AU2011211343A1 - Luminescence Module Inspection - Google Patents

Abstract

Abstract Luminescence-based methods are presented for improved detection of faults in photovoltaic (PV) cells and modules before, during and after the assembly of individual cells into modules. One or more images of luminescence generated by 5 electrical excitation, photo-excitation or photo-excitation with electrical bias are analysed or compared to identify faults such as cracks, shunts and series resistance problems. If the same defect is identified in many samples, indicating a systematic problem, steps can be taken to locate and rectify its cause.

Description

- 1 Luminescence Module Inspection Field of the Invention The present invention relates to luminescence-based methods for detecting defects in photovoltaic cells before, during or after the manufacture of photovoltaic modules. 5 Related Applications The present application claims priority from Australian Provisional Patent Application No 2010903557, filed on 9 August 2010, the contents of which are incorporated herein by reference. 10 Background of the Invention Any discussion of the prior art throughout this specification should in no way be considered as an admission that such prior art is widely known or forms part of the common general knowledge in the field. 15 In a typical photovoltaic (PV) module, approximately 70 PV cells are connected electrically in series. A number of failure modes can occur during module assembly, including shunts, series resistance problems and cracking or breakage of cells caused for example by thermal stress during soldering. Existing inspection methods for detecting cracks in cells in PV modules include simple visual inspection, machine 20 vision-based optical transmission methods that detect open cracks as bright features, and peel tests wherein cracks can be revealed by low force peeling off of the metal contacts bearing chunks of cell material. Existing methods for detecting electrical problems (which can also be caused by cracks) include I-V curves of series-connected cells, thermal infra-red imaging and electroluminescence (EL) imaging, as described 25 for example in A.M. Gabor et al, 'Soldering Induced Damage to Thin Si Solar Cells and Detection of Cracked Cells in Modules', Proceedings 21st European Photovoltaic Solar Energy Conference, pp 2042-2047, 4-8 September 2006, Dresden, Germany, and in D.L. King et al, 'Photovoltaic Module Performance and Durability Following Long Term Field Exposure', <http://photovoltaics.sandia.gov/docs/PDF/prmking.pdf>. 30 Existing crack detection methods such as visual inspection and optical transmission can miss closed cracks or micro-cracks in cells that can grow to cause bigger problems, -2 while methods requiring physical contact, such as I-V testing, EL imaging and peel tests, themselves risk damaging the PV cells. The earlier that cracks or other problems are identified in the module assembly process, the easier and cheaper it is to rectify the problem, so it is important to increase the likelihood of early detection. A need exists 5 therefore for further inspection methods for detecting failure modes of cells in PV modules. It is an object of the present invention to overcome or ameliorate at least one of the disadvantages of the prior art, or to provide a useful alternative. It is an object of the 10 present invention in its preferred form to provide improved methods for detecting failure modes in photovoltaic modules. Summary of the Invention A first aspect of the present invention provides a method for detecting failure modes in 15 a process for manufacturing a photovoltaic module, said module comprising a plurality of electrically connected photovoltaic cells, said method comprising the steps of: (a) acquiring one or more luminescence images of at least one photovoltaic cell associated with said photovoltaic module; and (b) processing said one or more luminescence images to identify the incidence 20 or growth of a defect in said at least one photovoltaic cell, wherein at least one of said luminescence images is a photoluminescence image. In one preferred form, the one or more luminescence images comprise two or more luminescence images of the at least one photovoltaic cell acquired at different 25 positions in the process, and the method further comprises the step of: (c) comparing the two or more luminescence images. The two or more luminescence images preferably comprise images acquired before and after a particular step of the process. The particular step is preferably selected from the group consisting of tabbing, stringing, lay-up, interconnection of strings, and 30 lamination. Preferably, the comparing step comprises a pixel-by-pixel generation of a difference or ratio image.

-3 In another preferred form, the one or more luminescence images comprise two or more luminescence images of the at least one photovoltaic cell acquired at the same position in the process, with different excitation conditions, and the method further comprises the step of: 5 (c) comparing the two or more luminescence images. The comparing step preferably comprises a pixel-by-pixel generation of a difference or ratio image. In one preferred form, the at least one photoluminescence image is generated by 10 laterally inhomogeneous illumination of the at least one photovoltaic cell. In another preferred form, the photoluminescence in said at least one photoluminescence image is additionally generated by application of an electrical bias to the at least one photovoltaic cell. 15 Preferably, the at least one photovoltaic cell comprises individual cells before tabbing, individual tabbed cells, individual cells incorporated in strings, several cells within one string or entire strings, single or multiple cells within lay-ups, and single, multiple or all cells within a completed module. 20 In one preferred form, the defect is a crack. In another preferred form, the defect is a shunt. Preferably, the method further comprises the comparison of two or more luminescence images acquired at different average carrier injection levels. It is preferred for the different average carrier injection levels to be generated by different illumination intensities, different bias voltages including reverse bias, or different 25 electrical bias conditions under illumination. In yet another preferred form, the defect is a series resistance fault. The one or more luminescence images preferably comprise a photoluminescence image generated by homogeneous illumination and operation at the maximum power point of the at least 30 one photovoltaic cell. Preferably, the method further comprises the comparison of two or more luminescence images acquired under different bias conditions.

-4 A second aspect of the present invention provides an apparatus when used to perform the method according to the first aspect. A third aspect of the present invention provides an article of manufacture comprising a 5 computer usable medium having a computer readable program code configured to conduct the method according to the first aspect, or to operate the apparatus according to the second aspect. Detailed Description 10 Preferred embodiments of the invention will now be described, by way of example only, with regard to PV modules containing silicon wafer-based cells. The invention is not limited to modules containing this particular type of cell. PV modules are typically manufactured by connecting tens of individual PV cells in 15 series and mounting them in a single package. A typical module manufacture process includes the following sequence of steps: 1) Tabbing: A piece of solder-coated copper ribbon approximately twice the length of an individual cell is soldered to the front contacts of each cell. 2) Stringing: A 'string' (or row) of ten or so cells is produced by soldering the ex 20 cess length of tab from each cell to the rear side of the next cell in the row. At the be ginning and end of the string some excess tab is provided to allow series electrical con nection of strings without soldering directly to the cells. 3) Lay-up: Strings (rows) are arranged on the module superstrate (typically a glass sheet) coated with an uncured layer of an adhesive such as poly(ethyl vinyl acetate) 25 (EVA), which may be slightly tacky, allowing strings to be positioned but also re moved if required. 4) Interconnection of individual strings. 5) Lamination: The complete panel is assembled, for example with a white Tedlar* backing film, additional EVA and a metal sheet, and a combination of heat and 30 pressure is used to cure the EVA.

-5 Steps 1) and 2) are often combined, in a process called 'combined tabbing and stringing' (CTS). As mentioned above, a number of failure modes can occur during module assembly, 5 including shunts, series resistance problems and cracking or breakage caused for example by thermal stress during soldering. Some causes and effects of these failure modes will now be described. Cracks 10 Soldering (e.g. during tabbing) can initiate cracks, typically close to the solder joint, because of the mechanical stress arising from the different coefficients of thermal expansion of copper and silicon, and also from local heating of the silicon. As the solder joint cools below the melting point of the solder, shear stress is transmitted to the silicon as the copper contracts, typically warping or bowing every cell. The 15 soldering stress can also cause pre-existing cracks, often at the edges of cells, to propagate further. These cracks may also be propagated by other sources of mechanical stress including handling, laminating and thermal cycling in the field. In extreme cases, a piece of the cell breaks off. As demonstrated in the abovementioned conference paper by Gabor et al, a large crack can electrically isolate a substantial part 20 of a PV cell, especially if it is substantially parallel to the bus bars. When the cell concerned is incorporated into a module, it may be more susceptible to 'hot spot' problems when shaded, i.e. it will be under reverse bias with a greater current density than an intact cell. 25 Although it is best to detect cracked cells as early as possible, the main economic benefit is retained if cracked cells are detected prior to lamination so that problem cells can be replaced. With the increased use of CTS, rejecting cells between the tabbing and stringing steps may in any case be impossible. 30 Shunts Localised shunting in an individual cell can cause hot spots in a module when that cell is shaded. Even if a given cell has an acceptable total shunt value as measured on an I- -6 V tester, it may subsequently cause 'hot spot' module failure if the shunting is due to a relatively small number of local shunts, caused for example by inclusions, rather than a more distributed source such as faulty edge isolation. 5 Series resistance Sometimes modules have a series resistance significantly greater than the sum of the series resistance of the cells, due to soldering faults, for example. The present invention discloses the use of luminescence imaging, and 10 photoluminescence (PL) imaging in particular, for the detection of failure modes such as cracks, shunts and series resistance faults in cells, tabbed cells, strings, lay-ups or modules at one or more stages of a PV module manufacturing process. Apparatus and methods for PL imaging, optionally with the application of electrical bias, to detect faults such as cracks, shunts and series resistance problems in PV cells and wafers 15 during and after manufacturing, have been disclosed in published PCT patent application Nos WO 2007/128060 Al, WO 2009/026661 Al, WO 2009/129575 Al, WO 2010/019992 Al, WO 2010/130013 Al and WO 2011/079353 Al. The contents of these documents are incorporated herein by reference. 20 In certain embodiments of the invention, faults are detected from a single PL image, where the photoluminescence is generated using any combination of homogeneous or non-homogeneous illumination, and with or without electrical bias to the cell. That is, luminescence generated by optical excitation of a PV cell will be termed photoluminescence irrespective of whether the cell is also subject to a forward bias that 25 may itself be sufficient to generate luminescence. In other embodiments, faults are detected based on a comparison or analysis of two or more PL images, again measured using any combination of homogeneous or non-homogeneous illumination and with or without electrical bias to the cell. In certain embodiments, the comparison is between two or more images of a sample at a given point in a module manufacturing process, 30 but with different excitation conditions. In alternative embodiments, the comparison is between images of a sample measured before and after a particular process step (e.g. tabbing, stringing, lay-up, interconnection or lamination) with the same excitation -7 conditions, for determining whether any faults are induced during that step. In yet other embodiments, the comparison is between two or more images of a sample measured before and after a particular process step, with different excitation conditions. In still other embodiments the comparison is between images of two or 5 more samples measured at the same point in a module manufacturing process; these embodiments may be particularly useful for determining whether a given process step is causing a certain type of defect in a disproportionate number of samples and needs remedial action. PCT patent application number PCT/AU201 1/ entitled 'Persistent Feature Detection', filed on 8 August 2011 and incorporated herein by 10 reference, describes some of these embodiments in more detail. In particular preferred embodiments of the invention, PL images measured with laterally inhomogeneous illumination are used to detect cracks at any stage of module fabrication, including individual cells (e.g. incoming cell inspection), individual tabbed 15 cells, individual cells incorporated in strings, several cells within one string or entire strings, single or multiple cells within lay-ups, and single, multiple or all cells within a completed module. As disclosed in published PCT patent application No WO 2010/130013 Al entitled 'Material or device characterisation with non-homogeneous photoexcitation' and incorporated herein by reference, inhomogeneous illumination 20 induces a pattern of local lateral electrical current flows within the emitter layer of individual cells, such that interruption of the current flow by a crack can be detected in a PL image. In embodiments involving the comparison of images measured before and after a 25 particular process step, PL imaging with substantially uniform illumination over a cell surface is generally advantageous compared to EL imaging for the detection of non series-resistance related faults, such as cracks or other faults that reduce the local minority carrier lifetime or the local cell voltage, because in PL the uniform injection of carriers over a cell surface reduces variation in luminescence intensity arising from 30 changes to the electrical connection of the cell during a process step, e.g. series resistance effects from tabbing/stringing of cells.

-8 Image comparison can include many different mathematical operations such as pixel by-pixel difference or ratio of the luminescence intensities from equivalent cell areas in the images. This may require automated pattern recognition to identify the cell position, followed by translation and/or rotation to match the cell position in the two 5 images. PL imaging also has several specific industrially important advantages over EL imaging at the incoming cell inspection stage, related to the fact that electrical contact is not required. Firstly, accurate alignment of the cell is unnecessary, enabling cheaper 10 robotics to be used or potentially allowing integration with existing robotic handling tools. Secondly, no additional mechanical stress is induced in the cell, e.g. from electrical probes and/or vacuum hold-down. Thirdly, no maintenance associated with probe change-out is required. Fourthly, the results can't be compromised by poor electrical contact. Finally, measurements can potentially be taken with the cells still in 15 the boxes they were shipped in, minimising the likelihood of cell damage at the module manufacturer site. While laser-based illumination is often used to generate PL, especially for indirect band gap semiconductors such as silicon that are weak PL emitters, laser illumination 20 is not essential and it may be advantageous to use non-laser sources for PL imaging at the module stage to simplify integration into module manufacturing equipment with respect to laser safety hazards. Suitable illumination sources (combined with suitable optics for filtering, collimation and/or homogenisation as required) include LED arrays, flash lamps and CW arc-lamps. Non-laser sources may be especially suitable 25 for PV cells and modules composed of direct band gap semiconductor materials that are much more efficient PL emitters than indirect band gap semiconductors. In further embodiments of the invention, local shunts are detected by the comparison/analysis of luminescence images of cells, typically at the start of a module 30 manufacturing process, i.e. incoming cell inspection, but alternatively at the end of a cell manufacturing process, taken at two or more average carrier injection levels generated for example using different illumination intensities (PL), different bias -9 voltages including reverse bias (EL), or different electrical bias conditions under illumination (biased PL). Based on a single image, shunted areas cannot be easily distinguished from areas of poor minority carrier lifetime. Our comparative method therefore exploits the different average injection level dependence of local voltage 5 variations in areas surrounding a local shunt compared to a local low lifetime region caused for example by a high dislocation density. The case where PL images are taken at different illumination intensities at open circuit is considered in more detail here. Regardless of whether the cause of contrast in a PL 10 image is due to a crack, local shunt, local low lifetime region etc, a PL image of a cell taken at higher illumination intensity will appear sharper than a PL image taken at low illumination intensity, because the effect of the emitter in sharing current between adjacent regions of a cell is reduced at higher intensities. An emitter layer is resistive in operation, so that the lateral current flowing in it is linear in the voltage drop across 15 it. However at open circuit the voltage of a solar cell is logarithmic in the incident light intensity, whereas the recombination current (dark diode current) is linear in the incident light intensity. This means that as the intensity of illumination is increased, the relative magnitude of any current added to a local dark diode from a neighbour via the emitter is reduced. In the case of a local shunt, the current through the shunt is 20 linear in the local voltage, which means that the shunt current can at most be proportional to the logarithm of the illumination intensity. Therefore as the illumination intensity is increased, there is a differential sharpening (reduction of the size of the effected region) in the case of a shunt-induced local low-voltage region compared to a low lifetime-induced local low-voltage region. 25 In still further embodiments of the invention, luminescence-based series resistance imaging of cells, tabbed cells, strings, lay-ups or modules can be used to identify series resistance faults such as poor electrical connection to cells based on incorrect soldering conditions, or inadequate cell bus bar metallisation. These types of imaging measure 30 ments are described for example in the abovementioned published PCT application No WO 2007/128060 Al. Series resistance imaging may also be performed on an entire module to identify poor connections within the module. Series resistance imaging can - 10 be performed with any existing luminescence-based series resistance technique, e.g. homogeneous illumination and operation of the module at the maximum power point. Alternatively, purely EL-based algorithms that rely on more than one image taken un der different bias conditions can be used. 5 Yet further embodiments of the invention comprise a sequence of tests involving two or more of the previously described embodiments. Once a fault has been identified, there are several actions that can be taken. For 10 example if a defective cell is identified early enough, it can be removed and replaced. Alternatively the cells can be connected in a different sequence, say to by-pass a defective cell. Modules known to be compromised by defective cells can be assigned to lower quality bins. On many occasions, an identified fault will be an isolated occurrence, for example a wafer with a shunt caused by a large inclusion, or a crack in 15 a particularly fragile wafer. However on other occasions a particular fault may be identified in many cells or modules, indicating a systematic problem that needs to be corrected. For example a large number of defective in-coming cells would suggest a problem at the cell manufacture stage, while a large number of cells cracking during module assembly may indicate a soldering or mechanical handling problem. 20 Although the present invention has been described with particular reference to certain preferred embodiments thereof, variations and modifications of the present invention can be effected within the spirit and scope of the following claims.

Claims (18)

1. A method for detecting failure modes in a process for manufacturing a photovoltaic module, said module comprising a plurality of electrically connected photovoltaic cells, said method comprising the steps of: 5 (a) acquiring one or more luminescence images of at least one photovoltaic cell associated with said photovoltaic module; and (b) processing said one or more luminescence images to identify the incidence or growth of a defect in said at least one photovoltaic cell, wherein at least one of said luminescence images is a photoluminescence image. 10

2. A method according to claim 1, wherein said one or more luminescence images comprise two or more luminescence images of said at least one photovoltaic cell acquired at different positions in said process, and wherein said method further comprises the step of: 15 (c) comparing said two or more luminescence images.

3. A method according to claim 2, wherein said two or more luminescence images comprise images acquired before and after a particular step of said process. 20

4. A method according to claim 3, wherein said particular step is selected from the group consisting of tabbing, stringing, lay-up, interconnection of strings, and lamination.

5. A method according to claim 1, wherein said one or more luminescence images 25 comprise two or more luminescence images of said at least one photovoltaic cell acquired at the same position in said process, with different excitation conditions, and wherein said method further comprises the step of: (c) comparing said two or more luminescence images. 30

6. A method according to any one of claims 2 to 5, wherein said comparing step comprises a pixel-by-pixel generation of a difference or ratio image. 42

7. A method according to any one of the preceding claims, wherein the at least one photoluminescence image is generated by laterally inhomogeneous illumination of said at least one photovoltaic cell. 5

8. A method according to any one of the preceding claims, wherein the photoluminescence in said at least one photoluminescence image is additionally generated by application of an electrical bias to said at least one photovoltaic cell.

9. A method according to any one of the preceding claims, wherein said at least 10 one photovoltaic cell comprises individual cells before tabbing, individual tabbed cells, individual cells incorporated in strings, several cells within one string or entire strings, single or multiple cells within lay-ups, and single, multiple or all cells within a completed module. 15

10. A method according to any one of the preceding claims, wherein said defect is a crack.

11. A method according to any one of claims 1 to 9, wherein said defect is a shunt. 20

12. A method according to claim 10 or claim 11, wherein said method further comprises the comparison of two or more luminescence images acquired at different average carrier injection levels.

13. A method according to claim 12, wherein said different average carrier 25 injection levels are generated by different illumination intensities, different bias voltages including reverse bias, or different electrical bias conditions under illumination.

14. A method according to any one of claims I to 9, wherein said defect is a series 30 resistance fault. 43

15. A method according to claim 14, wherein said one or more luminescence images comprise a photoluminescence image generated by homogeneous illumination and operation at the maximum power point of said at least one photovoltaic cell. 5

16. A method according to claim 14 or claim 15, wherein said method further comprises the comparison of two or more luminescence images acquired under different bias conditions.

17. An apparatus when used to perform the method according to any one of claims 10 1 to 16.

18. An article of manufacture comprising a computer usable medium having a computer readable program code configured to conduct the method according to any one of claims I to 16, or to operate the apparatus according to claim 17.