"The barriers for oil companies to adopting new AI technologies are many, ranging from resistance to change, a belief that what they already have is sufficient, and skepticism about whether new technologies will deliver," said Ray Hall, energy sector director at Tessella, a provider of engineering and consulting services that has helped global energy companies identify ways to improve drilling and operational efficiency with data.
"Many of our own customers have invested in large technology vendors promising the world from analytics, only to be left disappointed with the result," Hall said.
Oil companies for years have been using analytics approaches such as model predictive control (MPC) in supply chain platforms and linear programs in refinery planning, Hall said.
"They are no strangers to the use of structured data approaches and analytics," Hall said. "Often they just have not progressed and their approaches are now based around old technologies which lack the ability to incorporate newer analytics techniques such as machine learning to improve performance."
It's important, from a competitive standpoint, that oil and gas companies overcome the challenges in adopting AI and other emerging technologies, because the industry is facing challenges on many levels.
For one thing, there is long-term price uncertainty. "The days of the long-term barrel price of over a $100 are gone; pressures of global demand vs. production growth will keep the oil price lower, and mean oil companies have to get production and refining costs down considerably," Hall said.
That means taking significant steps in efficiency gains to reduce cost through the application of new technologies. "For example, could robotics take a role at a production asset to handle the drilling and completions processes?" Hall said.
In addition, the industry has an aging workforce. "We have seen various estimates on this but it's widely recognized that up to 40 percent of the workforce in oil and gas will retire in the next five to 10 years," Hall said. "Replacing this highly experienced workforce with people will be very challenging, but using increased automated decision support through cognitive and machine learning solutions is a means to reduce the reliance upon experience to make decisions."
Finally, as reliance on oil and gas diminishes, companies in the sector need to transform their businesses into being complete energy suppliers that embrace renewable energy sources.
"To do this, they would need their business to be able to operate across both fossil and renewables in a profitable manner," Hall said. "This will require a more modern technology environment able to meet the demands of their customers profitably and be able to optimize the source of the energy."
Tessella has helped several oil companies leverage the capabilities of AI to improve processes. For example, it worked with one oil customer to help the company improve its understanding of durability and levels of corrosion of existing wells, as part of the company's plan to get greater returns from wells.
"The company had a lot of historical data at its disposal, but not a huge understanding of it," Hall said. "We took the time histories of all the operational data for the wells and used a range of AI statistical techniques to find structure in the data. We looked for correlations between clusters in the historical data and the corrosion level, and then drilled down into the underlying variables to understand what in particular in the historical record was driving corrosion."
The work allowed the client to confidently make a risk-based decision to use the field, with full visibility of the uncertainties, risks, and sensitivities. "The impact was huge," Hall said. "Understanding corrosion was critical to enabling the decision to go ahead with the project."