Local Earthquake Tomography at Los Humeros Geothermal Field (Mexico)

Abstract

A passive seismic experiment using 25 broadband and 20 short-period stations was conducted between September 2017 and September 2018 at Los Humeros geothermal field, an important natural laboratory for superhot geothermal systems in Mexico. From the recorded local seismicity, we derive a minimum 1-D velocity model and obtain 3-D Vp and Vp/Vs structures of Los Humeros. We improved the classical local earthquake tomography by using a postprocessing statistical approach. Several inversions were computed and averaged to reduce artifacts introduced by the model parametrization and to increase the resolution of the investigated region. Finally, the resulting Vp and Vp/Vs structures and associated seismicity were integrated with newly acquired geophysical and petrophysical data for comprehensive interpretation. The recorded seismicity is mainly grouped in three clusters, two of which seem directly related to exploitation activities. By combining new laboratory measurements and existing well data with our Vp model, we estimate possible geological unit boundaries. One large intrusion-like body in the Vp model, together with neighboring high Vp/Vs anomalies, hints at a region of active resurgence or uplift due to the intrusion of new magma at the northern portion of the geothermal field. We interpret high Vp/Vs features as fluid bearing regions potentially favorable for further geothermal exploitation. Deep reaching permeable faults cutting the reservoir unit could explain fluid flow from a deeper local heat source in the area.

Geophysical characterization of geothermal fields: Inversion of temperature data and ground motions

“…Additionally, we find an elastic half-space a reasonable approximation, since the geological units do not have significantly different Poisson's ratios. Vp/Vs ratios based on seismic tomography (Toledo et al, 2020) near the activated fault segment also do not show any significant variations within the LHVC sediments, supporting the applicability of elastic half-space for this study.…”

“…Lermo Samaniego et al (2016) used a 1D P-wave velocity model (Lermo et al, 2008) and first arrivals from five seismic stations. The velocity model of Lermo et al (2008) adopts significantly different values from from the results of more recent studies, especially at shallow depth (Löer et al, 2020;Toledo et al, 2020). Also, due to the largely heterogenous subsurface of the study area, the hypocentral depth might be biased as a result of differences in seismic velocities from the 1D model.…”

“…The occurrence of the earthquake at a maximum depth of 1 km would require the presence of stress-resistant material in shallow depth. Seismic tomography results show a significant positive Vp anomaly near the seismic event, delineating a relatively rigid body between ~500-1500 m depth (Toledo et al, 2020), that could build up stresses for a Mw=4.2-4.35 rupture. Toledo et al (2020) estimated the top of the pre-caldera group (hosting the geothermal reservoir) to ~600 m depth in the vicinity of the earthquake, which is significantly shallower than the average reservoir depth of approximately 1000 m. This would suggest that the reactivated fault segment inferred from our study extends also within the upper part of the reservoir (Figure 6.4).…”

“…Seismic tomography results show a significant positive Vp anomaly near the seismic event, delineating a relatively rigid body between ~500-1500 m depth (Toledo et al, 2020), that could build up stresses for a Mw=4.2-4.35 rupture. Toledo et al (2020) estimated the top of the pre-caldera group (hosting the geothermal reservoir) to ~600 m depth in the vicinity of the earthquake, which is significantly shallower than the average reservoir depth of approximately 1000 m. This would suggest that the reactivated fault segment inferred from our study extends also within the upper part of the reservoir (Figure 6.4). The very shallow fault reactivation inferred from the InSAR inversion could suggest that the earthquake hypocenter is located near the bottom of the rupture that is often observed (e.g.…”

“…2019), connected to the reactivated Los Humeros fault is plotted with black dashed line. The average extent of the reservoir inferred from well data (Carrasco-Núñez et al, 2017) and seismic tomography (Toledo et al, 2020) is shown with blue (the shading indicates the uncertainty of reservoir extent). The trace of the section is shown in Figure 6.3b.…”