Cyclic performance of cascaded and multi-layered solid-PCM shell-and-tube thermal energy storage systems: A case study of the 19.9 MWe Gemasolar CSP plant

Abstract

A shell-and-tube heat exchanger which incorporates a sensible or phase change material (PCM) as the storage medium offers a potentially commercially viable alternative to the two-tank molten salt system. In particular, cascaded PCMs and multi-layered solid-PCMs (MLSPCMs) were investigated as proposed systems which can reduce the amount of storage material used and ensure optimal storage utilization. In this work, the performance of various thermal energy storage (TES) alternatives integrated into the 19.9 MWe Gemasolar concentrated solar power (CSP) plant (located in Seville, Spain) were compared with the conventional two-tank system. These alternative storage configurations were characterized by a single tank filled with a single, cascaded, or multi-layered storage media. Importantly, as a system-level study, this paper compared the performance of the design alternatives integrated with other CSP components in order to capture the effect of dynamic interactions between the storage system and other CSP components. Through a validated numerical investigation of the annual performance of the integrated systems, all the design alternatives were compared in the context of annual electricity generation, which represents the ultimate criterion to judge the true potential of each alternative. To conduct an apples-to-apples comparison, the storage capacity and geometric parameters were fixed. The design alternatives were categorized based on the storage materials involved and their percentages of occupancy in the TES tank (i.e. 12 storage groups and a total number of 45 design alternatives). It was found that the well-designed TES designs with cascaded PCMs performed similarly in charging and discharging (i.e. with a similar amount of total stored or delivered energy per cycle). This contrasts with a single PCM system, where there exists a significant difference between charging and discharging performance. The results of annual cyclic performance, under real-time operational conditions, indicated that a MLSPCM design configuration that was filled with a high melting point PCM in the top 25% of the tank, sensible concrete in the middle 50%, and a low melting point PCM in the bottom 25% of the tank had the best performance among all design alternatives studied. Moreover, it was found that changing the filler portions any one cascaded PCM group cannot significantly change the annual performance of the system. Contrary to much of the available literature – literature which does not consider system integration – it was shown that the shell-and-tube alternatives can only approach the annual performance of two-tank systems under ‘extended’ operational conditions (i.e. allowing temperature set points to float relatively far away from their fixed design points).