Complex non-linear effects associated to unconventional production can be represented using numerical models. Even for the simplest homogeneous equivalent model (and even more for complex fractured representations) these low-permeability environments require detailed, CPU intensive models. History matching and Monte-Carlo analyses become long and tedious processes.
In this paper, a series of efficient reduced-order numerical proxies is proposed to replace these detailed models:
For simple geometries, numerical SRV-bounded or trilinear models are used. These 1-D models are fast and the computation time is independent of the number of fractures. Nevertheless they accurately capture transients and nonlinear effects related to PVT, non-Darcy diffusion and stress-dependence.
For more complex geometries or longer time-scale simulations, other proxies are presented. They use of automatic coarsening or refining of control volumes using assumptions on the flow regimes. This preserves early-time transient responses and multiphase effects with a much smaller system size.
This collection of proxies significantly speeds up the simulation process with good accuracy, hence enabling the routine use of numerical models for history matching and Monte-Carlo simulations.