Computational Modeling of a Solar Thermoelectric Generator

Abstract

Solar thermoelectric generators (STEGs) convert solar energy to electricity. The solar energy is first used to heat an absorber plate that serves as the high temperature reservoir. Power is generated by connecting the hot reservoir and cold (ambient) reservoirs with a pair of p- and n-doped thermoelectric legs. Experimental studies have shown that the efficiency of a STEG can reach values of about 5% if the entire setup is placed in near-vacuum conditions. However, under atmospheric conditions, the efficiency decreases by more than an order of magnitude, presumably due to heat loss from the absorber plate by natural convection. A coupled fluid–thermal–electric three-dimensional computational model of a STEG is developed with the objective of understanding the various loss mechanisms that contribute to its poor efficiency. The governing equations of mass, momentum, energy, and electric current, with the inclusion of thermoelectric effects, are solved on a mesh with 60,900 cells, and the power generated by the device is predicted. The computational model predicts a temperature difference (ΔT) of 16.5 K, as opposed to the experimentally measured value of 15 K. This corresponds to a peak power of 0.031 W as opposed to the experimentally measured peak power of 0.021 W. When only radiative losses are considered (i.e., perfect vacuum), the ΔT increases drastically to 131.1 K, resulting in peak power of 1.43 W. The predicted peak efficiency of the device was found to be 0.088% as opposed to the measured value of 0.058%.