Browsing UA Faculty Publications by Journal
Now showing items 1-2 of 2
CPV generator with dish reflector and fly’s eye receiverWe describe a CPV generator in which an off-axis paraboloidal dish reflector powers a small receiver near the focus, housing many individually illuminated multijunction cells. The receiver entrance window doubles as a field lens that forms a reduced scale image of the reflector, at concentration of ∼30×. The image has a sharp boundary, and its position is stable against tracking errors. A fly’s eye lens array divides the image into equal portions, and further concentrates it to ∼500× onto the cells. This approach is in contrast to nearly all previous PV and CPV, where sunlight is equal apportioned (for simple series electrical connection) directly on entering the system. In our approach, small multijunction cells are packaged into a small receiver module that will be less expensive (per watt) to manufacture than large conventional PV or CPV modules, and can be economically upgraded for 40 year lifetime. Our concept differs from REhnu’s dish/receiver design  in its lack of obscuration and simpler cooling, using forced air convection rather than pumped liquid coolant, this made possible by the lower heat density at the cell array. In preliminary on-sun system data with a 2.4 m2 prototype powering 5.2 mm cells at 500× concentration, we demonstrate good tolerance to mispointing (90% at 0.5° off-axis), good air cooling (cell mounting plate at 19°C above ambient) and uniform division of light between the cells (scatter of 3.3% rms).
REhnu dish based CPV: Performance and reliability improvements based on a year of field experienceREhnu has now built up a year of experience operating two M-8 CPV generators with dish-receiver architecture described in Stalcup et al. 2017 . The M-8 generators use 8 primary collector mirrors made of back-silvered low-iron glass, 1.65 m square with a 1.5 m focal length. Small sealed receivers at each dish focus house 36 Solar Junction 3J cells operating at 950× concentration. Over the year of operation, a good measure of output power vs DNI and atmospheric conditions has been obtained. The efficiency varies through the day depending on air mass. The system is currently best matched to the solar spectrum at air mass 2. The daily peak CSTC efficiency averages 31%, with seasonal variation of ±1% through the year, peaking at 32% in August. Several improvements were made over the year to ensure reliability and to improve optical throughput. A dry air purging system has mitigated damage to the cells from moisture. Material upgrades to secondary reflectors in the receiver have resolved a runaway failure mode that went previously undetected due to inadequate testing time.