Equinor Research Fellowship

From 2018 to 2021 I held an Equinor Research Fellowship, working with Equinor’s North American Data Science team on production forecasting and well-placement optimization for unconventional reservoirs.

The production-forecasting methods developed during the fellowship were deployed in production at Equinor. That work is proprietary and unpublished; the methodological line of research it seeded is reflected in the papers below.

Publications from this work

• Scalable Risk-Averse Well-Placement Optimization Using Quadratic Knapsack Problem and Randomized Singular Value Decomposition. Farell, Bickel, Bajaj. SPE Journal 31(1), 480–496 (2026). [code]
• Amortized value-of-information screening for point-sample reservoir appraisal under spatial uncertainty. Farell, Bickel, Bajaj. Working paper, SSRN (2026).
• Field-Scale Bayesian Production Forecasting via Spectral Gaussian-Process Mixtures. Farell, Bickel, Bajaj. Unconventional Resources Technology Conference (URTeC) (2025). [code]
• Bayesian Port-Hamiltonian Surrogate for Three-Phase Reservoir Flow Simulation. Farell, Bickel, Bajaj. SPE Middle East Oil, Gas and Geosciences Show (MEOS GEO) (2025). [code]
• Estimating resources in unconventional assets: Spatial bootstrapping with n-effective. Farell, Pyrcz, Bickel. Journal of Petroleum Science and Engineering 212, 110174 (2022).

Acknowledgment

This work was supported in part by an Equinor Research Fellowship (2018–2021). I am grateful to Dr. Fakhri Landolsi (Equinor North American Data Science) for valuable insights during the early phases of this research.