Pilot Testing Issues of Chemical EOR in Large Fractured Carbonate Reservoirs

Abstract

Abstract Many world class large carbonate reservoirs leave behind at least half of the initial oil in place. Typically water injection is used to improve oil recovery while gas injection is used to maintain pressure or to promote oil gravity drainage. Immiscible gas injection, including injection of CO2, has been implemented on a large scale in only few fields for economic reasons. Furthermore, interest in using surfactants in large carbonate reservoirs has recently flourished. As a result, we began to investigate the viability of designing and conducting a manageable pilot test program in a large fractured carbonate reservoir using a single-well, dual-completion system to evaluate the efficacy of the surfactant oil mobilization and oil capture. However, pilot testing in large reservoirs is very expensive and requires a long time to complete. These issues are less problematic in pilot testing of small and thin reservoirs in onshore field. In this paper we present the results of a conceptual model to simulate the performance of surfactant flooding in the above-mentioned pilot test configuration. Three different model formulations, having different approaches to gridding and grid-refinement, were used. These include the dual-porosity, dual-permeability, and conventional variable permeability-porosity formulations to simulate fracture-matrix interactions. Simulation of pilot tests using dual-porosity models shows that gravity is most effective during waterflood but not as effective during the surfactant injection while in the variable permeability-porosity fine-grid model the surfactant oil recovery is greater because both gravity and viscous displacement contribute. We also present a study using our in-house numerical simulator for dual-porosity, dual-porosity refined matrix, dual-porosity with adjustments for viscous displacement and variable permeability-porosity fine-grid models. Dual-porosity models with matrix refinement and viscous displacement give more incremental oil recoveries than the conventional formulation. Results are also compared with variable permeability-porosity fine-grid and coarse-grid models. We will explain the reasons and will indicate which model is more reliable. In general the results of this study give an insight into the viability of using surfactant injection in thick carbonate reservoirs both in the pilot and production stage.