No nuclear power plants in the United States ordered since 1974 will be completed, and many dozens of partially constructed plants have been abandoned. What cut off the growth of nuclear power so suddenly and so completely? The direct cause is not fear of reactor accidents, or of radioactive materials released into the environment, or of radioactive waste. It is rather that costs have escalated wildly, making nuclear plants too expensive to build. State commissions that regulate them require that utilities provide electric power to their customers at the lowest possible price. In the early 1970s this goal was achieved through the use of nuclear power plants. However, at the cost of recently completed plants, analyses indicate that it is cheaper to generate electricity by burning coal. Here we will attempt to understand how this switch occurred. It will serve as background for the next chapter, which presents the solution to these problems.
Several large nuclear power plants were completed in the early 1970s at a typical cost of $170 million, whereas plants of the same size completed in 1983 cost an average of $1.7 billion, a 10-fold increase. Some plants completed in the late 1980s have cost as much as $5 billion, 30 times what they cost 15 years earlier. Inflation, of course, has played a role, but the consumer price index increased only by a factor of 2.2 between 1973 and 1983, and by just 18% from 1983 to 1988. What caused the remaining large increase? Ask the opponents of nuclear power and they will recite a succession of horror stories, many of them true, about mistakes, inefficiency, sloppiness, and ineptitude. They will create the impression that people who build nuclear plants are a bunch of bungling incompetents. The only thing they won’t explain is how these same “bungling incompetents” managed to build nuclear power plants so efficiently, so rapidly, and so inexpensively in the early 1970s.
For example, Commonwealth Edison, the utility serving the Chicago area, completed its Dresden nuclear plants in 1970-71 for $146/kW, its Quad Cities plants in 1973 for $164/kW, and its Zion plants in 1973-74 for $280/kW. But its LaSalle nuclear plants completed in 1982-84 cost $1,160/kW, and its Byron and Braidwood plants completed in 1985-87 cost $1880/kW — a 13-fold increase over the 17-year period. Northeast Utilities completed its Millstone 1,2, and 3 nuclear plants, respectively, for $153/kW in 1971, $487/kW in 1975, and $3,326/kW in 1986, a 22-fold increase in 15 years. Duke Power, widely considered to be one of the most efficient utilities in the nation in handling nuclear technology, finished construction on its Oconee plants in 1973-74 for $181/kW, on its McGuire plants in 1981-84 for $848/kW, and on its Catauba plants in 1985-87 for $1,703/kW, a nearly 10-fold increase in 14 years. Philadelphia Electric Company completed its two Peach Bottom plants in 1974 at an average cost of $382 million, but the second of its two Limerick plants, completed in 1988, cost $2.9 billion — 7.6 times as much. A long list of such price escalations could be quoted, and there are no exceptions. Clearly, something other than incompetence is involved. Let’s try to understand what went wrong.
Understanding Construction Costs1
|Fig. 1 — The EEDB cost of a 1,000,000 kW nuclear power plant as estimated by United Engineers in various years. M.E. is median experience; B.E. is best experience; Total is labor plus materials (see text for explanation). These costs do not include escalation or interest on funds used during construction. The EEDB cost would be the actual cost if the plant were built in a very short time.|
The Philadelphia office of United Engineers and Constructors (hereafter we call it “United Engineers”), under contract with the U.S. Department of Energy, makes frequent estimates of the cost of building a nuclear power plant at the current price of labor and materials. This is called the EEDB (energy economic data base), and its increase with time is plotted in Fig. 1. Circles are estimates based on the median experience (M.E.) for all plants under construction at that time, while squares represent the best experience (B.E.), based on a small group of plants with the lowest costs. Also shown in Fig. 1 are the separate contributions of labor and materials. For the M.E. estimates, we see that in 1976, labor costs were substantially less than those of materials, while by 1988 they were more than twice the materials cost. During this 12-year period, labor costs escalated at an average rate of 18.7% compounded annually, the total cost escalated by 13.6%, and the materials cost escalated by 7.7%. Meanwhile, the national inflation rate was 5.7%, and the EEDB for coal-burning power plants escalated by 7.7% per year. For the B.E. situations, the annual escalation for nuclear plants was 8.4%.
There is little difference between B.E. and M.E. plants with regard to materials. They purchased the same items from the same suppliers for the same price. Incidentally, the equipment for generating electricity is purchased from vendors and represents only a small part of the materials cost — 24% for the nuclear steam supply system, which includes the reactor, steam generators, and pumps, and 16% for the turbine and generator. They represent only 7.4% and 5.0%, respectively, of the total EEDB cost. The rest of the cost is for concrete, brackets, braces, piping, electrical cables, structures, and installation.
While there is little difference in materials cost, we see from Fig. 1 that the difference in labor costs between M.E. and B.E. plants is spectacular. The comparison between these is broken down in Table 1. We see that about half of the labor costs are for professionals. It is in the area of professional labor, such as design, construction, and quality control engineers, that the difference between B.E. and M.E. projects is greatest. It is also for professional labor that the escalation has been largest — in 1978 it represented only 38% of total labor costs versus 52% in 1987. However, essentially all labor costs are about twice as high for M.E. as for B.E. projects. The reasons for these labor cost problems will be discussed later in this chapter in the section on “Regulatory Turbulence.”
BREAKDOWN OF LABOR COSTS FOR NUCLEAR POWER PLANTS AND COAL-BURNING PLANTS FROM THE 1987 EEDB.*
|Type of labor||Median
*Figures are in hundreds of 1987 dollars per kilowatt of plant capacity.
Source. B.L. Cohen and I.S. Lee, “A Catalog of Risks,” Health Physics 36, 707 (1979).
The total cost of a power plant is defined as the total amount of money spent up to the time it goes into commercial operation. In addition to the cost of labor and materials which are represented by the EEDB we have been discussing up to this point, there are two other very important factors involved:
Items 1 and 2 depend almost exclusively on two things, the length of time required for construction, and the rate of inflation (interest rates, averaged over long time periods are closely tied to inflation). If there were no inflation, or if plants could be built very rapidly, these factors would be close to 1.0, having little impact on the cost.
|Fig. 2 — The product of the inflation and interest factors. This is the factor by which the EEDB from Fig. 1 must be multiplied to obtain total cost. The figures above the points are the estimated number of years for the project at its initiation date.|
The product of these two factors, ESC x INT, used in the United Engineers estimates at various project initiation dates, is plotted in Fig. 2. The number of years required for construction is given above each point. We see that ESC x INT was only 1.17 in 1967, when construction times were 5.5 years and the inflation rate was 4% per year. It increased to 1.45 in 1973, when construction times stretched to 8 years but inflation rates were still only 4% per year. It went up to 2.1 in 1975-1978, when construction times lengthened to 10 years and the inflation rate averaged about 7% per year, and jumped to 3.2 in 1980 when construction times reached 12 years and the inflation rate soared to 12% per year. That is, the cost of a plant started in 1980 would have been more than triple the EEDB cost; 69% of the final cost would have been for inflation and interest.
From this analysis we can understand two more important reasons, besides skyrocketing labor prices, that explain why costs of nuclear plants completed during the 1980s were so high: their construction times were much longer than in earlier years, and they were being built during a period of high inflation. We will now discuss the reason for the longer construction times.
The Nuclear Regulatory Commission (NRC) and its predecessor, the Atomic Energy Commission Office of Regulation, as parts of the United States Government, must be responsive to public concern. Starting in the early 1970s, the public grew concerned about the safety of nuclear power plants: the NRC therefore responded in the only way it could, by tightening regulations and requirements for safety equipment.
Make no mistake about it, you can always improve safety by spending more money. Even with our personal automobiles, there is no end to what we can spend for safety — larger and heavier cars, blowout-proof tires, air bags, passive safety restraints, rear window wipers and defrosters, fog lights, more shock-absorbent bumpers, antilock brakes, and so on. In our homes we can spend large sums on fireproofing, sprinkler systems, and smoke alarms, to cite only the fire protection aspect of household safety. Nuclear power plants are much more complex than homes or automobiles, leaving innumerable options for spending money to improve safety. In response to escalating public concern, the NRC began implementing some of these options in the early 1970s, and quickened the pace after the Three Mile Island accident.
This process came to be known as “ratcheting.” Like a ratchet wrench which is moved back and forth but always tightens and never loosens a bolt, the regulatory requirements were constantly tightened, requiring additional equipment and construction labor and materials. According to one study,4 between the early and late 1970s, regulatory requirements increased the quantity of steel needed in a power plant of equivalent electrical output by 41%, the amount of concrete by 27%, the lineal footage of piping by 50%, and the length of electrical cable by 36%. The NRC did not withdraw requirements made in the early days on the basis of minimal experience when later experience demonstrated that they were unnecessarily stringent. Regulations were only tightened, never loosened. The ratcheting policy was consistently followed.
In its regulatory ratcheting activities, the NRC paid some attention to cost effectiveness, attempting to balance safety benefits against cost increases. However, NRC personnel privately concede that their cost estimates were very crude, and more often than not unrealistically low. Estimating costs of tasks never before undertaken is, at best, a difficult and inexact art.
In addition to increasing the quantity of materials and labor going into a plant, regulatory ratcheting increased costs by extending the time required for construction. According to the United Engineers estimates, the time from project initiation to ground breaking5 was 16 months in 1967, 32 months in 1972, and 54 months in 1980. These are the periods needed to do initial engineering and design; to develop a safety analysis and an environmental impact analysis supported by field data; to have these analyses reviewed by the NRC staff and its Advisory Committee on Reactor Safeguards and to work out conflicts with these groups; to subject the analyzed to criticism in public hearings and to respond to that criticism (sometimes with design changes); and finally, to receive a construction permit. The time from ground breaking to operation testing was increased from 42 months in 1967, to 54 months in 1972, to 70 months in 1980.
The increase in total construction time, indicated in Fig. 2, from 7 years in 1971 to 12 years in 1980 roughly doubled the final cost of plants. In addition, the EEDB, corrected for inflation, approximately doubled during that time period. Thus, regulatory ratcheting, quite aside from the effects of inflation, quadrupled the cost of a nuclear power plant. What has all this bought in the way of safety? One point of view often expressed privately by those involved in design and construction is that it has bought nothing. A nuclear power plant is a very complex system, and adding to its complexity involves a risk in its own right. If there are more pipes, there are more ways to have pipe breaks, which are one of the most dangerous failures in reactors. With more complexity in electrical wiring, the chance for a short circuit or for an error in hook-ups increases, and there is less chance for such an error to be discovered. On the other hand, each new safety measure is aimed at reducing a particular safety shortcoming and undoubtedly does achieve that limited objective. It is difficult to determine whether or not reducing a particular safety problem improves safety more than the added complexity reduces safety.
A more practical question is whether the escalation in regulatory requirements was necessary, justified, or cost-effective. The answer depends heavily on one’s definition of those words. The nuclear regulators of 1967 to 1973 were quite satisfied that plants completed and licensed at that time were adequately safe, and the great majority of knowledgeable scientists agreed with them. With the exception of improvements instigated by lessons learned in the Three Mile Island accident, which increased the cost by only a few percent, there were no new technical developments indicating that more expenditures for safety were needed. In fact, the more recent developments suggested the contrary (see Chapter 6). Perhaps the most significant result of safety research in the late 1970s was finding that the emergency core cooling system works better than expected and far better than indicated by the pessimistic estimates of nuclear power opponents. Another important result was finding that radioactive iodine and other elements in a water environment behave much more favorably than had been assumed.
Clearly, the regulatory ratcheting was driven not by new scientific or technological information, but by public concern and the political pressure it generated. Changing regulations as new information becomes available is a normal process, but it would normally work both ways. The ratcheting effect, only making changes in one direction, was an abnormal aspect of regulatory practice unjustified from a scientific point of view. It was a strictly political phenomenon that quadrupled the cost of nuclear power plants, and thereby caused no new plants to be ordered and dozens of partially constructed plants to be abandoned.
We now return to the question of wildly escalating labor costs for construction of nuclear plants. They were not all directly the result of regulatory ratcheting, as may be seen from the fact that they did not occur in the “best experience” projects. Regulatory ratcheting applied to new plants about to be designed is one thing, but this ratcheting applied to plants under construction caused much more serious problems. As new regulations were issued, designs had to be modified to incorporate them. We refer to effects of these regulatory changes made during the course of construction as “regulatory turbulence,” and the reason for that name will soon become evident.
As anyone who has tried to make major alterations in the design of his house while it was under construction can testify, making these changes is a very time-consuming and expensive practice, much more expensive than if they had been incorporated in the original design. In nuclear power plant construction, there were situations where the walls of a building were already in place when new regulations appeared requiring substantial amounts of new equipment to be included inside them. In some cases this proved to be nearly impossible, and in most cases it required a great deal of extra expense for engineering and repositioning of equipment, piping, and cables that had already been installed. In some cases it even required chipping out concrete that had already been poured, which is an extremely expensive proposition.
Constructors, in attempting to avoid such situations, often included features that were not required in an effort to anticipate rule changes that never materialized. This also added to the cost. There has always been a time-honored tradition in the construction industry of on-the-spot innovation to solve unanticipated problems; the object is to get things done. The supercharged regulatory environment squelched this completely, seriously hurting the morale of construction crews. For example, in the course of many design changes, miscalculations might cause two pipes to interfere with one another, or a pipe might interfere with a valve. Normally a construction supervisor would move the pipe or valve a few inches, but that became a serious rule violation. He now had to check with the engineering group at the home office, and they must feed the change into their computer programs for analyzing vibrations and resistance to earthquakes. It might take many hours for approval, and in the meanwhile, pipefitters and welders had to stand around with nothing to do.
Requiring elaborate inspections and quality control checks on every operation frequently held up progress. If an inspector needed extra time on one job, he was delayed in getting to another. Again, craft labor was forced to stand around waiting. In such situations, it sometimes pays to hire extra inspectors, who then have nothing to do most of the time. I cannot judge whether all of these new safety procedures were justifiable as safety improvements, but there was a widespread feeling among those involved in implementing them that they were not. Cynicism became rampant and morale sagged
Changing plans in the course of construction is a confusing process that can easily lead to costly mistakes. The Diablo Canyon plant in California was ready for operation when such a mistake was discovered, necessitating many months of delay. Delaying completion of a plant typically costs more than a million dollars per day.
Since delay was so expensive, plant constructors often chose to do things that appeared to be very wasteful. Construction labor strikes had to be avoided at almost any cost. Many situations arose that justified overtime work with its extra cost. There was a well-publicized situation on Long Island where a load of pipe delivered from a manufacturer did not meet size specifications. Instead of returning it and losing precious time, the pipe was machined to specifications on site, at greatly added expense.
A major source of cost escalation in some plants was delays caused by opposition from well-organized “intervenor” groups that took advantage of hearings and legal strategies to delay construction. The Shoreham plant on Long Island was delayed6 for 3 years by intervenors who turned the hearings for a construction permit into a circus. The intervenors included a total imposter claiming to be an expert with a Ph.D. and an M.D. There were endless days of reading aloud from newspaper and magazine articles, interminable “cross examination” with no relevance to the issuance of a construction permit, and an imaginative variety of other devices to delay the proceedings and attract media attention.
But the worst delay came after the Shoreham plant was completed. The NRC requires emergency planning exercises for evacuation of the nearby population in the event of certain types of accidents. The utility provides a system of warning horns and generally plans the logistics, but it is necessary to obtain cooperation from the local police and other civil authorities. Officials in Suffolk County, where Shoreham is located, refused to cooperate in these exercises, making it impossible to fulfill the NRC requirement. After years of delay, the NRC changed its position and ruled that in the event of an actual accident, the police and civil authorities would surely cooperate. It therefore finally issued an operating license. By this time the situation had become a political football, with the governor of New York deeply involved. He apparently decided that it was politically expedient to give in to the opponents of the plant. The state of New York therefore offered to “buy” the plant from the utility for $1 and dismantle it, with the utility receiving enough money from various tax savings to compensate for its construction expenditures. This means that the bill would effectively be footed by U.S. taxpayers. As of this writing, there are moves in Congress to prevent this. The ironic part of the story is that Long Island very badly needs the electricity the Shoreham plant can produce.
The Seabrook plant in New Hampshire suffered 2 years of delay7 due to intervenor activity based on the plant’s discharges of warm water (typically 80°F) into the Atlantic Ocean. Intervenors claimed it would do harm to a particular species of aquatic life which is not commercially harvested. There was nothing harmful about the water other than its warm temperatures. The utility eventually provided a large and very expensive system for piping this warm water 2½ miles out from shore before releasing it.
But again with Seabrook, the most expensive delay came after the plant was completed and ready to operate. It is located in such a way that the 5-mile radius zone requiring emergency planning extends into the state of Massachusetts. Governor Dukakis of Massachusetts, in deference to those opposed to the plant, refused to cooperate in the planning exercises. After about 3 years of delay, which added a billion dollars to the cost, in early 1990 the NRC ruled that the plant could operate without that cooperation. Governor Dukakis is appealing that decision, but the plant is now operating.
A rather different source of cost escalation is cash flow problems for utilities. When they institute a project, utilities do financial as well as technical planning. If the financial requirements greatly exceed what had been planned for, the utility often has difficulty raising the large sums of extra money needed to maintain construction schedules. It may therefore slow down or temporarily discontinue construction, which greatly escalates the final cost of the plant. For plants completed in the 1980s, this source of cost escalation was to a large extent due to regulatory turbulence, which caused the original financial planning to be so inadequate.
In summary, there is a long list of reasons why the costs of these nuclear plants were higher than those estimated at the time the projects were initiated. Nearly all of these reasons, other than unexpectedly high-inflation rates, were closely linked to regulatory ratcheting and the turbulence it created.
But what about the “best experience” plants that avoided these horrendous cost escalations. For that matter there are many plants for which the costs were much higher than indicated by “median experience” data. Nuclear plant costs vary by large factors. Almost every nuclear power plant built in the United States has been custom designed. This is due to the fact that, when they were designed, nuclear power was a young and vibrant industry in which technical improvements were frequently made. Varied responses to regulatory ratcheting also caused big differences between plants. The variations in cost from one plant to another have many explanations in addition to difference in design. Labor costs and labor productivity vary from one part of the country to another. Some constructors adjusted better than others to regulatory ratcheting; some maintained very close contact with the NRC and were able to anticipate new regulations, while others tended to wait for public announcements. Several different designs were used for containment buildings, and reactors that happened to have small containments had much more difficulty fitting in extra equipment required by new regulations. Some plants were delayed by intervenors, while others were not. Some had construction delays due to cash flow problems of the utilities. Plants nearing completion at the time of the Three Mile Island accident were delayed up to 2 years while the NRC was busy absorbing the lessons learned from that accident and deciding how to react to them.
Perhaps the most important cause of cost variations was the human factor. Some supervisors and designers adapt better than others to a turbulent situation. Some, considering it to be a very interesting challenge, developed ingenious ways of handling it, while others were turned off by it and solved problems unimaginatively by lavish spending of money. Some made expensive mistakes, while others were careful enough to avoid them. Some were so overwhelmed by the innumerable regulations, codes, standards, quality control audits, formal procedures for making design changes, and general red tape that they became ineffective, while others kept these problems in proper perspective and used their energies in a productive way. In some cases, people were able to cope with turbulence, but in most cases the regulatory ratcheting and the turbulence it caused exacted a terrible toll.
As a result, the average cost of nuclear electricity in the United States is now somewhat higher than that of electricity from coal burning. This represents a reversal of the situation in the 1970s and early 1980s, when nuclear energy provided the cheapest electricity. It is also the opposite of the situation in most other countries where electricity from nuclear energy is the least costly available alternative.
Regulatory ratcheting is really the political expression of difficulties with public acceptance. In an open society such as ours, public acceptance, or at least non-rejection, is a vital requirement for the success of a technology. Without it, havoc rules.
It is clear to the involved scientists that the rejection of nuclear power by the American public was due to a myriad of misunderstandings. We struggled mightily to correct these misunderstandings, but we did not succeed.
By the mid-1980s the battle was over. Groups that had grown and flourished through opposition to nuclear power went looking for other projects and soon found them. Many of them learned to distinguish between trivial problems and serious ones like global warming and air pollution. Some of them have even made statements recognizing that nuclear power is a solution to some of those problems.
The regulatory ratcheting, of course, has not been reversed. But the nuclear industry is now developing new reactor designs that avoid most of the problems this regulatory ratcheting has brought. It is relatively easy to accommodate regulations in the initial design stages. Moreover, the new designs go far beyond the safety goals that drove the regulatory ratcheting. The nuclear industry absorbed the message that the public wants super-super safety, and they are prepared to provide it. The next chapter describes how this will be done.
But what about costs? It is useless to develop new plant designs if they will be too expensive for utilities to purchase. In fact, they must provide electricity at a substantially lower cost than that generated by coal burning. Nuclear power has an inherent disadvantage in this competition because most of its cost lies in plant construction, which the utility must pay for up front, while much of the cost of electricity from coal burning comes from buying fuel as the plant operates. Utilities have had no problem in obtaining approval from public utility commissions for charging fuel costs directly to customers in their rates. The subject of total cost of electricity will be covered in the next chapter. We will see there how recently completed nuclear power plants fail to compete with coal-burning plants, but how this situation will be remedied in the new reactor designs.
In this chapter we have pointed out lots of problems, and perhaps given the reader a gloomy outlook about the future of nuclear power in the United States. The next chapter will present the solution to these problems, and explain why the future looks bright from the standpoint of cost as well as safety.
CHAPTER 9 addendum
Breeder reactor development represents the principal current U.S. government investment in nuclear power. It is often claimed that our government is heavily subsidizing commercial nuclear power, but this is not true; it is contrary to explicitly declared policy. Some aspects of our nuclear power program are by law conducted by the government, notably disposal of high level waste and isotope separation of uranium, i.e., increasing the ratio of U-235 to U-238 in fuel material. But the full costs for these services, as well as for government regulation of the industry and other services, are charged to the utilities and through them, to the cost of electricity. The cost of waste disposal is paid by an 0.1 cent/kilowatt-hour tax on nuclear electricity, which is considerably more than it is now planned to spend. The isotope separation is carried out in plants constructed in the 1950s at very low cost by present standards to produce materials for military applications, but the charges for their services to utilities are computed as though they were constructed today. Government policy is to finance research and development of future nuclear technology, but operation of the industry with present technology is in no sense subsidized. On the contrary, it is very heavily penalized by the gross over-concern for safety and environmental impacts and the regulatory turbulence this engenders (see Chapter 8). This situation contrasts sharply with the case of solar energy where the U.S. government subsidizes 40% of the cost to all users through a tax credit, and where some states provide up
Breeder reactor development represents the principal current U.S. government investment in nuclear power. It is often claimed that our government is heavily subsidizing commercial nuclear power, but this is not true; it is contrary to explicitly declared policy. Some aspects of our nuclear power program are by law conducted by the government, notably disposal of high level waste and isotope separation of uranium, i.e., increasing the ratio of U-235 to U-238 in fuel material. But the full costs for these services, as well as for government regulation of the industry and other services, are charged to the utilities and through them, to the cost of electricity. The cost of waste disposal is paid by an 0.1 cent/kilowatt-hour tax on nuclear electricity, which is considerably more than it is now planned to spend. The isotope separation is carried out in plants constructed in the 1950s at very low cost by present standards to produce materials for military applications, but the charges for their services to utilities are computed as though they were constructed today. Government policy is to finance research and development of future nuclear technology, but operation of the industry with present technology is in no sense subsidized. On the contrary, it is very heavily penalized by the gross over-concern for safety and environmental impacts and the regulatory turbulence this engenders (see Chapter 8). This situation contrasts sharply with the case of solar energy where the U.S. government subsidizes 40% of the cost to all users through a tax credit, and where some states provide up to 30% additional, leaving only 30% for the user to pay.