Flotek/IBM partnership on ‘Reservoir Cognitive Consultant’ seeks to identify trends and make predictions
Big data came to the Oilfield Tech and Innovation Conference last week in Dallas with several presenters focusing on the access to big data as the avenue to take oil and gas field development to the next level.
Houston-based Flotek (ticker: FTK) CEO John Chisholm discussed the company’s partnership with IBM Watson for the purpose of analyzing well data and improving completion and reservoir performance. Flotek and IBM call it the ‘RC2’ system. It stands for Reservoir Cognitive Consult.
Flotek’s Reservoir Cognitive Consultant seeks to analyze Flotek’s inventory of data from client wells to discover trends and insights, with the ultimate goal of predictive data, rather than simply backward-looking information.
The downturn sent experienced employees out the door – average age at some big oil companies has dropped to 38
Industries need this sort of system now because the downturn shook a lot of experience people out of the oil and gas industry, creating a major brain drain in many companies, Chisholm pointed out.
“The past couple of weeks I’ve chatted with a couple of c-level executives with major oil companies, and I’ll give you a common statistic that really was enlightening to me. They said that their average age in their company is now 38 years old. And these are companies that you would kind of regard as the iconic oil and gas companies in this industry that have essentially been around forever, not some new private-equity-funded company that just got started.
“These are big, multinational, fully integrated oil and gas companies. The average age is 38 years old. Here’s the other statistic: 50% of their corporate staff has been there for five years or less,” Chisholm told the OTIC audience.
“Well, with those type of stats, one of the most, and I’m almost inclined to say there is only way you can bridge that — that is to be able to have cognitive capability to grab all your data and replace the knowledge that essentially has walked out the door.
“And I think that’s becoming more and more apparent with a lot of these E&P companies through the different downturns. So much of that knowledge walked when the people left the building that there has to be way to try to get that back. And the way you get it back is analyzing the data.”
Complicated problems vs complex problems.
Flotek’s RC2 system intends to address the complex problems posed in the oil industry, and eliminate human biases to determine what factors are the largest drivers of value. This will enable it to make predictions, not just comment on past results.
Chisholm described the difference between complex and complicated problems, saying “This has always been a complex problem of completing wells. But up until now, you tried to solve the problem with complicated solutions.”
An example of a complicated problem, Chisholm said, is building the Burj Khalifa in Dubai. “It was complicated to figure out how to get cement up to the 110th story, but they figured it out. And once you do it, it’s done.” A complicated solution can be applied numerous times to similar complicated problems, unlike complex situations.
“’Complex’,” Chisholm said, “is the highway driving pattern on the interstate 5 in Los Angeles at 5:00 every Friday night because it all changes. Every Friday is different than the previous Friday. And it doesn’t do any good for you to get an email two hours later and say, ‘Hey, James, if you’d taken this different route, you would have gotten to where you wanted to get to 30 minutes faster’.”
1 TB of data per rig per day
Flotek is not the only service company looking to big data to analyze the information the industry creates. There is a world of new information to examine every time a well is drilled. A
According to Noble Energy, data collection during drilling activities can generate 1 TB of data per rig per day, so there is a great deal of data to sift through. For reference, 1 TB is enough to hold 17,000 hours of music or 310,000 photos.
Schlumberger looks to automate repetitive tasks
Schlumberger has slightly different goals with its DELFI system, designed to enable cross-discipline work. DELFI combines information from every point in the life cycle of a well, from seismic to production. Schlumberger intends to move beyond this though, to automation of many complex functions of modeling and simulation. The company unveiled one application of this in late 2017, DrillPlan, which combines historical data on well plans and drilling activities and is used to automate the consistent aspects of designing wells.
Liberty leverages in-house data set to identify critical variables
Liberty Oilfield Services uses big data analytics on its significant completions dataset. The company has run thousands of frac jobs in its history, analyzing the parameters of each job and the productivity of the well in order to illustrate which variables can be improved. This allows the company to identify the most effective completions designs for a given location and play, significantly improving results, according to the company.
While “big data” may sound more like a buzzword or a technology that is perennially five years out, programs such as these have the potential to make significant changes in the oil and gas industry now, and they will show up in force in the next few years as the more predictable elements of the business are automated.
What results is big data producing now for oil and gas companies?
BP reported recently that its new supercomputer in Houston found 200 MMBO and saved the company 1,000 years in processing time.
“In 2017 an algorithm created by one of our team helped us to see through layers of salt to uncover a new ‘field’ within our existing Atlantis field in the Gulf of Mexico that holds 200 million barrels of oil. It took just two weeks to process the algorithm in our supercomputer in Houston – whereas with 20th century technology, we estimate it would have taken well over 1,000 years,” said BP’s Chief Operating Officer for Developments and Technology James Dupree.
Dupree said BP is also applying new well-bore strengthening technology that allows its teams to reduce the time it takes to drill a well.
In reservoir management Dupree said, BP is deploying new digital tools to transform operations across the complex network of 14 production platforms in Trinidad. “That new system allows us to simulate what will happen across the network during routine activities – such as wells starting up, valve testing and pipeline inspections – so we can optimize production.”
