Carbon Capture and Storage in Depleted Hydrocarbon Reservoirs: Lessons Learned from Projects in Asia Pacific

Abstract

Abstract Carbon Capture and Storage (CCS) is a lot harder than it seems. Storage capacity can be far lower than a back-of-the-envelope calculation would suggest and keeping Carbon Dioxide (CO2) in the desired phase requires careful engineering. As the adoption of carbon pricing mechanisms increase in Asia Pacific, more CCS projects are expected to be developed. In this paper, we share some technical lessons learned while working on several of these projects. The paper provides an overview of factors that should be considered for CO2 storage in depleted hydrocarbon reservoirs, primarily concerning capacity, injectivity and containment. We further discuss some of the challenges faced while assessing these factors. Firstly, we highlight the impact that regional geothermal gradients have on storage site selection. Particularly, we show that higher geothermal gradients due to the proximity to the ring of fire affects CO2 capacity negatively and CO2 injectivity positively. Building on this lesson, we propose a graphical approach which provides improved insights when screening depleted reservoirs for CO2 storage; the approach combines both density and viscosity data to screen reservoirs for capacity and injectivity. Next, challenges surrounding the dynamic modelling of CO2 injection into depleted reservoirs are discussed. We compare the use of large and small scale models, and their implications on simulation accuracy and well placement optimization. We also discuss simulation stopping techniques, and how the incorrect application of a stopping criterion may result in capacity overestimation and risk of CO2 leakage. Finally, we demonstrate the need for careful engineering at injection sites to accommodate complicated CO2 phase transitions. Two approaches were compared – injecting CO2 in liquid or supercritical state versus injecting CO2 in gas state. The trade-offs between both approaches suggested that neither options should be ruled out by default, and that both options should be evaluated against project specific constraints.