A strategy for choosing red-light thresholds to manage hydraulic fracturing induced seismicity in North America

Induced earthquakes caused by hydraulic fracturing (HF) are a growing concern, with risks that are often managed by traffic light protocols (TLPs). Here, we apply a risk-informed strategy for choosing TLP red-light thresholds. We utilize a combination of probabilistic maximum magnitudes, formation depth, site amplification, ground motion relationships, felt/damaging shaking tolerances, and population information to simulate the spatial distribution of nuisance and damage impacts. We apply this approach to many of the prominent North American shale plays that have caused earthquakes: the Horn River Basin, the Duvernay Formation, the Montney Formation, the Utica Shale, the Marcellus Shale, the SCOOP & STACK play in Oklahoma, the Eagle Ford Formation, and the Delaware Basin. We find that induced earthquake impacts are spatially heterogeneous, depending most strongly on population density. These spatial heterogeneities vary on different length scales, depending if they are related to nuisance (longer range, 100s kms) or damage (shorter range, 10s kms). Because of this variability, we suggest an approach that chooses red-light magnitudes that are based on tolerances to nuisance and damage risks (varying between Mw 2.0-5.0). Comparison of the results between North American shale plays, in particular those that already have TLPs enacted, suggests that nuisance risks have been an influencing factor in determining red-light thresholds. Our method provides a quantitative guideline for traffic light protocols designed in a risk-informed manner that retains the implementational simplicity of magnitude-based thresholds. In addition, our results provide a benchmark for jurisdictional nuisance tolerances that future TLPs could be guided by.