top of page

ACAP researchers develop 'gold standard' tech for diagnosis of defects in solar utility plants using daytime photoluminescence imaging


UNSW Research Fellow and ACAP grant recipient Dr Oliver Kunz has a vision. He wants to see solar panels reliably generating electricity for 30-50 years or more after installation. And, he says, crucial to this endeavour is effective and efficient quality control in the field.


To this end, Kunz and his team have developed a way to use photoluminescence (PL) imaging technology for solar plant surveillance in full sunlight without any contact or interfering with the plant’s operation – something that’s never been possible before.


Diagnosis of defects in solar utility plants using daytime photoluminescence imaging with a drone
Dr. Oliver Kunz with a custom drone taking daylight photoluminescence images on a utility scale solar farm. The first stage of development of the drone-based imaging was made possible by small project funding under ACAP.

PL imaging uses the emission of infrared light from a semi-conductor material to create a rich and detailed view of features in the material. The emitted light is the photoluminescence. It is produced when electrons in the material become excited by a light source and then relax back to their lower energy state. In the case of silicon solar cells, this light is not visible to the human eye.


The brightness of the emitted light is directly related to the quality of the material. Defects and degradation show up as darker spots in the images and expert analysis can determine the root cause of the problem.


PL imaging is considered the gold standard for assessing defects and degradation of solar panels.



Several thousand solar panels can be assessed with one image


Daytime Photoluminescence Imaging (DPL) of crystalline silicon photovoltaic panels is now possible on an entirely new scale.


Professor Thorsten Trupke (UNSW) says, "Using active control of system inverters our UNSW team succeeded with high quality luminescence image acquisition in full daylight of residential roof top systems and in utility scale solar farms, the latter from aerial drones."


"Single DPL images can contain up to several thousand modules, providing unique insight into module quality variations, including from degradation."


Diagnosis of defects in solar utility plants using daytime photoluminescence imaging
Daylight PL image of a section of a utility-scale solar farm showing approximately 1200 modules. Modules in the top section were connected to a separate central inverter. White arrows point to two specific modules that underwent additional DPL testing using UNSW's string modulation technique. (From https://onlinelibrary.wiley.com/doi/10.1002/pip.3807)

Detecting problems early is crucial to keeping solar plant costs down and ensuring the reliablity of solar panels on the market


Surveillance of solar farms for early detection and diagnosis of faults or degradation is crucial to speeding up the feedback loop and ensuring reliable, leastcost system power output, and the supply of panels that will go the distance.


The largest solar farms are now at gigawatt scale with several million modules installed. Global demand for panels is huge, and the pace of technological development is rapid. But this breathtaking pace of development and exponential scaling of production comes with risk.


“The current high efficiency PERC silicon technology will soon be replaced by even higher efficiency solar cell technologies that have been developed: TOPCon and heterojunction (HJT) solar modules.


“They’ll be rapidly rolled out at an industrial scale and the transition will happen in the next two to three years,” explains Kunz.


“With such a pace, you don’t have time to thoroughly test large quantities of all these new module types for years in the field.”


Modules are put through accelerated stress testing over a few months, but some things only go wrong with time or under broader environmental conditions. Logistics, manufacturing methods and supply chain pressures can also impact the quality of the modules. Solar manufacturers are contracted to supply high volumes of solar modules and may be forced to compromise on their choice of suppliers and the quality of production materials.


Kunz says, “We have seen panels failing after a short time.”
“If there are any problems with solar modules in the field, we need to detect them as early as possible and then inform the industry that something needs to be changed.”

PL imaging can identify: microcracks that are not visible to the naked eye; defects or impurities in the crystal structure of the solar cell; series resistance effects; and the presence of various degradation mechanisms. PL imaging can also be used to monitor degradation of the solar cells over time – helping the identification of modules that are degrading prematurely.


It’s worth noting that PL imaging was first successfully used on large area silicon wafers in 2005 at UNSW. Since then, it’s become a core enabling technology in solar PV R&D and is used at most PV research institutes and leading solar cell manufacturers.

Until now, though, PL imaging of solar cells has only been possible under laboratory conditions, usually using lasers and specialised filters in a darkroom. But sunlight is thousands of times brighter than the photoluminescence from the solar panel so it’s not possible to use the same technologies in the field in daylight.


To achieve outdoor PL imaging of installed PV modules during the day, Kunz and his team have developed a method that uses specific optical filtering and other techniques to tease out the tiny PL signal from the modules under bright daylight conditions.



It's fast, very informative and it doesn't impact the plant's power output


Another breakthrough has been to do this without the need for touching any of the electrical connections. The use of an optical modulation technique on single modules or strings of modules was paramount to this. Another method the team developed uses ultranarrow bandpass filters that image solar panels in a sub-nm wavelength bandpass channel where virtually no light from the sun reaches the earth’s surface, but where PL emission from solar panels is high.


Using these techniques from elevated platforms or drones significantly speeds up inspection times and allows surveillance of large sections of solar farms in a very short amount of time. And the inspection only minimally affects the power output of a solar farm.


If something interesting is picked up, then more detailed inspection can be carried out from the ground or via near-field drone-based inspection.


Large scale module inspection of solar farms typically involves drone-based infrared thermography which detects infrared heat emissions, but it has limitations. Infrared imaging cannot detect many electronic defects in solar panels and, in many cases, it does not have the image quality required to carry out root cause failure analysis.


Detailed testing of installed solar modules has previously involved removal and transport for testing offsite, or costly in-the-field electroluminescence imaging techniques performed at night-time.


Kunz says, “[In electro-luminescence testing] operators typically unplug strings of solar panels at night and then plug in generators and push current into the strings, which requires quite a bit of equipment, and personnel with very specific training.”


“In contrast, our outdoor technology is more efficient to use and does not require changes in system wiring or working at night.”


There are a number of commercial applications


The technology is not yet commercialised but future applications in the industry include assessments of the health of solar plants at commissioning stage, for insurance purposes, and when there’s a change of ownership. The technology can also be used for damage assessment after severe weather events such as hailstorms.


Other commercial opportunities arising from the technology include the provision of the monitoring service, analysis of the data, and selling or leasing the hardware to take the measurements.


The team is assessing all of these options in a current follow-on ARENA funded project. There’s also potential for the technology to be used with commercial and residential rooftop systems, although there are tight regulations around the flying of heavy drones capable of carrying the required equipment in built up areas.


In the future, there is potential to scale the application of the technology by combining it with automated piloting of the drones, and machine learning for automated processing and analysis of the data gathered.


Diagnosis of defects in solar utility plants using daytime photoluminescence imaging will assist with quality control of modules in the market


“The longer solar panels last, the more energy you get back from them and the less costly the energy is for everyone. But we also need to rapidly work towards a truly circular economy with as little waste streams as possible,” says Kunz. “The longevity of solar panels is an absolute core requirement of that and we hope our work can contribute to this goal."


This technology development is one of many world-firsts that have been achieved as a result of the establishment and support of the Australian Centre for Advanced Photovoltaics. Continuity of research, through ongoing funding, has ensured Australia maintains an international leadership position in solar research and development, with significant global impact in fundamental science and engineering, in technology transfer and commercialisation.   




Comments


bottom of page