There is a significant opportunity to enhance traditional reservoir management with new quantitative tools and technologies that allow integration of all kinds to data to create accurate predictive models while significantly reducing the cycle time from data to decisions. This talk describes Data Physics, a unique modeling approach which is the amalgamation of the state-of-the-art in machine learning and the same underlying physics present in reservoir simulators. These models can be created as efficiently as machine learning models, integrate all kinds of data, and can be evaluated orders of magnitude faster than full scale simulation models, and since they include similar underlying physics as simulators, they have good long term predictive capacity and can even be used to predict performance of new wells without any historical data. We present applications of Data Physics models to real steamflood and waterflood optimizations with thousands of wells, wherein, the injectant is redistributed to maximize/minimize multiple objectives. A significant increase in actual incremental oil production and reduction in operational cost is demonstrated.
Speaker: Pallav Sarma, Tachyus