Are reservoir earthquakes man-made?

Abstract

A debate on whether the Wenchuan Mw7.9 earthquake is triggered by the nearby Zipingpu reservoir has drawn the attention of both the scientific community and the general public. High performance computation provides a powerful new tool to quantitatively evaluate stresses produced by the weight of impoundment of reservoirs and the changes of pore pressure due to water diffusion along faults to the hypocenter. We calculated Coulomb stress changes of a number of reservoirs: including the hotly debated Zipingpu reservoir, the well known reservoir earthquake of Xinfengjiang in 1962, and the reservoir earthquake of Aswan in 1981. We have reached several main conclusions: Elastic energy increase due to the weight of impounded water is usually very small in comparison with seismic wave energy released by earthquakes of magnitude 6 or greater, these reservoir earthquakes are improper to be called man-made, but they are human activity triggered. Geological background and tectonic stresses control the occurrence of such large reservoir earthquakes. Stresses due to weight of impoundment may promote or prevent the occurrence of reservoir earthquakes determined by the location and nature of the earthquake fault. Pore pressure increase due to water penetration along permissive fault, however, always increases the risk of reservoir earthquakes. There seems no definite threshold value of Coulomb stress to trigger earthquake, it may varies from several kPa to 0.1MPa, depending on the magnitude of tectonic stresses and strength of the fault. The occurrence time of reservoir earthquakes after impoundment also varies from months to years depending on the permeability and stress of the fault zone. Geological surveys and numerical simulations may improve risk estimates before reservoir construction.