Reference is made to commonly-assigned International Patent Applications Nos. WO 2007/016363 to Miñano et al. and WO 2007/103994 to Benítez et al. which are incorporated herein by reference in their entirety.
Embodiments of the devices described and shown in this application may be within the scope of one or more of the following U.S. patents and patent applications and/or equivalents in other countries: U.S. Pat. Nos. 6,639,733, issued Oct. 28, 2003 in the names of Miñano et al., 6,896,381, issued May 24, 2005 in the names of Benítez et al., 7,152,985 issued Dec. 26, 2006 in the names of Benítez et al., and 7,460,985 issued Dec. 2, 2008 in the names of Benítez et al.; WO 2007/016363 titled “Free-Form Lenticular Optical Elements and Their Application to Condensers and Headlamps” to Miñano et al and US 2008/0316761 of the same title published Dec. 25, 2008 also in the names of Miñano et al; WO 2007/103994 titled “Multi-Junction Solar Cells with a Homogenizer System and Coupled Non-Imaging Light Concentrator” published Sep. 13, 2007 in the names of Benítez et al; US 2008/0223443, titled “Optical Concentrator Especially for Solar Photovoltaic” published Sep. 18, 2008 in the names of Benítez et al.; and US 2009/0071467 titled “Multi-Junction Solar Cells with a Homogenizer System and Coupled Non-Imaging Light Concentrator” published Mar. 19, 2009 in the names of Benítez et al.
Concentration-Acceptance Product (CAP)—A parameter associated with any solar concentrating architecture, it is the product of the square root of the concentration ratio times the sine of the acceptance angle. Some optical architectures have a higher CAP than others, enabling higher concentration and/or acceptance angle. For a specific architecture, the CAP is nearly constant when the geometrical concentration is changed, so that increasing the value of one parameter lowers the other.
Fresnel Facet—Element of a discontinuous-slope concentrator lens that deflects light by refraction.
TIR Facet—Element of a discontinuous-slope concentrator lens that deflects light by total internal reflection.
Primary Optical Element (POE)—Optical element that receives the light from the sun or other source and concentrates it towards the Intermediate Optical Element, if any, or to the Secondary Optical Element.
Intermediate Optical Element (JOE)—Optical element that receives the light from the Primary Optical Element and concentrates it towards the Secondary Optical Element.
Secondary Optical Element (SOE)—Optical element that receives the light from the Primary Optical Element or from the Intermediate Optical element, if any, and concentrates it towards the solar cell or other target.
Cartesian Oval—A curve (strictly a family of curves) used in imaging and non-imaging optics to transform a given bundle of rays into another predetermined bundle, if there is no more than one ray crossing each point of the surface generated from the curve. The so-called Generalized Cartesian Oval can be used to transform a non-spherical wavefront into another. See Reference [10], page 185, Reference [16].
Triple-junction photovoltaic solar cells are expensive, making it desirable to operate them with as much concentration of sunlight as practical. The efficiency of currently available multi-junction photovoltaic cells suffers when local concentration of incident radiation surpasses ˜2,000-3,000 suns. Some concentrator designs of the prior art have so much non-uniformity of the flux distribution on the cell that “hot spots” up to 9,000-11,000× concentration happen with 500× average concentration, greatly limiting how high the average concentration can economically be. Kaleidoscopic integrators can reduce the magnitude of such hot spots, but they are more difficult to assemble, and are not suitable for small cells.
There are two main design problems in Nonimaging Optics, and both are relevant here. The first is called “bundle-coupling” and its objective is to maximize the proportion of rays in a given input bundle that are transformed into a given output bundle. In a solar concentrator, that is effectively to maximize the proportion of the light power emitted by the sun or other source that is delivered to the receiver. The second problem, known as “prescribed irradiance,” has as its objective to produce a particular illuminance pattern on a specified target surface using a given source emission.
In bundle-coupling, the design problem consists in coupling two ray bundles Mi and Mo, called the input and the output bundles respectively. Ideally, this means that any ray entering into the optical system as a ray of the input bundle Mi exits it as a ray of the output bundle Mo, and vice versa. Thus the successfully coupled parts of these two bundles Mi and Mo comprise the same rays, and thus are the same bundle Mc. This bundle Mc is in general Mc=Mi∩Mo. In practice, coupling is always imperfect, so that Mc⊂Mi and Mc⊂Mo.
In prescribed-irradiance, however, it is only specified that one bundle must be included in the other, Mi in Mo. Any rays of Mi that are not included in Mo are for this problem disregarded, so that Mi is effectively replaced by Mc. In this type of solution an additional constraint is imposed that the bundle Mc should produce a prescribed irradiance on a target surface. Since Mc is not fully specified, this design problem is less restrictive than the bundle coupling one, since rays that are inconvenient to a particular design can be deliberately excluded in order to improve the handling of the remaining rays. For example, the periphery of a source may be under-luminous, so that the rays it emits are weaker than average. If the design edge rays are selected inside the periphery, so that the weak peripheral region is omitted, and only the strong rays of the majority of the source area are used, overall performance can be improved.
Efficient photovoltaic concentrator (CPV) design well exemplifies a design problem comprising both the bundle coupling problem and the prescribed irradiance problem. Mi comprises all rays from the sun that enter the first optical component of the system. Mo comprises those rays from the last optical component that fall onto the actual photovoltaic cell (not just the exterior of its cover glass). Rays that are included in Mi but are not coupled into Mo are lost, along with their power. (Note that in computer ray tracing, rays from a less luminous part of the source will have less flux, if there are a constant number of rays per unit source area.) The irradiance distribution of incoming sunlight must be matched to the prescribed (usually uniform) irradiance on the actual photovoltaic cell, to preclude hot-spots. Optimizing both problems, i.e., to obtain maximum concentration-acceptance product as well a uniform irradiance distribution on the solar cell's active surface, will maximize efficiency. Of course this is a very difficult task and therefore only partial solutions have been found.
Good irradiance uniformity on the solar cell can be potentially obtained using a light-pipe homogenizer, which is a well known method in classical optics. See Reference [1]. When a light-pipe homogenizer is used, the solar cell is glued to one end of the light-pipe and the light reaches the cell after some bounces on the light-pipe walls. The light distribution on the cell becomes more uniform with light-pipe length. The use of light-pipes for concentrating photo-voltaic (CPV) devices, however, has some drawbacks. A first drawback is that in the case of high illumination angles the reflecting surfaces of the light-pipe must be metalized, which reduces optical efficiency relative to the near-perfect reflectivity of total internal reflection by a polished surface. A second drawback is that for good homogenization a relatively long light-pipe is necessary, but increasing the length of the light-pipe both increases its absorption and reduces the mechanical stability of the apparatus. A third drawback is that light pipes are unsuitable for relatively thick (small) cells because of lateral light spillage from the edges of the bond holding the cell to the end of the light pipe, typically silicone rubber. Light-pipes have nevertheless been proposed several times in CPV systems, see References [2], [3], [4], [5], [6], and [7], which use a light-pipe length much longer than the cell size, typically 4-5 times.
Another strategy for achieving good uniformity on the cell is the Köhler illuminator. Köhler integration can solve, or at least mitigate, uniformity issues without compromising the acceptance angle and without increasing the difficulty of assembly.
Referring to
Despite the simplicity and high uniformity of illumination on the cell, the practical application of the Sandia system is limited to low concentrations because it has a low concentration-acceptance product of approximately 0.3 (±1° at 300×). The low acceptance angle even at a concentration ratio of 300× is because the imaging secondary cannot achieve high illumination angles on the cell, precluding maximum concentration.
Another previously proposed approach uses four optical surfaces, to obtain a photovoltaic concentrator for high acceptance angle and relatively uniform irradiance distribution on the solar cell (see Reference [9]). The primary optical element (POE) of this concentrator should be an element, for example a double aspheric imaging lens, that images the sun onto the aperture of a secondary optical element (SOE). Suitable for a secondary optical element is the SMS (Simultaneous Multiple Surface) designed RX concentrator described in References [10], [11], [12]. This is an imaging element that works near the thermodynamic limit of concentration. In this notation, the surfaces of the optical device are listed in the order in which the light beam encounters them: I denotes a totally internally reflective surface, R denotes a refractive surface, and X denotes a reflective surface that may be opaque. If a light beam encounters the same surface twice, it is listed at both encounters with the correct type for each encounter.
A good strategy for increasing the optical efficiency of the system (which is a critical merit function) is to integrate multiple functions in fewer surfaces of the system, by designing the concentrator optical surfaces to have at least a dual function, e.g., to illuminate the cell with wide angles, at some specified approximation to uniformity. That entails a reduction of the degrees of freedom in the design compared to the ideal four-surface case. Consequently, there is a trade-off between the selected geometry and the homogenization method, in seeking a favorable mix of optical efficiency, acceptance angle, and cell-irradiance uniformity.
There are two ways to achieve irradiance homogenization. The first is a Köhler integrator, as mentioned before, where the integration process is along both dimensions of the ray bundle, meridional and sagittal. This approach is also known as a 2D Köhler integrator. The other strategy is to integrate in only one of the ray bundle's dimensions; thus called a 1D Köhler integrator. These integrators will typically provide a lesser homogeneity than is achievable with in 2D, but they are easier to design and manufacture, which makes them suitable for systems where uniformity is not too critical. A design method for calculating fully free-form 1D and 2D Köhler integrators was recently developed (see References [13], [14]), where optical surfaces are used that have the dual function of homogenizing the light and coupling the design's edge rays bundles.
In all the embodiments of the present invention, the primary optical element is reflective. The use of reflective primaries is old in solar concentrators, since the parabolic mirror has been in the public domain since centuries. More recently, advanced high-performance free-form asymmetric mirror designs that use a free-form lens with a short kaleidoscope homogenizer protruding from it [14]. designs have been developed. Also recently, the use of two-mirror Cassegrain type concentrators, common in antenna and telescope design, has been extended to solar concentrators with the addition of a kaleidoscope homogenizer [6], and with radial Kohler integration [14] [15].
Embodiments of the present invention provide different photovoltaic concentrators that combine high geometric concentration, high acceptance angle, and high irradiance uniformity on the solar cell. In all the embodiments, the primary optical element is reflective in the sense that the light rays exit the primary on the same side that the light rays impinged from. Also in all the embodiments, the primary and secondary optical elements are each lenticulated to form a plurality of segments. In some embodiments, an intermediate optical element, not necessarily segmented, is used in between the primary and the secondary. A segment of the primary optical element and a segment of the secondary optical element combine to form a Köhler integrator. The multiple segments result in a plurality of Köhler integrators that collectively focus their incident sunlight onto a common target, such as a photovoltaic cell. Any hotspots are typically in different places for different individual Köhler integrators, with the plurality further averaging out the multiple hotspots over the target cell.
In some embodiments, the optical surfaces are modified, typically by lenticulation (i.e., the formation on a single surface of multiple independent lenslets that correspond to the segments mentioned before) to produce Köhler integration. Although the modified optical surfaces behave optically quite differently from the originals, they are macroscopically very similar to the unmodified surface. This means that they can be manufactured with the same techniques (typically plastic injection molding or glass molding) and that their production cost is the same.
An embodiment of the invention provides an optical device comprising: a primary optical element having a plurality of segments, which in an example are four in number; and a secondary optical element having a plurality of segments, which in an example are four lenticulations of an optical surface of a lens; wherein each segment of the primary optical element, along with a corresponding segment of the secondary optical element, forms one of a plurality of Köhler integrators. The plurality of Köhler integrators are arranged in position and orientation to direct light from a common source onto a common target. The common source, where the device is a light collector, or the common target, where the device is a luminaire, may be external to the device. For example, in the case of a solar photovoltaic concentrator, the source is the sun. Whether it is the common source or the common target, the other may be part of the device or connected to it. For example, in a solar photovoltaic concentrator, the target may be a photovoltaic cell. Further embodiments of the device, however, could be used to concentrate or collimate light between an external common source and an external common target.
The above and other aspects, features and advantages of the present invention will be apparent from the following more particular description thereof, presented in conjunction with the following drawings wherein:
A better understanding of various features and advantages of the present invention may be obtained by reference to the following detailed description of embodiments of the invention and accompanying drawings, which set forth illustrative embodiments in which various principles of the invention are utilized.
Two types of secondary optical elements are described herein: one comprising an array of refractors, the second an array of reflectors. Both exhibit overall N-fold symmetry. In the embodiments taught in this specification the primary reflective elements have the same N-fold symmetry as the secondary optic. In some embodiments the primary is asymmetric so the rest of elements are not located in front of the primary but on the side. Two types of intermediate optical elements are described herein: reflective type, and refractive type. The reflective intermediate optical element folds the ray path, permitting the removal of the secondary optical element and the solar cell (and heat-sink) from in front of the primary.
As may be seen from
Several Köhler integrating solar concentrators are described herein. They are the first to combine a non flat array of Köhler integrators with concentration optics. Although, the embodiments of the invention revealed herein have quadrant symmetry, the invention does not limit embodiments to this symmetry but can be applied, by those skilled in the art, to other configurations (preferably N-fold symmetry, where N can be any number greater than two) once the principles taught herein are fully understood.
Radial Köhler concentrators are 1D Köhler integrators with rotational symmetry. This makes the design process much easier than a 1D free-form Köhler integrator. Furthermore, rotational symmetry makes the manufacturing process as simple for a lenticular form as for any other aspheric rotational symmetry. The design process, however, first designs a 2D optical system, and then applies rotational symmetry.
Although the irradiance distribution produced by a radial Köhler concentrator has a hotspot, it is much milder than that produced by an imaging system. If α is the system acceptance angle, αs is the sun's angular radius, and k is a constant that depends on the shape of the cell's active area (where k=1 for a round cell and k=4/π for a square cell), it can easily be seen that the hotspot generated by a radial Köhler approach is proportional to k*(α/αs) times the average optical concentration, while the hotspot generated by an aplanatic device is proportional to k*(α/αs)2 times the average optical concentration. For instance, if α=1°, αs=¼° (the angular radius of the sun as seen from Earth), and k=1, the hotspot created by a radial Köhler is around 4 times the average concentration, while the aplanatic design produces a hotspot 16 times the average concentration. For a square cell (k=4/π) the corresponding hotspots are 5 and 20 times the average concentration.
The radial Köhler concept has been applied in CPV systems to a two-mirror Cassegrain-type reflective concentrator (see Reference [15] and above-referenced WO 2007/103994).
In this Radial Köhler design, the average concentration and the peak concentration can be high, so that it is necessary to introduce a further degree of freedom in the radial Köhler design, in order to keep the irradiance peak below 2000 suns. To perform the integration in a second direction, the present application comprises a concentrator with four subsystems (having quad-symmetry), hereinafter referred to as segments, that symmetrically compose a whole that achieves azimuthal integration, while keeping each of the four subsystems rotationally symmetric and thus maintaining ease of manufacture, since each is actually a part of a complete rotationally symmetric radial Köhler system, analogous to those of
Better homogenization is produced when using a two-directional free-form Köhler integrator instead of a rotational-symmetric one. A possible type of free-form Köhler system is the same XXR, comprising a primary reflector, and intermediate reflector and a secondary refractor, in which the Kohler integration is performed between the primary and secondary elements.
The photovoltaic receiver has preferably a square flat active area, and without loss of generality can be considered as located in a coordinate system in which the receiver plane is z=0 and the sides of the active area are parallel to the x and y axes, and the origin is in the center of the active area. Because of the symmetry, defining the unit in the region x>0, y>0 fully defines the primary optical element. The intermediate optical element will preferably have rotational symmetry around the z axis. The secondary optical element will preferably have the same four-fold symmetry as the primary. In the particular embodiment shown in
The design process has then three stages. First, the diagonal cross section profiles of the primary and intermediate mirrors are designed as in two dimensions using the SMS2D method (detailed below) with the conditions that the edge rays impinging on the entry aperture tilted +α and −α (α being the design acceptance angle) are focused in two dimensions (i.e., all the rays are contained in a plane) on close to the boundary points A and B of its corresponding lenticulation of the secondary lens, see
The first stage of the design is done with the following process, illustrated by
1. Choose β, which is the direction of the normal to the optical surface at B.
2. Choose the x coordinates of R (& R′), which are the corner points of the active area of the PV cell 43, the x and z coordinates of point B and of point E, which is the outer corner of the selected lenticulation of the primary 41, and the z coordinate of point D, which is on the rim of the intermediate optical element 44.
3. Calculate the x coordinate of D by tracing the reversed ray R′-B-D.
4. Calculate the optical path length R′-B-D-E.
5. Choose α.
6. Calculate the normal vector at E so as to reflect the known reversed ray D-E into the direction −α.
7. Choose the z coordinate zA of point A, Calculate the x coordinate of point A using the formula xA=(21/2−1)/(21/2+1)xB.
8. Calculate the line of the intermediate mirror from D to C as a “distortion-free imaging oval” so that there is a linear mapping between tilt (sin) angles of rays at E in the range +/−α and points along the straight segment A to B. (See
9. Calculate the points of the secondary lens, starting from B, so that the rays from E reflected off the intermediate mirror are focused by refraction to R′ (using the optical path length condition, if desired). This is most conveniently done at the same time each point of the intermediate mirror is calculated.
10. The secondary lens calculated in step 9 will usually not pass through the previously chosen point A. The intersection of the secondary lens with the line x=xA gives a better estimation of zA. So go back to step 7, substitute the new value of zA, and do an “iteration loop of zA,” repeating steps 8 and 9, and optionally repeating this step 10.
11. Calculate the primary and intermediate mirrors with SMS2D to form an image of the incident light from angle −α in B and of the incident light from angle +α in A. (See
12. When the primary arrives at the z-axis, if the ray from +α at G after refraction at A does not reach R but a different point R″ on the receiver surface, go back to step 5 and choose a better a with value α *|R′R|/|R′R″|. Then repeat the subsequent steps.
13. If the x-coordinate of the last calculated point of the intermediate mirror (i.e., the closest to the z-axis) is not properly allocated (for instance, is negative), go back to step 2 and choose a different value for the coordinate xB of point B. Then repeat the subsequent steps.
14. Generate the three-dimensional intermediate mirror by revolution of the profile with respect to the z-axis.
In the second stage of the design, illustrated in
In the third step of the design, the secondary free-form lens is designed to form an image of the paired section of the primary optical element, reflected in the intermediate optical element, on the solar cell. Again, such a free-form lens can be designed, for instance, as the Generalized refractive Cartesian oval that receives rays passing through corner point E of the primary and reflected on the rotational intermediate mirror, and focuses them in three dimensions on the corner point R of the cell.
Note that the calculation in three dimensions of the primary and secondary is consistent with the two dimensional design, which means that the curves 95 and 96 contained in the free-form mirror and lens at the intersection of the diagonal x=y plane 90 in
The contour of the primary mirror in three dimensions is given by the image of the photovoltaic cell projected by the secondary lens. A notional cell larger than the real cell can be considered here, to allow for cell placement tolerances. The minimum contour size of the secondary lens units is defined by the image of the three-dimensional acceptance area (that is, the cone of radius α).
The intermediate mirror designed as described in the first stage differs very significantly from the aplanatic two mirror imaging design used in reference [6]. The aplanatic design produced focusing of the on-axis input rays onto an on-axis point, while the focal region of the on-axis input rays in the intermediate mirror designed according to the present embodiment is approximately centered in the off-axis segment AB. The difference is specially clear if the three-dimensional design is done using the intermediate mirror described in reference [6] and both +α rays and −α rays are traced as in
In another preferred embodiment, the intermediate mirror is also free-form and the primary and intermediate mirrors are designed using the SMS3D method, so four edge rays of the acceptance angle cone are approximately focused on four points at the rim of its corresponding lenticulation of the secondary in 3D geometry.
Referring to
An embodiment of the XXR Köhler in
Tables 1 to 3 (placed at the end of the description) provide an example of a concentrator according to
The present embodiments are a particular realization of the devices described in the above-mentioned patent application WO 2007/016363 to Miñano et al.
Variations can be obtained by designers skilled in the art. For instance, the number of cells, also called sections or lenslets, on each of the primary and secondary optical elements can be increased, for instance, to nine. Also the cell can be rectangular and not square, and then the four units of the primary mirror will preferably be correspondingly rectangular, so that each unit still images easily onto the photovoltaic cell. Alternatively, or in addition, the number of array units could be reduced to two, or could be another number that is not a square, so that the overall primary is a differently shaped rectangle from the photovoltaic cell. Where each segment is further subdivided into lenslets, the desirable number of lenslets in each primary and secondary lens segment may depend on the actual size of the device, as affecting the resulting size and precision of manufacture of the lens features.
Examples of such variations are shown in
Although various specific embodiments have been shown and described, the skilled reader will understand how features of different embodiments may be combined in a single photovoltaic collector, luminaire, or other device to form other devices within the scope of the present invention. When the photovoltaic cell is replaced by an LED or an LED array, or other light source, the present embodiments provide optical devices that can collimate the light with a quite uniform intensity for the directions of emission, because all points on the source are carried to every direction. This can be used to mix the colors of different LEDs of a source array or to make the intensity of the emission more uniform without the need to bin the chips.
The preceding description of the presently contemplated best mode of practicing the invention is not to be taken in a limiting sense, but is made merely for the purpose of describing the general principles of the invention. The full scope of the invention should be determined with reference to the Claims.
This application claims benefit of U.S. Provisional Application No. 61/268,129 titled “Kohler Concentrator”, filed Jun. 8, 2009 in the names of Miñano et al., which is incorporated herein by reference in its entirety.
Number | Date | Country | |
---|---|---|---|
61268129 | Jun 2009 | US |