Finding the Key to CIGS PV Reliability
by Joseph McCabe, PE
This past week there was a photovoltaic (PV) workshop that probably wasn't on your radar. It was held at the National Renewable Energy Laboratory (NREL) and is called the PV Module Reliability Workshop (PVMRW). This is where the nerds of the PV industry get together to discuss the factors that influence how long a PV module will last and other factors which might influence the long-term performance of a PV system. It wasn't on your radar because it is not something that influences big business. Or is it?
If you track the PV industry it is more likely that you heard about the February 16th Photon CIS conference in San Francisco where high level people were discussing their company capacities and expected CIS efficiencies. But at the PVMRW meeting, held at the same time, people were discussing the challenges with copper indium gallium and (di)selenide PV (CIGS), which is very similar to CIS. CIGS has the promise of low cost manufacturing with high efficiencies. However many companies are taking quite a long time to develop large markets. CIGS products deposited on glass, like Solar Frontiers (100% subsidiary of Showa Shell [Tokyo:5002]), have had many years of advancements leading to the February 15th announcement indicating commercial production at their newest plant located in Kunitomi Japan.
The Promise of CIGS
CIGS holds the promise of low cost production and of being packaged in a flexible module. This month’s Department of Energy (DOE) SunShot Initiative announcement hopes to reduce PV systems costs by about 75 percent to roughly $1 per watt; flexible CIGS PV modules can provide a large system level price reduction towards this DOE goal. NREL specifically indicated the system level cost reductions could be from $0.17 to $0.94 per watt savings using flexible PV modules instead of traditional rigid glass.
The Hunt for the Culprit
The high efficiency, flexible PV module has been hampered by apparent susceptibility to moisture of the CIGS technology. Potential culprits range from the packaging of the modules that allow for moisture to enter into the PV cells to transparent conductive oxides (TCO). TCO are one of the layers in the CIGS thin film PV module.
At PVMRW companies like Dow Corning, DuPont, Saint-Gobain, Mitsubishi Plastics, and 3M were presenting how their materials can protect the PV product, specifically CIGS susceptibility to moisture. If the culprit causing the susceptibility to moisture inherent in today's CIGS technologies is the TCO, as NREL suggests, these expensive and unproven packaging solutions might not be needed.
Various CIGS companies provided reliability perspectives at the PVMRW. SoloPower, which just announced a conditional commitment for a $197M loan guarantee from the DOE for a new facility in Oregon, presented the effects of light soaking on shunts in their CIGS. Solarion compared reliability of their CIGS in a glass-glass encapsulation to a flexible encapsulation. Ascent Solar (ASTI) presented highly accelerated weathering of CIGS and Nanosolar presented their design for reliability on keeping the water out of CIGS. Companies like ADCO adhesives were supplying reliability information on edge seals and other building integrated PV (BIPV) adhesive attachment solutions appropriate for flexible CIGS.
One company's presentation was quite revealing. Sunpower (SPWRA) had quantified and presented various system failures to help understand reliability from their extensive historical field experience. SunPower's acquisition of PowerLight enabled them to compare various manufactures’ products over a number of years of performance data. This sharing of system failure data is indicative of the spirit of this unique PVMRW meeting. Our industry is learning from each others failures so that the industry as a whole will prosper. Just a note, SunPower’s modules were not necessarily those included in the system failures, but other manufactures modules.
Reliability is Location-Specific
For the first time I was hearing multiple discussions for location specific reliability evaluations. All modules are currently required to pass a set of tests that help build confidence in the safety and potential performance of the PV over time. However, there is not necessarily a correlation of those tests and the actual longevity of the PV product. It has only been assumed that these tests can represent a high probability of long-term performance. The tests reflect a general understanding of failure mechanisms for a relatively hot-humid location. New location specific reliability testing can open up hot dry markets for specific PV technologies, and can help to guarantee performance of PV products that might perform better in cold or humid climates. NREL’s Rommel Noufi suggested looking at today’s highways for what our PV industry might look like in the future. What he meant is that the highways are full of various manufactures and models of transportation solutions, and similarly, there will be many PV solutions for various locations and purposes in the future.
There were three tracks at the PVMRW; crystalline silicon, concentrating PV and thin film. CIGS discussions dominated the thin film track possibly due to the high interest in long term performance opportunities. Kudos has to go to NREL and the DOE for supporting this annual PVMRW workshop. And thanks should go out to the nerds of the industry who have worked for many years to build the current state of reliability in the PV industry. Over the past few decades it is these groups of people that have enabled PV systems to build the confidence in the PV market place that enables more than 20 years of reliable system performance.
DISCLOSURE: No positions.
Joseph McCabe is a solar industry nerd with over 20 years in the business. He is an American Solar Energy Society Fellow, a Professional Engineer, and is internationally recognized as an expert in thin film PV, BIPV and Photovoltaic/Thermal solar industry activities. Joe is a Contributing Editor to altenergystocks and can be reached at energy [no space] ideas at gmail dotcom.