Séminaire CETHIL - Sorption thermal energy storage: two approaches for system-scale investigations

Luca SCAPINO, Flemish Institute for Technological Research (VITO)

Résumé au format pdf

Sorption thermal energy storage (STES) has the potential to have higher energy densities and lower thermal losses compared to conventional thermal storage technologies, and it can contribute to increase the energy grid flexibility and the penetration of intermittent and distributed energy sources. However, STES is a technology still under research, and system-scale investigations are necessary to determine its potential in future energy systems.

In this presentation, two main topics are addressed. The first topic concerns the development of a STES reactor model suitable for dynamic energy system simulations. In the dynamic simulation of an energy system, the system components are co-simulated in a common modeling environment and, if an extensive temporal horizon has to be modeled, the temporal resolution of the simulation can be a bottleneck in terms of computational cost. Therefore, a relatively simple model for each system component, which can still give accurate prediction of the component outputs given the inputs, is desirable. In this presentation, a data-driven model that uses artificial neural networks to simulate the sorption reactor outputs given its inputs is shown. The main advantage of this type of models is that, once that the model is trained, the computational cost during a dynamic simulation is very low. However, the training process can be computationally expensive, and the availability of good quality and representative data is a crucial requirement.

The second topic concerns the techno-economic optimization of energy systems with a STES integrated. Techno-economic analyses are a crucial step towards the optimal integration of new technologies in future energy systems. The economic impact of an open solid STES in an energy system operating in different energy markets is investigated. The reference energy system analyzed consists of a main heating grid that extracts the thermal energy from a geothermal doublet, inspired by an existing installation in Belgium, and provides it to two consumers: a low temperature district heating network and an organic Rankine cycle. The optimization problems are formulated as a mixed integer linear program, and the impact of a STES on the energy system yearly profit is assessed for different storage sizes and at the presence of different market mechanisms.