Optimization of Well Placement for Geological CO2 Storage to Minimize Fault Slip Tendency

Abstract Geological carbon storage is an important technology for reducing CO2 emissions to the atmosphere. Storage operations can be improved by using computational optimization to maximize effectiveness and minimize risk. In this work, we apply computational optimization to determine the placement of CO2 injection wells such that a metric that quantifies the risk of induced seismicity is minimized. The simulation model includes a heterogeneous storage aquifer along with a large surrounding region, caprock, and basement rock. The storage aquifer contains two major faults. Fault slippage and seismicity can occur if the injection operation alters the stress state such that the fault slip tendency (FST) exceeds a particular value. The objective of the optimizations in this study is the minimization of the maximum FST observed along either fault in the model. The constraints imposed on the problem ensure that wells are spaced more than 1 km apart, that the target amount of CO2 is injected, and that all injected CO2 remains within the storage aquifer. The core optimizer is a differential evolution algorithm, and the simulations required for function and constraint evaluations are accomplished using the coupled flow-geomechanics simulator GEOS. Geometric constraints are handled using a repair procedure, and nonlinear output constraints are treated with a filter method. The setup involves three injection wells, each injecting 1.5 Mt CO2/year for 50 years. The optimization framework is shown to consistently provide well locations that lead to feasible solutions with maximum FST values that are less than those achieved in a set of heuristic cases. Although FST increases with time, the maximum value after 50 years in the optimized case corresponds to small likelihood for fault slip.