Abstract

The maximum concentration ratio achievable with a solar concentrator made of a single refractive primary optics is much more limited by the chromatic aberration than by any other aberration. Therefore achromatic doublets made with poly(methyl methacrylate) and polycarbonate are of great interest to enhance the concentration ratio and to achieve a spectrally uniform flux on the receiver. In this Letter, shaped achromatic Fresnel lenses are investigated. One lossless design is of high interest since it provides spectrally and spatially uniform flux without being affected by soiling problems. With this design an optical concentration ratio of about 8500× can be achieved.

© 2013 Optical Society of America

Since lenses exhibit chromatic aberration unlike mirrors, their applicability in solar concentration is limited [1]. Considering a spectrum of 380–1600 nm—a typical range for CPV since out of this range solar flux is low and the external quantum efficiency of a triple junction cell is about 40% [25]—the longitudinal chromatic aberration (LCA) of simple Fresnel lenses limits the concentration to about 1000× under normal incidence [6] while the theoretical limit on earth due to the acceptance angle of the solar semi-diameter (959 [7]) is about 46000× [8]. When combining a diverging lens in polycarbonate (PC) and a converging lens of poly(methyl methacrylate) (PMMA), higher concentration ratios can be achieved. Recently, a concentration ratio close to 2000× has been experimentally measured with a flat achromatic Fresnel doublet [9]. However, these lenses are either subject to soiling problems (if the outward surface is textured), or are hard to manufacture (if the interface between the optical materials is textured). Moreover, flat lenses are intolerant to tracking error. The tracking error angle, θerr, causes a displacement of the centroid of the collected flux following xC=ftanθerr (with f the paraxial focal distance), affecting particularly slow systems (i.e., with high f-numbers) since the maximum concentration ratio due to the angular size of the source is given by

Cmax=(1tanθs×f#)2.
To investigate the influence of the PMMA/PC-interface angle, we will first discuss the case of a double prism (one prism of PMMA fixed on a prism of PC). Then the dome-shaped achromatic Fresnel lens will be investigated and compared with an aspheric optimized interface.

As for previous publications [6,9,10], PMMA and PC will be used to design the achromatic lenses at 468 and 961 nm: the two wavelengths of achromatization, i.e., the wavelengths focusing exactly on the receiver and minimizing the LCA. Relevant parameters of the achromatic Fresnel prism are sketched in Fig. 1. The receiver is centered at the origin O(0,0) and is perpendicular to the y axis. β is the inclination angle of the PC/PMMA-interface and α and γ are the prism angle of the PMMA and PC, respectively. Finally, δ is the deviation angle and θS the angular radius of the source.

 

Fig. 1. Sketch of an achromatic prism made with PMMA (purple) and PC (cyan).

Download Full Size | PPT Slide | PDF

Let’s fix some parameters, say δ=30° and θS=2.5°. There is an infinite number of values possible to make one wavelength reach the center of the receiver. For a selected value of the inclination angle of the interface, β, we require that both 468 and 961 nm impact the receiver, which fixes the values of α and γ. More precisely, we impose a nonimaging condition in our algorithm: the edges rays are forced to symmetrically impact the center of the receiver, i.e., |x+x+|<ε, with ε close to zero. The minimum size of the receiver is therefore given by

Δx=|x+x|=2x+.

In order to obtain a value that depends only on the deviation angle, we will use the normalized value

Δx=Δx/xp.

As can be deduced from Fig. 2(b), Δx decreases when β tends to 45°. At this value, θ1+ (the angle between the normal of the PMMA and the positive incoming edge-ray) gets close to the limit of 90° [see Fig. 2(a)] and the Fresnel coefficient of reflection gets high. In Fig. 2(c), the evolution of Δx with the interface angle is also depicted for a deviation angle of 2°.

 

Fig. 2. Effects of the inclination angle on the incoming angle (a) and on the distance between the edge rays at the receiver level for a deviation angle of 30° (b) and 2° (c).

Download Full Size | PPT Slide | PDF

Two conclusions can be drawn: (1) curved achromatic segments can highly reduce the size of the receiver for high deviation angles but only a slight improvement is obtained for small angles of deviation, (2) a fully concentration-optimized lens is impossible to design since the incoming angle at the first interface must be 90° along the full lens (and in addition to a high reflection coefficient, this lens would have been intolerant to tracking errors). Finally, these conclusions are simply the same as for prisms made with a single material.

Even if curved lenses cannot improve the concentration factor, it can increase the flux uniformity and the tracking tolerance [1]. In 2003, Leutz and Ries explained that nonimaging (dome-) shaped lenses cannot be designed with commonly available materials [11]. This assertion appears to be valid if the outward surface is perfectly smooth. The idea of a textured surface facing the Sun (grooves-out) is usually immediately rejected in order to avoid soiling problems.

This problem of soiling exists in the case of dome-shaped lenses: two different dome-shaped lenses are shown in Fig. 3 with the PMMA component either above or below the interface. In both cases, the dome-shaped achromatic lenses were optimized in order to have the edge-rays impacting symmetrically the plane of the absorber.

 

Fig. 3. Dome-shaped lenses, i.e., with the interface, in blue, situated on a perfect circle. On the left, the top material is the PMMA (in purple) and on the right, the top material is the PC (in cyan).

Download Full Size | PPT Slide | PDF

In the first case, when the PMMA is above the interface, the distance between x+ and x decreases with the deviation angle (which corresponds to the previous analysis), meaning that the flux reaching the receiver will be highly nonuniform. Moreover, this design suffers from an important shadowing effect at the air/PMMA interface.

Concerning the second design, with the PC above the interface, the distance between x+ and x increases with the deviation angle, reducing drastically the concentration ratio and negatively affecting the uniformity of the flux reaching the receiver. The shadowing effect is however much lower than that of the first achromatic dome-shaped Fresnel lens.

To overcome the problem of uniformity, we have to move to free-shape lenses. We will force the slope of the interface to be such that Δx remains constant whatever the radial distance of the incoming edge-rays. As previously deduced from Fig. 2, the inclination angle of the left design of Fig. 3 must be increased in order to increase the distance between x+ and x, meaning that the radius of curvature of the interface must increase too. To the contrary, regarding the right design of Fig. 3, the inclination angle must decrease. Therefore the radius of curvature of the interface must also be decreased.

The optimized designs are shown in Figs. 4(a) and 4(b), both have been optimized for an angular radius of the source of 5° and has been achromatized for 468 and 961 nm. On the one hand, the first design [Fig. 4(a), with the PMMA on top] is prone to soiling and suffers from an important shadowing effect and therefore it will be no longer discussed. On the other hand, the second design [Fig. 4(b)] does not suffer from any shadowing effect and the soiling problem is avoided in spite of the texturization of the outer surface.

 

Fig. 4. Shaped Fresnel lenses optimized and illuminated with an angular diameter of 5°. Left and center: achromatic. Right: singlet.

Download Full Size | PPT Slide | PDF

The performance of this latest design will be compared to a nonimaging shaped Fresnel singlet [see Fig. 4(c)], only made with PMMA [1]. The comparison has been performed with a ray-tracing program with a simulated source whose characteristics match those of Sun. At the design wavelength(s), the (achromatic) shaped lens has a detector diameter given by Δx0=2ftanθS. We will investigate the energy enclosed in this radius as a fraction of the incoming energy (in other words the optical efficiency ηopt), to which corresponds an optical concentration ratio

Copt=(2rl/Δx0)2ηopt,
where rl is the lens radius: 0.46f and 0.70f for the singlet and for the doublet, respectively, whatever the incoming design angle [see Fig. 3]. Largest radii lead to total internal reflexions.

Table 1 presents relevant results for different design angles (θdsn) and different illumination angles (θS). Fresnel reflexion losses have been simulated, but the roughness has been neglected.

Tables Icon

Table 1. Nonimaging Shaped Lenses: Comparison Between Singlets and Achromatic Doublets

As can be expected, achromatic-shaped lenses concentrate much more energy in Δx0 than singlets, and the optical efficiency is always higher with doublets despite the supplementary interface. The advantage of the doublets is particularly pronounced for small angles of incidence where the chromatic aberration highly enlarges the focal spot produced by the singlet. In this condition, the flux reaching the receiver is almost perfectly uniform with the doublet while the distribution is Gaussian with the singlet, as shown on Fig. 5.

In conclusion, nonimaging achromatic shaped Fresnel doublets allow for very high concentration, up to 8500× for an angle of incidence of 0.26°, with a flux spectrally uniform (no chromatic aberration) and spatially uniform too (flat shape irradiance). To achieve such performance, a texturization of the outward surface was necessary while avoiding dust accumulation thanks to a profile without local minimum.

 

Fig. 5. Comparison of the concentration distribution for the singlet (left) and the doublet (right) with θdsn=θS=0.26°. The red and purple circles have the same diameter of Δx0=2ftanθS. They contain, respectively, 17.1% and 84.3% of the incoming irradiance (both squares receive about 88%), corresponding to the last line of Table 1.

Download Full Size | PPT Slide | PDF

References

1. R. Leutz and A. Suzuki, in Nonimaging Fresnel Lenses (Springer, 2001), Chap. 1.

2. A. W. Bett, F. Dimroth, and G. Siefer, in Concentrator Photovoltaics (Springer, 2007), Chap. 4.

3. Spectrolab data sheets: www.spectrolab.com/DataSheets/PV/CPV/CDO-100-C3MJ.pdf.

4. NREL’s AM1.5 Standard Dataset: http://rredc.nrel.gov/solar/spectra/am1.5/, accessed on 02/06/2011.

5. C. A. Gueymard, Sol. Energy 71, 325 (2001). [CrossRef]  

6. F. Languy, K. Fleury, C. Lenaerts, J. Loicq, D. Regaert, T. Thibert, and S. Habraken, Opt. Express 19, A280 (2011). [CrossRef]  

7. S. Puliaev, J. L. Penna, E. G. Jilinski, and A. H. Andrei, Astron. Astrophys. Suppl. Ser. 143, 265 (2000). [CrossRef]  

8. R. Winston, J. Opt. Soc. Am. 60, 245 (1970). [CrossRef]  

9. F. Languy, C. Lenaerts, J. Loicq, T. Thibert, and S. Habraken, Sol. Energy Mater. Sol. Cells 109, 70 (2013). [CrossRef]  

10. F. Languy and S. Habraken, Opt. Lett. 36, 2743 (2011). [CrossRef]  

11. R. Leutz and H. Ries, in Proceedings of the 2nd International Solar Concentrator Conference for the Generation of Electricity or Hydrogen (2003).

References

  • View by:
  • |
  • |
  • |

  1. R. Leutz and A. Suzuki, in Nonimaging Fresnel Lenses (Springer, 2001), Chap. 1.
  2. A. W. Bett, F. Dimroth, and G. Siefer, in Concentrator Photovoltaics (Springer, 2007), Chap. 4.
  3. Spectrolab data sheets: www.spectrolab.com/DataSheets/PV/CPV/CDO-100-C3MJ.pdf .
  4. NREL’s AM1.5 Standard Dataset: http://rredc.nrel.gov/solar/spectra/am1.5/, accessed on 02/06/2011 .
  5. C. A. Gueymard, Sol. Energy 71, 325 (2001).
    [CrossRef]
  6. F. Languy, K. Fleury, C. Lenaerts, J. Loicq, D. Regaert, T. Thibert, and S. Habraken, Opt. Express 19, A280 (2011).
    [CrossRef]
  7. S. Puliaev, J. L. Penna, E. G. Jilinski, and A. H. Andrei, Astron. Astrophys. Suppl. Ser. 143, 265 (2000).
    [CrossRef]
  8. R. Winston, J. Opt. Soc. Am. 60, 245 (1970).
    [CrossRef]
  9. F. Languy, C. Lenaerts, J. Loicq, T. Thibert, and S. Habraken, Sol. Energy Mater. Sol. Cells 109, 70 (2013).
    [CrossRef]
  10. F. Languy and S. Habraken, Opt. Lett. 36, 2743 (2011).
    [CrossRef]
  11. R. Leutz and H. Ries, in Proceedings of the 2nd International Solar Concentrator Conference for the Generation of Electricity or Hydrogen (2003).

2013 (1)

F. Languy, C. Lenaerts, J. Loicq, T. Thibert, and S. Habraken, Sol. Energy Mater. Sol. Cells 109, 70 (2013).
[CrossRef]

2011 (2)

2001 (1)

C. A. Gueymard, Sol. Energy 71, 325 (2001).
[CrossRef]

2000 (1)

S. Puliaev, J. L. Penna, E. G. Jilinski, and A. H. Andrei, Astron. Astrophys. Suppl. Ser. 143, 265 (2000).
[CrossRef]

1970 (1)

Andrei, A. H.

S. Puliaev, J. L. Penna, E. G. Jilinski, and A. H. Andrei, Astron. Astrophys. Suppl. Ser. 143, 265 (2000).
[CrossRef]

Bett, A. W.

A. W. Bett, F. Dimroth, and G. Siefer, in Concentrator Photovoltaics (Springer, 2007), Chap. 4.

Dimroth, F.

A. W. Bett, F. Dimroth, and G. Siefer, in Concentrator Photovoltaics (Springer, 2007), Chap. 4.

Fleury, K.

Gueymard, C. A.

C. A. Gueymard, Sol. Energy 71, 325 (2001).
[CrossRef]

Habraken, S.

Jilinski, E. G.

S. Puliaev, J. L. Penna, E. G. Jilinski, and A. H. Andrei, Astron. Astrophys. Suppl. Ser. 143, 265 (2000).
[CrossRef]

Languy, F.

Lenaerts, C.

F. Languy, C. Lenaerts, J. Loicq, T. Thibert, and S. Habraken, Sol. Energy Mater. Sol. Cells 109, 70 (2013).
[CrossRef]

F. Languy, K. Fleury, C. Lenaerts, J. Loicq, D. Regaert, T. Thibert, and S. Habraken, Opt. Express 19, A280 (2011).
[CrossRef]

Leutz, R.

R. Leutz and A. Suzuki, in Nonimaging Fresnel Lenses (Springer, 2001), Chap. 1.

R. Leutz and H. Ries, in Proceedings of the 2nd International Solar Concentrator Conference for the Generation of Electricity or Hydrogen (2003).

Loicq, J.

F. Languy, C. Lenaerts, J. Loicq, T. Thibert, and S. Habraken, Sol. Energy Mater. Sol. Cells 109, 70 (2013).
[CrossRef]

F. Languy, K. Fleury, C. Lenaerts, J. Loicq, D. Regaert, T. Thibert, and S. Habraken, Opt. Express 19, A280 (2011).
[CrossRef]

Penna, J. L.

S. Puliaev, J. L. Penna, E. G. Jilinski, and A. H. Andrei, Astron. Astrophys. Suppl. Ser. 143, 265 (2000).
[CrossRef]

Puliaev, S.

S. Puliaev, J. L. Penna, E. G. Jilinski, and A. H. Andrei, Astron. Astrophys. Suppl. Ser. 143, 265 (2000).
[CrossRef]

Regaert, D.

Ries, H.

R. Leutz and H. Ries, in Proceedings of the 2nd International Solar Concentrator Conference for the Generation of Electricity or Hydrogen (2003).

Siefer, G.

A. W. Bett, F. Dimroth, and G. Siefer, in Concentrator Photovoltaics (Springer, 2007), Chap. 4.

Suzuki, A.

R. Leutz and A. Suzuki, in Nonimaging Fresnel Lenses (Springer, 2001), Chap. 1.

Thibert, T.

F. Languy, C. Lenaerts, J. Loicq, T. Thibert, and S. Habraken, Sol. Energy Mater. Sol. Cells 109, 70 (2013).
[CrossRef]

F. Languy, K. Fleury, C. Lenaerts, J. Loicq, D. Regaert, T. Thibert, and S. Habraken, Opt. Express 19, A280 (2011).
[CrossRef]

Winston, R.

Astron. Astrophys. Suppl. Ser. (1)

S. Puliaev, J. L. Penna, E. G. Jilinski, and A. H. Andrei, Astron. Astrophys. Suppl. Ser. 143, 265 (2000).
[CrossRef]

J. Opt. Soc. Am. (1)

Opt. Express (1)

Opt. Lett. (1)

Sol. Energy (1)

C. A. Gueymard, Sol. Energy 71, 325 (2001).
[CrossRef]

Sol. Energy Mater. Sol. Cells (1)

F. Languy, C. Lenaerts, J. Loicq, T. Thibert, and S. Habraken, Sol. Energy Mater. Sol. Cells 109, 70 (2013).
[CrossRef]

Other (5)

R. Leutz and H. Ries, in Proceedings of the 2nd International Solar Concentrator Conference for the Generation of Electricity or Hydrogen (2003).

R. Leutz and A. Suzuki, in Nonimaging Fresnel Lenses (Springer, 2001), Chap. 1.

A. W. Bett, F. Dimroth, and G. Siefer, in Concentrator Photovoltaics (Springer, 2007), Chap. 4.

Spectrolab data sheets: www.spectrolab.com/DataSheets/PV/CPV/CDO-100-C3MJ.pdf .

NREL’s AM1.5 Standard Dataset: http://rredc.nrel.gov/solar/spectra/am1.5/, accessed on 02/06/2011 .

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (5)

Fig. 1.
Fig. 1.

Sketch of an achromatic prism made with PMMA (purple) and PC (cyan).

Fig. 2.
Fig. 2.

Effects of the inclination angle on the incoming angle (a) and on the distance between the edge rays at the receiver level for a deviation angle of 30° (b) and 2° (c).

Fig. 3.
Fig. 3.

Dome-shaped lenses, i.e., with the interface, in blue, situated on a perfect circle. On the left, the top material is the PMMA (in purple) and on the right, the top material is the PC (in cyan).

Fig. 4.
Fig. 4.

Shaped Fresnel lenses optimized and illuminated with an angular diameter of 5°. Left and center: achromatic. Right: singlet.

Fig. 5.
Fig. 5.

Comparison of the concentration distribution for the singlet (left) and the doublet (right) with θdsn=θS=0.26°. The red and purple circles have the same diameter of Δx0=2ftanθS. They contain, respectively, 17.1% and 84.3% of the incoming irradiance (both squares receive about 88%), corresponding to the last line of Table 1.

Tables (1)

Tables Icon

Table 1. Nonimaging Shaped Lenses: Comparison Between Singlets and Achromatic Doublets

Equations (4)

Equations on this page are rendered with MathJax. Learn more.

Cmax=(1tanθs×f#)2.
Δx=|x+x|=2x+.
Δx=Δx/xp.
Copt=(2rl/Δx0)2ηopt,

Metrics