(754f) Improving CO2 Enhanced Oil Recovery Performance through Data Analytics and Next-Generation Controllable Completions | AIChE

(754f) Improving CO2 Enhanced Oil Recovery Performance through Data Analytics and Next-Generation Controllable Completions

Authors 

Bosshart, N. - Presenter, University of North Dakota
Hamling, J. A., University of North Dakota
Sorensen, J. A., University of North Dakota
Azzolina, N., University of North Dakota
Patil, S., University of North Dakota
The Energy & Environmental Research Center (EERC) is conducting a 5-year project to field-test an advanced machine learning approach integrating controllable completions (interval control valves [ICVs]) to enable active (smart) well control during carbon dioxide (CO2) enhanced oil recovery (EOR). The concepts to be tested are that 1) ICVs can be used to direct fluids into selected zones of horizontal wells, improve conformance, and maximize enhanced recovery potential; and 2) real-time monitoring data can be integrated in a machine learning approach to develop a “human-in-the-loop” (semiautonomous) active control system for both injection and production wells. The concept will be validated through design and execution of a pilot CO2 EOR test in the carbonates of the Ordovician Red River Formation of southwestern North Dakota.

The pilot test focuses on a horizontal injection well paired with a horizontal production well, each completed with ICVs (up to ten ICVs per well). Rock properties are being assessed through laboratory investigations using core samples. Fast-flow pathways in the reservoir will be identified using a combination of downhole monitoring technologies. An active control system will be developed that uses machine learning to analyze real-time data sets and manage ICV operation (select which stages to open/close and their timing).

Geomodel construction and numerical simulation were conducted to predict key EOR performance metrics (e.g., incremental production and CO2 utilization) and inform pilot test operation. Sensitivity studies were conducted to investigate potential performance improvements in a range of scenarios and generate information applicable to other reservoir types. These studies included variations of permeability, heterogeneity, injection/production well ICV combinations, and operational schedule. These scenarios and their modeled impact on EOR performance will be discussed.

Successful completion of this field test will help reduce the uncertainty in CO2 EOR performance and improve overall EOR project economics. These improvements will help accelerate the adoption of the approach across existing and future CO2 EOR operations and will support the expansion of CO2 EOR into locations where current marginal economic outlook deters investment.