Effects Of CO2 Solubility in Brine on the Compositional Simulation Of CO2 Floods

Abstract

Summary A compositional simulator was modified to account for the effects of brineon CO2 solubility and aqueous-phase density and viscosity. Simulations of oildisplacements using continuous injection of CO2, CO2 slugs, and CO2water-alternating-gas (WAG) injection were conducted to determine the effectsof brine on the performance of CO2 EOR processes. Introduction The presence of an aqueous phase will influence the phase behavior of CO2/hydrocarbon systems. The goal of this study was to correlate these effectsaccurately and simply, to incorporate them into a reservoir simulator, and toevaluate their significance on the modeling of CO2 floods. This wasaccomplished by comparing simulations in which (1) the CO2 solubility in theaqueous phase was ignored, (2) the CO2 solubility in the aqueous phase wasequal to that of CO2 solubility in pure water, and (3) the effect of dissolvedsolids in the aqueous phase on CO2 solubility was considered. Henry's law andan empirical correlation for the effect of dissolved solids on solubility wereused to predict CO2 solubility in the aqueous phase. Effects of Dissolved Solids on the Equilibrium and Transport Properties CO2 Solubility in Pure Water. In a previous study, 110 solubilities along 12 isotherms between 25 and 250 degrees C [77 and 482 degrees F] were compiledover the 4.83 to 72.41 MPa [700 to 10.500 psia] pressure range to developcorrelations for the Henry's constant for pressure range to developcorrelations for the Henry's constant for the CO2/water system. One of theempirical correlations for the Henry's constant (used in the Krichevsky-Kasarnovsky equation) is given below: (1) The Lyckman et al. correlation was used in conjunction with this expressionto determine the partial molar volume of CO2 in water CO2 fugacities weredetermined from the Peng-Robinson equation of state (EOS). Effect of Dissolved Solids on CO2 Solubility. In a previous study, thereduction of CO2 solubility on a weight basis was correlated to the solubilityof CO2 in pure water on a weight basis, as suggested by Klins. Although thecorrelation is completely empirical, it is simple to use and is a function ofone parameter, total dissolved solids (TDS), rather than individual ionconcentrations. This correlation, also used in this study, is presentedbelow. (2) where C TDS is the TDS concentration in weight percent. The conversionbetween CO2 solubility in brine on a mass and molar basis was needed becausethe simulator was designed for molar component material balances. (3) where (4) Eqs. 3 and 4 represent the brine phase as a water/NaCl solution ofequivalent TDS concentration. Effect of Dissolved Solids on Aqueous-Phase Density. In reviewingaqueous-phase density data, we noted that the molar densities of brines areroughly equal to that of water at the same temperature and are relativelyunaffected by pressure. Therefore, a reasonable estimate for the mass densityof either NaCl or CaCl2 solutions was obtained by multiplying the molar densityof water by the molecular weight of the brine (assuming all dissolved solids tobe NACl). (5) Any other common, such as that of Rowe and Chou, could also be used beused. Effect of Dissolved Solids on Aqueous-Phase Viscosity. Several data sources for the viscosity of (H2O+NaCl) solutions, one tabularand one graphical, were used to develop the following expression, which relatesbrine viscosity to that of pure water for the 25 to 100 degrees C [77 to 212degrees F] temperature range under the assumption that the effects of pressureand dissolved CO2 may be ignored. (6) Although brine can contain many dissolved ions, we noted that on a weightbasis, (7) Although these particular expressions for brine density and viscosity and CO2 solubility in brine were used in this study, any correlations can be usedin the following procedure. These simple correlations were used because theywere functions of only one brine parameter, C TDS. If more accuratecorrelations based on specific ion concentrations are used, additional materialbalances and correlation parameters must be introduced for each dissolvedion. Governing Equations for Compositional Simulation The compositional model used an implicit-pressure, explicit saturation, and-composition (IMPESC) formulation. The transmissibilities, capillary pressures, and gravity terms are evaluated at the nth time level. The L superscript in thefollowing expressions refers to the Lth iteration of the pressure-convergenceloop within a time level. Components 1 through nc include CO2, gases, andhydrocarbons. Component nc + 1 refers to water and nc + 2 to the dissolvedsolids. Components 1 through exhibit solubility in the aqueous phase;components + 1 through nc are not soluble in the aqueous phase. Nghiempresented the basic equations used in an IMPESC compositional simulator, inwhich CO2 solubility in the aqueous phase is accounted for. To account fordissolved solids in the phase is accounted for. To account for dissolved solidsin the injected and reservoir brines, an additional component material balancewas used in which the solids were assumed to be soluble in only the aqueousphase.