Forecasting microseismic cloud shape as a proxy of stimulated rock volume is essential for the design of an energy extraction system. The microseismic cloud created during hydraulic stimulation is known empirically to extend in the maximum principal stress direction. However, this empirical relationship is often inconsistent with reported results, and the cloud growth process remains poorly understood. This study investigates microseismic cloud growth using data obtained from a hydraulic stimulation project in Basel, Switzerland, and explores its correlation with measured in-situ stress. We applied principal component analysis to a time series of microseismic distribution for macroscopic characterization of microseismic cloud growth in two- and three-dimensional space. The microseismic cloud in addition to extending in the direction of maximum principal stress expanded to the direction of intermediate principal stress too. The orientation of the least microseismic cloud growth was stable and almost identical to the minimum principal stress direction. Following the stimulation, the orientation of microseismic cloud growth was consistent with the in-situ stress direction. Further, microseismic cloud shape ratios showed good agreement when compared with in-situ stress magnitude. The permeability tensor estimated from microseismicity also presented good correlation in terms of direction and magnitude with microseismic cloud growth. We show that in-situ stress plays a dominant role by controlling the permeability of each existing fracture in the reservoir fracture system. Consequently, microseismic cloud growth can be scaled by in-situ stress if there is sufficient variation in the existing faults.

A wideband magnetotelluric survey was performed in the Sengan geothermal region of northeastern Japan. A conductor with resistivity of < 30 Ωm was found at a depth of -1.8 km in the Quaternary Kakkonda granite. Microseismic activity was not observed, suggesting a ductile zone with temperatures exceeding 370 °C. Under these conditions, H2O-NaCl fluids can exist as two-phase or single-phase types. The permeability of the upper reservoir surface is between 5.0E-17 and 5.0E-15 m2, indicative of an exploitable supercritical geothermal reservoir. Our results indicate that this is originated by heat and water supply system from deep magmatic fluids.

In geothermal development, microseismic monitoring is important technique to monitor the various phenomena in the reservoirs throughout location, activity, and magnitude of microseismicity. Picking P- and S- wave arrivals accurately from seismic data is inevitable process for subsequent seismic analyses. However, the manual phase picking is a time- and cost-consuming process and several automatic pickers still requires considerable quality checks and corrections by human analysts. Automatic pickers based on deep learning have recently been developed for natural earthquake analysis, whose accuracy has been confirmed to be comparable to that of human analysts. These phase pickers were mainly trained using natural earthquakes recorded by regional seismic networks. However, seismic networks and events in geothermal fields have features that differ from those of natural earthquakes. In such fields, seismic events with very low magnitudes occur immediately under the seismic network and are sometimes triggered by fluid activity. Therefore, the direct application of the existing deep learning phase pickers to such seismic networks may have difficulty. Here, we focus on developing a deep learning model specialized for local seismic networks in geothermal fields. We used microseismic data from four representative enhanced geothermal and hydrothermal fields and trained the model with deep learning. Based on the developed model, the hypocenter distribution was determined using continuous seismic waves in the Okuaizu geothermal field, Japan. These procedures were performed automatically without manual operations and we propose them as MAGMA: Machine learning Automatic picker for Geothermal Microseismicity Analysis. Subsurface fine structures were then revealed by relocating the hypocenters using a double-difference algorithm. The same procedures for the same data were then conducted using a deep learning model that trained by other field data, and the equivalent structures were successfully reveled. Thus, MAGMA is applicable to new fields even when data is lacking, such as green fields.