Solar power has one major advantage over its more ubiquitous cousin wind power: electricity that is generated during peak demand hours (hot, sunny, air conditioned afternoons). Such makes solar attractive to utilities that value such capacity for peak shaving, cost aside.
The problem of wind is shown by this example. The Electric Reliability Council of Texas (ERCOT) leads the nation with more than 8,000 MW of installed wind capacity, yet their resource planning–tasked with keeping the lights on–“counts 8.7 percent of wind nameplate capacity as dependable capacity at peak.”
The limited usefulness of wind and solar is reflected by their low system capacity factors. For example, the capacity factor of a typical utility-scale photovoltaic (PV) or concentrating solar project (CSP) is still limited to about 25% compared to the average for U.S. nuclear power plants of 91.5% in 2008, with many nuclear plants operating at or above 100%.
Also, given the lower capacity factors, the amortized cost of transmission per unit of energy carried is almost four times as high given the wide difference in capacity factors. We explored this systematic problem earlier.
The physics of solar energy production, without subsidy, will continue to conspire to keep the first cost and operating costs of the solar option higher than conventional approaches to producing electricity, especially when the cost of transmission is included in the equation. The capital cost of all the solar technologies are about $6,000/kW and higher (sharp-eyed readers will note that I’ve increased this number from the $5,000/kW estimate provided in earlier posts—the reason is discussed shortly) and projects are moving forward only in particular regions within the U.S. with tough RPS requirements and large subsidies from states and the federal government.
In Part I, we reviewed the enormous scale and capital cost considerations of PV projects and then introduced the standard taxonomy of central solar power generating plants. By far the favored technology for utility-scale projects is the CSP option that either produces thermal energy used to produce electricity in the familiar steam turbine process or by concentrating the sun’s thermal energy on an air heat exchanger to produce electricity via an air turbine. In Part II, we reviewed a sampling of recent solar projects.
This final post explores the latest cost solar project cost data and then rising interest in hybrid projects that combines these two solar energy conversion technologies with conventional fossil-fueled technologies. Hybrid projects offer the opportunity for utilities to reduce fuel costs, while simultaneously helping utilities cope with onerous renewable portfolio mandates.
Creative Electricity Accounting
Renewable energy does generate a larger portion of the world’s electricity each year but the reported numbers are misleading. The Solar Energy Industries Association (SEIA, a trade organization that promotes solar energy technologies) recently released its 2008 Year in Review report wherein the organization estimated the solar industry growth over the past year. According to SEIA’s number, the total capacity of the solar industry grew by 1,265 MW in 2008, up from 1,159 MW installed in 2007, a modest increase. However, since my first post in early October where I first referenced this report, a closer look at the numbers reveal much creative accounting in SEIA’s numbers. Their mistake, and it’s a doozie, is they sum the electrical production of a photovoltaic (PV) and concentrating solar power (CSP) systems that produce electricity with the thermal energy production of solar water heating. No can do.
According to SEIA, at year-end 2008, the U.S. had about 8,800 MW of installed solar capacity. This included about 1,100 MW of PV, 418 MW of utility-scale CSP, at least 485 MWth (megawatts thermal equivalent) of solar water heating systems, and over 7,000 MWth of solar pool heating systems. In sum, SEIA’s accounting of solar energy systems installed in the U.S. are 80% solar water heating systems. In the footnotes of their report, SEIA used a conversion factor of 0.7 kW per square meter to equate thermal energy with electricity, two distinctly different forms of energy.
This apples-to-oranges comparison obscures the true picture of solar electricity system growth and is just bad science. Overall, the total installed solar capacity numbers SEIA produces aren’t comparable to the nameplate values on conventional generators we historically use as our measure of installed electricity generation capability. We are fortunate that the EIA understands these differences and does not engage in similar creative accounting with their solar electricity energy generation reports (here). At a 25% capacity factor, the EIA numbers indicate about 1,800 MW equivalent of installed solar electricity capacity in 2008.
For those engineers reading this, yes, the conversion factors do have units consistency but the Second Law of Thermodynamics describes the quality of different forms of energy and tells us that you cannot merely add the energy content of warm water heating your swimming pool to that of electricity. As the SEIA data is compromised, where do we turn to get a better accounting of the economics and experience of utility-scale solar energy systems?
PV Installed Cost (Revisited)
Late in October, President Obama touted the value of solar energy while visiting a newly commissioned plant in Florida. “At this moment, there’s something big happening in America when it comes to creating a clean-energy economy,” the president said while visiting the country’s largest solar plant: Florida Power & Light’s new DeSoto Next Generation Solar Energy Center southeast of Tampa. “But getting there will take a few more days like this one, and more projects like this one.”
I truly hope that Obama’s words are not prophetic. The 25 MW DeSoto plant cost $152 million to build—that’s over $6,000/kW installed—the reason I updated the opening paragraph to $6,000/kW from $5,000/kW used in earlier posts. The FPL press release states that the plant is expected to produce about 42,000 MWh a year—which calculates out to a 19.1% capacity factor, less than the average of such plants in the U.S. If FPL uses a conservative levelized annual capital cost of 15%, then the raw cost of electricity coming from this plant is 54 cents/kWh. The EIA data for Florida shows that the average residential cost of electricity this year is 12.37 cents/kWh.
Since those early posts, other data points on the cost of solar energy have emerged.
A report released by researchers at Lawrence Berkeley National Laboratory (LBNL) in October titled Tracking the Sun II, The Installed Cost of Photovoltaics in the U.S. from 1998-2008 is perhaps the most comprehensive review available of actual installed costs of PV systems. In 2007 and 2008, the time when about half of the survey population’s rated power capacity was installed, the installed cost for systems greater than 750 kW ranged from about $6.00 to $8.00/WDC. Capacity-weighted average costs were $7.5/WDC in 2008. (The electrical output from the PV panels is in volts, direct current. The direct current must be converted to alternating current using an inverter and then the voltage stepped-up using a transformer to produce distribution-level voltage to connect with the grid.) Electrical losses from the panels, through the transformers to produce AC power suitable for the grid are typically about 20% making these historic estimates consistent with the EIA estimates of the installed cost of PV systems—close to $6,000/kW in 2011.
Large system costs (greater than 2 MW), according to the report, range from about $5.5/WDC to as high as $7.6/WDC, averaging $7.1/WDC which is equivalent to about $8,500/kW thereby illustrating the value of scale economics. These plant economics are difficult to directly extrapolate to similar projects given the expected capacity factors (averaging about 20%-25%) vary, whether the system uses a fixed, single, or double axis design, the level of state and federal subsidies and tax credits available, the forecasted renewable energy credit revenue, and the terms of the power purchase agreement. The report did quantify the distribution of the installed costs: 6% was for the inverter, 54% for the modules, and the remaining 39% was for “other” costs including overhead and transaction costs that don’t necessarily benefit from project economies of scale. Project specific costs, such as transmission, that can run up the real cost of the project much higher, were also omitted. In total, the number of utility-scale projects willing to report hard installation or operating costs was nil.
Table 3 from the LBNL report is included below. To estimate the first cost of these projects in $/kW, add 20% to the Installed Cost column and multiply by 1,000. Just remember that there are many project costs in the “Other” category that do not appear in these estimates.
CSP installation cost data is much more difficult to obtain given that there are few utility-scale projects. Note that there are only two projects other than the well-known SEGS projects in the U.S. (More project details on the Arizona Public Service Saguaro Plant is available here and Nevada Solar One here). While recently visiting another plant close to Acciona’s Nevada Solar One, I learned that the plant recently installed quite a few additional solar collectors in order to produce their contract-required electrical output. Details on the upgrades were not available.
Hybrid Solar Cycles Are More Economic
Beyond the cost penalty, another limitation of solar technologies is dispatchability. There are few renewable options available to a dispatcher on a still, overcast day when the public demands electricity. Fast-acting combustion turbines will have the advantage over renewable energy supplies when instantaneous matching of supply with demand is required — at least until some form of energy storage mitigates the intermittent nature of renewable energy sources. However, progress to commercialize large-scale energy storage technologies has been evolutionary, rather than revolutionary, and many technical and cost issues are yet to be resolved.
The EPRI, while taking note of the high cost of solar power alternatives, also suggested that a hybrid approach may have “the lowest-cost option for adding solar power to the generation fleet, as it utilizes existing plant assets. And because the highest-intensity solar energy typically is within a few hours of peak summer loads, it makes solar augmented steam cycles an attractive renewable energy option.”
The hybrid approach takes existing fossil-fueled power plant technologies and integrates renewable power sources, such as CSP or PV systems, with existing conventional steam plants (Figure 1) and existing or new combined cycle plants (Figure 2). The resulting hybrid plant will either increase its power generation (within existing permit limits) or more likely be used to reduce fossil fuel consumption (perhaps helping to justify the high capital costs of solar technology), mitigate the intermittent nature of most renewable technologies, remain dispatchable, and help many utilities with large fossil plant infrastructure meet their renewable energy mandates.
A typical combined-cycle plant is suitable for including in an integrated solar combined cycle (ISCC) configuration, where the solar energy portion of the plant can provide additional power at peak demand. The conversion of a combined-cycle plant to an ISCC begins with adding an additional source of heat, such as solar energy, to reduce natural gas consumption and thereby improve overall plant efficiency.
1. Conceptual design of a hybrid coal-fired pulverized coal power plant using steam produced by CSP equipment to improve the efficiency of the plant. Source: EPRI
2. Conceptual design of a hybrid natural gas-fired combined cycle plant using steam produced by CSP equipment to improve the efficiency of the plant. Source: EPRI
Dash for Gas Will Begin Anew
Many electricity industry observers are now projecting that most short-term plant construction will likely be natural-gas fired combined cycle plants that are quick to build, highly efficient, and relatively easy to permit. This new-found enthusiasm began with a report from the U.S. Potential Gas Committee that the natural gas reserves of the U.S. were much more than expected. Wholesale prices are less than $3 per million Btu compared to just a couple of years ago when natural gas prices peaked well above $10 per million Btu. I expect that this new-found interest in natural gas-fired plants will also push interest in ISCC modifications by utlities to meet expect rising electricity demand while simultaneously meeting RPS requirements.
There are other advantages of an ISCC, even when compared with standalone CSP-inspired plant designs. For example, the ISCC uses existing components (such as steam generators, steam turbine, and condensing system) that reduce the installation cost of a typical CSP system. Also, the potential for generation is increased because the steam turbine would be already synchronized to the grid when the solar energy contribution is added, thus avoiding lost generation during start-up. Another key advantage is gained during rising ambient temperatures, when gas turbine performance steadily drops. Operation of the solar energy portion of the ISCC compensates for that loss in efficiency and electricity production and improves the plant’s part-load performance.
Combining solar energy with conventional coal-fired plants is also possible in regions with reasonably good solar conditions. For these plants, where the steam pressures and temperatures are higher than for ISCC, the type of solar conversion technology used (Fresnel, parabolic trough, or tower) will dictate how solar is integrated into the plant.
Finally, don’t discount the possibility of hybridizing conventional plants with other, even multiple, forms of renewable energy such as biomass and wind. Our discussion of ISCC illustrates a single development path electric utilities could follow to efficiently and inexpensively bring multiple forms of renewable energy online in short order. Many other options are available, depending on the design of existing plants and their location particulars.
Two Studies and Two Plants Seeking Permits
EPRI has two solar augmentation projects underway—one at a natural gas-fired combined cycle plant and a second at a coal-fired plant. Tri-State Generation & Transmission Association, Progress Energy and Southern Company are participating in the project, with case studies to be conducted at Tri-State’s 245-MW coal-fired Escalante Generating Station in Prewitt, N.M., and at Progress Energy’s 742-MW Mayo Plant in Roxboro, N.C. “The studies involve adding steam generated by a solar thermal field to a conventional fossil fuel-powered steam cycle to offset some of the fuel required to generate electric power,” according to EPRI.
“The projects will provide a conceptual design study and two detailed case studies. Design options to retrofit existing plants will be analyzed and new plant design options will be identified. EPRI will rely on solar technologies, steam cycles and plant operation, as well as past solar and fossil plant studies. EPRI holds two patents in solar steam cycle optimization,” as EPRI describes the projects. The results of both studies are expected to be completed by the end of 2010.
According to the NREL press release, “During the case study, the system’s performance will be analyzed by solar thermal research engineers at the U.S. Department of Energy’s Sandia National Laboratory in Albuquerque, N.M., and the National Renewable Energy Laboratory (NREL) in Golden, Colo. NREL works with industry to analyze the cost and performance of solar systems, design new technology for solar electricity generation and look for ways to improve the performance, reliability and service life of systems and their key components.”
“One of the cost-effective aspects of concentrated solar power is that it can be fitted to an existing power plant to make electricity in tandem with fossil fuels,” said NREL’s Mark Mehos, CSP principal program manager. “And those existing plants already are connected to the transmission grid, making the technology relatively easy for utilities to work with. That can only help to expand the use of renewable energy.”
The hybrid plant approach is already getting traction with plant developers in California eager to build something, anything. Currently, the California Energy Commission (CEC) is reviewing a proposal by the City of Palmdale to build a new power facility similar to the EPRI concept, with a solar thermal facility providing 10% of the peak power generated by a 570-MW natural gas-fired power plant. If approved, the Palmdale Hybrid Power Project will start operating in 2013. Last year, the CEC approved a similar project in Victorville that will integrate 50 MW of solar thermal energy into a 563-MW natural gas-fired power plant. Victorville 2, developed by Inland Energy, was approved by the California Energy Commission in July 2008. A similar project, the Palmdale Hybrid Power Plant, has been proposed by the City of Palmdale.
One can guess how these projects will fare relative to tried-and-true generation alternatives that are market tested and market chosen–not very well. After all, the energy in the sun’s stock–oil, gas, and coal–is much more intense that the sun’s dilute flow. Technology has not and probably will not bridge this gap anytime soon.