Electrical conductivity of H[formula omitted]O-NaCl fluids under supercritical geothermal conditions and implications for deep conductors observed by the magnetotelluric method

Abstract

Magnetotelluric (MT) surveys have revealed the existence of subvertical conductors at depths of several kilometers in the volcanic and geothermal areas of northern Japan. The conductive anomalies suggest that saline magmatic fluids are trapped within or in the vicinity of granitic intrusions and potentially form supercritical geothermal reservoirs. Detailed interpretation of the observed conductivity is a challenging task, as estimating the contributions of the pore fluids to the bulk conductivity of the systems requires specific information. Such information includes the in situ pressure, temperature, and salinity conditions in the conductors.In this study, we modeled magmatic fluids as H2O-NaCl fluids and developed a new approach to estimate the electrical conductivity of the fluids at elevated temperatures up to 525 °C and salinity up to 25 wt% NaCl. Using our developed approach, the possible ranges of fluid electrical conductivity were calculated under supercritical geothermal conditions. Further, we examined the pressure, temperature, and salinity conditions required to explain these observations.Our study showed that H2O-NaCl fluids in the vapor and halite coexistence states likely have extremely low conductivity that could not explain the observed conductors. This finding indicated that relatively shallow conductors located above 4 km could have abnormally high pressures if the phase relations of in situ fluids were close to those of the H2O-NaCl fluids. Salinity of more than 0.5 wt% NaCl is necessary for single-phase fluids to account for such observation. Whether the fluids in the liquid and vapor coexistence state could have such high conductivity is currently uncertain owing to a lack of experimental data. Further, our predictions about the depth variation of fluid conductivity indicated that near-lithostatic pressures, or extremely high temperatures (above 550 °C) under sufficiently high pressures are required to reproduce the observed nearly uniform distribution with depth. However, it is currently difficult to rule out the possibility that a shallower part of the conductors is cooled to approximately 400 °C, as MT inversion could have difficulty in accurately capturing a small variation in conductivity within the conductors.