TAKING FEAR OUT OF NUCLEAR POWER Concern about the earth's rising temperature could turn a technological pariah into a savior -- if new reactor designs overcome worries about atomic safety.
By Edmund Faltermayer REPORTER ASSOCIATE Alicia Hills Moore

(FORTUNE Magazine) – JINXED by runaway construction costs and reviled for putting humanity at needless risk, nuclear power seemed destined for gradual abandonment. That was last year. Amid mounting evidence that the earth is warming because of the greenhouse effect, splitting atoms to generate electricity is getting new respect. Unlike coal or oil, it creates neither acid rain nor carbon dioxide, which is believed to be mainly responsible for the planet's rising temperature (FORTUNE, July 4). By happy coincidence, companies that build reactors are working on a medley of new models, some of which have begun to allay concerns about safety. With new humility following the accidents at Three Mile Island and Chernobyl, the nuke makers are relying heavily on ''passive'' design features to protect the public. In future versions of the water-cooled reactor, which dominates the industry, simple gravity flow from emergency tanks would protect the nuclear core for several days even if all else failed. One proposed version of a gas-cooled reactor would offer still greater peace of mind. In the worst imaginable situation, with the control room crew inattentive and the core entirely deprived of coolant, no significant radioactivity would be released -- ever. The new models are badly needed, if only for political and psychological reasons. Only giant strides in safety will overcome antinuclear passions and permit streamlining of the elaborate regulatory and judicial processes that have stopped atomic expansion cold. No U.S. utility has ordered a reactor $ since 1978. Why fight a battle in which any self-professed environmental group -- or just a few people scared of nukes in their community -- can tie up a project indefinitely merely by hiring a lawyer? Why risk the kind of strong political opposition, reaching up to Governor Mario Cuomo, that forced New York's Lilco to abandon the $5.3 billion Shoreham reactor last June before it lit a single lamp? The political climate might change, and the country might get the power it is going to need, if the industry's critics are moved by design improvements to alter their stance. The most respected watchdog group is the Union of Concerned Scientists, which has called for the phasing out of existing reactors. Robert Pollard, the group's senior nuclear safety engineer, is wary of all the newfangled nukes and wants to be shown that they will live up to the claims made for them. But he allows that at least one new model -- a gas- cooled reactor -- ''sounds eminently workable.'' Ah, but wouldn't the cost of added nuclear safety make the kilowatts prohibitively expensive? Not at all, insist the reactor builders. Safer reactors, they say, are also simpler and cheaper to build. Standardization would save money too. A major curse of U.S. nuclear power has been a plethora of custom-built plants, nearly all different. The Nuclear Regulatory Commission, which oversees reactor safety, hopes to okay a master design for each improved reactor in advance, thus speeding regulatory approvals for each project. SINCE MOST of the U.S. is awash in generating capacity these days, utilities can wait a few years before they place orders. Some buyers may not be utilities at all, but the emerging species of independent generating companies. It would take until the middle or late 1990s -- about the time electricity demand is expected to catch up with supply -- before detailed design work, regulatory approvals, and construction could be completed on the advanced nukes. The fact that safer reactors can be built does not mean that existing ones are unacceptably hazardous. The U.S. now gets 20% of its electric power from 108 reactors; 14 more are under construction and in no apparent danger of being idled like Shoreham and New Hampshire's Seabrook facility. Some other industrial countries rely more heavily on the atom -- among them France (70%), Belgium (66%), and South Korea (53%). None of the non-Communist world's power reactors are built like the tricky Chernobyl unit that blew its concrete roof in 1986, causing 32 deaths in the worst civilian nuclear disaster ever. Nevertheless, people near the plants live with potential danger in somewhat the same way that the Dutch who live below sea level face the possibility of flooding if a dike gives way.

The U.S. nuclear industry's answer has long been what it calls ''defense in depth.'' In the case of water-cooled reactors, four barriers plus automatically controlled standby pumps are supposed to keep the radioactivity of a crippled reactor from reaching the world outside. Three such ''dikes'' were breached or circumvented in history's second most serious power reactor accident, at Pennsylvania's Three Mile Island in March 1979. But the last and most formidable barrier, the thick steel-reinforced containment building, kept in nearly all the lethal stuff. No one was injured. For 40 minutes the uranium core was partly or almost entirely uncovered by coolant. More than half the fuel elements melted into a blob at the bottom of the reactor vessel -- a billion-dollar blob for General Public Utilities, which is spending that amount just to clean up. Yet while core damage was greater than anyone suspected at the time, the release of radioactive elements constituting a serious health hazard was negligible -- one-millionth of Chernobyl's or less in the case of radioactive iodine. Milk from nearby goat herds picked up two to three times as much radioactive iodine from the Russian disaster thousands of miles away as it did from the TMI mishap next door. Norman Rasmussen, an MIT nuclear engineer, ran a landmark government study before the TMI accident calculating the mathematical probability of releases from U.S.-type nuclear plants serious enough to cause ten or more premature cancer deaths. The reassuring odds: about once in three million reactor-years of operation. The TMI accident and more recent studies, Rasmussen says, suggest that nukes are ''substantially safer'' than previously thought. Certainly the operators are more on their toes than a decade ago. TMI-1, the surviving sister reactor of the ruined TMI-2, has a revamped control room. The Institute of Nuclear Power Operations in Atlanta, set up after TMI by utilities as a clearinghouse for safety information, reports a decline in safety-related incidents. Slack discipline gets harsh punishment. At Philadelphia Electric Co.'s Peach Bottom plant last year, some members of a night crew were found asleep or playing videogames. They were suspended, reassigned, or demoted; the company's chairman took early retirement. Water-cooled reactors, alas, function in such a way that no one can rule out a far more dangerous accident than TMI -- one with even worse consequences than Chernobyl. The odds are ludicrously small, but that no longer assuages the public. Vice president James Moore of Westinghouse, the biggest U.S. reactor builder, sounds almost like a nuke-basher when he declares: ''We gain very little when we explain that the potential for a radiation release is one in ten million if we have to acknowledge that it could theoretically happen tomorrow. We are selling our wares to nervous human beings, not to statisticians.'' CRITICS of water-cooled reactors point out that at full power the temperature inside their uranium-oxide fuel pellets reaches 3,300 degrees Fahrenheit, high enough to melt the zirconium alloy cladding that surrounds the fuel rods. Only the water holds it below melting temperature. In the course of keeping the reactor cool, the water picks up enough heat to produce the steam that drives generators. But if all the water suddenly vanished because of a pipe break, the cladding -- one of the barriers against radioactive releases -- could begin melting in seconds. That possibility, defenders say, is based on a wildly far-fetched combination of things gone wrong. Robert Long, a vice president at GPU's nuclear subsidiary, says that in the worst believable sequence of events a reactor's operators have at least 20 minutes before they must begin worrying about damage to the fuel core. At TMI, where the operators for a time did the opposite of what was necessary, it was more than an hour before core damage began. Instead of emphasizing multiple barriers and ''engineered systems'' -- a panoply of emergency pumps and automatic controls -- the new designs look to nature for a helping hand. One approach, used in rival proposals by General Electric and Rockwell International, is to abandon water as a coolant in favor of liquid sodium, which has a far higher boiling point and superior ability to absorb unexpected heat surges. A water reactor's coolant must be kept at high pressure -- up to 150 times atmospheric pressure -- lest it turn to steam. Liquid sodium needs no pressurization. If pumps circulating the coolant broke down, the liquid sodium itself, along with the natural circulation of air around the outside of the reactor, would prevent fuel damage indefinitely. One worry, however, is that liquid sodium reacts violently when exposed to air or water. The reactor vessel would have to be double walled, with the space between filled with nitrogen as an added precaution. PIUS is not a pope but another intriguing reactor design, developed at Asea Brown Boveri, the big Swedish electrical equipment maker. The letters stand for ''process inherent, ultimate safety.'' Based on the natural tendency of liquids of different densities to separate, as they do in a pousse-cafe cocktail, PIUS is a water-cooled reactor surrounded by a large pool of cold, heavier water containing boron. Under normal operation the borated water, which halts nuclear reactions by absorbing neutrons, would be kept at bay not by a valve but by the sheer pressure of the regular coolant. If the regular coolant were suddenly lost, the heavier fluid would rush in without human intervention, keeping the core safely cooled for a week while the operators deliberated their next move. THE MAIN drawback of both PIUS and sodium-cooled reactors is that prototypes have never been built. Says Chairman James J. O'Connor of Commonwealth Edison in Chicago, a major operator of nuclear plants: ''It would be unrealistic for a utility to order, in the near future, a reactor based on a technology that has yet to be proven commercially viable.'' The Electric Power Research Institute in Palo Alto, California, a utility-supported group, is putting most of its money for advanced reactor research into the familiar water-cooled workhorse. The reason, says Karl Stahlkopf, who runs the effort, is that ''we know where the devils are.'' Devils aplenty have been cast out of an improved water-cooled reactor design jointly developed by GE, Hitachi, and Toshiba. Tokyo Electric Power Co. has signed up for two units, and construction will begin in the early 1990s. Westinghouse, which is also working on an advanced model with Japanese partners, hopes to announce an order before long. Nothing radical is planned in either case -- only a host of evolutionary refinements in controls, pumps, and other equipment that cut the chances for trouble. The big leap in safety would come with compact new versions on the drawing boards at Westinghouse, GE, Combustion Engineering, and Babcock & Wilcox, a subsidiary of McDermott Inc. During the 1960s and 1970s reactors kept getting bigger, reaching a maximum capacity of 1,400 megawatts, enough to power a city of 700,000. The Westinghouse model AP600 -- the letters stand for ''advanced, passive,'' and the capacity is 600 megawatts -- is a more appropriate size for an era when utilities expect to grow in smaller increments.

The AP600 is a far more easygoing design than today's reactors. Fuel and water temperatures would both be lowered, reducing the concentration of heat that must be removed if there is an accident. When present reactors lose their coolant, pumps powered by emergency diesel generators are supposed to keep the core covered with water from an external tank. The AP600 would be safe even if the diesels failed to kick on. The reactor would sit directly beneath tanks containing a total of 400,000 gallons of water. In a severe accident, valves would open automatically from the change in pressure alone, and gravity would deliver a Niagara of emergency coolant. The AP600 would offer a prolonged breathing spell. Even if everyone in the control room died and all power was cut off, the core would remain undamaged for 72 hours. Richard Slember, general manager of Westinghouse's energy systems unit, says that three days is only the minimum expectation: ''Some of our engineers think we could go forever.'' The containment building, in fact, would be designed so that natural air circulation around the walls would prevent later overheating. After several weeks, the temperature of a crippled reactor tapers off to a less worrisome level. The AP600 and its cousins represent a retreat from past efforts to maximize economies of scale. Edwin Kintner, a GPU executive who heads an Electric Power Research Institute committee on advanced water-cooled reactor design, applauds the new trend. For too long, Kintner says, ''Westinghouse and GE were competing to see how much power they could get out of a reactor vessel. They pushed these things without really understanding the economics.'' The small reactors offer cost savings that offset their size limitations. Because they would be less temperamental, fewer unplanned shutdowns are likely. Partly because of passive safety features, the builders hope the NRC will require far less backup gadgetry of all kinds. Compared with a conventional reactor of the same capacity, the AP600 will need 50% fewer pumps and heat exchangers, 60% fewer valves and pipes, and 80% less control cable. According to Slember, orders could be placed as early as 1992 and the units could be supplying power to the grid in the late 1990s. Construction costs per kilowatt: roughly $2,000 at today's prices, counting interest expenses during construction, a third of the final bill at Shoreham and low enough to compete with coal. THESE REACTORS ought to be safe enough to dispel all the hobgoblins, their champions say. But every design mentioned so far has one limitation. The fuel core, whether cooled by liquid sodium or water, must be kept covered to prevent damage. Something could go awry: Those passive valves just might not open as they should. In the end the builders must fall back on probability calculations showing that the passive safeguards are virtually certain to work -- a proposition that may not impress critics. Enter the gas-cooled reactor. Depending on the type, its core can survive without coolant for periods ranging from hours to eternity. That is a mighty arresting fact. But like Shiite and Sunni Muslims, executives in the nuclear industry and the utilities are sharply divided over whether gas-cooled reactors are ready to play a big role in cranking out kilowatts. To nuclear engineer Lawrence Lidsky, who runs gas-cooled reactor research at MIT, the issue is already settled. ''The water-cooled reactor in any incarnation,'' Lidsky says, ''is inherently an exceedingly complex, unforgiving device.'' The U.S. embraced it prematurely and sidetracked other reactor concepts, say Lidsky and others in the gas-cooled gang, because it was a spinoff from the nuclear Navy. Such talk infuriates the water-cooled crowd, which worries that enthusiasm for the gas-cooled reactor could divert energies from improving their baby. Gas-cooled reactors will have their place early in the next century, says Stahlkopf of the Electric Power Research Institute. But only advanced water reactors will be ready within ten years, when the country will need more kilowatts. Nonsense, counter companies working on the gas-cooled concept. Among them is General Atomics of San Diego, once controlled by Chevron but in private hands since 1986. General Atomics has more than two decades' experience with gas reactors. Other players include West Germany's Siemens, which also builds water-cooled reactors, and HRB, a German company that designs the gas-cooled type. Asea Brown Boveri owns 55% of HRB and General Atomics the rest. The Soviet Union is getting interested in gas-cooled technology, and the Japanese government is weighing a proposal to build a $690 million research reactor. Gas-cooled reactors use radically different fuel elements, which hold in enormous amounts of radioactivity at the source. By an automated process, / General Atomics and HRB coat spheres of uranium fuel, no larger than grains of sand, with multiple layers of ceramic materials to a diameter of one millimeter (see diagram on page 106). The most heat-resistant and radiation- proof material is silicon carbide. The coatings perform the same containment role as that giant concrete structure at TMI. The ceramic layers cannot melt, although ultimately they vaporize. Not until the particles reach a temperature of 3,300 degrees -- well above the level encountered when a reactor suddenly loses all coolant -- does some radioactivity begin to leak out. The particles are clustered by tens of thousands in graphite fuel elements. General Atomics packs them into hexagonal prisms four feet high; the Germans embed them in ''pebbles'' the size of billiard balls. HARDLY ANYONE disputes the inherent safety of gas-cooled reactors. The big issues are their reliability and competitiveness. Britain and France together built dozens of them in the 1950s and 1960s, and they ran fine. But because they were cooled with carbon dioxide, efficiency was low; the plants converted only 28% of the heat from the chain reaction to electricity. France long ago switched to water-cooled reactors, which achieve efficiencies of around 34%, and Britain later followed. Helium-cooled gas reactors do even better. The U.S., West Germany, and Britain built experimental helium reactors in the 1960s. At the time the attraction was not safety but the high temperatures that are achieved, useful both in power generation and in such industrial processes as chemical manufacturing and coal gasification. All three reactors performed well; the West German one is still chugging away. By the mid-1970s, General Atomics had orders from utilities for ten commercial-size plants. THEN A SERIES of setbacks nearly killed off the helium reactor. When energy conservation hit, utilities canceled nine of the orders. The only commercial helium reactor in the U.S. -- Public Service Co. of Colorado's 330-megawatt plant at Fort St. Vrain, within view of the Rocky Mountains near Denver -- flopped economically. The fuel has performed splendidly as a containment device since the plant went into service in 1974. Because radiation within the plant is extremely low, a recent visitor was allowed to walk on top of the concrete-shelled reactor when it was at 80% power without the usual safety suit or dosimeter. A design flaw unrelated to safety has played hob with operations. In other gas-cooled units the circulators that keep helium moving in and out of the reactor are lubricated with oil. The Colorado circulators use water instead. From time to time water leaks into the reactor and the plant must be shut down to remove it. Though the reactor is up and running much of the time, the utility took a $101 million after-tax hit in 1986 and wrote off most of its investment. If the tale stopped there, gas-cooled reactors might rate a place among technology's also-rans. But in the past year the outlook has brightened. A motorist speeding down the Autobahn near Hamm, northeast of West Germany's Ruhr region, cannot miss the high cooling tower of the world's only other big helium reactor. The 300-megawatt facility, unromantically known as the THTR -- for thorium high-temperature reactor -- is on line and performing satisfactorily. A consortium including Asea Brown Boveri and West Germany's HRB affiliate built the plant for a group of utilities. Construction, begun in 1972, was repeatedly delayed as the government mandated additional safety features. ''They're built like bunkers,'' says a member of the plant's staff as he points to the main buildings, which can survive a crash by a stray airplane. Total cost, at the average dollar-mark exchange rate during construction: a far-from-cheap $2 billion. The West German government, intent on furthering the development of helium reactors, paid more than half. No water woes here. The reactor, using traditional oil-lubricated circulators, first generated current at 100% of capacity in September 1986. In the past 19 months it has run at full power roughly two-thirds of the time, converting heat to electricity at an impressive 41% efficiency. Klaus Knizia, who heads one of the utilities that own it, told a recent nuclear conference: ''As a prototype, the reactor has lived up to expectations. We are confident that it will prove to be reliable.'' HRB says it is ready to build a 550-megawatt model that would have generating costs comparable to those of large water-cooled reactors. The Hamm reactor's neighbors can sleep soundly. Martin Heske, a nuclear engineer who works at the plant, says that in the worst conceivable accident all devices for removing heat from the core would conk out. In that event, it would take eight hours before the fuel heated up to 3,600 degrees. The resulting radioactivity would be so small that the authorities would not require the local population to evacuate, Heske says, but locally grown lettuce and other vegetables might become unsafe to eat. By shrinking a gas-cooled reactor, designers can offer a grace period without limit. Both General Atomics and its West German collaborators are pitching modular reactors as the ultimate in atomic safety. In the American design, four 135-megawatt modules, each in an individual below-ground concrete silo, would make up a unit with a total capacity of 540 megawatts. The reactor vessels would be 22 feet in diameter -- just small enough so the ground can absorb sufficient heat to prevent fuel damage in a serious mishap. ''We get our safety from abject simplicity,'' says Linden Blue, one of two brothers who own General Atomics. ''The fuel temperature in the core can never exceed 2,900 degrees.'' In the worst accident, studies by General Atomics conclude, people living right at the plant fence would pick up a trivial dose of radioactivity -- less than air passengers, for example, get from natural sources on a round trip from coast to coast. Pretty heady stuff, all right, concedes the water-cooled faction. But what are the economics? Why rush to the modular reactor when no successful prototype exists? Gas Cooled Reactor Associates, an organization supported by utilities with about a third of U.S. generating capacity, has looked into the competitiveness question. Big sections of the modules, it says, could be built efficiently in factories and shipped to the silos. The organization's study claims a four-module nuke could produce electricity about 10% cheaper than a coal-fired plant. A four-module prototype would cost $1.5 billion to $1.7 billion to build. Utilities that would sell the power could raise about two-thirds of the amount, General Atomics says. But they probably could not come up with the rest -- the heavy additional costs in a first-of-a-kind undertaking. To make up most of the gap, lobbyists for the gas-cooled reactor cause are looking to the Department of Energy, which until now has been less enthusiastic about the idea than some Congressmen. True believers ask whether a prototype is absolutely necessary. ''You can't go out and kick the tires,'' admits Richard Dean, a senior vice president at General Atomics. ''But five helium-cooled reactors have already been built, and every system in the modular reactor has been proven.'' General Atomics says it is now prepared, under a joint arrangement with deep-pocketed ) Siemens, to offer a performance warranty to any customer who places an order. But skeptics will not be convinced until a commercial four-module plant is running, and that could take until 1997. IN THE HOT NEW design race, both gas-cooled and water-cooled models may survive and coexist. For buyers, the choice will come down to a trade-off between radiation risk and business risk -- between a proven performer with a smidgen of worry and an even safer design still in its youth. The power may be no bonanza in either case. But kilowatts from coal and oil will probably be just as costly. And new worries about the greenhouse effect are turning fossil-fuel plants into bad guys just when nuclear power is poking its nose out of the doghouse.