Optimizations of performance in a spectral beam splitting photovoltaic/thermal system using ideal optical windows

Abstract

Performance of a spectral beam splitting photovoltaic/thermal system depends on the characteristics of the splitting liquid. The ideal optical window directly guides the selection of splitting liquid. Most of the existing research established ideal optical windows in step function shapes and there is lack of systematic investigations. In this paper, the wavelength range to be studied is divided into discrete independent grids, thus introducing gridded transmissivity. A hypothetical liquid with unknown gridded transmissivity is used in a double-pass spectral beam splitting photovoltaic/thermal system model. Using particle swarm optimization algorithm, taking the best system performance as the target, the gridded transmissivity is determined. As the grid number increases, the determined distribution of gridded transmissivity converges to a step function. Based on the determined pattern, further optimizations are conducted. Moreover, considering the power consumption of the pump, the initial wavelengths, termination wavelengths, and flow rates to achieve the maximum efficiencies are also determined and discussed. Finally, metallic nanofluids are used to approach a determined ideal optical window. Combination of multiple kinds of metallic nanoparticles shows improvement over one kind in the system. A broadly tenable trend that being closer to the optimal window results in higher efficiency is also obtained.