The broader impact/commercial potential of this Small Business Innovation Research (SBIR) project is to replace existing time-consuming numerical tools and inaccurate statistical and analytical methods in unconventional oil and gas field development design with more accurate hybrid models that take advantage of physical properties as well as advanced machine learning methods. The initial tests suggest that such AI software can significantly increase the prediction accuracy and reduce the cost per barrel of produced oil and gas. The results of this research enable oil and gas producers to increase their ultimate oil recovery from their unconventional reservoirs (also known as shale reservoir) by exploring all possible well and completion designs and finding the most optimal one based on multiple criteria. This STTR Phase I project proposes to develop an intelligent decision support system (IDSS) to optimize unconventional oil and gas field development designs. It is for the first time that a comprehensive IDSS is being proposed in the oil and gas industry for this purpose. The technology behind this software improves the field development design process at three levels. First, a robust predictive model using a hybrid approach (physics-based reservoir engineering, advanced machine learning, and deep learning) makes an accurate and fast production forecast for every potential design. Then, at the optimization level, considering the local subsurface geological uncertainty, the model finds the most optimum field development designs. Eventually, at the highest level, the cognitive unit helps decision makers to find their final design based on their objectives. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
