Low Salinity Waterflooding for Enhanced Oil Recovery - Stochastic Model Calibration and Uncertainty Quantification

摘要: We focus on key aspects related to the quantification of uncertainty associated with modeling Enhanced Oil Recovery (EOR) through Low Salinity (LS) water injection in a reservoir. salinity waterflooding is an emerging EOR technique which injected controlled improve oil recovery, as opposed conventional where brine usually used. Several mechanisms have been proposed underpin processes leading additional mobility, but none them has conclusively identified driving cause. Literature results suggest that LS causes alteration wettability porous medium, more favorable conditions for recovery. In this context, simulation models represent process using salinity-dependent relative permeabilities developed. Here, we consider tertiary coreflood experiment performed at Eni laboratory facilities injection, following sea flooding. and permeability curves are parameterized Corey model. Model parameters their uncertainties estimated within stochastic inverse approach, upon relying classical reservoir simulator simulate measured The likelihood function maximized joint use Latin hypercube sampling Metropolis Hastings algorithm, while model coupled universal Kriging technique. posterior sample then employed quantify propagation sector selected North-East African sandstone This enables us impact parameter expected production resulting from field scale application under study. reveals potential recovery considered field.