Selected Recommended Practices for Increasing the Efficiency and Accuracy of CO2 Sequestration Models

Abstract

Abstract Computer simulation has become widely used to predict the life of CO2 plumes resulting from planned subsurface sequestration schemes. The process is very similar to what has been done in the oil and gas industry for decades, however, while reservoir simulations can be "history matched" to real-world performance, analog data is mostly missing in order to calibrate proposed CO2 project models. Therefore, it is critical to attempt to understand the impact of the various "default" input parameters that are often used in CO2 sequestration models. This paper attempts to highlight potential enhancements in the modeling process that can lead to improved accuracy of predictions. Many CO2 screening investigations use simple homogeneous block models in place of utilizing more complex realistic geologic models. Comparisons between homogeneous models of porosity and permeability with simple stochastically generated porosity and permeability (having the equivalent average values), could make significant differences. The use of Black Oil models instead of Compositional models will be investigated as a method to improve time efficiency allowing the investigation of more "what-if" scenarios. Guidelines for simple calibration of a black oil model to a compositional model are also presented. The importance of residual phase trapping as a potential primary trapping mechanism in addition to structural and stratigraphic trapping will be investigated. While including parameters for residual trapping, too often default values in simulation tools are used without recognizing the high variability of these parameters, especially considering the laboratory-based observations for some of them. Several scenarios are presented highlighting the importance of these uncertainties. End-point values of relative permeability curves can have significant effects on plume geometry and are modeled in various scenarios. Critical Gas Saturation is often overlooked in the modeling process. This value is often defaulted to an unrealistically low value that can lead to "runaway" low saturation plumes when models are run for extended periods (as required by government regulations). Critical gas trapping at the edge of plumes can be an often-overlooked additional trapping mechanism.