Few areas offer the perfect seismic acquisition setting. Weather, sea, topographical, geological, ecological, environmental and manmade conditions, among others present variables that must be managed during seismic acquisition and processing to produce unperturbed subsurface images. This task becomes even more difficult when a survey must be repeated as closely as possible for time-lapse (4-D) purposes - that is, use of 3-D seismic data acquired at different times over the same area to assess reservoir changes over time.

Statoil faced this exact challenge when contemplating using time-lapse monitoring to manage its Norne field, which presents numerous challenges to repeatable acquisition. The Q-Marine technology and services from WesternGeco were selected to meet these challenges. Norne field is the first location in which this technology has been used for both the baseline and subsequent monitor surveys.

Harvesting 4-D's potential

To date, the majority of all 4-D (time-lapse) seismic projects have been conducted in offshore clastic reservoirs, passive margins and rifts, and largely in the North Sea. Reportedly 75% of all seismic conducted in the North Sea today is time-lapse work, and that figure grows as excellent returns on investment continue to be realized. Worldwide 4-D represents less than 10% of the current total seismic market, but this figure is growing with projects in the Gulf of Mexico, West Africa and South America accelerating in pace. As technology and economic limitations are overcome, 4-D seismic is expected to play an increasingly important role in the management of reservoirs worldwide.

Today, 4-D projects continue to focus on improving recovery during a field's waning years. However, some operators, such as Statoil, are beginning to employ a 4-D strategy from production onset to ensure optimum management of a field over its lifetime.

Based on North Sea experience, fields with as little as 40 million bbl remaining recoverable reserves can benefit from use of the time-lapse seismic technique. This figure holds true for about 500 offshore reservoirs today. In the offshore market alone, there is room for significant growth in the application of
4-D seismic. In the deeper waters of the Gulf of Mexico and the Atlantic, where drilling is very expensive, 4-D reservoir monitoring may realize significant financial benefits to oil company asset teams.
In addition to being used to map water flood fronts for fluid saturation and pressure changes and more accurately placed wells, 4-D seismic has a wide number of industry applications. Time-lapse seismic helps to manage production strategies for hydrocarbons, water and reservoir voidage by identifying the connectivity between injection and production wells, locating untapped hydrocarbon compartments and flow barriers, and improving history matching and reservoir predictions. Other applications include monitoring gas storage, CO2 removal, and steam-assisted gravity drainage of heavy oil. It also provides additional interwell details, which may be missing because of today's ability to use only a few horizontal wells to drain a reservoir versus the many vertical wells required in the past.

Some industry analysts have estimated that the growing use of 4-D seismic should have a significant economic impact by 2020, with about 8 million to 11.5 million b/d additional oil production attributable to 4-D seismic by that time. A significant boost in activity will occur as application of 4-D in carbonate reservoirs advances.

Statoil's experience

Now in the declining stages of life, Norne field is being managed by operator Statoil with the help of 4-D seismic monitoring. Norne oil production first came onstream in 1997. From this early field stage Statoil's strategy has been to use time-lapse (4-D) seismic technology to optimize field development and manage production. However, the project has presented some special seismic survey design and execution challenges that had to be overcome to arrive at the high-quality, repeatable seismic images required.
The field was discovered by Statoil and partners in 1991. It is a one-billion-bbl oil in place reservoir that lies 125 miles (200 km) offshore the midwest coast of Norway. First oil flowed in 1997 and gas in 2001. It produces from 12 wells completed in lower and middle Jurassic sandstones that lie about 1.6 miles (2.5 km) below the seafloor in a northeast/southwest trending horst block. A 246-ft (75-m) thick gas cap rides atop a 361-ft (110-m) oil column. Sub-horizontal shale and calcite permeability barriers, as well as faulting, have a major impact on reservoir production, which now is being driven by gas and water injection.

Production flows from five subsea templates into a floating production, storage and offloading (FPSO) vessel moored at the field's center. Current daily oil production is about 132,000 b/d of oil.
Statoil's goal with the time-lapse seismic project has been to accurately map the fluid locations and movements through the reservoir over time, paying close attention to the oil/water contact (OWC). Specifically, detailed fluid and rock property data are being used to better understand the reservoir's current behavior and simulate future fluid movement, which helps with production planning and management.

The characteristics of the Norne project that have made the acquisition and processing of 4-D seismic a particular challenge include: the presence of an FPSO in the survey area, and the weather and sea conditions offshore Norway, which together contribute perturbations to the recordings masking the time-lapse response of the reservoir.

4-D acquisition

The Q-Marine seismic system (see sidebar) is designed to mitigate environmental perturbations resulting in high-resolution repeatable measurements of the reservoir, making it an excellent choice to manage Norne's acquisition challenges. In August 2001, a steerable-streamer seismic survey was conducted over Norne field utilizing this system. This was followed in June 2003 by a similar monitor survey. The Geco Topaz acquired both surveys, assisted by the Western Pacific and Western Inlet, which served as the source vessels required to undershoot the FPSO at midfield.

Low feather angles, straight streamers, constant streamer separation and consistent common midpoint (CMP) fold all have important benefits for 3-D and 4-D processing. Additionally, to achieve high 4-D signal quality, it is necessary to steer accurately to repeat the xyz positions of the sources and receivers (R.W. Calvert et al, EAEG, 2002, 2003). This is especially true on shallow and/or high-frequency targets and deep, lower-frequency targets contaminated by poorly repeatable sea surface multiples and diffractions, as well as on targets distorted by rapidly varying overburden.

A single-source array and six 10,500-ft (3,200-m) steerable, single-sensor streamers with 154-ft (50-m) lateral separations were used. The ability to steer the streamers allowed the receiver array to be towed within 131 ft (40 m) of the FPSO and acquire an image of the reservoir's center.

Very low feather was achieved with 79% of the 2001 data acquired with feathering of less than ±1°. This compares with a typical value of about 32% for nonsteered streamer data acquired in the area. In Figure 1, a comparison between the cumulative feather for a 1992 nonsteered streamer survey and the 2003 monitor survey, the effect that steerable streamers have on feathering is clear.

In 2001, the steering fins were spaced every 1,312 ft (400 m) down each streamer. In 2003, fin spacing was reduced to 984 ft (300 m) to give more steering capability. Added steering allowed efficient reacquisition of the low-feather parts of the 2001 baseline. In addition, some of the anomalies in the 2001 dataset caused by shooting for coverage were minimized to make it easier to acquire future monitor surveys. This was done by using low-feather acquisition on some of the 2001 lines with higher feather values. An added benefit of this strategy was that not all the baseline infill had to be reacquired and less than 1% of new infill was shot.

Figure 2 shows how well the 2001 feathering was repeated in 2003. Around 90% of the 2003 data were acquired with feathering within ±1° of the equivalent 2001 data.

In 2003, although variable crosscurrents of up to 18 in/s (45cm/s) caused the vessel to deviate from the preplot map, over half of the shots were within 9.9 ft (3 m) crossline from the preplot, and 95% of the shots were within 29.5 ft (9 m). As expected, because of the towing configuration, the repositioning was best at the head of the receiver array. However, even the tail-end differences between 2001 and 2003 were less than 65.6 ft (20 m) for much of the prospect.

Data processing

Two adjacent swaths were repeated when acquiring the 2001 baseline survey. Analysis of these data revealed that it would be possible to acquire technically comparable surveys in the future. Further, it was clear that results with low 4-D noise levels could be achieved by using deterministic processing methods. The normalized root mean square (NRMS) measured during these early tests indicated extremely good repeatability, with a value of about 13%.

The close similarity between the repeat lines and experience gained from the 2001 processing tests indicated that the subsequent monitor surveys could be processed independent of the baseline. This was significant, allowing the use of onboard-processed 4-D results only a few days after completing the acquisition of each monitor survey.

In 2003, the 2001 survey was reprocessed before acquiring the new monitor survey. The 2003 data were
then processed on the Geco Topaz through an identical sequence, including deterministic corrections for measured acquisition system variations. The two migrated volumes were differenced 10 days after the completion of the survey.

Single-sensor recording, calibrated source and receivers, and high positional accuracy played an important role in enabling rapid 4-D differencing, without which, neither the rapid response time nor the ability to detect subtle 4-D signals would have been possible. Statoil immediately analyzed the resulting fast-track onboard 4-D volume.

Within 3 weeks of survey completion, Statoil released a revised drilling plan that placed the next well 66 ft (20 m) above and to the side of the originally selected wellpath (Figures 3 and 4). The steered-streamer 4-D results showed a thin, heavily calcified horizontal permeability barrier that was not tight, and a waterfront that actually moved upwards rather than sideways. Had the original wellpath been drilled, at a cost of US $29 million, water would have been produced within a short time. Instead, a very successful horizontal well was subsequently completed and set into production by November 2003.
Onshore processing of the 32.8 sq miles (85 sq km) dataset employed time-consuming, computer-intensive and tailor-made techniques to further enhance the 4-D results. An earlier attempt to compare the 2001 4-D Q-Marine survey to a conventional survey previously acquired over Norne field failed in the free surface demultiple, despite the fact that the 2001 survey was acquired with steered-streamers. So, the 2003 monitor survey was acquired with free surface demultiple in mind. The lengths of the two-boat undershoot lines were extended to the southwest so that when joined with the single-boat deadhead lines shot in the opposite direction, towards the northeast, there would be sufficient data in the common receiver-station domain for the free surface demultiple to not leave a low-fold edge effect at the join between the undershoot and deadhead lines. The resulting excess CMP fold was regularized by coincident trace selection based primarily on similarity of source and receiver azimuth in 2001 and 2003.

A section of the 2003 monitor survey, the 1992-2003 4-D difference, and the 2001-2003 4-D difference from an inline in the northeast part of the area are shown in Figure 5. The 1992-2003 difference volume, derived by comparing a conventional survey image to a Q-Marine image, shows the effect of six years of production; however, there is a lot of coherent 4-D noise confusing the interpretation. Much of the noise is structural leakage - static geology that has not cancelled out. By comparison, the 2001-2003 difference, derived from two Q-Marine datasets, is much clearer even though it reflects only 22 months of production. There is a small amount of broadband remnant multiple energy, but no obvious sign of structural leakage.
Because the 2001 and 2003 acquisition systems were similar; key differences, such as shot and receiver variability, were measured and compensated for deterministically. Also, the source and receiver positions were repeated to a high degree of accuracy, and the certainty of the positions is unconventionally high because of the acoustic network positioning system used.

Cost and benefits

By avoiding what would have been a poorly producing Norne development well, Statoil saved $30 million, more than covering the cost of the added seismic survey. This savings can be added to the value of the additional and earlier production achieved.

Statoil also is benefiting from an improved understanding of the reservoir drainage pattern. Additional Norne field wells will be placed with more accuracy over time.

Comments