wells, like those drilled in the early days of discovery in Texas, it was possible to punch a hole in the ground and have the oil shoot up like a geyser. When capped, those wells could produce for years without further intervention.

Today, however, the management of reservoirs is a highly complex science. A reservoir may, for example, contain 100 million barrels of oil, but there may exist only enough natural pressure for that reservoir to produce 25 million barrels without intervention. To capture the next 25 million barrels, Exxon may need to use a variety of techniques, such as installing underground pumps, or injecting water or gas into the reservoir to increase pressure. Each action increases the cost of managing that resource and can, in turn, have a negative impact.

Pumping too much water into the reservoir could cause the oil and gas to mix, increasing the cost of recovery. Intervention can also impact the natural pressure that exists, creating a need for even more costly recovery measures.

Reservoir simulators, through complex algorithms that have been refined over the years and continually increasing computing power, allow oil and gas companies to best map out how to manage their reserves.

But not every oil or gas reservoir requires a supercomputer to best determine how to manage that resource. The cost, expertise required and the commitment of computational resources cannot always be justified.

Here again, Exxon used the strength of its in-house technology expertise to create an innovative answer. The upstream research organization created scaled-down versions of EMPower to run on Linux clusters and on a standard desktop or laptop. That way, the software can be utilized on everything from a 100,000-barrel-a-day reservoir to a 1,000-barrel-a-day reservoir.

Getting good results from a Linux cluster or grid computing system poses its own challenges (see "Gotcha,"

p. 52), but Exxon says it's a cost-effective alternative to using supercomputers.

Comstock declined to specify the business results EMPower has allowed the company to achieve since it was introduced four years ago. However, in a presentation to investors in August 2004, Steve Cassiani, president of Exxon Upstream Research, called EMPower the company's most heavily used engineering application. In a relatively small assessment of five fields where EMPower had been used to manage the resource, he says Exxon found it was able to recover an additional 100 million oil-equivalent barrels.

("Oil-equivalent" refers to the production of oil or natural gas to make up the equivalent energy contained in a barrel of oil.)

More important is the impact of the combination of all three key areas-an alignment of technology to pump out profits.

At an industry conference in Aberdeen, Scotland, last September, Robert Olsen, chairman of Exxon International, pointed to the Beryl field in the North Sea as a prime example. When the field was discovered in the 1970s, he said it was expected that Beryl would produce 600 million oil-equivalent barrels in its estimated 20-year life. Thirty years later, through a combination of better seismic analysis, more advanced drilling techniques and improved reservoir simulation, Exxon is still drilling at Beryl and now expects to recover more than 1.3 billion barrels-more than twice the original estimate.

At today's prices, that translates into a $42 billion payoff.



Story Guide:
Straight shooter: CEO Rex Tillerson Doesn't Play Games

Vendor Profile: A Look at IBM's Supercomputers and the Oil Business

Next page: Oil in the DNA: The Exxon Culture

Next: ' Oil in the DNA' >>

Discuss Behind Oil Profits: A Look At ExxonMobil's Technology Alignment   >>> Be the FIRST to comment on this article!

>>> More Projects: Processes Articles          >>> More By Mel Duvall