Solar Optical Design. It s an imaging problem. Mike Sullivan November 2008

Size: px

Start display at page:

Download "Solar Optical Design. It s an imaging problem. Mike Sullivan November 2008"

Rose Hart
5 years ago
Views:

1 Solar Optical Design It s an imaging problem Mike Sullivan November 2008

2 2

3 3 Outline Design Problem Solar Spectrum and Detectors Concentration Defined Direct Imaging Pupil imaging Concentrator Designs Fabrication

4 DesignProblem 1. Optics provide 100% fill factor with smallest cell footprint Concentration improves cell efficiency Reduces cost in module with discrete cell backplane 2. Design should allow +/- 1 degree Angle of Acceptance Sun is.5 degrees Field of View Associated with Tracker: step, alignment, wind, thermal 3. Design should provide 500x concentration or higher Concentration trades with Angle of Acceptance 4. Design should provide uniform illumination Cell inefficiencies due to hot spots: stay away from 2500x locally 5. Fabrication costs should be lower than $20 per square meter!!!!!!!! Goal for a $1 per Watt system installed Cost is for lens from vendor Very large volume consumer optics fabrication method needed 6. Operate for 20 years in the field Hail storms, UV exposure, thermal extremes

5 5 Solar Spectrum 50% in Vis, 50% in NIR Silicon 20% efficient and limited to 500x concentration Auger recombination Triple junction GaAs 40% efficient and can go 1500x or more Heat dissipation issues

6 Concentration Optical Invariant 1 A 1 2 A 2 Everyone wants to violate this constant! 1 A 1 2 Lens area Concentration defined as ratio of solid angles or ratio of areas between lens aperture and detector Geometric Ratio: Lens area/ Detector area A 2 Detector area Optical Ratio: Lens area/ Image size Concentrator Power = Independent of FL A 1 A 2 FL was determined by detector size and sun FOV

7 Direct Imaging Sun Angular view of sun from earth Direct radiation: 1000 W/m 2 = 1 sun 0.53 degrees Earth What Focal Length lens images the sun onto a 1 mm detector 0.53 deg D Marginal Ray Sin = h/f f Chief Ray h A 100mm FL lens makes a 920 micron diameter spot

8 Image the sun onto the detector with defocus to improve off axis performance Direct imaging Issues Cell size needs to be 9x larger area to accommodate +/- 0.5 deg. Geometric concentration lower than optical concentration Local high flux potential on Cell Stay away from 2500x hot spots Solar cell Sun Concentrator Optics For our 100mm FL lens with 1mm image Cell needs to be 3mm square to accommodate +/-.5 deg. Cost issue back on semiconductor again

9 Image the sun onto the detector with defocus to improve off axis performance Direct imaging w/defocus Defocus helps uniformity Spherical aberration more uniform spot before focus

10 10 Immersion Lens Secondary Optical Element shortens Focal Length Could make Equivalent 38mm FL in one lens though 75 mm 30 mm Single lens Power very high though Needs to be aspheric Curvature issue Fresnel lens, but have F/# issue with efficiency

11 Folded Optics Long Focal length can be folded to save space Large cells use long FL designs Concept is Galilean Telescope EFL Referred to as Cassegrain in Solar

11 11 Folded Optics Long Focal length can be folded to save space Large cells use long FL designs Concept is Galilean Telescope EFL Referred to as Cassegrain in Solar The "Classic" Cassegrain has a parabolic primary mirror, and hyperbolic secondary, there s RC and HK designs Obscuration Can it be made at low cost Use smaller cells

12 Pupil Imaging Pupil Concentrator Secondary Optic Pupil Image Solar cell Sun Concentrator Primary Optic Sun Image Pupil imaging gives uniform illumination on detector Secondary Optic forms image of pupil on cell Insensitive to sun position over secondary lens diameter Roland Winston calls this a Kohler design Kohler illumination used in microscopes

13 Pupil Imagers Concentration power trades with AOI sensitivity Sandia Labs Patent Roland Winston Patent Concentration up to 1500x feasible Secondary can be lens or light bucket.5 deg 1 deg

14 14 Fresnels

15 15 CPV - Fresnel Lens Ammonix Emcore Concentrix CS la Mancha Abengoa Green and Gold Daido Steel Opel Arima Sunrgi Sol3g Entech

16 16 CPV Reflective Optics Solar Systems Concentrating Technologies Menova Energy SolFocus GreenVolts

17 17 CPV Low Profile Energy Innovations Soliant Isofoton Pyron Solar

18 Plastic molding Mold costs, material, process: equal cost in 3 rd s Can get 100,000 from 1 mold UV and thermal issues Low melting point high cycle times Glass molding Can get 20,000 from 1 mold Cycle time high due to high melting point Can get large panels with current methods Rolled glass line like ornamental shower doors? AR coatings High cost for optics Can system take the efficiency hit without Low cost method needed- dip coat, sol gel, moth eye Hybrid Methods Silicone on glass Polymer on glass Untested Fabrication

19 Direct Imaging Lower cost Sun walks over cell so cell needs to be larger Potential hot spots Pupil imaging: Best for uniformity Allows beam walk or AOA Need 2 optical elements Concentrations up to 1500x feasible Fabrication Aggressive cost targets Fresnels nice but lossy AR coatings for improved throughput Summary

Multiband Solar Concentrator using Transmissive Dichroic Beamsplitting

Multiband Solar Concentrator using Transmissive Dichroic Beamsplitting Jason H. Karp and Joseph E. Ford Photonics Systems Integration Lab University of California, San Diego Jacobs School of Engineering