Two Alaska entrepreneurs are teaming up with partners in the Lower 48 and Europe to develop advanced computer-based analytic tools for the oil and gas industry.
The goal is to spot problems developing on drill rigs or in other complex mechanical systems. The system would alert rig operators and owners to potential problems earlier than sensor systems now in use.
Advanced warning on mechanical problems can save on maintenance and rig downtime. Enhanced safety will be another benefit.
Microsoft, the information services giant, is interested in the venture and is providing assistance.
Matt Larkin and Thor Kallestad are the Alaska partners in DataCloud International Inc. Mike Miller, a 30-year BP veteran, is drilling advisor to the group.
Larkin owns and operates Dittman Research, a survey firm he purchased from David Dittman in early 2011. Kallestad is a petroleum engineer formerly with Schlumberger.
Other partners, with backgrounds in industry or advanced computer analytics, are in Houston, London and Trondheim, Norway.
Philip Wade, based in Houston is formerly with Halliburton; Pablo Maestro is in Norway, Thibault Doulas is in Norway, Tom Mundheim is in Houston. Adrian King and Frode Sormo are in London and have significant background in analytics.
Pilot testing
The technology the company has developed is now being pilot-tested on a drill rig working for Hess Corp. in North Dakota. Hess is interested because rigs tend to burn through mud motors and other equipment quickly in the tough conditions in the Bakken, Larkin said.
Larkin and Kallestad are now in discussions with potential commercial customers in Alaska.
DataCloud’s technology doesn’t replace rig operators’ existing systems but is a supplement, Larkin said. “We’ve also designed it as a plug-in, so it doesn’t disrupt what the operators are now using. Our system just shows up as a new item on the drillers’ control panel,” Larkin said. Company managers in corporate offices who monitor the rigs will also have access.
Drilling rigs are one potential customer for DataCloud; operators of oil and gas process plants are another. Larkin is also in contact with trucking fleet operators.
On its current testing program with Hess Corp. in the Bakken, DataCloud has been receiving streamed data from a rig since January and analyzing it in real time. DataCloud writes algorithms that monitor the streaming data to predict future drilling problems. “The algorithms are able to ‘see’ problems before humans see them. It could be a bit wearing out or a vibration,” Larkin said.
The data shows the problems and the conditions that led up to them. “We can apply our algorithms to historic drilling mechanics data from legacy wells to show potential clients that our software would have spotted drilling tool failures before they occurred.” Kallestad said.
The partnership
DataCloud’s partners came together in an unusual way. Larkin became interested in potential applications of algorithm-based analyses in the energy industry based on his experience using advanced analytics in political polling to predict voter behavior.
Dittman Research was engaged in the “No on 1” campaign in 2013, an industry-supported effort to defeat an Alaska ballot proposition that would have been harmful to oil and gas development.
This was where Larkin first saw the effectiveness of using algorithm-based analysis to make predictions, and it was very successful. But applying it to mechanical systems, and Larkin’s first idea was in truck fleets, was an intuitive leap.
“The underlying technology was successful in showing up how voters will behave, but applying this to when a machine might break was an ‘aha’ moment for us,” Larkin said.
Potential applications
After local entrepreneur John Wannamaker introduced Larkin and Kallestad, the potential applications in the energy industry took priority.
Advanced analytics is widely used in financial markets, risk-analysis, health care and even parts of the energy, but applying it to machinery seems like a new idea.
It’s also a puzzle why oil and gas operators aren’t already doing this, but Kallestad puts this down, at least partly, to the conservative nature of the industry and its healthy skepticism of new ideas where there are perceived risks.
Kallestad said the risks are low, however. “This is really only a support tool. Humans are still in charge. The drilling personnel at the rig site ultimately decide whether or not to act on our recommendations,” he said.
Microsoft interested
Microsoft, meanwhile, has became interested in what DataCloud is doing, seeing potential for broadening that company’s reach into the energy industry. Larkin met with Microsoft officials and technical teams in Seattle and Houston, and a relationship developed.
In Houston the two discovered that others, some who are now the Texas and European partners of DataCloud, were pursuing the same idea and had developed a basic application, also persuading Hess to give it a try, which led to the pilot testing now underway.
“They (the other partners) brought what they had done and the relationship with Hess to the table. We brought the relationship with Microsoft and our contacts in Alaska,” Larkin said.
With Alaska’s industry buffeted by low oil prices it would not appear to be the best time to roll out new technology applications but these conditions also make operators acutely sensitive to costs and ways of reducing maintenance by better monitoring, Larkin said.
Also, seasoned industry managers know that prices will improve and industry activity will up again. However, in the meantime layoffs and workforce reductions have cut into the pool of experienced workers.
When activity picks up there will be fewer experienced people on the rigs, which is a concern to companies and drilling contractors.
Larkin sees the analytic tools being developed by DataCloud as a way of storing accumulated knowledge that will be “guide rails” for new-hires.
