212 Geophysics in Geothermal Exploration Another edifying study is the work by Sanchez-Pastor et al. (2021), who present an ANSI based method to monitor the Hengill geothermal field (Iceland). This site presents a challenge inherent to geothermal systems operations which is to probe the steam fluctuation inside the reservoir. During geothermal exploitation, the estimation of the steam content is key from both operational and economical perspectives. To quantify these quantities, the authors used an array of 50 stations and compute auto-correlation (ACs) on the vertical component. The nuance of the velocity variation measurement on ACs compared to cross-correlation between two sensors is essentially that the reconstructed wavefield in ACs will be much more sensitive to volume waves, and mainly to P waves. This means that the velocity variation measurement will be correlated with the fluid content, and not anti-correlated as is classically observed in studies targeting ground water table using surface wave (for example). The velocity variations measurements obtained on the ACs are compared with some rock physics model including hydrological and gas saturation information. The results of the various modelling work performed by the authors and the final comparison with the observations is shown in Figure 6.18. The subfigure (a) presents the pressure and temperature variation, the subfigure (b) the estimated steam cap evolution, the subfigure (c) the modelized Vp and Vs evolution over the monitoring time. Finally, the subfigure (d) shows the monitored dv/v versus the subsidence of the geothermal field. The author demonstrates than the seismic velocity values are decreasing over the monitoring time, when the steam ratio continues to rise which is consistent with the expected variation in Vp as water content decreases. This study highlights the economic possibilities to monitor the steam evolution during long periods with a low-cost method associated with robust modelling. The examples previously presented demonstrate than the measure of seismic velocities through ANSI-based approaches is feasible and is a useful tool to monitor changes in the state of stress in geothermal reservoirs (Taira et al., 2018; Muñoz-Burbano et al., 2024) or changes of fluid distribution (Sanchez-Pastor et al., 2021). Nevertheless, as for ANSI-based tomography, ANSI-based monitoring techniques also start evolving toward the study of other seismic attributes than seismic velocity. The work by Obermann et al. (2015) is an example of such evolution. The authors propose to survey the geothermal reservoir using another attribute called the decoherence of the reconstructed waveforms. Initially the decoherence is an indicator of the quality of the reconstructed signal between a reference waveform (e.g. at the beginning of the monitoring period) and each waveform reconstructed at a later time and on first order is affected by site-dependent noise conditions that change too abruptly. In the study by Obermann et al. (2015) though, a strong decoherence is observed that seems associated with a gas kick event which led to the failure of the St Gallen deep geothermal project. Figure 6.19 illustrates the evolution of decoherency time-series for multiple station pairs. For all the seismic stations pairs crossing the reservoir, a strong decoherence is observed between the 10th of July and the 14th of August, which period correlates with the injection
RkJQdWJsaXNoZXIy NjA3NzQ=