Optimization with genetic algorythms of PVT system global efficiency

Abstract

Photovoltaic (PV) solar panels generally produce electricity in the 6% to 12% efficiency range, the rest being dissipated in thermal losses. To recover this amount, hybrid photovoltaic thermal systems (PV/T) have been devised. These are devices that simultaneously convert solar energy into electricity and heat. A significant amount of research on PV/T collectors has been carried out over the last decade and water PV/T glazed flat plate collector systems turned out to be the most promising to develop (Zondag[1]). It is thus interesting to study the PV/T system as part of a closed loop single phase water CDU (coolant distribution unit) in laminar forced convection. In particular, the analysis was conducted on the optimal cooling performance of the thermal part, testing polynomial channel profiles of varying order (from zero to fourth) for channels of a real industrial module heat sink, under the following conditions: ideal flux of 1000 W/m² on one side, insulation on the opposite side, periodic conditions on the remaining sides, fully developed thermal and velocity profile in laminar flow of water. Through the use of a genetic algorithm, we have optimized the shape of the channel’s sidewalls in terms of heat transfer maximization. In terms of Nusselt number, results show that fourth order profiles are the most efficient. When limits to allowable pressure loss and module weight are introduced, these bring generally to a lower efficiency of the system than the unconstrained case.