Design and simulation of supercritical carbon dioxide recompression Brayton cycles with regenerators for recuperation

Abstract

The supercritical carbon dioxide recompression Brayton cycle is being considered for CSP applications due to its potential for lower cost and higher efficiency relative to steam Rankine cycles, but achieving high efficiency requires large, expensive recuperators. Regenerators may provide cost and performance advantages over traditional heat exchanger technology. A regenerator is a periodic heat exchanger in which the transfer of heat from the hot to cold fluid is temporally decoupled via a thermal energy storage media, such as a packed bed of spheres inside an insulated pressure vessel. Valves control the allocation of fluid into and out of such a packed bed to ensure quasi-steady flow. Regenerators are inherently transient devices, and their dynamic performance in sCO2 power cycles has never been quantified. This work presents a one-dimensional, transient regenerator model suitable for system simulation. The model is validated against 10 kWth experimental regenerator data. Simulations of 10 MWe sCO2 power cycles equipped with a high-temperature regenerator find that the pressurization and de-pressurization of regenerative heat exchangers result in compressor flow rate and net power fluctuations of up to 10% and 6%, respectively, every 22.5 seconds, demonstrating the importance of considering process switching when sizing regenerator packed beds and valves. Splitting the regenerator into a greater number of packed bed-valve sets operated in parallel could reduce the magnitude of these fluctuations. Part-load and load-following analysis finds that regenerators perform as well as traditional recuperators in load-following scenarios. Optimization of a 100 MWe regenerator-equipped sCO2 cycle for CSP finds that the potential economic benefit of regenerators relative to recuperators depends largely on the cost of the latter technology. Regenerators may have a greater advantage in higher-temperature applications, but this depends largely on the feasibility and cost of manufacturing large valves using nickel alloys.