We introduce an approach to estimate timeshifts and reservoir property changes in time-lapse seismic data using machine learning (ML). ML models were trained for each reservoir field of interest by generating synthetic seismic data using 4D reservoir and geomechanical models. We then apply the trained ML models to real 4D seismic data. The ML timeshifts show improved resolution and accuracy compared to traditional cross-correlation results.

Presentation Date: Monday, October 12, 2020

Session Start Time: 1:50 PM

Presentation Time: 2:15 PM

Location: 362D

Presentation Type: Oral

Keywords:
machine learning, upstream oil & gas, artificial intelligence, seismic data, difference seismic, neural network, production inc, exploration geophysicist 10, timeshift, time-lapse timeshift
Subjects:
Reservoir Characterization, Formation Evaluation & Management, Information Management and Systems, Seismic processing and interpretation, Four-dimensional and four-component seismic, Seismic (four dimensional) monitoring, Artificial intelligence
This content is only available via PDF.
2020. Society of Exploration Geophysicists
You can access this article if you purchase or spend a download.