American Scientist

A reprint from

American Scientist

the magazine of Sigma Xi, The Scientific Research Society

This reprint is provided for personal and noncommercial use. For any other use, please send a request to Permissions, American Scientist, P.O. Box 13975, Research Triangle Park, NC, 27709, U.S.A., or by electronic mail to [email protected]. ©Sigma Xi, The Scientific Research Society and other rightsholders

Technologue

A Path for Nuclear Power A novel but tested technology, the pebble-bed reactor, can make fission energy safe. Lee S. Langston

E

lectric power is like good health: When you have it, you don’t think about it. When you don’t have it, that’s all you think about. Certainly modern civilization isn’t going anywhere without power. Nuclear energy supplies a larger share of that power than many people realize. At present, about 20 percent of the electricity in the United States is generated by 104 nuclear power plants across the country, making it the leading nation in total installed nuclear capacity. There are now 440 reactors in operation around the world, providing about 14 percent of the overall electricity supply. France leads in nuclear use, which provides about 80 percent of the country’s electrical power. China, presently at 2 percent, has 30 or more new plants under construction. In recent years, political opposition and high costs have halted or even reversed the use of nuclear power, however. The first plant in the United States— the Shippingport Atomic Station, located on the Ohio River 25 miles from Pittsburgh—went online in 1957. Many more started up in the next two decades, but until a few years ago, no new nuclear plants had been licensed here since the 1970s. The commercial nuclear power plants now in operation were generally designed to have a 40-year life span. With new construction largely at a standstill, many plants are being granted 20-year license extensions by the U.S. Nuclear Regulatory Commission.

Lee S. Langston is a professor emeritus of mechanical engineering at the University of Connecticut. He received a Ph.D. in 1964 from Stanford University. He was a research engineer with Pratt & Whitney Aircraft working on fuel cells, heat pipes and jet engines from 1964 to 1977. For the past 10 years he has written an annual review of the gas turbine industry for Mechanical Engineering magazine. Email: [email protected] 90

American Scientist, Volume 102

Now the risks of climate change are prompting utilities and federal officials to take another look at nuclear. According to the Department of Energy’s Energy Information Administration, 37 percent of the electricity in the United States was produced by coalfired plants, which release high levels of carbon dioxide. Natural gas, which yields about half as much carbon dioxide as coal per unit of energy generated, accounted for another 30 percent. There are only two major zero-carbon components in the United States energy mix. Nuclear supplies 19 percent. And all renewables—including hydroelectric, wind, and solar—collectively account for 16 percent. The Southern Company has begun construction of two new nuclear units at the Vogtle site, where another two are already operating, on the Savannah River near Augusta, Georgia. They are the first new domestic fission-power plants in more than three decades. But a broader turnaround is unlikely unless nuclear engineers can address two major issues: safety fears (which were compounded by the Fukushima nuclear accident) and high construction and licensing costs. A tested technology, known as the pebble bed nuclear reactor, has the potential to solve both problems—and to make nuclear power a growing part of the carbon-free energy mix worldwide. Energy Economics 101

Nuclear power plants are expensive to build but relatively cheap to run. Current estimates for the capital costs of a nuclear plant run in the range of $5,000 to $6,000 per electrical kilowatt. By comparison, coal plant capital costs are about $2,000 per electrical kilowatt. Highefficiency, gas turbine combined cycle plant costs are in the $600 to $1,000 per electrical kilowatt range. Simple cycle gas turbine plant capital costs are as low as $300 to $700 per electrical kilowatt.

The Energy Information Administration’s Annual Energy Outlook for 2014 report provides more meaningful estimates of the averaged levelized costs for generating technologies to be brought online in 2018. (Levelized cost represents the per-kilowatt hour cost of building and operating a generating plant over an assumed financial life and duty cycle). The projected levelized cost for an advanced coal plant is 12.3 cents per kilowatt-hour, whereas that of advanced combined cycle gas turbine plant is 6.6 cents per kilowatt-hour. An advanced nuclear plant, such as the ones under construction at Vogtle, comes in between, at 10.8 cents per kilowatt-hour. Nuclear is competitive, then, and a future carbon tax on technologies that produce carbon dioxide pollution could make it even more so. One major controversy with nuclear power is what happens to spent fuel when a plant is retired. A good example of a cradle-to-grave story of a nuclear power plant is the Connecticut Yankee plant, located along the Connecticut River. It went online in 1968, and over its service life of 28 years it produced over 110 billion kilowatthours of electricity from its uranium fuel. The plant was decommissioned from 1998 to 2007, with all structures removed from three to four feet below ground level. The site is now a fairly flat field, with a low mound where the reactor building sat. Located on a hill above the plant site is the Independent Spent Fuel Storage Installation. It consists of 43 dry storage casks mounted on a concrete pad that is 30 meters long by 60 meters wide and a meter thick. Forty of the casks contain almost all of the fuel assemblies used over the life of the Connecticut Yankee plant; the other three contain reactor vessel parts. Each cask is a vertical concrete cylinder, about four meters in diameter and six meters high, containing

© 2014 Sigma Xi, The Scientific Research Society. Reproduction with permission only. Contact [email protected].

helium coolant in

5 millimeter graphite layer 0.0035 millimeter silicon carbide barrier coating

0.004 millimeter pyrolitic carbon layers

60 millimeter diameter fuel sphere

helium coolant out

pebble bed reactor core

0.92 millimeter diameter coated fuel particle

a cylindrical steel canister with ninecentimeter-thick walls. The casks, which cost $1 million apiece to construct, are licensed for a 20-year storage time, but could last 40 to 60 years or longer. In the future, a federal agency might take possession of the installation. If that happens, the casks might simply stay in place, considering the significant expense of moving them to a permanent storage site—if one is ever chosen. That might not be a popular decision politically, but financially the decommissioning process has been taken into account. The cost of maintaining the installation is about $3 million a year. All of the plant decommissioning and spentfuel storage costs had been factored into electric rates paid by customers during the plant’s 28 years of operation. Avoiding Another Fukushima

Overall, the safety record of the hundreds of nuclear power plants around the world has been outstanding. But as asserted by industry critic Arnold Gundersen, “Nuclear power is a technology that can have 40 good years and one bad day.” The industry had a very bad day on March 11, 2011, when magnitude 9.0 earthquake struck at 2:46 PM local time in the Pacific Ocean seabed, 80 miles east of the Japanese coastal city of Sendai. Nuclear power plants in the northeast part of Japan automatically shut down, with control rods inserted into the reactor cores. The 4,700-megawatt www.americanscientist.org

0.0095 millimeter porous carbon buffer

coated particles embedded in graphite matrix

0.5 millimeter uranium dioxide fuel kernel

A pebble bed nuclear reactor is schematically depicted (left) as a funnel-shaped container through which tennis-ball–size “pebbles” of fuel are circulated. Each pebble (middle) is made of graphite and carbon and contains about 15,000 kernels of fuel (right). Each kernel has a core of uranium that is coated with a layer each of porous carbon, pyrolytic carbon (which is similar to graphite), silicon carbide, and then a second layer of pyrolytic carbon. These layers are used to moderate the speed of neutrons, helping to control the nuclear reactions of the fuel, and to provide a fireproof seal.

Fukushima Daiichi nuclear power plant complex was one of these. The complex, located on the coast some 150 miles north of Tokyo, is made up of six separate boiling water reactor units. These units were designed to withstand a magnitude 8.2 scale earthquake, comparable to the 1906 San Francisco event. The March 11 earthquake was seven to eight times more powerful than that. Of the six Fukushima Daiichi reactors, Units 5 and 6 were offline for planned inspection and Unit 4 had been completely defueled. Unit 1, with a nominal output of 498 megawatts, and Units 2 and 3, both of which put out 796 megawatts, were in operation before their earthquake induced automatic shutdown. Even with control rods fully inserted, these three units still needed cooling from an electric-powered circulating water pump, due to the residual heat generated by intermediate radioactive elements created in uranium fission. At 3:44 PM, a 15-meter-high tsunami reached the Fukushima Daiichi complex, overtopping facilities designed to withstand a 5.7-meter tsunami. Both the offsite and onsite emergency diesel generators were knocked out by floodwaters, depriving the reactor cores of

cooling water. Once the backup battery power ran out, the reactor coolant water overheated, increasing pressures and temperatures in the reactor pressure vessel. Pressure was automatically relieved to a suppression pool designed to condense steam, but this reduced the amount of water in the reactor. The fuel rods continued to heat up, and a reaction between the zirconium fuel rod cladding and the remaining coolant generated hydrogen. The hydrogen eventually escaped from the reactor pressure vessel, causing explosions and further damage. The inability to provide reactor coolant, despite multiple-layered backup systems that all depended on AC power, led to the meltdown of the reactor cores in Units 1, 2, and 3. The root of the problem was that the backup systems were overwhelmed by disasters worse than the worst-case scenario they were designed to protect against. Similar damage occurred in 1979 at one of the two nuclear units at Three Mile Island in Pennsylvania, due to operator error when an emergency cooling system was turned off. In a New York Times interview, Lake H. Barrett, the senior U.S. Nuclear Regulatory Commission engineer for Three Mile


2014

March–April 91

The Pebble-Bed Alternative

Both Three Mile Island and Fukushima suffered melted-down reactor cores because of a loss of coolant. Multiple backup systems guarded against such a loss, but failure occurred anyway. In response, the newest nuclear power plants (called Generation III and III+ plants) incorporate passive safety features, which require no operator intervention or active controls in the event of an operational emergency. But as an engineer I would be more trusting of a nuclear reactor that would use its own nuclear reactions to shut itself down if reactor coolant is lost. A high-temperature reactor that uses helium instead of water can do just that. Uranium fuel—a mix of uranium-235 and uranium-238—is contained not in rods but in tennisball–size spheres called pebbles. The helium heated by the pebbles can be used to generate steam (as in watercooled reactors) to drive a turbine, or used directly to power a gas turbine to generate electricity. 92

American Scientist, Volume 102

Wikimedia Commons

Island, noted that cleanup of the site— which took 14 years and the removal of about 150 tons of radioactive rubble— “was a walk in the park compared to what they’ve got” in Japan. Barrett is now an adviser for Fukushima’s cleanup. Writing in the September 9, 2013, issue of the Bulletin of Atomic Scientists, he outlines the herculean measures the Japanese company is taking to cool the melted-down reactor cores in Units 1, 2, and 3, and the spent fuel pools in Unit 4. It involves the circulation and recycling of radioactive water—some 90 million gallons and growing—into more than 1,000 temporary water storage tanks, through tens of miles of piping throughout the four earthquake-damaged plants, each of which is also being affected by ground water leakage. Barrett notes that such a program of containment cannot continue indefinitely, bringing up questions about a long-term solution to the problems at Fukushima. Although the measureable health impact of the Fukushima accident has been thankfully small, the social, political, and economic effects have been vast. Experts estimate that a 40-year decommissioning of Fukushima could cost $100 billion. That figure doesn’t even account for the destruction of land values and productivity in the region and other less tangible economic factors.

A demonstration pebble bed nuclear reactor was built in Jülich, West Germany, and ran successfully from 1966 to 1988, but was not followed up with a commercial version.

One such high-temperature, gas cooled reactor was under development in South Africa by Pebble Bed Modular Reactor (Pty), a company with international participation by Westinghouse (from the United States), Mitsubishi (from Japan), and Nukem (from Germany). A series of modular reactor nuclear power plants using this design were to be built in South Africa, but due to the 2008 global financial crises, the program was canceled in 2010. In July 2007 I visited Pebble Bed Modular Reactor’s headquarters in Centurion, near Johannesburg, where they had an engineering and management team of about 700. I also toured their test facility at Pelindaba, the location of South Africa’s state-owned Nuclear Corporation. Pebble bed reactor nuclear fuel is formed in such a way as to moderate fission, contain pressure buildup, and accommodate fuel deformation. Uranium dioxide nuclear fuel, coated with mass diffusion and radioactive fission product containment layers of pyrolytic carbon and silicon carbide, is formed into poppy seed–sized fuel particles. Some 15,000 of these are embedded in a graphite sphere, which is encased in a thin carbon shell, sintered, annealed, and machined to a uniform diameter of 6 centimeters. In engineering designs, the cylindrical pebble bed reactor vessel, 27 meters high and 6 meters wide, is packed with about 450,000 pebbles. Due to gravity, the pebbles gradually filter down from the top

to the bottom of the packed bed. At the bottom, the pebbles fall into a chute and are moved pneumatically, from helium pressure, into inspection equipment that checks their physical integrity. Pebbles that are intact are returned to the reactor vessel, whereas those that are spent are routed to a holding tank. A typical pebble will make six loops through the reactor during its three years of fuel life. The process allows for refueling with new pebbles while the reactor is operating, limiting shutdown time. Inside the reactor vessel, helium gas coolant flows around and between the stacked pebbles, emerging at about 1,650 degrees Fahrenheit. Helium is inert, so it does not become radioactive from the exposure. The heated helium drives a gas turbine, connected to a 165 electrical megawatt generator. The helium flow continues through a number of other power-generating processes, then re-enters the pebble bed reactor to be heated again. This closed-cycle process has a predicted 41 percent efficiency— higher than the 30 to 35 percent efficiency of current pressurized-water and boiling-water power plants. In the event of a complete shutdown of helium flow in a pebble bed reactor, the temperature would rise at most to 2,900 degrees, well below the thermal limit of graphite. At such high temperatures, the uranium-238 nuclei, which are more plentiful than uranium-235, absorb more neutrons (due to an effect called Doppler broadening) and the reac-


tor output decreases, lowering the temperature until equilibrium is reached. The greatly reduced heat is then transferred passively by radiation, conduction, and convection to the steel reactor vessel, which is designed to eject the heat without human intervention. Nuclear Power’s Twisted History

Pebble bed reactors were first designed in 1944 in the United States, but the first to be built was developed by German physicist Rudolf Schulten. It began operation near Aachen, Germany in 1966 and ran successfully for 21 years, providing heat for a small steam power plant. Tests run during its life demonstrated safe operation in the event of a total shutdown of the helium coolant. Although the South African pebble bed reactor project has ended, General Atomics in San Diego recently applied for government funding to commercialize a similar, 240-megawatt, gas turbine–powered nuclear plant. The Chinese currently have two pebble reactors, but they are being used to generate steam for a conventional turbine, which local officials claim is less challenging. (This approach makes no sense to me, because it has lower thermal efficiency and retains the danger of water ingress into the reactor.) There are gas-cooled reactors in operation in the United Kingdom; the first was built at Calder Hall in 1956. High-temperature, gas-cooled reactors have also been attempted, notably at Fort St. Vrain in Colorado and at a 300-megawatt pebble bed unit in Germany. Both units were shut down in the 1980s after a few years of operation. Despite a handful of promising experiments, the overall rarity of pebble-bed nuclear reactors and similar gas-cooled reactors elicits an inevitable question: If this engineering approach is so much safer, why are the vast majority of the world’s nuclear power plant reactors water cooled? The answer lies buried in the history of atomic energy. After Enrico Fermi created the world’s first sustained nuclear chain reaction, in the University of Chicago’s squash courts on December 2, 1942, physicists considered

using helium as a coolant for future nuclear power plants. The idea was rejected as too complicated, but it soon returned in the work of Farrington Daniels, a chemistry professor from the University of Wisconsin. For some years, Daniels had been working on a process of fixing nitrogen from air using a novel furnace that heated the air with stone pebbles. Building upon this research, he proposed that a chain-reacting nuclear pile could be constructed along similar lines. The pile would consist of uranium oxide and carbide pebbles whose heat of fission would be removed by the flow of a cooling gas. Daniels filed a patent on his idea on October 11, 1945. In the patent (2,809,931, U.S.) he calls the pile a “pebble bed reactor,” claiming that the cooling gases be used to generate steam (to power a steam turbine), or “the heated gases can be used directly in gas turbines.” The next year, researchers at Oak Ridge National Laboratory started designing a helium-cooled reactor based on Daniels’ concept. The project didn’t get far. The Atomic Energy Commission was also formed in 1946, and one of its first acts was to cancel the Oak Ridge project. According to Alvin Weinberg, then the director of the physics division at Oak Ridge, the Daniels design team was reassigned and became the core of the group that designed the first naval nuclear power plant. Funded by the Navy under the direction of Hyman Rickover, the reactor for the submarine Nautilus was cooled by pressurized water. Water-cooled reactors made sense within the cramped confines of a submarine hull. Once companies such as General Electric and Westinghouse formed engineering teams and manufacturing facilities to produce water-cooled nuclear power plants for subs, the die was cast and they applied the technology to commercial power plants as well. Today, water-cooled reactors are used in most of the world’s operating nuclear plants.

As an engineer I would be more trusting of a nuclear reactor that would use its own nuclear reactions to shut itself down if reactor coolant is lost.

www.americanscientist.org

Back to the Path not Taken

The high-temperature, gas-cooled reactor technology path pioneered by

Daniels needs to be rejoined. Three Mile Island and Fukushima have shown the near-catastrophic effects that loss of coolant can have on water-cooled nuclear reactors. Nuclear power will not see a major rebound unless the new designs are clearly safer than what exists now. What is needed is an industrysponsored, public–private partnership to develop gas turbine, pebble bed–type reactors as truly failsafe sources of carbonfree electricity. There are still questions to answer, and problems to solve. Some experts in the nuclear industry feel that pebble bed reactors have not yet shown sufficient operational efficiency: It is unclear whether their costs would be competitive with other types of nuclear reactors, much less natural gas, over their full life span. Another criticism is that pebbles moving through the reactor could create contaminated graphite dust. Although pebbles can be decommissioned individually, breaking up the task of storing nuclear waste, the overall magnitude of waste from the combined mass of pebbles might be greater than from other reactor designs. Meanwhile, other nuclear technology is not sitting still. One intriguing, newer concept is to use molten salts instead of helium as the cooling agent. Liquid salts increase heat transfer and raise efficiency, and consequently reduce the size and the cost of the reactor, but are also inherently fail-safe. Physicists at the University of California at Berkeley and Oak Ridge have developed designs for a liquid-salt-cooled pebble bed reactor. There may be a few twists on the path to next-generation nuclear power. But it is clear to me that the path is worth following. Bibliography Barrett, L. H. 2013. Fixing Fukushima’s water problem. Bulletin of the Atomic Scientists September 9, pp. 1–9. Bodansky, D. 2004. Nuclear Energy: Principles, Practices, and Prospects. New York: Springer. Langston, L. S. 2013. The adaptable gas turbine. American Scientist July–August, pp. 264–267. Langston, L. S. 2011. The future of nuclear power in Connecticut. Bulletin of the Connecticut Academy of Science and Engineering 26, Spring, pp. 1-2, 8. Langston, L. S. 2011. PBMR—A future failsafe gas turbine nuclear power plant? Global Gas Turbine News, supplement to Mechanical Engineering, August, pp. 54, 59. Langston, L. S. 2008. Pebbles making waves. Mechanical Engineering February, pp. 34–38. Weinberg, A. M. 1994. The First Nuclear Era: The Life and Times of a Technological Fixer. Woodbury, NY: American Institute of Physics Press.


2014

March–April 93