Title: Mitigation and optimization of induced seismicity using physics-based forecasting

Abstract: Fluid injection can induce seismicity by altering stresses on pre-existing faults. Here, we investigate minimizing induced seismic hazard by optimizing injection operations in a physics-based forecasting framework. We built a 3D finite element model of the poroelastic crust for the Raton Basin, Central US, and used it to estimate time dependent Coulomb stress changes due to ~25 years of wastewater injection in the region. Our finite element model is complemented by a statistical analysis of the seismogenic index (SI), a proxy for critically stressed faults affected by variations in the pore pressure. Forecasts of seismicity rate from our hybrid physics-based statistical model suggest that induced seismicity in the Raton Basin, from 2001 - 2022, is still driven by wastewater injection. Our model suggests that pore pressure diffusion is the dominant cause of Coulomb stress changes at seismogenic depth, with poroelastic stress changes contributing about 5% to the driving force. Linear programming optimization for the Raton Basin reveals that it is feasible to reduce seismic hazard for a given amount of injected fluid (safety objective) or maximize fluid injection for a prescribed seismic hazard (economic objective). The optimization tends to spread out high-rate injectors and shift them to regions of lower SI. The framework has practical importance as a tool to manage injection rate per unit field area to reduce induced seismic hazard. Our optimization framework is both flexible and adaptable to mitigate induced seismic hazard in other regions and for other types of subsurface fluid injection.