Concept for an integral approach to explore the behavior of rock salt caverns under thermo-mechanical cyclic loading in energy storage systems

The fluctuating nature of renewable energy sources can be managed by storing the surplus of electrical energy in an appropriate reservoir. The excess electricity available during off-peak periods of consumption may be used to compress air or electrolyze hydrogen. Afterward, the pressurized gas is stored in the rock salt cavities and discharged to compensate the shortage of energy when required. During this process, the rock salt surrounding the cavern undergoes thermo-mechanical cyclic loading. In order to achieve a reliable geotechnical design, the stress–strain response of rock salt under such loading condition has to be identified and predicted. In order to investigate the rock salt behavior under such loading, a comprehensive study using three concepts of geotechnical engineering, i.e., experimental investigation, constitutive modeling and numerical analysis, is conducted. A triaxial experimental setup is developed to supplement the knowledge of the cyclic thermo-mechanical behavior of rock salt. The imposed boundary conditions in the experimental setup are assumed to be similar to the stress state obtained from a full-scale numerical simulation. The computational model relies primarily on the governing constitutive model for predicting the behavior of rock salt cavity. Hence, a sophisticated elasto-viscoplastic creep constitutive model is developed to take into account the dilatancy and damage progress, as well as the temperature effects. The contributed input parameters in the constitutive model can be calibrated using the experimental measurements. In the following, the initial numerical simulation is modified based on the calibrated constitutive model. However, because of the significant levels of uncertainties involved in the design procedure of such structures, a reliable design can be achieved by employing probabilistic approaches. Therefore, the numerical calculation is extended by statistical tools such as sensitivity analysis, optimum experimental design, back analysis, probabilistic analysis and robust reliability-based design to get final design parameters of paramount need for practice.