Uncertainty Analysis of Geomechanical Responses: China’s First Offshore Carbon Capture and Storage Project
Zhiqiang Wang,Shuyang Liu,Hangyu Li,Qizhi Tan,Wenyue Sun,Junrong Liu,Jianchun Xu,Xiaopu Wang
DOI: https://doi.org/10.1016/j.fuel.2024.133728
IF: 7.4
2025-01-01
Fuel
Abstract:Enping 15–1 project is China’s first Carbon Capture and Storage (CCS) project to store CO2 in the offshore deep saline aquifer formations. Due to the lack of detailed observed data, geomechanical responses are fraught with uncertainty, which can result in unanticipated risks during the implementation of CO2 injection. Thus, the uncertainty analysis of geomechanical responses shows great importance and necessity for CO2 storage projects. To bridge the gap, in this work, a Thermal-Hydraulic-Mechanical-Chemical (THMC) coupling model was built to study the uncertainty of geomechanical responses including surface uplift and rock damage. First, Monte Carlo sampling technology was used to create 616 cases. Then the sensitivity of 13 parameters (including reservoir, cap, and operating parameters) on geological responses was analysed using the Sobol sensitivity analysis method. The results show that 90.10 % of cases in the uncertainty analysis have a maximum surface uplift of less than 8 mm, and 80.84 % of cases have a safety factor higher than 0.5. However, it is the most likely to cause surface uplift over 100 mm and rock failure when the reservoir permeability is 1 mD. Therefore, the local low permeable zone is at more risk of rock damage and geological hazards. In addition, five parameters (i.e., reservoir permeability, reservoir Young’s modulus, initial horizontal/vertical stress ratio, cap cohesion, and well length) have large contributions to the surface uplift. Three parameters (i.e., Reservoir permeability, reservoir cohesion, and initial horizontal/vertical stress ratio) have considerable effects on the reservoir safety factor. Eventually, proxy models using Regression Trees (RTs) were proposed to predict the surface uplift and reservoir safety factor, which is efficient and saves computational resources. This work provides insights and suggestions into the geomechanical responses of CCS project.