PREDICTION OF CO2-CH4-H2S-N2 GAS MIXTURES SOLUBILITY IN BRINE USING A NON-ITERATIVE FUGACITY-ACTIVITY MODEL RELEVANT TO MEOR

Numerical simulations of microbial enhanced oil recovery (MEOR) require computationally rigorous iterative methods to solve resulting system of flow, transport and kinetic reaction equations. This includes additional iterative procedures to account for the solubility of gas mixtures in the aqueous phase, which has the risk of degrading the performance of MEOR simulation programs. This work proposes a new noniterative fugacity-activity thermodynamic model to predict the solubility of CO2-CH4-H2S-N2 gas mixtures in brine. This model can readily be implemented in MEOR simulation programs to account for the effect of CO2-CH4-H2S-N2 gas mixtures solubility on the kinetics of microbial reactions. Fugacity coefficients (ϕ_i) in the proposed model were calculated using Predictive Peng-Robinson 78 (PPR78) and Peng-Robinson (PR) equation of state (EOS). Predicted solubility values were compared against reported experimental data measured at temperature and pressure conditions relevant to MEOR. The proposed model was successful in predicting CO2 solubility in the aqueous phase with ϕ_i calculated either using PPR78 EOS or PR EOS. The proposed model with ϕ_i calculated using PR EOS accurately predicts H2O solubility values in the gas phase. However, the proposed model with ϕ_i calculated using PPR78 EOS overestimated the solubility of H2O and H2S in the gas phase. At temperature and pressure conditions relevant to MEOR, using calibrated binary interaction parameters (PR EOS) leads to more accurate predictions than binary interaction parameters estimated from the group contribution expression (PPR78 EOS). Comparison of predicted CO2-CH4-H2S gas mixtures solubility in the aqueous phase obtained with the proposed non-iterative fugacity-activity model and an iterative fugacity-activity model, confirmed that proposed model with ϕ_i calculated using PR EOS can be used as substitute for iterative fugacity-activity models that relays on the solution of Rachford-Rice equation.