Quantification of CO₂ flow and transport in the subsurface: Uncertainty due to equations of state algorithms

The purpose of this chapter is to evaluate uncertainty and variability generated from equations of state (EOSs) algorithms and to investigate how small errors generated from EOS algorithms propagate into evaluation of CO ₂ trapping mechanisms. In a previous study, we developed two integrated EOSs algorithms assembled for calculating thermophysical properties of multiphase CO₂ in deep brine reservoirs, including density, fugacity coefficient, enthalpy, viscosity, and solubility. In this chapter, we extended this previous work and focused on the uncertainty and variability associated with the choice of EOSs algorithms, using these two specific EOSs algorithms as an example case. For sake of example, we focused this uncertainty evaluation on a specific case study of the SACROC oil field in western Texas, a site notorious for 35 years of CO₂ injection activity for enhanced oil recovery. For the work described in this chapter, we adapted a 3-D high-resolution geocellular model of a specific section in the SACROC oil field, developed using 3-D seismic data and extensive well logging data. Comparison of these EOSs algorithms using the 3-D SACROC model indicates that thermodynamic properties (fugacity coefficient and solubility) are strongly coupled with transport properties (e.g., fluid density and saturation) in numerical codes. While differences in CO₂ solubility estimated by these two integrated EOSs algorithms are very small, the cumulative dissolved mass of CO₂ ultimately predicted by the two EOSs algorithms is extremely different. In simulations of SACROC oil field, for example, the total dissolved CO₂ mass differed by almost 90,000 tons over 2000 years.