Coupled thermal, hydraulic and geochemical processes in mine thermal energy storage at the Reiche Zeche underground mine (Freiberg, Germany)

Flooded and partially flooded mine workings are a promising but still under-quantified option for Underground Thermal Energy Storage (UTES), offering accessible volumes and well-constrained geometry for field-scale experimentation. We report a long-term Mine Thermal Energy Storage (MTES) demonstration in a fully instrumented test basin (≈20 m³) installed at ~147 m depth in the Reiche Zeche underground Geo-Lab (Freiberg, Germany). Three controlled heating–cooling cycles were operated over 504 days, combining dense thermometry in the surrounding gneiss, NaCl point-dilution tracer testing, hydrochemical monitoring, and in-situ heat exchanger fouling and material-performance assessment.

Across the three cycles, 38.0 MWh of heat was supplied. Basin temperatures reached ~26 °C in Cycles 1–2 and ~39 °C in Cycle 3. Wall-rock sensors recorded a delayed but persistent response, with the gneiss warming by 10.1 K at 1.8 m depth after the hottest cycle, consistent with a conduction-dominated regime and long-lived thermal memory. Energy-balance partitioning indicates that the surrounding rock mass stored ~90% of the injected energy, whereas the basin water primarily acted as a rapid heat carrier and exchanger interface.

Hydraulic exchange was quantified by conservative tracer decay, yielding a steady throughflow of ~79 L h⁻¹ (mean residence time ~10.5 days) and an advective heat-loss coefficient of 0.092 kW K⁻¹. This persistent throughflow represents the dominant loss pathway and explains the strong sensitivity of recoverability to hydraulic boundary control. Exchanger-based recovery metrics show a pair recovery fraction of ~0.53 for the actively discharged Cycle 2, while Cycle 3 exhibits multi-cycle conductive “memory” effects, with incremental recovery fractions reaching ~0.7 and a cumulative storage efficiency of ~0.56 over the full experiment.

Thermal cycling also induced pronounced hydrochemical and operational constraints. Warm phases triggered rapid Fe(II) oxidation and precipitation of Fe(III) oxyhydroxides, driving exchanger fouling; uncoated AISI 316L lost ~45% of initial conductance, whereas a hydrophobic coating limited losses to ~18% and a Fe-resistant alloy provided intermediate mitigation.

Overall, the dataset demonstrates reproducible MTES operation under mine conditions and identifies hydraulic isolation/throughflow reduction and oxygen control as the primary levers for improving MTES performance. The derived field metrics (advective-loss coefficient, conduction-driven storage depth response, and fouling resistance under acidic mine-water conditions) provide transferable guidance for designing and benchmarking MTES in post-mining UTES applications.