Passive-seismic sensor placement optimization for geologic carbon storage

Abstract

The implementation of passive-seismic monitoring is essential for the geological carbon storage projects. For secure and continuous observing of induced passive-seismic events by CO2 injecting, a ground geophone network would be required. By determining the ideal number of seismic stations within a regular network, recent research has enhanced monitoring capabilities while also accommodating budgetary limitations. The primary objective of the optimal placement strategy for the surface geophone network is to create a cost-effective monitoring system. Given the cost limitations, a restricted quantity of sensors can be deployed at each location to maximize monitoring performance. To address this challenge, our approach involves the P-median stochastic programming formulation. The formulation aims to minimize the expected value of a monitoring target metric, which ultimately results in the optimal placement of detectors exhibiting superior expected behaviors. Our methodology is designed to choose the arrangement of detectors with best performance to improve both the early alarm detection time and localization capabilities of the sensor network. We utilize site-specific passive-seismic scenarios that capture the uncertainty in the characteristics of carbon leakage. The sensor grids provided by the optimization method invariably improve the capacity to detect passive-seismic events in comparison to sensors positioned in a regular grid configuration. We test the effectiveness of our approach on synthetic data based on a carbon storage site named Kimberlina. In conclusion, our approach provides a cost-effective solution for the optimal placement of sensors to achieve superior monitoring performance for the detection of passive-seismic events. The incorporation of site-specific scenarios with stochastic uncertainty coverage allows for more accurate and reliable results, leading to better decision-making for long-term geological carbon storage.