Luka Tas , Niels Hartog, Martin Bloemendal, David Simpson, Tanguy Robert, Robin Thibaut, Le Zhang, Thomas Hermans

Aquifer Thermal Energy Storage (ATES) has great potential to mitigate CO2 emissions associated with the heating and cooling of buildings and offers wide applicability. Thick productive aquifer layers have been targeted first, as these are the most promising areas for ATES. Regardless, there is currently an increasing trend to target more complex aquifers such as low-transmissivity and alluvial aquifers or fractured rock formations. There, the uncertainty of subsurface characteristics and, with that, the risk of poor-performing systems is considerably higher. Commonly applied strategies to decide upon the ATES feasibility and well design standards for optimization may need to be adapted. To further promote the use of ATES in such less favorable aquifers an efficient and systematic methodology evaluating the optimal conditions, while not neglecting uncertainty, is crucial. In this context, the distance-based global sensitivity analysis (DGSA) method is tested. The analysis focuses on one promising thick productive aquifer, first used to validate the methodology, as well as a complex shallow alluvial aquifer. Through this method, multiple random model realizations are generated by sampling each parameter from a predetermined range of uncertainty. The DGSA methodology validates that the hydraulic conductivity, the natural hydraulic gradient and the annual storage volume dominate the functioning of an ATES system in both hydrogeological settings. The method also advances the state of the art in both settings. Darcy flux measurements can provide a first prediction of the relative ATES efficiency ahead of investing in more detailed studies. Insensitive parameters can be fixed to average values without compromising on prediction accuracy justifying streamlined models in the future. It also demonstrates the insignificance of seasonal soil temperature fluctuations for very shallow storage of thermal energy and it clarifies the thermal energy exchange dynamics directly above the storage volume in unconfined shallow aquifers. Analysis of the parameter distributions allowed us to gain more insights into favorable conditions for ATES and to propose a cut-off criterion for its application in alluvial aquifers with high natural hydraulic gradient. The nuanced understanding gained with this study contributes to the optimization of ATES systems, offering practical guidance for enhanced efficiency of feasibility studies, especially in challenging environments. The broad prior uncertainty strategy proves its value by expanding (while clearly delimiting) the applicability of the findings.