Seismic While Drilling as a Passive Source for Imaging Deep Geothermal Systems

Deep high-enthalpy geothermal systems are typically developed in geologically complex settings characterized by strong structural heterogeneity and crystalline reservoirs. In such environments, conventional subsurface imaging methods are severely limited. High-resolution geophysical techniques that perform well in sedimentary basins, such as active surface seismic reflection, are often impractical or ineffective and fail to provide reliable images of deep geological structures. In addition, acquisition costs become prohibitive, particularly for three-dimensional surveys. 

These limitations can be partly overcome by passive seismic imaging approaches, including ambient-noise tomography based on surface waves and earthquake-based passive seismic tomography. These methods have demonstrated their operational robustness in complex geological contexts and at depths beyond the reach of conventional active techniques. However, although generally reliable, their spatial resolution remains limited and typically degrades with depth. 

At the drilling stage of deep geothermal projects, improved subsurface characterization is essential to reduce geological uncertainty, support accurate well trajectory planning, and mitigate drilling risks. Enhancing the resolution and relevance of passive seismic imaging in the vicinity of the borehole therefore represents a key methodological challenge for geothermal exploration and development. 

In this contribution, we present results from a passive seismic acquisition conducted during drilling in a deep high-enthalpy geothermal field in southern Tuscany (Italy). The study investigates the potential of exploiting seismic energy generated by the drill bit (Seismic While Drilling, SWD) as an additional method to complement and enhance subsurface imaging. Although SWD is not a new concept, only a limited number of studies have investigated its application at such depths and in geologically complex crystalline environments. 

For this experiment, a total of 65 seismic nodes, including both single-component and three-component sensors, were deployed around the drilling site, with rig–receiver offsets ranging from 150 m to 1700 m. Continuous recordings were acquired over a 10-day period at a sampling interval of 2 ms, during which drilling progressed from 3,200 m to 3,700 m depth. 

Data processing followed workflows commonly used in ambient-noise tomography. However, the drilling operations generated strong surface waves that required specific processing strategies. Several beamforming and wavefield-separation approaches were therefore applied to suppress surface-wave energy and enhance body-wave signals associated with the drill bit. 

Preliminary results show that body waves generated by the drill bit at depths between 3,200 m and 3,700 m are clearly recorded by surface sensors. These observations enable the extraction of detailed P-wave velocity information, providing higher-resolution constraints that complement other passive geophysical surveys such as ambient-noise tomography.