An improved framework for discriminating seismicity induced by industrial activities from natural earthquakes

Abstract

Heightened concerns regarding induced seismicity necessitate robust methods to assess whether detected earthquakes near industrial sites are natural or induced by the industrial activity. These assessments are required rapidly, which often precludes detailed modeling of fluid pressures and the geomechanical response of the reservoir and nearby faults. Simple question‐based assessment schemes in current use are a useful tool but suffer from several shortcomings: they do not specifically address questions regarding whether available evidence supports the case for natural seismicity; they give all questions equal weighting regardless of the relative influence of different factors; they are not formulated to account for ambiguous or uncertain evidence; and the final outcomes can be difficult to interpret. We propose a new framework that addresses these shortcomings by assigning numerical scores to each question, with positive values for answers that support induced seismicity and negative values for responses favoring natural seismicity. The score values available for each question reflect the relative importance of the different questions, and for each question the absolute value of the score is modulated according to the degree of uncertainty. The final outcome is a score, the induced assessment ratio, either positive or negative (or zero), that reflects whether events were induced or natural. A second score, the evidence strength ratio, is assigned that characterizes the strength of the available evidence, expressed as the ratio of the maximum score possible with the available evidence relative to the maximum score that could be obtained if all desired data were available at a site. We demonstrate this approach by application to two case studies in the United Kingdom, one widely regarded as a case of induced seismicity, and the other more likely to be a series of tectonic earthquakes.