.

BRUGMAN, W A. 1986. Late Scythian and Middle Triassic palynostratigraphy in the Alpine realm. Albertiana, Vol. 5, 19–20.

BRUTON, C J. 1989. Predicting mineral dissolution and precipitation during burial: synthetic diagenetic sequences. 111–119 in JACKSON, K J, and BOURCIER, W L (editors). Proceedings of he Workshop on Geochemical Modelling, CONE-8609134, Lawrence Livermore National Laboratory, Livermore, California.

BURGESS, W G. 1981. Drill-stem tests on Permo-Triassic sandstones to investigate their geothermal potential. Investigations of the geothermal potential of the UK (Keyworth, Nottingham: Institute of Geological Sciences.)

BURGESS, W G, BURLEY, A J, DOWNING, R A, EDMUNDS, W M, PRICE, M, RICHARDSON, S W, and JONES, M Q W. 1981. The Marchwood Geothermal Borehole - a preliminary assessment of the resource. Investigations of the geothermal potential of the UK (Keyworth, Nottingham: Institute of Geological Sciences.)

BURGESS, W G, EDMUNDS, W M, ANDREWS, J N, KAY, R L F, and LEE, D J. 1980. The hydrogealogy and hydrochemistry of the thermal waters in the Bath-Bristol basin. Investigations of the geothermal potential of the UK (Keyworth, Nottingham: Institute of Geological Sciences.)

BURLEY, A J, EDMUNDS, W M, and GALE, I N. 1984. Catalogue of geothermal data for the land area of the United Kingdom. Investigation of the geothermal potential of the UK British Geological Survey. (Keyworth, Nottingham: British Geological Survey.)

BURLEY, S D. 1984. Patterns of diagenesis in the Sherwood Sandstone Group (Triassic), United Kingdom. Clay Mineralogy, Vol. 19, 403–440.

BURLEY, S D. 1987. Diagenetic modelling in the Triassic Sherwood Sandstones Group of England and its offshore equivalents, United Kingdom continental shelf. Unpublished PhD thesis, University of Hull.

BURNOTTE, E, PIRARD, E, and MICHEL, G. 1989. Genesis of grey monazites: evidence from the Palaeozoic of Belgium. Economic Geology, Vol. 84,1417–1429.

BUSHELL, T P. 1986. Reservoir geology of the Morecambe Field. 189–208 in Habitat of Palaeozoic gas in NW Europe. BROOKS, J, GOFF, J, and VAN HOORN, B (editors). Special Publication of the Geological Society of London, Vol. 23.

BUTLER, W C. 1992. A suggested rationale and selected bibliography for using rule-based expert computer systems (artificial intelligence) in recurrent petroleum resource assessments. United States Geological Survey Open-File Report, 126 pp.

CALVER, M A. 1968. The distribution of Westphalian marine faunas in northern England and adjoining areas. Proceedings of the Yorkshire Geological Society, Vol. 37, 1–72.

CALVERT, A F. 1915. Salt in Cheshire. (London: E & F N Spoon Ltd.)

CAMPBELL, J E. 1982. Permeability characteristics of the Permo-Triassic sandstones of the Lower Mersey Basin. Unpublished MSc thesis, University of Birmingham.

CAMPBELL, J E et al. 1981. Saline Groundwater investigation Phase 1 - Lower Mersey Basin, Final Report to the North West Water Authority, pp.81; Appendix Vol. 1, Geology, 163 pp. Vol. 3, Geophysical investigations, 77 pp. Birmingham University, Department of Civil Engineering.

CAMPBELL-SMITH, W. 1963. Description of the igneous rocks represented among pebbles from the Bunter Pebble Beds of the Midlands of England. Bulletin of the British Museum (Natural History), London, (Mineralogy), Vol. 2,1–17.

CARLON, C J. 1975. The geology and geochemistry of some British barite deposits. Unpublished PhD thesis, University of Manchester.

CARLON, C J. 1981a. The Gallantry Bank copper mine, Bickerton, Cheshire. With a review of mining in the Triassic rocks of the Shropshire-Cheshire Basin. British Mining, Vol. 16,1–50.

CARLON, C J. 1981b. The Eardiston copper mine, Shropshire. Bulletin of the Peak District Mines Historical Society, Vol. 8,29–42.

CARLON, C J. 1981c. Mineral deposits and metal mines in Cheshire. Proceedings of the Westmorland Geological Society, Vol. 3,15–21.

CAROTHERS, W W, and KHARAKA, Y K. 1978. Aliphatic acid anions in oil-field waters - implications for origin of natural gas. Bulletin of the American Association of Petroleum Geologists, Vol. 12, 2441–2453.

CARPENTER, A B, TROUT, M L, and PICKETT, E E. 1974. Preliminary report on the origin and chemical evolution of lead-and zinc-rich oilfield brines in central Mississippi. Economic Geology, Vol. 69,1191–1206.

CATHLES, L M, OSZCZEPALSKI, S, and JOWETT, E C. 1993. Mass balance evaluation of the late diagenetic hypothesis for Kupferschiefer Cu mineralization in the Lubin Basin of southwestern Poland. Economic Geology, Vol. 88,948–956.

CATHLES, L M, and SMITH, A T. 1983. Thermal constraints on the formation of Mississippi Valley-type lead-zinc deposits and their implications for episodic basin dewatering and deposit genesis. Economic Geology, Vol. 78,983–1002.

CERMAK, V, and BODRI, L. 1986. Two-dimensional temperature modelling along five East European geotraverses. Journal of Geodynamics, Vol. 5,135–163.

CERMAK, V, and RYBACH, L. 1982. Thermal conductivity and specific heat of minerals and rocks. 305–343 in Landolt-Bornstein numerical data and fiinctional relationships in science and technology, Group V Geophysics and space research, Volume 1/a Physical properties of rocks. ANGENHEISTER, G (editor). (Berlin: Springer Verlag.)

CHADWICK, R A. 1985. Permian, Mesozoic and Cenozoic structural evolution of England and Wales in relation to the principles of extension and inversion tectonics. In Atlas of onshore sedimentary basins in England and Wales: post-Carboniferous tectonics and stratigraphy. WHITTAKER, A (editor). (Glasgow: Blackie and Son.)

CHADWICK, R A. 1993. Aspects of basin inversion in southern Britain. Journal of the Geological Society of London, Vol. 150, 311–322.

CHADWICK, R A. 1997. Fault analysis of the Cheshire Basin, north-west England. 297–313 in The Petroleum Geology of the Irish Sea and adjacent areas. Meadows, N S, Trueblood, S, Hardman, M, and Cowan, G (editors). Special Publication of the Geological Society of London, No. 124.

CHADWICK, R A, and EVANS, D J. 1995 The timing and direction of Permo-Triassic extension in southern Britain. 161–192 in Permian and Triassic rifting in north-west Europe. BOLDY, S A R (editor). Special Publication of the Geological Society of London, No. 91.

CHADWICK, R A, EVANS, D J, SMITH, N J P, ROWLEY, W J, WALKER, A S D, BULAT, J, BIRCH, B, and YOUNG, S R. 1994a. The structure and evolution of the Cheshire Basin. British Geological Survey Technical Report, WA/94/72.

CHADWICK, R A, KENOLTY, N, and WHITTAKER, A. 1983. Crustal structure beneath southern England from deep seismic reflection profiles. Journal of the Geological Society of London, Vol. 140, 893–911.

CHADWICK, R A, KIRBY, G A, and BAILY, H E. 1994b. The post-Triassic structural evolution of northwest England and adjacent parts of the Irish Sea. Proceedings of the Yorkshire Geological Society, Vol. 50,91–102.

CHAMBERS, J M, CLEVELAND, W S, KLEINER, B, and TUKEY, P A. 1983. Graphical methods for data analysis. Wadworth International Group, Belmont, California. 395 pp. (Boston, Mass.: Duxbury Press.)

CHISHOLM, J I, CHARSLEY, T J, and AITKENHEAD, N. 1988. Geology of the country around Ashbourne. Memoir of the British Geological Survey, Sheet 124 (England and Wales.)

COCKS, L R M. 1993. Triassic pebbles, derived fossils and the Ordovician to Devonian palaeogeography of Europe. Journal of the Geological Scociety of London, Vol. 150,219–226.

COLEMAN, M L, and MOORE, M P. 1978. A method for the direct reduction of sulphates to sulphur dioxide for isotopic analysis. Analytical Chemistry, Vol. 50,1594–1595.

COLLINSON, J D. 1988. Controls on Namurian sedimentation in the Central Province basins of northern England. 85–100 in Sedimentation in a synorogenic basin complex: the Upper Carboniferous of Northwest Europe. BESLY, B, and KELLING, G (editors). (Glasgow: Blackie.)

COLLINSON, J D, JONES, C M, and WILSON, A A. 1977. The Marsdenian (Namurian R2) succession west of Blackburn: implications for the evolution of the Pennine Delta Systems. Geological Journal, Vol. 12,59–76.

COLTER, V S. 1978. Exploration for gas in the Irish Sea. Geologie en Mijnbouw, Vol. 51,503–516.

COLTER, V S, and BARR, K W. 1975. Recent developments in the geology of the Irish Sea and Cheshire basins. 61–73 in Petroleum geology and the continental shelf of North West Europe: Vol. 1, Geology. WOODLAND, A W (editor). (London: Applied Science Publishers.)

COLTER, V S, and EBBERN, J. 1978. The petrography and reservoir properties of some Triassic sandstones of the Northern Irish Sea Basin. Journal of the Geological Society of London, Vol. 135,57–62.

COLTER, V S, and EBBERN, J. 1979. SEM studies of Triassic reservoir sandstones from the Morecambe Field, Irish Sea, UK. Scanning electron microscopy, Vol. 135,531–538.

COOPER, D C, and READ, D. 1984. The distribution of monazite in panned stream sediment concentrates from Britain: palaeogeographic implications. Report of the British Geological Survey, No. 84/1,1–4.

COPE, F W. 1939. Oil occurrences in south-west Lancashire. Bulletin of the Geological Survey, No. 2,18–25.

COPE, J C W, GETTY, T A, HOWARTH, M K, MORTON, N, and TORRENS, H S. 1980. A correlation of Jurassic rocks in the British Isles. Part I: Introduction and Lower Jurassic. Special Report of the Geological Society of London, No. 14.

COPE, J C W, INGHAM, J K, and RAWSON, P F. 1992. Atlas of palaeogeography and lithofacies. Memoir of the Geological Society of London, Vol. 13.

COPESTAKE, P. 1989. Triassic. 97–124 in Strati graphical atlas of fossil foraminifira. 2nd edition. JENKINS, D G, and MURRAY, J W (editors). (Chichester: Ellis Horwood Ltd.)

COPESTAKE, P, and JOHNSON, B. 1989. The Hettangian to Toarcian (Lower Jurassic). 129–188 in Stratigraphical atlas of fossil foraminifira. 2nd edition. JENKINS, D G, and MURRAY, J W (editors). (Chichester: Ellis Horwood Ltd.)

COUSTEAU, H, RUMUAU, L, and SOURISSE, C. 1975. Classification hydrodynamique des bassins sedimentaires: utilisation combinee avec d'autres methodes pour rationaliser l'exploration des bassins non-productifs. Proceedings of the World Petroleum Congress, Vol. 2,105–109.

COWAN, G. 1993. Identification and significance of aeolian deposits within the dominantly fluvial Sherwood Sandstone Group of the East Irish Sea Basin UK. 231–245 in Characterisation of fluvial and aeolian reservoirs. NORTH, C, and PROSSER, D (editors). Special Publication of the Geological Society of London, No. 73.

COWARD, M P. 1990. The Precambrian, Caledonian and Variscan framework to NW Europe. 1–34 in Tectonic events responsible for Britain's oil and gas reserves. HARDMAN, R F P, and BROOKS, J (editors). Special Publication of the Geological Society of London, No. 55.

COWARD, M P. 1993. The effect of Late Caledonian and Variscan continental escape tectonics on basement structure, Paleozoic basin kinematics and subsequent Mesozoic basin development in NW Europe. 1095–1108 in Petroleum geology of Northwest Europe: Proceedings of the 4th Conference. PARKER, J R (editor). (London: Geological Society.)

CRAIG, J, FITCHES, W R, and MALTMAN, A J. 1982. Chlorite-mica stacks in low-strain rocks from central Wales. Geological Magazine, Vol. 119,243–256.

CULLERS, R L, CHAUDHURI, S, KILBANE, N, and KOCH, R. 1979. Rare earths in size fractions and sedimentary rocks of Pennsylvanian-Permian age from the mid-continent of the USA. Geochimica et Cosmochimica Acta, Vol. 43,1285–1302.

CURRY, D. 1992. Cainozoic. 389–411 in Geology of England and Wales. DUFF, P, McL D, and SMITH, A J (editors). (Geological Society of London.)

DARBYSHIRE, D P F, and SHEPHERD, T J. 1994. Nd and Sr isotope constraints on the origin of the Cornubian batholith, Southwest England. Journal of the Geological Society of London, Vol. 151, 795–802.

DARRAGI, F, and TARDY, Y. 1987. Authigenic trioctahedral smectites controlling pH, alkalinity, silica and magnesium concentrations in alkaline lakes. Chemical Geology, Vol. 63, 59–72.

DEAN, W E, DAVIS, G R, and ANDERSON, R Y. 1975. Sedimentological significance of nodular and laminated anhydrite. Geology, Vol. 3,367–372.

DELAIR, J B, and SARJEANT, W A S. 1985. History and bibliography of the study of fossil vertebrate footprints in the British Isles. Supplement 1973–1983. Palaeogeography Palaeoclimatology Palaeoecology, Vol. 49,123–160.

DEMAISON, G, and MURRIS, R J. 1984. Petroleum geochemistry and basin evaluation. Memoir of the American Association of Petroleum Geologists, No. 35.426 pp.

DEPARTMENT OF TRADE AND INDUSTRY. 1994. The Energy Report: Oil and Gas Resources of the United Kingdom. (London: HMSO.). 153 pp.

DE RANCE, C E. 1895. Results of the Salt Union boring at Marston near Northwich. Transactions of the Manchester Geological Society, Vol. 22,269–302.

DEWEY, H, and EASTWOOD, T. 1925. Copper ores of the Midlands, Wales, the Lake District and the Isle of Man. Special Report on the Mineral Resources of Great Britain, Memoir of the Geological Survey, No. 30,1–87.

DEWEY, J F. 1982. Plate tectonics and the evolution of the British Isles. Journal of the Geological Society of London, Vol. 139, 371–412.

D'LEMOS, R S, and BROWN, M. 1993. Sm-Nd isotope characteristics of late Cadomian granite magmatism in northern France and the Channel Islands. Geological Magazine, Vol. 130, 797–804.

DONNOT, M, GUIGUES, J, LULZAC, Y, MAGNIEN, A, PARFENOFF, A, and PICOT, P. 1973. Un nouveau type de gisement d'europium: la monazite grise a europium en nodules dans les schistes paleozoiques de Bretagne. Mineralium Deposita, Vol. 8,7–18.

DONOVAN, R N, and FOSTER, R J. 1972. Subaqueous shrinkage cracks from the Caithness Flagstone Series (Middle Devonian) of north-east Scotland. Journal of Sedimentary Petrology, Vol. 42, 309–317.

DOWNING, R A, EDMUNDS, W M, and GALE, I N. 1987. Regional groundwater flow in sedimentary basins in the UK. 105–125 in Fluid flow in sedimentary basins and aquifers. GOFF, J C, and WILLIAMS, B P J (editors). Special Publication of the Geological Society of London, No 34.

DOWNING, R A, and GRAY, D A. 1986. Geothermal energy - the potential of the United Kingdom, 187 pp. (London: HMSO for British Geological Survey.)

DOWNING, RA, and HOWITT, F. 1969. Saline groundwaters in the Carboniferous rocks of the English East Midlands in relation to the geology. Quarterly Journal of Engineering Geology, Vol. 1, 241–260.

DOWNING R A, and PENN I E. 1991. Groundwater flow during the development of the Wessex Basin and its bearing on hydrocarbon and mineral resources. British Geological Survey Research Report, SD/91/1.

DUNHAM, K C, and WILSON, A A. 1985. Geology of the Northern Pennine Orefield: Volume 2, Stainmore to Craven. Economic Memoir of the Geological Survey of Great Britain, Sheets 40,41,50 (England and Wales).

DURRANCE, E M. 1986. Radioactivity in geology: principles and applications. Chapter 6, The geochemistry of uranium and thorium. (Chichester: Ellis Horwood.)

DZOMBAK, D A, and MOREL, F M M. 1990. Surface complexation modelling. (New York: Wiley Interscience.)

EAKIN, P A. 1989. Isotopic and petrographic studies of uraniferous hydrocarbons from around the Irish Sea Basin. Journal of the Geological Society of London, Vol. 146, 663–673.

EARP, J R, and TAYLOR, B J. 1986. Geology of the country around Chester and Winsford. Memoir of the British Geological Survey, Sheet 109. 120 pp.

EBBERN, J. 1981. The geology of the Morecambe Gas Field. 485–493 in Petroleum geology of the continental shelf ofNW Europe. ILLING, L V, and HOBSON, G D (editors). (London: Institute of Petroleum.)

ECKSTRAND, 0 R (editor). 1984. Canadian mineral deposit types: a geological synopsis. Economic Geology Report of the Geological Survey of Canada, Vol. 36. 86 pp.

EDMUNDS, W M. 1975. Geochemistry of the brines in the Coal Measures of north-east England. Transactions of the Institution of Mining and Metallurgy, Vol. 84, B39-B52.

EDMUNDS, W M. 1986. Geochemistry of geothermal waters in the U.K. 111–122 in Geothermal energy: the potential in the United Kingdom. DOWNING, R A, and GRAY, D A (editors). (London: Her Majesty's Stationery Office.)

EDMUNDS, W M, BATH, A H, and MILES, D L. 1977. Pore fluid chemistry of the Bridport sands and the Bunter Sandstone intervals of the Winterbourne Kingston Borehole. Report of the Institute of Geological Sciences, WD/ST/77/3.

EDMUNDS, W M, MILES, D L, and COOK, J M. 1984. A comparative study of sequential redox processes in three British aquifers. 55–70 in Hydrochemical balances of freshwater systems. ERICKSSON, E (editor). IAHS Publication, No. 50.

EUGSTER, H P. 1980. Geochemistry of evaporitic lacustrine deposits. Annual Reviews in Earth and Planetary Sciences, Vol. 8, 35–63.

EUGSTER, H P. 1985. Oil shales, evaporites and ore deposits. Geochimica et Cosmochimica Acta, Vol. 49, 619–635.

EUGSTER, H P. 1989. Geochemical environments of sediment-hosted Cu-Pb-Zn deposits. 111–126 in Sediment-hosted stratiform copper deposits. BOYLE, R W, BROWN, A C, JEFFERSON C W, JOWETT, E C, and KIRKHAM, R V (editors). Special Paper of the Geological Association of Canada, Vol. 36.

EVANS, D J, REES, J G, and HOLLOWAY, S. 1993. The Permian to Jurassic stratigraphy and structural evolution of the Cheshire Basin. Journal of the Geological Society of London, Vol. 150, 857–870.

EVANS D J, REES, J G, HOLLOWAY, S, SMITH, N J P, and CHADWICK, R A. 1994. Reply to POOLE, E G. Discussion on the Permian to Jurassic stratigraphy and structural evolution of the central Cheshire Basin. Journal of the Geological Society of London, Vol. 151, 893–895.

EVANS, J A. 1989. Resetting of Rb-Sr whole-rock isotope systems during low-grade metamorphism, North Wales. Unpublished PhD thesis, University of London.

EVANS, L J, and ADAMS, W A. 1975. Chlorite and illite in some Lower Palaeozoic mudstones of mid-Wales. Clays and Clay Minerals, V 10, 387–397.

EVANS, W B, WILSON, A A, TAYLOR, B J, and PRICE, D. 1968. Geology of the country around Macclesfield, Congleton, Crewe and Middlewich. Memoir of the Geological Survey of Great Britain, Sheet 110. 328 pp.

EWBANK, G, MANNING, D A C, and ABBOTT, G D. 1995. The relationship between bitumens and mineralization in the South Pennine Orefield central England. Journal of the Geological Society of London, Vol. 152, 751–765.

FALCON, N L, and KENT, P E. 1960. Geological results of petroleum exploration in Britain, 1945–1957. Memoir of the Geological Society of London, No. 2. 56 pp.

FEIN, J B. 1991. Experimental study of aluminium-, calcium-and magnesium-acetate complexing at 80°C. Geochimica et Cosmochimica Acta, Vol. 55, 955–964.

FERRIER, G, PAN, S Y P, WADGE, G, and McDERMOTT, C. 1993. An integrated GIS and expert system for the geological interpretation of a sedimentary basin. Proceedings of the 3rd European Conference on GIS, Genoa, 738–745.

FERRIER, G, and WADGE, G. 1994. An integrated GIS and knowledge based system for the geological interpretation of a sedimentary basin. Proceedings of the 4th European Conference on GIS, Paris. 5 pp.

FISHER, M J. 1972a. A record of palynomorphs from the Waterstones (Triassic) of Liverpool. Geological Journal, Vol. 8, 17–22.

FISHER, M J. 1972b. The Triassic palynofloral succession in England. Geoscience and Man, Vol. 4 , 101–109.

FISCHER, R S, and KREITLER, C W. 1987. Geochemistry and hydrodynamics of deep-basin brines, Palo Duro Basin, Texas, USA. Applied Geochemistry, Vol. 2, 459–476.

FITCH, F, MILLER, J A, and THOMPSON, D B. 1966. The palaeogeographic significance of isotopic age determinations on detrital micas from the Triassic of the Stockport-Macclesfield district, Cheshire, England. Palaeogeography, Palaeoclimatology, Palaeoecology, Vol. 2, 281–312.

FLETCHER, S W. 1977. An examination of the hydrogeological characteristics of the Bunter Sandstone aquifer, in the area north of Shrewsbury. Unpublished MSc thesis, University of Birmingham.

FLINT, S. 1987. Diagenesis of Cainozoic playa sandstones of Northern Chile; implications for Andean uplift and metallogeny. Sedimentology, Vol. 34, 11–29.

FONTES, J-CH, and MATRAY, J M. 1993. Geochemistry and origin of formation brines from the Paris Basin, France. Chemical Geology, Vol. 109, 149–175.

FORSTER, S C, and WARRINGTON, G. 1985. Geochronology of the Carboniferous, Permian and Triassic. 99–113 in The chronology of the geological record. SNELLING, N J (editor). Memoir of the Geological Society of London, No. 10.

FOX-STRANGWAYS, C. 1905. The geology of the country between Derby, Burton-on-Trent, Ashby-de-la-Zouch and Loughborough. Memoir of the British Geological Survey, Sheet 141.

FRANKS, S G, and FORESTER, R W. 1984. Relationships among secondary porosity, pore-fluid chemistry and carbon dioxide, Texas Gulf Coast. 63–79 in Clastic diagenesis. MACDONALD, D A, and SURDAM, R C. Memoir of the American Association of Petroleum Geologists, No. 37.

FRASER, A J, and GAWTHORPE, R L. 1990. Tectonostratigraphic development and hydrocarbon habitat of the Caledonides in northern England. 49–86 in Tectonic events responsible for Britain's oil and gas reserves. HARDMAN, R F P, and BROOKS, J (editors). Special Publication of the Geological Society of London, No. 55.

FRASER, A J, NASH, D F, STEELE, R P, and EBDON, C C. 1990. A regional assessment of the intra-Carboniferous play of northern England. 417–440 in Classic petroleum provinces, BROOKS J (editor). Special Publication of the Geological Society of London, No. 50.

FRIEDMAN, I, and O'NEIL, J R. 1977. Compilation of stable isotope fractionation factors of geochemical interest. In Data of geochemistry. Chapter KK. FLEISCHER, M (editor). Professional Paper of the US Geological Survey, 440-KK.

FRITZ, P, and SMITH, D. 1970. The isotopic composition of secondary dolomites. Geochimica et Cosmochimica Acta, Vol. 34, 1161–1173.

FROSTICK, L E, and REID, I (editors). 1987. Desert sediments ancient and modern. Special Publication of the Geological Society of London, No. 35.

GALE, I N, EVANS, C J, EVANS, R B, SMITH, I F, HOUGHTON, M T, and BURGESS, W G. 1984. The Permo-Triassic aquifers of the Cheshire and West Lancashire basins. Investigation of the geothermal potential of the UK 1–44. (Keyworth, Nottingham: British Geological Survey.)

GALEHOUSE, J S. 1971. Point-counting. 385–407 in Procedures in sedimentary petrology. CARVER, R E (editor). (New York: WileyInterscience.)

GARLICK, W G. 1961. The syngenetic theory. 146–165 in The geology of the Northern Rhodesian Copperbelt. MENDELSOHN, F (editor). (London: MacDonald.)

GARLICK, W G. 1989. Genetic interpretation from ore relations to algal reefs in Zambia and Zaire. 471–498 in Sediment-hosted stratiform copper deposits. BOYLE, R W, BROWN, A C, JEFFERSON, C W, JOWETT, E C, and KIRKHAM, R V (editors). Special Paper of the Geological Association of Canada, No. 36.

GEBSKI, J S, WHEILDON, J, and THOMAS-BETTS, A. 1987. Investigations of the UK heat flow field (1984–1987). Investigation of the geothermal potential of the UK (Keyworth, Nottingham: British Geological Survey.)

GEINITZ, H B. 1889. Ueber die rothen and bunten Mergel der oberen Dyas bei Manchester. Ges. Isis in Dresden, Abh. Vol. 3, 48–57.

GEINITZ, H B. 1890. On the red and variegated marls of the Upper Dyas, near Manchester. Transactions of the Geological Society of Manchester, Vol. 20,537–554.

GIBBS, A D. 1984. Structural evolution of extensional basin margins. Journal of the Geological Society of London, Vol. 141,609–620.

GIBSON, W. 1925. The geology of the country around Stoke-on-Trent. Memoir of the Geological Survey of Great Britain. 112 pp.

GIORDANO, T H, and DRUMMOND, S E. 1991. The Potentiometric determination of stability constants for zinc acetate complexes in aquous solutions to 295°C. Geochemica et Cosmochimica Acta, Vol. 55, 2401–2415.

GLENNIE, K W, and BOEGNER, P. 1981. Sole Pit inversion tectonics. 110–120 in Petroleum geology of the continental shelf of north-west Europe. ILLING, L V, and HOBSON, G D (editors). (London: Heyden.)

GLENNIE, K W, and BULLER, A T. 1983. The Permian Weisliegend of NW Europe: the partial deformation of aeolian sand dunes caused by the Zechstein transgression. Sedimentary geology, Vol. 35, 43–81.

GLOVER, B W, HOLLOWAY, S, and YOUNG, S R. 1993a. An evaluation of coalbed methane potential in Great Britain. British Geological Survey Technical Report, WA/93/24.

GLOVER, B W, POWELL, J H, and WATERS, C N. 1993b. Etruria Formation (Westphalian C) palaeoenvironments and volcanicity on the southern margins of the Pennine Basin, South Staffordshire, England. Journal of the Geological Society of London, Vol. 150, 737–750.

GRAHAM, J I. 1922. Rock temperatures in the Coal Measures of Great Britain. Transactions of the Institution of Mining Engineers, Vol. 63, 343–399.

GREEN, G W, and MELVILLE, R V. 1956. The stratigraphy of the Stowell Park Borehole. Bulletin of the Geological Survey of Great Britain, No. 11,1–66.

GREEN, P F, DUDDY, I R, BRAY, R J, and LEWIS, C L E. 1993. Elevated palaeotemperatures prior to early Cainozoic cooling throughout the UK region: implications for hydrocarbon generation. 1067–1074 In: Petroleum Geology of North West Europe, Proceedings of the 4th Conference. PARKER, J R (editor). Geological Society of London.

GREER, R T. 1978. Evaluation of pyrite particle size, shape and distribution factors. Energy Sources, Vol. 4,23–51.

GUERROT, C, PEUCAT, J J, CANDEVILA, R, and Dosso, L. 1989. Archean protoliths within Early Proterozoic granulite crust of the west European Hercynian belt: Possible relics of the west African craton. Geology, Vol. 17,241–244.

GUIHAUMOU, N. 1993. Paleotemperatures inferred from fluid inclusions in diagenetic cements: implications for the thermal history of the Paris basin. European Journal of Mineralogy, Vol. 5, 1217–1226.

GUSTAFSON, L B, and Wnilitms, N. 1981. Sediment-hosted stratiform deposits of copper, lead and zinc. 139–178 in Economic Geology: 75th Anniversary Volume. SKINNER, B J (editor). (El Paso, Texas: Economic Geology Publishing Company.)

HALLAM, A, HANCOCK, J M, LABRECQUE, J L, LOWRIE, W, and CHANNELL, J E T. 1985. Jurassic to Paleogene: Part 1. Jurassic and Cretaceous geochronology and Jurassic to Paleogene magnetostratigraphy. 118–140 in The chronology of the geological record. SNELLING, N J (editor). Memoir of the Geological Society of London, Vol. 10.

HALLSWORTH, C. 1991. Heavy minerals from the Millstone Grit of the 1:50 000 Sheet 59 (Lancaster): implications for provenance. British Geological Survey Technical Report, WH/91/399R.

HANSLEY, P F. 1987. Petrologic and experimental evidence for the etching of garnets by organic acids in the Upper Jurassic Morrison Formation, northwestern New Mexico. Journal of Sedimentary Petrology, Vol. 57, 666–681.

HARDIE, L A. 1990. The roles of rifting and hydrothermal CaCl₂ brines in the origin of potash evaporites: an hypothesis. American Journal of Science, Vol. 290,43–106.

HARDING, T P. 1985. Seismic characteristics and identification of negative flower-structures, positive flower-structures and positive structural inversion. Bulletin of the American Association of Petroleum Geologists, No. 69/4, 582–600.

HARDMAN, M, BUCHANAN, J, HERRINGTON, P, and CARR, A. 1993. Geochemical modelling of the East Irish Sea Basin: its influence on predicting hydrocarbon type and quality. 809–821 in Petroleum geology of Northwest Europe. PARKER, J R (editor). Proceedings of the 4th Conference. (London: Geological Society.)

HARRIMAN, G, and MILES, J. 1995. New oil seep in the heart of Liverpool. Petroleum Exploration Society of Great Britain Newsletter, July, 1995,88–89.

HARRISON, W J, and SUMMA L L. 1991. Paleohydrogeology of the Gulf of Mexico basin. American Journal of Science, 291,109–176.

HARRISON, W J, and THYNE, G D. 1992. Predictions of diagenetic reactions in the presence of organic acids. Geochimica et Cosmochimica Acta, Vol. 56, 565–586.

HARTLEY, A, FLINT, S, and TURNER, P. 1991. Analcime: a characteristic authigenic phase of Andean alluvium, northern Chile. Geological Journal, Vol. 26, 189–202.

HASLAM, H W. 1982. The Marl Slate (Kupferschiefer) in the Southern North Sea Basin. Institute of Geological Sciences Mineral Reconnaissance Programme Report, No. 52,19–26.

HASLAM, H W, and SANDON, P T S. 1991. The geochemistry of some red-bed formations in the United Kingdom. British Geological Survey Technical Report, WP/90/2.

HASLAM, J, ALLBERRY, E C, and MOSES, G. 1950. The bromine content of the Cheshire salt deposit and of some borehole and other brines. Analyst, Vol. 75, 352–356.

HAUGHTON, P D W, TODD, S P, and MORTON, A C. 1991. Sedimentary provenance studies. 1–12 in Developments in sedimentary provenance studies. MORTON, A C, TODD, S P, and HAUGHTON, P D (editors). Special Publication of the Geological Society of London, No. 57.

HAWKINS, P J. 1978. Relationship between diagenesis, porosity reduction and oil emplacement in late Carboniferous sandstone reservoirs, Bothamsall Oilfield, E Midlands. Journal of the Geological Society of London, Vol. 135, 7–24.

HELGESON, H C, KIRKHAM, D H, and FLOWERS, G C. 1981. Theoretical prediction of the thermodynamic behaviour of aqueous electrolytes at high pressures and temperatures IV: calculation of activity coefficients, osmotic coefficients, and apparent molal and standard and relative partial molal properties to 5 kb and 600°C. American Journal of Science, Vol. 281, 1241–1516.

HEM, J D. 1989. Study and interpretation of the chemical characteristics of natural water. Water-Supply Paper of the United States Geological Survey, No. 2254.

HENNETT, R J C, CRERAR, D A, and SCHWARTZ, J. 1988. Organic complexes in hydrothermal systems. Economic Geology, Vol. 83, 742–764.

HENNEY, P J, PHARAOH, T C, CARNEY, J, GASKARTH, J W, and LENG, M J. 1993. A geochemical and Sr-Nd isotope study of the late Ordovician lamprophyric magmatism in the English Midlands. Volcanic Studies Group, 20 November 1993.

HESS, J C, and LIPPOLT, H J. 1986. ⁴⁰Ar/³⁹Ar ages of tonstein and tuff sanidines: new calibration points for the improvement of the Upper Carboniferous time scale. Chemical Geology (Isotope Geoscience Section), Vol. 59, 143–154.

HIGGO, J J W, CRAWFORD, M B, and METCALFE, R. 1993. The role of sorption in metal mobilisation from red beds into stratiform copper deposits. British Geological Survey Technical Report, WE/93/12.

HIRST, D M, and DUNHAM, K C. 1963. Chemistry and petrography of the Marl Slate of SE Durham, England. Economic Geology, Vol. 58, 912–940.

HOFMANN, B A. 1990. Reduction spheroids from northern Switzerland: mineralogy, geochemistry and genetic models. Chemical Geology, Vol. 81, 55–81.

HOFMANN, B A. 1991. Mineralogy and geochemistry of reduction spheroids in red beds. Mineralogy and Petrology, Vol. 44, 107–124.

HOFMANN, B A. 1992. Isolated reduction phenomena in red beds: a result of porewater radiolysis? 503–506 in Water-rock interaction. KHARAKA, Y K, and MAEST, A S (editors): Proceedings of the 8th international symposium on water-rock interaction. (Rotterdam: Balkema.)

HOLLIDAY, D W. 1993. Mesozoic cover over northern England: interpretation of apatite fission-track data. Journal of the Geological Society of London, 150, 657–660.

HOLLOWAY, S, MILODOWSKI, A E, STRONG, G E, and WARRINGTON, G. 1989. The Sherwood Sandstone Group (Triassic) of the Wessex Basin, southern England. Proceedings of the Geologists' Association, Vol. 100, 383–394.

HOLM, T R, and CURTIS, C D. 1990. Copper complexation by natural organic matter in groundwater. 509–518 in Chemical modelling in aqueous systems II. MELCHIOR, D C, and BASSETT, R L (editors).(Washington DC: American Chemical Society.)

HOLMES, I, CHAMBERS, A D, IXER, R A, TURNER, P, and VAUGHAN, D.J. 1983. Diagenetic processes and the mineralisation in the Triassic of Central England. Mineralium Deposita, Vol. 18, 365–377.

HUANG, W-L, and LONGO, J M. 1992. The effect of organics on feldspar dissolution and the development of secondary porosity. Chemical Geology, 98, 271–292.

HUDSON, J D. 1982. Pyrite in ammonite-bearing shales from the Jurassic of England and Germany. Sedimentology, Vol. 29, 639–667.

HUGHES, C D, and BRADLEY, A D. 1993. Trace elements in oils and tars - a preliminary investigation. Extended Abstracts, 1993 British Organic Geochemical Society meeting, Plymouth, UK, July

HULL, E. 1864. On the copper-bearing rocks of Alderley Edge, Cheshire. Geological Magazine, 1, 65–69.

HULL, E. 1869. The Triassic and Permian rocks of the Midland Counties of England. In Memoir of the Geological Survey of Great Britain. 127 pp.

HUMPHREYS, B, MORTON, A C, HALLSWORTH, C R, GATLIFF, R W, and RIDING, J B. 1991. An integrated approach to provenance studies: a case example from the Upper Jurassic of the Central Graben, North Sea. 251–262 in Developments in sedimentary provenance studies. MORTON, A C, TODD, S P, and HAUGHTON, P D W (editors). Special Publication of the Geological Society of . London, No. 57.

HUNT, J M, WHELAN, J K, EGLINGTON, L B, and CATHLES, L M. 1994. Gas generation - a major cause of deep Gulf Coast overpressures. Oil and Gas Journal, July 18, 59–63.

HUNTER, R E. 1977. Basic types of stratification in small eolian dunes. Sedimentology, Vol. 24, 361–387.

HYDEN, H J. 1961. Distribution of uranium and other elements in crude oils. Bulletin of the United States Geological Survey, 1100 Part B, metals.

ICKES, E L. 1923. Recent exploration for petroleum in the United Kingdom. Transactions of the American Institute of Mining and Mechanical Engineers, Vol. lxx, 1053–1075.

INTERNATIONAL ATOMIC ENERGY AGENCY. 1983. Chapter 7 (Tritium). 57–61 in Isotope techniques in the hydrogeological assessment of potential sites for the disposal of high-level radioactive wastes. Technical Report Series, IAF-A, Vienna, No. 228.

IRELAND, R J, POLLARD, J E, STEEL, R J, and THOMPSON, D B. 1978. Intertidal sediments and trace fossils from the Waterstones (Scythian-Anisian?) at Daresbury, Cheshire. Proceedings of the Yorkshire Geological Society, Vol. 41, 399–436.

IRWIN, H, CURTIS, C D, and COLEMAN, M. 1977. Isotopic evidence of source of diagenetic carbonates formed during burial of organic-rich sediments. Nature, London, Vol. 269, 209–213.

IXER, R A, TURNER, P, and WAUGH, B. 1979. Authigenic iron and titanium oxides in Triassic red beds (St Bees Sandstone), Cumbria, northern England. Geological Journal, Vol. 14, 179–192.

IXER, R A, and VAUGHAN, D J. 1982. The primary ore mineralogy of the Alderley Edge deposit, Cheshire. Mineralogical Magazine, Vol. 46, 485–492.

JACKSON, D I, JACKSON, A A, EVANS, D, WINGFIELD, R T R, BARNES, R P, and ARTHUR, M J. 1995. United Kingdom offihore regional report: the geology of the Irish Sea. (London: HMSO for the British Geological Survey.)

JACKSON, D I, and MULHOLLAND, P. 1993. Tectonic and stratigraphic aspects of the east Irish Sea Basin and adjacent areas: contrasts in their post-Carboniferous structural styles. 791–808 in Petroleum geology of Northwest Europe: Proceedings of the 4th Conference. PARKER, J (editor). (London: The Geological Society.)

JACKSON, D I, MULHOLLAND, P, JONES, S M, and WARRINGTON, G. 1987. The geological framework of the East Irish Sea Basin. 191–203 in Petroleum geology of North West Europe. BROOKS, J, and GLENNIE, K (editors). (London: Graham & Trotman.)

JAMES, J W C. 1983. The sand and gravel resources of the country around Prees, Shropshire. Description of 1:25 000 resource sheet SJ 53. Mineral Assessment Report of the Institute of Geological Sciences, No. 134.

JOHNSON, J W, OELKERS, E H, and HELGESON, H C. 1991. SUPCRT92: a software package for calculating the standard molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 to 5000 bars and 0° to 1000°C. (Berkeley: University of California.)

JONES, B F, and WEIR, A H. 1983. Clay minerals of Lake Albert, an alkaline saline lake. Clays and Clay Minerals, Vol. 31,161–172.

JONES, C M. 1980. Deltaic sedimentation in the Roaches Grit and associated sediments (Namurian R2b) in the south-west Pennines. Proceedings of the Yorkshire Geological Society, Vol. 23, No. 1,39–67.

JONES, D G, and HASLAM, H W. 1994. Lithogeochemistry of the Permo-Triassic rocks from the Cheshire Basin. British Geological Survey Technical Report, WP/94/7.

JONES, N S. 1994. Sedimentology of Permo-Triassic strata from selected boreholes in the Cheshire Basin. British Geological Survey Technical Report, Strati graphy Series, WH/94/57R. 29 pp.

JONES, R C B, TONKS, L H, and WRIGHT, W B. 1938. Wigan district. Memoir of the Geological Survey of Great Britain.

JOWETT, E C. 1986. Genesis of Kupferschiefer Cu-Ag deposits by convective flow of Rotliegende brines during Triassic rifting. Economic Geology, Vol. 81,1823–1837.

JOWETT, E C. 1989. Effects of continental rifting on location and genesis of stratiform copper-silver deposits. 53–66 in Sediment-hosted stratiform copper deposits. BOYLE, R W, BROWN, A C, JEFFERSON, C W, JOWETT, E C, and KIRKHAM, R V (editors). Special Paper of the Geological Association of Canada, No. 36.

JOWETT, E C. 1991. The evolution of ideas about the genesis of stratiform copper-silver deposits. 117–132 in Historical perspectives of genetic concepts and case histories of famous discoveries. HUTCHINSON, R W, and GRAUCH, R I (editors). Economic Geology Monograph, No. 8.

KACEWICZ, M. 1993. Mathematics between source and trap: uncertainty in hydrocarbon migration modelling. 69–84 in Computers in geology -25 years of progress. DAVIS, J C, and HERZFELD, U C (editors). (New York: Oxford University Press.)

KAKIMI, T. 1980. Magnitude-frequency relation for displacement of minor faults and its significance in crustal deformation. Bulletin of the Geological Survey of Japan, Vol. 31,467–487.

KARPETA, W P. 1990. The morphology of Permian palaeodunes - a reinterpretation of the Bridgnorth Sandstone around Bridgnorth, England, in the light of modern dune studies. Sedimentary Geology, Vol. 69,59–75.

KEMP, S J. 1994. The clay mineralogy of Permo-Triassic sedimentary rocks from the Cheshire Basin. British Geological Survey Technical Report, WG/94/22.

KENT, P E. 1949. A structure contour map of the surface of the buried pre-Permian rocks of England and Wales. Proceedings of the Geologists' Association, Vol. 60,87–104.

KHOURY, H N, EBERL, D D, and JONES, B F. 1982. Origin of magnesian clays from the Armargosa Desert, Nevada. Clays and Clay Minerals, Vol. 30,327–336.

KING, R J. 1968. Mineralization. 112–137 in The geology of the East Midlands. SYLVESTER-BRADLEY, P C, and FORD, T D (editors). (Leicester: Leicester University Press.)

KINSMAN, D J J. 1969. Modes of formation, sedimentary associations and diagenetic features of shallow water and supratidal evaporites. Bulletin of the American Association of Petroleum Geologists, Vol. 53,830–840.

KIRBY, G A, BAILY, H E, CHADWICK, R A, EVANS, D J, HOLLIDAY, D W, HOLLOWAY, S, HULBERT, A G, PHARAOH, T C, SMITH, N J P, AITKENHEAD, N, and BIRCH, B. In press. The structure and evolution of the Craven Basin and adjacent areas. Subsinface Memoir of the British Geological Survey.

KIRBY, G A, SMITH, K, SMITH, N J P, and SWALLOW, P. 1987. Oil and gas generation in eastern England. 171–180 in Petroleum Geology of Northwest Europe. BROOKS, J, and GLENNIE, K W (editors). (London: Graham & Trotman.)

KIRKHAM, R V. 1989. Distribution, settings, and genesis of sediment-hosted stratiform copper deposits. 3–38 in Sediment-hosted stratiform copper deposits. BOYLE, R W, BROWN, A C, JEFFERSON, C W, JOWETT, E C, and KIRKHAM, R V (editors). Special Paper of the Geological Association of Canada, No. 36.

KLEEMAN, J D, and LOVERING, J F. 1967. Uranium distribution in rocks by fission track registration in Lexan plastic prints. Atomic Energy Australia, Vol. 10, No. 4,3–8.

KLEMPERER, S L, and WHITE, N. 1989. Coaxial stretching or lithospheric simple-shear in the North Sea? Evidence from deep seismic profiling and subsidence. 511–522 in Extensional tectonics and stratigraphy of the North Atlantic margins. TANKARD, A J, and BALKWILL, H R (editors). Memoir of the American Association of Petroleum Geologists, No. 46.

KNOTT, S D. 1994. Fault zone thickness versus displacement in the Permo-Triassic sandstones of NW England. Journal of the Geological Society of London, Vol. 151,17–25.

KNORR, S D, BURCHELL, M T, JOLLEY, E J, and FRASER, A J. 1993. Mesozoic to Cenozoic plate reconstructions of the North Atlantic and hydrocarbon plays of the Atlantic margins. 953–974 in Petroleum geology of northwest Europe. Proceedings of 4th Conference. PARKER, J R (editor). (London: Geological Society.)

KNOX, R W O'B, BURGESS, W G, WILSON, K S, and BATH, A H. 1984. Diagenetic influences on reservoir properties of the Sherwood Sandstone (Triassic) in the Marchwood geothermal borehole, Southampton, England. Clay Minerals, Vol. 19, 441–456.

KOCUREK, G. 1981. Significance of interdune deposits and bounding surfaces in aeolian dune sands. Sedimentology, Vol. 28, 753–780.

KRUMGALZ, B. 1980. Salt effect on the pH of hypersaline solutions. 73–82 in Hypersaline brines and evaporitic environments. NISSENBAUM, A (editor). Developments in Sedimentology, Vol. 28.

KUCHA, H. 1982. Platinum group metals in the Zechstein copper deposits, Poland. Economic Geology, Vol. 77,1578–1591.

KUCHA, H. 1990. Geochemistry of the Kupferschiefer, Poland. Geologische Rundschau, Vol. 79/2,387–399.

KURZL, H. 1988. Exploratory data analysis: recent advances for the interpretation of geochemical data. Journal of Geochemical Exploration, Vol. 30,309–322.

KUSZNIR, N J, and ZIEGLER, P A. 1992. The mechanism of continental extension and sedimentary formation: a simple/pure shear flexural cantilever model. Tectonophysics, Vol. 215, 117–131.

LAKE, S D, and KARNER, G D. 1987. The structure and evolution of the Wessex Basin, southern England: an example of inversion tectonics. Tectonophysics, Vol. 137, 347–378.

LAND, L S, and MACPHERSON, G L. 1992. Origin of saline waters, Cenozoic section, Gulf of Mexico sedimentary basin. Bulletin of the American Association of Petroleum Geologists, Vol. 76, 1344–1362.

LAND, L S, and PREZBINDOWSKI, P R. 1981. The origin and evolution of saline formation water, Lower Cretaceous carbonates, south-central Texas, USA. Journal of Hydrology, Vol. 54, 51–74.

LANG, W H. 1980. Determination of prior depth of burial using interval transit time. Oil and Gas Journal Vol. 78, No. 4, 222–232.

LANGMUIR, D. 1979. Uranium geochemistry. 17–29 in Short course in uranium deposits: their mineralogy and origin. KIMBERLEY, M M (editor). Mineralogical Association of Canada Short Course Handbook, No. 3.

LEEDER, M R. 1982. Upper Palaeozoic basins of the British Isles -Caledonide inheritance versus Hercynian plate-margin processes. Journal of the Geological Society of London, Vol. 139, 479–491.

LEEDER, M R. 1988. Recent developments in Carboniferous geology: a critical review with implications for the British Isles and N.W. Europe. Proceedings of the Geologists' Association, Vol. 99, 73–100.

LEES, G M, and TAITT, A H. 1946. The geological results of the search for oilfields in Great Britain. Quarterly Journal of the Geological Society of London, Vol. 101 (for 1945) 255–317.

LENG, M J, and EVANS, J A. 1994. Provenance of late Ashgill (Hirnantian) fine-grained sediments and pebbles in the Welsh Basin: A Nd- and Sr-isotope study. Geological Journal. Vol. 29, 1–10.

LEWAN, M D. 1984. Factors controlling the proportionality of vanadium to nickel in crude oils. Geochimica et Cosmochimica Acta, Vol. 48, 2231–2238.

LEWIS, C L E, GREEN, P F, CARTER, A, and HURFORD, A J. 1992. Elevated K/T palaeotemperatures throughout Northwest England: three kilometres of Cainozoic erosion? Earth and Planetary Science Letters, Vol. 112, 131.L.145.

LINDGREN, W. 1933. Mineral deposits (4th edition). (New York: McGraw-Hill.)

LOGAN, A. 1967. The Permian bivalvia of northern England. Palaeontographical Society Monographs, No. 121, 1–72.

LONGIARU, S. 1987. Visual comparators for estimating the degree of sorting from plane and thin section. Journal of Sedimentary Petrology, Vol. 57, 791–794.

LOVELOCK, P E R. 1977. Aquifer properties of Permo-Triassic sandstones in the United Kingdom. Bulletin of the Geological Survey of London,No. 56, 49 pp.

LOVERING, T S. 1963. Epigenetic, diplogenetic, and lithogene deposits. Economic Geology, Vol. 58, 315–331.

LUCEY, P A. 1987. The hydrochemistry of groundwater in the west Cheshire aquifer, North-west England. Unpublished MPhil thesis, University of Lancaster.

LUNDEGARD, P D, and LAND, L S. 1986. Carbon dioxide and organic acids: their role in porosity enhancement and cementation, Palaeogene of the Texas Gulf Coast. 129–147 in Roles of organic matter in sediment diagenesis. GAUTIER, D L (editor). Special Publication of the Society of Economic Palaeontologists and Mineralogists, No. 38.

MAGARA, K. 1976. Thickness of removed sedimentary rocks, palaeopore pressure and palaeotemperature, southwestern part of western Canada Basin. Bulletin of the American Association of Petroleum Geologists, Vol. 60, No. 4, 554–565.

MAGRAW, D. 1961. Exploratory boreholes in the central part of the South Lancashire Coalfield. The Mining Engineer, Vol. 120, 432–448.

MANGE, M A, and MAURER, H F W. 1992. Heavy minerals in colour. (London: Chapman and Hall.)

MANNING, D A C. 1986. Assessment of the role of organic matter in ore transport processes in low-temperature base-metal systems. Transactions of the Institution of Mining and Metallurgy, Vol. 95, B195-B200.

MARIE, J P P. 1975. Rotliegendes stratigraphy and diagenesis. 205–211 in Petroleum geology and the continental shelf of Northwest Europe, V 1. WOODLAND, A W (editor). (London: Applied Science Publishers.)

MARRETT, R, and ALLMENDINGER, R W. 1992. Amount of extension on 'small' faults: An example from the Viking Graben. Geology, Vol. 20, 47–50.

MASTERS, J A. 1984. Elmworth - case study of a Deep Basin Gas Field. Memoir of the American Association of Petroleum Geologists, No. 38. 313 pp.

MAW, G. 1868. On the disposition of iron in variegated strata. Quarterly Journal of the Geological Society of London, Vol. 24, 370.

MAW, G. 1870. On the occurrence of the Rhaetic Beds in North Shropshire and Cheshire. Geological Magazine, Vol. 7, 203–204.

MAYALL, M J. 1983. An earthquake origin for synsedimentary deformation in a late Triassic (Rhaetian) lagoonal sequence, southwest Britain. Geological Magazine, Vol. 120, 613–622.

McBRIDE, E F. 1989. Quartz cement in sandstones: a review. Earth Science Reviews, Vol. 26, 69–112.

MCCREA, J M. 1950. On the isotopic chemistry of carbonates and a paleotemperature scale. Journal of Chemical Physics, Vol. 18, 849–857.

MCDERMOTT, C, WIGGANS G N, WADGE, G, and FERRIER, G. 1993. Expert systems and GIS as tools for the integration and analysis of geological data and mineral exploration. Proceedings of Artifical Intelligence in the Minerals Sector, Nottingham, 1–8.

MCHARDY, W J, WILSON, M J, and TAIT, J M. 1982. Electron microscope and X-ray diffraction studies of filamentous illitic clay from sandstones of the Magnus Field. Clay Minerals, Vol. 17, 23–39.

McKEE, E D, and BIGARELLA, J J. 1979. Sedimentary structures in dunes. 83–134 in A study of global sand seas. MCKEE, E D. (editor). Professional Paper of the United States Geological Survey, No. 1052.

MCKEE, E D, DOUGLASS, J R, and RITTENHOUSE, S. 1971. Deformation of lee side laminae in eolian dunes. Bulletin of the Geological Society of America, No. 82, 359–378.

MCKENZIE, D P. 1978. Some remarks on the development of sedimentary basins. Earth and Planetary Science Letters, Vol. 40, 25–32.

MCKENZIE, J A. 1981. Holocene dolomitization of calcium carbonate sediments from the coastal sabkhas of Abu Dhabi, UAE: a stable isotope study. Journal of Geology, Vol. 89, 185–198.

MCLENNAN, S M. 1989. Rare earth elements in sedimentary rocks: influence of provenance and sedimentary processes. 169–200 in Geochemistry and mineralogy of rare earth elements. LIPIN, B R, and McKAY, G A (editors). (Washington DC: Mineralogical Society of America.)

MEADOWS, N 5, and BEACH, A. 1993a. Controls on reservoir quality in the Triassic Sherwood Sandstone of the Irish Sea. 823–833 in Petroleum Geology of Northwest Europe: Proceedings of the 4th Conference. PARKER, J (editor). (London: The Geological Society.)

MEADOWS, N S, and BEACH, A. 1993b. Structural and climatic controls on facies distribution in a mixed fluvial and aeolian reservoir: the Triassic Sherwood Sandstone in the Irish Sea. 247–264 in Characterisation of fluvial and aeolian reservoirs. NORTH, C, and PROSSER, D (editors). Special Publication of the Geological Society of London, No. 73.

MEARNS, E W, KNARUD, R, RAESTAD, N, STANLEY, K 0, and STOCKBRIDGE C P. 1989. Samarium-neodymium isotope stratigraphy of the Lunde and Statfjord Formations of the Snorre Oil Field, northern North Sea. Journal of the Geological Society of London, Vol. 146, 217–228.

MERINO, E, MOORE, C, ORTOVELA, P, and RIPLEY, E. 1986. Mineral zoning in sediment-hosted copper-iron sulphide deposits - a quantitative kinetic approach. 559–571 in Geology and metallogeny of copper deposits: proceedings of the copper symposium, 27th International Geological Congress. FRIEDRICH, D H (editor). Special Publication of the Society for Geology Applied to Mineral Deposits, No. 4.

METCALFE, R and ROCHELLE, C A. 1994. Geochemical models for diagenesis and mineralisation in red beds and their implications for the Cheshire Basin. British Geological Survey Technical Report, WE/94/1R.

METCALFE, R, ROCHELLE, C A, SAVAGE, D, and HIGGO, J J W. 1994a. Fluid-rock interactions during continental red-bed diagenesis: implications for theoretical models of mineralization in sedimentary basins. 301–324 in Geofluids: origin, migration and evolution of fluids in sedimentary basins. PARNELL, J (editor). Special Publication of the Geological Society of London, No. 78.

METCALFE, R, ROCHELLE, C A, SAVAGE, D, HIGGO, J J W, and HOOKER, P J. 1994b. An investigation of fluid-rock interactions and groundwater hydrogeochemistry in the Cheshire Basin. British Geological Survey Technical Report, WE/94/2R.

METCALFE, R, ROCHELLE, C A, SAVAGE, D, HOOKER, P J, and GALE, I N. 1993. Hydrochemical modelling of red bed diagenesis and heavy metal mineralisation. In Proceedings of Geofluids '93, contributions to an international conference on fluid evolution, migration and interaction in rocks, held at Torquay, England. PARNELL, J, RUFFELL, A H, and MOLES N R (editors).

MILLOTT, J O'N. 1942. Regional variations in properties of the Eight Foot Banbury or Cockshead Seam in the N. Staffordshire Coalfield. Transactions of the Institution of Mining Engineeers, Vol. 101, 2–13.

MILLOTT, J O'N, COPE, F W, and BERRY, H H. 1946. The seams encountered in a boring at Pie Rough, near Keele, north Staffordshire. Transactions of the Institution of Mining Engineers, Vol. 105, 528–570.

MILODOWSKI, A E, STRONG, G E, SPIRO, B, KEMP, S J, HYSLOP, E K, LENG, M J, and HASLAM, H W. 1994. The diagenesis of the Permo-Triassic rocks of the Cheshire Basin. British Geological Survey Technical report, WG/94/34.

MILODOWSKI, A E, STRONG, G E, WILSON, K S, ALLEN, D 5, HOLLOWAY, S, and BATH, A H. 1986. Diagenetic influences on the aquifer properties of the Sherwood Sandstone in the Wessex Basin. Investigation of the geothermal potential of the UK (Keyworth, Nottingham: British Geological Survey.)

MILODOWSKI, A E, STRONG, G E, WILSON, K S, HOLLOWAY, S, and BATH, A H. 1987. Diagenetic influences on the aquifer properties of the Permo-Triassic sandstones in the East Yorkshire-Lincolnshire Basin. Investigation of the geothermal potential of the UK (Keyworth, Nottingham: British Geological Survey.)

MILODOWSKI, A E, WEST, J M, PEARCE, J M, HYSLOP, E K, BASHAM, I R, and HOOKER, P J. 1990. Uranium-mineralised microorganisms associated with uraniferous hydrocarbons in southwest Scotland. Nature, London, Vol. 347, 465–467.

MILODOWSKI, A E, and ZALASIEWICZ, J A. 1991a. Redistribution of rare earth elements during diagenesis of turbidite/hemipelagite mudrock sequences of Llandovery age from central Wales. 101–124 in Developments in sedimentary provenance studies. MORTON, A C, TODD, S P, and HAUGHTON, P D W (editors). Special Publication of the Geological Society of London, No. 57.

MILODOWSKI, A E, and ZALASIEWICZ, J A. 199 lb. The origin and sedimentary, diagenetic and metamorphic evolution of chlorite-mica stacks in Llandovery sediments of central Wales, UK. Geological Magazine, Vol. 128, 263–278.

MOLDOVANYT, E P, and WALTER, L M. 1992. Regional trends in water chemistry, Smackover Formation, southwest Arkansas: geochemical and physical controls. Bulletin of the American Association of Petroleum Geologists, Vol. 76, 864–894.

MOODY, G T. 1905. The causes of variegation in the Keuper Marls, and in other calcareous rock. Quarterly Journal of the Geological Society of London, Vol. 61, 431.

MOORE, C H. 1989. Carbonate diagenesis and porosity. Developments in Sedimentology Series, No. 46

MORTON, A C. 1984. Stability of detrital heavy minerals in Cainozoic sandstones of the North Sea Basin. Clay Minerals, Vol. 19, 287–308.

MORTON, A C. 1986. Dissolution of apatite in North Sea Jurassic sandstones: implications for the generation of secondary porosity. Clay Minerals, Vol. 21, 711–733.

MORTON, A C. 1993. Provenance of Permo-Triassic sandstones in the Cheshire Basin: Heavy mineral constraints. British Geological Survey Technical report, WH/93/284R. 34 pp.

MORTON, A C, DAVIES, J R, and WATERS, R A. 1992. Heavy minerals as a guide to turbidite provenance in the Lower Palaeozoic Southern Welsh Basin : a pilot study. Geological Magazine, Vol. 129, 573–580.

MORTON, A C, and HALLSWORTH, C R. 1994. Identifying provenance-specific features of detrital heavy mineral assemblages in sandstones. Sedimentary Geology, Vol. 90, 241–256.

MORTON, A C, and JOHNSSON, M J. 1993. Factors influencing the composition of detrital heavy mineral suites in Holocene sands of the Apure River drainage basin, Venezuela. 171–185 in Processes controlling the composition of elastic sediments. JOHNSSON, M J, and BASU, A (editors). Special Paper of the Geological Society of America, Vol. 284.

MURCHISON, R I. 1835. On an outlying basin of Lias on the borders of Salop and Cheshire, with a short account of the lower Lias between Gloucester and Worcester. Quarterly journal of the Geological Society of ondon, Vol. 2, 114–115.

NASH, J T, GRANGER, H C, and ADAMS, S S. 1981. Geology and concepts of genesis of important types of uranium deposits. Economic Geology, 75th Anniversary Volume, 63–116.

NAYLOR, H, TURNER, P, VAUGHAN, D J, BOYCE, A J, and FALLICK, A E. 1989. Genetic studies of red bed mineralization in the Triassic of the Cheshire Basin, northwest England. Journal of the Geological Society of London, Vol. 146, 685–699.

NEEDHAM, R S, and STUART-SMITH, P G. 1987. Coronation Hill U-Au Mine, South Alligator Valley, Northern Territory: an epigenetic sandstone-type deposit hosted by debris-flow conglomerate. BMR Journal of Australian Geology and Geophysics, Vol. 10, 121–131.

NEGLIA, S. 1979. Migration of fluids in sedimentary basins. Bulletin of the American Association of Petroleum Geologists, Vol. 63. 573–579.

NORTHOLT, A J G, and HIGHLEY, D E. 1973. Salt. Mineral Dossier, No. 73. 36 pp.

OAKES, C S, BODNAR, R J, and SIMONSON, J M. 1990. The system NaCl-CaCl₂-Fl₂O: 1. The ice liquidus at 1 atm. total pressure. Geochemica et Cosmochimica Acta, Vol. 54, 603–610.

OECD Nuclear Energy Agency and International Atomic Energy Agency. 1986. Uranium: resources, production and demand. (Paris: OECD.)

OHMOTO, H. 1986. Stable isotope geochemistry of ore deposits. 460–491 in Stable isotopes and high temperature geological processes. VALLEY, J W, TAYLOR, H P, and O'NEIL, J R (editors). Mineralogical Society of America, Reviews in Mineralogy, No. 16.

OHMOTO, H, and LASAGA, A C. 1982. Kinetics of reactions between aqueous sulphates and sulphides in hydrothermal systems. Geochimica et Cosmochimica Acta, Vol. 46. 1727–1745.

O'NIONS, R K, HAMILTON, P J, and HOOKER, P J. 1983. A Nd isotope investigation of sediments related to crustal development in the British Isles. Earth and Planetary Science Letters, Vol. 63, 229–240.

ORMEROD, G W. 1848. Outline of the principal geological features of the salt field of Cheshire and the adjoining districts. Quarterly Journal of the Geological Society of London, Vol. 4, 262, 268–288.

ORTLAM, D. 1980. Inhalt und bedeutung fossiler bodenkomplexe in Perm und Trias von Mitteleurope. Geologische Rundschau, Vol. 63, 850–884.

PAPELIS, C, HAYES, K, and LECKIE, J O. 1988. HYDRAQL: A program for the computation of chemical equilibrium composition of aqueous batch systems including surface-complexation modelling of ion adsorption at the oxide/solution interface. Technical Report of the Department of Civil Engineering, Stanford University, Stanford, California, No. 306.

PARNELL, J. 1983. The distribution of hydrocarbon minerals in the Welsh Borderlands and adjacent areas. Geological Journal Vol. 18, 129–139.

PARNELL, J. 1992a. Hydrocarbon potential of Northern Ireland: Part III - reservoir potential of the Permo-Triassic. Journal of Petroleum Geology, Vol. 15, 51–70.

PARNELL, J. 1992b. Metal enrichment in bitumens from Carboniferous-hosted ore deposits of the British Isles. Chemical Geology, Vol. 99, 115–124.

PARNELL, J, and SWAINBANK, I. 1990. Pb-Pb dating of hydrocarbon migration into a bitumen-bearing ore deposit, North Wales. Geology, Vol. 18, 1028–1030.

PASS, A. 1993. A summer of work experience at British Salt, Middlewich, Cheshire. Minerals Industry International, March, 28–30.

PATTISON, J. 1969. Some Permian foraminifera from northwestern England. Geological Magazine, Vol. 106, 197–205.

PATTISON, J. 1970. A review of the marine fossils from the upper Permian rocks of northern Ireland and north-west England. Bulletin of the Geological Survey of Great Britain, No. 32, 123–165.

PATTISON, J, SMITH, D B, and WARRINGTON, G. 1973. A review of late Permian and early Triassic biostratigraphy in the British Isles. The Permian and Triassic systems and their mutual boundary, No. 2, 220–260.

PATTRICK, R A D, and BOWELL, R J. 1991. The genesis of the West Shropshire Orefield: evidence from fluid inclusions, sphalerite chemistry, and sulphur isotope ratios. Geological Journal Vol. 26, 101–115.

PENN, I E. 1987. Geophysical logs in the stratigraphy of Wales and adjacent offshore and onshore areas. Proceedings of the Geologists' Association, Vol. 98, No. 4, 275–314.

PETTIJOHN, F J. 1975. Sedimentary rocks (3rd edition). (New York: Harper & Row.)

PINFOLD, E S. 1958. The search for oil in Lancashire. Liverpool and Manchester Geological Journal Vol. 2, 106–123.

PITZER, K S. 1973. Thermodynamics of electrolytes I: Theoretical basis and general equations. Journal of Physical Chemistry, Vol. 77, 268–277.

PLANT, J A, and TARNEY, J. 1994. Mineral deposit models and their primary geochemical signatures. 11–72 in Drainage geochemistry, handbook of exploration geochemistry, Vol. 6. HALE, M, and PLANT, J A (editors). (Amsterdam: Elsevier.)

POCOCK, R W, WHITEHEAD, T H, WEDD, C B, and ROBERTSON, T. 1938. Shrewsbury District, including the Hanwood Coalfield. Memoir of the British Geological Survey, sheet 152.

POCOCK, R W, and WRAY, D A. 1925. The geology of the country around Wem. Memoir of the Geological Survey of Great Britain, 125 pp.

POLLARD, J E. 1981. A comparison between the Triassic trace fossils of Cheshire and south Germany. Palaeontology, Vol. 24, 555–588.

POLLARD, J E. 1985. Isopodichnus, related arthropod trace fossils and notostracans from Triassic fluvial sediments. Transactions of the Royal Society of Edinburgh, Vol. 76, 273–285.

POOLE, E G, and WHITEMAN, A J. 1955. Variations in thickness of the Collyhurst Sandstone in the Manchester area. Bulletin of the Geological Survey of Great Britain, Vol. 9, 33–41.

POOLE, E G, and WHITEMAN, A J. 1966. Geology of the country around Nantwich and Whitchurch. Memoir of the Geological Survey of Great Britain, Sheet 122. 154 pp.

PUGH, W J. 1960. Triassic salt: discoveries in the Cheshire-Shropshire Basin. Nature, London, No. 187, 278–279.

PYE, K, and TSOAR, H. 1990. Aeolian sand and sand dunes. (London: Unwin Hyman.)

QUIRK, D G. 1996. Discussion on the relationship between bitumens and mineralization in the South Pennine Orefield, central England. Journal of the Geological Society of London, Vol. 153, 653–656.

RAAB, M, and SPIRO, B. 1991. Sulphur isotopic variations during seawater evaporation with fractional crystallization. Chemical Geology, Vol. 86, 323–333.

RAMSAY, J G. 1980. The crack-seal mechanism of rock deformation. Nature, London, Vol. 284, 135–139.

RAMSBOTTOM, W H C, CALVER, M A, EAGAR, R M C, HODSON, F, HOLLIDAY, D W, STUBBLEFIELD, C J, and WILSON, R B. 1978. A correlation of Silesian rocks in the British Isles. Special Report of the Geological Society of London, No. 10.

RAWSON, P F, and RILEY, L A. 1982. Latest Jurassic-early Cretaceous events and the late-Cimmerian Unconformity' in the North Sea area. Bulletin of the American Association of Petroleum Geologists, Vol. 66, No. 12, 2628–2648.

READ, D, COOPER, D C, and MCARTHUR, J M. 1987. The composition and distribution of nodular monazite in the Lower Palaeozoic rocks of Great Britain. Mineralogical Magazine, Vol. 51, 271–280.

REDWOOD, B. 1922. A treatise on petroleum, Vol. 1, 367 pp. (London: Griffin.)

REES, J G, and WILSON, A A. 1998. Geology of the country around Stoke-on-Trent. Memoir of the British Geological Survey, Sheet 123 (England and Wales).

Rtivs, G H. 1974. A proposed standard lithostratigraphical nomenclature for the southern North Sea and an outline structural nomenclature for the whole of the (UK) North Sea. Report of the Institute of Geological Sciences, No. 74/8.

RICKARD, D T. 1970. The origin of framboids. Lithos, Vol. 3, 269–293.

RICKARD, D T. 1975. Kinetics and mechanism of pyrite formation at low temperatures. American Journal of Science, Vol. 274, 636–652.

RITTENHOUSE, G. 1967. Bromine in oilfield waters and its use in determining possibilities of origin of these waters. Bulletin of the American Association of Petroleum Geologists, No. 51, 2430–2440.

ROBERTS, D G. 1989. Basin Inversion in and around the British Isles. 131–150 in Inversion tectonics. COOPER, M A, and WILLIAMS, G D (editors). Special Publication of the Geological Society, No. 44.

ROBERTS, R G, and SHEAHAN, P A (editors). 1988. Ore deposit models Geoscience Canada Reprint Series, 3.

ROBINSON, B W, and KUSAKABE, M. 1975. Quantitative preparation of sulfur dioxide, for ³⁴S/³²S analyses, from sulphides by combustion with cuprous oxide. Analytical Chemistry, Vol. 47, 1179–1181.

ROCHELLE, C A, WILLIAMS, L, HOOKER, P J, and METCALFE, R. 1993. The chemistry of modern groundwaters in the Cheshire Basin, UK. British Geological Survey Technical Report, WE/93/3R.

ROEDDER, E. 1971. Metastability in fluid inclusions. Special issue of the Society of Mining Geology of Japan, No. 3, 327–334.

ROEDDER, E. 1984. Fluid inclusions. Mineralogical Society of America, Reviews in Mineralogy, No. 12. 644 pp.

ROEDER, C. 1890a. Descriptions of fossils found in the Permian upper series, at Fallowfield. Transactions of he Manchester Geological and Mineralogical Society, Vol. 21. 13–16.

ROEDER, C. 1890b. Notes on the Upper Permians, &c., at Fallowfield, lately laid open. Transactions of the Manchester Geological and Mineralogical Society, Vol. 20, 615–630.

ROEDER, C. 1890c. On some newly discovered species in the Upper Permian of Manchester. Transactions of the Manchester Geological and Mineralogical Society, Vol. 20, 535–537.

ROEDER, C. 1892. Notes on the Permians and superficial beds at Fallowfield. Transactions of the Manchester Geological and Mineralogical Society, Vol. 21, 104–113.

ROSE, A W. 1976. The effect of cuprous chloride complexes in the origin of red-bed copper and related deposits. Economic Geology, Vol. 71, 1036–1048.

ROSE, A W. 1989. Mobility of copper and other heavy metals in sedimentary environments. 97–110 in Sediment-hosted stratiform copper deposits. BOYLE, R W, BROWN, A C, JEFFERSON, C W, JOWETT, E C, and KIRKHAM, R V (editors). Special Paper of the Geological Association of Canada, No. 38.

ROSE, A W, and BIANCHI-MOSQUERA, G C. 1985. Adsorption of Cu, Ag and other metals on Fe-oxides as a control on elemental composition of stratiform and redbed copper deposits. Abstracts with Programmes of the Geological Society of America, No. 17, 702.

ROSE, A W, and BIANCHI-MOSQUERA, G C. 1993. Adsorption of Cu, Pb, Zn, Co, Ni and Ag on goethite and haematite as a control on metal mobilisation from redbeds into stratiform copper deposits. Economic Geology, Vol. 88, 1226–1236.

ROSENBAUM, J, and SHEPPARD, S M F. 1986. An isotopic study of siderites, dolomites and ankerites at high temperatures. Geochimica et Cosmochimica Acta, Vol. 50, 1147–1150.

ROWLEY, W J, CHADWICK, R A, and HOLLIDAY, D W. 1993. Basin thermal modelling using HOTPOT version 3.0. British Geological Survey Technical Report, WC/93/04.

ROYDEN, L. 1986. A simple method for analyzing subsidence and heat flow in extensional basins. 49–73 in Thermal modelling in sedimentary basins. BURRUS, J (editor). (Paris: Editions Technip.)

SALES, R H. 1959. The White Pine copper deposit - a discussion. Economic Geology, Vol. 54, 947–950.

SALOMONS, W, GOUDIE, A, and MOOK, W G. 1978. Isotopic composition of calcrete deposits from Europe, Africa and India. Earth Surface Processes, No. 3, 43–57.

SANGSTER D F. 1993. Evidence for, and implications of, a genetic relationship between MVT and SEDEX zinc-lead deposits. 85–94 in World Zinc '93. MATTHEW, I G (editor). Australian I M M, No. 7/93.

SARJEANT, W A S. 1974. A history and bibliography of the study of fossil vertebrate footprints in the British Isles. Palaeogeography, Palaeoclimatology, Palaeoecology, Vol. 16, 265–378.

SARJEANT, W A S. 1984. British fossil footprints in the collections of some principal British museums. The Geological Curator, Vol. 3, 541–560.

SARJEANT, W A S. 1985. The Beasley Collection of photographs and drawings of fossil footprints and bones, and of fossil and Recent sedimentary structures. Geological Curator, Vol. 4, 133–163.

SAVAGE, D. 1993. Red beds, diagenesis and mineralisation: a literature review. British Geological Survey Technical Report, WE/93/1.

SAWLOWICZ, Z. 1991. The relationship between copper mineralisation and organic matter in the Polish Kupferschiefer. 589–592 in Source, transport and deposition of metals. PAGEL, M, and LEROY, J L (editors). (Rotterdam: Balkema.)

SAWLOWICZ, Z. 1992. Primary sulphide mineralisation in Cu-FeS zones of Kupferschiefer, Fore-Sudetic monocline, Poland. Transactions of the Institution of Mining and Metallurgy, Vol. 101, B1-B8.

SCHMIDT, V, and MCDONALD, D A. 1979a. The role of secondary porosity in the course of sandstone diagenesis. 175–207 in Aspects of diagenesis. SCHOLLE, P A, and SCHLUGER, P R (editors). Special Publication of the Society of Economic Paleontologists and Mineralogists, Vol. 26.

SCHMIDT, V, and MCDONALD, D A. 1979b. Texture and recognition of secondary porosity in sandstones. 209–225 in Aspects of diagenesis. SCHOLLE, P A, and SCHLUGER, P R (editors). Special Publication of the Society of Economic Paleontologists and Mineralogists, Vol. 26.

SCHNEIDERHOHN, H. 1932. The geology of the Copperbelt, Northern Rhodesia. Mineralogical Magazine, Vol. 46, 241–245.

SCLATER, J G, and CHRISTIE, P A F. 1980. Continental stretching: an explanation of the post-mid-Cretaceous subsidence of the Central North Sea Basin. Journal of Geophysical Research, Vol. 85, 3711–3739.

SCLATER, J G, and SHOREY, M D. 1989. Mid-Jurassic through mid-Cretaceous extension in the Central Graben of the North Sea-Part 2: Estimates from faulting based on a seismic reflection line. Basin Research, Vol. 1, 201–215.

SEASOR, R W, and BROWN, A C. 1989. Syngenetic and diagenetic concepts at the White Pine copper deposit, Michigan. 37–40 in Sediment-hosted stratiform copper deposits. BOYLE, R W, BROWN, A C, JEFFERSON, C W, JOWETT, E C, and KIRKHAM, R V (editors). Special Paper of the Geological Association of Canada, No. 36.

SEEWALD, J S, SEYFRIED, W E, and THORNTON, E C. 1990. Organic-rich sediment alteration: an experimental and theoretical study at elevated temperatures and pressures. Geochimica et CosmochimicaActa, Vol. 55, 659–669.

SELLEY, R. C. 1992. Petroleum seepages and impregnations in Great Britain. Marine and Petroleum Geology, Vol. 9, 226–244.

SENGER, R K. 1993. Paleohydrology of variable-density groundwater flow systems in mature sedimentary basins: example of the Palo Duro basin, Texas, USA. Journal of ydrology, Vol. 151, 109–145.

SHARMA, T, and CLAYTON, R N. 1965. Measurements of ¹⁸O/¹⁶O ratio of total oxygen of carbonates. Geochimica et CosmochimicaActa, Vol. 29, 1347–1353.

SHEARMAN, D J. 1966. Origin of marine evaporites by diagenesis. Transactions of the Institution of Mining and Metallurgy, Vol. 75, B208–215.

SHEARMAN, D J. 1970. Recent halite rock, Baja California, Mexico. Transactions of the Institution of Mining and Metallurgy, Vol. 79, B155–162.

SHEARMAN, D J, MOSSOP, G, DUNSMORE, H, and MARTIN, M. 1972. Origin of gypsum veins by hydraulic fracture. Transactions of the Institution of Mining and Metallurgy, Vol. 81, B149–155.

SHEPHERD, T J. 1994. Fluid inclusion study of diagenetic and hydrothermal events in the Cheshire Basin. British Geological Survey Technical Report, WG/94/24.

SHEPHERD, T J, and CHENERY, S R. 1995. Laser ablation ICP-MS elemental analysis of individual fluid inclusions: an evaluation study. Geochimica et Cosmochimica Acta, Vol. 59, 3997–4007.

SHEPHERD, T J, RANKIN, A H, and ALDERTON, D H M. 1985. A practical guide to fluid inclusion studies. 239 pp. (Glasgow and London: Blackie and Sons.)

SHOCK, E L. 1988. Organic acid metastability in sedimentary basins. Geology, Vol. 16, 886–890.

SHOCK, E L, and HELGESON, H C. 1990. Calculation of the thermodynamic and transport properties of aqueous species at high pressures and temperatures: Standard partial molal properties of organic species. Geochimica et Cosmochimica Acta, Vol. 54, 915–945.

SMITH, B. 1921. Lead and zinc ores in the Carboniferous rocks of North Wales. Memoir of the Geological Survey of Great Britain, Vol. 19.

SMITH, B, and GEORGE, T N. 1961. British regional geology: North Wales. (London: HMSO for the British Geological Survey.)

SMITH, D B, BRUNSTROM, R G W, MANNING, P I, SIMPSON, S, and SHOTTON, F W. 1974. A correlation of Permian rocks in the British Isles. Special Report of the Geological Society of London, No. 5. 45 pp.

SMITH, D B, and TAYLOR, J C M. 1992. Permian. 87–96 in Atlas of palaeogeography and lithofacies. COPE, J C W, INGHAM, J K, and RAWSON, P F (editors). Memoir of the Geological Society of London, No. 13.

SMITH, I F. 1986. Mesozoic basins. 42–83 in Geothermal energy-the potential in the United Kingdom. DOWNING, R A, and GRAY, D A (editors). (Keyworth, Nottingham: British Geological Survey.)

Small, I F, and BURGESS, W G. 1984. The Permo-Triassic rocks of the Worcester Basin. Investigations of the geothermal potential of he UK (London: HMSO for British Geological Survey.)

SMITH, K, SMITH, N J P, and HOLLIDAY, D W. 1985. The deep structure of Derbyshire. Geological Journal_s Vol. 20, 215–225.

SMITH, N J P. 1987. The deep geology of central England: the prospectivity of the Palaeozoic rocks. 217–224 in Petroleum geology of North West Europe. BROOKS, J, and GLENNIE, K W (editors). (London: Graham and Trotman.)

SMITH, N J P, CHADWICK, R A, WARRINGTON, G, KIRBY, G A, and JONES, D G, 1995. The hydrocarbon propectivity of the Cheshire Basin and surrounding areas. British Geological Survey Technical Report, WA/94/95C

SMITH, R T, COOPER, D C, and BLAND, D J. 1994. The occurrence and economic potential of nodular monazite in south-central Wales. British Geological Survey Technical Report, WF/94/1. (BGS Mineral Reconnaisance Programme Report, No. 130). 45 pp.

SOMERVILLE, I D, and STRANK, A. 1984. Discovery of Arundian and Holkerian faunas from a Dinantian platform succession in North Wales. Geological Journal, Vol. 19, 85–104.

SONNENFELD, P. 1984. Brines and evaporites. 613 pp. (London: Academic Press.)

SOPER, N J, WEBB, B C, and WOODCOCK, N H. 1987. Late Caledonian (Acadian) transpression in north-west England: timing, geometry and geotectonic significance. Proceedings of the Yorkshire Geological Society, Vol. 46, No. 3, 175–192.

STEEL, R J, and THOMPSON, D.B. 1983. Structures and textures in Triassic braided stream conglomerates (13unter' Pebble Beds) in the Sherwood Sandstone Group, North Staffordshire, England. Sedimentology, Vol. 30, 341–367.

STEIGER, R H, and JAGER, E. 1977. Subcommission on geochronology: convention of the use of decay constants in geoand cosmochronology. Earth and Planetary Science Letters, Vol. 36, 359–362.

STEIN, A M. 1988. Basement controls upon basin development in the Caledonian Foreland, north-west Scotland. Basin Research, Vol. 1, No. 2, 107–119.

STEPHENS, J. R. 1929. Wells and springs of Derbyshire. Memoir of the Geological Survey of Great Britain.

STEVENSON, I P, and MITCHELL, G H. 1955. Geology of the country betwen Burton upon Trent, Rugeley and Uttoxeter. Memoir of the Geological Survey of Great Britain, Sheet 140.

STOESSELL, R K, and PITTMAN, E D. 1990. Secondary porosity revisited: the chemistry of feldspar dissolution by carboxylic acids and anions. Bulletin of the American Association of Petroleum Geologists, Vol. 74, 1795–1805.

STONELEY, H M. 1958. The Upper Permian flora of England. Bulletin of the British Museum (Natural History) Geology, Vol. 3, 295–337.

STONELEY, R. 1983. Fibrous calcite veins, overpressures and primary oil migration. American Association of Petroleum Geologists, Vol. 67, 1427–1428.

STRAHAN, A. 1920. Mineral oil. Special reports on the Mineral resources of Great Britain, Vol. vii. Memoir of the Geological Survey of Great Britain.

STRONG, G E. 1993. Diagenesis of Triassic Sherwood Sandstone Group rocks, Preston, Lancashire, UK: a possible evaporitic cement precursor to secondary porosity? 279–289 in Characterisation of fluvial and aeolian reservoirs. NORTH, C P, and PROSSER, D J (editors). Special Publication of the Geological Society of London, No. 73.

STRONG, G E, and MILODOWSKI, A E. 1987. Aspects of the diagenesis of the Sherwood Sandstone of the Wessex Basin and their influence on reservoir characteristics. 325–337 in Diagenesis of sedimentary sequences. MARSHALL, J D (editor). Special Publication of the Geological Society of London, No. 36.

STRONG, G E, MILODOWSKI, A E, PEARCE, J M, KEMP, S J, PRIOR, S V, and MORTON, A C. 1994. The petrology and diagenesis of Permo-Triassic rocks of the Sellafield area, Cumbria. Proceedings of the Yorkshire Geological Society, Vol. 50, No.1, 77–89.

STRONG, G E, and PEARCE, J M. 1995. Carbonate spheroids in Permo-Triassic sandstone of the Sellafield area, Cumbria. Proceedings of the Yorkshire Geological Society, No. 50, 209–211.

STUART, I A, and COWAN, G. 1991. The South Morecambe Field, Blocks 110/20, 110/3a, 110/8a, UK East Irish Sea. 527–541 in United Kingdom Oil & Gas Fields. 25 Years Commemorative Volume. ABBorrs, I L (editor). Memoir of the Geological Society of London, No. 14.

STUMM, W, and MORGAN, J J. 1981. Aquatic chemistry, 2nd edition. (New York: John Wiley & Sons.)

SURDAM, R C, BOESE, S W, and CROSSEY, L J. 1984. The chemistry of secondary porosity. American Association of Petroleum Geologists, Vol. 37, 127–149.

SURDAM, R C, CROSSEY, L J, HAGEN, E S, and HEASLER, H P, 1989. Organic-inorganic interactions and sandstone diagenesis. Bulletin of the American Association of Petroleum Geologists, Vol. 73, 1–23.

SURDAM, R C, and SHEPPARD, R A. 1978. Zeolites in alkaline-lake deposits. In Natural zeolites: occurrence, properties, use. SAND, L B, and MUMPTON, F A (editors). (New York: Pergamon.)

SVERJENSKY, D A. 1984. Oil field brines as ore-forming solutions. Economic Geology, Vol. 79, 23–37.

SVERJENSKY, D A. 1987. The role of migrating oil field brines in the formation of sediment-hosted Cu-rich deposits. Economic Geology, Vol. 82, 1130–1141.

SVERJENSKY, D A. 1989. Chemical evolution of basinal brines that formed sediment- hosted Cu-Pb-Zn deposits. 127–134 in Sediment-hosted stratiform copper deposits. BOYLE, R W, BROWN, A C, JEFFERSON C W, JOWETT, E C, and KIRKHAM, R V (editors). Special Paper of the Geological Association of Canada, Vol. 36.

SWEENY, R E, and KAPLAN, I R. 1973. Pyrite framboid formation: laboratory synthesis and marine sediments. Economic Geology, Vol. 68, 618–634.

SWINTON, W E. 1960. The history of Chirotherium. Liverpool and Manchester Geological Journal, Vol. 2, 443–473.

TAYLOR, B J, PRICE, R H, and TROTTER, F M. 1963. Geology of the country around Stockport and Knutsford. Memoir of the Geological Survey of Great Britain, Sheet 98. 183 pp.

TAYLOR, S R. 1983. A stable isotope study of the Mercia Mudstones (Keuper Marl) and associated sulphate horizons in the English Midlands. Sedimentology, Vol. 30, 11–31.

TAYLOR, S R, and MCLENNAN, S M. 1985. The continental crust: its composition and evolution. 312 pp. (Oxford: Blackwell.)

TELLAM, J H. 1995. Hydrochemistry of the saline groundwaters of the lower Mersey Basin, Permo-Triassic sandstone aquifer, UK. Journal of Hydrology, Vol. 165, 45–84.

TELLAM, J H. 1996. Interpreting the borehole water chemistry of the Permo-Triassic sandstone aquifer of the Liverpool area, UK. Geological Journal, Vol. 31, 61–87.

TELLAM, J H, LLOYD, J W, and WATERS, M. 1986. The morphology of a saline groundwater body: its investigation, description and possible explanation. Journal of Hydrology, Vol. 83, 1–21.

THOMPSON, D B. 1966. The occurrence of an insect wing and branchiopods (Euestheria) in the Lower Keuper Marl at Styal, Cheshire. Mercian Geologist, Vol. 1, 237–245.

THOMPSON, D B. 1969. Dome shaped aeolian dunes in the Frodsham Member of the so-called 'Keuper' Sandstone Formation (Scythian-?Anisian: Triassic) at Frodsham, Cheshire, (England). Sedimentary Geology, Vol. 3, 263–289.

THOMPSON, D B. 1970a. Sedimentation of the Triassic (Scythian) Red Pebbly Sandstones in the Cheshire Basin and its margins. Geological Journal, Vol. 7, 183–216.

THOMPSON, D B. 1970b. The stratigraphy of the so-called Keuper Sandstone Formation (Scythian-?Anisian) in the Permo-Triassic Cheshire Basin. Quarterly Journal of the Geological Society of London, Vol. 126, 151–181.

THOMPSON, D B. 1985. Field excursion to the Permo-Triassic of the Cheshire, East Irish Sea, Needwood and Stafford basins. Poroperm Excursion Guide, No. 4, (Chester: Poroperm-Geochem Ltd.)

THOMPSON, D B. 1989. The geology of the neighbourhood of Chester - an essay review. Amateur Geologist, Vol. 13, No. I, 45–54.

THORNTON, E C, and SEYFRIED, W E. 1987. Reactivity of organic-rich sediment in seawater at 350°C, 500 bars: experimental and theoretical constraints for the Guaymas Basin hydrothermal system. Geochimica et Cosmochimica Acta, Vol. 51, 1997–2010.

THOROGOOD, E J. 1990. Provenance of the pre-Devonian sediments of England and Wales: Sm-Nd isotopic evidence. Journal of the Geological Society of London, Vol. 147, 591–594.

TISSOT, B P, and WELTE, D H. 1978. Petroleum formation and occurrence. 538 pp. (Springer-Verlag.)

TONKS, L H, JONES, R C B, LLOYD, W, and SHERLOCK, R L. 1931. The geology of Manchester and the South-East Lancashire Coalfield. Memoir of the Geological Survey of Great Britain, Sheet 85. 240 pp.

TOTH, J. 1963. A theoretical analysis of groundwater flow in small drainage basins. Journal of geophysical research, Vol. 68. 4795–4812.

TRESISE, G. 1989. Chirotherium: the first finds at Storeton Quarry, Cheshire, and the role of the Liverpool Natural History Society. Geological Curator, Vol. 5, 135–152.

TREWIN, N H, and HOLDSWORTH, B K. 1973. Sedimentation in the Lower Namurian rocks of the North Staffordshire Basin. Proceedings of the Yorkshire Geological Society, Vol. 39, 371–405.

TRUEBLOOD, S, BRYAN, C, and PICKERING, S. 1995. The Douglas Oilfield and its implications for exploration on the Irish Continental Shelf. 39–40 in The Petroleum Geology of Ireland's Offshore Basins. CROKER, P F, and SHANNON, P M (editors). Special Publication of the Geological Society of London, No. 93.

TRUSHEIM, F. 1961. Ober Diskordanzen im Mittleren Buntsandstein Norddeutschlands zwischen Ems und Weser. Erdol Zeitschrift fiir Bohr und Fordertechnik, Vol. 77, 361–367.

TRUSHEIM, F. 1963. Zur Gliederung des Buntsandsteins. Erela Zeitschrift Bohr und Fordertechnik, Vol. 79, 277–192.

TUCKER, R M. 1981. Giant polygons in the Triassic salt of Cheshire, England: a thermal contraction model for their origin. Journal of Sedimentary Petrology, Vol. 51, 779–786.

TUCKER, R M, and TUCKER, M E. 1981. Evidence of synsedimentary tectonic movements in the Triassic halite of Cheshire. Nature, London, Vol. 290, 495–496.

TUKEY, P A. 1977. Exploratory data analysis. 506 pp. (Reading, Mass.: Addison-Wesley.)

TUREKIAN, K K, and WEDEPOHL, K H. 1961. Distribution of elements in some major units of the earth crust. Bulletin of the Geological Society of America, Vol. 72, 186–194.

TURNER, D R, DICKSON, A G and WHITFIELD, M. 1980. Water-rock partition coefficients and the composition of natural waters: a reassessment. Marine Chemistry, Vol. 9, 211–218.

TURNER, P. 1980. Continental red beds. Elsevier Developments in Sedimentology, No. 29.

VAN DER am, J G L A. 1983. Aspects of Middle and late Triassic palynology. 6. Palynological investigations in the Ladinian and Lower Karnian of the Western Dolomites, Italy. Review of Palaeobotany and Palynology, Vol. 39, 189–300.

VAN HOORN, B. 1987. Structural evolution, timing and tectonic style of the Sole Pit inversion. Tectonophysics, Vol. 137, 239–284.

VELDE, B. 1977. Clays and clay minerals in natural and synthetic systems. Elsevier Developments in Sedimentology, No. 21.

VERMA, Y K. 1981. Studies in virgin strata temperatures, with special reference to the NCB's Western Area mines. The Mining Engineer, Vol. 141, 655–663.

VISSCHER, H, and BRUGMAN, W A. 1981. Ranges of selected palynomorphs in the Alpine Triassic of Europe. Review of Palaeobotany and Palynology, Vol. 4, 115–128.

VOSS, C I. 1984. A finite element simulation model for saturated-unsaturated, fluid-density dependant groundwater flow with energy transport or chemically-reactive single species solute transport. Report of the United States Geological Survey.

WADE, A. 1910. A deep boring at Heswall, Cheshire. Transactions Institute of Mining Engineers, Vol. xxxix, 163–176.

WADGE, A J, HUDSON, J M, PATRICK, D J, SMITH, I F, EVANS, A D, APPLETON, J D, and BATESON, J H. 1982. Copper mineralisation near Middleton Tyas, North Yorkshire. Institute of Geological Sciences Mineral Reconnaissance Programme Report, No. 54.

WADGE, G, YOUNG, P A V, and MASON, D C. 1991. Prospectivity mapping using an integrated image processing and expert system. 169–179 in Development of regional exploration criteria for buried carbonate-hosted mineral deposits: a multidisciplinary study in northern England. PLANT, J A, and JONES, D G (editors). British Geological Survey Technical Report, WP/91/1.

WADGE, G, YOUNG, P A V, and MASON, D C. 1992. Simulation of geological processes using an expert system. Journal of the Geological Society of London, Vol. 149, 455–463.

WALKER, T R. 1967. Formation of red beds in modern and ancient deserts. Bulletin of the Geological Society of America, Vol. 78, 353–368.

WALKER, T R. 1974. Formation of red beds in moist tropical climates. Bulletin of the Geological Society of America, Vol. 85, 633–638.

WALKER, T R. 1976. Diagenetic origin of continental red beds. 240–282 in The continental Permian in central, west and south Europe. FALKE, H (editor). (Dordrecht: Reidel Publishing Co.)

WALKER, T R. 1989. Application of diagenetic alterations in redbeds to the origin of copper in stratiform copper deposits. 85–96 in Sediment-hosted stratiform copper deposits. BOYLE, R W, BROWN, A C, JEFFERSON, C W, JOWETT, E C, and KIRKHAM, R V (editors). Special Paper of Geological Association of Canada, 36.

WALKER, T R, and HONEA, R M. 1969. Iron content of modern deposits in the Sonoran desert: a contribution to the origin of red beds. Bulletin of the Geological Society of America, No. 80, 535–544.

WALKER, T R, and RUNNELLS, D D. 1984. Potential role of smectite-illite conversion in the formation of haematite pigment in red beds. Abstracts with Programs, Geological Society of America, No. 16, 685.

WALKER, T R, WAUGH, B, and GRONE, A. 1978. Diagenesis in first-cycle desert alluvium of Cenozoic age, southwestern United States and northwestern Mexico. Bulletin of the Geological Society of America, Vol. 89, 19–32.

WALSH, J J, and WATTERSON, J. 1987. Analysis of the relationship between displacements and dimensions of faults. Journal of Structural Geology, Vol. 10, No. 3, 239–247.

WALSH, J J, and WATTERSON, J. 1992. Populations of faults and fault displacements and their effects on estimates of fault-related regional extension. Journal of Structural Geology, Vol. 14, No. 6, 701–712.

WALSH, J J, WATTERSON, J, and YIELDING, G. 1991. The importance of small-scale faulting in regional extension. Nature, London, Vol. 351, 391–393.

WALTON, N R G. 1982. A detailed hydrochemical study of groundwaters from the Triassic sandstone aquifer of south-west England. Report of the Institute of Geological Sciences, No. 81/5.

WAPLES, D W. 1983. Reappraisal of anoxia and organic richness, with emphasis on Cretaceous of North America. Bulletin of the American Association of Petroleum Geologists, No. 67, 963–978.

WARD, T. 1898. The rock salt deposits of Northwich, Cheshire, and the results of their exploration. Part 1. Transactions of the Manchester Geological and Mineralogical Society, Vol. 25, 274–305; Part 2, 530–567.

WARNER, M R. 1987. Seismic reflections from the Moho - the effect of isostasy. Geophysical Journal of the Royal Astronomical Society, Vol. 88, 425–435.

WARRINGTON, G. 1963. The occurrence of the branchiopod crustacean Euestheria in the Keuper Sandstone of Alderley Edge, Cheshire. Liverpool and Manchester Geological Journal Vol. 3, 315–320.

WARRINGTON, G. 1965. The metalliferous mining district of Alderley Edge, Cheshire. Mercian Geologist, Vol. 1, 111–129.

WARRINGTON, G. 1970a. The "Keuper" Series of the British Trias in the Northern Irish Sea and neighbouring areas. Nature, London, Vol. 226, 254–256.

WARRINGTON, G. 1970b. The stratigraphy and palaeontology of the "Keuper" Series of the central Midlands of England. Quarterly Journal of the Geological Society of London, Vol. 126, 183–223.

WARRINGTON, G. 1974. Les evaporites du Trias britannique. Bulletin de la Societe Geologique de France, Vol. 16, No. 6, 708–723.

WARRINGTON, G. 1980. The Alderley Edge mining district. Amateur Geologist, Vol. 8, 4–13.

WARRINGTON, G. 1981. The copper mines of Alderley Edge and Mottram St Andrew, Cheshire. Journal of the Chester Archaeological Society, 64, 47–73.

WARRINGTON, G. 1994a. Review of previous work on and related to sediment-hosted mineralization in the Cheshire Basin. British Geological Survey, Technical Report, WH/94/59R.

WARRINGTON, G. 1994b. A review of the structural and stratigraphical setting of sediment-hosted polymetallic mineralisation in the Cheshire Basin. British Geological Survey Technical Report, WH/94/134R.

WARRINGTON, G. 1994c. Review of the geological setting and stratigraphy of the Cheshire Basin. British Geological Survey, Technical Report, WH/94/58R.

WARRINGTON, G, AUDLEY-CHARLES, M G, ELLIOTT, R E, EVANS, W B, IVIMEY-COOK, H C, KENT, P E, ROBINSON, P L, SHOTTON, F W, and TAYLOR, F M. 1980. A correlation of Triassic rocks in the British Isles. Special Report of the Geological Society of London, No. 13, 77 pp.

WARRINGTON, G, and IVIMEY-COOK,  H C. 1992. Triassic. 97–106 in Atlas of palaeogeography and lithofacies. COPE, J C W, INGHAM, J K, and RAWSON, P F (editors). Memoir of the Geological Society of London, Vol. 13.

WAUGH, B. 1978. Authigenic K-feldspar in British Permo-Triassic sandstones. Journal of the Geological Society of London, Vol. 135, 51–56.

WEAVER, B L, and TARNEY, J. 1984. Empirical approach to estimating the composition of the continental crust. Nature, London, Vol. 310, 575–577.

WEDD, C B, and KING, W B R. 1924. Geology of the country around Flint, Hawarden and Caergwrle. Memoir of the Geological Survey, Sheet 108.

WEDD, C B, SMITH, B, KING, W B R, and WRAY, D A. 1929. The geology of the country around Oswestry. Memoir of the Geological Survey of Great Britain. 234 pp.

WEDEPOHL, K H (editor) 1978. Handbook of geochemistry. (Berlin-Heidelberg: Springer-Verlag.)

WHITE, D E. 1958. Liquid of inclusions in sulfides from Tri-State (Missouri-Kansas-Oklahoma) is probably connate in origin. Bulletin of the Geological Society of America, Vol. 69, 1660.

WHITE, N J. 1990. Does the uniform stretching model work in the North Sea? 217–239 in Tectonic evolution of the North Sea rifts. BLUNDELL, D J, and Gins, A D (editors). (Oxford: Clarendon.)

WHITE, P H N. 1949. Gravity data obtained in Great Britain by the Anglo-American Oil Company Limited. Quarterly Journal of the Geological Society of London, Vol. 104, 339–364.

WHITFIELD, M. 1979. The mean oceanic residence time (MORT) concept - a rationalisation. Marine Chemistry, Vol. 8, 101–123.

WHITTAKER, A (editor). 1985. Atlas of onshore sedimentary basins in England and Wales: post-Carboniferous tectonics and stratigraphy. 71 pp. (Glasgow and London: Blackie.)

WHITTAKER, A, HOLLIDAY, D W, and PENN, I E. 1985. Geophysical logs in British stratigraphy. Special Report of the Geological Society of London, No. 18, 1–74.

WHITTARD, W F. 1952. A geology of south Shropshire. Proceedings of the Geologists' Association, Vol. 63, 143–197.

WILLS, L J. 1950. The palaeogeography of the Midlands (2nd edition). 147 pp. (University Press of Liverpool.)

WILLS, L J. 1956. Concealed coalfields. (London: Blackie and Son.)

WILLS, L J. 1970. The Triassic succession in the central Midlands in its regional setting. Quarterly Journal of the Geological Society of London, Vol. 126, 225–283.

WILLS, L J. 1976. The Trias of Worcestershire and Warwickshire. Report of the Institute of Geological Sciences, No. 76/2. 211 pp.

WILLS, L J. 1978. A palaeogeological map of the Lower Palaeozoic floor below the cover of upper Devonian, Carboniferous and later formations. Memoir of the Geological Society of London, No. 7.

WILSON, A A. 1990. The Mercia Mudstone Group (Trias) of the East Irish Sea Basin. Proceedings of the Yorkshire Geological Society, Vol. 48, 1–22.

WILSON, A A. 1993. The Mercia Mudstone Group (Trias) of the Cheshire Basin. Proceedings of the Yorkshire Geological Society, Vol. 49, 171–188.

WILSON, A A, and EVANS, W B. 1990. Geology of the country around Blackpool. Memoir of the British Geological Survey, Sheet 66. (England and Wales). 82 pp.

WITHERSPOON P A, and NAGASHIMA K. 1957. Use of trace metals to identify Illinois crude oils. Circular of the Illinois State Geological Survey, No. 239, 1–16.

WOLERY, T J. 1992a. EQ3/6 package overview and installation guide. Lawrence Livermore National Laboratory Report.

WOLERY, T J. 1992b. EQ3NR, theoretical manual and user's guide. Lawrence Livermore National Laboratory Report.

WOLERY, T J, and DAVELER, S A. 1992. EQ6 theoretical manual and user's guide. Lawrence Livermore National Laboratory Report.

WOLF, K H. 1994. Artifical Intelligence (AI) and Expert Systems (ES). Ore Geology Reviews, Vol. 9, 241–243.

WOODCOCK, N H. 1984. The Pontesford Lineament, Welsh Borderland. Journal of the Geological Society of London, Vol. 141, 1001–1014.

WOODLAND, B G. 1985. Relationship of concretions and chlorite-muscovite porphyroblasts to the development of domainal cleavage in very low-grade metamorphic deformed rocks from north-central Wales, Great Britain. Journal of Structural Geology, Vol. 7, 205–215.

ZIEGLER, P A, and VAN HOORN, B. 1989. Evolution of the North Sea rift system. 471–500 in Extensional tectonics and stratigraphy of the North Atlantic margins. TANKARD, A J, and BALKWILL, H R (editors). American Association of Petroleum Geologists Memoir, Vol. 46.

ZIELINSKI, R A, BLOCH, S, and WALKER, T R. 1983. The mobility and distribution of heavy metals during the formation of first cycle red beds. Economic Geology, Vol. 78, 1574–1589.

ZIELINSKI, R A, BLOCH, S, and WALKER, T R. 1986. Mobility and distribution of heavy metals during the formation of first cycle red beds. 52–59 in Studies in Diagenesis. MUMPTON, F A (editor). Bulletin of the United States Geological Survey, No. 1578.

Figures, plates and tables

Median concentrations (in ppm) of Cu, Pb, and Zn in the MMG and the SSG in the Cheshire Basin, and the calculated total metal content (in Mt) of the basin fill, based on extrapolation of present-day concentrations to the volume of rocks prior to erosion.

(Figure 1) Preserved PermoTriassic sediment thickness (m) in southern Britain (modified from data in Whittaker, 1985).

(Figure 2) Locations of boreholes, quarries, mines and outcrops from which samples were collected, and of other, deep boreholes.

(Figure 3) Sample-handling procedure. At each question, if Yes, follow Y branch to conclusion; then return and follow C (continue) . If No, follow N branch.

(Figure 4) Caledonian geotectonic model for early Devonian time, from Soper et al. (1987). M: Midlands Massif.

(Figure 5) Distribution of the tectonic zones in the Variscides of northwest Europe, from Coward (1990).

(Figure 7) General geology of the Cheshire Basin.

(Figure 6) Location of the Cheshire Basin.

(Figure 8) Stratigraphy of the basin-fill succession of the Cheshire Basin, and its relationship to the youngest basin-floor rocks.

(Figure 9) The Cheshire Basin area with areas of basin-fill sediments free from superficial deposits shown in yellow. The form of the basin is outlined by areas of pre-Permian rocks (brown).

(Figure 10) The Cheshire Basin and adjacent areas in relation to BGS 1:50 000 geological sheets. 83, Formby; 84, Wigan; 85, Manchester; 86, Glossop; 96, Liverpool; 97, Runcorn; 98, Stockport; 99, Chapel en le Frith; 108, Flint; 109, Chester; 110, Macclesfield; 111, Buxton; 121, Wrexham; 122, Nantwich; 123, Stoke-on-Trent; 124, Ashbourne; 137, Oswestry; 138, Wem; 139, Stafford; 140, Burton upon Trent; 151, Welshpool; 152, Shrewsbury; 153, Wolverhampton; 154, Lichfield.

(Figure 11) Borehole geophysical logs (GR, gamma ray; BHCS, borehole-compensated sonic) and correlation of formation boundaries below the MMG between the Prees and Knutsford boreholes (based on Evans et al., 1993). Depths in metres below OD. CPBF, Chester Pebble Beds Formation; KSF, Kinnerton Sandstone Formation; MMF, Manchester Marls Formation; CB, Carboniferous; F, fault, faulted section; TSZ, top Silicified Zone.

(Figure 13)) for lithological key. From Wilson (1993)." data-name="images/P1000266.jpg">(Figure 12) The succession in the MMG of the Cheshire Basin and its corrrelation with the East Irish Sea and Southern North Sea basins. Column 1 is taken from Rhys (1974). See ((Figure 13)) for lithological key. From Wilson (1993).

(Figure 13) Sections in the Tarporley Siltstone Formation. Note difference in scale between Section 1 and Sections 2 and 3.

(Figure 14) Sections in the Bollin Mudstone Formation. All but section 4 are solely in the upper half of the formation. Section 3 is based on Arthurton (1980) with additions. See ((Figure 13)) for lithological key. From Wilson (1993).

(Figure 15) Isopachs (in m) of the Bollin Mudstone Formation. Rectangular outlines are of BGS 1:50 000 sheets (see (Figure 10)). From Wilson (1993).

(Figure 16) Sections in the Northwich Halite Formation. The key map also shows isopachs (in feet) of the formation. Based on Evans et al. (1968) with additions and alterations.

(Figure 17) Isopachs (in m) of the combined Byley Mudstone and Wych Mudstone formations. Rectangular outlines are of BGS 1:50 000 sheets (see (Figure 10)). From Wilson (1993).

(Figure 18) Sections in the Byley Mudstone Formation. Section 2 is based on Arthurton (1980) with additions. See (Figure 13) for lithological key. From Wilson (1993).

(Figure 19) Sections in the Wych Mudstone Formation. Section 5 is based on Evans et al. (1968), fig. 24, with additions. See (Figure 13) for lithological key. From Wilson (1993).

(Figure 20) Sections in the Wilkesley Halite Formation. See (Figure 13) for lithological key.

(Figure 21) Boreholes in the Brooks Mill Mudstone Formation and its equivalents in Cheshire, Nottinghamshire and Northern Ireland. See (Figure 13) for lithological key. From Wilson (1993).

(Figure 22) The Blue Anchor Formation and Penarth Group in the Wilkesley and Plattlane boreholes (biostratigraphy by H C Ivimey-Cook).

(Figure 23) Borehole geophysical logs (gamma ray and borehole-compensated sonic) of the upper MMG and the Penarth and Lias groups in the Prees Borehole. Depths in m below OD. LL1—LL3, geophysical log units in the Lias (after Penn, 1987; base of LL1 revised by Warrington, 1994c); BAF, Blue Anchor Formation; WF, Westbury Formation; CM, Gotham Member; LM, Langport Member; LF, Lilstock Formation.

(Figure 24) Lias succession and biostratigraphy of the Wilkesley and Plattlane boreholes (revised by H C Ivimey-Cook, 1993; after Poole and Whiteman, 1966).

(Figure 49), (Figure 50), (Figure 51)." data-name="images/P1000279.jpg">(Figure 25) Colour-shaded illuminated image of Bouguer gravity, with the locations of the sections in (Figure 49), (Figure 50), (Figure 51).

(Figure 26) The distribution of Permo-Triassic rocks in relation to principal basement structures. BF Bala Fault; CFD Clun Forest Disturbance; CSFZ Church Stretton Fault Zone; PLF Pontesford—Linley Fault; SA Shelve Anticline; WBRRFS Wem—Bridgemere—Red Rock Fault System.

(Figure 27) Shaded-relief image of the magnetic field (reduced to pole), illuminated from the north-west at an angle of 45°. BF Bala Fault; HF Hodnet Fault; BoF Boundary Fault; PLF Pontesford—Linley Fault ; CF Croxteth Fault; RRF Red Rock Fault; CSF Church Stretton Fault; TCF Titterstone Glee Fault; EMF East Malvern Fault; VCF Vale of Clwyd Fault; FPF Formby Point Fault; WF Wem Fault

(Figure 28) Sketch diagram to illustrate development of Wem Fault by repeated fault reactivation: a. Carboniferous basin development b. end-Carboniferous (Variscan) basin inversion c. Permo-Triassic extension and basin development, with reactivation of earlier fault.

(Figure 29) Permo-Triassic regional tectonic framework of the North Atlantic region. WB Worcester Basin; WC Wessex Channel Basin; VG Viking Graben; FR Faeroes Ridge; RT Rockall Trough; RH Rockall High; M Moray Firth Basin; NPB Northern Permian Basin; OG Oslo Graben; MNS Mid-North Sea High; RF Ringkobing—Fyn High; SPB Southern Permian Basin; EIS East Irish Sea Basin; LBM London—Brabant Massif; CSB Celtic Sea basins; WA Western Approaches Basin; PB Paris Basin.

(Figure 30) Tectonic framework of the North Atlantic region (modified from Knott et al., 1993): in late Triassic times - in late Jurassic times.a. in late Triassic times b. in late Jurassic times.

(Figure 31) Lithostratigraphy of the basin fill related to gamma-ray (GR) and borehole-compensated sonic (BHCS) logs of the Knutsford and Prees boreholes (mapped seismic reflectors in colour).

(Figure 32) Seismic stratigraphy of the basin fill (Sandbach—Knutsford Sub-basin), correlated with the geophysical logs from the Knutsford Borehole and a synthetic seismogram.

(Figure 33) Depth contours on Base Jurassic.

(Figure 34) Depth contours on Base MMG.

(Figure 35) Depth contours on Base SSG.

(Figure 36) Depth contours on Base Permo-Triassic.

(Figure 38)." data-name="images/P1000291.jpg">(Figure 37) Principal structural features and major faults of the Cheshire Basin, together with locations of cross-sections in (Figure 38).

(Figure 38) True scale cross-sections through the Cheshire Basin (locations of sections on (Figure 38)." data-name="images/P1000291.jpg">(Figure 37)). The strata classified as 'Permian' include all Permo-Triassic rocks beneath the SSG.

(Figure 39) Seismic-reflection profile across the Wem–Audlem Sub-basin. Note eastwards pinchout of Carboniferous strata beneath Permo-Triassic cover (data courtesy of Hamilton Brothers Oil and Gas).

(Figure 40) Seismic-reflection profile across the Edgerley Fault and the Milton Green Inlier (data courtesy of BP).

(Figure 41) Seismic-reflection profile in the northern part of the Cheshire Basin showing northward thinning of the Permian sequence. Note also the very thick Carboniferous succession with seismically reflective Coal Measures in its upper part (data courtesy of Amoco UK).

(Figure 42) Seismic-reflection profile across the Brook House Fault. Note thickening of the Permian sequence from the Lymm High into the Sandbach—Knutsford Sub-basin. Note also small, steep reverse fault on Lymm High (data courtesy of BP).

(Figure 43) Seismic-reflection profile across the Alderley High; Note stratigraphical thinning of Permian strata towards the north-east margin of the Cheshire Basin. Note also the thick underlying Carboniferous succession with seismically reflective Coal Measures in upper part (data courtesy of Britoil/BP).

(Figure 44) Seismic-reflection profile showing the Mobberley Fault and the Alderley High (data courtesy of Sovereign Oil).

(Figure 45) Seismic-reflection profile across the eastern margin of the Cheshire Basin, formed by the Red Rock Fault. Note the very thin Permian sequence at the southern end of the Alderley High, pinching out eastwards. Note also tight folding in hangingwall block of Red Rock Fault (data courtesy of Fina Exploration).

(Figure 46) Seismic-reflection profile across the south-east margin of the Cheshire Basin, marked by the Bridgemere and Wem faults. Note also the underlying (reactivated) Variscan thrust (data from Evans et al., 1993).

(Figure 47) Seismic-reflection profile across the King Street Fault (data courtesy of Enterprise Oil).

(Figure 48) Seismic-reflection profile showing reverse faulting and development of small pop-up structure in the vicinity of the King Street Fault (data courtesy of Amoco UK).

(Figure 49) 2-D gravity interpretation along Section A—A'. See (Figure 49), (Figure 50), (Figure 51)." data-name="images/P1000279.jpg">(Figure 25) for location of section. Orientation: left of section, north-west. Stipple, Lias Pale pink, Mercia Mudstone Group Deep pink, Sherwood Sandstone Group Black, Lower Permian Sandstones Pink, Silesian No ornament, basement (including Lower Carboniferous) Diagonal lines, magnetic basement.

(Figure 50) 2-D gravity interpretation along Section B—B'. See (Figure 49), (Figure 50), (Figure 51)." data-name="images/P1000279.jpg">(Figure 25) for location of section. Orientation: left of section, west. Pale pink, Mercia Mudstone Group; Deep pink, Sherwood Sandstone Group; Black, Lower Permian Sandstones; Pink, Silesian; No ornament, basement (including Lower Carboniferous) Diagonal lines, magnetic basement.

(Figure 51) 2-D gravity interpretation along Section C—C'. See (Figure 49), (Figure 50), (Figure 51)." data-name="images/P1000279.jpg">(Figure 25) for location of section. Orientation: left of section, northwest. Pale pink, Mercia Mudstone Group Deep pink, Sherwood Sandstone Group Black, Lower Permian Sandstones Pink, Silesian No ornament, basement (including Lower Carboniferous) Diagonal lines, magnetic basement.

(Figure 52) North—south sketch cross-section through the Dinantian outcrop in north-east Wales near the western margin of the Cheshire Basin. Cross-section along the outcrop of the Carboniferous Limestone of north-east Wales. The original section (Smith and George, 1961, fig. 25) noted the Dinantian syn-depositional rising of the early Palaeozoic anticlines (e.g. Cyrn-y-Brain) and the corresponding increase in thickness above the synclines (e.g. Llangollen). This is now interpreted as syn-depositional faulting on the Llanelidan (Bala), Aqueduct and Wein faults, forming half-graben basins. The Dinantian Cyrn-y-Brain High, with relatively thin strata, hosted Pb-Zn mineralisation (Minera). The Llanymynech Dinantian strata hosted Cu mineralisation discovered by the Romans. The ages of the lithostratigraphical units were revised by Somerville and Strank (1984).

(Figure 53) Borehole-log data plotted on standard compaction curves.

(Figure 54) Estimated eroded overburden (metres): a. values at the analysed boreholes b. interpretive contours.

(Figure 55) Structure-contour maps restored to maximum burial depth (c.60 Ma). Striped ornament denotes zone of structural attenuation by inclined normal faults: a. Base Jurassic b. Base SSG Base MMG d. Base Permo-Triassic.

(Figure 56) HOTPOT grids of total (decompacted) sediment thicknesses (metres):a. at 251 Ma, end of Permian deposition (thicknesses do not include eroded Permian strata) b. at 242 Ma, end of SSG deposition (thicknesses do not include eroded Permian strata and SSG) c. at 205 Ma, end of Triassic (thicknesses include estimated eroded Permo-Triassic strata) d.at c.97 Ma, end of late Cimmerian erosion (thicknesses include estimated eroded Permo-Triassic, Jurassic and lower Cretaceous strata) e. at c.60 Ma (thicknesses include total estimated eroded overburden i.e. Permo-Triassic, Jurassic, Cretaceous and Palaeocene strata).

(Figure 57) HoTPoT burial history plot from the Wem—Audlem Sub-basin.

(Figure 58) Super-regional model: HOTPOT grid of sediment-starved (tectonic) subsidence (metres) at 60 Ma.

(Figure 59) Regional analysis of fault strikes in the north-eastern, central and southern areas of the Cheshire Basin.

(Figure 60) Transects of the Cheshire Basin, parallel to the direction of extension. The transects, labelled AA to OO, are used in the assessment of fault heaves at Base Permo-Triassic.

(Figure 61) Frequency analysis of fault throws in the Cheshire Basin, displayed on log-log plot. Throws that fall on a linear trend obey power-law (fractal) distribution.

(Figure 62) Development of extension in Cheshire Basin, based on sequential restoration of corrected heaves. Arrows denote preferred dominant extension direction (083–263°), producing dip-slip displacement on WBRRFS at eastern margin, but sinistral transtension on WBRRFS at south-east margin.

(Figure 63) HOTPOT present-day temperature grids (Heat flow (Q) = 52 mWm⁻²): a. Base Jurassic b. Base MMG c. Base SSG d. Base Permo-Triassic

(Figure 64) Subsidence and heat flow v. time, based on uniform lithospheric extension model.

(Figure 65) HOTPOT temperature grids at c. 60 Ma (Palaeocene): a. Base Jurassic b. Base MMG c. Base SSG d. Base Permo-Triassic

(Figure 66) HOTPOT thermal history plot from the Wem—Audlem Sub-basin.

(Figure 67) HOTPOT pseudo-maturity grids at c.60 Ma (Palaeocene): a. Base Permo-Triassic b. Base MMG

(Figure 68) Compositions of typical detrital tourmaline assemblages in PermoTriassic sandstones of the Cheshire Basin. Each plot presents data for 50 individual tourmaline grains (shown as dots) in a single sample. Analyses have been normalised to 100%.

(Table 12)." data-name="images/P1000323.jpg">(Figure 69) Crossplot of apatite-tourmaline and monazite-zircon indices (ATi and MZi) for PermoTriassic sandstones of the Cheshire Basin. ATi and MZi are defined in the text and (Table 12).

(Figure 70) Downhole variations in heavy-mineral assemblages in the Kinnerton Sandstone of Stanlow Borehole.

(Figure 71) Stratigraphical variations in the range of means and standard deviations of the composition of detrital tourmaline assemblages in Permo-Triassic sandstones of the Cheshire Basin. Each mean and standard deviation value refers to a single tourmaline assemblage (as shown in (Figure 68)), characterised by the analysis of 50 individual tourmaline grains.

(Figure 72) Locations of pebble samples used for Sm-Nd isotope study, and the site of the Little Hay Borehole (examined for heavy-mineral work).

(Figure 73) Summarised Nd results for sedimentary rocks and pebbles from the Permo-Triassic of the Cheshire Basin and adjacent areas. Data from the literature are also plotted for potential source areas (see text for references).

(Figure 74) Explanatory diagram for box-and-whisker plots. See text for explanation.

(Figure 75) Box-and-whisker plots for selected elements in formations of the SSG of the Cheshire Basin.

(Figure 76) Box-and-whisker plots for selected elements normalised to alumina in formations of the SSG of the Cheshire Basin.

(Figure 77) Plot of K₂O against Al₂O₃ for the SSG of the Cheshire Basin.

(Figure 78) Box-and-whisker plots for Ni/Al₂O₃ and K₂O/Al₂O₃ for the Kinnerton and Wilmslow Sandstone formations from northern and southern parts of the Cheshire Basin.

(Figure 79) Box-and-whisker plots for selected elements plotted as a function of sedimentary facies.

(Figure 80) Typical REE plots for the SSG of the Cheshire Basin. The upper plots are chondrite normalised; the lower plots are normalised to Post-Archaean Average Shale (PAAS). Normalising values are from Taylor and McLennan, 1985.

(Figure 81) La and Lu plotted against selected major elements for SSG⁻and Tarporley Siltstone of the Cheshire Basin.

(Figure 82) Plots of La against other REE for SSG and Tarporley Siltstone of the Cheshire Basin.

(Figure 83) Box-and-whisker plots for La/Lu normalised to Al₂O₃ for formations of the SSG and the Tarporley Siltstone of the Cheshire Basin.

(Figure 84) Box-and-whisker plots for selected elements in formations of the MMG of the Cheshire Basin. Data for the Manchester Marls are included for comparison.

(Figure 85) Boxand-whisker plots for selected elements normalised to alumina in formations of the MMG of the Cheshire Basin. Data for the Manchester Marls are included for comparison.

(Figure 86) Variations of Cr/Al, Ni/Al and Zr/Al with depth in the MMG of the Wilkesley Borehole.

(Figure 87) Box-and-whisker plots of Cr/ Al₂O₃ Ni/Al₂O₃ and Zr/Al₂O₃ for the Byley and Wych mudstones, by borehole.

(Figure 88) Multi-element geochemical diagrams for laminated and structureless mudstone facies of the Byley Mudstones from the Crewe Borehole (normalised to upper continental crust).

(Figure 89) Variations of Cr/Al, Ni/Al and Zr/Al with depth in the Helsby Sandstone and Tarporley Siltstone of the Saughall Massie Borehole.

(Figure 90) Regional variations in the monazite-zircon index (MZi) shown by Permo-Triassic sandstones of the UK and adjacent continental shelf. The mean MZi of the Cheshire basin samples is shown by a single point.

(Figure 91) Box-andwhisker plots for the SSG, subdivided into reduced, oxidised and undifferentiated categories on the basis of rock colour. See (Figure 74) for explanation of box-and-whisker plots.

(Figure 92) Plot of δ¹³C against δ¹⁸O for diagenetic and fracture-mineralising carbonates from the Permo-Triassic rocks of the Cheshire Basin. Analyses are grouped (shaded areas) according to their paragenetic characteristics, and the summary diagram (right) illustrates the paragenetic sequence of carbonate cements and mineralisation.

(Figure 93) Diagrammatic illustration (not to scale) of the relationship between faults and mineralised sandstone bodies in West Mine, Alderley Edge and Clive Mine.

(Figure 94) Details of DE6 paragenetic and metalliferous mineralisation sequences from Clive Mine and West Mine (this study) compared with the Bickerton Hills and other localities from the Alderley Edge mining area (based on data by Ixer and Vaughan, 1982; Naylor et al., 1989; re-interpreted within the framework of Diagenetic Episodes proposed in the present study).

(Figure 95) Contoured distribution of δ¹³C data for DE6e calcite (or DE6E dolomite) compared with distribution of DE6 sulphide species. Contour based on minimum value of δ¹³C determined at each locality.

(Figure 96) Geochemical profiles for samples taken on traverses on the three levels at West Mine, Alderley Edge, approximately perpendicular to the south-west boundary fault.

(Figure 97) Geochemical profiles for samples taken on traverses on the three levels at West Mine, Alderley Edge, approximately perpendicular to the south-west boundary fault.

(Figure 98) Scatter plots of CaO and MgO against loss on ignition (at 1050°C) for the SSG. Many points fall on calcite and dolomite trends, showing that most CaO and MgO is associated with these minerals.

(Figure 99) Box-and-whisker plots for the MMG, subdivided into reduced, oxidised and undifferentiated categories on the basis of rock colour. See (Figure 74) for explanation of box-and-whisker plots.

(Figure 100) Box-and-whisker plots for the MMG, subdivided into reduced, oxidised and undifferentiated categories on the basis of rock colour. Data normalised to Al₂O₃ content. See (Figure 74) for explanation of box-and-whisker plots.

(Figure 101) Multi-element diagrams of analysed samples from oxidised and reduced laminae in the Byley Mudstone Formation, Crewe Heatflow Borehole (normalised to upper continental crust).

(Figure 102) Multi-element diagrams of analysed samples from oxidised and reduced laminae in the Tarporley Siltstone Formation, Saughall Massie Borehole (normalised to upper continental crust).

(Figure 103) δ³⁴S stable-isotope data for DE6 sulphides and baryte cements and fracture mineralisation in the SSG of the Cheshire Basin. Includes data obtained in the present study and data from Naylor et al. (1989).

(Figure 104) Stratigraphical variation in δ³⁴S stable-isotope values for primary (sedimentary) anhydrite and synsedimentary (DE1) anhydrite cements (and their secondary hydrated gypsum replacements) in the MMG of the Cheshire Basin .Data plotted include data obtained during this study, together with limited published data from the Cheshire Basin (Taylor, 1983; Naylor et al., 1989). These sulphates are compared with the Upper Triassic oceanic sulphate—sulphate mineral curve of Taylor (1983) and ranges of published δ³⁴S isotope data for stratigraphical horizon (represented by boxed areas) for European evaporitic sulphates of Triassic age (quoted by Taylor, 1983). Line A represents the best estimate for δ³⁴S. Line B represents an alternative for δ³⁴S during the Anisian if the Rot data are regarded as only representing a localised event.

(Figure 105) Element covariation for fluid inclusions in halite, Wilkesley Borehole. Element calculations normalised to 1000 ppm Sr. a. Mg v. K; b. Rb v. K; c. Pb v. K; d. Ba v.K;e. Mn v. K; f. Cu v. K; g. Cu v. Mn.

(Figure 106) Chromatograms of the alkane extract from samples from Grinshill Quarry, Mickle Trafford Borehole, Alderley Edge and Burton Point. A fraction with Peak Carbon Number of C17 to C20 is seen in all four samples. This fraction is present in many of the samples analysed, but is not necessarily present in the greatest concentration. The chromatograms also show a highly biodegraded fraction seen as a rising baseline, or hump. These four samples also show a third very light fraction (Peak Carbon Number approximately C10 to C14).

(Figure 108), (Figure 109), (Figure 110), (Figure 111), (Figure 112), (Figure 113), (Figure 114), (Figure 115), (Figure 116), (Figure 117), (Figure 118) (Figure 119) is shown." data-name="images/P1000361.jpg">(Figure 107) Groundwater flow in the Cheshire Basin. The line of the section illustrated in (Figure 108), (Figure 109), (Figure 110), (Figure 111), (Figure 112), (Figure 113), (Figure 114), (Figure 115), (Figure 116), (Figure 117), (Figure 118) (Figure 119) is shown.

(Figure 108) A section across the central Cheshire Basin, simplified for the purposes of coupled flow modelling.

(Figure 109) The freshwater flow field represented by logarithmically scaled Darcy flow vectors at the centre of each element.

(Figure 110) The freshwater flow field represented by a selection of particle pathlines.

(Figure 111) Steady-state density contours (ρ, in kg/m³) obtained by assuming a source zone in the lower part of the MMG with a brine density of 1200 kg/m3.

(Figure 112) The Darcy flow field represented by logarithmically scaled vectors for a source zone in the lower part of the MMG with a brine density of 1200 kg/m³.

(Figure 113) The Darcy flow field represented by a selection of particle pathlines for a source zone in the lower part of the MMG with a brine density of 1200 kg/m³.

(Figure 114) Steady-state density contours (ρ, in kg/m³) obtained by assuming a source zone in the lower part of the MMG with a brine density of 1100 kg/m³.

(Figure 115) The Darcy flow field represented by logarithmically scaled vectors for a source zone in the lower part of the MMG with a brine density of 1100 kg/m³.

(Figure 116) The Darcy flow field represented by a selection of particle pathlines for a source zone in the lower part of the MMG with a brine density of 1100 kg/m³.

(Figure 117) Steady-state density contours (ρ, kg/m³) obtained b_y assuming a source zone in the lower part of the MMG with a brine density of 1050 kg/m .

(Figure 118) The Darcy flow field represented by logarithmically scaled vectors for a source zone in the lower part of the MMG with a brine density of 1050 kg/m³.

(Figure 119) The Darcy flow field represented by a selection of particle pathlines for a source zone in the lower part of the MMG with a brine density of 1050 kg/m³.

(Figure 120) Schematic model for the evolution of red-bed formation waters, showing how contrasting formation-water chemistries may occur in different places within a red-bed sequence. The exact evolution of the formation waters will depend upon a wide variety of factors, such as the composition of the crystalline basement source areas of the sediments, occurrence of marine incursions, and fluid mixing. One possible evolutionary pathway for fresh, atmosphere-equilibrated recharge waters (A) is illustrated. In this example, evaporative concentration of pore waters leads to precipitation of calcite or dolomite; this increases Na/Ca, Na/Mg and (SO₄ + Cl)/(CO₃ + HCO₃) ratios, leading eventually to gypsum precipitation. This increases the Na/Ca and Cl/SO₄ ratios, and possibly causes development of Na-Cl waters. TDS = Total Dissolved Solids (based on Jowett, 1989).

(Figure 121) Logarithmic mineral-stability diagram for the system K₂O–SiO₂–Al₂O₃–H₂O, showing how variations in the activities of K+ and SiO₂ (aq.), and pH in diagenetic pore waters, can be major controls on authigenic phases during red-bed diagenesis. Phase boundaries were calculated for 25°C and 100°C at a pressure of 1 bar using the SUPCRT92 code (Johnson et al., 1991). As K+ activities increase due to mineral dissolution during burial, red-bed formation waters will move generally from the kaolinite stability field (A) to the muscovite stability field (B), and then to the K-feldspar stability field (C). Subsequent influx of dilute meteoric, or acidic basinal waters may then move the fluid composition back towards the kaolinite stability field.

(Figure 122) Eh-pH diagram for Cu at 25°C and a pressure of 1 bar. Tenorite, CuO; cuprite, Cu₂O; chalcocite, Cu₂S; covellite, CuS. The shaded field represents the most likely chemical limits of red-bed brines that are capable of transporting Cu as Cu⁺–Cl⁻ complexes. The field is defined by the area where hematite is stable, and is bounded by 9 > pH > 5 (considered a reasonable range for most ore-forming fluids) and the fields of tenorite, copper and chalcocite stability. 'A' is the area where hematite dissolution is possible at a pH within the range most likely to be encountered during diagenesis. '13' represents a typical shallow (< 300 m depth) groundwater from the Permo-Triassic fill of the Cheshire Basin. 'C' represents a Carboniferous Coal Measures water from a deep borehole near Chester. The diagram was constructed using data from the EQ3/6 thermodynamic database (Wolery and Daveler, 1992; Wolery, 1992a, b).

(Figure 123) Eh-pH diagrams for Fe, Zn, Pb and U at 25°C and a pressure of 1 bar. The field of 'Cu transport as CuCl₃²⁻, which is superimposed on each diagram is the same as the shaded field in (Figure 122). a. Fe (constructed using the EQ3/6 thermodynamic database (Wolery and Daveler, 1992; Wolery, 1992a, b)); b. Pb (phosgenite, PbCO₃.PbCl₂; cerussite, PbCO₃; galena, PbS) (adapted from Rose, 1989); c. Zn (smithsonite, ZnCO₃; sphalerite, ZnS) (adapted from Rose, 1989 ); d. U (uraninite, UO₂) (adapted from Brookins, 1984).

(Figure 124) Diagrammatic representation of the reaction of a single source of basinal brine with three different aquifer systems, resulting in Zn-rich, Pb-rich and Cu-rich deposits (adapted from Sverjensky, 1989).

(Figure 125) LogfO₂,_bars versus logfS₂,_b., at 125°C (adapted from Sverjensky, 1987). The relative stabilities of Fe- and Cu-bearing solid phases are shown. (Hematite, Fe₂O₃; magnetite, Fe₃O₄; pyrite, FeS₂; chalcocite, Cu₂S; bornite, Cu₅FeS₄; chalcopyrite, CuFeS₂; covellite, CuS). The solid diamond indicates the fO₂–fS₂ composition of a hypothetical basin brine. At physical conditions in the shaded area, at least 35 ppm Cu and SO₄² can be transported by fluids saturated with respect to hematite and chalcocite.

(Figure 126) Schematic illustration of the approach adopted when modelling red-bed formation waters during diagenesis and calculating the composition of a late burial diagenetic red-bed pore water.

(Figure 127) Schematic illustration of the approach adopted when modelling reducing organic-rich oilfield waters, such as those which might enter a red-bed aquifer.

(Figure 128) Illustration of the output from an EQ6 model of the mixing of evaporite-equilibrated late burial diagenetic pore water with reducing organic-rich oilfield water. a. there is a dramatic change in redox state when only c.5% of the mixture is reducing, organic-rich oilfield water; b. at the same time as this change in redox state, Cu is stripped from the solution, and then Pb, followed by Zn begin to be stripped from solution as progressively larger amounts of reducing oilfield water are added to the mixture; c. the model predicts a gangue-mineral assemblage which is dominated by dolomite and baryte, with trace quantities of quartz, muscovite, hematite and albite. The moles of minerals precipitating are those produced by a notional 1 kg of solution.

(Figure 129) A comparison between the fields of aqueous Cu- and Ag-chloride mobility in the absence of sorption (a), and calculated sorption edges for Pb, Cu, Zn, Co and Ag (b and c).

(Figure 130) Locations of observation boreholes in the Cheshire Basin region, used by the National Rivers Authority (NRA; now the Environment Agency). The larger symbols represent those boreholes used by the Northwest Region of the NRA (now Environment Agency), while the smaller symbols are boreholes used by the Severn Trent Region (now Midlands region of the Environment Agency). The approximate area underlain by Mercia Mudstone Group is shown, and it can be seen that very few of the observation boreholes are located upon these beds.

(Figure 131) A trilinear plot or 'Piper' diagram summarising the major component chemistries of groundwaters from 11 boreholes in the region of the Cheshire Basin.

(Figure 132) A plot of molar Cl concentrations against molar Br concentrations for selected groundwaters from the Cheshire Basin. The Cheshire Basin water analyses for selected shallow observation boreholes were obtained from the NRA; and analyses of waters produced during brine extraction were obtained from Haslam et al. (1950). For comparison, analyses of groundwaters from PermoTriassic aquifers from elsewhere in the United Kingdom are also shown. The data for these were taken from Allen and Holloway (1984); Burgess et al. (1980); Burgess (1981); Edmunds et al. (1977); Burgess et al. (1981); Walton (1982); Smith and Burgess (1984). CMC, Coal Measures from a borehole near Chester; HW, Halewood; MT, Mickle Trafford; RH, Rainhill; STW, Stanlow.

(Figure 133) Occurrences of polymetallic mineralisation in the Cheshire Basin. A, Alderley Edge; Be, Bearstone; Bi, Bickerton; C, Clive; E, Eardiston; H, Hawkstone; M, Mottram St Andrew; P, Pim Hill; W, Wixhill; Y, Yorton.

(Figure 134) The geology of the Alderley Edge area (from Warrington, 1980). Inset: schematic section, on line A—B, showing potential structural traps for hydrocarbons and other fluids beneath a cap rock of MMG sediments in the Alderley Horst (X) and an adjacent faulted block (XI); PSL, present surface level (all features in the section above that level have been removed by erosion and are projected from the geological map).

(Figure 136) are identified by the appropriate letter." data-name="images/P1000389.jpg">(Figure 135) Distribution of mine workings at Alderley Edge (from Warrington, 1980) . Sections shown in (Figure 136) are identified by the appropriate letter.

(Figure 136) Semi-diagrammatic sections of the mineralisation at Alderley Edge and Clive, based (a, b) on Warrington (1965), (c, d, e) on unpublished surveys and records held by G Warrington and (f) on field observations made by courtesy of the Shropshire Caving and Mining Club. For locations of sections a—e, see (Figure 136) are identified by the appropriate letter." data-name="images/P1000389.jpg">(Figure 135). (a) Engine Vein Mine; (b) Wood Mine; (c, d, e) West Mine;(f) Clive Mine. BLS Beacon Lodge Sandstones; EVC Engine Vein Conglomerate; GSF Grinshill Sandstone Formation;  NAS Nether Alderley Sandstones; TSF Tarporley Siltstone Formation; WMC Wood Mine Conglomerate; WMS West Mine Sandstones; WSF Wilmslow Sandstone; Formation; C Mudstone bed

(Figure 137) The geology of mining sites in the southern part of the Cheshire Basin (from British Geological Survey 1:63 360 geological sheet 138, Wem). 1, Pim Hill; 2, Yorton; 3, Clive; 4–6, Wixhill localities; 7–11, Hawkstone localities.

(Figure 138) General deposit-scale model for sediment-hosted stratiform Cu deposits. The upper, dashed line indicates the furthest advance of Cu into the basal grey bed unit (from Brown, 1993).

(Figure 139) Hydrocarbon shows and seeps.

(Figure 140) Subcrop at Base Permo-Triassic unconformity.

(Figure 141) Depth-maturity plot of Keele 1 borehole.

(Figure 142) Comparison of maturity gradients, Keele 1 and Knutsford.

(Figure 143) Relationship of oil shows to the structure of Westphalian rocks in the Madeley Coalfield.

(Figure 144) Computed average porosity of the lower part of the SSG, based on porosity loss with burial derived from density and sonic logs.

(Figure 145) Marsdenian isopachs, including speculative values beneath the Cheshire Basin (after Collinson et al., 1977).

(Figure 146) Main prospective areas for hydrocarbons in the Cheshire Basin.

(Figure 147) Aeromagnetic map of the Cheshire Basin and surrounding areas.

(Figure 148) Design model of the Cheshire Basin Expert-GIS.There are three main components: the Geographical Information System (ARC/INFO), the relational database (ORACLE) and the expert system shell (NEXPERT OBJECT). Note that the graphical interface that the user sees is controlled by ARC/INFO. The system is hosted on a Sun workstation runnning under SunOS.

(Figure 149) Screen display of the development system for NEXPERT OBJECT. In the top right-hand corner is an overview of the object-oriented structure of a knowledge base with a magnified version of the properties and their values (e.g. depth=209.08) for one object. In the top left is the rule 'checklow' in the knowledge base 'sepfin21.kb'. If the statements in the If box are true then the actions in the 'Then Do' box are carried out. The syntax, although appearing obscure, is not really difficult and there is a facility to explain the rule's meaning linguistically in readily understandable form (not shown here). At the bottom left is an overview of the rule structure for the knowledge base, part of which is displayed, magnified, at bottom right, showing relationships between rules and status indicators. Question marks indicate untested statements; they change to ticks once the rules are fired.

(Figure 150) The user interface of the Cheshire Basin Expert-GIS, displayed from the ARC/INFO GIS. The map is an overview of the geology of the basin. The top menu is the starting point for selecting which module to use. The bottom menu is one of a hierarchy that permits the user to select the 3-D rock mass to be investigated. The user can choose, using slider widgets, the upper and lower bounding formations to the stratigraphy of interest.

(Figure 151) Output from the diagenesis knowledge base showing a global paragenesis determined for the Kinnerton Sandstone Formation from samples in the northern half of the Cheshire Basin. Time is represented along the horizontal axis, and each arrowed line gives the relative range of a given mineral/textural event.

(Figure 152) Output from the diagenesis knowledge base showing the distribution of eo-, meso- and telodiagenetic characteristics of the SSG, summarised by borehole. The polygons are Thiessen polygons based on each borehole location (a Thiessen polygon contains all locations that are closer to the enclosed data point than to neighbouring data points). Where neighbouring polygons are of the same diagenetic type the polygons combine to form more complex shapes. Black areas represent polygons where no relevant data exists (e.g. the SSG is not present in the borehole).

(Figure 153) Output from the diagenesis knowledge base showing the sampled types of sulphide and baryte mineralisation. The shapes are Thiessen polygons as in (Figure 152). The more northerly blue triangle represents the Alderley Edge mines and the more westerly blue polygon the Gallantly Bank mineralisation.

(Figure 154) Output from the provenance knowledge base showing the possibility that the source of detritus in the Chester Pebble Beds comes from an Armorican source. The shapes are Thiessen polygons, as in (Figure 152), based on borehole location.

(Figure 155) Output from the mineralisation module. The surface is the present day structure of the Cheshire Basin at the level of the base of the MMG, viewed from the south-west. Superimposed on this are the zones (in red) of high prospectivity for sulphide mineralisation. See text for discussion.

Plates

(Plate 1) Desiccation crack in interlaminated dolomitic siltstones of the Bollin Mudstone Formation. Wilkesley Borehole, 1360.6 m depth.

(Plate 2) Pseudomorph in siltstone after halite. Loose block in road cutting, Manchester Airport.

(Plate 3) Haselgebirge: crystals of halite, which have grown in cross-laminated siltstone of the Northwich Halite Formation. Wardle Borehole, 217 m depth. From Wilson (1993)

(Plate 4) Gypsum nodules in reddish brown structureless mudstone of the Wych Mudstone Formation. Wilkesley Borehole, 781.50 m depth.

(Plate 5) Gypsum nodules in structureless brown mudstones of the Brooks Mill Mudstone Formation. Wilkesley Borehole, 222.50 m depth.

(Plate 6) BSEM photomicrograph of typical well-rounded detrital altered volcanic rock fragment composed of Kfeldpar laths in an aphanitic matrix of sericite (mid-grey) and quartz (dark grey). Kinnerton Sandstone Formation, Halewood Borehole (sample CHB 2).

(Plate 8) BSEM photomicrograph showing concentration of detrital heavy-mineral grains (bright) at the base of siltstone lamina. Tarporley Siltstone Formation, Saughall Massie Borehole (sample CHB 283).

(Plate 7) BSEM photomicrograph showing typical interlaminated mudstone (M) and siltstone (S) in repeated well-defined fining-upwards cycles. Tarporley Siltstone Formation, Mobberley Town Borehole BH5 (sample CHB 596).

crite or dolomicritic silt, interlayered with, and grading upwards into, 0.1–3 mm of anhydritic dolomicrite and sometimes finally anhydritic mudstone ((Plate 9)) or a thin layer of virtually pure aphanitic felted-lath anhydrite ((Plate 10)). Lenticles of felted-lath anhydrite (up to 10 mm long and resembling starved ripples) occur in low-angle ripple and cross-laminated dolomicrite and dolomicritic

(Plate 9) BSEM photomicrograph showing fine-laminated mudstones with small cycles grading from silty dolomicritic bases (dull) to finely laminated aphanitic anhydritic mudstone tops (bright). Locally patchy micronodular disruptive anhydrite horizons may develop (n). Bollin Mudstone Formation, Wilkesley Borehole (sample CHB 744).

(Plate 10) BSEM photomicrograph showing top of mudstone cycle: fine felted mass of anhydrite needles (bright grains) sedimented in an orientation parallel with disseminated micritic rhombs of non-ferroan dolomite (dark grains). Bollin Mudstone Formation, Wilkesley Borehole (sample CHB 744).

(Plate 12) BSEM photomicrograph showing fine gradational laminae of anhydritic dolomicrite (bright) and dolomicritic siltstone (dull). Byley Mudstone Formation, Crewe Heat Flow Borehole (sample CHB 242).

(Plate 11) BSEM photomicrograph showing localised development of patchy micronodular anhydrite (mid-grey) which disrupts the primary sedimentary dolomicrite (dark) fabric. Later diagenetic halite (bright) forms thin subhorizontal veinlets in the anhydritic-dolomicritic siltstone and mudstone. Bollin Mudstone Formation, Wilkesley Borehole (sample CHB 744).

(Plate 13) BSEM photomicrograph showing very fine-scale cross-bedding foresets in finely laminated anhydritic dolomicrite (bright) and dolomicritic siltstone (dark). The fabric has been disrupted by synsedimentary flow due to dewatering. Later irregular anhydrite nodules (bright) have developed within the disrupted fabric. Byley Mudstone Formation, Crewe Heat Flow Borehole (sample CHB 241).

(Plate 14) BSEM photomicrograph showing fine-grained DE1 b platey hematite (bright) precipitated between cleavage of partially exfoliated altered detrital biotite. Kinnerton Sandstone Formation, Mickle Trafford Borehole (sample CHB 27).

(Plate 16) BSEM photomicrograph of concentric-structured eodiagenetic DE1b dolomite nodule (d) showing late telodiagenetic dissolution.Note euhedral mouldic (secondary) dissolution porosity (p) where euhedral DE3 dolomite overgrowths have dissolved. The nodule is enclosed in mesodiagenetic DE6 ferromanganoan calcite which shows no evidence of dissolution. Chester Pebble Beds Formation, Mickle Trafford Borehole (sample CH 22).

(Plate 15) Photomicrograph (plane-polarised transmitted light) of very porous slightly compacted sandstone with DE 1 b eodiagenetic non-ferroan dolomite micronodules (d), showing slight dissolution following radial-concentric bands of inclusions of DEI a eodiagenetic iron oxide. Chester Pebble Beds Formation, Mickle Trafford Borehole (sample CHB 22). Width of field 4 mm.

(Plate 17) BSEM photomicrograph of early DE2 authigenic illitic clay grain coatings (i) and replacement of exfoliated and expanded altering detrital biotite (b). Note spalling of clay coating from detrital grain surfaces as a result of compaction. Chester Pebble Beds, Clotton Borehole (sample CHB 572).

(Plate 18) BSEM photomicrograph of recrystallised eodiagenetic DE1b non-ferroan dolomite micronodule (dl) with syntaxial rhombic euhedral overgrowths of DE3 nonferroan dolomite (d2), in turn overgrown by syntaxial ferroan dolomite (d3), which is further syntaxially overgrown by late ankerite (d4). Late poikilotopic pore-filling ferroan calcite (c) encloses both the dolomite-ankerite cement and euhedral quartz overgrowths (q). Detrital biotite partly replaced by hematite (b) has been deformed by compaction-impaction of euhedral ferroan dolomite-ankerite rhomb. Helsby Sandstone Formation, Hondslough Farm Borehole (sample CHB 650).

(Plate 19) Details of (Plate 18) showing detail of dolomiteferroan dolomite—ankerite zoning and relationship to calcite.

(Plate 20) BSEM photomicrograph of moderately compacted subarkose showing complete cementation of intergranular porosity by DE4 quartz cement. Quartz, dark grey; K-feldspar, white. Sherwood Sandstone Group, Prees Borehole (sample CHB 524).

(Plate 21) BSEM photomicrograph of well-developed euhedral DE4 quartz overgrowth cement (q) cut by compactional stress-cracks which are infilled with later DE6 weakly ferromanganoan calcite. Helsby Sandstone Formation, Hondslough Farm Borehole (sample CHB 650).

(Plate 22) BSEM photomicrograph of microgranular fault gouge cemented by chalcedonic silica. Chester Pebble Beds Formation, Mickle Trafford Borehole (sample CHB 22).

(Plate 23) BSEM photomicrograph of silicified fault gouge showing detrital quartz and feldspar grains in fine matrix cross-cut by later veinlet of DE6 baryte. Bridgnorth Sandstone Formation, Rock Hall road cutting, Shropshire (sample CHB 132).

(Plate 24) BSEM photomicrograph showing framboidal or globular inclusions of Ag₂Se (?naumannite) enclosed in microcolloform chalcedony. From silicified fault rock, Grinshill Sandstone, Clive Copper Mine (sample CHB 521).

(Plate 25) BSEM photomicrograph showing tiny inclusions (x) of cobaltite, galena, and Co-As-Ni sulphide trapped beneath quartz overgrowths (q) in fault-rock sandstone mineralised by later galena (g) which is partly altered to secondary mimetite. Helsby Sandstone Formation (West Mine Sandstone), West Mine Alderley Edge (sample CHB 719).

(Plate 26) BSEM photomicrograph showing octahedral crystal with complex intergrowth of galena (bright) and Co-(Ni)-As sulphide (probably siegenite or cobaltite gersdoffite). Shows core of galena with overgrowth of Co-Ni-As sulphide, in turn partly replaced by galena. Occurs as an inclusion within quartz overgrowth seen in (Plate 25).

(Plate 28) BSEM photomicrograph of clean aeolian sandstone with well-developed interlocking euhedral quartz overgrowths, cemented by secondary mimetite. West Mine Sandstone (Helsby Sandstone Formation) taken immediately adjacent to the White Barn Fault (Chain Shaft), West Mine, Alderley Edge (sample CHB 710).

(Plate 27) BSEM photomicrograph of sulphide inclusions in authigenic quartz. Shows quartz (dark) with mid-grey cobaltite, overgrown by galena (bright) towards surface of quartz crystal (cemented by altered galena). Helsby Sandstone Formation (West Mine Sandstone), West Mine Alderley Edge (sample CHB 720).

(Plate 29) BSEM photomicrograph of clean aeolian sandstone with well-rounded quartz (dark grey) and minor K-feldspar (mid-grey) showing minor compacted grain boundaries, and no evidence of quartz overgrowth development. Sandstone grains coated by fibrous and gel-like secondary chrysocolla and cemented by later secondary malachite (bright). West Mine Sandstone (Helsby Sandstone formation) taken 1–2 m up-dip from fault in middle level, West Mine, Alderley Edge (sample CHB 719).

(Plate 30) BSEM photomicrograph showing relationships between DE6b dolomite (d), DE6d pyrite (p) and DE6e ferroan calcite (c) in calcite vein containing scattered wall-rock detritus (quartz grains, q). Note that the dolomite has a non-ferroan dolomite core (dark) which has been corroded and overgrown by ferroan dolomite and ankerite (progressively lighter). This carbonate is also corroded by pyrite. Helsby Sandstone Formation, Thornton Borehole (sample CHB 47).

(Plate 32) BSEM photomicrograph showing DE6c anhydrite veinlet cutting porous sandstone, but showing little or no penetration into adjacent open porosity. This indicates that the porosity is secondary, resulting from postDE6c dissolution of an earlier (?DE5 halite) cement. Malpas Sandstone, Wilkesley Borehole (sample CHB 534).

(Plate 31) BSEM photomicrograph showing detail of fine zoning and corrosion surfaces within euhedral DE6b crystal of dolomite, ferroan dolomite and ankerite in DE6e calcite. Vein fill in Helsby Sandstone Formation, Thornton Borehole (sample CHB 47).

(Plate 33) BSEM photomicrograph showing edge of DE6 fault-vein containing pyrite and ferroan calcite and cutting porous sandstone. DE6d pyrite (p) lines the vein wall and shows the only limited penetration into and replacement of adjacent wall rock, with localised patchy development of pyrite cement. The later DE6e calcite (c) encloses the pyrite lining the wall and locally cross-cuts (x) the earlier pyrite lining, but similarly shows little penetration into the wall-rock. The small cross-cutting DE6e calcite veinlet (x) terminates against porous wall-rock. Helsby Sandstone Formation, Thornton Borehole (sample CHB 46).

(Plate 34) BSEM photomicrograph showing fine-grained early diagenetic pyrite (grey) with later chalcopyrite (bright) nucleated around the pyrite. Tarporley Siltstone Formation, Saughall Massie Borehole (sample CHB 307).

(Plate 36) BSEM photomicrograph of early mesodiagenetic porphyroblastic quartz crystals (dark) partially replacing anhydrite (bright) but preserving the original anhydrite fabric as corroded and partially replaced relict inclusions. The anhydrite host enclosing the quartz and palimpsest anhydrite fabrics has completely recrystallised. Anhydrite bed in Wych Mudstone Formation, Crewe Heat Flow Borehole (sample CHB 22).

(Plate 35) BSEM photomicrograph of cubic dissolution cavity after halite infilled by early mesodiagenetic dolomite (d) and later mesodiagenetic anhydrite cement (a). Byley Mudstone Formation, Crewe Heat Flow Borehole (sample CHB 241).

(Plate 37) BSEM photomicrograph of euhedral porphyroblastic celestite crystals (bright) associated with late gypsum-replaced anhydrite vein (grey) cutting recrystallised dolomicritic siltstone. Mercia Mudstone Group, Winsford W6 borehole (sample CHB 780).

Tables

(Table 1) Locations of boreholes sampled for geochemical, mineralogical and petrographical studies.

(Table 2) Locations of mines, quarries and outcrops sampled for geochemical, mineralogical and petrographical studies.

(Table 3) Boreholes selected for sedimentological study.

(Table 4) Boreholes with density logs used in depth-of-burial study.

(Table 5) Boreholes with sonic logs used in depth-of-burial study.

(Table 6) HOTPOT layer information for Cheshire Basin models.

(Table 7) Fault-heave analysis: observed displacement at Base Permo-Triassi.

(Table 8) Throw-frequency distribution of faults cutting Base Permo-Triassic.

(Table 9) Fault-heave analysis: observed and corrected displacements and extension factors at Base Permo-Triassic.

(Table 10) Fault-heave analysis: relative observed displacement at multiple horizons.

(Table 11) Fault-heave analysis: cumulative relative displacements, observed and corrected for compaction.

(Table 12) Provenance-sensitive indices used in the heavy-mineral study.

(Table 13) Summary of data for provenance-sensitive heavy-mineral indices.

(Table 14) Variations in detrital tourmaline composition.

(Table 15) Clasts sampled for isotopic study from the Chester Pebble Beds Formation.

(Table 16) Pebbles sampled from the Kidderminster Formation of the Knowle and Worcester basins.

(Table 17) Analytical results of Sm-Nd isotopic study.

(Table 18) Factor analysis results for the Sherwood Sandstone Group.

(Table 19) Factor analysis results for the Mercia Mudstone Group.

(Table 20) Samples selected for fluid inclusion examination.

(Table 21) Summary of the principal characteristics of each Diagenetic Episode recognised in the SSG in the Cheshire Basin.

(Table 22) Stable carbon and oxygen isotope data for diagenetic carbonates.

(Table 23) Summary of the principal characteristics of each Diagenetic Episode recognised in the MMG in the Cheshire Basin 154.

(Table 24) Stable sulphur isotope data for primary, diagenetic, and vein sulphate and sulphide minerals in Permo-Triassic rocks from the Cheshire Basin.

(Table 25) Samples selected for fluid inclusion analysis.

(Table 26) Thermometric analyses for inclusions in baryte and calcite.

(Table 27) Eutectic temperatures and minimum salinity estimates for various salt-water systems.

(Table 28) Laser ablation ICP-MS analyses for inclusions in halite.

(Table 29) Rocks (containing hydrocarbon) and oil (bitumen) samples analysed.

(Table 30) Hydraulic parameters used for the Cheshire Basin flow model.

(Table 31) Temperature, electrical conductivity (EC), porosity, and calculated total dissolved solids (TDS) of formation waters 183.

(Table 32) Summary of the example EQ model of the mineralogical evolution of red beds during eodiagenesis and mesodiagenesis.

(Table 33) Summary of the initial water compositions and product fluid compositions for example EQ3/6 model of mixing between late mesodiagenetic red-bed pore water and reducing oilfield water to the point where all Cu is stripped from the solution.

(Table 34) Mean analyses of modern groundwaters at specified horizons in selected boreholes in the Cheshire Basin 202.

(Table 35) Stratigraphical succession at Alderley Edge.

(Table 36) Principal characteristics of representative sediment-hosted stratiform Cu deposits.

(Table 37) Median values of ore-forming elements for the SSG and MMG compared with published averages.

(Table 38) Estimated total quantities of Cu, Pb and Zn in the SSG and MMG.

(Table 39) Metallogenic model for Cheshire Basin mineralisation.

(Table 40) Westphalian C-D formations of approximate equivalence in coalfields around the Cheshire Basin.

(Table 41) Dry wells around the Cheshire Basin.

Maps

(Map 1) Enclosure 1 Geological map of the Cheshire Basin at 1 :250 000 In pocket

Equations

(Equation 1) Average basin subsidence equation

(Equation 2) Equation. For 1-D sampling, the hidden cumulative throw for small faults can be determined by extrapolating the fractal size distribution (Marrett and Allmendinger, 1992)

Tables

(Table 1) Locations of boreholes sampled for geochemical, mineralogical and petrographical studies.

(Table 2) Locations of mines, quarries and outcrops sampled for geochemical, mineralogical and petrographicalstudies. The samples of Kidderminster Formation come from the Worcester and Knowle basins

(Table 3) Boreholes selected for sedimentological study

(Table 4) Boreholes with density logs used in depth-of-burial study

(Table 5) Boreholes with sonic logs used in depth-of-burial study

(Table 6) HOT POT layer informarion for Cheshire Basin models

(Table 7) Fault-heave analysis: observed displacements at Base Permo-Triassic

(Table 8) Throw-frequency distribution of faults cutting Base Permo-Triassic

(Table 9) Fault-heave analysis: observed and corrected displacements and extension factors at Base Permo-Triassic

(Table 10) Fault-heave analysis: relative observed displacements at multiple horizons

(Table 11) Fault-heave analysis: cumulative relative displacements, observed and corrected for compaction

(Table 12) Provenance-sensitive indices used in the heavy-mineral study (from Morton and Hallsworth, 1994)

(Table 13) Summary of data for provenance-sensitive heavy-mineral indices

(Table 14) Variations in detrital tourmaline composition

(Table 15) Clasts sampled for isotopic study from the Chester Pebble Beds Formation, Dunham-on-the-Hill (see (Table 2))

(Table 16) Pebbles sampled from the Kidderminster Formation of the Knowle and Worcester basins (see ((Table 2)))

(Table 17) Analytical results of Sm-Nd isotopic study

(Table 18) Factor analysis results for the Sherwood Sandstone Group