Solar Photovoltaics
Getting the Price Right on PVs
By: Brent Barker
Solar photovoltaics use a variety of solid-state technologies to convert sunlight directly to electricity with no moving parts. Properly sealed from the environment, a solar cell can function for many decades with minimal maintenance and no fuel cost. As a result, the economics of photovoltaics (PV) are almost entirely skewed to the up-front capital cost. Today that's a significant disadvantage, limiting PV to niche applications and highly subsidized residential and commercial markets. But in the longer term, this capital intensity will turn to PV's advantage because of its inexorably declining cost curve.
Over 30 years, the price of solar photovoltaic systems has dropped by about 20 percent for every doubling of installed capacity. Plotted on a graph, the price is a straight downward line with little variation until three to four years ago when a crippling worldwide shortage of silicon feedstock, coupled with exploding demand in places like Spain, kept prices from declining. But this year, with major new silicon capacity coming online, prices are expected to drop significantly—conservatively some 10 to 20 percent—which will move the PV price substantially back to its 30-year trend line.
"Grid parity is clearly in sight," said solar power consultant Terry Peterson. "PV economics will only get better, while conventional energy prices continue to rise."
A number of forces, from public policy to technology to global competition, are driving PV forward. In January, the $2,000 cap on the federal incentive for PV was removed, and the 30 percent federal tax credit extended for 10 years. Similarly, states and utilities facing renewable portfolio mandates continue to add their own subsidies to the mix to encourage residential and commercial application. Market expansion, which is moving along at a fast 25 percent per year, will continue to drive down prices which, in turn, will expand markets, which will drive down prices, on and on, in a circle driving PV toward true competitiveness within a decade.
Technologically, the two competing camps—silicon crystal technology and thin film technology—are fighting for dominance and spurring innovation and efficiency improvements across the board. PV has become the most recent darling of the U.S. venture capital community. With their eyes glued to the trillion-dollar energy markets of the future, venture capital companies poured some $3.5 billion into solar startups in 2008 alone. Most of the venture capital is pursuing thin-film processing technologies that involve depositing layers of semiconductor materials a few microns thick on a substrate of glass, metal foil or even roofing materials. The processing technology lends itself inherently to large economies of scale, running more like a printing press than discrete manufacturing.
The more traditional silicon crystal technology, which involves growing a silicon ingot, slicing it into thin sections and wiring them together, currently accounts for about 85 percent of the market. Silicon crystal manufacturing, even for U.S. firms, is rapidly moving to Asia. These global manufacturers in turn are feeding the highly subsidized markets in the United States, Europe and Japan. Germany, with more than 220 cloudy days per year, is now the world's largest market, followed by California and Japan.
Utility-managed rooftop installations depend almost entirely on subsidies to bring the homeowner's cost to a more reasonable level, and offer investment payback to the homeowner in the six- to 20-year range depending on a number of variables, including available sunshine, state incentives, utility programs and the prevailing electricity rates offset by a rooftop PV system. In high-cost areas, such as California and Hawaii, the PV investment payback is considerably faster.
Austin Energy in Texas started a solar roof program in 1999 and now has about 700 customers with rooftop PV systems. Austin is seeing a huge surge of interest now that the federal tax credit has been increased and publicized, said Leslie Libby, manager of solar programs. "We do everything we can not to oversell solar. We want people to understand what they are buying and what it can do for them." Austin is cautious with its customers because the economics of financing, billing and net metering can quickly get complicated. For starters, a typical rooftop PV system in Austin, she said, would be sized at around 3 kW and have an installed cost of around $23,000. "We provide a rebate of $13,500, bringing the cost to the consumer down to around $10,000. Then the federal government provides an investment tax credit of 30 percent, or in this case about $3,000. The net cost to the homeowner ends up at around $7,000, which is still a substantial investment for the average family." Because Austin's electricity rates are so low, the average payback on a PV investment is 12 to 15 years. The bottom line is that "the customer has to really want this," Libby said.
The Sacramento Municipal District in California has had a PV program in place for more than 25 years. "We have been fairly consistent over the years," said Mike DeAngelis, manager of advanced renewables and distributed generation technologies programs. "We installed a 1-MW installation back in 1984 that is still producing electricity. Sealed from moisture and other environmental corrosion, these systems will perform reliably for a very long time." Over the last 25 years, SMUD has installed about 1,500 PV systems for a total capacity today of about 12 MW (AC). Given California's mandates for renewables, this will accelerate dramatically. "Our goal is to get an additional 125 MW of installed PV over a 10-year span—a tenfold increase. All the rest of California has very similar plans. The tax incentives and net-metering program will be critical to our success. Feed-in tariffs may also be very effective in the future."
DeAngelis said an installed 2-kW PV system would cost around $16,000. The SMUD incentive combined with the federal and other incentives would reduce the homeowner's cost to around $7,000. "Now it begins to look like a reasonable investment, especially with utility net-metering programs, which allow the homeowner to receive retail rates at high-rate tiers rather than wholesale rates for the power it provides into the grid," he said. "I haven't personally run the numbers, but I've been told that many homeowners should be able to get a simple payback of about six to 10 years. Nevertheless, I would tell anyone considering a PV system to first do what they can to save electricity more cost effectively through energy-efficient appliances, [compact fluorescent light bulbs] and by changing their energy usage behavior, and then to check the financial investment numbers based upon their situation."
Salt River Project in Phoenix started its solar PV program in 2004 and now has about 550 residential customers with rooftop systems. "The average system in Phoenix is 5 kW and the installed costs are running about $35,000," said Lori Singleton, manager of sustainability initiatives and emerging technologies. SRP pays its customers an incentive of $2,700 per kilowatt. Combined with federal and state tax incentives, this brings the net cost to the homeowner down to around $14,050. Given our low electricity rates (averaging 9.8 cents/kWh), the homeowner's annual savings would be about $700 per year, meaning it will take them about 20 years to recoup their investment."
With paybacks this long, one wonders what kind of customers would be inclined to invest. Singleton said, "We are seeing systems in large homes where people have the wherewithal to do it. Other customers are attracted because of their strong environmental ethic and because they want a hedge against higher electricity rates in the future. Interest is clearly growing; with the new federal tax incentives, our phones are ringing off the hook. Most calls are educational: we send them information and we talk them through the various criteria to help them determine whether solar makes sense for them. We don't get involved in assisting customers with financing options, but we rely on third parties to do that. The solar installation industry in Phoenix is growing rapidly. A lot of new installers and retailers are getting into the business."
SRP is also providing incentives for commercial application of solar PV systems. The utility recently paid an incentive to Gatorade to install a 500-kW photovoltaic system on the roof of a local distribution facility. The installation is the largest customer-owned solar system in Arizona. SRP also recently installed a solar cooling system at an Eco building at the Arizona National Guard and is monitoring the performance of this innovative system for one year.
Innovation is not limited to new technology. New financing and leasing arrangements are emerging that could substantially increase the growth of solar PV in the coming decade. "There is a whole new business model, which is now being offered in Southern California, where the PV developer essentially comes in and leases your roof," said Ed DeMeo, president of Renewable Energy Consulting Services. "It's a direct takeoff of the wind model where developers lease the farm and ranch land, put up a turbine and pay the landowner an annual stipend. We've got all that urban and suburban real estate sitting idle."
Commercial real estate now represents the greatest opportunity for the new business models. "A very strong market is developing in the commercial sector for third-party financing and leasing," said Fred Jennings, principal and senior consultant with R.W. Beck. "The third party owns and maintains the system, receives the renewable incentives, and sells the power at retail rates. They are motivated to maintain and upgrade the system as required. Under the lease arrangements, the property owner benefits by receiving a long-term fixed rate for power. In short, they lock-in your utility rate in exchange for your roof. Some are now arguing that the third-party lease model will become the norm for commercial deployment of solar PV. And once established, it has potential applicability to the residential market. If you believe that utility rates will escalate in the future, these new financing options in many ways appear to have instantaneous payback."
Recent studies of solar deployment by R.W. Beck have drawn upon the well-established Bass Diffusion Model to predict patterns and rates of technology adoption. Much of it comes down to the payback period. "When the payback is over about 10 years, you lose most of the potential adopters. You keep the ones [who] would adopt under any circumstances. When payback drops to six to 10 years, you get a more complex set of adopters: those interested in conservation and sustainability, who are looking to do the right thing, but who also use economics in their decision-making process. For payback of five years and below, and especially in the three-year range, you pick up a whole section of adopters who look strictly at the personal economic advantages."
Solar cells emerged from the Bell Labs 50 years ago, giving the ingot-sliced, wafer-based, crystal silicon PV a head start over its chief competitor, the deposition of a thin film of solid-state chemicals on a substrate. The competitive advantages of crystalline silicon derive from the fact that it has higher conversion efficiencies than thin film, and can hitchhike on the manufacturing prowess of its cousin, the computer chip. The competitive advantage of thin film was that it could employ process technologies to ultimately roll out product in a continuous stream. Thin film was supposed to have overtaken traditional crystal silicon years ago, but it is chasing a very agile moving target.
The world of thin-film PV technologies has been under development in many locations for many years, including the National Renewable Energy Laboratory in Golden, Colo. "Thin film PV has significantly lower manufacturing costs than traditional silicon-wafer PV modules," DeAngelis said. "NREL has had programs to improve the sunlight-to-electricity conversion efficiency of thin films for over 25 years, but it has taken a long time for these efficiency gains to appear in the marketplace. Now there are a variety of thin film manufacturers for the three basic types: amorphous silicon (A-Si), with expected future commercial efficiencies to be as high as 10 percent, cadmium telluride (CdTe), with expected commercial efficiencies around 10 percent, and copper-indium-gallium-diselenide (CIGS), expected to be around 16 percent. Efficiencies for current thin-film products in the marketplace are considerably less.
The leading manufacturer of CdTe systems, First Solar of Tempe, Ariz., "has finally proven the long-time promise of thin film as more scalable and ultimately cheaper than the traditional crystal silicon approach," said Peterson. The jury is still out for CIGS to emerge as a strong commercial competitor. "CIGS has the enviable position of having the highest demonstrated lab cell efficiencies—around 20 percent, substantially higher than those for CdTe or A-Si—and their potential has attracted a lot of interest and money from the venture capital community. However, despite the infusion of capital, CIGS has yet to produce a commercial player." Claims to the contrary notwithstanding, Peterson sees no clear evidence of a commercial-scale production line with publicly demonstrated module performance. Everything is under wraps, but the suggestion is that these startups are finding it much harder to move from the laboratory to production with this particular type of film.
The venture capital community would like to see a clear winner emerge triumphantly from the competitive arena. However, hopes for a sudden research breakthrough in PV are misplaced, said DeAngelis, who spent part of his early career at NREL, and who ran the R&D program for the California Energy Commission before joining SMUD six years ago. "People always love to think about research breakthroughs. In the real world there are very few of these. The future improvements are likely to be incremental in solar PV development. This evolutionary reality argues strongly for incentives that support a marketplace where companies can make a profit and chip away at improving their products. A good example is SunPower of San Jose, Calif., one of the leaders in crystal silicon. Ten years ago, the efficiency of its crystal silicon modules was about 10 to 13 percent. Today, SunPower cites efficiencies of about 17 percent and believes it can get to 22 percent in the future.
Higher efficiencies mean less land or roof area, less square footage of modules, less framing, less wiring, less installation labor and fewer balance-of-system components, all adding up to lower cost," he said.
DeAngelis does not expect a real, near-term winner to emerge from this technological tumult. "In the 1980s when I was at NREL, we were always looking at these options and trying to pick out a winner," he said. "Even today I can't pick a winner between crystalline, thin film and the various innovative concentrator technologies [which use a lens or reflectors to concentrate a large area of sunlight onto a small area of highly efficient PV cell, increasing the output]. There are a lot of pathways to reduce costs, each with their own risk."
Peterson, who has followed the field for roughly 30 years, said he sees the horse race somewhat differently. "Thin film is undoubtedly going to edge out crystal silicon. The question is when. Twenty years ago I would have said by the year 2000. I've grown to appreciate the ingenuity of the people making crystal silicon, so I no longer predict a date."
After the PV module itself, the second highest cost in any PV system is the inverter used to convert DC to AC power. "It has also been the weakest link," said Libby. "Solar panels themselves are very robust, and carry 25-year warranties, but inverters are in the five- to 10-year range. I've been involved with solar since 1991 and I've seen the reliability of inverters go way up."
Libby said she is most excited about the new generation of micro-inverters. "Each panel is connected to its own micro-inverter. Right now, all the panels in a PV system are connected in series, so that if any panel in the series is shaded, it brings down the performance of the whole system. Also, for a series module to work, all the panels have to have the same orientation and tilt. We have a lot of trees in Austin that dapple shade and a lot of the new houses have all these different roof areas and variety of directions. Micro-inverters could solve these problems, and provide great flexibility. We'll be testing them to see how well they are designed, how they perform and how long they last. But if they work, the benefits will be enormous."
Technology is only a piece of the larger economic reality facing solar PV. There are a number of "softer issues that will enhance or hinder PV deployment," said Jennings of R.W. Beck. These include everything from codes and zoning regulations to the institutional awareness and acceptance of the value of solar within the finance and real estate communities. "We need to get to the tipping point where solar is the norm, not the exception. At that point, solar installations become an integral part of the mortgage, financial lenders make adjustments for reduced risk, codes and zoning ordinances support and encourage solar, solar becomes an integral part of all new home design and smart metering allows the full value of solar installation to be understood by the customer and utility alike. We have made strong recommendations to our clients to focus their efforts on greenfield development, particularly in new planned urban communities. One homebuilder in California told us that his new solar homes were now outselling non-solar homes, even with a price premium in a poor market. This, I think, is evidence of the broader market transformation that is under way."
After the economic tipping point is reached, does solar have the capacity to become mainstream the way coal-, nuclear- and gas-generated electricity are today? After all, these draw upon abundant, readily available natural resources that can produce electricity on a large scale and at a reasonable cost. DeMeo likes to point to an assessment he and others made more than 20 years ago. "We calculated at EPRI that you could power the entire U.S. electricity needs with a solar installation in Nevada the size of a circular land area 140 miles in diameter, and covering only 25 percent of that land with solar cells that were 10 percent efficient. The point is that the solar resource base is simply too large to ignore, and costs are moving in the right direction. The U.S. Southwest, including the Mojave Desert in California, is widely regarded as one of the richest solar resources in the world, and it is sitting there essentially untapped from an energy perspective."
Back to Top
Copyright © 2004-2010, American Public Power Association
