CSEM and Seismic Monitoring Studies of Deep Low-Enthalpy Reservoirs

Abstract

Achieving net zero in greenhouse gases emissions attributed to human activities depends on the transition to renewable energy resources. Low-enthalpy geothermal systems, characterized by their widespread geographical distribution and suitability for direct heating applications, represent a promising alternative to fossil fuels. However, maintaining the long-term efficiency and economic viability of such geothermal reservoirs requires the development of new methods to monitor subtle variations in their properties, particularly those induced by temperature changes during energy extraction and reinjection.
This thesis evaluates the feasibility and advances the methodology of two key geophysical approaches for reservoir monitoring: the controlled-source electromagnetic (CSEM) method and the full waveform inversion (FWI) of seismic data. The research is grounded in two study areas: the Delft campus geothermal project in the Netherlands and the Munich geothermal project in Germany.
A feasibility study of CSEM monitoring was carried out on the Delft site to assess its sensitivity to subtle resistivity variations corresponding to temperature changes in the reservoir. Surface-to-borehole CSEM survey configuration was modeled to optimize source frequency and offset, with results demonstrating the detectability of a 4 Ω・m resistivity increase calculated for a 25 ◦C temperature drop in the Delft Sandstone reservoir. The study systematically analyzed the impacts of environmental disturbances—random noise, repeatability errors, seasonal near-surface temperature fluctuations, and the presence of steel-cased wells—on the performance of CSEM monitoring data. It was shown that a careful survey design and adequate source parameters allow CSEM monitoring, which is robust against most undesired effects, although steel casings require careful consideration due to their strong field attenuation within a radius of 100 m for a frequency of 1 Hz.
For high-resolution seismic characterization, the thesis develops and validates a novel sequential FWI approach for reconstructing high-resolution models of P-wave velocity and impedance from vertical seismic profiling (VSP) data. The method incorporates traveltime tomography for starting models and introduces a temporal phase resemblance step to improve convergence and mitigate phase error propagation in impedance inversion. Inversion experiments of synthetic data demonstrate that this approach enables the detection of impedance variations greater than 2 %, directly linked to temperature-driven reservoir changes. Field application to baseline VSP data at the Munich geothermal site confirms the robustness of the approach. A comparative analysis of distributed acoustic sensing (DAS) and conventional geophone-based FWI of P-wave velocity further elucidates the operational benefits and challenges of fiber-optic deployments inside the casing for characterization of geothermal reservoirs.
The results presented in this thesis establish CSEM and advanced seismic FWI as promising and complementary tools for noninvasive monitoring of low-enthalpy geothermal reservoirs. The work concludes with a discussion of current limitations, practical considerations for field deployment, and recommendations for future research.