|
Gridless Solar Cell hits 20% efficiency mark [SunPower Corp] source: EPRI Journal While much of the rest of California’s Silicon Valley struggles to weather its worst recession in more than a decade, a 15-year-old Sunnyvale firm started by a former Stanford University electronics engineering professor sees a bright future ahead. SunPower Corporation is collaborating with Cypress Semiconductor Corporation, of San Jose, one of the valley’s leading integrated circuit manufacturers, to make solar photovoltaic (PV) cells that presently have the highest verified energy conversion efficiency—20.4%—of any silicon PV cell in commercial production.
The cell technology evolved from R&D work at Stanford—supported in large part by EPRI and, later, the U.S. Department of Energy (DOE)—that began more than a quarter-century ago. In subsequent applications developed by SunPower, the cell technology was used on the record-setting, high-altitude Helios solar-powered aircraft in 2001 and on the 1993 trans-Australian World Solar Challenge-winning racecar built by Honda Motor Company. Picture Somewhat ironically, the original focus of the PV cell R&D at Stanford was on small-area, high-efficiency cells for solar arrays that use highly concentrated sunlight in large, high-power generating systems. A version of the cell, built for 500-sun concentration, still holds the world record of 26.5% efficiency for silicon PV. But after many years of pursuing high-concentration PV cells, SunPower’s president (and former Stanford professor) Richard M. Swanson has concluded that the high-efficiency, silicon PV cell technology can be economically competitive for one-sun, or non-concentrating, flat-plate arrays. Such arrays are being installed on a growing number of residential and commercial building roofs or integrated into roof structures and exterior building surfaces. “The retail solar cell market has grown 30% a year over the past decade and is projected to reach $3 billion within three years, nearly double its current level,” says Swanson. “Flat-plate PV has progressed downward along a straight, classic learning curve since the late 1970s. The experience factor for PV is around 80%, which means that every doubling of cumulative production by the industry reduces product cost by 20%. “Worldwide solar module production capacity is increasing 10-fold every decade, and the price of modules decreases by half every decade. It’s now $3 per watt, and we can easily see it going to $1.50/W in the next decade, for a total installed system cost of $3/W,” says Swanson, meaning a 3-kW system would cost around $9,000. “The various technology and manufacturing elements driving the cost reduction of flat-plate solar modules can be expected to continue, so that current PV technologies will be cost-effective in most distributed power applications in 10 years and competitive with fossil fuel-generated electricity in 20 years,” Swanson adds. “Conceivably, someday concentrator systems could be a lower-cost PV alternative, but they are not now and they have a long way to catch up with continually improving flat-plate systems. Moreover, concentrators are not as well-suited for many small distributed, remote applications.” Picture Partnering with Cypress SunPower’s solar cells are being made in limited quantities in a pilot production line, capable of turning out about 2 megawatts’ (MW) worth of cells a year, that is located inside a Cypress Semiconductor silicon wafer fabrication and chip-making factory in Round Rock, Texas, near Austin. The close cooperation is a product of Cypress’ 2002 investment of $8.8 million for a 44% stake in SunPower. Cypress joined Honda, Japanese construction firm Sekisui Jushi, the Indiana-based energy company NiSource, and other investors with a stake in SunPower.
T.J. Rodgers, the president and CEO of Cypress and chairman of SunPower, has known Swanson since they were students at Stanford together in the early 1970s. In addition to Cypress’ corporate investment in SunPower, Rodgers reportedly also personally invested in the solar cell firm. Around the time of Cypress’ announcement of its SunPower investment, Cypress installed 300 kilowatts of PV modules on the roofs of two buildings in its San Jose headquarters campus that provide nearly half of the buildings’ electricity demand. Cypress is sharing its expertise in technology development and high-volume manufacturing, which SunPower says gives it a substantial competitive advantage in a silicon-intensive, highly cost-sensitive business that is remarkably similar to chip making. Both processes involve automated, precision sawing of very thin silicon wafers from single-crystal, high-purity silicon ingots, photolithography and other electronics manufacturing techniques, and very high quality control. “Our partnership with Cypress will be key in enabling our transition from a small-scale solar supplier to a world-class manufacturer of solar cells for high-volume applications,” says Swanson. “Cypress’ involvement brings an infusion of manufacturing expertise and synergy with the integrated circuit industry.” SunPower is planning to ramp up solar cell production in 2004 with a 25-MW factory, construction for which is under way, for possible siting next to a Cypress semiconductor test and assembly facility at a technology park south of Manila in the Philippines, or in India. Swanson envisions additional production capacity increases in 25-MW increments, with the goal of reaching 150 MW within three years. Back to the future Picture SunPower’s high efficiency cells incorporate many technical insights and innovations derived from detailed, atomic-level analysis and modeling of cell physics in order to optimize the creation of current-producing pairs of electrons and holes and to minimize their recombination in the bulk silicon. Most currently available PV cells convert 12-15% of the incident sunlight into direct current, while the most advanced type produced by industry leader Sanyo of Japan reportedly averages 19.5%. SunPower’s single-crystal A-300 cell was verified 20.4% efficient in May 2003 by DOE’s National Renewable Energy Laboratory (NREL) in Golden, Colorado. SunPower says that makes the 3-watt cell, which measures 12.5 centimeters on each side, the most efficient cost-per-watt PV solution. One thousand such cells, interconnected into modules or integrated into building surfaces, could provide three kilowatts of peak power (a typical residential peak demand) in less than 17 square meters of module area—about 170 square feet, well within the size of most residential rooftops, perhaps even fitting onto a garage roof. “The A-300 cell marks a major milestone both for SunPower and for the photovoltaic industry,” says Swanson. “The cell’s innovative design, high efficiency, and low manufacturing cost will enable our customers—the companies that design and build solar modules—to create superior products at a cost capable of accelerating the rate of conversion to clean, solar solutions.” Performance and cost are the critical factors to making solar PV a commercial success, says Rodgers. “The A-300 provides an entirely new class of cost-effective solutions to the clean energy industry. The solar cell business represents a great opportunity for several reasons. Even without the negative effects of political volatility, the production of conventional sources of energy, such as oil, is expected to peak over the next decade, requiring renewable forms of energy to begin to take its place. After more than 30 years of continuous improvement, the science of solar power has matured to the point where it has become highly cost-effective.” Concentrators’ potential persists
Solar System’s first concentrator PV station—the 10-dish receiver Anangu Pitjantjatjara station, rated 220 kilowatts, is located at Umuwa, South Australia. Solar Systems says it has contracts for three additional solar power stations in the Northern Territory that, together with the first one, will provide generating capacity of around one megawatt. Strong government support in Australia for a variety of measures to reduce the country’s emissions of heat-trapping greenhouse gases in the future have fueled perception of a vigorous, emerging market for distributed solar and other renewable power, particular in the sun-drenched country’s remote regions and villages, where diesel generators rather than a power grid are the principal source of electricity. In the United States, another California company that licensed the EPRI-Stanford high-efficiency silicon cell technology—Amonix, Inc., of Torrance—remains committed to developing a market for utility-scale applications of integrated high-concentration PV systems. EPRI supported development and commercialization work at both Amonix and SunPower in the 1990s that, in the case of Amonix’ modular, scaleable and sun-tracking concentrator arrays, included significant field-testing at a number of utility sites. One of the early utility hosts for Amonix concentrator arrays was Arizona Public Service (APS), one of five EPRI-member companies that supplementally funded high-concentration PV R&D in the 1980s. In 2001, APS installed some 500 kW of Amonix arrays at multiple sites in Arizona, including the Glendale Municipal Airport. Vahan Garboushian, president of Amonix, says the APS arrays and other units elsewhere have since logged a billion watt-hours—a gigawatt-hour—of operating experience. APS is now planning to install approximately two megawatts of Amonix’ new, 35-kW arrays at the Prescott, Arizona, airport, as part of a broader plan to deploy some five megawatts of additional distributed solar generating capacity on its system, he adds. To build those arrays and to support additional market development, Amonix is looking for investment partners to finance expansion of the present array manufacturing capacity from around one megawatt per year to five megawatts per year, by the end of 2004. EPRI, SunPower, Amonix, and DOE’s Sandia National Laboratory shared a 1994 R&D 100 Award for the 26.5%-efficient, 500-sun concentrator cell. Amonix employs a chip foundry approach to manufacturing its commercial cells that minimizes capital equipment requirements. Its Fresnel lens-equipped arrays that incorporate the concentrator cells convert sunlight into direct current at 19% efficiency. Amonix’ concentrator array is projected to cost less than $2/W when manufactured in large volume. “It’s always possible that future development could change the cost outlook for other PV technologies,” says Garboushian, “but we just keep plugging away with concentrator technology, because we think that in volume production, it could eventually be half the cost of what flat-plate can do.” Still a technology horse race “There are strong corporate leaders in each distinct PV technology, each of them making the terribly hard but incredibly rewarding uphill climb to success,” says Ken Zweibel, who manages the thin-film PV partnership program at NREL for DOE. “No doubt, the problems are daunting. But did we start this because we thought it would be easy? No. We are doing this because it is important." “I hear things from all sources, and I am strongly heartened that we will make it through the wilderness and succeed,” says Zweibel. “And our success will be greater for the difficulties we overcome.” Commitment to R&D for long-term “The road from laboratory to commercial success is always longer and more pot-holed than we technology optimists initially imagine,” says Terry Peterson, EPRI manager for solar power, “but the front-end R&D investment is indispensable to progress and—in the long run—tiny compared with its marketplace rewards.”  |
