Failure to find hydrocarbon prospects in the Solway Basin region has resulted in a lack of research into the local Sherwood Sandstone Group petrography, reservoir quality and depositional history compared to the analogous southern reservoirs in the EISB which will be utilized for carbon storage. A detailed petrographic study is presented which aims to understand if the Solway Firth could have similar utility. The Permo–Triassic Sherwood Sandstone Group is believed to be deposited in depocentres connected during the Early Triassic by the extensive ‘Budleighensis’ fluvial system. Here, the Solway and Carlisle basins are proposed as terminal sites for this endorheic system, with the Lower Triassic Annan Sandstone Formation ascribed to the distal region of a fluvial distributary zone and the overlying Kirklinton Sandstone Formation thought to mark a transition to a basinal zone, depositing aeolian sandstones and locally associated playa lake facies. Fluid inclusion, stable isotope burial history modelling and field observations have been used to assess the relative timing and importance of different diagenetic cements. Early diagenetic cements include grain-rimming haematite and patchy calcite cement, especially in the Annan Sandstone Formation. Later burial diagenesis sees further calcite cement, quartz overgrowths and, restricted to the Kirklinton Sandstone Formation, ferroan dolomite. Porosity and permeability show significant differences between fluvial Annan and aeolian Kirklinton facies associations. Despite the finer grain size, a reservoir with excellent porosity and permeability as well as no hydrocarbon charging or legacy hydrocarbon extraction is persevered, suggesting the Solway Basin could be a secure CO2 storage site.

This article is part of the Energy Geoscience Series available at https://www.lyellcollection.org/cc/energy-geoscience-series

Supplementary material: An overview of the primary and secondary data collected and utilised in this study, as well as raw data values are available at https://doi.org/10.6084/m9.figshare.c.5906677

The offshore and onshore NW England Triassic succession, comprising Early–Middle Triassic (Olenekian–Ansian) Sherwood Sandstone Group (SSG), and the Middle–Late Triassic (Anisian–Norian) Mercia Mudstone Group (MMG), represents part of a large-scale internal drainage system in the semi-arid to arid interior of the Pangaea supercontinent. NW Europe lay between 15°N and 25°N and was influenced by SW-directed subtropical trade winds giving rise to general semi-arid to arid conditions with an annual summer monsoon and intense seasonality (Kutzbach and Gallimore 1989; Parrish 1993; Szulc 1999; Preto et al. 2010).

The Middle Triassic sediments infilled extensional rift basins with Early Triassic fill ascribed to the action of a major northward flowing river system, first termed the ‘Budleighensis’ fluvial system by Wills (1951). This system carried material from the Variscan massifs of western and central Europe in >400 km northwards through the Wessex, Worcester, Stafford and Cheshire Basins of England, exiting into the East Irish Sea Basin (EISB) where it formed a sand-dominated, lower flow regime, low to moderate sinuosity braided river system (Fig. 1; Audley-Charles 1970; Warrington and Ivimey-Cooke 1992; Hounslow and Ruffell 2006; Tyrrell et al. 2012; Ambrose et al. 2014).

Within the arid Pangean climate, aeolian-derived facies are preserved amongst fluvial deposition, increasing northwards and becoming dominant at basin margins such as at Sellafield, West Cumbria (Jones and Ambrose 1994; Hounslow and Ruffell 2006). The reservoir properties and depositional framework of the SSG within the EISB has been extensively researched due to the region's abundance of hydrocarbon-prolific wells (Meadows and Beach 1993a, 63b; Meadows 2006). The EISB extends offshore from Liverpool Bay and the north coast of Wales northwards to Ramsey–Whitehaven Ridge, located between the west Cumbrian coast and the east coast of the Isle of Man, where it borders the Solway Basin to the north (Fig. 1). Failure of hydrocarbon exploration within the Solway Basin has meant little investigation has been conducted into its reservoir system, which is analogous to the proven hydrocarbon systems within the EISB, since the late 1990s (e.g. Newman 1999). As a consequence, the sandstone petrography of the Lower–Middle Triassic of the Solway Basin and its relation to the Budleighensis fluvial system is comparatively less well understood than that of the proximal EISB (Fig. 1; Akhurst et al. 1997; Meadows 2006; Tyrrell et al. 2012). It has been suggested that the Triassic sandstones deposited in the Solway Basin were isolated from the EISB by the Ramsey–Whitehaven Ridge; a NE–SW trending fault-bounded high throughout much of the Late Paleozoic and Mesozoic Era (Newman 1999; Quirk et al. 1999; Floodpage et al. 2001). Alternatively, Meadows (2006) suggests that the Budleighensis system diverted westwards into the Peel Basin, whilst Hounslow and Ruffell (2006) suggest that the system may have terminated in the Solway Basin or, which they deem more likely, diverted northwards into the proto-Atlantic.

In this study we attempt to characterize the distal section of the ‘Budleighensis’ fluvial system spatially and temporally during the deposition of the Early–Middle Triassic Annan and Kirklinton Sandstone Formations of the Solway Basin SSG, interpreting the nature of its termination. We present new results for the petrography and diagenesis of these sandstones, supported by offshore hydrocarbon data within the Solway Basin. Offshore hydrocarbon data from contemporaneous aged hydrocarbon reservoirs in the North and South Morecambe Fields of the EISB are additionally used to compare the porosity, permeability and petrography of the Solway Basin to its southern counterpart. This comparison will allow the porosity and permeability of the Solway Basin reservoir to be compared to benchmark proven hydrocarbon reservoirs and utilize any differences in petrography to determine the Solway Basin's depositional position in the overall Budleighensis system. The approach presented in this paper provides an important step towards identifying the distribution of facies, better understanding of this significant drainage system, its termination, and the effect this has had on the less-well constrained petrography and reservoir characteristics of the Solway and Carlisle Basins. Once better constrained, the utility of this reservoir system as a potential CO2-storage site is explored. Analogous reservoir systems in the EISB have already been chosen as carbon storage sites for the Hynet North West Project, a decarbonized industrial cluster project that will see simultaneous carbon capture and storage (CCS), utilizing Liverpool Bay depleted oil and gas reservoirs, and hydrogen production (HyNet 2021). This project is backed by the UK Government as one of two Track-1 projects, which will mean decarbonization will begin by 2025 and will both have access to £1 billion of state-funding (GOV.UK 2021).

Geological setting

The western side of England and Southern Scotland was subject to extensive faulting during the Permo–Triassic period and was concentrated in a roughly north–south trend from SW England to the Solway Firth, forming the large depocentres that would form the pathway to the ‘Budleighensis’ fluvial system and be filled by the SSG. This system would flow through a tectonically controlled landscape of topographically high massifs and deep pocket-like basins, with inter-massif channels connecting them (Newell 2018).

Such Permo–Triassic basins within Britain and NW Europe developed in response to east–west extension associated with the post-Variscan break-up of Pangaea and periodic rifting in the North Atlantic (Jackson and Mulholland 1993; Newman 1999; McKie and Williams 2009). The structural boundaries of these basins are formed by NE–SW trending thrusts and NW–SE trending transfer faults inherited from the Caledonian and Variscan Orogenies (Newman 1999). This strong control of Caledonian and Variscan structures upon the development of these basins is reflected in their present-day orientation, whereby the Carlisle, Solway and Peel Basins have a broad NE trend, whereas other nearby Permo–Triassic basins such as the Vale of Eden and EISB, have a marked NNW orientation (Fig. 1; Jackson and Mulholland 1993; Chadwick et al. 1995, 2001; Akhurst et al. 1997; Holliday et al. 2004).

Both the Solway Basin and its onshore extension, the Carlisle Basin, have undergone a complex tectono-stratigraphic evolution, involving several phases of extension and basin development followed by periods of uplift and erosion. Both basins crosscut structures of the underlying early Permian rift basins (Newman 1999; Floodpage et al. 2001; McKie and Williams 2009) and predates the late Triassic–Jurassic rifting phase, which began with the opening of the Central Atlantic Ocean (Manspeizer 1988). The late Permian to Jurassic sequences were deposited in response to regional post-rift thermal relaxation, which was initiated in the Early Carboniferous, forming an intracontinental basin 30–50 km wide and 125 km long (Quirk and Kimbell 1997; Newman 1999). In the basin centre, the fill comprises over 1600 m of Lower–Middle Triassic strata (Sherwood Sandstone Group) and interpreted to have been deposited in a variety of arid to hyperarid continental environments, differing from the general semi-arid to arid conditions of NW Europe (Brookfield 2004, 2008). Regional subsidence continued in the Late Triassic with the deposition of a thick succession of evaporates and shales which have been assigned to the MMG in the Solway Basin and laterally the Stanwix Shale in the Carlisle Basin (Fig. 2).

It is widely accepted that significant syn-depositional faulting and extension took place during the deposition of the SSG of the Lower–Middle Triassic within the EISB, as seen in seismic reflection data (Jackson and Johnson 1996; Chadwick et al. 2001), extension modelling (Rowley and White 1998) and facies distribution mapping (Meadows and Beach 1993a, 63b). This extension was broadly orientated east–west throughout much of the Triassic (Chadwick and Evans 1995; Jackson et al. 1995). In addition, Meadows and Beach (1993a) identify that basin evolution not only controlled facies distribution and depositional environment but this cumulatively influenced the distribution of grain populations. Grain size, which reservoir quality within the EISB has ultimately depended, is dependent on facies type (aeolian v. fluvial) and location along the fluvial system.

Despite evidence for a correlation between active Triassic faulting and facies development within the EISB, this does not seem to be the case in the Solway and Carlisle Basins. The SSG of the Solway Basin records a relatively uniform succession of fluvial and aeolian facies where uniform stratigraphic thicknesses are maintained across the Basin, including basin margins (Fig. 1; Chadwick et al. 1995; Jackson et al. 1995; Akhurst et al. 1997; Quirk and Kimbell l997; Newman 1999; Floodpage et al. 2001; Brookfield 2008).

Stratigraphy

The Permo–Triassic red bed succession lies unconformably upon all underlying units and structures. The Cumbrian Coastal Group is up to 190 m thick in boreholes within the Solway Basin and begins with a very thin (10 m) and variable breccia unit, known as the Basal Clastics (Fig. 2). Where the Basal Clastics overlie Carboniferous rocks, clasts in the basal strata are locally derived and have been interpreted to have accumulated in small fans around low hills and knolls (e.g. Holliday et al. 2004). A relatively thick (180 m) gypsum/anhydrite evaporite and red shale unit overlies this breccia, known as the Eden Shales. This unit is a direct correlative of the St Bees Shale and St Bees Evaporite of the EISB and West Cumbria (Brookfield 2008). However, solution removal of evaporites at the surface means only the upper clastic part is exposed at outcrop within the Solway Basin.

The Eden Shales sit below the fine-grained sandstones of the SSG (Fig. 2), which is the focus of this study and has a maximum recorded thickness of c. 1248 m offshore in well 112/15-1. The SSG comprises two distinct facies: the Annan Sandstone Formation (hereby ASF) and Kirklinton Sandstone Formation (hereby KSF). The lower junction of the ASF is transitional from the Eden Shales and this facies consists of mainly thick (c. 2 m) bedded multi-channel stacked sandstones, with relatively thick interbedded siltstone and mudstone units. The upper part of the ASF again consists of multi-storey channel units, but which reach up to 10 m thick and feature only thin mudstone and siltstone interbeds (Brookfield 2004). Facies associations within the ASF include flood plain fines and playa sediments, which are common features throughout, forming in association with both ribbon and sheetflood fluvial sandstones. The KSF sharply overlies the ASF, marking a sharp contrast in depositional style from stacked fluvial channels below to aeolian deposition above (Brookfield 2004, 2008). Specifically, onshore the KSF features dominantly dry facies, characteristically in the form of large (up to 2 m) cross beds with swept-out toesets as part of aeolian dune structures. Subordinate wet/damp facies with damp interdune and damp sandflat features are also present. Similar depositional facies to those described above are recognized within the correlative units of the EISB (Cowan 1993; Meadows and Beach 1993a, 63b; Meadows 2006).

The principal reservoir sequence within the EISB is formed by the upper St Bees Sandstone Formation (hereby SBSF) of the SSG and the overlying Ormskirk Sandstone Formation (hereby OSF) (Fig. 2; Meadows and Beach 1993b). These units are direct equivalents of the upper ASF and KSF within the Solway Basin, respectively (Fig. 2). The SBSF in the EISB was divided into the lower Rottington Sandstone Member (hereby RSM) (c. 550 m) and upper Calder Sandstone Member (hereby CSM) (c. 650 m) by Colter and Barr (1975) and Jackson et al. (1987), citing a shift in the geophysical log profile and creating the ‘Top Silicified Zone’ boundary between the two. The base of the OSF in the EISB is identified by a seismic marker thought to relate to the regional Hardegsen Disconformity (Barnes et al. 1994).

The KSF is overlain by the Stanwix Shales; the equivalent of the MMG of the EISB (Fig. 2). This unit comprises a thick succession of shales with evaporites, proven up to c. 813 m thick in the Solway Basin (well 112/15-1) and up to 3700 m in the EISB (Wilson 1990). The presence of halite beds within the MMG is critical to seal efficiency because of the multiple periods of fault reactivation during the Early Cretaceous (Late Cimmerian phase) and Tertiary (Floodpage et al. 2001).

Primary data collection for this study consisted of the extraction of onshore hand specimen samples from three exposures of the Lower and Upper ASF and the KSF, of the Solway and Carlisle Basins. These sections were chosen as the most well exposed, representative locations of the formations, where samples were collected on exposed surfaces by hammer and chisel (Brookfield 2004, 2008). Thirty-four samples of the Lower ASF were taken from Cove Quarry (Fig. 1, Location 1; Figs 3a and 4), six samples of the Upper ASF were taken from the gorge of the River Lyne at Glinger Burn (Fig. 1, Location 2; Fig. 4) and sixteen samples of the KSF were taken from the Cliff Bridge section (Fig. 1, Location 3; Fig. 3b). In each case, samples were taken at 0.5 m intervals to allow accurate, detailed documentation of the vertical change in facies and petrography. These samples were prepared into 76 mm by 26 mm blue epoxy-impregnated thin sections, revealing porosity. Where possible thin sections were cut perpendicular to bedding/lamination. As field samples, weathering could have influenced porosity and permeability, however samples were chosen with no weathering rind or from the immediate surface of the exposure to mitigate this effect.

Mineral composition, including authigenic minerals and cements, was determined using point-counting thin section micrographs on a PETROG stepping-stage and counting software under a Leica DM2500P microscope with 300 counts per sample. This data was used for QLF classification. Optical porosity was measured by using the digital image analysis technique, jPOR (Grove and Jerram 2011). The programme jPOR is a macro file (jPOR.txt), which is utilized within the Java-based image manipulation programme, ImageJ (Abramoff et al. 2004). By using jPOR, total optical porosity was determined in a short space of time and provided comparable results to more time-consuming point counting, but with significantly less ‘counting error’ and less variability than published point counting studies (Grove and Jerram 2011). For the measurement of permeability, a portable hand-held mechanical Tiny-perm II Air mini Permeameter was used for rapid in-situ determination of permeability at outcrop and on fresh hand specimens with >120 measurements taken (Chandler et al. 1989). Samples were collected in the field and then analysed in the laboratory on flat sections and where a good seal could not be achieved a new flat cut section was produced.

Petrography (including porosity) was used to calculate compactional and cementational porosity loss using the methodologies ‘Calculation of Compactional Porosity Loss (COPL)’ and ‘Cementational Porosity Loss (CEPL)’ outlined by Houseknecht (1987) and further explained by Stricker et al. (2016).

All thin sections (n:56) were then highly polished to 30 μm and coated with carbon prior to analysis by a Hitachi TM 1000 scanning electron microscope (SEM) and a Hitachi SU-70 field emission gun (FEG), both equipped with an energy dispersive detector (EDS). SEM analyses of thin section and bulk rock samples were conducted at 15–20 kV acceleration voltage with beam currents of 1 and 0.6 nA, respectively. Point analyses had an average duration of 2 minutes, whereas line analyses were dependent on length. SEM-EDS was used for rapid identification of chemical species. Constituents such as small macroquartz and porous clay fill were identified and quantified using a mixture of SEM and SEM-CL.

Oxygen and carbon isotope analysis was then conducted using the primary petrography samples. Calcite and ferroan dolomite cements were analysed from the ASF (n:23) and KSF (n:17), respectively, where samples were ground and reacted with phosphoric acid and the evolved gas for each carbonate fraction was analysed using a ThermoScientific MAT 253 mass spectrometer. Precision was monitored and was better than 0.1‰ PDB for both δ13C and δ18O.

Microthermometric fluid inclusion analysis of the ASF (n:10) from Cove Quarry and KSF (n:4) from Bridge Cliff was carried out using double polished wafers for conditions of cementation and formation waters, where quartz overgrowths and both quartz overgrowths and dolomite cements were sampled from each Formation, respectively. Analysis was conducted on a Linkam THM600/TS90 motorised X and Y, heating and cooling stage linked to a Leica DM2500P polarizing microscope. Accuracy and control over the temperature range of −196 to 600°C enables fluid inclusions to be characterized to better than ±0.1°C, over the range of temperatures reported here. Routinely available measurements are homogenization temperatures (Th) and final melting temperatures (Tm). Homogenization is the conversion of multiphase inclusion contents to a single phase (usually at temperatures above room temperature). Interpreting homogenization temperatures in carbonates, sulfates and halides can be complicated because aqueous inclusions can reset to higher temperature if they are overheated beyond a threshold which is dependent on the mineral strength and inclusion geometry (Goldstein and Reynolds 1994; Goldstein 2001). This can occur in the laboratory as well as through geological processes, so care is taken in the order in which analyses are made for each rock chip. If resetting has occurred, larger inclusions may give higher temperatures, homogenization temperature distributions may show a high temperature tail and data from paragenetically distinct settings may overlap. Final melting occurs at the disappearance of the last trace of solid in the inclusion on heating (usually after cooling an inclusion to well below room temperature).

Schlumberger's burial history simulation software PetroMod (V. 2012.2) was used in this study to help reconstruct the geological evolution of the Solway and Carlisle Basins. Maximum palaeotemperatures and timing obtained from apatite fission-track analysis and palaeotemperatures obtained from fluid inclusions in mineral cements were used to help calibrate the model (Newman 1999; Floodpage et al. 2001).

Primary data was supplemented by the following various supporting secondary data. Firstly, offshore data within the Solway Basin was sourced from the only two hydrocarbon wells drilled within the region (wells UK 112/15-1 and IOM 112/19-1, Fig. 1). Both wells penetrated the KSF and ASF equivalents (OSF and SBSF respectively; Fig. 2), with the OSF chosen as the primary reservoir target in both wells. The data was sourced from the Oil and Gas Authority (OGA) through the National Data Repository (NDR), presented mainly through Well/Geological Completion Reports. Data from a previously produced petrographic analysis was available for well 112/19-1, which consisted of the petrography of 8 core samples and core plugs of OSF, where core plugs were cut perpendicular to apparent bedding and thin sections were prepared and point-counted for 300-counts per sample. Porosity and permeability data was taken from both wells (112/15-1 & 112/19-1), where porosity is helium porosity at 1500 psi and permeability is air horizontal permeability, each taken from a total 111 core samples. Gamma and sonic wire-line data was similarly available from both wells (112/15-1 & 112/19-1).

Secondly, data for comparison to the EISB came from six wells in the North and South Morecambe fields (wells 110/2a-N1, 110/2a-8, 110/2a-7, 110/2-6, 110/2a-F1 and 110/3a-A3). All wells were drilled through the OSF and to the top SBSF (KSF and top ASF, respectively; Fig. 2). Petrographic analysis data was available for 443 samples across the numerous wells. The petrographic methodology was similar, utilizing core plugs and core samples. Porosity measurements were conducted using a Ruska parameter. Permeability is horizontal permeability calculated using dry nitrogen. Grain size data for the EISB came from well 110/2-6.

Finally, a further four thin section samples of SBSF from Fleswick Bay in West Cumbria (NX 945132) sourced from Durham University are used for an onshore EISB comparison.

Onshore analogue petrography

The depositional facies for the ASF and KSF of the SSG of the Solway Basin have been described by several previous authors (e.g. Akhurst et al. 1997; Newman 1999; Holliday et al. 2001, 2004; Brookfield 2004, 2008). The sandstones of the ASF are very fine-grained, sub-angular to sub-rounded, moderately mature and moderately to well sorted sublithic arenites with minor feldspathic subarkoses (Folk Classification; Figs 5a and 6). The detrital mineral assemblage is dominated by quartz, lithic clasts, K-feldspar, and illite and smectite matrix clays, with minor muscovite and biotite. Compositionally, the Lower and Upper ASF sampled from Cove Quarry and Glinger Burn, respectively, show a distinct compositional difference. The Upper ASF features a higher level of compositional maturity, where the average quartz abundance normalized to the total quartz, feldspar and lithic content of the Lower and Upper ASF is 82.0% and 87.5%, respectively (Figs 4 and 6).

Comparison of the ASF to the stratigraphically equivalent onshore outcrops of the RSM at Fleswick Bay, Cumbria, shows that the latter are coarser and cleaner (contain less clay), where the stacked channel sandstones contain only trace (<1%) matrix and pore filling clays (Fig. 7).

The sandstones of the KSF are fine-to medium-grained, sub-angular to well rounded, mature, well sorted, sublithic arenites to marginal quartz arenites (Folk Classification; Figs 5b and 6). The detrital mineral assemblage is remarkably similar to that of the ASF, and the only discernible difference is a general reduction in lithic clasts, feldspar and mica and an increase in detrital smectite (discerned from SEM). For both the ASF and KSF total feldspar is >95% K-feldspar with only trace plagioclase. Compositionally, the onshore KSF shows a significantly higher maturity (average relative quartz abundance of 88.3%) compared to the Lower ASF and a marginally higher maturity compared to the Upper ASF (Fig. 6). Compositional maturity between the onshore and offshore KSF/OSF (Cliff Bridge and well 112/19-1) within the Solway Basin corresponds well, where the offshore KSF/OSF (well 112/19-1) has an average relative quartz abundance of 88.8%. The analogous OSF in the EISB shows a noticeable reduction in maturity in comparison with an average relative quartz abundance of 78.5% (Well 110/2-6) (Fig. 6). Similarly, the KSF shows a prominently higher roundness and sphericity of its grains, especially those of medium sand grade.

The ASF and KSF are composed of fluvial and aeolian faces, respectively, forming lithostratigraphic boundaries (e.g. Holliday et al. 2004). The fluvial facies associations discerned here are dominantly fluvial sheetflood sandstones, characterized mainly by vertical lamination and ribbon channel sandstones, characterized by trough cross-bedded erosional based units. Both units form multistorey channel sandstone units with sheet and ribbon geometries. The fluvial ASF is also associated with flood plain facies associations, characterized by silt and mudstone deposits with rootlet traces. Playa facies are further discerned from flood plain facies because of their prominence in distal fluvial successions and in aeolian environments, where they are identified by desiccation structures and symmetrical ripples (Figs 3a and 4; Table 1). KSF aeolian facies associations identified are characteristic dune structures as described previously and damp interdune facies with wavy bedding and playa facies (Fig. 3b; Table 1).

Offshore Solway petrography

The SBSF within the offshore Solway Basin has a maximum proven thicknesses of c. 1072 m within well 112/15-1, where core sampling reports divide the SBSF into the ‘Lower St Bees’ and ‘Upper St Bees’. The ‘Lower St Bees’ is predominantly a red brown very fine to fine, locally medium grained sandstone that is very clay rich, with silt and chert beds at its base and anhydrite intermittently present throughout. The ‘Upper St Bees’ is a much cleaner very fine to fine, occasionally fine to medium grained sandstone that is very friable and ferruginous in parts. The OSF is proven to be c. 175 m thick in well 112/15-1. It is composed of red to orange-brown fine to medium, locally coarse-grained sandstones with frosted grains, that are very friable and ferruginous and contain halite and anhydrite evaporite beds at its top.

Lithological differences between the ‘Lower’ and ‘Upper’ St Bees are similar to those identified by Jackson et al. (1987) in the EISB and likely represent a change from the RSM to CSM of the SBSF, which is identifiable in the EISB and Sellafield area through geophysical data (gamma and sonic velocity values) (Fig. 2). The assumed ‘Top Silicified Zone’ was found within well 112/19-1 to prove thicknesses of 297 m and 420 m, respectively for the RSM and CSM.

Porosity and permeability

Porosities range from near 2% to 25% and from 14% to 28% for the ASF and KSF, respectively (Fig. 8a). These ranges are comparable to the equivalent facies in the EISB (Fig. 9; Meadows and Beach 1993a; Quirk et al. 1999). The permeability ranges for the sandstones are 0.5–300 mD for ‘wet’ facies (mainly ASF), including wet sandflat, wet sheetflood, and fluvial channels, whilst ‘dry’ facies range from 100–5000 mD and encompass dry sandflat and dry aeolian facies (mainly KSF) (see Newman 1999).

The extent of compaction observed in the primary outcrop-collected petrographic samples from the ASF and KSF varies markedly. The ASF has experienced significantly more compaction, with an average of 85% long contacts in the fluvial facies (Figs 5a and 7a). This contrasts with the sandstones within the KSF that have undergone considerably less compaction and have an average of 24% long contacts (Fig. 5b). Petrographic and SEM analysis has revealed no significant dissolution of grains in either formation. Point contact preservation shows a direct correlation with higher porosity preservation between the ASF and KSF (Fig. 8b).

It is recognized that the porosities of the ASF are noticeably lower than those of the KSF, this being a combination of porosity loss due to compaction and cementation effects (Fig. 10). The use of a division line between compactional and cementational porosity loss illustrates that porosity loss for most of Upper ASF is cementational, whilst porosity loss for most of the KSF is compactional (Fig. 10).

Analysis of data from the two Solway wells shows the OSF has an average porosity and permeability of 15% and 503 mD, respectively. However, significant variation was found within the values. Using core samples and geophysical wire-line values it was determined that ‘dry’ (mainly aeolian) facies had an average porosity and permeability of 18% and 1024 mD (range up to 7782 mD), respectively, and ‘wet’ (mainly fluvial channel but also playa and flood plain) facies had an average of 13% and 125 mD (range up to 833 mD), respectively. Specifically, it is reported that aeolian dune facies exhibit the best reservoir quality with a mean porosity and permeability of 19.8% and 1176 mD. Within the dry facies, permeability was compromised significantly in beds containing evaporites (dominantly anhydrite and minor halite), which additionally compromised the net:gross (N:G) of the top 26 m in well 112/19-1, giving it an overall value of 65% (Newman 1999). Despite also containing evaporites at the top of the Formation, the OSF in well 112/15-1 has a N:G of 81%. The limited data available for the SBSF shows a range of porosity and permeability values of 8% to 17% and >0.1 to 26 mD, respectively, with lower values interpreted to be related to mudstones and fine-grained floodplain facies. Permeability values are particularly low within wet facies and below the Top Silicified Zone, in the clay rich RSM. Within the upper ASF of well 112/19-1 however, which features stacked channels and a more ‘dry’ facies geophysical signature, the N:G is reported as 77%.

Grain size and sorting

Two distinct grain size populations are discerned. Both the Lower and Upper ASF are composed of very fine sand grains, which constitute c. 70% of detrital grains, with a minor proportional increase in fine sand grains within the Upper ASF (Fig. 11a). The second population displays a grain size bimodality within the KSF, with fine-sand sized detrital grains constituting 60–65% and the remainder medium-grained in size, forming a mean grain size that falls roughly between the two (Fig. 11a). Pore diameter values over the range 0.03–0.15 mm were measured and are micropores according to the classification scheme proposed by Katsube et al. (1999a, b) (Fig. 11b). The texturally mature, yet mineralogically sub-mature sandstones of the ASF and KSF are petrographically comparable to the contemporaneous SBSF and OSF of the EISB (Fig. 6; Meadows and Beach 1993b; Meadows 2006).

Much like the Ormskirk Sandstone in the EISB (Meadows and Beach 1993a), there is a tendency for grains of larger sizes to occur more commonly in the aeolian facies of the KSF, which can be found in the foresets of dunes and deflation surfaces. However, this is counter-balanced by the presence of very fine sand laminae in the same sequences. The sorting of the sandstones varies substantially from moderate to good for both the fluvial and aeolian facies, respectively, however the aeolian sandstones are well or locally very well sorted.

The difference in grain types encountered, in terms of angularity and sphericity, is significant between the ASF and KSF. The aeolian grains of the KSF are very well rounded and exhibit a high degree of sphericity, forming a major component of the facies (Fig. 5b). In comparison, the ASF grains are sub-angular to sub-rounded and exhibit a low degree of sphericity that is common for much of the fluvial facies (Fig. 5a).

The higher porosities in the ASF are encountered within the ribbon channel facies which preserve moderate reservoir quality. The lowest porosities and permeabilities within the ASF are accounted for by the very fine-grained nature of flood plain samples, which possess poor reservoir quality (Fig. 8; Table 1).

Similarly, within the offshore Solway Basin, high porosity and permeability values are found within the spherical, well-rounded grains of the aeolian dune facies, whilst lower values are found within wetter facies which are finer grained, with reduced pore size and higher clay content.

Diagenesis and reservoir quality

All previous studies of the detailed diagenesis for the SSG have been focused on the EISB (Burley 1984; Macchi et al. 1990; Meadows and Beach 1993a; Meadows 2006; Medici et al. 2019) or the Corrib and Slyne Basins to the West of Ireland (Schmid et al. 2004; Dancer et al. 2005). These two Basins were documented separately as detailed provenance studies using Pb isotopes of K-feldspars has shown they were separate, although the EISB and Solway were likely linked (Tyrrell et al. 2012). New observations and interpretations relevant to the impact of diagenesis on reservoir quality and porosity preservation, from the SSG of the Solway and Carlisle Basins, are described here.

The mineralogy of the ASF and KSF of the Solway Basin have undergone significant diagenetic alteration. Authigenic minerals occupy an average of c. 15% of the rock volume. The main diagenetic processes that have affected the sandstones include mechanical and chemical compaction, the precipitation of quartz, calcite, mixed layer smectite–illite clays, minor dolomite and kaolinite, as well as dissolution of unstable grains such as feldspar.

Kaolinite occurs as pseudohexagonal, vermicular booklets and is found in both the ASF and KSF, contributing to <1%. SEM evidence suggests kaolinite precipitation is linked to feldspar dissolution, where kaolinite is found precipitated within or around partially dissolved feldspar grains (Fig. 12a). Where kaolinite is present, it does exhibit a slight detrimental effect on porosity and tends to occur more widely in the fluvial ASF, mainly in the flood plain and sheetflood facies (Table 1).

Very little chlorite (<1%) is found in any of the studied sections unlike in other SSG sequences (e.g. Schmid et al. 2004; Dancer et al. 2005).

The red colour of the sandstones in this study reflects the presence of fine-grained haematite precipitated on the surface of the detrital grains. No noticeable differences in colour were documented between facies in either the ASF or KSF.

The major cement phases in the ASF and KSF are clay (mainly illite and mixed layer illite–smectite), calcite (non-ferroan and ferroan composition) and quartz, which are discussed in more detail below (Table 1).

Quartz cements

Quartz cements in the Solway Basin sandstones take the form of euhedral to prismatic, syntaxial macroquartz overgrowths, typically between 5–20 μm but can reach 80 μm in size (Fig. 12b). Thin section analysis clearly reveals that quartz cements precipitate directly upon, and therefore post-date, early haematite coatings which were observed coating detrital framework grains, in the ASF and KSF (Fig. 5a, b). There is a clear facies control upon the distribution of quartz overgrowths. Whilst these cements are found within the aeolian KSF (c. 4%), they are more common throughout the fluvial ASF (c. 8–12%), a finding that accords with a similar facies control on quartz diagenesis in the EISB (Meadows and Beach 1993a; Greenwood and Habesch 1997).

Calcite cements

Two stages of calcite cement are recognized within the ASF. These cements are:

  • (1) Early non-ferroan calcite cement forming aggregates of interlocking crystals directly coating detrital grains surfaces that are not found as isolated crystals within pore space. Crystals form blocky to euhedral micro-crystals and are typically c. 10–20 μm in length (Fig. 12c).

  • (2) Later stage non-ferroan and ferroan calcite cements are distributed as isolated, blocky to granular, sparite crystals and form both within pore spaces of detrital grains and on detrital grain surfaces. These cements are noticeably larger than earlier formed calcite cements and are typically c. 30 μm in length, forming well developed, euhedral crystals (Figs 12c and 13a).

Calcite cement forms a significant proportion (up to 13%) within the ASF and appears to be facies controlled. Greater percentages of this calcite cement are found in flood plain and subordinately sheetflood facies of the ASF and are only locally identified in the aeolian KSF, within damper aeolian facies (Table 1). Where found, these cements severely occlude pore space. The sandstones of the KSF only show minor evidence for late-stage calcite cementation, forming similar euhedral cements as described above within the ASF (Fig. 13a; Table 1). Calcite rarity within the KSF similarly correlates with the offshore KSF sample mineralogy from well 112/19-1, where calcite is found in abundances of <0.5%.

Ferroan dolomite only occurs in the KSF as c. 5–30 μm wide euhedral rhombs. Ferroan dolomite being the sole or dominant carbonate species is similarly seen within the offshore KSF (well 112/19-1) and within the equivalent OSF within the EISB Morecombe Bay. The ferroan dolomite is always pore-filling and can be associated with illite cements and fibres (Fig. 13b–d). BSEM, optical microscopy and CL-SEM analyses reveal that the dolomite occurs around detrital carbonate nuclei. The morphology shows little variation through the KSF and always occupies <2% of the rock by volume (Table 1). Interestingly, dolomite only forms a very minor component of the KSF but forms a major component in other SSG sequences, such as in the Corrib Field where it is the most abundant cement (up to 25% of the rock volume, with a mean of 10%) and has been found to be paramount in controlling reservoir quality (Schmid et al. 2004). Local abundances of ferroan dolomite were found in the offshore Solway Basin KSF however, with the greatest abundances located in the damp and wet sandflat facies.

Smectite and illite

Detrital smectite is abundant within the KSF, whilst minor amounts are found in the ASF, and was distinguished from mixed unresolved clays by SEM (Table 1). This is unsurprising, since previous studies demonstrate that smectite is the as this is the dominant weathering product in recent and Triassic desert environments (e.g. Weibel and Grobety 1999; Lybrand and Rasmussen 2018; Al-Juboury et al. 2020) and was likely a detrital and infiltrated clay. Authigenic smectite is found as mostly mixed-layer smectite–illite where it often forms clay coatings (Fig. 5b). This mixed layer smectite–illite additionally forms a honey-comb texture with protruding illite fibres and can be pore-occluding and pore-filling (Fig. 12d).

Illite is abundant in the fluvial ASF (c. 5%), with greater abundances found in the fine grained sheetflood and flood basin facies. Authigenic illite is observed on all clay-coated and uncoated detrital grains and on clay cutans and bridges, although rarely present on euhedral faces of quartz overgrowths (Fig. 12d). Authigenic illite tends to nucleate from a single thin veneer precipitated upon detrital grains and tend to grow as independent strands and fibres that extend into the pore space of host sandstones. Illite also commonly adapts a pore bridging habit, linking and connecting detrital framework grains. Where this is the case, pore space is not as severely occluded. Illite fibres generally have a length of c. 20 μm. This is the case both when illite occurs as isolated fibres and when illite is documented as mixed layer smectite–illite cement (Fig. 12d).

Stable isotope results for calcite and dolomite cements

Figure 14 illustrates the range of δ13C and δ18O for the calcite cement in the ASF and the ferroan dolomite cement in the KSF.

The early and later stage calcite cement have been analysed for the ASF and the distribution of data may reflect the two different generations of calcite cement. No systematic relationship to location or facies (e.g. ribbon v. sheet sandstones) is evident in the data. The calcite cement ranges from δ18O PDB −10‰ to – 3.5‰ and δ13CPDB – 5.8‰ to +0.3‰ (Fig. 15a). Ferroan dolomite only occurs in the KSF and shows no variation within the sandstone sections. The ferroan dolomite cement analysed ranges from δ18OPDB –10.6‰ to – 3.0‰ and δ13CPDB – 6.9‰ to – 0.8‰ (Fig. 14b).

Also plotted for comparison are the ranges of δ13C and δ18O from Naylor et al. (1989); Morad et al. (1998) and Greenwood and Habesch (1997), which are sampled from the Mid-Late Triassic Lossiemouth Sandstone Formation in the Moray Firth Basin, the Late Triassic Lunde Formation in the Snorre Field of the Norwegian North Sea and the OSF from the central EISB (blocks 110/13-110/15), respectively.

Burial history modelling and fluid inclusion data

The geological evolution of the Solway and Carlisle Basins is presented in Figure 15. Fluid inclusion microthermometric data show that the quartz precipitated at temperatures between 95 to 125°C from formation waters with salinities of 10–18 wt% NaCl. The measurable aqueous inclusions in the dolomite cements displayed homogenization temperatures of 100 to >135°C from formation waters with higher salinities 20–25 wt% NaCl. The salinity of formation waters encountered are broadly similar to those of the EISB where the salinity is suggested to have derived from the dissolution of Permian evaporites (e.g. Greenwood and Habesch 1997).

The burial history of the Solway and Carlisle Basins does differ from that of the EISB (Floodpage et al. 2001). This study demonstrates that the ASF and KSF reached maximum burial depth of c. 2800 m during the Late Jurassic and certainly before the influence of the Cimmerian uplift (Fig. 15). However, taking an average present-day geothermal gradient of 30.2°C km−1 would suggest the sandstones in this study only reached a maximum burial temperature of c. 85°C. This temperature is significantly less than that determined from fluid inclusion analysis. If the homogenization temperatures (Th) for the fluid inclusions are assumed to represent the minimum trapping temperature in the quartz and dolomite cements then an additional source of heat must be accounted for in the burial history.

Timing of diagenetic processes

The relative timing of diagenetic processes is presented in Figure 16a and b, as inferred from their textural relations as observed in thin section and using the SEM. Precipitation of haematite coatings onto detrital framework grains underneath quartz overgrowths shows early precipitation. As is common with recent sediments in hot, arid to semi-arid environments, this was likely a result of early near-surface diagenesis from the alteration of iron-bearing grains and smectitic clays (Burley et al. 1985), where greater haematite cementation is seen and expected within the aeolian KSF.

Early dissolution of detrital K-feldspar is common within continental red beds (Walker et al. 1978). Similar early dissolution of feldspar within the ASF and KSF is suggested as relative to the feldspar content of the sandstones (means of 7% and 4%, respectively), secondary porosity created by feldspar dissolution is minor (1% to 2%), suggesting that many K-feldspar grains did not undergo dissolution and/or dissolution was incomplete. This interpretation is also supported by SEM analysis, whereby partial dissolution of K-feldspar is common (Fig. 12a). Secondary porosity created by K-feldspar dissolution has not led to an increase in net porosity, since the dissolved material will have on the evidence of the quantity of the cements previously described, been locally re-precipitated.

Kaolinite occurrence within the Solway Basin SSG is minor (<1%) and in other SSG rocks within NW Europe it is found to be oddly absent (Schmid et al. 2004). Determining the exact source of silica for its formation is therefore difficult. Based on the evidence that feldspar dissolution is linked to later kaolinite precipitation within generated pore-space, kaolinite precipitation is suggested to have occurred after early feldspar dissolution, as is a typical open system eogenetic reaction (Fig. 12a; Bjørlykke and Jahren 2012).

In the ASF, early calcite cementation is suggested by the formation of overgrowths and cements upon detrital framework grains, as has been noted in other studies of the SSG in the Irish Sea area (e.g. Strong 1993; Greenwood and Habesch 1997). The carbon and oxygen stable isotope compositions for the ASF shows two generations of this early calcite formation (i.e. calcite I and II, Fig. 14a). Data points with a more negative δ13C and δ18O (calcite I) composition are similar to values for Triassic calcretes from the Inner Moray Firth as reported by Naylor et al. (1989). This would support calcite cements preserving an isotopic signature of the early groundwater in the fluvial sediments and reflects some recrystallization of caliche, the latter of which Meadows and Beach (1993a) also suggested for the EISB. A later second calcite cement shows more positive δ13CPDB values of −5.8‰ to +0.3‰, which is more positive than those usually associated with calcretes and may indicate a carbon source dominated by atmospheric CO2 (Cerling 1991). The more positive δ13CPDB compositions (from calcite II) also display a trend similar to that of the Triassic Lunde Formation of the Snorre Field, Norway (Fig. 14a, Morad et al. 1998). Carbon from a microbial methanogenesis oxidation of plant material and a possible contribution from atmospheric CO2 are attributed and is a probable cause for the spread of data for the ASF.

Disparity in δ13C and δ18O values between the EISB from Greenwood and Habesch (1997) and the Annan Sandstones samples, and mainly the negative δ13C signature of the EISB samples, are attributed to the effects of methanogenesis from a deep, dominantly thermally mature organic carbon source, related to hydrocarbon migration, seen within the EISB but which has not been identified within the Solway Basin (e.g. Newman 1999; Fig. 14a). Similarly, this was associated with a later stage calcite precipitate (calcite-III) in the EISB which is similarly not present in the Solway Basin (Greenwood and Habesch 1997).

These findings are therefore suggestive of eodiagenetic precipitation, pre-dating any significant compaction, similarly suggested by generally long and simple grain contacts (Fig. 5) with no evidence of sutured or complex contacts. During progressive burial, later stage calcite cements were precipitated over a more extended period of time, forming larger, well-developed crystals (Fig. 12c). This early cement acted as a nucleation site for later pore-filling carbonate phases, explaining the abundance of late-stage calcite cements in the ASF. Within the KSF, only a minor volume of calcite cement is found, and this is attributed to the absence of early calcite cement nucleation sites for later carbonate phases (Meadows and Beach 1993a).

Fluid inclusion data from quartz overgrowths show quartz cements were precipitated between 95–125°C meaning formation occurred mostly during the burial diagenesis/mesodiagensis stage (Fig. 16). Assuming an internal source, quartz overgrowths precipitated upon detrital quartz grains after the initial stage of mechanical compaction, where the bulk of a rocks intergranular volume is commonly lost (e.g. Paxton et al. 2002). Silica was sourced from the dissolution of early framework grains (predominantly K-feldspar, but also some detrital quartz). Internal silica sources have been attributed to quartz cementation in the Slyne Basin (Schmid et al. 2004). Quartz cementation is suggested to be more significant in the ASF simply because the ASF facies contain a greater amount of detrital K-feldspar.

The processes outlined above indicate early diagenetic modification following deposition, typical of red-bed style diagenesis in a semi-arid to arid setting (e.g. Burley et al. 1985). Infiltrated detrital clays percolated through the unsaturated zone, transported by groundwater and precipitated within intergranular pore spaces and around grains. Smectite is suggested to have been deposited as a detrital weathering deposit during early eodiagenesis, as is common in semi-arid to arid aeolian and, subordinately, fluvial deposits and within the Triassic. This deposition is therefore suggested to be greater in the aeolian KSF, as suggested also by detrital data (Fig. 16; Table 1; McKinley et al. 1999; Weibel and Grobety 1999; Lybrand and Rasmussen 2018; Al-Juboury et al. 2020). The transformation of this infiltrated smectite into mixed illite–smectite is suggested by the honey-comb texture, before a final transformation into the protruding pore-bridge illite fibres during progressive burial diagenesis (e.g. Weibel 1999; Stricker et al. 2016). The transformation of smectite to illite has been reported to occur below 90°C (Worden and Burley 2003), or may even form at a very low temperature between 20–30°C and depth of 500 m (Buatier et al. 1992).

Previous research has suggested that these sediments have been buried to c. 3500 m, and a temperature of c. 100°C (Newman 1999; Holliday et al. 2004). Burial history modelling undertaken in this study has demonstrated a shallower maximum burial depth of c. 2800 m and temperature of c. 85°C (Fig. 15). At this depth and temperature, it is likely that illite was sourced internally from within the SSG and represents the early alteration products of detrital and/or early diagenetic smectite, as has been suggested in the EISB (e.g. Schmid et al. 2004). Other potential potassium sources, such as feldspar dissolution and K-rich circulating pore waters are possible (Thyne et al. 2001).

Late stage ferroan dolomite forms microcrystalline rhombs (Fig. 13b) and is heterogeneously distributed throughout the KSF, as seen from local accumulations in the offshore KSF (well 112/19-1). The dolomite stable isotope data displays a variable δ18O signature, suggesting precipitation over a range of porewater isotope compositions (Fig. 14b). The positive correlation of δ18O and δ13C in the dolomite cement is similar to the trend of Naylor et al. (1989) for Triassic calcretes (Fig. 14b) and Spötl and Wright (1992) for Triassic groundwater dolocretes. Such trends can reflect evaporation of pore waters and a positive shift in the isotopic signature, although it seems unlikely that this has happened for this dolomite burial cement. Fluid inclusion data for dolomite suggest growth occurred between 100 to >135°C placing it firmly in the burial diagenesis realm. The high formation water salinities for dolomite are consistent with precipitation from basinal brines (Warren 2000). The suggested carbonate sources indicate a continued open system during diagenesis (Bjørlykke and Jahren 2012).

Burial history

The precipitation temperatures identified for the quartz and late burial dolomite does not support the burial history modelling in this study (Fig. 15). In order to achieve temperatures of 100°C with an average geothermal gradient, burial depths of >3000 m are required. Such depths are significantly greater than those modelled and either deeper burial has been experienced or elevated heat flow has occurred to explain the fluid inclusion data. Burial compaction fabrics in the ASF and especially the KSF are not compatible with such deep burial, as grains are generally point or long contacts, and pressure solution fabrics are very rare (Fig. 5). The fluid inclusion data is best explained by the pulsed migration of hot fluids through the reservoir sandstones, a suggestion also attributed to similar findings in fields in the EISB (e.g. Greenwood and Habesch 1997). Researchers have proposed that Early Tertiary Igneous activity increased geothermal temperatures and hydrothermal fluid throughout the region, explaining the raised fission-track temperature data found in the EISB, Peel and Solway Basins (Fig. 15) (Greenwood and Habesch 1997; Newman 1999; Quirk et al. 1999).

Stable isotope data and fluid inclusion analyses indicate that the later dolomite cements precipitated from evolved saline fluids, compatible with influxes of deep burial brine (e.g. Nguyen et al. 2013; Fig. 14). In the EISB, δ13C data for an equivalent dolomite cement phase indicate that the later cements with a slightly more negative δ18O signature incorporated greater proportions of 12C-enriched carbon, originating from organic maturation, during burial (Fig. 14; Greenwood and Habesch 1997). This occurrence is not shown in the Solway Basin and through burial history modelling may reflect the late time of oil migration in the EISB, postdating dolomite cement (Newman 1999).

Connection of the Solway Basin to the EISB

Within the EISB, the major northerly flowing ‘Budleighensis’ fluvial system supplied the vast majority of the basin fill during at least the early stages of Early–Middle Triassic basin evolution (e.g. Meadows and Beach 1993b; Tyrrell et al. 2007, 2012). Studies using Pb isotopic compositions of detrital K-feldspar grains from the Middle Triassic sandstones of the OSF within the EISB indicate a Variscan uplands source area (Tyrrell et al. 2012). Analysis of upstream analogous fluvial deposits from the Wessex Basin further define a dominantly Armorican source from Brittany and Normandy (Newell 2018).

The detailed petrography undertaken in this study, clearly show very similar detrital grain size distributions and burial diagenesis for EISB and Solway Basin (Fig. 6). Furthermore, paleocurrent analysis of channel facies in the ASF reveals a predominantly northward flow direction, similar to the northward paleocurrent direction of the stacked and amalgamated fluvial channels exposed at St Bees Head in West Cumbria, which sits near the boundary between the EISB and Solway Basin (Barnes et al. 1994). Based on the remarkable similarity in provenance characteristics between the EISB and the Solway Basin, regional palaeogeographical analysis based on published reconstructions (Newman 1999; Holliday et al. 2004) and fluvial palaeocurrent orientation, it seems highly likely that this northward oriented fluvial system flowed from the Cheshire Basin, through the EISB, and continued to flow north into the Solway Basin (Fig. 17a), at least throughout the deposition of the ASF. This contradicts previous studies that have suggested that the Ramsey-Whitehaven Ridge blocked any potential sediment supply from the EISB to the Solway Basin throughout the deposition of the Annan Sandstone (Fig. 1; Newman 1999; Quirk et al. 1999). No alternative local source can be invoked to explain such close similarities in provenance between the ASF and SBSF. This reconstruction complies with published accounts of SSG sedimentology in western Britain and suggests that fluvial sediment transport was sourced from areas to the south such as Wales, the English Midlands and Variscan Massifs of SW England (Audley-Charles 1970; Meadows and Beach 1993b).

There is, however, a distinct grain size difference between the fluvial and aeolian facies of the Solway Basin and their equivalents in the EISB (Meadows and Beach 1993a, 63b). Whilst samples from the SBSF and OSF of the EISB are medium-grained (mean 0.36 mm), the equivalent strata of ASF and KSF from the Solway Basin are very fine-grained and bimodal fine- and medium-grained in nature (mean 0.12 and 0.19 mm/0.28 mm, respectively, Fig. 11). This distinct variation in grain size and the variation in compositional maturity between the sediments of the EISB and those of the Solway Basin can both simply be attributed to fluvial reworking. Detrital grains become finer grained as they are transported further downstream in the fluvial system (Fig. 6), as is documented throughout other British deposits of the SSG (Medici et al. 2019).

Relationship to the ‘Budleighensis’ fluvial system

It has been suggested that the termination of this fluvial drainage system was endorheic and occurred within the distal Solway Basin (e.g. Hounslow and Ruffell 2006). Confining the termination characteristics is of regional importance to the fluvial system.

The Terminal Fan Model, presented by Kelly and Olsen (1993), is proposed to develop where evaporation rates exceed precipitation and runoff rates, establishing a moisture deficit and high levels of infiltration leading to discharge losses. A fluvial distributary zone then forms that is composed of a proximal, medial and distal zone that dissipates entirely downstream into flood basin, playa mudflat or aeolian facies at the basinal zone. Terminal Fans are presented as a feature of drylands or semi-arid to arid climates and systems that experience spatially and temporally fluctuating discharge. Facies trends show a decreasing down-fan grain size and channel body thickness, an increase in siltstone content and a shift to muddy flood basins (Friend 1977; Tunbridge 1984; Kelly and Olsen 1993).

Despite critiques of the Kelly and Olsen (1993) Terminal Fan Model regarding its occurrence in nature, as well as the lack of a distinct sedimentary facies succession (North and Warwick 2007), ‘terminal fan’ or ‘terminal fluvial systems’ models, based originally on the Terminal Fan Model, have continued to be adopted (e.g. Masrahy and Mountney 2015). Particularly, such models have been applied within analogous and/or contemporaneous Permo–Triassic dryland river successions to the Solway Basin SSG (Cain and Mountney 2009; McKie and Williams 2009; McKie 2014).

Within the ‘terminal fan’ or ‘terminal fluvial systems’ framework, the ASF can be categorized as the distal component of the distributary zone, with smaller channels that largely still dominate, alongside subordinate sheetflood, flood plain fines and aeolian facies (Fig. 18a). Downstream fining, increased clay/mud and silt content compared to the EISB (Fig. 7) and the presence of flood plain and playa type settings which have been described in detail at various localities within the onshore Solway Basin support this (Brookfield 2004, 2008; Holliday et al. 2004). At Cove Quarry, stacked channels are still present within the ASF, alongside abundant evidence of desiccation and ephemerality which suggests a distal distributary zone rather than a basinal zone, with channels representing periods of increased runoff in the hinterland. The EISB likely represents the medial component of the distributary zone, where its dominant stacked multistorey channel sandstones and interchannel sheetflood facies, as well as minor flood basin mudstones and aeolian components correspond well with the model (Kelly and Olsen 1993; Meadows and Beach 1993a; Cain and Mountney 2009). Aeolian reworking of interchannel sandstones is also characteristic of the model and is documented within the EISB (see Mckie and Williams 2009).

The question then remains as to why does the transition from the distal fluvial facies of the ASF to the dominantly aeolian facies of the KSF occur within the Solway Basin whilst fluvial deposition continues within the analogous OSF of the EISB?

The KSF suggests a transition from the distal to basinal zone within the Terminal Fan Model, where dry aeolian facies become dominant once fluvial influence wanes (Kelly and Olsen 1993) (Fig. 18b). Cain and Mountney (2009) propose an analogous transition within the Permian Organ Rock Formation within the Paradox Basin where it is proposed that the terminal fan system transitions to increasingly distal as the fluvial system retreats towards the hinterland and the basinal aeolian dune system advances up the system. The aeolian petrography is characterized by quartz arenites to sub-litharenites which show a reduction in mica, clay and feldspar whilst remaining texturally similar to the fluvial units due to reworking. Petrographically this is equivalent to the transition from the ASF to the KSF (Fig. 6).

The basinal region also characteristically includes flood plains, playa mudstones, evaporites and channels during extreme flood events, as well as aeolian sandstones (Kelly and Olsen 1993). This corresponds well to the blocky mudstone/claystone intervals, evaporite beds and periodic channels sands with wet geophysical characteristics found at the top of the SSG in both offshore Solway Basin wells (Fig. 18b).

One explanation for this transition is that at the time of KSF deposition, the Ramsey–Whitehaven ridge acted as a barrier to the northward oriented fluvial systems that were previously responsible for the deposition of the ASF. Evidence for this is considerable. As previously discussed, wells in the northern parts of the EISB similarly lack significant fluvial deposition (Meadows 2006). This westward shift is explained by regional basin evolution within the East Irish Sea area. Early rifting during the deposition of the SSG within the EISB allowed channels to flow northwards into the Solway Basin. However, as indicated by regional seismic lines (Quirk et al. 1999; Floodpage et al. 2001), the transition to thermal sag later during the deposition of the SSG re-established the Ramsey–Whitehaven Ridge as a barrier to this previously northward flowing fluvial system. The river then flowed westwards into the Peel Basin (Fig. 17b). Furthermore, dipmeter data in the OSF of the Morecambe Field demonstrates eastwards dipping cross bedding that have been tied to these channels (Cowan 1993). Finally, the increased aeolian deposition within the CSM and OSF within the Sellafield region is thought to be due to tectonism diverting the river system away from the eastern margin (Jones and Ambrose 1994).

Alternatively, a retraction of the ‘Budleighensis’ fluvial system itself could be the cause of an expansion of the aeolian basinal region within the Solway during the deposition of the KSF. This is supported by the regional upward increase in aeolian deposits and dry facies, as well as evidence of widespread regional evaporitic deposition throughout England and SW Scotland during the later stages of SSG deposition (Ambrose et al. 2014). This trend has been proposed to represent a period of increasing aridity in England and a more terminal character of the ‘Budleighensis’ fluvial system during OSF equivalent deposition (Mckie and Williams 2009; Tyrrell et al. 2012; Newell 2018).

Medici et al. (2019) quantifies this change, with aeolian facies proportions of 9%, 40% and 47% during ASF-equivalent deposition in the Worcester–Staffordshire Basins, Cheshire Basin and northern onshore EISB-Carlisle Basins, respectively. This changes to 35%, 60% and 100% during the KSF period (see Figure 4 in Medici et al. 2019).

Cain and Mountney (2009) propose the distal shift within the Permian Organ Rock Formation was due to an increase in arid conditions and denudation of the primary fluvial source leading to an increase in downstream discharge losses. The Permian Organ Rock Formation is a similar system that was likewise already experiencing strongly seasonal discharge change. Therefore, a regional drying of the ‘Budleighensis’ fluvial system could similarly be responsible for increased aeolian deposition. Equally, the aeolian shift could be a consequence of a reduction in gradient of the fluvial system (Hounslow and Ruffell 2006). The denudation of Variscan mountain chain is known to have continued throughout the Triassic (Warrington and Ivimey-Cooke 1992).

Decoding the exact cause of the facies change is challenging, with the impact of allocyclic factors difficult to distinguish within continental successions (Péron et al. 2005). The ‘Budleighensis’ fluvial system has already been characterized as sensitive to changes in the water table, with dry periods in the EISB consisting of only small streams, in a geological period prone to significant fluctuations in precipitation (Meadows and Beach 1993b; Sellwood and Valdes 2006). Any strengthening of the Pangean monsoon would cause increased seasonality and arid expansion within the continental realm (Parrish 1993). Meanwhile, any potential migration of the humid zone from the Variscan source region would likely significantly affect discharge at the terminal end of the river system (Newell 2018). Periodic return of fluvial facies to the Solway wells during the deposition of the OSF disputes suggestions of a complete separation of the Solway Basin from the EISB due to tectonism along the Ramsey–Whitehaven Ridge.

Aeolian provenance

The provenance for the aeolian KSF is problematical. Given the petrographic similarity between the ASF and KSF, it seems likely that the very fine-grained detrital grain population represents unroofed and reworked material, directly derived from the underlying ASF, as Brookfield (2004, 2008) suggests. Notwithstanding this, the medium-sized grains of the bimodal KSF grain population must be accounted for, as the size of this grain population appears to preclude the possibility of having been derived by wind reworking of the ASF.

Provenance for this period is disputed in literature. Brookfield (2008) states that there is little evidence for present day structural highs, such as the Southern Uplands and the Lake District, affecting Late Paleozoic sedimentation within the sedimentary succession. The Carboniferous palaeo-North Pennines (Askrigg and Alston Blocks) have alternatively been presented as an area of topographic relief during the deposition of the KSF, with the Millstone Grit Group specifically suggested (Meadows and Beach 1993a, 63b). Tyrrell et al. (2006) disputes this however, citing a radiogenic Pb population found within the Millstone Grit Group that is not present within the EISB. Instead, further Pb analysis of K-feldspars from the EISB (Tyrrell et al. 2012) suggests the Shap Granite as a contributory source, which therefore potentially applies to the Solway Basin. However, Tyrrell et al. (2012) also accepts an absence of K-feldspar data from the OSF of the Solway Basin. The Pennines/Durham area are suggested to have supplied the majority of sediment to the analogous OSF in the Sellafield area of west Cumbria, an area similar to the Solway Basin in its depositional model and petrography (Jones and Ambrose 1994).

Further research on the provenance of K-feldspars from the Solway Basin is required to determine if the palaeo-north Pennines/Durham area served as an aeolian source region during KSF depositional time and supplied sediment for the northern coarser-grained aeolian dune sediments (Fig. 18). Given that the dominant regional palaeowind was oriented from an East to West direction, this option is feasible (Figs 11 and 17).

Whilst Brookfield's (2004) suggestion of the North Sea region as a potential source area similarly fits the palaeowind direction, the uplifted palaeo-Pennines formed an area of positive relief and likely impeded any wind-blown sediment from the North Sea area (Figs 1 and 18). In addition, the most likely candidate for a source given the palaeo-wind patterns and the stratigraphy of the Central North Sea are the contemporaneous mudstones of the Smith Bank Formation (Goldsmith et al. 2003).

Reservoir quality

The sedimentary facies associations in the SSG of the Solway Basin exhibit differing reservoir quality, with the greatest disparity occurring between the flood basin/playa samples of the ASF and those of aeolian dune sandstones in the KSF. The facies control upon reservoir quality is important, but also grain size, sorting and diagenesis (both cementation and compaction) have played a key role within the Solway Basin.

The fine-grained nature of the fluvial sediments within the Solway and Carlisle Basins are due to the basin being situated at the terminal end of the ‘Budleighensis’ fluvial system. The variable porosities of the ASF (2% to 25%, with a mean of c. 13%) are a direct result of the heterogeneous nature of facies distribution because of this distal fluvial location. When ribbon fluvial channel facies and stacked channel facies are deposited, moderate porosity is preserved. A combination of matrix clays, cementation and the fine-grain size however has severely restricted the reservoir potential of the flood basin/playa and sheetflood sequences (Table 1). As a result, the upper ASF, which features fewer interbeds of flood plain and sheetflood fines has preserved the best reservoir quality of the ASF and a high N:G (77%) in well 112/19-1 (Fig. 8; Table 1).

The main matrix clay within the ASF is illite, which tends to adopt a pore-bridging habit; severely lowering permeabilities and increasing pore tortuosity (Figs 8, 9 and 12d) The widespread occurrence of illite has been well documented within the EISB (Woodward and Curtis 1987; Macchi et al. 1990; Meadows and Beach 1993a) and in this case, tends to form in box work or honeycomb type textures, severely bridging and infilling intergranular pores. However, illite, by possessing a dominant pore-bridging morphology within the ASF of the Solway Basin, is volumetrically less significant than examples of platy illite documented within the EISB, and its deleterious effect upon reservoir quality is far less.

Porosity loss from cementation has similarly reduced the reservoir quality of the ASF, with both calcite cementation and greater quartz cementation within the ASF compared to the KSF. Calcite cementation is identified as the primary mechanism of porosity loss, constituting up to 13%, with two stages identified (calcite I and II), whilst only minor quantities of the late-stage cement (calcite II) are found within the KSF. Calcite cementation is also tentatively found to be present in greater quantities within the flood basin and subordinately sheetflood facies of the ASF (Table 1).

In the worst affected instances (those where extensive calcite cements combine with illite), both porosities and permeabilities have been reduced to the level where these intervals are effectively non-reservoir, and these facies exhibit a permeability range an order of magnitude lower than the KSF aeolian sandstones (Newman 1999).

Excellent reservoir quality is maintained within the aeolian KSF, where the highest porosity and permeability values are preserved in the aeolian dune facies, where variation away from these higher values in the offshore KSF are found within the fluvial and ‘wet’ interdune facies (Fig. 9).

Variations away from the mean porosity at Bridge Cliff are largely due to the inherent variations in the quality of aeolian sands in response to differences in grain size and sorting as a result of pinstripe lamination and dune bounding surfaces (see Fig. 3b). Studies of packing parameters have identified that sands composed predominantly of high sphericity grains have lower primary porosities at the time of deposition compared to those grains of low sphericity (e.g. Dickin 1973). High sphericity however in this case has not impeded porosity preservation within the KSF when compared to the less porous ASF, which features less spherical grains, and has therefore not been the determining factor in the porosity preservation of these Formations.

The smectite and mixed layer illite–smectite grain coats abundant within the KSF (Fig. 5b) are suggested to represent a dominant method of porosity preservation. The presence of these detrital grain covers masked potential nucleation sites, preventing later stage growth of cements and crystals upon quartz grain surfaces. This phenomenon, of early infiltrated smectite later forming porosity preserving illite–smectite grain covers has been documented in many studies and is widely regarded as an excellent method of preserving primary intergranular porosity (e.g. Storvoll et al. 2002; Ajdukiewicz and Larese 2012). Specifically, Tang et al. (2018) reported that aeolian and interdune facies in the Upper Devonian deposits in the North Sea saw increased porosity preservation by this method as they featured greater amounts of mechanically infiltrated smectitic-rich clay bearing water at deposition. This process inhibited quartz cementation in the aeolian deposits and the fluvial deposits, which lacked illite–smectite grain coatings, saw poorer porosity preservation. As the KSF features a greater amount of smectite and this is characteristic of aeolian deposits in semi-arid to arid environments, then this could partially explain the increased quartz cementation and decreased porosity preservation in the fluvial ASF compared with the aeolian KSF (Table 1). It is likely that the present illite grain coatings still played a role in inhibiting quartz cementation within the ASF however the effects of abundant pore-occluding and bridging illite and calcite cementation detracted from much of this.

It has also been recognized that infiltrated clay coatings on framework grains may act as nucleation sites for the precipitation of other authigenic clays (Matlack et al. 1989), which aid in the complete coating of detrital framework grains to further prevent quartz precipitation. Although these clays do occlude pore space, the minor degree to which this is the case (<5%), far outweighs the potential porosity which may be lost by late stage authigenic quartz precipitation, up to 12% within fluvial sandstones in the EISB (Meadows and Beach 1993a).

It is assumed that the higher porosity, lower number of long grain contacts of the KSF and minimal compaction is due to an early framework stabilizing quartz cement. This accounts for the greater number of grain-to-grain point contacts and corresponding higher porosity compared to the ASF (Fig. 8b). The relatively minor degree to which this early cement has helped to resist compaction in the ASF has meant that ‘wet’ facies were more susceptible to compaction during early burial (Figs 5 and 8). At the initial critical stage of burial, this early cement phase also helped to resist the effects of compaction, lithifying the sandstones and allowing detrital grains to maintain c. 76% ‘point’ contacts (Fig. 8b). Consequently, pore throats have remained open, pore connectivity is maintained, and primary porosity is preserved.

The presence of this early framework stabilizing quartz cement in the KSF but absence in the ASF may be in part due to the cleaner composition of the aeolian sub-quartz arenite KSF, leading to fewer reactants being available during the onset of diagenesis. Alternatively, the grain contact and compaction differences could be a result of a difference in detrital composition between the ASF and KSF, however considering their remarkably similar composition a stabilizing quartz cement is perhaps more likely (Table 1).

Overall, whilst porosity is variable (14% to 28% and a mean porosity of 21%) in the aeolian KSF, from the samples taken and offshore correlation, the values are equivalent to, or better, than porosities of the equivalent strata in the EISB (Table 1; Figs 8 and 9). The basinal location of the KSF has meant that increased aeolian dune deposition has occurred, which preserve better quality reservoir sands than the mixed fluvial-aeolian sands of the EISB. A consequence of this distal location however is that the playa facies also present significantly reduce porosity and permeability and cause cause lateral anisotropy, as seen with the net:gross difference between wells 112/15-1 and 112/19-1 (Fig. 18b).

Reservoir utility for carbon storage

Hydrocarbon exploration of the Solway Basin was unsuccessful in identifying any prospects (Newman 1999; Floodpage et al. 2001). This was due to extensive erosion of the Carboniferous source rock as a result of Variscan Uplift (e.g. Newman 1999), with proven thicknesses of just c. 330 and c. 74 m in wells 112/15-1 and 112/19-1, respectively. Complete erosion of the Carboniferous Westphalian Group was found in well 112/15-1, with Namurian and Dianantian Group rocks lacking in any source rock lithologies. As a result, the Solway Basin features an excellent reservoir system with a thick MMG caprock that, unlike its analogous EISB counterparts, is uncharged by hydrocarbons (Fig. 19). This presents a possible novel opportunity to explore the use of the Solway Basin for CO2 storage. As an uncharged reservoir with no history of hydrocarbon (or other commercial) utility the Solway Basin reservoir is classed as a ‘saline aquifer’.

Hydrocarbon and CO2 storage systems are both characterized by the need for a porous and permeable reservoir unit with an overlying thick and laterally extensive impermeable seal unit. Additional characterization for CO2 storage reservoirs then must take into accounts the conditions required for successful physical and geochemical trapping of CO2. For CO2 storage a structural trap may not be required (e.g. Utsira Formation at Sleipner), just a stratigraphic trapping barrier to halt the migration of an upward migrating CO2 plume for a short period (<100 years) whilst more permanent trapping mechanisms take effect (Burnside and Naylor 2014). Physical residual/capillary trapping is dependent on pore-scale capillary heterogeneity and has been found to increase storage security significantly and rapidly by immobilizang CO2 and therefore reducing pressure on the caprock, working at timescales of <100 years (Ajayi et al. 2019). Geochemical solubility trapping, which dissolves buoyant supercritical CO2 into the formation water brine, is dependent on the solubility of CO2. It is therefore dependent mainly on pressure, temperature and salinity, where solubility increases with increasing depth but decreases with increasing temperature and salinity (Benson and Cole 2008). The timescale for such reactions varies but is likely in the magnitude of hundreds to thousands of years (Burnside and Naylor 2014). Mineral trapping then converts CO2 into the solid mineral phase, where reactions are dependent on reservoir conditions and the minerals present within the host rock and brine. The process is very advantageous for safe storage but is extremely slow and occurs over thousands to millions of years (Bachu et al. 1994).

For successful CO2 injectivity, the reservoir must be buried to a depth great enough that the temperature and pressure keep the CO2 in a supercritical phase (>31.1°C and >7.38 MPa), where it possesses a higher density than gaseous CO2 but still flows as a gas. At regular geothermal gradients this depth is c. 800 m depth (Johnson et al. 2004). Reservoir depth should be great enough as to not affect groundwater resources and adequately saline (total dissolved solids >10 000 ppm) as to be unsuitable for any other purpose (Benson and Cole 2008; Ajayi et al. 2019).

The current paradigm for CO2 storage site selection is that disused hydrocarbon sites’ proven ability to store hydrocarbons will similarly make them suitable for carbon storage with proven trap mechanisms reducing risk (Gammer et al. 2011). The presence of infrastructure and a bias of data being available for these regions is also more economical. These factors are inversely the disadvantages of exploiting saline aquifer targets. As such these sites are the primary CO2 storage targets in Britain despite saline aquifers making up the majority (c. 88%) of overall storage potential (ETI 2011). However, disused hydrocarbon sites have inherent disadvantages also. The integrity of legacy abandoned wells within the injection region must be investigated as these may have degraded structural integrity and could pose as sites of potential CO2 leakage (Ajayi et al. 2019). The likely presence of residual hydrocarbon such as natural gas will compete for pore-space with CO2 and is expected to have significant impacts upon multiphase flow during CO2 injection and affect residual trapping (Saeedi and Rezaee 2012).

The Solway Basin as a saline aquifer CO2 storage site has numerous advantages. In this study it has been identified that the reservoir quality is excellent. Porosity and permeability are better than or comparable to the EISB where CO2 injection is planned by 2025 as part of the Hynet North West Project. This project will target the analogous stratigraphy of depleted gas fields such as Hamilton, where the OSF reservoir target features a mixture of mainly mixed aeolian (dune, sandsheet and sabkha) and stacked fluvial sand facies with subordinate playa margin shale and playa lake facies (ETI 2016; GOV.UK 2021). The reservoir units in the Solway Basin are also located deeper than the c. 800 m TVDSS threshold for storing CO2 in the supercritical phase, where the primary store KSF is located within the optimum storage depth of 800–1000 m, after which there is no significant benefit of storing CO2 at greater depths (Ennis-King and Paterson 2002; Gammer et al. 2011).

The Solway Basin is classed as an ‘open’ saline aquifer as no specific stratigraphic/structural traps are identified, unlike ‘closed’ disused hydrocarbon reservoirs, and therefore CO2 can migrate laterally without boundaries. As the reservoir generally dips toward the basin centre, buoyancy driven up-dip lateral migration toward the basin margins could risk leakage (Fig. 20). However, as proven by Sleipner, structural confinement is not essential for secure CO2 storage and updip migration is limited by residual gas trapping and the rate of CO2 dissolution into formation pore waters, which prevents migration further than a few tens of kilometres (Ennis-King and Paterson 2002). Furthermore, fault-block structural traps have been tested at well sites 112/15-1 and 112/19-1, in addition to trench-collapse structures being found around the perimeter of the basin where the MMG seal is downfaulted against the reservoir unit, which would form barriers to lateral migration out of the basin (Newman 1999; Floodpage et al. 2001). A theoretical injection site could include an area around one of the two hydrocarbon wells where the structural traps and collapse-trenches could be utilized, where well 112/15-1 has a greater reservoir and caprock thickness. Alternatively, injection nearer the basin centre features the thickest extent of reservoir and caprock, allowing for greatest storage capacity and security, if lateral migration to the basin margin can be ruled out.

The Solway Basin has good potential for CO2 trapping. Physical trapping will be extensive due to the thick and very fine grained MMG caprock which has alternating layers of low permeability muds and excellent halite salt seals (Newman 1999). Residual trapping could be extensive due to heterogeneity between dry aeolian sands and wet interdune and playa deposits in the KSF, which could provide baffles and permeability barriers alike to the thin mudstone interbeds within the Utsira Formation in the Sleipner Project which trapped the bulk of the injected CO2 before it reached the reservoir–seal interface (Chadwick et al. 2004). This heterogeneity makes the Solway Basin KSF analogous to the ‘wet aeolian deposystem’ of the natural CO2 reservoir Middle Jurassic Entrada Sandstone Formation in Utah (e.g. Newell et al. 2019). Solubility trapping potential with the Solway Basin could be restricted by the shallow and low-temperature reservoir conditions and highly saline pore waters. Mineral trapping will be aided by the fine grain-size which will increase reaction rate but has not hindered porosity and permeability, and the abundance of rock lithics and feldspars, both of which serve as reactants for mineral-trapping reactions (Watson et al. 2003). Feldspars further provide secondary porosity and permeability generation from dissolution with acidic CO2-enriched pore waters, as shown by natural CO2 analogue sites (e.g. Teranaki Basin, New Zealand; O'Neill et al. 2020).

As a failed hydrocarbon target, the Solway Basin features very little risk of leakage and/or integrity failure from legacy exploration or production wells which are abundant within the EISB. Furthermore, the lack of hydrocarbon extraction avoids the risks of geomechanical failures of the caprock, jeopardising the structural trapping of CO2 through the reactivation of faults and induced seismicity because of repressurizing an already depleted reservoir (Ajayi et al. 2019). A lack of hydrocarbon charging means there are no disadvantages linked to residual hydrocarbon presence or continued generation expected.

A significant challenge is the lack of proximity to major sources of CO2 compared with the EISB. The closest large point sources are sporadic industry and power stations between the Cumbrian coast, Carlisle and Dumfries, with further coastal sources around Morcombe Bay, Lancaster and Blackpool (UK Emissions Interactive Map 2022). CO2 collection from these point sources and subsequent transport to a CO2 storage hub in the Solway Basin could be established, such as with the Norwegian Longship Project (Northern Lights 2022). Alternatively, the Solway Basin could serve as a further storage facility to the HyNet North West Project, where CO2 already plans on being sourced by CO2 shipping (Hynet 2021).

The Solway Basin also faces the issue of a lack of regional understanding and infrastructure due to a lack of previous economic interest, requiring further investment before injection, a problem all saline aquifers face despite their dominance in CO2 storage potential (Bentham et al. 2014). The Solway at least benefits from extensive research conducted on its analogous reservoir and seal units within the hydrocarbon producing EISB. In the short-term, CO2 storage in the Solway Basin may risk increased expenditure, however initial investment at this point will further our understanding of saline aquifers which will make up the majority of national and international CO2 storage sites. Additionally, with storage safety such a significant concern to stakeholders and the IPCC (2005) Special Report stating that reservoirs must be ‘likely’ to retain 99% of stored CO2 for 1000 years, stable long-term sequestration should be a priority, for which disused hydrocarbon fields may not be suitable.

Detailed petrography and field sedimentology reveal that the extensive ‘Budleighensis’ river system did exit north into the Solway and Carlisle Basins from the EISB during the deposition of the Lower Triassic Annan Sandstone Formation. This endorheic fluvial system was likely terminal and when applied to the Terminal Fluvial Model of Kelly and Olsen (1993), can be ascribed to the distal distributary zone, with a mix of intermittently ephemeral channels, sheetflood and flood plain facies. During the deposition of the overlying KSF, the Solway Basin and its onshore extension transitioned into the basinal zone, with deposition of aeolian dune and desiccating playa lake facies. Likely this transition was a result of tectonism, centred around the Ramsey-Whitehaven ridge, causing a preferential fluvial migration away from the Solway Basin into the Peel Basin. Evidence of an overprinted regional contraction of the entire fluvial system is additionally speculated.

Porosity, permeability and diagenetic properties were found to be controlled by the distribution of facies, driven by the dynamics of the ‘Budleighensis’ fluvial system.

Fluvial channels of the ASF preserve moderate porosities (c. 13%). Inherent facies heterogeneity in the form of fluvial channel facies but also sheetflood, flood plain/playa facies throughout has resulted in highly variable reservoir quality, especially in fine-grained, clay rich, sheetflood and flood plain/playa facies, which preserve poor reservoir quality. Porosity preservation was hindered by presence of quartz overgrowths, two phases of carbonate cementation and authigenic clays but was aided by the presence of illite which inhibited later stage burial diagenetic cements. Illite, however, decreases permeabilities quite significantly. Where calcite precipitation combines with illite cements, porosities and permeabilities are affected to the point where these intervals are non-reservoir. Deleterious effects of quartz overgrowths have a greater impact in the ASF than the KSF.

Despite the finer grain size of the Solway Basin fill, excellent porosity and permeability is preserved within the aeolian KSF. This is due both to primary sedimentological characteristics such as bimodal grain size distribution and packing, in addition to early diagenetic processes such as early calcite cement and smectite precipitation which created an early framework stabilizing cement and helped to prevent later stage precipitation products occluding pore space.

This paper presents a case for the Solway Basin as a possible saline target for CO2 storage and for exploring the reservoir properties of distal fluvial or basinal aeolian sequences. The excellent porosity and permeability and lack of legacy well leakage risks present it as credible and stable sub-surface reservoir for long-term sequestration. This potential is jeopardised somewhat by the lack of present infrastructure that will risk increased expenditure. Overall, this work further constrains the carbon storage saline aquifer catalogue of the British Isles, of which carry the greatest storage potential.

We thank former Centrica Energy for permission to use petrographic data from North and South Morecambe Fields of the East Irish Sea. This paper contains information provided by the Oil and Gas Authority and/or other third parties.

JRM: conceptualization (lead), data curation (lead), formal analysis (equal), investigation (lead), methodology (equal), resources (supporting), visualization (equal), writing – original draft (lead), writing – review & editing (equal); SJJ: conceptualization (supporting), formal analysis (equal), methodology (equal), resources (lead), supervision (lead), visualization (equal), writing – review & editing (equal); NSM: conceptualization (supporting), formal analysis (equal), methodology (equal), supervision (supporting), writing – original draft (equal), writing – review & editing (equal); JGG: conceptualization (supporting), formal analysis (equal), methodology (equal), supervision (supporting), visualization (equal), writing – review & editing (equal)

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

The offshore well datasets analysed during the current study are available from the Oil and Gas Authority, through the National Data Repository (https://ndr.ogauthority.co.uk). All other data generated or analysed during this study are included in this published article.

This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 License (http://creativecommons.org/licenses/by/4.0/)