Estimation of the Cumulative Fault Area Under Critical State Based on Microseismic Dataset at Hydraulic Stimulation for Seismic Risk Assessment

Occurrence of induced seismicity with large magnitude has been considered as critical environmental burden associated with fluid injection for shale gas/oil extraction, waste water disposal, and creation of engineered geothermal systems (EGS). Therefore, studies for understandings of the large events and risk assessment of induced seismicity has been activated (e.g. Ellsworth, 2013 or McGarr, 2014). Many of the risk assessment models is based on the seismological statistics and they use only basic information of the occurrence time and event magnitude.

In this study, we originally developed the risk assessment model based on the more physical interpretation of microseismic events. We newly defined new index reffered to as “Slip-able area (SAA)”, which describes the cumulative fault areas under critical state in a study area during/after the stimulation. The informations used in estimating SAA were given by the analysis of microseismic events. This model can provide the potentially slip-able area in semi real time nature. SAA can be directly converted into the event magnitude, therefore this model can contribute to the short (one or two days) or middle (week) term the risk assessment of the large seismic event.

SAA is defined as simple product of slipped area density and stimulated volume using the equation (1).

A_slip=K×V_stimu (1)

where A_slip is the cumulative SAA, K is slipped area density, and V_stimuis the stimulated volume by fluid injection.

K, slipped area density is defined as the summation of fault areas of the seismic events which occurred during given time step from given volume. V_stimu means the rock volume whose permeability was enhanced by the fluid stimulation. So, V_stimuis estimated by distribution of the hypocenter location.

In order to verify the feasibility of this model, SAA is estimated using the microseismic data set from Basel, Switzerland and compared the observed seismic fault area. In this process, slipped area density was calculated using 40m grid of cube for each time step (1day) and we selected the grid where the number of the seismic events is maximum. V_stimu were estimated by summation of 25m cubic cells, which contain at least one seismic event. The estimates of the SAA showed the good agreement with the observational fault area.