Development and Application of Capacitance Resistance Models (CRMs) for Enhanced Oil Recovery Processes

Abstract

Evaluating interwell connectivity provides critical insights for reservoir management, including the identification of flow conduits, barriers, and injection-production imbalances. Injection and production rates contain valuable information about connectivity, and various methods have been developed to extract this information. Among them, the Capacitance Resistance Models (CRMs) have been successfully applied in numerous waterflooding projects. CRM is based on a linear productivity index and assumes a pseudo–steady-state flow regime for slightly compressible fluids. However, its application to gas injection projects is limited due to the high compressibility of gas and its strong dependence on pressure. Therefore, it is essential to identify the range of fluid and reservoir conditions under which the CRM can still yield reliable results. The first objective of this work is to determine the applicability limits of the conventional CRM for immiscible gas flooding by conducting a sensitivity analysis on its predictive accuracy. Various reservoir and fluid properties—such as temperature, pressure, viscosity, density, and heterogeneity—were tested to evaluate their impact on CRM performance. Unlike water, gas exhibits significant compressibility, leading to nonlinear flow behavior, especially under changing pressure conditions. Consequently, the second objective of this study is to modify the conventional CRM to account for gas compressibility and nonlinear interwell flow dynamics during immiscible gas injection. The results indicate that CRM accuracy decreases in low-permeability reservoirs. Additionally, the model performs worse in light oil systems compared to heavy oil under immiscible gas injection. High pressure and low temperature conditions, which can lead the gas state be closer to liquid state, were found to improve CRM performance. Temperature variations had a more pronounced effect on CRM results during gas injection than pressure variations. Furthermore, by incorporating a pseudo-pressure term into the CRM framework, this study demonstrates improved accuracy in evaluating reservoir performance under immiscible gas flooding. The third objective extends the application of CRM to enhanced oil recovery (EOR) processes such as polymer flooding. By modifying CRM to account for viscosity-related changes in mobility ratio, the model showed improved performance in analyzing polymer injection. A synthetic cartesian reservoir models and a benchmark case study are evaluated to illustrate the effectiveness of the proposed approach. In conclusion, applying the conventional CRM in gas injection cases may result in misleading estimates of interwell connectivity and inaccurate rate predictions. The improved CRM framework proposed in this study enable better characterization of gas flow behavior and its influence on reservoir dynamics, supporting more effective reservoir management decisions.