LOVE, POKER & THE ECONOMICS OF SOLAR POWER PLANT ENERGY STORAGE

The Value of Concentrating Solar Power and Thermal Energy Storage
Ramteen Sioshansi and Paul Denholm, February 2010 (National Renewable Energy Laboratory)

THE POINT
In poker, love and energy, what is apparently true usually is. But isn’t, necessarily.

Case in point: Almost anybody who knows anything about operational or planned solar power plants (SPPs) incorporating concentrating solar power (CSP) technology will likely say that one of their big advantages is the potential to capture and store heat for use in generating electricity at the plant when the sun disappears behind a cloud cover or goes down at the end of the day.

Like the assumptions that the guy isn’t calling because he’s just not that into the girl and the poker player with the biggest pile of chips is the smartest bettor at the table, hard analysis shows this intuitively apparent assumption about SPPs to be true enough. The Value of Concentrating Solar Power and Thermal Energy Storage, from Ramteen Sioshansi of Ohio State University and Paul Denholm of the National Renewable Energy Laboratory, shows that energy storage usually adds value to SPPs.

click to enlarge

An SPP has 3 “independent but interrelated” parts: (1) the power block (the turbine), (2) the solar field (the mirrors), and (3) the thermal storage tank (that holds the heat). Sizing each effectively is crucial to making the plant profitable.

Unlike the solar photovoltaic (PV) technology commonly seen on rooftops that turn the sun’s light into electricity, SPPs capture the sun’s heat and use it to boil water and create steam that drives turbines. This generates electricity in the same way that electricity is generated by steam-driven turbines at plants that use heat from nuclear energy or the burning of coal and natural gas to boil water and create steam - without radioactive waste, greenhouse gas emissions or toxic byproducts.

The energy captured for storage at SPPs is therefore heat energy and the storage of it is called thermal energy storage (TES). Confirming the intuitively obvious, the paper concludes that TES adds value to SPPs (1) by allowing more thermal energy from an SPP’s solar field to be used, (2) by allowing an SPP to put a larger solar field to work, and (3) by allowing SSP generation to be shifted to times of use when energy prices are higher.

The study of SPPs in four states in the U.S. Southwest looks at how TES impacts SPP operations in a variety of ways and even reports that, though it is as counter intuitive as the possibility that the guy IS just too busy to call the girl or the poker player IS just lucky, there are ways that the expense of TES in the wrong circumstances can be too much of a burden on SPP economics to pay off.

Generally speaking, the cost of SPP technologies without storage is too high to be profitable in most electricity markets but adding TES makes them profitable by allowing for the production of more energy for longer hours. On the other hand, where the price of electricity is low and the value of additional energy sales does not boost revenues adequately, the cost of TES makes the technology inequitable.

Schematic of thermal energy storager (TES) (click to enlarge)

THE DETAILS
A big part of the renewed interest in solar power plants (SPPs) that utilize concentrating solar technology (CSP) comes from breakthroughs in thermal energy storage (TES). Without storage, the economics of SPPs are more tentative.

CSP is a method of using mirrors to concentrate the heat of the sun on a focal point. At the focal point, liquid is heated. Although some SPP concepts heat water directly, the 2 SPP concepts now most widely in use pipe molten salts to the focal point and carry them away to where the heat is transferred to boil water and generate steam.

Either the steam, heated water or hot molten salts can be stored for use when there is no sun. Molten salts are more commonly used as a heat-transfer fluid (HTF) because they more efficiently carry and hold the heat.

click to enlarge

With TES, SPPs can (1) hold their electricity for sale at the times of highest demand and highest prices, (2) more effectively replace conventional power sources instead of just being a peak demand supplement, and (3) serve as reserves quickly brought on line at times of fluctuating demand.

An SPP has 3 “independent but interrelated” parts: (1) the power block, (2) the solar field, and (3) the thermal storage tank.

The power block is essentially the turbine of the SPP.

The input of the power block is measured in megawatts of thermal energy (MW-t) and the output of the power block is measured in megawatts of electric energy (MW-e).

The solar field is essentially the mirrors of the SPP.

click to enlarge

The size of the solar field determines the MW-t that goes to the power block for any given area’s amount of solar energy per measured area (insolation or solar irradiance).

The size of the solar field and the potential of the power block set the SPP’s capacity. Matching them prevents under use of the turbine or the waste of reflected heat.

TES allows an SPP to have a larger solar field because if the power block does not use the heat the field generates, it can be stored.

The size of storage can be measured in MWh-t or simply in hours of stored operational capacity. The roundtrip efficiency of the best TES technologies is 98.5%, meaning only 1.5% of stored heat is lost as it comes back from storage to boil water to generate steam to drive the turbine (power block).

click to enlarge

The right combination of the 3 elements hinges on pricing in the electricity markets into which the SPP is selling. Where even the peak demand price of electricity is low, there is no marginal value to incurring the capital costs of building TES. There are electricity markets where the price is so low that the 1.5% loss from storage eliminates the profit margin from adding TES to an SPP.

The marginal value of SPPs with and without TES is also significantly affected by (1) anticipated improving technology and integrating its capabilities into the grid(optimization), (2) varying availability of sun, (3) the accuracy of energy price forecasting, (4) the ability of SPPs with TES to provide more value to the grid as spinning and/or back-up reserves, and (5) the impact on SPP operastional costs of saving water through dry cooling of the power block. Assessments of all these variables are included in the paper.

The 4 SPP locations used to make cost and profit estimates: (1) Gila Bend, Arizona, (2) Daggett, California, (3) Southern New Mexico, and (4) Western Texas. These 4 locations incorporate 4 separate electricity managers: (1) the Arizona Public Service (APS) utility, (2) the California Independent System Operator (CAISO), (2) PNM Resources, the New Mexico utility, and (4) the Electric Reliability Council of Texas (ERCOT).

click to enlarge

The paper goes into break-even and return on investment numbers for SPPs with and without TES in a variety of scenarios at the 4 locations.

Primary variables include (1) the unknowability of energy prices and solar resources over the 25-or-more year lifetime of SPPs and (2) the unknowability of future costs and capacities of SPP and TES technology.

Variables determining the marginal value of TES for SPPs: (1) 2009 cost estimates for present costs, 2015 estimates for future costs; (2) A solar field 1.5 times the capacity of the plant’s power block becomes twice the power block’s capacity with a 6- hour TES; (3) A 110 MW-e power block leads to an estimated present cost (2009) of $156.3 million for TES and a $117.6 million future cost (2015); (4) An SPP investment tax credit (ITC) could be 30% (the current credit) or 10% (a future scenario).

The break-even cost: The maximum overnight cost of TES that is justified by the increase in year-1 operating profits of the SPP.

click to enlarge

The return on investment (ROI): The percentage of the annual cost of the TES component recovered by the increase in year-1 operating profits. An ROI of 100% or greater means TES pays for itself by bringing in more energy revenues.

ROI variables: the SPP site, the ITC rate, and cost of TES.

SPP operating profits can be increased by selling spinning reserves and backup reserves services to the grid. By doing so, even with a 10% ITC, an SPP with TES can provide a greater than 100% ROI in markets with a moderate or above electricity price.

The research showed that locational profit differences are mainly due to differences in energy prices in the different system operator and utility systems. Differences in solar resource and longer term dispatch decisions were also found to be less significant.

click to enlarge

QUOTES
- From the paper on solar power plants and storage: “For a merchant CSP developer, the decision to build and the choice of the size of a CSP plant will be governed not only by the amount of solar energy available but also by the pattern (coincidence) of solar resource and by electricity prices. Clearly, high electricity prices and an abundance of solar resource are necessary for CSP to be economic, but a lack of correlation between solar availability and electricity prices can make CSP economically unattractive. TES can improve the economics by shifting generation to higher-priced hours, but this adds capital costs and some efficiency losses in the storage cycle.”

click to enlarge

- From the paper on solar power plants and storage: “…Using a model that optimally dispatches CSP with TES into existing electricity markets [in the Southwestern U.S.], we examine the potential operating profits (i.e., net revenues from energy sales, not accounting for fixed capital costs) that a CSP plant can earn. We show how these profits vary as a function of plant size. We also show the sensitivity of operating profits to different assumptions, including the possibility of selling ancillary services, the optimization process used, and the use dry rather than wet cooling. We show that while the current cost of CSP technologies make them uneconomic based on energy value alone, addition of TES improves the economics of CSP. We also show that when the value of ancillary services and capacity are taken into account, TES and CSP can be economic even with current technology costs.”

click to enlarge

- From the paper on solar power plants and storage: “The price of large-scale CSP is still highly uncertainty because of large fluctuations in commodities prices and the potential for substantial cost reductions from engineering and manufacturing improvements. Furthermore, the overall cost competitiveness of CSP will depend on changes in fuel prices and carbon policies…[T]he value of TES above its cost can help to increase the relative cost competitiveness of the entire CSP plant…[W]hether the increased revenues from TES will justify the cost of TES will be highly sensitive to the site of the CSP plant, ITC rate, and cost of the TES, but substantial cost reductions appear to be necessary to justify the addition of TES based on energy sales alone…the return on investment of the TES components, considering the ability to provide both spinning reserves sales and capacity credit…[and] only a 10% ITC…the return on investment (ROI) is greater than 100%, meaning that the value of TES is greater than its incremental capital cost. As a result, the incremental value above the cost of TES will improve the cost competitiveness of the entire CSP plant.”