The risk of gas release in a formation under reservoir pressure depletion conditions can lead to a productivity drop. This is particularly critical for carbonates with highly non-uniform properties. Identifying intervals with high gas-oil ratios using conventional PLT methods is challenging when initial gas saturation is low, or wellbore pressure is close to the bubble-point pressure. This paper presents an approach for early gas detection based on machine learning analysis of passive acoustic data as part of PLT data.

Keywords:
geology, geologist, inflow control device, well logging, machine learning, artificial intelligence, production monitoring, plt acquisition, production logging, log analysis
Subjects:
Well & Reservoir Surveillance and Monitoring, Fluid Characterization, Formation Evaluation & Management, Information Management and Systems, Production logging, Phase behavior and PVT measurements, Open hole/cased hole log analysis, Artificial intelligence
Copyright 2024, Society of Petroleum Engineers DOI 10.2118/223396-MS
You can access this article if you purchase or spend a download.