Allegra Hosford Scheirer, PhD | Stanford University
Machine Learning for Understanding Petroleum Systems
The evaluation of petroleum systems via basin modeling, mapping, geochemical analyses, etc., is now a mature field due to codification of the petroleum system concept coupled with the development of sophisticated basin modeling software. Even so, evaluating petroleum systems with numerical techniques remains primarily a deterministic process resulting in non-unique solutions. In this talk, both synthetic analyses and real world case studies illustrate how machine learning is deployed in exploration workflows to reduce uncertainty. These examples demonstrate the mathematical and scientific reasoning for workflow design and the challenges encountered. The main goal of this approach is to demystify machine learning by showing how it can be effectively used in an exploration context if domain experts work together to integrate results. In this way, the value of machine learning in the evaluation of Earth resources enhances geologic understanding, one that moves beyond buzzwords and proprietary algorithms.
Allegra Hosford Scheirer is a Research Scientist at Stanford University, where she co-directs the Basin and Petroleum System Modeling Industrial Affiliates Program. Prior to Stanford, Allegra was a Research Geophysicist at the USGS. She is the editor of USGS Professional Paper 1713 and a past Associate Editor of Journal of Geophysical Research. Allegra’s degrees are from the MIT-WHOI Joint Program in Oceanography (Ph.D.) and Brown University (Geology-Physics/Math).