A reduced numerical model for predicting temperature dynamics in flooded mine galleries under seasonal heat loads and storage conditions

Mining activities occur on all inhabited continents, which is why many countries are faced with the question of how exhausted mines can be utilised sustainably. One promising option is their use as energy sources, with flooded structures acting as sub-surface heat exchangers. The large water volumes offer significant potential for heating and cooling. However, the exploitation of abandoned mine is associated with high costs due to preliminary investigations and drilling. In order to minimise the economic risk, a reliable forecast of the mine water temperature is essential to ensure long-term economic viability. As complex thermal-hydraulic simulations require specific expertise, they are difficult to access for many energy suppliers and mine operators. This underlines the need for user-friendly models that allow an initial assessment of the potential without in-depth knowledge of modelling.

Only a few reduced models for predicting mine water temperature exist in the literature. While analytical models impress with extremely short calculation times (Milliseconds for decades), they are not able to take seasonal storage effects into account. Reduced numerical models from the literature can consider these transient effects, but require significantly longer calculation times (Minutes for decades), especially for turbulent flow regimes. This is too time-consuming for a comprehensive parameter study. In order to combine the advantages of both approaches, a new model is required that takes seasonal storage effects into account and works in an acceptable computing time (< 1 minute for decades).

The newly developed model combines two coupled sub-models: The heat transfer in the rock is solved numerically using an implicit finite volume scheme, whereby an irregular grid enables an efficient calculation. The energy transport in the fluid is modelled using an analytical solution. The verification using a reference case shows a high accuracy of the model. At a constant reinjection temperature, the deviation of the outlet temperature after twenty years is 3 % compared to the benchmark (fully numerical axisymmetric CFD simulation). With cyclical heat loads, the maximum deviation occurs for the inlet temperature with 3,5 %. At the same time, the model is around 750 times faster than the benchmark calculations with a run time of 1 s for two decades.

In order to test the model with practical operating modes, it is compared with an existing CFD simulation of the north-west field of the mine in Ehrenfriedersdorf, Germany. The comparison focuses on the temperature development of the mine water at the outlet, assessing how well the model captures seasonal variations and contributes to optimizing operational decisions.