In the Okuaizu geothermal field, we carried out continuous self-potential (SP) measurements in 2000, 2002 and 2004 to monitor field-wide shut-ins associated with regularly-scheduled maintenance of the Yanaizu-Nishiyama geothermal power plant. Thirteen stations for SP measurements were distributed in the central area of the Okuaizu field, at each of which two or three non-polarizing electrodes were set at one meter depth to check the station reliability. SP changes of 5 to 10 mV associated with the shut-in and re-start of production wells were detected at a few stations located in a low resistivity area in the vicinity of the Sarukurazawa fault. In order to evaluate the observed SP variations, numerical simulations of changes in reservoir conditions due to fluid production and successive “SP-postprocessor” calculations were performed based upon a simplified reservoir model. The reservoir region of the model is divided into deeper and shallower parts, both of which are represented as MINC double-porosity media. In the final model, the matrix permeability of the deeper reservoir needs to be smaller than 10-20 m2 to reproduce the rapid increase of the produced fluid enthalpy as observed in the first a few years of the power plant operation. In contrast, it is required to assign the matrix permeability to be 10-18 m2 or more for the shallower reservoir to reproduce the observed SP changes associated with the short-term shut-ins. This combination of deeper and shallower reservoir properties also reproduces long-term SP changes revealed by repeated SP measurements in 1996 and 1998, and long-wavelength micro-gravity decline between 1997 and 2000. In conclusion, continuous SP measurement to detect short-term SP changes such as associated with field-wide shut-ins is thought to provide additional useful constraints against which reservoir models can be calibrated.
Based on the empirical formula proposed by the authors for estimating the hydraulic aperture of a single fracture, the hydraulic conductivity of a fracture with an arbitrary direction under rock stress and pore pres-sure at depths of 500 m, 1000 m and 2000 m was estimated for an arbitrary direction of fluid pressure gradi-ent. The parameters necessary for determining the hydraulic aperture and the mechanical properties of the fracture were determined from previous experimental and analytical results for a tensile fracture created in Inada granite in the laboratory. The mean aperture of the fracture is mainly governed by shear dilation, and there is a ridge of the local maxima in the mean aperture in the stereographic projection of the normal direc-tion of the fracture, which is parallel to that in the shear stress but inclined toward the minimum principal stress. The mean aperture is maximum for the fracture with the normal direction that lies in the maximum principal stress - the minimum principal stress (σ1 - σ3) plane with an inclination angle of 20° to 33° from σ3 for all stress conditions used in this study. The fracture with the maximum hydraulic aperture for a given di-rection of the fluid pressure gradient has a normal direction at which the angle between the macroscopic wa-ter flow and the shear displacement is maximum nearby the ridge of the local maxima in the mean aperture. Thus, the normal direction of the fracture with a large hydraulic aperture is distributed around that of the fracture with the maximum mean aperture and is extended along the ridge of the local maxima in the mean aperture. The maximum hydraulic aperture is produced in the fracture with the maximum mean aperture when the pressure gradient is given in the direction of the intermediate principal stress (σ2), and this fracture has the normal direction within the σ1 - σ3 plane with an inclination angle of 6° to 13° from the normal direc-tion of the fracture with the maximum ratio of the shear stress to the effective normal stress.