Photovoltaic technology produces electricity from solar radiation, but the cost of this electricity is typically 2—4 higher than the cost of electricity from conventional fuels. Two reasons contribute to this high cost: the need to use large amounts of expensive semiconductor material, and the low conversion efficiency. The first issue can be resolved by concentration: collecting solar radiation over a large area with a mirror or lens, and concentrating it to a small target where only a small area of semiconductor cells is needed. Several types of concentrating systems have been proposed and tested in recent years with varying sizes and levels of concentration. The second issue is the conversion efficiency: silicon PV cells convert 10–20% of the collected radiation to electricity, while the more sophisticated (and very expensive) multi-junction cells still reach less then 40% efficiency. Most of the collected solar energy is then rejected to the environment. A way to achieve a better overall efficiency is to capture the rejected heat as well, and use it as an additional energy product. This is possible today with Photovoltaic/Thermal (PV/T) flat collectors; but these provide heat at relatively low temperature, about 40–60°C. This is suitable only for the limited applications of domestic water heating or space heating. According to available end-use statistics, thermal energy for heating and cooling accounts for 50—70% of the total energy use in many types of buildings. However, in sunny regions much of that requirement is for air conditioning and refrigeration that cannot be provided by low-temperature heat. Concentrating photovoltaic (CPV) systems can achieve a synergy that addresses both the above-mentioned issues. CPV systems can greatly reduce the needed area of cells, and they can operateat higher temperatures than flat collectors. The result is the CPV/Thermal (CPVT) system as proposed in the UPP-Sol project, which provides both electricity and medium-temperature heat. The collected heat is therefore suitable for a wide range of thermal applications, including absorption cooling and air-conditioning, steam production, desalination, and industrial process heat. The CPV array heat exchanger can be easily adjusted to provide a wide range of temperatures by regulating the flow rate of the coolant. Therefore, a CPVT system can be used for polygeneration- providing electricity, heating, cooling, and possibly other products such as steam or desalinated water. A schematic of a typical CPVT system, including an array of small collectors and the electrical and thermal components, is shown in Figure1.

Figure 1: Layout of a CPVT polygeneration plant and its connections to the building’s utility systems

An important requirement for a CPVT polygeneration system is the location of the solar collectors. Since heat cannot be transported over long distances, the collectors have to be close to the end-user, and therefore have to be small enough to integrate into an urban environment, for example on the rooftop of the building where the end-user is located. Small units can be arranged in a modular way, even in irregular spaces that may be available in many buildings. Examples of small CPVT collector arrays on typical rooftops are shown in Figure 2. Another advantage of the location at the end-user site is the fact that the solar energy replaces conventional energy bought at retail cost, which is much higher than the production cost to the utility. Therefore, the same solar technology may be non-competitive at the utility’s power station, but competitive at the end-user site. In addition, the end-user may enjoy government incentives given to renewable energy systems, making the system even more competitive.

Figure 2: Small CPVT collectors on typical buildings: freestanding collectors (left) and roof-integrated collectors (right); collectors are computer generated over images of real buildings.

CPVT collector technology for urban applications has been developed by the project participants (TAU, DiSP, SHAP and UF-CREAR) under national and bi-national programs. The collectors are small, about 1 m is diameter, to permit easy integration and installation in buildings. Concentration is by a factor of a few hundreds, so the needed area of PV cells is much smaller than the collector aperture area. Triple-junction cells with nominal efficiency of 35% are used, to obtain the highest possible system efficiency, and also since this type of cells is capable of operating under higher temperatures relative to silicon cells. Two collector versions with different geometries have been developed for applications in different types of buildings. One is a roof-integrated stationary collector, and the second is a freestanding tracking parabolic dish type collector. The stationary Building Integrated Spherical Collector (BISC) has a fixed primary reflector that is installed permanently in the roof structure and an internal tracking system that moves the CPV receiver and a smaller secondary reflector. This is possible due to the symmetry of the spherical geometry: the optics is the same regardless of the direction of the incident solar radiation. Movement of the sun shifts the location of the focus, and tracking requires moving the receiver along the trajectory of the focal point. The secondary reflector corrects the spherical aberrations of the primary, thus allowing high concentration on the CPV receiver. A BISC collector developed at TAU is shown in Figure 3. A similar collector with some variations in design details has been constructed at UF-CREAR. A single collector will produce about 70 W of electricity and 160 W of thermal energy under nominal conditions. The amount of solar radiation collected and converted by the stationary collector is smaller than for a fully tracking collector of the same aperture area, since only a part of the reflective area is active at any given time. However, the cost of the stationary collector is lower than a fully tracking one of the same aperture area, and therefore the cost of the energy should be similar. Stationary collectors can be installed side by side, while tracking collectors need some empty space to avoid shading; therefore roof area usage may be better with stationary collectors.

Figure 3: BISC collector at TAU: hemispherical reflector, tracking mechanism and roof-integration flange

The second CPVT solution, the freestanding dish collector developed by TAU and DiSP, is shown in Figure 4. It consists of a concentrating glass mirror, a tracking mechanism that follows the apparent motion of the sun, and a CPV receiver with high-efficiency triple junction cells and a heat exchanger for extraction of the thermal energy. Each collector unit, with aperture area of about 1 m2, can produce about 200 W of net electrical power and 480 W of thermal power. The thermal power can be used directly as heat, or can be converted, for example, to 330 W of cooling power by the use of an absorption cooler.

Figure 4: The MCPV dish collector is shown under solar testing

In a recent analysis, we have estimated the cost of cooling with a CPVT system with electric net metering, coupled to a single effect absorption chiller. This is an attractive candidate for polygeneration, since cooling requirements have a good correlation with the availability of sunlight. This was compared to the cost of cooling with a conventional electric compressor, and to solar cooling with flat-plate thermal collectors. The results are shown in Figure 5 for a certain conventional (grid) electricity cost. The solar thermal cooling is more expensive than conventional cooling, but cooling with the CPVT polygeneration system is preferable when its installation cost is less than $4.5/Wp (€3.75/Wp). Several recent analyses have estimated the cost of CPV systems, when produced on large scale, to reach the range of $1—3/Wp. Therefore the potential for competitiveness against conventional energy sources, even without government incentives, is very promising.