Seismic Characterization of Thin Ratawi Limestone Reservoir in Ladaira-Shaham, NW Kuwait using Machine Learning

Summary Ratawi Limestone formation in Kuwait belongs to the Early Cretaceous age of Thamanna group. Significant Oil discoveries were found in Ratawi Limestone and is a producing reservoir in many fields in Kuwait. However, it remains a challenging reservoir to develop because of its low permeability, substantial porosity heterogeneity, and thin nature of the reservoir units. So, it is important to understand the reservoir quality distribution away from the drilled well for effectively appraise and develop the reservoir. Seismic is an important tool to decode such complex reservoirs. Elastic attributes derived from seismic inversion are used popularly for this purpose. However, these products are limited to seismic resolution which leads to challenges in characterizing the thin reservoir units from seismic. Nevertheless, this challenge can be overcome if we can produce high resolution elastic attributes well-correlated with the reservoir properties. The paper presents a novel and robust workflow integrating Machine Learning and seismic inversion to characterize thin reservoir units, which are otherwise challenging to identify from seismic data using conventional approaches. This study aided in identifying the spatial distribution of porous zones within Ratawi Limestone, which combined with the structural interpretation results, will help significantly in placing the future wells.