Probabilistic net present value analysis for designing techno-economically optimal sequential CO2 sequestration and geothermal energy extraction

• Published in: Journal of hydrology (Amsterdam) 2022-09, Vol.612, p.128237, Article 128237
• Main Authors: Mahdi Rajabi, Mohammad, Chen, Mingjie, Reza Hajizadeh Javaran, Mohammad, Al-Maktoumi, Ali, Izady, Azizallah, Dong, Yanhui
• Format: Article
• Language: English
• Subjects: CO2 plume geothermal system; Depleted oil reservoir; Hybrid Monte Carlo-genetic algorithm; Neural network; Risk-aware design
• ISSN: 0022-1694
• DOI: 10.1016/j.jhydrol.2022.128237

•Combined brine-produced and CO2-circulated geothermal energy harvest is analyzed.•A depleted oil reservoir in North Oman is employed as the case study.•Methodology is based on numerical modeling, ANN, and hybrid MCS-GA.•The proposed approach reduces the risk of non-profitability in presence of geological uncertainty. The use of CO2 as the heat transmission fluid, increases the efficiency of geothermal energy extraction from low-enthalpy resources such as depletion oil and gas reservoirs. In the resulting so-called CO2 plume geothermal (CPG) systems, the optimal choice of well position and operational parameters represents a strategic decision problem, due to its profound effect on efficiency. Combined simulation–optimization (SO) schemes have been recognized as a valuable tool in making these strategic decisions. Noting that the total lifespan of a CPG system consists of a ‘sequestration’ and a ‘circulation’ stage, past CPG SO studies only focus on the circulation stage, assuming that the reservoir is initially saturated with CO2. Hence they neglect the realistic state of the reservoir following CO2 sequestration, ignore brine-based power generation, and either neglect the sequestration costs or assume that the sequestration costs are part of the fixed initial investment. This study aims to fill this gap by developing a SO algorithm that takes into account both the sequestration and circulation stages of a CPG system lifespan in choosing optimal well location and operations. We frame the problem as a probabilistic risk-minimization scheme to allow for the consideration of geological uncertainty, and solve it through the combined application of a multi-phase numerical model, artificial neural networks, and a hybrid Monte Carlo-genetic algorithm method. Under this context, we successfully minimize the probability of having a negative net present value from the operation. We also examine the influence of economic factors on the profitability of the proposed system, and show that the net CO2 storage income is the economic variable that most affects the risk of non-profitability. Our case study involves a homogeneous, fault-blocked, inclined thin formation that is commonly present in oil and gas fields, but has been the subject of a very limited number of CPG studies.