Electrical resistivity at the Travale geothermal field (Italy)

Abstract

Electromagnetic methods have been tested in the Travale geothermal field (Italy), a high-enthalpy field where the geothermal reservoir is located at a depth of 2-4 km in metamorphic and magmatic rocks, characterized by a high degree of heterogeneity and anisotropy and by the presence of superheated steam.
In order to determine whether magnetotelluric method is capable to define different reservoir features, and in particular a steam-dominated system, magnetotelluric (MT) data were collected during different surveys. After detailed robust analysis of these very noisy data, we have carried out 1D, 2D and 3D forward and inverse modeling in order to establish the resistivity data distribution at depth, its robustness and the sensitivity of MT data to different features. Resistivity log data, zones of high temperatures and productive fractures from drillholes measurements, a bright-spot type horizon identified in seismic profiles and estimated resistivity values of the lower crust and mantle were considered. The results have shown at the depth of the exploited geothermal reservoir a significant reduction in electrical resistivity whose large volume embrace but cannot be restricted to productive fractures from drillholes measurements and the bright-spot type horizon.
The observed reduction in resistivity may be interpreted taking into account the lithology and heterogeneities of reservoir rocks, their alteration and/or to the presence of a liquid phase within a fracture net sufficiently interconnected to produce electrolytic conduction. Since this last hypothesis is in disagreement with the known physical characteristics of the Travale geothermal fluid ( superheated dry steam), a study on cores and cuttings was carried out in order to identify the types and abundance of primary and alteration minerals and to compare these data with resistivity values. An inverse correlation between resistivity of the encountered formation and phyllosilicates amounts has been observed. However, the quantity of phyllosilicates and their effect on resistivity may account only for the small variation observed in the resistivity logs, but not for the main resistivity anomaly defined by magnetotelluric data in the area. We conclude that lithology and conductive minerals appears to be not the main cause of resistivity reduction, and that the natural state of fluids in the microfractures remains an important matter that needs further investigation.