Data based sensing of Shale Oil yield in Oil Shale Retorting process

Abstract Oil shale is sedimentary organic rocks that are being converted into useful shale oil and shale gas. North American regions, Canada and China are exploring the oil shale reserves to accommodate the depletion of natural oil and gas resources. Oil shale retorting technology is being utilized to convert the shale rocks into shale oil and shale gas. The major product is oil that is further treated to convert it into gaseous form. In this study, machine learning techniques like ensemble learning (least square boosting and bagging) and artificial neural network (ANN) are employed for data sensing of oil shale retorting process and being compared. Data is generated for ensemble models through MATLAB-Excel-Aspen interfacing. The proposed framework shows that ANN provides higher accuracy as compare to other models for oil shale retorting process for efficient oil recovery.