In this study we examine the ability of statistical models to simulate the magnitudes of seismic events induced by hydraulic fracturing (HF). The motivation for doing so is that, if operators are able to forecast potential event magnitudes, then they may be able to discriminate at an early stage between the normal case, where HF does not induce large, felt seismic events, and the abnormal case, where larger events do occur. By doing so, an operator may be able to take mitigating actions for the abnormal cases in order to avoid induced seismicity. Here we examine two statistical models for induced seismic event magnitudes, and apply them to two case studies consisting of microseismic datasets collected during hydraulic fracturing in the Horn River Shale. At both sites, some stages showed evidence for fault reactivation and larger induced seismic events (MMAX > 1.0). We find that these statistical models are capable of identifying the fault-reactivation stages relatively early on, and provide reasonably accurate forecasts of the eventual largest events. Had such modelling been performed in real time, it could have formed the basis for a strategy to mitigate induced seismicity.
Presentation Date: Tuesday, October 16, 2018
Start Time: 1:50:00 PM
Location: 208A (Anaheim Convention Center)
Presentation Type: Oral