Mapping Sand Injectites with Seismic Attributes and Machine Learning

This technical report presents a methodology for detailed mapping of sand injectites in the Northern Viking Graben, Norwegian North Sea, utilizing seismic attributes and machine learning techniques. The study highlights recent exploration activities that have confirmed the injectite play in the region, with notable discoveries such as Kveikje and Heisenberg. The proposed methodology involves analyzing seismic data to identify injectite bodies, which are characterized by complex geometries and high permeability, making them significant for hydrocarbon exploration. The report outlines the integration of RGB blending and machine learning algorithms to enhance the identification and mapping of these injectites. By employing dual-azimuth seismic data and advanced imaging techniques, the study successfully identifies potential areas for further exploration. The identification of injectite geo-bodies is emphasized as a critical factor in understanding subsurface reservoirs, ultimately facilitating more efficient hydrocarbon resource exploitation. The findings contribute to the ongoing efforts in improving exploration strategies within this mature oil and gas region.