Supervison: J. Kinscher (Ing.. INERIS), E. Klein (Ing. INERIS), O. Lengliné (MCF UdS, EOST), J. Schmittbuhl (DR CNRS, EOST)
- INERIS, Campus ARTEM, 92 rue Sergent Blandan, BP 14234, F-54042 Nancy Cedex, France (main attachment)
- EOST, Université de Strasbourg/CNRS, 5 rue René Descartes, 67084 Strasbourg Cedex, France (working periods at EOST)
- Enhanced geothermal systems, induced seismicity, machine learning
Deep geothermal energy is a renewable and sustainable underground energy source in full development, particularly in geological contexts where it is necessary to artificially develop the deep reservoir to achieve economic profitability (EGS technology). It represents an important sector in the energy transition for a reduction in carbon and greenhouse gas emissions. Only a small part of the world's geothermal potential is currently being exploited. Many countries, including France, are aiming to strengthen the development of this technology in the coming years. However, there are obstacles. Indeed, like most underground industrial activities, geothermal exploitation can lead to risks for populations and the environment, which can in some cases even lead to the temporary or permanent termination of the project. One of the main difficulties is induced seismicity, the origin of which is not yet well understood, as demonstrated, for example, by the recent high magnitude induced event (M 5.4) in Pohang, South Korea in 2017. Micro-seismic monitoring not only of the largest events but also of the development of the smallest events through new detection and location techniques (e.g. template matching, cross-correlation of waveforms, double-difference, array coherency back projection, etc.), is one of the most promising approaches to controlling these related risks, and provides good practice guidelines for monitoring and regulatory developments that are often controversial.
Objective of the PhD
The scientific objective of this thesis is to understand the evolution of seismicity generated during the stimulation and operation of a deep geothermal reservoir by combining seismological analysis, forcing (injection) data, artificial intelligence tools and geological and geomechanical modelling. Through the approaches developed, tested and applied using the database of the Soultz-Sous-Forêts geothermal reservoir (Alsace), hosted by the EOST (https://cdgp.u-strasbg.fr/), it will follow an approach linking seismology and geomechanics to understand and predict the probable evolution of the structures identified within the reservoir and establish the basis of a predictive tool using machine learning techniques (e.g. TensorFlow). Results of this work may potentially help to improve rapid analysis and decision-making to face induced microseismic activity and thus revisit tools like "Traffic light systems".
Master in Geosciences/Geophysics/Physics
The candidate must be motivated by massive data processing, have good knowledge of seismology and geomechanics, have programming and signal processing skills and a good level of English. He/she will be able to share his/her time between INERIS in Nancy and EOST in Strasbourg as part of the collaboration between the two institutions.
How to apply
Cornet, F. H. (2016). Seismic and aseismic motions generated by fluid injections. Geomechanics for Energy and the Environment, 5, 42-54.
DeVries, P. M., Viégas, F., Wattenberg, M., & Meade, B. J. (2018). Deep learning of aftershock patterns following large earthquakes. Nature, 560(7720), 632.
Grigoli, F., Cesca, S., Priolo, E., Rinaldi, A. P., Clinton, J. F., Stabile, T. A., ... & Dahm, T. (2017). Current challenges in monitoring, discrimination, and management of induced seismicity related to underground industrial activities: A European perspective. Reviews of Geophysics, 55(2), 310-340.
Lengliné, O., Boubacar, M., & Schmittbuhl, J. (2017). Seismicity related to the hydraulic stimulation of GRT1, Rittershoffen, France. Geophysical Journal International, 208(3), 1704-1715.