Maximum Magnitude Forecast in Hydraulic Stimulation Based on Clustering and Size Distribution of Early Microseismicity

作者：Moein Mohammad Javad Afshari; Tormann Thessa; Valley Benoit; Wiemer Stefan

来源：Geophysical Research Letters, 2018, 45(14): 6907-6917.

DOI：10.1029/2018GL077609

We interpreted the spatial clustering and size distribution of induced microseismicity observed during the stimulation of an enhanced geothermal system beneath Basel by comparison with scale-invariant synthetic data derived from discrete fracture network models. We evaluated microseimic specific influential factors including the effect of hypocentral location uncertainties, existence of a fractured zone and repeating events on the observed spatial organization. Using a dual power-law model originally developed in the context of discrete fracture network modeling, we developed theoretically the relationships among spatial clustering and magnitude distributions. We applied this model to the Basel data set and showed that the spatial clustering characteristics presented stationary properties during the hydraulic stimulation. Based on this observation, we proposed a statistical seismicity model calibrated on the scaling of early stimulation spatial patterns that is capable of forecasting the maximum magnitude of induced events with increasing injection time and stimulated volume.
Plain Language Summary Developing enhanced geothermal systems requires permeability enhancement by hydraulic stimulation, in which pressurized fluid is circulated between injection and production wells. This operation induces microseismicity that may be large enough to be felt by public and result in destructive events. Here we studied the possibility to forecast the maximum magnitude of induced events using the early seismicity patterns during hydraulic stimulation operations. Therefore, we analyzed the spatial clustering and size distribution of induced microseismicity observed during the stimulation of an enhanced geothermal system beneath Basel by comparison with synthetic data derived from fracture network models. We proposed a statistical model, originally developed in the context of discrete fracture network modeling, to represent the clustering and magnitude distribution of induced events. We calibrated the model using the scaling of early stimulation spatial patterns on Basel data set and successfully reproduced the rupture radius distribution, which was also capable of forecasting the maximum magnitude of induced events with increasing injection time and stimulated volume.

• 出版日期2018-7-28