Enhanced recovery

Physics-Informed ML Improves Forecasting, Connectivity Identification for CO₂ EOR

The authors of this paper propose hybrid models, combining machine learning and a physics-based approach, for rapid production forecasting and reservoir-connectivity characterization using routine injection or production and pressure data.

Fig. 1—Permeability distribution of the field case.
Fig. 1—Permeability distribution of the field case.
Source: SPE 221057.

Routine well injection and production data contain significant information that can be used for closed-loop reservoir management and rapid field decision-making. Traditional physics-based numerical reservoir simulation can be prohibitive computationally for short-term decision cycles. As an alternative, reduced physics models often have a limited range of applicability. Pure machine-learning (ML) models often lack physical interpretability and can have limited predictive power.

×
SPE_logo_CMYK_trans_sm.png
Continue Reading with SPE Membership
SPE Members: Please sign in at the top of the page for access to this member-exclusive content. If you are not a member and you find JPT content valuable, we encourage you to become a part of the SPE member community to gain full access.