Abstract :
[en] Continued demand for new applications from spacecrafts leads to larger power needs. For most of them, the primary energy sources are solar panels. However, these solar panels are heavy and expensive, mainly due to solar cells and their coverglass. The power need increase requires thus new concepts of solar panels. Furthermore, space environment is highly constraining: the vacuum limits heat trans-
fer since convection is not allowed, there is contamination modifying the light spectrum
reaching cells, ultraviolet light induces a yellowing of silicone glue, radiations degrade pho-
tovoltaic cells, etc. Usually, multi-junctions (MJ) cells are used, that are strongly sensitive to spectral modification in their incident spectrum due to their series connected structure (the worst cell defines the whole output performance). The power dimensioning of solar panels is then based on end of life (EOL) PV cells expected performance. Reducing the sensitivity to increase EOL output power is then another challenge for space applications.
This thesis aims to answer to both problems, by the proposition and study of a new lightweight solar concentrator with spectral splitting. This conceptual concentrator is composed of a Fresnel lens for sunlight concentration, coupled to a surface relief diffraction grating to spectrally split incident light, forming a single optical element in silicone. The concentration behavior allows a reduction of solar cells area (including its coverglass), replaced by a lightweight silicone lens, reducing the global cost and mass. The lateral spectral separation will permit other types of cells than the usual MJ cells. To demonstrate the concept, two single junctions (SJ) solar cells are placed side by side, the first collecting visible light, the second collecting near IR light. Since cells are electrically independent, sensitivity is lower and EOL output power can surpass standard MJ cells systems. Moreover, cells combination is nearly free: all photovoltaic technology can be used and combined, opening the solar concentration field to other technologies than MJ.
This work studies and evaluates the pros and cons of diffraction grating/Fresnel lens combinations as solar concentrator with spectral splitting. The analysis includes the choice of material, optical developments and optimizations (grating selection, freeform lenses, grating period optimization along the lens, etc.), electrical modeling and methods to find the best cells combination as well as the modeling of expected output power, a thermal simulation, weight and deployment considerations, and approaches of sensitivity with spectral modifications. Some experimental results complete the study. Two main configurations were developed: a first with a blazed grating, and a second with a symmetrical lamellar grating. Both configurations, after optimization, show similar performances: a global geometrical concentration ratio around 5-6× (ratio between lens width and the total width of the two cells), a tracking error tolerance up to 0.7◦, no drastic degradation with respect to deformations, fabrication errors, etc., an output power at begin of life (BOL) better than a classical concentrator focusing on a SJ cell or a planar solar panel composed of SJ cells. Both configurations present also a BOL specific mass [kg/W] lower than a classical planar solar panel covered by MJ cells, and a lower sensitivity to space environment, with theoretical larger EOL output power depending on chosen PV cells combinations. Existing deployment systems are proposed and discussed. Also, an experimental prototype of the optical element was realized, with optical results close to simulations. Some adaptations to terrestrial concentrators, other spectra, or using more than 2 different SJ cells are also introduced, highlighting the versatility of our concept.
Results demonstrate thus the coherence of the concept, leading to experimentally feasible and quite tolerant concentrator, with interesting cost reduction thanks to concentration and specific weight reduction. However, optical losses due to the non-perfect spectral splitting of the grating is too high to be able to surpass MJ cells systems BOL up to now, and the lower sensitivity of our concentrator cannot completely compensate this lack of performance at EOL. To reach better performances than MJ systems, from cost, mass, radiation resistance, etc. point of view and to exploit the obvious gain in sensitivity, the concentrator needs thus further investigations mainly related to cells combinations. Especially as our concept opens a lot of opportunities thanks to the complete independence of cells: other technologies are allowed, specific coatings can improve performances, ... Applications for Earth or places with different/changing incident spectra is another plausible perspective. Our concentrator with spectral splitting can also be a major advantage for specific scientific space missions like deep space missions.
