Can the number of interconnect wires per cell in asolarpanel be patented? According to Japan's patent office, the answer is yes.

Panels from the 1970s and before often had just one flat wire connecting each small cell. For decades afterward, a configuration based on two busbars dominated the industry. Then, around 2004, the concept of using three busbars started to gain market share, with Kyocera and Mitsubishi leading the way. Mitsubishi jumped to four busbars in 2012, and a few others such as Canadian Solar have since followed suit. In a parallel development, around 2002, Day4Energy invented the first wire array solution where a much larger number of round wires connects the cells, although the firm never scaled the technology to large production volumes. 

Other wire array solutions were later developed by Evergreen Solar (multi-wire), Schmid (multi-busbar), and GT Advanced Technologies (Merlin). Meyer Burger has recently had some success selling the rebranded Day4Energy technology under the name SmartWire. Last year, GTM Research published a report detailing the motivation and cost analysis for increasing the number the number of wires. Reasons include:

  • Greatly reducing the amount of silver needed in the fingers
  • Improving the robustness of the modules should cracks occur in the cells
  • Reducing the resistive power loss in the wires
  • Reducing the effective wire shading losses when using round wires

It’s hard to believe, but somehow Kyocera was granted a patent in Japan for solar cells incorporating three busbars. Kyocera is now entering into litigation against Q-Cells, which has been selling modules in Japan with the offending cell configuration. Since most crystalline silicon PV companies now use three-busbar designs, future litigation may follow. 

On one hand, litigation against patent infringement could be considered a good thing for the industry. Why invest in R&D if IP theft is rampant and goes unpunished? The problem here is that the patented idea seems to fall under the “obvious” category, and this author found multiple examples of what might be considered prior art published before the priority date of the Kyocera patent application. It is telling that the equivalent patent applications in other regions of the world were not granted. But it is still an uphill battle fighting against an issued patent from a giant electronics company in its home country. 

So what are the options?

  1. Ignore Japan, one of the world's top markets.
  2. Revert to two-busbar designs. This is a clear step backward in terms of module performance and silver paste costs.
  3. Move forward to four- and five-busbar designs. Stringers from vendors such as Meyer Burger and NPC are now available to solder four or five flat wires onto each cell, and these designs enable lower silver costs.
  4. Stick with three busbars, but use light-trapping interconnect wire developed by 1366 Technologies. Due to the grooved surface of this wire, most of the light that hits the wires is reflected at steep angles back down to the cell, reducing wire-shading losses. A key point is that the Kyocera patent is only valid for wires with widths between 0.5 and 2.0 millimeters. Typical wire width is ~1.5 millimeters, and so going beyond 2.0 millimeters would normally increase the shading losses. However, with light-trapping wire, the penalty in efficiency for going to 2.1 millimeters is very small, and in so doing, the module power can be increased by more than 2 percent relative (resulting in 6 extra watts from a 300 watt module). Ulbrich Solar Technologies and Schlenk both manufacture variations of this wire, and relatively simple modifications to stringer soldering equipment can enable use of the wires. 

Source: Ulbrich Solar Technologies

  1. Similar to the light harvesting solution shown in the figure above, a technique developed by the University of Stuttgart's Institute for Photovoltaics covers standard wires with white paint, which can scatter some light back to the cell surface. Similarly, placing 3M's light-trapping films on the wires could have the same effect.
  2. Wire array solutions have the best potential in the long term, and it will be interesting to see which of the variations emerges as the strongest in terms of materials costs, “stringer” equipment costs and uptime, mechanical yield during module assembly, module power, module energy delivery, and module reliability. 
  3. Move on to narrower fingers (gridlines). The Kyocera patent is only valid for fingers with widths between roughly 0.05 and 0.1 millimeters. Advanced metallization techniques such as stencil printing, inkjet printing, and plating have the potential to produce narrower fingers.
  4. Cut the cells in half. Mitsubishi has again led the industry in this shift for mainstream-sized panels, with other companies such as REC swiftly following suit. The Kyocera patent is limited in scope; in this case, the width of the cell must be >100 millimeters.  Cutting a 156-millimeter cell in half can help companies get around this claim, and this approach has additional benefits as well. Since the current running through the wires is lower, the resistive power losses in the wires are also lower, and the wire thickness can be reduced to save on wire cost, and more significantly, to reduce the amount of soldering-induced damage imparted to the silicon. This reduction in microcracks under the wires can substantially reduce the likelihood of cracked cells in the field. 

Ultimately, the Kyocera litigation could be just the motivation that the industry needs to adopt new technologies that will boost module performance and reliability.


Andrew Gabor is the founder of Gabor Photovoltaics Consulting and a consulting analyst to GTM Research.