With 1000 "little cuts", SunShot aims to drop solar costs 75% in less than 10 years.
It’s easy to be skeptical about the Dept of Energy’s (DOE) SunShotInitiative. The goal, installing utility-scale solar at $1 per watt by2020, would bring solar costs down to 6 cents per kilowatt-hour (kWh),roughly the cost of coal-fired electricity.
The name, SunShot, is a play on President Kennedy’s 1961 pledge to land a man on the moon by the end of the decade. DOE Secretary Chu announcedthe $27 million program in February, the funds for which will be spreadamong nine companies. How can $27 million make such a big impact?
To achieve $1 per watt, the solarindustry needs to streamline in a big way. It will need considerablemodule efficiency gains and slashed costs for installation, operationsand maintenance and all other system components. Photovoltaic (PV)modules will need to come down 70%, inverters 55% andconstruction-related costs nearly 75%.
"We’ll get there," says Frank van Mierlo, CEO of SunShot grant recipient 1366 Technologies. "Look at the historical cost curve of solar. The production costs come down 10% every year."
At 1366, van Mierlo and fellow co-founder Emanual Sachs are working tocommercialize a method for manufacturing crystalline silicon wafers that halves the need for hyper-pure silicon. The process, which virtuallyeliminates silicon waste (about 50% is currently lost), sets the stagefor a dramatic wafer price drop. "If the total installed cost is goingto be $1, that means wafer costs have to come down to about $0.25, from$1 today.," says van Mierlo. "And I think it’s extremely doable. We’rejust one of many companies working on this, and our technology alonewill take one of the largest cost components and slash production costsby a factor of three – at least."
Here’s a closer look at what four of the SunShot grant recipients are doing to help achieve these results.
Target: Thin-Film Efficiency, Durability
Thin-film PV technology has its advantages – it’s lightweight,production is relatively cheap and simple, and it’s notsilicon-dependent.
Durability, however is a disadvantage right now. Until nanotechcompanies make headway on a durable, exposed cell, commercial thin-filmmodules will continue to require robust encapsulants.
As First Solar and other thin-film manufacturers began to ramp up, that need caught the attention of PPG Industries, a Pittsburgh-based glass and glass-coating maker with roots in the 19th century. PPG’s contribution actually predates the industry: Back in the 1930s, the company pioneered the first low-iron glass. Now, backed by$3.1 million in SunShot funding, PPG is working to perfect a glassencapsulant for modules made with cadmium telluride (CdTe) – the mostadvanced thin-film technology in mass production.
PPG will try to maximize CdTe module efficiency by pulling threeinnovations onto their high transmission glass. "About 4.5% of the sun’s energy is lost on the outside of the module," says Jim McCamy, PPG’smanager of solar technology. "By reducing losses from reflectivity,we’re increasing the number of watts. By creating better conductivelayers, we improve the number of watts."
The new encapsulants, still unnamed, will separate the first layer ofCdTe from the underlying transparent conducting oxide (TCO) glasssubstrate with a buffer layer. A third component, an anti-reflectivecoating, will be applied to the module front side over the TCO glass."These technologies exist, but they’ve never been combined," McCamysays. Research is taking place at PPG’s research center outsidePittsburgh, at Colorado State University (whose research formed thebasis for Abound Solar’s CdTe technology), and at Oak Ridge National Lab in Tennessee.
During a 2009 First Solar conference, glass was named as the CdTeindustry’s single largest cost component. Two years later, CdTe isproviding the best value proposition for utility-scale PV developers.
Target: Silicon Wafer Manufacturing, Expense
For all the hype over thin-film’s future dominance of the solarindustry, crystalline-silicon technologies still account for 80% of theglobal market.
Silicon’s everywhere – it’s the second most common element in theearth’s crust and is found in everything from iPhones to your favoritelager. But when refined into hyper-pure silicon, necessary for wafermanufacturing, it costs around $350 per kilogram.
In North Lexington, Mass., cleantech innovator 1366 Technologies isplanning to slash that expense for module manufacturers through itsDirect Wafer manufacturing process.
Besides attracting $3 million from Sunshot and $4 million from AdvancedResearch Projects Agency-Energy (ARPA-E), 1366 has secured backing fromGE, VantagePoint Venture Partners and Hanwha Chemical. In total, $46million is being committed to help commercialize Direct Wafertechnology.
Co-founder van Mierlo likens the manufacturing breakthrough to theBessemer process, the steel mass production process that was thefoundation for the Carnegie fortune. Direct Wafer reduces waferproduction steps from four to one by fashioning wafers directly frommolten silicon. The process eliminates silicon waste by negating theneed for sawing and grinding hardened silicon.
Relative to legacy technology, Direct Wafer is thousands of times faster post-melt, four times more capital efficient and uses just half theamount of silicon. "We eliminate the waste and we streamline theproduct," says van Mierlo. "That’s where we get our cost savings."
To meet SunShot’s goals, 1366 is trying to bring down the cost ofsilicon wafers from $1 per watt today to 25 cents. "You never know until the factory’s built, but from a technical viewpoint, we’ve met thatgoal," he says.
As soon as 1366 completes engineering and construction of its high speed manufacturing machine, it will break ground on a 100 MW demonstrationplant – by as early as year’s end. The plan is to scale up to a gigawatt plant. "Everything we see now leads me to believe we can do this," hesays.
Target: Thin-Film Installation Costs, Versatility
A number of roofing companies now partner with installers to providebuilding integrated PV (BIPV) systems that low-lying, lightweight,non-penetrating and – depending on your preferences – more physicallyattractive. Copper-indium gallium (di) selenide (CIGS) thin-filmtechnology is on the cutting edge of this expanding niche market.
In the past, CIGS module manufacturers trying to tailor a product to the BIPV market have been restricted by the limited availability offlexible polymer encapsulants, used in lieu of conventional rigid glassencapsulants. No manufacturer has mass-produced a proven polymer basedproduct for CIGS panels – anything that’s made its way onto a rooftopcame off a pilot manufacturing line.
Now, propped up by $4.4 million in SunShot funding, 3M is planning to scale up and fill that hole. The Minnesota-based company is on schedule to begin mass production of a multi-layer,fluoropolymer-based encapsulant early next year, leading the thin-filmindustry’s shift toward versatile, flexible modules.
"3M has been working on this for over a decade," says Arnie Funkenbusch, thin-film solar program manager. "But it’s only recently that we’vebegun to focus on the solar application." For years, 3M’s flexiblepolymer films have been used in organic light emitting diodide displays, found on cell phones and other hand-held electronic devices. Inadapting its film for the solar industry, 3M has laser-focused itsresearch on improving weatherability. "Unlike with electronics, in asolar application our film is subject to sunlight and temperatureextremes," he adds.
Manufacturing of a first-generation film, Ultra Barrier Solar Film, isunderway at a pilot line in Minnesota, and what’s been produced to datehas been sold to CIGS module manufacturers. The final product will betweaked slightly, Funkenbusch says, as accelerated life-time testing isstill underway at National Renewable Energy Labs.
Using feedback from NREL and the CIGS module manufacturers, Funkenbuschexpects the redesign to have increased light transmission, betterdurability and lower cost. "We want to make sure we
have a product that lasts 25 years," he says. "That’s where SunShot helps us out."
The new manufacturing line will be at an existing 3M factory in Columbus, Missouri. According to the Columbus Daily Tribune, the line will add 120 solar jobs to a plant that’s shed jobs for years.
Construction of the line is underway and production Construction of theline is underway and production is set to begin in early 2012. There are indications that CIGS module manufacturer
SoloPower, which recently secured a $197 million DOE loan guarantee for a new factory in Wilsonville, Oregon, will be the first customer.
Target: IBC Cell Production, Cost
Recent NREL tests have measured SunPower’s interdigitated-back-contactsolar (IBC) cells at the highest conversion efficiencies to date. Theonly factor preventing SunPower’s technology from becoming the industrystandard is cost.
The principle is very simple: IBC cells eliminate frontside metallicconductors, which typically cover 5-8% of the physical surface of awafer. Move the wiring to the backside, more sunlight gets absorbed, and cell efficiency spikes.
"Currently, IBC solar has the highest efficiencies, but it’s veryexpensive to make," says Jim Mullin, general manager for solar productsat Varian Semiconductor. "In solarmanufacturing, every step adds cost, adds complexity, and can result inpotential yield loss." In Gloucester, Mass., Varian is using $4.8million in SunShot funding to develop an ion-implant tool that promisesto slash the number of manufacturing steps from 21, necessary in the IBC process today, to eight.
Ion implantation is far from a novel concept – implanters are integralto semiconductor device fabrication and have long been used to makecomputer chips. Varian’s been selling the tools to the electronicsindustry since 1975 and has an ion implanter, the Varion Solion, forconventional PV cells that’s already running in high-volume production.SunShot backing will be used to redesign the Solion so that it can to do backside wafer patterning necessary for IBC cell production.
"We’re the first ones to do this," Mullin says. "We are in a uniqueposition because we’re the only ones that can do it without impactingthe tool’s productivity – because we own the intellectual property."
Mullin says that Varian has matched its DOE funding with internalinvestments well north of $20 million. He expects the tool to becommercially ready within three years.
It’s unclear how much the reduction in production complexity will reduce IBC costs, however. Module manufacturers have some ground to make upbefore IBC solar is cost-competitive with other PV technologies.
SunPower is the only company commercially producing IBC modules todayand is operating at a production cost of around $1.70 per watt, whileChinese crystalline-silicon manufacturers are producing at about $1.20per watt, and First Solar, a CdTe thin-film manufacturer, is now downbelow the $0.75 mark.
NREL Incubator Program
Not all SunShot funding is going toward projects at 1366Technologies, 3M, PPG, Varian Semiconductor and Veeco, a fifthrecipient. Another $7 million is being injected into the NationalRenewable Energy Lab’s Photovoltaic (PV) Technology Incubator program.
Started in 2009, the incubator program is primarily tasked withhastening nascent PV technologies’ transition from the drawing boardinto the lab. There are two funding tiers: The first supportsdevelopment of commercially viable prototypes; the second helpscompanies scale up to the pilot-project manufacturing stage.
SunShot funding targets four companies:
Caelux (Pasadena, CA) – $1 million: developing aflexible PV-manufacturing process that minimizes the amount ofsemiconducting material used.
Solexant (San Jose, CA) – $1 million: developing a new printable nanoparticle thin-film ink from common nontoxic substances.
Stion (San Jose, CA) – $1 million: developing atechnology consisting of two stacked high-efficiency thin-film devices,offering improved absorption of light.
Crystal Solar (Santa Clara, CA) – $4 million: commercializing single-crystal silicon wafers, four times thinner than standard cells.