On the Importance of Correct Inclusion of Capillary Pressure in Reservoir Simulation

Abstract

Copyrig ht 1996 . Stee rirq Committee of th e Europea n IOR -Symposium. This peper was pt esented at the 8th. Eueopeen IO R -Symposium in Vbnna, AueMa , May 1 6 -17 . 1 99 6 This paper was aebcted tor presenta tion by the Stee rinp Comm~, following re view of infortn a tion conte in ed in an abstraet submitted by the authortsl. The pepar, as precented hes not been reviewed by t h e Steerinp C ommittee. ABSTRACTCapillary pressure is frequently omitted from large stele reservoir simulation , rationalized by arguments that may not always be valid . A number of papers have shown that correct modelling of capillary pressure may be as important for the simulation result as properly accounting for relativa permeabilities . Especially in strongly heterogeneous modale it is important to study the sensitivity to all petrophysical parameters before simplifications are introduced in the simulation model . In a previous paper from Norsk Hydro [1]. it was demonstrated and explained why stochastic oil / water relativa permeabilities (varying from cell to cell) in many cases can be replaced with one pair of "average" relativa permeabilities without significantly changing the simulation results. Our current study indicates that a similar procedure is not always valid for capillary pressure .A number of simulations exemplifies the effects of stochastic capillary pressure in various displacement regimes in a realistic geostatistical reservoir model : -Gravity dominated water/oil displacement -Viscous dominated water / oil displacement Varying water/oil mobility ratios -Including/excluding capillary pressur e -Stochastic vs . constant capillary pressure -Grid sensitivities Even if the simulation results can be explained by well understood physical theory , our study emphasizes that the net effect of capillary pressure on the oil production curve is in general difficult to predict without numerical simulation on a fine grid : Front smearing lende to give earlier water breakthrough while suppressing of channelling / fingering will have the opposite effect . These true physical effects c an be obscured by the numerical diffusion introduced by the finite difference schema applied in the reservoir simulator . INTRODUCTIO N

Numerical Evaluation of the Combined Effect of Wettability and Heterogeneity on Waterflood Performance

“…This is not surprizing. Effects of capillary pressure on large scale simulation have already been studied (Bu and Haoy, 1995) and the underlying physical mechanisms are well understood.…”

Domain decomposition methods for flow in heterogeneous porous media

“…We have neglected gravity forces, but the methods described can also be applied for models where gravity effects are included [17]. The following analytical form for the capillary pressure is chosen [35]:…”