Even though the proposed screening criteria for low salinity water flooding (LSWF) are fulfilled, improve recovery is not always obtained. The LSWF mechanisms are therefore still discussed. The objective for the study was to describe the brine-rock interactions at high and low salinity. Reservoir core plugs were flooded either by formation water, sea water and low salinity waters in succession or by low salinity water directly from initial water saturations. Effluent samples were analyzed for ionic concentrations and pH. Relative permeability (kr) and capillary pressure (Pc) curves were obtained at the core scale by history matching the experimental production and differential pressure across cores using a simulation tool. A developed two-phase model was used to predict the release of divalent cations from the rock during LSWF, and to relate this to the oil recovery. The high salinity water flood with formation water was found to give close to piston-like displacement, while the oil was produced over much longer periods in LSWF. The estimated kr- and Pc-curves indicated that the rock was water-wet in the high salinity floods and mixed-wet in the low salinity floods. The experimental results were in accordance with the modeling of the brine-rock interactions. When the formation water was replaced with the low salinity water, increase in the concentrations of divalent cations onto clay surfaces was predicted for the selected brine compositions. Higher concentrations of polar oil components can then be bonded to the clay surfaces by the divalent cations and make them less water-wet. It is concluded that the low salinity water altered the wetting state of the rock. The direction of alteration can be explained by ion exchange taking place on the clay surface. The low salinity water potential for improving recovery should be considered on a case by case basis based on the interactions between the formation brine, injected brines, oil components, and rock type.

The potential of low salinity water flooding (LSWF) to improve oil recovery in sandstone cores has been well documented. The objective of this paper is further development and examination of a one-dimensional mathematical model for the study of water flooding laboratory experiments. The model describes how dissolution/precipitation of various carbonate minerals and multiple ion-exchange (MIE) will have an impact on the water-oil flow function. The model is formulated such that the total release of divalent cations from the rock surface will give rise to a change of the relative permeability such that more oil is mobilized. The combined effect of MIE and dissolution/precipitation was modelled to determine how this will affect pH and the total release of divalent cations The effects of solubility of carbonate minerals on the ion exchange process during the LSWF have been studied. Dynamic experiments were carried out using brines with different compositions and reservoir rocks. The ion concentrations and pH in effluent samples were determined. In experiments, pH was found to increase during the interaction between low-salinity brine and reservoir rocks. The ion composition of the low salinity brine was also found to change due to interaction with the reservoir rocks. The model developed successfully matched the experimental results. The simulation results show that dissolution of carbonate minerals can occur and this will then alter the composition of the injected low-salinity brine, the concentration of divalent ions on the rock surface and hence the potential of low salinity to improve recovery. It is important during the evaluations of the potential for LSWF to take into account that dissolution of minerals can change the composition of the injected low salinity brine. The potential should be evaluated using the composition that the brine will have in the main part of the reservoir.