This paper presents a load-based optimization approach for improving the efficiency of a packed bed. The optimization is based on splitting the work-cycle of the thermal store into two frequency components: low and high. A packed bed is designed for each one of the two profiles. A packed bed can be customised much better for a duty-cycle that contains a narrow range of frequencies.
The case study presented considers a 24 h working-cycle (12 h charge / 12 h discharge) with a 10 MW peak power and an exergy storage requirement of 33.3 MW h (76.3 MW h of heat). A packed bed was optimized for this duty-cycle using a one dimensional model that varies the aspect ratio and the rock size. This packed bed is the ‘reference case’ for the study. The aim of the load-based optimization is to create a two-bed system that achieves lower exergy losses than the reference case while keeping the overall storage capacity constant.
A sign-preserving filter is used as the signal-splitting tool. Numerous different work-cycle “splits” are explored. Results show that the exergy losses of the packed bed can be considerably reduced. The optimum work-cycle split considers a low-frequency packed bed that supplies 85% of the storage capacity and a high-frequency packed bed that provides the remaining 15%. The combined losses of the two packed beds are 644 kW h, which represents a reduction of 25.5% in comparison to the exergy losses of the reference case. The study demonstrates that the “load-based optimization” allows replacing a packed bed with an equivalent but more efficient two-bed system at almost no additional cost.
Cárdenas, B., & Garvey, S. D. (2019). A load-based approach for optimizing a packed-bed thermal store. Journal of Energy Storage, 25, Article 100835. https://doi.org/10.1016/j.est.2019.100835