Oil & Gas

This paper illustrates a robust methodology for achieving detailed mapping of sand injectites in the Greater Fram area using seismic attribute analysis and integrating machine learning techniques that are specifically targeted at injectites and fault prediction. Key to our methodology is the use of newly processed dual-azimuth seismic data, characterized by advances in pre-processing, noise mitigation, and velocity model building (VMB) technologies. Applying ML predictions to enable the identification of injectite geobodies provides deeper insights into these subsurface reservoirs, allowing a better understanding of the geometries of the injectites and helping to increase the efficiency of hydrocarbon exploration and production.