ECCOMAS 2024

Multiphysics Simulation of CO2 Geological Storage by Mimetic Finite Differences and Virtual Elements

  • Borio, Andrea (Politecnico di Torino)
  • Castelletto, Nicola (Lawrence Livermore National Laboratory)
  • Gazzola, Thomas (TotalEnergies)
  • Hamon, François (TotalEnergies)
  • Ju, Xin (Stanford University)
  • Karimi-Fard, Mohammad (Stanford University)
  • Settgast, Randolph R (Lawrence Livermore National Laboratory)
  • White, Joshua A (Lawrence Livermore National Laboratory)

Please login to view abstract download link

Safe and efficient operation of CO2 geological storage projects requires numerical simulation of a multi-physics problem in which compositional multiphase flow and transport can be tightly coupled with the porous medium deformation. To simulate these processes, one needs to solve a set of coupled, nonlinear, time-dependent partial differential equations governing the mass conservation of each component and linear momentum of the solid-fluid mixture. We focus on the use of polyhedral meshes that prove particularly useful to discretize complex geological structures. We present a scalable fully implicit framework based on a displacement/total velocity/pressure/component density formulation that combines a virtual element approach for mechanics with mimetic finite differences for fluid flow and transport. A Newton-Krylov approach is used to advance the solution in time, preconditioned by a multigrid reduction method. The solver is implemented in GEOS, an open-source, exascale-compatible, research-oriented simulator for modeling fully coupled flow, transport and geomechanics in geological formations. We demonstrate performance and robustness of the proposed framework on a variety of challenging test problems, including realistic field cases.