Uncertainty analyses of CO2 plume expansion subsequent to wellbore CO2 leakage into aquifers

Abstract This study focused on CO 2 plume expansion subsequent to wellbore CO 2 leakage into a shallow unconfined aquifer post-CO 2 injection. The target response variables included CO 2 plume size, as well as flux to the atmosphere. Many processes contribute to CO 2 plume expansion in the aquifer; here we considered process and model parameters including those affecting the abandoned well leak rate, aquifer hydraulic properties, and aquifer geochemistry. In order to identify the significant factors affecting leakage, we adopted an uncertainty quantification framework to quantify input uncertainty, generate exploratory samples effectively, perform scalable numerical simulations, visualize output uncertainty, and evaluate input–output relationships. We combined quasi-Monte Carlo and adaptive sampling approaches to reduce the number of forward calculations while fully exploring the input parameter space and quantifying the output uncertainty. The CO 2 migration was simulated with STOMP-CO2 (water–salt–CO 2 module). Response surfaces of model outputs were built with respect to input parameters to determine the individual and combined effects. Four most significant parameters were identified to be dominating the CO 2 plume expansion process subsequent to wellbore CO 2 leakage: distance between the leaky and injection wells, maximum leakage rate, porosity, and hydraulic conductivity.