From Oil Field to Geothermal Battery: Derisking the DeepStor HT-ATES Project via Advanced Seismic Attributes and Machine Learning

The DeepStor project in the Upper Rhine Graben, southwest Germany, represents a pioneering initiative for high-temperature aquifer thermal energy storage (HT-ATES) within deep sedimentary reservoirs. This study presents an integrated geophysical and data-driven workflow for evaluating the possible DeepStor reservoirs and optimizing the location of the DeepStor-1 exploration well through advanced seismic interpretation of a pre-stack depth migrated (PreSDM) 3D seismic dataset. Structural and stratigraphic constraints characterize the geometries, fault development, and compartmentalization of reservoir formations within the Cenozoic sedimentary graben filling. Particular emphasis was placed on delineating porous sandstone successions as most promising target reservoirs for heat storage.
To enhance reservoir imaging beyond conventional seismic interpretation, spectral decomposition techniques were applied to identify subtle stratigraphic heterogeneities, channel architectures, and frequency-dependent reservoir signatures linked to lithological variations and fluid distribution. The spectral attributes significantly improved the delineation of reservoir continuity and lateral facies variability, thereby reducing geological uncertainties.
... mehr
Furthermore, machine learning algorithms were integrated with seismic attributes, spectral decomposition volumes, and geological constraints to perform reservoir characterization and predict spatial distributions of reservoir quality indicators. The combined workflow enabled the identification of structurally and laterally continuous reservoirs with favorable petrophysical conditions for HT-ATES deployment. The results demonstrate the effectiveness of integrating PreSDM seismic interpretation, spectral decomposition, and machine learning for geothermal reservoir evaluation and exploration well optimization in complex rift-related sedimentary systems.