Evaluating the long-term evolution of induced seismicity sequences with implications for induced seismicity hazard assessment

We examine the temporal evolution of induced seismicity at a compilation of over 20 individual cases where earthquakes have been caused by long term, low pressure subsurface fluid injection. These cases serve as useful proxies for future commercial-scale CCS projects, given the similarities in injection rates and volumes. We quantify the induced seismicity by comparing the temporal evolution of seismogenic index and seismic efficiency - these parameters relate the rates of seismic activity to the rates of fluid injection. We find that seismicity evolves through time: in general, rates initially accelerate and then stabilise, typically after a period of 1 - 3 years. After this, rates of seismicity generally decrease gradually. We use the seismogenic index and seismic efficiency parameters to make forecasts of induced earthquake magnitudes. We find statistically significant correlations between observed and modelled seismic magnitudes, showing that the models do provide useful forecasting information. However, in some cases the observed magnitudes significantly exceeded the forecast values, showing that larger magnitude events cannot entirely be precluded. Our findings for WWD-induced seismicity can be carried forward to inform induced seismicity hazard assessment, management and mitigation at future commercial-scale CCS projects.