Combining entropy weight and TOPSIS method for selection of tank geometry and filler material of a packed-bed thermal energy storage system

Thermocline thermal energy storage systems are promising alternatives for recovering waste heat lost by industry around the world. The aim of this work is to extend the methodology presented in previous work, by optimising an existing industrial packed-bed storage system on two geometric optimisation variables, considering exergy, environmental and economic aspects. Seven filler materials are compared for the same heat transfer fluid, to include discrete variables in the model. The multi-objective optimisation problem is solved using the NSGA-II multi-objective genetic algorithm. For each filler material, a Pareto set is obtained. The non-dominated solutions within the union of the different Pareto sets are then selected, which give a new single set of optimised solutions. A multi-criteria decision-making method (TOPSIS) is then applied to obtain the optimal solution. To avoid any subjective choice from the decision-maker by determining the objective weights of each of the optimisation criteria, the Shannon entropy is used. The combination of TOPSIS and Shannon entropy led to the selection of a recycled ceramic obtained from hard coal ashes as the best filler. This solution has a stocky tank shape (2.4 m diameter, 2.1 m height) and a small particle diameter (7 mm). The exergy and environmental performance is improved compared to the reference storage. They reach 98.0% (vs 95.6%) and 58 hab.year (vs 67 hab.year) respectively. The levelised cost of energy is close to that of the reference tank (3.35 vs 3.31 ceuro/kWhth).