Solexant is the latest thin film solar startup to line up some nice government subsidies for building a new factory. I wrote about its basic plan for a 100-megawatt plant near Portland, Ore., in today’s DailyFinance.
Before I wrote the story, I asked Solexant’s CEO Damoder Reddy for more information about the company, particularly since I saw his presentation at a Dow Jones conference last December. At the time, Damoder said his startup had just completed a 2-megawatt pilot line and raised two rounds totaling $22.5 million. The company had 34 employees plus six contractors.
The company, which has since raised a new round of $41.5 million, is using a rather novel technology first developed at the Lawrence Berkeley National Lab. The process involves printing cadmium-telluride nanocrystals on rolls of flexible metal foil. This method is different from the cadmium-telluride solar panels made by First Solar, which deposits cadmium-telluride on sheets of glass. The roll-to-roll process is supposed to be faster – and presumably cheaper – than putting materials on pieces of glass in a conveyor belt-kind of process.
I asked Damoder by email about the size of the panels rolling out of the pilot line and their efficiencies. A pilot line is where you test your manufacturing process to make sure it’s sound enough for you to start mass production. It’s also where you assess whether the solar panels could hit the efficiencies you were expecting when you were in the research and development stages.
Damoder was coy. Well, he was willing to divulge that the pilot line is producing 1.2 square-meter panels. But he wouldn’t tell me the power rating of the panels or any efficiency numbers. He wrote that the factory in Oregon would roll out 2-square-meter panels with over 200-watt of power ratings after it starts mass production (a factory takes time, usually about 6-9 months, to go from initial production to mass production).
At the Dow Jones conference, Damoder said he was expecting the pilot line to produce panels with more than 10 percent efficiencies. He also showed a chart projecting some sort of savings if a project developer decided to pick Solexant’s panels over First Solar’s. The chart showed that a 10-megawatt power project would require 106,000 First Solar panels but only 33,660 Solexant panels. The cost would be $1.5 million versus $900,000.
During the first quarter of this year, First Solar was churning out panels with 11.1 efficiency, at a manufacturing cost of $0.81 per watt.
COUNTING ELECTRONS 
First Solar’s most efficient module, the FS-280, is 80W, so a 10MW system requires 125,000 First Solar panels, not 106,000. Based on what you are telling us, Solexant’s module is 200W, so a 10MW system requires 50,000 modules, not 33, 600.
A 10MW system cannot cost $1.5 million, much less $0.9 million. If it did, we would have been at grid parity a long time ago, as it implies 15c per Watt (or 1c per kWh).
Of course, all this is a moot point. To be competitive, Solexant’s cost of manufacturing need to be below $1.10 per Watt. It isn’t. So Solexant is just another company that will run out of money soon.
Solexant said its commercial panels would be rated at more than 200-watt. Didn’t disclose the exact number.
Thanks for the cost calculations. Should’ve pointed out that the figures probably contained assumptions not stated in the chart. The company said it could do less than 10 cents per kilowatt hour. But that was the projection last year.
By the way, why do you say that Solexant needs to manufacture at below $1.10 per watt but Solar Frontier needs to do it at $0.90 per watt?
First, Solexant’s module efficiency cannot be much more than 10% – I would say 11% at most. The reason: flexible modules have 70%-80% of the efficiency of glass modules, for any given chemistry, and since First Solar’s most efficient modules are less than 12% efficient currently, even 10% is pushing it for Solexant. For 10MW to result in 33,600 modules that have 2 square meters area each, Solexant’s efficiency needs to be 14.8%, which is currently impossible for any CdTe commercial module.
If Solexant can produce flexible modules, then this gives them a 30c per Watt advantage. See, flexible modules can be simply glued to a membrane roof, without the need for any rack mounting systems, which results in about 30c per Watt savings on the balance of system costs (that saving can be between 10c to 50c per Watt depending on the type and size of the system, but 30c per Watt is a reasonable average number). That is why an 11%-efficient flexible Solexant module needs to be made at about $1.10 per Watt (81c + 30c) to be competitive. A 12%-efficient non-flexible Solar Frontier module needs to be made at 85-90c per Watt (as the 1% better efficiency vs First Solar results in 5-10c per Watt savings on the balance of system costs) to be competitive.
Thanks for explaining the numbers — really helpful.
I don’t think we can assume that Solexant’s modules will be the flexible laminates that you refer to. In his presentation at Dow Jones, he mentioned the use of glass as the front sheet. The setup sounds like what CIGS companies are doing — they still use glass on the front, or even glass-glass encapsulation, depending on what they customers want. Damoder also said there would be no need to cut the cells into strips and then reassemble them for modules, as some CIGS companies do after the roll-to-roll process.
If Solexant uses glass, then they won’t command the 30c premium – required cost of manufacturing pricing goes back to First Solar’s “benchmark.”
Not all CIGS players use glass. SoloPower, Ascent, and GlobalSolar for example all make flexible CIGS modules – unfortunately, neither one of them has achieved either UL or IEC certification (and neither one of them has the cost structure to compete in today’s PV market).
Roll-to-roll vs. batch, monolithic vs. reassembled cells, CIGS vs. CdTe – all these things are irrelevant (as for each process there are examples of companies that have competitive costs of manufacturing and companies that will fail to ever achieve competitive costs). All it matters is the cost of manufacturing per Watt adjusted for efficiency (meaning, a 10%-efficient module needs to be manufactured at a cost of 5-10c per Watt below an 11%-efficient module), and then adding the 30c per Watt credit for flexibility, if required.
Ah, I didn’t mean to say that CIGS companies that use glass can’t do flexible modules. But those that can make flexible modules have been selling the glass variety instead, partly because there are no good flexible front sheet materials to replace glass. I thought Global Solar doesn’t make modules but sells its cells to module makers (or it could hire module makers), and its May announcement about that big installation in Italy used modules with the glass front sheet. But Global Solar is looking into doing its own packaging.
SoloPower claims to be shipping flexible modules in “sample quantities to several key customers in different geographical areas, with general availability of the new modules expected in the second half of 2010:”
http://www.pv-tech.org/news/_a/solopower_launches_flexible_cigs_line_ships_modules_ramps_production889/
Ascent’s customer FTL claims to be shipping Ascent modules aka “chargers” (although, they are showing as “currently unavailable”):
Global Solar claims to be shipping flexible “chargers” as well:
http://globalsolar.com/index.php?option=com_rubberdoc&view=doc&id=47&format=raw
None of those guys can actually ship any of their flexible modules for the rooftop market yet, as they are not certified.