That's not because it's easy, or even because it's a good idea to try, but when well over a billion dollars in investment pours into a given technology, something is bound to come out the other side - eventually. A seductively high efficiency potential technology with very low potential materials costs, CIGS has been just over the horizon for a decade or more, but has enjoyed a huge influx of capital and increase in the number of programs chasing in over the last 5 years. Similar to other solar thin film technologies, device complexity, effective yield, throughput, and process control issues are always the bugaboo.
Given its seductivenes, its somewhat capricious nature, and the siren filled history of the technology, perhaps we should think of CIGS like a woman, and all men need a few rules of thumb to keep in mind before we jump in. Here are mine (for CIGS, not women):
Number one, like most thin film technologies, $100 mm in investment is the ante up to play the game. Just because you spend it doesn't mean you get real product out, and with CIGS, you tend not to know whether anything is workable until oh, say $50 to $100 mm is already spent.
Number two, what you think you know, you don't. Until the pilot plant has been operating for a few years, companies generally really underestimate what they don't know.
Number three, remember those experiments and great idea you sold your investors on, the hard part is not there, the hard (read risky) part is ALL in the "it's just engineering" end of the scale up process you told the investors was "fairly straightforward". This isn't IT, it's deposition with a very commoditized end product.
Number four, whatever the projection as far as timing, add 3 years, maybe 5. I'm not kidding here, I said years.
Number five, when the words "fast", "roll to roll", "reel to reel" or anything else equating to speed in the process are in the pitch deck, translate that to read excruciatingly slow in the development timeline, and lots of "issues" popping up in those nasty yield and process control areas.
Number six, when investing, be very careful about that "yield" number and the "capacity" numbers they made up based on it. All thin film development companies keep "little black books" with the data and charts on every process run they've ever made. Read every single one of those charts, and ask lots of stupid questions about why only 4% of the total square footage produced is above 6% efficiency in run XYZ. Think in terms of "effective total average yield". That's where the problems are hiding.
CIGS watchers have a number of darlings to follow. There's Miasole, which now under new management is rumored to have substantially tightened down its development discipline to take it's shot, Nanosolar, another Silicon Valley venture darling that has been described by many observers along the lines of, "never met hype they didn't like", but with a seductively low cost printable process if they can get it to work, Solyndra, the "stealth" company with the big sign on I-880, Heliovolt, the Texas-based hot CIGS deal of last year, which burst on to the fundraising scene on the back of it's still extremely early stage "FASST" technology. And those are just the largest of the US based venture backed deals, without including Honda, IBM, DayStar, Ascent Solar, Solopower, and literally dozens upon dozens of others around the world with significant backing (though all at a very, very early stage). Wikipedia has a decent cut at a list, though by no stretch of the imagination comprehensive.
My best estimate is that most of the venture investors in each of those deals personally looked in depth at the manufacturing process of single digit numbers of competing approaches before investing. And only read the little black book on two of them. That strategy was tried, with ahem, "mixed" results, in fuel cells a few years back. We'll see how well it works in thin film solar.
And of course, as with most things in solar, the major players should probably be watched more carefully than the startups. I've always liked larger companies to crack thin film issues, in no small part because the term "stage gate" tends to mean something to them.
But my personal favorite for front runner currently is Arizona based Global Solar, a solar company I have been following for years. Their announcement a few months ago of 10% efficiency in production runs, was pretty much lost in the crush of press around solar, for reasons unfathomable to me.
While admittedly not yet proven in a full production environment (they are working on the scale up to 30 MW plants) they do have the massive advantage of having run virtually the only operating CIGS pilot plant in the world - and I believe have shipped more volume of CIGS product than anyone if not everyone else. True to form, that technology, which originally came out of the Tuscon Electric backed ITN Energy Systems labs in Colorado which later did Ascent Solar, has had an estimated $150-$200 mm plus invested in it over the last decade, before Solon AG bought the company for a reported $16 mm. Though to be fair, current management under CEO Mike Gering was brought on well into that process. So while I'll keep my fingers crossed that some one will crack the CIGS nut, and continue to be flabbergasted at the $1 Bil plus valuations estimated to have been achieved by some of the startups named here for very large science projects, when it comes to the one to watch, Global Solar is my personal pick.
Neal Dikeman is a founding partner at Jane Capital Partners LLC, a boutique merchant bank advising strategic investors and startups in cleantech. He is the founding CEO of Carbonflow, founding contributor of Cleantech Blog, a Contributing Editor to Alt Energy Stocks, Chairman of Cleantech.org, and a blogger for CNET's Greentech blog.
AUSTIN, Texas--"It's like a great big grilled cheese sandwich machine."
That's how solar power company
"You have the Gruyere cheese on one side and the Swiss cheese on the other; we make them react with one another and stick on the bread," Langdon continued when CNET News.com visited the company's headquarters in Austin.
HelioVolt's "grill sandwich" pilot line
(Credit: Hanna Sistek)Replace the cheeses with chemicals specifically prepared for the process and the toast with a thin layer of photovoltaic absorber, and you'll have the essence of HelioVolt's manufacturing technique, called FASST. When the pre-cursed materials react, they form the CIGS, a combination of cadmium, indium, gallium, and selenium.
That "cheesy" layer is put on a sheet of weather-sealing glass and then covered with another glass layer. The company claims that FASST is 10 to 100 times quicker than the last step of the manufacturing process used by some of its competitors. This is because the CIGS reaction is a one-step process, instead of first "making the cheeses" and getting them to react after. A solar cell takes six minutes to produce in this fashion.
Last year, HelioVolt raised $101 million in B-round funding, which it will use to build its first factory, with an initial capacity of 20 megawatts per year.
One of the markets for its thin-film cells is building integrated photovoltaics (BIPV). Earlier this month, HelioVolt announced a partnership with construction firm Architectural Glass and Aluminum to enter that market. AGA makes curtain walls--non-weight-bearing facades--for buildings, and HelioVolt will incorporate thin-film cells in the spandrels of AGA's curtain walls.
When it starts production, HelioVolt will begin selling its panels to large solar farms. The company aims to start shipping CIGS panels by the end of the year. The next step is to use BIPV for commercial buildings. The company hopes it will have prototypes for spandrels by next year and products by 2010.
HelioVolt believes that integrated photovoltaics is the future as far as construction materials are concerned, and claims that an average American home would be able to generate 3 kilowatts of electricity during peak hours. This might, in the best-case scenario, be enough to cover the energy needs of the average household.
Solar cell roof tiles
But to get a real breakthrough, BIPV prices need to come down from today's average of $7 or $8 per watt. The price tag for equipping an average American home lies around $20,000, according to HelioVolt. That includes a 25 percent savings from not having to retrofit the solar panels on the roof (the mounting actually makes up half of the cost), the company claims.
The Department of Energy's Solar America Initiative program has set a goal of making solar electricity cost-competitive with grid electricity by 2015. Developing BIPV is a part of that program. Dow Building Solutions, a business unit of The Dow Chemical Company, got a three-year $9 million grant a year ago to do just that.
Dow is developing flexible solar roof shingles as part of the program, and teamed up with CIGS company Global Solar a couple of weeks ago.
Global Solar is ahead of HelioVolt, with production running at a 40-megawatt capacity from their plant in Tucson, Ariz., and another that's expected to produce 30 megawatts should be up and running in Berlin by the end of 2009. Instead of printing, the company uses a co-evaporation technique to deposit the CIGS material onto 1,000-foot-long rolls of thin stainless steel. They can ship the rolls directly to clients, who themselves may cut out the desired size of the solar cells, and then coat them with glass. This gives Global Solar the advantage of not having to ship the heavy and fragile glass themselves, but creates more work for the customers.
Global Solar is not involved in the BIPV creation process, it just sells the PV material to Dow, and is open to more partnerships with building companies. Tim Teich, vice president of sales and marketing at Global Solar, claims that building companies are rushing to team up with solar corporations. "I get approached every single day," he said.
HelioVolt doesn't have any products--such as roof tiles, shingles, or wall coating--ready yet. It didn't even have any test samples of its CIGS cells to show when I visited. The company still has to figure out how to print the photovoltaics on materials other than glass, which can be expensive. The next materials in line are metal foil and plastics.
BIPV is currently being produced by companies like Suntech Power Holdings, or the much smaller Open Energy Corporation, integrating classic silicon wafers into roof tiles. Other players in the area include Sharp, BP Solar, and SunPower PowerLight.
CIGS are cheaper to manufacture than silicon wafers, which today constitute about 90 percent of the solar market. By 2012, the CIGS market share will increase to 25 percent, according to Lux Research.
Solar upstart HelioVolt on Monday will announce that it has reached 12.2 percent efficiency with its CIGS solar cells, setting another mark in the race against competitors and silicon.
Company CEO BJ Stanbery will present a paper at the IEEE Photovoltaics Specialists Conference where he will disclose the efficiency threshold which HelioVolt has reached in its labs. The well-financed company expects it can go much further in converting light to electricity.
The little solar cell that could. HelioVolt says CIGS cells are approaching silicon in effciency.
(Credit: HelioVolt)More significant than the actual number is the fact that HelioVolt hit 12 percent efficiency with its manufacturing process, which it says can turn out a cell in six minutes.
The National Renewable Energy Laboratories (NREL) earlier this year said that it attained 19.9 percent efficiency for a solar cell made out of CIGS (copper indium gallium selenide), an alternative to traditional silicon.
NREL used a technique called co-evaporation where active chemicals are immersed in a solution, which then gets removed. That process can take 40 to 50 minutes, according to HelioVolt.
Another CIGS company, Global Solar Energy, also uses co-evaporation. Earlier this year, it said it broke the 10 percent mark with its commercial products and expects to hit 13 percent or 14 percent this year.
What matters more than the efficiency record is the speed with which solar companies can manufacture. Ultimately, high scale is what brings costs down, said John Langdon, HelioVolt's vice president of marketing.
In fact, when HelioVolt first delivers product at the end of this year or early next year, Langdon said that the efficiency will be between 10 and 12 percent because it's a more efficient process.
"Everybody has known for years that the cost of CIGS film is much less than silicon--something like 3 cents on the dollar. The issue has always been making it fast enough," he said.
Langdon said the company's FASST process is like making a grilled cheese sandwich where two pieces of bread contain different chemicals. Under heat, the two layers of "cheese" bind together to form a cell.
At first, the company plans to put the cells onto a glass substrate for solar panels. Later, it intends to put the flexible cells onto a plastic substrate so it can be integrated into building materials, like roofs or awnings.
Competitor Nanosolar claims an efficiency in the nine to ten percent range for its commercial products which started shipping at the end of last year and higher results in its labs.
For comparison, commercial silicon cells convert light to electricity at about 14 percent to 20 percent. But efficiency doesn't translate into commercial viability.
Companies are pursuing alternative materials to silicon to get around the high prices and demand associated with it. Thin-film cells use less material than traditional cells.
Solar high-flyer First Solar uses thin film solar cells made out of cadmium telluride that are less efficient than silicon.
But working with CIGS is notoriously difficult, particularly manufacturing at large scale.
Another venture-backed CIGS company, Miasole, had to set back its initial plans after technical difficulties and efficiency levels that were only 4 percent to 6 percent.
Solar company HelioVolt and Architectural Glass & Aluminum on Tuesday announced a partnership to produce glass windows capable of generating electricity.
HelioVolt is one of several new solar manufacturers using different materials to produce thin-film solar cells.
HelioVolt's solar cell which it will put into solar panels and embe into building materials.
(Credit: HelioVolt)The company intends to make solar cells for rooftop panels and later get into building-integrated photovoltaics (BIPV), where cells are embedded onto roof shingles, blinds, awnings, or other building components.
The deal with Architectural Glass & Aluminum calls for the companies to design solar-enabled curtain walls, the glass facades on the outside of buildings, or architectural glass in the interior of buildings.
Citing a Department of Energy study, HelioVolt said that solar cells integrated into buildings can produce about half of a building's energy usage.
Last week, another thin-film solar producer, Global Solar Energy, announced a partnership with Dow to make solar shingles.
Another company doing solar-enabled roofing is DRI Energy, a division of a construction company that has developed roof shingles and solar cells that glue onto flat roofs of commercial buildings.
In its coverage, Greentech Media pointed out that BIPV has a number of technical challenges, making the days of power-generating windows a few years away.
Specifically, solar cells typically have a shorter warranty--at 20 or 25 years--than many building materials. Thin-film cells made from CIGS (copper indium gallium selenide), as HelioVolt is making, corrode more in water than traditional silicon cells.
Renewable energy provider EDF Energies Nouvelles has plunked $50 million into Nanosolar and will get access to output from Nanosolar's factories in 2009.
Nanosolar specializes in copper-indium-gallium-selenide (CIGS) solar cells. CIGS cells aren't as efficient as silicon solar cells, but they cost less and can be integrated into building materials. Nanosolar's cells can be printed on thin, flexible sheets of metal.
The company became one of the first to start producing CIGS commercially when it started cranking out production in December. Some other competitors have had to delay production. Right now, only Global Solar in Arizona also produces CIGS commercially.
A few weeks ago, we heard consistent rumors that Nanosolar had been seeking additional investment and has been telling investors that the company is worth around $2 billion. The company hasn't confirmed the $2 billion part (though we've heard it from a lot of people), but this confirms the "seeking investment" part, I guess. In 2006, Nanosolar raised $100 million. Investors include Mohr Davidow Ventures, Larry Page, and Sergey Brin.
EDF serves nine European countries and has 1.4 gigawatts of installed capacity and is building 1.1 more gigawatts.
CIGS looks like it could pay off.
Nanosolar and Solyndra, which both develop copper-indium-gallium-selenide (CIGS) solar cells, are looking at raising additional funds, according to sources, and both companies have put large valuations on themselves.
According to sources, Nanosolar is telling investors it will have a valuation, after another round of funds, of around $2 billion. Solyndra says it is worth $1 billion. Not bad for companies with combined current revenues at the moment that probably would have difficulty rivaling the take of a reasonably located convenience store. Nanosolar just started shipping a few solar cells to customers at the end of 2007, and Solyndra is ramping up toward production.
I haven't confirmed these rumors, and they might be wrong, but they have been consistent.
Nanosolar CEO Martin Roscheisen has said that Nanosolar does not have a term sheet, a document that provides details about business prospects and a funding proposal, at the moment. Roscheisen, however, has not discussed valuation.
Both Nanosolar and Solyndra were contacted for official comment, but no response has yet been received. Companies generally don't comment this early on financial issues such as valuation.
The high valuations seem to be driven by the current dynamics of the solar business. Demand continues to outstrip supply. The shortage of silicon continues to hamper manufacturers of silicon solar panels. CIGS solar panels aren't as efficient for converting sunlight into electricity as silicon panels, but advocates say they will cost less. The active materials in CIGS panels also aren't in dire supply at the moment either.
The love affair that investors have had with First Solar, which makes thin film solar panels with cadmium telluride, also persists, which lends some glow to CIGS companies. First Solar went public at $20 per share in late 2006 and now trades at $207. (Before the recent swoon on Wall Street, it hit $283.) First Solar has also seen tremendous growth in revenue and earnings with each passing quarter.
When it went public, First Solar was valued at close to $2 billion. It currently is valued at $16.3 billion. Some believe the company is overvalued, but those are the numbers.
First Solar, however, was not your ordinary start-up. The company's founders started tinkering with cadmium telluride technology in the 1980s. By the time the company went public, First Solar had already begun mass production.
HelioVolt, which plans on producing thin film solar panels made from copper indium gallium and selenide (CIGS), will build a 20-megawatt factory in Austin, Texas, that will start popping out panels in 2008.
The factory will employ about 150 people. The company then hopes to move into mass manufacturing by the first quarter of 2009 as well as expand production capacity.
CIGS solar panels aren't as efficient at converting sunlight into electricity as silicon solar panels, but advocates say that they cost far less to produce. CIGS solar panels can also be placed on glass or polymer sheets. Ideally, a plastic sheet coated with CIGS solar cells could cover the roof of a giant retail outlet and provide the building with a huge percentage of its electricity.
Getting CIGS into mass production, however, has taken some work. HelioVolt, Miasole, and DayStar Technologies have all experienced delays. NanoSolar, in Silicon Valley, just began producing solar cells out of its new factory.
Although each of the CIGS companies will make the same basic product, they each employ a different manufacturing process. He who comes up with a cheap, reliable way to produce finicky CIGS solar cells will be the winner, say analysts. First Solar, which makes cadmium telluride thin film solar cells, can attribute a lot of its success to its manufacturing process, which it has honed for the last few decades.
Earlier this year, HelioVolt raised $101 million. Investors include New Enterprise Associates and the Masdar Clean Tech Fund, an investment group formed by the government of Abu Dhabi.
Venture capitalists have poured more than $344 million into five CIGS companies in the last few years--Nanosolar, Miasole, Solopower, Solyndra, and HelioVolt.
- prev
- 1
- next
