The early 1980s was a dynamic period for technology development, during which all the major oil companies had substantial research labs. Arco’s lab in Plano, Texas, was one of the best, and researchers there recorded the first time-lapse surveys over the Holt reservoir in North Texas in 1982 and ’83. In this tertiary recovery project, combustion was initiated around an injection well within the reservoir, and the combustion gases were used to drive the remaining oil toward the producer wells.

The process was monitored with 4-D—a baseline 3-D survey followed by two repeats. The results, when interpreted, clearly illustrated the combustion progress (which generated a “bright spot” due to the increase in gas saturation). They were described by Greaves & Fulp in a paper in Geophysics in 1987, earning them the prestigious Society of Exploration Geophysicists’ Kauffman gold medal.

Stanford University, always at the forefront of seismic technology, had begun to propose 4-D as a means of improving the woefully low recovery factors from oil reservoirs. Several of Stanford’s students became advocates of 4-D technology and initiated 4-D projects (and companies) of their own in different parts of the world after their graduation. Research projects and early 4-D work in Chevron, BP, Norsk-Hydro and Statoil can be traced back to Stanford’s influence, and papers began to appear in the technical journals of the professional societies.

FIGURE 1. Cross sections are shown in milliseconds through the original 2005-1993/1992 (upper) and the new 2012-2005 (lower) 4-D acoustic impedance change volumes. Producers are in green and injectors in blue. In 2005, the hardening response was limited to the upper part of the Tor reservoir. In 2012, the hardening response penetrates deeper into the Tor reservoir, indicating an improved vertical sweep compared with 2005. (Source: Calvert et al, 2014, “Insights into sweep efficiency using 4-D seismic at Halfdan field in the North Sea.” Courtesy of Maersk Oil and The Leading Edge. Copyright 2015 by the Society of Exploration Geophysicists. Used with permission.)

Watery shift

But wells are cheap on land and seismic is expensive, so the focus shifted to marine, where multistreamer techniques were reducing seismic costs in an environment in which well costs were escalating. Some marine 3-D surveys from the ’80s had been re-recorded in the early ’90s to improve the image quality, and these were natural candidates for 4-D research. Those on Statoil’s Gullfaks and BP’s Magnus fields were especially influential.

Once the commercial viability of 4-D had been realized, the technology blossomed, particularly in the North Sea, where the growth was dramatic, moving from an R&D-driven activity in 1995 to an almost totally commercial activity by 2000. Growth was slower in some other regions, notably the Gulf of Mexico (GoM), where the widespread use of relatively inexpensive “multiclient data” made proprietary 4-D surveys seem excessively expensive. The technology was also more difficult there owing to the circulating ocean currents and the prevalence of stacked reservoirs.

Steerable streamers

Early time-lapse work could be described as the subtraction of two conventional 3-D seismic surveys and making sense of any amplitude and time differences that resulted. It was apparent very quickly that accurate repeatability of source and receiver positions was a key quality factor, and a major shift in the market occurred with the arrival of steerable streamers in 2002. Pioneered by WesternGeco and ION, these new streamers offered improved positioning control.

This has led to a methodology that includes a detailed analysis of the tidal and positioning data from the earlier survey and a very carefully planned repeat. Current best practice includes denser streamer coverage to improve the spatial repeatability. The most-used quality measure of residual noise (thus of nonrepeatability) is normalized root mean square, which in the late ’90s would be high (around 80%) and is now below 10% on the best towed-cable marine repeats and can be 5% or less on emplaced systems.

By the mid-2000s additional repeat surveys were being performed on several reservoirs, particularly in deeper water where subsea well completions make for very expensive well interventions and drilling. A fairly extreme example is in the West Shetlands area of the U.K. Continental Shelf, which has seen close to a dozen towed cable 4-D surveys since the first exploration 3-D in 1993. A more common repeat interval for clastic reservoirs in the North Sea is four to five years. Time-lapse seismic has been a major integrator of disciplines within oil companies—drillers, reservoir engineers, petrophysicists, interpreters and other seismic staff meet together in “immersive visualization” rooms. A common feature of 4-D projects is the inclusion of a contractor’s processing team embedded within the oil company offices.

FIGURE 2. PRM systems are gaining traction, but it has been a slow process due to the expense. (Source: Ian Jack)

PRM

R&D activity continues to focus on repeatability; sampling and thus sensitivity and image quality; and impedance inversions and interpretation of the results, allowing better discrimination between pressure and fluid effects. A recent example is shown in Figure 1. Interpretation has been moving gradually from the “qualitative” toward the “quantitative,” although the greatest value is often gained from the early qualitative interpretations. Geomechanics became strongly linked to 4-D as depletion-induced compaction became seismically evident on reservoirs such as Genesis in the GoM. (It had long been associated with the massive compaction taking place within soft chalk reservoirs such as Ekofisk and Valhall).

The best repeatability to date is achieved with permanent reservoir monitoring (PRM) systems, in which the receivers are trenched into the seabed. The first trial of an emplaced system was a BP pilot in 1995, in which six cables totaling 25 km (15.5 miles) in length were trenched 1 m (3 ft) into the seabed in 450 m (1,476 ft) of water on the Foinaven reservoir. This successful proof of concept was followed by the first major PRM installation on BP’s Valhall reservoir in the Norwegian sector in 2003. This consisted of more than 120 km (74.5 miles) of emplaced cables covering about 40 sq km (15.4 sq miles), 70% of the reservoir’s area. Once the initial installation had been made, repeat surveys were very cheap, and thus the six surveys originally sanctioned over a two-year period have mushroomed to 17 (at the time of writing).

The Valhall installation has been followed by others (Figure 2). The operators appear to be happy with their systems, extolling their virtues in impressive conference papers. However, the growth curve has been slow, for several reasons. Firstly, the upfront cost is high—it will be more than $1 million/sq km ($2.6 million/sq mile) for the equipment and its installation, so although the surveys themselves are cheap, several have to be acquired in order to be cost-effective; thus the projected longevity of the reservoir has to be substantial, with a significant drilling program over several years. The cost crossover point for emplaced vs. repeat streamer surveys occurs at about five surveys.

Secondly, technology has been evolving, and this has discouraged decision-making. For example, the successful and proven wired electrical systems developed in the early 2000s now compete with newer fiber-optic systems and more recently with redeployable ocean-bottom node systems.

The physics of ocean-bottom recording has been hugely beneficial to the quality of the seismic images. Pressure sensors with particle motion sensors and also the freedom of the source vessel, which allows full azimuthal sampling and unconstrained source-receiver distances, have produced irresistible images, removing many of the problems imposed by safety distances when towing cables adjacent to platform/FPSO installations.

New techniques

So where is 4-D heading? Out of a total marine seismic market of about $10 billion, 4-D (including PRM and nodes) constitutes about 13%, with strong regional variations, and it has been fairly steady at this percentage. Most 4-D seismic surveys continue to be recorded “conventionally” with towed streamers. Meanwhile, the technology has continued to develop into broadband techniques, which themselves have produced much improved images. The 4-D market is moving slowly and carefully into this technology, the difficulty being the careful matching of all the parameters of the earlier surveys. There is a slow trend toward the techniques that produce the best images—ocean-bottom methods and towed-streamer broadband.

The ocean-bottom 4-D node market has been strengthening consistently over the last five years. These surveys are expensive (especially in deepwater with ROV deployment), but engineering research is beginning to reduce costs via “nodes on ropes” for shallow to intermediate water depths and long-life nodes for deepwater.

Greater recoveries

The great majority of 4-D projects have been applied to the easier (i.e. clastic) reservoirs. However, examples have emerged on more difficult reservoir types such as carbonates. These have been enabled by improved repeatability of the seismic data and thus the sensitivity of the method and by improvements in data processing and impedance inversions of the time-lapse data. Since currently the industry leaves typically 70% of the reservoir’s oil in the ground, technologies such as 4-D can hardly fail to gather impetus.