As producing fields enter late life, successful mitigation of subsurface risk is critical in order to justify continued infill well drilling. Despite penetrations by more than 100 wells and the availability of modern 3D seismic across the Scott and Telford fields (Outer Moray Firth, UK Central North Sea), reservoir absence has led to repeated development well failures. A new model now attributes reservoir attenuation to Late Jurassic footwall uplift and erosion.

The 2015 Scott J40 well encountered a reduced Upper Jurassic reservoir section of Lower Scott sands, with Upper Scott and Piper sands absent. Reinterpretation ascribed reservoir truncation to the effect of Late Jurassic footwall uplift and erosion rather than fault cut-out as previously interpreted. The failed 2016 Telford F6 well was shown to have drilled a thick Kimmeridge Clay hanging-wall section north of the Telford Fault, rather than the southern footwall section targeted. Planning for the Scott J43 and four subsequent wells implemented lessons learned from these failures and mitigated reservoir risk while optimizing reserves.

The Scott and Piper sand distributions are likely to reflect early growth folding before the main Mid-Kimmeridgian–Early Tithonian phase of NE–SW extensional faulting and footwall uplift, which saw the Piper and Scott reservoirs eroded at footwall crests and locally reworked as deep-water Claymore sands. Later Mid-Tithonian–Early Berriasian east–west faulting during the opening of the Witch Ground Graben saw major crestal synsedimentary erosion at Telford and continued erosion at the western crest of Scott Block 1b, while the footwall to the east was partially downfaulted.

The continued development of mature fields is likely to require the consideration of increasingly challenging drilling locations that carry inherent subsurface risk: for example, in targeting smaller and more structurally complex attic locations or previously undrilled blocks at the margins of proven accumulations.

Despite the drilling of more than 100 wells in the Scott and Telford fields in the Outer Moray Firth (UK Central North Sea) and the availability of multiple generations of towed streamer and ocean-bottom 3D seismic data, a number of development well failures have continued to deliver surprises, with some 25% of wells encountering missing or significantly attenuated reservoir sections. It is therefore critically important to understand the mechanisms responsible for reservoir development and preservation if geological risks are to be managed to an acceptable level and future target selection errors avoided.

Previous interpretations regarded missing reservoir sections as a result of complex cut-out by subseismic faults that were not mappable and which therefore presented an irreducible subsurface risk. Nevertheless, following the depletion of early well stock, updip attic areas may provide the largest remaining incremental reserve targets and these have continued to be prioritized as infill well targets with mixed results.

An initial review of the distribution of failed Scott Field wells showed the reservoir risk to increase towards major faults in some cases while noting complex fault sets in multiple orientations. Seismic interpretation uncertainty was seen as implicit from the acoustic response of the Upper Jurassic sedimentary units, since neither the Piper nor the Scott reservoir showed a strong seismic response, and a proxy near-top reservoir surface at the Base Kimmeridge Clay Formation is a complex unconformity with truncation below and onlap above.

This paper presents a stratigraphic and structural study of the Scott and Telford area, aiming to identify the controls on reservoir presence and preservation in order to de-risk future drilling while allowing enhanced reservoir modelling to assist efficient reserves recovery and the targeting of water-injection reservoir support.

While the structural history of the Central North Sea is commonly considered in the context of Middle Jurassic regional uplift (Underhill and Partington 1993), interpretations presented here emphasize variations in stratigraphic development as a result of Late Jurassic faulting and footwall uplift, with tectonics influencing the Oxfordian–Kimmeridgian deposition of Scott and Piper reservoir sands and their localized erosion associated with Late Jurassic extensional faulting and footwall crestal erosion.

Interpretation of multiple regional 2D and field-scale 3D towed streamer and ocean-bottom seismic surveys has been integrated with stratigraphic correlation of historical exploration, appraisal and development wells in the Scott and Telford area. Previous interpretations are continuously updated as new seismic and well data become available. Core samples from initial exploration and development drilling formed the subject of several major sedimentological studies of the key reservoir units (Thickpenny and Russell 2000, 2010, 2012), with these results continuously updated and correlated with new data as additional wells are drilled.

These datasets have been used to build static and dynamic reservoir models in Petrel™ for the Scott and Telford fields, which again are updated as new seismic and well data become available. This paper is based on seismic interpretation and well correlation work carried out between 2015 and 2023, incorporating recent results from a 2021 Scott and Telford ocean bottom node (OBN) 3D seismic survey and the Scott J47 well that came on stream in early 2024.

The Scott Field (Guscott et al. 2003) lies in the Outer Moray Firth of the UK Central North Sea (Fig. 1) within a series of fault blocks forming a large high in the eastern Witch Ground Graben. The Telford Field (Syms et al. 1999) lies 9 km to the south, hosted within a major east–west fault block south of the Telford Fault. A small separate closure in the west of this block forms the Marmion sector of the field.

Discovery

The first well on the Scott structure was 15/22-3, drilled in 1977 on a crestal location and encountering a major NE–SW fault beneath the Kimmeridge Clay Formation without finding any reservoir (Guscott et al. 2003). The Scott Field was subsequently discovered by well 15/22-4 in 1983, which encountered some 200 ft (approx. 60 m) of Upper Jurassic sandstones in two packages: the Upper Oxfordian Scott Member of the Sgiath Formation and the Lower Kimmeridgian Piper Formation (referred to hereafter as Scott and Piper sands). These reservoir units were separated by a 50 ft (approx. 15 m) shale interval (Harker et al. 1993). The lower Scott sands were oil bearing, while the Piper sands were water wet. Following the drilling of a disappointing downdip well penetrating only thin and poor-quality water-wet sands at 15/22-5 in 1985, the potential of the field for a major hydrocarbon accumulation did not become clear until the 15/21-15 well in 1987, which encountered almost 400 ft (approx. 120 m) of net pay. This was subsequently followed by a successful field appraisal, and discovery by well 15/22-9 in 1990 of the South Scott accumulation, which in turn was confirmed and appraised by well 15/21a-43 in 1991.

Although the Telford Field was first drilled in 1974 by well 15/22-1, which encountered 31 ft (10 m) of good-quality gas-bearing sand, a second well drilled at 15/22-2 found over 200 ft (approx. 60 m) of water-wet sands, and the presence of a commercial discovery was accordingly delayed until drilling of the Marmion discovery well 15/21a-44 in 1991.

Development and production history

The Scott Field was developed via a series of subsea and platform wells, supported by water injection and producing via two linked platforms into the Forties Pipeline System and the Scottish Area Gas Evacuation (SAGE) gas export line. Following first oil in September 1993, production peaked at 220 000 bopd in October 1995, declining to 40 000 bopd by 2002 as the water cut rose. The 2023 production was approx. 6000 bopd (Fig. 2) at a water cut of more than 95% and supported by water injection at rates of up to 160 000 bwpd. Total production to July 2024 was 450 MMbbl of oil plus 348 Bcf of gas, with total water injection of 1.84 Bbbl.

The Telford Field was developed as a subsea tieback to the Scott Platform (Jewell and Ward 1997) via a series of injector–producer well pairs, coming on stream in October 1996. Production peaked at approx. 42 000 bopd in October 1997, declining to 10 000 bopd by 2005 as the water cut increased to 90%. The 2023 production was 1500 bopd (Fig. 2) at a water cut of approx. 95% and supported by water injection rates of up to 40 000 bwpd. Total production to July 2024 was 109 MMbbl of oil plus 286 Bcf of gas, with total water injection of 331 MMbbl.

As production from the Scott and Telford fields has matured over time, the Scott Platform maximum fluid handling capacity has remained broadly constant, with peak total liquid rates in 2023 of approx. 220 000 blpd, which is close to initial peak production rates but with produced water progressively substituting for produced oil.

Structure

NE–SW and east–west faulting in the area was associated with the Jurassic opening of the Central North Sea and Witch Ground Graben, respectively (Guscott et al. 2003). A structure map of the area is shown in Figure 3, while Figure 4 provides a representative seismic section.

Scott Field

The main closure of the Scott Field comprises a large composite fault block high at the intersection of the east–west-trending Witch Ground Graben with the NE–SW-trending Theta Graben (Boldy and Brealey 1990; Hibbert and Mackertich 1993). A variety of fault orientations define the structure of the component fault blocks as they wrap around the Scott High. Facies distributions and fault-block rotation geometries indicate that the Scott structure was subject to extensional fault movements through the Oxfordian–Berriasian interval, with several fault sets showing evidence of displacement in different senses during this period.

A number of fault block panels have been mapped (Fig. 3), with multiple development wells drilled in each:

  • Block 1b: the updip, crestal sector of the field;

  • Block 1: a contiguous fault block located to the south of Block 1b;

  • Block 1w: a small, fault-bounded and pressure-isolated compartment to the west of Block 1;

  • Block 1s: a hanging wall terrace in the south of the Scott Field and north of the Telford Fault;

  • Block 2: a SW-dipping fault block to the west of Block 1b;

  • Block 2a: a small fault block north of Block 2; and

  • blocks 3, 3A and 4: three small fault blocks located to the north of Block 1b.

Telford Field

The field is located in the crestal area of a major east–west footwall block south of the Telford Fault (Fig. 3). NE–SW faulting is limited to the eastern and western sectors of the Telford Ridge, with segmentation in the west defining a separate closure forming the Marmion sector of the field. Development wells have been drilled in all sectors of Telford (including in Marmion).

Key units and seismic markers

This subsection describes the main seismic units recognized and mapped in the area. Key reflectors and sequence stratigraphic correlations are listed in Table 1 and represented diagrammatically in Figure 5.

A Carboniferous section was penetrated by several wells including 15/22-J17, although the unit is weakly reflective and not easily mappable on seismic data. By contrast, the Zechstein (Permian)–Carboniferous interface forms a high-amplitude doublet that is laterally persistent on structural highs and forms the acoustic basement across the Scott and Telford area. The acoustic impedance of the Top Zechstein anhydrite is lower where more deeply buried in the hanging walls of major faults. Major overburden thickness changes across faults may see this reflector confused with overlying high-amplitude events formed by the Jurassic Rattray volcanics or Skene Coal (see below).

A Triassic section of Smith Bank Formation shales occurs in the Witch Ground Graben (Harker et al. 1993) and was found in the Scott J17 well. This section is overlain by a series of igneous extrusives of the Middle Jurassic Rattray Member, defined by Richards et al. (1993) as part of the Middle Jurassic Pentland Formation. Variable Rattray thicknesses reflect the development of a complex Bajocian–Bathonian (and potentially basal Callovian) succession of prograding volcanics sourced from the Buchan–Glenn Fissure System, together with associated reworked volcaniclastics (Quirie et al. 2019, 2020). This package reaches a thickness of 1079 ft (329 m) in the Ivanhoe Field discovery well 15/21a-3, while thinning to the NE and locally filling erosive relief (as in the north of Scott Block 4).

Figure 6 illustrates the Upper Jurassic–lowermost Cretaceous sequence stratigraphy of the Scott area (after Copestake and Partington 2023b, c, showing correlations with the lithostratigraphic scheme of Rattey and Hayward 1993).

Resting unconformably on the Rattray volcanics (Slater et al. 2024), the Sgiath Formation (Harker et al. 1993) commences with the Skene Member, comprising thinly bedded shales, sandstones and a coal marker that is locally traceable on seismic data. These non-marine facies in the transgressive systems tract (TST) of the lower J52 sequence pass up across a probable transgressive surface into dark grey offshore marine shales of the Saltire Member, which correlate with the upper part of J52 (Copestake and Partington 2023c). Although less evident in downfaulted areas to the west, a weak angular discordance is evident in crestal locations, whereas in central and eastern areas the lower J52 Skene Coal is absent with a northward-dipping/-prograding Top Rattray reflector directly overlain by laterally continuous upper J52 Saltire Member shales, resulting in subtle variability in the Top Rattray seismic surface. Harker et al. (1993) correlated the Skene and Saltire members with the Alness Spiculite Member developed in the Inner Moray Firth Basin, an interpretation supported by Copestake and Partington (2023c).

The Saltire Member is in turn overlain by two shallow-marine reservoir packages of Late Jurassic age: the Scott and Piper sands (Fig. 7). The Scott Member of the Sgiath Formation (‘Scott sands’; basal Upper Oxfordian, J54) is present across the Scott Field and locally in west Telford. Lower and Upper Scott units are commonly separated by a regionally persistent intraformational mudstone termed here the ‘Intra-Scott Shale’, alternatively referred to as the G Shale (Guscott et al. 2003). The Scott sands are subdivided into correlatable units designated as Lower Scott A, B and C, and Upper Scott A and B.

The Scott sands are overlain by the ‘Mid Shale’, a regionally correlatable horizon corresponding to the Rosenkrantzi maximum flooding surface (MFS) of latest Oxfordian age, J55 (Hesketh and Underhill 2002; Copestake and Partington 2023c) (Figs 6 and 7), which is in turn overlain by a further package of shallow-water sandstones assigned to the Piper Formation (‘Piper sands’; Basal Kimmeridgian, J56). The Piper sands are reduced or missing in some Scott Field wells and on the crest of the Telford Field (Syms et al. 1999). Lower and Upper Piper Formation units are commonly differentiated in thicker sections where they are separated by the ‘Intra-Piper Shale’.

The upper, shalier part of the Piper Formation of Copestake and Partington (2023c) comprises siltstones and shales previously described as the Kimmeridge–Piper Transition unit or ‘KPT’ (Guscott et al. 2003) and displays marked thickness variations. The ‘Top KPT’ surface provides an acoustically hard, low-amplitude and laterally discontinuous seismic reflection at the base of the acoustically soft Kimmeridge Clay Formation (KCF; see Fig. 5). Neither the Top Piper sands nor the Top Scott sands provide significant seismic reflections but the Top KPT presents a mappable near proxy horizon for a ‘Top Reservoir’ pick across the area, although it is critical to recognize that this reflector provides no evidence of the lithologies beneath it. Guscott et al. (2003) postulated correlation of the KPT with the ‘Transgressive Unit’ in the Rob Roy Field (Boldy and Brealey 1990) and the ‘Hot Sand Unit’ in the Tartan Field (Coward et al. 1991).

In places, the lowermost 200 ft (60 m) of the KCF shows moderately faster velocities, while locally generating a higher-amplitude seismic reflector that passes laterally into and merges with the Top KPT reflector, creating complex seismic tuning effects across the contact. This configuration can result in a mispick of the KPT.

The overlying KCF section is dominated by acoustically soft mudstones (Kadolsky et al. 1999, 2005), and usually displays a slow seismic velocity and complex, low-amplitude internal reflections. Cemented silts and carbonate layers punctuate the succession, together with a unit of cemented turbidite sands associated with the Hudlestoni MFS. This sand unit creates a prominent mappable seismic reflector and a useful time marker in the Theta Graben area to the west of the Scott Field. The KCF can therefore be broadly divided above and below this level into two seismic units: a lower ‘KCF1’ package, which is present within, and thickens into, structural lows, and an overlying and passively onlapping ‘KCF2’.

Although not seismically mappable across structural highs, continuation of the KCF1 unit in condensed sections updip and onlapping onto footwall relief is suggested by the presence of a distinctive gamma-ray spike observed on well logs within the lowermost horizons resting on the Top KPT unconformity (Coward et al. 1991). Detailed log correlation allows mapping of this package and reveals its absence towards the margins of Scott Blocks 1 and 2 facing the Theta Graben (see Fig. 3). A Top KCF1 (Hudlestoni) reflector is seismically mappable and has been dated as Early Tithonian by Copestake and Partington (2023c).

The overlying KCF2 package locally commences with deep-water turbidites of the Claymore sands (see Partington et al. 1993). These sands and the package overlying them appear to onlap footwall relief on the Top KPT surface. The base of this KCF2 package is assigned a later Early Tithonian age by Copestake and Partington (2023c). The Top KCF2 forms a regionally mappable seismic marker that is more commonly described as the Base Cretaceous Unconformity (BCU), characterized by truncation below it and onlap above. This event, dated as being of Late Berriasian age, is unconformably overlain by onlapping strata of the Lower Cretaceous Cromer Knoll Group (Copestake and Partington 2023c) in basinal areas and by the Chalk on structural highs.

Lateral variations in stratigraphy

While the Scott Field has major reserves in both the Scott and Piper sands, Telford reserves are mostly hosted in the Piper sands. Broadly, Scott sands thin to the east and north, while Piper sands thin to the west and south. Even with good well control and modern 3D seismic data, mapping the stratigraphy in detail is challenging, noting as above that neither the Scott nor Piper sands present a clear acoustic reflection.

Marked differences in sequences drilled across Scott and Telford partly reflect two distinct unconformities:

  • a Middle Oxfordian unconformity (Harker et al. 1993) evident from variable onlap of the Sgiath Formation onto Rattray volcanics (Fig. 7); and

  • a Mid-Kimmeridgian unconformity that corresponds with the complex and composite Top KPT reflector seen on seismic data. This surface shows significant erosional truncation relief, with onlap of the KCF above (see Fig. 5). At the crest of the Scott Field, the Base KCF–Top KPT surface forms a high-amplitude seismic event that is recognizable above an erosionally truncated stratigraphy, ranging from Piper, Scott, Saltire and Rattray to the Zechstein and Carboniferous beneath.

Sedimentation and tectonics

The Top Rattray Member forms an important unconformity in the Scott and Telford area, and shows a variable and commonly poor seismic response associated with the onlap of overlying shales onto a range of Rattray lithologies. Despite some evidence of erosional truncation at this surface, there is no evidence of volcaniclastic material in the Scott and Piper sands, consistent with a principal sediment source area for these deposits from the Fladen Ground Spur to the NE (see Fig. 1b) and with more minor input from the Halibut Horst to the west (Harker et al. 1993). Figure 8 is a schematic sketch showing the sedimentary architecture of the Scott and Piper sands across the Scott and Telford area.

Figure 9 presents facies maps for the Scott and Piper sands, compiled from detailed sedimentological core analysis (Thickpenny and Russell 2000, 2012; Guscott et al. 2003). Based on analogy with modern coastal clastic systems in Texas and North Carolina, USA (e.g. see Susman and Heron 1979; Heron et al. 1984), this earlier work interpreted the Scott sands as sand-rich barrier island and tidal facies in the NW, with protected back-barrier, lagoonal and washover facies towards the SE. Shallow-water sands are developed in the Lower Piper in the north of the field, while the Upper Piper comprises shallow-water sands passing progressively offshore towards the NW.

Piper sands are absent at the crest of the field and, while Figure 9 is reconstructed to show the deposition of Scott sands across the entire field, their absence from the crest of Block 1b with shallow-water facies surrounding it means that partial emergence of this area during Scott deposition cannot be entirely excluded.

The primary sediment supply in coastal and shallow-marine sedimentary environments is mainly from rivers, with clastic material redistributed effectively along coastlines over considerable distances by longshore drift. Modern barrier islands in the Saloum Delta of Senegal migrate at rates of up to 120 m a−1 (Went et al. 2013) and accordingly the preservation potential of barrier island structures may be limited. Reflecting the mobile dynamics of shoreface environments, ancient barrier island facies commonly show a complex sedimentary architecture reflecting changes in sea level and climate, as well as subsequent pervasive reworking (Mulhern et al. 2021). Although fossil barrier islands may be difficult to identify, their existence is therefore interpreted from facies associations containing back-barrier fill, lower- and upper-shoreface, proximal upper-shoreface and tidal-channel deposits. The deposition of stacked and overstepping successions typically occurs in transgressive intervals during periods of relative sea-level rise.

Scott and Piper facies distributions in Figure 9 indicate a greater marine influence in the NW, suggesting synsedimentary control through early NE–SW fault movements from Mid-Oxfordian times onwards. Subtle variations in accommodation space, with greater subsidence to the NW and relative uplift to the SE, are likely to have controlled the location of barrier island (Scott) and shoreface (Piper) environments, as well as the locations of thin and condensed sections in crestal areas. A viable interpretation is therefore that the stratigraphic architecture (Fig. 8) and facies distributions (Fig. 9) of Scott and Piper sands may reflect sedimentation during early growth folding prior to the onset of surface faulting (cf. Gawthorpe et al. 1997).

As initial high production rates from the Scott and Telford fields declined in the late 1990s (see Fig. 2), infill drilling began to be considered. The Scott Platform has 28 well slots and the strategy adopted was to reutilize these by sidetracking high water cut platform wellbores to areas where dry oil remained, either in previously undrilled panels or in updip and previously undrained attic locations. The latter approach implicitly carries increased subsurface risk in targeting smaller volumes in more structurally complex areas.

2003–04

A new OBC seismic survey was acquired in 2001 and reprocessed in 2003 to integrate with an existing 1996 towed streamer dataset. Following seismic reinterpretation, a drilling campaign in 2003–04 targeting reserves towards the margins of the field saw mixed success (Brook et al. 2010). Structural complexities were encountered in marginal areas of the seismic volume where migration was suboptimal.

2005–07

Further seismic interpretation in 2005 focused on improved fault mapping, ahead of an infill campaign that drilled new producer wells (J30, J32z and J34). These wells targeted unswept areas in Blocks 1 and 1b, away from the margins of the field. In addition, the J31 well was drilled as a downdip injector supporting J16 in Block 1w and the J33 well aimed to drain updip attic reserves in Block 2. The J35 well was another updip attic producer in Block 1b, encountering thinner than expected Scott sands overlain by Piper sands, with this configuration thought to reflect faulting through the wellbore at this location.

This selection of updip well targets proved effective in boosting the Scott Field production rates, which recovered from 15 000 bopd back towards 35 000 bopd. However, early water breakthrough followed in wells J30, J32z and J35, and the overall production uplift from the new wells was relatively brief and below forecasts (Brook et al. 2010).

The J36 well was drilled in 2007 from the Scott Platform to the Marmion sector of the Telford Field. At the time, this was the longest well in terms of measured depth, and the well with the longest step-out on the asset. The initial borehole J36 and first sidetrack J36z both failed in the overburden, owing to the shallow drilling angle through the overburden that was required to achieve such a step-out. A second sidetrack J36y then encountered a dry attic Piper reservoir in the Marmion block in the Telford Field.

2009–10

The next Scott infill campaign comprised wells J37, J38 and J39. Results were mixed and production uplift was modest. The 2009 Scott J37 well and sidetrack J37z targeted attic volumes in Block 4; however, both wells failed to find any reservoir. A second sidetrack J37y was then drilled as a production accelerator from Piper sands in Block 1b. The well location was downdip of the existing production offtake and a swept reservoir was encountered with only 0.03 MMbbl of incremental production achieved.

The Scott J38 well (also 2009) targeted attic volumes in Block 1 and failed to find any reservoir before being sidetracked downdip as J38z and successfully encountering hydrocarbon-bearing sands. This well produced 1.95 MMbbl, yielding production acceleration but without adding material new reserves.

Finally, the Scott J39 well (also 2009) targeted by-passed hydrocarbons mapped against the main intrablock fault in Block 1b. The well encountered a full reservoir sequence but water cut rose swiftly. The well produced 0.25 MMbbl before being sidetracked as the Scott J40 well in the 2015 drilling campaign.

2015–24

After acquisition of a new towed streamer seismic survey in 2010 and refurbishment of the Scott drilling rig in 2014, a further infill well campaign began in 2015. Target selection aimed to strike a balance between low- and higher-risk targets, including twins of some wells that had been abandoned early when water cuts reached 50% and now presented low-risk targets, as well as structurally complex, and therefore higher-risk, attic locations.

Figure 10 and Table 2 show the location and result of wells drilled in the Scott Field during this time, with the Telford F6 well result also included (discussion of F6 follows below).

Scott J40 well

This first well in the new campaign targeted an area updip of the J16 producer in Block 1w. Pre-drill prognosis was for 198 ft (60 m) of Piper sands and 507 ft (154 m) of Scott sands. However, the well encountered only 373 ft (114 m) of predominantly Lower Scott sands. Nevertheless, when brought online in December 2015, the well produced at high initial rates of up to 25 000 bopd before declining relatively rapidly due to a lack of reservoir support following a mechanical failure in the downdip injector well J31.

Telford F6 well

This well aimed to provide an infill producer in west Telford and reached total depth (TD) in early 2016, targeting attic reserves updip of the Telford F5 well, itself an updip sidetrack of the previous producer F3z. The F6 well was sited based on a combination of reprocessed 2001 OBC seismic data and a spectrally whitened volume from the 2010 towed streamer survey. The section drilled did not match pre-drill prognosis, instead comprising a thicker than expected KCF section and only 8 ft (2.4 m) of basal Lower Scott sands above Saltire Formation shales. The well was suspended to allow reuse of the top hole as a donor well for a future sidetrack if required.

Following the surprise results of these two wells, a choice of conservative locations for the next two infill targets provided time for failure analysis before further drilling.

Scott J41z well

J41 was conceived as a downdip injector in Block 1w. After a mechanical sidetrack, the J41z injector was successful in sustaining production from the J40 well, which had produced 4 MMbbl of oil by July 2023.

Scott J42 well

J42 was designed as an updip twin of the successful Block 2 well J33, which had ceased production following a failed workover in late 2015. The new well successfully delivered initial rates of 13 000 bopd.

Scott J40 well

The J40 well result revealed a fault-block structure in Block 1w that was more complex than anticipated before drilling. The absence of Upper Scott sands in the well was initially thought to show the borehole trajectory crossing briefly into a hanging-wall KCF section before drilling a thinner Lower Scott section in a second downthrown block. An alternative interpretation could see the deviated well trajectory as having grazed intra-KCF Claymore sands in the hanging wall before entering Scott sands reservoir in the footwall (Fig. 11).

Telford F6 well

A range of alternative structural scenarios was considered in order to explain the absence of almost the entire reservoir section in the Telford F6 well (Fig. 12).

Fault section cut-out

The interpretations of several previous wells, such as the Scott J35 well described above, invoked fault cut-outs of various portions of the Scott and Piper sands, and this was the first scenario considered. In the case of the Telford F6 well, the Piper reservoir interpreted at the well location (Fig. 12a) proved to be missing and a complex two-fault configuration was initially proposed to explain this (Fig. 12b). However, robust seismic interpretations could not be constructed on this basis and the hypothesis was rejected as unlikely.

Sedimentary slumping

Previous studies of the Northern North Sea fields Statfjord (Hesthammer and Fossen 1999), Gullfaks (Yielding et al. 1999), Snorre (Berger and Roberts 1999), Brent (McLeod and Underhill 1999) and Ninian (Underhill et al. 1997) demonstrated sedimentary slumping from the footwall crests of major fault blocks. A scenario was constructed envisaging transport of the key reservoir target section from the footwall to the hanging wall (Fig. 12c). However, with the seismic data showing continuous reflectivity in the hanging-wall section, there was no clear evidence of a displaced block and this hypothesis was also seen as unlikely.

Seismic processing artefacts and related interpretation issues

Mapping of the F6 target area had made use of a 2012 reprocessing of a 2010 towed streamer survey across the west Telford area. However, the original seismic data as acquired contained relatively high noise. The data had been reprocessed using a spectral whitening routine with the aim of improving reflection continuity, with apparently positive results obtained. It was on this dataset that the F6 well was sited.

A post-drill review of multiple seismic volumes allowed comparison of this reprocessed towed streamer seismic volume with a 2001 OBC seismic survey, which displayed a lower frequency range and lower noise. Near the Telford Fault, the 2012 towed streamer reprocessing showed footwall reflectors extending for a significant distance further north than their equivalent events on the 2001 OBC data, as subsequently confirmed in a 2021 OBN re-shoot (Fig. 12d). The previous interpretation had therefore implied a larger gross rock volume (GRV), while misleadingly extending the reservoir target area further to the north.

Interpretation of the OBC and OBN data at the F6 location now suggested that the northward continuations of seismic reflectors on the reprocessed towed streamer data were seismic artefacts, with the well initially penetrating the hanging wall of the Telford Fault up to 200 m off structure to the north (Fig. 12d). This explained the presence of only a thin reservoir section, directly beneath a much thicker than expected KCF.

Findings

While earlier interpretations had suggested that failed wells had drilled into fault planes that were not identifiable on seismic data, the number of wells where fault cut-out was inferred to explain incomplete stratigraphy appeared unreasonably high. Accordingly, the following two alternative explanations were put forward.

Misplaced well target due to inaccurate seismic imaging

As outlined above, the Telford F6 well was reinterpreted as a structural near miss, drilling the hanging wall rather than the footwall of the Marmion to West Telford Fault. A comparison of imaging results from the towed streamer and OBC seismic surveys served to confirm that the well location was incorrectly placed.

Reservoir truncation due to footwall erosion

Reinterpretation of the Scott J40 well and several earlier well failures suggested that erosion at fault-block crests was a key cause of reservoir truncation. An accurate description of the Base KCF unconformity was therefore critical to define the preservation of the Scott and Piper sands at any location in the area. With neither the Scott nor Piper sands presenting clear seismic reflectors, structure mapping is commonly carried out on the Base KCF–Top KPT event, with a proxy ‘Top Reservoir’ horizon created by contouring down below it. However, the Base KCF–Top KPT surface forms a significant unconformity that carries a risk of reservoir erosion on structural highs at the frontal culmination of footwall blocks. Accurate mapping of this unconformity and understanding of its structural and stratigraphic controls is therefore essential in order to reduce the risk of well failure due to reservoir truncation or removal.

Footwall erosion of the reservoir along fault block crests is described in detail in the following subsections.

Depth of erosion

Well correlations and seismic mapping demonstrate that a range of strata are truncated at the Base KCF–Top KPT surface across the area. On the leading edge of footwall crests east and west of the Theta Graben, the KCF may rest on Middle Jurassic Rattray volcanics (e.g. in Scott wells J13 and J12: Figs 13 and 14) or on the Saltire Formation (e.g. in well 15/22-3 and Scott wells J38, J37z, J5 and J3). The Scott J37 well found the KCF resting directly on the Carboniferous. In each of these locations, the Scott, Piper and KPT sections are entirely absent, while downslope from footwall crests there is variable truncation with the Scott, Piper and then KPT sections progressively preserved.

It is likely that this erosional truncation was largely subaerial, with the fault-block crests forming ‘footwall islands’ during part of the Early–Mid-Kimmeridgian interval. A similar evolution of Late Jurassic rift-margin emergence was explored in the Northern North Sea by Roberts et al. (2019).

Onlap onto the unconformity

While a relatively complete section of the KCF is present downdip, progressive onlap onto footwall crests sees only a thin and condensed basal KCF section draping the unconformity in updip settings. In more crestal footwall locations, a condensed KCF1 section may be present, although not resolved seismically, with seismic sections appearing to show the unconformity overlain by an onlapping KCF2 section.

Controls on reservoir preservation

A west–east transect and well correlation (Fig. 15) from Block 2 to Block 1b across en echelon fault segments in the Theta Graben illustrates the variation in stratigraphy between the hanging wall in the west and the eroded footwall in the east. The Base Saltire–Top KPT thickness reduces towards the footwall crest from 420 ft (128 m) in the Scott D4 well to zero in the Scott J3 well, where a condensed KCF interval with a prominent basal gamma-ray spike overlies a truncated section of Rattray volcanics. Meanwhile, the Top KPT is more than 2500 ft (760 m) deeper in the downdip hanging-wall well A1z than in Scott wells J11 and J17, which were drilled on the footwall.

These changes reflect footwall uplift and erosion of the Piper, Scott and Saltire units at the crest of Block 1b in association with NE–SW fault-controlled subsidence of the hanging wall in Block 2. Although present and forming a prominent reflector in the hanging wall, the Skene Coal is not recognized in the footwall, reflecting eastward onlap onto the Top Rattray surface (see Fig. 7). This configuration potentially provides the conditions for a seismic mispick between the Skene Coal in the west and the Top Zechstein/Top Carboniferous in the east (Fig. 15b).

Figure 16 provides a south–north well correlation and seismic line from Block 1 into Block 1b, showing progressive truncation towards the north at the KPT unconformity. Block 1 appears to have developed within a fault relay zone where maximum displacement on the NE–SW fault bounding Block 1b was transferred westwards to a fault extending northeastwards from the northern limit of Block 1 (see Fig. 3).

Scott J43 well

Following a reassessment of subsurface risks, the Scott J43 well was conceived to target an infill location in Block 3 (Fig. 17). A structural model was developed to assess the timing of fault movements and reservoir erosion (Fig. 18) in order to define a suitable target that would optimize updip volume while assuring drilling success.

Structural model

The Scott J43 well area is compartmentalized by multiple fault sets (see Fig. 17) trending NE–SW, parallel to the Theta Graben, east–west, parallel to the Witch Ground Graben, and NW–SE, stepping down from the Scott Field culmination in Block 1b to the south. The well was sited as far updip as possible in order to maximize reserves while simultaneously mitigating against reservoir erosion risk at the footwall crest.

Detailed block-specific structural, stratigraphic and reservoir models were constructed before drilling, paying particular attention to the correlation of key Upper Jurassic marker horizons within and above the reservoir package (Figs 19 and 20). Seismic picks were tied to offset Scott wells J9 (which encountered a reduced reservoir section of Lower Scott only) and J15 (which penetrated a full Piper and Scott succession; see Fig. 17 for well locations).

Earlier interpretations had shown Upper Scott and Piper sands faulted out at the J9 well location (Fig. 20a). Following the drilling of the Scott J40 and Telford F6 wells, this interpretation was revised, now attributing the reduced reservoir section in the Scott J9 well to footwall erosion (Fig. 20b).

An assessment of the footwall erosional geometry shown in Figure 20b provided critical de-risking inputs for the decision to drill at this location – not only in order to mitigate the risk of reservoir attenuation but also because, as shown in Figure 20c, the target volumetrics and reservoir offtake rate predictions were dependent on the detailed distribution of reservoir units which was incorporated into the reservoir model for the structure.

Scott J43 well result

On drilling, the Scott J43 well penetrated a near complete and mildly truncated Scott and Piper reservoir package, preserving all but the top 2 ft (0.6 m) of the KPT below the Base KCF unconformity. This successful result verified the geological model while vindicating the rigorous approach to well planning.

Following the mixed results of earlier infill drilling, the positive result of the Scott J43 well provided confidence that the reasons for previous well failures could now be better understood, therefore reducing subsurface risk in future wells.

The Scott Joint Venture planned and drilled a further four infill wells between 2018 and 2023, although progress was delayed by an extended rig recertification operation in 2019 and by the COVID-19 pandemic of 2020–21, as well as by power generation issues on the Scott Platform during 2022–23. A summary of results from this subsequent infill drilling programme is given below.

Scott J44 well

Drilled in 2019–20, this relatively aggressive step-out well targeted undrained Scott sands in Block 1s. The well followed the positive outcome of the Scott J43 well and was also successful, with reservoir thicknesses on prognosis.

Scott J45z well

This 2020 injector well was sited in a low-risk downdip location in central Block 2. While requiring a sidetrack following mud losses during drilling, the well was successful and encountered reservoir thicknesses on prognosis, providing additional water injection to Block 1.

Scott J46 well

This downdip Block 2 producer well successfully drilled a full Scott and Piper section and came online in 2021.

Scott J47 well

This well targeted attic reserves near the crest of Block 2. Subsurface risk was seen as high in this updip location and a fallback downdip target was selected for a geological sidetrack in the event of failure.

The KPT and Upper Piper sands were absent but the well found Lower Piper sands as prognosed, above a thin Mid Shale and a condensed interval of Scott sands that displayed a high net/gross consistent with sedimentary onlap towards the east of Block 2. The Intra-Scott Shale appeared to be thin or absent at this updip location (Fig. 21).

The positive result of the Scott J47 well confirmed the eighth consecutive successful Scott infill well, raising confidence that subsurface risk could be effectively managed. The future strategy is again likely to use a combination of low risk near ‘twin’ well opportunities with moderately higher-risk attic locations mapped on the new seismic survey acquired in 2021 and step-outs towards the margins of the field development area (FDA).

Timing and style of faulting

The mapping of the Scott and Telford field area showed two key fault orientations (see Fig. 22).

NE–SW

This fault trend delineates the western edges of Scott blocks 1 and 1b (Fig. 22), and defines Block 2 as an intermediate relay structure stepping down towards the Theta Graben (Boldy and Brealey 1990). NE–SW faults show no expression at the Top KCF2–BCU, indicating that displacement had ended by the Berriasian. Facies and stratigraphic changes across the area (see above) suggest that this fault trend saw extensional synsedimentary movement from Mid-Oxfordian times onwards, consistent with general evidence from the seismic data that suggest a degree of Oxfordian growth folding on blind NE–SW faults (e.g. see Fig. 15).

During an initial phase of deposition, the Skene Coal was laid down in a defined area in the west of the Scott Field, consistent with the infill of subaerially exposed topography at the Top Rattray surface (Quirie et al. 2019). Subsequently, facies trends in both the Scott and Piper sands record deeper-water conditions towards the NW; with shallower-water facies deposited towards the footwall crest of Block 1 (see Fig. 9).

Truncated reservoir sections at the western edges of blocks 1b, 1, 1w, 3 and 4 (Figs 11 and 15) are interpreted as recording erosion during subaerial exposure (see Fig. 19). Reservoir sections are missing in the NW of Block 1b, reflecting strong footwall erosion. Piper and partial Upper Scott sections are eroded on crestal Block 1 in Scott wells A3, J38z and J4, and the absence of Piper and Upper Scott sections in the Scott J40 well suggests that blocks 1 and 1w were subaerially emergent as footwall islands and eroded to a similar level before Block 1w was downthrown to its current position.

The KCF shows a marked thickness increase across the NE–SW fault bounding Block 1 to the east and Block 2 to the west (Fig. 23). The Hudlestoni MFS between KCF1 and KCF2 provides a useful seismic marker, showing that most synsedimentary growth occurred during the deposition of KCF1. By contrast, KCF2 shows only differential subsidence while passively onlapping footwall relief.

Sands within the KCF succession appear to onlap Lower Scott sand in the Scott J40 well in the footwall and are revealed on an amplitude map (Fig. 23a) as filling a re-entrant cut into the footwall crest of Block 1w, while forming fan-shaped anomalies in the hanging wall to the NW. It therefore appears likely that these intra-KCF Claymore sands are at least locally derived as second-cycle deposits from the erosion of Scott and Piper reservoir sands in the footwall followed by sedimentation in the hanging wall, a conclusion supported by onlap of the Hudlestoni reflector onto the KPT in several locations (e.g. see Figs 9 and 15).

East–west

A second set of faults parallel to the Witch Ground Graben includes those which separate blocks 1b, 1, 3 and 4. These faults are downthrown to the north in each case, with the KCF2 unit onlapping footwall relief and thickening markedly downdip (Fig. 24). Differential subsidence continued locally into the Early Cretaceous, with relief across the major fault separating Block 1b and Block 4 being onlapped by Valhall Formation deposits. This fault and the large-offset east–west fault defining the Telford Field structure show displacement at all levels from Top Carboniferous to the KCF, although offset is variable along discrete fault segments and linked splays that are discontinuously connected along strike.

Figure 24 shows substantial erosion at the Top KPT horizon at the crest of Block 1b where this area forms the footwall to NE–SW faults at the edge of the Theta Graben (see Fig. 22a). This erosion began in the Mid-Kimmeridgian and may have continued locally during the Tithonian on the northwestern crestal flank of Block 1b, which remained uplifted during later east–west faulting. Fault traces in this area show segments of differing orientations, suggesting formation via the linking of multiple fault splays.

The geometries shown in Figure 24 also allow an approximate estimate of the amount of Late Jurassic footwall erosion. The two-way travel time (TWTT) offset between the reconstructed and eroded crests of Block 1b and Block 1 is approx. 128 ms, equivalent at a seismic velocity of 3000m s−1 to a vertical offset of 192 m (630 ft).

The presence of local highs at the BCU horizon (see the central part of seismic line B–B′ in Figs 22b; see also 23b) might initially be considered to provide evidence of structural inversion. However, these features coincide with footwall crests below, which present a high at every stratigraphic level from the Zechstein through to the KPT. The presence of these highs at the BCU level is therefore interpreted as reflecting continued post-tectonic differential subsidence during the deposition of Lower Cretaceous shales.

Footwall uplift and subaerial erosion

The concept that the footwalls of large extensional fault blocks could experience relative uplift and crestal erosion was first developed from models of basin formation (Jackson and McKenzie 1983; Barr 1987a; Yielding 1990; Roberts et al. 1993). A combination of extensional fault displacement and high regional heat flow during rifting (McKenzie 1978; Friedmann and Burbank 1995) can see the faulted margins of rift basins raised above the regional base level due to flexural uplift of the lithosphere (Fernández-Blanco et al. 2019) and isostatic effects following ductile crustal thinning. During periods of sea-level fall, emergent footwall fault blocks may be subject to subaerial erosion, resulting in highly complex sedimentary and stratigraphic architecture around ‘footwall islands’ (Barr 1987b; Roberts et al. 1993; Elliott et al. 2011). Shallow marine erosional processes may also result in the peneplanation of uplifted fault blocks (Martinez et al. in press).

Synsedimentary fault-block movement, footwall uplift and erosion may all exert a strong influence on sedimentation in coastal depositional systems, with ancient analogues including Middle Jurassic clastic deposits in the Hebrides Basin of NW Scotland (Archer et al. 2019) and mixed carbonate–clastic sediments in the Neogene of the Gulf of Corinth (Gawthorpe et al. 2018) and the Gulf of Suez (Azabi 2024), where outcrop studies illustrate the morphology and scale of footwall relief created (Sharp et al. 2000).

Another modern outcrop analogue illustrating rotated fault-block geometries comes from the southern Black Mountains of Death Valley, USA (Fig. 25a). In this example, an extensional normal fault cuts a series of laterally continuous stratigraphic units. Extension is associated with rotation, relative uplift and crestal erosion of the footwall fault block. Although the section has been eroded to the right of the fault, the relative elevation of the footwall crest is higher and this area has seen erosion to a deeper stratigraphic level than on the downdip flanks, causing a characteristic erosional sculpting of the footwall crest similar to that recognized on many regional seismic lines from the Northern North Sea (e.g. see Platt 1995).

This pattern of crestal sculpting results in an unconformity with significant erosional relief and across which the stratigraphic separation following later flooding is greatest immediately adjacent to the footwall. An imaginary well drilled on the footwall crest to the right of the red fault in Figure 25a would encounter a thinner section beneath the unconformity because the upper three stratigraphic units below it are now missing due to subaerial erosion, in comparison with the more complete section beneath the unconformity that would be encountered on the downdip footwall area towards the right-hand side of the image.

Footwall uplift in the North Sea

The geology of the North Sea has been studied for six decades, with the acquisition of regional 2D and 3D seismic surveys and the drilling of more than 5000 wells providing stratigraphic control (Underhill and Richardson 2022). In adapting and building on the simple geometrical model for footwall uplift developed by Yielding and Roberts (1992), the role of footwall uplift in the tectonic and sedimentary evolution of the North Sea has been reviewed in regional structural models of rift evolution (Yielding et al. 1992; Roberts et al. 1993, 2019; Ter Voorde et al. 1997, 2000; Odinsen et al. 2000a, b) and in detailed structural studies in the Northern North Sea, including in the Gullfaks Field and at the Frøya High in Norway (Fossen and Hesthammer 1998; Nesse 2019; Gresseth et al. 2023a, b; Nakken et al. 2023), and in the Ninian and Brent fields in the UK (Underhill et al. 1997; McLeod and Underhill 1999).

Figure 25b presents a schematic footwall erosion model for the Scott and Telford area based on the geometry shown in Figure 25a, indicating the stratigraphic development of the Scott and Piper sands in key wells.

Fault movement and block rotation

A NE–SW section across the Scott Field (see Fig. 24) shows that, following an initial phase of footwall uplift and erosion on NE–SW faults during the Mid-Kimmeridgian, a period of east–west fault growth with block rotation ensued during deposition of the Late Kimmeridgian–Early Tithonian KCF1.

The pattern of thickness changes in the KCF1 section and in the KCF2 section overlying it suggests some later displacement on the NE–SW faults as well. Given the strong evidence for NE–SW-orientated extension during this period, these later movements could not have been simply strike-slip but instead were likely to have been accommodated by the development of relay zones and complex fault linkage geometries such as those seen around Scott Block 2 and at the margins of the Theta Graben.

Where these complex fault trends intersect at the strongly uplifted western footwall crest of Block 1b, erosion initiated in the Mid-Kimmeridgian appears locally to have continued (or to have been periodically reinitiated) for a considerable interval during the deposition of KCF2 unit. The complex patterns of erosion and preservation of the Scott and Piper sands can therefore be seen to reflect the influence and complex interaction of both principal fault trends.

Geological history

Abundant well and seismic data over the Scott and Telford fields allow a detailed reconstruction of the effects of footwall uplift in an area affected by multiple phases of faulting. Figure 26 provides a summary of the structural evolution according to the five stages described below.

Regional uplift

Regional Middle Jurassic uplift in the Central North Sea (Underhill and Partington 1993) was associated with the extrusion of Rattray volcanics and the deposition of volcaniclastics. Reworking and subaerial erosion led to the formation of irregular relief that was transgressed and unconformably overlain by Late Jurassic sediments (Quirie et al. 2020), commencing in the Scott area with Saltire Formation mudstones.

Late Oxfordian–Early Kimmeridgian extension, fault-block tilting and shallow marine deposition

This period of high Middle Jurassic heat flow and regional uplift was followed in Oxfordian times by the onset of extension. The Scott and Piper sands show marked lateral facies and thickness variations that are likely to have reflected growth folding on blind faults beneath. In each case, marine influence is interpreted to have been greatest towards the NW (Thickpenny and Russell 2000, 2012). Eustatic sea-level fluctuations during deposition of Scott shallow-water sands led to the successive deposition of shallowing-upward sedimentary packages that are correlatable on logs and in core as Lower Scott C, B and A.

In the Scott Field, depocentres of these successive Scott sand units were progressively displaced to the SE, potentially reflecting onlap in this direction during progressive transgression onto a Middle Jurassic high. Piper sand thicknesses decrease to the west, consistent with a proximal–distal transition and increasing distance from a sediment source area in the east, as well as with the effects of subsequent footwall erosion.

In the Telford Field, Scott sands are present only to the west of a NW–SE fault separating west and central Telford (Figs 27a and 28). This is consistent with basal onlap onto a fault-bounded high towards the east.

Mid-Kimmeridgian NE–SW extensional faulting, footwall uplift and crestal erosion

An important phase of extension during the Mid-Kimmeridgian resulted in displacement on NE–SW faults. This was accompanied by footwall uplift of the flanks of the Theta Graben, which resulted in subaerial exposure of the footwall crests of fault blocks and the localized erosion of the recently deposited Piper and Scott sands, particularly on the NE–SW-trending southwestern flanks of blocks 1, 1b and 1w.

Late Kimmeridgian–Mid-Tithonian east–west faulting

Continued extension during the Late Kimmeridgian–Mid-Tithonian saw a second, lesser phase of movement on east–west faults parallel to the Witch Ground Graben. Continued subaerial exposure of the pre-existing footwall high in the northwestern parts of blocks 1b, 1 and 1w was associated with further erosion of the Piper and Scott sands and resedimentation within localized Claymore sand deposits downdip.

There is also evidence of footwall erosion at Telford, where the crestal wells 15/22-1, 15/22-13, G12, G12z and G15 to the south of the Telford Fault show eroded reservoir sections containing no uppermost Piper Formation or ‘KPT’. The areas affected lie to the SE of NE–SW faults or NE–SW fault offsets in the Telford Fault trace (Figs 27b and 28). An interpretation of this area as a structural high is supported by the presence of the Sola Formation resting directly on Lower Piper deposits, with the Valhall Formation absent here and potentially pinched out downdip (see also Fig. 4).

Mid-Tithonian–Late Berriasian faulting and passive sedimentary onlap

Mid–Late Tithonian transgression and local reactivation of NE–SW faults saw renewed submergence of a faulted and erosionally sculpted unconformity surface and passive onlap of the KCF onto truncated sections of older Jurassic strata beneath. The widespread presence of at least a thin veneer of the youngest KCF2 shales above this unconformity across the Scott and Telford fields indicates infill and complete burial of earlier erosional relief by Late Berriasian times, although the Top KCF–BCU reflector (Kyrkjebø et al. 2004) and the overlying Valhall Formation show local deflections (see Fig. 24), recording further minor structural adjustments above footwall crests and continuing hanging-wall differential subsidence into the Early Cretaceous.

The east–west faults show shallower dips in places where their upper segments cut post-Mid-Kimmeridgian KCF2 strata, indicating local fault reactivation that was most likely to have been extensional in origin, while also potentially reflecting continuing adjustment to differential subsidence. By contrast, the upper segments of reactivated NE–SW faults in the Theta Graben trend record late movements on splays developed at a higher angle. These are commonly associated with a small positive structure at the BCU level, suggesting wrench movement on NE–SW faults during this time.

A strongly reduced KCF section encountered in East Telford well G20 reflects the continuing influence of footwall crestal relief during this interval. By contrast, wells located further south and lower on the Telford structure show thicker Piper and KCF sections.

Figure 29 shows an integrated geological model illustrating the structural and sedimentary history of the Scott Field. Facies distributions, thicknesses and stratigraphic development in the Scott and Piper sands record an element of early tectonic control on Oxfordian–Early Kimmeridgian sedimentation. This was followed by a major phase of extension and footwall uplift that led to the subaerial exposure and erosion of fault-block crests orientated parallel to the Theta Graben. The relative sequencing of fault displacements as described above places this principal phase of footwall uplift and NE–SW Theta Graben-parallel fault movement within a relatively tightly constrained interval between the later part of the Early Kimmeridgian and the Mid-Kimmeridgian.

Later east–west faulting during the opening of the Witch Ground Graben saw compartmentalization of the crest of blocks 1b, 1 and 1w, leading to further local reservoir erosion and the resedimentation of second-cycle Claymore sands downdip.

Additional lessons learned

While the Top KPT–Base KCF event forms a useful proxy for the Top Reservoir in the Scott and Telford area, previous interpretations failed to recognize the scale of faulting and the extent of footwall erosion. Revised mapping and log analysis now allows the identification of a NE–SW-orientated zone along the flanks of the Theta Graben (see Fig. 13), where 15 wells have encountered attenuated or missing reservoir sections. Ten of these wells display a gamma-ray spike at the base of the KCF, suggesting condensed sedimentation.

Fault-controlled footwall uplift conditioned a zone of reservoir erosion across an area from the western margin of Block 3A in the north to Block 1w in the south, with a correlative conformity observed in areas downdip. Later sea-level rise and a long period of thermal subsidence saw slow deposition of a condensed and diachronous basal KCF1 interval in crestal locations, which were subsequently overstepped and onlapped by upper KCF2 shales (see Fig. 26).

Future wells drilled in the footwall of NE–SW faults can expect to find attenuated reservoir sections. Many wells which were previously interpreted to have intersected faults without encountering drilling losses are now considered instead to have been drilled through the Mid-Kimmeridgian unconformity. Previous positive assessments of drilling risk will now also be recalibrated for locations drilled near to faults where fractures may be expected.

Attenuated reservoir sections on the flanks of footwall crests may have experienced feldspar or calcite cement dissolution due to weathering and meteoric water influx (Shanmugam and Higgins 1988), as reported from the Brent and Magnus sandstones of the Northern North Sea (Emery et al. 1990; Bjørlykke et al. 1992). However, thin sections from the Scott Field show examples of kaolinite filling pore spaces created by feldspar dissolution, with cementation overprinting through the formation of quartz overgrowths and carbonate also resulting in reduced reservoir quality. Given a clear variability in reservoir quality between different sedimentary facies in the Scott Field, and a decline in macroporosity with increasing depth, any variations in reservoir quality at elevated footwall locations are nevertheless difficult to differentiate from broader sedimentological and burial diagenetic effects (e.g. see Worden and Burley 2003).

An improved understanding of the structural and sedimentological controls on reservoir presence has proved effective in reducing subsurface risk while increasing confidence in infill drilling. While the longevity of continuing production from the Scott and Telford fields also depends on a range of facilities engineering, commercial and economic factors, the future infill target inventory can now be risked more effectively. This has already proved important in extending field life through the success of the Scott infill wells drilled since 2015, which by early 2024 had already added circa 18 MMbbl to the field's ultimate recovery. Current subsurface work continues to evaluate potential infill well targets in the various Scott Field fault blocks, as well as in west and central Telford, with the aim of prolonging production life while maximizing economic recovery and deferring asset decommissioning towards the latest 2020s or early 2030s.

The Scott and Telford fields of the UK Central North Sea produce oil and gas from high-quality Scott and Piper sand reservoirs. Data from more than 100 wells and multiple seismic surveys allow stratigraphic correlation, with the limitation that neither reservoir displays a mappable seismic reflector. The fields show a complex structure and variable reservoir development and preservation as a result of two phases of Late Jurassic faulting.

Following Mid-Jurassic regional uplift and volcanism, the onset of extension in the early part of the Late Jurassic saw an interval of growth folding during deposition of the Upper Oxfordian Scott and Lower Kimmeridgian Piper shallow-marine sandstones on structural highs in the Scott and Telford fields, with marine influence increasing to the NW. While Scott sand thicknesses are greatest in the west, reflecting progressive basal onlap, Piper sands are thicker in the east, reflecting a transition to offshore facies in the west.

These reservoir units were displaced during two intervals of extension on NE–SW and later east–west fault trends, respectively. First, a Mid-Kimmeridgian phase of NE–SW faulting coincided with low relative sea level, leading to subaerial exposure and the erosion of Piper and Upper Scott sands on footwall crests. Following sea-level rise and deposition of the lower part of the Kimmeridge Clay Formation (KCF1), Late Kimmeridgian–Mid-Tithonian faulting saw segmentation of footwall blocks and further erosion and fluvial incision at the crest of Scott Block 1w, with resedimentation of second-cycle Claymore sands in submarine fans downslope. Piper sands were also eroded from crestal areas of the Telford Field at this time. Finally, a later period of thermal subsidence and the local reactivation of linking NE–SW extensional faults saw significant deepening of the basin areas and passive onlap of the upper part of the Kimmeridge Clay Formation (KCF2) onto significant erosional and fault relief on the now submerged footwall crests.

A clearer understanding of this complex Late Jurassic–earliest Cretaceous sedimentary and structural history now explains previously puzzling development well failures and will help to reduce subsurface risk in planning future infill wells through the remaining late field life.

We thank the operator of the Scott and Telford fields, CNOOC Petroleum Europe Ltd, and Joint Venture partners Waldorf Operations Ltd, Dana Petroleum (E&P) Ltd, Energean UK Ltd and Neo Energy Exploration UK Ltd, as well as Rockflow Resources Ltd for their support of this work and permission to publish. The seismic data shown in Figures 12a–c are reproduced courtesy of TGS.

We are keen to acknowledge the technical insights gained and the vast volume of detailed work carried out by successive subsurface operator teams working on the Scott and Telford area over the past five decades, and especially CNOOC colleagues in Uxbridge and Aberdeen during infill drilling from 2015 to 2024. Geoff May and Michael Overstolz (CNOOC) and Fiona Goodfellow (Energean) kindly reviewed previous manuscript drafts.

We are grateful to CNOOC for supporting the cost of Open Access publication and our thanks go to Graham Yielding for editorial support, to Tim Needham for a thoughtful review, and, in particular, to Stuart Archer for providing many detailed and constructive suggestions which greatly improved the paper. N.H. Platt is grateful for inspiration from Joe Cartwright, Peter Philip, Mike Badley and Alan Roberts.

NHP: conceptualization (lead), formal analysis (lead), investigation (lead), writing – original draft (lead), writing – review & editing (lead); AM: formal analysis (supporting), investigation (supporting), writing – original draft (supporting), writing – review & editing (supporting); RKM: conceptualization (supporting), formal analysis (supporting), investigation (supporting), writing – original draft (supporting), writing – review & editing (supporting); AJA: conceptualization (supporting), formal analysis (supporting), investigation (supporting), writing – original draft (supporting), writing – review & editing (supporting); RA: conceptualization (supporting), formal analysis (supporting), investigation (supporting), writing – original draft (supporting), writing – review & editing (supporting).

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.

All data generated or analysed during this study are included in this published article (and, if present, its supplementary information files).