Burlington Resources brings high-end visualization strategies to the desktop.

Volume interpretation tools, designed for detailed analysis of multi-dimensional data volumes, have evolved in recent years from technologies intended primarily for visualization to tools that can be used for both interpretation and visualization. Energy companies have deployed visualization systems typically in large visionariums visualization centers or theatres. The very names suggest that viewing, not interpreting data, have been their main purpose.

In addition, access to these centers has often been limited to asset teams with extremely large datasets working on high-profile offshore exploration and development projects. Geoscientists and engineers with "ordinary" 3-D seismic volumes in producing fields, especially onshore, had few if any opportunities to use the new tools and techniques. However, all of that is about to change.

Due to dramatic improvements in price, performance and functionality of hardware and software, scalable volume interpretation and visualization can now be incorporated into every phase of the upstream workflow. The result is that energy companies gain access to solutions for deployment in more practical and cost-effective ways than ever before. Burlington Resources Inc. is currently implementing volume interpretation and visualization technology in a rather unique way. Their strategy and use of Magic Earth's GeoProbe volume interpretation software will be summarized later in this article.

From visualization to interpretation

The evolution from visualization to volume interpretation has gone through three basic phases. Phase 1, which started in the early 1990s, mainly involved data review using simple visualization techniques on standard UNIX workstations.

Geoscientists would evaluate small 3-D seismic surveys, or mere subsets of their total data volumes. They would fan through time slices or cross sections like a deck of cards, looking for gross changes in seismic character, pausing to overlay amplitudes on a surface or identify problem areas. However, most of the real work took place within traditional seismic interpretation applications, which still used 2-D extracts from the 3-D volume on a coarse line-by-line basis. While faults, horizons and wells could be viewed within a transparent 3-D wire frame display, interpretation was largely independent of visualization.
Phase 2, from the mid-1990s to more recent times, was the dawn of true volume interpretation.
Geoscientists were no longer satisfied with simply looking at data. They wanted to reach in, interpret and modify 3-D volumes. They also wanted to scan all of the data, quickly identify and isolate areas of interest, and begin to perform more detailed analyses within the same software suite. Previously, most volume interpretation techniques focused on rendering voxels; i.e., 3-D data cubes, as opposed to traditional 2-D trace-based methods. By the late '90s, voxel tracking tools enabled geoscientists to identify discrete 3-D geobodies, usually based on seismic amplitudes, and to visualize them in proper spatial context by adjusting the opacity of the seismic volume. Nevertheless, to build the structural framework, users still resorted to conventional fault and horizon interpretation software.

During Phase 2, evolving visualization resources focused primarily on the largest and most strategic exploration projects because the technology required to store, render and manipulate complex data volumes was very costly. The era of large visualization theatres began during this phase with wall-sized curved screens, often using 3-D stereoscopic projection systems to create an immersive visual experience, powered by multi million-dollar hardware platforms with massive computing capabilities. Most oil companies were using their new centers for team, management and partner review meetings, discussions or well planning sessions; i.e., for the visualization and communication of multi-dimensional, multi-discipline data, but not yet for everyday interpretation. A few very large corporations boldly deployed hundred thousand-dollar UNIX visualization workstations to their asset teams, but the price tag remained out of reach for the majority.

Phase 3 is taking place today, with the introduction of desktop solutions that any company can afford, the maturing of 3-D voxel-based interpretation tools and their incorporation within the same software of more familiar 2-D line-based techniques of conventional 3-D seismic interpretation.

With the advent of inexpensive Linux desktop hardware that can accommodate the large amounts of random access memory required to load and manage geoscience data, volume interpretation is entering the industry's mainstream. The price/performance of today's desktop has effectively removed the barrier to widespread adoption.

Any geoscientist, on any project in a company - large or small, offshore or onshore, exploration or production - can now access applications previously available only in the large visualization centers. Sophisticated PC technology also has spawned a new multi-volume rendering technique based on high-performance graphics cards designed specifically for 3-D computer games. In effect, this soon-to-be-released technology will enable a geoscientist to illuminate structural and stratigraphic features in surprising new ways using different 3-D seismic volumes at the same time. Not available on UNIX workstations, this is one more reason why desktop volume interpretation solutions are likely to become commonplace in the upstream oil and gas industry.

Interpreters no longer have to pick horizons and faults and build structural frameworks in standard interpretation software before switching to more specialized volume rendering technology. Software developers have realized it made no sense to proliferate completely independent tools in order to complete a single 3-D interpretation workflow. As a result, new versions of classical tools are being built into next-generation volume-based applications, turning them into practical, everyday solutions.

Three technologies are worth mentioning:

An advanced, easy fault framework construction tool that allows users to pick a single fault segment, and then drag a probe through the data volume that automatically builds the remainder of the fault plane. This creates a more spatially consistent fault interpretation.

Waveform trackers that automatically generate complex faulted surfaces with just a few input seed points.
In conditions where autotracking inevitably breaks down, a new manual tracking capability which augments these other tools.

Soon the volume interpretation and visualization systems that have produced stunning 3-D images for years will also provide fully integrated 2-D map and cross-section displays. Once those are available and geoscientists can perform both conventional line-by-line interpretation and sophisticated voxel rendering within a single environment, volume interpretation will have reached maturity.

Burlington's strategy

Burlington Resources Inc., one of the world's largest independent oil and gas exploration and production companies, ranks among North America's leading hydrocarbon producers. As a natural extension of a 5-year plan aimed at enhancing the collaborative capabilities and technical proficiency of its geological, geophysical and engineering (GG&E) staff, Burlington recently decided to evaluate the GeoProbe volume interpretation software.

Taking a conservative, phased approach to the evaluation, Burlington leveraged the new desktop solution to avoid starting with a costly visualization infrastructure. After carefully building a business case for volume interpretation technology, including expected cost/benefits, for about 3 months, Burlington pilot-tested and learned how to use the software on dedicated Linux desktop PCs. Their goal was to gain experience through application on actual projects and to establish expert users who could act as a local resource before moving forward with wider deployment in the company.

Recently, Burlington purchased three software licenses that their E&P professionals throughout the world can access from existing Sun workstations, either in their offices or asset team workrooms. The company had plans to install an 8-CPU SGI machine by the end of 2003 in its first "collaboration room" located in its new Houston, Texas, office building.

"We don't use the term visualization around here because it carries the connotation of a fancy 'planetarium' style facility that rarely gets used," G.M. Byrd Larberg, vice president, Geosciences, said. "We'll certainly have a large screen with projectors, but it is more of a conference room than a movie theatre." Burlington's GG&E group is now collaborating in ongoing interpretations; not just 3-D views in project reviews.

According to Larberg, Burlington expects the new software to enhance its current work process in three primary ways:

• Interpreting the full 3-D volume, instead of just the horizon or target of interest.
• Reducing substantially the interpretation cycle time for new 3-D surveys on a short time fuse.
• Minimizing traditional hand-offs between disciplines and encouraging better collaboration.
"Collaboration between a geologist and geophysicist was momentary; i.e., only at the hand-off points, and punctuated by lots of individual work," Steve Western, Burlington's chief geophysicist, said. "With today's software, volume interpretation takes place so quickly, they can actually collaborate in real time - calibrating, interpreting and editing seismic and well data on the fly. That should improve efficiency and effectiveness, even for an asset team that already works well together."

Burlington initially tested the new volume interpretation software on a tight-deadline drilling project in a Canadian onshore producing field. "In addition to exploration expertise, we're a strong production-oriented company," said Terry Cosban, senior staff geophysicist with Burlington's corporate Geophysical Technology Group, which worked the project with the Canadian team. "We had substantial well control to integrate with our 3-D seismic volume. Furthermore, we needed to look at very detailed infill targets and step-outs for our winter drilling program."

Cosban had just received a 60-sq-mile (155-sq-km) 3-D survey out of reprocessing with which he was not previously familiar. Well locations had to be validated or adjusted before drilling could commence. "I had to pick some rather difficult reflectors, and I needed to understand very quickly how the spatial distribution of seismic attributes related to the geology," he said.

Using the latest volume interpretation tools, he was able to complete work in 3 days that would have taken weeks using the conventional method of slicing up the volume and tracking the events. Cosban not only mapped the top and base of the target interval along with the distribution of amplitudes within the reservoir, but he mapped half a dozen reference horizons above and below the area of interest and identified a dozen or more "interesting events" for further investigation. In the process, two well locations were optimized.

Cosban did, in fact, draw upon the software's visualization strengths to help communicate his results to management and team members, in addition to the familiar 2-D maps and sections. As a result, he quickly recognized the value of the tool for conversation and collaboration with others. Burlington's expectation is that this and similar programs will also become standard interpretation tools for many other E&P companies and that it will not take place primarily in big visualization centers but on the desktops of ordinary geoscientists.

During pilot testing he identified three main benefits that the user derives:
• A better understanding of spatial relationships between seismic attributes, geologic concepts and production data.

• The ability to make faster, more timely decisions.
• The ability to work in both two-dimensional and three-dimensional worlds at the same time and capture the final results in familiar 2-D displays which are easily communicated and shared among standard interpretation tools.

Summary

Today, with the price/performance of desktop PCs, volume interpretation will enter the mainstream. Applications previously only available in large visualization centers have now become accessible to any geoscientist, for use on any company project, large or small, inland or offshore. In the very near future, volume interpretation and visualization systems will provide fully integrated 2-D maps and cross-section displays. At that time, volume interpretation will be a technique used by every interpreter in the industry.