Low salinity water flooding of North Sea reservoirs: Physical and chemical effects

In some sandstone reservoir rocks low salinity water flooding (LSWF) has been reported to improve the oil recovery compared to high salinity water flooding. Even though the proposed criteria for having improved oil production by LSWF are fulfilled, other sandstone reservoir rocks have not shown any improvement of oil recovery. In the project experiments and simulations have been carried out to establish knowledge about rock-brine-oil interactions and recovery mechanisms in LSWF by varying the composition s of brines (formation water, secondary high salinity brine, low salinity brine), oils and rocks. The oil recovery in LSWF can be improved due to reduction of residual oil saturation and /or improvement of flow functions. It is the improvement of the tot al sweep efficiency (macroscopic and microscopic) in LSWF that will determine whether the process is cost efficient. When low salinity water is injected, interactions with the rock/minerals, e.g. ion-exchange and solubility of minerals, can change the com position of the water. It is the water contacting the main part of the reservoir that should have the potential to improve the oil recovery. The composition of the injected low salinity water should therefore be optimized by taking into account both the c ation exchange and solubility of minerals. The crude oil composition has also been found to be important for the rock-brine-oil interactions, both for the initial wettability conditions and the wettability alteration during LSWF. An one-dimensional mode l describing the effects of rock-brine interactions on the oil production, has been developed and evaluated. The improved oil recovery in LSWF has in most cases been found to be due to reduction in the concentration of divalent cations onto clay surfaces and thereby alteration of wettability to more water-wet. Interpolation between saturation functions at high and low salinity by the divalent cation concentration onto clay surfaces was presented. Good match between experimental and simulation results have been obtained. The model has been used for screening of oil reservoirs for their LSWF potential. The LSWF potential for promising reservoirs has then been evaluated in experiments. Promising low salinity water compositions have also been identified using this model. In addition, it has been shown that the potential for tertiary LSWF will depend on the composition of the secondary high salinity brine. Reservoir simulators should be able to simulate the main recovery mechanisms in LSWF. Change in the satu ration functions should in most cases be simulated by interpolation between saturation functions at high and low salinity by using the concentration of divalent cations onto clay surfaces. Reservoir simulators with LSWF options are available using the con centration of ions in the water phase or the total salinity of the water phase in the interpolation between saturation functions, i.e. they can?t interpolate by using the concentration of divalent cations onto clay surfaces. LSWF options in Computer Model ing Group?s GEM? and UTCOMP - IPhreeqc (combination of Phreeqc and UTCOMP) have been presented with the possibility to do this type of interpolations. The option in GEM was presented by comparing with experimental results from the project. The first scr eening of the potential for LSWF in oil reservoirs can be carried out by geochemical simulation of the rock-brine interactions using the compositions of rock, formation water and earlier injection waters as input. The potential to reduce the concentration s of divalent cations onto clay surfaces should then be determined using available low salinity water compositions (from natural sources and/or from LSW-plants). These simulations should be confirmed by flooding experiments. Unsteady-state flooding experi ments should be performed to determine the recovery potential for the most promising low salinity water compositions and then the saturation functions at high and low salinity brine compositions should be determined, including determination of the interpo lation function. Since the LSWF potential for improving oil recovery depends of formation water composition, secondary water composition, low salinity water composition. rock composition and oil composition, it has to be evaluated on a case by case basis .

A routine/1D-model for simulation of low salinity floods at laboratory scale will be established. A function describing the effect of salinity reduction on the flow saturation functions will be included in the model. Experiments will be carried out to est ablish descriptions of the main mechanisms in low salinity water flooding. The model will be validated by comparing the modelling results with core flooding experiments. Special designed experiments will be important in validation of the model. Reservoir sandstone rocks with different mineral compositions and distributions will be selected. The oil composition, brine composition, core dimensions and temperature will be varied. Rocks will be characterised (mineral composition and distribution) before and after core floods. More detailed studies of mechanisms will be carried out by flooding of columns of minerals and mineral mixtures, and cores. The effects of adsorption and desorption of polar oil components, and ion-exchange (minerals and reservoir rock) on zeta potential and contact angle will be studied. Chemical modelling will include: Incorporate surface chemistry in the existing chemical modelling software EQAlt (Cathles 2006). Predict surface potential as a function of the aqueous chemistry, pres sure and temperature. Include oil rock interactions in the chemical model, guided by experiments performed Predict water flooding behaviour by the chemical model. Validation of the 1D-model will be carried out to verify that the model is capable of captur ing important effects of low salinity on oil recovery. Larger scale: A simple extension to a water flood simulator that can be used to translate laboratory results into field scale estimates of low salinity water flooding processes will be worked out. Po tential at larger scale will be evaluated using real or realistic sandstone reservoir models. A tool/best practice suitable for evaluation of potential for sandstone reservoirs will be established.