Abstract

Ambient noise correlation techniques are of growing interest for imaging and monitoring deep geothermal reservoirs. They are simple to implement and can be performed continuously to follow the evolution of the reservoir at low cost. However, these methods rely on assumptions of spatial and temporal uniformity of seismic noise sources. Violating them can result in misinterpretation of seismic velocities owing to preferential noise propagation directions. Using several years of seismic data recorded around the two geothermal sites of Soultz-sous-forets and Rittershoffen in northern Alsace, France, we propose a detailed characterization of the spatial and temporal properties of the high frequency seismic noise (0.2 to 5Hz). We consider two fundamental properties of the cross correlation functions (CCFs) of ambient noise. Firstly, the reliability of the Green's function reconstruction, an important indicator for tomographic studies. Secondly, the temporal repeatability of the CCFs between 0.2 and 0.5 seconds. At periods below 1s, we observe a sharp decrease in signal to noise ratio resulting from the non uniform distribution of anthropogenic sources. At periods above 1s, we show that the high directivity of the northern Atlantic microseismic peak biases the CCFs' phase significantly. We show that nocturnal noise is the most suited for temporal analysis of the CCFs. Using nocturnal noise, we should be able to monitor temporal variations induced by the geothermal activities inside the reservoir.

Highlights

  • Ambient noise correlation techniques are of growing interest for imaging and monitoring deep geothermal reservoirs

  • We band-pass filter the cross correlation functions (CCFs) in three period ranges and represent them according to the inter-station spacing

  • The increased symmetry indicates that, in this period range, noise sources are more uniformly distributed around the network or that the noise energy is efficiently distributed through the medium via scattering

Read more

Summary

Introduction

Ambient noise correlation techniques are of growing interest for imaging and monitoring deep geothermal reservoirs They are simple to implement and can be performed continuously to follow the evolution of the reservoir at low cost. Active seismic sounding is a commonly used approach; its dense spatial and temporal sampling provides high-resolution images of the reflectivity of the subsurface layers and of fault geometry Such seismic data are not readily available everywhere, and acquisition of new data, especially in 3D, is often very expensive compared to the profitability of geothermal resources. Images of the upper crustal structure can be obtained from tomographic inversion of arrival times of natural or induced local earthquakes These inversions can be repeated over time to map velocity changes (e.g., Calò et al 2011; Calò and Dorbath 2013) much more cheaply than active seismic sounding. The need to minimize seismic risk excludes using time-lapse arrival-time tomography for continuous reservoir imaging

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call