NATURAL EARTHQUAKES AS PROXIES FOR INDUCED SEISMIC HAZARD AND RISK: COMPARING PEAK AND CYCLIC INELASTIC RESPONSE

Over the last decade, human-induced earthquakes have become more common in the world due to the increased number of operations involving hydraulic fracturing (fracking) of tight hydrocarbon reservoirs, geothermal energy, and wastewater disposal. Although such events rarely exceed magnitude 5, some instances of damage have been noted. Even the smaller events pose risks, as they can cause concern in the local communities, lead to high insurance claims and damage the reputation of operating companies, especially in areas unaccustomed to seismicity.<br/><br/>Some evidence suggests that induced earthquakes cause ground motions that are different from those of natural ones: anthropogenic tremors are shallower and appear to generate smaller stress drops. Because of their shallow depths, induced earthquakes can cause amplitude motions in the immediate vicinity of the epicenter that are larger than those expected from tectonic quakes. Such shaking mainly affects low-rise buildings. There has not been any systematic study, yet, on how the above characteristics of human-induced earthquakes affect the structural responses of buildings and whether patterns of differences exist compared to corresponding natural response that could be used for improved induced seismic risk assessment.<br/><br/>This study presents a framework for assessing the aforementioned differences. First, three datasets of natural, fracking and coal mining earthquake records are collected. Ground motion characteristics are compared. Then, peak and cyclic elastic and inelastic response parameters for different Single-Degree-of-Freedom Systems (SDOFs) are statistically compared through hypothesis testing, setting low magnitude natural earthquakes in the United Kingdom (UK) as the benchmark. Comparison is made in terms of normalized response measure such as the inelastic deformation ratio (CR) and the equivalent number of cycles (Ne). Results suggest a preliminary conclusion that human-induced events exhibit a trend of higher cyclic normalized inelastic response with respect to natural ones, at least for selected waveforms.<br/><br/>The implications of identifying and quantifying such differences in structural response between natural and human-induced earthquakes is relevant to design and retrofit of buildings in non-seismic prone areas, as well as to damage predictions for insurance purposes. Conclusions may be of interest to UK regulatory bodies in the preparation of recommendation documents for areas potentially affected by human-induced earthquakes.