Intelligent Production Optimization in Real-Time by Implementing Hybrid Data-Physics Simulation

In day to day operations, oil producers need to optimize their production workflow to reduce operational costs. Building a physics-based reservoir model is costly and time-consuming and is not suitable to generate and compare many scenarios. The alternative procedure, Data-Driven Model, is fast enough; however, testing and validating the model is controversial. In this project, a hybrid data-physic framework will be developed, which meets the requirement to quantitatively describe the physical behavior of the reservoir and also predict its mechanism statistically. The model is fast and can be feed by real-time data and create many potential operational scenarios and recommend the most favorable ones.

Raya Matoorian
Faculty Supervisor: 
Roman J Shor;Roberto Aguilera
Partner University: