BEST BET YET
Our odds of extracting more oil just went up—thanks to one little meeting of some very different minds.
Cocktail hour started the evening pleasantly enough, but an uncomfortable feeling crept over me at the dinner table, when keynote speaker Chris Farmer uttered the following words with a hearty chuckle: “I love quantum mechanics!”
I watched Farmer’s face carefully for the next few seconds just to make sure. But no, he wasn’t joking. Whereupon my mind was seized by a complex emotion that only two words can describe: “uh oh.” I briefly considered running for the nearest exit, but that would have been rude. Instead, in the interest of enhanced concentration, I swallowed a large gulp of wine and braced for the next puzzler.
Scientific advisor, theoretical physicist and Schlumberger Visiting Professor in Mathematics at the University of Oxford, Farmer was here to address the 60-odd participants of the first-ever Schlumberger Uncertainty & Intelligent Well Control Workshop, which took place in Oxford, England earlier this month. An acknowledged expert on the subject of uncertainty, Farmer was going to help us understand how the human mind makes decisions—specifically those made quickly about how to get the most oil out of a given hydrocarbon reservoir. This is important because apparently a lot of mediocre decisions of this sort get made in our industry, and naturally we want to do it better than any other company.
“We can only judge a decision by the quality of the decision making,” Farmer intoned, ducking and bobbing over his notes like an orchestra conductor. “This means that we MUST consider probability, and subjective probability in oil reservoirs is very VERY complicated.”
I could make no objection here, and let Farmer continue—about “Kalman filtering,” Bayes’ Theorem of Reverse Probability (published in 1748), and an enigmatic gambling puzzle called the “St. Petersburg paradox.” Finally, raising his arm for emphasis, Professor Farmer said, “We MUST maximize our expected utility! The arguments for doing this are OVERWHELMING! And this workshop is important for precisely this reason.”
Well, I couldn’t argue with that. Chris Farmer was clearly right. But now I was feeling an acute need to maximize my expected utility, primarily by trying to find out what he was right about.
I got help in this endeavor from workshop organizer Anne Gerd Raffn, a reservoir engineer product analyst at our nearby Abingdon Technology Center, and from Aubrey O’Callaghan, a reservoir engineering instructor for SIS. Looking at their subject with great kindness, understanding and a degree of pity, they explained the problem at hand: The people in Schlumberger who design our reservoir modeling software (ECLIPSE, Petrel, Cougar, etc) and the people who design and sell our intelligent (real-time) completions tools have not been talking to each other. Because they don’t talk, they haven’t been able to maximize the potential working relationship between their respective tools, and our ability to make good, fast production decisions has not improved.
“By bringing these groups together, we are effectively linking our understanding of the reservoir with our capacity to use sophisticated completions effectively in real time,” said Aubrey O’Callaghan. “As the quality of this link improves, our ability to make better decisions improves, and oil production goes up.”
Even I was able to guess what happens next: Schlumberger sells more completions.
“Not more completions—just better ones,” O’Callaghan corrected, grinning. “Better completions will enable us to differentiate Schlumberger from the pack. In effect, we’re arming our sales people with more certainty about what our solutions are able to get out of the ground. This is important.”
Terje Moen, looking relaxed after a problem-solving workshop the previous afternoon, agreed. As one of the founders of Reslink, the Norwegian completions company Schlumberger acquired two years ago, and as the original architect of Reslink’s high-end completions gizmo known as the Inflow Control Device, Moen had bumped into the reservoir modelers’ Tower of Babel more than once in his career. “For a long time, modelers viewed hardware (completions) as little more than a disturbance in the reservoir model,” he said, wincing. “Events like this create value for all parties by improving communications.”
Undoubtedly true, but I was still hungry for a grittier, rawer explanation of how our Oxford experience would benefit the world. I got it from Timo Jokela, a Finnish completions architect at our Completions Center in Stavanger, Norway.
“I’m looking for an edge to sell my hardware,” said Timo, punching the air. “If I have to slug it out with Baker and Halliburton based on price alone, I’m going to get a bloody nose. I came here for ammunition!”
And Timo got it. “I met some guys here from SIS and DCS who I’m sure can help me,” he said. “They’ve got this software that I’d never heard of before, but that is going to help my clients make better decisions to optimize their production. That’s real value.”
The software Jokela discovered, called RTPM for Real Time Production Management, is designed to provide valuable decision-building answers to production engineers, “not just spit out numbers they don’t have time to understand.” Developed at our Beijing Research Center over the last two years, RTPM will be released this October strictly for Schlumberger internal use only—“as value-enhancing ammunition for our reservoir engineers and completions people,” says Raj Banerjee, DCS Production and Reservoir Engineering product champion, based in Abingdon.
Raj is the guy who tipped off Timo Jokela about RTPM. Now they know each other. And with a little luck, they will keep talking.