Sign In

Communications of the ACM

ACM TechNews

Positive Reinforcements Help Algorithm Forecast Underground Natural Reserves


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
An oil well blowout.

Texas A&M University and University of Oklahoma researchers have designed a reinforcement-based algorithm that automates the prediction of underground oil and gas reserves.

Credit: Michael Reed

Texas A&M University (TAMU) and University of Oklahoma researchers have developed a reinforcement-based algorithm that automates forecasting of subterranean properties, enabling accurate prediction of oil and gas reserves.

The algorithm focuses on the correct characterization of the underground environment based on rewards accumulated for making correct predictions of pressure and flow anticipated from boreholes.

The TAMU team learned that within 10 iterations of reinforcement learning, the algorithm could correctly and rapidly predict the properties of simple subsurface scenarios.

TAMU's Siddharth Misra said, "We have turned history matching into a sequential decision-making problem, which has the potential to reduce engineers' efforts, mitigate human bias, and remove the need of large sets of labeled training data."

From Texas A&M Today
View Full Article

 

Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA


 

No entries found