Computational framework for modeling, simulation, and optimization of geothermal energy production from naturally fractured reservoirs

Abstract

We describe an open-source computational framework for the automated search for deviated multi-well lay- outs in hot fracture-controlled reservoirs that sustainably optimize geothermal energy production. This search is performed via 3D simulations of groundwater flow and heat transfer. We model the reservoirs as geologi- cally consistent, randomly generated discrete fracture networks (DFNs) in which the fractures are 2D manifolds with polygonal boundaries embedded in a 3D porous medium. The wells are modeled as line sources and sinks. The flow and heat transport in the DFN-matrix system are modeled by solving the balance equations for mass and energy, while expressing the momentum balance by the Darcy law. The spatial discretization is based on the finite element method stabilized via the algebraic flux correction. For the time discretization, we use a semi-implicit approach to enhance the solver efficiency. The optimization is performed via a gradient-free global optimization algorithm. By employing the immersed boundary method and a non-matching discretization strategy, the need for computationally expensive remeshing when altering well configurations within the reser- voir is effectively eliminated, thereby enhancing the robustness of the proposed framework and enabling fully automated optimization. We present the results of our optimization tests for randomly generated DFNs consisting of thousands of fractures, considering realistic values of physical parameters. To demonstrate the analytical capabilities of our open-source framework, we use it to analyze and visualize the above optimization results and the structure of the above DFNs. The developed framework was verified and validated using a set of simplified yet purpose-specific fracture configurations relevant to geothermal energy extraction in naturally fractured reservoirs.