New insights into the structure, geology and hydrocarbon prospectivity along the central-northern Corona Ridge, Faroe – Shetland Basin

: The Faroe – Shetland Basin (FSB) is one of the only significant exploration frontiers remaining on the UK Continental Shelf. Over half of the basin ’ s discovered reserves and resources lie along two intra-basinal highs, the Corona Ridge and the Rona Ridge. In contrast to the Rona Ridge, the central-northern Corona Ridge has received much less attention. To reveal new insights into the geology, structural configuration and hydrocarbon prospectivity of the central-northern Corona Ridge, we analyse 3D seismic data and data from the exploration wells 213/23-1 (Eriboll), 214/21a-2 (South Uist) and 213/25c-1V (North Uist). This study extends the Colsay T40 – T45 sub-and intra-basaltic play concept from the Rosebank Field NE along the Corona Ridge, at least into well 213/23-1. Analysis also suggeststhat no Triassic strata are present within well 213/23-1, challenging the previous understanding of Triassic distribution within the central FSB. Our findings show that the central-northern Corona Ridge is structurally complex, comprising a series of discrete basement-bounding faults, downflank fault terraces and faults that are oblique to the dominant NE – SW-striking structural fabric of the FSB.

The Faroe-Shetland Basin (FSB), located west of the Shetland Isles on the NE Atlantic continental margin (Fig. 1), is one of the only remaining significant exploration frontiers on the UK Continental Shelf (UKCS), and will also arguably be one of the most important areas to contribute to indigenous oil and gas supply in the next 40 years (Austin et al. 2014;Ellis and Stoker 2014). As of September 2021, the region is estimated to hold c. 3.7 Bboe (billion barrels of oil equivalent) of discovered reserves and contingent resources, around 30% of the UK's total (2P and 2C: North Sea Transition Authority 2019). In 2019, the UKs North Sea Transition Authority (NSTA) estimated mean yet-to-find resources of 5.8 Bboe within the West of Shetland region, totalling 38% of the estimated yet-to-find resources across the entire UKCS (encompassing lead, prospect and play-level prospective resources: North Sea Transition Authority 2019).
Particularly prospective areas of the FSB include intra-basinal highs; in total, over half of the discovered reserves and contingent resources in the basin are situated on the Corona Ridge and the Rona Ridge ( Fig. 1) (Austin et al. 2014;Scotchman et al. 2016; North Sea Transition Authority 2019). Significant discoveries, some now developed and in production, have been made along the Rona Ridge ( Fig. 1b) within early Cretaceous shallow to marginal-marine (Victory Field: Goodchild et al. 1999) and marine (Edradour Field: Clark et al. 2020) sandstones, fractured Precambrian crystalline basement rocks (e.g. Clair Ridge: Robertson et al. 2020), and Devonian-age siliciclastic rocks of the Lower Clair Group (Clair Field: Robertson et al. 2020) (Fig. 2).
Exploration focused along the Corona Ridge has provided encouraging results (e.g. Rosebank and Cambo fields, and Lochnagar and Blackrock discoveries: Fig. 1b). However, in contrast to the Rona Ridge (on which the Clair and Lancaster fields are developed), as of August 2022, none of the discoveries made along the Corona Ridge have been sanctioned for development. The 300 MMboe (million barrels of oil equivalent) (recoverable) Rosebank Field, reservoired within the upper Paleocene-lower Eocene-aged fluvial-deltaic intra-and sub-basaltic Colsay Member (T40-T45) sandstone units, is currently awaiting a final investment decision (FID: estimated in 2022), 18 years since its discovery in 2004 (Siccar Point Energy 2020a). Planned sanction of the Cambo Field (30 km SW of Rosebank) was expected within 2022, with first oil projected for 2025, which will be 23 years after the discovery of 800 MMboe (in place) within Hildasay Member (T45) sandstones in 2002 (Siccar Point Energy 2020a, b).
Whilst hydrocarbon shows and discoveries prove a working petroleum system is present both along the Corona Ridge and within the basins bound by the Corona Ridge, as highlighted by Scotchman et al. (2006Scotchman et al. ( , 2016, Mark et al. (2018a, b) and Gardiner et al. (2019), many aspects of the region's petroleum geology are still not fully understoodas emphasized by recent disappointing results of wells drilled along the Corona Ridge (Lochnagar appraisal, well Within the last two decades, exploration focus and success along the Corona Ridge has been largely confined to the post-rift Paleocene-Eocene-aged reservoirs around the Rosebank Field, located along the southern part of the Corona Ridge (Smallwood and Kirk 2005;Loizou 2014). Consequently, published literature of the structure, evolution and geology of the central-northern Corona Ridge is sparse, especially when compared with the prolific Rona Ridge. This is largely the result of basaltic lava cover that extends from the Faroe Islands into the FSB over much of the Corona Ridge, resulting in challenging sub-basaltic seismic imaging (Schofield and Jolley 2013). Published research has largely overlooked the possible distribution of intra-and sub-basaltic Paleocene-Eoceneaged (sequence T40-T45) reservoirs NE of the Rosebank Field and older pre-rift (Devonian-Jurassic) strata present along the centralnorthern Corona Ridge. Many of the exploration wells drilled along the Corona Ridge were drilled within the late 1990s and early 2000s (e.g. 213/23-1, 214/04-1 and 214/09-1: Fig. 1b), prior to key advancements in the understanding of the evolution and lithostratigraphy of the FSB and specifically the Corona Ridge.   and Mark et al. (2018b) with structural lineaments (also referred to as transfer zones) and intrabasinal highs modified from Ellis et al. (2009) and Hardman et al. (2018).

Fig. 2.
Faroe-Shetland Basin tectonic history and lithostratigraphy with various aspects of the basin's petroleum system and associated discoveries. Numerical ages and chronostratigraphy are taken from the International Chronostratigraphic Chart 2020 (Cohen et al. 2013). The figure and lithostratigraphy are modified from Scotchman et al. (2006) and stratigraphic units (groups) from . Plate tectonic events and UK Atlantic margin events are based on Scotchman et al. (2006),  and Schöpfer and Hinsch (2019). The sea-level curve is modified from Haq et al. (1987). Potential and proven source rocks are from both Scotchman et al. (2006) and . Play types are modified from numerous authors individually cited (Goodchild et al. 1999;Herries et al. 1999;Duncan et al. 2009;Larsen et al. 2010;Trice 2014;Ogilvie et al. 2015;Watson et al. 2019). Note: wells and discoveries located along the Corona Ridge are shown with an asterisk (*); Eriboll (213/23-1) contained only oil shows and is added for context. At the time of publication, over 100 MMboe of condensate and 656 bcf (billion cubic feet) of stranded gas lies undeveloped along the central-northern Corona Ridge, currently viewed as uneconomic for development without a substantial discovery nearby to warrant the development of gas infrastructure as a tieback option (P 50 postdrill in-place volume estimates: Tobermory 464 bcf, ExxonMobil 2000; Bunnehaven 192 bcf, ExxonMobil 2001;North Uist 104 MMboe, CNOOC Ltd 2017).
The objective of this paper is to provide new insights into the structural configuration, geology and hydrocarbon prospectivity of the central-northern Corona Ridge, which lies within an area in need of a substantial discovery, through the interpretation of 3D seismic data integrated with post-well analysis of three exploration wells (213/23-1 (Eriboll), 214/21a-2 (South Uist) and 213/25c-1V (North Uist): Fig. 1b).
By analysing well data, we interpret that no Triassic-age strata are present within well 213/23-1 (Eriboll) and possibly the central FSB entirely, challenging the previous understanding of the distribution of Triassic strata within the centre of the basin Stoker et al. 2014Stoker et al. , 2017b. We also extend the Colsay Member intra-and sub-basaltic play concept NE along the Corona Ridge from the Rosebank Field into well 213/23-1 (Eriboll). Our findings also suggest that the central-northern Corona Ridge is structurally complex, with evidence of north-south-and NNW-SSE-striking normal faults that are oblique to the basinal NE-SWstriking structural trends.

Geological history
The Faroe-Shetland Basin (FSB) is located between the Shetland and Faroe Islands on the NE Atlantic passive continental margin of NW Europe (Fig. 1), bound by the Wyville-Thomson Ridge (UKCS) to the SW and the Møre Marginal High (Norwegian Continental Shelf ) to the NE (Ziska and Varming 2008;. The FSB is up to 400 km long and 250 km wide, and constitutes a series of smaller rift-related sub-basins that formed throughout several main phases of extension between the early Cretaceous and the early Cenozoic . The subbasins are separated by NE-SW-trending, intra-basinal structural highs comprising Precambrian crystalline rock (Figs 1 and 3) that are commonly unconformably capped by incomplete occurrences of pre-rift Paleozoic-Jurassic sedimentary rocks (Lamers and Carmichael 1999;Sørensen 2003;Peacock and Banks 2020).
Precambrian crystalline basement, observed onshore Scotland as the Lewisian Gneiss Complex, is understood to have formed after Neoarchean pluton emplacement and cooling between 2800 and 2700 Ma (Bonter and Trice 2019;Holdsworth et al. 2019;Peacock and Banks 2020). Offshore within the FSB, basement rocks are generally considered to be equivalent to the Lewisian Gneiss Complex (Holdsworth et al. 2019), and have been cored and dated to the Neorchean . Wells that have penetrated crystalline basement within the FSB have typically encountered Neoarchean quartzofeldspathic granodioritic orthogneisses (Gardiner et al. 2019;Holdsworth et al. 2019). Towards the north of the FSB, on the West Shetland and North Shetland platforms, the nature of the basement rock changes from crystalline to Neoproterozoic Moine and Dalradian rocks and Paleozoic Caledonian rocks (Kinny et al. 2005;Holdsworth et al. 2019).
Since Neoarchean pluton emplacement, crystalline basement has been subject to the complex structural evolution of the FSB. Prominent NE-SW-striking structural trends present throughout the FSB are thought to have been inherited from pre-existing structural fabrics within basement rocks. Structural fabrics are thought to have initially developed as Precambrian shear zones during the Archean and Proterozoic, which were later reactivated as both thrusts and deep-rooted strike-slip faults during the Cambrian-Devonian Caledonian Orogeny (Coward 1990;Doré et al. 1997). During Devonian orogenic collapse, strike-slip faults were then reactivated as normal faults (Moy and Imber 2009;. The predominant NE-SW-striking structural fabric has since been periodically exploited during multiple phases of crustal extension from the Permo-Triassic to the early Cenozoic (UK Atlantic margin events: Fig. 2), separating the once contiguous crystalline basement into a series of intra-basinal highs (Figs 1a and 3) (Doré et al. 1997(Doré et al. , 1999Ellis and Stoker 2014;Schofield et al. 2020).
The initial phases of sedimentary input within the FSB are attributed to the erosion of Caledonian topography into postorogenic Devonian-Carboniferous intermontane basins, thought to have been generated by the extensional collapse of the Caledonian fold-and-thrust belt (Underhill 2003;Smith and Ziska 2011;Ellis and Stoker 2014;Stoker 2016). Devonian-Carboniferous-age basins (e.g. the Clair Basin on the Rona Ridge: Nichols 2005) accommodated the deposition of the Clair Group, dominated by fluvial sequences with lacustrine and aeolian influences (Underhill 2003;Smith and Ziska 2011). These deposits rest unconformably on Precambrian basement rock and form the main reservoir sequences of the Clair Field (Ogilvie et al. 2015;Robertson et al. 2020). Where penetrated by wells within the FSB on the Rona Ridge, the proven thickness of the Clair Group varies from absent (e.g. wells 205/21-1A and 206/12-1: Fig. 1b) to c. 1000 m thick (Ziska and Anderson 2005), reflecting a varied history of deposition and post-depositional erosion. Subsequent basin development of the FSB, after the Devonian-Carboniferous, primarily relates to the Atlantic rift system active throughout the Triassic-Cretaceous, contributing to the fragmentation of Pangaea and the opening of the Atlantic (Roberts et al. 1999;Ellis and Stoker 2014;Stoker 2016;Stoker et al. 2017b). Initial rifting is indicated by large thicknesses of Permian-Triassic strata deposited within the West Orkney and Papa basins (e.g. 2436 m in well 205/27a-1: Fig. 1b)   and inferred from seismic profiles Stoker et al. 2017b). Permo-Triassic extension is thought to be concentrated within the Papa, West Shetland, West Orkney and Unst basins ( Fig. 1) where Permian strata are proven (e.g. well 205/27a-1). This early rifting event is thought to have exploited existing basement structural fabrics, leading to the formation of NE-SW-trending half-graben basins filled with aeolian, fluviatile and lacustrine sequences, although differentiation of Permian-Triassic sequences is often challenging due to the lack of diagnostic biostratigraphy (Swiecicki et al. 1995;Štolfová and Shannon 2009;Stoker et al. 2017b). The presence of Permian-Triassic strata within the FSB is mostly interpreted to be concentrated around the present-day margins of the basin (e.g. wells 204/19-1, 204-19-9 and 204/29-1: Fig. 1b), with the exception of 213/23-1 (Eriboll) and 214/09-1 (Bunnehaven) which are located within the central FSB along the Corona Ridge . Semi-arid fluvial-lacustrine conditions prevailing throughout the Triassic are expressed within the FSB as variable thicknesses of the Papa Group, ranging from 426 m within the Judd Sub-basin (204/ 29-1: Fig. 1b)  Post-rift thermal subsidence coupled with global sea-level rise initiated a marine incursion during the early Jurassic, leading to the deposition of the shallow-marine Skerry Group sequences, although well penetrations of Lower Jurassic strata are sparse within the FSB (Dean et al. 1999;Ritchie and Varming 2011). Thermal subsidence continued until the middle-late Jurassic, when renewed extension of existing Permo-Triassic half-graben occurred, thought to be related to synchronous rifting in the North Sea  The present-day geometry of the FSB, along with the adjacent Rockall and Møre basins, was mainly developed by widespread extension initiated within the early Cretaceous, which is largely responsible for the present-day structure of the Corona Ridge Hardman et al. 2018). Extension along the NE Atlantic margin during the Cretaceous was concurrent with northward propagation of the Central Atlantic rift system, which became relatively inactive after the late Cretaceous period (Doré et al. 1999;Fletcher et al. 2013;Stoker 2016;Schofield et al. 2020).  Deposition of thick (up to 5 km) deep-marine Cretaceous successions occurred on the hanging walls of major fault systems (Larsen et al. 2010;Stoker 2016;Gardiner et al. 2019).
Post-rift thermal subsidence was terminated by a rifting phase during the early Paleocene (concentrated 400 km to the west) which exploited the established Mesozoic normal rift system and occasionally manifests on the Corona Ridge as minor normal faulting on reactivated major bounding faults only (Doré et al. 1999;Smallwood and Gill 2002;Ellis and Stoker 2014). The NE Atlantic margin experienced significant igneous activity during the Paleocene-Eocene, prior to (c. 62-58.5 Ma) and during (c. 57-54 Ma) continental break-up between Greenland and NW Europe (Watson et al. 2017;Jolley et al. 2021).
This study employs the British Geological Survey (BGS) lithostratigraphy from , which adopts the BP Paleocene T-sequence framework of Ebdon et al. (1995) (Fig. 4) for consistency with previous studies. The regional onset of volcanism within the FSB occurred prior to continental break-up and is recorded by Danian (66-61.6 Ma) to Selandian (61.6-59.2 Ma) basaltic tuffs of sequences T10-T36 (Watson et al. 2017;Jolley et al. 2021). Marine conditions prevailed throughout the Paleocene, accompanied by the deposition of a series of major turbidite fans in adjacent basins (e.g. the Flett, Judd and Nuevo sub-basins) during the mid-Paleocene (sequence T31-T36) (Stoker and Varming 2011).
Rifting was followed by uplift that began during the late Paleocene (Ebdon et al. 1995). Subsequent emergence of areas adjacent to the FSB (e.g. the East and West Shetland platforms) facilitated the deposition of Thanetian (59.2-56 Ma) to Ypresian (56-47.8 Ma) major prograding shelf-margin systems (sequences T38-T40) during the late Paleocene-early Eocene (Stoker and Varming 2011).
The main phase of extrusive volcanism occurred within the early Eocene during continental break-up, with the eruption of a widespread extrusive basaltic component comprised predominantly of thick flood basalt (lava) sequences of the Faroe Islands Basalt Group (FIBG) (Bell and Jolley 1997;Passey and Hitchen 2011;Schofield et al. 2017;Jolley et al. 2021). Flood basalt eruption during the deposition of Colsay Member sandstone units (sequence T40-T45) led to the formation of intra-basaltic plays of the Rosebank Field located along the Corona Ridge (Jolley and Bell 2002;Passey and Jolley 2009;Schofield and Jolley 2013;Hardman et al. 2018). The emplacement of an extensive suite of mafic intrusions, the Faroe Shetland Sill Complex (FSSC), occurred before and during continental break-up between c. 58 and 55 Ma Watson et al. 2017;Jolley et al. 2021).
A migration of the Central Atlantic rift system extension locus to the west of the present-day Faroes occurred during the early Eocene, synchronous with the onset of North Atlantic seafloor spreading between c. 56 and 55 Ma (Ellis and Stoker 2014;Stoker et al. 2017a, b;Jolley et al. 2021). Supra-basaltic Hildasay Member (sequence T45) sandstones of the Flett Formation are overlain by the Balder Formation (sequence T50), which acts as a regional seismic marker (Ebdon et al. 1995;Watson et al. 2017). A return to deepmarine conditions in the early Eocene resulted in the deposition of the Stronsay Group (sequence T60-T98: Stoker et al. 2012), which contains a number of submarine fan complexes including the mid-Eocene Breydon, Caledonia, Rothbury and Strachan fans, that have proven to be sandstone rich along the Corona Ridge, with the latter a reservoir for the Tobermory gas field (Mobil North Sea Ltd 1999a, b; Smallwood and Gill 2002;Stoker and Varming 2011).
The FSB was subject to multiple further episodes of tectonism during the late Paleocene, early-mid-Eocene and Oligo-Miocene (Smallwood and Maresh 2002;Smallwood 2004;Ellis et al. 2009). Throughout this time, inversion along intra-basinal highs facilitated the growth of elongate anticlines and domal structures with fourway closures within Cretaceous-Cenozoic-age strata (Sørensen 2003;Doré et al. 2008;Ritchie et al. 2008). This inversion is most likely to have been responsible for many of the traps present along the Corona Ridge but is also considered to be responsible for the breach of pre-existing top seals and subsequent leakage Ellis and Stoker 2014;Stoker et al. 2017a).

Previous structural interpretation of the Corona Ridge
The most prominent structural elements in the FSB are the NE-SWtrending intra-basinal highs composed of Precambrian crystalline basement including the Corona Ridge (and other structural highs such as the Rona Ridge). In this study we follow the definition of basement high proposed by Peacock and Banks (2020, p. 18), 'an area in which the basement rocks [Precambrian crystalline rocks] are significantly higher than in the surrounding areas'. Within the Peacock and Banks (2020) definition, the word 'significantly' is used in the context of influencing the petroleum system (e.g. source rock maturation and subsequent migration). The term 'centralnorthern Corona Ridge' is used herein to refer solely to the Precambrian crystalline rocks that form the intra-basinal high itself. Rumph et al. (1993) introduced the term Corona Ridge, although the ridge was previously variably referred to as the Mid-Faeroe Ridge (Mudge and Rashid 1987;British Geological Survey 1996;Hughes et al. 1997), Central High Complex (Hitchen and Ritchie 1987), Axial Opaque Zone (Ridd 1981(Ridd , 1983 and North Westray Ridge (Iliffe et al. 1999). Early conceptual models for the development of the Corona Ridge include a Cenozoic igneous complex similar to the Wyville-Thomson Ridge (Ridd 1981;Hitchen and Ritchie 1987) or, as the Corona Ridge is now known, a Lewisian-age crystalline basement-cored structural high analogous to the Rona Ridge (Mudge and Rashid 1987).
Numerous authors extend the ridge SW to the Cambo High (Ellis et al. 2009;Hardman et al. 2018) and NE towards Quadrant 217 (Mudge and Rashid 1987;BGS 1996;Hughes et al. 1997;Dean et al. 1999;Ellis et al. 2009;. The complexity of the Corona Ridge varies between publications. Many publications (Mudge and Rashid 1987;Rumph et al. 1993;Ebdon et al. 1995;BGS 1996;Hughes et al. 1997;Dean et al. 1999;) describe the Corona Ridge as comprising one or two linear NE-SW-trending basinal highs, totalling c. 100-150 km in length, bound by NE-SWstriking faults and delineated only by the Victory Lineament transfer zone (Rumph et al. 1993;Ebdon et al. 1995;. Within other publications the Corona Ridge is presented as a more complex structure with three structurally separate blocks ( Although the characterization along the southern Corona Ridge has improved through time due to the drilling of the three deep penetrating Rosebank exploration and appraisal wells (213/27-1Z, 213/27-2 and 213/27-3Z), knowledge of the structural configuration, geology and hydrocarbon prospectivity of the centralnorthern Corona Ridge is still enigmatic. The work presented here shows that the along-strike structural configuration of the Corona Ridge is much more complex than previously thought, as will be discussed below.

Data and methodology
The main seismic data used for interpretation throughout this project, and shown here in Figures 5-8, is a broadband 3D seismic survey acquired by PGS (Petroleum Geo-Services) for Total E&P UK Ltd and partners in two phases; phase 1 in 2013 (Galloway) and phase 2 in 2015 (Corona) (Joseph et al. 2017). The key parameters of each acquisition phase are summarized in Table 1, and the area covered by the survey is shown in Figure 1.
Data are displayed at zero-phase negative standard (Sheriff and Geldart 1982) European polarity, with a downward increase in acoustic impedance (hard) corresponding to a negative amplitude (trough) displayed in red and a downward decrease in acoustic impedance (soft) corresponding to a positive amplitude ( peak) displayed in blue. Near the Top Paleocene, at 3.5 s two-way time (TWT) (c. 3.5 km depth), the dominant frequency is 22 Hz, giving a vertical resolution of c. 37.5 m (assuming a velocity of 3300 m s −1 for the Paleocene: Rippington et al. 2015). Towards the base Cretaceous at 6.5 s TWT (c. 12 km depth) within the Flett Sub-basin the dominant frequency drops to 8 Hz, giving a vertical resolution of c. 147 m (assuming a velocity of 4700 m s −1 for the Cretaceous: Rippington et al. 2015). Towards the base of the survey at 8 s TWT (c. 18 km depth) the dominant frequency drops further to c. 6 Hz, giving a vertical resolution of c. 270 m (assuming a velocity of 6500 m s −1 for the crystalline crust: Rippington et al. 2015). Seismic data extend down to 10 s TWT.
Three exploration wells drilled along the central-northern Corona Ridge were used throughout this study: 213/23-1 (Eriboll), 214/ 21a-2 (South Uist) and 213/23c-1V (North Uist). Well data were downloaded from the NSTA National Data Repository (North Sea Transition Authority 2022), where it is freely available. The wells within this paper were drilled as near-vertical exploration wells, and throughout this paper the depths are quoted in measured depth (MD (m)), unless otherwise specified. We have completed a detailed reinterpretation of well data including end of well reports, wireline data, composite logs, image logs, core and biostratigraphy (where available) to allow the wells to be placed in the stratigraphic context of the NE Atlantic margin (see Jolley et al. 2021), and, in particular, in the context of the detailed stratigraphy of the Rosebank Field (after Schofield and Jolley 2013).

213/23-1 (Eriboll)
213/23-1 (Eriboll) was the first well drilled along the Corona Ridge, in July 1998 by Mobil. The well targeted three stratigraphic intervals: the primary Mesozoic-age Eriboll oil prospect on the footwall of a tilted fault block (the Corona Ridge) with three-way dip closure, and two shallower secondary targets within the upper Paleocene (Solway prospect) and middle Eocene (Caledonia prospect) (Mobil North Sea Ltd 1999a). Pre-drill concerns regarding timing, migration and lack of charge of the secondary targets were confirmed by the presence of water within the upper Paleocene Solway and middle Eocene Caledonia prospects. The well encountered Eocene-Paleocene strata, Cretaceous strata, a section postulated to be Triassic strata (referred to as 'Triassic?' in the end of well report) and Devonian-Carboniferous strata. No Jurassic strata were penetrated in the well (Mobil North Sea Ltd 1999a). According to the end of well report, 213/23-1 (Eriboll) penetrated 1.5 m of rock assigned as Precambrian Lewisian gneiss before reaching total depth (TD) at 4344 m MD. However, cuttings of the 1.5 m of Lewisian basement are described within the end of well report as both 'crystalline and porphyritic' and to contain 'lithic fragments' (Mobil North Sea Ltd 1999a). Numerous oil shows were encountered within Upper Cretaceous, 'Triassic?' and Devonian-Carboniferous strata.
213/23-1 (Eriboll) -Paleocene-Eocene: identification of the T40-T45 Colsay Member and Colsay units 1-4 Our reinterpretation of various types of well data (specifically biostratigraphy suggestive of depositional environments associated with Colsay units 1-4) obtained from 213/23-1 has allowed sequences penetrated within the well to be assigned to the current Paleocene-Eocene lithostratigraphic framework Schofield and Jolley 2013) and T-sequences (Ebdon et al. 1995) used throughout the FSB (Fig. 4), not assigned at the time the well was drilled.
By re-evaluating biostratigraphy, we found that the Horda Formation (sequence T60, Ypresian age), not previously identified, lies conformably on the Balder Formation (sequence T50), which comprises claystone with thin beds of volcanic tephra. The upper Paleocene Solway prospect was prognosed between the Balder and Flett formations. The Solway prospect, which was thought to be hosted in shallow-marine sandstones, was also found to be dry (Mobil North Sea Ltd 1999a).
We found that the well (213/23-1, Eriboll) encountered 422 m of Flett Formation; more than double the original thickness interpreted post-drill (194 m) (Fig. 9a). No oil shows were observed within the 42 m-thick upper section of the Flett Formation, which we assigned to the Hildasay Member (sequence T45). The well (213/23-1) penetrated a 380 m-thick section that we assigned to the Colsay Member (sequence T40-T45) of the Flett Formation, which is comparable in thickness to the Colsay interval throughout the Rosebank Field (c. 400 m: ChevronTexaco Upstream Europe 2004; Chevron Upstream Europe 2007a). Within 213/23-1 (Eriboll), the Colsay interval is described in the end of well report as a series of sandstones interbedded with mudstones and siltstones (Mobil North Sea Ltd 1999a). Numerous traces of residual oil within the lower (sequence T40) Colsay Member were observed (Fig. 10).
According to the end of well report, the well (213/23-1) encountered the Lamba Formation (sequence T36-T38) and an undifferentiated Paleocene-aged tuffaceous unit below the Flett Formation ( Fig. 9a) (Mobil North Sea Ltd 1999a). However, our biostratigraphic reinterpretation suggests that both intervals actually represent the T40 Colsay 4 unit interval (Fig. 10) and that no T10-T38 sequences are present within the well. The Colsay 4 unit comprises material eroded and reworked from Paleocene-aged (sequence T10-T38) sediments.
Reinterpretation of the Colsay Member (sequence T40-T45) within 213/23-1 (Eriboll) using the end of well report, composite log, biostratigraphy and resistivity image logs has facilitated the identification of a series of sequence T40 volcanic rocks, and sub-basaltic and intra-basaltic sandstones. Biostratigraphic interpretation of two Rosebank wells (appraisal well 213/27-2 and exploration well 213/27-3Z) was used to correlate the Colsay intervals 21.3 km along the central-northern Corona Ridge from the Rosebank Field to 213/23-1 (Eriboll) (Fig. 11). Descriptions of each Colsay unit (units 1-4) within 213/23-1 are given within Table 2. 213/23-1 (Eriboll) -Paleocene-Eocene: identification of T40-T45 Colsay Member volcanic intervals According to the original composite log, two 2-3 m-thick sections of 'intrusive volcanics' identified as 'dolerite? volcanics?' were penetrated between 2870 and 2890 m MD (2840-2860 m TVDSS (true vertical depth subsea)) (Figs 9 and 10), within an 18 m-thick limestone host rock. Cuttings of the 'intrusive volcanics' are described within the composite log as 'black-grey, translucent, opaque, brown, locally green, black mafic groundmass, hard, crystalline to very finely crystalline, equigranular, trace mica and pyrite, no porosity'. However, a core plug taken within the upper 'intrusive volcanic' section is contrastingly identified as claystone in the end of well report (Mobil North Sea Ltd 1998, 1999a. Core plugs taken within the proposed limestone section are also contrastingly identified as siltstone and dolerite within the end of well report (Fig. 10). As a result of these discrepancies, we reinterpreted lithologies throughout the entire Colsay Member (T40-T45) interval using wireline, composite log, core plug, core and resistivity image log data (Fig. 10). We interpreted four volcanic intervals (Fig. 10) that were previously interpreted as either sandstone or limestone intervals in the original composite logs (Mobil North Sea Ltd 1998).
The reinterpreted volcanic intervals are typified by low gammaray values and fast sonic values (Fig. 10). Resistivity image logs of the volcanic intervals contain c. 10 cm-wide conductive zones that are likely to represent open fractures or cooling joints (Fig. 10c, e), whilst other volcanic intervals (Fig. 10a, c) have thinner (<1 cm wide) vertical conductive (open) fractures ( possibly drilling induced). The thickness of laminated claystone overlying the volcanic intervals appears to have been influenced by the underlying rough top surface of lava observed within the image logs (Fig. 10f ), suggesting that the intervals within well 213/23-1 (Eriboll) are lava flows, not igneous intrusions, as suggested for the two 2-3 m-thick sections described as 'intrusive volcanics' within the original composite log (   213/23-1 (Eriboll)does 213/23-1 penetrate a Triassic sequence?
The Triassic interval (c. 3585-3763 m MD) is assigned with uncertainty according to the end of well report and is referred to as 'Triassic?', although there is no biostratigraphic evidence for the assignation of a Triassic age to this section. The original biostratigraphic analysis states that 'palynological evidence suggests Carboniferous aged sediments may be present, but the results are not conclusive' in regard to the section assigned 'Triassic?' (Mobil North Sea Ltd 1999b, p. 3). Within the 'Triassic?' section (at 3736 m MD: Fig. 12), 'in situ Carboniferous taxa' are recorded, although considered sparse, and the interval remained lithostratigraphically unassigned as such (Mobil North Sea Ltd 1999b, p. 45). Two separate interpretations of the interval are provided in the postwell biostratigraphic analysis: either a continuation of the Lower Cretaceous Cromer Knoll Group or the section being middle Carboniferous in age (Mobil North Sea Ltd 1999b, p. 2). A continuation of Lower Cretaceous strata is unlikely due to the seismic observation that the Cromer Knoll Group lies at an apparent angular unconformably to the proposed 'Triassic?' section ( Fig. 6). The absence of any unconformity within resistivity logs at and proximal to the proposed 'Triassic?'-middle Carboniferous boundary ( pink dashed line, 3763 m MD: Fig. 12a) and the similarity of facies and bedding (orientation and dip) within both the 'Triassic?' and middle Carboniferous section suggests that the 'Triassic?' section may instead be a continuation of underlying middle Carboniferous strata (Mobil North Sea Ltd 1999a). We therefore interpret the 'Triassic?' interval within 213/23-1 (Eriboll) as a continuation of the underlying middle Carboniferous Clair Group strata identified within the original end of well report (Fig. 13) and this is discussed in the following subsection.

213/23-1 (Eriboll) -Carboniferous
The middle Carboniferous (Clair Group) is described within the end of well report as a series of homogenous sandstones with good (c. 15%) porosities, conglomerates, laminated mudstones and one 5 m-thick interval of coal (Mobil North Sea Ltd 1999a). Fining-up sequences containing flat to low-angle bedding and cross-bedding with localized mud clasts and burrows are evident within core obtained from the interval (Mobil North Sea Ltd 1999c). We interpret the presence of coal, laminated mudstones, conglomerate, and cross-bedded sandstones containing mud clasts and burrows as having been deposited within a fluvial depositional environment (Mobil North Sea Ltd 1999c). Coal is also present within the ageequivalent fluvial-fluviodeltaic Carboniferous sequences of the Clair Field (Robertson et al. 2020). Post-drill petrography of middle Carboniferous core (Fig. 12) suggests that lower porosities (<2.5%) are present, the result of c. 1 m-thick sections exhibiting extensive dolomite and calcite cement (Fig. 12), and the presence of matrix clay and quartz overgrowth cement (Mobil North Sea Ltd 1999c). Oil shows were encountered within the middle Carboniferous interval, with oil described as 'bubbling' from core ( Fig. 12) (Mobil North Sea Ltd 1999a) and traces of oil were seen in mud at the shakers (3876-3884 m MD) (Mobile North Sea Ltd 1999a). Lower Carboniferous strata are characterized within the composite log by sandstones with fair to good porosity, recorded as containing oil shows (Mobil North Sea Ltd 1998). No image logs or core were obtained through the lower Carboniferous section. Average petrophysical properties throughout the Carboniferous section are shown in Table 3.
Despite some Carboniferous fractures (Fig. 12) exhibiting residual brown and black oil stains, the reservoir quality of the unit is considered poor due to the presence of dolomite cement, calcite cement and clay smears (Mobil North Sea Ltd 1999c). We interpret fractures present within resistivity image logs of the middle Carboniferous strata to have dominant strike orientations of NE-SW and ENE-WSW (Fig. 12b). Fracture density is higher within the uppermost section of the middle Carboniferous (1 m −1 ) previously interpreted as 'Triassic?' (3585-3763 m MD) compared with the lowermost middle Carboniferous section (0.5 m −1 at 3763-3884 m MD), although borehole washouts impaired image log interpretation within the lower section (Fig. 12). We interpret borehole breakouts with NE-SW orientations and drilling-induced tensile fractures (DITFs) with NW-SE orientations within middle Carboniferous strata as suggesting a present-day maximum horizontal stress (S H max ) orientation of NW-SE.
A drill stem test (DST) was later performed on the lower-middle Carboniferous prospective 'Eriboll Middle Zone' from depths of 3814 to 3914 m MD (Figs 9a and 13) after re-entry in August 1999, almost 9 months after the well was initially suspended (Mobile North Sea Ltd 1999a). The DST perforated 100 m of the Clair Group from which the 6.5 m of 'oil-bubbling' core had been recovered the previous year (Fig. 13). The DST produced only water and did not flow oil to surface, with no technical issues reported during the DST (The Expro Group 1999).

213/23-1 (Eriboll) -Devonian
The Devonian Lower Clair Group (interpreted post-drill as Upper-Middle Old Red Sandstone: Fig. 13) is described as a fine-to coarse-grained, moderately sorted sandstone with poor porosity and poor oil shows within the end of well report (Mobil North Sea Ltd 1999a). No image logs were obtained within the Devonian interval of 213/23-1 (Eriboll) (Fig. 13). The average petrophysical properties throughout the Devonian section are shown in Table 3.
Post-drill interpretation of wireline data by Mobil suggested that a significant oil column was present in the Devonian-Carboniferous sequences, with average net-to-gross (NTG) of 42% and high water saturation (S w ) of 71% (Mobil North Sea Ltd 1999a, p. 7).

214/21a-2 (South Uist)
Well 214/21a-1 was drilled as part of a licence commitment in February 2008 by Shell and partners to target the South Uist gas prospect, located along an elongated (c. 4 km by 10 km) four-way dip-closed structure, around 10 km from the Corona Ridge structural high (Shell UK Ltd 2010). Pre-drill in-place estimates were 2.1 tcf (trillion cubic feet of gas) (Shell UK Ltd 2014). After extended operational challenges (e.g. blowout preventer (BOP) repairs and waiting on weather) the well was suspended and redrilled as 214/21a-2 (South Uist) in July 2009 (Shell UK Ltd 2010). Before drilling, prognosed risks and uncertainties were the absence of reservoir and the age of the target interval, which was debated by the well partners. Shell interpreted the target interval as the lower Paleocene Sullom Formation (T20); however, partner ConocoPhillips interpreted the target as Maastrichtian-age Cretaceous sedimentary rocks (Shell UK Ltd 2010). On drilling, the Sullom T20 target was encountered 173 m above the prognosed target depth. The target interval (>3713 m MD) consisted of Cretaceous mudstone and limestone stringers with no sandstones present, proving the stratigraphic interpretation of ConocoPhillips to be correct (Shell UK Ltd 2010

213/25c-1V (North Uist) -Paleocene-Eocene
The T60 middle Eocene Stronsay Group comprises mudstone with massive (c. 50 m-thick) interbeds of sandstone. Beneath the Stronsay Group, the Balder Formation was encountered, and is described as tuffaceous mudstone with siltstones, trace sandstone and trace limestone within the end of well report (BP Exploration Operating Company Ltd 2013b). Although not prognosed, an 8 mthick (subseismic resolution) section of the Flett Formation was encountered within the secondary target (Cardhu) interval beneath the Balder Formation. Within well 213/25c-1V (North Uist), the Flett Formation is described as predominantly mudstone, with traces of sandstone observed as rock flour in cuttings. Although also not prognosed, the Lamba and Vaila formations were both penetrated below the Flett Formation (BP Exploration Operating Company Ltd 2013a). The Vaila Formation lies unconformably on the Upper Cretaceous Shetland Group.

213/25c-1V (North Uist) -Cretaceous and Jurassic
According to the composite log, the Upper Cretaceous strata consist of mudstone, siltstones and occasional limestone stringers (BP Exploration Operating Company Ltd 2013a). In well 213/

213/25c-1V (North Uist) -Devonian-Carboniferous
Within the composite log, the Carboniferous UCG is characterized by sandstones with mudstone interbeds (BP Exploration Operating Company Ltd 2013a). The lower, Devonian-Carboniferous, section of the Clair Group comprises sandstone, siltstone and thin mudstones. Sedimentological observations of the Devonian-Carboniferous Clair Group are comparable to well 213/23-1 (Eriboll), suggesting a fluvial depositional environment, although this is speculative due to the lack of conventional (continuous) core obtained in 213/25c-1V (North Uist).
213/25c-1V (North Uist) discovered gas condensate (42.5°API: CNOOC Ltd 2017) in two reservoirs: the upper (Carboniferous, 3974-4069 m MD) and lower (referred to as Devonian-Carboniferous within the end of well report, 4069-4773 m MD) Clair Group reservoirs (Fig. 9b). No gas-water contact (GWC) was identified in the composite log or the end of well report, although neutron-density crossover implies a GWC at c. 4630 m MD. The much thicker (704 m) lower reservoir is referred to as 'tight', with very low mobilities (<0.1 mD/cP) suggestive of low permeabilities, which alongside high S w (82%) resulted in no net pay interpreted by BP within the lower reservoir. Properties of the upper and lower reservoirs are shown in Table 3.
After drilling, the upper reservoir (North Uist discovery) was estimated to contain 104 MMboe (P 50 case) of condensate in place, just under 10% of the pre-drill in-place P 50 estimates (1067 MMboe) (CNOOC Ltd 2017). No well testing was performed, and the well was plugged and abandoned.

Seismic interpretation along the central-northern Corona Ridge
Top Crystalline Basement time-structure maps (Fig. 14) show the central-northern Corona Ridge as one structurally high (at c. 5 s depth TWT) feature within the SW of the study area, which splits into two structural highs towards the NE that have less relief (c. 6 s depth TWT). The southerly branch splits further into two components, creating a series of lows between the highs of crystalline basement within the NE of the dataset (Fig. 14b). The Corona Ridge is bound by a series of mapped normal faults that strike NE-SW (collectively termed the Flett-Corona Fault: Rippington et al. 2015) within the Flett Sub-basin. Two faults that strike north-south and NNW-SSE (red faults, Figs 5 and 14), oblique to the main series of NE-SWstriking faults (Fig. 14), are mapped within the centre of the dataset, where the Corona Ridge bifurcates. The time-structure map of the Top Crystalline Basement (Fig. 14) horizon clearly shows that the structure of the Corona Ridge is far more complex than previous interpretations (see Fig. 14b for comparison). Downflank fault terraces of Jurassic and/or Devonian-Carboniferous strata (Figs 7, 15a and 16a) were also identified, downthrown from the main Corona Ridge structure along the Flett-Corona Fault, which are overlain by Cretaceous and/or Jurassic sedimentary rocks.
Towards the base of the seismic survey (between 7 and 9 s TWT) a series of high-amplitude, hard (negative amplitudes/troughs 1 Siliciclastic successions of interbedded sandstones, siltstones and claystones with occasional volcaniclastic claystones and sandstones. Average gross thickness across the field is c. 56 m. Reservoir sandstones are described as very fine to medium grained (rarely coarse), subrounded to sub-angular, poorly cemented and moderately to well sorted with average porosities of 21-23%. Interpreted as braided fluvial sandstones within the T40 Colsay 1 interval and fully marine within the T45 Colsay 1 interval.
Siliciclastic and volcaniclastic sandstones, siltstones and claystones. Sandstones are very fine to medium grained, angular to sub-rounded, poor to well sorted, grain supported, friable, with occasional calcareous cement. Highest point-counted porosities seen are 15% with others as low as 5%. Reservoir quality is considered poor, reflecting high proportions of clay matrix, calcite cement and grain compaction. 10 m of T45 Colsay 1 core (Fig. 10) contains flat to low-angle lamination, waveripple lamination with local hummocky cross-stratification, bivalves, burrows and plant fragments interpreted as shallow-marine shelf deposits. 2 Thin siltstones, lignitic material and occasional volcaniclastic lithologies (siltstones and sandstones). Sandstones are described as fine (rarely coarse), well sorted and angular with visible porosities described as poor to moderate.
Sandstones, claystones and volcaniclastic claystone. Sandstones are very fine to medium (locally coarse), sub-rounded to rounded, moderately sorted and friable with intermediate porosities. 3 Successions of sandstones, claystones and siltstones with influxes of volcaniclastic sandstone indicative of an estuarine environment, overlain by fluvial sediments sourced from the SW. Average gross thickness across the field is c. 28 m. Reservoir sandstones are very fine to fine grained, occasionally coarse, sub-angular to angular, well sorted and poorly cemented with average porosities between 21 and 23%.
Sandstones, siltstones, claystones and volcaniclastic rocks (siltstone and claystone). Sandstones are very fine to fine, sub-angular to sub-rounded, moderately to poorly sorted, friable, very micaceous and grain supported with no visible matrix or cement present and poor porosity. One residual oil show was observed within volcaniclastic siltstone at c. 2955 m MD (Fig. 10). Porosities of sandstone cuttings are described as intermediate to poor; no values are given. 4 Interbedded clastic and volcaniclastic claystones and siltstones, interpreted as prodelta deposits. Average gross thickness of c. 22.5 m across the field. Sandstones are described as very fine, sub-angular to angular, well sorted and thin.
Sandstones, claystones and volcaniclastic lithologies (sandstone, siltstone and claystone). Sandstones are very fine to fine grained, subangular, moderately sorted, grain supported, with occasional calcareous cement and oil shows. Oil show observed (3079 m MD) within a 25 mthick, friable, very fine to fine, sub-angular, moderately sorted reworked sandstone with 'loose' quartz grains.
Porosities given for Colsay units 1 and 3 within 213/23-1 are based on petrography of Colsay 1 core (Fig. 10)  displayed as red reflections on Fig. 5a), semi-continuous reflections (Fig. 5) can be observed beneath the central-northern Corona Ridge (labelled in Figs 5 and 7). The high-amplitude reflections appear to coalesce and have upwards concave geometries. The highamplitude reflections are present at these depths throughout most of the 3D survey, although a much higher quantity is present in the centre of the 3D seismic dataset (e.g. Fig. 5) compared with the northern part of the dataset where there are none (Fig. 8). Assuming velocities based on Rippington et al. (2015) for the crystalline continental crust, this would place these reflections at depths of c. 16-20 km. The minimum vertical resolution at c. 18 km depth (c. 8 s TWT) is c. 270 m. Based on our mapping of the Faroe-Shetland Sill Complex, emplacement of magma into Cretaceous and lower Paleocene sedimentary host rocks was generally confined to the Flett and Sissal sub-basins (Fig. 17). Only two igneous intrusions were interpreted within the seismic data to have been emplaced within the Cretaceous host rock directly above the central-northern Corona Ridge (Figs 5 and 17), implying a lack of magma emplacement.

Crystalline basement structure of the central-northern Corona Ridge
Whilst the crystalline basement rocks of the Rona Ridge have been extensively studied (due to the Clair Field and other basement plays: e.g. Bonter and Trice 2019;Holdsworth et al. 2019), the Corona Ridge has received much less attention, and currently there are no detailed crystalline basement maps published. Work along the Corona Ridge has largely focused on prospectivity within overlying strata (e.g. Paleocene-Eocene plays: Hardman et al. 2018;Duncan et al. 2020) and tends to adopt simplified structural maps of the Corona Ridge. The NE-SW-trending structure of the central-northern Corona Ridge mapped as part of this study (Fig. 14b) shows that its structure is more complex than the older, poor, sub-basaltic seismic imaging previously allowed authors to image (e.g. Hughes et al. 1997;Ellis et al. 2009). Previous interpretations show the structural high as one crystalline block within the study area (see map bottom right in Fig. 14b), whereas we show that the central-northern Corona Ridge splits into three structural highs towards the NE, creating a series of lows (minibasins: Fig. 14a) between the highs of crystalline basement. We also show that the NE-SW-striking Corona Ridge bounding normal faults (Fig. 14), often simplified as one feature within regional maps of the basin (e.g. Mudge and Rashid 1987;BGS 1996;Ellis et al. 2009;) and collectively termed the Flett-Corona Fault by Rippington et al. (2015), in fact constitute a series of discrete faults that instead should be termed the Flett-Corona Fault System.

Oblique faults and downflank terraces of the Corona Ridge
The two north-south-and NNW-SSE-striking faults (red faults in Figs 5, 14 and 15a) mapped within the centre of the study area are oblique to the NE-SW-strike of basement bounding faults of the collectively termed Flett-Corona Fault System. The 'oblique' faults are expressed at both the Top Crystalline Basement and Base Cretaceous levels (red faults in Figs 5, 14 and 15a), and are located where the Corona Ridge is interpreted to bifurcate into two parts mapped within the NW of the dataset. Although much of the NE-SW-striking structural fabric in the FSB (and, indeed, broader NE Atlantic margin) is thought to reflect inheritance from pre-existing crystalline basement trends, the north-south-striking oblique faults do not appear to reflect structural inheritance from NW-SE-trending Precambrian shear zones (Knott et al. 1993), NE-SW-striking Caledonian compressional fault systems (Doré et al. 1997) or NW-SE-trending lineaments or 'transfer zones' (Rumph et al. 1993;Moy and Imber 2009). Dean et al. (1999)   et al. 2020). It has also been suggested that the north-south structural trends seen across the NE Atlantic margin may reflect the reactivation of pre-Caledonian or Caledonian suture zones, although direct evidence for this is sparse (Schiffer et al. 2020). Assessing the origins of oblique faults mapped here (red faults in Figs 5, 14 and 15a) and any further extent of north-south-striking faulting across the FSB is challenging, as the imaging of basement structures are often obscured by overlying Paleogene lava flows and the Faroe-Shetland Sill Complex (Hardwick et al. 2010). However, the identification of north-south structuration within the crystalline basement of the central-northern Corona Ridge suggests that such features could be more extensive than previously realized within the basin.
Downflank fault terraces of Jurassic and/or Devonian-Carboniferous strata (Figs 7, 15a and 16a) overlain by Cretaceous and/or Jurassic sedimentary rocks were also identified downthrown from the main Corona Ridge structure along the Flett-Corona Fault System. Barr et al. (2007) described several similar terraces that extend the Clair Field from the main Rona Ridge structure, which also comprises Devonian and Jurassic strata (referred to as the Clair South development: Robertson et al. 2020).

High-amplitude reflections beneath the central-northern Corona Ridge
A series of hard (negative amplitudes/troughs displayed as red reflections on Fig. 5a) high-amplitude reflections are observed within the 3D seismic data (Figs 5 and 7) beneath the centralnorthern Corona Ridge between 7 and 9 s TWT (c. 16-20 km depth). The high-amplitude reflections have upward-concave geometries and often appear to coalesce with other high-amplitude reflections. The minimum vertical resolution at c. 18 km depth (c. 8 s TWT) is c. 270 m. Seismic refraction data and associated velocity modelling within the study area (e.g. iSIMM, Mobil-1 and Mobil-2 lines: White et al. 2005;Makris et al. 2009;Roberts et al. 2009) indicate that the depth to the Moho beneath the Corona Ridge is c. 17-20 km Petersen and Funck 2017). It is therefore likely that the high-amplitude reflections are located within the lower crust.
Whilst there is no direct evidence as to the origin of these highamplitude reflections, through comparison with analogous highamplitude reflections published throughout the NE Atlantic margin that have the same reflectivity, geometries and are located at similar depths (c. 15 km, White et al. 2008; c. 7-9 s TWT, Abdelmalak et al. 2017 andWrona et al. 2019), we tentatively propose that the high-amplitude reflections could be caused by basic intrusive bodies within the lower crust.

Devonian-Carboniferous Clair Group
The Clair Field, located on the Rona Ridge (c. 70 km from the central Corona Ridge), is expected to produce over 940 MMbo (million barrels of oil) from Devonian-Carboniferous Clair Group reservoirs (Robertson et al. 2020). The discovery of 104 MMboe in place ( post-drill P 50 case) of condensate in North Uist (213/25c-1V) proves that the Devonian-Carboniferous is a working play on the Corona Ridge (CNOOC Ltd 2017). However, no well tests were performed within well 213/25c-1V, which limits the interpretation of the reservoir volume and accurate estimation of the post-drill connected in-place volume of North Uist.
Remnants of an intermontane basin are mapped on the Corona Ridge (Fig. 14b) that contain Devonian-Carboniferous strata penetrated by wells 213/23-1 (Eriboll) and 213/25c-1V (North Uist) (Fig. 16a). We also note the presence of the Clair Group below the Rosebank Field to the SW of the study area, confirmed by wells 213/27-1Z (95.8 m) and 213/27-2 (90 m) (Fig. 1b) Smith and Ziska 2011). No Devonian-Carboniferous Clair Group was reported to have been penetrated to the NE of the study area in well 214/09-1 (Bunnehaven: Fig. 1b), and hence most previous work (e.g. Smith and Ziska 2011;Stoker et al. 2014) does not extend its distribution north of 213/25c-1V (North Uist). Smith and Ziska (2011) interpreted 213/23-1 (Eriboll) as the only well to penetrate the entire Devonian-Carboniferous sequence on the Corona Ridge (totalling 753 m) before terminating in crystalline basement. The lack of penetrations of the entire Devonian-Carboniferous sequence on the Corona Ridge compared with the Rona Ridge does not allow for comparison of thicknesses across the two ridges. However, our mapping of the thickness of the Devonian-Carboniferous sequences (Fig. 16a) does reflect varying thicknesses of the Devonian-Carboniferous strata (0-1.6 s TWT), comparable to the variable thicknesses within the Clair Field (c. 150-900 m: Robertson et al. 2020) and the Rona Ridge as a whole (c. 1000 m: Ziska and Anderson 2005).
Early Clair Field wells (drilled from 1977 to 1985) failed to confirm economical amounts of recoverable hydrocarbons within the Clair Group on the Rona Ridge (Robertson et al. 2020). Ogilvie et al. (2015) attributed the low flow rates (<1000 bopd (barrels of oil per day)) observed in early appraisal drilling to vertical wells being unlikely to penetrate potentially productive open fractures, compared with later-stage horizontal well sections (i.e. during the 1990s) that penetrated vertical fractures, proving that required flow rates could be achieved. We note that well 213/23-1 (Eriboll), much like early Clair Field appraisal wells, had a vertical trajectory and was therefore not orientated optimally to intersect any open fracture sets that may be present. The DST performed within 213/23-1 (Eriboll) also tested only 100 m of lower-middle Carboniferous strata, which is 10% of the length which was needed to obtain successful flow rates (18 000 bopd) from a 950 m horizontal section within the later-stage Clair Field appraisal wells (206/8-10 and 206/ 8-10z: Fig. 1b) (Robertson et al. 2020). Core obtained within the tested interval of 213/23-1 (Eriboll) (Fig. 12c) exhibits evidence of abundant calcite cementation and is not representative of the 753 mthick Devonian-Carboniferous section that showed positive early signs of oil present (3876-3884 m MD: e.g. positive flow tests and oil shows 'bubbling' from core). We therefore suggest that Eriboll may be under appraised.
Devonian-Carboniferous Clair Group: is there fracture enhanced permeability along the central-northern Corona Ridge?
According to critical stress theory, fractures optimally orientated in relation to the present-day stress field (usually parallel to the orientation of S H max ) are more likely to be hydraulically conductive and thus potentially allow the transmission of fluids (Hillis 1997;Rogers 2003;Tassone et al. 2017). On the central-northern Corona Ridge, we interpret an S H max orientation of NW-SE, which is consistent with general trends of present-day maximum horizontal stress observed across NW Europe (Holford et al. 2015). Therefore, according to critical stress theory, NE-SW-and ENE-WSWstriking Carboniferous fractures that we interpret within 213/23-1 (Eriboll) (Fig. 12) are not optimally orientated for enhanced fracture permeability, much like the Carboniferous strata of the Clair Field on the Rona Ridge (Robertson et al. 2020).
Within the Devonian Lower Clair Group reservoir of the Clair Field, according to Robertson et al. (2020), enhanced fracture permeability has been instrumental in achieving commercial flow rates. In contrast to the Clair Field, only poor oil shows were encountered within the Devonian strata penetrated by well 213/23-1 (Eriboll), and subsequently no image logs were obtained and no DST was completed within the Devonian section of 213/23-1, which precludes the analysis of any potential fracture permeability that may be present within Devonian strata along the centralnorthern Corona Ridge.
Triassic strata in the FSB are typically entirely barren of all fossil groups (Swiecicki et al. 1995). According to Morton et al. (2010) and Morton and Milne (2012), the Devonian-Carboniferous strata of the Clair Field are also entirely barren, making age determination between Triassic and Devonian-Carboniferous sections challenging. Age determination of the Carboniferous UCG section in 213/ 25c-1V (North Uist) using biostratigraphic and heavy mineral data was also challenging according to the end of well report, comparable to the challenges associated with the proposed 'Triassic?' interval in 213/23-1 (Eriboll) and the indeterminate interval in 214/09-1 (Bunnehaven).
If, like 213/23-1 (Eriboll), the indeterminate interval in well 214/ 09-1 (Bunnehaven) is Devonian-Carboniferous in age and not Triassic (or Jurassic), this could provide renewed potential for the presence of exploration plays of this age ( provided that there is Devonian enhanced fracture permeability and large enough structural closures) and the possible presence of gas-prone source rocks (Carboniferous coals: Figs 2 and 13) along the northern Corona Ridge, which have not previously been interpreted this far north Stoker et al. 2014Stoker et al. , 2017b. This may also have wider implications in understanding the early geological evolution of the FSB subsequent to post-Caledonian orogenic collapse, as Devonian-Carboniferous depositional systems may have been far more extensive than previously thought within the north of the FSB (Stoker et al. 1993;Nichols 2005;Smith and Ziska 2011).
Our interpretation of no Triassic strata present within 213/23-1 suggests an unconformity between middle Carboniferous and Lower Cretaceous strata of at least 154 myr. Implications of our interpretation, the absence of Triassic strata, are that: either (1) Permian-Triassic rocks were deposited but then eroded from the Corona Ridge and recycled into the adjacent Corona, Flett and Sissal sub-basins as synrift reservoir units throughout the late Jurassic-early Cretaceous evolution of the Corona Ridge; or (2) no Permian-Triassic rocks were deposited as the Corona Ridge was a prominent structural high throughout the Permian-Triassic.
Although  2018) is supported by the lack of Permian-Jurassic strata within 213/23-1, the preservation of Lower-Middle Jurassic within 213/ 25c-1V (North Uist) and Upper Jurassic-age strata within 213/27-1Z does not support this interpretation (ChevronTexaco Upstream Europe 2004;BP Exploration Operating Company Ltd 2013b). If the latter interpretation is correct, the Corona Ridge may have been a structurally high feature during the Permian-Triassic (hence no deposition) which was then drowned during the proposed Jurassic marine incursion (Dean et al. 1999), and then re-emerged during the late Jurassic-early Cretaceous rift period, which is understood to be largely responsible for the present-day structure of the Corona Ridge Hardman et al. 2018). During late Jurassicearly Cretaceous extension there may have been some localized shedding and recycling of Carboniferous and Jurassic strata from the Corona Ridge (forming localized synrift targets) into adjacent suband mini-basins (Figs 14 and 15a). This may explain the absence (in 213/23-1) and relatively thin (

The Lower Cretaceous synrift play
The possible recycling of 'pre-rift' Devonian-Carboniferous, Permian-Triassic and Jurassic strata from emergent structural highs (such as the Rona Ridge: Clark et al. 2020), with the addition of extrabasinal sediment (Stoker and Ziska 2011), during widespread early Cretaceous extensional deformation is thought to be responsible for the deposition of Lower Cretaceous reservoirs of the Cromer Knoll Group, which act as reservoirs for the Edradour Field and Glendronach discovery (Stoker 2016;Clark et al. 2020). Evidence of this widespread early Cretaceous extension can be observed within Figures 7 and 8, where synsedimentary deposition of Cretaceous sediments appears to relate to normal faults. Within the West Shetland  (Fig. 15a), with relevant structural elements (oblique fault shown in dark red). Due to the absence (e.g. 213/23-1) and thin (e.g. 213/25c-1V) presence of Jurassic encountered on the central-northern Corona Ridge along the structural high, the isochron can be used to infer Devonian-Carboniferous thicknesses. Note: we interpret no Triassic present in the study area; (b) the Cretaceous, (c) the Paleocene and (d) the Flett Formation within the study area. Note: (1) wells shown in red reflect where wells have penetrated the interval; and (2) where thicknesses of both the Paleocene (c) and Flett Formation (d) are less than 100 ms, intervals may be absent (e.g. the Paleocene T10-T38 interval is absent in well 213/23-1).
Basin, the Lower Cretaceous Victory Formation is also a proven gas reservoir for the Victory Field (Figs 1b and 2) (Goodchild et al. 1999). Rift-related topography is thought to have controlled synrift depositional systems throughout the early Cretaceous along major relay ramps into inherited half-graben (Larsen et al. 2010).
Localized recycling of Carboniferous and Jurassic strata during the late Jurassic-early Cretaceous emergence of the Corona Ridge (as discussed in the previous subsection) is possibly reflected by the thin presence of the Lower Cretaceous Cromer Knoll Group within wells 213/23-1 (Eriboll: Fig. 9a), 213/25c-1V (North Uist: Fig. 9b) and 214/09-1 (Bunnehaven) (c. 120 m: ExxonMobil 2001), and the irregular occurrence of Jurassic rocks along the Corona Ridge (e.g. absent in 213/23-1 but present in 213/25c-1V and 214/09-1). Within neighbouring basins such as the Flett Sub-basin, into which localized shedding of eroded sediment may have occurred from the Corona Ridge, Lower Cretaceous intervals are much thicker, reaching up to 1168 m (Stoker 2016). The Cretaceous significantly thickens (>2000 ms) from the relief present along the centralnorthern Corona Ridge at the Base Cretaceous level (Figs 15a and b) into Cretaceous depocentres of the Sissal Sub-basin and other minibasins identified between the sections of the Corona Ridge within the NE of the dataset (mini-basins: Fig. 14b). This suggests that there may be the potential for packages of Lower Cretaceous synrift plays to be present in the NE of the study area, with play fairways located off structural highs.
At the time of publication, no well has specifically targeted the Lower Cretaceous synrift play near the central-northern Corona Ridge within the sub-basins (Corona, Flett and Sissal) in and around the northern area of the FSB (Larsen et al. 2010;Stoker and Ziska 2011). In general, few wells have specifically targeted Lower Cretaceous strata in the northern FSB. Well 208/24-1A (NE part of the Flett Sub-basin: Fig. 1b) encountered 64.7 m of Lower Cretaceous sandstone that contained no oil shows (Stoker and Ziska 2011). Wells that have encountered Lower Cretaceous sands within the northern FSB whilst targeting other intervals are confined to the Corona Ridge (214/09-1, Bunnehaven) and the Erlend High (209/12-1) (Fig. 1b). Thin Lower Cretaceous sands encountered in both wells (214/09-1 and 209/12-1: Fig. 1b) were only observed as rock flour at the surface. The bias resulting from wells drilled upon structural highs restricts the understanding of how synrift Lower Cretaceous reservoirs may be distributed within the deeper sub-basin areas such as the Corona, Flett and Sissal sub-basins. The possible distribution of Lower Cretaceous reservoirs is therefore largely implied by underlying structures (Figs 14 and 15a), which may have influenced the deposition of synrift Cretaceous sediments into structural lows off the Corona Ridge within the NE of the study area.

Igneous intrusions emplaced within Cretaceous strata along the central-northern Corona Ridge
There is a lack of igneous intrusion emplacement within the Phanerozoic sedimentary sequences directly above the centralnorthern Corona Ridge itself (Fig. 17). This contrasts with the Flett Sub-basin where significant amounts of igneous intrusions have been emplaced within Cretaceous and lower Paleocene sedimentary rocks (Figs 7,8 and 17). This possibly suggests that magma was less able to exploit the thick crust via basement-bounding faults (e.g. the   (2) the dark grey area within the study area corresponds to the crystalline basement structure of the central-northern Corona Ridge (CR) interpreted by this study (see Fig. 14b).
Along the Cambo High and within the south Rosebank Field, the T40-T45 interval has a very distinct seismic response that is characterized by a series of stacked high-amplitude reflections (Poppitt et al. 2018, p. 378, fig. 6;Duncan et al. 2020, p. 985, fig. 6) that are tied to individual T40-T45 Colsay units (Hardman et al. 2018, p. 77, fig. 7a). Throughout most of the dataset used in this study, no similar Flett Formation seismic response is observed, except for the area south of 213/23-1 (Eriboll) (Fig. 5) which does not appear to correlate with areas containing the thickest presence of the Flett Formation (Fig. 16d). Schofield and Jolley (2013) proposed that where a clastic intrabasaltic depositional system develops, lava-field morphology plays an important role in controlling and diverting sediment input, drainage systems and subsequent reservoir deposition within topographical lows in and around the margins of the lava fields. Understanding the structure present along the Corona Ridge (e.g. palaeohighs and lows underlying sequence T40) prior to (Fig. 15b) and during (Fig. 15c) the deposition of such complex sequences is also, according to Hardman et al. (2018), key to identifying sequence T40-T45 intra-basaltic play fairways within the centralnorthern Corona Ridge area. Our mapping of structures (Fig. 15) underlying such complex depositional systems and our identification of volcanic intervals previously missed within 213/23-1 (Eriboll) (Fig. 11) has facilitated a more detailed understanding of the possible distribution of T40-T45 intra-and sub-basaltic play fairways along the central-northern Corona Ridge. Sequence T40 intra-basaltic Colsay sediment is thought to have predominantly been deposited within topographical lows (Schofield and Jolley 2013) parallel to palaeohighs (NE-SW) present along the Corona Ridge, where drainage systems are thought to have developed (Hardman et al. 2018). This distribution of T40-T45 sediments is also reflected within our mapping (Figs 15 and 16). The NW-SE trend present within the T40-T45 isochron (Fig. 16d) runs parallel to the central-northern Corona Ridge and is thickest within the topographical lows present at the near Top Paleocene level (Fig. 15c). Significant thinning of the Flett Formation from 422 m within well 213/23-1 (Eriboll) to 8 m within 213/25c-1V (North Uist) (over 18 km: Fig. 16d) also reflects the complex distribution of sequence T40-T45 along the Corona Ridge discussed by Schofield and Jolley (2013). This is possibly attributed to the palaeohigh present to the north of 213/25c-1V (North Uist) at the Top Cretaceous (Fig. 15b) and near Top Paleocene levels ( Fig. 15c) that may have acted to divert sediment into the topographical low adjacent to the central-northern Corona Ridge.
Although published research has largely overlooked the possible distribution of intra-and sub-basaltic Paleocene-Eocene-aged (sequence T40) reservoirs NE of the Rosebank Field, an examination of prospectivity by OMV and partners c. 50 km north of the study area, within now relinquished licence area P1997 (Fig. 1b), suggests that reservoir presence and quality is high risk within the Paleocene-Eocene Flett interval and that seismic imaging is too poor to allow calibration with offset wells (OMV 2015).

Extending the intra-and sub-basaltic plays of the Rosebank Field along the central-northern Corona Ridge
The reinterpretation of the Flett Formation within well 213/23-1 (Eriboll) has facilitated the identification of four volcanic intervals and intra-basaltic and sub-basaltic sandstones of the Colsay Member (sequence T40-T45). These intervals were not previously interpreted within 213/23-1 (Eriboll) (Mobil North Sea Ltd 1999a) and were not previously thought to extend this far north along the Corona Ridge from the Rosebank area (Passey and Hitchen 2011;Hardman et al. 2018).
Log signatures of the four volcanic intervals (Fig. 10) we interpret within the Colsay Member section of 213/23-1 (Eriboll) are not representative, as suggested within the composite log ('intrusive volcanics': Mobil North Sea Ltd 1998), of igneous intrusions encountered within the FSB as described by Mark et al. (2018b, p. 741, fig. 6). Instead, the intervals have comparable log characteristics with lavas present across the FSB as described by Nelson et al. (2009) and Watson et al. (2017), and are reinterpreted as such by this study (Fig. 10). The four intervals of volcanic rocks we interpret within 213/23-1 (Eriboll) are also analogous to volcanic image log facies of lava flows from within the Rosebank Field described by Watton et al. (2014, pp. 180-182, fig. 6) and Millett et al. (2021Millett et al. ( , p. 2897 fig. 9).
Our reinterpretation of 213/23-1 (Eriboll) has proved the presence of T40-T45 lava flows and possible intra-and subbasaltic plays north of the Rosebank Field, not previously interpreted this far north (Mobil North Sea Ltd 1999a; Passey and Hitchen 2011;Hardman et al. 2018). Our results may facilitate further understanding of the true extent and distribution of the intrabasaltic Colsay Member play fairway along the northern Corona Ridge.
The results presented in this study emphasize the need for a reevaluation of wells drilled along the northern Corona Ridge: that is, 214/09-1 (Bunnehaven) and 214/04-1 (Tobermory) (Fig. 1b). Specifically, our results highlight the need for a re-evaluation focused on the possible identification and characterization of Paleocene-Eocene intra-and sub-basaltic Colsay Member sandstone (sequence T40-T45) plays that may have previously been missed (as we demonstrate within 213/23-1, Eriboll). The reinterpretation of wells, integrated with the knowledge gained from new wells (e.g. Lyon 208/02-1 and Blackrock 204/05b-2: Fig. 1b) and recent publications (e.g. Jolley et al. 2021), combined with an appreciation of the complexities of sediment-lava interaction, could not only assist with further locations of sequence T40-T45 intra-and sub-basaltic sand fairways but may also facilitate the development of complex reservoirs such as the Rosebank Field. This approach could provide a strategic method to identifying Paleocene-Eocene targets within the underexplored gas-prone area of the northern FSB, an area that (as of September 2022) holds c. 656 bcf of stranded gas as a potential tieback option for any nearby commercial discoveries ( post-drill in-place volume estimates: Tobermory 464 bcf, ExxonMobil 2000; Bunnehaven 192 bcf, ExxonMobil 2001).

Conclusions
The Faroe-Shetland Basin is one of the only remaining exploration frontiers on the UK Continental Shelf and will also arguably be one of the most important areas to contribute to indigenous oil and gas supply during the next 40 years. Particularly prospective areas of the FSB include the Rona Ridge and Corona Ridge, intra-basinal highs along which the Clair Field (Rona Ridge), Rosebank Field and Cambo Field (Corona Ridge) have been discovered. Exploration and research have largely focused on both the Rona Ridge and postrift Paleocene-Eocene-aged prospectivity along the southern Corona Ridge. Whilst hydrocarbon shows and discoveries prove that a working petroleum system is at play within the sub-basins bound by the central-northern Corona Ridge, many aspects of the region's geology are still not fully understood.
Through the interpretation of 3D seismic data integrated with an analysis of three exploration wells (213/23-1 Eriboll, 214/21a-2 South Uist and 213/25c-1V North Uist) we have revealed new insights into the structural configuration, geology and hydrocarbon prospectivity of the central-northern Corona Ridge. We show that: • The Rosebank intra-and sub-basaltic play concept extends to the NE of the Rosebank Field, along the central-northern Corona Ridge, at least into well 213/23-1 (Eriboll). • Based on a re-evaluation of available well data, we suggest that Triassic strata are absent along the Corona Ridge, challenging the previous understanding of Triassic distribution within the central Faroe-Shetland Basin. This interpretation also possibly opens further Devonian-Carboniferous-aged exploration plays of the Clair Group ( previously interpreted as Triassic) and the possible presence of gas-prone Carboniferous coals within the northern Faroe-Shetland Basin. • Using 3D seismic data, we present detailed time-structure maps of the northern-central Corona Ridge. Our results show that the structural configuration of the central-northern Corona Ridge is far more complex than previously understood, including north-south-and NNW-SSE-striking faulting oblique to the basinal NE-SW-striking structural trends within the Faroe-Shetland Basin.