Nuclear Power: The Unaffordable Option (Today Magazine, 1982)
Note: This article was published in the weekend magazine supplement Today, in June, 1928]
In 1953 President Dwight Eisenhower proposed an international Atoms for Peace program under which the United States would turn over to the United Nations its entire peace-time nuclear industry for the benefit of humanity at large. Lewis Strauss, Eisenhower’s atomic energy advisor, enthusiastically predicted that nuclear-generated electricity would be “too cheap to meter.” That same year, Ontario Hydro chose a nuclear future.
Eisenhower’s offer was deceptive, because the United States then had essentially no nuclear industry – and Strauss’s soaring fantasies have crashed and shattered. In 1980, 16 planned nuclear plants were canceled in the United States, about 60 more postponed, and Americans were deriving more energy from firewood than from nuclear power. Southern California Edison had renounced nuclear expansion in favor of conservation, solar and wind power. Donald Cook, former chairman of American Electric Power, the largest privately owned U.S. utility, was publicly regretting that his company had been sent “down the wrong road [by] an erroneous conception of the economics of nuclear power. The economics that were projected…never materialized – and never will materialize.” Financial analysts on Wall Street, steering investors away from utilities with heavy nuclear commitments, were predicting that nuclear power would push some companies into bankruptcy.
And in Canada?
In 1980 Ontario Hydro was generating more than 30% of its power from CANDU reactors, and Hydro’s officials were exuberantly announcing 90% effectiveness and hundreds of millions of dollars in savings from nuclear power. In June 1981 Hydro won approval to build the world’s largest nuclear station at Darlington, near Oshawa. The new plant would cost $6.64 billion and would ultimately generate 3,600 megawatts. It would begin coming on stream in 1988. Of the world’s 250-plus commercial nuclear reactors, six of the 10 most efficient in 1981 were Ontario Hydro CANDUs. Their success had stimulated new nuclear stations in Quebec and New Brunswick. Almost alone among heavily nuclear utilities, Hydro retained its triple A credit rating in New York.
What is Hydro’s secret? The whole future of nuclear power hangs on the answer.
Archie Robertson says the secret is CANDU, the best reactor in the world. Robertson is assistant to the vice-president of Atomic Energy of Canada Limited, who is in charge of the Chalk River Nuclear Laboratories. The world’s second nuclear reactor, ZEEP (for Zero Energy Experimental Pile), started up here in 1945. One of ZEEP’s descendants is Atomic Energy’s NRU (National Research Universal reactor), which went critical in 1957. Standing in front of NRU, Robertson explains that its essential purpose is to support the power reactor industry and its regulators. With its sister reactor, NRX (National Research Experimental), NRU also allows some basic atomic research and creates radio-isotopes for medical use. Descendants of these reactors are found in India, Pakistan, Argentina, Taiwan and Korea. The commercial version is CANDU.
Robertson points to the top of the NRU reactor vessel, where a robotlike crane hangs poised. Among power reactors, only CANDU can be refueled without shutting down. CANDU is admittedly more expensive to buy than competitive American reactors, but unlike them it operates on natural uranium, omitting the complex and expensive process of fuel enrichment. Supplies of natural uranium are running out nearly as fast as supplies of oil, but CANDU can be modified to run on plutonium or thorium, which would extend fuel supplies remarkably. CANDU is the only reactor developed explicitly for peacetime use, by a country that has foresworn nuclear weapons – although an NRX-type reactor gave India a bomb in 1974, and a CANDU in Karachi may soon do the same for Pakistan.
Ken Biron worked on the Karachi reactor and now works on the commissioning of New Brunswick Power’s first nuclear station at Point Lepreau, on the Bay of Fundy. “If this reactor had been developed in the United States,” Biron declares, “it would be the world reactor. No question.”
Viewed from the deck of the Grand Manan ferry, 25 kilometres away, Point Lepreau shimmers on the horizon like a vast, illusory oil tank. Closer up, the reactor building towers above the spruce-clad shore, concrete, dominant, immovable. Inside the tower, the roof soars like the vaulted ceiling of a technological cathedral, hexagonally braced with I-beams, bristling with sprinklers, laced with color-coded tubes and pipes for steam, heavy water and light water. A forest of control rods stands poised above the reactor vessel, the calandria. Pierced by 380 pressure tubes, the end of the calandria looks like a Chinese checkerboard. Each tube will soon receive a dozen 50-pound bundles of nuclear fuel. Each of these bundles is the energy equivalent of 400 tons of coal or 2,000 barrels of oil. The heat of the nuclear reaction is carried by heavy-water coolant to four steam generators. The steam drives a turbine, which spins an electrical generator to produce 600 megawatts of electricity. Lepreau’s output will be comparable to one-quarter of the energy from Newfoundland’s Hibernia oil field.
Back in the control room, Biron demonstrates the computers that monitor the whole station. Computer X continuously reports hundreds of flows, temperatures, pressures, voltages and levels of radioactivity. Lights wink on and off all over the vast control panel. Computer X completes an electronic tour of 5,000 information points every couple of minutes. Other lights trace the progress of Computer Y, which continuously tracks Computer X. If X fails, Y takes over. If Y fails, its last instruction shuts down the station.
CANDU is a “technical wonder,” says Hans Bethe, an American Nobel laureate. “Not only is it very conservative in fuel, but it works with a regularity and reliability that are absolutely fantastic.”
But if CANDU is so superior a mousetrap, why is no one at our door? Why are desperately importuning the Romanians, the Mexicans and the Koreans. Why can’t we sell, for instance, to the Americans?
American utilities are in trouble – and nuclear power is a major cause. Ask the customers of Washington Public Power Supply System, a West Coast consortium of 23 power companies that face rate increases of up to 300% in order to pay for two reactors that will never be built.Washington Power intended to build five reactors; federal financing would cover three of them. The other two were to be funded by 88 small utilities serving 2 million customers in four northwestern states.
When the plan was hatched 17 years ago, electrical demand was rising steadily. By 1976, though, growth had slowed dramatically: 100 U.S. utilities predicted average annual growth of 8.8%, but actually increases ran at 4.3%, and one company experienced a 7% decline. Meanwhile, the cost of the Washington consortium’s project exploded from $4.1 billion to $23.8 billion. With $2.25 billion already spent on the last two plants and $95 million outstanding in construction loans, the 88 utilities found they could not even afford to mothball the plants; they had to scrap them outright. In January 1982, the bond market already awash in the consortium’s bonds, they faced imminent bankruptcy. If they survived, they could only pay their debts by doubling or tripling their rates – unless government bailed them out.
Such mammoth increases, however, would leave the companies open to the “cost-price-demand spiral,” which threatens many othre U.S. utilities as well. Consumers respond to stiff rate increases by turning off lights, turning down water heaters, switching away from electric heat. A Carnegie-Mellon University study states that doubling the price of electricity will cut demand by 74%, so the higher rates will produce less income for the company. Rates must rise again, and consumers conserve further. Caught in such a spiral, the company eventually goes bankrupt.
The spiral even menaces companies that do complete their reactors. In California, Pacific Gas and Electric began its Diablo Canyon reactors in 1968. They were essentially complete five years ago, but they lie only about six kilometres from a major active earthquake fault. Nuclear opponents have frustrated Pacific’s application for an operating licence, and its $2.5 billion investment is producing not a nickel of revenue, while interest charges are costing it hundreds of thousands of dollars a day.
In Michigan, Consumers Power brought its Palisades plants on line in 1971 at twice its projected cost. By 1978 Palisades had spent two-thirds of its life closed for repairs, and it faced two full years of costly rehabilitation. Power rates had risen six times, from 1.7 cents to 3.5 cents per kilowatt-hour. And General Public Utilities Corporation, which owns the disabled plant at Three Mile Island in Pennsylvania, has trembled on the brink of bankruptcy ever since its notorious accident in 1979.
CANDU has no divine immunity from these problems. The first commercial CANDU, at Douglas Point, Ontario, does not even pay the interest charges on its construction costs. Until recently it could not run at over 70% of its designed capacity. Two later CANDUs at Pickering were shut down 22 months for retubing in 1975-76, and more retubing will be needed on others. In February this year, Hydro took its new Bruce-2 reactor out of service because of a small pressure-tube leak, and Hydro officials admit an accident like the one at the Ginna station earlier this year in Rochester, New York, could also occur with a CANDU.
And consider Quebec’s first CANDU, Gentilly-1, completed in 1971 at a cost of $87 million. It has never operated for more than four weeks at a time. It has produced no power for four years and could not produce any without four years of further modifications. Its full staff of 160 is required to operate it at “zero power,” in keeping with its current licence. Its costs $10 million annually. But Gentilly-1 does not hurt Hydro-Quebec any more than Douglas Point hurts Ontario Hydro. Both dud reactors belong to the federal taxpayer.
Point Lepreau does hurt New Brunswick. The average American nuclear station’s price tag went up 142% between 1971 and 1978. Similarly, Point Lepreau was committed in 1974, slated for completion in 1979 at a cost of $466 million. With luck, it may actually go into production in 1982 or 1983, at a cost of well over $1.2 billion. (New Brunswick’s totally provincial budget is only $1.7 billion.) New Brunswick electricity rates have risen 180%, and Ottawa has already had to forgive interest charges of up to $102 million.
Massive support from the federal government has so far insulated the provincial utilities from the cost-price-demand spiral. Norman Rubin of Toronto’s Energy Probe argues that Ontario Hydro may soon face the spiral anyway. Historically, Hydro’s planning has been based on a 7% annual rise in demand – but the figure has been steadily dropping. Hydro recently lowered its forecast for the next few years from 3.4% to 3.1%, and most independent estimates are closer to 2.5%.
Hydro now has about 44% excess capacity. About 25% is considered an optimum reserve, including plenty of allowance for daily and seasonal peaks. But Hydro insists it will need still more capacity soon, and since nuclear plants typically take a decade or more to construct, Hydro intends to keep building. It will double its debt to roughly $25 billion by 1990.
“We can supply electricity at essentially the same price anywhere,” declares Archie Robertson. “We can guarantee an abundant supply at a ceiling of 150% of the current price in constant dollars, and the resource will be available for centuries.” If Robertson were talking only about fuel this might be true, though it makes large assumptions about the smooth and easy development of thorium and plutonium fuel cycles. Fuel is so small a component of nuclear power that tripling its costs would add only 10% to the price of electricity.
But Robertson is not just talking about the cost of fuel. He is talking about the price of electricity, and his claim is a striking example of the romantic fantasies of nuclear advocates. Can he guarantee – for all time – the interest rates? The price of steel and concrete? The safety features required by regulatory commissions? The cost of the oil required in the construction of plants?
Such considerations determine a plant’s capital costs, which in turn are the largest part of the price of nuclear generation – especially with CANDU. As Michael McCracken, an Ottawa economic consultant, says, a nuclear plant is “the interest rate cast in concrete.” It makes no difference whether Lepreau is finished or unfinished, operating or idle; it still costs $3.5 million a week. You can’t turn off the interest rate. Power from Lepreau will cost about the same as power from oil-fired plants at today’s prices.
The high capital cost, says Norman Rubin, is the result of “trying to build safety into an inherently unsafe technology.” According to a 1974 study for the U.S. Atomic Energy Commission, a severe accident might cause 3,300 fatalities, 45,000 early illnesses, 1,500 latent cancers, $14 billion in property damage and an irradiated wasteland covering 464 square kilometres. Not surprisingly, the safety requirements on the industry are extremely stringent.
Charles Komanoff is a New York energy economist whose latest book is a sophisticated mathematical study of power plant economics. It concludes that the high cost of safety will keep nuclear power uncompetitive for the foreseeable future.
“The number of new safety problems being detected each year isn’t falling, it’s climbing – steeply,” says Komanoff. “The design criteria for nuclear power won’t stabilize for many years – if ever – simply because there are so many things the Nuclear Regulatory Commission still says are potentially dangerous.” In 1970 the NRC issued four regulatory guides. In 1977, it issued 250. New Brunswick Power’s public affairs manager, Terry Thompson, calls this process “regulatory ratcheting,” and Three Mile Island has made it much worse. Building a nuclear power plant is like trying to build a car while the designer keeps changing his mind about the number of wheels it should have.
Despite the ratcheting, unnerving accidents persist. When heavy water was fed into Lepreau for the first time last December, two 15,000-gallon tanks buckled under the weight. “They can’t design an adequate container for water,” says Tom Murphy, a nuclear critic from New Brunswick, “but they’ll bet people’s lives that they can design adequate containment for radiation.”
The high capital cost of safety is no deterrent to Ontario Hydro, whose financial arrangements constitute another level of subsidy. As Hydro’s chairman, Hugo Macaulay, points out, Hydro is forbidden by law to make a profit. Like most North American utilities, Hydro has a “declining block” rate structure.The first kilowatt-hour you buy is the most expensive, and the price goes down as consumption rises. By encouraging demand, these rates create a need for new generating capacity. Hydro generates power from oil, coal, fission and falling water, notably at Niagara Falls. Many of these facilities are long since paid for, and the fuel is cheap. Lumping all these sources together, Hydro charges customers an average cost, so that low-cost hydro subsidizes high-cost nuclear power.
Private industry, by contrast, sells each unit it produces for at least the cost of producing the next unit. If a new widget will cost $2 to produce, you don’t sell the one on the shelf for $1. Two dollars is the “opportunity cost” or “marginal cost.” In a study for the Ontario Legislature’s Select Committee on Hydro Affairs, Jack Gibbons, an economist, found that nuclear electricity sells for about 38% of its marginal cost. These bargain prices make Ontarians into profligate wastrels of power, using 11,114 kilowatt-hours apiece in 1977 compared to 5,461 for the West Germans and 3,981 for the French, who are paying close to the full marginal cost.
Hydro can sell electricity for less than it costs because, in addition to its freedom from profit, it also pays no taxes – and it pays rock-bottom interest rates on borrowed money because its loans are guaranteed by the province. As a result, says Michael McCracken, “they may pick a project that wouldn’t be economic for a privately owned utility.”
Jack Gibbons contends that these uneconomic choices distort the whole economy of Ontario. Hydro projects create fewer than half as many jobs as average investments elsewhere and produce a 3.3% return, as opposed to a taxable 10% or more in private industry. Ontario’s total “economic waste” is estimated by Glenn Jenkins, an assistant deputy minister of finance, at $1.3 billion annually. Gibbons puts it at $1.4 billion and adds 900,000 man-years of employment. These amounts, says Jenkins, “are a complete waste of resources, which can only result in a reduction in the total wellbeing of Canadian residents.” This past winter, 93 university economists joined the chorus calling for marginal cost pricing.
At first, such pricing would swamp Hydro with cash, to Macaulay’s dismay. “What would you do with the money?” he asks. McCracken would give it to Hydro’s owner, the province. Rubin suggests a proportional rebate to industry and a flat rebate to residences. For a conserving household, the rebate might cover the whole electric bill. Macaulay charges that Rubin’s plan would use a household necessity to enforce social change. Rubin replies that it’s simply “textbook economics – passing the right signal along.” McCracken agrees: “When you price electricity at half its cost, you’re simply encouraging the misuse of a scarce resource.”
Beyond Ontario’s built-in subsidies, nuclear power benefits from a host of other federal supports and uncalculated future costs. A study for the U.S. Department of Energy valued American subsidies to nuclear power at $37 billion overall and reckoned they doubled its apparent cost. For openers, the government of Canada has put up about $2 billion to develop CANDU, lost $130 million selling a CANDU to Argentina, and written off $816.9 million in loans for Atomic Energy’s heavy water plants.
The Nuclear Liability Act, proclaimed in 1976, requires Canadian reactor operators to take out $75 million in liability insurance for each site. Ottawa pays for losses beyond that amount. The U.S. Price-Anderson Act, sets out a similar limit of $560 million, though the actually damage from an accident might reach $40 billion. Herbert Denenberg, former Pennsylvania commissioner of insurance, estimates that full liability insurance – if companies could get it – would cost about $23 million a year, which would roughly double a plant’s operating budget.
Finally, the garbage. Uranium mining and milling has already given Canada more than 100 million tons of “tailings,” which though not immediately fatal are radioactive and harmful. Cleanup costs may total somewhere between $300 million and $18 billion. Spent fuel from power reactors is immediately lethal and will remain so for several hundred years. At present spent fuel is stored in “swimming pools” at reactor sites.
Archie Robertson considers it a resource. It its remaining plutonium and uranium were extracted, it could be recycled as fuel after the year 2000, when uranium will be getting scarce. But the byproduct of reprocessing is a highly corrosive, highly radioactive liquid waste, much harder to isolate than the original fuel rods.
Hugh Macaulay says that the technological problems of waste disposal have been solved, but not the “sociological and political implications.” Macaulay is mistaken. Proposals have been made to dump the wastes in the ocean, rocket them into the sun, bury them in salt deposits, or sink them in the polar ice caps. Each of these is now discredited. Canada’s current idea is to encase them in glass and bury them in plutons – inert, geologically stable rock formations in the Canadian Shield. The Atomic Energy research program has four phases, and the first – “concept verification,” to confirm the soundness of the plan – was once expected to be concluded by 1980. The Ontario and federal energy ministers now expect it “probably in the late 1980s.” The overall timetable yields a date around 2010 for full-scale operation, assuming that everything goes smoothly. In 1981, after 15 years of nuclear generation, Ontario Hydro set up a sinking fund devoted to the unknown cost.
After 30 years or so, a nuclear plant itself becomes irradiated and must be “decommissioned.” Atomic energy claims that a 600 megawatt CANDU can be dismantled and removed for $30 million over a six-year period. Well, maybe. The only reactor of consequence yet decommissioned is a 58 megawatt research reactor in Elk River, Minnesota. It had to be cut apart underwater by divers using laser tools, and the price exceeded the original construction cost.
A crippling accident could send costs much higher. Dismantling Three Mile Island is a totally innovative operation, and estimates range as high as $2 billion. Meanwhile, the plant’s owners have had to import power to serve their own customers. Metropolitan Edison has now paid more for replacement power than it paid to build Three Mile Island. That power is supplied by none other than Ontario Hydro.
Energy Future is the report of the Energy Project at Harvard Business School. Irving C. Bupp, who wrote its chapter on nuclear power, is a former employee of the Atomic Energy Commission and an authority on utilities. After examining much expert opinion on the economics of nuclear power, Bupp concluded that “the genuine uncertainty about this extraordinarily complex issue and the dedication of critics intent on proving that nuclear plants are both uneconomic and unsafe make the economic calculations an increasingly time-consuming and thankless task.” He saw, “little chance of an unbiased scientific consensus.”
The Harvard study’s surprising conclusion is that we do have an amazingly cheap, proven, effective solution to the energy problem: stop wasting it. “We favor financial incentives to encourage consumers to use conservation and solar energy, not because there is anything intrinsically virtuous about these energy sources, but because they make good economic sense.” The slackening demand for oil and electricity suggests that consumers have already figured this one out. Interestingly, Energy Future singles out as a model Ontario Hydro’s own head office, which is heated by waste heat from appliances, lights and human bodies as well as by heat pumps.
Conservation pays. America could have enjoyed the same standard of living in 1973 while using 40% less energy. Britain could triple its gross national product by 2025, using the energy of 1975. In Zero Energy Growth for Canada, David Brooks makes much the same case for this country: a richer, cleaner, thriftier nation living within its energy income. Whatever the cost of new power plants, nuclear of otherwise, it is more than we need to pay.
Surely that is the bottom line.
– 30 –
Power Crazy: What Do You Do with a Reactor Nobody Needs?
The CANDU industry is in crisis. Unless it finds two reactor orders a year, it will be defunct by 1986. A worldwide slump in power demand means the international market is vicious and cutthroat. We have some wistful hopes of selling (at heavy losses) to Korea, Japan, Mexico or possibly Egypt. But 20 years of export efforts in a much less forbidding market have sold only five CANDUs abroad. And the domestic market is saturated: Canadian utilities will not need further nuclear stations before the 1990s.
The deep thinkers at Energy, Mines and Resources, however, found a nifty solution: build CANDUs along the border and export the power to the United States. They proposed – apparently seriously – “five or six 630 megawatt reactors in New Brunswick and from five to seven 850 megawatt reactors in Ontario,” plus a web of new transmission lines, all serving an American market that may develop between 1990 and 2000. The Energy Department admits that Canadian power rates would rise as Canadians shouldered the capital costs, the waste-disposal and decommissioning costs, the industrial dislocation, the safety risks and the pollution.
“The only nice thing that comes out of a nuclear power plant is electricity,” remarks Energy Probe’s Norm Rubin, “and the Americans would get that.” But CANDU would be saved.
Should it be saved? About 20% of CANDU components are imported, and two-thirds of the “Canadian” suppliers are subsidiaries of firms like Westinghouse and General Electric. The industry, notes a waggish federal government study, “is fragmented, concentrated in Quebec and Ontario, comprised of many small companies, short of research and development and engineering capability, and controlled largely by foreign interests. In a word, it is typically Canadian.”
Common sense is not totally extinct at Energy, Mines and Resources. Sources now say the proposal will likely be limited to “prebuilding” a reactor or two, which Canada will eventually need anyway, and exporting power in the meantime. New Brunswick is doing that now. One-third of Point Lepreau’s output is committed to New England. New Brunswick Power hopes to sell 54% and to prebuild Lepreau-2. But prebuilding may not be enough to save CANDU.
What do you do with a reactor nobody needs? Buy a dozen yourself, says the Energy Department’s deep thinkers. Shallower thinkers might suggest something simpler: scrap it and build something we do need.
– 30 –