Reservoir Quality of Upper Jurassic Corallian Sandstones, Weald Basin, UK

: The Upper Jurassic, shallow marine Corallian sandstones of the Weald Basin, UK, are signiﬁcant onshore reservoirs due to their future potential for carbon capture and storage (CCS) and hydrogen storage. These reservoir rocks, buried to no deeper than 1700 m before uplift to 850 to 900 m at the present time, also provide an opportunity to study the pivotal role of shallow marine sandstone eodiagenesis. With little evidence of compaction, these rocks show low to moderate porosity for their relatively shallow burial depths. Their porosity ranges from 0.8 to 30% with an average of 12.6% and permeability range from 0.01 to 887 mD with an average of 31 mD. The Corallian sandstones of the Weald Basin are relatively poorly studied; consequently, there is a paucity of data on their reservoir quality which limits any ability to predict porosity and permeability away from wells. This study presents a potential ﬁrst in the examination of diagenetic controls of reservoir quality of the Corallian sandstones, of the Weald Basin’s Palmers Wood and Bletchingley oil ﬁelds, using a combination of core analysis, sedimentary core logs, petrography, wireline analysis, SEM-EDS analysis and geochemical analysis to understand the extent of diagenetic evolution of the sandstones and its effects on reservoir quality. The analyses show a dominant quartz arenite lithology with minor feldspars, bioclasts, Fe-ooids and extra-basinal lithic grains. We conclude that little compactional porosity-loss occurred with cementation being the main process that caused porosity-loss. Early calcite cement, from neomorphism of contemporaneously deposited bioclasts, represents the majority of the early cement, which subsequently prevented mechanical compaction. Calcite cement is also interpreted to have formed during burial from decarboxylation-derived CO 2 during source rock maturation. Other cements include the Fe-clay berthierine, apatite, pyrite, dolomite, siderite, quartz, illite and kaolinite. Reservoir quality in the Corallian sandstones show no signiﬁcant depositional textural controls; it was reduced by dominant calcite cementation, locally preserved by berthierine grain coats that inhibited quartz cement and enhanced by detrital grain dissolution as well as cement dissolution. Reservoir quality in the Corallian sandstones can therefore be predicted by considering abundance of calcite cement from bioclasts, organically derived CO 2 and Fe-clay coats. Current ripple lamination, wave ripple lamination, sand-mud heterolithics beds. Cementation variable.


Introduction
Diagenesis involves physical, chemical and biological processes which act on texturally and compositionally unstable sediments, causing them to reach mineralogical and textural maturity and converting sediment into rock [1]. Diagenetic processes alter sedimentary rock's porosity and permeability (reservoir quality) as burial progresses [2]. Understanding the controls and processes involved in diagenetic alteration is therefore useful in predicting reservoir quality evolution [3].
Under some circumstances, differences in reservoir properties, such as porosity and permeability, can be linked to variations within their depositional systems [1]. The specific characteristics of primary depositional systems are paramount for reservoir quality  [7]; Butler and Pullan [19].
During the middle part of the Mesozoic, thermal subsidence, associated with block faulting, created accommodation followed by the Lower Jurassic Rhaetic transgression and the subsequent deposition of the Lower Jurassic White Lias, Lower Lias, Middle Lias and Upper Lias units ( Figure 2) [7,15,21]. A broad, shallow, carbonate bank developed due to tectonic uplift at the end of the early Jurassic on which Middle Jurassic Bajocian Inferior Oolite and Bathonian Great Oolite were deposited ( Figure 2) [7,[20][21][22]. Continued rifting, thermal subsidence and transgression from the Middle Jurassic (Callovian) to Lower Cretaceous caused the deposition of marine mudstones and limestones of the Oxford Clay Formation, Corallian Group, Kimmeridge Clay Formation and Portland Group [14,15]. The deposition of these marine mudstones and limestones was interrupted by the  [7]; Butler and Pullan [19].

Figure 2.
Generalised stratigraphic section for the Jurassic of the Weald Basin as adapted from [7]. Showing lithologic units as well as selected hydrocarbon fields and their reservoir sections. The Corallian sandstone Formation is highlighted in green.
Cenozoic fault reactivation and basin inversion occurred in response to compressive forces from the south during the opening of the North Atlantic in the Paleogene, with a  [7]. Showing lithologic units as well as selected hydrocarbon fields and their reservoir sections. The Corallian sandstone Formation is highlighted in green. Cenozoic fault reactivation and basin inversion occurred in response to compressive forces from the south during the opening of the North Atlantic in the Paleogene, with a second phase in the Miocene possibly associated with the Alpine tectonism [24]. The resulting inversion caused a regional uplift, reported to be as much as 2134 m (7000 ft) [13]. Basin inversion led to the formation of broad, dome-shaped, hanging-wall anticlines (with the London Basin to the north and the Hampshire-Dieppe Basin to the south), subsiding to form flanking basins in the late Paleogene-Eocene [14]. The Cenozoic inversion led to erosion and removal of sediments younger than the Valanginian Wealden Group [15].

Trap Formation and Hydrocarbon Occurrence
The Upper Jurassic to Lower Cretaceous rifting caused the development of extensional trapping geometries, trending east-west along low angle faults, which originated from earlier Variscan thrusts [9,14].
Organic-rich candidate source rocks in the Weald Basin include the Liassic shales [14,19,25], the Oxford Clay Formation and the Kimmeridge Clay Formation [7,19,25,26]. It has been suggested that the Lias and Oxford Clay reached the gas generation window in the centre of the Weald Basin while the Kimmeridge Clay may have only reached the oil generation window [26].
Others have suggested that the Lower Jurassic source rocks reached maturity and expelled hydrocarbons in the early Cretaceous [14,19,25]. Expulsion of hydrocarbons terminated during Cenozoic fault reactivation, inversion and cooling of source rocks [9,14].

Bletchingley Field
The relatively shallow Bletchingley field was discovered in 1963, near the village of Bletchingley, when seismic surveys showed two faulted structures within existing mining leases [6]. The seismic survey resulted in the drilling of four wells between December 1965 and 1966; wells BL1 and BL2 on the eastern flank both tested gas from the Upper Jurassic Corallian limestone with a maximum flow rate of 9.65 MMSCF/day [6]. Gas in place was estimated to be between 1 to 6 BCF [6]. Wells BL5 and BL6 were drilled in 2008 and the Corallian sandstone, the subject of this paper, reportedly flowed at about 250 b/d in BL5 at a true vertical depth of about 850 m. Well BL5 was drilled at a low angle so that measured depths (m MD) are significantly greater than true vertical depths (m TVD). In this study, logs and core sample from BL5 have depths reported in terms of the measured depths (m MD).

Palmers Wood Field
Similar to Bletchingley, the relatively shallow Palmers Wood field is also a seismicallydefined prospect located on the northern flank of the Weald Basin [6]. In 1983, the first well was drilled at Rook's Nest farm and tested oil at unstable rates of 540 b/d from a 9.1 m (30 ft) Corallian sandstone section [6], at a true vertical depth of about 900 m. Further testing, in new appraisal wells, from 1984 to the 1990s tested oil [6]. Average porosity is about 17% with mean permeability ranging from 5 to 7 mD, explaining the low production rates. The stock tank oil initially in place (STOIIP) was initially estimated to be 11.73 MMSTB [6].

Sedimentary Core Logging
The sedimentary cores were originally measured in feet, as all well depths (logs and core) were recorded in feet, but the depths have here been converted to metres. High resolution sedimentary core logging of three wells: Bletchingley 5 (BL5), Palmers Wood 3 (PW3) and Palmers Wood 7 (PW7), was carried out at the British Geological Survey (BGS) core store in Keyworth, Nottinghamshire. The wells were logged on a scale of 1:24, recording grain size, lithology, sedimentary structures, bioclasts, ichnofabrics, cementation, Geosciences 2021, 11, 446 6 of 38 fractures, hydrocarbon stains and well deviation. About 17 m of core was available from PW3, 17 m of core was available from PW7 and nearly 20 m of core was available from BL5.
Ichnofabrics and depositional environments were determined using methods outlined by Hampson and Storms [27,28]. Sedimentary logs have been summarised and digitised to a scale of 1:240 and related to wireline and routine core analysis data.

Wireline Analysis
Wireline logs used in this study include neutron logs, density logs and gamma ray logs. Gamma ray logs were used to determine shale proportions (Vshale) using the formula: (1) where GR meas is the measured gamma ray value for the depth of interest, GR max is the maximum gamma ray log value for the rock section of interest and the minimum gamma ray value for the rock section of interest is denoted by GR min . Wireline neutron-density cross-over plots were used to define pay and non-pay zones as is conventional in petrophysical analysis [29]. Core-to-log depth shifts were made to match density log porosity to core analysis porosity and Vshale to mudstones identified during core description.

Petrographic Analysis
Petrographic analysis was carried out to determine mineralogy, modal composition, textural relationships, pore space morphology, and types of cements present, as well as the relative timing of diagenetic events. In this case, 51 polished sections were prepared from plugs used for core analysis from PW7, PW3 and BL5, in the standard manner for water-sensitive plugs. Polished thin sections were used for both optical microscopic examination and SEM analysis. The polished sections were all impregnated with blue-dyed epoxy resin to aid in the characterisation of pores. Light microscopy was undertaken using an Olympus BX51 microscope. In this case, 48 samples were point counted to determine modal composition using an Olympus BX53M microscope fitted with a Conwy Valley systems Petrog stage and software. For point counting, appropriate grid spacing was chosen to enable representative sampling of 300 compositional points with a X10 magnification objective. Attention was paid to cement types, cement stratigraphy and replacive cements to help reveal the order of diagenetic events (i.e., paragenetic sequence). A further 250 points were analysed with a X5 objective to measure grain size and sorting.
Backscattered Scanning Electron Microscopy (BSEM) was carried out using a Hitachi TM3000 table-top SEM. This was undertaken to ascertain mineralogy, mineral/cement growth relationships and textural relationships.
Automated mineralogy data were acquired using SEM-EDS technology developed by FEI [30]; SEM-EDS is a scanning electron microscope (SEM) equipped with high-speed Energy Dispersive X-Ray Spectroscopy detectors (EDS) [31]. SEM-EDS identifies minerals using Species Identification Protocols (SIPs) which represents an extensive mineral chemical database archived in a library [32,33] to produce a quantitative evaluation of mineral proportions. SEM-EDS analysis produces quantitative mineral proportions, grain and pore space morphology and distribution to a minimum resolution of about 1 to 2 µm (i.e., equal to the diameter of the smallest beam-sample interaction volume). SEM-EDS analyses were undertaken with a user-defined resolution depending on the level of detail required and analysis time. In this study, SEM-EDS analyses were undertaken at the University of Liverpool, using a tungsten-filament, operating at 15 kV, equipped with two Bruker EDS detectors [34]. A minimum spacing of 2 µm was used for higher resolution and 20 µm spacing resolution was used where average mineralogy was required across a whole polished section.
X-ray diffraction (XRD) analyses were undertaken at the University of Liverpool's Diagenesis Research Group lab. The analysis was carried out using randomly oriented powders with a PANalytical X'Pert Pro MPD X-ray diffractometer (Malvern PANalytical, Malvern, UK). Minerals were quantified by the relative intensity ratio (RIR) [35] method using the PANalytical HighScore Plus software. XRD analysis was carried out to determine bulk rock mineralogy with specific interest on identifying berthierine. Presence of berthierine was determined by heating samples to temperature from ambient to 300 • C, 400 • C and 550 • C.

Routine Core Analysis
Porosity and permeability were measured by industry core laboratories, from one-anda-half-inch diameter core plugs, cleaned in hot refluxing solvents to remove hydrocarbons and residual brines. Helium porosity was determined using a Boyle's Law porosimeter. The porosity of six additional core plugs from PW3 and PW7 was measured at the rock deformation laboratory University of Liverpool using a Quantachrome Pycnometer. Permeability was measured using a nitrogen permeameter, with samples mounted in a "Hassler" type core holder. Permeability was measured using a steady state flow of nitrogen gas through core plugs; flow rate, temperature and differential pressure were recorded and used in conjunction with core dimensions to calculate permeability from Darcy's equation.

Burial History Modelling
Burial history modelling was carried out with BasinMod software, courtesy of Platte River Associates. The stratigraphic data are from well completion reports for BL5, where depths for modelling are reported as true vertical depths subsea (TVDss) in metres and formation tops defined from cuttings, measurement while drilling (MWD) gamma ray and Rate of Penetration (ROP) data. The TVDss were converted to true vertical depths (TVD) in metres to account for surface elevation above mean sea level before burial analysis was carried out. The stratigraphic sections eroded during the Aptian to Turonian (Cenozoic) inversion was modelled using the Dartford area north of the Weald Basin as Upper Cretaceous (Chalk) and Cenozoic deposits above the Weald Clay have been eroded, hence its absence in the vicinity of the two fields in question. The ages of the various stratigraphic intervals were derived from the British Geological Survey's (BGS) lexicon of named rock units [36] and the age for erosion inferred by calculating the duration of erosion from the deposit thicknesses and a maximum erosion rate of 14.7 m/Ma. The erosion rate was deduced by dividing the total eroded thicknesses by the number of years since erosion started (55 Ma). Bottom-hole temperatures used here were from BL5 well temperature profile as provided in the well report.

Sedimentary Core Logs
The Corallian Group sandstones are dominated by light grey to brown sandstone, siltstone and minor dark brown to black mudstone ( Figure 3). The results from core logging are summarised in Figure 4 and Table 1. The sandstones are generally well-sorted (Figure 3a,b) with many poorly-sorted argillaceous intervals (Figure 3c,d).
The sandstones have a variety of sedimentary structures which underscore the changing hydrodynamic conditions during deposition ( Figure 4). These sedimentary structures include planar cross-bedding, low angle cross-bedding, trough cross-bedding, low angle lamination, mud drapes, massive bedding, and hummocky cross-bedding, lenticular bedding, flaser bedding, cross lamination, current ripple lamination, wave ripple lamination and climbing current ripple lamination ( Figure 4).   . Shows sedimentary log, Vshale log, neutron (NPHI) and density (RHOB) cross-over plot, core analysis porosity log and core permeability log for wells, (a) PW7, (b) PW3, and (c) BL5. Sedimentary logs show short-range heterogeneity in sedimentary structures and intensity of cementation. They also show variation in hydrocarbon stains with depth. The wireline logs (Vshale, NPHI and RHOB) track the variation in shale content from the sedimentary logs with the core porosity and permeability logs showing the best reservoir quality sections in the sand-rich portions of the logs. Even though the logs show the highest porosity and permeability in the sandier zones. They also have low porosity and permeability in the sandier zones suggesting that lithology is not the dominant control on reservoir quality. The intervals with high calcite cement and clay content show lower porosity and permeability giving an initial indication of possible reservoir quality controls. . Shows sedimentary log, Vshale log, neutron (NPHI) and density (RHOB) cross-over plot, core analysis porosity log and core permeability log for wells, (a) PW7, (b) PW3, and (c) BL5. Sedimentary logs show short-range heterogeneity in sedimentary structures and intensity of cementation. They also show variation in hydrocarbon stains with depth. The wireline logs (Vshale, NPHI and RHOB) track the variation in shale content from the sedimentary logs with the core porosity and permeability logs showing the best reservoir quality sections in the sand-rich portions of the logs. Even though the logs show the highest porosity and permeability in the sandier zones. They also have low porosity and permeability in the sandier zones suggesting that lithology is not the dominant control on reservoir quality. The intervals with high calcite cement and clay content show lower porosity and permeability giving an initial indication of possible reservoir quality controls. The core is locally bioturbated. Bioturbation is most evident in argillaceous intervals (Figure 3d,e) but it also homogenises sandy and clay-rich intervals ( Figure 3d). Ichnofabrics present include Thalasinoides, Teichichnus, Planolites and Chondrites ( Figure 3). Rhizocorallium, Diplocraterion, Rosellia, Palaeophucus, Skolithos and Ophiomorpha.
The core is locally hydrocarbon stained. Sandstones typically react with hydrochloric acid indicating the occurrence of calcite, present as both recognizable bioclasts and in cemented intervals. Calcite cementation is of variable intensity, hydrocarbon staining is most intense where calcite cement seems to be least abundant (Figures 3 and 4) and weak or no hydrocarbons staining occurs in intensely cemented areas (Figure 4a

Wireline Analysis
The Vshale log (derived using Equation (1)) reveals the presence of radioactive clay mineral zones as well as the presence of non-radioactive, sand-dominated reservoir sections in the Corallian sandstones ( Figure 4). The neutron-density cross-over plot [29] typically is assumed to differentiate pay from the non-pay intervals ( Figure 4). There is good agreement between the Vshale logs and the neutron-density cross-over diagram. However, not all the net-sand intervals show the wide separation in neutron-density cross-over typical of pay zones (e.g., BL5, Figure 4c

Routine Core Analysis Porosity and Permeability
Core porosity and permeability seem to show a good relationship with each other (Figures 4a-c and 5a). The whole of the cored section, including sandstones and much finer-grained lithologies from BL5, has a porosity range of 2.7 to 26.6% with a mode of 10% and with values skewed towards the lower porosity range (Figure 5a). Core from BL5 has a permeability range of 0.02 to 887 mD with a mode of 0.08 mD (Figure 5c). Core from PW3 has a porosity range of 0.8 to 23.1% with bimodal porosity with modes at 9.4% and 18.5% (Figure 5b). PW3 has a permeability range of 0.01 to 508 mD and a mode 0.95 mD (Figure 5c). Core from PW7 has a porosity range of 2.6 to 30% with a mode of 10.5% ( Figure 5b) and a permeability range of 0.01 to 401 mD ( Figure 5c). PW7 has bimodal permeability with modes at 0.1 mD and 4 mD (Figure 5c).

Routine Core Analysis Porosity and Permeability
Core porosity and permeability seem to show a good relationship with each other (Figures 4a,b,c and 5a). The whole of the cored section, including sandstones and much finer-grained lithologies from BL5, has a porosity range of 2.7 to 26.6% with a mode of 10% and with values skewed towards the lower porosity range (Figure 5a). Core from BL5 has a permeability range of 0.02 to 887 mD with a mode of 0.08 mD (Figure 5c). Core from PW3 has a porosity range of 0.8 to 23.1% with bimodal porosity with modes at 9.4% and 18.5% (Figure 5b). PW3 has a permeability range of 0.01 to 508 mD and a mode 0.95 mD ( Figure 5c). Core from PW7 has a porosity range of 2.6 to 30% with a mode of 10.5% (Figure 5b) and a permeability range of 0.01 to 401 mD ( Figure 5c). PW7 has bimodal permeability with modes at 0.1 mD and 4 mD (Figure 5c).     Figure 4c 2193.0-2197.0 m MD) suggesting that lithology is not the only control on sandstone porosity and permeability. In addition, permeability varies over two orders of magnitude for a given porosity (Figure 5a) suggesting that permeability is affected by factors other than porosity.

Detrital Composition and Fabric
The Corallian sandstones are composed of moderately-to moderately well-sorted quartz arenites with dominant quartz (Q) grains and minor feldspar (F) and lithic (L) grains. The proportion of quartz to feldspars and lithic grains is in the range of 83.2 to 99.5%, with an average composition of Q 94 F 3 L 3 ( Figure 6). for a given porosity ( Figure 5a) suggesting that permeability is affected by factors other than porosity.

Detrital Composition and Fabric
The Corallian sandstones are composed of moderately-to moderately well-sorted quartz arenites with dominant quartz (Q) grains and minor feldspar (F) and lithic (L) grains. The proportion of quartz to feldspars and lithic grains is in the range of 83.2 to 99.5 %, with an average composition of Q94F3L3 ( Figure 6).   showing extensive grain dissolution with intergranular porosity between grains as well as moldic and microporosity from dissolved grains. The photomicrograph also shows dissolved calcite cement (bottom-left). The feldspar grains have been intensely dissolved and altered to kaolinite with bitumen present within the altered feldspar grains (centre and bottom-right). Bitumen also is also shown lining quartz overgrowths (bottom-left) suggesting that hydrocarbon charge post-dated quartz cementation. Inclusions of bitumen in the micropores as well as at the edge of kaolinite at the top-left suggest some contemporaneous hydrocarbon charge and kaolinite precipitation. Definition of abbreviations used on the images: plagioclase feldspar (pl), k-feldspar (kf), polycrystalline quartz (pqtz), monocrystalline quartz (mqtz), quartzite (qte), iron ooid (Fe-ooid), phosphate ooid (P-ooid) bivalve shell (bi), bitumen (bt), echinoderm (ech), foraminifera (fm), pyrite (py). showing extensive grain dissolution with intergranular porosity between grains as well as moldic and microporosity from dissolved grains. The photomicrograph also shows dissolved calcite cement (bottom-left). The feldspar grains have been intensely dissolved and altered to kaolinite with bitumen present within the altered feldspar grains (centre and bottom-right). Bitumen also is also shown lining quartz overgrowths (bottom-left) suggesting that hydrocarbon charge post-dated quartz cementation. Inclusions of bitumen in the micropores as well as at the edge of kaolinite at the top-left suggest some contemporaneous hydrocarbon charge and kaolinite precipitation. Definition of abbreviations used on the images: plagioclase feldspar (pl), k-feldspar (kf), polycrystalline quartz (pqtz), monocrystalline quartz (mqtz), quartzite (qte), iron ooid (Fe-ooid), phosphate ooid (P-ooid) bivalve shell (bi), bitumen (bt), echinoderm (ech), foraminifera (fm), pyrite (py).     There is 13.3 to 54.6% monocrystalline quartz (average 36.1%) and about 0.7 to 16.4% polycrystalline quartz (average 9.0%) (Table A1. Quartz is texturally mature with subrounded to rounded grains (Figures 7-10). Quartz is most abundant in the proximal upper shoreface facies association but shows no clear relationship from the foreshore to the offshore facies associations (Figure 12k). There is 13.3 to 54.6% monocrystalline quartz (average 36.1%) and about 0.7 to 16.4% polycrystalline quartz (average 9.0%) ( Table A1). Quartz is texturally mature with subrounded to rounded grains (Figures 7-10). Quartz is most abundant in the proximal upper shoreface facies association but shows no clear relationship from the foreshore to the offshore facies associations (Figure 12k).
The feldspars are mainly K-feldspar, which comprises trace to 4.0% of the total rock with an average of 1.0%. Plagioclase feldspar comprises trace to 1.7%, with an average of 0.2% (Table A1). The feldspars occur in various stages of dissolution (Figures 7a, 9a,b and 10a) with dissolved grains resulting in secondary porosity. Both feldspar types increase in proportion from the foreshore to offshore depositional environment (Figure 12a  The feldspars are mainly K-feldspar, which comprises trace to 4.0% of the total rock with an average of 1.0%. Plagioclase feldspar comprises trace to 1.7%, with an average of 0.2% (Table A1). The feldspars occur in various stages of dissolution (Figures 7a, 9a,b and 10a) with dissolved grains resulting in secondary porosity. Both feldspar types increase in proportion from the foreshore to offshore depositional environment (Figure 12a,b) possibly suggesting primary depositional controls on their abundance.
Lithic grains in these sandstones are composed of extrabasinal mudstones, quartzite, sandstone and chert. In the QFL analysis, bioclasts, Fe-ooids and muscovite are considered as lithic grains and have here been added to the lithics ( Figure 6). Bioclasts are dominated by bivalves with minor echinoids, foraminifera and ostracods; bioclasts collectively represent about 0 to 4.0% with an average of 0.8%. Fe-ooids, described later, make up between about 0 to 2.7% with an average of 0.12%. Trace amounts of chlorite are present in the Corallian sandstones ( Figure 13); however, it is not certain if chlorite is authigenic or detrital as chlorite is not evident as a cement or pore filling phase in optical petrographic analyses. Lithic grains in these sandstones are composed of extrabasinal mudstones, quartzite, sandstone and chert. In the QFL analysis, bioclasts, Fe-ooids and muscovite are considered as lithic grains and have here been added to the lithics ( Figure 6). Bioclasts are dominated by bivalves with minor echinoids, foraminifera and ostracods; bioclasts collectively represent about 0 to 4.0% with an average of 0.8%. Fe-ooids, described later, make up between about 0 to 2.7% with an average of 0.12%. Trace amounts of chlorite are present in the Corallian sandstones ( Figure 13); however, it is not certain if chlorite is authigenic or detrital as chlorite is not evident as a cement or pore filling phase in optical petrographic analyses. quartz or heavy minerals (Figure 7c). Optical microscopy and SEM-EDS images reveal the presence of Fe-rich clay; XRD analysis confirmed the presence of berthierine ( Figure 13).
Fe-ooids have a floating grain fabric where they are separated from other grains by pores or cement; cement growth locally truncates their grain boundaries, proving their detrital origin (Figures 7c, 9a and 10c). Fe-ooids increase in quantity from the foreshore to the proximal lower shoreface but there are none in the lower shoreface and offshore (Figure 14f). Bioclasts are dominated by bivalves (Figures 7b, 8a and 11a) and show evidence of intense bioerosion by micritisation (Figure 11a). Bioclast concentration is highest in the upper shoreface facies association and is lowest in the offshore facies association ( Figure  14b); the variable degree of bioclast neomorphism makes interpretation of point count abundance difficult. If all calcite was derived from bioclasts, then total calcite from SEM-EDS analysis may be used as a proxy for total initial quantity of bioclasts; in this case offshore sediment had the lowest quantity of bioclasts (Figure 12l).
The mineralogy of fine-grained sediment was resolved using SEM-EDS data (Figures  8d and 9c). Fine-grained sediments are mostly composed of illite, kaolinite, muscovite, biotite, minor quantities of pyrite, quartz, K-feldspar and plagioclase, as well as trace quantities of smectite (Figures 7d and 9b,c). SEM-EDS data show that biotite (Figure 12f Heavy minerals are dominated by apatite, which makes up about 0 to 1.0% (average 0.1%) of total rock volume. Other heavy minerals include zircon, rutile and iron oxide, which occur in minor quantities (<1%). Apatite occurs as fine grains visible in SEM-EDS (Figure 10b,c) and does not show a relationship with facies associations (Figure 12n). Fe-ooids have brown to olive-green cortices surrounding a nucleus composed of either quartz or heavy minerals (Figure 7c). Optical microscopy and SEM-EDS images reveal the presence of Fe-rich clay; XRD analysis confirmed the presence of berthierine ( Figure 13).
Fe-ooids have a floating grain fabric where they are separated from other grains by pores or cement; cement growth locally truncates their grain boundaries, proving their detrital origin (Figures 7c, 9a and 10c). Fe-ooids increase in quantity from the foreshore to the proximal lower shoreface but there are none in the lower shoreface and offshore (Figure 14f).
Bioclasts are dominated by bivalves (Figures 7b, 8a and 11a) and show evidence of intense bioerosion by micritisation (Figure 11a). Bioclast concentration is highest in the upper shoreface facies association and is lowest in the offshore facies association ( Figure 14b); the variable degree of bioclast neomorphism makes interpretation of point count abundance difficult. If all calcite was derived from bioclasts, then total calcite from SEM-EDS analysis may be used as a proxy for total initial quantity of bioclasts; in this case offshore sediment had the lowest quantity of bioclasts (Figure 12l).
The mineralogy of fine-grained sediment was resolved using SEM-EDS data (Figures 8d and 9c). Fine-grained sediments are mostly composed of illite, kaolinite, muscovite, biotite, minor quantities of pyrite, quartz, K-feldspar and plagioclase, as well as trace quantities of smectite (Figures 7d and 9b,c). SEM-EDS data show that biotite (Figure 12f Heavy minerals are dominated by apatite, which makes up about 0 to 1.0% (average 0.1%) of total rock volume. Other heavy minerals include zircon, rutile and iron oxide, which occur in minor quantities (<1%). Apatite occurs as fine grains visible in SEM-EDS (Figure 10b,c) and does not show a relationship with facies associations (Figure 12n).
Calcite cement occurs as acicular and isopachous cements lining bivalve shells ( Figure  7b) and poikilotopic crystals (Figure 11a) that fill pore spaces. Calcite represents the dominant cement by volume in the Corallian sandstones, locally constituting more than 40% of rock volume (Figure 14c). Pore-filling calcite constitutes about 1.6% in high porosity samples and up to 42.0% in samples with the lowest porosity (Table A1). A significant proportion, up to 23.7%, of calcite cement is replacive (Figure 10b) (Table A1). Some calcite cement shows evidence of neomorphism of bivalve shells (Figure 11a), identifiable by the presence of earlier micrite envelopes.
Dolomite occurs as rhombic crystals (Figures 7a and 10a). Dolomite occurs in samples with evidence of calcite cement and grain dissolution (Figures 7a, 9a and 10a,b,d) and is also common in proximity to iron-rich minerals, e.g., siderite (Figure 9b). SEM-EDS analysis shows Figure 14. Boxplots relating point count data to facies associations: (a) intergranular porosity, (b) bioclasts, (c) pore-filling calcite, (d) grain size, (e) pore-filling berthierine, (f) Fe-ooids, and (g) pore-filling quartz. Intergranular porosity and pore-filling calcite show an inverse relationship while bioclasts and grain size show a good relationship with both reducing basinward from foreshore to offshore facies association. Pore-filling berthierine show an increasing trend from the foreshore to the proximal lower shoreface (e) similar to Fe-ooids (f) suggesting a genetic relationship. However, the absence of Fe-ooids in the distal lower shoreface and offshore facies association suggests other sources of authigenic berthierine apart from reprecipitation of Fe-ooids.
Calcite cement occurs as acicular and isopachous cements lining bivalve shells (Figure 7b) and poikilotopic crystals (Figure 11a) that fill pore spaces. Calcite represents the dominant cement by volume in the Corallian sandstones, locally constituting more than 40% of rock volume (Figure 14c). Pore-filling calcite constitutes about 1.6% in high porosity samples and up to 42.0% in samples with the lowest porosity (Table A1). A significant proportion, up to 23.7%, of calcite cement is replacive (Figure 10b) (Table A1). Some calcite cement shows evidence of neomorphism of bivalve shells (Figure 11a), identifiable by the presence of earlier micrite envelopes.
Dolomite occurs as rhombic crystals (Figures 7a and 10a). Dolomite occurs in samples with evidence of calcite cement and grain dissolution (Figures 7a, 9a and 10a,b,d) and is also common in proximity to iron-rich minerals, e.g., siderite (Figure 9b). SEM-EDS analysis shows that non-ferroan dolomite constitutes up to 6%, with an average of 0.7%, while ferroan dolomite constitutes up to 13%, with an average of 1.0% (Figure 12c,d, respectively).
Siderite occurs as yellow-brown rhombohedra under plain polarised light (Figure 7a). Siderite is heterogeneously distributed. It is pervasive in areas with calcite cement (Figure 7a) and occurs filling moldic pores from detrital grain dissolution (Figures 7a and 10b). Siderite is present up to 4.3% as pore-filling siderite, with an average value of 1.9% (Table A1). Grain replacive siderite locally constitutes up to 15.4% with a mean value of 0.49% (Table A1).
Berthierine occurs as pore-filling (Figure 9a), grain replacive (Figure 7b,d) and grain coating cement (Figure 10c). Berthierine grain coats form discontinuous rims around detrital grains and range from trace to 2.6% with an average of 0.2% of total rock volume (Table A1). Grain-coating berthierine has locally inhibited the growth of quartz cement (Figure 10c). Pore-filling berthierine occurs next to iron-rich minerals, for example, in Figure 7d, berthierine occurs adjacent to biotite-rich matrix and pyrite cement. Partly dissolved Fe-ooids, with their initial concentric laminae, contain replacive structureless berthierine (Figures 7a and 9a). Berthierine ( Figure 12h) and biotite (Figure 12f) show an increase from foreshore to the distal lower shoreface, where they have the highest quantity, and then a reduction in quantity in the offshore, suggesting a genetic relationship between biotite and berthierine.
Illite is present as a porous, fiber-like mass of irregular crystals that fill pore spaces (Figure 10d). Rare grain-replacive illite is present (Figure 7c) where illite has grown in place of detrital muscovite. Pore-filling illite has an average of 2.7%; replacive illite represents an average of 0.2% (Table A1).
Pyrite occurs in three main forms. Pore-filling pyrite surrounds detrital quartz and other grains (Figure 7d) and has equant outlines where it has grown into open pores (Figure 7d). Aggregates of framboidal pyrite (Figure 10d), and bioclast-replacive pyrite (Figure 7b) are also present in these sandstones. Pore-filling pyrite represents an average of 1.3% of total rock volume with a range between trace to a maximum of 14.8% of total rock volume (Table A1).
Apatite seems to be present as a cement as well as a detrital grain as it occurs as an unusual grain-coating phase (Figure 9a) but it only represents 0.1% of the sandstones, on average. Incipient syntaxial quartz overgrowths have formed relatively early as they are surrounded by calcite cement (Figure 10d). Quartz cement is present as small (11 to 20 µm) euhedral outgrowths on detrital quartz grains (Figure 10c). Fluid inclusions seem to be absent in the quartz cements (Figure 7c,d). Authigenic quartz ranges from trace quantities to 7.9% with an average of 0.6% (Table A1). Quartz cement does not have rounded edges suggesting that it is not recycled from a previously deeply buried sandstone; quartz cement has grown in situ. For example, Figure 7b displays perfectly euhedral multiple quartz outgrowths on a single grain that would not have survived weathering, erosion, transport, and deposition. Locally, quartz cement growth seems to have been inhibited by berthierine grain coats (Figure 10c). Quartz cement is more common in the cleaner, higher porosity sandstones that have less calcite cement (Figures 7c,d and 10d). Quartz cement increases from the foreshore facies association to the proximal lower shoreface facies association (Figure 14g).

Modal Porosity
Point count porosity ranges from trace to 28.2%. The main porosity type is intergranular porosity, which constitutes up to 23.0% of total rock volume, followed by moldic porosity which makes up to about 14.0% of total rock volume. Intergranular porosity tends to be lowest in samples with the greatest quantity of authigenic cements (Figure 11a). Moldic porosity occurs due to the dissolution of detrital grains such as Fe-ooids (Figure 7a,b) and feldspars (Figures 7a and 9b). Other porosity types include cement dissolution porosity (Figure 7d), intercrystalline porosity (Figure 10b), matrix porosity ( Figure 7c) and fracture porosity (Figure 9d).

Burial and Thermal History Modelling
Burial curves show that the base of the Corallian sandstones (which is also the top of the Ampthill Clay Formation) was buried to a maximum depth of~1700 m TVD before uplift to the present depth of~840 m TVD.
Compactional porosity-loss was modelled as a function of depth and clay fraction following Ramm, Forsberg [38] with core analysis porosity data added from BL5, PW3 and PW7 (Figure 15b). The core porosity values were also projected to maximum burial, as modelled in Figure 15a. At the current depths, porosity is low compared to the modelled compaction curves (Figure 15b). If we assume that the sandstones had approximately 20% clay, then the best reservoir quality sandstones accord to rocks buried approximately 1000 m deeper than they are at the present time, if we assume that compaction was the dominant porosity-loss mechanism in the best reservoir quality sandstones. However, most of the core porosity cannot be accounted for by compactional processes, alone, implying that cement growth had a major impact on porosity. Overall, Figure 15b suggests that the porosity values at current depths are not just the result of compactional porosity-loss at maximum burial.

Burial and Thermal History Modelling
Burial curves show that the base of the Corallian sandstones (which is also the top of the Ampthill Clay Formation) was buried to a maximum depth of ∼1700 m TVD before uplift to the present depth of ∼840 m TVD.
Compactional porosity-loss was modelled as a function of depth and clay fraction following Ramm, Forsberg [38] with core analysis porosity data added from BL5, PW3 and PW7 (Figure 15b). The core porosity values were also projected to maximum burial, as modelled in Figure 15a. At the current depths, porosity is low compared to the modelled compaction curves (Figure 15b). If we assume that the sandstones had approximately 20% clay, then the best reservoir quality sandstones accord to rocks buried approximately 1000m deeper than they are at the present time, if we assume that compaction was the dominant porosity-loss mechanism in the best reservoir quality sandstones. However, most of the core porosity cannot be accounted for by compactional processes, alone, implying that cement growth had a major impact on porosity. Overall, Figure 15b suggests that the porosity values at current depths are not just the result of compactional porosityloss at maximum burial. There are two possible petroleum source rocks in the modelled section: the Oxford Clay and the Kimmeridge Clay Formations. The top of the Oxford Clay Formation was buried to ∼1859 m TVD (Figure 15a). The younger Kimmeridge Clay Formation was buried to ∼1707 m TVD (Figure 15a). In this well, neither source rock reached thermal ma-  (Figure 15a). The younger Kimmeridge Clay Formation was buried to~1707 m TVD (Figure 15a). In this well, neither source rock reached thermal maturity sufficient for oil generation but note that the deeper (and thus hotter) Lower Jurassic Liassic source rock was not included in the model as it was not penetrated by the well.

Timing of Diagenetic Events
Diagenesis in the Corallian sandstones occurred only in the eogenetic realm as the Corallian sandstones have only reached about 60 • C (Figures 15 and 16) with the boundary between eo-and meso-diagenesis defined as being about 60 to 70 • C [5]. The timing of diagenetic events is summarised in Figure 16. We interpret the first event to be feldspar dissolution (Figure 7a,b) which may have begun during weathering and sediment transport as well as continuing after burial [39,40]. Shell bioerosion (Figure 7b) is interpreted to have occurred early, under open marine photic conditions which are needed for the vital activities of microscopic boring algae [41,42]. Locally intense bioturbation (Figure 3e) occurred soon after burial, when animals were either seeking shelter or food [28]; it is likely that bioturbation led to any remaining organic matter associated with the bioclasts being destroyed. The initial stages of compaction seems to have occurred next (Figures 7b and 10a). There was locally abundant growth of framboidal pyrite soon after deposition (Figure 10f) that also probably occurred during the initial stages of burial [43]. Framboidal pyrite was followed by the development of apatite, berthierine and calcite cement. Apatite occurs as a relatively unusual grain-coating phase (Figure 9a). Berthierine occurs as a grain-coating and pore-filling cement phase (Figure 10c,d). Poikilitopic calcite (calcite-2; Figures 7b and 11a) overlies an earlier generation of acicular calcite cement (calcite-1; Figure 7b); the latter possibly formed on or near the sediment surface. Poikilotopic calcite, the most important of all the pore-filling cements in the Corallian sandstones, locally filled most pore spaces with intergranular volumes of up to about 43%, confirming that such cement formed before significant compaction had occurred [44].
Subsequent diagenetic processes included quartz overgrowth formation (Figures 7d,  8b and 10a,c). Partial dissolution of Fe-ooids (Figures 8a and 10c) and feldspars (Figure 7a) occurred, possibly after quartz growth as quartz was preceded by berthierine coat formation (Figure 10c). Quartz growth and Fe-ooid dissolution were followed by hydrocarbon charge as shown by the presence of bitumen staining on quartz cement (Figure 7d) and on the internal dissolution surfaces of ooids (Figure 7b). These events were followed by minor quantities of siderite and dolomite cement, replacive calcite (labelled calcite-3), and illite and kaolinite. The timing of siderite and dolomite cannot be differentiated (Figure 10b) but they occurred after Fe-ooid dissolution (Figure 10b). Replacive calcite-3 developed after siderite and dolomite (Figure 10b) but at about the same time as pore-filling illite (Figure 10d). Kaolinite seems to have been a late-stage diagenetic mineral that grew at the expense of detrital feldspars (Figure 10c).
Compaction occurred during burial, but these sandstones are typically cementationdominated rather than compaction-dominated ( Figure 17); this was caused by the extensive development of locally-dominant poikilotopic calcite (Figure 11a) and other porefilling minerals (Figure 7a). The least calcite-cemented sandstones display long grain to grain contacts evidencing the occurrence of compaction (Figure 7d). Figure 17. Intergranular volume (%) versus cement (%) diagram after the method developed by Houseknecht [45] showing the role of cementation and compaction in initial porosity-loss with the distal lower shoreface and offshore facies associations having the highest initial porosity-loss.
The dissolution of calcite cement (Figure 7d), Fe-ooids (Figures 7b and 8a) and feldspars (Figure 7a,d), towards the end of the succession of events, led to the creation of secondary porosity. Based on XRD data, chlorite is present in trace quantities ( Figure 13) but it is not apparent in optical images so we could not determine if it had detrital or authi- Figure 17. Intergranular volume (%) versus cement (%) diagram after the method developed by Houseknecht [45] showing the role of cementation and compaction in initial porosity-loss with the distal lower shoreface and offshore facies associations having the highest initial porosity-loss.
The dissolution of calcite cement (Figure 7d), Fe-ooids (Figures 7b and 8a) and feldspars (Figure 7a,d), towards the end of the succession of events, led to the creation of secondary porosity. Based on XRD data, chlorite is present in trace quantities ( Figure 13) but it is not apparent in optical images so we could not determine if it had detrital or authigenic origins. Chlorite is not present in sufficient quantity to influence reservoir quality or to ascertain its diagenetic relationships.

Pyrite
Pyrite formed early and precipitated throughout burial in the Corallian sandstones (Figures 8d and 10f). In addition, pyrite partly replaced bioclasts (Figure 7b) suggesting early sulfide mineral formation, before neomorphism of bivalves to calcite cement. Pyrite-filled pores (Figure 10d) formed adjacent to detrital grains and completely occluded pore spaces suggests that early diagenetic growth continued during subsequent burial. Framboidal pyrite is characteristic of early diagenesis while the coarser morphologies of pyrite, e.g., Figure 8d, typically represent later diagenesis [1,43]. Pyrite in marine sediments requires organic-rich, anoxic, reducing conditions to convert marine aqueous sulphate to sulfide and ferric iron to ferrous iron [43,46]. The iron in pyrite in the Corallian sandstones may have been supplied by depositional biotite, Fe-ooids, and detrital or seabed-related glauconite (Figure 8b,d). The Corallian sandstones were deposited in sulphate-rich sea water and the decomposition of organic matter potentially consumed oxygen and reduced the redox potential (Eh) producing sulfide and ferrous iron [47]. In the absence of carbonates, sulfide and ferrous iron combine to form iron sulfide.

Apatite
Apatite requires a source for phosphorus; apart from heavy mineral deposits, apatite is commonly associated with sediments that experienced much biological activity. Detrital apatite may be derived from bone and teeth fragments [47][48][49]. Apatite can also be enriched in coprolites [47] and can occur at a trace concentration in bioclasts [47,50]. It is likely that much of the diagenetic apatite found in the Corallian sandstones is recrystallised detrital apatite derived from bone fragments [48].
Berthierine forms as a result of reduction of iron at low Eh conditions [51][52][53]; these conditions commonly occur in nutrient-rich coastal waters [55]. The presence of detrital Fe-rich biotite (Figure 9b) and Fe-ooids in these sediments (Figure 7c) indicates that the primary sediment was enriched in iron. The presence of diagenetic siderite (see later and Figure 9b) and pyrite (Figures 8d and 10f), as well as berthierine, confirms that Fediagenesis was an important aspect of the evolution of these rocks. The specific dominant Fe-phase was influenced by the supply of sulfide and bicarbonate, both probably from organic sources [49,56]. Berthierine grows in sulfide-and bicarbonate-limited pore waters, influenced by a suite of bacterial processes [57]. The degradation of Fe-ooids led to the formation of thin berthierine coats ( Figure 10) and pore-filling berthierine (Figure 7a) similar to berthierine in the Wealden sandstones of the Paris Basin [58].
Another possible source of berthierine is detrital biotite (Figure 7d). Authigenic berthierine tends to be physically close to detrital biotite-rich matrix (Figures 8d and 9a). There is a significant correspondence between biotite and berthierine enrichment patterns as a function of facies association (Figure 12f,h). Biotite can be a source of the main ingredients for berthierine; however, the latter has a higher Al/Si ratio than the former suggesting either that (i) silica was exported to other parts of the Corallian sandstone (e.g., as quartz cement) or (ii) Al was imported from other parts of the Corallian sandstone (e.g., from alumino-silicate mineral breakdown).

Calcite
Early calcite cement is here interpreted to have formed from neomorphism of bivalve tests, starting soon after deposition and continuing during burial. Petrographic evidence shows bivalve shells with various degrees of neomorphism (Figures 7b, 8a and 11a) to calcite [42,59,60]. Neomorphism of aragonitic bivalve shells is emphasised in Figure 11a where pervasive calcite cement cannot be distinguished from bivalve shells except where micrite envelops outline the boundary of bivalve shells which neomorphosed into calcite. In contrast, Figure 11b shows a porous rock, with unaltered bivalve shells and very little calcite cement as the shells have only been slightly micritised, with no significant neomorphism (Figure 11b), leading to a low calcite cement volume.
Calcite-1 forms acicular rims on bivalve shells (Figures 7b and 11a) and does not significantly project into pore space, indicating early formation. Acicular rims are surrounded by poikilotopic calcite (calcite-2) which pervasively fills pore spaces indicating development of calcite as neomorphism progressed (Figure 11a).
Secondary calcite (calcite-3) is here interpreted to have formed after local dissolution of detrital or eogenetic calcite following addition of CO 2 -rich waters during burial, possibly by source rock decarboxylation. There is evidence for dissolution of calcite cement (Figure 7d), Fe-ooids (Figures 8a and 10c,e) and feldspars (Figures 7a,d and 9a,b). There is evidence of calcite development following grain dissolution (Figures 9a and 10b). The development of calcite-3 after dissolution is here interpreted to occur due to reduction in the partial pressure of CO 2 (PCO 2 ) during grain and cement dissolution causing a reduction in acidic conditions and a return to more neutral conditions suitable for calcite development.

Quartz
Typically, quartz cementation is a kinetically-driven process whose rate depends on a silica source, mass transfer to the site of cementation and, finally, cement precipitation [61]. Quartz cement requires a silica source, a mode of transport and a site for deposition or crystallisation; these depend on detrital mineralogy, temperature, grain size and degree of grain coat coverage [62]. Silica for quartz cementation is here interpreted to be sourced from feldspar dissolution (Figures 7a,d and 9a,b) [39,63,64] as well as the alteration of other detrital alumino-silicates such as Fe-ooids (Figure 10c,e), degradation of biotite ( Figure 9a) and muscovite (Figure 8c).
The low temperature (<60 • C), shallow depth ( Figure 15) of the quartz cement in the Corallian sandstones is significant as it shows that incipient quartz cementation occurs at shallow depths, similar to Lower Cretaceous Wealden sands in the Paris Basin [58]. This is noteworthy as it shows that quartz cement does not only develop at temperatures greater than 80 • C as has been widely reported in literature [62,63,65,66]. Furthermore, this low temperature quartz cement was locally inhibited by berthierine coats (Figure 10c). Inhibition of quartz cement is commonly linked to chlorite grain coats with berthierine serving as a possible precursor mineral for chlorite [55,57]. The inhibition of quartz overgrowth here proves that the preservation of reservoir quality by grain-coats starts during eodiagenesis with berthierine coats. Berthierine coats are therefore potentially important in shallow reservoirs where chlorite coats have not developed. As has been recently shown for shallow buried sandstones in NW France [58], berthierine coats can therefore preserve reservoir quality by inhibiting quartz cement at temperatures much lower than 70 • C.

Siderite
The precipitation of siderite is interpreted to be a late-stage event, as seen in Figure 10b, where siderite precipitated in moldic pore. Siderite was probably derived from iron supplied by alumino-silicates (e.g., berthierine in Fe-ooids, Figure 7a, or biotite) and carbonates. This is similar to models of siderite origin published by Macquaker, Taylor [49], Gehring [67].

Dolomite
Dolomite is here interpreted to have developed following the dissolution of calcite cement and biotite or Fe-ooids. Dolomite rhombohedra locally occur close to dissolved Fe-ooids (Figures 8a and 10a,d). As magnesium and iron ions are present in berthierine [53], alteration of berthierine in ooids releases Fe 2+ and Mg 2+ to combine with carbonate ions to produce dolomite [52]. In addition, dolomite and Fe-dolomite show a similar pattern, by facies association, to biotite (Figure 12c,d,f) which suggests a genetic relationship where iron and magnesium released from biotite alteration can also combine with carbonate ions to produce dolomite.

Illite
Illite cement is here interpreted to develop from the alteration of phyllosilicate minerals such as muscovite ( Figure 8c) and K-feldspar (Figure 10a) [68]. Illite is also interpreted to result from the dissolution of feldspars (Figure 7a) [59].

Kaolinite
Similar to illite, kaolinite is interpreted to have precipitated following alumino-silicate mineral breakdown, e.g., Emery, Myers [39], Orem and Finkelman [69]. Meteoric water flushing has widely been discussed as causing the growth of kaolinite in shallow sandstones [39,40], with the leaching of feldspars and mica occurring at depths as shallow as 10 to 20 m [70]. As kaolinite is mostly in pore spaces close to dissolved silicates (Figure 10c), it is interpreted to have formed as a result of dissolution of K-feldspar [59] (Figures 7d and 9a,b) in iron-deficient acidic pore-fluids after CO 2 flux, possibly from source rocks.

Role of Compaction and Cementation on Porosity
Detrital mineralogy in the Corallian sandstones is dominated by quartz ( Figure 6). Quartz-rich sandstones undergo less compaction than ductile-rich sandstones [45,71]. As the modelled burial curves in Figure 15b illustrate, porosity-loss did not follow the expected trend for mechanical compaction. The deviation from the modelled trend suggests that mechanical compaction is not the dominant control on porosity evolution.
The degree of compactional and cementational porosity-loss was quantified using point count data using the approach described by Houseknecht [44] employing an assumed initial porosity of 45% ( Figure 17). The Houseknecht plot shows that cementation is the dominant control on reservoir quality for the Corallian sandstones. All but three samples plot in the quadrant for porosity-loss being dominated by cementation. The three samples that lie in the field of dominant compactional porosity-loss either have a relatively low cement content (i.e., BL5-2209.46 m MD, cement content of 11.34%; BL5-2208.0 m MD, cement content of 16.35%; BL5-2211.6 m MD, cement content of 16.36%) or have a high matrix clay content (PW3-1098.3 m MD, 61% clay).

Controls on Diagenesis and Reservoir Quality
The Corallian sandstones do not have a simple relationship between depositional facies and reservoir quality. For example, there is no clear relationship between reservoir quality and either grain size or sorting (Figure 18a). In addition, there is no apparent relationship between facies association and reservoir quality (Figure 18b). The best reservoir quality samples, i.e., those that plot in the top-right corner of Figure 18c, tend to have little calcite or clay minerals. The worst reservoir quality samples have high calcite and/or clay contents (Figure 18c). This shows that there is a strong diagenetic and cement-related control on reservoir quality. Samples with moderate to high porosity, low calcite but high clay content, have lower permeability, by up to an order of magnitude or more, than samples in their porosity range (Figures 5 and 18c). This indicates that permeability is strongly affected by clay mineral content, even in high porosity samples. Reservoir quality was partly preserved by berthierine grain coats which inhibited the development of quartz cement (Figure 10c). Porosity was significantly enhanced by grain dissolution and the creation of secondary porosity (Figures 7a,b, 8a and 10c,d,e). However, reservoir quality was minimally affected by grain replacive cementation which filled secondary pores created by dissolved grains without enhancing or destroying porosity (Figure 10b).

Reservoir Quality Model
A reservoir quality evolutionary path, showing the effects of calcite and clay minerals, is illustrated in Figure 19 illustrating the controls on reservoir quality using SEM-EDS images Reservoir quality was partly preserved by berthierine grain coats which inhibited the development of quartz cement (Figure 10c). Porosity was significantly enhanced by grain dissolution and the creation of secondary porosity (Figures 7a,b, 8a and 10c,d,e). However, reservoir quality was minimally affected by grain replacive cementation which filled secondary pores created by dissolved grains without enhancing or destroying porosity (Figure 10b).

Reservoir Quality Model
A reservoir quality evolutionary path, showing the effects of calcite and clay minerals, is illustrated in Figure 19 illustrating the controls on reservoir quality using SEM-EDS images for selected points. The lines connecting the SEM-EDS images show the reservoir quality evolutionary path for various sandstone types. Pattern 1 shows the evolutionary path of clean sandstones and illustrates how an increase in overall clay mineral content and calcite destroyed reservoir quality. Pattern 2 shows the effect of dissolution of cements and grains on reservoir quality, whereby there is an increase of porosity and permeability. Pattern 2 has a higher permeability range than sediments on pattern 1, possibly because of extensive creation and connection of pores from dissolution at the current shallow depths without significant compensative pore size reduction from compaction. Pattern 3 shows how clays disproportionally decrease permeability in calcite-poor samples indicating that calcite-poor sandstones will not have high permeability if they are clay-rich. for selected points. The lines connecting the SEM-EDS images show the reservoir quality evolutionary path for various sandstone types. Pattern 1 shows the evolutionary path of clean sandstones and illustrates how an increase in overall clay mineral content and calcite destroyed reservoir quality. Pattern 2 shows the effect of dissolution of cements and grains on reservoir quality, whereby there is an increase of porosity and permeability. Pattern 2 has a higher permeability range than sediments on pattern 1, possibly because of extensive creation and connection of pores from dissolution at the current shallow depths without significant compensative pore size reduction from compaction. Pattern 3 shows how clays disproportionally decrease permeability in calcite-poor samples indicating that calcite-poor sandstones will not have high permeability if they are clay-rich. Corallian sandstones rich in bioclast-derived calcite presumably were deposited in shallow, limpid ocean waters with a great abundance of biological activity. The addition Figure 19. A model for reservoir quality evolution of the Corallian sandstones with SEM-EDS images of selected samples from points in Figure 18c. The arrows are a schematic for reservoir quality evolution patterns from a clean sand-rich. Corallian sandstones rich in bioclast-derived calcite presumably were deposited in shallow, limpid ocean waters with a great abundance of biological activity. The addition of a quartz-dominated sand to a site of active bioclast accumulation led to compromised reservoir quality in the subsurface due to the propensity for such bioclast-quartz sand hybrids to become calcite cemented [12,60,72]. Conversely, sandstones rich in matrix clay, presumably in turbid ocean waters, near to the site where rivers fed sediment into the ocean, have far fewer bioclasts and thus much less calcite cement [72]. However, these clay-rich Corallian sandstones have their subsurface reservoir quality strongly affected by the presence of the same clay minerals that inhibited bioclast development at the site of deposition.

1.
Upper Jurassic Corallian sandstone reservoirs in the Weald Basin, UK, were buried to a maximum of about 1700 m and achieved a maximum temperature of about 60 • C before being uplifted to present depths of about 800 to 900 m and temperature of about 40 • C. 2.
The Corallian sandstones have moderate to poor reservoir quality. The diagenetic processes that modified these sandstones include the growth of various cements (in approximate order of abundance: calcite, siderite, dolomite, Fe-dolomite, quartz, pyrite, apatite, berthierine, kaolinite and illite) as well as minor compaction. Reservoir quality was also influenced by dissolution of Fe-ooids, K-feldspar, plagioclase feldspar and calcite cement.

3.
Reservoir quality does not seem to have strong depositional controls. Corallian sandstones are cementation-dominated, as opposed to compaction-dominated, calcite is the dominant cement and the main control on reservoir quality. Calcite cement was derived from detrital bioclasts co-deposited with the sand grains.

4.
Kaolinite and illite occur as pore-filling clays while berthierine occurs as pore-filling, grain replacive and grain coating clay.

5.
Quartz cement has a low temperature, early diagenetic origin but is volumetrically minor, at least, in part, due to its inhibition by berthierine grain coats. Berthierine clay coats locally preserved reservoir quality by preventing quartz overgrowths. 6.
Clay minerals (kaolinite and illite) locally reduced reservoir quality, especially decreasing permeability for only a small decrease in porosity. 7.
Secondary porosity, following dissolution of calcite cement and detrital feldspars and Fe-ooids, locally increased porosity and permeability. 8.
The best reservoir quality in Corallian sandstones is found in the cleanest, most clayfree sandstones with the least amount of calcite cement. Detrital bioclast content is the over-riding control on reservoir quality in these sandstones.