Optimal Sensor Network Design to Detect CO2 Leakage at a Geologic Storage Site Considering Three-Dimensional Aquifer Heterogeneity

Surface or near-surface environment can be seriously damaged when carbon dioxide leaks from its storage in geologic formations. For this reason, near-surface monitoring is necessary at a CO2 storage site to meet requirements of regulations, to reduce potential risk, and to obtain public acceptance. In this monitoring, it is necessary to design a monitoring network to detect CO2 leakage aiming effectiveness and low uncertainty. In general, this objective for loosely characterized (especially, for the overlying aquifer) site can be achieved in a stochastic way, usually via Monte Carlo simulations. However, the computational cost of existing three-dimensional multiphase flow simulation is prohibitively expensive considering the scale of the influence of hypothetical CO2 leaks. In the present study, a computationally cheap three-dimensional rule-based percolation model was applied to several scenarios with different aquifer heterogeneities. From the applications, an optimal CO2 monitoring network design per scenario that maximizes detectability with relatively low uncertainty was obtained. The ensemble results showed that adoption of near-surface permeability information, including the spatial correlation structure, may be beneficial for optimizing monitoring network design with respect to sensor-wise detectability of CO2 leakage monitoring at the surface level. In addition, given detection likelihood uncertainty across the entire sensor deployment, regular spacing deployment of sensors is considered the most relevant method when a sufficient number of sensors are available.