Maintaining the daily production rate, and hence prolonging the field life, of a mature asset is largely dependent upon locating and producing the oil that has been bypassed by the existing well stock. At the Alba field (Central North Sea, UK Continental Shelf (UKCS)), recent efforts to identify the bypassed reserves and subsequently optimize future well paths have focused on the use of dedicated time-lapse (4-D) seismic to better understand the field-wide variation in remaining oil thickness.

Background

NRMS values, in a window above the reservoir, between the 2008 monitor and 2002 baseline co-processed datasets. Note the larger NRMS values in the vicintiy of the three surface production facilities (red squares). (Images courtesy of Chevron)

The Alba reservoir is a high net-to-gross, high-porosity, unconsolidated Eocene turbidite channel at an average depth of 6,400 ft (1,950 m). Since production began in 1994, more than 360 MMbbl of heavy oil (19° API) have been extracted via horizontal wells with waterflood support. Daily production currently sits at 32,000 bbl from 28 wells.

Efficient drainage of the Alba reservoir, and hence the magnitude of ultimate field recovery, is strongly dependent on locating production wells as high in the reservoir as possible. However, due to a near-zero compressional P-wave impedance contrast between Alba oil sands and overlying shales, picking the top reservoir on conventional seismic data (downgoing P wave, upgoing P wave) has proven to be extremely difficult. This has been somewhat remedied by the acquisition in 1998 of a converted-wave seismic survey (downgoing P wave, upgoing shear or S wave), as the sands and shales are more easily distinguishable on the resultant P/S reflectivity dataset.

Since then, the technical focus has been on accurately locating the oil-water contact (OWC) such that, for new producers, the standoff from the OWC (i.e., the oil column thickness) is maximized.

Data gathering and processing

Example of the quadrature difference data across Alba. The 2008 OWC (in blue) is obtained by picking the zero crossing at the top of the 4-D signal. Also shown are Top Reservoir (green), Base Reservoir (cyan), and the original OWC (orange). The region between the 2008 OWC and Top Reservoir is the Remaining Oil Column (in two-way travel time).

In July 2008, the Alba partnership acquired the first dedicated monitor survey over the field, replicating the survey parameters of a 2002 towed-streamer spec survey. Throughout the planning and execution of the survey, significant effort was put in to minimizing the combined source and receiver errors while repeating the 2002 vessel headings (even when undershooting the surface production facilities). A post-survey repeatability analysis confirmed that this effort had been rewarded, as 95% of the source crossline positions and 95% of the source inline positions were within 33 and 1.6 ft (10 and 0.5 m), respectively, of the planned locations.

The 2008 monitor dataset was then co-processed through a 4-D sequence focused on imaging fluid movement with the 2002 preferred baseline dataset and an additional pre-production baseline dataset from 1991. A fast-track volume, which, in due course, influenced drilling activities in 2009, was available within two months of acquisition, and the final 4-D products were available within six months of acquisition.

As with acquiring the data, the significant effort put in to the planning, execution, and quality control during processing of the data was duly rewarded with low normalized root mean squares (NRMS) values (the standard industry measure of repeatability noise within 4-D datasets). Between the 2002 baseline and 2008 monitor datasets, the field-wide average of NRMS in a window above the reservoir is ~15%, as shown in Figure 1. This is an excellent result for streamer acquisition in the North Sea and indicates that a high-quality dataset has been produced.

Remaining oil thickness

Recoverable oil thickness (ROT) across the Alba field as of 2008. Areas with red indicate the thickest columns of recoverable oil (up to 40 ft or 12 m). The volumes of produced oil (white spheres) and injected water (blue spheres) between 2002 and 2008 are indicated by the size of the spheres.

Pre-survey synthetic seismic modeling indicated that any changes observed in P-wave impedance over time would be largely due to changes in fluid saturation (i.e., injected seawater replacing oil), with only a minimal contribution from pressure changes. Therefore, due to the high data quality, it was possible to map the 2008 OWC directly from the seismic quadrature difference data (see Figure 2).

Combining the 2008 OWC with the converted-wave top sand interpretation provided the remaining oil column in two-way time, which was then converted to a remaining oil column thickness using the seismic velocity of the oil-bearing sands. From this a recoverable oil thickness (ROT) was calculated by assuming an average porosity and oil saturation for the sands.

The resulting 2008 ROT map across the Alba field is shown in Figure 3, together with the cumulative volumes of produced oil and injected water up to 2008. The clear match between water injector locations and areas of low/zero ROT within Figure 3 is positive evidence that, as predicted, the observed time-lapse signal is a reliable indicator of fluid movement within the reservoir.

A more quantitative assessment of this reliability was obtained through comparison of estimated and actual volumes of oil production in the “12 Area” of the field (indicated on Figure 3). Using a similar workflow to that which produced the ROT map, an estimated oil production for the “12 Area” of 28 MMbbl was obtained from the two-way travel time difference between the original OWC and the 2008 OWC as interpreted on the seismic difference data. This estimate is within 4% of the actual “12 Area” oil production of 27 MMbbl and therefore provides further confidence in the ROT map.

Optimizing future well locations

Making use of the ROT map to optimize the location of future wells requires that the production-induced movement of the OWC observed in the time-lapse data is first rationalized with the reservoir fault framework, top reservoir interpretation, production/injection data, and reservoir simulation models.

This multidisciplinary process has subsequently led to the creation of a revised simulation model that better reflects the post-4-D characterization of the reservoir. Significant differences that may influence the placement of future production and injection wells between this latest model and previous simulation models include the following:

• Certain faults were identified as baffling fluid flow based on the vertical offsets in the OWC. Historically, the transmissibility of faults has been uncertain;

• Water cones beneath producers were observed to be steeper than previously modeled, and the assumed longevity of these cones is also now questioned; and

• Top reservoir interpretation was revised based on inconsistencies with the flushed zone derived from the quadrature difference data.

Revised model

Successful integration of the time-lapse seismic with other pieces of static and dynamic data has led to a revised simulation model, which has, in turn, led to the identification of a set of bypassed oil drilling targets that may otherwise have been discounted. This has all been achieved in less than 24 months.

Acknowledgements & suggested reading

The authors would like to thank Chevron management and the ALBA co-venturers (BP, CIECO Energy (UK) Ltd., ConocoPhillips, Endeavour, Statoil, and Total) for their permission to publish this work. The reader is referred to “Remaining oil thickness and well positioning using 4-D at Alba field, North Sea” by Tura et al. (SEG 2009 Expanded Abstracts) for further information on this project.