Nuclear follies. (nuclear power is a failing due to internal, not external problems)
This was a Forbes magazine cover article. The issue contained another article, "The Best. (Duke Power Co. successfully owns and builds its own atomic power plants)." which should also be read by anyone wishing to understand why construction of new nuclear reactors stopped in the US decades ago.
This is the article Al Gore cherry picked a quote from for his book "Our Choice". Gore couldn't resist this: "for the U.S., nuclear power is dead--dead in the near term as a hedge against rising oil prices and dead in the long run as a source of future energy. Nobody really disputes that" even though the article does not support Gore's case that nuclear power is too expensive to use anywhere in the world in 2010.
What Al ignored is that Cook, when writing in Forbes, stated that he wanted to answer this question, which he wrote was the "most important": "why did the U.S. fail where the French, Germans and Japanese succeeded?".
Here is the text, minus the first part of the first sentence, [ sorry, some text eater got that part ] of the entire article:
Nuclear Follies by James Cook
...and, in little more than a decade, transformed what elsewhere in the word is a low-cost, reliable, environmentally impeccable form of energy into a power source that is not only high in cost and unreliable, but perhaps not even safe.
The scale of U.S. nuclear power program's collapse is appalling: 75 plants canceled since 1978, including 28 already under construction, with another half-dozen or so cancellations in prospect. Even in its death throes, the U.S. nuclear program looks ambitious: 16 new stations put into commercial operation since 1979, another 3 this past year, and 40 more units by the end of the decade. By that time nuclear power will account for something like 20% of the U.S.' total generating capacity up from 14% today. Altogether, these stations will be producing 129, 725 megawatts, more than the nuclear capacity of France, West Germany, Japan and the Soviet Union put together. But that's it. Once these plants are completed, the U.S. nuclear power program is at an end.
Elsewhere in the world, some 148 nuclear power plants are under construction, 9 more are on order and 157 are in the planning stage. By 1990 the Japanese should be getting close to 20% of their electricity production from nuclear power, the Taiwanese 30%, the Belgians 40%, the French 55%. That's low-cost energy, all of it, and, according to one study, 30% to 50% lower in cost than coal. The newest French nuclear plant, at Cruas in the south of the country, produces power for under 4 cents a kilowatt-hour, and that's cheap by almost any standard.
In the U.S., however, nuclear power makes less and less economic sense. Production costs of existing coal and nuclear plants run neck and neck, but as the new nuclear plants go into service, coal should gain ground decisively. In upstate New York, New York State Electric & Gas' newest coal-fired station at Somerset began producing power last August at a cost of 7.5 cents per kwh. A hundred miles away, Niagara Mohawk's $5.1 billion Nine Mile Point nuclear plant will produce power two years from now for 18 cents per kwh.
At such levels, nuclear power can't compete with oil, much less with coal. That's why, for the U.S., nuclear power is dead--dead in the near term as a hedge against rising oil prices and dead in the long run as a source of future energy. Nobody really disputes that. Not the Nuclear Regulatory Commission (NRC) or the Office of Technology Assessment. Not even the Atomic Industrial Forum (AIF), the trade organization that represents the nuclear power industry. And certainly not the 200 or so executives who head the U.S.' investor-owned electric companies. Nuclear power is an option nobody in his right mind would now seriously consider. "The costs and risks of nuclear development," an Edison Electric Institute study concluded last month, "have become unacceptably high."
What destroyed the nuclear option in the U.S.? How could U.S. nuclear power costs run so outrageously out of control? And why didn't the NRC or the state regulators, never mind the utility managements, do something about them? Most important of all, why did the U.S. fail where the French, Germans and Japanese succeeded? The answers to such questions are important, and they suggest a good deal about where the U.S. is going to go--or perhaps ought to go--in the future. Without an understanding of what caused the nuclear debacle, the U.S. will lack any clear sense of what it must do not only to restore nuclear energy as a viable energy option but also to assume that history does not repeat itself. 1 The incredible price
It wasn't technology that doomed nuclear power in the U.S. As experience everywhere demonstrates, the technology is as sound and productive as its promoters always have claimed it would be.
Nor was it the dedicated and imaginative obstructionist tactics of the antinuclear forces, effective though they were. The opponents of nuclear power have hampered and harassed it, inflated its costs and stretched out construction times to unconscionable lengths, but they could not, unassisted by events, have prevailed.
The truth is that nuclear power was killed, not by its enemies, but by its friends:
* The federal government and the Nuclear Regulatory Commission, which not only botched the day-to-day management of the program but also failed to consider the economic cost of the regulations it imposed.
* The equipment manufacturers, who maintained that nuclear power was just another way of boiling water.
* The contractors and subcontractors, the designers, and engineers and construction managers who, insulated by their own cost-plus contracts, had little incentive to question the cost-effectiveness of the NRC's dictates.
* The utility executives, who believed that no matter what happened to cost and construction schedules, the rate commissions would somehow provide the revenues to bail them out.
* And the state regulatory commissions themselves, whose grossly inadequate oversight of the schemes, ambitions and monstrous expenditures for nuclear projects made it easier for all of the above to betray consumers and investors alike.
These collective failures are reflected in the accompanying table. It lists 35 nuclear power projects that had not yet gone into commercial operation at the beginning of 1984. These new plants cost six to eight times more than originally projected, three to four times more on average than nuclear plants already operating. As a result, more than half the plants on the list are no longer competitive with coal nor, some of them, even with oil.
These plants, some conceived as early as the late Sixties, were subjected to all the problems that afflicted the industry over the past decade. All suffered the delays and cost overruns that doubled the average construction time from 6 to 12 years. Yet for all their common technical, social and political environment, the costs of these plants differ widely, ranging from a commendable $932 a kilowatt for Duke Power's McGuire 2 station to a grotesque $5,192 a kilowatt for Long Island Lighting's Shoreham plant.
The pressures of regulation and antinuclear interveners created more problems for some plants than for others, but they are not enough to explain so wide a range of differences. Labor costs are not appreciably lower in Chicago than they are in New York. Opposition to nuclear power is no more intense on Long Island than it is in northern California.
The disparities in cost are so great as to make a prima facie case for mismanagement in the first degree. At least a dozen projects still going forward--more than one in four--are problem plants by any measure, suffering from quality, construction and design problems as well as cost overruns.
The mismanagement extends beyond design and construction to financial management. At least half a dozen utilities have abandoned plants already well on their way to completion--Marble Hill and Midland among them--because they simply did not have the financial resources to complete them. Other projects still alive--Shoreham, Seabrook and second units everywhere--are similarly imperiled. The financial pressures have been so extreme that at least six major utilities--including Lilco, Public Service of New Hampshire, Consumers Power in Michigan, Public Service of Indiana--are in serious danger of going into bankruptcy.
But for legal reasons--never mind pride--managements can admit no wrong. Any admission of mismanagement is tantamount to an admission of "imprudence," and some state regulatory commissions, growing increasingly worried about the effect these plants will have on the cost of electricity, have begun disallowing at least part of their costs on just such ground. The staff of the New York State Public Service Commission has proposed disallowing close to $1.5 billion in Lilco's nuclear expenditures as imprudent. Such a sum, Lilco Chairman William Catacosinos readily concedes, would be enough to bankrupt the company.
Bankruptcy would not put the companies out of business. They are certain to survive as operating, if not corporate, entities. But a price will have to be paid--indeed, already is being paid--in the form of higher rates.
The failure of the U.S. nuclear power program is changing the economic geography of the U.S., encouraging business to exchange high-cost power for low, weakening areas in the East, the Midwest and the Northwest, strengthening the already resurgent South. Lukens, Inc. in Coatesville, Pa. already has tried (unsuccessfully, thus far) to escape Philadelphia Electric and its high-cost Limerick power in favor of Pennsylvania Power & Light, just as U.S. Steel's Gary Works tried unsuccessfully to shift from Illinois Power to Commonwealth Edison. To avoid the large rate increases in prospect with the completion of Middle South Utilities' Grand Gulf nuclear plant, the governor of Arkansas, anxious to keep his state competitive, has threatened to take Arkansas Power & Light out of the Middle South system altogether.
What is so heartbreaking is that if things had been different--if the times were not what they were, if the regulators had been more responsible, if the utility managements and nuclear contractors had been more competent, and, of course, if the antinuclear obstructionists had not been so imaginative and successful--none of it need have happened. But it did. And the U.S. will pay the price for decades to come. 2 The setup
The electric utilities burst into the nuclear age with an enormous and far from unjustified confidence in their own technical abilities and those of the experts they committed their fortunes to, an unshakable optimism about their growth prospects and a conviction that they and the commissions that regulated them had a community of interest. They would soon learn otherwise.
The first commercial reactors in the U.S. were turnkey plants, built at fixed prices and turned over, ready to go, to their owners. In the Fifties and Sixties, General Electric and Westinghouse built 21 of them. The first plants cost more than their sponsors intended--as much as $800 million, according to one estimate--but they did their job well enough to persuade the utilities that nuclear power would be cost-effective. Nuclear power was not an unmixed blessing. It involved capital costs higher than for oil and coal, but its fuel costs were low, so low that promoters claimed it could practically be given away.
Impelled by evangelism, optimism and seemingly irresistible economics, the nuclear bandwagon began to roll, and in the late Sixties and again in the early Seventies the orders began flooding in. But the suppliers had learned their lesson. The new generation of plants would be built under reimbursable-cost-plus-fixed-fee contracts. Without that, the nuclear power program would probably have sputtered out in the mid-Seventies, when costs lurched out of control.
At the time, these new plants seemed priced reasonably enough, costing maybe $200 million to $300 million apiece, a sizable but hardly insupportable commitment for a utility to take on. And the technology seemed not all that exotic. As reactor manufacturers explained it, a power plant was a power plant, and nuclear fission was just another way of heating water.
Well, even in the late Sixties it should have been clear that heating water with nuclear fuel was infinitely more complex and sophisticated than doing it with coal, and complexity meant cost. Perhaps the clearest signal was that coal remained competitive, so competitive that, as the nuclear program gathered momentum, most of the new plants had to be scaled up--to maybe 1.2 million kilowatts--to retain their cost advantage. In scaling up, however, the utilities were moving into unexplored territory. Nobody knew what the design problems were or, worse, what they were going to become. Says J. Christopher Young, senior vice president of Donaldson, Lufkin & Jenrette, "The assumption was you had a mature technology when in fact it was still evolving."
The new nuclear plants, according to Atomic Industrial Forum estimates, embody 50% to 85% more design work, more hours of craft labor and more materials than the earlier generation of nuclear plants--5 million to 6 million feet of cable, 30,000 pipe hangers, 300,000 cubic feet of concrete, millions of parts altogether. "In a big coal plant," says Duke Power Executive Vice President Warren H. Owen, "you would have no more than 200 critical pipe hangers. In a nuclear plant, you need 30,000. In a coal plant, you used a half-hour of engineering time and 2 hours perhaps to put those hangers up. In a nuclear plant, you spend 15 hours on the hanger design, 75 to 150 hours to put them up. So you've got more than 100 hours per hanger and 30,000 to put up."
The industry, however, had little doubt about its ability to meet any challenge presented to it. As the Sixties ended, after all, electric utilities had closed out the most prosperous decade in their 100-year history. Year after year, the scale economies of power production enabled them to reduce the costs of producing power, and those economies were so accessible that even mediocre managements could partially achieve them.
Thanks to their declining costs, the utilities came to be thought of as a growth business, showing a return on equity, despite regulation, pretty much in line with the rest of U.S. industry. And with consumers getting the benefit of steadily declining power costs, most utilities succeeded in establishing a long-term community of interest with the politically sensitive state commissions that regulated them. Until the late Sixties at least, the commissions had little to do other than step in from time to time and order a rate reduction when returns got unconscionably high.
As these cost reductions began to tail off and reverse themselves in the late Sixties, the utilities and the commissions were able to work out a series of accounting accommodations--deferred-tax and flow-through accounting, for example--that enabled the utilities to keep their rates low (taking the political heat off the commissions) and their reported earnings high (protecting the utilities in the stock market). Later on, in the early Seventies, when energy costs first began rising rapidly, the commissions turned more and more to fuel adjustment clauses to enable higher fuel costs to be quickly passed along in higher rates. Later still, when rising construction costs began posing financial problems, a number of commissions even permitted the utilities to include a portion of their construction work in progress--CWIP, as it became known--in the rate base, on which they could earn a cash return. Consumerist pressures to hold rates down were rising, but the community of interest between the utilities and the commissions endured.
This genial climate was to play a considerable part in the miscalculations that followed. Thus, when demand slackened in the middle Seventies and the utilities' financial problems threatened to become acute, many companies did not hesitate to stretch out their nuclear programs. "They had every confidence that the old system would work," AIF President Carl Walske says," and that when the plant was finished, and it was time to get it in the rate base, the regulators would put it into the rate base."
The utilities, after all, were a cost-recovery industry, whose executives did not think about cost the way executives in other industries did. When accused of committing $50 million of stockholders' money to a nuclear project the government was prepared to finance, the chairman of Consolidated Edison once shrugged and said, "Hell, it's only $50 million." And what was $50 million to a $3 billion company? But shrug often enough, and the cost begins to add up.
But why worry even then? Few utility managers thought demand growth would stay low. Electrical demand had been growing at 7% a year for over 40 years, and despite some fits and starts in the early Seventies, there was no reason to think that anything fundamental had changed. The energy crisis, in fact, made nuclear power more vital than ever. The Nixon Administration's 1974 Project Independence counted on nuclear power to provide 40% of U.S. electricity by 1990, 50% by 2000. "You were dealing with a mind-set that had no idea of what was down the road," says Salomon Brothers analyst Neal Kurzner. "If you said, 'Maybe you're a bit off,' they didn't care. They went ahead with their construction program."
In a program geared to a 7% annual growth pattern, however, the mathematics of miscalculating the market was awesome. At 7% annual growth, you needed to double your capacity every 10 years. At 3%, every 24 years. At 1%, every 70 years. After the second runup in oil prices, in 1978, utility executives had to accept the fact that something basic might be happening to demand, that all the nuclear capacity they were planning might not be needed until a decade or more later than they thought. The Tennessee Valley Authority canceled 8 out of 17 nuclear projects, Public Service Electric & Gas 5 out of 8, Duke Power 6 out of 13, Detroit Edison 3 out of 4. The cancelations continue.
Of all the woes that descended on the U.S. nuclear industry beginning in 1978--high inflation, high interest rates, slackening demand--none was to prove more traumatic than the accident at Three Mile Island on Mar. 28, 1979. On that day, a stuck valve and inadequately trained operating personnel caused the nuclear reactor to overheat dangerously. Though the reactor core was damaged, no one was hurt or even seriously endangered, but the accident demonstrated the fallibility of the utilities, the NRC and the manufacturers, and produced a public relations disaster from which nuclear power has not yet recovered.
After Three Mile Island--TMI, as everyone in the industry antiseptically calls it--safety became an obsession. The NRC promulgated hundreds of new safety regulations between 1978 and 1983 to cope with contingencies ranging from earthquakes to missile strikes. Says engineering consultant James O. Love: "An implicit 'perfection standard' was emerging as the NRC's expectation and measuring stick." The utilities had no choice but to adjust, in some instances tearing apart nearly completed plants to conform to the changes, or redesigning plants just under way. The process seemed unending. "At the McGuire plant," says Bob Dick, Duke Power's construction vice president, "we were really putting in with one hand and taking out with the other."
Some utilities had risked starting construction with designs only 20% complete, rather than the 40% considered minimal, only to discover, with the NRC changes, that they were not even that far along. Florida Power & Light's much admired St. Lucie 2, which started out with a design 70% complete, wound up with a design only 30% complete. The logistical and morale problems were overwhelming. "These plants are so congested you can't be sure all the pipes are going to go where they're designed," says Bob Dick. "You design, you attempt to erect, you change, you analyze, and it may recycle two or three times."
"The work force becomes inefficient as a result of the delays," says Stone & Webster Senior Vice President John Landis. "You lose workers, usually the very best workers, and when you restart, the workers that remain make mistakes. The delays are, in my opinion, the most costly factor in the nuclear power economic picture."
There was another cost consequence of TMI, more subtle than delays as such but scarcely less profound. "After TMI," as AIF's Carl Walske sums it up, "the balance of decision-making power shifted sharply from the utilities to the regulators." The NRC itself didn't worry about the cost impact of the regulations it imposed; it wasn't supposed to. And under cost-plus contracts, of course, suppliers and contractors had virtually no incentive to be efficient; they were reimbursed for the costs of retrofitting NRC-dictated changes as they were for any others. The industry itself was on the ropes, and most companies were prepared to do whatever was asked of them, never mind the cost. When the PSC charged Detroit Edison with wasting money out of a "near-paranoid concern for environment and safety," utility Chairman Walter J. McCarthy readily conceded: "This is a charge to which we gladly plead guilty."
Many utilities were totally unprepared for projects requiring management skills of such a high order. The NRC shut down Marble Hill and South Texas for over a year because construction managers there had problems pouring concrete correctly. Contractors had difficulties welding metal almost everywhere--Marble Hill, Zimmer, Shoreham, Clinton, Byron, Limerick, Nine Mile Point, take your pick. They couldn't always install electrical equipment properly, either, or follow the specifications for steel. At Commonwealth Edison's Byron plant, the NRC chewed out the contractors who did the electrical, piping and duct work, and even banned one from supplying safety-related components to the industry. At Nine Mile Point, rejection rates in final welding and mechanical aspects for poor workmanship ran around 38%. Poor craftsmanship, an NRC study concluded, "was an effect, not the cause. The principal underlying cause . . . was found to be poor utility and project management."
"We developed the whole nuclear technology in the U.S.," laments New York consulting engineer Gerard C. Gambs, "and now the nuclear program is falling apart. Not on nuclear technology. It's falling apart on conventional construction, which I think is absolutely incredible."
Many companies even failed to comply with the NRC's requirements for documenting that a plant had met the proper quality standards. Granted, those requirements were exceptionally onerous, resulting in paperwork by the carload, microfilmed records by the millions. But the duty was clear. The penalty for failure to comply was severe--lacking that documentation, the NRC refused to accept the quality of the plant. Paperwork failure, nevertheless, was endemic--at Illinois Power's Clinton plant as at Duke Power's Catawba, at Texas Utilities' Comanche Peak as at Gulf States' River Bend, at Kansas Gas & Electric's Wolf Creek as at MSU's Grand Gulf.
The utility business may not have attracted the most talented managers in U.S. business, but most of them recognized that they didn't have the skills to handle the job and hired what they thought were the best in the business to do it for them--big guns like Stone & Webster, Bechtel, Ebasco, Sargent & Lundy, GE, Westinghouse, Combustion Engineering, Babcock & wilcox.
But the presence of any one of them on the job was no guarantee of efficient performance. Too many hands play a part in a nuclear project for any one of them to be decisive--as many as a dozen contractors and subcontractors in most projects, ranging from the architect/engineer who oversees the design and engineering of the project to the construction manager who implements the design, presiding over everything from procurement and the pouring of concrete to the installation of the mechanical and electrical systems.
Thus, both Cincinnati Gas & Electric and Public Service of Indiana hired Sargent & Lundy, the architect/engineers of Commonwealth Edison's highly successful nuclear program, to duplicate that success at the Zimmer and Marble Hill plants. But it was Edison's role as construction manager that made the big difference. PSI acted as its own construction manager at Marble Hill, while CG&E hired the Henry J. Kaiser Co., which had no previous commercial nuclear experience, to build Zimmer. Both plants were abandoned, Zimmer because its safety and quality could not be established, Marble Hill, after some quality problems, because it ran out of money.
Houston Lighting & Power turned to Texas' own Brown & Root, which offered "full responsibility" contracts to engineer and construct nuclear plants. But B&R's engineering couldn't keep up with construction, or so HL&P complained in its suit against the company. The project was halted for two years while B&R corrected faulty concrete and welding work, and the NRC complained constantly about the quality of its work. HL&P complained that, after nine years, B&R had completed less than half the engineering and one-third of the construction. An outside consultant found failures of the most basic kind: defective systems integration and overview function in the design process, lack of a well-thought-out, consistent basis for the work, failure to see that subcontractors had met design specifications, failure to document the work done.
It might be noted, in fairly assessing the dismal record, that Brown & Root had never built a nuclear plant on its own before. Neither had Kaiser. But what is to be said for Stone & Webster, which took credit for some of the most successful nuclear projects in the Sixties and early Seventies, yet went on to preside over three of the most costly and troubled plant projects still under construction--Lilco's Shoreham, NM's Nine Mile Point and Duquesne Light's Beaver Valley?
The staff of the New York Public Service Commission chewed out Stone & Webster for major deficiencies in its role at Shoreham, while the NRC complained abut quality control deficiencies at Nine Mile Point. Shoreham had plenty of problems (see box, p. 88), but S&W's John Landis suggests that they were pretty much par for the course. "We didn't have the same magnitude of
problems elsewhere that we did at Shoreham, but there were similar problems: a lot of intervention, a lot of starts and stops. Some lack of communication among the parties involved. High labor costs and, of course, construction in a period when the effects of TMI were particularly harmful."
The bungling the industry was capable of boggles the mind. In the Zimmer control room, according to a study for the Ohio PUC, the control panel would catch fire when the alarm nodule lights went on close together, so that in an emergency the panel would have knocked itself out and the staff would have been unable to control the plant. But nobody worried about that. Many of the lights had burned out, and the staff had unplugged others to decrease the risk of fire.
The ineptitude had no pattern, and virtually anything could go wrong, and did. How could an experienced contractor like Bechtel have prepared the Midland plant site so poorly that the diesel generator building began settling excessively? How could Bechtel have installed the reactor backwards at San Onofre? How could Brown & Root have got the reactor supports 45 degrees out of whack at Comanche Peak? How could experienced operators pour defective concrete at Marble Hill and the South Texas project? How could the NRC itself approve designs for the Mark II reactor when what Grand Gulf was building was a Mark III? How could design control have been so lax that PG&E used the wrong drawings in calculating seismic response for the steel in the Diablo Canyon containment building?
The NRC has a partial answer. "In some cases," an NRC study concluded, "no one was managing the project, the project had inertia, but no guidance and direction." The NRC's diagnosis may seem self-serving, but an Office of Technology Assessment study last spring came to the same conclusion: "Inadequate management has been one of the major causes of construction cost overruns and erratic operation."
In the end, things sometimes got so bad the utility managers themselves were forced to intervene. Lilco took over from Stone & Webster as construction manager at Shoreham. Houston Lighting dumped Brown & Root for Bechtel and Ebasco. Public Service of New Hampshire moved in on United Engineers. CG&E replaced Kaiser with Bechtel. PG&E itself had bungled.
Even the most carefully managed projects ran into trouble. At Georgia Power's Vogtle plant, for example, 12,000 socket welds had to be reworked and reinspected. Duke Power, the most successful of the lot, has had problems with quality control like everyone else.
And so have some of the industry's most respected suppliers--most notably the General Electric Co. Seven utilities committed themselves to GE's Mark II containment system, only to discover, back in the mid-Seventies, that the reactor containment structure was defective. In the interest of safety, the containment had to be redesigned to contain the violent forces that could be generated in the event of an accident, and that took time and money. How much money the utilities decline to say, but the modified design was not approved by the NRC until 1981. And the changes were certainly costly. A study done for the Ohio PUC urged that Cincinnati Gas & Electric sue GE for the $360 million the Mark II modifications eventually cost the company.
Most disheartening of all, perhaps, are the instances of mendacity and worse in U.S. nuclear projects. A number of major contractors allegedly rigged bids at the Washington Nuclear, Diablo Canyon and Marble Hill projects, and some have gone to jail for it. Quality control inspectors have been intimidated, documents have been forged, operator training records have been falsified. Metropolitan Edison pleaded guilty to using faculty test procedures at TMI. Wolf Creek even managed to document as inspected a weld that didn't exist.
One conclusion seems in inescapable: Any utility that failed to involve itself directly in every aspect of the project was likely to end up with a mismanaged project. And as Lilco and PSI made clear, even that may not be enough. Look at the nuclear table on page 85: On five of the eight projects likely to come in for less than $2,000 per kilowatt, the utilities involved handled their own engineering and construction. So it's a catch 22 dilemma. Lacking knowhow, utilities hired experts to manage nuclear projects. But the experts themselves all too often bungled the job.
Apologists argue that most of the nuclear industry's troubles were simply beyond the industry's control. Can't keep your costs under control? Can't meet your construction schedules? Blame the NRC, blame the antinuclear forces, blame inflation, blame interest costs. Don't blame the fact that you haven't the slightest idea of how to control costs or schedule production. "'Regulatory problems' is a kind of a catchall for all the cost overruns," says one consultant who understandably wishes not to be named. "They made mistakes and could have done a better job. The construction mentality is to build the plant and don't let paperwork slow you up."
But the NRC's requirements, while burdensome and sometimes outrageous, were not insupportable, and some utilities and construction firms had considerable success in coping with them. Arizona Public Service even anticipated many of the design changes the NRC imposed after TMI--simply because it had committed itself to building the best plant it could. And Duke Power has consistently brought in the lowest-cost and most efficient plants in the industry.
The project that serves as a model--and a reproach--to almost everyone is Florida Power & Light's St. Lucie 2 project. Despite post-TMI design changes and a brutal hurricane, St. Lucie 2 was built in six years at a cost of only $1,795 per kw. FP&L had nuclear experience, and much of the design was complete when construction began. But management made the difference. "They expected the NRC to respond to the schedule," says Russell J. Christesen, president of Ebasco Services, the engineering firm that designed and built the plant, "and whatever the NRC had to do, the company made happen. Current regulatory procedures need not necessarily preclude meeting construction schedules and budgets." 3 points of no return
What the utilities, the contractors and the suppliers began, the state regulatory commissions finished. High inflation and higher interest rates exploded the costs of the nuclear program and ultimately dissolved the community of interest between the regulators and the utilities.
Even before TMI, the costs of the industry's nuclear program had been rising faster than prices in the economy at large, faster even than costs in the construction industry as a whole. The effect of everything that went wrong after TMI--every schedule that was missed, every design that was reworked, every pipe hanger that had to be ripped out and reinstalled, every quality problem, every stretch-out, every postponement, every legal or antinuclear delay--served to double and triple these costs. Cincinnati Gas & Electric's Zimmer plant, for instance, started out in 1969 costing $240 million, was budgeted at $500 million by 1975, $1 billion by 1980 and $1.7 billion in 1982, with an additional $1.4 billion yet to be spent when construction was abandoned in 1984.
NRC regulation was responsible for at least part of these cost increases. but the accounting practices prescribed by the state regulatory commissions were responsible for even more.
If an industrial company builds a plant, the interest cost of any debt it incurs generally goes into the profit-and-loss statement as a charge against income. Not in utility accounting. Today's customers, the regulators' argument runs, shouldn't be required to pay even the interest cost of tomorrow's power. A power plant usually has to be "used and usable" before the investment can go into the rate base and earn a return.
Instead of charging off interest costs, the utility imputes an interest charge to the money it has invested in a project based on its average cost of capital, capitalizes the amount (that is, adds it to the total investment) and reports it as a noncash credit in its income statement. The credit is called AFUDC, Allowance for Funds Used During Construction. AFUDC increases reported net income, but it also increases the cost of the project. Impute a 10% or 15% return to a $2 billion investment and then capitalize it, and the cost increases by $200 million, $300 million a year. Worse, the capitalized amount is then imputed a return of its own. And so the costs of an unfinished plant expand and compound, and the longer a project takes, the more the cost compounds.
The FP&L's St. Lucie 2 plant. "The interest costs were $300 million," says project manager William Derrickson (see box, p. 91). "If we had taken two or three years more to build it, even if the bricks and mortar were the same, we would probably have incurred up to a billion dollars more."
As construction dragged on, AFUDC came to provide a major part of utility earnings--close to 100% for outfits like Lilco, 113% for Kansas Gas & Electric, and 126% for Public Service of New Hampshire. But, though the companies looked healthy, there was no cash behind these paper earnings. Utilities in such a bind had to finance their construction programs out of the cash their operating properties were generating, and when that wasn't enough, borrow money, sell more common or preferred or even cancel the dividend. When inflation, high interest rates and delay push the cost of a project into the billions, that often gets pretty difficult.
The Wall Street firm of Donaldson, Lufkin & Jenrette calculates that current investment in ongoing nuclear projects accounts for 78% of total net plant for PSNH, 75% for El Paso Electric, 64% for Louisiana Power & Light/New Orleans Public Service, 50% for Kansas City Power & Light and for Union Electric, 49% for Toledo Edison. "If you only have half of your capital generating cash," says DLJ's Christopher Young, "you're not really solvent."
For many companies in many states the cash flow problem has been eased a little by partial CWIP--construction work in progress--providing at least some return on a small portion of construction cost allowed in the rate base. But on an investment of several billions, even 20% is far from enough and not much more relief from that quarter can be expected. Regulators saw what was coming. As costs began to rise into the billions and the utilities began arguing for more generous CWIP, the state commissions become increasingly unsympathetic. Washington State, for instance, recently banned CWIP by law.
In such circumstances, poor management is not simply failing to develop's a construction project properly. It is also failing to recognize that you lack the financial resources to fund a project or, having recognized it only belatedly, failing to shut the project down. Some utility managers accepted their financial limitations and acted accordingly--Boston Edison and Dominion Resources, for example. Many more did not. "In terms of its financial viability, Public Service of New Hampshire should never have built Seabrook," DLJ Vice President Mariel Clemensen says bluntly about one case in point. "It simply could not support close to 80 of its capital in a construction project. We can't blame regulatory reluctance for the problems the companies got themselves into."
The chairman of Michigan's PSC, Eric Schneidewind, couldn't agree more: "We don't manage the utilities. We manage ultimately the amount of money they can collect, but we don't tell them you must or must not use it in certain ways. That's the reason they get paid."
Perhaps so. YEt the commissions do manage to a degree, and so they are not completely off the hook. Many commissions, after all, can approve or disapprove the financing for utility construction programs. Just as a good many companies were derelict in failing to monitor what they contractors were doing, a good many commissions were equally derelict in failing to monitor what the utilities were up to. If completion of a project was in the best interest of the ratepayers, even the normally difficult Massachusetts commission has had to concede, a utility was entitled to expect regulatory support. But all too often, it failed to get it.
The commissions are now in a box no matter what they do. With interest capitalized rather than written off, rates have to rise sharply when a new and very much higher-cost plant representing a major proportion of assets enters the rate base. It's not the plant as such that makes the difference, but the cost of the plant in relation to the rate base. Thus Niagara Mohawk may need only a 19% rate increase to account for its $5.1 billion Nine Mile Point plant, whereas Union Electric may need 42% and more to offset Callaway, which cost only half as much. Big companies like PG&E or Duquesne Light can absorb such increases with ease. But for smaller companies, the projected impact is chilling.
To mitigate the rate shock of these new plants, some commissions have been turning to various schemes for "phasing in" higher costs over a period of years. Rather than posting a single large rate increase, Kansas Gas & Electric will probably bring in Wolf Creek in stages--39.4%, 11.5%, 10.8%, 8.2%, 4.6%. Lilco has a similar plan, and Connecticut utilities are considering a similar phase-in for Millstone 3 and Seabrook 1.
For the utilities and the commissions, the honeymoon is over at last. "This is supposedly a cost-recovery industry," says Edison Electric Institute Vice President Richard Braatz. "Yet the commissions practice cost deferral. They've deferred so much by now that the pigeons are coming home to roost. What you're seeing is a realization that management has no assurance they can recover their costs. As a result, you are going to see an eroding in the commitment to service."
For some utilities it already may be too late for improvisations like phase-ins. Once they passed the one-third or halfway point in the project--the point of no return--these utilities could not afford to write off an unfinished plant. So they pushed ahead, hoping to finish the plant as quickly as possible, get it into the rate base and begin recovering their money. In several states--Ohio, Oregon, Missouri among them--they had no choice at all, because by law they could not recover the cost of "unused" properties under any circumstances.
All too often, as the utilities raced for completion, construction schedules kept lengthening and costs went on rising with no end in sight. But even when they did get the plant completed, the utilities were beginning to discover that the commissions were no longer the rubber stamps they once were. You could not put a plant into the rate base unless it was "usable"--that is, built to serve an existing demand. And so, last August, for example, the Pennsylvania Public Utility Commission in effect disallowed $287 million of the cost of PP&L's Susquehanna 1 on the grounds that the plant provided excess capacity, and it may look no more favorably on Unit 2, which recently went into operation. The Texas PUC disallowed $166 million of the cost of Houston Lighting's abandoned Allen Creek project, claiming the company should have canceled the project two years earlier, and then cut HL&P's allowed rate of return as a penalty for poor management. In New York, as noted, the PSC staff wants to disallow $1.5 billion of Lilco's Shoreham costs on the grounds that they were imprudently incurred. In California the $2.5 billion overruns in PG&E's Diablo Canyon costs over the past eight years may never find their way into the rate base.
And that was hardly the worst. A number of companies lost the gamble and ran out of cash before they could finish their construction programs--Consumers Power, Lilco, Public Service of Indiana, Public Service of New Hampshire. "The markets tended to close," says DLJ's Young. "The only way you can fund it yourself is to stop paying your common dividend. Rate relief doesn't do it because it takes too long to recover those dollars. And once you cut the dividend, the market really closes. You can't sell common or preferred, and you can only marginally sell bonds. So now you can't fund the completion of the project, and you have to abandon it. But if the PSC doesn't accept its value, you have to write off the rest, and the writeoff can be so substantial you're brought to the brink of bankruptcy."
But in the end some companies had no choice. Over the past three years, the Washington Public Power Supply System stopped work on two units it had under construction, canceled two others and defaulted on $2.2 billion in bonds. A year ago, CG&E and its partners in the Zimmer project gave up and decided to convert the site to coal. Public Service of Indiana first suspended and then, last year, canceled Marble Hill, 50% complete. In June Consumers Power halted Midland 2,85% complete. Except for Zimmer, which was halted because it could not meet the NRC's quality standards, all these plants were canceled, because their owners could not raise the money to finance them.
Those are extreme cases, all of them, but they embody the problem companies everywhere face. Utility analysts expect that most of the second units still nominally under construction will be canceled as well--Grand Gulf 2, Limerick 2, Perry 2, Seabrook 2, WPPSS 3. Clinton, Grand Gulf 1, Waterford and River Bend, among others, still suffer recurring cash shortages. Lilco appears to have solved its financing problems at the moment, but whether its Shoreham station, now virtually complete, will ever go into commercial operation is anybody's guess (see box, p. 88).
And so for the first time the threat of bankruptcy has become a reality for at least a handful of major electric companies. A bankrupt utility is obviously not going to suspend operations, but established procedures for utility bankruptcies, unlike the railroad industry, don't exist. Nobody is eager to deal with such questions as who--a utility regulator or a bankruptcy court--has primary jurisdiction over utility property, who has the authority to dispose of assets, repudiate power or fuel contracts or even, conceivably, set rates.
Nobody knows either what will happen to a utility's costs in bankruptcy. "While the lights may stay on in bankruptcy," according to Jay Worenklein and Glenn Gerstell, utility and bankruptcy experts at Milbank Tweed, the New York law firm, "the costs may be higher, the service less reliable and the economic environment less attractive to new business." With inordinately high-cost plants such as Shoreham and Seabrook, it's not inconceivable that in the short run the consumer would be better off if bankruptcy forced a drastic scaling down of all the companies' debts. Their debt burdens aside, after all, both Lilco and PSNH are demonstrably viable companies.
Nobody wants to see the utilities go bankrupt. The states would lose tax revenue and probably jobs. A Consumers Power bankruptcy would even jeopardize the $108 million in pension fund money that the state of Michigan has invested in CP bonds. Even the tough-minded New York PSC ordered a rate increase to keep Lilco from bankruptcy. Large industrial customers have been paying their utility bills early. Suppliers have been letting utilities pay their bills late. And last December a group of industrial customers even supported Consumers Power's petition for a rate increase to avert bankruptcy. So chances are the utilities in trouble will somehow be kept afloat.
The repercussions of even a single bankruptcy could be fairly widespread, undermining the integrity of utilities everywhere and demonstrating that the unthinkable needed to be thought about after all. "If New Hampshire were to fail," asks DLJ's Chris Young, "and the capital market were really to come to grips with the fact that a utility can fail, could Gulf States finish the financing required for the River Bend project? Or Middle South that needed for Grand Gulf or Waterford?"
There is one other avenue left the utilities. If they can't recover their costs through the rate structure, some state commissions have argued, they conceivably could recover them through the courts. That's why everybody in the industry these days lives under the threat of litigation. Dow Chemical is suing Consumers for the $60 million Dow says it wasted on cogeneration equipment for the aborted Midland 1 plant. Consumers, in turn, is countersuing for the $460 million Dow was obligated to provide as its share of the cogeneration deal. Houston Lighting & Power is suing Brown & Root for its alleged mismanagement of the South Texas project, and the city of Austin is suing HL&P for misrepresenting B&R's competence as a manager. B&R is suing HL&P in return. Wabash Valley is suing PSI to recover its $466 million investment in Marble Hill (see box, p. 95).
"A lot of these companies could sue some of their big contractors, but they're not going to," one consultant explains. "Termination payments for some of these canceled plants will have to be made, and the utilities would like to defer them, to work out an accommodation."
One way or another, then, the utilities are likely to come through. Indeed, once the current construction program is at an end, they should recover their traditional earning power, if not necessarily their traditional growth. In the very long run--how's this for irony?--their biggest problem may turn out to be finding an outlet for their rising cash flow (FORBES, Jan. 28, p. 31).
But there still will be a price to be paid for the nuclear industry's miscalculations, inefficiencies and mismanagement over the years. When the last nuclear plants go into operation, by one authoritative reckoning, power costs in the U.S. will be 5% higher on average than they might have been otherwise. That's hardly any burden in itself, perhaps, but in many parts of the country the price increases should be powerful enough to reshape the economic patterns of the region, weakening regions where power costs will go sharply higher and strengthening those where they will not. Lilco worries that high-cost nuclear power will drive industry away from Long Island, while Commonwealth Edison is counting on low-cost nuclear power to make northern Illinois increasingly attractive to industry. And there you have it: both the penalty and the promise of the U.S. nuclear program. The penalties will remain. The promise expires with the program itself.
Can the U.S. afford to let the nuclear option die?
Sooner or later--in the mid-Nineties, in the late Eighties even--the U.S. will almost certainly need to begin planning large central power stations again. As the Atomic Industrial Forum points out, assuming a 3% annual growth in electrical demand, the U.S. will need 300,000 megawatts of additional capacity by 2000. Only 100,000 megawatts--implying a 1% annual growth expectation--is currently being planned. Given the long lead times involved in building new capacity, someone must get moving in a timely way. If no one does, as Harvard utility expert Peter Navarro points cut, "A shortage of electricity generation could become the binding restraint on GNP growth."
There are ways to temporize, of course, between now and the end of the century even if demand does grow by as much as 3% annually. Since any shortages are likely to be regional, the utilities could strengthen their interconnections with companies and regions with surpluses. Then, too, there's plenty of excess power in Canada. And the utilities could even use their existing capacity more efficiently through various load management schemes. Finally, various small power sources--small dams, geothermal or solar energy--may prove more effective in providing capacity than most utility men expect.
But there's little likelihood of much of this happening, as matters now stand. Protective of their own territories, many utilities are reluctant to wheel power from one interconnection to another, while the U.S. itself may be reluctant to become overdependent on Canadian power supplies, even if the Canadians are willing to sell their surplus at reasonable prices.
So, after the temporizing, what kind of generation will we choose? Hardly anyone believes oil and gas prices will be low enough or supplies abundant and secure enough for oil to be a real possibility. There's coal, of course, but nobody willingly uses coal unless there is no alternative. And conceivably the acid rain problem could one day limit coal usage considerably. But if not coal, and if not oil and gas, then what?
Everywhere else in the world the answer is nuclear. Could it not be so for the U.S. if only its cost, regulatory and management problems could be solved? As Commonwealth Edison's Chairman James O'Connor points out, "American engineering, American equipment, American constructors are building pants all over the world and bringing them in at roughly one-quarter to one-third the cost of plants in the U.S. We can do it technically. We have to learn to do it institutionally, rationalizing the process to eliminate the adversarial system that we have presently. And if we learn how to do it institutionally, I think the nuclear option is very much alive."
The federal--NRC--regulatory problem that has produced open-ended costs, rework and delay, may be on the way to solving itself. The torrent of NRC regulations since TMI suggests that the NRC believes the present generation of plants is not really safe. But that may be changing. Recent research into what would happen in an accident involving a reactor core indicates that the radiation hazards are so much less than previously thought that the risk to the public is virtually nil. The China Syndrome is, like the movie that popularized the phrase, just a fiction after all.
Even without any new assessment of nuclear risk, NRC Chairman Nunzio Palladino recently urged the industry to develop one standardized design, get a site approved and demonstrate just how quickly and cheaply a standard plant can be built. Booz, Allen & Hamilton, the management consultants, have taken up the challenge and proposed an industry consortium to do just that.
Standardized designs are the industry's great panacea these days. Superficially, at least, the argument seems sound enough. You want a plant design that is 100% complete, one that the NRC could approve once and for all and so end all those delays involving construction permits and operating licenses, rework and quality concerns, at a considerable saving. In nuclear construction, the learning curve really is there. Using a standardized design, Palo Verde 2 was built in 15% less time than Palo Verde 1, for instance, and Palo Verde 3 in 30% less time than 2.
The model everyone cites is the French nuclear program, with its two large-scale standardized designs--in 900- and 1,300-megawatt sizes. The government designs and builds all the plants, regulates them, even controls the prices at which it buys the equipment. The French have the same two-stage regulatory process that the U.S. does, but they permit no public participation in the regulatory process once the project gets under way. (The Canadians go them one better and permit no public participation at all.) That prohibition may be half the battle, but whether such a system could be implemented in a far more open society like the U.S. is problematic.
Conceived as a response to the 1973 oil crisis, the French program has built 30 reactors so far, with 27 more under way. At roughly $1,000 per kw, the newest French reactors cost considerably less than most new U.S. reactors, but they probably run just about neck and neck with Commonwealth Edison's LaSalle 2 plant and Duke Power's McGuire 2 plant, especially when you consider that one-third of McGuire's costs and one-fifth of LaSalle's are capitalized interest, costs that the French reactors do not carry. U.S. nuclear costs would start coming down by at least a third if the utilities could put CWIP into the rate base.
What many advocates of the French program really seem to admire is the government's total control of the industry. At least one NRC commissioner, James Asselstine, thinks the U.S. would have been better off if the government had never lost control of the program. "The earlier reactors were built under a cooperative government-industry program, and the early reactors turned out very well," he notes. "The problem started when we scaled up, and more and more private companies got involved."
Standardization has its own drawbacks, in any case. It tends to discourage innovation, and it freezes design defects as well as solutions. As the French themselves are uncomfortably aware, their nuclear program could one day be in real trouble if they have mistakenly standardized on the current equivalent of TMI or General Electric's troubled Mark II containment structure.
And standardization is not enough in itself to solve all cost problems. Union Electric's Callaway and Kansas Gas & Electric's Wolf Creek are twins, but for all that, both are fairly high-cost twins. Indiana Public Services' Marble Hill and Commonwealth Edison's Byron were twinned plants as well, but what made Byron a success and Marble Hill a failure was not the standardized design but the managerial talents of the companies that undertook them. Management talent, or the lack of it, is what has made most of the difference in virtually every other nuclear project as well. It would be unfortunate if the technocrats forgot the central lesson of the nuclear follies.
A return to the turnkey designs of the Sixties might offer a solution to both the standardization and management problems, but no one seems to have the stomach for that right now. Properly handled, however, standardization should help reduce plant costs. Scaling down--from 1.2 million kw to 600,000 say--apparently isn't likely to help much. Nuclear power needs economies of scale to compete with coal. At a time when the growth in demand has slowed, this means that even large utilities may have to pool their resources in joint ventures, regional companies and generating companies, in order to balance capacity with consumption.
But none of these promising tactics will be easy to effect without some hard bargaining having gone before. State regulatory commissions should be able to justify the inclusion of CWIP in the rate base--after all, those are costs of an ongoing business, and consumers could reasonably be expected to finance them. But the commissions will almost certainly want something in return. "That means that management's going to have to give up some of their prerogatives," says Milbank Tweed's Jay Worenklein, "and utility companies aren't eager to have utility commissions looking over their shoulders every step of the way."
Look at the trouble Philadelphia Electric has been having. For several years now the Pennsylvania PUC has been trying to force PE to postpone Limerick 2 for lack both of a market and of adequate financial resources. PE finally suspended construction last January after the PUC refused to approve an issue of $1.1 billion in bonds, but it continues to push for completion. An exasperated PUC finally ordered it to show cause why the unit should be completed. That's the kind of interference most utilities figure they don't need.
As for joint ventures, even assuming they aren't in effect ruled out by the Federal Energy Regulatory Commission (FORBES, Dec. 31, 1984, p. 70), state regulatory agencies aren't likely to take kindly to the emergence of local generating companies to produce power for local consumption that would be regulated by the FERC. That was why the New York PSC refused to approve Empire State Power Resources, a proposed utility-owned generating company that would have taken over Shoreham. "The PSC turned it down," Jay Worenklein says, "on the basis that they would not tolerate the transfer to federal regulation of the rates relating to those plants." In short, the federalization or regionalization of regulation would stir up states' rights sensibilities.
Such are the possibilities and the problems. The means of reviving the nuclear option are available, but they cannot be implemented without sacrificing consumer interests to those of the community at large, corporate interests to those of the states, state interests to those of the federal government. Is the U.S. really so inflexible?
In the end the problem may boil down simply to this: Can a technology as rigorous and demanding and, for all that, as useful as nuclear power find a place in a society as open as the U.S.'?