Geometrical Optimisation of Receivers for Concentrating Solar Thermal Systems

Abstract

In concentrated solar thermal technologies, the receiver converts concentrated solar radiation into high-temperature heat. Solar receivers are commonly simulated with a stochastic integration method: Monte-Carlo ray-tracing. The optimisation of the geometry of receivers is challenging when using existing optimisation methods for two reasons: each receiver evaluation using Monte-Carlo ray-tracing requires significant computational effort and the outcome of a simulation involves uncertainty. A series of novel optimisation techniques are proposed to enable gradient-free, stochastic and multi-objective optimisation adapted to such problems. These techniques address the computational load difficulty and the challenge of conducting stochastic optimisation based on uncertain evaluations by introducing the concepts of “Progressive Monte-Carlo Evaluation (PMCE)”, “Intermediate Ray Emission Source (IRES)” and adaptive view-factor calculation. A new “Multi-Objective and Evolutionary PMCE Optimisation (MOEPMCE-O)” method is then built around PMCE to enable multi-objective geometrical optimisation of receivers. PMCE is shown to be able to reduce the computational time of a random search optimisation by more than 90% and is used in the geometrical design of a new receiver for the Australian National University SG4 dish concentrator that achieved 97.1% (±2.2%) of thermal efficiency during on-sun testing. MOE-PMCE-O is applied to a multi-objective tower receiver problem where liquid sodium is used as the receiver heat-carrier in a surround configuration heliostat field. A series of useful geometrical concepts emerge from the results, with geometrical features able to maintain high efficiency while keeping acceptable incident peak flux values with a moderate receiver total mass. Finally, a more fundamental look at the impact of the interaction of concentrating optics on the exergy of radiation available at the receiver location highlights the major role played by concentrator surface slope error in lowering the exergy in concentrated solar thermal systems and quantifies the exergy loss associated with non-ideal match between flux and surface temperature in receivers.