Blog Post

Was the DOE Loan Guarantee for Solyndra a Mistake?

President Obama was out touting green jobs at Solyndra’s Fremont, Calif-based factory Wednesday morning, which a $535 million loan guarantee from the Department of Energy is helping to build. The TV news coverage of the event has been calling Solyndra’s factory an example of how successful the Department of Energy’s clean power investment programs can be. But is it really?

The reality is that it’s not yet clear how successful Solyndra will be. Compared to many of the thin film solar companies around, including other CIGS players (which stands for the four materials that make up the solar cell, copper, indium, gallium, selenium), Solyndra has one of the highest manufacturing and capital factory costs around. Solyndra discloses in an amendment to its S-1 that its per-watt manufacturing cost was $4.00 per watt for the fourth fiscal quarter ended January 2, 2010 (page 4). That means Solyndra is spending more money than some competitors to make its solar gear at this stage, which could lead to higher operating costs than competitors, and could make it difficult for Solyndra to compete.

In comparison MiaSole’s CEO Joseph Laia told me last week that MiaSole will be able to produce its solar material for 85 cents per watt in manufacturing costs and 50 cents in pure capital expenses in its current facility by the end of the year. Thin film solar darling First Solar (s FSLR) says it can manufacture solar for under $1 per watt, and most recently said it has hit 85 cents per watt.

CIGS competitor HelioVolt says on its website that its production roadmap can get it to under $1 per watt (though HelioVolt has had its own problems). Nanosolar has said it is already shipping its CIGS solar material with a cost of less than $1 per watt.

The former CEO of Nanosolar, Martin Rocheisen wrote a rant on the company’s blog two years ago that basically suggested that Solyndra’s costs were particularly high. Rocheisen wrote: “[W]e find that tubular panels are worse on most if not all metrics,” and included a picture of a prototype of a tube-shaped solar rack that Nanosolar had made. Rocheisen said in the post that Nanosolar could make its own tube design for “less than a third of the cost achievable,” by an unnamed competitor, which was clearly Solyndra. Rocheisen has since left Nanosolar (some in the industry say he was pushed out).

When I asked Laia of MiaSole in an interview last week why he thought the DOE had given Solyndra the loan guarantee when it has such high manufacturing costs, he said the DOE was examining the CIGS solar industry when it was at an early stage of development.

Last month Solyndra auditor PricewaterhouseCoopers said that Solyndra’s operating losses, negative cash flow, and $532.3 million deficit “raise substantial doubt about its ability to continue as a going concern,” reported Reuters.

Of course, manufacturing cost-per-watt doesn’t tell the entire story of a solar company. One of Solyndra’s big innovations is that it says its installation costs are lower than competitors because its racks can be installed more quickly, more easily and with less skilled labor. According to a report on SharesPost, installation costs for Solyndra can range between $0.50 and $0.75 per watt compared to several dollars per watt installed cost for some competitors.

On HelioVolt’s website in the FAQ section it says:

Although widely used, this [manufacturing cost per watt] is really a measure of the initial cost of just one part of the entire electricity-generating system. It ignores some very important factors: the efficiency of photovoltaic devices, their lifetime, and the cost of the rest of the system in which they are installed. The best measure is the actual cost of the electricity that the system itself generates over its lifetime, measured in cents per kilowatt hour.

In addition Nat Bullard, analyst with New Energy Finance, Bloomberg, explained to me that even with a higher manufacturing cost than some competitors, Solyndra will be able to reduce its cost significantly in the short term. “Early on production prices can drop steeply. With a doubling of output costs can drop by 10 percent.” Solyndra is trying to boost its manufacturing capacity from 54 MW annually for the year ended January 2, 2010, to 110 MW annually by the fourth fiscal quarter of 2010.

Even if the DOE made its loan guarantee award to Solyndra without having a full picture of the thin film solar industry, the DOE does plan to offer loan guarantees to competitors in various industries, explained Jonathan Silver, the former venture capitalist whom the Obama administration tapped to lead the Department of Energy’s loan guarantee and green car loan programs, to us in an interview. Silver said the DOE plans to give out up to four loan guarantees to companies in June, so we’ll see if any of them go to thin film solar competitors.

Still it would be pretty ironic if the DOE’s flagship loan guarantee for clean power turned out to have gone to a company that had one of the least economical thin film solar manufacturing business models around.

To learn more about clean tech financing deals see GigaOM Pro (subscription required):

How EV Battery Startups Can Cross the Valley of Death

Report: Cleantech Financing Trends: 2010 and Beyond

19 Responses to “Was the DOE Loan Guarantee for Solyndra a Mistake?”

  1. Per-Ola

    Government subsidies can be good in many casdes, but it also throws the market forces “off kilter”, and in this case maybe pushes one technology at the expense of others.

    Consider this instead. Don’t dole out any subsidies, but instead tax the bad forms of consumption (essentially anything related to fossil fuel). Then let the market decide which of the many new and emerging technologies that will actually deliver.

    Also realize that this is probably (due to polotics) an impossibility in a country where we do have a domestic supply of fossil fuels and where tons of jobs are directly linked to the extraction of said fuels.

    In countries w/o any domstic supply, consumption taxes of this type have proven to be very effective and drive both innovation, and costs down.

  2. TankThink

    That’s why the government needs to set the stage and then let the companies play, not pick the winners. Obviously, the government officials don’t study engineering or physics.

  3. CTYankee

    Wasn’t it obvious to everyone that is takes Pi*m^2 of active material to collect 1 m^2 of sunlight using the Solyndra collectors? The round profile also means that most of the area is off axis to normal solar radiation. I dunno, maybe diffuse or reflected light is OK, but if that’s the case then why space the tubes at all? If light can get in to the back if the panel behind the tubes it can escape too? The less light the panel captures the less potential energy it can harvest.

  4. When you consider that a parabolic dish based concentrated PV system uses 1/1000 the PV material as that of flat-plate PV systems for the same power output it makes all flat plate PV systems look antiquated.

    If you can generate the same amount of electricity with 0.23m² of PV solar cell material as would normally require approximately 350m² of expensive PV cells…. you’d be using a redundant method and would be wasting a hell of lot of money building a system comprised of only the later.