Utilities Take a Shine to Solar Power

Solar energy, such as the 250 MW SunPower PV facility, increasingly is being developed at utility scale. Bolstered by lower costs (due in part to market imbalances that currently favor buyers), state renewable portfolio standards, federal incentives and even a bit of creative thinking, solar energy is gaining a foothold in many utility companies’ generation portfolios.

Ron Kenedi, vice president-Americas for Sharp Solar, calls this the “beginning of the utility era” in the U.S. solar market. Utility demand for PV could be “huge” and may grow to be the company’s largest segment. “It’s happening all over at once,” he says. (For caption and credit information, click on this image in the image gallery below.)

Others agree. “Our opinion is it appears to be a booming market segment,” says Matt Cheney, CEO of Renewable Ventures, a financial firm which has been in the market since 2006. The company is part of Gemini Solar Development, which is building a 30 MW PV facility for Austin Energy east of the Texas state capital. “Utilities are waking up to the importance of solar,” Cheney says.

Solar adoption may be advancing, but the counter is that “capital markets are in disarray,” says Tom Fair, vice president of renewable energy for Las Vegas, Nev.-based NV Energy. Last year’s financial market shock, the lingering recession and lower demand for electricity all are complicating factors. Project finance has changed markedly with tax equity investors now in short supply. That’s bad news for most forms of renewable energy, which rely on tax-benefit-driven investors for capital. Purchased power agreements with utilities are now coin of the realm. But opportunities for utilities to take advantage of enhanced federal tax credits–available only since late 2008–may at last be having an effect.

A Year To Remember

Last year was a good one for utility-scale solar. According to the Solar Electric Power Association’s (SEPA’s) latest ranking of top solar utilities, many utilities doubled the solar megawatts in their generation portfolio during 2008. Overall, California continued to lead the country in solar megawatts. And top utilities included Pacific Gas & Electric, Southern California Edison, San Diego Gas & Electric, Public Service Co. of Colorado, Public Service Electric, Arizona Public Service, Hawaiian Electric Co., Portland General Electric, Sacramento Municipal Utility District and the Long Island Power Authority.

Uncertainties over future carbon costs have led many utilities to view solar as a risk mitigation option, says Julia Hamm, executive director of the Washington, D.C.-based SEPA. Hamm acknowledges that solar technology may not be the lowest-cost form of generation. But it offers utility capacity planners certainty when it comes to calculating lifetime operating costs.

Utility-Scale Technologies

Utilities have two primary solar technologies from which to choose: photovoltaic (PV) and concentrating solar thermal. PV technologies use a photosensitive material to generate electricity directly from sunlight. PV can produce at least some electricity under less than ideal conditions, such as low sun angles and overcast skies. That characteristic is why utilities in places like Massachusetts, Michigan and New Jersey can become solar energy players alongside their brethren in the desert Southwest. (For caption and credit information, click on this image in the image gallery below.)

PV also offers the advantage of modularity and scalability. Utilities and large energy users can start relatively small and scale up as demand grows and as they become more comfortable with the technology. But scalability can have its drawbacks, too. For one thing, the larger the deployment the greater the likelihood that cloud cover will affect output.

“During cloudy periods, the output from PV can get noisy with spikes,” which can have an effect on the grid, says Kelly Beninga, global director of renewable energy for WorleyParsons. PV installations around 20 MW in size can be managed without too much trouble. Larger than that and portions of the grid can be affected by passing clouds.

For a system the size that SunPower is developing for PG&E in California a solution might be to install lead acid batteries to cope with spikes caused by clouds. What’s more, the extent to which clouds pose a problem depends in part on where the nearest substation is located. “If you’re at the end of the transmission system it’s not good to have transients,” Beninga says.

To better understand the issue, NV Energy is studying power output variations that may result from deploying PV in and around Las Vegas. The study won’t be complete for another year, but Tom Fair says early data suggest that geographic dispersion helps dampen variability. A second finding is that solar facilities need to be placed on strong parts of the grid. “That leads us away from having huge amounts of PV at any one site,” Fair says. Ten to 20 MW at any one site might be the limit.

Southern California Edison already plans to scatter 1 MW and 2 MW rooftop PV installations across its service territory, part of its goal to deploy 250 MW of PV over the next five years. Minimizing transient spikes is one reason. A second is that transmission remains the No. 1 barrier to renewable energy growth in California, says Mike Marelli, the utility’s director of renewable and alternative power contracts. “We can implement smaller systems with little or no transmission” additions, he says.

In contrast to PV, concentrating solar thermal (CST) technologies use mirrors or lenses to focus solar radiation on a central receiver or pipe to heat, typically, a fluid. In turn, this fluid drives a more conventional steam cycle to generate electricity. To be effective, however, these systems require a consistent supply of high quality solar radiation. They also are less scalable than PV technology, largely because of the fixed size of the turbine.

A recent report from the World Resources Institute, “Juice from Concentrate: Reducing Emissions with Concentrating Solar Thermal Power,” says that direct insolation of around 5.5 kWh/m2/day is a minimum requirement for CST development. “Significantly higher DNI (available sunlight) is much preferred if costs are to be kept to an acceptable level,” the report said. Conditions in North America favorable enough to support CST are in the U.S. Southwest. Elsewhere, South Africa, Australia, Northern Africa, Spain, Brazil and parts of India and China have suitable conditions for CST development.

“The main thing making concentrating solar thermal bankable is the quality of the sunlight,” says Britt Childs Staley, one of the report’s authors. “You can’t site CST in Maine; you need a much higher quality of sun with higher radiation.”

Even Florida offers somewhat limited quality insolation, the Institute report says. That’s because higher levels of atmospheric water vapor disperse the radiation, reducing a Florida-based solar plant’s potential output. For example, a CST in Phoenix with six hours of storage would have a capacity factor of around 40 percent. An identical facility in Tampa would have a 25 percent capacity factor. The difference affects both facilities’ bottom-line economics. The Phoenix plant’s long-term real cost of electricity would be around 14.4 cents/kWh. The report said it would be around 23 cents/kWh for the Tampa plant.

“Given the reduced output and lower profitability of CST plants located outside the Southwest, it is unlikely that significant capacity will be installed in other parts of the country,” the report said.

Perhaps, but that hasn’t stopped Florida Power and Light from developing the 75 MW Martin Next Generation Solar Energy Center, which will be one of largest solar plants of any kind outside of California. The facility, near Indiantown, Fla., will also be among the first hybrid facilities to connect a solar facility to an existing combined-cycle power plant, providing solar thermal capacity that directly displaces fossil fuel usage. The project will consist of around 180,000 mirrors over 500 acres at the existing FPL Martin Plant site. Construction began late in 2008 with an in-service date expected in mid-2010.

(The World Resources Institute report pegged the long-term cost of electricity for a 500 MW pulverized coal power plant with an 85 percent capacity factor and a $2,290/kW capital cost at about 6.26 cents/kWh. A 200 MW trough CST with six hours of storage, a 40 percent capacity factor and capital cost of $6,044/kW would have a long-term cost of electricity of 15.36 cents/kWh. Using the federal investment tax credit, the same plant would have a long-term cost of electricity of 11.37 cents/kWh.)

Not surprisingly, CST developers focus their attention on sun-rich locations where utilities and large-scale developers can choose between both PV and CST technologies. “We don’t see it as an either-or,” says NV Energy’s Tom Fair when asked if his utility favors either technology. “Both are equally proficient at finding sites” around the state. (For caption and credit information, click on this image in the image gallery below.)

NV Energy has a purchased power agreement to take 20 MW of PV generated at a site in southwest Nevada. The utility is also looking at installing 250 MW of CST capacity with molten salt storage northwest of Las Vegas. It also is considering adding 80 to 100 MW of solar capacity at its gas-fired Harry Allen and Chuck Lenzie stations. If built, the integrated solar combined cycle power plant would essentially swap solar Btu’s for natural gas Btu’s, similar to FPL’s Martin Next Generation scheme.

The idea of incorporating a solar field with a natural gas-fired combined cycle power plant is gaining momentum, Fair says. “When you have direct normal insolation (shadow-producing sunlight) it’s a pretty interesting resource.” One benefit is that peak demand typically occurs on hot, sunny days, which is a good fit with the solar resource. Utilities also gain by making use of existing power blocks and transmission infrastructure. That helps minimize permitting scuffles, gets renewable capacity into the utility’s generating portfolio in a hurry and helps control capital costs.

And that’s no small thing, since until recently utility scale solar has been expensive; in some cases prohibitively so.

Falling Costs

“Costs have been coming down faster,” says WorleyParsons’ Kenny Beninga. Eighteen months ago, a utility-scale PV facility cost around $6,000/kW. Today that cost is closer to $3,500/kW. Some of the decline is due to technology improvements. But a lot more is the result of market dynamics.

A global shortage of polysilicon led many companies to start manufacturing PV. Then the recession hit. The economic downturn led to product oversupply and price declines of around 30 percent, says Renewable Ventures’ Matt Cheney. The market slump has been tough for many companies in the PV supply chain but offers opportunities for utilities.

“There has been a dramatic decrease in cost and prices over the past 12 months, which puts utilities in a strong position to bargain,” says Chris O’Brien, head of market development for Oerlikon Solar. The Swiss company launched its solar group in 2007, offering end to end manufacturing lines for thin film PV. O’Brien says first-generation thin film customers in Europe have a manufacturing cost of approximately $1.50/W for a thin film PV module. His company’s goal is to drive costs to around $0.70/W by the end of next year. With the current cost structure, including federal incentives, a 10 MW PV plant in California can have a delivered cost of electricity of around $0.15/kW. O’Brien says he expects that to fall below $0.10/kW by 2012.

The company, meanwhile, currently guarantees a 9 percent efficiency level for its PV system. By next year, it hopes to guarantee efficiencies in excess of 10 percent. That level is a threshold necessary to achieve economies of scale, says Sharp’s Ron Kenedi. Current industry efficiencies range from 5 percent up to 10 percent, with some companies aiming for 15 percent efficiencies. Kenedi says Sharp’s PV products are in the range of 9 to 10 percent efficient.

O’Brien says the current market oversupply is temporary as suppliers in some cases sell below cost to clear inventories. “It’s a tactical opportunity” for utilities right now, he says, adding that current conditions are not an indicator of where prices will end up once the market rights itself. A delivered cost of electricity in the range of $0.12/kW for utility-scale PV projects in California could be more realistic in the long term, he says.

The Finance Question

Utilities may also have a short-term advantage over independent power producers and developers due to the dearth of financing available for many types of electric power projects, including solar. What had been a $5 to $7 billion market in 2008 among banks and investors looking for tax credits generated by renewable energy projects has largely disappeared. Utilities are one of the few big sources of project capital, representing an investment pool that O’Brien says could be anywhere from $8 to $10 billion.

Current market conditions offer a “real win-win opportunity” for utilities to take advantage of federal tax credits aimed at boosting renewable energy deployment, he says. Matt Cheney isn’t so sanguine. He says credit is tight even for utilities and suggests it may take as long as four years for financial sector volume and pricing to recover from last year’s banking shock.

“Financing using traditional project financing is nearly impossible due to technology risk,” says Robert Rogan, senior vice president of North American markets for eSolar. Traditional tax equity financing no longer exists as it did 18 months ago as banks and insurance companies have either disappeared altogether or exited the market due to the broader financial market turmoil. “The key now is that for any plant the debt/equity markets have been upended in the past year,” he says.

eSolar successfully raised $170 million from Google.org, Oak Investment Partners, Idealab, NRG Energy and ACME Group of India. In early August it began delivering power to the grid from a 5 MW CST demonstration project in Lancaster, Calif. known as Sierra SunTower (pictured on our cover). And earlier this year it announced a string of project contracts to build scaled-up version of the SunTower for Southern California Edison, Pacific Gas & Electric and El Paso Electric.

“These are project we have been working on for some time,” Rogan says. He declined to discuss specific pricing, but said all of the deals are based on purchased power agreements.

eSolar’s commercial-scale CST is 46 MW and includes 16 power towers, which feed heat to drive a single steam turbine. The projects can be scaled up in chunks of 46 MW, which could appeal to the utility power generation market. Projects are also designed so they don’t require a 230 kV transmission connection; instead they can use 115 kV or 135 kV lines, which may make project siting somewhat easier. The projects also can accommodate storage, but Rogan says current U.S. market signals show little demand for that feature. He says utilities and grid operators feel confident they can manage resources and not pay for storage.

That may change as renewable energy penetrates farther into generating portfolios. “Storage is a huge piece of solar thermal in the future,” Rogan says.

Storage may present its greatest value to solar and wind power developments due to the intermittent nature of their power production. The wind doesn’t blow all the time and darkness is an effective deterrent to solar power generation. As larger amounts of these intermittent resources are added, the likelihood grows that a drop in wind velocity, passing clouds or darkness will have an effect on grid stability. Storage is one way that intemittency can be bridged and–for solar, at least–generation extended into the evening hours.

Kelly Beninga of WorleyParsons estimates that adding four hours of storage at a CST could raise construction costs by 10 to 15 percent. Of course, storage also adds to the amount of time the power facility is producing electricity so the costs can be recovered at least in part through extended operating hours.

The World Resources Institute report says the 64 MW Nevada Solar One CST was built without storage for around $4,200/kW. A comparable plant with six hours of storage could cost $6,400/kW, the report says. With storage, the levelized cost of generation can actually decline, the report said, since storage increases the annual generation output over which to spread the initial capital outlay. The report also confirms Rogan’s view that most planned projects in the U.S. do not include storage.

“Although storage might allow a plant to generate later in the day, if CST cannot compete with the cheaper plants that bid power at that time, it has no market and thus no revenues for such generation.”

Power towers are not the only form of utility-scale CST. Parabolic trough and linear Fresnel trough are also being deployed, for example, at Nevada Solar One. The technology may offer the most similarities to a conventional steam electric power plant, since the mirrors concentrate solar radiation to heat a fluid that ends up driving a fairly conventional steam cycle.

“That’s why we went with it,” says Christopher Huntington, vice president of business development for SkyFuel Inc., which designs and sells solar trough collector equipment. What’s more, with around 20 years of operating experience at utility scale, bankers see that the technology has an established track record. “They know it will work and that they can get their money back,” Huntington says.

Current parabolic trough technology uses oil as a circulating heat transfer fluid. Research is underway to substitute molten salt, which holds the possibility of pushing temperatures even higher. That’s good news, Huntington says, since “the bigger the temperature drop the more power you get.”

One operational issue with CST is the daily cycling the steam plant must endure. CSP plants face a daily startup cycle and some energy is lost to warm the turbine. This can result in perhaps as much as a 30-minute delay in terms of availability, says WorleyParsons’ Beninga. On the other hand, thermal inertia means the solar power system can continue to generate power into the evening hours, following sunset.

“The main issue is the lifetime reliability stress on the turbine” and in particular low-cycle fatigue in the turbine blades, Beninga says. One strategy is to keep the turbine hot overnight with an auxiliary boiler or a steam blanket, possibly increasing parasitic load.

Whether or not utility-scale solar can ever really be a part of baseload power generation remains to be seen. Beninga says its niche will be daytime and evening peak and mid-peak load. As such it could play a role as a natural gas displacer.

For a utility like Southern California Edison, the goal is not to favor one renewable energy technology over another. The state’s renewable portfolio standard is agnostic, says power procurement director Mike Marelli, although the utility finds more value in summer or peak load resources.

“We’ve done contracts for all five renewable technologies and we feel comfortable we’re getting good resource diversity.”

David Wagman is Chief Editor of Renewable Energy World North America Magazine.