Search sponsor
Search sponsor
The 37th edition of the IEEE Photovoltaic Specialists Conference shattered previous records for participation and presentations, with more than 1800 attendees from 40 different countries and over 1000 oral and poster papers—more information and networking opportunities than any person with half a life could possibly digest. In the final blog on the annual PV techs-travaganza and inadvertent celebration of the summer solstice, I wrap things up with some solar short takes from the voluminous program offerings and the small-but-mighty exhibit hall.
Although most of the commercial exhibitors were purveyors of analysis, metrology, characterization, and test equipment and services, there were some notable exceptions. The leading US PV company, First Solar, made a rare expo-hall booth appearance, with one thing on its corporate mind: recruitment of fresh blood. Company recruiter Megan Adams told me she had talked to a lot of promising prospects, both undergraduate and graduate students on the verge of seeking/needing gainful employment and more veteran types checking their options for a possible move.
First Solar was also the source of a compelling utility-scale power-plant case study, a presentation by Alex Panchula about Sarnia 20’s performance during its first year of operation. He noted that although the actual performance of the southern Ontario 20MW(AC) site was 2.1% above the predicted performance ratio modeling—at 81.8%—the model used suffered from an inability to accurately account for the white stuff so prevalent in those climes during the winter months. In one case, a huge snowstorm knocked the performance ratio down to 48%—literally off the chart that Panchula showed.
“The random nature and large variation during the winter makes snowfall a particularly difficult parameter to model/forecast,” he explained in his paper. “As PV power plants collect more data on snowfall events, long-term averages should improve, although the year-to-year variability will remain.” During his presentation, he also called for more industry cooperation in the area of data collection and analysis on the effects of snow and soiling on PV system performance.
Another paper from the University of Arizona presented a decidedly smaller scale but longer-term, more wide-ranging set of performance data than the First Solar study. The team showed various performance data from 20 different fixed 1-2KW PV module arrays deployed in the Tucson Electric Power test yard, some installed as far back as 2003, others hooked up as recently as 2010.
Unlike some reports that leave out the actual brand and model number of the panels, the TEP study names names, including some companies, like Astropower [corrected] and Solarex, that are no longer in business. The study looked at several different parameters, including “measured conversion efficiencies, final yields, performance ratios, temperature de-ratings, and degradation rates” for the systems.
One finding that caught my eye was the degradation rates, with and without irradiance data, of the older modules that had been under sun for several years. This seasoned crew includes BP Solar crystalline- and amorphous-silicon units, Uni-Solar multijunction a-Si laminates, a pair of different Sanyo HIT a-/c-Si panel set, a group of 40W Shell CIS plates, some vintage GSE CIGS, and Solarex a-Si.
In addition to showing that the annual degradation percentage and standard deviation both depend somewhat on the inclusion of irradiance data—the number spreads tighten up for the most part—the study revealed a wide range of performance among the modules.
The arrays with the least degradation (factoring in irradiance) over the test period were the Uni-Solar US64 (+0.2 ± 0.7% per year), the Sanyo HIPJ54BA2 (–0.2 ± 0.2%), and the Solarex MST-43MV (–0.2 ± 0.6%). The worst performers were the BP MST-50 a-Si (–4.5 ± 0.3%) and the GSE and Shell panes (both at –2.9 ± 0.5%).
In a few years, when ever-growing data sets are collected from newer arrivals like Solyndra, Skyline, Semprius, SunPower, and Prism, the TEP field should become an even-more intriguing source of side-by-side test information.
One curious talk I heard came from the Tokyo Institute of Technology, which discussed research using HAADF-STEM to characterize the interface between the Ag(In,Ga)Se2 and the CdS in an AIGS solar cell. My first reaction, other than trying to sort through the acronyms (HAADF-STEM is “high-angle annular dark field—scanning tunneling microscopy”), was to ask why on earth, rare or otherwise, would someone be looking to develop and commercialize a silver (Ag)-based thin-film solar cell?
The trend in crystalline silicon has been to seek alternative elements to minimize the amount of—or better yet replace—the expensive silver pastes used, such as Suntech’s move toward plated copper in its Pluto architecture. The irony of thin-film PV researchers taking an opposite tack, especially a team from a country known for some rather arcane though fascinating research excursions, was not lost on me.
To be fair, apparently AIGS “is recognized as a promising candidate for the top cell” of a tandem-junction device that would use CIGS as the bottom cell, according to the Tokyo side. The silver-imbued absorber stack “has the same chalcopyrite structure as CIGS, which makes it easier to be fabricated on a CIGS solar cell, and the bandgap of AIGS can be changed from 1.24eV (AgInSe2) to 1.83eV (AgGaSe2) by changing Ga/(In+Ga) (Ga/III), which can satisfy the bandgap requirement for the top cell.”
The problem with AIGS has been low efficiencies, hence the interface study carried out by the institute. As it turns out, the HAADF-STEM images of the test cells revealed that the AIGS film has both silver-rich and silver-poor layers, and that the AIGS and CdS lattice was distorted and the buffer layer had some voids—all of these defects contributed to a poor performing silvery cell.
The folks at EAG are no slouches when it comes to CIGS-related characterization and analysis, though their work relates to the “real” commercial world, where companies are in active development (little r, big D) and/or production modes. Gary Mount had double-duty at IEEE, working the company booth as well as fielding questions in front of a poster presentation discussing the use of surface analysis to determine CIGS composition measurements.
He said that until about a year, year and a half ago, the thin-film crowd sought out EAG for comparative work, but recently accuracy has become much more important, with an emphasis on particular compositional and stoichiometric profiles and related information. Companies want to correlate composition to defect structures, get a better understanding of actual compositions, and understand how certain electrical characteristics correlate to composition. Given the lack of standards for such measurements in the thin-film PV arena, quantification is a key to moving from relative measurements to something more precise and accurate.
The Evans group throws down a serious alphabet soup of nondestructive and destructive analytical techniques: ICP-MS, GI XRD, Auger, STEM, XRF, SIMS, RBS, etc. One of those tools—secondary ion mass spectrometry—has seen an uptick in interest of late, according to Mount.
He said that the issue of contamination control is being taken more seriously, so SIMS is used to look for trace elements, transition metals, and other possible performance inhibitors. Yet not everyone is convinced of the connection of contamination levels to efficiencies—there is no consensus either way on which transition metals matter, for example—and the companies see such “do they or don’t they” correlations as trade secrets, according to the EAG technical specialist.
Although the status of contamination control efforts in c-Si cell manufacturing circles is light years behind the efforts of their cousins in the IC fabrication world, a growing number of concerned PV citizens seem to be connecting the “contamination can lead to poorer efficiencies” dots.
Studies have shown, for example, that iron contaminants can negatively impact minority carrier lifetimes, bringing down conversion efficiencies. Other research has revealed that the wrong cleaning or surface conditioning approach, pre- or postdeposition, can reduce passivation-layer quality.
Yet many in the industry remain in denial. After describing the shockingly filthy state of supposedly “high-purity” wet processes he had witnessed in some major PV cell factories, one insider wondered if the industry really “has a handle on bath chemistry and contamination control,” suggesting that there needs to be a “come to Jesus moment” for the cellmakers to clean up their acts.
A poster from Adi Gildor and his colleagues at Entegris presented a straightforward study of how wet processes contribute to the contamination of silicon wafers. Bluntly put, prediffusion and other wet bath processes are often pretty dirty, with little or no filtration, relying on materials used in the delivery systems that can be major sources of contaminants. The paper offered a model for PV manufacturers to use, which will help them evaluate their own wet processes and the impact of particles and other contaminants found in those baths on their wafer-purity levels—and their efficiencies.
Despite the seemingly critical importance for the PV specialists to drive conversion efficiencies up and bring production and system costs down, sometimes the world outside the solar labs and fabs makes those priorities seem transitory. The empty spaces left by a pair of missing poster presentations at the Seattle conference brought this home.
The absent papers, titled “Libyan Experiences and Future Prospects in Utilizing Photovoltaic Systems” and “Power Degradation of Thirty One Years Stand Alone Photovoltaic Modules,” were written by Ibrahim Saleh and his colleagues at the Renewable Energy Authority of Libya and Alfateh University, both located in Tripoli.
One of the conference management team told me that the two abstracts had been accepted, but there had been no word from the Libyans and no attempt to upload the final manuscripts. It’s unknown whether the researchers were unable to get out of their civil war-torn country or were refused visas to enter the US or a combination of the two. Let us hope these members of the global PV community have remained out of harm’s way.
