While most of the seismic industry is still waiting for a full-fledged oil company spending recovery, one side of the business has seen demand climb dramatically since mid-1999.
The demand for prestack depth imaging, particularly outsourced services, merely dimmed rather than plummeted during the downturn and was one of the first market segments to return to health. Indeed, prestack depth imaging is possibly the only segment of the seismic business where spending has quickly exceeded preslump levels.
What is driving this rapidly increasing demand for depth imaging services? Financial and technological causes are responsible for what could be the industry's most sustainable boom since that for 3D time seismic.
High-risk, high-reward plays
Oil companies are accelerating their uptake of prestack depth migration, and one of the most important reasons is the industry is increasingly turning to areas that require it. The Lehman Brothers Original E&P Spending Survey of 320 oil and gas companies, released in December 1999, found operators remain conservative in their budgeting process, projecting an average oil price of $19.25/bbl for WTI crude in 2000. With the memory of the recent downturn obviously still lingering, and with large, unexplored, two-way time structures all but gone, the seismic industry is embracing the use of innovative, cost-effective depth imaging technology in an attempt to reveal significant fields obscured by complexity of geology or velocity.
The routine use of 3D time seismic data since the late 1980s has been instrumental in reducing E&P risk around the world, but over an area exhibiting complex velocity or geology, even a well-defined 3D subsurface model in time is not sufficient to accurately characterize the geology. Such plays benefit from prestack depth migration because the velocity model derived during the process is inherently more accurate than that for time processing. Using this velocity model and prestack time seismic traces as input, prestack depth migration enhances the imaging or focusing of the resulting seismic data by placing reflections in more accurate depth space. As a result, the structure and trap volumetrics are also more accurately revealed and positioned.
Both 2D and 3D prestack depth imaging are making it possible to reduce the risk associated with more complex plays, maximizing the opportunity for successful discoveries and more economical development. Such plays are found around the world and commonly include those related to salt, volcanics, thrust faults, fault shadows, reefs, shallow gas and others where lateral velocity variation is significant. Two examples serve to illustrate the types of geology that are increasingly benefiting from prestack depth migration.
First is the Gulf of Mexico deepwater subsalt play. In an area where a dry hole can cost US $40 million - $50 million or more, oil companies routinely use prestack depth imaging to mitigate their leasing and drilling risk. Due to their strong velocity contrast with surrounding sediments, salt bodies can strongly bend downgoing and upgoing seismic energy, resulting in zones of poor seismic illumination or dead zones beneath and adjacent to salt, as well as mislocation of the reflections on two-way time data (Figure 1). Prestack depth migration moves the seismic energy into more geometrically accurate locations - laterally and vertically in depth - improving the subsalt reflection image and the accuracy of the structure (Figure 2). In this case, the apparent subsalt time high is revealed through prestack depth imaging to be a much flatter feature than originally imaged.
Reef plays also can benefit from prestack depth imaging. As an example from western Java, Indonesia illustrates, prestack depth migration has significantly improved the imaging of the reef (Figure 3). The top of the reef now exhibits a sharp crest with steep dips on either side and is much more clearly discerned than in time. The base and
flanks of the reef, as well as the associated onlapping sediments, are also better imaged in depth.
These examples illustrate that prestack depth imaging can significantly improve the quality and location of reflection imaging, opening new, complex plays to more cost-effective exploration and development and lower risk. Because of its increasingly successful application, the demand for depth imaging has been growing at such a pace during the past 2 years that it has strained the ability of oil companies and some service providers to meet it. Those meeting the demand have been able to do so by leveraging recent improvements in computing price/performance and imaging technology.
Computing price/performance breakthrough
Of course, regardless of its benefits, prestack depth imaging will only be used if it is cost-effective; in the not-too-distant past, this was a significant issue. Three-dimensional prestack depth migration was and is one of the most compute-intensive tasks in the world, and it certainly wins top honors in the world of oil industry computing. Initially developed in the early 1980s, prestack depth migration was done on the fastest and most expensive computers of the day: massively parallel, shared memory computers produced by such companies as Cray and Convex. Because of the huge costs associated with the hardware and specialized software development, many such computers are still used for prestack depth migration.
At prices as high as tens of millions of dollars, these computers were so expensive that 3D prestack depth migration was applied to only the highest-profile, highest-risk prospects. This has all changed, however, with the recent introduction of depth imaging systems based on clustered supercomputers, consisting of a network of commodity-priced PCs or Unix-based workstations with software that enables them to work together as one computer. As an example, at a cost of as little as $10,000 to $15,000 per node, a clustered supercomputer can run prestack depth migration twice as fast as a Cray T3E with an equivalent number of processors. Stated another way, this constitutes a price/performance advantage of as much as 20 to 30 times.
Another advantage of clustered supercomputing for depth imaging is scalability. Additional computer nodes can be added to the system at a low incremental cost and made available for processing in a matter of hours. Nodes of varying architecture and performance may be mixed easily, allowing the clustered supercomputer to retain and extend its value well into the future. As computer price/ performance improve, new nodes may easily be rolled on and old ones rolled off as needs and finances dictate.
With vastly greater performance per unit cost, today's clustered supercomputers are a key factor in meeting and driving the industry's demand for more depth imaging. Service providers who have invested in this technology are able to offer superior turnaround at an attractive price. Oil companies in turn are able to depth-image more projects more cost-effectively without suffering delays in their E&P programs.
A final point about price/performance is that in the world of prestack depth imaging, quality, turnaround and cost-effectiveness are, in practice, each a function of the other. The time savings wrought by faster, more cost-effective computers can be invested in one or more ways: higher-quality, although more time-consuming, migration; faster overall project turnaround; more iterations and sensitivity analysis; or maintaining the status quo at a lower price. Indeed, some of the innovations in quality are at least partially a byproduct of the time surplus created by recent price/ performance breakthroughs in computing.
Higher quality, faster turnaround
Even though time processing has a history of nearly 40 years, results still vary widely from one provider to another. Depth imaging, at less than 20 years old, is a youngster by comparison. Technology and expertise differ significantly from vendor to vendor and oil company to oil company, so it should be no surprise that results differ even more markedly than in the world of time processing. The more innovative service providers and oil company specialists have recently made significant improvements to the quality and performance of prestack depth imaging, and these improvements are helping drive the rising demand for depth solutions.
The velocity model and the migration impact quality. No matter how good the migration algorithm, if the velocity model input to depth migration is inaccurate or oversimplified, the classic computing dictum will hold true: garbage in, garbage out. On the other hand, even if the velocity model is accurate, a poorly conceived or inefficiently implemented migration will produce inferior results or take so much time as to be practically unusable.
Recent improvements to quality and turnaround have come in the areas of velocity modeling and migration. The first example of these is 3D tomography. Most velocity modeling techniques assume a vertical ray path when locating the velocity update in 3D space, a simplification that results in misplaced velocities when the geology is more complex. Three-dimensional tomography results in a more accurate velocity model by locating the velocity along the modeled 3D ray path (Figure 4). It also simulates in minutes or hours the depth migration process that would otherwise consume days or weeks. The combined quality and turnaround benefits can be significant, particularly in cases where shallower velocity complexity has disrupted underlying reflections.
Another improvement to velocity model refinement is alpha focusing. Most depth imaging makes an assumption similar to the normal moveout common to time processing, a shortcut whereby the degree of flatness of migrated gathers is the standard for an accurate velocity model. By comparison, alpha focusing forgoes the shortcut and uses true migration velocity analysis, resulting in a more accurate velocity model for depth migration, particularly in areas of steep dip.
There are also improvements to the migration process, and one example is illumination imaging. Bending of seismic energy results in uneven subsurface illumination, with some areas producing
more data than required and others less.
Most algorithms do not evaluate the quality
of subsurface illumination prior to migration and therefore use more data or more substandard data than desirable. Illumination imaging automatically inspects the illumination quality of the data and uses the optimum amount to ensure higher quality and better turnaround.
While a few of the large oil companies continue to develop their own proprietary capabilities, much of the innovation in prestack depth migration is taking place inside specific geophysical service providers, particularly those focused on this field. There is every indication that this innovation will continue to produce significant additional improvements to quality and turnaround.
Rise in outsourcing
Over the past several years, an increasing amount of prestack depth imaging has shifted from the oil company specialist to the geophysical service provider. There are three basic reasons. First, budget constraints have made it imperative that oil companies eliminate unnecessary capital purchases. While there is pressure even to reduce the purchase of equipment considered mainstream and mission-critical, there is an even greater push to reduce purchases of specialty items such as depth imaging software and associated computer hardware. This is possible because oil companies have the cost-effective alternative of outsourcing to the depth imaging service provider, which in turn absorbs the burden of all capital expenditure; even though the price of such computers and related equipment has come down significantly, a system capable of production work on many simultaneous projects can still cost several million dollars.

Downsizing
The second contributor to the outsourcing trend is the downsizing of oil company staff required to develop or effectively use depth imaging technology. To be sure, several of the largest oil companies have retained at least some of these highly trained personnel, but many have left the oil companies and joined the geophysical service providers. This migration of geophysical specialists is helping service providers meet the increasing demand for depth imaging, but it is also a reflection of an outsourcing trend that was already growing before industry downsizing began.
Finally, even those large oil companies
that have sizable depth imaging staffs are experiencing a growth in demand for depth imaging that internal resources cannot fully meet. Faced with staffing up, restricting their use of risk-reducing technology or outsourcing, the trend is to outsource at least the overflow work to depth imaging providers.
An increasing worldwide E&P focus on complex, high-risk plays and the accelerating price/performance of computing are driving the worldwide demand for more depth imaging. The requirement to reduce oil company capital and personnel expenditure, coupled with the rapid pace of innovation in vendor-supplied quality and turnaround, are responsible for shifting more of this work to depth imaging service providers.
By amortizing the cost of computers and personnel over many more projects than those encountered by even the largest oil companies, the innovative service provider is able to supply a timely, high-quality depth solution at a price attractive to any oil company. For the foreseeable future, it appears the demand for depth imaging and the demand for an outsourced solution will continue to rise at a rapid rate.