OR/15/072 Deeper seismic interpretation

Gent, C M A¹, Stewart, M A¹, Evans, D J¹, Lamb, R², Alcalde, J^{3, 4}, Heinemann, N⁴ and Akhurst, M C¹. 2015. A summary of the methodology for the seismic stratigraphic interpretation for the 'GlaciStore' bid to IODP. (Energy and Marine Geoscience Programme) British Geological Survey Internal Report, OR/15/072. 1 British Geological Survey 2 University of Manchester 3 University of Aberdeen 4 University of Edinburgh

The interpretation of the deeper part of the succession investigated was undertaken using Petrel software. Initial interpretation of fine-scale seismic packages and first-pass sequence stratigraphic interpretation of three selected lines was undertaken on paper records and digitised in the graphic design software package CorelDraw.

Available data and its origins

For the interpretation of the Mid Eocene-Quaternary section, the project had access to a large volume of data covering the central and northern North Sea area. Both 2D and 3D seismic reflection data were available, alongside a number of hydrocarbon exploration and production wells. However, due to time constraints, the volume of 3D data and identified miss-ties between 2D and 3D datasets, it was agreed that the focus of the deeper seismic interpretation should be on seismic lines from the regional CNST82RE 2D survey and the Statoil NSR05-statoil-st501 line (‘crooked line’).

The vast majority of the data were contained within a large project file derived from a Geographix interpretation project. These data included seismic reflection SEG-Y files, well headers and traces, geophysical well logs, well tops and velocity data (Table 3). There were also some existing seismic ‘picks’ that provided a useful framework from which to extend the current interpretation.

There were 426 wells loaded to the project, including 261 wells from the Norwegian sector. Of the wells available, 84 had no depth values, only coordinates and 50 wells were loaded in Two Way Travel Time (TWTT), mostly located in UK quadrants 15 and 16 and Norwegian quadrants 15, 16 and 17. UK-sector wells were most likely to have been acquired as released wells. The origin of Norwegian sector well data is unknown, although is most likely as a result of previous projects associated with the Sleipner facility.

Table 3 File origins for data used (excluding seismic data)
File Name	Created	Last Modified	Contents and Origin
TIME-DEPTH.xls	15/10/1996	16/08/2007	Created for Sleipner Project, it contains checkshot values for Norwegian and UK sector wells. Norwegian data most likely supplied by Statoil in 1996. UK data origin unknown (likely from old Wellog data). Also a global velocity curve has been calculated using the checkshot values.
Time Depth Pairs from Query3.xls	24/02/2014	29/04/2014	One-way travel time (OWTT) and two-way travel time (TWTT) values for Formation Tops in wells across the CNS and NNS, 9934 formation tops recorded (including sea bed). Recent nomenclature has been assigned to old wells in column ‘LithoStratUpper_NEW’ by Sam Holloway. Values (TWTT and Formation Tops) derived from composite logs with both marked on. File created from a DTI Access database.
cut down time depth pairs fewer columns.xls	24/02/2014	27/05/2014	As above but with 2 fewer columns and empty rows separating each well entry. Wells with only one time-depth pair for an unknown formation top (no nomenclature) have been removed.
Sleipner_SACS2.pet	13/08/2007	13/11/2014	Master project with large volumes of data loaded over the CNS/NNS in UK and Norwegian sectors (wells, seismic, checkshots, formation tops). Probably created by importing a Geographix Sleipner Project into Petrel. Contains all 2D surveys of use and 3D Mega merge. Also contains several interpretations of 2D lines by D Evans (from Feb–May 2014).
Utsira_Picks	--	13/11/2014	Picks of the Top and Base Utsira Sand, most likely from Sleipner Geographix Project. Points converted to lines (surfaces) and exported.

Data quality

Seismic reflection data

The quality of the seismic reflection data available was taken into consideration. Ideally, a good quality seismic line would be processed to a useful depth, eliminating as many sea floor multiples as possible whilst retaining the shallower events. Furthermore, as this study is being undertaken across the width of the North Sea, it should have been shot and recorded at a suitable station spacing to image medium- to large-scale features (kilometres to tens of kilometres) on a regional scale. A summary of the available seismic reflection surveys is provided in Table 4. As can be seen (Table 4), the CNST82RE 2D survey and the Statoil NSR05-statoil-st501 lines have Common Depth Point (CDP) spacing of 12.5 m which is most suitable for the purposes of this project.

Table 4 Available seismic reflection data
Survey	Type	CDP spacing (m)	No. Lines/Cubes	Comments
CNST82	2D survey	12.5	22 Lines	Survey over the Central North Sea
NNST84	2D survey	25.0	18 Lines	Survey over the Northern North Sea
NVGT88	2D survey	12.5	24 Lines	Survey over the North Viking Graben (Norwegian Sector)
VGST89	2D survey	12.5	9 Lines (17 sections)	Survey over the Viking Graben
CNST86	2D survey	25.0	26 Lines (95 sections)	Survey over the Central North Sea (most coverage in UK sector). Formed of many short 2D sections of line
NSR05	2D survey	12.5	1 Line	Statoil line across the Central North Sea. Cropped to 2000 ms TWTT. Also known as the ‘crooked line’
PGS MEGA SURVEY	3D survey compilation	12.5	28 Cubes	Coverage over the UK sector of the Central North Sea

Geophysical well log quality

There were a number of wells with loaded geophysical logs in Petrel. The geophysical logs used were compared to the composite logs to verify the curves, namely the gamma ray and sonic curves. Some of the logs were run over sections of the well with casing and casing shoes set within the zones of interest. These may be from surface to a certain depth or sections set in deeper parts of the borehole that may have been prone to breakouts or deterioration and they disrupt the geophysical logging. They are generally recognised by subdued log responses or straight line gaps between two logging runs. Due to the straight line nature of the geophysical log curves especially in casing shoes, they were easily recognisable, verified by cross-referencing the composite log with the disrupted sections of the geophysical logs and consciously omitted.

Selected data

The seismic survey which showed the greatest regional coverage whilst also comprising whole lines, rather than numerous segments was the regional CNST82RE survey. The CNST82RE survey is used because it has fair to good quality data across the area of interest and E–W lines run close to parallel to the Statoil NSR05-Statoil-st501 (‘crooked line’) line. The previous interpretations are also a good starting point from which interpretations could be extended with a good degree of confidence into the rest of the regional survey. The 3D cubes were tested but due to their limited regional extent (per cube) and identified miss-ties between 2D and 3D data, producing several regional 2D interpretations across multiple seismic cubes would have been impractical in the given time of the project. Surveys NNST84 and NVGT88 were considered to cover areas too far to the north, VGST89 is not of regional extent as well as being split into multiple sections for one line. CNST86 covers mainly the UK sector with a few more regional lines but some lines are also cut into much smaller segments (kilometres scale), which is unusable for a regional interpretation. To make the short lines more useful, they would have had to have been merged, which wasn’t deemed necessary as the CNST82RE was already seen as suitable for a regional interpretation.

The well data loaded into Petrel, are from vertical or near vertical hydrocarbon exploration, appraisal and production wells, the majority have gamma ray and sonic logs. Using, if available, velocity data (e.g. time-depth pairs — see Velocity analyses), enabled some of the wells to be displayed in TWTT. However, the majority of the wells were not converted into TWTT because many were outside of the area of interest (Quadrants 15–16 and Norwegian Quadrants N 15, 16, 17). Where possible, composite logs were also used to verify the geophysical well logs, although much of the Eocene and younger units, particularly in the Quaternary were not well subdivided or absent.

In addition, the CNST82RE line 04 contains a 20 kilometre section where the navigation data was disrupted; there was also a similar smaller 2.3 kilometre zone of distortion on line CNST82RE-05.

Interpretation method

The CNST82RE-04 line was interpreted by Drs G Kirby and S Holloway in 2003 for the CO₂ NGCAS (Next Generation Capture and Storage Project) as it is in the vicinity of the Forties structure (20 kilometres to the north) and used to model migration of CO₂ injected into the Forties Sandstone in 2D. This interpretation provided a useful guide and starting point for interpretation of the remainder of the CNST82RE survey, although the ages of the strata representing the shallower reflections have subsequently been revised.

There was a slight (1–2 wavelets) discrepancy between the seismic reflection data and the corresponding geological logs and interpreted formation tops in the well data. This could be due to a number of reasons, e.g. inaccuracies in the velocity data, or the borehole being offline. Therefore, when interpreting the seismic data, the loaded formation top data was used as a guide rather than as rigid tie-in points.

The interpretation was initially undertaken for the major reflection events in the shallower sections, including the Mid Quaternary down to, roughly, the Oligocene section. However, after starting the interpretation it was agreed to divide the activity into a shallower and deeper interpretation, appropriate to the fields of expertise and experience of the interpreters. For the ‘deeper seismic’ interpretation, and the window of interest shifted to deeper strata, including any major reflection events from the Top Utsira Sand (or equivalent) down to the Top Balder Formation (or equivalent) (Figure 5).

The general interpretation workflow steps were to identify any major horizons, using the loaded formation tops as a guide, and extrapolating them across several lines: both East-West and North-South. Once these packages had been identified then the interpretation was further refined, highlighting more discrete events or sequences. Using published literature and the formation top information, these units were assigned possible ages or to lithostratigraphical units (e.g. Top Horda Unit, Top Mid-Miocene Unconformity) and described by their seismic reflection character. What these minor events or internal boundaries correspond to, on a local regional or global scale, has not been investigated in this study.

Results

The interpretation

Several of the surfaces (reflections) were interpreted over the majority of the survey and some of the more subtle boundaries were also interpreted over a large area. However, the majority of the more subtle boundaries and packages were interpreted only in detail on three of the regional lines (CNST82RE-04, 05 and 06). A summary of the interpretation is given in Table 5.

Table 5 Selected interpreted horizons and their areal extent
Survey	Interpretation Extent	Lines Interpreted	Comments
Top ‘Chaotic’ Zone	Regional	6	Quaternary feature
Top and Base Utsira Sands	Regional	7
Mid Miocene Unconformity	Regional	13	Unconformity is very subtle in places
Near Top Polygonal Faulting	Regional	6
Top Intra-Eocene ‘Quiet’ Zone	Regional	13
Eocene Prograding surfaces	Local	4	Complicated prograding sequence
Internal Eocene Intra-‘Quiet’ Surface	Local	4
Top Horda ‘Noisy’ Zone	Regional	13	Possibly Top Horda Formation
Top Balder Formation	Regional	16	Base of interpretation

During the earlier phases of the interpretation, horizons subsequently incorporated in the ‘shallower seismic’ interpretation (see Shallower seismic interpretation) were included; Figure 5a shows an interpretation including the shallower features. Of note is the aerially restricted Chaotic Zone within what is now classified as the Quaternary succession. This sedimentary package was mappable on a number of lines and shows some features which could be interpreted as shelf channel erosion and basin floor deposition. After initial interpretation, the shallower part of the sequence was reassigned (see Shallower seismic interpretation). Interpretation then highlighted units at greater depth, initially, the strongest reflectors were picked, subsequently, any interior reflectors or minor packages were identified. The stratigraphical intervals that are the focus of the interpretation are:

The Base Utsira to Mid Miocene Unconformity (MMU)
The MMU to the Top Eocene
The Mid Eocene prograding units and basin interactions

A preliminary sequence stratigraphic interpretation of the western part of seismic line CNST82RE-06 (inset in Figure 5a) from CDP 4397 to 8198 was undertaken, as much as could be achieved within project resources. The interpretation (Figure 5b) sets the sequence stratigraphic and depositional framework for scientific drilling sites within the UK sector of the North Sea and the method followed and terminology used is that of Vail (1987)^[1] and was applied to the horizons identified in Table 5. The preliminary sequence stratigraphic interpretation is not constrained by palaeontological data, and data from wells 15/17-7 and 15/19-1 was used as a guide rather than certain formation boundaries. CNST82RE-06 (inset in Figure 5a) from CDP 4397 to 8198 was undertaken, as much as could be achieved within project resources. The interpretation (Figure 5b) sets the sequence stratigraphic and depositional framework for scientific drilling sites within the UK sector of the North Sea and the method followed and terminology used is that of Vail (1987)^[1] and was applied to the horizons identified in Table 5. The preliminary sequence stratigraphic interpretation is not constrained by palaeontological data, and data from wells 15/17-7 and 15/19-1 was used as a guide rather than certain formation boundaries.

The same prograding units were reviewed by all ‘deeper seismic’ interpreters to determine confidence in the pick and to assess against possible noise in the 2D seismic reflection data. As a result, more likely or more ‘real’ events were chosen for further work. Some of the more obvious horizons were mapped over a larger regional scale, whereas interpretation of the finer-scale, lower-order boundaries was restricted for the most part to lines 04, 05 and 06 (Figure 5b). The restriction to three lines is to allow for an interpretation which would highlight the complexity of the sedimentation and lateral continuity of sequences that may be encountered in any appraisal of a storage site.

Velocity analyses

Velocity data from previously compiled spreadsheets (Table 3 and Available data and its origins) and well checkshot files for 13 wells in close proximity to the regional lines was interpreted and plotted to create a ‘regional’ velocity curve (Figure 6). The interpretation and subsequent maps (Figure 7) were all produced in the time domain, however, the depth conversion was not completed so all work has remained in TWTT.

References

↑ ^1.0 ^1.1 VAIL, P R. 1987, Seismic stratigraphy interpretation using sequence stratigraphy: Part 1: Seismic stratigraphy interpretation procedure, in Bally, A W, ed., Atlas of seismic stratigraphy, American Association of Petroleum Geologists Studies in Geology 27, p.1–10.

[Vail_1987-1] 1.0 ^1.1 VAIL, P R. 1987, Seismic stratigraphy interpretation using sequence stratigraphy: Part 1: Seismic stratigraphy interpretation procedure, in Bally, A W, ed., Atlas of seismic stratigraphy, American Association of Petroleum Geologists Studies in Geology 27, p.1–10.

[1]