A beam-splitting photovoltaic thermal receiver for solar concentrators
Sunlight is an abundant source of energy that can be converted into heat and electricity using photothermal and photovoltaic technologies, respectively. Usually these devices are separate from each other, and occupy significant space on a rooftop or in a solar park. Combining thermal and electrical output in a single package would achieve several advantages, such as more efficient use of available space and light collection.
Photovoltaic-thermal hybrid (PVT) collectors deliver solar-generated electricity and heat concurrently. They capture light in silicon photovoltaic (PV) cells and partially convert it to electricity with 15–20% efficiency. A running heat transfer fluid at the back of the device captures heat dissipated in the cells from the rest of the incoming light energy, and delivers the heat as useful energy. However, this kind of PVT collector is not yet commercially viable. One shortcoming is that the outlet temperature of the thermal component is low, since the silicon cells need to be kept cool to function efficiently.
We have developed a low-cost hybrid PVT system capable of delivering a thermal output of 150°C or greater, without increasing the temperature of the PV cells. The system is designed for linear solar concentrators, which operate using optics to focus sunlight onto a linear receiver tube to heat a fluid, creating steam to run a turbine. Using our system produces high-temperature thermal energy output suitable for industrial processes and absorption chillers.
The core innovation of our design is a low-cost spectral beam-splitting device that divides the concentrated sunlight into different wavelength bands and directs each to different receivers. Previous research on spectral splitters describes mainly thin-film wave interference filters.1 These comprise multiple layers of refracting materials, tens of nanometers thick, which are deposited on a transparent substrate to create a wave interference effect. The thickness of each layer is finely tuned, so that incoming rays are caused to overlap each other. This process then generates constructive and destructive waves, which increase reflection in specific bands and transmit the rest of the spectrum. However, manufacturing such filters requires sophisticated and costly processes that normally include vacuum deposition. We have overcome this problem using volumetric filters, which exploit the intrinsic properties of one or more materials to absorb specific wavelengths and transmit the rest of the incoming light. Figure 1 depicts the wave interference and volumetric processes.1
The solar spectrum consists of wavelengths in the range 400–2500nm, while silicon solar cells function most efficiently for the range 700–1200nm. Our beam-splitting mechanism separates light in the 700–1200nm range and directs it to the PV cells, sending the rest of the solar spectrum to a thermal absorber. Since the thermal and PV receivers are now fed by two separate beams of light, we can thermally decouple them from each other and control their temperatures independently. The thermal absorber can operate at 150°C, whereas the silicon cells remain cool at ambient temperatures.
We designed our system for a commercial parabolic trough: a partially curved, mirror-lined solar collector. We constructed a detailed ray tracing model (used to calculate the path of light waves through a system) for the proposed integration of our device in the solar concentrator, and optimized the dimensions of the receiver to maximize the energy yield of the system. It is also possible to adapt the proposed receiver for linear concentrators with different concentration ratios and aperture areas.
Our device enables the production of electricity and heat to supply a building using a single package that maximizes the energy yield from a limited rooftop area. Furthermore, the amount of silicon used in the PV component is small, and therefore it would be possible to retrofit existing linear thermal concentrators with PVT technology without significant extra cost. As a further development, we will analyze the energy production of such a receiver for different load conditions and electricity-to-heat ratios.