Search sponsor
Search sponsor
With its annual solar summit event, Greentech Media intermingles the various food groups of the PV and CSP communities for a tasty buffet. The market, technology, manufacturing, project development, and financial comestibles might come in the form of small plates of newsy nibblets or larger entrees that intrigue and challenge the intellectual palate. Like any fine culinary encounter, the conversations between mouthfuls are an important part of the experience. You walk away from the two-day info-sharing and networking feast satiated but not stuffed, unlike the often-painful mental indigestion that happens after a 15-course technical conference gorging or weeklong mega tradeshow binge.
Here’s a sampler of newsworthy appetizers from this year’s event, which took place last week in Indian Wells, CA.
During a lively panel on PV in the industrial age, Canadian Solar’s Alan King mentioned that the company had shipped its first container of modules the previous week from its Ontario manufacturing facility, which is ramping toward 200MW capacity. He also said that CanSolar was shipping an average of a truckload (double-wides or pick-ups?) of modules every hour of every day, 24/7/365, from its plants in China.
Abound Solar also had a milestone of sorts take place the week before the Greentech event. The company’s Mark Chen told me that the CdTe thin-film manufacturer had received “the metal” that will be used to fabricate its process chambers for its second 65MW production line, which should come online and begin ramping in the third quarter of this year.
Abound’s first line is running at an annualized rate of 50MW, with costs in the $1.25-$1.50/W range and module efficiencies distributed between 9.5% and 10.5%, he said. The company’s output was about 30MW in 2010, with an expectation of at least 80MW being produced this year.
Another thin-film representative who sat on a panel with Chen, Solar Frontier’s Greg Ashley, offered interesting head-to-head performance-ratio test data as well as some head-scratching numbers about the Japanese CI(G)S company’s shipments and production ramp.
Not surprisingly, the data representing actual output vs. targeted yield showed the Showa Shell subsidiary’s panels in a good light, hitting the mid- to high 90s and outperforming multi- and monocrystalline c-Si modules by at least 10% absolute and CdTe panels by 5% or more in tests set up at the company’s Atsugi Research Center.
But when Ashley spoke of Solar Frontier’s production ramp and expected shipments from its new Miyazaki fab in southern Japan (which was not directly affected by the recent quake/tsunami in the country), the numbers didn’t quite add up.
After noting that the thin-film firm shipped about 80-90MW of modules in 2010, he updated the status of the 900MW factory buildout, saying that the first “block” of 300MW was running, the second block would come online in April, and the third block would be operational by June, bringing the Solar Frontier’s cumulative run rate—including the two existing production facilities—to 980MW later this year.
When questioned about whether the company would indeed reach that fully utilized 900MW capacity and production output this year, Ashley said it would. Such a feat would mean Miyazaki would then become one of, if not the, fastest and most successful solar fab ramps in history, achieving maximum capacities at high-90s yields in a matter of months and producing at full run rates for a near-term pipeline that had suddenly blew up to nearly a gigawatt.
But the math just doesn’t work. Let’s say, for the sake of argument, that the first block had achieved 300MW capacity by January 1, that the second block does the same by April 1, and the third by June 1, an amortized estimate of the run rate (again at full-tilt boogie utilization) still comes out to between 600-700MW.
I followed up with Bryan Harland, who works the Solar Frontier account for Edelman, and he thought that Ashley may have misunderstood my question. He asked the company what it expects to produce this year, and the number came back as >500MW—still a very high, shall we say, “optimistic” figure, given the givens of large-scale production-ramp (AKA process optimization and yield learning) reality and still comparatively low market demand pull for those CI(G)S modules.
On the crystalline-silicon side of the table, Nick Cravalho of Innovalight spoke about the nanosilicon-ink purveyor’s current status and future plans. The company has seen its efficiency-enhancing screen-printed secret sauce adopted and integrated into cell production by several Chinese manufacturers.
He said that the Silicon Valley firm will announce more customers soon, including some in South Korea and India. The biz dev/marketing VP also offered a glimpse of Innovalight’s roadmap to bumping up efficiencies another couple of notches, with P-type inks used in back-contact area joining the existing N-type selective-emitter-friendly materials this year and combo P/N inks coming onstream in 2012.
But a slide that Cravalho showed during his presentation but later removed from the final “public” PDF presentation revealed a bit of data from a key Innovalight customer—JA Solar—which has been manufacturing with the nanomaterial since Q1 2010, has four lines running with it, and is evidently ramping more ink-stained capacity in the second quarter.
An efficiency distribution chart from a day’s production run at JA showed a few champion outlier devices reaching as high as 19.2%, with the densest concentration in the low-to-mid 18% range. But what was also revealed was a distribution curve in need of tightening up, with a small but not insignificant amount of trailing-edge cells’ efficiencies creeping below 18% into the 17s—numbers that are on par with results achieved without the inks.
Floating far downstream from the production floors, Southern California Edison’s Jessica Ritchey said the utility’s program to place PV systems on commercial rooftops and in some ground-mounted locations had 11 installations (28MW) completed, with another five projects (18MW) under construction and a total of 25MW planned for this year.
She also shared a laundry list of lessons learned from the project development side of the program. The rooftop can be a challenging environment to install a PV array, with wrinkles in the membranes, puddling and pooling of moisture, and potentially dangerous hatch locations all potentially thwarting the smooth running of the operation.
SCE also found out through the experience of developing and building the first few megawatts in the program that there were many improvements that could be made to the design of system components, according to Ritchey.
Design modifications to simplify the interconnection architecture, combiner boxes, raceway design, master fuse box or recombiner, fuse clips, cable access in the inverter enclosure, and transformer have been implemented and deployed on the recent installations.
For example, the original combiners had no blown-fuse indicators, required a full inverter shutdown when undergoing maintenance, and every string disconnect had to be opened for work to be done on the boxes. The improved design features the missing indicators and a window in the door of the box for better visibility, and incorporates isolation switches.
