A New Framework for Compositional Simulation Using Reduced Order Modeling Based on POD-DEIM Conference Paper uri icon

abstract

  • Abstract Fast and reliable reservoir simulation is a key for the successful decision making in integrated reservoir studies. Large and complex multiphase reservoir models usually require expensive computational infrastructure. Physics-based model order reduction (MOR) methods have been introduced and applied (POD-DEIM, POD-TPWL), especially for mitigating the computational cost of black oil models in workflows that require multiple calls of the reservoir simulator. However, only a limited number of methods have looked deeper at the effectiveness of these techniques to multiphase and compositional simulation where expensive phase equilibrium calculations are added to the level of complexities associated with obtaining robust solutions. In this work, we develop the physics-based MOR techniques for rapid compositional simulations that accelerate calibrating of system of equations and phase equilibrium. The combination of proper orthogonal decomposition (POD) and discrete empirical interpolation method (DEIM) has been used extensively in two-phase flow systems. POD reduces the size of the system to be solved, and DEIM contributes to approximate the nonlinear terms for faster computation. These snapshot-based methods work in a two-step process. In the training case one can obtain the snapshots, which are the solutions at each time step, to derive the POD basis and DEIM basis. Then the test case is utilized to validate if the reduced model works for the different well control schedule cases, and to compare the speed of the simulation run time. Results show that the POD-DEIM technique enables us to approximate the conventional model with high levels of accuracy up to more than 99%. And it also enables a faster simulation owing to the reduced order system. In this study, we show the robustness of POD-DEIM method to reduce the computational cost for multi-phase, multi-component 3D reservoir model.

author list (cited authors)

  • Lee, J. W., & Gildin, E.

citation count

  • 0

publication date

  • July 2020

publisher

  • SPE  Publisher