Optimization of Recovery by Huff ‘n’ Puff Gas Injection in Shale-Oil Reservoirs Using the Climbing-Swarm Derivative-Free Algorithm

Abstract

Summary Recent improved-oil-recovery and enhanced-oil-recovery (EOR) methods in shale reservoirs use huff ‘n’ puff gas injection (H&P). Investigating the technical and economic impact of this technology for one well is challenging and time consuming, and even more so when the petroleum company is planning H&P and refracturing (RF) jobs in multiple wells. Thus, in this paper we present an original methodology to learn how to perform these tasks faster and at lower cost to improve oil recovery. The procedure is explained with the use of an actual H&P gas-injection pilot horizontal well in the Eagle Ford Shale, the performance of which is matched using the methodology developed in this paper. The methodology includes use of an original climbing-swarm (CS) derivative-free algorithm that drives, without human intervention, desktop computer or laptop material-balance (MatBal) and net-present-value (NPV) calculations. The code was written in open-source Python programming language. Following history match, the methodology demonstrates that significant improvements in oil recovery can be obtained by injecting gas at larger rates during shorter periods of time (as opposed to injecting gas at lower rates during longer periods of time). Once oil-recovery improvement in a pilot horizontal well is demonstrated, the methodology is extended to the analysis of H&P gas injection and RF in horizontal wells and shale reservoirs that have not yet been developed or are in initial stages of development; this provides a preliminary assessment of H&P and RF potential. Results indicate that oil recovery and NPV from multiple wells can be improved significantly by a strategic combination of H&P gas injection and RF. A combination of derivative-free optimization algorithms, MatBal calculations, and NPVs permits optimizing when to start the H&P gas-injection project, the optimum gas-injection rates (GIRs) and time span of injection, the reservoir pressure at which gas injection should be started in each cycle, and the time span during which the well should produce oil, before starting a new cycle of gas injection. The development strategy of shale-oil reservoirs could be improved significantly if the possibility of H&P gas injection is considered before field development. This could be the case of the Eagle Ford Shale in Mexico, La Luna Shale in Colombia and Venezuela, Vaca Muerta Shale in Argentina, and other shale-oil reservoirs worldwide. The paper contributes the development of an original methodology, which includes use of a derivative-free algorithm we call CS. CS drives the desktop computer or laptop to perform MatBal and NPV calculations, without human intervention, once the optimization process is started. The methodology improves oil recovery and NPV from a single horizontal well or from multiple horizontal wells operating under H&P gas injection.