Estimation of response of fracture system to hydraulic stimulation by induced microseismic multiplet analysis

Development of a commercial geothermal power plant using an artificially stimulated reservoir is underway at Cooper Basin, Australia. In 2003, they have made the first stimulation where openhole section from 3667-4421m inside granitic basement was stimulated. Researchers in Tohoku University have collected more than 32,000 microseismic events while the stimulation using a monitoring network consists of 4 near surface stations (horizontal offset around 5-7 km from the injection well) and one downhole station near the injection well. Sub-horizontal hypocentral cloud with a thickness around 100-200m have been obtained by JHD in the previous studies (Asanuma et al., SEG Exp. Abst., 2005). In this study, the authors clustered and re-located microseismic multiplets using coherence evaluation in the frequency domain and DD relocation technique. Evaluation of coherency of the seismic traces at the onset of P-wave revealed that 99.9% of the events, which are not saturated and overlapped two events, were clustered into one multiplet cluster. This is because of strong low-pass characteristics of the earth transfer function and simple rupture process. The events were then sub-clustered into 26 groups by their polarity of the first break of P wave at each station. Integrated interpretation of the multiplets using fault plane solutions (FPS) by the composite focal mechanism, spatio-temporal distribution of the hypocenter, and critical stress state for shear slip was made in this study. Results from the analyses revealed that most of the multiplets occurred on sub-horizontal fractures, and some of the multiplet events (a few percent) are very likely to origin from sub-vertical fractures which connect the sub-horizontal fractures. It has been also showed that the events from sub-vertical fracture started to occur after seismic activity in one sub-horizontal fracture became high, and other sub-horizontal fracture was seismically activated after appearance of sub-vertical seismic structure. Kumano et al. (SEG Exp. Abst., 2006) have made coherence analysis of multiplets including coda, and reported that two or three sub-parallel and sub-horizontal fractures were mainly stimulated at this site. It has been reported that several sub-horizontal permeable fractures have been found inside the basement rock while drilling of the injection well, and most of the pre-existing fractures were plugged by cement except for one. Integrated interpretation of the observed facts and results from the microseismic analysis is that (a) an existing fracture connected to the injection well was firstly stimulated, then (b) increased pore-pressure in the fracture induced shear slip on sub-vertical fractures and permeability improved, and (c) horizontal fractures, which were hydraulically connected by the sub-vertical fractures, were stimulated and seismically activated. A combination of multiplet analysis, FPS, and critical pore pressure was effectively used to interpret behavior of simulated fracture system.