This is the third in a series of blog posts discussing the water-energy nexus.
Large-scale solar power is coming to the United States. After much debate about water conservation and land preservation, the California Energy Commission (CEC) recently approved plans for nine concentrating solar power (CSP) plants in the state. Worldwatch found that this group of proposed plants will consume much less water per megawatt-hour than if California was to build typical coal, natural gas, or nuclear power plants instead. These CSP plants minimize water use through dry cooling.
The CEC expedited its commissioning process this fall to approve the CSP projects before their eligibility for a 30-percent cash grant from the U.S. Treasury’s 1603 Program expired at year’s end. Luckily for U.S. solar developers, Congress has since extended the program through 2011, as I discussed in a previous post. The approved projects represent a total of 4.4 gigawatts (GW) of new installed solar power capacity (or 4.6 GW if a tenth project, which should be approved shortly, is included).
The new CSP projects would increase California’s grid-connected solar power capacity fourfold from the 2009 level. Considering that California has the largest solar capacity in the country—about 10 times larger than any other U.S. state—this is all the more impressive. The development will dramatically assist California in meeting its renewable portfolio standard (RPS) target of generating 20 percent of its electricity from renewable sources by 2010, and 33 percent by 2020. More broadly, these projects would increase the nationwide supply of non-hydroelectric renewable power by over 9 percent, based on 2009 figures from the U.S. Energy Information Administration.
Water consumption has been a central issue in the commissioning process. Mindful of water usage, the CEC has recommended the use of dry cooling systems for the new CSP projects, which in some cases could reduce water consumption by up to 90 percent with minor efficiency losses and added costs (as I discussed in a previous post). Of the ten CSP projects in the pipeline, eight rely on dry cooling.
Compared to conventional sources of electricity, CSP with dry cooling is more water efficient. Our calculations suggest that the average lifecycle water efficiency of the 10 new plants would be 120 gallons per megawatt-hour (gal/MWh), even including the two water-intensive plants that use wet cooling towers. [Water consumption during construction, which accounts for a small portion of lifetime use, was not included in this estimate.]
For contrast, estimates from a recent Worldwatch briefing paper indicate that a typical Powder River Basin coal power plant with wet cooling uses 523–1,084 gal/MWh, and a typical natural gas combined-cycle power plant with conventional gas and wet cooling uses 152–525 gal/MWh. Meanwhile, a Congressional Research Service study indicates that a typical nuclear power plant with wet cooling uses 475–900 gal/MWh.
Clearly, CSP plants with dry cooling systems offer California commercial-scale power with minimal water usage, but not all of the new projects are created equal (see Full CSP Spreadsheet). Below, I discuss the differences in cooling and generation technology for the 10 proposed projects:
When the CEC approved NextEra’s 250 MW Beacon Solar Energy Project at the end of August, it was the first all-concentrating solar plant to be commissioned in California in 20 years. Like most of the other proposed projects, the Beacon plant will use parabolic trough technology. Here, sunlight is concentrated into a focal point at the center of the collectors, and vacuum tubes that run the length of the trough transfer a heating fluid (usually oil or water heated to 750 degrees Fahrenheit) to the boiler to make the steam needed to generate electricity.
Yet Beacon is one of the two proposed CSP plants that will not use a dry cooling system. Instead, it will have a wet cooling tower, where the low-pressure steam exhaust from the Rankine cycle generator-turbine is cooled by evaporation after passing through a heat exchanger. Potentially, half or more of the water in the cooling tower is lost to evaporation, creating the need for additional (makeup) water, but any remaining water will be condensed and collected for reuse. Over 90 percent of the project’s water needs will be met with recycled water from nearby wastewater treatment facilities in the towns of California City or Rosamond. Nonetheless, the plant will still require 456 million gallons annually to operate and cool, equivalent to the annual water consumption of 5,600 families in the U.S. Southwest.
The Abengoa Mojave Solar Project was the second commercial-scale CSP plant to be commissioned this fall. Like the Beacon project, it uses a 250 MW parabolic trough system; however, it will be substantially more water efficient, consuming some 277 million gallons per year due to advanced water recycling and reuse techniques. The plant will collect the condensate from the generator’s reject streams after being run through reverse-osmosis filters in a service water-storage tank for recycling. Further water storage will be available in a clear well when discharge exceeds the treatment system demand. Lastly, the treatment system will draw recycled service water first rather than pumping virgin makeup water from the site’s groundwater wells.
Developers at Solar Millennium LLC have been involved in several CSP projects in the U.S. Southwest. Solar Millennium’s recently approved Blythe Solar Power Project in southeastern Riverside County comprises four adjacent 250 MW parabolic trough systems with a combined capacity of 1 GW—roughly the capacity of an average nuclear power plant. The project is so large that it will nearly double the current U.S. installed CSP capacity. Additionally, Blythe is the first CSP plant in California to propose using dry cooling—specifically, a closed-loop apparatus known as an air-cooled condenser (ACC). Despite its size, Blythe will consume less water than either Beacon or Abengoa Mojave. Every year, Blythe will use 196 million gallons of water from 10 groundwater wells for feedwater makeup, mirror washing, and on-site domestic use. Each of the four power plants will also have an auxiliary closed-loop wet cooling system for ancillary equipment.
Unlike most of the other CSP projects, the 370 MW Ivanpah Solar Electric Generating Station will use a central power tower rather than a parabolic trough system. Sunlight is reflected from the mirrors to a single receiver atop the tower in order to heat a transfer fluid that is used to create steam to run the generator-turbine below. The Ivanpah project will use dry cooling via an ACC, enabling the plant to consume only some 33 million gallons of water per year. It is worth mentioning that Ivanpah was the first project to file for CEC approval (back in 2007), but it hit resistance from environmental activists who were concerned that the project would displace a large population of desert tortoises from their native habitat in the Mojave Desert.
The Imperial Valley Solar Energy Project will use a third type of CSP technology. The 750 MW project, formerly known as the Stirling Energy Systems (SES) Solar Two Project, will rely on 30,000 mirrored parabolic dishes that power the solar Stirling engines to produce electricity. Each dish concentrates sunlight on a receiver that is housed in a power conversion unit (PCU) that sits on a boom above the dish and heats the working fluid—in this case, hydrogen rather than steam—to run the Stirling generator-turbine, also located in the PCU. The project uses SES’s unique dish-engine units, called SunCatchers™, which operate independently and generate electricity without water. The only water consumed at the plant will be primarily for mirror washing and demineralization processes in the treatment system. The Stirling engine’s small closed-loop dry cooling system also may require minimal water to compensate for leakage. The hydrogen will be separated from water through electrolysis, requiring only negligible water input. Total annual water consumption at the plant is estimated at 11 million gallons, roughly the usage of 131 Southwestern families. However, Imperial Valley has the largest land footprint of the new CSP projects, occupying 6,500 acres.
The Genesis Solar Energy Project originally proposed using a wet cooling system, but the CEC reviewers deemed that dry cooling would be feasible for the project and called for the developers to replace the wet cooling tower with an ACC. This decision will reduce the plant’s annual water usage more than 87 percent, from 521 million gallons to only 65 million gallons. But the switch will also increase construction costs and decrease the plant’s electricity output, especially during the summer when hotter ambient temperatures strain the dry cooling system’s efficiency.
Originally part of a larger group of projects that included the Imperial Valley plant, the Calico Solar Power Project (formerly the SES Solar One Project) was taken over by different developers and has a new operating capacity of 850 MW. Like Imperial Valley, Calico will use Stirling engine and dish technology, but it will have 34,000 units. Because Stirling engines are more efficient than the Rankine engines used at other proposed solar thermal plants in California, a Stirling plant can generate more electricity per year with the same capacity and acreage. Calico also has the lowest estimated water consumption of any of the approved projects, at 6.8 million gallons per year, which is even lower than Imperial Valley despite being larger in size. One major drawback of Calico is that it requires 4,600 acres of land, which will have negative effects on the local environment including the clearing of vegetation and the dislocation of wildlife.
A few weeks ago, it was doubtful that the Palen Solar Power Plant would even be able to start construction or pay the 5 percent of its equipment costs required to receive the U.S. Treasury 1603 cash grant before the December 31 deadline. But when Congress extended the 1603 program, it meant the Palen project would be able to move ahead with funding from the cash grant. As a sister project to Blythe, the Palen plant is also being developed by Solar Millennium in Riverside County, but it will have only half the capacity, at some 500 MW. Like Blythe, Palen will use an ACC for dry cooling but will consume no more than 96 million gallons of water per year for operation and cooling.
The Rice Solar Energy Project also might have died without support from the 1603 program, but both the Rice and Palen projects were approved on December 15, and now stand to receive the incentives from the 1603 extension. Rice will be the second new CSP plant to have a central power tower (like the Ivanpah plant), but it will use molten liquid salts as a heat-transfer fluid rather than synthetic oil or super-heated water. Because the salts retain heat extremely well, this gives the plant thermal storage ability, making it the only plant out of the 10 proposed to be able to generate electricity at night. Rice also has the smallest capacity of the proposed plants, at only 150 MW, although its longer operating hours should make Rice’s net generation comparable to the others. It will rely on a dry cooling system, consuming an estimated 49 million gallons of water annually.
The Ridgecrest Solar Power Project is the only project currently in the CEC’s pipeline that has yet to be approved. Developed by Millennium Solar, it proposes using a parabolic trough system and has an installed capacity of 250 MW—about a quarter the size of Blythe and half the size of Palen. Ridgecrest will be located in northwestern Kern County and will use an ACC for dry cooling, consuming roughly the same amount of water as the Rice project, at 48 million gallons per year. Considering how quickly the other Solar Millennium projects have moved through the commissioning process, the Ridgecrest project will likely be approved by the CEC in the next month or so.
Figure 1 provides a comparison of the 10 CSP projects. For reference, my full spreadsheet of data and calculations is available here.
The massive proposed development of CSP in California will have vast implications for the U.S. renewable energy industry. The CEC has done its best to ensure that water resources in the arid Southwest will not be exacerbated by this CSP boom. California has not just laid out a model for renewable energy growth—it has created a paradigm for sustainable energy development that minimizes water consumption.