Comparison of Four Numerical Simulators for SCAL Experiments
- Roland Lenormand (CYDAREX) | Kjetil Lorentzen (PRORES AS) | Jos G. Maas (Consultant) | Douglas Ruth (University of Manitoba)
- Document ID
- Society of Petrophysicists and Well-Log Analysts
- Publication Date
- February 2017
- Document Type
- Journal Paper
- 48 - 56
- 2017. Society of Petrophysicists & Well Log Analysts
- 10 in the last 30 days
- 216 since 2007
- Show more detail
There have been several benchmarks to test and compare reservoir simulators, but so far, there are not the equivalent exercises for numerical simulators used in the design and interpretation of SCAL laboratory experiments. In this study, we have compared four simulators used for the determination of relative permeability and capillary pressure from SCAL experiments. Several tests are performed in direct simulation (no history matching) with one or two fluids injected, generally called unsteady-state (USS) and steady-state (SS), either without or with capillary pressure corresponding to a mixed wettability (positive and negative Pc in imbibition) sample. In addition, a centrifuge drainage experiment is included in the comparisons.
After discussion, the latest versions of the four simulators use the same boundary conditions and give similar results.
An important point that concerns both inlet and outlet is the notion that in the laboratory the plugs are in equilibrium with fluids in the endpieces at the beginning of most experiments. If out-of-equilibrium conditions (spontaneous imbibition) occur, this phenomenon must be clearly identified because it leads to countercurrent flow at the inlet and/or outlet, and possibly to negative pressure in the water phase. Normally, for SS and USS relative permeability measurements and centrifuge experiments, we assume that fluids are at capillary equilibrium at the beginning of the experiments.
For boundary conditions, two simulators use an extra grid block with Pc = 0 to represent the fluids in the endpieces. One simulator uses directly the boundary condition on the first and last grid block within the plug and the fourth simulator uses a zero-width grid-block set to a fixed saturation condition. We show that the three approaches lead to the same results for pressures and saturation inside the plug.This study does not cover all the types of displacements, and we recommend that providers of SCAL simulators give more details concerning the type of boundary conditions and the way they are coded. Tabular data for Kr, Pc and for the results will be available on the websites of the authors, and on the SCA website.
|File Size||5 MB||Number of Pages||9|