AUTOMATE OPTICAL INSPECTION FOR THIN FILM SOLAR A crucial step in the quest for grid parity

Solar panel manufacturing has grown at an astounding pace: 33% on average since 2002 i. It is projected to continue to grow by 8% per annum from 2012 to 2017 ii. Achieving grid parity, without special government incentives, has been the goal of the solar industry. Grid parity is the point at which the generation of electricity from solar power produces power at a levelized energy cost that is equal to or less than the price of purchasing power from the electrical grid.
A dramatic reduction in the cost of solar panel manufacturing took place between 1990 and 2010, when the price of electricity from solar power dropped about 25 times. The price continues to drop; between late-2009 and mid-2011 the wholesale cost of solar modules dropped approximately 70% iii. This rapid reduction in cost shows no signs of abating and improved manufacturing processes will be a major contributor to the reductions in cost.
Photovoltaic modules use solar glass for protection, performance and as a substrate for thin film modules. With glass panels being a major component of thin film solar panels, optimizing current glass manufacturing processes provides an opportunity for additional cost reduction and improved panel efficiency.
Most solar glass manufacturers currently inspect their glass manually with human inspection. Some do no inspection at all. This means that reduced quality glass is shipped to solar panel manufacturers. The impact on the solar panel manufacturer is lost production due to glass breakage and other complications, resulting in higher warranty claims and reduced panel efficiency.
The use of Automated Optical Inspection (AOI) for solar glass manufacturing provides major benefits to both the solar glass manufacturer (processor) and the solar panel manufacturer. Solar glass processors face a number of challenges to ensure that the end product meets not only customer specifications, but also pricing requirements as the industry continues to move toward grid parity. Defects in the glass come from a number of sources: either they were not detected at the float when the glass was produced or they were produced from transporting the glass to different factories or they came from the processing equipment. In addition, some glass has special features such as drilled holes or special shapes, such as a corner cut (dub) that requires verification. Another challenge is to be able to accurately measure the size and also the shape (e.g.: square) of the panel.

There are many opportunities on glass production lines to deploy AOI systems. In extreme cases, even where incoming glass quality is poor, AOI has increased production line yields by over 10%. In this case, AOI systems can be justified at the very beginning of the production line for surface inspection of various float glass defects, such as stones, bubbles and also scratches. Typically, the first inspection point would be after cutting, grinding and drilling. At this point, there is the opportunity to inspect for all surface defects as well as inspection of the edge of the glass for grinding defects. Inspection systems can detect and, most importantly, classify a variety of defects, including float glass defects and process generated defects. Examples of these defects include, bubbles, stones, scratches, edge chips, shiners produced by poor edge grinding, missing holes, chips on holes, etc. Typical specifications for point defects are 0.50 mm but the requirement seems to be moving toward 0.20 mm, which is barely detectable with the naked eye. One method used to objectively size scratches is based on the ASTM standards, with ASTM F428-5 being the typical acceptable requirement but again, there seems to be a movement to ASTM F428-3.
The heart of an inspection system is the Classifier, the system’s ability to differentiate the various types of defects and to size them accurately and consistently. However, the image of the glass captured by the AOI system is just as important since this is the information that the Classifier uses to determine if there is a defect and of what size. Typically, Bright Field configurations work better in an industrial environment than Dark Field. Bright Field is defined by the glass blocking the light; there is a corresponding intensity drop on the CCD of the camera and a defect is flagged. Bright Field provides much more information about the defect, which allows for better classification. Bright Field with collimated illumination is the preferred optics and illumination arrangement for clear solar glass inspection. At this point, the size of the glass can also be measured and, if holes have been drilled, they can be measured for size and position. There are two major issues that need to be considered when implementing a precision measurement system for solar glass panels: the stability of the glass transport system (the conveyor) and the imaging angles of the image of the glass panel, which also affect proper sizing. Most standard conveying systems, using line scan cameras, achieve an accurate measurement of the glass panel at right angles to the conveyor, but tend not to be reliable with measurements running parallel to the conveyor. Therefore, if accurate glass measurement and shape is important, the use of area scan cameras as opposed to a line scan camera is a must for this function. Since directional illumination is used for measurement, telecentric imaging ensures that the measurement is taken from the correct spot and there are no compensation algorithms required, which are typically a source of error. By implementing these techniques, it is possible to achieve measurement accuracies of +/- 0.1 mm.
Other important requirements of an AOI system include the ability to identify clusters of defects, each of which may be smaller than the individual specification for that defect but which, when viewed together, are very evident and should be rejected. This type of defect is usually a repetitive defect that is process induced and if not detected, could result in hundreds of defective panels to be processed. The same applies to repetitive defects, even if they are under spec. If the production line has multiple legs, then the AOI system must be able to accommodate multiple panels side by side. Many parts or models can be processed on a production line. Therefore, a system must be able to store model definitions with all the corresponding specifications for that model and, when the model is selected, make all the necessary system adjustments for that model. Of course, the ability to store all of this data and to produce reports to ship with the product is a valuable by-product of AOI.
Once other value-added processes, such as tempering or coatings, have been applied to the glass, another inspection can be performed before shipping to the customer to ensure that there are no defects. This can facilitate automatic packing.
The use of AOI in solar glass manufacturing processes ensures that the part per million defect specification can be met by the solar glass processor. Applying human inspection is just not consistent enough to meet these objectives. In addition to this primary benefit, AOI is also a process management and yield management tool that can positively affect the profitability of the glass processing factory. The AOI system is one of the best data collection points for managing the glass processing environment. Through the collection of this data, the manufacturing process can be monitored in real time, avoiding costly product runs that do not meet the specification. For instance, there are real time tools that track all the production steps of the glass process. A specific example of this is a tool that tracks two separate production legs of solar glass. The four grinders for each leg are monitored and the defects generated from each of the grinders are tracked in real time and displayed on the monitor. The same applies to the two different drills for holes for each leg and the dub grinding for each leg of the process. In addition, typical defects such as inclusions, distortion and scratches are being tracked. The additional benefit of such a tool is that this information can then be interfaced to the company’s production monitoring system, allowing the company to compare the performance of their factories.
AOI systems in glass manufacturing plants produce a number of important benefits to the solar panel manufacturer. High quality glass reduces breakage in the production process which means higher production availability. Higher quality glass will reduce warranty claims due to breakage in the field. AOI clearly represents an opportunity to reduce costs and improve solar panel efficiency.
Stephane Lemieux is the President and CEO at SynergX Technologies Inc. SynergX Technologies is a global AOI Systems supplier to glass manufacturers in the solar, automotive, appliance, and architectural industries. Based in Laval, Quebec, the Canadian firm has sales and support offices in Europe, United States, China, South Korea and Japan.

i. L. Setar & M. MacFarland. “Top 10 Fastest-Growing Industries”. IbisWorld Inc. 2012. Web. 29 June 2012.

ii. Setar & MacFarland.

iii. a b K. Brankera, M.J.M. Pathaka, J.M. Pearce. "A review of solar photovoltaic levelized cost of electricity", Renewable and Sustainable Energy Reviews, Volume 15 Issue 9 (December 2011), pp. 4470-4482