Unveiling Challenges and Enhancing Robustness of CO2 Storage Capacity Estimation in a Depleted Gas Reservoir for Storage Development Planning

Abstract The strategic imperative of PETRONAS in addressing the escalating gas demand in Malaysia necessitates the development of gas fields characterised by high CO2 content. However, this strategy poses a formidable challenge, as the resulting CO2 emissions are poised to exceed targeted reduction limits necessary for achieving net zero carbon emissions by 2050. To ensure sustainable development, the focus shifts to the critical imperative of decarbonisation through CO2 sequestration, either within depleted gas reservoirs or saline aquifers. This paper systematically examines the feasibility of utilising depleted gas reservoirs as CO2 sequestration sites, primarily investigating the anticipated relationship between the estimated ultimate recovery (EUR) and storage capacity under various aquifer strength drive mechanisms through simulation modelling for well and field levels. Accordingly, a robust compositional simulation model is formulated, incorporating specialised CO2 storage features to assess the viability of CO2 injection within the same geological formation post the field's developmental stages. An uncertainty study is conducted to evaluate EUR and potential CO2 injection as pivotal factors in any development plan and CO2 handling management. The study findings reveal a discernible correlation between the EUR/Storage ratio and Aquifer Strength Index (ASI) offering a valuable tool for accurately estimating the storage capacity of depleted gas reservoirs. Simulation results underscore the critical importance of understanding aquifer performance and strength during the production field life of a depleted gas reservoir for effective planning of CO2 storage projects. The paper further suggests enhancing the conventional volumetric storage calculation approach by incorporating a multiplier (E-factor) to account for the impact of aquifer strength and improve storage capacity estimation during the early stages of development planning. Machine learning models including Gradient Boosting Regressor, Random Forest Regressor and Linear Regression shows a prediction of 90 per cent accuracy for the E-factor using different water-gas ratio (WGR) and ASI numbers. Simulation outcomes indicate that this E-factor, ranging from 0.3 to 1.3 is contingent upon reservoir aquifer strength, with higher aquifer strength resulting in a smaller E-factor and vice versa. Neglecting or underestimating this factor during early planning stages may lead to inaccurate anticipation of CO2 storage potential, escalating the risk of development and potentially resulting in unsuccessful project implementation with substantial economic consequences. This study introduces an innovative approach considers uncertainties associated with aquifer strength and its consequential impact on CO2 injection plans. Notably, the findings of our research illuminate a compelling correlation between the EUR/Storage ratio and aquifer strength, providing a valuable tool for precise estimation of storage capacity in depleted gas reservoirs. This strategic approach is fortified by meticulous CO2 management practices, ensuring the successful implementation of projects, and mitigating economic risks.