Nuclear Power in an Age of Uncertainty February 1984

Nuclear Power in an Age of Uncertainty

February 1984

NTIS order #PB84-183953 Recommended Citation: Nuclear Power in an Age of Uncertainty (Washington, D. C.: U.S. Congress, Office of Technology Assessment, OTA-E-216, February 1984).

Library of Congress Catalog Card Number 84-601006

For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 Foreword

This report responds to a request by the House Committee on Science and Tech- nology, with the endorsement of the Senate Committee on Energy and Natural Resources, to assess the future of nuclear power in this country, and how the technology and institutions might be changed to reduce the problems now besetting the nuclear option. The report builds on an earlier OTA report, Nuclear Powerplant Standardization, and complements a current report, Managing Commerical High-Level Radioactive Waste. The present nuclear era is drawing to a close. The cover of this report show Unit I of the Washington Public Power Supply System, which was indefinitely, perhaps per- manently, deferred even though it was 60 percent complete and $2.1 billion had been invested. This plant and others such as Zimmer and Marble Hill epitomize the difficuIties facing the nuclear industry. It is important to remember, however, that other nuclear plants have been very successful and produce reliable, low cost electricity. The future of nuclear power poses a complex dilemma of policy makers. It has advantages that may prove crucial to this Nation’s energy system in the coming decades, but at present it is an option that no electric utility would seriously consider. OTA examined questions of demand growth, costs, regulation, and public acceptance to evaluate how these factors affect nuclear power’s future. We reviewed research directions which could improve conventional light water reactor technology and opportunities to develop other types of reactor concepts that might enhance safe and reliable operation. In addition, the crucial role of utility management in constructing and operating nuclear powerplants is examined at length. The controversy about nuclear safety regulation is also analyzed, and is presented with a review of current proposals for regulatory reform. Finally, the study discusses policy approaches that could assist a revival of the nuclear option should that be a choice of Congress. In the course of this assessment, OTA drew on the experience of many organizations and individuals. In particular, we appreciate the generous assistance of our distinguished advisory panel and workshop participants, as well as the efforts of the project’s consuItants and contractors. We would also like to acknowledge the help of the numerous reviewers who gave their time to ensure the accuracy and comprehensiveness of this report. To all of the above goes the gratitude of OTA, and the personal thanks of the project staff.

Director The Future of Conventional Nuclear Power Advisory Panel

George Rathjens, Chairman Center for International Studies

Jan Beyea Arthur Porter National Audubon Society Belfountain, Ontario Richard Dean David Rose General Atomics Corp. Massachusetts Institute of Technology George Dilworth Lee Schipper Tennessee Valley Authority Lawrence Berkely Labs Linn Draper James Sweeney Gulf States Utilities Energy Modeling Forum Stanford University Fritz Heimann General Electric Co. Eric Van Loon Union of Concerned Scientists Leonard Hyman Merrill Lynch, Pierce, Fenner & Smith Observers Robert Koger North Carolina Utilities Commission James K. Asselstine ‘U.S. Nuclear Regulatory Commission Myron Kratzer International Energy Associates, Ltd. Thomas Dillon U.S. Department of Energy Byron Lee Commonwealth Edison Victor Gilinsky U.S. Nuclear Regulatory Commission Jessica Tuchman Mathews World Resources Institute

Reviewers

Institute for Nuclear Power Operations Arizona Public Service Co. Ralph Lapp, Lapp Inc. William R. Freudenburg, Washington State Yankee Atomic Electric Co. University Long Island Lighting Co. Paul Slovic, Decision Research Public Service Co. of Indiana Steven Harod, Department of Energy Cincinnati Gas & Electric John Taylor, Electric Power Research Institute Pacific Gas & Electric Atomic Industrial Forum Washington Public Power Supply System Stone & Webster, inc. Office of the Governor, Washington State Bechtel Power Corp. Oregon Department of Energy Save the Valley Bonneville Power Administration Congressional Research Service Portland General Electric

iv OTA Future of Nuclear Power Project Staff

Lionel S. Johns, Assistant Director, OTA Energy, Materials, and International Security Division

Richard E. Rowberg, Energy and Materials Program Manager Alan T. Crane, Project Director Mary Proctor (Economic and Commercial Issues) Patricia S. Abel (Technical, Operational, and Regulatory Issues) Margaret Hilton (Public Acceptance) Jenifer Robison (Regulations)

Administrative Staff Lillian Quigg Edna Saunders

Contributors Warren Donnelly, Congressional Research Service Thomas Cochran, Natural Resources Defense Barbara G. Levi, Consultant, Colts Neck, N.J. Council John Crowley, United Engineers & Contractors, Todd LaPorte, University of California at Berkeley Inc. E. P. Wilkinson, Institute for Nuclear Power Alfred Amerosi, Consultant, Downers Grove, Ill. Operations James J. MacKenzie, Union of Concerned Arvo Niitenberg, Ontario Hydro Scientists Mark Mills, Science Concepts Joanne Seder, Office of Technology Assessment Richard Lester, Massachusetts Institute of Technology

Contractors Institute for Energy Analysis, Oak Ridge, Term. Terry Lash, Weehawken, N.J. Sanford Cohen & Associates, McLean, Va. NUS Corp., Gaithersburg, Md. Pickard, Lowe & Garrick, Inc., Washington, D.C. ENSA, Inc., Rockville, Md. Komanoff Energy Associates, New York MHB Technical Associates, San Jose, Calif. Science Concepts, Vienna, Va. Eliasanne Research, Inc., Washington, D.C. William J. Lanouette, Washington, D.C.

OTA Publishing Staff

John C. Holmes, Publishing Officer John Bergling Kathie S. Boss Debra M. Datcher Joe Henson Glenda Lawing Linda A. Leahy Cheryl J. Manning Workshop l—The Economic Context for Nuclear Power

Clark Bullard, Chairman Mark Mills University of Illinois Science Concepts Jack Barkenbus David Moulton Institute for Energy Analysis Energy Conservation Coalition Charles Berg Keith Paulson Buckfield, Maine Westinghouse Electric Corp. Carleton D. Burtt Doan Phung The Equitable Life Assurance Society of the Oak Ridge Associated Universities United States Paul Riegelhaupt Gordon Corey Stone & Webster, Inc. Evanston, Ill. Marc Ross John Crowley University of Michigan United Engineers & Constructors, Inc. Philip Schmidt James Edmonds University of Texas at Austin Institute for Energy Analysis Milton Searl Victor Gilinsky Electric Power Research Institute U.S. Nuclear Regulatory Commission James Sweeney Eric Hirst Energy Modeling Forum Oak Ridge National Laboratory Stanford University Leonard Hyman Vince Taylor Merrill Lynch, Pierce, Fenner & Smith Richford, Vt. Henry Kelly Jon Veigel Office of Technology Assessment Alternative Energy Corp. Robert Koger James Walker North Carolina Utilities Office of the commissioners Charles Komanoff Alvin Weinberg Komanoff Energy Associates Oak Ridge Associated Universities Mark Levine John Williams Lawrence Berkeley Laboratory Annapolis, Md. Lynn Maxwell Tennessee Valley Authority

Workshop n-Technology, Management and Regulation

Harold Lewis, Chairman Arthur Fraas University of California at Santa Barbara Institute for Energy Analysis James K. Asselstine Jerry Griffith U.S. Nuclear Regulatory Commission U.S. Department of Energy Lynne Bernebei Saul Levine Government Accountability Project NUS Corp. Dale Bridenbaugh Fred Lobbin MHB Technical Associates SC&A, Inc. Robert J. Budnitz James MacKenzie Future Resources Associates, Inc. Union of Concerned Scientists Sanford C. Cohen Mark Mills SC&A, Inc. Science Concepts John Crowley Keith Paulson United Engineers & Constructors, Inc. Westinghouse Electric Corp. Peter R, Davis Daniel Prelewicz Intermountain Technologies, Inc. ENSA, Inc. Thomas G. Dignan, Jr. George Rathjens Ropes & Gray Center for International Studies Richard Eckert Robert Renuart Public Service Electric & Gas CO. Bechtel Power Corp. Colin R. Fisher David Rose General Atomics Corp. Massachusetts Institute for Technology

vi Alan Rosenthal Alvin Weinberg U.S. Nuclear Regulatory Commission Institute for Energy Analysis R. P. Schmitz Abraham Weitzberg Bechtel Power Corp. NUS Corp. Irving Spiewak Bertram Wolfe Institute for Energy Analysis General Electric Sharon Thompson Edwin Zebroski SC&A, Inc. Institute for Nuclear Power Operations Robert E. Uhrig Florida Power & Light

Workshop Ill—Institutional Changes and Public Acceptances

Stephen Gage, Chairman John D. Long Northbrook, III. Indiana University James Asselstine R. P. MacDonald U.S. Nuclear Regulatory Commission Alabama Power Co. Jack Barkenbus Roger Mattson Institute for Energy Analysis U.S. Nuclear Regulatory Commission Edward A. Brown Eugene W. Myer New England Power Service Co. Kidder Peabody & Co., Inc. Michael Faden Keith Paulson Union of Concerned Scientists Westinghouse Electric Corp. Eldon Greenberg Irving Spiewak Galloway & Greenberg Institute for Energy Analysis Jay Hakes Linda Stansfield Office of the Governor of Florida N.J. League of Women Voters Fritz Heimann Linda Taliaferro General Electric Co. Pennsylvania Public Utilities Commission Roger Kasperson Alvin Weinberg Clark University Institute for Energy Analysis Terry Lash Weehawken, N.J.

Utility Executives Attending Workshop Ill

E. F. Cobb Hugh G. Parris Chief of Nuclear Planning and Control Manager of Power Georgia Power Co. Tennessee Valley Authority Austin J. Decker II Cordell Reid Manager, Operational Analysis and Vice President Training Commonwealth Edison Co Power Authority of the State of New York Sherwood H. Smith, Jr. James W. Dye Chairman and President Executive Vice President Carolina Power & Light Co. Long Island Lighting Co. Norris L. Stampley Jack Fager Senior Vice President for Nuclear Vice President, Engineering and Mississippi Power & Light Co. Construction Richard F. Walker Louisiana Power & Light Co. President and CEO Wendell Kelley Public Service Co. of Colorado Chairman and CEO William O. Whitt Illinois Power Co. Executive Vice president T. C. Nichols Alabama Power Co. Senior Vice President, Power Operations Russell C. Youngdall South Carolina Electric & Gas Co. Executive Vice President Warren Owen Consumers Power Co. Senior Vice President Duke Power Co.

vii Contents

Chapter Page Over view . . . . , . . . , ...... , ix

1. Introduction: The Seven-Sided Coin ...... 3

2. Summary ...... 13

3. The Uncertain Financial and Economic Future ...... 29

4. Alternative Reactor Systems ...... 83

5. Management of the Nuclear Enterprise...... 113

6. The Regulation of Nuclear Power ...... 143

7. Survival of the Nuclear Industry in the United States and Abroad. . . 179

8. Pu lic Attitudes Toward Nuclear Power ...... 211

9. Po icy Options...... 251

Appendix: Acronyms and G ossary ...... , ...... 283

Index ...... 289

ix OVERVIEW AND FINDINGS

Without significant changes in the technology, management, and level of public acceptance, nuclear power in the United States is unlikely to be expanded in this century beyond the reactors already under construction. Currently nuclear powerplants present too many financial risks as a result of uncertainties in electric demand growth, very high capital costs, operating problems, increasing regulatory requirements, and growing public opposition. If all these risks were inherent to nuclear power, there would be little concern over its demise. However, enough utilities have built nuclear reactors within acceptable cost limits, and operated them safely and reliably to demonstrate that the difficulties with this technology are not insurmountable. Furthermore, there are national policy reasons why it could be highly desirable to have a nuclear option in the future if present problems can be overcome. Demand for electricity could grow to a level that would mandate the construction of many new powerplants. Uncertainties over the long-term environmental acceptability of coal and the adequacy of economical alternative energy sources are also great and underscore the potential importance of nuclear power. Some of the problems that have plagued the present generation of reactors are due to the immaturity of the technology, and an underestimation by some utilities and their contractors of the difficulty of managing it. A major commitment was made to build large reactors before any had been completed. Many of these problems should not reoccur if new reactors are ordered. The changes that have been applied retroactively to existing reactors at great cost would be incorporated easily in new designs. Safety and reliability should be better. It is also likely that only those utilities that have adequately managed their nuclear projects would consider a new plant. While important and essential, these improvements by themselves are probably not adequate to break the present impasse. Problems such as large cost overruns and subsequent rate increases, inadequate quality control, uneven reliability, operating mishaps, and accidents, have been numerous enough that the confidence of the public, investors, rate and safety regulators, and the utilities themselves is too low to be restored easily. Unless this trust is restored, nuclear power will not be a credible energy option for this country. It appears possible, however, that additional improvements in technology and the way nuclear power is managed and regulated might be sufficient to restore the required confidence. Technological improvements, while insufficient by themselves, can nevertheless be very important in that effort. One approach would be to focus research and development (R&D) on improving current light water reactor (LWR) designs. The goal would be standardized designs representing an optimal balance of costs, safety, and operability. Private industry is unlikely to undertake all the R&D needed, so a Federal presence is probably necessary. It is also possible, however, that even greatly improved LWRs will not be viewed by the public as acceptably safe. Therefore, R&D on alternative reactors could be essential in restoring the nuclear option if they have inherently safe characteristics rather than relying on active, engineered systems to protect against accidents. Several concepts appear promising, including the high temperature gas- cooled reactor (HTGR), the PIUS reactor, and heavy water reactors. Such R&D should also be directed toward design and developing smaller reactors such as the modular HTGR. Improvements in areas outside the technology itself must start with the management of existing reactors. The Nuclear Regulatory Commission, as well as the Institute for Nuclear Power Operations, must ensure a commitment to excellence in construction and operation at the highest levels of nuclear utility management. Improved training programs, tightened procedures, and heightened awareness of opportunities for improved safety and reliability would follow. If some utilities still prove unable to improve sufficiently, consideration could be given to the suspension of operating licenses until their nuclear operations reflect the required competence, perhaps by employing other utilities or service companies, Similarly, certification of utilities or operating companies could be considered as a prerequisite for permits for new plants in order to guarantee that only qulaified companies would have responsibility. These are drastic steps, but they may be warranted because all nuclear reactors are hostage, in a sense, to the poorest performing units. Public acceptance, which is necessary if the nuclear option is to revive, depends in part on all reactors performing reliably and safely.

xi Nuclear safety regulation also can be improved even without substantial new legislation. Several utilities recently have shown that current regulatory procedures need not preclude meeting construction budgets and schedules. The regulatory process, however, is more unpredictable than necessary, and there is no assurance that safety and efficiency are being optimized. Encouraging preapproved standardized designs and developing procedures and the requisite analytical tools for evaluating proposed safety backfits would help make licensing more efficient without sacrificing safety.

The improvements in technology and operations described above should produce gains in public acceptance. Additional steps may be required, however, considering the current very low levels of support for more reactors. Addressing the concerns of the critics and providing assurance of a controlled rate of nuclear expansion could eliminate much of the reason for public disaffection. An important contribution to restoring public confidence could be made by a greater degree of openness by all parties concerned about the problems and benefits of nuclear power. If progress can be made in all these areas, nuclear power would be much more likely to be considered when new electric-generation capacity is needed. Such progress will be difficult, however, because many divergent groups will have to work together and substantial technical and institutional change may be necessary.

xii

Chapter 1 Introduction: The Seven-Sided Coin

Photo credit: Atom/c Industrial Forum Steam generator for a pressurized water reactor (PWR) arrives by barge Contents

Page The Policy Problem ...... 3 Nuclear Disincentives ...... 4 The impasse ...... 5 The Purpose of This Study ...... 8

Figure

Figure No. Page A. The Seven Sides to the Nuclear Debate ...... 7 Chapter 1 Introduction: The Seven-Sided Coin

THE POLICY PROBLEM

The nuclear power industry is facing a period coal. Such a dependence, however, would leave of extreme uncertainty. No nuclear plant now the Nation’s electric system vulnerable to price operating or still under active construction has increases and disruptions of supply. Furthermore, been ordered since 1974, and every year since coal carries significant liabilities. The continued then has seen a decrease in the total utility com- combustion of fossil fuels, especially coal, has the mitment to nuclear power. By the end of this dec- potential to release enough carbon dioxide to ade, almost all the projects still under construc- cause serious climatic changes. We do not know tion will have been completed or canceled. Pros- enough about this problem yet to say when it pects for new domestic orders during the next could happen or how severe it might be, but the few years are dim. possibility exists that even in the early 21st century it may become essential to reduce sharply Such a bleak set of conditions has led some the use of fossil fuels especially coal. Another po- observers to conclude that the industry has no tentially serious problem with coal is pollution future aside from operating the existing plants. in the form of acid rain, which already is caus- Some conclude further that such an end is en- ing considerable concern. Even with the strictest tirely appropriate because they believe that current control technology, a coal plant emits nuclear reactors will not be needed due to the large quantities of the oxides of sulfur and nitro- low growth in demand for electricity, and that gen that are believed to be the primary source the present problems are largely a result of the of the problem. There are great uncertainties in industry’s own mistakes. our understanding of this problem also, but the If nuclear power were irrelevant to future potential exists for large-scale coal combustion energy needs, it would not be of great interest to become unacceptable or much more expen- to policy makers. However, several other factors sive due to tighter restrictions on emissions. must be taken into account. While electric growth has been very low over the last decade There are other possible alternatives to coal, (in fact, it was negative in 1982), there is no of course. Improving the performance of existing assurance that this trend will continue. Even powerplants would make more electricity avail- growth that is quite modest by historical stand- able without building new capacity. Cogenera- ards would mandate new plants—that have not tion and improved efficiency in the use of elec- been ordered yet–coming online in the 1990’s. tricity also are equivalent to adding new supply. Replacement of aging plants will call for still more These approaches are likely to be the biggest con- new generating capacity. The industrial capability tributors to meeting new electric service require- already exits to meet new demand with nuclear ments over the next few decades. Various forms reactors even if high electric growth resumes. In of solar and geothermal energy also appear prom- addition, reactors use an abundant resource. Oil ising. Uncertainties of economics and applicabili- is not a realistic option for new electric-generating ty of these technologies, however, are too great plants because of already high costs and vulner- to demonstrate that they will obviate the need ability to import disruptions which are likely to for nuclear power over the next several decades. increase by the end of the century. Natural gas Therefore, there may be good national-policy may also be too costly or unavailable for gener- reasons for wanting to see the nuclear option ating large quantities of electricity. preserved. However, the purpose of the preced- The use of coal can and will be expanded con- ing discussion is not to show that nuclear power siderably. All the plausible growth projections necessarily is vital to this Nation’s well-being, It considered in this study could be met entirely by is, rather, to suggest that there are conditions

3 4 ● Nuclear Power in an Age of Uncertainty under which nuclear power would be the pre- accident or neglect. This report analyzes the tech- ferred choice, and that these conditions might nical and institutional prospects for the future of not be recognized before the industry has lost nuclear power and addresses the question of its ability to supply reactors efficiently and ex- what Congress could do to revitalize the nuclear seditiously. If the nuclear option is foreclosed, option if that should prove necessary as a national it should at least happen with foresight, not by policy objective.

NUCLEAR DISINCENTIVES

No efforts-whether by Government or the in- major changes in plant designs. Although the dustry itself–to restore the vitality of the industry reasons for these changes often are valid, will succeed without addressing the very real they lead to increases in costs and schedules problems now facing the technology. To illustrate that are unpredictable when the plant is this, consider a utility whose projections show ordered. In addition, the paperwork and a need for new generating capacity by the mid- time demands on utility management are 1990’s. in comparing coal and nuclear pIants, much greater burdens than for other gener- current estimates of the cost of power over the ating options. pIant’s lifetime give a small advantage–perhaps ● Once a plant is completed, the high capital 10 percent–to nuclear. Fifteen years ago, that costs often lead to rate increases to utility advantage would have been decisive. Now, how- customers, at least until the plant has been ever, the utility managers can see difficulties at partially amortized. This can cause consid- some current nuclear projects which, if repeated erable difficulty with both the customers and at a new pIant, would eliminate any projected the public utility commission (PUC), If rate cost advantage and seriously strain the utility: increases are delayed to ease the shock, net ● The cost projections may be inaccurate. payback to the utility is postponed further. Some plants are being finished at many times ● Most of the money to pay for a plant has to their originally estimated cost. Major portions be raised from the financial market, where of a plant may have to be rebuilt because nuclear reactors increasingly are viewed as of design inadequacy, sloppy workmanship, risky investments. The huge demands for or regulatory changes. Construction lead- capital to pay construction costs (and the times can approach 15 years, leaving the util- high interest costs on this capital) make un- ity dangerously exposed financially. The precedented financial demands on utilities severe cash flow shortages of the Washing- at a time when capital is costly. ton Public Power Supply System (WPPSS) are ● There are many opportunities for opponents an extreme example of this problem. of a plant to voice their concerns. Some ● Demand growth may continue to fall below pIants have been the focus of suits over spe- projections. A utility may commit large sums cific environmental or safety issues. in the of capital to a plant only to find part way licensing process, critics may raise a wide through construction that it is not needed. variety of issues to which the utility has to If the plant has to be canceled, the utility and be prepared to respond. These responses call its shareholders must absorb all the losses for a significant legal and technical effort as even though it looked like a reasonable in- well as long delays, regardless of the ultimate vestment at the beginning. The long con- disposition of the issue. struction schedules and great capital de- ● Plant operation may not meet expectations, mands of nuclear pIants make them especial- Some reactors have suffered chronic reliabili- ly risky in the light of such uncertainty. ty problems, operating less than 50 percent ● The Nuclear Regulatory Commission (NRC) of the time. Others have had to replace ma- continues to tighten restrictions and mandate jor components, such as steam generators, Ch. 1—Introduction: The Seven-Sided Coin ● 5

at a cost of tens of millions of dollars because any other choice. The future of nuclear power of unexpectedly rapid deterioration. While would appear to be bleak. there is no specific reason to think a new plant would not operate its full life expect- Yet there is more to nuclear power than the ancy without major repairs, no reactor is yet well-publicized problems affecting some reactors. old enough to have demonstrated it. There In fact, many have been constructed expeditious- also is the possibility of long-term shutdowns ly, and are operating with acceptable reliability. because of accidents such as Three Mile Some have enjoyed spectacular success. For in- Island. Furthermore, a nuclear utility is stance, the McGuire unit 2 of Duke Power in vuInerable to shutdowns and major modifi- North Carolina was completed in 1982 at a cost cations not only from accidents at its own of $900/kW, less than a third of the cost of the facility, but also from accidents at any other Shoreham plant in New York. The Vermont Yan- reactor. kee plant operated in 1982 at 93 percent avail- ● Public support for nuclear power has been ability, one of the best records in the world for slipping, largely due to concerns about safety any kind of generating plant. Calvert Cliffs sup- and costs. Public concerns can manifest plies electricity to Baltimore Gas & Electric cus- themselves in political opposition. Several tomers at 1.7¢/kWh. Finally, safety analyses are states have held referenda banning nuclear improving steadily, and none has indicated that power or restricting future construction. nuclear plants pose a level of risk to the public None has passed that would mandate shut- as high as that accepted readily from other tech- ting down operating reactors, but some have nologies. These well-managed plants have oper- come close. Furthermore, State and local ated safely whiIe providing substantial econom- governments have considerable control over ic benefits for their customers. the plant through rate regulation, permitting, Such examples, however, are insufficient to transportation of waste, and approval of counterbalance the problems others have en- emergency plans. If the public does not want countered. Nuclear power has become entangled the plant, all these levers are likely to be used in a complex web of such conflicting interests and against it. emotions that matters are at an impasse. The util- Given all these uncertainties and risks, few ity viewpoint discussed above shows that there utilities would now consider nuclear reactors to is little advantage and a great many disadvantages be a reasonable choice. Moreover, the pressures to the selection of a nuclear plant when new ca- arising from virtually continuous interactions with pacity is needed. Therefore, there will be few– contractors, NRC, the PUCs, financial institutions, if any—more orders for reactors in this century and perhaps lawsuits by opponents, make nucle- without significant changes i n the way the indus- ar power far more burdensome to a utility than try and the Government handle nuclear power.

THE IMPASSE

Consider now the perspective of those Federal sial, because there are at least seven parties with energy policy makers who believe the nuclear op- distinct–and often conflicting–interests: tion should be maintained in the national interest. It is unlikely that the U.S. Government will heav- ● utilities, iIy subsidize the purchase of reactors by utilities ● nuclear safety reguIators, or that it will build and operate reactors itself. ● critics of nuclear power, Therefore, new orders will be stimulated only by ● the public, alleviating those concerns and problems that now ● the nuclear supply industry, preclude such orders. Any policy initiative that ● investors and the financial community, and is proposed, however, is likely to be controver- ● State public utility commissions.

25-450 0 - 84 - 2 : QL 3 6 ● Nuclear Power in an Age of Uncertainty

To illustrate how these interests pull in different believe that the technology has so many directions for different reasons, consider just one uncertainties that much greater margins of issue. Changes in plant licensing and safety reg- safety are warranted. Thus, nuclear critics ulation often are cited as necessary elements of strenuously oppose any changes in the NRC any strategy to revitalize the option, but there is regulations that might limit their access to little agreement on either the type or extent of the regulatory process or constrain the im- reform that should be instituted. plementation of potential improvements in reactor safety. ● Before utilities will make a commitment to ● The public is yet a fourth side. Public opin- invest several billion dollars in a nuclear ion polls show a long-term trend against nu- plant, they want assurances that extensive clear power. The public demands that nu- modifications will not be necessary and that clear reactors pose no significant risks, is the regulations will remain relatively stable. frustrated by the confusing controversy sur- Utilities contend that such regulatory rounding them, and is growing increasingly changes delay construction and add greatly skeptical about any benefits from nuclear to costs without a clear demonstration of a power. These conditions do not give rise to significant risk to public health and safety. a clear mandate for regulatory reform in To the utilities, such assurances do not ap- order to facilitate more reactor orders. Such pear to be impossible to grant. They point a mandate will depend largely on improved out that NRC has licensed 80 plants and public confidence in the management ability should know what is necessary to ensure of utilities and their contractors, in the safe- operating safety. Therefore, they would sup- ty of the technology, in the effectiveness of port revisions to the regulatory process that the regulatory process, and on a perception would make it more predictable and stable, that nuclear energy offers real benefits. ● However, there is another side to this coin. ● The nuclear supply industry’s interests are No plant design has been analyzed exhaus- not synonymous with the utilities’ and thus tively for every possible serious accident se- represent a fifth side of the coin. The util- quence, and operating experience is still too ities need to meet demand with whatever limited for all the potential problems to have option appears least expensive. If that op- been identified. Accidents at Three Mile tion is not nuclear power, something else will Island and at the Browns Ferry reactor in- suffice. The supply industry, however, has volved sequences of events that were not un- a large vested interest in promoting nuclear derstood clearly enough until they occurred. reactors, and the careers of thousands of if they had been, both could have been pre- industry employees may hinge on policy vented easily. As the NRC and the industry changes to revitalize the nuclear option, in- recognize different accident sequences, cluding regulatory reform. backfits are needed to prevent future occur- ● Investors may be ambivalent about licens- rences. Proposals to reduce NRC’s ability to ing reform. Lengthy and uncertain licensing impose changes in accordance with its engi- makes nuclear power a riskier investment neering judgment will be seen by safety reg- during construction, but any accident dur- uIators as hampering their mission of ensuring operation can have the same, if not great- ing safety. er, effect. Insofar as more stringent licens- ● But there is a third side to this coin. Not only ing makes accidents less likely, it reduces the do the industry and NRC see regulatory re- financial risk. However, investors probably form very differently, but critics of nuclear will be more concerned with the near-term power find much to fault with both the util- risks involved in getting a plant online and ities and the NRC. In particular, they feel that would be more supportive of streamlined the NRC does not even enforce its present licensing if it reduced those risks. rules fully when such enforcement would be ● As representatives of consumers’ economic too costly to the industry. Furthermore, they interest, public utility commissions’ share Ch. 1—introduction: The Seven-Sided Coin ● 7

the investors’ ambivalence, but they might whether a nuclear plant will be built, how much give more weight to operating safety because it will cost, and how well it will work. Each of an accident that shuts down a reactor for a these parties has its own agenda of conditions that prolonged period usually will mean the sub- must be met before it would support a decision stitution of more expensive sources of elec- by a utility to order a reactor. These conditions tricity. are listed with each party. Those conditions that are common to all are listed at the bottom of the Thus, there are at least seven different parties figure. For instance, nuclear power must be very in each policy debate on nuclear power: seven safe, with a very low risk of core meltdowns or sides to the coin of each issue. No doubt others major releases of radioactivity. Disputes over this could be added, but those described above rep- point relate to the degree of safety required, the resent the major positions. Each party is a col- adequacy of the methodology in determining lection of somewhat differing interests, and each safety, the assumptions of the analyses, and the will look for different things in any policy initia- actual degree of compliance with regulations. in tive. Given such a multiplicity of interests, it is any case, however, existing reactors must be not surprising that the present impasse has demonstrably safe, and future reactors probably developed. will be held to even higher standards. Figure 1 illustrates these concepts. Utilities are at the center because they make the ultimate de- A closely related issue is reliability. A smoothly cision about whether to order a nuclear plant or operating reactor is more productive for its own- something else. The other parties have considers, and it also is likely to be safer than one that erable, sometimes decisive, influence over frequently suffers mishaps, even if those mishaps

Figure A.— The Seven Sides to the Nuclear Debate

Public Utility Commissions Nuclear critics Stable construction costs Confidence in the technology Minimal operating risks Confidence in regulators and utilities Adequate financing Economic advantage Public support Liabilities of other fuels proved

Utilities . Nuclear industry investors Adequate return on Investment Healthy utility Adequate financing Stable Iicensing Stable construction costs Minimal opposition National policy Minimal operating risks predictable construction costs Public acceptance Minimal political risk Public and political acceptance Favorable risk/reward . Predictable regulation ● Nuclear Regulatory Commision // Public Confidence in technology Confidence in safety Confidence in utilities confidence in regulators and utilities Public support I Lees controversy Economic advantage National policy - Noncontroversial, necessary conditions No major accidents Reactors prove reasonably reliable Additional generating units needed Cost advantage for nuclear power Convincing waste disposal program

SOURCE Off Ice of Technology Assessment 8 ● Nuclear Power in an Age of Uncertainty

have no immediate safety consequences. Thus, Many of the conditions in figure 1 are neces- it also will be considerably more reassuring to the sary before enough of the participants in the public. debate will be satisfied that nuclear power is a viable energy source for the future. It is much Other common criteria are that there must be more difficult to know how many must be met a clear need for new generating capacity and a to be sufficient. All the groups discussed above significant cost advantage for nuclear power. In have considerable influence over the future of addition, a credible waste disposal program is a nuclear power. Efforts to revive the option— prerequisite for any more orders. whether initiated legislatively, administratively, Other conditions are especially important to or by industry—are unlikely to be successful if some groups but less important to others. Some some of the interests find them unacceptable. The of these conditions already are met to some task of breaking the impasse therefore is formid- degree. The arrows in figure A drawn to the con- able. ditions under utilities indicate the major areas that are related to the other parties.

THE PURPOSE OF THIS STUDY

This report responds to requests from the aged plants, there would be much less concern House Committee on Science and Technology over the future prospects for the nuclear option. and the Senate Committee on Energy and Natural Resources asking OTA to “assess how nuclear It is the intent of this study to explore these technology could evolve if the option is to be possibilities in the light of the different interests made more attractive to all the parties of con- and different concerns discussed above. The cern” and to identify possible technical and in- report details the various difficulties facing the stitutional approaches for the Congress “that future of nuclear power and the measures that could contribute to the maintenance of this im- would be useful and practical in overcoming portant industry.” The report describes the ma- these difficulties if the Nation wishes nuclear jor impediments to nuclear power relative to power to once again be a well accepted, viable other types of generating capacity, identifies op- energy option. The technological options are re- tions that might be considered to remove those stricted to converter reactors similar to those now impediments in light of the problems and con- available on the international market. These are flicts discussed above, and explores the conse- the reactors that could be deployed in the United quences of not maintaining the nuclear option. States by the end of the century. Breeder reactors are not included because their development Changes could be made in the technology and program will not make them commercially avail- in the institutions that manage it. If a reactor were able until sometime in the next century. The to be developed that physically could not suffer other elements of the fuel cycle—uranium re- a major accident or pose health and safety risks sources and enrichment, reprocessing and waste for the public, it might allay some of the concerns disposal–are not included either. Waste has of the regulators, the interveners, and the public. been considered in great detail in a recent OTA Such a reactor might not require the ever more report. The other elements need not pose con- stringent standards of quality required for current straints to reactor orders, which is the key issue light water reactors (LWRs), thus reducing the addressed in this report. economic risks. Improvements also could be considered in management of the construction, op- This assessment was carried out with the as- eration, and regulation of reactors. If all reactors sistance of a large number of experts from all sides were to match the experiences of the best man- of the nuclear debate—utilities, nuclear critics, Ch. 1—Introduction: The Seven-Sided Coin ● 9

reactor vendors, consumer groups, NRC, aca- costs of nuclear plants, and other elements demics, State PUCs, nuclear insurers, executive in utility planning. branch agencies, the financial community, archi- ● Chapter 4 considers the technological alter- tect-engineering (AE) firms, and interested mem- natives to today’s light water reactor: im- bers of the public, As in all OTA studies, an ad- proved LWRs; the high-temperature gas re- visory panel representing most of these interests actor as evolving from the demonstration met periodicalIy during the course of the assess- plant at Fort St. Vrain, Colo.; the heavy water ment to review and critique interim products and reactor as developed in Canada; the PIUS this report. Contractors supplied analyses and concept—an LWR redesigned to make catas- background papers in support of the assessment trophic accidents essentially impossible; the (these are compiled in vol. II). In addition, OTA effects of standardization and sealing down held three workshops to review and expand on reactor size. the contractors’ reports and to ensure that all the ● Chapter 5 examines the human element in relevant interests on each issue would be con- building and operating reactors and ways to sidered. The first workshop examined the energy improve the quality of these efforts; it ana- and economic context for nuclear power, includ- lyzes the wide range of experiences in con- ing projections of electricity demand, capital costs struction costs and schedules and in reac- for powerplant construction, and the financing tor operation, new measures that may im- and rate regulation of electricity generation. The prove quality control (e.g., the Institute for second workshop focused on the technological, Nuclear Power Operations), and further managerial, and regulatory context for nuclear steps that could be implemented if existing power, identifying the problems with current efforts prove inadequate. LWRs and the licensing process for them, and as- ● Chapter 6 describes the present regulatory sessing alternative reactor technologies and pro- process and the various concerns with it, and posals for licensing revision. The third workshop evaluates the major proposals for revision. examined institutional changes, public accept- ● Chapter 7 reviews the long term viability of ance, and policy options for revitalizing nuclear the nuclear industry if no new orders are power. Based on these and other discussions, the forthcoming to see if the option would be OTA staff developed a set of policy options. Ad- foreclosed without stimulation, and how the visory panel members, contractors, and workshop operation of existing reactors would be af- participants are listed at the front of the report. fected; and examines the management of The nuclear debate long has been character- nuclear power in other countries to see what ized by inflexible, polarized positions. We see lessons can be learned from alternative some evidence that this polarization is softening approaches. For the most part, the OTA workshop participants ● Chapter 8 focuses on trends and influential and advisory panel members showed a willing- factors in public acceptance as one of the ness to compromise, including admissions by in- key elements in a revival of nuclear power, dustry representatives that many mistakes had and evaluates measures designed to improve been made, and by nuclear critics that nuclear public acceptance. power could be a viable source of electricity if ● Chapter 9 analyzes a series of policy options managed properly. that Congress might consider. Depending on one’s views of the desirability of and necessi- Volume I of this report is organized as follows: ty for nuclear power, a policy maker might • Chapter 2 presents a summary of the report. see little need to do anything, want to im- ● Chapter 3 sets the context for decisions on prove the operation of existing reactors, or the future role of nuclear power–factors af- make the option more attractive so that it can fecting electricity demand, financial consid- play an expanded role in the Nation’s energy erations including rate regulation and the future. The options are analyzed for effec- 10 • Nuclear Power in an Age of Uncertainty

tiveness and for acceptability by the various in support of the assessment, will be available parties to the debate. Packages of options are through the National Technical Information Serv- considered to see if compromises might be ice, 5285 Port Royal Rd., Springfield, Va. 22161. possible. Volume II of the report, which includes contractor reports and background papers prepared

Chapter 2 Summary

Photo credit: Department of Energy

Oconee nuclear units owned by Duke Power Co. Contents

Page The Uncertain Financial and Economic Future ...... 13 Nuclear Reactor Technology...... 15 Management of Nuclear Powerplants ...... 18 Regulatory Considerations ...... 21 Survival of the Nuclear industry in the United States and Abroad ...... 22 Public Attitudes on Nuclear Power ...... 23 Policy Options ...... 25

Tables

Tab/e No. Page I. Additional Capacity Required by 2000...... 13 2. Comparison of Construction and Reliability Records for Selected U.S. Light Water Reactors ...... 19

Figures

Figure No. Page l. Conceptual Design of a Modular, Pebble-Bed HTGR ...... 17 2. Trends in Public Opinion on Nuclear Power ...... 24 Chapter 2 Summary

THE UNCERTAIN FINANCIAL AND ECONOMIC FUTURE

Future orders for nuclear plants depend in Table 1 .–Additional Capacity Required by 2000 part on electricity demand and on the financial (gigawatts) comparisons that utilities will make with alter- Levels of replacement Electricity demand growth natives to nuclear power. Utilities ordered far of existing plants 1.5%/yr 2.5%/yr 3.5%/yr more generating capacity in the early 1970’s than Low: 50 GW to replace all plants they turned out to need, and have canceled many over 50 years old ...... 9 144 303 Moderate: 125 GW to replace all of their planned plants. Nuclear plants have plants over 40 years old, plus borne the brunt of the slowdown in construction. 20 GW of oil and gas capacity ...... 84 219 379 There has been a pronounced decline in the High: 200 GW to replace all growth rate of electricity demand. Demand plants over 40 years plus two thirds of the oil and gas growth has averaged about 2.5 percent annual- capacity ...... 159 294 454 ly since 1973, compared to about 7.0 percent NOTES: 1. Planned generating capacity for 1991 is 740 GW, 158 GW more than from 1960 to 1972. Utility executives contem- 1982 generating sources of 582 GW. Starting point for demand calculations Is 1982 summer peak demand of 428 GW. plating the construction of long Ieadtime coal 2. The calculations assume a 20-percent reserve margin, excess of or nuclear powerplants must contend with con- planned generating resources over peak demand. siderable uncertainty about the probable future SOURCE Office of Technology Assessment growth rates in electricity demand. With certain assumptions about the future, it is reasonable to expect fairly slow growth rates of 1 to 2 percent financial picture is improving as external financ- per year. Very few large new powerplants would ing needs decline and allowed rates of return in- be required to meet this demand. With other crease, current rate structures still may not pro- plausible assumptions, electricity load growth vide adequate returns for new investment in large could resume at rates of 3 to 4 percent per year, nuclear projects. Without changes in rate regu- which would require the construction of several lation, utilities may not be able to attract capital hundred gigawatts* of new powerplants by the when they need it for construction, because in- year 2000. The actual need for new powerplants vestment advisers associate construction with will depend on the growth rate of the economy, a deterioration in financial health. the rate of increase in the efficiency of use of elec- The primary targets for rate reform include tricity, price increases for electricity vis a vis other the current lag between allowed and earned energy sources, new uses for electricity, and the rates of return, the “rate shock” which results rate of retirement of existing plants. None of these in the first few years after a large, capital- variables can be predicted with certainty. The ef- intensive plant is added to the rate base, and fects of the electric growth rate and the replace- the absence of explicit incentives to reduce fuel ment rate on the capacity that would have to be costs. Options for resolving these problems ordered in time to be completed by 2000 are assume that the investors and State public utility shown in table 1. commissions will take a long-term perspective in addition to the slowdown in electric load and will maintain a particular method of deter- growth, powerplants have also been canceled mining revenue requirements for several dec- and deferred due to deterioration in the finan- ades. Yet when commissioners may only remain cial condition of utilities. Although the industry’s in office for a few years, or when State legislatures adopt a short-term perspective, methods that take *One gigawatt equals 1000 MW (1 ,000,000 kW) or slightly less a long-term view of rate regulation are difficult than the typical large nuclear powerplant of 1100 to 1300 MW. to achieve.

14 ● Nuclear Power in an Age of Uncertainty

Although ratemaking changes to increase the 1970’s and are expected to increase by another attractiveness of capital investment would 80 percent for plants now under construction. eliminate some disincentives, utilities and their Some of this increase has come from new regu- investors and ratepayers would still face sub- latory requirements which are applied to all stantial financial risks from nuclear power. plants, whether operating or under construction. These risks include the unpredictability of the Some utilities, however, have adapted better to capital costs of a nuclear plant at the beginning these new regulatory conditions, as shown by the of construction, the difficulty of predicting con- increasing variability in capital costs. Of the group struction Ieadtimes, the very high costs of cleanup of plants now under construction, the most ex- and replacement power in the event of a major pensive is expected to cost more than four times accident, and the possibility of future regulatory the least expensive. The variation in cost has been changes. due in part to regional differences in the cost of labor and materials and the weather, but more Nuclear plant average construction costs more to differences in the experience and ability of than doubled in constant dollars during the utility and construction managers. Only the best

Photo credit: United Engineers & Constructors Managing a multibillion dollar nuclear construction project is difficult, complex, and subject to uncertainty. The most expensive nuclear plant under construction is estimated to cost about four times the least expensive (per unit of generating capacity) Ch. 2—Summary ● 15 managed construction projects are now compet- value of a nuclear plant. The accident at Three itive with new coal plants. Mile island will have cost the owner $1 billion in cleanup costs alone, plus the cost of replace- Average nuclear plant construction Ieadtimes ment power, the carrying costs and amortization doubled over the decade (from about 60 to about of the original capital used to build the plant, and 120 months) and are now about 40 percent long- the cost of restarting the plant (if possible). Only er than coal plant Ieadtimes. Very long Ieadtimes $300 million of the cleanup cost was covered by increase interest costs and the difficulty of match- property insurance. Nuclear plants can also be ing capacity with demand. Average plant con- closed by referenda such as the narrowly destruction costs and leadtimes could be reduced feated vote in 1982 that would have closed Maine in the future in several ways: 1) Plants could be Yankee. built only by experienced and competent utilities and contractors, who would work under con- Utility executives have other options to meet tracts with incentives to control costs and use in- future load growth than constructing new gen- novative construction techniques. 2) Standardiza- erating plants including: converting oil or gas tion of design and licensing could bring the low- plants to coal, building transmission lines to fa- est U.S. construction costs down by another 20 cilitate purchase of bulk power, developing small to 25 percent. 3) Further reductions in plant car- hydro, wind or cogeneration sources, or load rying costs could come about if Ieadtimes were management and energy conservation programs. cut by 25 to 30 percent and interest rates were Some of these alternatives may prove more at- reduced. It should also be recognized, however, tractive to utilities than nuclear plants given the that there are circumstances under which costs uncertain demand and financial situation. Even might increase. In particular, further serious ac- if rate regulatory policies across the country were cidents or resolution of important safety issues to shift to favoring longer Ieadtime capital-inten- could lead to a new round of costly changes. sive technologies, smaller coal-fired powerplants would be preferred because they have shorter Utility executives are also aware that single Ieadtimes, lower financial risk, and greater public events couId occur causing the loss of the entire acceptance than current nuclear designs.

NUCLEAR REACTOR TECHNOLOGY

Virtually all nuclear powerplants in this coun- There is no standardized LWR design in the try and most in other countries are light water United States. This is due to two major factors. reactors (LWRs). This concept was developed for First, the different combinations of vendors, archi- the nuclear-powered submarine program, and tect-engineers (AEs), constructors, and utilities was adapted to electric utility needs. Since then, produced custom-built plants for each site. In many questions have been raised over the safe- addition, the reactor designs themselves have ty and reliability of LWRs in utility service, costs changed greatly since introduction of LWR tech- have risen dramatically and regulatory require- nology. The pace of development from proto- ments have proliferated. There is no specific in- type to nearly 100 commercial reactors was very dication that LWRs cannot operate safely for their rapid. Large, new reactors were designed and expected lifetimes, but it appears that current construction started prior to significant operating LWR designs are unlikely to be viable choices in experience of their predecessors. As hardware the future unless concerns over costs, regulatory problems developed or new safety issues sur- uncertainties and safety can be alleviated. Either faced, changes had to be made to existing reac- LWR designs will have to be upgraded, or alter- tors, rather than integrating them into new de- native reactor concepts will have to be consid- signs, As regulatory agencies improved their ered. understanding of nuclear power safety, criteria 16 . Nuclear Power in an Age of uncertainty

changed, and many features had to be mandated It is clear that, before they order new nuclear as retrofits. Thus the light water reactors under plants, utilities will want assurances that the construction and in operation today do not rep- plants will operate reliably and will not require resent an optimized LWR design. expensive retrofits or repairs due to unanticipated design problems or new Nuclear Regula- Utilities’ experience with the LWR range from tory Commission (NRC) regulations which may excellent to poor. Some reactors have operated be needed to solve such problems, and that they at up to an 80-percent capacity factor for years will not run an unacceptable risk of a TMI-type with no significant problems, while others have accident that could bankrupt them. been plagued with continual hardware problems that lead to low-capacity utilization. While the Many of the nuclear industry’s concerns safety record to date is very good, the accident about the current generation of LWRs are beat Three Mile Island (TMI) and other potentially ing addressed in designs for advanced reactors. severe incidents raise concerns about the ability An advanced pressurized water reactor is being of all utilities to maintain that record. designed to be safer and easier to operate than Many of the concerns over safety and reliability the present generation, and to have improved have been fueled by the seemingly constant fuel burnup and higher availability (90 percent stream of hardware problems and backfits asso- is the goal) through resolution of some of the ciated with LWRs. Many of those in the nuclear more critical hardware problems. An advanced industry feel that such problems reflect normal boiling water reactor is being designed to oper- progress along the learning curve of a very com- ate at a relatively high capacity factor and to in- plex technology, and they assert that the reac- corporate advanced safety features that will tors are nearing a plateau on that curve. Nuclear reduce the risk of core-melting even if the primary critics observe that there are still many unresolved cooling system fails. safety issues associated with LWRs, and the tech- If the utilities and the public cannot be con- nology must continue to change until these are vinced that new LWRs would be acceptably safe addressed adequately. and reliable, however, renewed interest may de- The design and operation of LWRs has un- velop in using alternative reactor technologies. questionably improved over time. The training Among the more promising near-term possibili- of operators has been upgraded, human factors ties are high temperature gas-cooled reactors, considerations have been incorporated into con- LWRs with inherently safe features, and the heavy trol room design, information on operating ex- water reactor. perience is shared, and numerous retrofits have been made to existing reactors. The high temperature gas-cooled reactor (HTGR) has attracted considerable interest Whether these steps have made LWRs safe because of its high thermal efficiency (nearly 40 enough cannot be demonstrated unambiguous- percent—compared to 33 percent for an LWR) ly, however. There is no consensus on how to and its inherent safety features. The core of the determine the present level of safety, nor on the HTGR is slow to heat up even if coolant flow is magnitude of risk represented by particular prob- interrupted; this reduces the urgency of the ac- lems or the cost-benefit criteria for assessing possi- tions that must be taken to respond to an acci- ble solutions. In some cases, retrofits in one area dent. In addition, the entire core and the primary can possibly reduce safety in other areas, either cooling loops are enclosed by a vessel which because of unanticipated system interactions, or wouId prevent the release of radioactive materi- simply because the additional hardware makes als even after an accident. Lessons learned on the it difficult to get into part of the plant for only U.S. operating HTGR are being applied to maintenance or repairs. Even if all the parties to the design of a 900-megawatt (MW) prototype. the debate could agree that the risks are accept- Small, modular versions (fig. 1) also have been ably small, the public still might not perceive proposed that might have very attractive safety nuclear power as safe. characteristics and be especially suitable as a Ch. 2—Summary ● 17

Figure 1 .—Conceptual Design of a Modular, Pebble-Bed HTGR Top-mounted helium circulator

Steam Feedwater outlet inlet

Reflector control rod Steam

/“’generator

200-MWt pebble-bed r reactor , core

Photo credit: Gas-cooled Reactor Associates (GCRA) I The fuel in a high temperature gas-cooled reactor is inserted into graphite blocks like the one shown above. The fuel form is one of the key features of Pebble the HTGR, since graphite can absorb a great deal transport of heat before melting system source of process heat. While HTGRs appear to be potentially safer than LWRs, there are still many questions concerning HTGR reliability and economics. Continued research and development (R&D) of the HTGR is necessary if these questions are to be resolved. fuel \ /

The heavy water reactor (HWR) has attractive Spent fuel safety and reliability features, but there are SOURCE: Office of Technology Assessment. several roadblocks to its adoption in this coun- market, unless the Canadian technology can be try. The HWR has performed well in Canada, but easily adapted, or the U.S. experience with the process of adapting it to the American envi- HWRs in the weapons program can be utilized. ronment might introduce modifications which would lower its performance. In addition, much The process inherent ultimately safe (PIUS) re- of its good performance may be the result of skill- actor, a new LWR concept being developed in ful management and not a consequence of the Sweden, is designed with safety as the primary reactor design. Without significant evidence that objective. protective against large releases of the reactor is inherently superior to other options, radioactivity would be provided by passive means the HWR is not a strong candidate for the U.S. that are independent of operator intervention and 18 ● Nuclear Power in an Age of Uncertainty mechanical or electrical components. Because would probably be required to achieve designs the PIUS is designed so that a meltdown is virtu- that exploit the favorable characteristics of small ally impossible, it might be the reactor most suited reactors. to restoring public confidence in nuclear power. The potential benefit of a standardized design The PIUS reactor is still in the initial design phase, appears to be especially high in view of the prob- however, and has not yet been tested, although lems of today’s nuclear industry. Many of the computer simulations have been initiated to ad- problems with construction and operation stem dress questions about operational stability. Exten- from mismanagement and inexperience, and a sive R&D is needed to narrow the uncertainties standardized plant would help all utilities learn about cost, operation, and maintenance. This from those who have been successful. France R&D and eventual deployment of the PIUS, and Canada seem to have done well with build- would be expedited by its similarity, in some ing many plants of one basic design. Still, the im- respects, to conventional LWRs. plementation of standardized plants in the United Features that might be applied to any reactor States faces many obstacles. Reactor system de- technology include smaller sizes and stand- signs differ from vendor to vendor and grow fur- ardized designs. Smaller nuclear plants would ther apart when coupled with the different bal- provide greater flexibility in utility planning— ance of plant designs supplied by the numerous especially in times of uncertain demand growth AEs. They are additionally modified by the re- —and less extreme economic consequences quirements of NRC, the utilities, and the specific from an unscheduled outage. The shorter con- sites. Despite these obstacles, the industry may struction periods and lower interest costs during be motivated to converge on one or two stand- construction would reduce the utilities’ financial ardized designs if that path seemed to offer exposure. The ability to build more of the sub- streamlined licensing, stabilized regulation, systems in the factory rather than onsite might faster construction, and better management. reduce some construction costs, offsetting the lost The help of the Federal Government may be re- economies of scale. Moreover, smaller reactors quired to develop and approve of a common de- might be easier to understand, more manageable sign, especially if it is significantly different from to construct, and safer to operate. Federal R&D the LWR.

MANAGEMENT OF NUCLEAR POWERPLANTS

The management of commercial nuclear pow- There are many special problems associated erplants has proven to be a more difficult task with managing a nuclear powerplant. Nuclear reac- than originally anticipated by the early pro- tors are typically half again as large and con- ponents of nuclear technology. While the overall siderably more complex than coal plants. The job safety record of U.S. plants is very good, there of building and operating a nuclear powerplant has been great variability in construction times has been further complicated by the rapid pace and capacity factors (see table 2). Some utilities of development. As new lessons were learned have demonstrated that nuclear power can be from the maturing technology, they had to be in- well managed, but many utilities have encoun- corporated as retrofits rather than integrated into tered difficulties, Some of these problems have the original design. The regulatory structure was been serious enough to have safety and finan- evolving along with the plants, and the additional cial implications. Since the entire industry is engineering associated with changing NRC reg- often judged by the worst cases, it is important ulations and with retrofits strained the already that all nuclear utilities be able to demonstrate scarce resources of many utilities. Some utilities the capability to manage their powerplants safe- have also had difficulty coordinating the various ly and reliably. participants in a construction project. Ch. 2—Summary ● 19

Table 2.—Comparison of Construction and Reliability Records for Selected U.S. Light Water Reactors

a --- b Construction—“. Reliability Lifetime Date of commercial Years to Date of commercial capacity operation construct operation factor (0/0) Best: Best: St. Lucie 2 ...... 1983 6 Point Beach 2 ...... 1972 79 Hatch 2 ...... 1979 7 Connecticut Yankee . . . . 1968 76 Arkansas Nuclear One 2 1980 7 Kewaunee ...... 1974 76 Perry 1 ...... 1985 8 Prairie Island 2 ...... 1974 76 Palo Verde 1 ...... 1984 8 Calvert Cliffs 2 ...... 1977 75 Byron 1 ...... 1984 8 St. Lucie 1 ...... 1976 74 Callaway 1 ...... 1984 8 Worst: Worst: Watts Bar 1 ...... 1984 12 Brunswick 1 ...... 1977 48 Sequoyah 2 ...... 1982 12 Indian Point 3 ...... 1976 46 Midland 1 ...... 1985 13 Salem 1 ...... 1976 46 Zimmer 1 ...... 1985 13 Brunswick 2 ...... 1975 41 Salem 2...... 1981 13 Davis Besse 1 ...... 1977 40 Diablo Canyon 2 ...... 1984 14 Palisades ...... 1971 39 Diablo Canyon 1 ...... 1984 16 Beaver Valley 1 ...... 1977 34 alnclude~ only ~lant~ licensed t. operate after the accident at Three Mile Island in March of 1979 Dlnclude~ only ~lant~ greater than IOIJ MWe in Opf3C3tiOfI for k3n9er than 3 years.

SOURCE Nuc/ear News, February 1983 and U S Nuclear Regulatory Commwslon

Both technical and institutional changes are The most important improvement required is needed to improve the management of the nu- in the internal management of nuclear utilities. clear enterprise. Technical modifications would Top utility executives must become aware of the be useful insofar as they reduce the complexity unique demands of nuclear technology. They and sophistication of nuclear plants and their not only must develop the commitment and sensitivity to system interactions and human skills to meet those demands, but they must error. More substantial design changes, such as become directly involved in their nuclear proj- the PIUS reactor concept, might be considered ects and they must impress on their project as an option since they have the potential to ad- managers and contractors a commitment to ditionally decrease the sensitivity of nuclear plants safety that goes beyond the need to meet regu- to variations in management ability. latory requirements. They also need to establish clear lines of authority and specific respon- Technological changes, by themselves, how- sibilities to ensure that their objectives will be ever, cannot eliminate all the difficulties involved met. INPO could be instrumental in stimulating in building and operating nuclear units since they an awareness of the unique management needs cannot replace commitment to quality and safe- of nuclear power and in providing guidance to ty. It is important that design changes be sup- the utilities. plemented with institutional measures to improve the management of the nuclear enter- It is also important that utilities be evaluated prise. One example is the Institute of Nuclear objectively to assure that they are performing Power Operations (INPO), * which is attempting well. Both NRC and INPO have recognized the to improve the quality of nuclear powerplant need for such evaluations, and currently are en- operation, and to enhance communication gaged in assessment activities. INPO attempts to among the various segments of the industry. assess the performance of utility management in order to identify the root causes of the problems *The Institute for Nuclear Power Operations is a self-reguIatory as well as their consequences. The NRC conducts nonprofit organization organized by the electric utilities to establish several inspection programs with the purpose of industrywide standards for the operation of nuclear powerplants, including personnel and training standards, and to ensure that util- identifying severe or recurrent deficiencies. ities meet those standards. NRC’s program is more fragmented than IN PO’s,

20 ● Nuclear Power in an Age of Uncertainty

and the relationships among its various inspec- cases, actual construction could alleviate this tion activities appear to be uncoordinated. problem. Increased vendor responsibility might encourage fixed-price contracts for nuclear Enforcement activities also can be important plants, but it could detract from utilities’ ability in encouraging better management. Both NRC to manage the plant if they are not involved ac- and INPO can take actions to encourage utilities tively in all stages of construction. to make changes or penalize them if their performance is below standard. If measures taken by NRC and INPO prove to be ineffective in pro- A second means of centralizing responsibility moting quality construction and safe operation is through nuclear service companies, which of nuclear powerplants, however, more ag- already offer a broad range of regulatory, engi- gressive action might be required. A future for neering, and other services to utilities. Nuclear nuclear power could depend on institutional service companies could help strengthen the ca- changes that demonstrate the ability of all util- pabilities of the weaker utilities by providing all ities with nuclear powerplants to operate them the services needed to build and/or operate a safely and reliably. It is not yet clear whether nuclear plant. However, utilities may be reluc- these efforts will prove adequate. tant to forego their responsibility for safety and quality while retaining financial liability. Also, Another approach might be for the NRC to re- without some mechanism that required weaker quire that a utility be certified as to its fitness to utilities to hire service companies, their existence build and operate nuclear powerplants. Certifica- might have little effect on the overall quality of tion could force the poor performers to either im- nuclear management. prove their management capabilities, obtain the expertise from outside, or choose other types of Privately owned regional or national nuclear generating capacity. power companies would extend the service com- Many of the current management problems pany concept into the actual ownership of nu- can also be traced to the overlapping and con- clear powerplants. Such companies could be flicting authority of the the utility, the reactor ven- owned by a consortium of utilities, vendors, AEs, dor, the AE, and the constructor. Centralized and/or constructors and would be created ex- responsibility for overall design and, in some pressly to build and operate nuclear powerplants. Ch. 2—Summary ● 21

REGULATORY CONSIDERATIONS

Nuclear power is one of the most intensively tion of public health and safety and national se- regulated industries in the United States, and curity, and only secondarily on additional ben- the scope and practice of regulation is a volatile efits, such as reducing the cost of nuclear plants. issue, Strong—and usually conflicting—opinions It is also important to recognize that licensing abound among the actors in the nuclear debate changes alone cannot resolve the problems of on the adequacy and efficiency of the current reg- the nuclear industry. All parties to nuclear ulatory system. regulation must commit themselves to excellence in the management of licensing, construc- The utilities and the nuclear industry have been tion, and operation, as well as to resolving out- outspoken critics of the current system of nuclear standing safety and reliability issues. plant regulation, claiming that neither the criteria nor the schedules for siting, designing, building, Many nuclear utilities are adamant that they and operating nuclear plants are predictable will not order another reactor until licensing is under the current licensing scheme. They argue more predictable and consistent. These charac- that public participation has been misused to pro- teristics should also be welcomed by the critics long licensing hearings unnecessarily. They believe since they are prerequisites for uniformly high that these factors have been the primary cause safety standards. The primary source of current of higher costs and longer construction Ieadtimes uncertainty is the potential for imposition of back- and may have been detrimental to safety. fits. Backfits serve an important safety function, since unanticipated safety problems do arise after Nuclear critics, on the other hand, have been construction permits are granted. But careful less critical of Federal regulation of nuclear revision of the backfit rule could make the proc- powerplants than of the industry that designs, ess more rational and ensure that plant safety constructs, and operates them. They argue that is not inadvertently decreased by installation or the lack of predictability and the increase in lead- maintenance problems or by unexpected inter- times were due to the immaturity of the technol- actions with other systems. Proposed changes ogy and growing pains due to rapid escalation. to the backfit rule focus on making the criteria They attribute many safety concerns to utility and for ordering backfits more explicit, such as the constructor inattention to quality assurance, and use of cost-benefit analysis. While a cost-bene- inconsistent interpretation and enforcement of fit approach would “improve” consistency, it regulations within the NRC. While some critics should not be used as the sole criterion since feel that nuclear plants will never be safe enough, the available methodologies are inadequate to others believe that the current regulatory process fully quantify safety improvements. A process could ensure safety if it were interpreted consist- to review proposed backfits could also involve ently and enforced adequately, but that limiting a centralized group either within the NRC or as the opportunities for interested members of the an independent panel drawn from utilities, the public to participate in licensing will detract from public, and the nuclear industry to ensure that safety. criteria and standards are consistently applied. As a result of these concerns, a number of mod- Legislative amendment of the Atomic Energy ifications in reactor regulation have been pro- Act is not necessary to reform the backfit regu- posed, either through legislation, rulemaking, or lations, since the changes discussed above can better management of the regulatory process. The be accomplished through rulemaking. Moreover, primary targets of the various packages are back- legislative definitions and standards may actual- fitting, the hearing process, siting, and the licens- ly reduce flexibility needed to adjust to changing of designs and plants. The evaluation of pro- ing construction and operating experience. Leg- posals for regulatory revision must depend first islative action would be more likely, however, on whether they will ensure adequate protec- to ensure predictability.

25-450 0 - 84 - 3 : QL 3 22 ● Nuclear Power in an Age of Uncertainty

Another issue in regulatory reform is the use ing process would not be even lengthier and of formal trial-type hearings in reactor licensing. more uncertain. Because adjudicatory hearings can be long and Two other proposals for changes to the current costly, proposals have been made to replace regulatory system would allow for binding pre- them with hybrid hearings, which would be more approval of reactor designs and sites. Proapproval restricted. A hybrid hearing format might be at- of standardized plant designs could make the li- tractive to the owners of nuclear powerplants, censing process more predictable and efficient but it might also limit the opportunities for public by removing most design questions from licens- inquiry and foreclose debate on safety issues. The ing. It also raises new issues, such as the degree hearings could be made more efficient without of specificity required for proapproval and the changing the format if they were managed bet- conditions under which a utility and its contrac- ter. They could also be improved by making tors could deviate from a preapproved design. greater use of rulemaking to resolve generic is- Proapproval of reactor sites is a less controver- sues and by eliminating issues not germane to sial proposal. As long as safety issues related to safety. Only the last of these changes would re- the combination of a site and a proposed plant quire legislative amendment of the Atomic are considered in subsequent licensing process, Energy Act. binding site approval would not detract from It has also been suggested that construction plant safety. Moreover, it would contribute to permits and operating licenses in the current shorter construction Ieadtimes since it would take system be combined into a single step to improve siting off the critical path for licensing. This pro- predictability and efficiency. One-step licensing, cedure, which is followed in France, Great Brit- however, raises questions on how to manage out- ain, and Japan, could even enhance public par- standing safety issues and backfits during con- ticipation by encouraging in-depth analysis at an struction without any guarantee that the licens- earlier phase.

SURVIVAL OF THE NUCLEAR INDUSTRY IN THE UNITED STATES AND ABROAD The bleak outlook for nuclear power, at least The companies that supply nuclear compo- in the near future, raises concern about the nents may keep going by supplying parts for back- long-term viability of the nuclear industry in the fits and repairs, but their numbers are expected United States and its ability to compete inter- to shrink by two-thirds in the next 3 to 5 years. nationally. Utilities will have increasing difficulty purchasing parts when needed and at expected costs. Reactor vendors may remain busy for many The cessation of new-plant orders has already years by providing operating plant services and caused some shortages in parts and services fuel loading. These companies are also expand- needed by operating plants. ing their scope and competing with the service contractors for jobs. However, in the absence of Shortages are also developing in some person- at least a few new-plant orders each year, the nel areas. The industry has vacancies for health vendors will not survive in their present form. physicists and for reactor and radiation-protection engineers, but it has a surplus of design en- The AEs will also have substantial work fin- gineers. Enrollment for nuclear engineering de- ishing construction on plants now in progress, grees has declined since the mid-1970’s, and the installing retrofits and dealing with special prob- graduate levels will barely be enough to fill the lems such as replacement of steam generators. anticipated need for 6,000 engineers for opera- The AEs may find additional activity by “recom- tion of plants by 1991, even if enrollments drop missioning,” or extending the useful life of older no more. With no fresh orders, the industry is plants. not likely to attract the best students. Ch. 2—Summary ● 23

The ability of the nuclear industry to respond In most nations with nuclear power programs, to an influx of new orders depends on the length the public has expressed some opposition. In of time before those orders arrive. If utilities re- several cases (e.g., Sweden), this has been strong quest new powerplants within 5 years, the in- enough to stop new plants from being ordered. dustry could supply them, although perhaps All nuclear industries have experienced delays with delays of a year or so to restart design in building plants, but the costs have typically teams and manufacturing processes. If the been lower. The licensing process in West Ger- hiatus in plant orders lasts 10 years, the recov- many is as complex as that in the United States, ery would be slow and not at all certain-espe- but licensing in nations such as France is stream- cially if U.S. vendors have not been selling reac- lined by strong government control and support tors abroad in that period. In that case, U.S. util- and the use of standardized designs. ities may have to buy components, if not entire The efforts by the major reactor vendor in West powerplants, from foreign suppliers. Germany to standardize its plants might prove Many of the problems that have beset the U.S. to be a useful model for the U.S. nuclear industry. nuclear industry have hampered the nuclear in- The German vendor plans to produce a series of dustries abroad, but with less severe impact in powerplants in groups of five or six whose stand- general. Worldwide forecasts of the future role ardized features will reduce delays, engineering of nuclear power have experienced the same workhours, and paperwork. Each series of stand- boom and bust that they have in the United ardized designs would build on the experience States. West Germany, France, Japan, and Can- of the previous group of plants. ada all are intending to compete with the United States in what is expected to be a very competitive international market for nuclear plants.

PUBLIC ATTITUDES ON NUCLEAR POWER

public attitudes towards nuclear power have A central factor in public concern is the fear become increasingly negative over the past of a nuclear accident with severe consequences. decade, largely because of growing concern over Surveys indicate that most people view death due safety and economics. The most recent polls in- to a nuclear accident as no worse than other dicate that only a third of the public supports causes of death, but they fear nuclear plants construction of nuclear plants, while over so because the technology is unfamiliar and fore- percent are opposed (see fig. 2). boding. Much of the loss of public confidence is a result of a series of safety-related incidents Public support is an essential ingredient in any at several reactors, especially the accident at TMI, strategy for recovery of the nuclear power op- and the evident mishandling of these incidents tion. Negative public attitudes are most directly by utilities. The likelihood of a catastrophic ac- manifested through referenda. Although all bind- cident is perceived as greater than that estimated ing referenda that would have shut existing plants by safety analysts in industry and government, have been rejected to date, some have been creating a credibility gap. close. Referenda and legislation have been approved in 11 States that will prevent construction Another factor in public concern about nucle- of any new nuclear plants unless prescribed con- ar power is the ongoing debate about nuclear ditions are met. Indirectly, public worry over plant safety among scientists and other experts. nuclear risks has been a principal reason for As the public has listened to the experts debate, NRC’s imposition of safety backfits to existing re- they have grown increasingly dubious about plant actors. State public utility commissions are unlike- safety. If the experts cannot agree, the public con- Iy to adopt rate structures favoring nuclear proj- cludes, then there must be a serious question ects unless a majority of the public is in favor. about the safety of nuclear power. 24 ● Nuclear Power in an Age of Uncertainty

Figure 2.—Trends in Public Opinion on Nuclear Power 100 100

90 90

20 10

0 0 1975 1976 1977 1978 1979 1980 1981 1982 1983 Year of survey

Question asked” “Do you favor or oppose the construction of more nuclear powerplants?" SOURCE: Cambridge Reports, Inc.

Concerns other than accidents have caused The credibility of both the industry and NRC some people to turn away from nuclear power. is low, so words and studies alone will have lit- Perhaps the largest concern after the possibility tle impact on the public. Steps to improve public of an accident is the disposal of high-level wastes attitudes towards nuclear power must rely on an generated by nuclear reactors. In addition, the actual demonstration of the safety, economics, potential esthetic and environmental damage and reliability of nuclear power. If the reactors caused by nuclear plant construction also raises currently under construction experience contin- objections. Some groups see a link between the ued cost escalation, the next generation will have military and commercial applications of atomic to be much more economic to gain public sup- power. Finally, distrust of large government and port. Alternative reactor concepts that have in- institutions has carried over, to some degree, to herent safety features, and studies of other energy both the nuclear industry and NRC. sources, including analysis of the environmental costs and benefits, also might help change People are prepared to accept some risk if public attitudes, though other concerns such as they see a compensating benefit. The high cost over waste disposal would still remain. of some nuclear plants and current excess generating capacity, however, lead many to ques- One of the most important steps in reducing tion if there is any advantage. public fears of a nuclear accident would be to improve utility management of the technology. While media coverage of nuclear power has Improved management could greatly reduce the become more extensive in recent years, there likelihood of accidents which the public views is no evidence of overall bias against nuclear as precursors to a catastrophe. While making power. The spectrum of opinion among reporters every effort to minimize both minor incidents is the same as that for the population as a whole. and more serious accidents, however, the nu- Their coverage is more likely to reflect than to clear industry should be more open about the determine society’s concerns. possibility of accidents. Ch. 2—Summary ● 25

Improved communications with nuclear respond to the substance of critic’s concerns critics might also alleviate public concerns about could reduce acrimonious debate which con- reactor safety. A concerted effort to identify and tributes to negative public opinion.

POLICY OPTIONS

Further orders for nuclear powerplants are unlikely without some government action and support. If Congress chooses to improve the chances of nuclear reactors being ordered in the future, Federal initiatives could be directed to the following goals:

● reduce capital costs and uncertainties, ● improve reactor operations and economics, ● reduce the risks of accidents that have public safety or utility financial impacts, and ● alleviate public concerns and reduce political risks. These general goals are neither new nor as controversial as the specific steps designed to achieve the goals. The initiatives discussed in this report that-are likely to be the most effective are: 1. Support a design effort to re-optimize reactor designs for safety, reliability, and overall economy. This initiative would extend the efforts of the reactor vendors. De- signs would incorporate the backfits that have occurred in existing plants and address the outstanding safety issues, thereby significantly reducing the possibility of costly changes during construction and the concern for safety in current LWRs. It would be Photo credit: William J. Lanouette expensive, however, especially if a demon- This photo shows the campaign headquarters of the Project Survival Group which supported stration plant is necessary. Proposition 15, a California referendum prohibiting 2. Improve the management of reactors under further construction of nuclear plants without a construction or in operation. Inadequate definitive resolution of nuclear waste storage. The referendum was defeated in 1976, but a management has been one of the major similar California law was recently upheld causes of construction cost overruns and er- by the Supreme Court ratic operation. Efforts are underway by the NRC and INPO to upgrade reactor management, and they should show results in im- 3. Revise the regulatory process. Many of the proved training programs, better quality con- difficulties experienced with licensing would trol and more reliable performance. The be avoided with future plants through im- congressional role in improving reactor man- proved technology and management. Im- agement would be oversight of the NRC, and proving the predictability and efficiency of support for improvements in analytical tech- licensing is a prerequisite for any further niques and resolution of the remaining safety orders, however. To a large extent this could issues. be done administratively by the NRC without 26 ● Nuclear Power in an Age of Uncertainty

legislation. Some of the elements such as may be the only way to assure the public consistent backfit evaluations and preap- that safety concerns are being addressed proval of reactor designs and sites discussed adequately. in the regulatory section above will probably 7. Control the rate of nuclear construction. require at least strong congressional over- Many of the concerns over nuclear power sight and possibly legislation. Legislation that originated from the early projections of makes the process inflexible or restricts pub- rapid growth, and expectations of a per- lic access could be counterproductive. vasive “nuclear economy. ” The present 4. Certify utilities and contractors. If efforts to modest projections appear less threatening, improve reactor management are only par- but some people will oppose all nuclear tially successful, stronger measures could be power as long as a major resurgence is warranted. A poorly performing utility can possible. Controlling the growth rate might affect the entire nuclear industry through the alleviate these concerns, thus reducing the response of the public and the NRC to inci- controversy, rebuilding public acceptance, dents with safety implications. The NRC and making some new construction might consider withdrawing the operating possible. licenses of utilities that do not demonstrate None of the options described above will be competence or commitment in managing very effective by itself. Some could be very dif- their nuclear plants. Evidence of capability ficult to implement. It appears at least possible, of the utility and its contractors might be however, that made a prerequisite for a new construction combinations of these options permit to alleviate concerns of the public, could contribute to a much more favorable envi- investors, and critics about the quality and ronment for nuclear power. cost of the plant. Whether any of these strategies would “work” 5. Support R&Don new reactors. Some new is a function of several factors including: reactor concepts have features that, if the extent to which Federal policy strategies proven out, could make them inherently resolve the problems and make nuclear safer than current operating plants, thus power more attractive, alleviating some of the concerns of the util- the electricity demand growth rate and the ities and the public. If advanced LWRs do eventual need for new powerplants, and not appear adequate to overcome these the improvement in designs and operations concerns, then the availability of an alter- - in the absence of policy initiatives. native reactor, such as the HTGR, would be important. Research, development, and The future of the nuclear industry will be demonstration of these technologies will be shaped by the evolution of these factors. The necessary to make them available. degree to which the Federal Government should 6. Address the concerns of the critics. Im- become involved (the first factor above) depends proved public acceptance is a prerequisite on an assessment of the uncertainties surround- for any new orders. At present, the public ing the other two factors. Under some conditions is confused by the controversy over safety (e.g., relatively rapid growth in demand for elec- and is therefore opposed to accepting the tricity, reliable operation of existing plants and risks of new reactors. The best way to improved technology available) a revival is quite reduce controversy would be to resolve possible. Under other conditions, even a strongly some of the disagreements between the supportive policy strategy could fail. Successful nuclear industry and its critics. This could implementation of any strategy will depend on be initiated by involving the critics more how well the concerns of all interested parties directly in the regulatory process. Involv- have been addressed. ing knowledgeable critics in regulation or in the design and analysis of new reactors

Chapter 3 The Uncertain Financial and Economic Future

Photo credit: United Engineers & Constructors Contents

Page The Recent Past: Utilities Have Built Far Less 8. Numbers of 1,000 MW Powerplants Needed Than They Planned ...... 29 in the Year 2000 ....., ...... 46 9. Additions to Overnight Construction Cost The Uncertain Outlook for Electricity Demand 31 Due to Inflation, Escalation and Interest The Top-Down Perspective on Electricity During Construction ...... 64 Demand–Sources of Uncertainty ...... 33 10. Nuclear Plant Property and Liability The Bottom-Up Perspective on Electricity Insurance ...... 68 Demand—Sources of Uncertainty ...... 35 11. Six Sets of Alternative Utility Conclusion ...... , ...... 41 Construction Plans ...... 73 Reserve Margins, Retirements, and the Need 12. Cost and Volume of Various Load for New Plants ...... 42 Management Devices ...... 74

Rate Regulation and Powerplant Finance . . . . . 46 Figures Utilities’ Current Financial Situation ...... 46 Implications of Utilities’ Financial Situation . . 50 Figure No. Page Possible Changes in Rate Regulation ...... 52 3. Planned Construction of Coal and Nuclear-Generating Capacity, 1982-91 . . . . 30 The Cost of Building and Operating 4. Comparison of Annual Ten-Year Forecasts Nuclear Powerplants ...... 57 of Summer Peak Demand ...... 30 The Rapid Increase in Nuclear Plant Cost . . . 58 5. A Comparison of the Growth Rates of Reasons for increased Construction Cost . . . . 60 Real GNP and Electricity Sales, 1960-82 . . . 32 The Increase in Nuclear Construction 6. Penetration of Air-Conditioning and Leadtimes ...... 62 Electric Heating in Residential Sector . . . . . 39 The Impact of Delay on Cost ...... 63 7. Projected Generating Capacity and The Cost of Electricity From Coal and Alternative Projections of Peak Demand . . . 43 Nuclear Plants...... 64 8. The Energy Source and Age of Existing Future Construction Costs of Nuclear Electricity Generating Capacity ...... 44 Powerplants...... 66 9. The Electric Utility industry is Dependent The Financial Risk of Operating a on Externally Generated Funds to a Greater Nuclear Powerplant ...... 68 Extent Than Other Large Capital Users . . . . 48 Impact of Risk on the Cost of Capital ...... 70 10 Since 1967, the Fraction of Utility Assets Nuclear Power in the Context of Tied Up in “Construction Work in Progress” Utility Strategies ...... 71 Has Steadily Increased ...... 48 Implications for Federal Policies , , ...... 74 11. Beginning in 1972, the Utilities’ Real Return on Equity Has Been Eroded by Inflation . . . 48 Chapter 3 References ...... 76 12. Since 1973, the Average Utility Stock Has Sold Below its Book Value...... 49 13. In 1970, a Majority of Utilities Were Rated Tables AA or Better; in 1982, a Majority Are Table No. Page Rated A or Worse ...... 49 3. Growth Rates in Electricity Demand 14. Over Time, the Share of AFUDC in Total Given Different Price Responses ...... 34 Earnings Has Increased and the Share of 4. Estimated Impact of Industrial Electrotech- Cash Has Fallen ...... 49 nologies in the Year 2000 ...... 38 15. Costs of Nuclear Units With Construction 5. Efficiency Improvement Potential From Typical Permits Issued, 1966-71 ...... 59 to Best 1982 Model: Household Appliances 40 16. Total Capital Costs for Nuclear Plants 6. Possible Needs for New Electric-Generat- Completed in 1971, 1978, and 1983-87 ing Capability to Replace Retired Powerplants, in 1982 Dollars Without Interest Loss of Powerplant Availability, and Oil and During Construction ...... 59 Gas Steam Powerplants ...... 44 17 Trends in Material Requirements in 7.installed Capacity and Net Electricity Gen- Estimates of PWR Construction Cost ...... 61 eration by Type of Generating Capacity, 18 Construction Leadtimes for Nuclear 1981-82 ...... 45 Powerplants ...... , 63 Chapter 3 The Uncertain Financial and Economic Future

If significant new electric-generating capacity ecutives will compare the risks and returns from is needed in the 1990’s and beyond, and if nu- construction of more nuclear powerplants with clear power is to provide a major fraction of that the risks and returns of several other options for capacity, utilities must order reactors some time meeting their public obligation to provide ade- before the end of the decade. Utility executives quate electric power. will compare the reliability, safety, and public ac- This chapter will explore the elements of stra- ceptance of nuclear power (issues that are ex- tegic choice for utilities that arise from the uncer- plored in other chapters in this report) relative tainty of future rates of growth in electricity de- to other types of generating capacity, especially mand, the uncertainty of economic return on in- to coal. But above all, they will treat the ques- vestment, and the uncertainty of construction and tion of whether to order a nuclear plant as a stra- operating costs of nuclear powerplants. The chap- tegic decision to be taken in the context of ex- ter will also assess the regional and national im- pected demand for electricity and the current and plications of individual utility choices. future financial status of the utility. Utility ex-

THE RECENT PAST: UTILITIES HAVE BUILT FAR LESS THAN THEY PLANNED

Utilities have built far less new electric-generat- costs are expected to be canceled and another ing capacity in the 1970’s and early 1980’s than 5 units with $3 billion to $4 billion costs may be they expected to a decade ago. In 1972, the peak- canceled (35,53). year of generating capacity forecasts, utilities overestimated construction in the last half of the Of 246 GW of orders for nuclear plants ever 1970’s by 25 percent and overestimated construc- placed, about 110 GW have been canceled. All tion in the first half of the 1980’s by 60 percent of the 13 orders placed since 1975 have been (46). Utility predictions of future generating ca- canceled or deferred indefinitely and no new pacity declined over the decade but still greatly orders have been placed since 1978. exceeded actual construction. The actual gener- Utilities expect to complete many of the nu- ating capacity of 580 gigawatts (GW) * in the sum- clear plants under construction but have not er of 1982 was about 75 GW lower than what planned to build any more. Utilities still, however, had been forecast as recently as 1977 (70). are planning to build new coal plants. A total of Both nuclear and coal plants were canceled in about 43 GW of coal construction is planned for the late 1970’s and early 1980’s, but nuclear completion from 1988 to 1991 (see fig. 3) but only plants were canceled in far greater numbers and 11 GW of nuclear capacity is planned (much of with far greater cost in sunk investments. Over which is likely to be canceled) (68). 100 nuclear units were canceled from 1972 to One obvious reason for so many canceled and 1983, more than twice the number of coal units deferred plants is that from 1973 to 1982, elec- (29,35). Almost $10 billion (1982 dollars) in invest- tricity load grew at less than half the pace (2.6 ment costs was tied up in the 26 sites (some with percent per year) that it had grown from 1960 2 units) where at least $50 million per site had to 1972 (7.1 percent per year). Most utilities in already been spent (35). Another 11 nuclear units the early 1970’s used simple trend-line forecasts totaling $2.5 billion to $3 billion in construction — that took neither gross national product (GNP) *One gigawatt equals 1,000 MW (1,000,000 kW) or slightly less or response to electricity price into account. They than the capacity of the typical large nuclear powerplant of 1,100 to 1,300 MW. GW as used in this chapter always refers to GWe were unprepared for the change in electricity or gigawatts of electricity. growth rates. Figure 4 shows a remarkable down-

29 30 ● Nuclear Power in an Age of Uncertainty

Figure 3.– Planned Construction of Coal and Nuclear- ward trend in utility forecasts of future load Generating Capacity, 1982=91 growth. it is no surprise that utilities, making plans in the late 1960’s and early 1970’s, and unable to foresee the economic changes of the 1970’s, planned to build more generating capacity than was eventually needed. Even with all the cancellations and deferrals, there was almost a 50-percent increase in electric-generating capacity from 1973 to 1982 while the average reserve margin* increased from about 21 percent in 1973 to about 33 percent (68,70). In between, the reserve margin peaked at about 37 percent as utilities completed and brought online plants that had begun construction before the slower load growth was recognized as the norm rather than 1982- 1984- 1986- 1988- 1990- as an anomaly (58). 1983 1985 1987 1989 1991 In addition to the slowdown in electric load growth, powerplants also have been canceled

SOURCE: North American Reliability Council, Electric Power Supply and and deferred due to the widely acknowledged Demand 1982-1991, August 1982. deterioration in the financial condition of utilities. At the beginning of the 1970’s, utilities enjoyed good financial health. Almost 80 percent had Figure 4.—Comparison of Annual Ten-Year Forecasts of Summer Peak Demand (contiguous U.S.) bond ratings (Standard and Poors) of AA or AAA. 800 Utility stock sold well above book value so that there was little difficulty financing new generating 700 capacity from new issues of either debt or equity. By 1981, however, utilities were in a greatly 650 weakened financial condition. Currently, there are no electric utilities with AAA bond ratings and less than a fourth with AA ratings. At its low point in 1981, utility stock sold on average at only 70 percent of book value. This meant that any issue of new stock to pay for new generating plant would dilute the value of the existing stock (see box A later in this chapter). The financial deterioration was influenced in part by the general financial conditions of the decade, especially the rapid

400 rates of inflation and the poor performance of the stock market. Beyond the influence of general economic conditions, however, the financial 350 status of utilities deteriorated because of the enor- 300 74 75 767778798081828384858687 8889909192 *“Reserve margin” is defined as the percent excess of “planned resources” over “peak demand” where “planned resources” in- SOURCE: North American Electric Reliability Council, Electric Power Supply and Demand, 1981-1991, August 1982 and advanced release of the cludes instaIled generating capacity plus scheduled capacity pur- projections for 1983-1992 in April 1983. chases less sales. Ch. 3—The Uncertain Financial and Economic Future ● 31 mous strain of financing requirements for new rate regulation that discourage those types of plants. Despite many cancellations and deferrals generating capacity that are of high capital cost of powerplants, new plant financing increased and high risk relative to other types. Over the long from about $10 billion in 1970 to over $28 billion run such rate regulation would discourage fur- in 1982, of which more than two-thirds had to ther construction of large coal and nuclear plants be financed externally (45). even when increased load and replacement of existing plants would make it sensible to construct The biggest incentive for utilities to cancel and central station plants, for which capital costs are postpone more nuclear than coal plants, was the high relative to fuel cost. more than fivefold increase in the constant dollar cost of nuclear plants from plants completed in The fifth section of the chapter lays out the un- 1971 to plants scheduled for completion in the certainties involved in constructing and operating 1980’s compared to the approximately threefold a nuclear plant which discourage utilities from increase in the constant dollar cost of coal plants ordering more nuclear plants even when they (55,56). Another incentive to favor coal plants decide to order more central station powerplants. was their shorter Ieadtime, an average of 40 to Construction costs have risen much faster than 50 percent fewer months than for a nuclear plant general price increases and vary severalfold from (l). plant to plant even when built the same year. In Looking ahead, the prospects for substantial addition, there is a financial risk of at least several numbers of new central station powerplants ap- billion dollars from an accident that disables a pear fairly uncertain. The prospects for more powerplant and more from one that causes dam- nuclear plants appear even more uncertain. The age to public health and property. To date insur- reasons why this is so are laid out in the rest of ance is available to cover only a fraction of this the chapter. The next section describes the uncer- risk. tainty about the future growth rates in electrici- Given the uncertainties of demand and nuclear ty demand. With some assumptions about the construction cost, utility decisions to cancel nu- future it is reasonable to expect that the fairly slow clear powerplants and some coal plants have growth rates (1 or 2 percent per year) of the past been sensible, and in the short-term interests of few years will continue. With equally plausible the ratepayers. Over the long run, however, if assumptions, however, electricity load growth ratemaking discourages electric-generating tech- couId resume at rates of 3 to 4 percent per year. nologies of greater capital cost and greater risk, The sources of uncertainty are described in the further investment in nuclear powerplants could next section. be discouraged even if it were in the longer term The third section explains why utilities can af- interests of ratepayers. ford to wait awhile before ordering powerplants The final section of the chapter describes the in large numbers, since reserve margins are now choices utilities have and the choices they seem so high. Sooner or later, however, as this section to be making. Under a few specific assumptions points out, some number of new powerplants will about changes in outside circumstances and rate need to be built to replace aging powerplants and regulation incentives, utilities could order nuclear meet even modest increases in electric load. plants again. It appears now, however, that they The fourth section of the chapter presents an will avoid central station construction as long as argument that there may be systematic biases in possible and then build coal plants.

THE UNCERTAIN OUTLOOK FOR ELECTRICITY DEMAND

From 1973 to 1982, annual increases in elec- growth rates were to continue for the next 20 tricity demand averaged 2.6 percent. If these years, they would provide no more than a weak 32 ● Nuclear Power in an Age of Uncertainty

stimulus to further building of central station GNP faster than an average of 3.0 percent per powerplants, including nuclear powerplants. (See year are not likely. the detailed discussion of capacity requirements The projections disagree more significantly in the next section.) It would be possible for most about the likely future relationship between utilities, with some effort, to avoid building cen- growth rates in GNP and growth rates in elec- tral station powerplants altogether until the late tricity demand. Projections of faster electricity 1990’s by encouraging conservation, load man- growth assume that electricity will increase faster agement, cogeneration, and small sources of than GNP. Projections of slower electricity power from hydro and wind; or by purchasing growth asume that electricity demand will in- from U.S. utilities with excess capacity or from crease significantly less fast than GNP. Canadian utilities; or by keeping existing plants online past normal retirement age. These strate- The ratio between electricity growth rates and gies are discussed later in the chapter. GNP growth rates has indeed dropped since the 1960’s. As is shown in figure 5, electricity growth What are the chances that the average elec- rates were about double GNP growth rates in the tricity demand growth rate will be significantly 1960’s and approximately equal to GNP growth higher or lower than 2.6 percent per year? A sig- rates (except for recession years) in the 1970’s. nificantly lower growth rate would make it diffi- Those expecting fast growth rates in electricity cult to justify any major construction of central regard the late 1970’s as an anomaly and expect station powerplants. A significantly higher growth a resurgence of a ratio of electricity to GNP rate would make a strategy of little or no power- growth of more than 1.0. Those expecting slow plant construction difficult to sustain. growth rates in electricity assume that the ratio Published projections of electricity demand of electricity growth to GNP growth will fall still reflect considerable uncertainty about future further, well below 1.0. They expect that electrici- growth rates. As is clear from figure 4 above, the ty will continue to behave like other forms of en- utilities’ own estimates of future peak demand ergy for which ratios to GNP have dropped stead- have dropped each year since 1974 and now ily since 1973. average an annual increase of 2.9 percent from The sources of uncertainty in electricity de- 1983 to 1992 (70). Some studies (e.g., the Energy mand forecasts are very evident from a look at Information Agency and Starr and Searl of EPRI) project higher rates of electricity growth than the electric utilities do, although none project more Figure 5.—A Comparison of the Growth Rates of Real than 4 percent annual growth through 1990 (27, GNP and- Electricity Sales, 1980=82 51 ,83). Only one (Edison Electric Institute) projects more than 5 percent from 1990 to 2000. Sev- eral studies (e.g., Audubon and the Solar Energy Research Institute) on the other hand, project very low rates of annual electricity growth of O to 1.5 percent (77,84). One of the reasons for this range is different assumptions about the future growth rates in GNP. The projections of faster electricity growth assume a range of 2.5 to 3.0 percent annual GNP growth per year (51). The projections of slower electricity growth assume somewhat slower growth rate in real GNP, a range of 2.0 to 2.8 1 1 1 1 percent per year (77). In general, however, all 1960 62 64 66 68 70 72 74 76 78 80 82 these projections assume that the United States SOURCE: Craig R. Johnson, “Why Electric Power Growth Will Not Resume, ” Public Utilities Fortnigfhtly, Apr. 14, 1983 from data in the EIA, Annual has a “mature” economy and increases in real Report to Congress 1982 and the Edison Electric Institute, Ch. 3—The Uncertain Financial and Economic Future ● 33

the uncertainty surrounding the factors under- Regional uncertainties about economic growth lying the forecasts. The uncertainty exists both are more extreme than national ones. Income in conventional macroeconomic approaches to and industrial output have fallen in some regions forecasting which relate electricity growth to ex- as a result of the recent recession and the extent pected changes in GNP, electricity prices and the of long-term recovery from the recession in these prices of competing fuels. (This is sometimes re- regions is unclear. Rapid population growth is ex- ferred to as the top-down approach.) There is pected to occur in the South and Southwest. comparable uncertainty about the factors underlying engineering or end-use projections of elec- Electricity Prices.– Future electricity prices and tricity use. This approach (sometimes called bot- their impacts are a second source of uncertainty tom-up analysis) identifies possible technical about electricity demand growth. This is both be- changes in the use of electricity that are econom- cause there is disagreement about future change ically feasible—such as improvements in appli- in electricity prices and because there is uncer- ance and electric motor efficiency, opportunities tainty about how electricity demand responds to for fuel switching, and new electricity-using in- electricity prices. dustrial technologies–and then estimates the likely market penetration of these changes. From 1970 to 1982, average electricity prices increased in constant dollars at about 4 percent The Top-Down Perspective on per year, reversing a 20-year trend of decreas- Electricity Demand—Sources ing real prices (26). There is considerable disagreement about the future course of electricity of Uncertainty prices even though they should be easier to pro- One of the advantages of top-down analysis of ject than oil or gas prices because they are largely electricity demand is that uncertainty is confined determined by regulatory rules that are predict- to only a few powerful variables—future growth able. The Energy Information Administration in in GNP, changes in electricity prices, and changes the Department of Energy (DOE) has consistent- in the prices of competing fuels and the respon- ly projected very slow increases in real electrici- siveness of electricity demand to each. ty prices of less than 0.5 percent per year until 1985 and 1.4 percent per year after that (27). The The Influence of Economic Growth. -Future Office of Policy Planning and Analysis, also of growth of GNP is a major source of uncertainty, DOE, projected somewhat more rapidly increas- both because income and industrial production ing electricity prices, at 2.4 percent per year un- are assumed by economists to have major im- til 1995 with level prices after that (20). Finally, pacts on electricity demand, and because of some Data Resources Inc. (DRI) has projected sharply deep uncertainties about the future direction of increasing electricity prices for both industrial and the economy. Even the fairly narrow range of residential users of 3.7 percent per year until the GNP growth rates of 2 to 3 percent that has been mid-1 980’s, slower increases until 1990 and less assumed by the major electricity demand projec- than 0.5 percent per year from 1990 to 2000 (18). tions implies a range of electricity demand growth rates (assuming no price influence) of about 2 to Forecasts of electricity prices disagree principal- 3 percent over the long run if electricity demand ly about the future cost of coal for electricity, the follows the income response patterns identified future construction cost of nuclear and coal pow- in the past (79). Many observers concede the erplants and the future rate regulation policies range is even wider, Those with private misgiv- of Public Utility Commissions. Stabilizing of elec- ings about the future health of the economy ac- tricity prices in the 1990’s is expected to occur cept the possibility of an annual rate of GNP because of a growing share of partially depreci- growth lower than 2 percent. Optimists about ated plants in the rate base and little new con- economic renewal and increased productivity struction. (See the discussion of these factors in suggest the potential for a higher rate of growth. later sections of this chapter.) 34 ● Nuclear Power in an Age of Uncertainty

Response of Electricity Demand to Electricity somewhat slower increases of about 3 percent Prices.—There is generally less agreement about per year through the 1980’s and 1990’s (18). the impact of electricity prices on electricity de- In some areas, especially New England, oil is mand than there is about the impact of changes the chief competitor to electricity and the chief in GNP. Most analysts agree that the short-run source of uncertainty. Oil prices are now higher response of electricity demand to an increase in than natural gas prices and are projected to in- electricity prices is very limited—l O to 20 percent crease but more slowly than natural gas prices. of the price increase. Based on comparisons from State to State, however, analysts estimate the The Impact of Prices on Demand .–The com- long-run response to be much greater–50 to 100 bined effect of uncertainty about future electric- percent of the price increase. (The response to ity and natural gas (and oil) prices and uncertain- a price increase is always a decrease in demand.) ty about how electricity demand responds to For example, an increase of 2 percent in elec- changes in electricity prices is enough to explain tricity prices would be expected to result in a a range of uncertainty in electricity demand from short-run decrease in electricity demand of 0.2 very slow growth to quite rapid growth. This is to 0.4 percent (from what it would have been illustrated in table 3. If GNP is projected to grow otherwise), but a long-run decrease in electrici- at 2.5 percent (the midpoint of the range assumed ty demand of 1 to 2 percent. in current forecasts) and the long-run response to increases in income is assumed to be 100 per- Prices of Competing Fuels.–Very few analysts cent, the effect of price and price response as- have attempted to estimate the long-run response sumptions is to produce a projection of 1.1 per- of electricity demand to changes in the prices of cent annual increase in electricity demand, at the other (and competing) forms of energy of which low end, and of 4 percent at the high end. It the principal competitor is natural gas. Of these would be possible, for example, for electricity de- attempts, the consensus is that electricity demand mand to grow at 4 percent per year, if electricity would be expected to increase (over the long run) prices increase at no more than 1 percent per about 0.2 percent for every 1 percent increase year (in constant dollars) while natural gas prices in natural gas prices. increase at 10 percent per year, and there is a relatively small long-run decrease in demand in The Energy Information Administration projects response to an electricity price increase (defined that natural gas prices will increase at more than as a long-run elasticity of – 0.5). 10 percent a year (in constant dollars) until 1985 and more slowly, at 5 percent per year after that Timing of Response to Prices.–Unfortunately until 1990 (27). DRI, on the other hand, projects no analyses have been published of the long-run

Table 3.—Growth Rates in Electricity Demand Given Different Price Responsesa Annual increase in electricity demand given: Rates of electricity and gas Low long-run High long-run price increase price response price response 1. High electricity prices (2%/yr) Moderate gas prices (3%/yr) ...... 2.1%/yr 1.1%/yr 2. Low electricity prices (1‘\O/yr) Moderate gas prices (3%/yr) ...... 2.6%/yr 2.1%/yr 3. Low electricity prices (1%/yr) High gas prices (5%/yr)...... 3.0%/yr 2.5%/yr 4. Very high gas prices (10%/yr) Low electricity prices (1%/yr)...... 4.0%/yr 3.5%/yr aAl~O ~alled price elaStiClt&. See Assumptions.

Assurrrptlons: GNP increases at 2.5 percent par year; Income elasticity of electricity demand - 1.0; response of electricity demand to gas price (cross-elasticity) - 0.2; response of electricity demand to electricity price (own-elasticity): a) low = 0.5, b) high = –1.0.

SOURCE: Off Ice of Technology Assessment based on a presentation by James Sweeney to an OTA workshop. Ch. 3—The Uncertain Financial and Economic Future ● 35 response of electricity demand to electricity from the national average. A recent regional anal- prices in the crucial decade since the 1973 oil ysis of projected changes in electricity prices embargo. The estimates mentioned above were shows a mixture of declining electricity prices in all based on data up to 1972. The chief reason some regions and increasing electricity prices in is because this analysis cannot be done without others (48). Real electricity prices in the Moun- consideration of the timing of the long-run retain region are forecast to drop by an average of sponse. Not only have electricity prices (in con- more than 3 percent per year (in constant dollars) stant dollars) changed from decreasing to increas- until 1987 and then stay nearly stable until 2000. ing, but competing fuel prices (which also affect Meanwhile, in the West South Central region electricity demand) changed from decreasing to electricity prices are projected to increase by an increasing even more dramatically. average of 4.6 percent per year until 1991 and then taper off slowly until the year 2000. Price The length of time it takes for the long-run price changes as different as these will inevitably in- response to be felt is crucial to making any esti- duce a wide regional variation in demand growth mate of this response. If the “long run” is 3 to rates. Declining rates in the Mountain region 4 years, we have already seen much of the re- should eventually stimulate increases in electricity sponse to the price increases of the 1970’s. If the demand while the opposite occurs in the West “long run” is 10 years or longer, we are just now South Central region. This regional variation in beginning to witness the effects of actions taken both present and projected electricity prices will in response to those price increases. complicate and perhaps delay industry’s invest- This lack of understanding of how long it takes ments to improve efficiency. for the full long-term response to increases in electricity prices makes it difficult to predict what The Bottom-Up Perspective on still remains of consumer and industrial responses Electricity Demand—Sources to the increasing electricity prices of the last of Uncertainty decade. If the response takes 10 years or longer, the effects will last until the early 1990’s. Within an overall framework of economic Electricity Rate Structure.-The uncertainty of growth rates and changes in relative energy price impacts is further complicated because fore- prices, bottom-up or end-use analysis offers a casts of average electricity price do not fully cap- closer look at how electricity customers might ac- ture the potential price changes that will influ- tually change their patterns of electricity use in ence electricity demand. Decisions of industry response to prices and income changes. Industrial and consumers are also influenced by the price customers purchase the most electricity, about of an additional unit of electricity, that is the mar- 38 percent of the approximately 2.1 billion kwh ginal price of electricity. Utilities have recently sold in 1981. * Residential customers are close begun to shift from “declining block rate” struc- behind with 34 percent of all sales in 1981. Com- tures (in which each additional block of units of mercial customers and other customers pur- electricity costs less than previous blocks) to in- chased 24 and 4 percent, respectively. creasing block rate structures (in which each ad- Given a range of plausible assumptions about ditional block costs more than previous blocks). how customers are likely to change their patterns There is no current survey of data on utility rate of electricity use over the next two decades, structures, but a crude estimate can be made that growth rates in electricity demand that range from as many as one-fifth of all utilities may have in- 1 percent per year to as high as 4 percent per creasing block rates for households. The number year are possible. of such utilities is likely to continue to grow. From a close look at each sector it is clear Regional Differences in Demand Response to that a few variables are far more important than Price Increases for Individual Utilities. -Another source of uncertainty is that individual utilities will * 1981 data are used because industrial purchases had fallen to have very different experience in electricity prices 35 percent of the total in 1982 as a result of the economic recession. 36 ● Nuclear Power in an Age of Uncertainty others. Industrial electricity sales will be strong- Industrial Demand for Electricity .-The elec- ly influenced by the output experience of several tric-intensity of U.S. industry (electricity used per key industries such as steel and aluminum, by the unit of output) has held steady since 1974. This market penetration of greater efficiency in the use is despite a steady decrease for more than two of electric motors, and by the success of several decades in the overall use of energy per unit of important new electrotechnologies for replacing industrial output. The steady relationship of elec- oil and natural gas sources of industrial process tricity to industrial output is the product of two heat. opposing trends: a steady increase in electricity- use in industries which are heavy users of elec- The future rate of household formation will also tricity, and a steady decrease in the proportion strongly influence residential sales. How fast the of output from those same industries relative to use of electric heat and central air-conditioning total U.S. industrial output (58). Looking ahead, spreads into new and existing housing units is also there is considerable uncertainty about both important, as is the success of high-efficiency ap- these trends. pliances, air-conditioners, and heat pumps. Electricity use in industry is concentrated. Just For future commercial sales of electricity the 13 specific industrial sectors* consume half of all important future influences will be: the rate of industrial electricity. These are: primary alumi- construction of new commercial buildings; the num, blast furnaces, industrial inorganic and or- prospects for significantly more efficient air-con- ganic chemicals not elsewhere classified, petro- ditioning and lighting; and the potential for signifi- leum refining, papermills, miscellaneous plastic cant increase in electricity used for computers products, industrial gases, plastics materials and and other automation. resins, paperboard mills, motor vehicle parts, al- Utilities, in fact, are beginning to monitor such kalis and chlorine, and hydraulic cement. variables more closely in the effort to reduce Future rates of output growth in several of these some of the uncertainty about future growth rates industries are highly uncertain. For example, the in electricity. Utilities are increasingly turning to industrial output of primary metals, which in- end-use modeling of future demand. This is in cludes both aluminum and steel, decreased by part because such models can be used to assess about 1.0 percent per year between 1972-74 and the impact of utility conservation and load man- 1978-80 (57). Capacity expansion plans for both agement programs, but it is also possible to moni- steel and aluminum are down about two-thirds tor important indicators of future customer be- from their level a decade ago (50). havior. Some utilities and State governments are indeed undertaking load management and con- Within the chemicals industry, electricity is servation programs in order to directly influence used heavily in the production of several basic customer behavior and reduce future uncertainty. chemicals such as oxygen (where it is used for refrigeration), and chlorine (where it is used for There is general acceptance that the high and electrochemical separation). Although the chemi- low end of the bottom-up range is influenced by cals industry as a whole grew by an average of the key variables of price and income (GNP) used 4.5 percent per year from 1974 to 1980, the in top-down analysis. Slow increase in electrici- production of these basic electrically produced ty prices is a weak stimulus to improvements in chemicals grew only 1 to 2 percent per year (ox- the efficiency of electricity use even if they are ygen and chlorine) or actually decreased (acet- technically feasible, while rapid electricity price ylene and phosphorus) (1 1). Observers of the increases are a strong stimulus to efficiency im- chemical industry expect this trend to continue provements. Similarly, rapid increases in natural as demand for basic chemicals becomes saturated gas prices relative to electricity prices will stimu- and some production moves overseas. The U.S. late the adoption of new electrotechnologies that substitute for natural gas. Less rapid price in- *As identified by four-digit Standard Industrial Classification (SIC) creases will provide less stimulus. codes. Ch. 3—The Uncertain Financial and Economic Future ● 37 chemical industry is expected to concentrate in- tion in the United States competitive with alumi- creasingly on small volume specialty chemicals, num production overseas (81). which use relatively little electricity for production. Electric process heating in industry accounts for only about 10 percent of current uses of electrici- Since chemicals, iron and steel and primary ty but has great potential to become much more aluminum alone account for more than a third important as new electric process heating tech- of electricity use in industry, a 2 percent lag in niques are developed that make better use of growth of these electricity-using industries behind electricity’s precision and ability to produce very general industrial output would alone cause a lag high temperatures. In some important high tem- of about 0.5 percent in electricity demand behind perature industries such as cement, iron and overall growth in industrial output. steel, and glassmaking, electricity makes up 20 In addition to being concentrated by type of to 35 percent of all energy use and could as much industry, electricity use in industry is also con- as double its share. centrated by function. Uncertainty about the di- Some techniques being developed could have rection of trends in electricity use in electric- very large impacts on electricity demand. These intensive industries comes about because elec- include plasma reduction and melting processes tricity use will probably become more efficient for primary metals production, and induction in two of the functions (electric motors and elec- heating for shaping and forging. Other techniques trolysis) and is likely to expand into significant such as lasers and robotics, however, are likely new uses in the third function (electricity to supp- to have small impacts on electricity demand be- ly or substitute for process heat). cause only small amounts of electricity are used About half of all electricity use in industry is for each application. The biggest lasers, for ex- used for electric motors, including compressors, ample use only about 1 to 2 kW. Most use under fans and blowers, and pumps (3). Improvements 100 W. Table 4 shows an assessment of the rela- in the efficiency of electric motors are likely to be tive impact likely from each of the newer tech- continuous for 10 to 15 years through improve- niques (81). ments in the motors themselves and through im- Great uncertainty surrounds the contribution proved efficiency of use which takes advantage to industrial electricity demand from the most im- of new semiconductor and control technology. portant new electrotechnologies. The iron and Thus, electricity use per unit of output for these steel industry could experience the greatest in- purposes could decrease by 5 percent (if there crease in electricity demand. Production and is little price stimulus) or up to 20 percent (if there profit levels, however, in that industry are uncer- is significant price stimulus). Some of this im- tain. Overall growth in the steel industry is pro- proved efficiency should come about as a result jected to be only about 1.5 percent per year to of past price increases, as capital stock turns over. 2000 (81 ). The contribution of electric arc steel- impetus for the rest will depend on future elec- making from scrap iron is expected to increase tricity price increases, and the cost of installing from approximately one-fourth to at least one- the control technologies. third of the total, but the potential for plasma Another 15 to 20 percent of all industrial elec- reduction and melting is more uncertain partly tricity is used for electrolysis of aluminum and because there may be too little capital to take ad- chlorine (3,81 ). Aluminum electrolysis is more vantage of technological advances (81). If new likely to decrease than increase as a fraction of technologies penetrate slowly, the impact on industrial use, because efficiency improvements electricity demand could be minimal until the late of 20 to 30 percent are technically possible from 1990’s. It is conceivable, however, that rapid pen- several technologies and are probably necessary etration could occur with a few very successful (given sharply increasing prices for electricity in new techniques which in turn may increase the the Northwest, Texas, and Louisiana where plants U.S. steel industry’s competitive position and have been located) to keep aluminum produc- prospects for growth. In such circumstances,

25-450 0 - 84 - 4 : QL 3 38 . Nuclear Power in an Age of Uncertainty

Table 4.—Estimated Impact of Industrial Electrotechnologies in the Year 2000

Rough estimates of GW of capacity Change Load Level of 1983 2000 to 2010’ factor uncertainty Arc furnace steelmaking...... 3.5-4.5 5-7 + High Low Plasma metals reduction ...... 0 2-4 + High High 3. Plasma chemicals production . . . 0 3-5 + High High 4. High-temperature electrolysis (aluminum and magnesium) . . . . 8-9 9-12 O or – High Low 5. Induction melting 3-4 4-5 o Low and Moderate off peak (casting) ...... b 6. Plasma melting ...... 0.05 ? + High Very high 7. Induction heating (forging) . . . . . 5-7 8-10 o Moderate Moderate 8. Electro slag remelting...... 0.075 0.15 + Moderate Moderate 9. Laser materials processing . . . . . 0.0005-0.001 0.001-0.002 + Moderate Moderate 10. Electron-beam heating ...... 0.006-0.008 0.015-0.02 + Moderate Moderate 11. Resistance heating and melting ...... 0.3-0.5 0.4-0.6 O or – High Low 12. Heat pumps ...... 0.1 1-3 + High Moderate “ 13. Electrochemical synthesis, electrolytic separation, chlorine/caustic soda ...... 4-5 4-5 o High Low 14. Microwave and radiofrequency heating ...... 0.5 7 + Moderate High 15. Ultraviolet/electron-beam coating ...... 0.01-0.02 0.5-1 + Moderate Moderate 16. Infrared drying and curing ...... 0.4-0.5 0.5-1 o Moderate Low 17. Electrically assisted machining and forming...... <0.001 <0.001 — — 18. Low-temperature plasma processing ...... <0.001 <0.001 — — — 19. Laser chemical processing . . . . . 0 <.5GW C — — — 20. Robotics ...... <0.001 <0.001 — —

aExplanation. “ + “ = Increase to 2010; 4’0” = stable to 2010; “ – “ = decrease to 2010, bAny capacity in plasma melting directly competes with electric arc furnace steelmaking. cThis will probably only be used in U.S. Government-owned f’CilltleS for Uranium isotope Separation. SOUR C E: Table prepared by Philip Schmidt, based on research for a repori publicized by EPRI in early 19S4, Electricity and Industrkl Productivity: A Technica/ and Economic Perspective (81).

there could be a substantial impact on electrici- in cogeneration of this magnitude would reduce ty demand by the early 1990’s. the growth rate of purchased electricity by about 0.5 percent per year. Cogeneration is being The role of cogeneration* in satisfying future adapted more slowly than the market potential industrial electricity demand is a final source of indicated earlier partly because the success of in- uncertainty. (Cogeneration is discussed further in dustrial energy conservation programs has re- the section on utility strategies and was a sub- duced industrial requirements for steam. ject of a recent OTA report Industrial and Com- mercial Cogeneration (72).) A recent study by For the 1980’s, the highest growth rate in indus- DOE estimated that cost-effective opportunities trial electricity demand is likely to be no more existed in industry for about 20 GW of self-con- than equal to industrial output, and growth rates tained cogeneration without sales to the outside could fall as low as 2 percentage points below electrical grid, a potential increase of slightly more industrial output growth. For the 1990’s, it is quite than the current installed capacity of about 14 possible that electricity demand could grow faster GW with about 3 GW additional capacity in the than industrial output (20). The reasons for possi- planning stage (72). if fully realized, an increase ble faster growth in industrial electricity demand in the 1990’s than in the 1980’s should be clear *The combined production of electricity and useful steam (or hot from the above discussion. The output of such water) in one process. electricity-using industries as steel and aluminum Ch. 3—The Uncertain Financial and Economic Future ● 39 is likely to continue to grow very slowly during Figure 6.—Penetration of Air-Conditioning and Electric the 1980’s, but may grow more rapidly in the Heating in Residential Sector 1990’s. Technologies for improving the efficiency of use of electric motors are also likely to have the greatest impact in the 1980’s, while new electrotechnologies for substituting electricity for process heat are not likely to have a big impact in the 1980’s because they are now a small share of total electricity use. Even if they grow rapidly, they cannot have a major impact on overall industrial electricity demand until they are a larger fraction of the whole, sometime in the 1990’s. Residential Demand for Electricity .-From a bottom-up perspective, what happens to future electricity demand from households depends both on how fast the total number of households increases and on what happens to electricity use per household. There is uncertainty, first of all, about the rate of household formation. Over the decade from 1970 to 1980, the U.S. population formed households at a rate much faster than population growth. In current census projections, this trend 1940 1950 1960 1970 1980 is expected to continue through the 1980’s, re- Percentage of new Percentage of new sulting in a fairly rapid rate of household forma- r households with air households with conditioning central air tion of 2.2 percent per year and a further drop conditioning in household size from about 3.2 people per household in 1970 to 2.8 people in 1980 to 2.5 Percentage of new Percentage of new households with I households with people per household in 1990. On the other electricity for heat pumps hand, were the U.S. taste for living in smaller and primary heating smaller households to become less important, the Percentage of total - - - Percentage of total households with households with growth rate in household formation could fall to air conditioning electricity for 1 percent per year or less. primary heating

SOURCE: Department of Energy, The future of Electric Power in America. The potential for increased use of electricity per Economic Supply for Economic Growth, Report of the Electricity household largely comes about because the num- Policy Project, June 1983, This graph was prepared from data in the EIA residential energy consumption surveys ber of households with air-conditioning and electric heating is still increasing. This is shown in which will in turn be reduced by increases in figure 6. As of 1979 only 16 percent of all house- electric heating efficiency. About 70 percent of holds had electric heat, but about half of new new households have air-conditioning compared dwelling units were heated electrically. As new to about 55 percent of existing households, so dwelling units replace existing ones, the percent modest increases in electricity use from air- of total dwelling units with electric heat could conditioning are also likely. double or triple. A doubling of the share of all households heated electrically (assuming no in- The use of electricity to heat water may expand crease in the efficiency of space heat) would add beyond the 30 percent of households that now about 0.7 percent per year to household growth use it and could as much as double if there is a in electricity demand. The potential for increased big decrease in the relative cost of electric and use of electric space heating, however, will be gas-heated hot water. The demand for other elec- influenced by the relative cost of electric heat tric appliances is considered largely saturated and 40 . Nuclear Power in an Age of Uncertainty

unlikely to expand substantially beyond the de- ment in appliance efficiency will occur. With mand caused by increases in new households. modest increases in electricity prices, refrigerators Most (98 percent) of all households have refriger- are expected to reduce their average per unit ators, 45 percent have freezers, and sO percent electricity use by 10 percent. With stronger price have electric ranges (10). incentives and perhaps utility market induce- ments electricity for refrigerator use may decrease What makes projecting household electricity by so percent both because of greater efficien- use highly uncertain is the difficulty of knowing cy and smaller volume refrigerated. There is a how much electricity use per household will be similar range of possibilities for other appliances. reduced because of increases in appliance and lighting efficiency. From a comparison of the most Commercial Demand for Electricity .-The energy-efficient appliances and lighting available commercial sector is the smallest of the three but in 1982 with the more typical appliances and is growing the fastest in electricity use. Sales of lighting available (table s), it is clear that efficien- electricity to the commercial sector increased 3.9 cy increases of more than so percent can be percent per year over the decade from 1972 to achieved for all types of appliances, except 1982, faster than sales to either the industrial or freezers for which efficiency improvements since residential sector, although less than half the rate 1973 have already increased almost 50 percent of increase in commercial-sector sales of the pre- (42). Some of these highly efficient appliances vious decade (26). cost up to 100 percent more than the typical ap- One of the reasons for uncertainty in project- pliance, but the extra cost would normally be ing future commercial demand for electricity is paid back in electricity savings in 4 to 8 years. that there is no reliable source of data on how Continued increases in electricity prices will in- fast the commercial building stock is increasing. crease the demand for these high-efficiency prod- From the one available source (a 1979 survey of ucts. Since some regions will have much larger nonresidential building energy consumption electricity price increases than others, these may (33)), it appears from the number of recently built well create a large enough market to bring the buildings that commercial building square foot- prices of the high-efficiency products down. In age increased by about 2.7 percent per year some regions, market incentives will be aug- from 1974 to 1979, somewhat slower than GNP mented by local utility programs: rebates to pur- growth of 3.8 percent per year over the same pe- chasers of energy-efficient appliances (e.g., Gulf riod. Over the same period, electricity sales in- Power Co. in Florida) or rebates and bonuses to creased about 4.2 percent per year, a rate faster dealers who sell them (e.g., Georgia Power Co.) than GNP and much faster than the increase in (42). Most observers agree that some improve- building square footage. If the same trends con-

Table 5.-Efficiency Improvement Potential From Typical to Best 1982 Model: Household Appliances

Percent Typical Most efficient increase in 1982 model 1982 model efficiency. Heat pump: C. O.P.a ...... 1.7 2.6 + 53 Electric hot water heater: C. O.P.a ...... 0.78 2.2 + 182 Room air-conditioner: EERb ...... 7.0 11.0 + 57 Central air-conditioner: SEERC ...... 7.6 14.0 +84 No-frost refrigerator-freezer: energy factord ...... 5.6 8.7 +55 Chest freezer: energy factord ...... 10.8 13.5 +25 Bulb producing 1,700 lumens: efficacy (lumens/watt)e. . . 17 40 + 135 ac,o,p, ,s the coefficient of performance, kWh of thermal output divided by kWh of electrical inPUt. bEER ls the energy efficient ratio obtained by diwding Btu/hr of cooling Power by watts Of dedrical POwer InPut. CSEER IS a seasonal energy-efficient ratio standardized In a DOE test procedure. dEnergy factor ,s the corrected volume divided by dally electricity consumption, where corrected volume IS the refrigerated space PIUS 1.63 times the freezer sPace for refrigerator/freezers and 1.73 times the freezer space for freezers. el ,7~ lumens iS the Output of a l(lfl-watt incandescent bulb. SOURCE: Derived from Howard S. Geller, Effmenl Res/derrtm/ Appllarrces ” Overwew of Performance afrd POIICY Issues, American Council for an Energy Efflctent Economy, July 1983. Ch. 3—The Uncertain Financial and Economic Future • 41 tinue, and commercial square footage continues and therefore, considerable uncertainty remains to grow more slowly than GNP, commercial elec- about the potential. tricity use will only increase as fast or faster than The results of the bottom-up analysis for the GNP as long as electricity use per square foot commercial sector indicate a demand for electric- continues to increase. ity that will increase for a few years at about the Electricity use per square foot in commercial rate of increase of GNP but with wide ranges of buildings may continue to increase for several uncertainty. It could be higher as a result of faster reasons. Only 24 percent of existing commercial penetration of electric space heating and more building square footage but almost half (48 per- air-conditioning and big loads from office ma- cent) of the new building square footage is elec- chines, or lower as a result of big improvements trically heated (33). If these trends continue the in the efficiency of commercial building electric- share of buildings that are electrically heated ity use. By the 1990’s, however, when the de- could double. mand for electric heat in commercial buildings will be largely saturated, the trend rate of growth Air-conditioning in commercial buildings is in electricity demand is likely to settle to the probably saturated. About 80 percent of all build- growth in commercial square footage, somewhat ings have some air-conditioning. Small increases less than the growth rate of GNP. in electricity use per square foot can come about by air-conditioning more of the building and by replacing window air-conditioners with central Conclusion or package air-conditioners. Window air-condi- Utility executives contemplating the construc- tioners cool 20 percent of the existing building tion of long Ieadtime powerplants must contend stock, but only 9 percent of the newest buildings with considerable uncertainty about the probable (33). future growth rates in electricity demand, The Greater use of office machines and automation range of possible growth rates encompasses low might increase electricity use both to power the average growth rates of 1 or 2 percent per year, machines and to cool them in office buildings, which would justify very few new large power- stores, hospitals, and schools. Machines, how- plants, up to fairly high growth rates of 3.5 to 4 ever, are less likely in churches, hotels, and other percent per year, for which the pressure to build categories of commercial buildings. several hundred gigawatts of new large powerplants is great, as will be clear from the discus- The potential for increased efficiency of elec- sion in the next section. tricity use in commercial buildings is less well known than for residential buildings because The sources of uncertainty are many. Future commercial buildings are very diverse and the trends in electricity prices are viewed differently potential for increased efficiency depends part- by different forecasters, because there is disagree- ly on success in balancing and integrating the ment about future capital costs of generating ca- various energy loads: lighting, cooling, heating, pacity, future rates of return to capital and future refrigeration, and machines. OTA analyzed the prices of coal and natural gas for electricity. There theoretical potential for reductions in electricity also is uncertainty about how consumers and in- and fuel use in commercial buildings in a recent dustry will respond to higher prices, given many report, Energy Efficiency of Buildings in Cities (71), technical opportunities for improved efficiency and found that electricity use for lighting and air- in appliance use and industrial electricity use and conditioning in commercial buildings can be re- increasing numbers of promising new electro- duced by a third to a half. Heating requirements technologies that could substitute for the use of also can be reduced substantially by recycling oil and natural gas for industrial process heat. Sev- heat generated by lighting, people, and office eral of the industries, such as iron and steel, how- machines from the building core to the periphery. ever, where the new electrotechnologies could There is still very little verified documentation of have the greatest impact, face an uncertain fu- energy savings in commercial buildings, however, ture. 42 ● Nuclear Power in an Age of Uncertainty

At present, utility strategy for supplying ade- ficulties and regulatory disincentives for large cap- quate electricity is influenced heavily both by the ital projects. The influence of utility rate regula- recognition of uncertainty about future growth tion and current utility strategies are discussed in electricity demand and by recent financial dif- Iater in the chapter.

RESERVE MARGINS, RETIREMENTS, AND THE NEED FOR NEW PLANTS

Powerplants are planned and ordered many The electric utility industry currently projects years in advance of when they are needed. in average growth in peak summer demand of 3.0 order to produce power for sale by 2000, nuclear percent per year between 1982 and 1991 (68). * powerplants on a typical schedule would have The industry has also planned for an increase in to be ordered by 1988, or 1990 at the latest, and electric-generating resources of 158 GW by 1991 coal plants or nuclear plants on an accelerated bringing the total generating capability up to 740 schedule, must be ordered by 1992 or 1993. GW (68). Only 13 GW of scheduled retirements Sources of generating capacity with shorter lead- have been included in the 1991 estimate (69). At times, such as gas turbines or coal conversion, the same time the current reserve margin** of need not be ordered before 1996 or 1997. 33 percent is forecast to fall to 20 percent. As a rule, utilities like to maintain a reserve margin There is considerable disagreement about how of 20 percent to allow for scheduled maintenance many new powerplants will be needed by 2000. and repair and unscheduled outages. Individual Those who believe that large numbers of new regions may require higher reserve margins if they powerplants will be needed (several hundred are poorly connected to other regions, if they are GWs) anticipate rapid growth of electricity de- dependent on a small number of very large plants mand (3 or 4 percent per year), expect that large or if they are dependent for a large share of gen- numbers of existing plants will be replaced erating capacity on older plants or plants that because of deterioration in performance or retire- burn expensive oil and natural gas. ment due to age and economic obsolescence, and expect only modest contributions from small As shown in figure 7, the planned resources of power production (20,83). 740 GW scheduled for 1991 would allow a reserve margin of 20 percent to be maintained un- On the other hand, some believe that no new til 1996 if electricity peak demand grows only at powerplants or very few (a few dozen GW) will 2 percent per year, or until 2000 if electricity de- be needed before 2000 because they anticipate mand grows only at 1 percent per year. On the only slowly growing electricity demand (1 or 2 other hand, the average reserve margin will fall percent per year), expect little or no need to below 20 percent by 1987 if electricity demand replace existing generating capacity, and expect increases at 4 percent per year. substantial contributions to generating capacity from small sources of power such as cogenera- However, the number of new powerplants that tion, geothermal, and small-scale hydropower must be built to maintain a given reserve margin (83). does not depend only on the rate of increase in This section lays out the range of possibilities from no new powerplants to several hundred *This had dropped to 2.9 percent per year for 1983 to 1992 in the 1983 North American Electric Reliability Council Forecast of gigawatts of new powerplants arising out of dif- Electric Power Supply and Demand (70). ferent combinations of growth in electric demand **’’ Reserve margin” is defined as the percent excess of “planned resources” over “peak demand” where “planned resources” in- and varying utility decisions about powerplant cludes: 1 ) installed generating capacity, plus 2) scheduled power retirements and use of small power sources. purchases less sales. Ch. 3—The Uncertain Financial and Economic Future • 43

Figure 7.—Projected Generating Capacity and Alternative Projections of Peak Demand Load growth Load growth Load growth Load growth Load growth of 1%/yr of 2%/yr of 3%/yr of 3.5%/yr of 4%/yr —. ● ...... ------— --—

1,400

1,200

1,000

800

600

400

200

0 1980 1985 1990 1995 2000 2005 2010 Year

NOTE. “Planned resources” IS defined by NERC to include: 1) Installed generating capacity, existing, under construction or in various stages of planning; 2) plus scheduled capacity purchases less capacity sales; 3) less total generating capacity out of service in deactivated shutdown status. “Reserve margin” given here IS the percent excess of “planned resources” over “projected peak demand. ” SOURCE” North American Electric Reliability Council, Electric Power Supply and Demand 1982-1997, August 1982 and Off Ice of Technology Assessment electric peak demand. It also depends on how 40 years old, more than a quarter of the total much existing generating capacity must be re- generating capacity that was in service 40 years placed because powerplants are retired, due to ago. age or to economic obsolescence, or because ex- In fact, the bulk of the current generating ca- isting powerplants must be derated to lower elec- pacity of the United States is comparatively new. tricity outputs. Over half has been built since 1970, as shown Retirements Due to Age.–The “book lifetime” in figure 8. The number of plants that would be of a powerplant, used for accounting purposes, retired by 2000 varies greatly with the assumed is usually 30 to 40 years. Over this period the plant life. In the unlikely event that a 30-year life plant is gradually depreciated and reduced as a would be used, over 200 GW would be retired recorded asset on the utility’s books until, at the by 2000 (see table 6). A 50-year schedule would end of the period, it has no more book value and retire only 20 GW. produces no return on capital. However, in practice powerplants may continue to operate for 50 Economic Obsolescence. -From 1965 to 1979 years or more. As of 1982 there were about 10 a large number of steam-generating plants using GW of generating capacity that were more than oil or natural gas were built (see fig. 8). They were 44 ● Nuclear Power in an Age of Uncertainty

Figure 8.–The Energy Source and Age of Existing natural gas to utilities is expected to increase Electricity Generating Capacity steadily through the 1980’s as the long-term contracts for natural gas sold at relatively low prices Energy source expire and are replaced by contracts for more ex- Other pensive gas. Nuclear Water As of 1981, there were 152 GW of oil and natu- Gas ral gas steam-generating capacity. Together they Coal totaled 27 percent of all generating capacity but Oil produced only 22 percent of all electricity. As shown in table 7, oil-fired steam plants produced only half as much electricity relative to their share of generating capacity. Natural gas-fired steam pIants, on the other hand, produced a greater share. Even though oil and natural gas will be expensive, plants burning these fuels can be used as Before 1950- 1955- 1960- 1965- 1970- 1975- 1980- part of the reserve margin. Oil and gas are, in fact, 1950 1954 1959 1964 1969 1974 1979 1982 just about 20 percent of two regions, the South- SOURCE: Energy Information Administration. Inventory of Power Plants in the Ufnited States, 1981 Data from the ‘Generating Units Reference File: east (SERC)* and the West (WSCC). The fraction of oil- and gas-generating capacity, however, is much larger than 20 percent in three regions: Table 6.—Possible Needs for New Electric. Texas (ERCOT) about 72 percent, Southwest Generating Capability to Replace Retired Powerplants, Loss of Powerplant Availability, and Power Pool (SPP) about 56 percent, and the Oil and Gas Steam Powerplants Northeast Power Coordinating Council (NPCC) about 51 percent (68). If oil and gas steam plants Cumulative replacement capacity were retired continuously in these regions until (GW) needed by: they formed no more than 20 percent of total 1995 2000 2005 2010 generating capacity, the total retired would be If existing powerplants about 55 GW. are retired after: 30 years...... 155 230 395 510 Loss of Availability of Generating Capacity.– 40 years...... 55 105 155 230 The percent of time that nuclear and fossil base- 50 years...... — 20 55 105 Ioad plants were available to generate electrici- If all 011 and gas steam capability is retired as ty averaged around 70 percent** over the decade follows: of the 1970’s (67). If there were a reduction from All ...... 152 152 152 152 70 to 65 percent in the average availability of Half ...... 76 76 76 76 All oil and gas capacity nuclear and coal powerplants this would be the above 20 percent of region equivalent of a loss of 21 GW out of a total cur- (3 regions) ...... 55 55 55 55 rent coal and nuclear-generating capacity of 294 If average coal and GW (see table 6). nuclear availability slips from 700/0 to: A recent study for DOE assesses the prospects About 65% ...... 21 21 21 21 for changes in average availability (82). Statistical About 600/0 ...... 42 42 42 42

SOURCE: Office of Technology Assessment. *These are the regions of the Northeast Electric Reliability Council (NERC). ordered before the 1973 oil embargo. Under cur- **The availability figure used here is equivalent availability and rent price forecasts, oil prices are expected to re- includes service hours plus reserve hours less equivalent hours of main fairly stable until the late 1980’s or early partial outages. (66) From 1971-80, nuclear pIants and coal plants over 575 MW averaged 67.8 percent in equivalent availability and 1990’s and then increase substantially (by about coal plants from 200-574 MW averaged 74.3 percent in equivalent 60 percent) from 1990 to 2000 (27). The price of availability. Ch. 3—The Uncertain Financial and Economic Future • 45

Table 7.—installed Capacity and Net Electricity Generation by Type of Generating Capacity, 1981-82 Net Installed electrical generating energy capabiIity, Percent generation, Percent summer of 1981-82 of 1981 (GW) total (billion kWh) total Steam — coal ...... 243 42 1,177 52 Steam — oil ...... 89 16 190 8 Steam — gas ...... 63 11 320 14 Nuclear power ...... 51 9 274 12 Hydro electric ...... 66 12 261 12 Combustion turbine — oil ...... 34 6 3 — Combustion turbine — gas...... 6 1 7 — Combined cycle oil ...... 3 — 2 . Other ...... 17 3 26 1 Total ...... 572 100 2,260 100

SOURCE North American Electric Reliability Council, Electric Power Supply and Demand 1982-1991, August 1982. evidence from the past two decades would sup- Summary-The Need for New Powerplants.– port an estimate of a loss of 3 to 5 percentage the need for new powerplants depends on both points of average availability for every 5-year in- the growth rates in electricity demand and on the crease in average age of coal plants. Looking need for replacement of existing generating ca- ahead, there could also be losses in availability pacity. Table 8 summarizes most of the range of of several percentage points due to emission con- disagreement and its implications for new power- trols and requirements for low sulfur coal that is plants. Estimates of growth in electricity demand at the same time of lower combustion quality. range from 1 to 4 percent per year. (The table shows the implications of electricity demand Offsetting these tendencies to reduced availa- growth rates of 1.5 to 3.5 percent.) bility, however, there are also forces that might increase average availability. The utility industry judgments about replacement of existing plants has completed a period of construction of coal can, somewhat arbitrarily, be divided into high, plants with poor availability, and the newest medium, and low replacement. A high-level re- plants (from the late 1970’s and early 1980’s) placement of about 200 GW by 2000 would be should have substantially higher average availa- necessary to: offset a slippage of about 5 percent- bility. If this were to continue, overall availabili- age points in availability, meet a schedule of ty could increase. If utilities invest in higher avail- 40-year life expectancy for all powerplants, and abilities (e. g., by converting forced draft boilers retire about half the oil and gas capacity in this to balanced draft), this will also increase availabili- country (see table 6). A low-level replacement of ty (82). It is clear that attention to fuel quality and 50 GW would meet a 50-year schedule, retire a good management also can raise availabilities. little oil and gas capacity and would assume no Some Public Service Commissions (e.g., Michi- slippage or an actual increase in average availa- gan) are including incentives to improve availability. bilities in utilities’ rate of return formulas. If these alternative replacement assumptions On balance, it is unlikely that availability will are combined with alternative growth rate as- increase or decrease dramatically. If a change in sumptions (table 8), they lead to a wide range of availability should occur, however, it would have needs for new plants. About 454 GW of new ca- a noticeable impact on the need for new capaci- pacity would be needed, for example, by 2000 ty. A 10-percentage-point change could imply an (beyond NERC’s planned resources for 1991) to increased (or reduced) need for powerplants of meet a 3.5 percent per year increase in peak de- more than 40 GW by 2000. mand for electricity and the high replacement re- 46 . Nuclear Power in an Age of Uncertainty

Table 8.—Numbers of 1,000 MW Powerplants quirements, as is pointed out in recent work at Needed in the Year 2000 the Electric Power Research Institute (83). On the (beyond current utility plans for 1991) other hand, a low replacement requirement com- Levels of replacement Electricity demand growth bined with only 1.S percent demand growth per of existing plants 1.5%/yr 2.5%/yr 3.5%/yr year would require almost no new capacity. Low: 50 GW; Replace all plants over 50 years old . . 9 144 303 This then is the dilemma for utility strategists. Moderate: 125 GW; Replace A shift of only 2 percentage points in demand all plants over 40 years old; plus 20 GW of oil and growth combined with a more stretched out gas capacity ...... 84 219 379 retirement schedule can reduce the requirement High: 200 GW; Replace for new powerplants from hundreds of gigawatts plants over 40 years plus 95 GW of oil and gas (a number requiring a capital outlay of $0.5 capacity ...... 159 294 454 trillion to $1 trillion 1982 dollars) to almost NOTES: 1. Planned generating capacity for 1991 is 740 GW, 158 GW more than nothing. Some of the factors affecting utility 1982 generating sources of 582 GW. Starting point for demand calculations is 1982 summer peak demand of 428 GW. choice of strategy, given this situation, are dis- 2. The calculations assume a 20- percent reserve margin, excess of plann- cussed in the next sections. ed generating resources over peak demand. SOURCE. Office of Technology Assessment.

RATE REGULATION AND POWERPLANT FINANCE

Although the national average reserve margin of utility assets became tied up in construction will not dip below safe limits until well into the of new generating capacity (fig. 10), which equaled next decade, individual utilities may consider a quarter of all utility assets as of 1981. At the ordering powerplants before 1990 for any of three same time, even with high rates of inflation the reasons: to replace expensive oil or natural gas nominal return on equity was kept constant. Thus generating capacity, to anticipate growth in their the real value of utilities’ return on equity de- region, or to start a long Ieadtime plant well be- clined sharply (fig: 11). fore it may be needed in the 1990’s. This section As a consequence of high inflation, enormous describes the framework of rate regulation with- amounts of investment and large fractions of in which such a decision is made. The next sec- assets under construction, there was weakening tion explores the broader strategic options for of many indicators of financial health that are utilities. watched closely by investors. The amount of earnings paid out as dividends increased, leav- Utilities’ Current Financial Situation ing less for retained earnings to finance future projects. The ratio of operating income to interest Although the financial situation of utilities is im- on debt—pretax interest coverage—fell to disturb- proving slowly, they are still in a greatly weak- ingly low levels. The cost of capital from issues ened financial condition compared to their situa- of new stock and bond sales rose accordingly. tion in 1970. The series of figures 9 through 14 After 1973, far more utility bonds were down- prepared for the Department of Energy shows the graded than upgraded; the number of utility origins of the financial difficulties of the 1970’s bonds rated only “medium grade” BBB, in- until the relative improvement of 1982 (described creased from 10 to 43 (fig. 12). A lower rating below). usually means that investors require a higher yield Utilities raised enormous amounts of capital in in order to purchase the bond, and institutional external financing in the 8 years from 1973-81, investors may not be willing to purchase the bond more than double the requirements of the tele- at all (45). Similarly, the average market value of phone industry-the next most capital-intensive utility stock fell steadily from its high of about 2½ industry (fig. 9). In the process, more and more times book value in 1965, to a level equal to book Ch. 3—The Uncertain Financial and Economic Future ● 47 value in 1970 to less than book value in 1974 struction (AFUDC). (See box B.) One result is that (fig. 13). When stock sells below book value, utilities have retained less and less of their earn- there are two consequences. The utility must ings. The share of earnings paid to stockholders issue more stock, (and pay out more dividends) as dividends—the dividend payout ratio— to raise each new unit of capital than the value increased from 68 percent in 1970 to 77 percent of each existing unit of capital, and the value of in 1981, For some companies it was above 100 stock for the existing stockholders is diluted. (See percent, which means they paid out more than box A for a discussion of market-to-book ratio and they earned (45). stock dilution. ) In 1982 and 1983, there was some improve- One of the most serious problems for utilities ment in utility financial health. As of December has been a steady squeeze on cash flow. As 1983, market-to-book ratios were up to an aver- shown in fig. 14, almost 50 percent of utilities’ age 0.98, and 51 out of 103 utilities had market- nominal return on equity was paper earnings in to-book ratios of more than 1.0 (59). Although the form of allowance for funds used during con- more bonds are being downgraded than up- 48 . Nuclear Power in an Age of Uncertainty

The History of the Deterioration in the Financial Health of Electric Utilities, 1960=82

Figure 9.—The Electric Utility Industry is Dependent on Externally Generated Funds to a Greater Extent Than Other Large Capital Users 173 Total External Financlng,a 1973-81 90 External Financing’ as a Proportion of Total Capital 80 F Expenditures, 1973-81

36 35 15 14 1

aNet of debt retirements. blncludes e~Ploratlon and prduction by companies Prlfnadb lnvolv~ ‘n ‘efining.

Figure 10.–Since 1967, the Fraction of Utility Assets Tied Up in "Construction Work in Progress" Has Steadily increased Other assets

Other assets (8%)

225 Other assets

150

(6%)

75 .

1967 1973 1981 fconstruction work In progress (cwlP) refers to plant IXIlntl built, but not Yet In sewice.

Figure 11.— Beginning in 1972, the Utilities' Real Return on Equity Has Been Eroded by Inflation (utility Industry annual average return on common equity (end of year))

14% Nominal Roe 12%

10%

40%

–2% 1 1 1 i i i i 1 i 1 1 1 i 1967 1969 1971 1973 1975 1977 1979 1961 eNomlna\ Roe deflated by GNP implicit price deflator (measured at fourth quarter SnnuallY). Ch. 3—The Uncertain Financial and Economic Future • 49

Figure 12.—Since 1973, the Average Utility Stock has Sold Below its Book Valuea

2.21 I

1.39 1.41 1.29 1.09 0.91

II 62 64 66 68 70 72 74 76 78 60 62 cAverage. of high and low value for the Year. ‘August 1962.

Figure 13.—ln 1970, A Majority of Utilities Were Rated AA or Better in 1982, a Majority Are Rated A or Worse

Key 69 I I I 1970 1962

43 41 !

20 10 3 1 I 0 1 0 AAA AA A BBB BB B

%e box A for an explanation of the implications of stock selling below book value.

Figure 14.-Over Time, the Share of AFUDC in Total Earnings Has Increased and the Share of Cash Has Fallen

Utility Industry Annual Average Return on Equity (end of year) 14 r . Nominal Roe

1967 1969 1971 1973 1975 1977 1979 1961 SOURCE: Booz-Allen & Hamilton Inc., The Financial Health of the Electrlc Utility Industry, prepared for the Department of Energy October 1962, using data from Utility Compustat; U.S. Department of Commerce, Survey of Current Business; Edison Electric Institute, Statistical Yearbook of the Utility Industry; Standard and Poor’s Bond Guide; Energy Information Administration, Statistics of Privately Owned Utilities. 50 . Nuclear Power in an Age of Uncertainty

graded, the number of net downgradings has a market-to-book ratio of 1.0. For the first part been reduced sharply. By late 1983, the average of 1982, the return on investment of the electric earned return on common equity for the 100 utilities ranked 14th out of 39 industries, well largest electric utilities was up to 14.1 percent, above the average for such industries as chemi- more than 200 basis points higher than the cals, appliances and paper (41). earned return in 1980 (59). From this point of view, there is no need for further concern about near-term utility solven- Implications of Utilities’ cy. The worst of the utilities’ problems are com- Financial Situation ing to an end, and their financial situation should improve gradually. Public utility commissions There are two ways of looking at the current (PUCs) have responded by increasing the allowed financial situation of the utilities and the incen- rate of return to utilities, and the Federal Govern- tives to build more plants. From one perspective ment has provided additional relief from cash the electric utilities have shared in general eco- flow shortages in the 1981 Economic Recovery nomic problems and have not fared as badly as Tax Act through liberalized depreciation allow- some industries. The market-to-book ratios of all ances. The tax law further mandates that these industrial stocks fell over the 1970’s, although on must be “normalized” (retained by the utility) average the industrial stocks stayed well above rather than “flowed through” to the consumer Ch. 3—The Uncertain Financial and Economic Future Ž 51

in lower electricity rates. Further relief for those had increased greatly, and the unit costs of pro- utilities, with problems selling their stock to ducing electricity had begun to increase. From finance large construction programs, has also 1973 to 1981, the return on equity earned by been available from the 1981 Tax Act benefits for utilities (10.9 percent) fell substantially below the reinvestment of dividends in purchases of new return on equity allowed by State PUCs (1 3.3 per- utility stock. Over the next few years, external cent). Although the allowed return came close financing needs should diminish gradually, as to the return on equity earned by the 400 indus- total construction expenditures decrease slight- trials (14.3 percent), the earned return fell well ly and internal funding–from retained earnings, below. If the return on equity is designed to ex- depreciation and deferred Federal taxes–include AFUDC-deferred paper earnings, it fell far creases. below the return on industrials (see fig. 14). From another perspective, however, there is For utilities to earn a return on equity signifi- still cause for concern. Inflation has brought cantly below that earned by industrial stocks rep- about several distortions in the ratemaking proc- resents a change from past regulatory practice. ess that may need to be corrected before the next The basis for determining rate of return has its round of orders for new generating capacity. legal foundation in two cases–the 1923 Bluefield From this point of view, the issue is not whether Water Works* case and the 1944 Hope Natural utilities need rate relief to keep from going bank- Gas Co. case.** These cases established three rupt, but whether the treatment of capital in rate principles: regulation needs to be adjusted for the impact 1. the utility can charge rates sufficiently high of inflation in order to prevent allocation of utility to maintain its financial integrity; resources away from capital-intensive electrici- 2. the utility’s rates may cover all legitimate ex- ty-generating processes—beyond the point where penses including the cost of capital; and it would be beneficial to the ratepayer. 3. the utility should be able to earn return at It is not enough, from this perspective, for utili- a rate that is comparable to companies of ties to recover their financial strength during the comparable risk. coming decade of slowed construction programs. Although in principle PUCs allowed rates of In the 1990’s, when the utilities need capital again return on equity comparable to companies of for construction, investment advisors will have comparable risk, in practice they failed to adapt models that predict a deterioration in financial rate regulation practices to accommodate infla- health associated with a large construction pro- tion. The practice of using an historical test year gram unless rate regulation can be expected to rather than a future year to determine income give an adequate return on capital right through and expenses is one example. Politically, it often the construction period. was difficult for PUCs to grant full rate increases The concern that electricity rate regulation requested by utilities when inflation caused them needs to be adjusted for the impact of inflation to return year after year. As precedents accumu- can be summed up in several points. The first lated in each State it became harder for individual problem is erosion in the definition of cost of Commissioners to argue for a restoration to the capital. Back in 1962, electricity demand was full Hope/Bluefield definition of cost of growing rapidly. There was relatively little infla- capital. * * * tion, and unit costs of generating electricity were decreasing. Utilities were allowed an average ● Blue field Water Works and Improvement Co. v. West Virginia Public Service Commission, 262 ‘US 679, 692 PUR 1923D il. 11.1 percent return on equity and actually earned **Federal Power Commission v. Hope Natural Gas Co. (1944), slightly more than that—about 11.3 percent— 320 US 591, 60351 PUR NS 193. right in line with the average of Standard & Poor’s ***It is in this context that one Wall Street partici ant in an OTA workshop recommended a high-level commission o rrate regulators, 400 industrial stocks (45). utility executives, and investment advisors to reexamine the Hope/Bluefield principles, determine the problems of implement- By 1973, this situation had changed. Demand ing the principles in times of inflation, and make recommendations for electricity was growing more slowly, inflation that take into account the political realities of a Federal system. 52 . Nuclear Power in an Age of Uncertainty

A second problem, exacerbating the first, is that beginning to say openly (seethe later discussion of “rate shock” which arises from the front-end of utility strategies) that they are attempting to loading of rate requirements for large capital in- minimize capital requirements rather than total vestments. This phenomenon (explained in box revenue requirements, to protect the interests of C) is noticeable at low rates of inflation and is their stockholders to the possible long-term detri- striking at high rates of inflation. Assuming 7 per- ment of the ratepayers. cent inflation, for example, the cost of electricity in constant dollars from Nuclear Plant X shown Possible Changes in Rate Regulation in box C would be 9.5¢/kWh the first year and There have been many specific proposals for only 1.5¢/kWh the 20th year. For large plants utility rate regulation. Some are designed to en- entering the rate base of small utilities, the in- courage conservation, load management, or the crease can be 20 to 30 percent the first year. This rehabilitation of existing powerplants. Others are problem is exacerbated the more AFUDC (see designed to encourage the construction of new box B) is included in the cost of the plant as it powerplants, especially when they are intended enters the rate base. (AFUDC is described further to displace powerplants now burning oil or nat- in the next section on the risks of constructing ural gas. nuclear plants because the impact of AFUDC is largest for plants with the high capital cost and This assessment does not deal with the com- long duration of nuclear plants.) plex subject of rate regulation reform in any detail. However, it is useful to describe briefly some The combination of the very high rate require- of those reform proposals that are specifically in- ments in early years and low rate requirements tended to offset those aspects of rate regulation in subsequent years discourages multidecade that discourage capital-intensive or risky projects. planning of generating capacity. While short-term These reform proposals would be most likely to rate increases must be tolerated to realize long- improve the prospects for more orders of nucle- term reductions in real electricity rates, there are ar powerplants. intense political pressures on public utility commissioners to hold down these short-term rates Construction Work in Progress (CWIP).–The (88). simplest of the proposed changes in rate regulation is to allow a large fraction or all of a utility’s A final cause for distortion in utility rate regula- CWIP in the rate base. CWIP is advocated be- tion is the generally practiced fuel pass-through cause it reduces or eliminates AFUDC. This in which allows utilities to pass changes in fuel turn increases utility cash flow and the quality of prices onto consumers without going back to the earnings and reduces the likelihood of rate shock PUC for an increase in rates. This has been a use- because the rate base of the new plant includes ful device for avoiding damage to utility cash flow little or no AFUDC. from the volatile changes in fuel (oil and natural gas) prices in the 1970’s but it has had the in- One argument for including CWIP in the rate advertent effect of shielding utilities from the ef- base is that electricity rates come closer to reflect- fects of inflation in fuel costs while they have not ing the true cost of incremental electricity de- been shielded from increases in the cost of capi- mand, providing a more accurate incentive for tal. For a utility faced with capital expenditures conservation. Opponents of CWIP in the rate to avoid fuel costs—through rehabilitating a plant, base, however, fear that utilities will lose the in- building a new one, or investing in load manage- centive to keep plant costs down and to finish ment—there is a theoretical incentive to stick with them on time. Furthermore, opponents claim, the fuel-burning plant as long as fuel costs are utilities may return to overbuilding. Many PUCs recovered immediately and capital costs are re- have responded to these concerns by including covered late and not fully. only a portion of CWIP in the rate base (62). Many utilities continue to base their generating Phased-in Rate Requirements. -At least six capacity decisions on what will minimize lifetime States have developed methods of phasing in the costs to ratepayers. However, some utilities are rate requirements for large new nuclear power- Ch. 3—The Uncertain Financial and Economic Future ● 53

Box C.—Utility Accounting and the Origins of “The Money-Saving Rate Increase”1 The conventions of utility accounting have created a dilemma that affects all investment choices between a capital-intensive plant (e.g., a nuclear or baseload coal plant) and a fuel-intensive plant (e.g., a combustion turbine or an oil or gas steam plant}. if two plants (one of each type) have equal Ievelized annual cost, and equal Iifecycle cost, the capital-intensive pIant will cost consumers far more in early years and the fuel-intensive plant will cost far more in later years. This situation causes a dilemma for oil- and gas-using utilities who wish to substitute coal or nuclear plants for their oil or gas plants.When their analysis convinces them that the Iifecycle cost of the coal or nuclear plant will be less, they still must face a “money-saving rate increase” to cover the early years extra cost of the capital-intensive plant. An Example. The dilemma is illustrated by a specific example in table Cl. A nuclear plant called Nuclear Plant X has been constructed for $2 billion (including accumulated AFUDC) and is about to be placed in service to replace an oil plant that produces the same amount of electricity. The oil plant is old and fully depreciated and earns no return to capital. When the nuclear plant goes into service there will be a net fuel saving the first year of $263 million, which equals the value of oil saved less the cost of nuclear fuel for the nuclear plant. At the same time, accounting and ratemaking conventions dictate that the first-year capital charge for the nuclear plant will be $471 million, or $208 million more than the fuel savings. In the fifth year the capital charge has dropped to $364 million, less than the fuel savings for the first time (if the cost of fuel escalates at 9 percent per year), and by the eighth year the cumulative capital charge of the nuclear plant will be less than the cumulative savings resulting from lower fuel costs. In the 15th year, the nuclear plant costs only $268 million in rate requirements and saves $878 million in fuel costs. If fuel costs escalate more slowly (at only 5 percent per year), the results are shown in the right- hand columns in table Cl. Annual capital charges for Nuclear Plant X drop below annual fuel savings during the sixth year and Nuclear Plant X breaks even on a cumulative basis by the 12th year. In both cases, Nuclear Plant X will cost less over the 30-year life of the plant (if a discount rate of 12 percent is used). In the first case (with 9 percent fuel escalation), Nuclear Plant X will cost about $3.1 billion in lifetime discounted rate requirements and will save $5 billion. In the second case (with 5 percent annual fuel cost escalation), Nuclear Plant X will save $3.5 billion. In both cases the electricity ratepayers would be better off with Nuclear Plant X over the long run. However, consumers would be worse off in the short run, because of the high capital charges at the beginning of plant operation, which translate into high electricity rates. Two Other Examples. To take another example of this phenomenon, which is sometimes referred to as “front-end loading” of capital costs, suppose another Nuclear Plant Y, with identical construction cost (in 1982 dollars) as Nuclear Plant X had been placed in the rate base 8 years before, in 1974. By 1982, the capital charge for Plant Y in the eighth year of operation would have diminished so much (using the same schedule of capital charges) that it would cost $0.03/kWh while the first-year capital charge for Nuclear Plant X, put in service in 1982, would be $0.09/kWh. Part of the reason for the difference is that the book value (see explanation below of Plant Y is only $1.1 billion, the equivalent in 1982 dollars to the $2 billion cost of Nuclear Plant X. The rest of the difference is that the capital charge for the eighth year is only 0.15 of the original cost; compared to 0.24 in the first year. In still another example, if Nuclear Plant X is replaced in its 30th year of operation by another identical plant, the first-year capital charge for that plant will& $3.6 billion, more than 20 times the capital charge of $170 million in the 30th year of Nuclear Plant X. Why Utility Accounting Practice Produces This Result. T he main reason for this result is that the value of a plant is carried at original cost (book value) not at replacement cost (market value) on a utility’s books. The annual capital charge used in computing rate requirements has a series of components,

‘The analysis in this box is bed on two ankles by My Hunt Waiter, ‘~rendhgthe Rate be;’ PuLrk Utilities Fbttnightiy, May 13, 1982, and “Avokiingthe h40ney Saving Rate Increase,” Public Uti/ities hrtnighdy, June 24, 1962.

25-450 0 - 84 - 5 : QL 3 54 . Nuclear Power in an Age of Uncertainty

143

2,263

8,779 Ch. 3—The Uncertain Financial and Economic Future . 55

Figure Cl .—Rate Base and Revenue Requirements Under Original Cost Accounting

Current earnings Depreciated rate base x 0.12

Rate base under original cost accounting Straight line depreciation — beginning of each year Original rate base x 0.05

— — General price level Including fuel costs

It--k—

Total revenue requirements = . earnings plus depreciation

—

$56

3 4 5 6 7 8 9 10 12 14 16 18 20 Year

Figure C2. -Rate Base and Revenue Requirements Under Trended Original Cost Accounting

Current earnings Trended depreciated rate base x 0.05 — Depreciation formula that Increases depreciation charges with nominal Interest rate

— Rate base under trended original cost accounting

General price level k Includlng fuel costs

Total revenue requl rements = earnings plus I depredation I l-r - — m iiii 1 34567 9 10 12 14 16 18 20 Year 56 ● Nuclear Power in an Age of Uncertainty

plants which would cause significant early year lower than market rate of return as interest rates rate increases under conventional rate treatments are going up, in return for enjoying a somewhat- (75). Two of these (New York and Illinois) have higher-than-market rate of return as interest rates developed plans for “negative CWIP.” Under are coming back down. these schemes, some CWIP is allowed in the rate It is just this implicit agreement that seems to base for several years before a plant comes on- be missing from today’s rate regulation proce- line and then an equal amount is subtracted from dures. in some cases, the PUC may be willing to the rate base for the first few years after the plant work out a sensible approach to rate determina- comes online. After the first few years the rate tion over the long term, but is blocked by the base returns to what it would have been in the State legislature. The Indiana PUC, for example, absence of any CWIP. introduced a graduated rate increase incor- Pay-As-You-Go for Inflation Schemes.—Re- porating trended original cost principles to bring cently, a series of proposals have been made to Marble Hill, a large nuclear plant, into the rate adapt the sequence of rate requirements for a base of Indiana Public Service. The plan was ex- capital-intensive plant, (e.g., a nuclear plant) plicitly blocked by the State legislature. Eight more explicitly to inflation (54,88). In effect, these States, by vote of the State legislature or by proposals would eliminate the front-end loading referendum, have banned CWIP inclusions in of capital return for capital-intensive plants in time utility rate bases for just the reasons described of inflation and replace the “downward slope” above (41). of annual rate requirements in constant dollars Furthermore, commissioners may lack the time (shown in fig. Cl in box C) with a horizontal or or motivation to grasp the long-term view. Penn- gentle upward slope more like the sequence of sylvania is one of the few States with 1()-year annual rate requirements for an oil plant (shown terms for its appointed commissioners. Many in fig. C2 in box C). States have reduced PUC terms of 6 or longer to Some of these proposals would do this direct- 4 or 5 years. An increasing number of States have ly by using a rate of return net of inflation and elected rather than appointed commissioners. For adjusting the rate base for inflation (this is called most commissioners, electric utility rate cases are “trended original cost” ratemaking in contrast to only a few among hundreds or thousands of cases “original cost” ratemaking), Others would do it from local as well as statewide utilities, that pro- indirectly by deferring certain operating or de- vide water, sewer, telephone, and gas as well as preciation expenditures until later so as to ap- electricity. Often, electric utilities and their con- proximate an upward slope of rate requirements. sumers must take time to educate commissioners about the issues surrounding electricity supply, The Obstacles to a Long-Term Commission demand and rates over the long term. Perspective.– In principle, these latter proposals would all make it easier for PUCs to increase It is interesting to note (see ch. 7) that the the authorized real return on equity because rate United States is the only one of all the major de- increases for new powerplants are less likely. veloped countries with a Federal system in which Increasing the authorized return should retail electricity rates are regulated at the State in turn improve the incentives for constructing level. In many countries, electricity rates are un- new powerplants when it is in the long-term in- regulated. In West Germany, State electric au- terests of the ratepayers. All these proposals, how- thorities set their own rates subject to Federal ap- ever, rely on an implicit agreement between in- proval. In the United States, State regulation leads vestors and PUCs that a particular way of deter- to the result that the cost of utility capital (return mining revenues will be maintained over dec- on equity) varies among the different States from ades. Under Trended Original Cost schemes, utili- 12 to 17 percent, even though the market for cap- ty investors accept a lower rate of return in early ital generally is recognized as national. Because years with the promise that the rate base will be of the strong U.S. Federal tradition, however, any fully adjusted for inflation. Under indexed rate proposals for regional or Federal determination of return schemes, investors accept a somewhat of the cost of capital or other regulation on State Ch. 3—The Uncertain Financial and Economic Future ● 57 regulation must be developed in the context of 14.2 percent for the Midwestern utility. Eight longstanding legal traditions about the Federal years later, however, in 1990, the rates would be regulation of commerce. only 0.4 percent higher than without CWIP for The Impact of Changes in Rate Regulation in the Southeastern utility and would actually be Electricity Prices.–Changes in rate regulation to lower for the Midwestern utility. Although short- increase the return to capital, utility cash flow, Iived, the increase in rates is large enough that and/or quality of earnings, in turn would increase there would be a substantial impact on electrici- electricity rates. How much rates would increase ty demand, spread out, to be sure, over a number is important to know for two reasons. It would of years. help to weigh current consumer interests against Before PUCs can tackle fully the long-term im- future consumer interests. It would help also to plications of possible rate regulatory changes, it identify the likely future course of electricity would be useful to have a more complete under- prices and the resulting impact on electricity de- standing of the impacts on rates and potential demand. Uncertainty about future electricity price mand responses. increases is a key source of uncertainty about how much electricity demand will increase. Conclusion.– Because of the financial deterioration experienced in the 1970’s, utilities do not Most of the attempts to estimate the impact of have the financial reserves that they had in the changes in rate regulation on electricity rates have late 1960’s and must therefore pay more atten- focused on regions (48,85). There appears to be tion to the impact of their future construction pro- minimal impact on average regional electricity grams on their future financial health. Although prices from increasing the return to capital and their finances are improving, utilities are likely including CWIP in the rate base—an increase in again to find themselves in weakened condition average regional electricity prices of 2 to 3 per- similar to that experienced in the 1970’s if they cent. Regional analysis of rate impacts, however, embark on another round of large-scale construc- combine the experiences of quite different util- tion later this century. This is especially true if in- ities. flation increases again and exacerbates the im- For two individual utilities, examined as case pact of AFUDC and the front-end loading of rate studies in a recent analysis, the impacts of rate requirements for such capital-intensive projects reguIation changes wouId be significant but fair- as nuclear plants. Iy short-lived (2). Increasing the average rate of The last section of the chapter discusses utility return in 1982 from 12 to 16 percent, for exam- strategies. One element of choice for both utilities ple, would have caused a 1 -year increase of 3.3 and PUCs is the tradeoff between short-term price percent in the rates of a Southeastern utility and increases from rate regulation policies designed a 6.2-percent increase in rates for a Midwestern to be more attractive to capital and longer term utility. Rates would have stabilized in the follow- price decreases projected to come about from ing years. construction of central station powerplants (in- The impact of CWIP on rates is estimated to cluding nuclear) which are expected to be the be greater but also fairly short-lived. Including lowest cost source of baseload power over their CWIP in rate base in 1982 would have increased lifetimes. rates 5 percent for the Southeastern utility and

THE COST OF BUILDING AND OPERATING NUCLEAR POWERPLANTS

In addition to the bias against capital-intensive major financial reasons to be wary of investing generation caused by current ratemaking prac- in nuclear powerplants. First, the cost of building tice, investors and utility executives cite several a nuclear powerplant has increased rapidly over 58 ● Nuclear Power in an Age of Uncertainty

the past decade. The estimated cost of the aver- Several attempts have been made to document age nuclear plant now being completed is so high and understand the increase in cost of nuclear that the average coal plant, in most cases, would power over the past decade (6,7,1 7,37,55,61 ,76). produce electricity more cheaply over a lifetime The task is difficult because the cost data cited (although in most regions the lowest cost nuclear above cannot be used for comparing plants over plants are still competitive with coal plants). Sec- time. The above estimates are composites of con- ond, the average construction time* of a nuclear struction expenses paid in different years with dif- plant has increased much faster than the average ferent dollar values, referred to as “mixed” dol- Ieadtimes for a coal plant, and this makes it hard- lars. Most also include some interest that has been er for nuclear plants than coal plants to match deferred during construction, capitalized, and demand. Third, since the Three Mile Island ac- added to the total capital cost (see box B in the cident, it has been widely recognized that a ma- previous section on CWIP and AFUDC). The jor accident can disable an entire plant for an ex- amount of interest that is capitalized varies from tended period of time and require more than $1 State to State and interest rates vary from year billion in cleanup costs, as well as other expenses to year. to restart the plant and pay for replacement pow- The increase in costs of nuclear powerplants er. Major disabling accidents at coal plants are through the 1970’s was analyzed in a carefully both less likely and less costly to cleanup and documented study by Charles Komanoff (55). The repair. Finally, the current political climate for costs exclude interest during construction and nuclear (described in ch. 8) is a major source of were adjusted for inflation, permitting com- financial risk. The output of a several billion dollar parisons from year to year. * Figure 15 is a plot investment in a plant could be lost for a year or of the costs per kW (expressed in 1982 dollars) more if regulatory commissions refuse to put it of individual powerplants with construction per- in the rate base, or if a statewide anti-nuclear mits issued from 1967 to 1971. (It is more accu- referendum passes and the plant is shut down. rate to group the different generations of nuclear Nuclear accidents or near-accidents in plants powerplants by start date than by completion owned by other utilities can lead to a new series date. Later completion dates, by definition, will of safety regulations (discussed further in ch. 6) have a disproportionate share of the delayed, and requiring backfits that may cost a sizable fraction therefore probably more expensive, plants.) of the original cost of the plant. For plants with construction permits issued The Rapid Increase in Nuclear around 1967 (and generally completed in 1972- 74), the direct costs in 1982 dollars ranged from Plant Cost $400 to $500/kW. For plants with construction In the early 1970’s, nuclear powerplants were permits issued 3 years later, in 1970-71 (and com- completed for a total cost of about $150 to $300/ pleted in 1976-78) the direct cost in 1982 dollars kW. ** As of 1983, seven nuclear powerplants al- had more than doubled to $900 to $1,300/kW. most complete and ready to come online will cost A comparable analysis by Komanoff of the costs from $1,000 to $3,000/kW, an increase of 550 of plants currently under construction has been to 900 percent. General inflation alone would ac- completed but will not be published until early count only for an increase of 115 percent from 1984 (56). Preliminary results show that the cost 1971 to 1983. Inflation in components of (labor of a typical plant continued to increase, and the and materials) used to build nuclear power- range of cost experiences also has increased since plants*** would account for a further increase the early 1970’s. Figure 16 compares “typical” of about 20 percent. plants completed in 1971 and 1978 (these are

*Defined as follows: for nuc/carp/ants, issue of construction permit to commercial operation; for coal plants, order of boiler to com- *As described in Komanoff’s Powerplant Cost Escalation (55) app. mercial operation. C, a standard pattern of cash payments was assumed for each plant **In “mixed” dollars, see explanation below. and then deflated using the Handy-Whitman index developed to * **As measured by the Handy-Whitman index. See below. make inflation estimates of components used in powerplants. Ch. 3—The Uncertain Financial and Economic Future • 59

Figure 15.–Costs of Nuclear Units With Figure 16.—Total Capital Costs for Nuclear Plants Construction Permits Issued, 1966-71 Completed in 1971, 1978, and 1983=87 (estimated) in (without interest during construction) 1982 Dollars Without Interest During Construction 4,000

● * 3,540 highest

3 plants 3,000 2,770 5 plants

‘ 2,340

2,000 8 plants 1,725 8 plants Typical 1,350 plant 5 plants , ,20 1,000 1,020 3 plants ‘ 840 lowest

0 1967 1968 1969 1970 1971 0 1971 1978 1983-87 Construction permit issue date (by midpoint of year) Date of completion NOTE: Plant costs in mid-1979 dollars were escalated to mid-1962 dollars using the Handy-Whitman index for nuclear plant components (multi- NOTES: “Plant” is defined as a single nuclear site or station with one or more plying by a factor of 1.276). reactors, The plant costs in-mid-1979 dollars from Komanoff’s book for SOURCE: Updated by OTA from data In Charles Komanoff, Power Plant Cost 1971 and 1978 plants were stripped of interest and escalated to mid-1962 Escalation, Komanoff Energy Associates 1961, republished by Van dollars using the Handy-Whitman index for nuclear plant components Nostrand Reinhold, 1962. (multiplying by a factor of 1.276). The costs for the plants to be completed in 1963-87 are based on mid-1963 utility estimates for a group of 32 sites (stations) with a total of 50 reactors and excludes: Marble Hill, Waterford 3, Susquehanna and all Washington Public Power System constructed from a composite of characteristics (WPPS) plants except WPPS 2. associated with average and not high or low SOURCE: Charles Komanoff, Power Plant Cost Escalation, Komanoff Energy costs) with the full range of costs estimated for Associates, 1981, republished by Van Nostrand Reinhold 1982; unpublished analysis from forthcoming report by Charles Komanoff a group of 32 plants under construction for com- and Irving C. Bupp to be published in the winter of 1984, and the pletion in the 1980’s. The cost (in 1982 dollars) Off Ice of Technology Assessment. of a typical plant increased from about $430/kW in 1971 to $1,020/kW in 1978, to a range of $840 $1,150/kW (in 1982 dollars) will cost about 90 to $3,540/kW for plants under construction in percent more than the company’s Zion 1 and 2, 1983. The median plant of this group is expected completed in 1973 and 1974. to cost $1,725/kW and the average cost is some- Until the early 1980’s, nuclear plant costs in- what higher ($1,880/kW) reflecting the wide vari- creased steadily with each generation of plants ation in costs at the upper end of this wide range. (55,61,75). However, Komanoff’s analysis shows The same increase also is evident in pairs of no tendency for plants scheduled for completion plants built by the same company and intended later in the 1980’s to have significantly greater to be identical except for regulatory changes and expected costs than plants being completed in some construction management improvements. 1983-84. In part, this may be due to underestima- The cost of Florida Power & Light’s St. Lucie 2 tion of costs for plants still far from completion, when completed in 1983 ($1,700/kW in 1982 dol- but it also is probable that factors other than time lars) was about 50 percent more than that of St. now are more influential on powerplant cost. (A Lucie 1 completed in 1976. Commonwealth Edi- complete list of the mixed-dollar cost for plants son’s Byron 1 and 2, to be completed in 1984 in various stages of completion is given in app. and 1985 at an estimated cost of $1,100 to table 3A.) 60 ● Nuclear Power in an Age of Uncertainty

The variation from lowest to highest cost nuclear powerplants in the current generation is striking. Construction costs per kilowatt are expected to be over four times higher (in 1982 dollars) for Long Island Lighting Co.’s Shoreham at $3,500/ kW than they are for Duke Power’s McGuire 1 and 2 at $840/kW. For the current generation of pIants, Komanoff found some variables that explain much of this large variation (56). For example, estimated plant cost decreases about 15 percent for every doubling of the number of megawatts at a single site. Estimated plant cost also decreases 8 to 10 percent for each previous plant site built by the same utility. Based on these results, a utility that had built on 5 previous plant sites should be able to construct its next plant site for 30 to 40 percent less than a utility with no experience. Plant cost also varies by manufac- turer (as much as 15 percent) and is significantly higher (30 to 40 percent) for plants located in the Northeast region due primarily to higher labor cost and shorter construction seasons. The importance of utility experience and site- related experience for this latest group of pIants is evidence of the impact of some of the utility managerial experience described in chapter 5. Photo credit: Duke Power Co. There seems to be a site-specific and company- Capital costs per kilowatt of identical plants at a single specific learning-curve for bringing plant costs site are usually lower than average. This photo shows down. Catawba nuclear station, owned by Duke Power Co., which is expected to produce among the lowest cost electricity of any plant in its timeframe when Reasons for Increased it comes online in 1985 Construction Cost for the average of eight plants under construc- Several different kinds of increase contributed tion in 1982-85. Figure 17 shows similar increases to the dramatic increase in average cost described in the use of concrete, piping and cable raceway above. To begin with, these comparative cost esti- (supports for electric cable) (1 7). Increased ma- mates exclude the influence of nuclear-compo- terial requirements are due both to direct in- nent inflation that compares the cost of equal- creases in structural and electrical complexity and quality nuclear components and materials over to the increased rework necessary to meet more time. Nuclear component prices increased 1 or stringent quality-control requirements. 2 percentage points faster than inflation. * Materials also have become more complex. A Several changes account for the increase in whole set of seismic requirements to restrain pip- constant dollar cost. According to several related ing systems during earthquakes was introduced DOE studies, materials used in nuclear plants in the late 1970’s. Simple cast or machined pipe have increased, e.g., from an estimated 2,000 supports (costing several hundred dollars) have ft/MW of cable for a typical plant to be con- been replaced with very sophisticated restraints structed in 1971 to about 5,000 ft/MW of cable called “snub bers,” with shock-absorbers, costing many thousands of dollars. Pipe supports have *This is measured by the Handy-Whitman nuclear index (55). become more massive and have had to be fitted Ch. 3—The Uncertain Financial and Economic Future ● 61

Figure 17.—Trends in Material Requirements in Estimates of PWR Construction Cost

225

PWR Wash 1230 (1971)

PWR Nureg-0241 (1976) m

Data of 8 PWR’s (1982-85 comm. OP)

151

7323

Highest

Average

Lowest 4889 570 . 3966 70

147 I 340 263

Concrete Piping Cable Raceway (yd3 per MW) (ft. per MW) (ft. per MW) (ft per MW)

NOTE The first two columns come from engineering estimates prepared for the NRC (formerly the AEC) based on construction data from plants under construction or complete at that time The last column comes from a survey of eight plants under construction See source article for references and more description SOURCE John H. Crowley and Jerry D. Griffith, “’U S Construction Cost Rise Threatens Nuclear Option “ Nuclear Engineering International, June 1982 with much tighter tolerances to the pipes they Finally, there has been a steady increase in the support. amount of labor required per kW, both manual (craft) and nonmanual. For a series of typical Quality-control procedures and paperwork plants costed out over 15 years in a study for have added to the cost of materials and com- DOE, craft labor requirements increased from 3.5 ponents. Although there has been no compre- workhours/kW for a plant starting construction hensive study, there are individual examples and in 1967 to 21.6 workhours/kW for the average anecdotes to illustrate the claim that quality con- of 16 plants under construction for completion trol represents a bigger and bigger share of in 1982-85 (17). Nonmanual field and engineer- nuclear materials cost. In one such example (86) ing services also have increased dramatically. For structural steel supports now required for nuclear a slightly different series of typical plants, plants cost between two and three times the cost estimates of field and engineering services in- of the same quality steel supports that are still creased from 1.3 workhours/kW in 1967 to 9.2 used for general construction projects and that workhours/kW in 1980 (16). were permitted on nuclear projects until 1975. Of this amount, the quality control procedures The increase in labor per kilowatt of capacity account for virtually all the increased cost. is the result of complex interactions resulting from 62 • Nuclear Power in an Age of Uncertainty

increasingly demanding regulations, quality-assur- utilities because of slow growth in electricity de- ance requirements and the subsequent utility mand and financing difficulties. There also ap- management response to these. These are de- pears to be important differences in the regulatory scribed in more detail in chapters 4 and 5 and environments for different generations of plants in several case studies (92). that must also be taken into account. There is large variation in material and labor A recent study of Ieadtimes for EPRI took both requirements from plant to plant, just as there is deliberate delays and regulatory stage into ac- large variation in overall capital cost. For a group count* (1). The study identified from published of 16 plants scheduled to be completed in sources those plants that had been delayed delib- 1982-86, craft labor varied from a low of 15 erately more than a year by their utilities and workhours/kW to a high of 33 workhours/kW. analyzed their Ieadtimes in a separate group. In Similarly, for a group of eight plants, linear feet a more detailed case study of 26 of these plants of cable varied from a low of about 3,300 ft/MW EPRI found that 8 had been delayed significant- to a high of about 7,300 ft/MW (see fig. 17). ly (averaging 27 months) while 22 had only been delayed an average of 2.5 months. Grouped by date of permit, it is plausible to The Increase in Nuclear identify three generations of nuclear plants. For Construction Leadtimes the first generation, for which construction permits were issued from 1966 to 1971, leadtimes* * Nuclear construction Ieadtimes also increased increased steadily from about 60 to 80 months. over the decade, making it increasingly difficult This appears to reflect an increase in the designed to match nuclear plants to demand, adding to in- complexity of nuclear powerplants and possibly terest and escalation costs, and exacerbating the strains of rapid growth as well. problems with cash flow. At the same time, leadtimes for coal plants increased very little (from A second group of plants had their construc- an average of 58 to about 60 months) (1). tion permits issued from 1971 to 1974. Leadtimes for that group were much higher than the first, Documenting the increase in Ieadtimes for nu- averaging 120 months and ranging from 100 to clear plants is made difficult by the fact that some 160 months. Leadtimes for plants without signifi- plants have been delayed deliberately by their cant deliberate delays averaged about 10 months less than those with significant deliberate delays. It appears that this group of plants suffered a major increase in regulatory complexity (including the 1974 Calvert Cliffs decision, the regulations following the 1976 Browns Ferry fire and the 1979 accident at Three Mile Island) without the opportunity to develop construction and regulatory planning techniques to handle the increased complexity. They may have suffered as well from some of the effects of rapid growth in the industry, such as incomplete designs and inexperienced supervisors. Although the data are sketchy, it is possible that a third generation of nuclear plants is now emerg-

*The study grouped the plants by date of construction permit to avoid the obvious problem that later completion dates, by definition, include a larger proportion of Iong-leadtime plants. * *Leadtimes for this analysis are defined as time from date of construction permit to commercial operation. Ch. 3—The Uncertain Financial and Economic Future ● 63

ing, with construction permits issued later, in their dealings with the NRC. (Case Study 2 in ch. 1975-77. After adjusting for deliberate delays and 6.) excess optimism in time estimates, EPRI found that this latest group of plants appears to have somewhat shorter Ieadtimes than the 1971-74 The Impact of Delay on Cost group. Leadtimes for all plants in the group av- In a period of substantial general inflation char- erage about 100 months and range from 65 to acteristic of the last 5 years, a delay in nuclear 120 months. Plants without significant announced plant construction can cause an alarming increase deliberate delays average 10 to 15 months less in the current dollar cost of the plant. Increases than the average. The plants with shortest lead- in the current dollar cost, however, must be dis- times in this group are already in operation and tinguished carefully from increases in the real or were completed faster than the shortest Ieadtime constant dollar cost of the plant (after the impact plants in the earlier group (see fig. 18). The num- of general inflation has been eliminated). These bers are so small, however, that it is too early to in turn must be distinguished from increases due tell if these plants are anything but anomalies. to changes in regulations or other external influ- Those plants with longer Ieadtimes in this latest ence during the period of delay. group still may experience significant delays beyond the adjusted estimates calculated by EPRI. For a hypothetical plant that has been expected At the same time there is some case study evi- to be completed in 8 years but instead has been dence that the plants that were started later were delayed to 12 years, with no increase in complex- able to compensate for increased regulatory com- ity or scope, there are two sources of increases plexity in the plant design and construction man- in total capital cost in constant dollars. One is that agement and were also able to plan systematically nuclear components, materials and labor may

Figure 18.—Construction Leadtimes for Nuclear Powerplants

I I

● ✍ ● ✍✍ ●

✎ ● ●

✼ ● I

●

● ● ● ● I

1965 1970 1975 1980 Construction permit issue date

NOTES. The Ieadtimes are based on estimated times to commercial operation for those plants not yet in service. The gaps correspond to periods of Iicensing inactivity in the industry Leadtimes are calcuIated from construction permit issue date-to-date of commercial operation SOURCE Applied Decls!on Analysis, Inc An Ana/ys/s of Power Plant Construction Lead T/rrres, Volume 7” Analys/s and Results, EPRI. EA.2880 February 1983 Graph based on Nuclear Regulatory Commlsslon data. 64 Ž Nuclear Power in an Age of Uncertainty have increased 1 or 2 percent faster than general increase of 30 percent. For this case, shortening inflation (escalation). The second is real interest the plant’s Ieadtime would save about a third of during construction* that is capitalized and its current dollar cost but only about 12 percent added to general total plant cost as AFUDC (Al- of its constant dollar cost. * lowance for Funds Used During Construction). (See box B above.) The Cost of Electricity From Coal and Table 9 shows the increases in constant dollars Nuclear Plants and in current dollars of several different cases: The steadily increasing capital costs of nuclear 5, 7, and 9 percent general inflation with no real power (including the increasing costs brought escalation in nuclear components and with 2 per- about by increasing Ieadtimes) leads to a crucial cent real escalation and real interest rates of 3 question: at what point does the increasing cap- percent and 5 percent above general inflation ital cost of nuclear plants make nuclear power (1 3). Several of the examples in table 9 can serve a more expensive source of electricity compared as illustrations of the difference between increases to alternative generating sources, especially coal? in current and constant dollars. For example, in As long as it is likely that utilities will avoid the case 3, if a plant takes 12 years to build during use of oil and gas for base load electricity gener- a period of general inflation of 7 percent, escala- ation, the chief competitor to nuclear is coal. tion of nuclear components of 9 percent (2 percentage points faster than general inflation) and Initially (for most nuclear plants completed by an interest rate of 12 percent (a rather high real the early or mid-1970’s) there was no doubt that interest rate of 5 percent), the “mixed current electricity generated from these plants was sub- dollar cost” of the plant will be 233 percent high- stantially cheaper than coal-generated electrici- er than its overnight construction cost. Two-thirds ty. Because of the way capital charges are recov- of the increase, however, is general inflation. The ered in the rate base (see box C above), the cost real constant dollar increase in the plant cost is of electricity from these plants has become steadi- only 48 percent. Construction of the same plant ly cheaper relative to electricity from coal plants in 8 years time would cause a current dollar in- built at the same time. As the capital cost of crease of only 123 percent and a constant dollar nuclear plants has risen, however, the relative ad-

*Real interest is the nominal rate of interest less the rate of general inflation, e.g., real interest is 5 percent for nominal interest rates *In this case 2.23 (8 years) is about 67 percent of 3.33 (12 years) of 12 percent when general inflation is 7 percent. and 1.30 (8 years) is about 88 percent of 1.48 (12 years).

Table 9.—Additions to Overnight Construction Cost Due to Inflation, Escalation and Interest During Construction (in constant and current dollars) Percent increase in Percent increase in constant dollars current dollars 8-year 12-year 8-year 12-year Ieadtime Ieadtime Ieadtime Ieadtime Case 1: 7% general inflation, 0 escalation, 100/0 interest rate . . . . . + 12 + 19 +92 + 167 Case 2: 7% general inflation, 0 escalation, 12% interest rate . . . . . +21 +33 + 107 + 200 Case 3: 7% inflation, 2% escalation, 12% interest rate...... +30 +48 + 123 + 233 Case 4: 5% inflation, O escalation, 100/0 interest rate...... — +34 — + 140 Case 5: 90/0 inflation, O escalation, 14% interest rate...... — +33 — + 273

NOTE: “Escalation” is defined as the increase in the unit costs of labor and components used in nuclear plants (with no change in quality) above the rate of general inflation. For inflation of 7 percent, an interest rate of 10 percent corresponds to a real interest rate of 3 percent, an interest rate of 12 percent corresponds to a real interest rate of 5 percent. SOURCE: For the calculation, Wilfred H. Comtois, “Escalation Interest During Construction and Power Plant Schedules,” Westinghouse Power Systems Marketing, September 1975. Ch. 3—The Uncertain Financial and Economic Future ● 65 vantage of nuclear power has diminished. For for nuclear plants. Fuel cost, however, may range plants presently under construction, the average from less than $0.01/kWh in regions where plants capital cost is now so high that the typical nuclear can be built near the coal mine to almost four plant probably would produce more expensive times that in regions located far from coal fields electricity over its life time than the typical coal (assuming rapid increases in coal prices). The rate plant. Only electricity from the least expensive at which coal prices are likely to escalate over nuclear plants still may be competitive with av- several decades has a significant influence on the erage cost coal-generated electricity. Average cost forecast average cost of electricity from the plant nuclear plants, however, still can compete with over its lifetime. Levelized electricity prices will more expensive coal plants. be about $0.015/kWh (in constant dollars), higher, on average, if coal prices escalate at a real Comparing the costs of nuclear and coal-fired annual rate of 4 percent than if they don’t escalate electricity is made difficult by the different im- at all in real terms (36). pact of fuel and capital cost components for each type of plant. Capital cost is a far more impor- Unfortunately, there is no study of recent plants tant component of nuclear-generated electricity using actual reported capital cost of coal plants. than for coal plants. While the Ievelized cost of In a DOE study (36) using coal and nuclear plant fuel (uranium ore, enrichment, storage shipment model data on capital cost, the cost of electrici- and disposal) and operations and maintenance ty is about equal in five of the ten DOE regions, each run about $0.0075/kWh, the capital charge slightly lower for nuclear in two of the regions (Ievelized charge* over the life of the plant) per and considerably lower for coal in two of the re- kWh for new plants may range from as little as gions. There is reason to believe, however that $0.01/kWh for older reactors to $0.10/kWh or nuclear capital costs are higher and coal capital even more for the most expensive of today’s reac- costs may be lower than the study results. Capi- tors. The capital charge per kWh increases with tal costs of the typical nuclear plant reported in higher total construction cost (including the im- the study are about 15 percent lower than the pact of longer Ieadtimes), with higher interest average (in constant dollars) of the plants now rates, and with a shorter capital recovery period. under construction. On the other hand, the capi- The capital charge (as well as operations and tal cost of the typical coal plant reported in the maintenance) per kWh also increases as the plant DOE study is more than 40 percent higher than capacity factor* * is reduced because there is less the capital cost of the typical 1978 coal plant (in- output among which to apportion the annual cluding a flue-gas desulphurization scrubber) in capital cost. Since the earliest nuclear plants were the 1981 Komanoff study updated to constant built, there have been significant increases in all 1982 dollars. While it is possible that the capital the categories that increase annual capital cost of coal plants may have increased 40 per- charges. The capacity factors of nuclear plants cent since 1978, several factors make it unlike- also have been less than expected. (See ch. 5 for ly. Coal plant construction Ieadtimes (unlike nu- more discussion of nuclear capacity factors.) clear plant Ieadtimes) have not increased since 1978. Since the cost of scrubbers already is in- For coal-fired electricity, on the other hand, the cluded in the 1978 typical coal plant capital cost, cost of the fuel is at least as important as the it is unlikely that further pollution control im- capital cost in determining the price of electrici- provements and design improvements would add ty over the life of the plant. Operations and main- more than 20 percent. If, indeed, actual nuclear tenance of coal plants cost somewhat less than construction costs are higher and actual coal plant construction costs are lower than the plant *Various techniques are used to “levelize” costs over a plant’s model results, the typical nuclear plant would be lifetime. One simple method, used in the EIA study of coal and expected to produce more expensive electricity nuclear costs, is to take the present discounted value of the stream of costs and divide it by the number of years to get an annual level- in all regions. ized cost (36). **Capacity factor equals the number of hours of actual opera- Low-cost nuclear plants, however, still would tion divided by the hours in the year. be competitive with the average coal plant. Com- 66 ● Nuclear Power in an Age of Uncertainty

However, there is a further element of the competition between the costs of electricity from nuclear and from coal. The pattern of costs over the life of the plants are substantially different. Under current accounting rules capital charges are highest in the early years and decrease as depreciation charges are deducted from the asset base. Coal costs on the other hand will increase at, or faster than, the rate of general inflation over the life of the plant. Thus for plants with the same Ievelized cost of electricity, electricity from the nuclear plant will cost more in the early years and electricity from the coal plant will cost more in later years (see box C above). The higher cost of nuclear plants in the early years could be discouraging to an electric utility that had faced much opposition to rate increases.

Future Construction Costs of Nuclear Powerplants It is very unlikely that there will be any future for nuclear plants of current average capital cost. Nuclear plants, on average, are now so costly that they are no longer likely to produce electricity more cheaply than coal over their lifetimes. Given the pattern of front-end loading of capital costs,

Photo credits: Atomic Industrial Forum even with equal lifetime costs, nuclear-generated electricity would not be cheaper than coal-gen- Nuclear fuel comes in pellets and is assembled into fuel rods and then into fuel assemblies. Fuel for a erated electricity for 10 to 15 years. For this nuclear powerplant is compact and only transported reason, utilities are not likely to order more every 12 to 18 months. It is inexpensive compared to nuclear plants if the capital cost of newly ordered the cost of alternative fuels such as coal, natural gas, and oil. Each Vi-inch-long pellet of enriched uranium plants is expected to repeat that of the current shown here can generate approximately the same average plant. amount of electricity as 1 ton of coal There is considerable evidence that, with effort, the cost of an average nuclear plant can be monwealth Edison Byron plants are expected to reduced substantially from the current level. The cost $1,100 to $1, 150/kW in 1982 dollars (without lowest cost nuclear plants already cost substan- interest during construction)* and Duke Power’s tially less than the average. Within the present McGuire and Catawba plants are expected to cost framework of regulatory requirements for plant $900 to $1,200/kW in 1982 dollars. When the design and quality control, a few utilities have construction cost of a nuclear plant is no more built enough plants to take advantage of a con- than 20 to 40 percent above that of a coal plant, struction learning curve and have developed it can be expected to produce electricity more techniques for minimizing delays, rework, and cheaply, sometimes substantially, over the plant’s worker idleness. These techniques are described lifetime. in more detail in chapter 5. They involve careful and complete engineering, careful and thorough *Costs of the Byron plants may go up, however, following a planning and project management (including the January 1984 NRC decision to deny the plants an operating license. use of sophisticated computerized tracking and Ch. 3—The Uncertain Financial and Economic Future ● 67 inventory planning), and attention to motivation ● Reduction in Construction Workhours. A for productivity and cooperation. If a separate fully standardized pre-certified design and company is used to manage the project, there emphasis on multi-unit sites could reduce must be explicit incentives to complete projects construction workhours from 14 million to on time and at budgeted cost. A reasonable goal 12 million per 1,300-MW plant, a number for such efforts could be to equal the capital cost which is still about 25 percent higher than of Commonwealth Edison’s Byron and Braid- estimated construction workhours for a com- wood plants and Duke Power’s McGuire/Cataw- parable French plant. This would come ba plants of $1,100/kW (1982 dollars) direct con- about through progress up a learning curve struction cost plus another $200 to $250/kW for of construction management techniques. interest during construction. ● Reduction in Engineering Workhours. Stand- ardization and regulatory predictability also Looking overseas, there is evidence that a dif- could cut engineering workhours per plant ferent approach to certain aspects of safety regu- roughly in half from about 9 million to about lation and quality control could bring construc- 4.5 million, by reducing construction engi- tion costs down still further. Constructing a French neering support by more than 80 percent plant requires about half as many workhours/kW and quality assurance workhours similarly. as constructing an American plant (86). As is de- Engineering workhours would still be about scribed in more detail in chapter 7, the French 60 percent higher than those in France, re- have standardized their plants and have built two flecting the differences in U.S. construction basic types of reactors (925 MW and 1,300 MW). project organization. This avoids much of the rework that occurs in ● Eight-Year Project Schedule. Reducing American plants because the engineering is es- average project schedules from 11 to 8 years sentially complete before each plant is started, would reduce interest and real escalation and because the first plant of each type functions costs. Total construction cost in constant dol- as a full-scale model to help avoid piping and lars would be reduced by 7 to 15 percent cable interferences and other problems of two (see table 9). dimensional design. Within a far more centralized The desirability of standardization in nuclear and controlled approach to safety regulation (see plant design and construction is a complex ques- ch. 7) the French have also taken an approach tion that would affect far more than the capital to earthquake protection that minimizes the im- cost of nuclear plants. It was the subject of a pact on construction. They also have a different previous OTA report, Nuclear Powerplant Stand- approach to quality control that minimizes delays ardization (April 1981 ) and is discussed further during construction (described in chs. 4 and 5). in chapter 4. One issue is whether standardization can be achieved without sacrificing the adap- A specific estimate of possible reductions in av- tability of nuclear technology to new information erage plant cost was made in a recent study of about designs that would benefit nuclear plant the U.S. nuclear industry viability (87). According safety or operation. A second issue is the institu- to this estimate, typical plant costs (including in- tional obstacles to standardization in an industry terest during construction) could be reduced from with more than 60 nuclear utilities, 4 reactor ven- about $2,220/kW (in 1982 dollars) to $1,700 to dors, and more than 10 AE firms. $1,800/kW (20 to 25 percent less). This estimate assumes that the United States can go only part While opportunities exist for reducing nuclear way towards constructing plants with as few construction costs, it should be recognized that workhours as is now done in France. The French costs might also increase. Further serious ac- regulatory environment, and utility management, cidents could lead to a new round of major and construction tradition are sufficiently different changes in regulation. There are still important that much is unlikely to be duplicated in the unresolved safety issues (discussed in ch. 4) that United States. The proposed steps to bring con- could lead to costly new regulations. utility ex- struction costs down would include: ecutives are well aware of this possibility. 68 ● Nuclear Power in an Age of Uncertainty

The Financial Risk of Operating a Since Three Mile Island, property insurance Nuclear Powerplant coverage for nuclear pIants has increased to $1 billion, about half in primary insurance and the rest in excess insurance (once a $500 million ac- The Risk of Property Damage.–The accident cident cost has been reached). Some of the pri- at Three Mile Island was a watershed in U.S. mary insurance and most of the excess insurance nuclear power history because it proved that have been provided by two new mutual insur- serious accidents could indeed occur and cause ance companies formed by groups of utilities, enormous property losses even without causing Nuclear Mutual Ltd. (NML) and Nuclear EIectric any offsite damage. The total cost of the cleanup Insurance Ltd. (NEIL) –(see table 10). NEIL also is estimated at $1 billion, not counting the carry- provides almost $200 million in insurance for pur- ing costs and amortization of the original capital chases of replacement power while a plant is used in building the plant nor the cost of restart- disabled. ing the plant. Of this $1 billion, $300 million was covered by insurance from the insurance pool Despite this threefold increase in insurance, (see table 10 for explanations). General public however, some of the expenses incurred in the Utilities (GPU) is now in the process of negotiating Three Mile Island accident still have not been in- the financing of the rest from various sources in- sured; namely, maintenance of the disabled plant cluding the utility industry through the Edison and carrying costs and amortization of the capital Electric Institute, the rate payers as approved by tied up in the disabled plant. For the moment the Pennsylvania and New Jersey PUCs, the States these are being paid by GPU stockholders who of Pennsylvania and New Jersey and the Federal have not received a dividend since the accident Government. (44).

Table 10.—Nuclear Plant Property and Liability Insurance

Description Coverage ANI-MAERP: Commercial insurance consortium Reactors at 34 sites. “Primary” of about 140 investor-owned insurance (responds initially companies (American Nuclear to a loss) $500 million/site Insurers - ANI) and 120 mutual $68 million excess companies (Mutual Atomic Energy Reinsurance Pool — MAERP) NML: Nuclear Mutual Limited is a Reactors at 27 sites, $500 mutual insurance company created million primary insurance/site by several investor-owned utilities and located in Bermuda NEIL-I: Nuclear Electric Insurance Limited — Reactors at 36 sites, $2.3 million/ extra expense insurance to pay week 1st year; $1.15 million/week for replacement power from an 2nd year up to $195 million accident covered in primary insurance NEIL-II: Nuclear Electric Insurance Limited — Reactors at 32 sites. “Excess” property damage excess damage insurance (covers damage above above limit of primary insurance limit of primary insurance) $415 million/site Liability insurance: ANI-MAERP: Liability insurance required by All reactors. $160 million available Price-Anderson from premiums, plus $400 million/ accident available from retroactive assessments of $5 million/reactor/ accident SOURCE: John D. Long, Nuclear Property Insurance: Status and Outlook, report for the U.S. Nuclear Regulatory Commission, NUREG-0891, May 1982; papers presented at Atomic Industrial Forum Conference on Nuclear Insurance, Feb. 14-16, 1983. Ch. 3—The Uncertain Financial and Economic Future ● 69

Further increases in insurance capacity are de- Ironically, this pressure to increase the limit for sirable, given the increasing replacement value public liability comes just as the private insurance of plants. However, they are limited by the assets resources of the industry have increased enough of the insurance companies and the utilities in to cover the current statutory limit fully. About this country and by the reluctance of reinsurers, $160 million is available from the insurance such as individual syndicators in Lloyd’s of Lon- pools, and the rest (about $400 million) is avail- don, to assume any more American nuclear risk. able from a $5 million per reactor retroactive assessment required in the Price-Anderson Act. The adequacy of current insurance is threat- Under current law, the Federal Government has ened further by several other issues. Under public guaranteed to provide any liability damages be- pressure, Congress could stipulate that all prop- yond the nuclear industry’s resources and up to erty insurance be used first to pay cleanup costs the statutory limit. Currently, because of the avail- and only then to pay carrying costs and the costs ability of private insurance funds and the increase of restarting the plant. This would mean that the in the number of plants available to pay the $5 utility would have to turn to the PUC to obtain million assessment, the Federal Government has higher electricity rates to cover all the other costs no liability. if the limit were raised or eliminated, of an accident or obtain the funds by withholding there would be pressure for the Federal Govern- dividends from shareholders. Another source of ment to again assume the excess liability. uncertainty is the heavy reliance of the utility-run From the standpoint of the insurance industry, mutual insurance systems on retroactive assess- the most serious issue is the growing pressure ments. These are premiums that the utility com- from citizen’s groups to allow damage from nu- mits itself to pay in the event of an accident. NML, clear accidents to be covered in homeowners’ for example, may call on retroactive payments policies. Currently, by consensus of all insurers, up to 14 times annual premiums in the event of all homeowners’ insurance policies specifically a serious accident that depletes existing insurance exclude a nuclear accident as an insurable risk. reserves. The willingness and ability of utilities Homeowners, in effect, may make claims only to pay these assessments has not yet been tested. against the responsible utility and be paid from Some observers have expressed the fear that the utility’s own insurance resources. For insurers, PUCs may balk at allowing utility insurance ex- this characteristic of nuclear insurance channels penses “to pay for the other guy’s accident” (57). the risk into a single category which can be identified and assessed. Since the potential damages The Risk of public Liability .–Since 1957, the are both large and of unknown probability, insur- Price-Anderson Act has limited public liability, in ers are much more willing to provide fairly large the event of a serious accident, to $560 million. sums if the structure of risk is simple because a Pressure to increase this limit has been mounting. given accident will result only in a claim from a Inflation alone would justify raising the limit to single utility, and not in hundreds of individual about $1.9 billion assuming $560 million was an homeowner claims through multiple insurance appropriate figure in 1957. Pressure, however, companies. Were the homeowner’s insurance to go beyond keeping up with inflation, or even exclusion to be removed by law, one probable to eliminate the limit altogether, arises from result would be a reduction in total private in- several studies published over the last 10 years, surance resources available for a single accident. the Rasmussen Report (WASH-1900) in 1974 and This is because the resulting multiple sources of the 1982 Sandia siting study. Both analyze the liability (from millions of homeowner policies) in- consequences from low probability accidents and creases the perceived risk to the insurers, who are described more fully in chapter 8. Part of the in turn respond by reducing the total amount Price-Anderson Act is due to expire in 1987. The available. first round of debate in Congress on this issue may begin soon, stimulated by the recent publication Another financial risk of unknown size is the of an NRC report on public liability. possibility that workers exposed to radiation may

25-450 0 - 84 - 6 : QL 3 70 . Nuclear Power in an Age of Uncertainty file occupational health suits in future years, pays fixed sums for each type of injury. Since in- based on a statistical link between low-level radia- jury in this case is based on a probability link to tion exposures and various diseases, such as levels of radiation exposure the level of prob- cancer, with long periods of development. In ability would be included in the program, much some respects the nuclear power industry is bet- as has been proposed for compensation for vic- ter prepared for such suits than industry has been tims of nuclear weapons testing. for comparable suits arising from exposure to asbestos because detailed records are kept on Impact of Risk on the Cost of Capital every worker’s exposure to radiation. Current- ly, any court settlements due to workers’ expo- As previously noted, from an investor’s point sure to radiation would be paid out of ANl- of view there are several reasons to be wary of MAERP pooI insurance. If the size of such settle- utilities with substantial nuclear operations: ments becomes at all substantial, however, utili- ● since nuclear-generating plants take longer ties and their insurers may move to establish a to build, it is more likely that they will be fixed compensation program, comparable to the poorly matched to actual demand; basic workman’s compensation program, which ● the cost of constructing them is harder to estimate and control than it is for coal plants; and ● there is a small but finite risk of a major disabling accident. Under current insurance coverage and PUC rate decisions, a large fraction of the many costs of such an accident would have to be borne by the stock- holder. A few attempts have been made to estimate the impact of these three elements of risk on the cost of capital to nuclear-owning utilities but the results are not clear-cut. As of 1981, the highest bond ratings belonged to utilities operating nuclear plants, although the lowest bond ratings belonged to utilities with nuclear pIants under construction. After Three Mile Island, there was an immediate effect on the relative stock market prices of nuclear and non-nuclear utility stocks. The price of non-nuclear stock increased over 50 cents a share relative to nuclear stock. The effect persisted for at least 2 years (46). Financial experts and utility executives agree, however, that another serious accident could have very serious financial consequences. In the year following the Three Mile Island incident, a study of investor attitudes towards nuclear utilities (1 1,46) showed that investors ranked the risks associated with nuclear power Photo credit: Westinghouse as a serious problem but less than problems One source of uncertainty about the future cost caused by regulation, high interest rates and in- of nuclear power is the possibility that workers exposed to radiation may sue the nuclear plant flation. Twenty-five percent of institutional inves- owners to recover damages from health effects of tors said the Three Mile Island accident had a neg- radiation exposure ative impact on the weighting of electric utilities Ch. 3—The Uncertain Financial and Economic Future • 71

in their investment portfolios. I n general, port- and late November 1983, stock prices dropped folio managers showed increased concern if utili- in 36 out of 50 companies with nuclear plants ties had a high dependence on nuclear, but over under construction. 82 percent said they recommend companies with In summary, utilities assume greater risks when some nuclear component in their fuel mix. It they build nuclear pIants than when they build would seem that investors are more concerned coal plants. Even if, due to standardization and currently about the risk of construction cost over- careful construction, the cost of nuclear power runs and delays and the financial strains of plant over 30 years is estimated to be substantially less construction than they are about the financial risk than coal-fired electricity, utilities still might hesi- of a disabling accident (47). The recent indica- tate to order more nuclear plants unless they are tions that several nuclear plants such as Zimmer, compensated in some way for the additional risks. Midlands, and Marble Hill may never be completed and licensed to operate has caused another round of investor concern. * Between October

*See Nuc/ear phobia, a research report by Merrill Lynch, Pierce, Fenner & Smith, Dec. 15, 1983.

NUCLEAR POWER IN THE CONTEXT OF UTILITY STRATEGIES

The decision to order a nuclear powerplant is utilities do not seem to be avoiding capital invest- only one of many choices utility executives can ment at the risk of providing inadequate electric make given their companies’ load forecasts and supplies, nonetheless they are taking financial present and future financial situations. They could considerations heavily into account, especially instead order one or more coal plants; convert the ability to earn a return on CWIP. Some ex- an oil or gas plant to coal; build a transmis- ecutives say their companies have deliberate sion line to facilitate purchase of bulk power policies of providing generating capacity with from Canada or from elsewhere in the United either minimum capital cost or short Ieadtimes States; develop small hydroelectric sources, wind or both. sources, or other small-scale sources of power; One possible option is to delay any powerplant or start a load-management, cogeneration, or construction as long as possible and then meet energy conservation program (or some combina- any need for new capacity with combustion tur- tion of all of these). bines which can be constructed in 3 to 4 years What utility executives choose depends on the and cost only $200 to $300/kW (in 1982 dollars). reliability of their load forecasts, the options for Such a choice is now less risky for future elec- retiring oil and natural gas plants, the availabili- tricity rates because of apparent softening of na- ty of reliable sources of purchased power, the natural gas markets. Combustion turbines cost so ture of rate regulation in the State in which they little that they could in theory be written off operate, and their companies’ abilities to manage quickly and replaced by longer Ieadtime plants large construction projects on the one hand or if electricity demand were increasing enough to successful load management and conservation warrant longer Ieadtime generating capacity with programs on the other. lower fuel costs. From a recent survey of utility executives (90) In a recent study six utilities described two alter- and the results of two OTA workshops, it is clear native sets of construction plans: one plan that that utility executives are now considering a they would follow under financially generous rate much wider variety of alternatives to construc- regulation and the other under financially con- tion of new large generating plants. Although strained rate regulation (64). There is a somewhat 72 . Nuclear Power in an Age of Uncertainty

exaggerated difference between the two sets of less vulnerable to changes in regulation and ad- circumstances*–but they do illustrate the range verse public reactions and thus more attractive. of utility choice. Several utilities are members of a gas-cooled reactor council that supports research and develop- Under financially constrained circumstances, ment of high temperature gas-cooled reactors the six utilities expect to rely on more use of pur- (HTGR) (described inch. 4). One executive testi- chased power. One will invest in more transmis- fied for Florida Power & Light Co. (FP&L) in March sion lines. The utilities also expect to keep old 1983 that for FP&L’s crowded site in Dade Coun- plants on line and will defer the retiring of oil and ty, an HTGR is the only option. Shallow water natural gas pIants. It is interesting that none ex- and delicate ecology hinder coal transportation pect to build combustion turbines to catch up to and the closeness of the City of Miami and prob- demand growth.** lems with raising the water temperature rule out The preferred generating choice of the six utili- a light water reactor. FP&L does not want all the ties under financially generous circumstances is possible headaches of building the lead HTGR medium-sized coal units of 500 to 600 MW which but might be willing to build the second plant the six utilities plan to build in substantial num- (91). bers, over 17 GW by the year 2000. Although sev- If utilities wish to avoid building powerplants eral utilities expect to finish nuclear powerplants altogether they have several options (14). One under construction, only one of the six expects of these, featured in several of the six utility strat- to start a new nuclear plant. Utility executives at egies described in table 11, is to make better use the OTA workshops and in the survey now per- of existing powerplants. Ironically, the financial ceive nuclear power to be too risky to include weakness and excess capacity that had led utili- in future construction projects even under a less ties to cancel new construction has also fostered financially constrained future. Even for those neglect of maintenance of existing powerplants. utilities that have experience in keeping construc- In one recent survey of 80 GW of coal power- tion costs under control, there are important plants there had been a decline of more than 12 perceived risks from lack of public acceptance percent in availability from 1970 to 1981 (1 5). and the possibility of one or more Three Mile Pennsylvania is one of several States investigating Island types of accidents. Utility executives said changes in regulatory policies that would en- they expect to make use of “cookie-cutter” coal courage more efficient use of existing powerplants because of the greater predictability of their plants (78). operating and construction costs. Major investment may be needed to extend the It is conceivable that other types of nuclear life of an existing plant well beyond the normal plants than those currently available in the United retirement age of 30 to 35 years. A plant that has States would be more attractive to utility execu- been operating effectively for 40 to 50 years may tives, Executives reported in an EPRI survey of not have any of the same components as the orig- utility executives’ attitudes towards nuclear pow- inal plant. Nonetheless, substantial investment to er that smaller nuclear plants would be desirable upgrade an existing plant will in almost all cases because they would require a smaller total capital cost far less than building an entirely new plant commitment and could more easily be fit to un- of the same capacity. certain load growth (22). Utilities can also substitute programs to reduce Some executives also believe that significant peak demand for building new capacity. A recent safety improvements would make nuclear plants EPRI survey identified over 200 utilities that were working with their customers on conservation *For example, the cost of capital is assumed to differ by 300 basis and load management programs (23). While points between the generously treated case which results in an AAA bond rating and the constrained case which results in a BBB bond some of the programs were demonstrations, rating. others represented major corporate commitments * * Building a combustion turbine for use more than 1,500 hours a year is still prohibited under the Fuel Use Act. In practice, to load control. For example, the New England however, an increasing number of exemptions are being allowed. Electric System’s successful experiments with on- Ch. 3—The Uncertain Financial and Economic Future ● 73

Table 11 .—Six Sets of Alternative Utility Construction Plans

Projected Utility and type growth in Plan A Plan B (timeframe of plan) load (%/yr.) capital discouraging capital attraction Utility A. Coal conversion 1.5 Cost $4.1 billion Cost $9.9 billion (1982-2000) ● Sell part of share of nuclear plant . Finish nuclear plant on schedule • Convert 200 MW oil to coal • Convert oil plant to coal (800 MW) • Reduce sales to outside ● Build four 600 MW coal plants (two- customers thirds ownership) Ž Purchase 175MW . Retire no old plants Utility B. Coal/nuclear 2.0 Cost $23.9 billion Cost $40.6 billion (1982-2005) . Delay two-thirds of a new nuclear ● Complete several nuclear plants plant for 4 years without delay . Double the amount of purchased . Build four 500 MW coal plants in power 1990’s • Consume more oil and gas ● 75 ‘/0 share i n two 1,100 MW nuclear . No existing plants retired units in 2000 ● All plants retired on scheduIe ● Intermediate load coal plant Utility C. Gas/coal 4.0 Cost $64.8 billion Cost $62.8 billion (1982-2000) ● 3,000 MW coal capacity 1982-97 ● 5,000 MW new coal capacity in 600 ● 6,000 MW coal capacity MW increments 1982-97 1998-2009 ● 5,000 MW more coal 1998-2009 • Reserve margins 13°/0 1990; 9.60/0 . Reserve margin over 20°/0 in 2000 Utility D. Gas displacement 3.0 Cost $8.9 billion Cost $30.5 billion (1982-2001) ● Finish large nuclear plants near ● 2,500 MW of coal capacity 1988-97 completion to displace gas . No plants retired . Three large coal plants in . Meet incremental demand mid-1 990’s to meet additional load through purchased power ● Oil and gas capacity retired on schedule ● Finish large nuclear plants near completion Utility E. Oil displacement 1.5 Cost $12.9 billion Cost $22.1 billion (1982-2000) ● No construction projects . Purchase share in large coal project ● Defer 2,000 MW of natural gas under construction and oil capacity ● Joint owner of coal plant online early 1990’s . Build transmission lines to purchase power Utility F. Purchase/coal 3.0 Cost $4.8 billion Cost $7.6 billion (1982-2001) • No construction . Four new coal units of 500 MW ● Increase purchased power to 1988-97 36°/0 of total ● Purchased power shrinks to 7 ‘/0 ● Spend $1 biIIion on transmission of total capacity lines to wheel in power

SOURCE Peter Navarro. Long Term Impacts of Electricity Rate Regulatory Policies for DOE Electricity Policy Project, February 1983 site thermal storage of electric heat and home management and energy conservation by their energy conservation led it to develop a 15-year customers. As described in the EPRI report and plan aimed at reducing peak demand by over 500 others (14,23,42), these include programs to pro- MW and average demand by another 300 MW vide rebates for the purchase of energy-efficient over the 1980-95 period. The plan is expected refrigerators, air-conditioners, or heat pumps, low to save utility customers about $1.2 billion over interest or interest-free loans and energy index- that period (65). ing programs. For a utility interested in conserv- ing capital, some utility-controlled load manage- Utilities have successfully used a wide variety ment technologies offer strikingly low capital cost of techniques to encourage investment in load ($110 to $200/kW). It takes thousands of installa- 74 . Nuclear Power in an Age of Uncertainty tions of such devices in buildings owned by hun- generation and wind totals about 63 GW and the dreds of different customers to equal one 100 technical potential for small hydroelectric is MW powerplant (see table 12). estimated at over 45 GW (14). OTA recently estimated the full technical potential for cogenera- Utility executives interviewed for the Theodore tion as even higher, 200 GW (72). Barry survey cited above report that they are relying more and more on non-powerplant options It is worthy of note that the estimated market to meet the needs of future growth but they still potential for cogeneration alone of about 42 GW have concern about their long-term effectiveness is one measure of the market for using HTGRs (90). Without extensive metering of components for cogeneration. As is discussed further in of individual buildings, and extensive data collec- chapter 4, HTGRs operate at far higher temper- tion on occupancy and patterns of use, it is dif- atures than do light water reactors and can be ficult to determine how much of a given build- used to supply steam and pressurized hot water. ing’s change in electricity use is due to load man- The resulting high efficiency of operation and pro- agement devices and how much is due to other duction of both electricity and steam for sale can reasons that may be short-lived or unpredictable. offset their somewhat higher capital cost. Utilities seeking to avoid costly capacity additions might also work to encourage power pro- Implications for Federal Policies duction by cogenerators and other small power producers in their service territories. (The poten- As long as electric utilities are regulated there tial for cogeneration to reduce electricity demand is great potential to influence the strategic choices was discussed above in the section on electrici- they make by adjusting the way expenses and in- ty demand.) vestments are handled in electricity rate determination. This report does not analyze all the However, current estimates of market poten- possible ways in which utility strategies other than tial for small power producers are several times central station powerplant construction can be higher than estimates of the central generating influenced, but it should be recognized that capacity that could be displaced by small produc- Federal and State regulation can be structured ion. This is because utilities can use small power to encourage cogeneration, conservation and production to displace additional central station load management, and upgrading of central sta- generating capacity only if it can be counted on tion powerplants. to occur at times of peak demand. A recent CRS study estimates the total capacity displacement Utilities have several reasons to wait before potential from cogeneration, wind, and small hy- ordering more powerplants. The current high droelectric as ranging from about 5 GW to about reserve margins are one reason, and uncertain- 22 GW, even though the market potential for co- ty about the future growth of the economy and

Table 12.-Cost and Volume of Various Load Management Devices (controlled by utilities)

Estimated number of installations to equal 100 MW reduction in peak Cost/ Approximate Device demand installation cost/kW Water heater time switch , . . . . 91,000 $130-$240 $118-$218 Radio and ripple control 71,000 Radio $95-$108 (cycles water heater, air- (water heaters) conditioners) ...... $67-$107 93,000 Ripple $100-$115 (air-conditioners) SOURCE: Table published in OTA study Energy Efficiency of Buildings in Cities; based on John Schaefer, Equipment for Load Management 1979; and other sources in contract report to the Office of Technology Assessment by Temple Barker & Sloane. Ch. 3—The Uncertain Financial and Economic Future ● 75 its impact on electricity demand is another. Al- cation of industries, which might benefit over the though electricity demand forecasting is a tricky long run from the declining rate phase. business at best, there is reason to believe, from At the moment, if rate regulatory policies across both a “top-down” and a “bottom-up” approach the country were to shift to favoring longer lead- to demand forecasting, that electricity demand time, capital-intensive technologies, coal-fired growth will be slower in the 1980’s than it will generation would be encouraged far more than in the 1990’s. nuclear powerplants because utility executives Each of the utilities, influenced by its State PUC seem to prefer the smaller size, shorter Ieadtimes, will make decisions about the proper rate and lower financial risk, and greater public accept- type of generating capacity to add. What these ance of coal. The implications for the nuclear in- individual decisions add up to in terms of a na- dustry are bleak, and read as such by the industry tional electric grid depends on the balance (87). Only a handful of orders for central station among individual utility construction decisions. powerplants of any kind is likely before 1990. For example, if all utilities choose to minimize After that, if a modest number of powerplants are capital requirements and depend on purchase needed, 10 to 15 GW/yr, coal may seem ade- power arrangements, the national reserve margin quate (unless there is a dramatic change in atti- would drop dangerously low and there could be tudes and public policy about the impacts of coal- a scramble to build plants quickly. There could burning on acid rain and carbon dioxide buildup, be a short period of unreliable electricity supply. and no significant improvement in coal-burning On the other hand, if all utilities chose to build technology). long Ieadtime generating capacity to meet fore- If the amount of new capacity needed is much casts of rapid growth in electricity demand, and larger, however (up to 30 GW/yr), utilities may demand growth failed to increase as forecast, the look to nuclear again as a way of diversifying their current high reserve margins might reappear. dependence on a single technology (coal). In ad- From a national point of view, it might be best dition, now that natural gas shortages appear less if the different States and utilities adopted a mix- likely over the next decade, it is also possible that ture of these approaches. combustion turbines, particularly high-efficiency From a perspective of long-range industrial poli- combined cycle plants will seem to be accept- cy, there may be reason for the Federal Govern- able sources of diversity. For those reluctant to ment to encourage steps that make electricity rate continue such reliance very long, or to depend regulatory policies handle inflation better, and in- heavily on so-called new technology there could sure stable electricity prices over the long term be renewed interest in new nuclear plants begin- (see box C). Although average electricity rates are ning in the 1990’s. By that time if the construc- forecast to increase smoothly and slowly over the tion and operating risks of nuclear are significantly next one or two decades, this masks a set of off- better than they seem to utilities now, or if alter- setting roller coaster rides for individual utilities, natives, namely coal, are significantly worse, util- that is reflected clearly in the differences among ities may place orders for more nuclear plants. forecasts of regional electricity rates (mentioned In particular, if nuclear plants can “match the in the section on electricity demand above). In market more” (in the words of one OTA work- times of high inflation, utilities bringing new shop participant) and come in smaller sizes, with powerplants online have rapidly increasing rates shorter Ieadtimes and predictable costs, they for several years and then slowly decreasing rates. might be a competitive option again for supply- The increasing rate phase may discourage the lo- ing electric-generating capacity. 76 . Nuclear Power in an Age of Uncertainty

Appendix Table 3A.– Estimated Costs of Nuclear Plants Under Active Construction in the United Statesa

A. 95°/0 or more complete: Bellefonte 1, 1,213 MW, TVA ...... $2,300/kW Diablo Canyon 1, 1,064 MW, Pacific Gas & Electric . .$1,700/kW Byron 2, 1,120 MW, Commonwealth Edisonb ...... $1,500+/kW b Shoreham, 819 MW, Long Island Lighting Co. . . . .$4,500 +/kW Seabrook 1, 1,500 MW, P.S. Company of New Wm. H. Zimmer, 810 MW, Cincinatti E & G ...... $3,900/kW Hampshire...... $2,800/kW Grand Gulf 1, 1250 MW, Mississippi P&L ...... $2,300/kW Watts Bar 1, 1,177 MW, TVA ...... $1,500/kW Palo Verde 1, 1270 MW, Arizona PS Co...... $2,300/kW Braidwood 1, 1,177 MW, TVA...... $1,500/kW McGuire 2, 1160 MW, Duke Power Co...... $830/kW Comanche Peak 2, 1,150 MW, Texas Utilities...... $1,050/kW B .90-95 °/0 complete: Harris 1, 900 MW, Carolina P&L ...... $2,500/kW Waterford 3, 1,104 MW, Louisiana P&L...... $2,400kW. . Beaver Valley 2 MW, 644 MW, Duquesne Diablo Canyon 1, 1,106 MW, Pacific Gas & Electric . . $1,700/kW Lighting Co. (Pa.) ...... $3,700/kW Limerick 1, 1,1055 MW, Philadelphia Electric Co. . . . .$2,900/kW Susquehanna 2, 1,050 MW, PA Power & Light Co. . . .$1,900/kW La Salle 2, 1,078 MW, Commonwealth Edison ...... $1,100/kW Nine Mile Pt. Pt. 2, 1,065 MW, Niagara Mohawk . . . . .$1,900/kW Catawba 1, 1,145 MW, Duke Power Co. , ...... $1,700/kW Hope Creek 1, 1,67 MW, Public Service E&G (N.J.) . . .$3,600/kW WPPSS 2, 1,100 MW, WPPSS ...... $2,900/kW Braidwood 2, 1,120 MW, Gulf States Utilities...... $1,400/kW Comanche Peak 1, 1,150 MW, Texas Utilities (Dallas) .$1,700/kW River Bend 1, 934 MW, Gulf States Utilities ...... 3,700/kW San Onofre 3, 1,100 MW, Southern California Edison .$1,900/kW Millstone 3, 1,150 MW, Northeast Utilities ...... $3,000/kW Fermi 2, 1,100 MW, Detroit Edison ...... $2,800/kW E. 49-59°10 complete: C. 80410°A complete: South Texas 1, 1,1250 completed ...... $3,000/kW Clinton, 950 MW, Illinois Power Co...... $3,000/kW Marble Hill 1, 1,130 MW, PS Co. of Indianab ...... $3,100 +/kW Byron 1, 1,120 MW, Commonwealth Edisonb ...... $1,500+/kW Palo Verde 3, 1,270 MW, Arizona P.S Co...... $2,300/kW Watts Bar 1, 1,177 MW, TVA ...... $1,500/kW Perry 2, 1,250 MW, CAPCO Group (Ohio) ...... $2,200/kW Palo Verde 2, 1,270 MW, Arizona OS Co...... $2,300/kw Catawba 2, 1,145 MW, Duke Power Co...... $1,700/kW Callaway, 1,150 MW, Union Electric of Missouri . . . . .$2,500/kW Vogtle 1, 1,150 MW, Georgia Power Co...... $2,700/kW Bellefonte 1, 1,213 MW, TVA ...... $2,300/kW F. Around 20°/0 complete: Wolf Creek, 1,150 MW, Kansas G& E/K.C. P&L ...... $2,300/kW Marble Hill 2, 1,130 MW, P.S. Co. of Indianab ...... $3,100 + /kW Perry 1, 1,250 MW, CAPCO Group (Ohio) ...... $2,200/kW South Texas 2, 1,250 MW, Houston L&P...... $3,100t/kW Midland 1, 522 MW, Consumers Power, Michigan . . .$2,700/kw Vogtle 2, 1,150 MW, Georgia Power Co...... $2,700/kW D .50-600/’ complete: Midland 2, 811 MW, Consumers Powerb ...... $2,700 +/kW

~ost data;s of December 1983, in mixed current dollars. Construction completion as of October 1982. +” Is added where costs are likely to go higher than utility estimates. cphysically g50/~ complete, but potentiality subject to major rework.

SOURCE: Data comoiled bv Chartes Komanoff for a DaDer bv 1. C. BUDD and Charles Komanoff. The source of data is utilitv estimates of cost of comdete nuclear plants. The costs are in “mixed current dollars,” the sum of dollars spent In each year plus applicable capitalized inierest. They are not mutuall~ comparable due to different accounting conventions for items such ss “constmction work in Pro9ress” and interest capitalization and different time oerlods. See ch. 3 for a discussion of comparable costs of these Plants. FuIIY comparable data wlli be published in early lw by Komanoff Energy Associates.

CHAPTER 3 REFERENCES

1. Applied Decision Analysis, Inc., An Ana/ysis of ar Power, ” Address to the American Ceramic so- Powerplant Construction Leadtimes, Vo/un7e 1: ciety, May 4, 1982, Atomic Industrial Forum Re- Ana/ysk and Resu/ts, Electric Power Research insti- print. tute, February 1983. 7. Budwani, Ramesh N., “Powerplant Capital Cost 2. Asbury, J. G., Mueller, R. O., and Starron, J., A Analysis,” Power Engineering, May 1980. Case Study of Rate of Return Regulatory Policy Al- 8. Bupp, Irvin C., and Derian, Jean-Claude, Light ternatives for E/ectric Uti/ities, Argonne National Water (New York City: Basic Books, 1978). Laboratory, ANL/EES-TM-198, August 1982. 9. Bureau of the Census, 1980 Annua/ Survey of 3. Boerker, Sara Wood, Garacterization of /ndus- Manufactures: Fuels and Electric Energy Con- tria/ Process Energy Services, Institute for Energy sumed, August 1982. Analysis, ORAU/lEA 79-9(R), May 1979. 10. Bureau of the Census, Statistic/ Abstract of the 4. Booz Allen & Hamilton, The Financia/ Hea/th of United States, 1981. the Electric Uti/ity /ndustry, prepared for the De- 11. Burwell, C. C., /ndustria/ Electrification: Current partment of Energy, October 1982. Trends, Institute for Energy Analysis, ORAU/lEA- 5. Borcherding, J. D., and Pal meter, S. B., “Nuclear 83-3 (M), February 1983. vs. Fossil Bulk Commodity and Productivity Trends, 12. Cambridge Reports, Inc., Investors Look at the How Bad Are They?” GPU Nuclear Corp., presen- E/ectric Uti/ity /ndustry, April/May 1980, prepared tation to the Edison Electric Institute Construction for Merrill Lynch White Weld Capital Markets Committee, May 3, 1982. Group. 6. Brandfon, William W., “The Economics of Nucle- 13. Comtois, Wilfred H., “Escalation, Interest During Ch. 3—The Uncertain Financial and Economic Future ● 77

Construction and Powerplant Schedules,” West- 31 Energy Information Administration, /mpacts of Fi- inghouse Power Systems Marketing, September nancial Constraints on the Electric Utility indus- 1975. try, December 1981. 14 Congressional Research Service, A Perspective on 32 Energy Information Administration, Inventory of Electric Utility Capacity Planning, August 1983. Powerplants in the United States, 1981 Annual. Committee print of the Subcommittee on Energy 33 Energy Information Administration, Nonresiden- Conservation and Power, Committee on Energy tial Building Energy Consumption Survey: 1979 and Commerce. Consumption and Expenditures Part 1: Natural 15. Corio, Marie, “Why Is the Performance of Elec- Gas and Electricity, March 1983. tric Generating Units Declining?” Pub/ic Uti/ities 34. Energy Information Administration, Nonresiden- Fortnight/y, Apr. 29, 1982. - tial Buildings Energy Consumption Survey: Fuel 16. Crowley, John H., “Cost Drivers for Nuclear and Characteristics and Conservation Practices, June Coal-Fired Powerplants,” United Engineers& Con- 1981. structors, Inc., September 1981. 35, Energy Information Administration, Nuc/ear P/ant 17, Crowley, John H., and Griffith, Jerry D., “U.S. Cancellations: Causes, Costs, and Consequences, Construction Cost Rise Threatens Nuclear Op- April 1983. tion, ” Nuc/ear Engineering International, June 36. Energy Information Administration, Projected 1982. Costs of Electricity From Nuclear and Coal-Fired 18 Data Resources, Inc., Energy Review, summer Powerp/ants, vol. 1, August 1982, 1982, spring 1983, and summer 1983. 37, Energy Information Administration, Projected ?9. Department of Energy, Office of Nuclear Reactor Costs of Electricity From Nuclear and Coal-Fired Programs, Update, April/June 1983. Powerp/ants, vol. 2, August 1982. 20. Department of Energy, Office of Policy Planning 38, Energy Modeling Forum, E/ectric Load Forecast- and Analysis, The Future of Electric Power in ing: Probing the Issues With Models, VOIS, 1 & 2, America: Economic Supply for Economic Growth, EMF Report 3, March 1980. DOE/PE-0045, June 1983. 39, Energy Modeling Forum, Aggregate Elasticity of 21, Edison Electric Institute, Statistic/ Yearbook of the Energy Demand, vol. 2, EMF Report 4, Novem- E/ectric Uti/ity Industry 1980, November 1981. ber 1981.

22< Electric Power Research Institute, Summary of Dis- 40, Federal Energy Regulatory Commission, Quarter- cussions With Utilities and Resulting Conclusions: ly Repori on Qualifying Small Power Production Preferred Characteristics of New LWR’S, June and Cogeneration Facility Filings, Oct. 1, 1982. 1982. 41, Foley, Michael, “Electric Utility Financing–Let’s 23 Electric Power Research institute, 1980 Survey and Ease Off the Panic Button, ” Pub/ic Uti/ities Fort- Evaluation of Utility Conservation, Load Manage- night/y, Jan. 6, 1983. ment and Solar End-Use Projects, January 1982. 42. Geller, Howard S., Efficient Residential Appli- 24. Electric Power Research Institute, Technica/ As- ances: Overview of Performance and Policy sessment Guide, May 1982. /ssues, American Council for an Energy-Efficient 25. Electric Power Research Institute, Workshop Pro- Economy, July 1983. ceedings: Environmental and Social Impacts of an 43. General Energy Associates, /ndustria/ Cogenera- Electricity Shortage, 1978. tion Potential: Targeting of Opportunities at the 26. Energy Information Administration, 1982 Annua/ P/ant Site, 1982. Energy Review, April 1983. 44 Graham, John G. (Treasurer, GeneraI Public Util- 27. Energy Information Administration, 7982 Annua/ ities Corp.), Three Mi/e /s/and Status Report, pre- Energy Outlook, April 1983. sented to the Atomic Industrial Forum Nuclear in- 28. Energy Information Administration, Capacity Uti/- surance Issues Conference, Washington, D. C., izati;n and Fuel Consumption in the Electric Feb. 13-16, 1983. Power /ndustry, 1970-1981, July 1982. 45. Hyman, Leonard S., America’s E/ectric Uti/ities: 29 Energy Information Administration, De/ays and Past, Present, and Future,, Public Utilities Reports, Cancellations of Coal-Fired Generating Capacity, Inc. and Merrill Lynch, Pierce, Fenner & Smith, JUIY 1983. Inc., 1983. 30 Energy Information Administration, EfLects of New 46. Hyman, Leonard, S., “Testimony-Congressional P/ants and Equipment Expenditures on Electricity Hearings on Nuclear Energy, October 26, 1981 ,“ Prices, July 1982. Merrill Lynch, Pierce, Fenner & Smith, Inc., 1981. 78 . Nuclear Power in an Age of Uncertainty

47. Hyman, Leonard S., and Kelley, Doris A., The Fi- 61. Mooz, William E., Cost Analysis of LWR Power- nancial Aspects of Nuclear Power: Capital, Credit, p/ants, (Santa Monica, Calif.: Rand Corp., 1978). Demand and Risk, presentation to the American 62. Muhs, William F., and Schauer, David A., “State Nuclear Society 1981 Winter Meeting, San Fran- Regulatory Practices With Construction Work in cisco, Dec. 3, 1981. Progress: A Summary,” Public Utilities Fortnight- 48. ICF, Inc., An Ana/ysis of the E/ectric Uti/ity /ndus- /y, Mar. 27, 1980. try in the United States: An Assessment of the Na- 63. Myers, Stewart C,, et al., Inflation and Rate of Re- tional Residential Regional Cost and Rate Impacts turn Regulation unpublished April 1981. of Restoring Uti/ity Financia/ F/eakh, prepared for 64. Navarro, Peter, Long-Term Consumer Impacts of the DOE Electricity Policy Project, winter 1982/83. Electricity Rate Regulatory Po/icies, for DOE Elec- 49. ICF, Inc., An Assessment of the Economic and tricity Policy Project, January 1983. Rate Impacts of Building Additional Powerplants, 65. New England Electric System, NEESPLAN, Octo- report to the DOE Electricity Policy Project, fall ber 1979. 1982. 66. North American Electric Reliability Council, 12th 50( Johnson, Craig R., “Why Electric Power Growth Annual Review of Overall Reliability and Ade- Will Not Resume,” Public Utilities Fortnightly, quacy of the North American Bulk Power Systems, Apr. 14, 1983. August 1982. 51. Kaufman, Alvin, Do We Really Need All Those 67. North American Reliability Council, Ten Year Re- E/ectric P/ants? Congressional Research Service, view 1971- 1980: Report on Equipment Avail- August 1982. ability. 52. Kelley, Doris A., E/ectric Uti/ity /ndustry: Nuc/ear 68. North American Electric Reliability Council, E/ec- Powerplants–Another Look, Merrill Lynch, tric Power Supply and Demand 1982-1991, Pierce, Fenner & Smith, May 1982. August 1982. 53. Kelley, Doris A., E/ectric Uti/ity /ndustry-Nuc/ear 69. Letter from a staff member of the North Ameri- Power, Annual Powerp/ant Review, Merrill Lynch, can Reliability Council to the Office of Technol- May 1983. ogy Assessment, October 1982. 54. Kolbe, A Lawrence, “Inflation-Driven Rate Shocks: 70. North American Reliability Council, E/ectric Power The Problem and Possible Solutions,” Pub/ic Uti/i- Supply and Demand 1983-1992, July 1983. ties Fortnightly, Feb. 17, 1983. 71. Office of Technology Assessment, Energy Eficien- 55. Komanoff, Charles, Powerp/ant Cost Escalation: cy of Bui/dings in Cities, OTA-E-168, March 1982. Nuclear and Coal Capital Costs, Regulation and 72. Office of Technology Assessment, /ndustria/ and Economics (New York City: Komanoff Energy As- Commercial Cogeneration, OTA-E-1 92, February sociates, 1981 ), republished by Van Nostrand 1983. Reinhold, 1982. 73. Office of Technology Assessment, Technology and 56. Komanoff, Charles, letter dated July 13, 1983, in- Stee/ /ndust~ Competitiveness, OTA-M-122, June corporating analytical results to be published in 1980. 1984. 74. Olds, F. C., “A New Look at Nuclear Power 57. Long, John D., Nuclear Propetty Insurance: Status Costs,” Power Engineering, February 1982. and Outlook, report for the U.S. Nuclear Regula- 75. O’Neill, John, “Rate Shock: The States Map Out tory Commission, NUREG-0891, May 1982. Phase-Ins of Raises,” Nuc/ear /ndustry, June 1983. 58. Marlay, R., Office of Planning and Analysis, U.S. 76. Peal, Lewis J., NERA, “The Economics of Nuclear Department of Energy. Analysis based on data Power,” presented at NERA Conference on “The from the Board of Governors of the Federal Re- Economics of Nuclear Power: Should Construc- serve System, the U.S. Department of Commerce, tion Proceed?” New York, June 3, 1982. Published the Energy Information Administration and on “in- by Atomic Industrial Forum Public Affairs and industrial Energy Productivity, 1954-1980,” Massa- formation Program. chusetts Institute of Technology, Cambridge, 77. Putnam, Hayes & Bartlett, Inc., Alternative Energy Mass. Futures: A Review of Low Energy Growth Fore- 59. Merrill Lynch, Pierce, Fenner & Smith, Inc., Elec- casts and Least-Cost Planning Studies, a report for tric Utilities Utility Research Group, Financia/ the Department of Energy, November 1982. Statistics, Dec. 7, 1983. 78 Reilly, John, and Domingos, Alvaro, Extended 60. Messing, Mark, Nuclear Powerplant Construction Summary: Electric Powerplant Productivity Re- Costs, contractor report for the Office of Technol- /ated to P/ant Avai/abi/ity, Pennsylvania Public ogy Assessment, March 1983. Utilities Commission, December 1980. Ch. 3—The Uncertain Financjal and Economjc Future ● 79

79, Resources for the Future, Inc., F%ce Elasticities of 86, Stevenson, J. D., and Thomas, F. A., Se/ected Re- Demand for Energy: Evacuating the Estimates, review of Foreign Licensing Practices for Nuclear port for the Electric Power Research Institute, prin- Powerp/ants, for the Nuclear Regulatory Commis- cipal author Douglas Bohi, September 1982. sion (N UREG/CR-2664), April 1982. 80. Resources Planning Associates Co., The Effects of 87. S. M. Stoner Corp., Nuclear Supply Infrastructure Three Nuclear Powerplants on Future Electricity Viabi/ity Study, for the Argonne National Labora- Prices, a report for the DOE Electricity Policy Proj- tory, contract No. 31-109-38-6749, November ect, February 1983. 1982. 81. Schmidt, Philip, analysis for forthcoming report, 88. Streiter, Sally Hunt, “Trending the Rate Base” and Electricity and Industrial Productivity: A Technical “Avoiding the Money Saving Rate Increase,” two and Economic Perspective, to be published by of a three part series in Pub/ic Uti/ities Fortnight- EPRI in 1984. /y, May 13, 1982, and June 24, 1982. 82. Science Applications Inc., Case Studies of E/ectric 89. Strom, David, The Potentia/ for Reducing E/ectric Generating Plant Demographics: Efficiency and Utility Oil Consumption During 1985, unpub- Avai/abi/ity Enhancements, for the Department of lished working draft for OTA report on Technol- Energy, January 1983. ogies for Response to a Major Oil Disruption, Dec. 83. Searl, Milton, and Starr, Chauncy, “U.S. Generat- 6, 1982. ing Capacity Requirements for Economic Growth, 90. Theodore Barry & Associates, A Review of Ener- 1990-2000,” Pub/ic Uti/ities Fortnight/y, Apr. 29, gySupp/y Decision Issues in the U.S. Electric Uti/i- 1982. ty Industry, 1982. 84< Solar Energy Research Institute, A New Prosperi- 91. Uhrig, Robert, testimony in March 1983 before the ty: Buildinga Sustainable Energy Future, Commit- House Committee on Science and Technology tee Print of the Committee on Energy and Com- and telephone communication with OTA staff, Oc- merce, April 1981. tober 1983. 85. Southwest Energy Associates, Inc., The Financia/ 92 United Engineers & Constructors, Inc., “Regula- Condition of the Electric Utility Industry and Pos- tory Reform Case Studies,” August 1982. sibilities for Improvement, a report for the Edison Electric Institute, March 1982.

Tennessee Authority Boiling water reactor vessel being hoisted into a containment building at Browns Ferry nuclear plant Contents

Page Introduction ...... 83 Some Basics in Nuclear Powerplant Design ...... 84 The Safety and Reliability of Light Water Reactors ...... 87 Overview of U.S. Reactors ...... 87 Safety Concerns ...... 87 Reliability Concerns ...... 88 Examples of Specific Concerns ...... 90 Lessons Learned From Operating LWRs ...... 93 Advanced Light Water Reactor Design Concepts ...... 94 Heavy Water Reactors ...... 96 The High Temperature Gas-Cooled Reactor ...... 99 Inherently Safe Reactor Concepts ...... 102 The Small Reactor ...... 105 The Standardized Reactor ...... 107 Conclusions ...... 107 Chapter 4 References ...... 109

Tables

Table No. Page 13. Comparisons of Lifetime Capacity Factors for U.S Reactors ...... 89 14. Unresolved Safety Issues ...... 91 15. World Comparison of Reactor Types ...... 96 16. World Power Reactor Performance ...... , ...... 97

Figures

Figure No. Page 19. Pressurized Water Reactor ...... 85 20. Boiling Water Reactor ...... 86 21. Comparison of Fossil Units (400 MWe and Above) to All Nuclear Units ...... 89 22. Heavy Water Reactor ...... 98 23. Schedules for Alternative Reactors ...... 99 24. High Temperature Gas-Cooled Reactor ...... 100 25. Comparative Fuel Response Times...... 101 26, Process-Inherent Ultimately Safe Reactor ...... 104 27. Nuclear Reactor Availability ...... 106 Chapter 4 Alternative Reactor Systems

INTRODUCTION

Nuclear power in the United States achieved to investigate designs of alternative reactors that some remarkable successes in its early years and have different and potentially desirable charac- experienced dramatic growth in the late 1960’s teristics. It is possible that a new design could im- and early 1970’s. While this rapid growth was prove safety and reliability at an acceptable cost seen as a measure of the success of the technol- and within a reasonable timeframe. ogy, in retrospect it may have been detrimental. This provides the basic technical reason for re- As discussed in chapter 5, the size and complexity evaluating current nuclear technology as em- of reactors increased rapidly and there was little bodied in LWRs. It is important to question, opportunity to apply the experience gained from however, the justification for actual changes to older plants to the design of newer ones. In ad- the current system. Are there any indications that dition, the regulatory framework was incomplete the current generation of LWR is less than ade- when many of the plants were designed. As new quately safe or reliable? The public appears to be regulations were formulated, the designs had to increasingly skeptical that nuclear reactors are be adjusted retroactively to accommodate to good neighbors. As discussed in chapter 8, more changing criteria. With the rush of construction than half of those polled expressed the belief that in the mid-l 970’s, it was difficult to fully integrate reactors are dangerous. The same percentage of these new requirements into the original designs; the public opposes the construction of new hence some portions of the reactor designs plants. While this is not an absolute measure of emerged as a patchwork of nonintegrated and the adequacy of today’s reactors, it does reflect often ill-understood pieces. a growing concern for their safety. Several changes in design requirements have The nuclear utilities also have assessed the cur- had far-reaching effects in today’s reactors, even rent reactors in view of their special needs and though they were not originally expected to have interests. While they do not believe that LWRs such an impact. For example, new criteria on fire are seriously flawed, the utilities have expressed protection in nuclear powerplants have spawned a desire for changes that would make plants eas- new features and systems to prevent, contain, ier to operate and maintain and less susceptible and mitigate fires. This led to greater separation to economically damaging accidents (1 3). Some of safety systems, changes in cable-tray design, movement has already been initiated within the requirements for more fire-resistant materials, and nuclear industry in response to utility needs. Most changes in civil structures to prevent the spread of these efforts focus on evolutionary changes to of fires. Clearly, these modifications can have the current designs and thus represent normal ramifications for other plant systems. Other reg- development of LWR technology. ulatory actions concerning seismic design, decay heat removal systems, and protection of safety The increasing levels of concern for safety systems from other equipment failures have also among the public and the utilities has contributed had extensive impacts. to an interest in safety features that are inherent to the design of the reactor rather than systems A fresh look at the design of light water reac- which rely on equipment and operators to func- tors (LWRs) could be useful if it more fully in- tion properly. The emphasis on inherent safety tegrated the cumulative changes of the past and is reflected to some extent in evolutionary designs reexamined the criteria that stimulated those for LWRs, and to a much greater degree in inno- changes. In addition, a new design could incor- vative designs of alternative reactors. In this chap- porate analytical techniques and knowledge that ter, LWRs as well as several proposed alternatives have been acquired since the original designs first will be examined and their relative advantages were formulated. In fact, it could be beneficial and disadvantages assessed.

83 84 ● Nuclear Power in an Age of Uncertainty

SOME BASICS IN NUCLEAR POWERPLANT DESIGN

To assess the safety and reliability of current materials, and no single combination has reactors and compare them with alternative de- emerged as being clearly superior to the others. signs, it is important to understand the basic prin- Several designs have been developed for pro- ciples involved in generating power with nucle- ducing electricity commercially. The most com- ar technology. At the center of every nuclear mon reactors are known as light water reactors, reactor is the core, which is composed of nuclear which use ordinary water as both coolant and fuel. Only a few materials, such as uranium and moderator. LWR fuel is slightly enriched uranium, plutonium, are suitable fuels. When a neutron in which the percentage of fissionable material strikes an atom of fuel, it can be absorbed. This has been increased from its naturally occurring could cause the nucleus of the heavy atom to be- value of 0.7 percent to about 3 or 4 percent. After come unstable and split into two lighter atoms enrichment, the fuel is shaped into ceramic known as fission products. When this occurs, en- pellets of uranium dioxide and encased in long, ergy in the form of heat is released along with thin fuel rods made of a zirconium alloy. two or three neutrons. The neutrons then strike other atoms of fuel and cause additional fissions. Another commercially feasible reactor is the With careful design, the fissioning can be made heavy water reactor (HWR), which is moderated to continue in a process known as a chain reac- by heavy water and cooled by ordinary water. tion. The fuel in an HWR is similar in form and com- position to LWR fuel, but it need not be enriched A chain reaction can be sustained best in ura- to sustain a chain reaction. Another design is the nium fuels if the neutrons are slowed before they high temperature gas-cooled reactor (HTGR), strike the fissionable materials. This is done by which uses helium as a coolant and graphite as surrounding the fuel with a material known as a moderator. The HTGR can use uranium as a a moderator that absorbs some of the energy of fuel, but it usually is enriched to a greater con- the neutrons as they are released from the fission centration of fissionable material than found in process. Several different materials are suitable LWR fuel. The fuel form is very different from as moderators, including ordinary water, heavy LWR and HWR fuel, with the uranium shaped water, * and graphite. into small coated spheres, mixed with graphite The heat from the fission process is removed to form fuel rods, and then inserted into hex- from the core by a continuous stream of fluid agonal graphite blocks. called the primary coolant. The reactors exam- In addition to selecting a fuel, moderator, and ined in this chapter use water or helium as the coolant, reactor designers also must devise a coolant, although other fluids have been consid- means to transfer the heat from the core to the ered. The heat in the coolant can be used directly turbines. In the United States, two different steam to produce electricity by driving a turbine-gen- cycles have been developed for LWRs. The pres- erator, or it can be transferred to another fluid surized water reactor (PWR) shown in figure 19 medium and then to a turbine-generator. Both maintains its primary coolant under pressure so methods have been used effectively in U.S. nu- that it will not boil. The heat from the primary clear powerplants. system is transferred to a secondary circuit There are many possible combinations of fuel, through a steam generator, and the steam pro- coolant, and moderator that can be used in the duced there is used to drive a turbine. design of nuclear reactors. There are advantages The second type of LWR that is in commercial and disadvantages associated with the various use is the boiling water reactor (BWR), shown in figure 20. It eliminates the secondary coolant cir- *A molecule of light water is made from one atom of oxygen and cuit found in a PWR. In the BWR, the heat from two atoms of the lightest isotope of hydrogen. By contrast, a molecule of heavy water is made with the isotope of hydrogen called the core boils the coolant directly, and the steam deuterium, which has twice the mass. produced in the core drives the turbine. There Ch. 4—Alternative Reactor Systems ● 85

Figure 19.—Pressurized Water Reactor

Generator

Containment building

SOURCE: “Nuclear Power from Fission Reactors, ” U.S. Department of Energy, March 1982 is no need for a heat exchanger, such as a steam vent minor events from developing into major generator, or for two coolant loops. In addition, problems. This is accomplished in part by incor- since more energy is carried in steam than in porating inherent features into the design to en- water, the BWR requires less circulation than the sure stable and responsive operation. For exam- PWR. ple, the physics of the core dictates that most reactors will internally slow down the fission proc- The two LWRs described above can be used ess in response to high coolant temperatures, and to illustrate another crucial part of reactor design. thus dampen the effects of problems in remov- Since nuclear reactors produce highly radioactive ing heat from the core. Both PWRs and BWRs materials as byproducts of the fission process, it have been designed to respond in this way. is essential that the design incorporates enough safety features to ensure the health and safety of Other features, known as engineered safety sys- the public. During normal operation, the radio- tems, operate in parallel with, or as a backup to, active materials are safely contained within the the inherent physical safety features. They are de- fuel rods and pose no threat to the public. The signed to ensure that the chain reaction is inter- concern is that during an accident the fuel may rupted promptly if there is a problem in the plant become overheated to the point that it melts and and to remove heat from the core even under releases the fission products that accumulate dur- extreme circumstances. This is necessary because ing normal operation. radioactive decay continues to produce heat long after the reactor has been shut down. If decay Safety is designed into a nuclear reactor on heat is not removed, the core can overheat to several levels. First, every effort is made to pre- the point of melting. In the event that the shut-

25-450 0 - 84 - 7 : QL 3 86 • Nuclear Power in an Age of Uncertainty

Figure 20.—Boiling Water Reactor

nser

Containment vessel

SOURCE: “Nuclear Power from Fission Reactors,” U S. Department of Energy, March 1982. down or decay heat removal systems fail, addi- They are dependent on human action only inso- tional safety systems prevent the escape of fission far as they must be designed, constructed, and products to the atmosphere. maintained to function correctly.

Rapid interruption of the nuclear chain reac- The final step in the design for the safety of a tion is accomplished by inserting control rods nuclear powerplant is to incorporate features that which contain neutron-absorbing boron into the prevent the release of fission products in the core. The control system is designed to shut down event of a fuel-melting accident. This is done the reactor automatically in the event that ab- using the concept of “defense in depth,” that is, normal conditions develop in the core or primary providing successive barriers that radioactive ma- coolant system. Even after the chain reaction is terials must breach before endangering the pub- interrupted, however, the coolant must continue lic. The barriers in LWRs are the fuel cladding, to circulate to remove decay heat. If the coolant the heavy steel of the reactor pressure vessel, and pressure drops in a BWR or PWR–indicating that the thick concrete of the containment building some of the coolant has been lost from the pri- that encloses the pressure vessel and other com- mary system—the core is automatically flooded ponents in the coolant system. by an emergency core cooling system (ECCS). If the secondary cooling system fails in a PWR, an These examples necessarily oversimplify the auxiliary feedwater system is designed to take complex designs and interactions of safety sys- over. Other backup cooling systems in these tems. Many safety systems play a role in the plants include high- and low-pressure injection routine operation of the plant as well. This sam- pumps and spray systems. These safety systems pling serves as background for the subsequent are designed to operate automatically, with no discussions of safety features of LWRs and of alter- requirement for action by the plant operators. native designs. Ch. 4—Alternative Reactor Systems ● 87

THE SAFETY AND RELIABILITY OF LIGHT WATER REACTORS Overview of U.S. Reactors Safety Concerns Of the 84 nuclear reactors with operating li- The occurrence of several widely publicized censes in the United States today, about two- accidents such as those at Three Mile Island and thirds are pressurized water reactors. They are Browns Ferry nuclear plants have underscored offered by three companies—Babcock & Wilcox the potential for a catastrophic accident. These Co., Combustion Engineering, Inc., and Westing- accidents shook some of the confidence in our house Electric Corp. The remaining reactors (with understanding of nuclear reactors. For example, the exception of one HTGR) are boiling water the scenario that unfolded at Three Mile Island reactors, sold by General Electric Co. These four had not been stressed prior to the accident: it in- companies all supply the nuclear steam supply volved the loss of coolant through a small leak system (NSSS), or the nuclear-related compo- in a pressure relief valve, whereas safety analysis nents of the reactor. The balance of the plant con- had previously concentrated on large loss-of-cool- sists of such items as the turbine-generator, the ant accidents. Most studies of these serious ac- auxiliary feedwater system, the control room, and cidents have faulted the plant operators more the containment building. The balance of plant than the reactor hardware (1 O), which indicates design typically is supplied by an architect-engi- that LWR designs are not as forgiving of human neering (AE) firm, any one of which might team error as they might be. up with a vendor to provide a reactor plant that Safety concerns also arise because nuclear meets the needs of a particular utility at a specific powerplants have encountered hardware mal- site. So far, no completely standardized plant functions in virtually every system, including con- design has emerged, although some convergence trol rods, steam generators, coolant pumps, and has occurred among the designs of each nuclear fuel rods. The majority of these hardware prob- steam system vendor. There is still a great deal lems have been resolved by retrofits, changes of difference among the designs of similar com- in methods of operation, and redesign. Some ponents (e.g., steam generators) and system con- problems are expected as a new reactor matures, figurations. This is not surprising considering the but many of the LWR problems have persisted. various combinations of vendors and AE firms Others continue to surface, some because of the that have been involved in powerplant design. intense scrutiny of plants following the Three Mile Furthermore, the utilities themselves may custom- Island accident and others because of the aging ize a reactor design to meet specific site require- of the earlier reactors. Most of the difficulties ments. probably have technically feasible solutions, but it is not always clear that they would be cost ef- Even without the benefits of a standardized de- fective to implement. Meanwhile, the discovery sign, the LWRs that have operated in the United of new problems and the slow resolution of old States for more than 20 years have had good safe- ones continues to erode confidence in the safety and reliability records. There never has been ty of LWRs. an accident involving a major release of radioactivity to the environment, and the operating Confidence in LWRs might be enhanced if performance, while not spectacular, has been there was an objective standard for judging the comparable to that of coal-fired powerplants. Still, safety of these plants. As a step in this direction, doubts linger about both the safety and reliabili- the Nuclear Regulatory Commission (NRC) has ty of these LWRs. This section examines the rea- proposed a set of qualitative and quantitative safe- sons for such concerns, including particular fea- ty goals for nuclear powerplants on a 2-year trial tures of these reactors that contribute to concern. basis (4). These safety goals will provide a means 88 • Nuclear Power in an Age of Uncertainty

for answering the question, “How safe is safe malfunctions or threatens to do so, the plant must enough?” be shut down for a lengthy and often expensive period of maintenance. On the other hand, There is a fundamental problem with specify- chronic reliability problems are likely to indicate ing standards for safety: there is no technique for or contribute to fundamental difficulties that quantifying the safety of a nuclear powerplant in could reduce safety. an objective and unambiguous way. One attempt to define nuclear safety is probabilistic risk assess- The reliability of LWRs is easily quantifiable, in ment (PRA), which outlines sequences of events contrast to the difficulties in defining safety. De- that could lead to accidents and then assigns tailed data on reactor performance have been probabilities to each basic event (12). PRA is be- collected since the beginning of the nuclear era, coming a useful tool for such tasks as compar- and they can be analyzed to determine trends. ing certain design options in terms of their safe- Two measures of performance are commonly ty impact. However, the technique is still in its used—availability and capacity factor. The avail- infancy and the results vary widely from one prac- ability is defined as the percentage of a time pe- titioner to the next. The variations occur because riod during which the reactor was available for the users of PRA must put in their own assump- operation (whether or not it was actually in serv- tions about factors contributing to accidents and ice). The capacity factor is the ratio of the actual their probabilities of occurrence. More research amount of electric generation to the total theo- is required to establish reasonable and standard retical output of the plant during the same time assumptions and to refine the process of assess- period. Each of these quantities has some drawing risk. backs as a measure of plant reliability: the capacity factor is affected by the demand for electrici- Another important component of safety anal- ty and the plant availability is insensitive to the ysis is the consequence of an accident. This capability of the plant to operate at full power. depends on the amount of radioactive material Since nuclear powerplants usually are base- that can be released to the environment follow- Ioaded, the capacity factor is generally a better ing a nuclear reactor accident, otherwise known measure of reliability. Both capacity and availa- as the nuclear source term. Recent findings indi- bility are shown in figure 21 as a function of time cate that the source terms now used in regula- for all years from 1971 through 1980 (’17). To pro- tion and risk analysis may overestimate the mag- vide a basis for comparison, reliability records are nitude of potential fission product releases (5). also shown for coal-fired plants larger than 400 Only further analysis can tell whether reductions megawatts electrical (MWe). It can be seen that in the source terms can be fully justified, and, if the average availability for the two types of plants so, the magnitude of the appropriate reduction has been nearly identical at about 69 percent. The for each fission product and for each accident average capacity factor for nuclear plants over scenario. Modeling and analysis programs are the same time period was 60 percent, which was now being conducted by NRC and by the Elec- 3 percentage points better than for coal. Thus, tric Power Research Institute (EPRI), the American nuclear plants operate reliably enough compared Nuclear Society, and by the Industry Degraded with their closest counterparts, even though the Core Rulemaking Program. These studies should average performance has not been as outstanding eventually produce realistic estimates of fission as anticipated by the original nuclear powerplant product releases, but the task is complex and like- designers. ly to be lengthy. it is instructive to reexamine performance data Reliability Concerns for groups of reactors as well as the industry as a whole. Capacity factors are shown for each Reliability and safety concerns are closely reactor type and vendor in table 13 (27). When related, since the same factors that create con- comparing the data on a lifetime or cumulative cern about the safety of LWRs also raise ques- basis, it can be seen that there are only slight dif- tions about their reliability. If a safety system ferences among reactor vendors or types. It also Ch. 4—Alternative Reactor Systems • 89

Figure 21.—Comparison of Fossil Units (400 MWe and Above) to All Nuclear Units Equivalent availability Capacity factor 100 1 100 I

I I 1 I I I J ‘W71 72 73 74 75 76 77 78 79 80 71 72 73 74 75 76 77 78 79 80 Year of operation Year of operation

Year

Equivalent availability 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 Avg.

Fossil 72.2 67.8 73.2 68.3 67.7 67.2 66.7 67.4 68.0 69.5 68.6

Nuclear 71.6 73.7 74.4 64.0 66.7 65.6 69.8 74.3 64.8 64.5 68.9 b

Year

Capacity factor 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 Avg.

Fossil 60.0 59.6 62.4 57.7 57.3 69.4 57.4 55.3 56.3 59.5 56.6

Nuclear 58.9 63.1 64.1 53.6 59.4 59.0 65.2 68.3 52.9 59.5 59.8

SOURCE. National Electric Reliability Council, “Ten Year Review 1971-1980 Report on Equipment Availability “

Table 13.—Comparison of Lifetime Capacity are as low as 40 percent while those of the best Factors for U.S. Reactors are as high as 80 percent. Lifetime capacity factor The hardware problems discussed above have Reactor type: contributed to low availabilities in some plants. PWR ...... 61 .OO/o These and other hardware problems have been BWR ...... 58.70/o Reactor supplier: responsible for lengthy periods of downtime as Westinghouse ...... 62.7% discussed in detail in volume Il. It is concluded Combustion Engineering ...... 59.70/0 there that most of the these problems have been Babcock & Wilcox...... 57.0 ”/0 General Electric Co...... 58.7% or soon will be resolved (27). All plants ...... 60.2% Despite signs of progress, LWRs still are not SOURCE: A. Weitzberg, et al , “Reliability of Nuclear Power Plant Hardware — Past Performance and Future Trends,” NUS Corp., Jan.15, 1983. operating trouble-free. The steam generators in several plants have degraded to the point that it should be noted that these averages can mask has been necessary to replace them. This repair substantial spreads in the performance of in- is estimated to cost between $60 million and $80 dividual plants. As discussed in chapter 5, the million in addition to the cost of purchasing re- cumulative capacity factors of the worst plants placement power. Other plants may have to un- 90 . Nuclear Power in an Age of Uncertainty dertake expensive retrofits or modify operation Examples of Specific Concerns to mitigate concerns over pressurized thermal shock (26). Since 1978, NRC has been required by Con- gress to prepare a list of generic reactor problems. Another impediment to achieving high availa- This list is revised annually to reflect new infor- bility is the stream of retrofits that has followed mation and progress toward resolution. Each time the accident at Three Mile Island. The Three Mile a new safety issue is identified, NRC assesses the Island action plan contains about 180 require- need for immediate action. In some cases, ac- ments for changes in operational plants; these tion such as derating (reducing the approved changes, of course, could not be incorporated operating power) certain reactors, is taken to into the basic powerplant design, but had to be assure public health and safety. in other cases, added to existing systems. This type of retrofit- an initial review does not identify any immediate ting is seen as a problem by both the nuclear in- threat to the public, and further research is con- dustry as well as its critics since it introduces the ducted. Many generic safety issues have been possibility of adverse safety consequences. In fact, resolved and removed from NRC’s list of signifi- in some cases, new requirements might reduce cant safety items (26). rather than enhance safety. This could happen if unanticipated interactions arise or if there is an Table 14 summarizes the 15 most important un- inadequate understanding of the system the re- resolved safety issues as determined by NRC in quirement is intended to improve. 1982. A few of the items on that list will be ex- The revision in NRC requirements for seismic amined here as examples of the types of concerns restraints on piping is often cited as an example that remain about LWRs and some of the factors of retrofit problems. The restraints in nuclear preventing their resolution. powerplants are designed to preserve the integrity One of the most widely publicized safety prob- of pipe by limiting vibrations even if an earth- lems is the potential in PWRs for fracture of the quake should occur. Many pIant operators and reactor vessel from pressurized thermal shock. designers complain that these restraints are ex- Reactors are designed to be flooded with relative- pensive to install and that they hold the pipes too ly cold water if a loss of coolant accident occurs. rigidly to allow for thermal expansion. Further- The sudden temperature differential causes sur- more, some critics of the current seismic require- face strains, known as thermal shock, on the thick ments feel that piping actually may be more metal wall of the reactor vessel and imposes prone to failure in an overconstrained system. severe differential stress through the vessel wall. These critics assert that today’s requirements for While plant designers have understood and ac- seismic restraints result from an attempt to make counted for this phenomenon for ‘years, they it easier to analyze conditions in plants rather have only recently discovered that two other fac- than from an identifiable need (l). tors may make the effect more acute than antici- On the balance, retrofits probably have im- pated. One is that the emergency cooling system proved the safety of operating nuclear power- is likely to be actuated following a small-break plants. in fact, one assessment of plants before accident (e.g., the one at Three Mile Island) when and after the Three Mile Island retrofits concludes the reactor vessel is still highly pressurized. In that the probability of an accident has been such a situation, the stresses due to thermal shock reduced by a factor of 6 in PWRs and by a factor would be added to those due to internal pressure. of 3 in BWRs, with the core melt probability for The second factor is that the weld and plate ma- PWRs now only slightly higher than for BWRs. terials in some older reactor vessels are becom- These improvements are attributed primarily to ing brittle from neutron exposure faster than had higher reliability of feedwater systems and regu- been expected. Such embrittlement increases the latory and inspection procedures that reduce the vulnerability of the vessel to rupture following probability of human error (19). pressurized thermal shock. While the possibility Ch. 4—Alternative Reactor Systems • 91

Table 14.—Unresolved Safety Issues issue/Descridion issue/DescrilMion Water hammer: Since 1969 there have been over 150 reported Containment emergency sump performance: Following a loss incidents involving water hammer in BWRs and PWRs. The of coolant accident in a PWR, water flowing from a break incidents have been attributed to rapid condensation of in the primary system would collect on the floor of con- steam pockets, steam-driven slugs of water, pump startup tainment. During the injection mode, water for core cool- with partially empty lines, and rapid valve motion. Most of ing and containment spray is drawn from a large supply the damage has been relatively minor. tank. When the tank water is depleted, a recirculation mode Steam generator tube integrity: PWR steam generators have is established by drawing water from the containment floor shown evidence of corrosion-induced wastage, cracking, or sump. This program addresses the safety issue of the reduction in tube diameter, and vibration-induced fatigue adequacy of the sump and suppression pool in the recir- cracks. The primary concern is the capability of degraded culation mode. tubes to maintain their integrity during normal operation Station blackout: The loss of A.C. power from both off site and under accident conditions with adequate safety and onsite sources is referred to as a station blackout. In margins. the event this occurs, the capability to cool the reactor core Mark I containment long-term program: During a large-scale would be dependent on the avail ail it y of systems which do testing program for an advanced BWR containment system, not require A.C. power supplies and the ability to restore new suppression pool loads associated with a loss of A.C. power in a timely manner. There is a concern that a coolant accident were identified which had not been ex- station blackout may be a relatively high probability y event plicitly included in the original design of the Mark I con- and that this event may result in severe core damage. tainment systems. In addition, experience at operating Shutdown decay heat removal requirements: Many im- plants has identified other loads that should be recon- provements to the steam generator auxiliary feedwater sidered. The results of a short-term program indicate that, system were required after the accident at Three Mile for the most probable loads, the Mark I containment system Island. However, an alternative means of decay heat would maintain its integrity and functional capability. removal in PWRs might substantially increase the plants’ Reactor vessel material toughness: Because the possibility capability to deal with a broader spectrum of transients and of pressure vessel failure is remote, no protection is pro- accidents and thus reduce the overall risk to the public. vided against reactor vessel failure in the design of nuclear Seismic qualification of equipment in operating plants: The facilities. However, as plants accumulate service time, design criteria and methods for the seismic qualification neutron irradiation reduces the material fracture toughness of equipment in nuclear plants have undergone significant and initial safety margins. Results from reactor vessel sur- change. Consequently, the margins of safety provided in veillance programs indicate that up to 20 operating PWRs existing equipment to resist seismically induced loads may will have materials with only marginal toughness after com- vary considerably and must be reassessed. paratively short periods of operation. Safety implications of control systems: It is generally believed Fracture toughness of steam generator and reactor coolant that control system failures are not likely to result in the pump supports: Questions have been raised as to the loss of safety functions which could lead to serious events potential for Iamellar tearing and low fracture toughness or result in conditions that cannot be handled by safety of steam generator and reactor coolant pump support ma- systems. However, indepth plant-by-plant studies have not terials in the North Anna nuclear powerplants. Since similar been performed to support this belief. The purpose of this materials and designs have been used on other plants, this program is to define generic criteria that may be used for issue will be reassessed for all PWRs. plant-specific reviews. Systems interactions in nuclear powerplants: There is some Hydrogen control measures and effects of hydrogen burns question regarding the interaction of various plant systems, on safety equipment: Reactor accidents which result in both as to the supporting roles such systems play and as degraded or melted cores can generate large quantities of to the effect one system can have on other systems, par- hydrogen and release it to the containment. Experience ticularly with regard to the effect on the redundancy and gained from the accident at Three Mile Island indicates that independence of safety systems. more specific design provisions for handling large quan- Determination of safety relief valve pool dynamic loads and tities of hydrogen releases maybe appropriate, particular- temperature limits for BWR containment: Operation of BWR ly for smaller, low-pressure containment designs. primary system pressure relief valves can result in Pressurized thermal shock: Neutron irradiation of reactor hydrodynamic loads on the suppression pool retaining pressure vessel weld and plate materials decreases frac- structures or structures located within the pool. ture toughness. This makes it more likely that, under cer- Seismic design criteria: While many conservative factors are tain conditions, a crack could grow to a size that might incorporated into the seismic design process, certain threaten vessel integrity, aspects of it may not be adequately conservative for all plants. Additional analysis is needed to provide assurance that the health and safety of the public is protected, and if possible, to reduce costly design conservatism.

SOURCE: US. Nuclear Regulatory Commission, “Unresolved Safety Issues Summary,” Aug. 20, 1982, 92 ● Nuclear Power in an Age of Uncertainty

of a pressure vessel failure is peculiar to only a conditions. NRC estimates that steam generator few older reactors, it is of concern that such a degradation has accounted for about 23 percent potentially severe condition was not recognized of the non refueling outage time in nuclear reac- sooner. Measures to mitigate the problem of pres- tors. The corrosion has been attributed to a com- surized thermal shock include reducing the neu- bination of inappropriate water-chemistry treat- tron flux near the outer walls, increasing the ment and poor quality materials in the steam gen- temperature of emergency cooling water, heating erators. The result has been wastage, cracking, the reactor vessel at very high temperatures to reduction in tube diameter, separation of clad- reduce brittleness, and derating the plant (1 5,27). ding from the tube sheet, and deterioration of the metal plates that support the tubes inside the gen- BWRs are not susceptible to pressurized ther- erators. The severity of the problem varies with mal shock, but they have been plagued by a steam generator design and water treatment problem known as intergranular stress corrosion methods. Much of the corrosion has been cracking. This problem, which involves defects brought under control, but plant operators con- in the reactor coolant piping, is now listed by tinue to inspect their steam generators regularly NRC as resolved, but it continues to be the sub- and plug the degraded tubes when necessary. ject of extensive and costly research programs Operators of several nuclear units–Surry 1 and throughout the industry. Most of the service pip- 2 in Virginia and Turkey Point 3 and 4. in Florida— ing sensitive to such cracking has been designed have already had to replace their steam gener- out of the later BWRs, but reactors currently ators. Other units may face expensive and lengthy under construction will have recirculation loop overhauls in the future. The fatigue cracking ap- piping with some susceptibility to this phenom- pears to result from flow-induced vibrations, and enon (15,27). resolution of this problem may require design Another problem on the list of unresolved safe- modifications. ty issues deals with the corrosion or fatigue crack- Two general safety issues deal with uncertain- ing of steam generator tubes (15). This is of con- ties over the behavior of the complex systems cern because these tubes separate the primary found in nuclear powerplants. One concern is coolant from the secondary system, and there is that the interactions among the various systems some question whether degraded tubes will be are not fully understood and could contribute to able to maintain their integrity under accident an accident under some circumstances. in particular, it is possible that some system interactions

Photo credit: Atomic Industrial Forum These four steam generators are awaiting installation in a large pressurized water reactor. In this type of reactor, water flows through the core to remove heat and then through narrow tubes in the steam generator to transfer it to a secondary coolant loop. Ch. 4—Alternative Reactor Systems ● 93 could eliminate the redundancy or independence safety. While no single issue has been identified of safety systems. Another concern is that the fail- as a fatal flaw, neither is there a clear indication ure of a control system might aggravate the con- that current LWR technology is fully mature and sequences of an accident or prevent the operator stable. from taking the proper action. These concerns relate to the need to understand the entire reac- Lessons Learned From tor and its interrelated systems rather than to any Operating LWRs specific feature of its hardware. Resolving either of these concerns probably would require anal- The utilities operating LWRs have gained ysis on a plant-specific basis, but NRC is attempt- knowledge of both the strengths and weaknesses ing to identify generic criteria. Both of these issues of their plants. A recent survey of the utilities contribute to a larger question: are LWR designs reflects the concerns about safety and reliability more complex than necessary and could a sim- mentioned above and gives specific recommen- pler, but equally safe, reactor be designed? dations about features that might mitigate the problems (13). The following recommendations Another safety concern is that a nuclear plant were made: may develop a serious problem and fail to shut down automatically in an incident known as an ● safety and control systems in new LWRs “anticipated transient without scram. ” In such should be simpler and easier to operate, test, a situation, the emergency cooling system would and maintain. Utility personnel expressed have to remove not only the decay heat, but also preferences for passive and fail-safe charac- the heat generated by full-power operation. NRC teristics whenever possible; has removed this item from its list of generic safety ● safety systems should be separated from issues, but many critics feel that it continues to nonsafety systems and dedicated to single represent a valid concern. This issue was high- functions. plant operators worry that over- lighted by the discovery at the Salem plants in lapping functions may lead to adverse im- New Jersey that faulty circuit breakers prevented pacts in a complex accident scenario; the scram systems from activating automatically ● the response of existing plants to abnormal (16). occurrences should be slowed. Because of the low inventory of primary and secondary The resolution of several of the unresolved cooling water in current LWRs, the time be- issues may require adding new equipment to LWRs tween the start of a transient and the onset in operation or under construction. One of these of melting in the core can be short. If all safe- items is the provision for an alternative means of ty equipment works as designed, it is likely removing decay heat. NRC is examining whether that transients would be brought under con- an additional system might substantially increase trol quickly, regardless of the cause. If cer- the capability of the reactor to deal with a broader tain important safety components fail, melt- range of accidents or malfunctions. Another issue ing could begin after 40 minutes in a PWR. is whether the installation of a means to control The failure of all safety systems, which is not hydrogen is required to prevent the accumula- considered a plausible scenario, could cause tion of dangerous levels of the gas in the contain- core melting in less than 1 minute; ment vessel. These concerns are an outgrowth ● containment buildings should be larger to of lessons learned from the Three Mile Island ac- provide adequate space for maintenance. cident. Either measure discussed above would be Current structures do not allow enough very costly to retrofit on existing plants but might room to easily handle equipment that is be- be easily designed into a new plant. ing disassembled for maintenance; and These examples indicate that there are still un- ● the potential for retrofits on future plants resolved generic safety questions concerning the should be reduced as much as possible by LWR, and that the resolution of these issues will taking a fresh look at the LWR. Such an ef- involve a complex tradeoff between cost and fort should integrate the retrofits into a new 94 Ž Nuclear Power in an Age of Uncertainty

design rather than piece them on top of an standing of their behavior; complexity has been existing one. increased by the large number of safety-mandated retrofits; and current reactors are somewhat Both NRC’s list of unresolved safety issues and unforgiving of operator errors. Despite this less- the survey of utility executives provide concrete than-perfect record of the LWR, many in the in- examples of reasons for concern over the safety dustry are reluctant to abandon it. They argue and reliability of current plants. Existing LWRs that they have made appreciable progress along have serious, although resolvable, problems with the learning curve that would have to be repeated important hardware components; the interrelated with an alternative reactor concept. safety and control systems hinder a deep under-

ADVANCED LIGHT WATER REACTOR DESIGN CONCEPTS Improvements could be made to the current will be completed by 1984, and there are plans generation of LWRs by redesigning the plants to to initiate verification testing by 1986. Westing- address the safety and operability problems out- house is reviewing this design effort with NRC lined above. Furthermore, the entire design could so that the advanced PWR could be readily li- be integrated to better incorporate various censed in the United States. changes made to LWRs in the past decade. If such a redesign effort were successful, LWRs would The Westinghouse design focuses on reducing probably continue to be the preferred option in risks, improving daily operations, anal controlling the future. LWRs have the advantage of being fa- costs (8,11). It attempts to reduce economic risk miliar, proven designs with a complete infrastruc- to the owner and health risk to the public by in- ture to support manufacturing, construction, and corporating several new features. The coolant operation. piping has been reconfigured and the amount of water in the core has been increased to reduce Advanced LWR designs are being developed by both Westinghouse Electric Corp. and General the possibility that a pipe break could drain the Electric Co., and they should be available to U.S. primary coolant enough to uncover the core. Pro- utilities before any new reactors are ordered. tection against other accidents that could uncover General Electric is designing a new BWR that is the core has been provided by safeguarding an evolution of the most advanced plants cur- against valves that could fail in an open position. rently under construction. The newer reactor has Additional risk reduction efforts have been fo- been modified to enhance natural circulation of cused on the response of plant operators and sys- the primary coolant and hence improve passive tems following an accident, with the goal of requiring less operator action and more automatic cooling. This increases the ability of the coolant responses. system to remove decay heat in the event that the main circulation system fails. The design fur- Improvements in normal plant operations are ther provides for rapid depressurization of the another important feature of the Westinghouse primary system so that both low- and high-pres- advanced design. The reactor has been rede- sure pumps can supply water to the reactor ves- signed to operate 18 to 24 months on a single sel. This safety feature provides additional assurance that emergency coolant would be batch of fuel, rather than the current 12 month available in the event that primary coolant is lost cycle. The longer fuel cycle is made possible by (6). enlarging the diameter of the reactor and reducing the power density and neutron flux. This is Westinghouse Electric Corp. is developing an combined with a different way to moderate the advanced version of its PWR. The company has neutrons at the beginning of the cycle so that undertaken a joint program with Mitsubishi more plutonium is produced; the extra fuel is Heavy Industries, the Westinghouse licensee in burned at the end of the cycle when the fission- Japan. It is expected that the design development able uranium is depleted. Other efforts have been Ch. 4—Alternative Reactor Systems ● 95 made to reduce the amount of time required to refuel and inspect the reactor when it is shut down. The combined effect of these final two features is to increase the availability of the reactor, perhaps to as much as 90 percent. Efforts have also been directed toward increasing the reliability of the advanced PWR. For example, the steam generators have been redesigned so that there will be fewer stresses on the tubes and less potential for contaminants to collect. Other improvements have been made to increase the reliability of the nuclear fuel and its support structure. Operational improvements have been made in the area of occupational radiation exposure. The new plants have been designed for easy access to reduce worker exposure. In addition, radiation shielding has been added to areas where high radiation fields can be expected. Finally, ease of maintenance and repair have been fac- tored into the redesign of the steam generators; this should reduce the large occupational exposures that have resulted from steam generator maintenance in current PWRs. An area that is closely related to both safety and operation is the effort to increase the design margins of the advanced PWR. The increased amount of coolant in the core and the reduced core power density make the reactor less sensitive to upsets. Furthermore, the physics of the design dictates that the core will be even better at slow- ing the fission process in response to a rapid rise in temperature. The capital cost of the advanced PWR proposed by Westinghouse probably would be greater than that of current PWRs. An effort has been made to hold down the capital cost by sim- plifying fluid systems, making more use of multi- plexing, and completing at least three-quarters of the design before construction is initiated. Moreover, other features of the design should compensate for the increased capital costs, resulting in lifetime costs that should be comparable to those of today’s LWRs. Significant fuel savings are expected, with reductions in both uranium and enrichment requirements. This is due to the reduction in specific power, the more effective use of plutonium created within the core, 96 . Nuclear Power in an Age of Uncertainty

An intermediate step in the redesign of the the number and capacity of pumps in the emer- PWR has been taken by the Central Electricity gency-cooling system, and increasing the number Generating Board (CEGB) in Great Britain (22). of independent diesel generators to provide elec- CEGB did not redesign the basic PWR; rather, it trical power if the normal supply is interrupted. added safety features to the Standardized Nuclear These changes are likely to reduce the probability Unit Power Plant System (SNUPPS), which was of a core meltdown, and the probability for the developed and marketed in the United States in release of fission products is much lower still. the 1970’s. Most of the changes increase the sep- These safety improvements will not be inexpen- aration and redundancy of safety systems, such sive; the new design is estimated to cost 25 per- as adding a steel containment shell to the nor- cent more than the original SNUPPS plant. mal concrete containment structure, increasing

HEAVY WATER REACTORS

If the new designs for LWRs are perceived as Table 15.—World Comparison of Reactor Types being less than adequate to ensure safe and (150 MWe and larger) reliable operation, it is possible that alternative Average annual designs will become more attractive to utilities Number of load factor and investors. The HWR is a potential candidate Type of reactor reactors (percent) because it is the only other type of reactor that Pressurized heavy water reactor ...... 13 71 has been deployed successfully on a commercial Gas-cooled reactora scale. The HWR has been developed most ex- (Magnox) ...... 26 57 Pressurized (light) water tensively in Canada, where the CANDU (Canada reactor ...... 106 59 deuterium uranium) reactors produce all the nu- Boiling water reactor . . 57 60 clear-generated electricity. In addition, HWRs aThe natural. uranium gas-cooled reactors referenced here differ Significantly have emerged as competitors with LWRs in sev- from the HTGR discussed in this report. SOURCE: “Nuclear Station Achievement 19S3,” Nuclear Engmeerirrg /nterna- eral other nations, including India, Korea, and fiona/, October 1983, Argentina. The interest in this type of reactor originally derived from the effective way in which heavy The design of an HWR is somewhat similar to water moderates neutrons, with a resultant in- a PWR in that primary coolant transfers heat from crease in fuel economy when compared to LWRs. the core to a secondary coolant system via a There are also various inherent safety features of steam generator (3,6,21 ). In many other ways, the the HWR that make it an attractive alternative to design of an HWR differs significantly from that the LWR. in addition, the current generation of of a PWR. As implied by its name, heavy water HWRs has compiled an excellent operating rec- is used to moderate the neutrons generated in ord. In fact, the HWRs in operation worldwide the chain reaction. This is a more effective mod- have the most impressive reliability record of any erator than ordinary water, and so the core can commercial reactor type. As shown in table 15, be composed of less concentrated fissionable ma- HWRs operated at an average capacity factor of terial than in a PWR. As a result, the HWR can 71 percent in 1982, which far exceeds the records operate with unenriched or natural uranium. Na- of either PWRs or BWRs. Moreover, in 1982, 5 tions that have not developed enrichment tech- of the top 10 best performers internationally were nology perceive this as an advantage, but in the HWRs, even though this type of plant accounts United States uranium enrichment is readily avail- for only 5 percent of the total nuclear capacity. able. it is likely that U.S. utilities would elect to Both lifetime and annual capacity factors for the operate HWRs with uranium that is slightly en- world’s best power reactors are shown in table riched. On such a fuel cycle, the HWR would 16. require only 60 percent of the uranium used in Ch. 4—Alternative Reactor Systems • 97

Table 16.—World Power Reactor Performance A disadvantage of the pressure tube configura- (150 MWe and larger) tion is that the thin walls of the tubes restrict the Nominal temperature of the coolant more than the heavy rating Capacity steel pressure vessel of an LWR. Hence, the over- Reactor gross factor all efficiency of energy conversion is somewhat type Unit (MWe ) (percent) a less in an HWR. Efficiency is further limited in Annual PHWR Pickering 2 542 96 the HWR by the heat that is deposited in the PHWR Bruce 3 826 96 heavy water moderator. Current HWRs achieve PHWR Bruce 4 826 95 an efficiency of only 29 percent, while LWRs PHWR Pickering 4 542 91 BWR Muehlberg 336 91 typically achieve a 33-percent efficiency. PWR Turkey Point 3 728 90 PWR Beznau 2 364 89 The pressure tube configuration has some ad- BWR Garona 460 88 vantages in that it separates the moderator from PWR Grafenrheinfeld 1299 88 the coolant. The reactivity control devices and PHWR Bruce 1 826 87 b safety systems are located outside the primary Cumulative PHWR Bruce 3 826 88 coolant loop and cannot be affected by a loss of PHWR Bruce 4 826 85 coolant accident. Moreover, the moderator acts PWR Beznau 2 364 85 as a backup heat sink that could cool the fuel if PHWR Pickering 2 542 83 GCR Hunterston A 169 83 the primary coolant system fails. This reduces the PWR Stade 1 662 82 necessity to develop elaborate systems to provide PHWR Pickering 4 542 81 BWR Muehlberg 336 80 emergency core cooling and decay heat removal, PWR Obrigheim 345 80 which have been a primary concern in LWRs. aErOm July 1!382 through June 1983. bFrom start of operation through June 1983 The HWR differs from the LWR in its method SOURCE “Nuclear Stat Ion Achievement 1983, ” Nuclear .Errgmeerlng /nterrra- of refueling. In an LWR, the reactor must be shut tlorral, October 1983 down, the cover of the pressure vessel removed, and the fuel rods exchanged in an operation that lasts for several weeks. HWRs can utilize online an LWR to produce the same amount of elec- refueling because they use pressure tubes rather tricity. than a single pressure vessel. This means that the The use of heavy water instead of ordinary reactor can continue to operate while depleted water as a moderator and coolant provides a fuel- fuel assemblies are removed and replaced with economy advantage, but it also suffers from a fresh fuel. This method of refueling contributes disadvantage. Heavy water is expensive, and the to the overall availability of the reactor since there initial inventory of heavy water represents about is no need to shut it down. It also enhances rapid 20 percent of the capital cost of HWRs. Total identification and removal of fuel elements that capital costs of HWRs are probably comparable leak radioactive materials into the cooling water. to those of LWRs, but fuel cycle costs over the As discussed above, the HWR appears to have life of the plant should be lower. certain safety and operational advantages with Another feature of the HWR that distinguishes respect to the LWR. However, there are other it from a PWR is the use of hundreds of small factors that make it unlikely to be considered as pressurized tubes instead of a single large pres- a viable alternative to the LWR in the United sure vessel. The pressure tubes in an HWR en- States. As with all alternatives to the LWR, the close the fuel assemblies and heavy water coolant heavy water reactor suffers from lack of familiarity which flows through the tubes. They are posi- and experience in the United States. There are tioned horizontally in a large unpressurized vessel no vendors in the United States which offer known as a calandria, as shown in figure 22. The HWRs, and hence there is no established domes- calandria is filled with heavy water, which acts tic infrastructure to build and service them. Fur- as a moderator and is kept at low temperature thermore, there are no utilities with HWR ex- and pressure. The heavy water in the calandria perience. These factors would not necessarily surrounds the coolant tubes but is isolated from prohibit the introduction of an HWR into the the fuel. United States, since the manufacturing require- —

98 . Nuclear Power in an Age of Uncertainty

Figure 22.—Heavy Water Reactor

Generator

Turbines

Pump

inment building

SOURCE: Office of Technology Assessment.

ments, design, and operational skills for the HWR capital debts without a clear demonstration of the and LWR are similar. However, any alternative ad-vantages of an alternative reactor. design would have to overcome barriers before Another issue relates to uncertainties in the being accepted. It would have to be clearly licensing process. The HWR is a fully commer- superior to the LWR, but in some areas that have cial and licensable reactor in Canada and other proved to be most troublesome for LWRs for in- nations. In the United States, however, it would stance, operational complexity, the HWR may be necessary to modify the licensing procedures actually add to the uncertainty. to match a new reactor. It also is likely that design modifications would have to be made to the In addition to lack of familiarity, the HWR of- HWR in order to accommodate to the U.S. reg- fers no capital cost advantages to the LWR. Even ulatory structure; there is a spectrum of opinions though lifetime costs may be lower, it is unlikely on how extensive these changes would be (24). that utilities would be willing to assume large Cost and availability uncertainties would be in- Ch. 4—Alternative Reactor Systems ● 99

troduced into a system that already is less predictable than desired. Initial capital costs might be increased by design modifications to improve efficiency or meet stringent seismic requirements. The commercialization process could be extended if significant changes are required in the licensing and design of this reactor. One possible schedule for development was developed by EPRI and is shown in figure 23. In the United States, the HWR is not perceived as offering enough advantages to abandon LWR technology. Nonetheless, the HWR, may become a source of electricity for U.S. utilities. It is possible that Canadian reactors operating near the U.S. border might significantly increase the export of power to U.S. grids if the HWRs in Canada continue to operate safely and reliably.

Figure 23. —Schedules for Alternative Reactors

1980 1990 2000 2010 2020 Heavy water reactor 1,000 MW prototype A 1,000 MW commercial

Commercialization uncertainty

High temperature gas-cooled reactor 330 MW demonstration (Fort St, Vrain) 900 MW prototype

1,500 MW commercial

Commercialization uncertainty

Photo credit: Ontario Hydro

In a heavy water reactor, natural uranium is used as the fuel. The fuel bundles are loaded horizontally, each in its own tube containing pressurized water. The pressure tubes in a heavy water reactor replace the large steel commercialization target vessel in light water reactors SOURCE: “Alternative Nuclear Technologies, ” Electric Power Research Institute, October 1981.

THE HIGH TEMPERATURE GAS-COOLED REACTOR

The HTGR is cooled by helium and moderated be transformed into useful fuel when it is ir- by graphite, and, as shown in figure 24, the en- radiated. Because helium is used instead of water tire core is housed in a prestressed concrete reac- as a coolant, the HTGR can operate at a higher tor vessel (PCRV) (2,6,25). The reactor uses en- temperature and a lower pressure than an LWR. riched uranium along with thorium, which is This results in a higher thermal efficiency for elec- similar to nonfissionable uranium in that it can tricity generation than can be achieved with the 100 . Nuclear Power in an Age of Uncertainty

Figure 24.—High Temperature Gas-Cooled Reactor

SOURCE: GA Technologies and Office of Technology Assessment. other alternative reactors discussed in this The design of the fuel and core structure for chapter; it also makes the HTGR particularly the HTGR also has inherent safety features. The suited for the cogeneration of electricity and proc- core is characterized by a low power density in ess steam. the fuel, only about one-tenth that of a light water reactor. As discussed above, the core has a large The HTGR has several inherent safety charac- thermal capacitance due to the presence of teristics that reduce its reliance on engineered graphite in the core and support structures. As devices for safe reactor operation. First, the use a result, temperatures would rise very slowly even of helium as a primary coolant offers some ad- if the flow of coolant was interrupted, The oper- vantages. Because helium is noncorrosive in the ators of an HTGR would have a great deal of time operating temperature range of the HTGR, it to diagnose and correct an accident situation be- causes little damage to components. Further- fore the core is damaged. Figure 25 compares the more, it is transparent to neutrons and remains 10-hour margin to fuel failure for an HTGR with nonradioactive as it carries heat from the core. conventional LWRs, in which the margins are The use of graphite as a moderator also has some measured in minutes. advantages. It has a high heat capacity that greatly reduces the rate of temperature rise following a Another safety feature of the HTGR is the severe accident, and hence there is less poten- PCRV, which contains the entire primary coolant tial for damage to the core. As a result, HTGRs system. The PCRV provides more shielding from do not require a containment heat removal the radioactive materials in the core than a steel system. vessel since the thick concrete naturally attenu- Ch. 4—Alternative Reactor Systems . 101

Figure 25.—Comparative Fuel Response Times oped. The concept was successfully demonstrated in 1967 when a 40-MWe reactor was placed in commercial operation. This prototype unit, Peach Bottom 1, was constructed by Phila- delphia Electric Co. and was the world’s first nuclear station to produce steam at 1,000° F. The plant operated at an average availability of 88 percent before being decommissioned in 1974 for economic reasons.

Research and development (R&D) leading to commercial-sized systems continued after the Peach Bottom 1 demonstration. A cooperative — 0.01 0.1 1.0 effort of Public Service Co. of Colorado, the Hours after instantaneous loss-of-cooling Atomic Energy Commission, and General Atomic SOURCE: Gas-Cooled Reactor Associates. Co. led to the construction of the 330-MWe Fort St. Vrain reactor in Colorado. This plant intro- ates radiation. in addition, catastrophic failure of duced several advanced features relating to the a PCRV is regarded to be much less likely than design of the fuel, steam generators, helium cir- failure of a steel vessel. The vertical and circulators, and PCRV. The plant first generated cumferential steel tendons that are used to keep power in 1976 but experienced problems in its the concrete and liner in a state of compression early years that contributed to a disappointing are isolated from exposure to damaging neutron availability record. However, the majority of the fluxes and extreme temperatures. Furthermore, systems in the Fort St. Vrain reactor have per- they are independent of one another and can be formed well. Furthermore, radiation exposures readily inspected; it is extremely unlikely that have been the lowest in the industry, even though many of the tendons wouId fail simultaneously. extensive modifications were made to the pri- The entire primary coolant loop, including the mary system after the plant started operating. The steam generators, helium circulators, and other Fort St. Vrain experience also demonstrates that auxiliary equipment, is included within cavities the HTGR is manageable and predictable, even in the PCRV. The advantage of such a configura- under extreme conditions. In the past 7 years, tion is that pipe breaks within the primary loop forced circulation cooling has been interrupted cannot result in a rapid loss of coolant. As a result, 17 times at the Fort St. Vrain reactor. The opera- the only engineered safety system needed to pro- tors generally were able to reestablish forced cir- tect the core from overheating in the event the culation within 5 minutes, with the longest inter- main core cooling fails is a forced circulation, ruption lasting 23 minutes. Even with so many decay heat removal system. In the HTGR, this is loss of flow incidents, there has been absolutely provided by a core auxiliary cooling system, no damage to the core or any of the components. which is dedicated to decay heat removal and Fort St. Vrain also experienced some unex- incorporates three redundant cooling loops for pected technical difficulties-slow periodic fluc- high reliability. This is enhanced by a PCRV liner tuations in certain core exit temperatures were cooling system that provides an additional heat observed. The fluctuations were associated with sink for decay heat. small movements of fuel in the core, caused by Because of the high thermal efficiency of this differential thermal expansion of fuel blocks. The reactor and the safety features discussed above, problem was resolved at Fort St. Vrain by main- the HTGR has long been considered as a possi- taining the spacing between fuel regions with ble alternative to the LWR for commercial power core restraint devices. The next generation of generation. Work on the HTGR was initiated in reactors will avoid such problems by redesigning the United States soon after the LWR was devel- the fuel block.

25-450 0 - 84 - 8 : QL 3

102 ● Nuclear Power in an Age of Uncertainty

high efficiency or be applied to the cogeneration of electricity and process steam. This design incorporates the lessons learned from the operation of Fort St. Vrain, experience from the earlier design of commercial HTGRs, and information obtained from cooperative international programs. Key design changes have been made to simplify the plant, improve its licensability and reliability, correct specific component-design deficiencies, and increase performance margins (7). The design work has been sponsored by the U.S. Department of Energy (DOE). Another design effort is directed toward developing a much smaller gas-cooled reactor, known as the modular HTGR (14). This concept capi- talizes on small size and low power density to produce a reactor that may be able to dissipate decay heat with radiative and convective cooling. In other words, it would not require emergency cooling or decay heat removal systems. Such a reactor would be inherently safe in the sense that no operator or mechanical action would be required to prevent fuel from melting after the reactor is shut down. The design for this type of reactor is still in a preliminary stage, but its potential for walk-away safety may warrant further investigation. In addition to the domestic effort to develop

Photo credit: Gas-Cooled Reactor Associates (GCRA) the HTGR, there are also international programs to design and construct gas-cooled reactors. The The high temperature gas-cooled reactor uses helium as a coolant instead of water. Helium Federal Republic of Germany has operated a 15- offers some advantages since it is less corrosive MWe prototype plant since 1967 with great suc- than water and does not become radioactive. The cess, achieving an average availability of more helium circulator for the Fort St. Vrain reactor is shown above than 85 percent. A 300-MWe demonstration plant is scheduled for startup in 1984, and work has been initiated on the design of a 500-MWe Since 1977, the HTGR program has focused on HTGR. Japan also has been involved in the devel- the development and demonstration of a com- opment of a very high temperature gas-cooled mercial-sized plant. The emphasis is currently on reactor for process heat applications. The Japa- designing a four-loop, 2240 megawatt thermal nese program is directed at designing a 50-MWth (MWth) reactor that can generate electricity at reactor to begin construction in 1986.

INHERENTLY SAFE REACTOR CONCEPTS

incentives for developing a more forgiving reac- a number of unresolved safety issues. The acci- tor arise from several sources. As previously dis- dent at Three Mile Island increased the incentive cussed, the design of current LWRs has devel- to develop a foolproof reactor when it became oped in a patchwork fashion, and there are still clear that LWRs are susceptible to serious ac- Ch. 4—Alternative Reactor Systems ● 103 cidents arising from human error. A more forgiv- pool and core are enclosed in an imposing con- ing reactor design became desirable in terms of crete vessel (13 meters in diameter and 35 meters investment protection as well as public health and high) that is reinforced with steel tendons. In the safety. PIUS reactor, the other primary components are submerged in the same pool of water that con- The modular HTGR discussed above is an ex- tains the fuel, and all penetrations to the PCRV ample of an effort to develop an inherently safe are located at the top of the vessel. With such reactor. Another example of a reactor that at- a configuration, it is impossible for any type of tempts to improve safety dramatically is based on leak or equipment malfunction to drain off the LWR technology. [t is known as the Process In- cooling water. herent Ultimately Safe (PIUS) reactor, and it is being developed by the Swedish nuclear firm ASEA- During normal operation, primary coolant is ATOM (6,9,23). The design goal is to ensure safe- pumped through the core and primary loop. At ty, even if the reactor is subjected to human er- the bottom of the plenum under the reactor core ror, equipment failure, or natural disaster. In the there is a large open duct extending down into PIUS reactor, this goal was translated into two the pool. The pool water ordinarily is prevented primary design objectives: first, to ensure safe from flowing into the core circuit by the dif- shutdown and adequate decay heat removal ference in density between the hot water within under any credible circumstances; and second, the core and the cool pool water. At the static to use the construction and operating experience hot-cold interface in the duct, a honeycomb grid of current LWRs as a basis for development. This helps prevent turbulence and mixing between the eliminates some of the uncertainties associated hot and cold fluids. If for any reason the core with a new design. temperature should rise to the point at which steam is formed, the pressure balance at the in- The safety goal was paramount in the design terface would be upset, and the pool water would of the PIUS reactor concept. The nuclear island automatically flow upward and flood the core. was designed to ensure that the fuel would never Natural thermal convection through the pool melt, even if equipment failure were to be com- would provide enough circulation to cool the pounded by operator error. To accomplish this, core, and the pool water would keep the core two conditions have to be met. First, it is neces- covered for about a week. This system relies com- sary to ensure that the core always remains cov- pletely on thermohydraulic principles and is total- ered with water. In LWRs, engineered safety sys- ly independent of electrical, mechanical, or tems ensure that cooling water is available to the human intervention. The principal uncertainty in core. However, confidence in these systems was this reactor concept is in the stability of the pres- shaken by the accident at Three Mile Island when sure balance between the hot primary circuit the fuel was exposed long enough to melt. water and the cold berated water. The basic configuration of the PIUS reactor is The PIUS reactor is designed to automatically similar in many ways to that of a conventional shut down as soon as the pool water begins to PWR. It employs a primary loop of pressurized flow through the core. This is guaranteed by water that transfers heat to a secondary steam maintaining a high concentration of boron in the loop through a steam generator. The main dif- pool water; boron is a strong neutron absorber ferences between the two reactors are the size and automatically interrupts a chain reaction. This of the pressure vessels and the location of the feature of the PIUS reactor is attractive because primary system components. In a PWR, the fuel the possibility of a transient without a subsequent is surrounded by water and enclosed in a pres- reactor scram is virtually eliminated. sure vessel that is slightly larger than the core; the primary system pump, steam generator, and If the PIUS reactor proves to be reliable, it might piping are located outside the vessel. In the PIUS resolve some of the troublesome problems with reactor, the core is located at the bottom of a very LWRs. With inherent mechanisms for automatic large pool of water. As shown in figure 26, the shutdown, natural convective cooling, and a 104 • Nuclear Power in an Age of Uncertainty

Figure 26.—Process-inherent Ultimately Safe Reactor

Generator

Upper c interface region

Pool of berated water generator

Recirculation Prestressed pump concrete reactor vessel

Reactor core Lower open interface region

SOURCE: ASEA-ATOM and Office of Technology Assessment large heat removal capacity, there do not appear vessel. The steam generator in the PIUS reactor to be any credible mechanisms for uncovering also will require additional development since it or overheating the core of the PIUS reactor. How- will be of a different configuration than in con- ever, the changes from the standard LWR design ventional PWRs. More importantly, maintenance create new problems and uncertainties. The de- may be very difficult since the components will sign of the PIUS is different enough from current be submerged in a pool of berated water. It is LWRs that further development will be required therefore essential that all primary system com- for components, materials, and procedures. The ponents perform reliably and with little main- design and construction of the PCRV will pose tenance. The submerged components and pip- a problem, since it is larger than any other similar ing pose another problem—since the pool water Ch. 4–Alternative Reactor Systems • 105 will be about 380° F less than the primary system Overall, the PIUS reactor represents a fairly coolant, it is necessary to insulate the primary dramatic departure from conventional LWRs. coolant loop from the borated pool water. The One consequence of this reconfiguration is that insulation for such an application has not yet the economics become far less certain. Because been fully developed or tested. Another poten- of the requirement for a large PCRV, the nuclear tial difficulty relates to fuel handling. Nuclear fuel island of the PIUS is likely to be significantly more assemblies are removed and exchanged routinely expensive than that of an LWR. Furthermore, the in current LWRs, but a similar operation in the remaining technical uncertainties relating to com- PIUS reactor is complicated by having to work ponent and materials development could be cost- at a distance of 80 feet. ly to resolve. The originators of the PIUS design suggest that other plant costs might be reduced Another serious concern relates to the thermal by easing or eliminating the safety qualifications hydraulic response of the reactor. There is con- for the balance of plant systems because reactor siderable uncertainty about the flow patterns in safety would not be dependent on them. I n cur- the lower interface region between the pool rent LWRs, safety qualification of secondary and water and the primary coolant. If the boundary auxiliary systems contributes significantly to the is not stable, normal operations could be inter- overall cost of the plant. If nuclear regulators rupted unnecessarily by the inflow of borated agree to such a reorientation, it is conceivable pool water through the core, shutting the reac- that the overall cost of the PIUS reactor would tor down. Computer simulations have been per- be comparable to today’s PWRs or BWRs. It formed by ASEA-ATOM to determine the char- should be noted, however, that many nuclear- acteristics of the interface region. The Tennessee grade systems would still be required to remove, Valley Authority has supplemented this with a process, and return radioactive gases and liquids small-scale test to observe flow patterns. The from the pressure vessel. uncertainties associated with the liquid-interface region can only be resolved with larger and more Technical and economic uncertainties are sig- definitive experiments, such as th 3MWm test nificant factors in the decision to develop any planned by ASEA-ATOM. new reactor concept. In spite of these unknowns, further development of the PIUS reactor might In many respects, the PIUS design builds on be warranted due to its potential safety advan- demonstrated LWR technology. The fuel for the tages. These advantages might restore public con- PIUS reactor is essentially the same as for LWRs. fidence in the safety of nuclear power, but they The PIUS core is designed to use burnable poi- must be tested further before any final judgments sons to maintain a constant reactivity throughout can be made. the fuel cycle. These have been used extensively in BWRs, and the experience is directly appli- Regardless of the merits of this particular de- cable to the PIUS reactor. The water chemistry sign, the PIUS concept illustrates that innovative and waste handling systems for the PIUS are also revisions of the standard LWR design can emerge very similar to today’s LWRs. Finally, most of the if designers are not constrained in their thinking. materials that would be used in the PIUS are iden- Even if the PIUS concept itself turns out to have tical to those in conventional LWRs. In fact, the some insurmountable problems, exploratory re- temperature, pressure, and flow conditions in search might continue concerning the basic con- the PIUS reactor would be less severe than in cept of inherent safety. LWRs.

THE SMALL REACTOR

Considerable sentiment is often expressed in norm. When the current generation of reactors favor of reactors that are smaller than the 1,000- was being designed in the late 1960’s, it seemed to 1,300-MWe LWRs that now represent the natural to continue scaling up the size because 106 ● Nuclear Power in an Age of Uncertainty larger nuclear units were cheaper to build per kilowatt of capacity. Moreover, utilities were growing rapidly and needed large increments of new power. The situation is very different now. Many seem to feel that a carefully designed small reactor might be easier to understand, more manageable to construct, and safer to operate. Al- though many of these claims seem intuitively convincing, they are difficult, if not impossible, to substantiate. OTA sponsored a search for evidence that small plants have any advantages over large plants in terms of safety, cost, or operability (20). This search revealed no firm statistical data in support of the small reactor, although it summarized some of the arguments that make it an attractive concept (see vol. II). Utilities may find small pIants especially appealing today because they allow more flexibility in 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 planning for the total load of the utility. In addi- Years after start of commercial operation tion, the consequences of an outage would have a smaller impact on the overall grid. Furthermore, B.—As a Function of Calendar Year reducing the size of plants would limit the financial exposure of the utility to loss and increase overall system reliability. Initially, small plants appear to suffer a disadvantage in unit construction costs since they cannot realize the full benefits of economies-of-scale. However, more of the plant could be fabricated in the factory rather than constructed in the field, and this could result in large cost savings if the market is large enough to justify investment in new production facilities. Moreover, the construction times for small plants would probably be much less than for their larger counterparts. Overall, it is not clear that small plants would necessarily be more expensive than today’s large ones. 676869 70 71 72 73 74 7576 77 78 79 8081 The operability of different sized plants maybe compared on the basis of availability. As shown Calendar year in figure 27A & B, the availability of small plants generally exceeds that of currently operating may be easier to control or maintain. However, larger plants, although only by about 5 percent. this comparison is not conclusive since most of This trend surfaces both when availabilities are the smaller reactors were designed and built in plotted as a function of the number of years after the 1960’s and have not been affected by as start of operation (which compares plants of the many design changes as the newer, larger plants. same age) and as a function of calendar year (which compares plants operating in the same en- It is difficult to compare the safety of small and vironment). These differences could be due to large reactors, but one indication is the occur- either the number or duration of outages at rence of events that could be precursors to severe smaller plants, which indicates that small plants accidents. It appears that the frequency of such Ch. 4—Alternative Reactor Systems . 107

events is independent of reactor size. However, number of small units needed to comprise the the initiating events that do occur at small plants same generating capacity. may be easier to manage than similar events at Small reactors are unlikely to be able to com- large plants. This result is based on too small a pete commercially with their larger counterparts sample to be conclusive, but it may warrant fur- unless R&D that specifically exploits the poten- ther study. Another safety comparison can be tial for modular, shop fabrication of components made on the basis of the consequences of an ac- is sponsored. This would allow small plants to cident. The worst-case accident in a small plant take full advantage of the increased productivity would be less damaging because the fission prod- and quality of work in a factory setting. uct inventory is much less in a smaller plant. This effect, however, might be offset by the larger

THE STANDARDIZED REACTOR

The concept of a standardized reactor has been A major barrier to designing a standard plant widely discussed for years in the industry but has is the difficulty in marketing identical reactors, yet to become a reality (18). The advantages are given the current industry structure and regula- many: more mutual learning from experience tory climate in the United States. There are many among reactor operators, greater opportunity for opportunities for changes in today’s plants, such indepth understanding of one reactor type, and as to match a particular site, to meet the needs sharing of resources for training operators or of a specific utility, and to accommodate NRC developing procedures. Since much of the con- regulations. In addition, the existing institutional cern over current reactors centers on their man- structure does not lend itself easily to industry- agement rather than on their design, the oppor- wide standardization. There are currently five tunity to concentrate on learning the correct ap- reactor suppliers and more than a dozen AE firms. plication of one well-understood design is While each reactor vendor is moving toward a appealing. single standardized design, balance of plant designs by the AEs continue to vary. It is unlike- Utilities and vendors would be especially en- ly that a single dominant plant design will arise thusiastic about standardized designs if that con- out of all combinations of vendors and AEs, cept were coupled with one-step or streamlined which implies that there may not be industrywide licensing. The simplification of the licensing proc- standardization. However, it is possible that a few ess might bring concomitant benefits in reduced prominent combinations of the more successful capital costs. If the plants were smaller than those reactor suppliers and AEs will join forces to proof the current generation, larger numbers of small duce a more manageable number of stand- plants would be built to meet a given demand, ardized designs. and this would facilitate standardization.

CONCLUSIONS

No single reactor concept emerges as clearly shutdown, decay heat removal, and fission prod- superior to the others since the preferred design uct containment are provided by simple, passive varies with the selection criterion. If safety is of systems which do not depend on operator ac- paramount concern, the reactors that incorporate tion or control by mechanical or electrical means. many inherent safety features, such as PIUS or The full-scale HTGR is also attractive in terms of the modular HTGR, are very attractive. In such safety since it provides more time than any of the reactors, the critical safety functions of reactor water-cooled concepts for the operator to res- 108 Ž Nuclear Power in an Age of Uncertainty

pond before the core overheats. The remaining since the HWRs have lower fuel costs. Standard- reactors appear to be roughly comparable regard- ization of any of the reactors discussed would ing safety features. The HWR has the lowest in- reduce costs, if the reactors could be licensed and ventory of radioactive materials, and the indepen- constructed more quickly. The HTGR appears to dent moderator loop serves as a passive, alter- be comparable in cost to LWRs, but there are native decay heat removal system. In addition, greater and different uncertainties associated with the HWR has compiled a superb record in Can- it. It is premature to estimate the cost of a PIUS- ada. Advanced LWRs incorporate the benefits ac- type reactor for several reasons. First, it is still in crued from many years of extensive operational the conceptual design phase, so types and experience. Finally, small and/or standardized amounts of materials cannot be determined reactors may have operational advantages precisely nor can construction practices and resulting from a better understanding of and con- schedules be accurately anticipated. In addition, trol over their designs. the PIUS designers are relying on low costs in the balance of plant to compensate for the higher If the reactors were to be ranked on the basis costs of the nuclear island. It is not clear whether of reliable operation and easy maintenance, a dif- the balance of plant systems can be decoupled ferent order results. The advanced LWR is very from their safety functions; the regulatory agen- attractive because these criteria have heavily in- cies obviously will have a major impact on this fluenced its design. Small reactors also appear decision, and hence the cost of a PIUS-type plant. high on the list because their size and shop- fabricated components may facilitate operation, A final criterion applied to these reactors might maintenance, and replacements. HWRs rate high be the certainty of our knowledge of them. How because they have performed well to date, and predictable will their performance be? The rank- they do not require an annual refueling shut- ing here is almost the reverse of that for safety. down. The few HTGRs that are in operation have Advanced LWRs are clearly superior in terms of had mixed performance records, but the newest familiarity because they have evolved from plants design addresses some of the problems that con- that have operated in the U.S. for more than 20 tributed to poor reliability. One factor enhanc- years. HWRs have also compiled a Icing record, ing overall performance is the ease of mainte- but design modifications might have to be made nance in an HTGR resulting from inherently low before the reactor could be licensed in the United radiation levels. There are many uncertainties States. There is much less experience with HTGRs associated with the PIUS concept. It is likely to in the United States, with only a single facility in pose maintenance problems. It is also possible operation. The PIUS concept lags far behind the that the behavior of the PIUS will be erratic in other reactors in terms of certainty since it has normal transients, thus increasing the difficulty never been tested on a large scale. of operation. In other ways, however, the PIUS This survey has examined many reactor con- could be simpler to operate. cepts and found that none were unambiguously Any attempt to rate these reactor concepts on superior in terms of greater safety, increased the basis of economics is very difficult. Experience reliability, and acceptable cost. Most represent with LWRs indicates that the price of facilities of a compromise among these factors. A few could the same design can vary by more than a factor not be adequately compared because so many of 2, so estimates of costs of less developed reac- uncertainties surround the design at this stage. tors are highly suspect. Only a few speculative The present lull in nuclear orders provides an op- comments can be made. Small reactors suffer a portune time to reduce the uncertainties and ex- capital-cost penalty due to lost economies-of- pand our knowledge of the less well-tested con- scale, but it is possible that this could be reduced cepts. A demonstration of advanced LWRs may by fabricating more components in factories and soon occur in Japan, and the results should be keeping construction times short. HWRs are ex- valuable input to future decisions on the LWR pected to have comparable capital costs, but their concept. If continued, the Department of Ener- lifetime costs may be lower than those of LWRs gy’s development program on HTGRs will con- Ch. 4—Alternative Reactor Systems Ž 109 tinue to provide information and experience that a big enough market to support new research could make the HTGR a viable alternative to the programs. LWR. It may also be valuable to examine the Until the results of future investigations are in, operation of Canada’s HWRs to determine if any nothing on the horizon appears dramatically bet- of their experience can be applied to U.S. reac- ter than the evolutionary designs of the LWR. tors. If considerable sentiment continues to be There is a large inertia that resists any move away expressed in favor of small reactors, some initial from the current reactor types, in which so much design work may be appropriate. Finally, a time has been spent and from which so much preliminary investigation of the PIUS reactor experience has been accrued. However, if to- would teach us still more about a concept that day’s light water reactors continue to be plagued is very promising. Work on this or another “fresh by operational difficulties or incidents that raise look” design would require government support safety concerns, more interest can be expected since the existing reactor designers do not see in alternative reactors.

CHAPTER 4 REFERENCES

1. Abel, P., Workshop Discussion on Technological 11, Leggett, W. D., “Advances in Nuclear Power,” Changes, Dec. 8, 1982, U.S. Congress, Office of presented at the Second Joint Nuclear Engineer- Technology Assessment, Jan. 3, 1983. ing Conference of the American Society of Me- 2. Agnew, H. M., “Gas-Cooled Nuclear Power Reac- chanical Engineers and the American Nuclear So- tors, ” Scientific American, vol. 244, No. 6, June ciety, July 26, 1982. 1981. 12. Lewis, H., et al., “Risk Assessment Review Group 3. Atomic Energy of Canada Limited, “CANDU-The Report to the U.S. Nuclear Regulatory Commis- Facts,” November 1982. sion,” NUREG/CR-0400, September 1978. 4. Burnham, D., “Safety Goals Set by Nuclear 13, Martel, L., et al., “Summary of Discussions With Panel,” New York Times, Jan. 11, 1983. Utilities and Resulting Conclusions,” Electric Pow- 5. Civiak, R. L., “Potential for Reduction in the Pre- er Research Institute, June 1982. dicted Release of Radioactive Material Following 14. McDonald, C. F. and Sonn, D. L., “A New Small a Severe Nuclear Accident, ” Congressional Re- HTGR Power Plant Concept With Inherently Safe search Service, Mar. 21, 1983. Features–An Engineering and Economic Chal- 6, Fraas, A. P., “Survey and Assessment of the Tech- lenge,” presented at the American Power Confer- nological Options Available to the Nuclear indus- ence, Apr. 18-20, 1983. try in the 1980 to 2000 Period,” Institute for Energy 15 MHB Technical Associates, “Issues Affecting the Analysis, January 1983. Viability and Acceptability of Nuclear Power 7, Gas-Cooled Reactor Associates and General Usage in the United States,” Dec. 28, 1982. Atomic Company, “High Temperature Gas- 16 “Mounting Reports of Failures in Electrical Break- Cooled Reactor Steam CycIe/Cogeneration Design ers With Undervoltage Trip,” Nuc/eonics Week, and Technology Development Plan for Nuclear vol. 24, No. 14, Apr. 7, 1983. Steam Supply System, Volume II Summary,” 17. National Electric Reliability Council Generating GCRA82-011, JUIY 1982. Availability Data System, “Ten Year Review 8. George, R. A. and Paulson, C, K., “A Nuclear Plant 1971-1980 Report on Equipment Availability.” Design for the 1990’s—Meeting Tomorrow’s 18. Office of Technology Assessment, U.S. Congress, Needs,” presented to the American Power Con- Nuclear Powerplant Standardization: Light Water ference, 1983. Reactors, OTA-E-134, April 1981. 9. Hannerz, K., “Towards Intrinsically Safe Light 19. Phung, D. L., “Light Water Reactor Safety Since Water Reactors,” Institute for Energy Analysis, the Three Mile Island Accident,” Institute for En- June 14, 1982. ergy Analysis, july 1983. 10. Kemeny, J. G., Chairman, “Report of the Presi- 20. Prelewicz, D. A., et al., “Nuclear Powerp!ant dent’s Commission on the Accident at Three Mile Size,” ENSA, Inc., Nov. 19, 1982. island,” Washin~ton, D. C.. October 1979. 21. Robertson, J. A. L., “The CANDU Reactor System: 170 ● Nuclear Power in an Age of Uncertainty

An Appropriate Technology,” Science, vol. 199, 25. U.S. Department of Energy, “High Temperature Feb. 10, 1978. Gas-Cooled Reactors, Preliminary Safety and En- 22 Smith, D., “A Reactor Designed for Sizewell,” vironmental Information Document, Vol. IV, ” New Scientist, vol. 98, No. 1329, Oct. 28, 1982. Nonproliferation Alternative Systems Assessment 23 Tiren, l., “Safety Considerations for Light Water Program, January 1980. Reactor Nuclear Power Plants: A Swedish Perspec- 26. U.S. Nuclear Regulatory Commission, “Unre- tive, ” Institute for Energy Analysis, solved Safety issues Summary,” NUREG-0606, ORAU/lEA-83-7, May 1983. VOI. 4, No, 3, Aug. 20, 1982. 24 U.S. Department of Energy, “Heavy Water Reac- 27. Weitzberg, A., et al., “Reliability of Nuclear Power tors, Preliminary Safety and Environmental infor- Plant Hardware—Past Performance and Future mation Document, Vol. 11,” Nonproliferation Trends,” NUS Corporation for the Office of Tech- Alternative Systems Assessment Program, January nology Assessment, NUS-4315, Jan. 15, 1983. 1980.

Chapter 5 Management of the Nuclear Enterprise

Photo credit: Atomic Industrial Forum The control room at a typical nuclear plant Contents

Page Introduction ...... 113 Variations in Quality of Construction and Operation ...... 114 Construction ...... $ ...... 114 Operation ...... 116 Management Challenges in Construction and Operation ...... 118 Technological Factors That Influence Construction and Operation ...... 118 External Factors That Influence Operation and Construction ...... 121 Internal Factors That Influence Construction and Operation ...... 124

improving the Quality of Nuclear Powerplant Management ...... 127 Technical Approach ...... 127 Institutional Approach ...... 128 Assessment of Efforts to Improve Quality...... 132 New Institutional Arrangements ...... , 134 A Larger Role for Vendors in Construction ...... 135 Service Companies ...... 136 Certification of Utilities...... 137 Privately Owned Regional Nuclear Power Company...... 138 Government-Owned Regional Nuclear Power Authority...... 138 Chapter 5 References ...... 139

Tables Table No. Page 17. U.S Reactor Types, Suppliers, Architect/Engineers, and Constructors ...... 114 18. Construction Records of Selected U.S Light Water Reactors ...... 115 19. Performance of Selected U.S. LWRs...... 117 20. Early Operating Experience of U.S. Commercial Light Water Reactors ...... 124 21. Classifications for INPO Evaluations ...... 132

Figures

Figure No. Page 28. Comparison of Commodity Requirements for Coal and Nuclear Powerplants . . . . . 119 29. Comparison of Manpower Requirements for Coal and Nuclear Powerplants ...... 122 30. Historical Labor Requirements in the Nuclear Power Industry ...... 123 31. NRC Regulatory Guidelines Issued From 1970 to 1980 ...... 124 Chapter 5 Management of the Nuclear Enterprise

INTRODUCTION

In the previous chapter, alternative reactor have misjudged the level of effort required to types were reviewed in terms of safety, operabili- manage nuclear power operations successfully. ty, and economics. While light water reactors This is not surprising, considering the variability (LWRs) lack some of the inherent safety features in the nuclear utility industry. Forty-three utilities that characterize the alternatives, this compari- operate 84* nuclear powerplants, and 15 addi- son yielded no compelling reason to abandon tional utilities are in the process of constructing LWR technology in favor of other, reactor types. their first nuclear units (40). Among these vari- The excellent performance records of some of ous organizations can be found a wide variety the pressurized water reactors (PWRs) and boil- of management structures and philosophies, ex- ing water reactors (BWRs) indicate that LWRs can perience, commitment, and skill. While utilities be very reliable when properly managed. Large are not the only organizations that seem to have and complex nuclear units can also be built with- underestimated the difficulties involved with nu- in budget and on schedule, as proven by some clear powerplants, they must assume the ultimate recent examples. These cases indicate that it is responsibility for the safety and financial success possible to construct and operate nuclear power- of their plants. plants efficiently and reliably. The diversity of the utility industry has not Unfortunately, not all utilities perform to the created major difficulties in managing nonnuclear same high standards. Numerous examples of con- generating plants. Many different organizational struction malpractice and operating violations styles and structures have been used successful- have surfaced in recent years, and many of these ly to construct and operate fossil fuel stations and problems are serious enough to have safety and distribution systems. With the advent of nuclear financial implications. Utilities evidently need to technology, however, several new questions can depend on more than government safety require- be raised: ments and the conservatism of nuclear designs ● Is the technology so sensitive to its manage- to compensate for errors. As the accident at Three ment that it is not adequately safe or reliable Mile Island so vividly illustrated, LWRs are not when poorly managed? entirely forgiving machines; they are susceptible ● If so, can the quality of management be im- to certain combinations of human error and me- proved to a uniformly acceptable level? chanical failure. Although LWRs are built to ac- ● Alternatively, can the technology be modi- commodate to some problems in construction, fied so that it is less sensitive to its manage- maintenance, and operation, there is a limit to ment? the extent of malfunctions and operational error that can be tolerated. The construction and oper- Management and quality issues will be ad- ation of nuclear powerplants are highly sophis- dressed in this chapter by illustrating the sensitivi- ticated processes. Because nuclear technology is ty of nuclear power operations with a few recent very complex and has the potential for accidents examples, a look at factors that contribute to such with major financial and safety implications, man- problems, and a review of current efforts to en- agement of the nuclear enterprise must be of an sure uniformly high levels of performance. intensity that is seldom required in other utility operations, or indeed, in most other commercial endeavors. Many utilities readily grasped the unique characteristics of nuclear technology and “Includes all plants with operating licenses, even though some devoted their best management resources to its (Three Mile Island Units 1 and 2, Dresden 1, and Diablo Canyon development. Others, unfortunately, seem to 1 ) are not currently in operation.

113 114 ● Nuclear Power in an Age of Uncertainty

VARIATIONS IN QUALITY OF CONSTRUCTION AND OPERATION

The following discussion will address variations Table 17.—U.S. Reactor Types, Suppliers, in quality of construction and operation on three Architect/Engineers, and Constructors levels: Reactor types: Pressurized water reactors (PWR) 1. An overview of the nuclear industry will be Boiling water reactors (BWR) presented to demonstrate that various proj- High temperature gas-cooled reactors (HTGR) Reactor suppliers: ects differ significantly from one another. Babcock & Wilcox Co. (PWR) This will provide a qualitative basis for assess- Combustion Engineering, Inc. (PWR) ing the sensitivity of the technology to its General Atomic Co. (HTGR) General Electric Co. (BWR) management. Westinghouse Electric Corp. (PWR) 2. Some of the more successful pIants will be Architect engineers and/or constructors: examined to identify the conditions under American Electric Power Service Corp. (AE, C) Baldwin (C) which it is possible for nuclear powerplants Bechtel Power Corp. (AE, C) to be constructed and operated to the high- Brown & Root, Inc. (C) est standards of quality. Burns & Roe, Inc. (AE, C) Daniel Construction Co. (C) 3. Some less successful plants will be examined Ebasco Services, Inc. (AE, C) to identify the factors that contribute to poor Fluor Power Services (AE, C) management and to understand the cost and General Atomic Co. (C) Gibbs & Hill, Inc. (AE, C) safety implications. Gilbert Associates, Inc. (AE) Kaiser Engineers (C) The examples that have been selected for dis- J.A. Jones Construction Co. (C) cussion are not intended to fully span the range Sargent & Lundy Engineers (AE) Stone & Webster Engineering Co. (AE, C) of good and bad practices; they are, however, United Engineers & Constructors, Inc., (AE, C) useful in illustrating the differences in the ways Wedco (a subsidiary of Westinghouse Electric Corp.) (C) in which nuclear power has been implemented SOURCE: “World List of Nuclear Power Plants,” Nuclear News, February 1983. in recent years.

Construction tion time and cost, which differ widely among the utilities shown in table 18. Only plants begin- The construction of nuclear powerplants in the ning commercial operation after the accident at United States is far from being a standardized Three Mile Island were included, so all of these process. As shown in table 17, a utility must units were affected to some degree by the regu- choose among several reactor types and vendors latory changes that have occurred since 1979. and among an even larger selection of architect These data should be interpreted with some engineers (AEs) and constructors. Wide differ- care. Several of the longer construction times may ences in design and construction practices can reflect inordinate licensing delays or a utility’s result from these various combinations. A utility decision to delay construction in response to slow can superimpose additional changes on the basic growth in the demand for power. In addition, design to customize its plant according to its spe- some of the projections for very short construc- cial needs or to accommodate to specific site re- tion times may be overly optimistic. It is also dif- quirements. Such factors partially explain the var- ficult to make direct comparisons of construction iations in construction time and quality discussed costs since they are based on different account- below. ing schemes. Furthermore, both estimates and ac- There are no simple measures of quality in con- tual expenditures are reported by the utilities in struction, and no attempt will be made to develop “current dollars. ” Annual expenditures are then comprehensive measures. But efficiency in con- summed without accounting for the time value struction can be partially indexed by construc- of money, with the total construction costs ex- Ch. 5—Management of the Nuclear Enterprise ● 115

Table 18.—Construction Records of Selected U.S. Light Water Reactors

Construction Year of timea commercial operation Costc (actual or Plant (years) (actual or expected) expected, $/kWe) Shortest construction times: St. Lucie 2 ...... 6 1983 1,800 Hatch 2 ...... 7 1979 607 Arkansas Nuclear One 2 . . . . 7 1980 308 Perry 1 ...... 8 1985 2,200 Palo Verde 1 ...... 8 1984 1,900 Byron 1 ...... 8 1984 1,500 Callaway 1 ...... 8 1984 2,500 Longest construction t/roes: Diablo Canyon 1...... 16 1984 1,700 Diablo Canyon 2 ...... 14 1984 1,700 Salem 2 ...... 13 1981 704 Zimmer 1 ...... 13 1985 2,400 Midland 1 ...... 13 1985 2,700 Sequoyah 2 ...... 12 1982 740 Watts Bar 1 ...... 12 1984 1,500

aeased on construction permtt to commercial operation. b“World List Of Nuclear Power Plants,” Nuclear News, February 1983, and other recent updates. ‘Komanoff Energy Associates, 1983, expressed In mixed current dollars. pressed in terms of “mixed current dollars.” This Two of the more notable nuclear powerplant accounting system tends to further distort actual construction projects are Florida Power & Light costs. Co.’s St. Lucie 2 unit at Hutchinson, Fla. and the Palo Verde 1 plant at Wintersburg, Ariz. owned It is interesting to note that the best and worst by Arizona Public Service Co. As shown in table construction schedules from table 18 differ by an 18, both units are projected to be completed with average of about 6 years. In fact, the plants with relatively short construction schedules. Neither the longest construction times took twice as long utility has encountered significant regulatory dif- to complete as those with the shortest schedules. ficulty nor much opposition from interveners (6). Dramatic differences also can be observed in the Both units had to accommodate to the wave of costs, even when the construction schedules are backfit and redesign requirements of the Nuclear similar. For example, the Callaway 1 unit is pro- Regulatory Commission (NRC) that followed the jected to cost $2,500 per kilowatts electrical accident at Three Mile Island, and yet no signif- (kWe) after 8 years of construction, while the icant delays have been experienced at either site. Byron 1 plant is projected to cost only 60 per- These examples indicate that nuclear power- cent of that with the same construction schedule. plants can be constructed expeditiously, even in the most difficult regulatory environment. A recent study by the Electric Power Research Institute (EPRI) attempts to identify the reasons In contrast to these examples, other plants have for the variations noted here (3). In a statistical had a long history of problems. Quality control analysis of all nuclear powerplants, it was found in nuclear powerplants, as in other commercial that 50 to 70 percent of the variation in Ieadtime endeavors, is important in assuring consistency could be accounted for by regulatory differences, and reliability. In industries such as nuclear power deliberate delays, and variations in physical plant and aerospace, where the consequences of fail- characteristics, EPRI ascribed the remaining var- ure can be severe, quality is guaranteed by su- iation to management practices and uncontrol- perimposing a formalized, independent audit lable events. To more fully understand the im- structure on top of conventional quality control portance of utility management in the construc- measures in design, procurement, manufacturing, tion phase, it is valuable to examine a few specific and construction (30). Deficiencies in the quality examples. control procedures at nuclear reactors are cause 116 . Nuclear Power in an Age of Uncertainty

tices have led to an NRC stop-work order. The \ . . . . NRC has uncovered a number of problems at the plant resulting from what it calls a “widespread breakdown of Cincinnati Gas and Electric’s management . . . “ (20). The Zimmer plant has been troubled for years by poor construction practices and an inadequate quality assurance program. * NRC cited deficiencies in 70 percent of the structural steel welds, inadequate documentation and qualification of welders and quality assurance personnel, unauthorized alteration of records, and inadequate documentation of quality for materials in safety-related components (4).

The examples discussed above represent the extremes of good and bad construction experiences. They indicate that nuclear powerplants can be constructed with varying emphasis on quality, and that such differences in management approach result in noticeable differences in the Photo credit: Atomic industrial Forum plants. Florida Power & Light used experience gained in building St. Lucie 1 to construct St. Lucie 2 in the record time of 68 months Operation As with construction, it is difficult to identify for serious concern because they may make it specific measures of safety or quality in plant impossible to verify that the plants are safe. operations. There is, however, an intuitive cor- Deficiencies in the quality assurance program relation between safe and reliable plant opera- at the two Diablo Canyon nuclear powerplants tion. Two parallel arguments for this connection were uncovered in 1981 (18). These deficiencies can be made. First, a safe plant is one which is had gone undetected by NRC and the plant constructed, maintained, and operated to high owner, Pacific Gas & Electric Co. (PG&E), for standards of excellence. Such a plant is also likely years. They did not surface until after NRC had to be a reliable performer. Conversely, a reliable granted PG&E a preliminary license to operate plant that operates with few forced outages is less one of the reactors at low power. Since the prob- likely to tax its safety-related equipment by fre- lems have surfaced, a number of errors in seismic quent cycling. Some caution must be used in design have been identified, and it is not yet cer- equating safety with reliability; it is possible that tain that the plants will be able to withstand a a plant could be operated outside of its most con- design basis earthquake. Diablo Canyon’s license servative safety margins in the interest of increas- has been suspended and will not be reinstated ing its capacity factor. But in general, good plant until NRC can be convinced that the safety equip- availabilities (or capacity factors for base-loaded ment provides adequate protection to the public plants such as nuclear units) are reasonably good and that the quality assurance weaknesses have indicators of well-run plants. The average cumu- been corrected. (The Diablo Canyon plants are lative capacity factor of all U.S. reactors is cur- discussed further in ch. 8.) rently 59 percent. This means that all of the reactors in the United States have operated an aver- Other management control failures have re- age of 59 percent of their design potential sulted in lengthy construction delays. A recent example is the 97-percent-complete Zimmer 1 plant owned by Cincinnati Gas & Electric Co., in *On Jan. 20, 1984, Cincinnati Gas & Electric announced that Zim- which alleged deficiencies in construction prac- mer would be converted to a coal-fired facility. Ch. 5—Management of the Nuclear Enterprise ● 117 throughout their lifetimes. As discussed in chapter ities with previous nuclear experience as well as 4, this is comparable to the capacity factors of those with only a single plant. Although there are base-loaded generating coal units. more PWRs than BWRs in both lists, this merely reflects the fact that there are nearly twice as The average data conceal the more interesting many PWRs as BWRs in operation. It should be variations in individual plants. A number of LWRs noted that although size does not appear to be have operated for years at much lower capacity a dominant characteristic of either good or poor factors, while some plants have consistently ex- performers, there is a tendency for the best per- ceeded the average by wide margins. Table 19, formers to be smaller than their less successful which lists lifetime capacity factors for the best counterparts. While 20 percent of all mature reac- and worst plants in the United States, illustrates tors are larger than 1,000 megawatts electrical the wide range of performance that can be found (MWe), 4 of the 10 plants with the worst capaci- among reactors of comparable design, Note that ty factors are larger than 1,000 MWe; only 1 plant the best plants have lifetime load factors that are of this size is in the list of the best performers. 20 points greater than the 59-percent average capacity factor discussed above, while the poorest Three of the best plants shown in table 19 have performers have capacity factors 10 points less been in operation for more than a decade–Point than the industrywide average. The management Beach 2, Connecticut Yankee, and Vermont Yan- of maintenance and operations is one of several kee. It is clear from the performance records of factors that contributes to these differences. these units that a nuclear powerplant can be a very reliable source of electricity over many The data in table 19 suggest an important point: years. Other less fortunate plants have experi- no single external characteristic can be identified enced considerable operating difficulties, as in- that unambiguously distinguishes between good dicated by the worst performers listed in table and poor performers. The lists of best and worst 19. Four of these plants have operated at less than plants each contain both PWRs and BWRs, small 50-percent capacity factor throughout their life- and large reactors, new and old units, and util- time.

Table 19.—Performance of Selected U.S. LWRs

Lifetime Design Number of reactors capacity capacity Type of Years of in operation by Plant factor (MWe) reactor operation a same utility Best capacity factorsb: Point Beach 2 ...... 79 497 PWR 11 2 Connecticut Yankee . . . 76 582 PWR 15 1 Kewaunee ...... 76 535 PWR 9 1 Prairie Island 2 ...... 76 530 PWR 8 3 Calvert Cliffs 2...... 75 845 PWR 6 2 St. Lucie 1 ...... 74 830 PWR 6 3 Prairie Island 1 ...... 71 530 PWR 9 3 Monticello ...... 71 545 BWR 12 3 Vermont Yankee ...... 70 514 BWR 10 1 Calvert Cliffs 1 ...... 70 845 PWR 8 2 Worst capacity factorsb: Beaver Valley 1 ...... 34 852 PWR 7 1 Palisades ...... 39 805 PWR 11 2 Davis Besse 1 ...... 40 906 PWR 5 1 Brunswick 2 ...... 41 821 BWR 7 3 Salem 1 ...... 46 1090 PWR 6 2 Indian Point 3 ...... 46 965 PWR 6 2 Brunswick 1 ...... 48 821 BWR 6 3 Rancho Seco ...... 50 918 PWR 8 1 Duane Arnold...... 51 538 BWR 8 1

aBy the end of January 1983. blncludes only plants greater than 100 MWe in operation 3 years or 10n9er. SOURCE “Licensed Operating Reactors, Status Summary Report, data as of 01-03 -83,” U S Nuclear Regulatory Commlsston, February 1983

25-450 0 - 84 - 9 : QL 3 118 ● Nuclear Power in an Age of Uncertainty

Operating difficulties also can be inferred from (5). NRC based its fine on evidence of lax man- NRC’s system of fines and enforcement actions. agement, deficiencies in the training of staff, and NRC recently proposed several fines for alleged inattention to certain safety procedures. safety violations which it claims resulted from NRC also has proposed stiff fines at other util- breakdowns in management controls. The largest ities for difficulties related to management of these penalties is a proposal for an $850,000 controls. Carolina Power & Light Co. has paid fine to be collected from Public Service Electric $600,000 because it failed to develop certain pro- & Gas Co. for problems at its Salem 1 nuclear cedures and conduct tests at its Brunswick pIants powerplant in New Jersey (21). On two occasions in North Carolina, and because its quality-assur- in February 1983, the circuit breakers in the reac- ance staff failed to detect the problem. Boston tor’s automatic shutdown system failed to operate Edison Co. was fined $550,000 for management as designed to shut the reactor down safely. In problems at its Pilgrim 1 plant. These and other both cases, the plant operators initiated a manual examples demonstrate that there are serious man- shutdown and avoided damage to the plant. agement difficulties in some operational plants While the problem can be partially attributed to and that poor management can have important a design flaw in the shutdown equipment, it safety and economic implications. might have been avoided if the automatic shutdown equipment had been properly maintained

MANAGEMENT CHALLENGES IN CONSTRUCTION AND OPERATION

The nuclear utilities have identified a number of obstacles to maintaining quality in the construction and operation of nuclear powerplants; other organizations such as NUS Corp. and EPRI also have investigated the difficulties involved in nuclear operations (1 7,38). For discussion purposes, the factors that adversely affect nuclear power operations can be categorized according to their sources. Some problems arise from the nature of the technology itself; others can be attributed to the external conditions that influence all utilities. While these factors affect the management of all nuclear powerplants, there appears to be a great deal of variation in the ability of util- Photo credit: Atomic Industrial Forum ities to cope with them. A third source of prob- These workers at the Connecticut Yankee nuclear lems arises from the utility management itself. plant are making underwater adjustments to equipment in the reactor vessel. While cumbersome, submersion of the equipment is a protection against radiation Technological Factors That Influence Construction and Operation As presently utilized, nuclear technology is sion process and efforts to sustain, control, and much more complex than other conventional monitor it during normal operation. Nuclear reac- generating sources; this creates difficulties in con- tors have other unique features to contain radio- struction and operation beyond those experi- active material produced as a result of the fission enced in fossil units. Most of the unique charac- process and to protect the work force and the teristics of nuclear powerplants arise from the fis- public. Finally, and most importantly, nuclear Ch. 5—Management of the Nuclear Enterprise ● 119 powerplants are equipped with many levels of tain tens of thousands of elaborate pipe hangers, safety equipment that prevent the release of ra- supports, and restraints that must be designed dioactive material in the eventofanaccident(19). and installed according to highly specific criteria. In contrast, a comparable coal plant might con- Overall, nuclear powerplants are more sophisti- tain only about 5,000 pipe supports whose design cated than fossil-fuel stations. This has obvious and installation is not subject to the same restric- implications for the construction of nuclear units. tive standards found in nuclear powerplants (35). As shown in figure 28, nuclear powerplants re- In view of this, it is not surprising that the con- quire greater quantities of most of the major construction of a nuclear reactor is considerably struction materials than coal plants (34). Certain more labor intensive than that of a coal plant. As components are also more numerous in nuclear shown in figure 28 a new nuclear unit might re- plants than in fossil units. For example, a large quire 64 percent more workhours than a similarly nuclear plant may have 40,000 valves, while a sized coal unit. fossil plant may have only 4,000 (1 7). Another case in which requirements for nuclear reactors The design effort for nuclear reactors also be- exceed those of coal plants is in the area of piping comes increasingly difficult as the plants become supports. Because nuclear plants must be able more complex. A particularly challenging aspect to withstand an earthquake, they are equipped of nuclear powerplant design is anticipating po- with complex systems of piping supports and tential interaction among systems or unexpected hangers designed to absorb shock without dam- failure modes within a single system. As discussed aging the pipes. A typical nuclear unit might con- in chapter 4, it is of vital concern to ensure that

Figure 28.—Comparison of Commodity Requirements for Coal and Nuclear Powerplants

Coal Reinforcing steel and structural steel Nuclear

Coal Concrete Nuclear

Coal Piping Nuclear

Coal Wire and cable Nuclear

Coal Labor Nuclear

SOURCE. United Engineers and Constructors, Energy Economic Data Base, September 1981 (Based on 1,139 MWe PWR and 1,240 MWe high. sulfur coal plant) 120 ● Nuclear Power in an Age of Uncertainty

the smooth operation of safety systems is not im- Coal plants, on the other hand, are much less sen- paired by malfunctions in unrelated and less crit- sitive to scaling factors, and most are less than ical areas. The risk of such adverse interaction half the size of the newest nuclear plants (7). Thus has increased with the steady growth in the num- nuclear construction projects not only involve ber and complexity of nuclear plant components. more sophisticated systems, but also larger ones, In fact, it is not always easy to predict the overall with a work force of 2,000 to 4,000 per unit. This impact of changes that superficially seem to con- can significantly increase the difficulties in coor- tribute to safety. For example, NRC requested dinating and monitoring the activities of the vari- one operating utility to install additional pumps ous parties involved in nuclear construction. to reduce the risk of a Three Mile Island-type ac- In addition to being complex, nuclear technol- cident. An extensive analysis of the system using ogy is very exacting. As discussed above, the safe- probabilistic risk assessment indicated that adding ty of nuclear powerplants is ensured by sophisti- the extra pumps would not necessarily reduce cated control systems that must respond rapidly the risk. It was discovered that the location, not and reliably to prevent an accident or mitigate the number, of pumps was the key to enhanc- its consequences. These control and safety sys- ing safety (8). tems must be constructed, maintained, and oper- Finally, the operation of a nuclear plant be- ated according to the highest standards of excel- comes more difficult as the plant increases in lence. This is so important that NRC has devel- complexity. Many of the routine actions that must oped detailed procedures for monitoring and ver- be taken to control the reactor during normal op- ifying quality. During the construction process, eration or to shut it down during an accident are NRC requires extensive inspection and documen- handled automatically. There are, however, un- tation of all safety-related materials, equipment, usual combinations of events that could produce and installation (13). In response to such require- problems with these automatic safety systems and ments, construction practices have become in- which cannot be precisely predicted. For this rea- creasingly specific and inspection procedures son, nuclear reactor operators are trained to re- have become more formalized. An undesirable spond to unusual situations in the plant. This is consequence of this is that it is extremely difficult not extraordinarily difficult in very simple, small to construct nuclear powerplants in accordance reactors, such as research or test reactors, which with very rigid and explicit standards. One ex- can be designed with a great deal of inherent safe- ample of the complications that can result is in ty and operated with less sophisticated control the installation of pipe supports and restraints. systems. In today’s central power stations, how- It is not uncommon for field engineers to have ever, there are many complex systems that have to work to tolerances of one-sixteenth to one- the potential to interact, making it difficult for thirty second of an inch, which can be more re- operators to respond correctly and rapidly to ab- strictive than the fabrication tolerances used in normal situations. Furthermore, if the control manufacturing the equipment to be installed (35). room design is poor, operators may not receive This results in greater labor requirements than in pertinent data in a timely and understandable fossil plants and can increase the level of skill re- manner. This was part of the problem at Three quired. In addition, various levels of checks, au- Mile Island, where important indicators were on dits, and signoffs are required for most construc- the back of a control panel and unimportant tion work in nuclear powerplants, adding to the alarms added to the confusion of the accident labor requirements necessary to complete instal- sequence (15). lation. One NRC publication has estimated that these checks add 40 to 50 percent to the basic Nuclear units also differ from fossil plants in engineering and labor costs (30). Figure 29 com- their size. The latest generation nuclear pares labor requirements, including quaIity con- powerplants are very large, on the order of 1,000 trol and engineering, for typical coal and nuclear to 1,300 MWe. It was expected that nuclear units plants. Note that for all the items listed here, would be cost-sensitive to size changes and nuclear reactors require at least half again as would be most economical in very large units. much labor as coal plants. Ch. 5—Management of the Nuclear Enterprise ● 121

External Factors That Influence Operation and Construction Certain other factors appear to be less related to the technology than to the environment in which commercial nuclear plants must operate. For example, the nuclear industry has experienced problems with shortages of trained personnel. The commercial nuclear power industry requires engineers, construction crews, and operating teams to be qualified in very specialized and highly technical areas. As shown in figure 30, the demand for technical personnel with nuclear training grew rapidly during the 1970’s (2). At the Photo credit: OTA Staff beginning of the 1970’s, the nuclear work force To keep track of the many components of a nuclear was very small, but many reactors had been or- plant under construction, each of which maybe subject dered and were entering the construction phase. to modification, each pipe and pipe support is labeled with a number that corresponds to Labor requirements grew steadily and peaked in a number on a blueprint the 4-year period 1973 to 1977, when the number of people employed in the nuclear industry increased at the rate of 13 percent a year. In the Another consequence of strict quality control early years of the commercial nuclear industry, is that a large amount of paperwork is generated. the greatest shortages were found among reac- According to one recent estimate, approximate- tor designers. This contributed to the practice of ly 8 million pages of documents have been pro- initiating construction with incomplete designs. duced to support the quality assurance program While it was recognized that 60 percent or more for a nuclear unit that began operation in 1983 of the design should be completed before con- (32). In the midst of such massive requirements struction was initiated, some utilities began with for paperwork, it can be difficult, if not impossi- half that or less. As plants have progressed from ble, to maintain a positive attitude toward quali- the design phase, through construction, and into ty for its intrinsic value. This becomes even more operations, the emphasis on personnel has also difficult in an environment where rework is re- shifted. Reactor designers are no longer in short quired frequently, since this adds to the paper- supply, but there is a need for more people qual- work burden and decreases the morale of the ified in plant operations, training, and certain en- workforce. gineering disciplines (1 2). The exacting nature of nuclear technology A second external problem is inadequate com- manifests itself somewhat differently during op- munication among utilities. Only a few utilities eration. It often is necessary to maintain extreme- had any experience with nuclear power before ly tight control of sensitive systems to keep the the 1970’s. A structured method for transferring plant running smoothly. For example, the water learning might have accelerated the overall pro- chemistry system in LWRs must be carefully mon- gress by providing warnings about common er- itored and adjusted to prevent corrosion and re- rors and transmitting effective problem-solving move radioactive materials from the cooling approaches. Such communication networks did water. Failure to maintain these systems within not exist in any formal manner until the accident narrow limits can lead to severe damage in such at Three Mile Island stimulated an industrywide major components as steam generators or con- effort to improve the transfer of nuclear opera- densers and this can ultimately curtail plant tions information. Even today there is little struc- operations (36). As discussed in chapter 4, cor- ture in sharing information regarding reactor con- rosion has been a serious problem in many oper- struction, with the primary mechanism being the ating PWRs and has led to replacement of steam transfer of trained personnel from one utility to generators in four nuclear units. another. 122 ● Nuclear Power in an Age of Uncertainty

Figure 29.—Comparison of Manpower Requirements for Coal and Nuclear Powerplants

Reinforcing steel Coal and structural steel Nuclear

Coal Concrete Nuclear

Coal Piping Nuclear

Coal Wire and cable Nuclear

Coal Welding Nuclear ( + 480/o)

Coal Hangers and restraints Nuclear

SOURCE. J D. Stevenson and F. A. Thomas, Selected Review of Foreign Licensing Practices for Nuclear Power Plants, April 1982

An additional consideration is that nuclear reac- dle. As a consequence, the utilities had to divert tor owners have had to deal with a rapidly chang- scarce engineering forces from design and review ing regulatory environment throughout the past activities to deal with NRC (37). decade. Frequent revision of quality and safety The utilities had to deal with more than a steady regulations and backfit requirements have greatly increase in regulatory requirements: a series of affected construction and operation patterns. As regulatory “shocks” was superimposed on the shown in figure 31, NRC issued and revised reg- cumulative effect of “normal” regulation. A study ulatory guidelines at an average rate of three per by EPRI identifies three major events that were month in the mid-1970's (33). followed by a flurry of NRC activity: the Calvert Plants that were under active construction dur- Cliffs decision in 1971 to require Environmental ing this time had to continually adjust to the Impact Statements for nuclear plants, the fire at changing regulatory environment. While no the Browns Ferry nuclear plant in 1975, and the single NRC requirement overtaxed the utilities accident at Three Mile Island in 1979 (3). The with plants under construction, the scope and aftermath of these incidents has created an at- number of new regulations were difficult to han- mosphere of regulatory unpredictability that has

Ch. 5—Management of the Nuclear Enterprise ● 123

Figure 30.—Historical Labor Requirements in the Nuclear Power Industry

Nuclear facility engineering, design, operation, and 150 maintenance

100 Miscellaneous (including consulting, lab testing, radiography, etc.)

50 Fuel cycle

Manufacturing and production o I III IIIIIIIII1 I 1970 1972 1974 1976 1978 1980 1982 Year

SOURCE ‘ Occupational Employment in Nuclear-Related Activities, 1981 .“ Oak Ridge Associated Universities for the Department of Energy, April 1982 particularly affected plants in the construction Finally, rapid technological changes have fur- phase. In some cases, major portions of nuclear ther complicated nuclear powerplant construc- construction projects have had to be reworked tion and operation. Nuclear reactors were scaled to comply with changing requirements. For ex- up from the earliest demonstration plants of sev- ample, after the fire at the Browns Ferry plant, eral hundred megawatts to full-scale 1,000-MWe NRC issued new requirements to fireproof all plants within a decade, By 1968, most orders trays carrying electrical cables. While this was not were placed for units greater than 1,000 MWe. an unreasonable request, it did disrupt many con- As shown in table 20, there were only three LWRs struction schedules. with a generating capacity greater than 100 MWe in operation in the United States when the first In many cases, changes in NRC regulations ob- of these orders was placed. Thus the designs for viously enhance plant safety. In other cases, it the larger plants were not built on the construc- is not clear that safety is increased by adding or tion and operating experience of gradually scaled modifying systems and components. As discussed units. By the time the first 1,000-MWe units began in chapter 6, the adverse impacts of certain reg- operation in 1974, an additional 70 large plants ulations include equipment wearout due to ex- had been ordered. There was little opportunity cessive surveillance testing and restrictions on ac- for orderly, deliberate design modification and cessibility to vital equipment as a result of fire or transfer of knowledge in this rapid scale-up. security barriers (37). The factors discussed above have contributed Piping system design provides another exam- to the complicated task of maintaining rigid stand- ple of possible adverse effects of regulation. The ards of excellence in nuclear powerplants. As a current trend in NRC guidelines is to require more result, the construction and operation of nuclear rigidly supported systems. This is not necessari- reactors has demanded the full resources, both ly because flexible systems are less safe, but ana- technical and financial, of the utilities. Many util- ytical techniques cannot unequivocally prove ities have failed to fully meet these challenges. them safe. Rigid systems are easier to analyze, Others, however, have managed to cope with all but can present serious operational difficulties of these complications—plants have been con- during routine changes in temperature (23). structed with few major setbacks and operated 124 ● Nuclear Power in an Age of Uncertainty

Figure 31.— NRC Regulatory Guidelines Issued From 1970 to-1980 40(

300

. Photo credit: Atomic Industrial Forum Cable trays increased in weight and complexity because of fire-proofing and separation of function requirements following the fire in the electrical system at Browns Ferry in 1976

Table 20.—Early Operating Experience of U.S. Commercial Light Water Reactors

Date of commercial Unit Size (MWe) Operation Type Dresden 1 ...... 100 207 8/60 BWR Yankee Rowe ...... 175 6/61 PWR Big Rock Point . . . . . 63 12/62 BWR Humboldt Bay 3 . . . . 63 8/63 BWR Connecticut Yankee . 582 1/68 PWR San Onofre 1 ...... 436 1/68 PWR La Crosse ...... 50 11/69 BWR Nine Mile Point 1 . . . 610 12/69 BWR Oyster Creek 1 ...... 620 12/69 BWR SOURCE “Update — Nuclear Power Program Information and Data, October- 0 December 1982, ” U S. Department of Energy, February 1983 1970 1975 1980 Year ending clear technology from that of other conventional power technologies. The complexity of the reactor and the demands for precision and documen- well. This suggests that some of the variability among utilities can be attributed to differences tation provide significant challenges to utility in factors internal to the utility. managers. Even more important are the difficulties in deal- Internal Factors That Influence ing with a changing environment. Successful utili- Construction and Operation ty managers have had to maintain a great deal of flexibility to keep up with the rapid growth in Factors related to utility management are dif- the size and design of nuclear plants and changes ficult to assess since they are less visible than ex- in regulatory structure. In fact, some utilities have ternal factors; moreover, they are not easily quan- reorganized several times in an attempt to con- tified. Nonetheless, they can influence the finan- trol their nuclear projects better. The most com- cial success of a nuclear project or plant safety. mon changes have been away from traditional As discussed above, there area number of char- line management and towards matrix or project acteristics that distinguish the management of nu- management (3). While this has been successful Ch. 5—Management of the Nuclear Enterprise ● 125

in some cases, it is not always sufficient to im- Corporate commitment also can be indicated prove the quality of utility management. Other by the way in which a utility responds to changes factors are also very important, as discussed or problems. The more successful utilities have below. a history of responding rapidly and with adequate financial resources to resolve problems and adapt Managing nuclear power projects requires a to new situations. Other utilities with less eager- commitment to safety and quality that is less ness to confront their problems directly have ex- essential in other electric utility operations. This perienced construction delays and operational implies far more than a concern for schedules and difficulties (3). budgets, which pervades all commercial endeavors. Because there is some possibility that an ac- An important factor in the management of any cident could occur in a nuclear reactor, every ef- powerplant is the distribution of responsibility fort must be made to protect the investment of and authority. This is particularly vital in the con- the utility and the safety of the public. It is im- struction of nuclear plants because of the com- portant that nuclear managers adhere to the spirit plexity of the technology and the need to co- as well as the letter of safety and quality-assurance ordinate the activities of vendors, architects, engi- reguIations. neers, construction managers, consultants, quali- The Palo Verde plants are good examples of commitment to quality (6). When Arizona Pub- lic Service announced its nuclear program in 1972, it thoroughly studied all aspects of designing and constructing nuclear powerplants. Many advanced safety features were incorporated into the Palo Verde design from the beginning of the project. One unexpected consequence of this attention to safety is that Arizona Public Service anticipated many of the Three Mile Island backfit and redesign requirements. As a result, regulatory changes in response to Three Mile Island had less impact on the Palo Verde projects than on many other plants which had not originally planned to incorporate the extra safety features. Sincerity of commitment can be observed in several ways. Highly committed senior managers can impress their commitment on project managers, who in turn can communicate it to designers, manufacturers, and constructors, The strength of utility commitment is also indicated by the level of quality required in the utility’s contractual and procedural arrangements with suppliers of material, equipment, or personnel, For example, if a contract primarily emphasizes the schedule for physical installation, the message from project management is production. On the other hand, if the contract also emphasizes Photo credit: OTA Staff owner-acceptance and adequate documentation, As a plant nears completion, responsibility is gradually the message is quality as well as production. The transferred from the construction managers to the operating division. These tags give an idea of latter case provides the proper incentives for high- the detailed level at which explicit quality work (1 3). responsibility is assigned 126 ● Nuclear Power in an Age of Uncertainty

ty inspectors, test engineers, operators, and crafts- These four plants were troubled by the inability men. In this environment, it is vital that a utility to assure continued financing and the decreas- establish clear lines of authority and specific re- ing need for power in the Northwest. The man- sponsibilities to ensure that its objectives will be agement restructuring came too late to graceful- met. ly reverse the effects of early planning decisions (l). When authority and responsibility are diffused throughout an organization rather than focused The example discussed above is only one of in a few key positions, widespread problems can many in which utilities learned that it is in their result. This occurred at the Washington Public best interests to monitor and control the activities Power Supply System (WPPSS) in the late 1970’s, of their constructors and architect/engineers. It when extensive construction delays and regula- was common in the 1970’s for utilities, especial- tory difficulties plagued the WNP-2 plant in Rich- ly those with little experience, to relinquish most Iand, Wash. (9). Responsibility for design and of the responsibility for design, cost, and schedule construction was distributed between the owner to their contractors. As problems developed, the and the construction company in an obscure utilities gradually became more involved with manner, with neither the owner nor the construc- their constructors; this resulted in shifting respon- tion manager claiming authority in decisionmak- sibility to the plant owners. The same oversight ing nor accepting responsibility for mistakes. Ad- could be applied to architect/engineers, and there ditional problems arose when authority was fur- are recent signs that this is happening in the larger ther delegated to the contractors without suffi- and more experienced nuclear utilities. cient provisions for monitoring feedback. While Once a utility has developed a workable orga- most contractors assumed the responsibility for nizational structure, it is further challenged to maintaining quality, others were less conscien- coordinate and motivate the many diverse tious. In some cases unqualified people were groups of people involved in nuclear construc- used for construction work, and documentation tion. At the peak of construction on a large nu- was incomplete and inconclusive. The project clear unit, as many as 6,000 craftsmen, engineers, management effectively lost control of the sub- and support personnel may be working together. contractors and was in no position to detect and In such situations, scheduling can become a correct these problems. logistic nightmare, and a sense of teamwork and having common goals can vanish. These prob- When this situation came to the attention of lems are exacerbated by changing regulatory re- NRC in 1980, a stop-work order was issued for quirements that can result in construction rework WNP-2. WPPSS tackled its problems directly by and delays and by the lack of continuity that re- completely restructuring its project management. sults from long construction times (1 7). It established clear and direct lines of responsibility for design and construction by creating the new Coordination can be equally challenging within position of Project Director and by specifying the a utility’s management structure, especially if the role of the construction manager reporting to utility is undergoing organizational changes. This him. New review and surveillance procedures occurred at Commonwealth Edison Co. in Chi- were initiated by the construction manager and cago, where a matrix-management structure was overseen by WPPSS. Construction finally was re- replaced with formal project management. In the sumed when NRC was satisfied that the major new organization, each nuclear construction proj- deficiencies had been resolved, and WNP-2 is ect was given its own staff, including engineer- nearly complete. The four other WPPSS plants ing, construction, testing, and startup personnel. under construction were less fortunate. Two units An independent staff of quality and safety in the early stages of construction were moth- engineers was maintained in a central office to balled in 1981, and WPPSS has since defaulted provide oversight and ensure uniformity among on repaying the outstanding debt on those plants. the project teams. There are some overlapping Subsequently, construction was halted on 2 other and conflicting functions in the new system, and plants that were more than 60 percent complete. strong leadership from the management within Ch. 5—Managemnt of the Nuclear Enterprise Ž127

Commonwealth Edison has been required to in- construction times (3). That study further con- still a sense of teamwork among the various cludes that a utility can compensate for lack of groups (25), experience by relying on an architect/engineer or constructor that has previously dealt with nu- Another problem that some utilities face is that Iear projects. nuclear projects may make excessive demands on their resources. Some utilities have not been Lack of experience was a major source of able to hire qualified management personnel, and Houson Lighting& Power Co.’s problems in con- they have not had sufficient time to gain the man- structing its South Texas projects. It selected the agement and technical experience independent- AE firm, Brown & Root, Inc., even though that ly. As previously discussed, this often resulted in firm was inexperienced with large-scale nuclear construction being started with incomplete de- plants. After a number of quality problems came signs. Furthermore, limited resources have made to light, NRC issued a stop-work order in 1980 it difficult for some utilities to provide for ade- for all safety-related construction. Houston Light- quate training in quality inspection and reactor ing & Power is attempting to resolve these dif- operations while simultaneously constructing a ficulties by replacing Brown & Root with the more nuclear plant. experienced firms of Bechtel Corp. and Ebasco Services, Inc.; they are also acquiring in-house A final consideration is experience in construc- capability by hiring well-trained engineers (28). tion and operation. It is more difficult for a util- This latter approach has been used successfully ity with no experience to cope with nuclear by other utilities. When Arizona Public Service power’s unique characteristics than it is for a util- Co. initiated construction of its first nuclear pow- ity with several nuclear plants. In fact, a recent erplants, it expanded its staff with engineers and EPRI study concludes that nuclear experience is managers experienced in nuclear power. one of the most significant variables influencing

IMPROVING THE QUALITY OF NUCLEAR POWERPLANT MANAGEMENT

The problems discussed above indicate that attempt to raise the quality of management great dedication is needed to manage nuclear through institutional controls. powerplants. This technology presents many challenges to successful management, and not all utilities have demonstrated that they have the Technical Approach skill, resources, and commitment to meet the The technical improvements that could be challenge. made to the current generation of nuclear reac- Several different approaches can be taken to tors range from minor evolutionary modifications alleviate management problems. One possibility to clean-sheet designs. As discussed in detail in for reactors that will be sold in the future is to chapter 4, a recent EPRI survey indicates that redesign them to be simpler and safer, and hence many utilities would like to see at least minor less susceptible to management control failures. changes in new LWRs to enhance conservatism, This approach suggests reactors that are more reliability, operability, and maintainability (1 7). “forgiving” of human errors than current LWRs, It was proposed that the next generation of LWR as discussed in chapter 4. A parallel effort might designs focus on simplicity, reduced sensitivity 128 ● Nuclear Power in an Age of Uncertainty

to anticipated transients, and reduced system in- must accompany the construction and operation teractions. Modifications of this sort could relieve of at least the nuclear island of any reactor. Nor some of the pressures on management by requir- would a new design eliminate the demands for ing less precision during both construction and management skills during the construction proc- operation. ess—effective distribution of responsibilities would continue to be important, as would the Management benefits also could be derived need to coordinate and motivate the construc- from standardizing the LWR, with or without tion work force. In short, inherently safe reactors modifications. This would allow the utilities to might markedly reduce the problems that arise transfer learning from one unit to another, and from technological considerations, but new de- all plants could gain from the experience of any signs cannot alone ensure high-quality construc- plant owner. Another potential benefit of stand- tion and safe and reliable operations. ardization is that the regulatory climate is likely to be less active once an industry-wide design has been selected and approved. As a result, stand- Institutional Approach ardization should reduce the frequency of NRC- inspired design changes (22). Institutional approaches focus on internal and external factors that affect performance rather The modifications discussed above fit within than on technical considerations; in this sense, a pattern of evolutionary development of LWR they complement the activities taken to reduce technology. They do not involve dramatic changes the complexity and sensitivity of nuclear reactors. to components or to the basic reactor design, but Institutional measures can be divided into two focus on reconfiguring the current system. These types of activities: changes would be welcomed by the nuclear utilities, and they could made nuclear reactors some- ● those that create a favorable environment what easier to manage. It is unlikely, however, for successful utility management of con- that they would significantly reduce the overall struction and operation. Such activities level of management intensity required to han- would focus on external problems, including dle nuclear projects. It is possible that this could efforts to enhance communications, increase be accomplished by a more innovative and dras- the supply of trained personnel, and stabilize tic alteration to the present technology. the regulatory environment; and ● those that monitor utility operations to detect More extreme technical alternatives include re- management failures and elicit better per- designing the LWR to optimize it for safety or re- formance from the less successful utilities. placing the LWR with an alternative design. These Such efforts would focus on problems that reactors might prove to be less sensitive to man- are specific to individual utilities. agement control failures than current LWRs. For example, the Fort St. Vrain high temperature gas- Two principal organizations are now involved cooled reactor (HTGR) has experienced several in institutional controls that monitor operations incidents that have had no significant conse- and improve communications. The NRC has long quences, but which might have been serious in been involved in programs designed to regulate an LWR. In one incident, forced circulation cool- the nuclear industry and to protect the public. ing was interrupted for more than 20 minutes. In the past, its initiatives were focused primarily Because the HTGR has a high heat capacity and on design and licensing issues for reactors in the the fuel has a high melting point, there was no construction phase. As the nuclear industry con- damage to the core or components. tinues to mature and more plants enter operation, it is expected that the emphasis gradually Even drastic changes in technology, however, will shift to operating plants. cannot eliminate all of the problems that face managers of nuclear projects. In particular, modi- The Institute of Nuclear Power Operations fication of the technology cannot replace high- (IN PO) is a more recent participant in this area, Ievel commitment to quality and safety, which and its influence is growing rapidly. INPO is spon-

Ch. 5—Management of the Nuclear Enterprise ● 129

Photo credit: Pennsylvania Power & Light

Site-specific control room simulators, used increasingly by utilities to train nuclear plant operators, are duplicates of actual control rooms. Possible nuclear operating events are simulated on control instruments by computer. The simulator shown here trains operators of PP&L’s Susquehanna plant and cost $6 million. sored by the nuclear utility industry, and every INPO has been active in collecting, evaluating, utility with a nuclear plant in operation or under and redistributing utility reports of operating construction is a member. In addition, utility experience. This is particularly important in view organization in 13 other nations participate in of the massive amount of information that is gen- IN PO. It was formed in 1979 in response to the erated by operating nuclear powerplants. It is a accident at Three Mile Island. challenging task to distinguish the vitally important from the more mundane reports. INPO de- Creating the Right Environment veloped the Significant Events Evaluation and In- formation Network (SEE-IN) to handle informa- An area that has received considerable attention on an industrywide basis. In 1982, more than tion is the improvement of communication 5,000 event reports were screened, and approxi- among utilities. The utilities have combined mately 100 significant reports relating directly to forces to form the Nuclear Safety Analysis Center plant reliability and safety were identified (41). (NSAC) to analyze technical safety problems and The most frequently cited problems involve solutions. NSAC has already addressed a number valves and electrical and instrument controls, of issues, including the accident at Three Mile closely followed by the reactor coolant system, Island, NRC’s unresolved safety items, and de- steam generators, diesel generators, and piping. graded cores. After additional review, INPO distilled this infor- 130 . Nuclear Power in an Age of Uncertainty

mation into a few important recommendations. portant items by linking all operating nuclear Among these recommendations are measures to plants in a single computer network (1 1). preclude equipment damage, reduce prolonged INPO also is active in developing guidelines in outages, and minimize radiation exposure. the areas of training accreditation, emergency re- In addition to identifying generic problems in sponse, and radiation protection (39). These activ- the industry, the SEE-IN program also checks to ities are coordinated with NRC needs and re- see if its recommendations are implemented. quirements. There is currently a great deal of vari- Thus far it has been very successful in encourag- ability in the ways in which utilities handle probing utilities to make voluntary changes to com- lems in these areas. As greater uniformity devel- ply with its recommendations. Nearly half of all ops, the utilities should be better able to learn operating plants implemented every “immediate from one another and raise the level of perform- attention” recommendation within a year, and ance on an aggregate basis. many other plants are making progress in this Another important activity within NRC is the direction (41). effort to moderate regulatory activity by control- NRC is planning a similar program in which it ling requirements for changes during construc- will systematically analyze the information that tion and operation. In 1981, NRC established the it collects through various reporting mechanisms. Committee for Review of Generic Requirements NRC plans to computerize its data base and search to assess backfitting proposals and to try to reduce methods so that it can better detect generic prob- the burden of shifting requirements. As discussed lems (13). in chapter 6, this is a difficult task since safety is not easily quantifiable, and many technical uncer- A related NRC activity is focused on construc- tainties remain. However, this effort has the po- tion rather than operation. A long-term effort to tential for enhancing the ability of utilities to con- review quality-assurance problems and to pro- struct new plants in a timely and efficient manner. pose changes that could improve quality in design, construction, testing, and operation has Detecting and Improving been initiated. This review will start with an exam- Poor Performance ination of nuclear plants at Diablo Canyon, South Texas, Midland, Marble Hill, and Zimmer to iden- Improvements have been made in certain as- tify specific problems and causes. At various times pects of the commercial nuclear industry in re- in the past, NRC has issued stop-work orders at cent years. The accident at Three Mile Island each of these plants due to concerns about the convinced many utilities of the importance of at- quality of construction. NRC will also examine tention to quality, and some have made volun- projects with good records to identify the positive tary changes to improve management. For exam- measures that could be applied generically (13). ple, a number of utilities have modified their organizational structure in an attempt to find one The data analysis efforts by both NRC and better suited to building and operating nuclear INPO should enhance the formal transfer of powerplants (3). learning among the utilities. Other INPO activities attempt to improve communication less formal- The utilities that are sensitive to quality con- Iy by providing a forum for the exchange of ideas. cerns and responsive to NRC and INPO initiatives Managers involved in nuclear plant construction are probably not a source of great concern; and operation are encouraged to meet at work- rather, the concern is centered on those utilities shops and conferences to share their experiences that do not seem to be responding to the same in detecting and solving problems. Another way motivation. In fact, the most successful utilities in which INPO encourages the exchange of in- claim that they are being “held hostage” by the formation is through its electronic communica- least capable and least committed utilities (23). tion system known as “Nuclear Notepad.” This They fear that another major nuclear accident in system provides timely transfer of news on im- any commercial reactor would have disastrous Ch. 5—Management of the Nuclear Enterprise • 131 consequences for all nuclear plant owners in INPO evaluation reports have proven to be val- terms of public acceptance. It is, therefore, in the uable in several ways. First, they form the basis interest of the best performers to ensure that poor for INPO’s “good practice” reports, which sum- practices are detected and eliminated, wherever marize effective approaches used throughout the they occur. This concern has Ied to INPO’s ex- nuclear industry. These reports are particularly tensive evaluation and assistance programs to at- useful to utility managers who want to identify tain excellence. NRC also evaluates utility per- problem areas and adopt approaches that have formance, but with a different perspective. Its in- been used successfully in other plants. tent is to ensure that construction and operation of all nuclear units meet minimum standards, as The second major contribution of INPO evalua- defined by NRC. tions is to highlight problem areas in individual utilities and make recommendations to improve INPO’s evaluation programs appear to be well- performance. In the event there is some reluc- structured and in logical relation to one another. tance on the part of a utility to comply with INPO The first INPO efforts were devoted to establish- recommendations, a number of actions can be ing a comprehensive system for evaluating op- taken to encourage cooperation. These actions erating nuclear reactors. In an operating plant are designed to raise the performance of all util- evaluation, special teams of up to 15 people are ities by applying peer pressure, from other leaders sent to each nuclear unit to assess the perform- in the industry. These pressures could be applied ance of the utility in many different areas, in- through utility chief executive officers, boards of cluding those shown in table 20. A final report directors, and insurers. Such actions have not yet is prepared to summarize the findings, make rec- been required. ommendations for improvements, and identify good practices. This report is reviewed with the The NRC has its own series of plant inspections. highest levels of utility management, who devel- Starting in 1978, resident inspectors were located op a plan of action for implementing INPO rec- at each nuclear plant to monitor day-to-day oper- ommendations (26). ations. These inspectors provide much of the in- INPO has completed two rounds of operating formation that is used to develop the Systematic plant evaluations and has initiated a program to Assessments of Licensee Performance (SALP), evaluate construction projects. Construction eval- which are prepared by NRC’s five regional of- uation procedures have been developed and fices. The SALP’s analyze performance in each have been applied to 18 near-term operating of 10 categories, which are similar to the INPO licensee plants. The first phase of the construc- categories shown in table 21. The goal of this tion assessment program was completed in 1982 evaluation is to identify areas in which manage- when 22 utilities with nuclear plants under con- ment excels, areas which call for minor improve- struction performed self-evaluations. The second ments, and those in which major weaknesses are phase of evaluations is being conducted by either evident. NRC uses this assessment to direct its in- INPO or independent organizations under con- spection efforts and to suggest changes to the plant owners. tract to the utilities and monitored by IN PO. NRC has been following INPO’s evaluation efforts A more comprehensive NRC evaluation effort closely and may restructure some of its own quali- involves the Performance Assessment Team ty initiatives around the industry program (39). (PAT). This team operates from NRC headquar- In addition to evaluating nuclear plants, INPO ters, and its inspections provide a check on the is assessing the corporate support of nuclear util- NRC regional offices and the INPO evaluations. ities. Corporate evaluation criteria have been de- Although the PAT evaluations overlap the SALP’s, veloped by a task force of senior utility execu- they are broader in scope, with assessments of tives. These criteria have been field-tested and management controls and broad recommenda- are in use in INPO evaluation programs (26). tions for change (24). 132 ● Nuclear Power in an Age of Uncertainty

Table 21.—Classifications for INPO Evacuations 3-to 4-year cycle. They were limited to only two to three inspections per year after 1982. Organization and administration: Station organization and administration; management objectives; management assess- Another phase of the NRC inspection program ment; personnel planning and qualification; industrial safe- focuses on near-term operating licensees (NTOL). ty; document control; station nuclear-safety review group; quality programs To increase its confidence in the quality-assur- Operations.K)perations organization and administration; con- ance programs at plants that will soon begin op- duct of operations; plant-status controls; operator knowl- eration, edge and performance; operations procedures and docu- NRC now requires a self-evaluation of mentation; operations facilities and equipment quality-assurance programs for design and con- Maintenance: Maintenance organization and administration; struction at these plants (13). This includes a plant material condition; work-control system; conduct of maintenance; preventative maintenance; maintenance pro- review of management involvement, audits, sig- cedures and documentation; maintenance history; main- nificant problems, and corrective actions. The tenance facilities and equipment; materials management self-evaluations are supplemented by NRC re- Technical support: Technical-support organization and administration; surveillance-testing program; operations- gional office reviews. These assessments examine experience review program; plant modifications; reactor the inspection and enforcement history of the engineering; plant-performance monitoring; technical- project to determine whether additional inspec- support procedures and documentation Training and qua//f/cation: Training organization and admin- tions are needed. In addition, NRC encourages istration; licensed and nonlicensed operator training and independent design reviews at all NTOL utilities. qualification; shift-technical-advisor training and qualification; maintenance-personnel training and qualification; The purpose of the NRC inspection activities training for technical staff; training for managers and is to identify severe or recurrent deficiencies. engineers; general employee training; training facilities and There is less effort made to analyze the structure equipment Radiological protection: Radiological-protection organization and commitment of utility management than to and administration; radiological-protection personnel train- identify problems that might arise from the failure ing and qualification; general employee training in radiological protection; external radiation exposure; internal of management controls. Thus, NRC evaluations radiation exposure; radiological-protection instrumentation serve the purpose of indirectly monitoring the and equipment; solid radioactive waste; personnel dosim- sources of problems by directly monitoring their etry; radioactive-contamination control Chemistry:Chemistry organization and administration; chem- manifestations. In contrast, the INPO evaluations istry-personnel training and qualification; chemistry con- focus directly on weaknesses in management sys- trol; laboratory activities; chemical and laboratory safety; tems and controls. radioactive effluents Emergency preparedness: Emergency-preparedness In the event that any of the NRC inspections organization and administration; emergency plan; emer- uncover major problems, NRC has recourse to gency-response training; emergency facilities, equipment and resources; emergency assessment and notification; a series of progressively severe penalties. Enforce- emergency-personnel protection; personnel protection; ment actions include: formal notification of a vio- emergency public information. lation; imposition of a fine if the licensee com- SOURCE: Institute of Nuclear Power Operations mits a major violation, willfully commits a violation, or knowingly fails to report a violation; and In 1982, NRC decided to limit the number of finally, the modification, suspension, or revoca- PAT inspections in recognition of the similarity tion of a license. In the most extreme cases, NRC to INPO’s programs. The PAT program original- can refer the case to the U.S. Department of jus- ly was scheduled to evaluate each reactor on a tice for investigation of criminal violations.

ASSESSMENT OF EFFORTS TO IMPROVE QUALITY The management of commercial nuclear of U.S. plants is quite good, the reliability of the powerplants has proven to be a more difficult task plants has been less than hoped for, and several than originally imagined by the early proponents accidents have occurred which have reduced the of nuclear technology. While the safety record confidence of the public in the industry. Further- Ch. 5—Management of the Nuclear Enterprise Ž 133

more, construction projects have been plagued requirements; INPO could further improve the with cost and schedule overruns and questions situation by enhancing communication among about quality. the utilities. The more difficult and important changes, however, relate to the internal manage- Nuclear power is not so intractable that it can- ment of nuclear utilities. Utility managers must not be managed in an exemplary fashion; this has become aware of the unique demands of nuclear been demonstrated by the records of the most technology, and they must develop the commit- capable utilities. However, utilities with only ment and skills to meet them. INPO could be in- average skills and commitment have been much strumental in stimulating this awareness and in less successful in managing nuclear projects. Bet- providing guidance to the utilities. INPO recog- ter approaches are needed to improve the opera- nizes this point and is striving to develop such tion of today’s plants and to establish public con- utility management awareness. It is likely that the fidence that the utility industry could manage utilities will tend to be more receptive to INPO new reactors if they are needed in the future. than to an outside organization since INPO is a Both technical and institutional changes could creation of the nuclear industry. help improve the management of nuclear power It is equally important that the nuclear utilities operations. Technical modifications would be be evaluated objectively to assure that they are useful insofar as they decrease the complexity performing well. NRC and INPO have recognized and sensitivity of nuclear plants. Some of these the need for such evaluations, and both organiza- changes are relatively simple to make and have tions currently are engaged in assessment activ- been incorporated in the design of new LWRs. ities. The INPO assessments, which now cover It is likely that other more drastic design changes many areas, continue to evolve, and so far ap- could further decrease the sensitivity of nuclear pear to have been handled with sensitivity and reactors to their management by making them insight. The INPO evaluations attempt to assess inherently safer and less dependent on engi- the performance of the utility management to neered systems. identify the root causes of the problems and rec- While a technical solution to all management ommend corrective actions. The NRC inspection problems would be welcome, it is not likely to program is more fragmented and somewhat unfo- be forthcoming. Even if an ultrasafe reactor could cused. The relationships among the various in- be developed, the demands on its operators to spection activities are not always clarified, ensure reliable performance would still be greater although these activities should complement one than in a fossil plant. Furthermore, even drastic another. Furthermore, most of NRC’s inspection changes in reactor design would not significantly activities concentrate on the consequences of decrease the sophistication or complexity of the quality problems rather than on the sources. It nuclear island, even though they might allow a should be noted that NRC does try to identify reduction in the safety requirements for the re- management control failures once a problem sur- maining of plant systems. Overall, nuclear con- faces, but that this is not a part of its standard struction managers still would be taxed to coor- evaluation procedure. dinate massive projects amid exacting require- INPO and NRC communicate with one another ments for high levels of quality and extensive concerning their respective evaluation and in- documentation. spection activities, and they are attempting to Since technological changes cannot by them- coordinate their programs. In establishing their selves eliminate all the difficulties involved in con- respective roles, it should be noted that the INPO structing and operating nuclear units, it is impor- evaluation teams may be better able to commu- tant that they be supplemented with institutional nicate with utility managers and discover the measures to improve the management of the source of problems. But this does not imply that nuclear enterprise. For example, NRC could re- NRC should turn over its inspection activities to duce pressures on utility managers by exercising IN PO; the public must have confidence that the as much care as possible in expanding regulatory utilities are being evaluated objectively and ac-

25-450 0 - 84 - 10 : QL 3 134 . Nuclear Power in an Age of Uncertainty

curately, and only a government organization can However, the next few years should provide the guarantee such objectivity. NRC currently accom- evidence needed to evaluate these initiatives. It plishes this by carefully monitoring INPO activi- is not yet clear that significant improvements will ties and by performing independent assessments occur in management of construction since there on a limited basis. However, NRC also performs are few formal mechanisms for transferring learn- a variety of other detailed evaluations that are not ing or developing more successful approaches. well-coordinated with INPO activities. Some du- It is possible that operational reliability and safe- plication of effort maybe appropriate, since NRC ty will improve noticeably if the NRC and INPO must remain objective and informed in its over- initiatives are successful. Improvements in plant sight role. However, it should be possible to bet- reliability should be reflected in increased capaci- ter coordinate the activities of the two organi- ty factors and availabilities and in decreased zations to make better use of limited resources forced outage rates. Industry efforts to improve and relieve the utilities of redundant inspections. component reliability and enhance maintenance and operation should start showing results soon. Enforcement activities also can be very important in raising the level of the poorest utility per- Improvements in safety are more difficult to formers. Both NRC and INPO have a number of measure, but one indication of plant safety is the measures at their disposal to encourage utilities frequency and severity of events that could lead to make changes or penalize them if they don’t to accidents. These are known as precursor cooperate. INPO operates through peer pressure, events, and NRC requires that they be reported and it is not clear that it would actually invoke routinely. If there is a significant decline in the its strongest measures. INPO has not yet found number or severity of precursor events in the it necessary to exercise all its options. coming years, it is likely that private and Govern- ment efforts are achieving some measure of suc- NRC operates on a different level with a series cess in increasing safety. Conversely, if incidents of enforcement actions that can be taken if it such as the loss of the emergency shutdown sys- detects an unwillingness of the nuclear industry tem at the Salem nuclear plant continue to oc- to regulate itself with sufficient stringency. NRC cur, it will be difficult to place much confidence has proven willing to exercise the option of fin- in the effectiveness of the efforts to improve ing utilities when it detects breaches of security safety. or safety regulations. It may be very difficult to achieve significant It is difficult at this time to assess the effec- gains in performance in an industry with so many tiveness of the efforts of the nuclear industry and different actors and such diverse interests and its regulators to improve plant performance. talents. The industry’s support for IN P<) is a ma- There is not yet any clear evidence that the util- jor step in generating a uniform level of excel- ities have been able to translate NRC and INPO lence. However, only time will tell if INPO can programs into better reliability and fewer safety- remain both strong and objective and if all util- related incidents. INPO is still in the process of ities will commit themselves to high standards in establishing its guidelines and evaluation proce- construction and operation. dures, and NRC is just starting to assume a more active role in evaluating management controls.

NEW INSTITUTIONAL ARRANGEMENTS

In the previous discussions, we saw that many be a significant step in this process of improve- utilities have built and operated nuclear power- ment. NRC also is encouraging quality in nuclear plants safely and reliably, and others are now power management as a way to achieve safety. working to improve the quality in construction However, all of these efforts from within and out- and operation. The creation of INPO appears to side the utility industry may not be sufficient to Ch. 5—Management of the Nuclear Enterprise . 135 provide assurances to the public and investors The trend toward greater vendor responsibili- that adequate levels of economy and safety are ty for construction management may be helped being achieved. indirectly by the current financial problems in the nuclear industry. Cost uncertainties make it un- in the introduction to this report the many ac- likely that utilities will order new nuclear plants tors and institutions involved in nuclear energy unless they can be assured of a fixed price. If in- were described. One of the keys to breaking the flation were more moderate and licensing uncer- present impasse among these institutions is a clear tainties reduced (perhaps through the use of demonstration that all utilities with nuclear reac- standardized and preapproved designs), vendors tors are operating them safely and reliably. If the might offer some type of fixed price as they did efforts to improve utility management described with the turnkey contracts of the early 1960’s. thus far are insufficient to satisfy all these actors, However, it is unlikely that vendors would grant it is unlikely that new plants will be ordered. In this type of contract unless they were assured of this case, a future for conventional nuclear power greater control over construction. It has been sug- may require changing the existing relationships gested that a single person within the NSSS com- among the various actors or creating new insti- pany be given point-source responsibility for safe- tutions. ty, quality control, and construction management It should be noted that the potential advantages of the nuclear island. Westinghouse assumes of these new entities are only speculative at this these responsibilities in its international projects, time. Some industry problems such as the overall (where licensing is less complex) and has had shortage of qualified personnel would not be good experience with the approach. Because it helped by simply rearranging people and institu- has greater control, the vendor is able to offer tions. However, if current efforts to improve util- fixed-price contracts to its customers abroad. ity management have little impact, these alternatives might be worth further consideration. The various changes are discussed briefly below and A greater role for vendors in construction management offers a number of potential advantages the implications of the changes are explored in chapter 9. Some are only incremental adjust- in addition to those just described. The NSSS ments to the current structure of the nuclear in- companies have a long history of experience in dustry, while other are major reorganizations re- nuclear energy, highly trained personnel, and the quiring legislation. However, they share the com- financial incentive to build the plant well. The mon goal of improving overall management of major potential disadvantage of this approach is both construction and operation of nuclear pow- that vendors currently have little experience in erplants. construction management. If the vendors do not build up their construction capabilities, this approach may not be an improvement over using A Larger Role for Vendors in a qualified AE and constructor. Other problems Construction could arise after construction, when utilities with Many of the current problems in plant con- little knowledge of their plants must assume re- struction can be traced to the overlapping and sponsibility for maintenance and operation. conflicting authority of the utility, the nuclear steam supply system (NSSS) vendor, the AE, and Another approach to integrating responsibilities the constructor. To overcome this problem, one for construction management is used in Belgium contractor (probably the NSSS vendor) could as- for all large construction projects, including sume greater responsibility for overall design, nuclear powerplants. There, the construction and, in some cases, construction management. company assumes financial liability for the This change already is occurring to some extent. nuclear plant for a decade after it is completed. For example, Wedco, a subsidiary of Westing- The construction company is able to assume this house, has acted as the constructor for nuclear risk because it can purchase insurance after an plants in New York State, and Westinghouse itself independent assurance company has certified the is constructing a plant in the Philippines. quality of its work. 136 . Nuclear Power in an Age of Uncertainty

Service Companies cessful Florida Power & Light Co. This organization has been hired by Public Service Co. of New Currently, nuclear consultants and service com- Hampshire to speed up the construction of the panies provide a broad range of services to util- troubled Seabrook projects (27). ities, including: interactions with NRC; systems design and engineering intergration; operational While all of the entities just described are and maintenance tasks; fuel procurement; and owned by utilities, it is possible to envision others quality assurance. These firms can help strength- that would be independent. A number of busi- en the capabilities of the weaker utilities in both nesses might be interested in offering their serv- construction and operation of nuclear plants. For ices to utilities as nuclear operating companies. example, many utilities are now calling on con- Duke Power Co., a successful nuclear utility, has sulting firms to conduct independent audits of expressed interest in operating plants for other construction quality and make recommendations utilities. Some present nuclear service ‘companies for improvements. Teledyne, Inc., has audited the also might be interested, if it were clear that they two Midland units in Michigan and the two Dia- were being given management responsibility. The blo Canyon plants in California, C. F. Braun eval- fundamental shift in the present relationship be- uated the LaSalle generating station in Illinois, and tween utilities and service companies would have the Quadrex Corp. was called in to examine the to be clarified for both parties. Finally, high-tech- two South Texas plants (18). nology companies, especially those already involved in the nuclear business, might want to pro- While services such as these can be useful, they vide operating services. currently are provided only at certain times for one or more specific tasks. To resolve safety and Service companies are commonly used at Gov- management problems among the weaker utili- ernment-owned facilities. The successful use of ties, it may become desirable for service compa- contractors at armament plants, whose opera- nies to play a much larger role. This might also tions involve careful attention to safety and quali- be attractive to a disaffected public living near ty control, suggests that the complexities of nu- a troubled nuclear powerplant. These roles could clear powerplants could be handled by an inde- range from handling all quality assurance or all pendent service company. It has been estimated engineering work to actual management of con- that the Departments of Defense and Energy and struction or plant operations. the National Aeronautics and Space Administra- tion have contracts for Government-owned, con- Currently, service companies belonging to util- tractor operated facilities amounting to $5 billion ity holding companies such as Southern Co., Mid- to $10 billion per year (29). An analysis of these dle South Utilities, Inc., American Electric Power facilities indicates that operations are most suc- Co., Inc., and General Public Utilities Corp. act cessful when either the owner or the contractor somewhat like the service organizations dis- has the dominant managerial and technical role. cussed above. For example, a centralized nuclear In addition, financial liability has not been a prob- engineering group provides services to all of Mid- lem in these contracts: all liability rests with the dle South’s member utilities. In the 1950’s, a con- facility owner, and the threat of replacement pro- sortium of New England utilities formed Yankee vides the incentive for quality operations by the Atomic Electric Co., which built and operates service company. Such arrangements also might Yankee Rowe in Massachusetts. Three other cor- work well in service contracts between a utility porations, owned by many of the same utilities, and a nuclear powerplant operating company. were subsequently formed to build and operate Vermont Yankee, Maine Yankee, and Connecti- The nuclear service company alternative pro- cut Yankee. The service division of Yankee vides a way to pool nuclear expertise and make Atomic provides a broad range of services to all it available to many utilities at once. During con- of these plants (except Connecticut Yankee) and struction, the service company could play the others in New England. A more recent entrant vital role of system integrator. In addition, im- is Fuel Supply Service, a subsidiary of the suc- plementation of this alternative would be great- Ch. 5—Management of the Nuclear Enterprise ● 137

Iy simplified by the fact that it does not require panels could be similar in makeup and activities a major change in the other institutions, such as to those used in accreditation of colleges and the utilities, NRC, or the vendors. universities. Like accreditation panels, the reviewers could issue warnings and allow the utilities However, the proposal also has some disad- time to improve their management prior to de- vantages. First, it seems unlikely that utilities nying certification. Because of the unique diffi- would be willing to give up responsibility for safe- culties of nuclear plant construction, the require- ty and quality while retaining financial liability. ments for certification of utilities proposing to Secondly, unless the roles of the actors were build new plants could be made particularly strin- made very clear, the arrangement could simply gent. Based on the review panel’s findings, NRC add to the confusion that already exists in nucle- could either grant or deny the construction cer- ar powerplant construction and cause continu- tificate. ing disagreements. In addition, depending on where the owning utility and service company were headquartered, the arrangement could Once a plant is completed, another review by cause problems in dealing with State public Serv- the panel could determine the company’s abili- ice Commissions. Finally, without some mecha- ty to manage it. Depending on the results of the nism that required the weaker utilities to hire serv- review, the company might be required to hire ice companies, the existence of these entities an outside service company to take over or sup- might have little impact on the overall quality of plement operations. Thereafter, the utility and/or nuclear power management. the service company could be reviewed period- ically to make sure that changes in personnel had Certification of Utilities not diminished their management capabilities. Utilities presently operating nuclear plants also An independent review and certification of util- would be subject to the certification review. One ities as capable of constructing or operating nu- model of such a review- and certification-process Iear powerplants could provide the “stick” is the accreditation procedure for utility training needed to make the service company alternative programs currently being developed and imple- work. NRC currently has the authority to revoke mented by IN PO, the operating license of any utility the agency feels is not capable of safely operating nuclear The primary advantage of a certification proc- plants. However, since the utility industry rec- ess is that it could force the weaker utilities to ognizes that the agency is very reluctant to take improve their nuclear management capabilities, such drastic action, this authority may not be suf- obtain independent and external expertise, or re- ficiently convincing to assure high-quality opera- frain from entering the nuclear power business. tion of all nuclear plants. * Certification might pro- Such a step would be very convincing to the vide a more politically feasible alternative. It public and skeptics of nuclear power. The pri- might be more acceptable to the utilities because mary barrier to implementation is that the utility the certification review could come from an inde- industry would be reluctant to accept it. Nuclear pendent panel of experts, rather than from NRC. utilities already feel overburdened by NRC and If certain management characteristics were found INPO inspections, and the certification panel’s to affect safety negatively, utilities with those review could add yet another layer of “regula- characteristics could be decertified until those tion. ” Another disadvantage of certification is that characteristics were changed. its success depends on the existence of an entity Certification might involve periodic review of (e.g., a service company) which has the exper- utility management by an independent panel, in- tise the utility lacks. Unless such entities are avail- cluding representatives of NRC and INFO. The able and have the appropriate management characteristics, the certification procedure will not im- *The recent refusal by the Atomic Safety Licensing Board to grant an operating license to Commonwealth Edison’s Byron plant may prove the construction and operation of nuclear indicate a change in NRC’s reluctance on this issue. powerplants. 138 ● Nuclear Power in an Age of Uncertainty

Privately Owned Regional Nuclear class of electricity (in that case, power from small Power Company producers). The law has been upheld as constitu- tional over objections from State government. Since the 1920’s, electric utilities has become increasingly coordinated through horizontal in- While the initial attraction of the RNPC model tegration and power pooling. This trend has cap- may be these financial benefits, such companies tured economies-of-scale, fostered the sharing of also could be expected to improve nuclear power expertise, and eased the task of planning (31). The management by their larger staffs, allowing a regional nuclear power company (RN PC) dis- greater concentration of expertise. The proposed cussed here is one approach to increased integra- change would leave nonnuclear utility operations tion that does not involve restructuring the whole untouched, and would avoid the complications industry. It is seen as a logical extension of the of mixed financial liability and authority in the current trend toward multiple utility ownership service company scheme. In addition, the greater and single utility management of many existing expertise of the larger company could make it nuclear plants. less reliant on vendors and AEs during construction. The RNPC would be created expressly to finance construction and/or operate nuclear pow- The size of the RNPC could prove as much a erplants. It could be owned by a consortium of disadvantage as an advantage. A bureaucratic utilities, vendors, and AEs, and might place an operation could decrease the sense of individual order for several plants at once, based on a stand- responsibility for the reactors, which in turn could ardized, preapproved design. All RNPC proposals lead to a decrease in safety and reliability. Addi- currently under discussion call for a confined sit- tionally, while shared utility ownership of the ing policy to take advantage of the benefits of RNPC could help share the financial risks and clustered reactors. While some analysts feel the burdens, it might be difficult to obtain financing RNPC should be applied only to new construc- for a company whose only assets were nuclear tion, others think that existing plants could be powerplants. in the past 2 years, the utilities own- transferred to RNPC authority. Federal legislation ing the Yankee nuclear corporations in New Eng- probably would be required to transfer owner- land have had to back financial offerings with ship of existing plants because of the tax and their full utility assets. financial complications (10)0 One possible advantage of an RNPC from a fi- Government-Owned Regional nancial perspective is that its electricity output Nuclear Power Authority might not be subject to some of the difficulties This option is basically the same as that just posed by State rate regulation discussed in described, except that the Government would chapter 3. Presumably, the RNPC would sell either own the entity or provide financial assist- power to the utility grid at wholesale rates, and the utilities in turn would distribute the power ance to it. The Federal Government has previously assumed this role in the creation of the Ten- to their customers. Interstate wholesale rates are regulated by the Federal Energy Regulatory Com- nessee Valley Authority and the Bonneville Power Administration (B PA) to tap hydropower. Ontario mission (FERC). With appropriate legislation, the Hydro in Canada and TVA, which have succesful- power generated by an RNPC could be deregu- lated totally or granted special treatment in rate- Iy built and operated nuclear plants, are the making. Congress could exempt the RNPC from closest models to such an authority. However, FERC price regulation, and the electricity pur- both of these entities own nonnuclear as well as nuclear power. The RNPA envisioned here would chased by local utilities could be exempted from State rate regulation when sold to customers. The be involved only in nuclear projects, Public Utility Regulatory Policies Act (PURPA) Several factors would justify the creation of one provides the precedent of a special pricing mech- or more Government-owned RNPAs. First, it may anism, legislated by Congress, for a particular be the only way to maintain nuclear power as Ch. 5—Management of the Nuclear Enterprise ● 139 a national energy option. Given utilities’ current ernment intervention in the private sector. How- reluctance to order new nuclear plants, the Fed- ever, this problem could be alleviated if the RNPA eral Government might use the RNPA as a vehi- were primarily a financing entity, assisting private cle to demonstrate that newly designed, stand- utilities with construction of new nuclear plants. ardized plants could be built and licensed eco- Another model would be RNPA ownership of the nomically. Second, because of nuclear power’s plants, with operations handled by private serv- unique characteristics, the Federal Government ice companies. A Government-owned RNPA has always had a major role in the development could also face the problem of public opposition and regulation of this technology, and Gov- encountered by the RN PC. In addition, there is ernment ownership might be a logical extension no guarantee that management by the Federal of that role. Finally, the advantages of large-scale Government would be superior to private-utility operations cited for the privately owned regional management. Finally, if existing plants were to utilities would apply to RNPAs as well. be included in the RNPA scheme, the transfer of these plants to Government ownership could be The primary barrier to creation of a Govern- difficult. ment-owned RNPA is that it involves greater Gov-

CHAPTER 5

1. Abel, P., “Meeting With Washington Public Power and Operation of Nuclear Power Plants, San Fran- Supply System Personnel, Mar. 15, 1983,” OTA cisco, Jan. 17, 1983. memo, Mar. 24, 1983. 10. Golub, B., and Hyman, L., “The Financial Dif- 2. Baker, J., and Olsen, K., “Occupational Employ- ficulties and Consequences of Deregulation ment in Nuclear-Related Activities, 1981 ,“ Oak Through Divestiture,” Public Utilities Fortnightly, Ridge Associated Universities for the U.S. Depart- Feb. 17, 1983. ment of Energy, ORAU-197, April 1982. 11. Institute of Nuclear Power Operations, “Foremost 3. Bauman, D. S., et al., “An Analysis of Power Plant Safety and Reliability Issues,” December 1981. Construction Lead Times, Volume 1: Analysis and 12. Institute of Nuclear Power Operations, “1982 Results,” Applied Decision Analysis, Inc. for Elec- Survey of Nuclear-Related Occupational Employ- tric Power Research Institute, EPRI EA-2880, ment in U.S. Electric Utilities, ” December 1982, February 1983. INPO Report 82-031. 4. “Cincinnati G & E Bows to Stop-Work Order,” 13. Jordan, E. L., “NRC’s Design and Construction Nucleonics Week, vol. 23, No. 46, Nov. 18, 1982. Quality Assurance Initiatives,” presented at the 5. “Commissioners Okay Salem-1 Restart After Set- Atomic Industrial Forum Workshop on Achieving ting Some Prerequisites,” Nucleonics Week, vol. Excellence in the Construction and Operation of 24, No. 17, Apr. 28, 1983. Nuclear Power Plants, San Francisco, Jan. 17, 6. “Doing It Right in the Nuclear Business,” The 1983. Energy Daily, vol. 11, No. 14, Jan. 20, 1983. 14. Kelley, D., “Electric Utility Industry; Nuclear 7. Friedlander, G. D., “Coal or Nuclear? Economics Plants–Another Look,” Merrill Lynch, May 1982. Govern,” Electrical World, vol. 195, No. 10, Oc- 15. Kemeny, J. G., “Report of the President’s Com- tober 1981. mission on the Accident at Three Mile Island,” Oc- 8. Garrick, B. J., “Advances in Using Probabilistic tober 1979, Washington, D.C. Risk Assessment as a Risk Management Tool,” 16. Komanoff, C., “The Bottom Line–Nuclear Ca- presented at the Nuclear Power Reactor Safety pacity Factors: Another Bad Year,” Critical Mass, Course, Massachusetts Institute of Technology, April 1983. July 18, 1983. 17. Martel, L., Minnek, L., and Levy, S., “Summary 9 Glasscock, R., “Responding to Quality Concerns,” of Discussions With Utilities and Resulting Conclu- presented at the Atomic Industrial Forum Work- sions, ” Electric Power Research Institute, June shop on Achieving Excellence in the Construction 1982. 140 . Nuclear Power in an Age of uncertainty

18. MHB Technical Associates, “Issues Affecting the 32. Uhrig, R., Letter to the Office of Technology As- Viability and Acceptability of Nuclear Power sessment, Aug. 1, 1983. Usage in the United States,” Dec. 2, 1982. 33. United Engineers & Constructors, Inc., Case Study 19. Nero, A., A Guidebook to Nuclear Reactors, Reports for the U.S. Department of Energy, 1981. University of California Press, Berkeley, 1979. 34. United Engineers & Constructors, Inc., “Phase IV 20, “NRC Pulls the Plug on Zimmer Nuclear Power Final Report and Fourth Update of the Energy Plant,” The Energy Dai/y, vol. 10, No. 215, Nov. Economic Data Base (EEDB) Program,” prepared 16, 1982. for the U.S. Department of Energy, [JE&C-ANL- 21, “Nuclear Safety Fines,” Washington Post, No, 810930, COO-641 1-1, September 1981.

159, May 13, 1983. 35< United Engineers and Constructors, Inc., “Regu- 22, Office of Technology Assessment, U.S. Congress, latory Reform Case Studies (Piping) -Draft,” UE&C Nuclear Powerplant Standardization–Light Water 820212, February 1982. Reactors, OTA-E-1 34, June 1981. 36. U.S. Department of Energy, “How to Prolong 23, Office of Technology Assessment Workshop, Steam Generator Life in Nuclear Plants,” Update Washington, D. C., Dec. 9-10, 1982. —Nuclear Power Program Information and Data, 24. Partlow, j., U.S. Nuclear Regulatory Commission, February 1983, DOE/NE-0048 11. personal communication, Feb. 3, 1983. 37. U.S. Nuclear Regulatory Commission, “A Survey 25. Peterson, E., Commonwealth Edison, Co., per- of Senior NRC Management to Obtain Viewpoints sonal communication, August 1983. on the Safety Impact of Regulatory Activities From 26, Procter, M., “Trip Report-Visit to INPO on Jan. Representative Utilities Operating anc{ Construc- 26, 1983,” OTA memo, Feb. 15, 1983. ting Nuclear Power Plants,” NUREG-0839, August 27. “PS New Hampshire Hires Florida P&L Subsidiary 1981. to Help Bring Seabrook on Line,” E/ectric L)ti/ity, 38. Weitzberg, A., et al., “Reactor Quality—Factors Aug. 1, 1983. Affecting Managerial and Institutional Compe- 28. “Revamped South Texas Project Showing Strong tence of Utility Nuclear Operations, ” NUS Corp., Signs of Recovery,” Nucleonics Week, vol. 24, Jan. 15, 1983. No. 12, Mar. 24, 1983. 39 Wilkinson, E., “lNPO/NRC Interactions, ” pre- 29, Ring, Leon, “Use of Nuclear Service Companies,” sented at the American Nuclear Society Topical An Acceptable Future Nuclear Energy System, Meeting on Government and Self-Re,gulation of Firebaugh and Ohanian, (eds.) Oak Ridge, Term.: Nuclear Power Plants. Institute for Energy Analysis, 1980. 40< “World List of Nuclear Power Plants,” Nuc/ear 30. Stevenson, J. D., and Thomas, F. A., Se/ected News, vol. 26, No. 2, February 1983. Review of Foreign Licensing Practices for Nuclear 41 Zebroski, E., “Analysis of Operating IEvents and Power P/ants, NUREG/CR-2664. Priorities,” presented at the American Nuclear 31 Theodore Barry & Associates, A Study ofAggregate Society Topical Meeting on Government and Self- Alternatives in the U.S. Electric Utility Industry, Regulation of Nuclear Power Plants, DOE-RG-1O295-1, U.S. Department of Energy, 1982.

Chapter 6 The Regulation of Nuclear Power

Photo credit: Florida Power & Light One of aisles in the document vault at St. Lucie 2 Contents

Page Federal Regulation ...... 143 The Role of Emergency Planning...... 150 State and Local Regulation . . . . . 150 Need for Power ...... 151 Environmental Policy ...... 151 State Siting Activities...... 153 Issues Surrounding Nuclear Plant Regulation ...... 153 Backfitting ...... 154 Specific Concerns ...... 155 Proposals for Change and Evaluation ...... 157 Hearings and Other NRC Procedures ...... 160 Hearings ...... 160 Proposals for Change in the Hearing Process and Evaluation ...... 161 The Two-Step Licensing Process ...... 168 Other NRC Responsibilities. 171 Safety Goals ...... 172 NRC Safety Goal Proposal ...... 172 Licensing for Alternative Reactor Types ...... 174 Chapter 6 References ...... 175

Table

Table No. Page 22. State Siting Laws ...... 152

Figures

Figure No. Page 32. NRC Responsibilities in Nuclear Powerplant Licensing ...... 146 33. Utility Responsibilities in Nuclear Powerplant Licensing and Construction ...... 147 Chapter 6 The Regulation of Nuclear Power

Nuclear power is one of the most intensively reviews the various criticisms of that process rais- regulated industries in the United States, and the ed by the different parties in the nuclear debate; scope and practice of regulation are among the and discusses proposals for substantive and pro- most volatile issues surrounding the future of nu- cedural changes in nuclear power regulation. The clear power. Strong—and usually conflicting— chapter focuses on the health and safety and en- opinions abound among the participants in the vironmental regulation of nuclear powerplants; nuclear debate on whether the current regula- financial and rate regulation are discussed in tory system is adequate to ensure safe and reliable chapter 3. powerplants or is excessive, and whether it is en- It should be emphasized that this chapter pri- forced adequately or is interpreted too narrowly. marily reports on the existing regulatory process Every aspect of the nuclear industry—from the and on proposals for changes in that process. establishment of standards for exposure to radia- Arguments for and against the existing system and tion to the siting, design, and operation of nuclear proposed changes are presented as they appear powerplants and the transportation, use, and dis- in the literature or as OTA determined them in posal of nuclear materials–is regulated at the the course of this study. Such criticisms of the reg- Federal, State, or local level. In general, the Fed- ulatory system can reflect the biases and vested eral Government retains exclusive legislative and interests of the commentators. In light of this, it regulatory jurisdiction over the radiological health is important to examine the arguments critically and safety and national security aspects of the from a safety and efficiency perspective. Where construction and operation of nuclear reactors, OTA found sufficient documentation to support while State and local governments share the reg- a particular argument, the basis for the conclu- ulation of the siting and environmental impacts sions is identified. In instances where OTA could of nuclear powerplants and retain their traditional not make such a determination, the arguments responsibility to determine questions of need for are presented without conclusions to illustrate power, reliability, user rates, and other related the scope of the controversy and the wide diver- State concerns. gence among the parties’ perceptions of the current role of regulation and of the need for This chapter describes the existing regulatory changes in the regulatory system. process at the Federal, State, and local levels;

FEDERAL REGULATON The primary forms of regulation under the has been the target of so much criticism by the Atomic Energy Act (see box D) are: 1 ) the issu- nuclear industry, utilities, nuclear critics, and reg- ance of licenses for the construction and opera- ulators. In addition, this section discusses the way tion of reactors and 2) inspection and enforce- in which this licensing process might operate in ment to ensure that nuclear plants are built and the current climate. Although the basic regula- operated in conformance with the terms of a li- tions have not changed substantially since the cense. This section describes the licensing proc- 1970’s, the way those regulations are applied to ess that was put in place during the 1970’s when construction permits or operating licenses might the last group of plants received construction per- be very different if an application were filed mits. This is precisely the licensing process that today.

143 144 ● Nuclear Power in an Age of Uncertainty Ch. 6—The Regulation of Nuclear Power ● 145

In the 1970’s, a utility would undergo an ini- that many of the construction problems evident tial planning phase before it would apply to the in today’s plants could have been prevented if Nuclear Regulatory Commission (NRC) for a con- more complete designs had been available dur- struction permit, It would select a site in according CP review. I n recognition of this argument, ance with NRC (and State and local) policies and NRC officials have indicated that they now would guidelines; choose an architect/engineering (AE) require an essentially complete design with a CP firm; solicit bids for the nuclear steam supply sys- application, a move that has widespread support. tem (NSSS) and the balance of the plant; award contracts; and assemble data to be submitted to In the next step of the process, the NRC Of- NRC with the construction permit (CP) applica- fice of Nuclear Reactor Regulation would com- tion. During this planning phase, the utility also pare the details of the permit application with the would ensure compliance with State and local NRC’s Standard Review Plan (SRP) and usually laws and regulations, which could require a vari- would submit two rounds of questions to the ap- ety of permits for approval of the facility. plicant. These questions often would result in changes in the plant design. The staff then would The utility then would file an application for prepare a Safety Evaluation Report (SER) docu- a CP, as indicated in figures 32 and 33. The ap- menting the review and listing “open issues, ” plication would include: 1 ) a Preliminary Safety which are changes dictated by NRC but disputed Analysis Report (PSAR) that presents in general by the applicant. Concurrent with the prepara- terms the plant design and safety features and tion of the SER, the Advisory Committee on Reac- data relevant to safety considerations at the pro- tor Safeguards (ACRS) would review and composed site; 2) a comprehensive Environmental ment on the application, and the NRC staff could Report (ER) to provide a basis for the NRC evalua- issue supplements to the SER to respond to issues tion of the environmental impacts of the pro- raised by ACRS or to add any information that posed facility; and 3) information for use by the may have become available since issuance of the Attorney General and the NRC staff in determin- original SER. During the 1970’s, this review proc- ing whether the proposed license would create ess culminating in SER might have taken 1 to 2 or maintain a situation inconsistent with the an- years. The review period could potentially be titrust laws. shortened if an application were filed now with essentially complete design information or a NRC regulations require the antitrust informa- standardized design. Detailed design information tion be submitted at least 9 months but not more would be likely to meet the minimum criteria for than 36 months prior to the other portions of the acceptance of the application with little delay. CP application. A hearing might be held at the A standardized design could indirectly accelerate completion of the antitrust review, but it would the process even more because it is unlikely that not be mandatory unless requested by the Attor- many new questions would be raised by the ney General or an interested party, The NRC also ACRS or about the SRP after approval of the first must make a finding on antitrust matters in each plant using that design. case where the issue is raised before the Com- mission. During this period, the NRC staff also would be reviewing the proposed plant’s environmen- Upon receipt of a CP application, the NRC staff tal impacts and preparing a draft Environmental would review it to determine if it is complete Impact Statement (E IS) to be issued for review by enough to allow a detailed staff review, and re- the relevant Federal, State, and local agencies and quest additional information if necessary. The ap- by interested members of the public. After com- plication would be formally “docketed” when ments on the draft EIS were received and any it met the minimum acceptance criteria. questions resolved, the staff would issue a final EIS. In the past, the PSAR included very incomplete design information (only 10 to 20 percent in some Soon after a CP application was docketed, NRC cases). Most parties in the nuclear debate agree would issue a notice indicating that it would hold 146 ● Nuclear Power in an Age of Uncertainty

Figure 32.-NRC Responsibilities in Nuclear Powerplant Licensing I

Operating license Operating license issued Hearing (if required)— Review by ACRS—

NRC safety evaluation

I Radiological safety review (based on I FSAR)

I I I I I QA construction and testing inspections

Review by NRC safety I evaluation 1 Final environmental I

acceptance review I I Draft environmental (based on CP I Antitrust decision I Environmental acceptance review I I I QA program review Antitrust hearing I I I

1 Antitrust review I I I Antitrust Radiological QA program review safety review review Ch. 6—The Regulation of Nuclear Power ● 147

Figure 33.—Utility Responsibilities in Nuclear Powerplant Licensing and Construction I I

I — Begin commercial operation I — Initiate power operation test program I I — Complete construction, load fuel I

I I Update environment report and file FSA I I

I Onsite QA program —Start plant construction

I —Start site preparation I State water quality I certificate ion —Make construction arrangements File State PUC File CP application application I Prepare PSAR and I provide site inform, —Design plant I to NRC Prepare environmental Issue reactor specifications I Develop QA program provide antitrust information to NRC I State review NRC review Construction-related activities 148 ● Nuclear Power in an Age of Uncertainty a hearing on safety and environmental issues standpoint, and after Commission approval. A li- raised by the application. Interested parties could censing board may begin hearings on an LWA provide written or oral statements to the three- within a maximum of 30 days after issuance of member Atomic Safety and Licensing Board the final EIS. (ASLB) as limited participants in the hearings, or When construction of a plant had progressed they could petition for leave to intervene as full to the point where final design information* and participants, including the right to cross-examine plans for operation were ready, an application all direct testimony in the proceeding and to sub- for an operating license (OL) would be prepared. mit proposed findings of fact and conclusions of The OL process has been very similar to that for law to the hearing board. a CP. The applicant would submit a Final Safety NRC regulations provide an opportunity at an Analysis Report (FSAR), which sets forth the perti- early stage in the review process for potential in- nent details on the final design of the facility, in- tervenors to be invited to meet informally with cluding a description of the containment, the the NRC staff to discuss their concerns about the nuclear core, and the waste-handling system. The proposed facility. This provision has not been FSAR also would supply information concerning commonly used; as a result, the safety concerns plant operation, including managerial and admin- of the critics have not been considered serious- istrative controls to be used to ensure safe opera- ly and formally until the hearings. The problem tion; plans for preoperational testing and initial with this timing is that it places the critics in the operations; plans for normal operations, includ- position of attempting to change or modify a deci- ing maintenance, surveillance, and periodic test- sion that already has been made rather than in- ing of structures, systems, and components; and fluencing its formulation. plans for coping with emergencies. The applicant also would provide an updated ER. Amendments The environmental hearings could be con- to the application and reports could be submitted ducted separately to facilitate a decision on a from time to time. The staff would prepare an- Limited Work Authorization (LWA) or could be other SER and EIS and, as at the CP stage, ACRS combined with the safety hearing. The SER and would make an independent evaluation and pre- any supplements to it plus the final EIS would be sent its advice to NRC by letter. The ASLB would the major pieces of evidence offered by the NRC also review the OL application and issue a deci- staff at the hearing. The ASLB would consider all sion. Until recently, the ASLB has granted all re- the evidence presented by the applicant, the staff, quests for OLS. However, in January 1984 the and interveners, together with proposed findings ASLB refused to grant an OL to Commonwealth of fact and conclusions of law filed by the parties, Edison Co. for the two-unit Byron station. As in and issue an initial decision on the CP. ASLB’s the procedure for a CP, this decision will be re- initial decision would be reviewed by the Atomic viewed by the ALAB and the Commission. Safety and Licensing Appeal Board (ALAB) on exceptions filed by any party to the proceeding or, A public hearing is not mandatory prior to is- if no exceptions were filed, on ALAB’s own in- suance of an OL. However, soon after accept- itiative (“sua sponte”). Since Three Mile Island, ance of the OL application, NRC would publish all ASLB decisions must be approved by NRC notice that it was considering issuing a license, before they take effect. NRC also considers peti- and any person whose interest would be affected tions for review of appeals from A LAB decisions. by the proceeding could petition NRC to hold a hearing. The hearing would apply the same ad- NRC regulations provide that the Director of judicatory procedures (e.g., admission of parties the NRC Office of Nuclear Reactor Regulation and evidence, cross-examination) and decision may issue an LWA after ASLB has made all of the process that pertain to a CP. environmental findings required under NRC reg- uIations for the issuance of a CP and has reason- *The final design illustrates how the plant has been built and thus able assurance that the proposed site is a suitable reflects all amendments and variances issued and backfits ordered location from a radiological health and safety by NRC since the CP.

Ch. 6—The Regulation of Nuclear Power ● 149

Photo credit: Nuclear Regulatory Commission

The members of the Nuclear Regulatory Commission are meeting with the licensing staff of the NRC to review an upcoming operating license. The Commission’s meetings are open to the public

A stated goal of NRC (under normal circum- ulations designed to protect the public health and stances and barring any important new safety safety and the environment. issues) is to conclude ACRS and Office of Nuclear Other Federal agencies with statutory or regu- Reactor Regulation reviews and the hearing proc- latory authority over some aspects of the con- ess before the utility completes construction of struction and operation of nuclear powerplants the plant. Current NRC regulations authorize the include: Environmental Protection Agency, U.S. staff to issue an OL restricted to 5-percent pow- Army Corps of Engineers, National Oceanic and er operation; full power operation must be ap- Atmospheric Administration, Department of En- proved by the Commissioners themselves. Upon ergy, Department of the Interior, U.S. Geological receipt of the low-power OL, the utility could Survey, Department of Agriculture, Department begin fuel loading and initial startup. The plant of Housing and Urban Development, Advisory then would have to undergo extensive testing Council on Historic Preservation, Department of before it could begin commercial operation. Transportation, Federal Aviation Agency, Depart- Through its inspection and enforcement program, ment of Defense, Council on Environmental NRC maintains surveillance over construction Quality, River Basin Commissions, and Great Lakes and operation of a plant throughout its service Basin Commission. These agencies review, com- life. As discussed in chapter 5, this surveillance ment on, and administer specific issues under is intended to assure compliance with NRC reg- their jurisdiction.

25-450 0 - 84 - 11 : QL 3 150 ● Nuclear Power in an Age of Uncertainty

The Role of Emergency Planning and supplies are kept up to date. Finally, the utility must develop preliminary criteria for deter- NRC requires license applicants and licensees mining when, following an accident, reentry of to specify their plans for coping with emergen- the facility would be appropriate and when cies. NRC regulations established in 1980 specify operation could be resumed. minimum requirements for emergency plans for use in attaining “an acceptable state of emergen- The role of emergency planning has become cy preparedness”, including information about increasingly controversial since the accident at the emergency response roles of supporting orga- Three Mile Island. Local governments must par- nizations and offsite agencies (16). For plants just ticipate in the preparation of the emergency plan starting construction, these plans have to be and reach agreements with the utility on public stated in general terms in the PSAR and submitted notification, evacuation, and other procedures, in final form as part of the FSAR. Detailed pro- and they may intervene in the hearings on the cedures for emergency plans would have to be plan. This is the principal leverage a local govern- submitted to NRC no less than 180 days prior to ment has over the operation of a nuclear plant, the scheduled issuance of an OL. Licensees with and it can hold up the issuance of an OL. For ex- operating plants in 1980 were also required to ample, at the Shoreham nuclear powerplant, sig- submit detailed emergency plans to comply with nificant differences in scope between the emer- post-TMl regulations. gency plan proposed by the utility and that developed by the county are the primary issue that NRC regulations specify a broad range of in- must be resolved before the utility can obtain an formation that must be included in emergency OL. There is a possibility that those differences plans. Utilities must develop an organizational might not be resolved. The adequacy of the util- structure for coping with radiological emergen- ity’s emergency planning also has become an cies and define the authority, responsibilities, and issue at the Indian Point Station due to its prox- duties of individuals within that structure as well imity to densely populated New York City. as the means of notifying them of the emergency. Second, the utility must specify the criteria Such situations are of great concern to nuclear on which they determine the magnitude of an utilities, their investors, and the surrounding com- emergency and the need to notify or activate pro- munities. If emergency plans are not developed gressively larger segments of the emergency or- in a timely and satisfactory fashion, the plant ganization (including NRC, other Federal agen- owners will not be granted a license to operate cies, and State and local governments). Third, the their plant. Moreover, emergency planning prob- utility has to reach agreements with State and lems are difficult to anticipate, and their resolu- local agencies and officials on procedures for tion is not necessarily assured by prudent man- notifying the public of emergencies for public agement. Thus, they tend to increase the uncer- evacuation or other protective measures and for tainties associated with nuclear plant schedules. annual dissemination of basic emergency plan- This source of uncertainty might be eliminated ning information to the public. Fourth, programs if final approval of emergency plans were re- must be established to train employees and other quired much earlier in the licensing process or persons to cope with emergencies, to hold peri- even as a condition of State issuance of a Certifi- odic drills, and to ensure that the emergency plan cate of Public Convenience for a nuclear plant. and its implementing procedures, equipment,

STATE AND LOCAL REGULATION

A wide range of State and local legislation, reg- During the last decade, more and more States ulations, and programs affect the licensing, con- moved to a more thorough consideration of need struction, and operation of nuclear powerplants. for power and choice of technologies, environ- Ch. 6—The Regulation of Nuclear Power ● 151 mental policy, and energy facility siting. Table 22 Ient guarantee that it will be allowed to earn a identifies the various State authorities in these return on its investment. Furthermore, it is un- areas. In most cases, NRC requires State approv- likely that a utility would apply to NRC for a CP als to be obtained before the Commission will without already having obtained a CPCN or at take any action on a CP or OL application. least being confident of receiving it. Several States (e.g., California, Oregon, Ver- mont, Wisconsin) also have enacted special re- Environmental Policy strictions on the construction of nuclear power- Several Federal statutes, including the National plants on the basis of economic, environmental, Environmental Policy Act (N EPA), the Clean Air or waste-disposal considerations. The U.S. Su- Act (CAA), the Clean Water Act (CWA), and the preme Court recently upheld State authority to Housing and Urban Development (HUD) 701 restrict nuclear power development when it ruled Comprehensive Planning Assistance program, in favor of California’s siting law, which bans new emphasize the State role in regard to environ- nuclear powerplants pending a method to dis- mental issues. Moreover, many States have en- pose of nuclear waste. The Court held that the acted their own environmental legislation. Thus, Atomic Energy Act does not expressly require the the same environmental aspects of a proposed States to construct or authorize nuclear power- nuclear powerplant often are reviewed by both plants or prohibit the States from deciding, as an the State and the NRC, and in some cases, joint absolute or conditional matter, not to permit the NRC-State hearings may be held on matters of construction of any further reactors. concurrent jurisdiction. The State regulation of environmental and siting Traditionally the States have been responsible issues discussed below adds to the complexity for land use, and many States have comprehen- and uncertainty in planning and licensing nuclear sive land-use planning programs. Energy facility powerplants. In a State with multiple layers of siting also can be affected by States, local govern- review within numerous agencies, dozens or ments, or regional organizations through compre- even hundreds of State approval steps may be hensive planning activities under federally ap- involved. However, the State regulatory process proved coastal-zone management programs and generally has been far less difficult to manage under the HUD program. than the Federal regulatory aspects related specifically to nuclear health and safety. State water management agencies must approve a proposed nuclear powerplant, examin- Need for Power ing issues related to both the quality and quanti- ty of water supply and to effluent discharge limita- Primary responsibility for regulating electric util- tions. The States have programs to review water ities has been vested for many decades in State withdrawals from streams and structures placed public utility commissions (PUCs). PUCs set rate in water, and they issue Water Quality Certificates schedules designed to meet the cost of service under CWA, which include any effluent limita- and to earn the utility stockholders an appropriate tions, monitoring, or other requirements neces- return on investment, as discussed in chapter 3. sary to assure that the plant will comply with ap- Many PUCs also approve financing for new facili- plicable Federal and State water-quality stand- ties deemed necessary to supply service and issue ards. These conditions become part of the NRC Certificates of Public Convenience and Necessi- permit or license. In addition, if a nuclear facility (CPCN), which certify that when the facility ty will discharge into navigable waters, it must goes into service the capitalized cost will be obtain a National Pollutant Discharge Elimination added to the rate base. System (NPDES) permit. CWA establishes special procedures for NPDES permits dealing with ther- Although the procedures for determining need mal discharges. for power and issuing a CPCN vary from State to State, no utility will proceed beyond engineer- Nuclear plants can have air-quality impacts ing to construction without a CPCN or an equiva- from cooling tower plumes, but neither the States 152 ● Nuclear Power in an Age of Uncertainty

x x w

xx xx xxx

x xxx ...... Ch. 6—The Regulation of Nuclear Power ● 153

nor the Federal Government have standards gov- ety of State agencies, each concerned with a sep- erning such emissions. Rather, the primary effects arate aspect of the construction or operation of of CAA and State air-quality programs under that a plant; State licensing through a “one-stop” act are through restrictions on the siting of and agency charged with determining the suitability emissions from fossil-fueled plants, which may in- of all aspects of a particular site on behalf of all crease the attractiveness of the nuclear option for State regulatory bodies; or State ownership of the electric utilities. site, with a single agency empowered to administer the terms of a lease with the utility or con- State Siting Activities sortium that owns the plant. Twenty-five States currently have siting laws. These include “multistop” regulation by a vari-

ISSUES SURROUNDING NUCLEAR PLANT REGULATION For the last decade, nuclear plant regulation such participation prolongs hearings unnecessari- has been slow, unpredictable, expensive, and ly without adding to safety. They believe that frustrating for many involved in licensing. More- these factors have contributed to higher costs and over, it has failed to prevent accidents such as longer construction times and may have reduced those at Three Mile island and Browns Ferry as safety by requiring the applicant and the regula- well as construction problems like those experi- tors to focus more of their efforts on the proce- enced at Diablo Canyon and Zimmer. Even in dural aspects of licensing to the detriment of the case of the Byron plants where the OL was substance. denied by the ASLB, the problems were not acted Nuclear critics, on the other hand, argue that on until the two plants were nearly complete. The the lack of predictability and construction diffi- frustrations, costs, and uncertainties have resulted culties were due to the immaturity of the tech- in extensive criticisms of the regulatory process nology and a “design-as-you-go” approach. The and a variety of proposals for changes in that critics feel that many of their safety concerns process. The focal points for such criticisms are would not have arisen had it not been for the backfitting, * hearings and other NRC procedures, rapid escalation in plant size and number of the current two-step licensing process, NRC re- orders that occurred in the 1960’s and 1970’s, sponsibilities not directly related to safety, the use utility and constructor inattention to quality of rulemaking, and safety goals. assurance, and inconsistent interpretation and en- The principal concerns about nuclear plant reg- forcement of regulations within NRC. While some ulation expressed by utilities and the industry are critics say that nuclear plants will never be safe that neither the criteria nor the schedules for sit- enough, others believe that the current regula- ing, designing, building, and operating nuclear tory process could ensure safety if it were inter- plants are predictable under the current licens- preted consistently and enforced adequately. ing scheme. The industry and some regulators Most critics agree that limiting the opportunities also complain about the extensive opportunities for interested members of the public to partici- for public participation in licensing, arguing that pate in licensing will detract from safety. This section will describe in detail the various *Although “backfitting” technically refers only to design or reg- parties’ views (as determined by OTA) on NRC ulatory changes ordered by NRC during plant construction, and regulation—what they believe works, what they “ratcheting” to changes imposed after a plant goes on line, “backfit- believe doesn’t, and why they hold their views– ting” usually is used in the literature to refer to both types of changes. Modifications requested by the permit or license holder and how they think the regulatory process could are termed “amendments” or “variances. ” be improved. These views were solicited by OTA 154 ● Nuclear Power in an Age of Uncertainty at workshops and panels involving a broad spec- health and safety goal–or at least not detract trum of interested parties, including nuclear crit- from it–then secondary policy goals might be: ics and representatives from utilities, vendors, and ● to provide a more predictable and efficient AEs. These meetings were supplemented by a sur- licensing process in order to assure license vey conducted for OTA in which a small sample applicants that a plant, once approved, can of qualified individuals responded to an exten- be built and operated as planned; sive questionnaire(l 5). Suggestions for revisions ● to increase the effectiveness of public par- to current NRC regulations and procedures have ticipation in licensing; and been made by a number of interested parties. ● to improve the quality of NRC decisions in They will be presented in the following text as order to increase public confidence in plant originally proposed and then evaluated on the safety. basis of the information available to OTA. The discussion of regulatory revisions will focus on Achieving these secondary policy goals prob- two legislative packages currently before Con- ably is a necessary condition in ensuring (whether gress: The Nuclear Powerplant Licensing Reform for national security, economic, or other reasons) Act of 1983, submitted by NRC (23), and the Nu- that nuclear power remains a viable option for clear Licensing and Regulatory Reform Act of utilities in choosing their mix of generating tech- 1983, proffered by the U.S. Department of Energy nologies. However, it should be emphasized that (DOE) (21). these goals cannot be accomplished through licensing changes alone. Rather, they also will re- The evaluation of proposals for changes in NRC quire a commitment to excellence by all parties regulation must depend first on an assessment of in the management of plant licensing, construc- the goals to be served by regulation and by the tion, and operation, as well as a commitment to individual changes. The primary goal of NRC reg- resolving outstanding safety and reliability issues. ulation as defined in the Atomic Energy Act is to ensure that the utilization or production of special Another consideration in evaluating proposals nuclear material will be in accord with the com- for change in the licensing process is whether mon defense and security and will provide ade- amendment of the Atomic Energy Act is necessary quate protection to the health and safety of the to accomplish a particular change, or whether public. Therefore, in analyzing proposals for it can be accomplished through rulemaking or changes in licensing, the first consideration must even simply more effective implementation of the be whether they are necessary to further the ful- existing regulations. fillment of this goal. If changes would further this

BACKFITTING The utilities’ and the nuclear industry’s com- is discretionary) in ordering a backfit is whether plaints about lack of predictability in reactor reg- it will “provide substantial additional protection ulation focus principally on the potential for which is required for the public health and safe- changes in regulatory and design requirements ty or the common defense and security.” during plant construction and operation (“backfitting”). NRC never has invoked the backfit definition formally to amend a permit, license, rule, regu- The present NRC regulations define backfitting lation, or order. Rather, it has changed its require- as” . . . the addition, elimination or modification ments through a variety of less formal procedures, of structures, systems or components of the facili- such as bulletins, circulars, regulatory guides, and ty after the construction permit has been issued” informal meetings. While NRC has justified the (16). Under the present regulations, the stand- changes on a safety basis, the decisionmaking ard NRC may use (the language in the regulations process has not been as transparent nor as pre- Ch. 6—The Regulation of Nuclear Power ● 155 dictable as desired by the industry or its critics. To review these claims about backfitting, OTA The industry would like to have the backfit rule undertook a survey of people of all viewpoints rewritten and the procedure for invoking it re- connected with nuclear power, including indus- vised so that it would provide greater certainty try representatives, regulators, and critics. The and stability in terms of costs and schedules and results of this survey form the basis for the anal- greater flexibility in implementation. The critics ysis presented here. would like to see NRC follow an established and There are certain ways in which backfits have documented procedure in ordering backfits to fa- the potential to adversely affect the safety of cilitate evaluation of safety considerations. nuclear plants. First, additional equipment can impair normal operations or safety functions; Specific Concerns backfits related to seismic protection often are cited as examples of these problems. As discussed Until recently, NRC’s Office of Nuclear Reac- in chapter 4, requirements for additional pipe tor Regulation has been responsible for review- hangers and restraints increasingly have con- ing and coordinating backfit proposals. The nu- strained the layout of the piping systems in nuclear industry has perceived that review to be unclear powerplants. This could contribute to ther- systematic, haphazard, and without reference to mal stresses in normal operation and make the any regulatory standard. The industry does not system more prone to failure in accident situa- believe that all of the changes made to plants over tions. Another adverse consequence associated the years have contributed significantly to safe- with seismic backfits is that they can lead to over- ty. In fact, it considers some of these changes to crowding, making some equipment virtually inac- have made plant design and operations more cessible for inspection or maintenance. complicated, less predictable, and possibly less Backfits also can affect safety by disrupting nor- safe (8). Moreover, these changes have absorbed mal plant operations while new equipment is be- a large share of the utilities’ financial and technical resources. For example, after a decade of ing installed. While this potential problem is less operation, there were still hundreds of people a result of NRC’s management of backfits than working to make changes at the Browns Ferry of utility planning and expertise, it is important nuclear plants (5). At another utility, the backfits to recognize that there are safety implications in 1980 alone required $26 million and 10 to 12 associated with installation. Such an incident oc- staff-years of engineering. In addition, the long- curred at the Crystal River plant in 1980 when term expenditures associated with Three Mile the utility attempted to install a subcooling mon- island backfits are estimated to cost $74 million itor while the unit was still operating. This action at the same plant (26). Thus, there are powerful triggered a series of unplanned events, eventually followed by safe shutdown of the plant. incentives for the nuclear industry to try to have the backfit rule and its implementation changed. It should be noted that while examples can be Nuclear critics counter that the rule would be found in which backfits have adversely affected adequate if it were implemented consistently and safety, this does not imply that the overall impact understandably. They contend that backfits serve has been negative. In fact, one recent study in- an important safety function, since many prob- dicates that modifications made to plants after Three Mile Island may have reduced the proba- lems arise only after construction or operation has been initiated. The critics, however, agree bility of a large-scale accident by as much as a factor of six at some plants (1 3). However, the with the industry that it would be more appropriate to allocate resources to the design phase overall gain or loss in safety due to backfits has rather than using them to satisfy safety concerns not been analyzed in any comprehensive fashion. with backfits. Unfortunately, this is not an option OTA concludes that, while most backfits rep- for existing plants, but can be applied only to the resent safety improvements, they can have a next generation of nuclear reactors. negative impact when deployed in a manner 156 ● Nuclear Power in an Age of Uncertainty that does not allow for sufficient analysis of the It is OTA’s conclusion that the requirements consequences of installing or modifying equip- associated with the design, construction, and ment and its interaction with other systems. A operation of nuclear powerplants are prescrip- more rational and less hurried approach could tive and, in some cases, internally inconsistent improve this situation for current plants. if the or in conflict with other good practices. However, next generation of reactors is an evolutionary de- while the inconsistencies and contradictions are velopment of today’s light-water reactors (LWRs), problematic, the prescriptive nature of the rules new plants should be even less troubled by back- should not pose insurmountable difficulties for fits. New LWR designs will incorporate the les- plant owners and designers. Some utilities have sons learned from Three Mile Island and the ac- been able to accommodate to the same prescrip- cumulated experience of current reactors, and tive requirements that govern all nuclear con- they will address unresolved safety issues with struction and still complete their plants efficiently state-of-the-art technology. However, if an alter- and with few regulatory difficulties. Moreover, native reactor design is selected for commercial NRC is not wholly responsible for prescriptive deployment, it maybe impossible to avoid exten- regulation. The nuclear industry has developed sive backfitting until the technology is fully a large and growing set of voluntary standards mat ure. to provide guidance in interpreting NRC criteria. These standards were expanded greatly in the NRC backfit requirements also have been criti- mid-1970’s to match the growth in NRC require- cized by the nuclear industry as being overly pre- ments and often were written with little consid- scriptive to the point of being incompatible with eration of their impact. In addition, many of the practical design, construction, or operating tech- early standards were written too rapidly to reflect niques. Rather than establishing general guide- field experience and a convergence of accepted lines or safety criteria and allowing individual util- practices. NRC magnified these problems by in- ities some flexibility in applying them, NRC gen- voking the standards precisely as written rather erally issues detailed and specific requirements. than allowing them to evolve gradually (19). Nuclear powerplant designers must conform to the regulations and appendices in 10 CFR Part Another concern about backfitting is that there 50 as well as 10 other major parts to title 10, over are no clear and consistent priorities. Permit and 150 regulatory guides, three volumes of branch license holders argue that they have not always technical positions, numerous inspection and en- been given consistent and stable criteria by which forcement circulars and bulletins, proposed rules, to construct and operate a plant and, as a result, and over 5,000 other voluntary codes and stand- some less important backfits have been imposed ards that may be invoked at any time by regula- before more critical ones. The prime example of tory interpretation. During construction, these this cited by utilities and the industry is the Three standards and codes often are interpreted in the Mile Island action plan, in which the NRC gave strictest sense possible, with no allowances for the utilities no guidance on the relative priorities engineering judgment. For example, the fillet among approximately 180 requirements of vary- weld, which is commonly used in field construc- ing importance. The action plan was developed tion, varies in width along the length of the weld. with little comprehensive analysis. As discussed Plant designers recognize that some variation will above, if the next generation of plants incorpo- occur and set the design requirements according rates a clean-sheet design based on past experi- to an average width. An inspector, by strict in- ence with LWRs, backfits should not be as serious terpretation of an industry code, may not look a problem as they have in the past. While a lack at the average width, but reject an otherwise ac- of priorities has been troublesome for plants cur- ceptable weld if it is slightly less in width than rently under construction or in operation, it is called for by the designer at any point along the unlikely that future LWRs will experience the length of the weld. Constructors compensate for same degree of difficulty. such anomalies by overwelding, which entails A final concern about backfitting is its poten- considerable time and expense (19). tial contribution to increases in construction lead- Ch. 6—The Regulation of Nuclear Power ● 157 times and plant costs, These issues are dis- through strong management. All plants had to ac- cussed in greater detail in chapters 3, 4, and 5 commodate to some backfits that resulted from and are only summarized here. The most recent the immaturity of the technology and the overly plants to obtain CPs from NRC required 30 to 40 rapid scale-up of plant size. In these cases, regu- months after docketing (i.e., not including the latory delays must be considered positive. More- preliminary utility planning phase) to obtain their over, in some plants that have experienced reg- permits, compared to 10 to 20 months between ulatory delays, such as Cincinnati Gas & Electric 1960-70. Similarly, construction (the time be- Co.’s Zimmer plant, regulatory actions were an tween issuance of the construction permit and appropriate response to evidence of improper the operating license) typically takes 100 to 115 construction practices(9). NRC should not be ar- months, up from the 32 to 43 months in 1960-70 bitrarily limited from imposing backfit require- (18). Backfitting has been suggested as one of the ments that lead to long delays in such cases, since sources of delay, along with deliberate delays due interest in the public health and safety should to a decrease in the need for power and difficul- supercede concern for minimizing Ieadtime and ties in financing construction. cost . In order to examine the impact of regulation In general, OTA concludes that, as in other as- on nuclear powerplant construction Ieadtimes, pects of quality control, skillful management by OTA analyzed case studies of the licensing proc- the utility, its contractors, and NRC is the key ess, which are detailed in volume 2 of this report to avoiding delays that otherwise might result (1). Because it is difficult to separate the effects from the licensing process. Thus, licensing is of backfitting from other regulatory activities, they most likely to proceed without hitches with ex- were considered in the context of the entire li- perienced, committed utility and contractor man- censing process. Based on these case studies, on agement personnel; a clear need for power from published analyses of the causes of increases in the plant; and a constant and open dialogue nuclear plant construction Ieadtimes, and on ex- among NRC staff, nuclear critics, and utility and tensive discussions with parties from all sides of construction managers. Since skillful manage- the nuclear debate, OTA has concluded that the ment has not been a hallmark of NRC administra- regulatory process per se was not the primary tion, changes to make the organization more resource of delay in nuclear plant construction. sponsive and efficient should enhance the li- Rather, during the 1970’s (when Ieadtimes esca- censing process and reduce unnecessary delays. lated the most), utilities delayed some plants de- However, such changes cannot substitute for liberately because of slow demand growth and good utility management and a commitment to financial problems, Plant size was being scaled safety in construction and operation. up very rapidly and construction was begun with incomplete design information. The increasing- Proposals for Change and Evaluation ly complex plant designs meant that more materials—concrete, piping, electrical cable—were re- In 1981, NRC created the Committee for Re- quired, and constructors often experienced de- view of Generic Requirements (CRGR) to respond lays in delivery of equipment and materials. At to some of industry’s concerns and to reduce the same time, worker productivity declined sub- some of the burdens that the utilities felt backfit- stantially, at least in part because plants were ting had imposed on them. The CRGR review more complicated and thus more difficult for the should guide the industry in assigning priorities, utilities to manage and build (3). even if it does not solve some of the more fundamental problems with backfitting. The NRC Backfits did lead to delays in some plants, es- and DOE proposals for reform attempt to address pecially those subject to the extensive regulatory the larger issues, changes that followed the accidents at Browns Ferry and Three Mile Island, but in others the ef- In evaluating the proposals outlined below, it fects of regulatory changes were moderated is important to recognize that backfitting cannot 158 ● Nuclear Power in an Age of Uncertainty be eliminated entirely, but will continue to be ap- ty,” and thus excludes important institutional and plied to plants under construction or in opera- management changes. tion as long as there are outstanding generic safe- One alternative definition has been put forward ty questions. As discussed in chapter 4, the cur- by the NRC Regulatory Reform Task Force (RRTF) rent generation of LWRs still is troubled by a num- in its proposed revisions to the NRC rules: “the ber of potentially serious safety issues even imposition of new regulatory requirements, or the though they have been studied extensively by modification of previous regulatory requirements NRC and industry groups. Nuclear critics are con- applicable to the facility, after the construction cerned that the resolution of problems such as permit has been issued” (23). Prior to the invoca- steam-generator degradation and cracking in pri- tion of a backfit, NRC would set approved de- mary system components might be compromised sign and acceptance criteria for the protection in the interest of limiting backfits. Therefore, they of public health and safety and national securi- are skeptical about proposals that would restrict ty. Once a licensee embarked on the design, con- NRC’s freedom to impose legitimate backfit re- struction, or operation of the reactor and had quirements or emphasize cost and efficiency at committed substantial resources to and was act- the expense of safety. ing in accordance with the NRC criteria, then, The debate about backfitting centers on four according to the definition above, any proposed main considerations of the backfitting rule: 1 ) change in those criteria should be considered a whether the current definition and standard need backfit and should trigger a special decisionmak- to be revised or simply invoked and enforced; ing process. 2) if they do need to be changed, how should A second definition (proposed in the DOE legis- the new definition and standard be phrased, 3) lative package) is “an addition, deletion, or mod- what criteria should be applied by NRC in order- ification to those aspects of the engineering, con- ing a backfit; and 4) whether any changes that struction or operation of a . . . facility upon which may be needed should be made legislatively or a permit, license or approval was issued” (21). through rulemaking. As discussed in the previous This definition may be slightly narrower than the section, some change in the manner in which RRTF definition in that it applies backfit criteria backfits are managed and enforced within NRC only to the conditions in a license rather than to probably is necessary so that the primary regulatory goal of ensuring safety is achieved. More- the full range of regulatory requirements applicable to a facility. over, to provide license applicants with more stability and certainty, and to increase the effective- The most important attributes of any NRC re- ness of public participation in licensing, the back- quirement are explicit criteria and consistent ap- fit procedures and criteria at least must be made plication of these criteria by NRC management. more explicit. Thus, either NRC or DOE proposed definitions would be preferable to the current one, under Definition which it is unclear when a change ordered by The present definition of a backfit in the NRC NRC should be considered a backfit, provided rules includes any design or technological change that the application of the definition by NRC is ordered after issuance of the CP. In doing so, it consistent and clear to all interested parties. Such ignores the reality that much design information a change should contribute to more predictability is not available when the CP is issued, and not about backfits. all evaluations and modifications of designs The definition of a backfit would be particularly should be considered backfits merely because important if it were coupled with a threshold they are postpermitting. From another perspec- standard for triggering it. One approach would tive, however, the present definition may be too require a backfit to result in a substantial increase narrow in that it focuses only on changes in in public protection, with benefits from the in- “structures, systems, or components of the facili- creased protection exceeding both the direct and Ch. 6—The Regulation of Nuclear Power ● 159 indirect costs of the backfit. While such a cost- rects NRC to employ a lower standard of safety benefit standard would presumably assure con- for older plants with shorter remaining operating sistency, OTA concludes that the available meth- lives, even though these are often the plants most odologies are inadequate to fully quantify im- in need of upgrading. Further, the DOE bill would provements in safety. Thus it is likely that a cost- apply to breeder reactors and reprocessing plants benefit standard alone (or the use of quantitative where backfitting is likely to lead to significant safety goals to justify backfits, as discussed in improvements in safety. detail below) would be unworkable until such These procedural changes in the DOE bill methodologies are developed further. Rather, would have the effect of making it more difficult some combination of engineering judgment cou- for NRC to order safety-related improvements pled with cost-benefit analysis, as has been used after a construction permit has been issued. Such in the past, will be necessary. changes will be controversial without other assur- Within this context, however, NRC could im- ances—absent in the DOE bill—that safety can prove the process of evaluating and imposing be assured. backfits by making its standards more explicit and A more general and fundamental change has by specifying the relative consideration to be been proposed by the nuclear industry, which given to factors such as the effects for ordering would like to see NRC’s prescriptive rules re- backfits on public and occupational exposures placed with a few general criteria. In such a sys- to radioactivity; the impact on safety given overall tem, each utility could determine how it might plant system interactions, changes in complexi- best satisfy NRC’s criteria, subject to concurrence ty, and relationship to other regulatory require- by NRC. OTA finds that the latter proposal has ments; the cost of implementing the backfit, in- some merit in that it might encourage innovative cluding plant downtime; the resource burden im- approaches among the more capable utilities and posed on NRC; and, for backfits applicable to vendors. Treating the problems generically rather multiple plants, the differences in plant vintage than prescriptively also might reinforce the use and design. While these factors probably are con- of owners’ groups and data pooling. However, sidered in some form in NRC’s current delibera- it should be noted that such an approach also tions, the decisionmaking process is frequently could pose severe resource problems for NRC. inscrutable. If NRC staff had to review numerous different pro- Other changes could be made in the backfit posals for changes rather than devise a single review process to ensure that criteria and stand- solution of its own, it would severely tax a system ards are applied consistently. A centralized group that already has difficulty with coordination and such as CRGR or ACRS could review backfits rou- organization. tinely and judge them according to standards es- It generally is agreed that the key to a shift to tablished by NRC. Alternatively, an independent performance standards is to make the industry panel of experienced engineers drawn from util- (including utilities, vendors, and AEs) accept full ities, the public, and industry (but not from the responsibility for safety and to design and build organization that did the design) could be set up plants according to a consistent regulatory philos- for centralized review. ophy rather than making numerous modifications as problems rise. Acceptance of this responsibility General Procedures could be demonstrated in part by industrywide Changes in overall procedures and guidelines improvements in management practices, quali- for backfitting also have been proposed. The DOE ty control, performance records, and event-free bill would shift the burden of proof from industry operations. If the evidence indicates that the in- to NRC by requiring NRC to demonstrate that a dustry has matured sufficiently to be able to con- backfit is cost effective. Moreover, the DOE bill struct and operate plants safely and reliably, NRC would restrict the information that NRC can re- may be able to allow more flexibility in the inter- quire from licensees. In addition, it implicity di- pretation of its guidelines. However, as long as 160 ● Nuclear Power in an Age of Uncertainty

any industry participants demonstrate an inabili- tratively, through rulemaking. This would allow ty to guarantee safe operations, OTA believes greater flexibility in adjusting to changing con- that the current level of detail in the backfit reg- struction and operating experience and in apply- ulation probably is necessary to fulfill NRC’s ing risk assessment and cost-benefit analysis than primary legislative mandate of protecting public a backfitting standard rigidly determined by leg- health and safety. islative action. Because of the extensive public comment process associated with rulemaking, it also might permit greater participation in devel- Legislation opment of a backfit rule by all parties. This was OTA found that congressional action is not the rationale followed by NRC in drafting its leg- necessary to change the backfit rule. Changes islative regulatory reform proposal, which did not that would contribute to reactor safety, and lend include provisions related to backfitting. NRC per- stability and certainty to, and increase the effec- sonnel reportedly are working on a draft revision tiveness of public participation in this aspect of of the current rule, which will appear as a notice regulation can be accomplished better adminis- of proposed rulemaking.

HEARINGS AND OTHER NRC PROCEDURES Hearings and other procedural aspects of NRC it would be difficult, if not impossible, to inter- licensing and safety regulation, including the con- pret the act as allowing anything less than a for- duct of safety reviews, management problems mal adjudicatory hearing. Furthermore, trial-type within NRC, the use of rulemaking, and some as- hearings are required under the principles of ad- pects of enforcement, are highly controversial. ministrative law to the extent that the examina- The industry and the utilities perceive the hear- tion of evidence is necessary to resolve questions ings and other procedures as contributing mini- of fact, as opposed to issues of law or policy, mally, if at all, to plant safety, but requiring over- which can be resolved in legislative-type hearings. whelming amounts of paperwork and manage- Part of the debate concerning hearings has fo- ment resources. Nuclear critics, on the other cused on the appropriateness of using an adju- hand, see these procedures as their only means dication process to resolve technical issues. In- of raising safety concerns, and they strongly ob- dustry and utility representatives claim that the ject to any attempts to limit the process and their current system leads to unnecessary delays and participation. inefficient allocation of resources. On the other hand, nuclear critics view the hearing process as Hearings an opportunity to examine NRC records and raise The current licensing process includes adjudi- issues that might have been overlooked. In this catory hearings, * with public participation, before sense, the adjudicatory hearings are appropriate a CP is issued and optional hearings (generally for NRC licensing because they are designed, le- requested) at the OL stage. Formal adjudicatory gally, to illuminate the contested issues of fact and hearings probably are not required under the cause the utility and NRC to justify their technical Atomic Energy Act, which does not specify the decisions more thoroughly than they might in a type of hearing that the Commission must hold. legislative-type hearing. However, they have been granted for so long that Closely associated with the issue of appropriate- *A formal adjudicatory hearing is similar to a trial, in that the par- ness is the efficiency argument. The industry ties present evidence subject to cross-examination and rebuttal, claims that hearings have been too long (spread and the tribunal or hearing officer/board makes a determination out over a year or more in extreme cases) and on the record. The key ingredient is the opportunity of each party to know and meet the evidence and the arguments on the other costly due to the highly technical and complex side; this is what is meant by “on the record. ” nature of the subject matter and the inclusion of Ch. 6—The Regulation of Nuclear Power ● 161 issues not directly germane to safety, such as are addressed in a public forum before the CP need for power and alternative means of gener- or OL hearing. ating that power, It is possible that changes could The industry and some regulators would like be made to the hearing process to reduce ineffi- to see the hearings restructured to a hybrid for- ciencies while preserving the right of the critics mat that would combine some of the elements to participate effectively. As discussed below, leg- of adjudication and legislative-type hearings. In islative action would not necessarily be required, a hybrid hearing, all testimony and evidence since most of the problems could be ameliorated wouId be presented first in written form, as in a by strict conformance to the NRC rules of admin- legislative hearing. Adjudicatory hearings would istrative procedure. be granted on issues that present genuine and A final important consideration is the degree substantial factual disputes that only could be to which critics participate in the decisionmak- resolved with sufficient accuracy by the introducing process and their effectiveness in raising safety tion of evidence in a trial-like setting. concerns. Timing is a central issue concerning Both the NRC and DOE legislative packages participation. In the past, plant designs have been would amend the Atomic Energy Act to provide so incomplete at the CP hearing stage that it has for hybrid hearings. Under the NRC proposal, been virtually impossible to make constructive hearings on CPS would be optional rather than criticisms about them. But by the time the OL mandatory, and the Commission could substitute hearings are held, the final design is complete, hybrid hearings for adjudication, after providing it has been reviewed and approved by the NRC the parties an opportunity to present their views, staff, and the plant is built. Therefore, any con- including oral argument on matters determined cerns the critics raise are directed toward a group by the Commission to be in controversy. Such that has already decided upon the plant’s safety. arguments would be preceded by discovery, and Another issue related to participation is the ef- each party, including the NRC staff, would sub- fectiveness of the interaction between critics and mit a written summary of the facts, data, and the NRC staff. Industry representatives interact arguments to be relied on in the proceeding. The with the staff prior to hearings and reach agree- hearing board then would designate disputed ments on the major safety issues. When the critics questions of fact for resolution in an adjudicatory question these resolutions at the hearings, they hearing based on the standard described above feel that the staff does not give adequate atten- and on whether the decision of the Commission tion to their complaints. Furthermore, the critics is likely to depend in whole or in part on the res- feel that they have even less influence with the olution of a dispute. staff when they are not in an adjudicatory set- The hearings as proposed by NRC and DOE ting. The critics cite occasions on which they would be limited to matters that were not and were ignored by the NRC staff when they infor- could not have been considered and decided in mally raised issues such as emergency core cool- prior proceedings involving that plant, site, or ing, environmental qualification, and fire protec- design unless there was a substantial evidentiary tion. These issues later proved to be major showing that the issue should be reconsidered concerns. based on significant new information, The NRC bill defines “substantial evidentiary showing” as Proposals for Change in the Hearing one sufficent to justify a conclusion that the plant Process and Evaluation no longer would comply with the Atomic Energy Act, other Federal law, or NRC regulations (23). There have been several proposals to address the industry’s and critics’ complaints about the The DOE bill would require hybrid hearings to NRC hearing process, including changing the for- be substituted for adjudication. This maybe con- mat of the hearings, improving management of tracted with the NRC bill, in which the shift to the hearings and other procedures, and chang- a hybrid hearing would be discretionary. The ing the structure of licensing so that safety issues DOE bill would allow anyone to introduce writ- 162 ● Nuclear Power in an Age of Uncertainty ten submissions into the record. Interested par- highly controversial changes in the structure and ties could petition the hearing board for oral ar- scope of the hearings themselves. Furthermore, guments, which would be granted on contentions because proposals for a shift to hybrid hearings that had been backed up with reasonable speci- include more opportunities for requesting hear- ficity. As in the NRC bill, oral argument would ings than under the present licensing process, and be preceded by discovery and submission of writ- more administrative decisions subject to appeal, ten facts and arguments. After oral argument, it is likely that these proposals actually would in- each party could file proposed findings that set crease the amount of time taken up by hearings. forth the issues believed to require formal hear- Other changes in NRC regulations or in man- ings. Under the DOE bill, the hearing board’s de- agement of the hearings could contribute to more cision as to which issues required adjudication efficient hearings. Such changes include: vigor- would be reviewed by the Commission. ously enforcing existing NRC regulations that im- The DOE bill specifies that issues raised and pose time limits in hearings; excluding issues not resolved by NRC in other licensing proceedings raised in a timely manner without a showing of could not be heard again unless “significant, new good cause; requiring all parties to specify the information has been introduced and admitted factual basis for contentions; resolving generic which raises a prima facie showing that action issues through rulemaking once they have been is needed to substantially enhance the public litigated in a licensing proceeding; using summary health and safety or the common defense and disposition procedures for issues not controverted security.” New issues would not be admitted by other parties; excluding issues that were raised unless they were “significant, relevant, material and resolved in earlier proceedings unless a and concerned the overall effect of the plant” showing of significant new information can be on health, safety, or security (21). made; and eliminating consideration of issues not germane to safety that are best considered in Efficiency improvements other forums. Only the last of these changes would require legislative action. Hybrid hearings are intended to increase the efficiency of the hearing process and to improve the effectiveness of public participation in that improvements in Public Participation process. In terms of efficiency, proposals for hybrid hearings are directed toward complaints that The hybid hearings proposed by NRC and DOE hearings are too long and costly and tie up too also can be assessed in terms of the effectiveness much of staff and industry resources without con- of public participation. DOE and the industry tributing to plant safety. Although it is true that argue that these proposals would provide more the hearings can be unduly long and expensive, opportunities for critics to influence the decision OTA found, based on extensive discussions with process. As stated by Secretary of Energy Donald utility and industry representatives, regulators, Hodel: and nuclear critics and public interest groups, that After a plant is essentially complete, with many if management of the hearings were tighter, hundreds of millions–or billions–of dollars either through enforcement of the existing reg- already spent, the view of the public cannot, as ulations or through changes in those regula- a practical matter, be considered as effectively tions, a formal shift to hybrid hearings would as it could be at the beginning of the licensing be unnecessary. Most of the problems cited by process. Therefore, [DOE is] proposing a system the industry that contribute to unnecessarily long with multiple opportunities for public participa- hearings can be remedied through better man- tion early in the process, before firm decisions agement control by the utilities and NRC to en- are made by the Commission and the applicant sure that safety issues are resolved early in the (6). licensing process and through tighter manage- Under the DOE bill, these opportunities would ment of the hearings by the licensing boards or occur if and when standardized plants are con- hearing officers without making fundamental and sidered for approval, when the specific site is con- Ch. 6—The Regulation of Nuclear Power ● 763 sidered for approval, and when the issuance of written submissions and oral presentations, and a combined CP/OL is being considered. The NRC affirm or reverse the board’s designation on each bill would allow hearings at these points as well issue. The critics are especially cautious about as on construction permits, operating licenses, NRC dictating to the hearing boards which issues and preoperational reviews for plants with CP/ to consider. They cite quality assurance at Zim- OLs. mer and the steam generators at Three Mile Island as issues NRC previously has taken away from In analyzing whether hybrid hearings would hearing boards on the grounds that the staff was improve the effectiveness of public participation working on them. In the critics’ view, all of the in licensing, it is important to distinguish the tim- points listed above are serious defects that would ing and number of hearings from the scope of seriously erode public confidence in the effec- those hearings. To the extent that the NRC and tiveness of NRC safety regulation. DOE bills would increase the number of opportunities for public involvement in nuclear plant OTA concludes that the effectiveness of public licensing before final decisions have been made, participation in licensing can be improved with- they would improve the effectiveness of public out causing the hearing process to negatively af- participation in the hearing process. However, fect costs or construction schedules. First, the if those opportunities are not provided when de- proposals for early design and site approvals sign decisions are being made and safety issues would permit extensive public participation in are being raised and resolved—all of which cur- hearings on safety issues prior to the start of con- rently occurs in industry-staff interactions from struction of any particular plant. Then, when a which members of the public are excluded—then utility applied for a CP based on an approved de- the public’s ability to have its safety concerns sign and site, the only questions that would re- heard will not be improved, and the critics still main to be heard in the CP hearings would be will feel that decisions will have been made prior the combination of the site and the design, plus to the hearings. any safety issues that were not resolved in the Nuclear critics contend that the means pro- design approval. This might alleviate the critics’ posed by DOE and NRC to increase the efficien- concern that design-related safety issues are re- cy of the hearings would serve to undercut the solved in private industry-staff interactions. Allow- effectiveness of public participation by severely ing public involvement early and often in utility limiting the scope of that participation. They note planning for nuclear power also would enhance that both bills would weaken the rights of the the effectiveness of public participation in the public to cross-examine NRC and utility wit- licensing process. nesses, which they argue is often the only way to uncover safety problems and uncertainties that Second, a funding mechanism for public par- could not be revealed through examination of ticipation in licensing would ensure that the critics written testimony. Furthermore, the critics feel could make a substantive contribution to design that both bills (but especially the DOE bill) may and safety issues by enabling them to devote make it more difficult for members of the public more resources to the identification and analysis to raise serious safety issues by raising the stand- of reactor engineering and safety. This would re- ards for admission of evidence. spond to the industry’s complaint that the critics are not sufficiently knowledgeable about reac- Nuclear critics also point out that, under the tor engineering and safety, as well as to the critics’ bills’ provisions for hybrid hearings, the hearing view that the utility, and to a lesser extent the board would have to decide in each case which NRC staff, can devote extensive resources to de- evidence is subject to cross-examination—a deci- fending design decisions. sion that often would be appealed, thus lengthen- ing the process rather than shortening it. Under Funding of public participation has been a part the DOE bill, the Commission itself would have of the rulemaking proceedings in the Federal to review the hearing board’s decision, plus the Trade Commission, the National Highway Traf- 164 . Nuclear Power in an Age of Uncertainty

fic Safety Administration, DOE, and the Environ- and an agency decisionmaker that take place out- mental Protection Agency, and of both rulemak- side the hearing and off the record. The Admin- ing and public hearings in the National Oceanic istrative Procedures Act (APA) prohibits such and Atmospheric Administration and the Con- communications once a notice of hearing has sumer Product Safety Commission. In the mid- been published for a particular proceeding. 1970’s, NRC considered an intervener funding When an improper off-the-record contact does program but did not implement one, arguing that occur, the A PA requires that it be placed on the NRC adequately represents the public interest in public record; if it was an oral communication, reactor safety and that the present method of a memorandum summarizing the contact must funding through citizen contributions is a more be prepared and incorporated into the record. democratic measure of public confidence in how Strict interpretation of the ex parte rule effec- well NRC does its job. Given the extent of the tively cuts off communication between the Com- criticism of NRC management and expertise ex- missioners and some parts of NRC during a licens- pressed to OTA by all parties, any policy pack- ing determination or requires that the commu- age intended to revitalize the nuclear option nication be made public. The Rogovin Report should include reconsideration of an intervener recommended more active involvement by the funding program and alternatives such as an of- Commissioners in individual licensing determina- fice of public counsel within NRC. tions, but implementation of this recommendation is constrained by the ex parte rule (14). In Changes in the Role of the NRC Staff its rulemaking options, NRC’s RRTF argued that The NRC staff currently participates as an ad- the Commissioners should be allowed to talk to vocate of the license application in the hearings. staff supervisory personnel who are not partici- This role is a consequence of the detailed involve- pating directly in a particular hearing. The ex ment of the staff in licensing issues and the resolu- parte rule could be interpreted more liberally to tion of most issues to the satisfaction of the staff allow such Commission/staff interaction—espe- and the applicant prior to the hearings. The disad- cially if the role of the staff in hearings is limited— vantage of this situation is that the staff may be as long as true ex parte communications continue perceived as being less effective in resolving safe- to be made public. ty problems than it might be. This concern could be addressed by limiting the staff’s participation Other NRC Procedural Issues in contested initial licensing proceedings to those Additional issues related to NRC procedures in- issues on which it disagrees with the applicant’s clude the role of ACRS in the conduct of safety technical basis, rationale, or conclusions. The staff reviews; management problems within NRC and then would not be perceived as defending a par- other aspects of safety reviews; the use of rule- ticular plant in a hearing and might be more ef- making; and NRC enforcement methods. fective in aiding ASLB. A related issue is the ex parte rule. Like a court Advisory Committee on trial, an agency adjudication is supposed to be Reactor Safeguards decided solely on the basis of the record so that The Atomic Energy Act requires ACRS to review a participant in an adjudicatory hearing will know each license application referred to it, at both the what evidence may be used and will be able to CP and OL stages, even if the Committee does contest it. These rights can be nullified if agency not judge the review to be merited. Many observ- decisionmakers are free to consider facts outside ers consider ACRS to be particularly adept at re- the record without notice or opportunity to re- vealing previously unrecognized safety problems, spond. but because its members devote only part of their The most common problem of extrarecord evi- time to ACRS activities, it has few resources to dence occurs when there are ex parte contacts– pursue such problems in depth. Both the Rogovin communications between any interested party Report and the President’s Commission report on Ch. 6—The Regulation of Nuclear Power ● 165

Three Mile Island recommended that ACRS be small problems can be given proportionately relieved of its mandatory review responsibilities more attention than is warranted. Furthermore, and be allowed to participate in hearings (7,14). the decision path within NRC is virtually untrace- A 1977 NRC review of the licensing process also able, making it difficult to knowledgeably critique recommended that the ACRS be given discretion- the staff’s analysis and resolution of safety ary authority to decide which license applications concerns. merit its review (22). Another concern that is shared by the NRC staff The DOE bill would make ACRS discretion very and the industry is that the regulatory process is broad; it would amend the Atomic Energy Act too cumbersome and legalistic in an area that to make ACRS review of applications to grant, is primarily technical. This produces requirements amend, or renew a CP, OL, combined construc- for an inordinate amount of paperwork and may tion and operating license (COL), or site permit divert attention away from the primary mission discretionary unless the Commission specifical- of ensuring plant safety. For example, the Sholly ly requested a review. Only ACRS review of de- Rule (which requires that a notice be put in the sign approvals or amendments and renewals Federal Register before any change—no matter would be mandatory under the DOE bill. Further, how trivial–is made in a plant’s technical specifi- the DOE proposal specifies that neither the ACRS cations) requires extensive staff attention and re- decision to review nor the NRC decision to refer sources, but produces little accompanying benefit an issue to ACRS would be subject to judicial re- to the public. Similarly, the industry thinks it has view. Under the NRC legislative proposal, ACRS to report too much to NRC, and that significant review of CPS, OLs, site permits, design approv- safety issues may get lost in the resulting paper- als, and amendments to any of these would con- work. tinue to be mandatory. Another problem concerns consistency of re- OTA concludes that the ACRS review of de- views; the SRP helps to even out reviews, but it signs should be mandatory to ensure that safe- is limited by the resources of the NRC staff. This ty problems are identified early in the licensing review is, of necessity, an audit review, with the process. If proapproval of standardized designs ratio of hours spent on the design to those spent were implemented, only discretionary ACRS re- in review on the order of 10,000 to 1. Consistency view of a CP application should be required be- will be increasingly difficult to guarantee as more cause it would be based on a thoroughly studied review functions are shifted to the NRC regional design, Similarly, if site-banking were imple- offices. mented, ACRS reviews of sites also could be dis- It is unclear how to address these concerns. cretionary. Another mandatory ACRS review Good management cannot be legislated. Adding might be appropriate before granting an OL or more technically qualified staff probably would deciding to allow a plant to begin operation improve the quality of substantive reviews but under a COL, since at this stage significant safe- would not necessarily improve management. Fur- ty issues can arise about compliance with the ther, it is generally agreed that the ultimate original design. responsibility for safety rests with the utilities and the industry. Even the most competent and effec- Management Control tive NRC could not make an incompetent or un- According to the industry, the primary prob- willing utility operate safely short of shutting lem with NRC procedures is lack of management down a plant if the utility did not accept this control within NRC, as reflected in uneven safe- responsibility. Financial sanctions other than ty and other reviews, in a lack of priorities, and fines, such as might be imposed through insurers in the problems with backfitting discussed pre- or financiers, may be the most effective in this viously. There does not appear to be any true de- regard. cisionmaking process; rather, NRC appears to re- As noted above in the discussion of backfitting, act to immediate, pressing problems. As a result, it is important that NRC procedures be explicit,

25-450 0 - 84 - 12 : QL 3 166 ● Nuclear Power in an Age of Uncertainty

workable, and applied consistently, but even the mixed oxides. However, considerably more prog- best written regulations or legislation cannot ress could be made in resolving generic issues achieve this if there is not a firm commitment by through rulemaking. top NRC management to ensure that the regulations are implemented properly. For example, the OTA concludes that resolution of generic questions through administrative rulemaking current NRC rules of administrative procedure, for the most part, are adequate to increase the would remove a source of regulatory inefficiency if the rulemaking procedure were improved. effectiveness of the hearing process but are not Moreover, it also would improve the effectiveness enforced by NRC. of participation by the public (including the industry, nuclear critics, and other interested par- The Use of Rulemaking ties) on these issues because of the opportunities for review and comment through publication in In other agencies, increasing the use of rule- the Federal Register and, often, for public hear- making, as opposed to bulletins, circulars, no- ings on a proposed rule. Furthermore, a rule is tices, and regulatory guides, has improved the an enforceable regulation, and thus is a stricter quality of management decisions due to the ex- means of instituting requirements than notices, tensive opportunities for external review and circulars, regulatory guides, or bulletins. Also, as comment by all interested parties. However, NRC discussed previously, generic treatment of safe- is not perceived as being particularly good at ty concerns would facilitate industry use of own- rulemaking. Many NRC rules are considered in- ers’ groups and other management tools. comprehensible due to the poor wording that results from the cumbersome internal review NRC is not particularly adept at rulemaking, process: a rule drafted by the technical staff is producing poorly worded regulations that are dif- revised by numerous others culminating with the ficult to interpret by those who must implement legal staff-by which time it may be unrecogni- and enforce them. But, because of the regulatory zable—but the technical staff is reluctant to and enforcement problems posed by the use of change the wording lest it has to start the review alternatives to rulemaking for generic issues, NRC process all over again. Thus, the staff tends to would be better off to improve its ability to write avoid rulemaking because fulfilling the review re- comprehensible rules than to continue to devel- quirements is likely to make the final product look op solutions to generic problems through licens- much different than the initial intention. ing or notices on individual plants. If NRC could streamline its rulemaking proce- RRTF suggested that a generic question be dure, it might be particularly useful in resolving heard once in a license proceeding and then be generic issues–those common to more than one published as a proposed regulation within 45 plant. As discussed earlier, one of the factors that days after resolution in that proceeding. If the reg- can contribute to inefficiency in the regulatory ulation were adopted by NRC following the req- process is the consideration of generic questions uisite public comment period, it could not be re- during the licensing or oversight of a particular Iitigated in subsequent licensing proceedings un- plant. The Rogovin Report and the President’s less “special circumstances” were shown. If the Commission on the Accident at Three Mile Island “hearing” of the initial rulemaking were in an ad- recommended the increased use of administra- judicatory setting, then this proposal would pro- tive rulemaking procedures to resolve issues that vide for comprehensive discussion of generic is- affect several licensees or plants, as opposed to sues for all parties. However, if the hearing were considering such issues during the licensing or limited to a legislative-type proceeding, critics of oversight of individual plants (7,14). NRC has the regulation may not feel that they really have been heading in this direction over the past dec- been heard. ade with its rulings on emergency core cooling If the RRTF suggestion is not implemented, systems and its environmental statement on some other means of involving as many parties Ch. 6–The Regulation of Nuclear Power ● 167

as possible in the development of rules—before the recipient would be subject to fines or other they are published in the Federal Register for enforcement action if he did not respond. Al- cement—should be devised. NRC is involving the though many observers would prefer to see a industry in the development of the generic rule greater use of rules to change requirements, for radiation protection but not the critics. As a NRC’s current problems with wording could lead result, when the draft rule is published, the critics to enforcement problems. inspectors in the field may view it with skepticism and distrust of the have to enforce a rule based on what was writ- process. Their only recourse would be the public ten, which may differ from what was meant. In- comment process and, ultimately, a petition to dividual judgment on the inspectors’ part as to change the rule after it has been finalized. whether the intent of a rule is being met is discouraged to prevent the matter from ending up In addition to the concern for ruIemaking pro- in court. Yet, inspection and enforcement staff cedures, there is another issue relating to con- rarely are asked to participate in the formulation tent of NRC rules. The current NRC technical reg- of regulations, and thus have little contribution ulations have evolved over a 25-year period with to their enforceability. each new rule devised on a largely ad hoc basis. The relative contribution of each of the numerous Second, there is general agreement that the regulations to safety is undetermined, although current system of fining utilities for violations does it is likely to be highly variable. As one utility rep- not work, although the range of opinions about resentative expressed it: why it doesn’t work is quite broad. The utilities contend that they are less inclined to identify safe- Many codes and standards were contrived and ty concerns when they know that a fine is likely written by well-qualified, well-meaning individuals projecting ideal situations. They never had to follow. Further, they state that the present sys- any idea that in this day and age of rigid quality tem of fines does not distinguish between a one- assurance and quality control, the codes and time simple human failure and continual inatten- standards would be enforced to the letter (1 2). tion to problems or negligence. Utilities would prefer to see a system in which they could begin Some industry representatives and regulators by informally negotiating solutions to safety con- argue that it is now time for a wholesale revamp- cerns with NRC. If the problem is not remedied ing of NRC’s technical regulations to reflect the immediately, the Commission then could resort current state-of-the-art and the accumulated op- to fines and press releases. This procedure cur- erating experience. Before such a radical step, rently is followed by some NRC Regional Admin- however, what is needed is a detailed analysis istrators. of the existing technical rules. A possible starting point would be to initiate a thorough revision of Nuclear critics agree that the present system the technical regulations related to licensing. Any of fines is inadequate, but they cite different rea- such effort also should examine the relative ad- sons. They point out that a large fine ($500,000) vantages and disadvantages of a shift away from is equivalent to a single day’s outage cost for a hardware-based (prescriptive) standards and to- major utility and, in some cases, can be passed ward performance criteria, the role of safety goals on to the ratepayers. They would like to see NRC and PRA, and the source-term work. change its enforcement policy to include the option of shutting down a plant or denying an OL and making it clear that those options will be invoked. The current perception that NRC does not Enforcement enforce the regulations already in place does not With regard to enforcement, nearly all parties bode well for convincing the critics or the indus- to the nuclear debate agree that the procedures try that strong enforcement is a real threat. The could be improved. First, there are some 80 or recent ASLB action in denying an OL for the more means by which NRC can transmit informa- Byron plants may contribute to a change in the tion to a utility or the industry, but only two, perception of NRC’s willingness to enforce its orders and rules, are mandatory in the sense that reguIations. 168 ● Nuclear Power in an Age of Uncertainty

THE TWO-STEP LICENSING PROCESS The current two-step licensing process (CP and and licensing practices could alleviate many of OL) was instituted before the nuclear industry was the nuclear industry’s difficulties in verifying the fully mature. There were many first-time license safety of individual plants. In addition, standard- applicants, designers, and constructors with un- ization could facilitate the transfer of safety proven and incomplete design concepts; at that lessons from one reactor to another and could time, plant designs needed a final evaluation prior help reduce the rate of cost and Ieadtime escala- to operation. Now, reactor engineering may have tions (1 O). As discussed in detail in chapter 4, it matured to the point where final designs for most is likely that any new plants would try to maxi- plants can be described at the CP stage. There- mize these advantages by standardizing designs fore, the industry argues that a two-step Iicens- to the greatest extent possible. ing process no longer is necessary. The NRC legislative proposal specifies that to The utilities and the nuclear industry contend get a COL, an application must contain “sufficient that the two-step procedure exacerbates con- information to support the issuance of both the struction scheduling problems because the plant construction permit and the operating license. ” design, regulatory design review, and hearings The NRC staff analysis of the proposal interprets all occur during construction. They would like this to mean that the application must include to change this to a one-step process that would an essentially complete design. Under the NRC place all three activities before construction bill, an optional hybrid hearing could be re- begins. They believe this would improve the pre- quested before the COL is issued and again be- dictability and efficiency of the licensing process fore the plant goes into operation for matters that by making scheduling more certain. Also, an OL were not considered in the first hearing. The final is perceived in many cases to be pro forma, but review before a plant goes on line would end it still requires a full EIS and optional but usually with NRC issuing an “operation authorization” requested hearings. * They suggest that a one-step that would be the regulatory equivalent of a li- procedure might encourage earlier identification cense for purposes of inspection and judicial and resolution of licensing issues while continu- review (23). ing to accommodate participation by interest This proposal would eliminate the duplication groups and State and local governments. of detailed environmental and safety reviews that There are two ways to achieve the equivalent are currently needed for an OL; otherwise, it is of a one-step NRC licensing process: by combin- the equivalent of the present two-step process ing the CP and OL, and by banking reactor de- with a new name. It is likely that hearings would signs and sites. The NRC and DOE legislative still be held before construction and again before packages include proposals for both of these operation. Moreover, if the plant were a unique measures. it should be noted that neither the rather than standardized design, this procedure DOE nor the NRC bills is tied to the use of stand- could take even more time than the current two- ardized designs, either in the provisions for com- step process. bined CP/OLs or for design banking. However, In the DOE legislative proposal, NRC would in the following discussion of these proposals, it provide an expedited procedure for COL holders is assumed that plants will be much less custom- to start operation by allowing the licensee to cer- ized, relying on only a few standardized and com- tify safety when the plant is virtually complete. plete designs. An earlier OTA study, Nuclear NRC would publish notice of the certification Powerplant Standardization, found that standard- with a 30-day comment period, and the staff ization of designs and construction, operation, would have 45 days from the date of that notice *The ASLB refusal to authorize an OL for Commonwealth Edison to review the plant for safety, consider the pub- Co.’s Bryon plants may indicate a change in approach at NRC. Even if the decision is overturned by the ALAB, it is unlikely that utilities lic comments, and recommend action to NRC. will ever again consider the OL to be a formality. There would then be an additional 30-day peri- Ch. 6—The Regulation of Nuclear Power ● 169 od in which NRC could take action to prohibit visions for “design-banking.” Debate continues, or limit operation if the certification was found however, on the degree of specificity that should to be incorrect. If NRC did not prohibit opera- be required for proapproval of designs and tion during that period, the plant could goon line. whether such approval would act as a disincen- The only opportunity for public hearings would tive to the continued improvement of designs. be at the issuance of the initial COL. Under the NRC legislative proposal, a binding The COL proposal is controversial because of design approval valid for 10 years could be uncertainty about the level of design detail that granted without reference to a particular site and would be required to obtain a combined license, could be renewed for 5 to 10 years unless NRC since this is left up to NRC to specify through found that significant new safety information rele- rulemaking. In addition, neither bill directs NRC vant to the design had become available. The to resolve all outstanding safety issues prior to public would have an opportunity to request hy- licensing. Nuclear critics argue that the number brid hearings on the design before NRC granted of design changes still being made between a CP approval. Issues related to the design could not and an OL and the critical safety issues still be- be raised in a subsequent CP, OL, or COL hearing uncovered at the OL stage indicate that the ings unless the combination of a design with a industry and NRC are not yet ready for one-step particular site resulted in new issues that had not licensing. Such a procedure could reduce atten- been addressed in the design approval or there tion to unresolved safety issues raised at the CP was convincing evidence that reconsideration of stage and could be used to restrict NRC’s ability design issues was necessary. to order backfits. Regulators and critics especially object to the DOE bill because it allows the li- The DOE bill also would allow utilities to censees themselves to certify safety, with a limited choose a preapproved plant or major subsystem time for the NRC staff to verify that certification, design as an alternative to selecting a unique and no real opportunities for citizen participation. plant design. Design approvals would be subject Some utilities are not convinced that a one-step to hybrid hearings. Once approved, a design process would be any more predictable than the would be valid for 10 years and then could be current two-step process in terms of requirements renewed for 10 years but would be subject to the for a license and backfits. Furthermore, it is possi- same backfitting requirements as normal plants ble that the proposed COL procedure, when cou- under the DOE bill. Preapproved designs would pled with hybrid hearings, would take longer than be incorporated into a CP or COL application, the current CP and OL process. Using procedural and the review of design issues in the hearings changes to improve the management of the hear- would be strictly limited. The DOE bill would re- ings and implementing site- and design-banking, quire NRC to define the level of detail necessary which together would serve as a surrogate for for design approvals through the normal rulemak- one-step licensing, probably would do more to ing process. increase the efficiency and predictability of licens- Proapproval of standard designs might make a ing than a switch to a COL. substantial contribution to a more efficient and Current NRC regulations allow for design re- predictable licensing process by removing most view prior to the filing of the CP application, but design questions from the licensing of a particular the results of the review are not binding upon plant, but it is likely to be as controversial as the the CP determination. Alternatively, reactor ven- proposal for a COL. Issues include the degree of dors can submit generic designs for approval specificity required for design approval, the con- through rulemaking. Many industry analysts ar- ditions and procedures under which the utility gue that reactor engineering has matured suffi- or its contractors could deviate from a preap- ciently to allow proapproval of standardized plant proved design once construction has begun, and designs, or of major system or subsystem designs, the ability of NRC to order backfits on approved and both the NRC and the DOE bills include pro- designs. 170 . Nuclear Power in an Age of Uncertainty

Nuclear critics are concerned that discussion ply for site approvals, thus encouraging broader of new or previously unresolved safety issues planning. In the NRC bill, a site approval would would be foreclosed in the CP or COL hearings not preclude the use of the site for an alternative on preapproved standardized designs, especial- or modified type of energy facility or for any other ly in light of the provisions that prohibit the rais- purpose. However, other uses not considered in ing of generic safety issues in the licensing of par- the original approval may invalidate the site per- ticular plants and of the provisions that shift the mit, as determined by NRC. The public would burden of proof to the public to show that a pre- have an opportunity to request hybrid hearings approved design does not meet current safety on the site approval, but issues related to the site standards. Proponents of this change argue that would be excluded from further licensing pro- proapproval of designs could improve the effec- ceedings unless matching the site with a particular tiveness of public participation in that it would plant design raised issues that were not consid- allow earlier and more detailed discussion of ered at the time of the site approval. design issues in hearings without the time constraints imposed by the licensing of a particular The DOE proposal is similar to NRC’s, except plant. the site-approval procedure in the DOE bill would The critics also object to the length of time for not allow alternative uses and would allow CP which a design approval would be valid, given applicants to perform limited construction activ- the frequency with which design changes have ities before issuance of a permit. A site approval been instituted in the past, and to the subsidy would be valid for 10 years, with 10-year renew- granted by deferral of the application fee until als. Under the DOE legislative proposal, the pub- the design is used. Furthermore, there is concern lic could request hybrid hearings prior to NRC that once a design has been approved, the vested approval of a site. interest in it would remove any incentives to improve it. However, as discussed in chapter 7, the As with design approvals, OTA concludes that industry argues that its need to remain competi- site-banking could improve the efficiency and tive with foreign countries should be incentive predictability of the licensing process by taking enough. siting out of the critical path entirely. As long as the site-approval process allows adequate op- In the present system, NRC approval of site suit- portunity for public participation and ensures ability is not initiated until the CP application is docketed, which places site review on the “crit- consideration of issues related to the combina- ical path” for reactor licensing. The existing NRC tion of a particular site and design prior to issu- regulations permit review of site suitability prior ance of a CP, binding early site approval should not be a controversial change. In fact, severing to filing of the CP application, but the outcome of this review is not binding in the final CP deci- site approval from the CP could facilitate earlier and more substantive public participation. The sion unless a special ASLB decision is obtained. principal objections nuclear critics have to these Both the NRC and the DOE legislative packages recommend a procedure for binding early site ap- bills are the length of time for which approvals proval that would be independent of a CP appli- are valid (including renewals, 20 years in the NRC cation. bill and an indefinite period in the DOE) and the subsidy introduced by deferring the application In the NRC legislative proposal, a site approval fee until the site actually is used or the approval that does not reference a particular nuclear plant expires. Furthermore, the selection of particular could be granted for up to 10 years, with renewal sites—whether they are matched with a plant or possible for 5 to 10 years. Federal, State, regional, not—will remain controversial, as discussed in and local agencies, as well as utilities could ap- chapter 8. Ch. 6—The Regulation of Nuclear Power ● 171

OTHER NRC RESPONSIBILITIES

In licensing a nuclear powerplant, NRC is re- Although NRC weighs the opinion of the justice quired to make several determinations that are Department heavily in its determination, the not related directly to safety. These include cer- Commission remains responsible for the final an- tification of the need for power from the plant titrust decision. As in need for power, it may not (required under NEPA) and of compliance with be appropriate for NRC to devote staff resources antitrust laws, to antitrust law. One option is for NRC to adopt the Justice Department’s decision on antitrust un- There is general agreement that NRC is poorly less an affected party objects within a specified equipped to judge need for power on a local or time after notice of the decision. If the objection regional basis, and therefore that it is a waste of is found to have merit, then NRC could remand staff resources to make such a determination. to Justice for further consideration or do an in- Moreover, at least 45 States already require other dependent review. Legislative action would be agencies to determine the need for power either required to delegate this authority to the Justice in the certification or licensing of powerplants, Department. in rate cases, in the approval of financing, or in an independent planning process (20). Further- The U.S. General Accounting Office (GAO) more, evaluations of the need for power and the also has recommended that the NRC provide bet- choice of alternative types of generating technol- ter coordination with State and local governments ogies can take up hearing time and staff time that in NEPA reviews. At least 23 States have statutes could be better spent in the analysis of safety and requiring preconstruction environmental reviews design issues. similar to those required under NEPA, but NRC’s NEPA regulations make no provision for coordi- Both the .NRC and the DOE legislative packages nation with the States or for eliminating duplica- provide for binding NRC acceptance of a need tion of efforts. GAO recommends NRC work for power determination made by a Federal, State jointly with all the States to identify common legal or other agency authorized to do so. The NRC and procedural requirements as a first step in bill also provides for acceptance of other agen- coordinating environmental reviews (2). cies’ rulings on alternative sources of generating capacity. Only where no other agency is required Finally, it has been suggested that introducing to make such a determination would NRC per- a little flexibility into the concept of exclusive form a de novo review of the need for power. Federal jurisdiction over reactor regulation would In both bills, these provisions are embedded with- go a long way toward alleviating State and local in the section on a one-step licensing process, concerns and improving public acceptance. For but they could be separated out. Because each example, Oregon has a memorandum of un- agency is required under NEPA to make these derstanding with NRC that sets forth “mutually determinations, legislative action would be re- agreeable principles of cooperation between the quired to delegate that authority to the States or State and NRC in areas subject to the jurisdiction other Federal agencies. it is possible that this proof the State or the NRC or both. ” This memoran- vision would result in expanded opportunities for dum is intended to minimize duplication of ef- public participation in the discussions of the need fort, avoid delays in decisionmaking, and ensure for power and choice of technology. However, the exchange of information that is needed to neither bill sets minimum standards for public make the most effective use of the resources of participation in delegating this authority to the the State and NRC. To accomplish these ends, States, nor do the bills mandate consideration of the memorandum provides for potential future the full range of alternatives, as required in NEPA. subagreements in areas of mutual concern, including siting of nuclear facilities, water quality, Under current practice the Department of Jus- nuclear plant operation, radiological and environ- tice performs a comprehensive review of license mental monitoring, decommissioning of nuclear applications for compliance with antitrust laws. plants, emergency preparedness, personnel train- 172 . Nuclear Power in an Age of Uncertainty ing and exchange, radioactive material transpor- exchange of information, and an agreement on tation, and other areas. Subagreements adopted resident inspectors at the Trojan plant, the only to date include a protective agreement for the nuclear powerplant in Oregon (11).

SAFETY GOALS One concept that has attracted much attention approach allows it to capture the benefits of qual- in discussions of backfitting and other changes itative goals and quantitative guidelines in meas- in the NRC technical regulations is the use of safe- uring performance while avoiding the vagueness ty goals* to establish safety requirements and of qualitative goals without numerical guidance. gage the need for changes in those requirements. It does not lock NRC into quantitative goals that may not be able to yield technically supportable NRC currently is developing a safety goal poli- results given the uncertainties inherent in quan- cy, and the DOE legislative proposal emphasizes titative risk assessment. the importance of this effort by endorsing the Commission’s efforts. The DOE bill would require The qualitative safety goals established in the NRC to report to Congress within 1 year on its NRC policy statement are: progress in developing and implementing a safety Individual members of the public should be goal policy. The NRC proposal is described be- provided a level of protection from the conse- low. quences of nuclear powerplant accidents such that no individual bears a significant additional risk to life and health NRC Safety Goal Proposal Societal risks to life and health from nuclear powerplant accidents should be as low as rea- NRC has issued a policy statement on safety sonably achievable and should be comparable goals for nuclear powerplants that is being used to or less than the risks of generating electricity on an experimental basis (25). It currently plays by viable competing technologies (25). no part in licensing decisions, and license applicants do not have to demonstrate compliance The intent of the first safety goal is to require with it. If the proposed policy receives sufficiently a level of safety such that individuals living or favorable response, NRC will consider amending working near nuclear powerplants should be able its regulations to include safety goals in licens- to go about their daily lives without special con- ing decisions. cern by virtue of their proximity to such plants. The second safety goal limits the societal risks In developing a safety goal policy, NRC consid- posed by reactor accidents and includes an im- ered qualitative goals that would interpret the plicit benefit-cost test for safety improvements to Atomic Energy Act’s standard of adequate pro- reduce such risks. tection of public health and safety, as well as quantitative goals that could provide a more ex- These goals focus on nuclear powerplant acci- act standard against which risks could be meas- dents that may release radioactive materials to ured. Qualitative goals were adopted to lend NRC the environment. They do not address risks from safety decisions “a greater coherence and pre- routine emissions, from other parts of the nuclear dictability than they presently appear to have,” fuel cycle, from sabotage, or from diversion of supported by numerical guidelines as goals or nuclear material. The policy statement notes that benchmarks (25). The NRC report notes that this the risks from routine emissions are addressed in current NRC practice through environmental im-

*NRC defines a safety goal as “an explicit policy statement on pact assessments that include an evaluation of the safety philosophy and the role of safety-cost tradeoffs in the NRC radiological impacts of routine operation of the safety decisions” (25). plant on the population around the plant site. For Ch. 6—Regulation of Nuclear Power ● 173 all plants licensed to operate, NRC has found that prompt fatalities that might result from reactor routine operations will have no measurable radio- accidents should not exceed 0.1 percent of the logical impact on any member of the public. sum of prompt fatality risks resulting from other Therefore, the object of the experimental policy accidents to which members of the U.S popula- is to develop safety goals that limit to an accept- tion are generally exposed. able level the additional potential radiological risk The risk to an individual or to the population in the area near a nuclear powerplant site of that might be imposed on the public as a result cancer fatalities that might result from reactor ac- of accidents at nuclear powerplants. cidents should not exceed 0.1 percent of the In establishing the numerical guidelines to sup- sum of cancer fatality risks resulting from all port these safety goals, NRC noted that progress other causes. in developing probabilistic risk assessment (PRA) The benefit of an incremental reduction of risk below the numerical guidelines for societal mor- techniques and in accumulating relevant data tality risks should be compared with the associ- since the 1974 Reactor Safety Study (24) has led ated costs on the basis of $1 ,000/man-rem to recognition that it is feasible to begin to use averted. quantitative assessments for limited purposes. The likelihood of a nuclear reactor accident However, because of the sizable uncertainties still that results in a large-scale core melt should nor- present in the methods and the gaps in the data mally be less than 1 in 10,000 per year of reac- base–essential elements in gaging whether the tor operation (25). guidelines have been met–NRC indicated that In its experimental safety goal proposal, NRC the quantitative guidelines should be viewed as left open a number of questions for future con- goals or numerical benchmarks that are subject sideration. These include: whether the benefit to revision as further improvements are made. side of the tradeoffs should include the economic Many of the participants in the Safety Goal Work- benefit of reducing the risk of economic loss due shops held by NRC agreed that quantitative goals to plant damage and contamination outside the were not feasible at this time, but numerical plant; whether a numerical guideline on availa- guidelines could be used to support qualitative bility of containment systems to mitigate the ef- goals. Finally, in setting the numerical guidelines, fects of a large-scale core melt should be added; NRC specified that no death attributable to a reac- and whether there should be a specific provision tor accident ever will be “acceptable” in the for risk aversion and, if so, what it should be. In sense that the Commission would regard it as a addition, the proposal sought further guidance routine or permissible event. NRC intends that on developing a detailed approach to implement- no such accidents occur but recognizes that the ing the safety policy, including decision making possibility cannot be eliminated entirely. under uncertainty; resolving possible conflicts With these caveats, NRC established four ex- among quantitative aspects of issues; the ap- perimental numerical guidelines: two for individ- proach to be used for accident initiators that are ual and societal mortality risks for prompt and difficult to quantify (e.g., seismic events, sabotage, delayed deaths; a benefit-cost guideline for use human and design errors); the terms for defini- in decisions on safety improvements that would tion of the numerical guidelines (e.g., median, reduce those risks below the levels specified in mean, 90-percent confidence); and identifying accordance with the longstanding regulatory prin- the individuals to whom the numerical guidelines ciple that risks from nuclear power should be “as should be applied (e.g., the individual at greatest low as reasonably achievable”; and a plant per- risk, the average risk). formance guideline that proposes a limitation on Shifting from prescriptive regulation to a safe- on the probability of a core melt as a provisional ty goal approach could have far-reaching conse- guideline for NRC staff use in reviewing and eval- quences. Such a change might contribute to a uating PRAs of nuclear powerplants. These guide- more favorable regulatory environment for the lines are: nuclear utilities since the number and unpredict- The risk to an individual or to the population ability of regulatory actions probably would be in the vicinity of a nuclear powerplant site of reduced. Furthermore, utilities would be allowed 174 Ž Nuclear Power in an Age of Uncertainty to select the least costly route to compliance, with redundant and diverse means of protection resultant gains in efficiency. Another result of the against core damage, sound containment, suffi- safety goal approach might be to encourage di- cient distance from populated areas, effective versity and innovation in developing alternatives emergency preparedness, and, of course, careful for improving safety. Such activities, however, attention to quality assurance in construction may not be consistent with the standardization and operation. To provide guidance to the NRC technical staff and the nuclear industry, and to of nuclear powerplants (4). inform the public, the Commission should distill The proposed safety goals and numerical guide- its experience and state clearly and succinctly lines are not free of controversy. The proposed that each of these [engineering] principles must guidelines have been criticized as being “too be satisfied separately, and how this is to be remote from the nitty-gritty hardware decisions done. Unfortunately the Commission seems to be on an opposite course (25). that have to be made every day by designers, builders, operators, and regulators to be of much The nuclear critics object more strongly to the use” (25). Most regulators and industry repre- safety goal proposal, arguing that to adopt goals sentatives agree that while, in principle, it would with no viable means of confirming their achieve- be nice to be able to use overall goals to supplant ment is a useless exercise. They do not believe the myriad specific decisions NRC must make there is any immediate prospect of PRA being de- about the adequacy of hardware and procedures, veloped sufficiently to provide a means of con- they find the proposed goals too generaI and firmation. Therefore, the critics argue that it is not abstract to provide specific guidance for dealing feasible to use quantitative guidelines for limited with practical questions, and withhold judgment purposes, and NRC only misleads the public in on whether they will prove useful. As one Com- saying that PRA calculations will be used to sup- missioner noted, the only reliable guides to reac- port qualitative goals. tor safety remain time-tested engineering principles:

LICENSING FOR ALTERNATIVE REACTOR TYPES

Nuclear powerplant licensing experience in the Small LWRs contributed greatly to the original United States, for the most part, is based on the development of commercial nuclear power in the LWR design concept. The exceptions are the Fort United States. However, as operating experience St. Vrain high-temperature gas-cooled reactor grew, apparent economies of scale motivated util- (HTGR), which achieved full power in 1981, and ities to purchase larger reactors. Today the norm the Clinch River breeder reactor. Yet variations is over 1,000 megawatts electric (MWe), but in- on the LWR and other reactor design concepts terest in smaller reactors is reemerging, primari- are attracting attention for their possible safety ly for financial and system flexibility reasons. A and reliability advantages over the LWR, as shift to smaller reactors could not be accom- discussed in chapter 4. Given the extent of the plished by replicating existing small plants licensing and regulation experience with LWRs, because the designs of those plants do not meet it is reasonable to question whether a shift to a all current safety requirements. NRC has estab- different design would entail substantial changes lished a systematic evaluation program specifical- in the regulatory process, such that the same ly to review these older designs and improve their problems encountered in the regulation of LWRs safety where possible. New small reactors would would be repeated with alternative reactors, and require new designs based on current NRC reg- whether the development of a licensing process ulations, although such designs would not for such reactors would delay their implementa- necessarily differ substantially from large LWRs tion. except in the size of the core and other plant Ch. 6—The Regulation of Nuclear Power ● 175 components. Thus the regulatory process prob- rience of the Fort St. Vrain plant and the 900 ably would be similar to that for current large MWe prototype that DOE is sponsoring. LWR designs, including the potential for backfits, The heavy water reactor, as represented by the unless small LWRs were standardized within the well-proven Canadian CANDU design, has attrac- context of proapproval of designs. tive safety and reliability features, but licensability The high temperature gas reactor has little is a major constraint on the adoption of this de- operating experience in the United States. The sign by U.S. utilities. The NRC requirements for primary safety concerns are quite different from seismic protection and thicker pressure tube walls the LWR and have not been studied as intensive- would require design changes in the CANDU that ly. As a result, the potential for the emergence might reduce its efficiency and could lead to of significant unforeseen safety concerns prob- backfits until these changes were proven. NRC ably is higher than for the LWR. On the other would have to establish new design criteria and hand, inherent characteristics of HTGRs make standard review plans before a heavy water reac- them less susceptible to certain types of accidents tor could be licensed. that can progress more quickly or have more serious consequences in a LWR. This eventually may The Process Inherent Ultimate Safety (PIUS) simplify the licensing process after any initial reactor is the least developed of the alternative problems are resolved. design concepts. It has readily visible safety advantages, but they might not be accounted for During the early 1970’s, several utilities made in the regulatory process until significant CP applications for HTGRs. As a result, NRC operating and construction experience is estab- made a significant effort to formalize design re- lished. If PIUS is forced to include all the engi- quirements and establish review plans for the neered safety features of the LWR, it is not likely HTGR. Nevertheless, several years would be re- to be competitive. Successful development of quired to make the regulatory process for this PIUS, therefore, depends on NRC determining design as mature as that for LWRs. Backfitting re- what level of safety is appropriate and crediting quirements for the HTGR are uncertain but the inherent safety features of PIUS during the should be reduced through the operating expe- design approval and licensing process.

CHAPTER 6

1. Cohen, S. C., Lobbin, F. B. and Thompon, S. R, 6. Hodel, D., statement before the Subcommittee on “The Future of Conventional Nuclear Power: Nu- Nuclear Regulation of the Committee on Environ- clear Reactor Regulation, ” SC&A, Inc., Dec. 31, ment and public Works, U.S. Senate, May 26, 1982. 1983. 2. Comptroller General of the United States, “Reduc- 7. Kemeny, j. G., “Report of the president’s Commis- ing Nuclear Powerplant Leadtimes: Many Obsta- sion on the Accident at Three Mile Island, ” Oc- cles Remain, ” EMD-77-15, Mar. 2, 1977. tober 1979, Washington, D.C. 3. Crowley, j. H., and Griffith,, j. “U.S. Construction 8. Matiel, L., Minnek, L. W., and Levy, S., “Summary Cost Rise Threatens Nuclear Option,” /Vuc/ear of Discussions with Utilities and Resulting Conclu- Engineering /ntemationa/, June 1982. sions, ” Electric Power Research Institute, June 4. Fischoff, B., “Acceptable Risk: The Case of Nucle- 1982. ar Power,” Journal of Policy Analysis and Manage- 9. “NRC Pulls the Plug on the Zimmer Nuclear Pow- ment, vol. 2, No. 4, 1983. er Plant,” The Energy Dai/y, vol. 10, No. 215, Nov. 5. Freeman, S. D., “The Future of the Nuclear Op- 16, 1982.

tion,” Environment, vol. 25, No. 7, 1983. 104 Office of Technology Assessment, U.S. Congress, 176 ● Nuclear Power in an Age of Uncertainty

Nuclear Powerplant Standardization–Light Water 20. U.S. Department of Energy, “The Need for Power Reactors, OTA-E-1 34, June 1981. and the Choice of Technologies: State Decisions 11, Oregon Department of Energy and the U.S. Nucle- on Electric Power Facilities, ” DOE/EP/l 0004-1, ar Regulatory Commission, Memorandum of Un- June 1981. derstanding, Jan. 4, 1980. 21. U.S. Department of Energy, “Nuclear Licensing 12, Patterson, D. R., “Revitalizing Nuclear Power and Regulatory Reform Act of 1983,” draft bill sub- Plant Design and Construction: Lessons Learned mitted to Congress, March 1983. by TVA,” October 1981. 22 U.S. Nuclear Regulatory Commission, “Nuclear 13, Phung, D. L., “Light Water Reactor Safety Since Power Plant Licensing: Opportunities for improve- the Three Mile Island Accident,” Institute for En- ment,” NUREG-0292, June 1977. ergy Analysis, July 1983. 23. U.S. Nuclear Regulatory Commission, “Nuclear 14. Rogovin, M., et al., “Three Mile Island: A Report Power Plant Licensing Reform Act of 1983,” draft to the Commissioners and to the Public,” U.S. Nu- bill submitted to Congress, February 1983. clear Regulatory Commission, NUREG/CR-l 250, 24. U.S. Nuclear Regulatory Commission, “Reactor February 1980, Safety Study,” WASH-1400, October 1975. 15. SC&A, Inc., “Nuclear Reactor Regulation and Nu- 25. U.S. Nuclear Regulatory Commission, “Safety clear Reactor Safety, ” Sept. 6, 1983. Goals for Nuclear Power Plants: A Discussion 16. Title 50 of the Code of Federal Regulations, Part Paper,” NUREG-0880, February 1982. 50.1 09(a). 26. U.S. Nuclear Regulatory Commission, “A Survey 17. Title 10 of the Code of Federal Regulations, Part by Senior NRC Management to Obtain Viewpoints 50, Appendix E. on the Safety Impact of Regulatory Activities From 18. United Engineers & Constructors, Inc., “Nuclear Representative Utilities Operating and Construct- Regulatory Reform,” UE&C 810923, September ing Nuclear Power Plants,” NUREG-0839, August 1981. 1981. 19. United Engineers & Constructors, Inc., “Regula- tory Reform Case Studies,” UE&C 820830, August 1982,

Chapter 7 Survival of the Nuclear Industry in the United States and Abroad

Photo credit: Electricite de France I Four nuclear units at Dampierre, France Contents

Page Introduction ...... 179 The Effects in the U.S. Nuclear lndustry of No, Few or Delayed New-Plant Orders 1983 to 1995...... 179 Reactor Vendors ...... 180 Nuclear Component Suppliers ...... 182 Architect-Engineering Firms ...... 183 The lmpact on Nuclear Plant Operation...... 183 The lmpact on Future New Construction of Nuclear Plants ...... 186 The Prospects for Nuclear Power Outside the United States ...... 189 The Economic Context for Nuclear Power ...... 189 Public Acceptance ...... 190 Foreign Technical Experience: Plant Construction and Reliability ...... 191 Licensing and Quality Control ...... 191 Overall Nuclear Development...... ,...... 194 Implications for the U.S. Nuclear Industry...... 197 The U.S. Industry in an International Context ...... 199 Nuclear Proliferation Considerations ...... 200 Conclusion ...... 203 Chapter 7 References ...... 207

Table No. Page 23. NSSS/AE Combination of Light Water Reactors Under Construction or On Order As of 1981 ...... 180 24. Forecasts of Installed Nuclear Capacity in OECD Countries ...... 189 25. Sample Construction Times for Nuclear Plants in Various Countries...... 192 26. Operating Performance by Country ...... 195 27. Categories of Foreign Nuclear Programs in Western Europe and the Asia-Pacific Region 194 28. Structure of Electric Utility in Main Supplier Countries and the Distribution of Authority Over Key Decisions Influencing Financial Health of Utilities ...... 196 29. Estimated Reactor Export Market in Units ...... 201 30. No-Nuclear Weapons Pledges in Effect in 1981 ...... 202 7A. Onsite and Offsite Nuclear-Related Job Vacancies at lNPO Member Utilities, Mar. 1,1982204 7B. Nuclear-Generating Capacity Outside the United States ...... 205 7C. Usability of Nuclear Materials for Nuclear Weapons or Explosives...... 206

Figures

Figure No. Page 34. Engineering Manpower ...... 181 35. Estimates of Additional Manpower Requirements for the Nuclear Power industry, 1982-91186 36. Nuclear Engineer Degrees: Foreign Nationals and U.S. Citizens ...... 187 37. Average Annual Increase in Electricity Demand: France, West Germany, and United Kingdom ...... 190 38. Engineering and Construction Man-Hours per Megawatt of Capacity ...... 194 Chapter 7 Survival of the Nuclear Industry in the United States and Abroad

INTRODUCTION

Whether or not utility executives order more Although there are no strict parallels between powerplants (given all the uncertainties and dis- the U.S. nuclear industry and that of any other incentives described in earlier chapters) has direct country, there nonetheless is much to be learned implications for the U.S. nuclear industry and its from foreign experience. Many of the same probability to remain viable as a source of nuclear lems faced by the U.S. industry are being faced powerplants both within the United States and elsewhere: public opposition to nuclear power, abroad. This chapter examines the consequences slow demand growth, and the difficulty of con- for different parts of the U.S. industry of a long trolling cost and time overruns in nuclear plant period with no orders for new plants or a period construction. Understanding how these and other in which orders for new plants follow a long de- problems are being coped with in each country, lay. The chapter then surveys the prospects for provides some perspective on the U.S. situation nuclear power abroad and the likelihood of U.S. and information on approaches that might be exports as well as the possibility that the United successful in the United States. States might be able to turn to foreign suppliers as future sources of the technology.

THE EFFECTS IN THE U.S. NUCLEAR INDUSTRY OF NO, FEW OR DELAYED NEW-PLANT ORDERS 1983 TO 1995 The nuclear industry may be portrayed as a Of about 90,000 employees of the nuclear in- monolith by its critics. I n fact, however, it has dustry, about half operate and maintain commer- always been a loose-knit group of several hun- cial power reactors (as well as some test and dred businesses and organizations, given what research reactors), a quarter are engaged in reac- cohesion it has by the demands of a difficult tech- tor and reactor component manufacturing, and nology and the need to develop a coordinated a quarter are engaged in design and engineer- response to critics. Today, the industry consists ing of nuclear facilities (other than design of the 59 public and private utilities that are the associated with reactor manufacture) (4). principal owners of nuclear powerplants in opera- Companies and organizations in each of these tion or under construction, 4 reactor manufac- sectors must develop strategies for coping with turers also known as nuclear steam supply sys- the likelihood of no new orders for nuclear plants tems (NSSS) vendors, 12 architect-engineering for 3 to 5 years and the possibility of no or very (AE) firms with a specialty in nuclear design and few orders for 5 or more years after that. In a construction, about 400 firms in the United States comprehensive study for the U.S. Department of and Canada qualified to supply nuclear compo- Energy (DOE), the S. M. Stoner Corp. (37) as- nents, and several hundred nuclear service con- sessed the impact on NSSS vendors and compo- tractors. Table 23 shows the combinations of nent suppliers of three possible futures: reactor manufacturers and AE firms for plants under construction or on order as of the spring ● a slowly increasing projection of: no orders of 1981. until 1986, an average of two to three a year

179 180 ● Nuclear Power in an Age of Uncertainty

Table 23.—NSSS/AE Combination of Light Water Reactors Under Construction or On Order As of 1981 Reactor vendors Architect/ General Combustion Babcock & engineering firms Westinghouse Electric Engineering Wilcox Bechtel ...... 6 10 6 5 Burns & Roe ...... — 1 1 — Black & Veatch...... — 2 — — Brown & Root ...... 2 — — — Ebasco ...... 4 1 4 — Gilbert/Commonwealth. . . . 1 2 — — Gibbs & Hill ...... 2 — — — Gilbert Associates...... — — — — Utility Owner ...... 7 — 6 — Fluor Power Services . . . . . — — — — Sargent & Lundy ...... 8 7 — — Stone & Webster ...... 5 6 2 2 United Engineers ...... 2 — — 2 Tennessee Valley Authority, ...... 3 6 2 2 SOURCE Nuclear Powerplant Standardization Light Water Reactors (Washington, D.C. U.S. Congress, Office of Technology Assessment, OTA-E-134, April 1981)

until 1989, and six to eight orders a year after quiring sophisticated skills and a sound knowl- that; edge of nuclear physics. Used fuel rods are re- ● no orders until the early 1990’s; and moved and new fuel rods are inserted among par- ● no orders until 1988 or 1989 and an average tially used fuel rods, and the array of both fresh of one a year for 5 years after that. and older fuel rods is then reconfigured to provide maximum nuclear energy. Vendors expect The findings of the Stoner study are echoed in also that spent fuel management will also be a the results of 35 interviews conducted by OTA continuing source of business. with representatives of reactor vendors, nuclear suppliers, AE firms, utilities with nuclear plants, Vendors are now competing for the nuclear and industry analysts and regulators. Further in- service business in an arena once dominated by sights are available from several assessments of the nuclear service consultants. The Stoner report personnel needs for the industry (4,9,16,18). estimates further that backfits and rework may require 30 to 50 man-years of contracted engi- Reactor Vendors neering work per operating plant with a total demand of 3,000 to 6,000 technical people per year. No new nuclear reactors are now being built. (37) The vendors are uncertain, however, if the The nuclear business for the four reactor vendors current level of backfits, stimulated largely by currently consists of assembling at site, fuel load- requirements following the Three Mile Island ac- ing and services, and the latter two will continue cident, will continue beyond the next few years, regardless of what happens to new orders. Figure and, at the same time realize that over the long 34 shows one vendor’s prediction of the need run the continued cost of backfits will discourage for engineering manpower through the 1980’s. new orders. Engineers will be needed for services to operating The only current new plant design activities are plants and fuel loading. Manpower to handle joint ventures by both GE and Westinghouse with changes in existing plants, and rework in plants Japanese companies. The Westinghouse project under construction, will initially increase but then is being aimed at both the domestic and export diminish. The need for engineering manpower markets and is being developed in consultation to design new NSSS will practically disappear. with the Nuclear Regulatory Commission (NRC). Refueling, which occurs in each plant approx- (See ch. 4.) The GE project is being developed imately every 18 months, is a demanding task re- for Japan only and not for future U.S. licensing.

Ch. 7—Survival of the Nuclear Industry in the United States and Abroad • 181

Figure 34.-Engineering Manpower

Operating plant services

100 Nuclear fuel

Changes and delays

Original nuclear steam supply system scope 75

Year supplier,

SOURCE A S Program (Washington, D C.: Department of Energy, November 1981)

Both companies hope that an export market dence that some orders have already been lost will sustain some of their design and manufac- because U.S. vendors are losing their reputation turing capability. The likely export market (de- for up-to-date technology. As a Finnish source scribed later in this chapter), however, has shrunk told Nucleonics Week: “Why should Westing- to only a fraction of what was projected 5 years house put in millions (of dollars) for R&D if they ago and is currently substantially less than what don’t have business prospects. That is one reason can fully use worldwide manufacturing and de- why we are not studying their (U. S.) reactors in sign capability. a [plant-purchase] feasibility study” (29). For U.S. companies to compete successfully re- Moreover, future export orders are likely to in- quires not only continued technical success (as volve reduced U.S. manufacturing demand since is being attempted in these joint ventures) but also many of the countries most likely to pursue nu- possible modifications in U.S. export financing clear programs have nuclear import policies de- and nonproliferation policies (25). There is evi- signed to promote domestic industries. Other ad-

25-450 0 - 84 - 13 : QL 3 182 . Nuclear Power in an Age of Uncertainty vanced countries with nuclear programs, espe- At present the number of component suppliers cially Japan, are also likely to bid successfully for appears to be declining slowly. One clear sign component manufacturing business (25,37). is the decision by suppliers not to renew the “N- stamp,” a certificate issued by the American Soci- The current backlog of NSSS manufacturing ety of Mechanical Engineers (ASME) for the man- work is scheduled for completion in 1984. All ufacturer of nuclear plant components. N-stamps U.S. vendors have taken steps to close or moth- are not specifically required by the NRC for the ball many of their manufacturing facilities, or to manufacture of safety-related nuclear-plant com- convert them for other uses. It has been estimated ponents. However, they are required by some that announced facilities closings and consolida- States, and their use certifies that certain NRC tions have already reduced by two-thirds the U.S. quality-assurance requirements have been met. capacity to supply nuclear powerplants. Some vendors are maintaining their technical capabil- The number of domestic firms holding N- ity with nuclear work for the U.S. Navy, DOE, stamps has dropped by about 15 to 20 percent or research and development (R&D) sponsored since 1979, the year of the accident at Three Mile by the Electric Power Research Institute (EPRI) Island, and the drop would probably be greater (37). if the renewal were annual instead of triennial (21). By contrast, foreign N-stamp registration has Vendors are already feeling the effects of a held steady. By the end of 1982, some 400 com- shrinkage in nuclear component suppliers on panies in the United States and Canada held which they base future standardized NSSS de- about 900 N-stamps, according to ASME. An ad- signs. Currently, each vendor purchases compo- ditional 50 companies held about 100 certificates nents from about 200 qualified nuclear suppliers. for Q-system accreditation on nuclear-grade ma- Several vendors estimate that the number of sup- terials. Overseas, about 70 companies held about pliers will drop by two-thirds in 3 tos years, leav- 100 N-stamps, and about 20 companies held ing the vendors dealing with a much higher pro- about so Q-system certificates (21). portion of sole source suppliers (37). Vendors faced with this situation are considering various Maintaining an N-stamp requires both person- responses, such as manufacturing their own nel and money. Thus, in the absence of new nu- components, encouraging less qualified suppliers clear business, many smaller companies have de- to upgrade their products and get them certified, cided they cannot justify the costs. In addition and developing new sources of foreign supply. to the $5,000 to $10,000 that must be spent for ASME certification (renewable at the same cost every 3 years), there is also the need to dedicate Nuclear Component Suppliers part of the plant and at least one or two The impact of the shrinkage in new orders is employees to the intricate paperwork that accom- panies each N-stamp component. In total, cost most dramatic on component suppliers. Some estimates for maintaining a stamp range from companies supply components used both for $25,000 to $150,000 a year (21). Suppliers say new plants and for backfit and spare parts for that no other work, including contracts for the plants in operation. These companies expect to keep their businesses going. Many companies, National Aeronautics and Space Administration (NASA) and the U.S. Navy nuclear program, re- however, supply only components for new quires such a detailed paper trail. “1 make a valve plants. Some of these produce nuclear compo- that sells for about $300” one supplier said. “If nents that are identical or very similar to non- it has an N-stamp I have to charge $4,000 for the nuclear components except for quality-control documentation. These companies can be ex- same valve. And with low volume, I suppose I should charge even more” (21). pected to maintain their nuclear supply lines. Others, however, produce very specialized So far, the reduction in N-stamps has not been nuclear components that require separate testing as rapid as the lack of new orders might suggest. and manufacturing facilities. Many of these facil- Part of the reason may be a habit of looking to ities are now closed or mothballed (37). the future that has been characteristic of the Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 183 nuclear industry since the beginning. Some sup- cept for interim survival involves “recommis- pliers evidently believe that the N-stamp imparts sioning” nuclear stations—installing some new a certain status to a nuclear supplier’s operations, components to extend their operating lives by 10 and those who must consider letting their certifi- to 20 years. Like the reactor vendors, AE firms cation expire say they would do so reluctantly. complain of reduced sources of supply for nu- “It’s a nice marketing tool,” one supplier said, clear-grade components and materials. And, like “even when you’re selling non-nuclear items. the reactor vendors, they are moving outside their And it’s good discipline for a company to have specialties to bid on nuclear services (e. g., emer- it” (21). Some suppliers and utilities report that gency planning) and rework proposals. they must persuade their subcontractors to keep Most AE firms also have large amounts of busi- the stamp. “We’re giving companies [with N- ness stemming from major construction projects stamps] our nonnuclear business, just to help other than nuclear: cogeneration, geothermal, them along, ” one utility executive said. Another and coal technologies; petrochemical plants; in- challenge is to prevent market entry by foreign dustrial process heat applications; and conven- companies. “If equipment from overseas be- tional fossil powerplants. During the 1981-83 comes standard,” one supplier said, “we’ll never recession, business in these areas was no more get that business back” (21). robust than the firms’ nuclear work. One AE ex- For many suppliers it will be almost impossi- ecutive said, “As it is now, we can’t move our ble to obtain nuclear qualification for new prod- nuclear people to nonnuclear projects just to uct lines. For some product lines, 1 to 5 years keep them in-house. There isn’t much nonnucle- wouId be needed to carry out the necessary tests. ar work around either” (21). Several firms Maintaining an older nuclear-qualified product reported they expected their nonnuclear work line alongside a newer nonnuclear product line to pick up long before their nuclear work (21). will be difficult for those suppliers with a prepon- Much of the project management and con- derance of nonnuclear business. Since nonnucle- struction skills used on other types of large con- ar business is likely to respond more quickly to struction projects are also required for nuclear an increase in general business investment of the projects. These skills will be available as long as recession than is nuclear business, there may be the AE firms have experience in major construc- pressure to drop the nuclear product lines. The tion projects. The design and project manage- existence of nuclear components in 35 gigawatts ment skills unique to nuclear projects area small (GW)* of partially completed but canceled nu- proportion of the total work force. clear plants is viewed as a further damper on the nuclear component business even though only Some firms are taking losses to keep their some of this equipment is expected to be suf- skilled nuclear people employed because they ficiently maintained and documented enough to estimate that retraining would ultimately cost be usable (see advertisement). For all these more. Architects are working as draftsmen, for reasons, there may be a far more rapid decrease example, and skilled machinists are cutting and in suppliers over the next 3 to 5 years than over stacking sheet metal. Layoffs have not been nec- the past 3 years, possibly down to a third of the essary, one AE executive said, because employ- present number (37). ees are retiring early or quitting to move to fields with more growth potential and less regulation, Architect= Engineering Firms such as military R&D (21). AE firms have substantial work for the next few years finishing the plants under construction, in- The Impact on Nuclear stalling backfits and dealing with special problems Plant Operation ———.such as steam generators. One promising con- “One GW = 1 GWe = 1,000 MWe (1 ,000,000 kWe) or slightly less The halt in nuclear plant orders and uncertain than the typical large nuclear powerplant of 1,100 to 1,300 MW. prospects for new orders have had discernible 184 Ž Nuclear power in an Age of Uncertainty

ONE OF THE WORLD’S LARGEST AUCTIONS!!! 50 MILLION DOLLAR ACQUISITION COST OF SURPLUS MATERIALS, SUPPLIES CONSTRUCTION EQUIPMENT Beginning April 4 and Continuing Daily 10:00 A.M. to 6:00 P.M. (Each Day) Phipps Bend Nuclear Plant Site Surgoinsville (Kingsport), Tennessee

By order of the Tennessee Vahey Authofrty. Mllwr ~ Miller Auctioneers, Inc WIII sell at pubkc auction unused ma!erials, unused suppltes and used construction aqwpmenl purchased tor the mnstrucmon of the Pmpps Bend Nuclear PlanL Everything WIII posmvely sell to the hghest bWer. no mmtmum or resewatlor, Auction tobe conducteo at the Phtpps Bend Nuclear Plant Srte located near Surgomsvrlle (Kmgspoflj, Tennessee The auction will be m an audrtonum srt down fashion Blddem are wwrted and urged to respect the propefty to be sold before bidding as property WII! be sold AS IS, Materials, supphes ano equlprnenl may be respected bagmnmg March 21 through APril 1 trom8-OOA M to330 P.M Monday mrough Friday All pmspecove purchasers will be mquwc! 10 regster and sign a holdihamless agreement prior to emerrng sale srte Ouantrhes shown Wow are approximate, shoukl be verrfted by mspechon

PIPE” 392,491 of ductile concrete, corrugated culven, SPA300VH 1,550CFM alr compressors. each pwered cast Iron carbon steel coamct, stainless steel. PVC by 400HP etectnc motor. m 40’ x 130’ metal buddq sewer, alummum, galvanized 8 clay Assorted sizes with (2) hols: 8 trolieys (4) 5HP au compressors trom %“ to 66” VALVES (562, 1- to 12C’ motor & at’ OVERHEAD CRANES (9; 250-ton to 1 %-ton e}ectnc operalec butterfly globe gale. swing checu and manual Overheac travellqg bridge aanes (4) 5-ton to 1 ~n-ton operalec DUttef’flY PUMPS (128 I centrrtqal booster electrlc hots: OTHER RELATED ITEMS Enwrovac vemca’ f~rome, transfer, teed & vacuum, 200GfW tc vacuum sewage collection syslem COolmg tower 125.000GPM, 3450HP tc 5HP electrlc motors FANS strainers Assortec sizes of meta’ bwlomgs Bl~epnnl (1051 Zum clarage fans 500SCFM tc 39,90CSCFM mac~w 8 copier Large amoun: of ha~c toots pipe AIR HAN~LING UNITS (151) Govemar. 26,000SCFM benoers, pipe threaders welchrm sucrpilers, rmefntmos. to 1500SCFM, vertca~ 8 hortontal. sirge zone WATER scaffoldincj CON STRUCTIOk EOUIPMENT C:s CHILLERS (8) Carrier 400-ton, (2) Trane 1500-ton, (4) Johnson 12-varr3 concrete batch plant ADpcQ 6-yard ~rane 180-ton (2,1 Trane 2-s!age absorphon cold gen- concrete baic~, plant Concrete “pumps Dumtiete erators AIR CONDITIONING UNITS (94) Carrier trucks Cement bulk trahers Concrete terms CraAe: WELDERS (239) 30Gamp to 1500-amp. (5) Panjms iractors Scrapers Loadem Cranes Hydraulic back- automatic honzomal & verma’ welders SWfTCti GEAR hoes & cranes Tower cranes End dwnps” Compactors (17) 460-vofl swrtch gear & transformers (7) 6900-voIT Trenchers Alr compressors Tru~ & trailers. swrtch gear 8 transformers (13) 480-volt motor control APPROXIMATE TWO BILLION BL LAR~ worth d centers ELECTRICAL 2940,646’ of copper cable, rnatene}s and equipment Smilar to above, plus other I AWG to 20AWG, 600-volt 12.771 M tray (7) rtems, assmaled Wf7 lame a2nstructum ormects are Neutral grounding resistors Assomd sizes of ccm- avakbie tor sale For I nton%hon cnrtfacI tiA, itWEST. dulets (4) G.E. 250KVA generators TANKS (I 1 j MENT RECOVERY PROJECT at 615632-7750 78,000-gallon to 15.000-gallon stainless steel steel /3 alumlnum tanks AIR COMPRESSORS Statlonao Contact Auchoneers for Free Oescnptwe Brochure compressor plan: ccwmstmg ot, (5! lngersoll-Rana hStlMJ terms and prope~ to be sold at auctton ~-- —.——.— —---.-—------: .%?nC to Wkr L MUleT Auchon,orx hbc. 2525 Rulgrlw aoutevarc For! Wonh Texas 76116 UC017 7324ese Ye- I WOUK! ke a oexcnowv bwc%re w. TVA ● uclIoc WS4W pnni NAME – — 2525 Fhdfynar Boulevard COMPANY — 76116 Fon Worth. Texas ADDRESS —- ——. _.—.— -— — 017.7324KM 10KI 7SS440 C!TY —- -. -ATE -~ TN LIC ho 4?3 L.. ---.. -—-- —------.. —.-— .—.— ----- Photo credit: Engineering News Record Advertisement for nuclear component auction appearing in Engineering News Record, Feb. 21, 1983 Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 185 effects on the utilities’ experience in keeping their consultant said. “We’ve got to try for anything existing nuclear plants staffed and maintained. out there, just to survive” (2]). The effects are most noticeable in two areas: obtaining component parts and services, and filling Skill Shortages.–Utilities are also having trou- certain key jobs. ble recruiting certain categories of employees and this may get worse in the future. According to Component Costs and Delays. -With the a personnel study by the Institute of Nuclear Pow- decrease in nuclear component suppliers de- er Operations (INPO), the overall vacancy rate scribed above, utilities report an increase in the was 12.5 percent of all nuclear-related positions. number of sole source suppliers and a resulting However, for nuclear and reactor engineers, for upward pressure on prices. One utility reported radiation protection engineers, and health physi- that sole source suppliers received 40 to 50 per- cists (technical specialists in health effects of radia- cent of 1982 contract dollars. A more typical tion), the vacancy rate was more than 20 percent range reported by utilities was 25 to 30 percent, (see app. table 7A). The average turnover rate for an increase from 15 to 20 percent a decade ago engineers is almost 7 percent a year, and, for most (21). categories of engineers, quitting their jobs in util- In a few cases, utilities report that prices of serv- ities means leaving the industry altogether (16). ices and components are falling because of in- For the nuclear utilities as a group, an estimated creased competition. Generally, however, prices 6,000 additional engineers will be needed be- are expected to rise partly because of lack of com- tween now and 1991. Almost 5,000 of these will petitive pressures on the increased number of be needed to replace those that leave the indus- sole source suppliers and partly because the fixed try (see fig. 35). About 3,000 technical level health cost of nuclear quality assurance must be spread physicists will be needed, about 2,000 of these over dwindling sales. for replacement (18). Delays are also expected to be more of a prob- Despite the availability of ample jobs for nucle- lem for similar reasons. With less nuclear work ar specialists, degrees and enrollment in nuclear- to do, suppliers are more likely to arrange pro- related fields are stable or declining (see fig. 36 duction schedules to use qualified craftsmen and for nuclear engineering degrees). There is some their ‘special machinery only when a number of evidence that students are being discouraged orders are in hand, postponing work on some from enrolling in programs leading to employ- projects for months. Or they could require pre- ment in nuclear power by a perception that the miums for deadlines that are more convenient industry is declining and by parental concern and for the utilities. “He’d get the part for you, when some peer pressure against nuclear power ca- you wanted it,” one utility executive said of a sup- reers. A recent DOE study of personnel for the plier, “but you’d have to pay for a whole shift nuclear industry contrasted steadily increasing to go on overtime” (21). Suppliers report that util- enrollment in medical radiation physics programs ities are placing more “unpriced” orders, for with declining enrollments in technically similar which the supplier alone sets the costs, and health physics and radiobiology programs aimed choosing other than the lowest bid to get the at work in the field of nuclear electricity genera- schedule and quality they need (21). tion (9). In addition to possible increased prices and de- INPO which has developed demanding train- lays, utilities are also experiencing some greater ing requirements for utility personnel has also confusion in the bidding process for rework and taken some modest steps to help with recruiting nuclear services as more and more firms attempt by setting up a fund for graduate nuclear train- to diversify in the face of falling profits. “Anything ing (see ch. 5). Individual utilities have also taken in an RFP [request for proposal], that’s at all re- steps to fund nuclear programs at local communi- lated to our business, we’ll bid on it,” one nuclear ty colleges. More, however, may be needed if

186 ● Nuclear Power in an Age of Uncertainty

Figure 35.–Estimates of Additional Manpower Requirements for the Nuclear Power Industry, 1982-91

I Professionals I Technicians I Operators I Skilled craft I

SOURCE: Ruth C. Johnson, Manpower Requirements in the Nuclear Power Industry, 1982-91, September 1982, ORAU-205.

current turnover and recruiting trends continue, for the nuclear contracts to roll in. I maybe do- or worsen. ing something else in the meantime” (21). Some effects of a hiatus in new plant orders are inevitable even if the optimistic projection of nu- The Impact on Future New clear orders after 3 to 5 years occurs. There are Construction of Nuclear Plants likely to be fewer reactor manufacturers (two or There are several ironies in the current situa- three rather than four) and some initial delays for tion of much of the nuclear industry. Many com- vendors in securing all the necessary suppliers panies are sustaining themselves on backfits and and encouraging their renewal of N-stamps. Even rework, which over time increases the cost of nu- under this optimistic projection, foreign sources would probably be used for some specific areas clear power to utilities and consumers alike and makes it less likely that utilities will place orders of supply. soon for more nuclear powerplants. Some com- Component suppliers estimate a delay of 1 or panies are maintaining their nuclear business 2 years in obtaining N-stamp qualifications and because the rest of their business has not yet been additional delays once they are operating be- affected by the improvement in the economy. As cause of unfamiliarity with support services. business investment picks up, the rate of com- “Right now,” one said, “my people know just panies leaving the nuclear business may accel- who to call for an interpretation of the regula- erate. The recovery is likely to create work and tions. They know which seismic stress labs are jobs in most other industries before utilities see the best. If we had to start over [several years their reserve margins shrinking and begin order- hence], a question that takes an afternoon on the ing again (see ch. 3). As one supplier said, “If the phone to answer today would take three to four economy revives, I’m not sure I can wait around months to answer” (21). Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 187

Figure 36.—Nuclear Engineer Degrees: Foreign Nationals and U.S. Citizens

Key 800

700 B.S.

600

500

400

300

200

100

( 1973 1974 1975 1976 1977 1978 1979 1980 Year SOURCE: A Study of the Adequacy of Personnel for the U.S. Nuclear Industry (Washington, D. C.: Department of Energy, November 1981).

Despite these difficulties, it is probable that vide designs for initial plants that were developed even after a hiatus, a realistic 8-year project overseas in joint ventures with foreign countries, schedule (under conditions assuming no hiatus) perhaps even as licensees of foreign companies would be delayed only about a year, and perhaps (37). With a longer passage of time, there will be less, if utilities were to freely allow overseas pur- more critical areas with no qualified U.S. supplier chasing. If utilities insisted on U.S. sources for all and more dependence on overseas component or most components, the delay could be longer suppliers. Since licensing and quality-control re- (37). quirements for foreign nuclear programs are quite different it could prove time-consuming and dif- With a hiatus of 10 years or longer, there will ficult to obtain nuclear qualification and licens- be a much bigger problem in new plant construc- ing for the design and components of the initial tion. Unless there have been at least some over- plant (37). Under these circumstances it is unlike- seas orders, the reactor vendors are likely to pro- ly that there would be more than two U.S. ven- 188 ● Nuclear Power in an Age of Uncertainty dors. It is also conceivable that foreign companies power are predictable. It is quite possible that might bid directly for the design and supply of foreign vendors might offer turnkey plants in the the initial units (37). United States. It is perhaps more likely that U.S. vendors may attempt to form consortia with for- After a period of 10 years or more without nu- eign designers and component suppliers to of- clear plant construction some skills would still be fer turnkey plants. available from other industries. Control and instrumentation designers and workers would prob- Conclusion.–As of 1983 the nuclear industry ably be available from the electronics and aero- is still intact although somewhat reduced from space industries. Construction contractors expect 3 to 4 years ago. The industry probably would that semiskilled construction and maintenance survive a short hiatus of 3 to 5 years in new orders workers would also be available. Of the construc- with only some increase in costs and delays in tion skills, a shortage of welders with nuclear cer- obtaining some components from U.S. sources tification might pose the greatest staffing prob- and perhaps little or no increase in costs and lem. But one AE executive said that the biggest delays if foreign component sources are used. difficulty “would be administrative;” learning Predicting the consequences of a hiatus of 10 again to control “the thousands of decisions and years or more is more difficult but it is unlikely tasks” needed to construct and test a nuclear to mean the end of the nuclear option in the plant (21). United States. If vigorous, economical, and safe Still another possibility, a very low volume of nuclear programs survive in several foreign coun- orders beginning in the late 1980’s, is the situa- tries, they are likely to provide designs and some tion most likely to encourage evolution in the nu- components for the initial plants of a new round clear industry structure to permit the necessary of nuclear construction if one occurs. (The sur- economies of scale in design and construction vival of foreign nuclear programs is the subject management experience (37). In this situation, of the next section of this chapter.) Many U.S. it is likely that utilities or others will form regional businesses would still supply nuclear components or national nuclear generating companies to ob- because they supply very similar nonnuclear tain the economics of scale from multi-unit sites components. Many others probably would get and standardized construction. This is also the recertified to supply nuclear components. U.S. situation that is likely to encourage “turnkey” vendors of NSSS will still have large nuclear serv- construction, a practice used for the earliest ice and fuel-loading businesses and probably plants constructed in the late I %0’s and still used some foreign nuclear work as well. They are likely for some exports of nuclear plants (e.g., a plant to be active in any consortia or joint ventures in- being constructed by Westinghouse in the Philip- volving foreign sales of nuclear powerplants in pines). In turnkey construction, a company or the United States. consortium, often headed by an NSSS vendor, Under some circumstances, AE firms could end offers to construct and warranty an entire nuclear up with less nuclear business after a long hiatus, island, * or even a complete nuclear plant, for a depending on what restructuring might occur in fixed price, ready for the operating utility to “turn the industry. A shift to turnkey construction of the key” and operate the plant. Such fixed-price entire plants or the formation of a few generating agreements may be the only way for vendors to companies with their own design and construc- convince utilities that their costs for nuclear tion management staffs would sharply reduce the *Nuclear island refers to all the equipment that directly or indi- role of the architect-engineer. The number of util- rectly affects the safety of nuclear operations. In addition to the ities directly involved in nuclear construction reactor vessel itself and the primary cooling system it usually includes the secondary cooling system and the steam generators (in would also be drastically reduced under such cir- a pressurized water reactor). cumstances. Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 189

THE PROSPECTS FOR NUCLEAR POWER OUTSIDE THE UNITED STATES Many of the problems that have threatened the the OECD countries forecast for the same year nuclear power industry in the United States have had fallen by 75 percent, to only 455 GW of nu- also weakened nuclear power prospects abroad. clear power (table 24). The reasons for this dras- However, a few countries–with somewhat differ- tic reduction in expected nuclear capacity are fa- ent institutional structures for producing electrici- miliar to anyone acquainted with the U.S. nuclear ty and stronger motivation to avoid dependence industry: slower-than-expected electricity de- on energy imports-may be able to nurture their mand growth, high interest rates that increased nuclear industries to survive the 1980’s in stronger the cost of capital for nuclear powerplants and condition than the U.S. industry. This section stronger-than-expected public opposition in surveys the highlights of the foreign nuclear ex- many countries. perience–economic, technical, and political– Just as in the United States, the rate of growth and points out a few aspects of foreign experience of electricity demand slowed from the 1960’s to that provide a perspective on U.S. experience. the 1970’s in France, West Germany, and the The section also assesses the likely competitive United Kingdom (except for demand from French situation of the U.S. industry vis a vis its com- households) (10) (see fig. 37). petitors abroad. Given the slower-than-expected growth rates The Economic Context for in electricity demand, many countries are now Nuclear Power lowering their forecast growth rates for 1990 and 2000 and finding themselves with adequate gen- Worldwide forecasts of the future role of nu- erating capacity, West Germany expects to need clear power have experienced the same boom new powerplants only if oil and gas capacity is and bust cycles as have U.S. forecasts. In 1975, to be replaced (24). A Government commission OECD* countries forecast a total of 2,079 GW in France estimated that completion of the pres- of nuclear power by the year 2000. As of 1982 ent construction program should provide most *OECD means Organization for Economic Cooperation and De- of the electricity forecast to be needed before velopment, a Paris-based organization of industrialized countries. 2000. As of mid-1983, the Government had not

Table 24.–Forecasts of Installed Nuclear Capacity in OECD Countries (GW)

Installed public generating Forecasts for the year 2000 Nuclear capacity installed capacity of all types Regions and countries OECD, 1975 INFCE, 1980a OECD, 1982a 1980 1990b 1979C Western Europe: 798 341 214 43.8 126.6 371 France ...... 170 96 86 16.1 56.0 47 Germany...... 134 63 34 8.6 23.5 66 United Kingdom...... 115 33 31 6.5 11.1 69 North America: 1,115 384 173 60.2 131.2 671 Canada ...... 115 59 22 5.2 15.0 72 United States ...... 1,000 325 151 55.0 116.2 599 West Pacific: 166 130 68 15.8 28.0 153 Japan ...... 157 130 68 15.8 28.0 126 Australia/New Zealand . . . 9 0 0 0.0 0.0 27 OECD total 2,079 855 455 119.8 285.8 1,195 aFlgu~~S are the averages of high and low estimates. bcapaclty installed or due to be commissioned by Jan 1, 1990 Source SpRIJ turbine 9enerator data bank csource IJnlted Nat Ions, 1981.

SOURCE Mans Lonnroth, “Nuclear Energy In Western Europe,” a paper for the East-West Center, Honolulu Conference on Nuclear Electric Power In the Asia Paclf!c Region, January 1983.

190 ● Nuclear Power in an Age of Uncertainty

Figure 37.–Average Annual increase in Electricity broader philosophical concerns about the future Demand: France, West Germany, and United Kingdom direction of a society based on high technology which requires extensive central control (24,30). In a few countries, public opposition to nuclear power has become directly involved in political and administrative decisions that affect the growth of nuclear power. The most dramatic of these is Sweden. in a referendum held in March 1980, 57 percent of the public voted that 3 plants under construction should be completed but the total of 12 reactors then in existence should be operated only until economical means are found to replace them and should not be replaced with more nuclear powerplants if there is any feasible alternative. Parliament subsequently adopted

Industry House- Industry House- Industry House- this position as official Government policy. A sim- holds holds holds ilar referendum, which halted nuclear power-

1980-70 1970-80 SOURCE: OECD. yet accepted the report because of the drastic implications for the future of the nuclear power industry. Other conditions, familiar to observers of the U.S. nuclear industry, were described by Mans Lonnroth, and William Walker in a 1979 paper, The Viability of the Civil Nuclear Industry (26). Although the rapid increases in oil prices after 1973 made nuclear power appear relatively less expensive over the long run, it made it harder to finance in the short run because the resulting high rates of inflation increased interest rates. In those countries where the government approves electricity rates it became harder to get political support for rate increases to compensate for inflation. Inflation and the several recessions of the decade also put pressure on governments providing financing to electric utilities to restrict the extent of their support (26).

Public Acceptance

In most European countries and some other Photo credit: OTA Staff countries, there has been considerable public op- In West Germany, construction of nuclear plants position to nuclear power. This sometimes arises has been stopped in the courts while in France the more centralized decision-making system has kept from specific concerns about plant siting, design nuclear construction going without delays, despite and management, and sometimes from much public opposition. Ch. 7—Survival of the Nuclear Industry in the United States and Abroad . 191

plant construction in Austria, was passed follow- Typical nuclear plants in several foreign coun- ing a period of widely publicized debates among tries can also be constructed more cheaply than nuclear experts. In West Germany, citizens can typical U.S. plants. Based on information ob- sue in State courts to stop the construction of tained in an NRC survey of foreign licensing prac- powerplants. Four plants were stopped in the tices, U.S. costs are comparable with those of mid-1970’s and brought nuclear construction in Sweden and West Germany but about 30 per- Germany to a virtual halt. Subsequent extensive cent higher than in Japan and about 80 percent Parliamentary Commissions have recommended higher than in France (38). In France, a nuclear caution but not a halt. In the United Kingdom, plant can be constructed for about half the man- a public inquiry was conducted throughout 1983 hours/kW required for a nuclear plant in the to consider the general adoption of a modified United States. The other three countries use Westinghouse light water reactor (LWR) design. about 30 percent fewer man-hours/kW than in (This is the Sizewell B design, discussed in ch. the United States (fig. 38). Most of the savings oc- 4.) This technology would be in addition to the curs in two categories: the nuclear increment existing series of advanced gas-cooled reactors over man-hours needed for constructing a non- that until now have formed the basis of the nuclear powerplant, and the engineering man- United Kingdom’s nuclear technology. hours used during construction, which is almost 10 times higher in the United States than in any In France, there have been infrequent Parlia- other country (38) (see ch. 3). mentary discussions of policy with respect to nuclear power; otherwise decision making has Performance of U.S. reactors falls at, or slight- been treated as a technical and administrative ly below, average in cumulative load factors for rather than a political matter (31). Public opposi- world reactors* (see table 26). Several countries, tion has been expressed in anti-nuclear demon- such as Finland, Switzerland, Belgium, and the strations, in demonstrations at particular sites, and Netherlands, with only one to four reactors, have in the formation of ecology parties which have very high average cumulative load factors. Can- challenged candidates in local and regional elec- ada’s nine heavy water CANDU reactors have the tions. None has had any substantial impact on highest load factors of all, averaging over 80 per- the French nuclear program. In part, this appears cent (see ch. 4). In other countries, however, with to be because public opinion surveys have shown large numbers of reactors, average reactor per- increasing support for nuclear power (36). formance does not differ substantially from the United States. At the same time, a smaller share Foreign Technical Experience: of U.S. reactors can be found among the top- Plant Construction and Reliability ranking reactors. Among the top 25 percent are 96 percent of Canada’s reactors, 32 percent of Many foreign countries have experienced de- Sweden’s reactors, 44 percent of West Germany’s lays in building nuclear powerplants similar to reactors, 16 percent of Japan’s reactors, but only those experienced in the United States, but some 10 percent of U.S. reactors. Many U.S. reactors have built all their plants as fast or faster than any can be found at the bottom of the list; 27 per- U.S. plant (see table 25). In France, the slowest cent of U.S. reactors rank in the lowest 25 per- plants have taken 7 or 8 years from reactor order cent of world reactors. to commercial operation, while the fastest plants take 5 to 6 years (14). In Japan, slower plants have Licensing and Quality Control taken 9 or 10 years while faster plants have also been built in 5 or 6 years. In Japan most of the West Germany has as complex a licensing delay occurs at the site-approval stage, prior to process as the United States. Licensing of nuclear the start of construction. For the other countries plants is governed by seven State (Lander) licens- with at least several nuclear plants the fastest construction times are comparable (6 to 8 years) with *World reactors excluding reactors in Eastern Europe, the the fastest construction schedules in the United U. S. S. R., and several third-world countries for which cumulative load factor data is not available (see appendix table 7B for listing States. of nuclear capacity in all countries). 192 . Nuclear Power in an Age of Uncertainty

Table 25.—Sample Construction Times for Nuclear Plants in Various Countries Faster Slower Date of com Years Sine; Date of com Years since operation reactor order operation reactor order France: St. Laurent B1, B2 ...... 1981 — — Gravelines C5 ...... 1984 — — Dampierre 2,3,4 ...... — 1981 Cattenom 2 ...... — — 1986 Germany: KKU, Unterweser...... 1978 7 — KKI-1, Ohu ...... 1977 6 — KKG, Grafenrheinfeld . . . 1981 6 — KKK, Krummel ...... — — 1983 11 KBR, Brokdorf ...... — — 1987 12 Italy: Caorso ...... 1978 8 — Montalto 1 ...... — — 1986 12 Japan: Ikata 2...... 1982 5 — Fukushima 11-2 ...... 1984 6 — Fukushima 11-3 ...... 1985 5 — Genkai 2 ...... 1981 5 — Ohi 1,2 ...... — — 1979 9 Fukushima 11-1 ...... — — 1982 10 Sweden: Barseback 1 ...... 1975 6 — Barseback 2...... 1977 5 — Ringhals 3,4 ...... — — 1982 10 Forsmark 1 ...... — — 1980 9 Spain: Almaraz 1 ...... 1981 9 — Valdecaballeros 2 ...... — — 1988 13 Lemoniz 1 ...... — — 1983 11 Switzerland: Goesgen...... 1979 6 — Leibstadt ...... — — 1984 14 Taiwan: Kuosheng 1 ...... 1981 8 — Chin-Shan 1 ...... 1979 9 — Maansham 2 ...... — — 1985 11 Canada: Pickering 1 ...... 1971 6 — — Bruce 5 ...... 1983 7 — Gentilly 2 ...... — — 1983 9 Darlington 1 ...... — — 1989 12 United Kingdom: Torness 1, 2 ...... 1986 8 — — Heysham 3,4 ...... 1986 7 — Dungeness B-1, B-2. . . . . — — 1982 17 Heysham 1,2 ...... — — 1983 13 United Statesa: St. Lucie 2 ...... 1983 6 — Hatch 2...... 1979 7 — — Diablo Canyon 1 ...... — — 1984 16 Midland 1 ...... — — 1985 13 %arting time is construction permit rather then reactor order. SOURCES: December 19S1 Atomic Induatrlal Forum List of Nuclear PowerPlants Outside the United States; ch. 5, table for U.S. nuciear plants.

Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 193

Photo credit: Electricite de France

Each of the four identical 925-MW units at Tricastin in France averaged 6.5 years construction time. Work force and engineering experience with the standard design and with the site accounted for the short construction time 194 . Nuclear Power in an Age of Uncertainty

Figure 38.-Engineering and Construction Man-Hours is clearly separated from the consideration of safe- per Magawatt of Capacity (typical nuclear plants) ty issues. In all three countries the only oppor- 19,160 tunity for public intervention occurs at the earliest site-approval stage. This process can take 2 to 3 years in Japan (17). In France, the site-approval process can take place as much as 8 years before application for a construction license. In fact, sites have been approved for all nuclear construction until 1990 (36). In England, there is an option to hold a public inquiry at this early stage. The public inquiry for the Sizewell B covers a broad range of issues concerning development of pressurized water reactors (PWRs) in England (32). Once site approval has been obtained, the licensing process in all three countries involves a technical exchange of information between licensing entity and licensee. In Japan and the United Kingdom, there is only one license (a construction license) but there is a series of technical requirements—for tests, safety analyses, and operating procedures—that must be met before the plant is allowed to operate. In practice, this amounts to a multistage licensing procedure. In France, there is an additional operating license, but the process is similar. The plant must pass United West Sweden Japan France a series of tests and have proposed operating pro- States Germany Engineering: cedures approved before an operating license can be issued (36). Construction: Design & Normal analysis Overall Nuclear Development There are significant differences among coun- SOURCE: J. D. Stevenaon end F. A. Thomee, Selected Review of Forwign Llcemw tries in the probable future course of nuclear Ing Practlcea for Nuclear Power P/8nt8, NUREG/CR-2e64. Theae estimates were made from estlmatea of staffing pattema, project dura- development. Table 27 classifies countries with tion, end other Information obtained from Intervlewa with about 50 pao. ple in the Unltad States end foreign countrlea. The Intewlews are listed actual or possible nuclear power programs into In the front of the raport. six categories based on the prospects for further nuclear construction. ing agencies each of which in turn depends on Five countries–France, Japan, Canada, Tai- one of seven independent inspection agencies wan, and Korea—have Government-supported (TUV) for technical licensing requirements. In the programs of constructing and operating nuclear 1970’s, West German licensing and quality-con- plants. Three—France, Japan, and Canada–have trol requirements imposed on nuclear plants in- emphasized standardization to minimize con- creased in much the same fashion as they did in struction cost and increase reliability. All three the United States (19). Engineering work tripled, have ambitions for major export programs in the quality-assurance documentation quintupled, 1990’s when demand picks up again, although and there was a 50-percent increase in the time Canada has had several setbacks in its efforts to required for component manufacture. make the CANDU heavy water reactor (HWR) In France, Japan, and the United Kingdom, the a viable export, when sales failed to go forward consideration of siting and environmental issues in Mexico, Korea, and Rumania. In addition, Can- Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 195

Table 26.—Operating Performance by Country (average reactor cumulative load factor, to end of 1933)

Reactor types PWRs BWRs Magnoxs Heavy water HTRs &AGRs Load Number of Load Number of Load Number of Load Number of Load Number of

Country factor, 0/0 units factor, % units factor, 0/0 units factor, % units factor, 0/0 units United States ...... 54.8 48 56.1 23 — — — — 17.9 1 Argentina ...... — — — — — — 75.3 1 — Belgium ...... 75.6 3 — — — — — — — Brazil...... 2.3 1 — — — — — — Canada ...... — — — — — — 80.1 9 — — Finland ...... 75.0 2 62.7 2 — France ...... 56.0 23 — — — Germany, West ...... 71.2 7 44.1 4 — Great Britain ...... — — — — 58.5 18 - - 40.6– 4 India ...... — — 49.1 2 — 33.4 2 — Italy...... 44.1 1 30.3 1 59.3 1 — — — — Japan ...... 56.2 11 61.0 12 62.6 1 49.8 1 — — Netherlands ...... 77.9 1 — — — — — — — South Korea ...... 55.2 1 — — — — — — Spain ...... 36.8 2 62.0 1 72.0 1 – — — — Sweden...... 38.3 2 65.1 7 — — — — — Switzerland ...... 75.1 3 80.5 1 — — — — — Taiwan ...... — 51.9 4 — — — — — Yugoslavia ...... 53.1 1 — — — — — — — World ...... 56.7 106 56.6 57 57.8 26 76.0 13 38.0 5 Notes: 1. All plants operating less than a year as of July 1983 were excluded from this calculation. 2. The USSR and several countries in Eastern Europe with substantial numbers of nuclear plants are not listed, (See appendix table 7B) 3. Graphite-water and fast breeder reactors are not included, 4. Plant load factors were weighted by plant rated capacity to get country and reactor-type averages. SOURCE: Nuclear Engineering International, “Nuclear Station Achievement 1983,” October 1983.

Table 27.—Categories of Foreign Nuclear Programs in Western Europe and the Asia-Pacific Region

Category Countries in category I. More nuclear plants planned and under- France, Japan, Taiwan, construction backed by Government policy Canada, Korea Il. More nuclear plants planned but may be United Kingdom, West stopped by public opposition Germany, Italy Ill. More nuclear plants planned but delayed Spain, Yugoslavia, Greece due to economic difficulties Portugal, Turkey IV. Nuclear plants in existence with Sweden, Switzerland de facto and de jure halt on further construction v. Nuclear plants begun but indefinitely halted Philippines, Indonesia VI. Government policy prohibits nuclear plants Australia, Austria, Norway Ireland, Denmark SOURCES: Off Ice of Technology Assessment categorization based on papers presented at conference on Nuclear-Electrlc Power In the Asia Pacific Region Jan 24-28, 1982; and Mans L6nnroth and William Walker Nuc/ear Power Sfrugg/es. /ndustr/a/ Corrrpet/t/orr and Pro//ferat/on Corrtro/, George Allen and Unwln, London, 1983.

ada has entered into negotiations with U.S. util- in France and three nuclear-owning utilities in ities to build plants whose output is primarily in- Canada, the institutional coordination for an ef- tended for the U.S. market (20,24). Several char- fective standardization program is fairly easy. In acteristics of the nuclear industry in Canada, Japan, the nine utilities are investor-owned and France, and Japan make it far easier to maintain depend on private financing. However, planning momentum in the nuclear industry than it is in for nuclear power development takes place with- the United States. In Canada and France, the in the overall framework of private-public coop- nuclear-using utilities are Government-owned eration established by the Ministry of interna- (see table 28). With only one nationalized utility tional Trade and Industry (MITI). Of these coun- 196 . Nuclear Power in an Age of Uncertainty

Table 28.—Structure of Electric Utility in Main Supplier Countries and the Distribution of Authority Over Key Decisions Influencing Financial Health of Utilities

Ownership Choice of External financing Rate of utilities generating mix of investments regulation Comments United States: Large number of In principle utility, Capital market (bonds, Public utility Fragmented authority utilities, mainly but, State governments stocks). Bonds rated by commissions in each over utility financial privately owned tend de facto to influence independent rating State health. Role of Federal decisions agencies; State finance Government very weak for public utilities France: State owned, Government, at National budget, National Government National Government EdF recommendation from international capital approves rate controls financial health EdF market (e.g. US) for change bonds West Germany: Several utilities, Utility, but regional govern- Capital market, Federal Government Local, regional, and Federal the larger ones hav- ment makes final regional government has to approve rate governments all ing mixed State licensing decisions changes influence, and have (land) and private interest in, financial ownership health of utilities due to ownership and rating responsibilities Canada: Mainly provincially Utility recommendation, Budget of provincial Provincial government Provincial government owned (Ontario, provincial government government, capital approves rate controls financial health Quebec, etc.) final decision market (bonds) changes United Kingdom: State owned, National Government, after Budget of National National Government National Government CEGB and SSEB recommendation from Government approves rate controls financial health generating boards changes Japan: Private investor Safety assessments carried Mixed (bonds, Ministry of Financial health generally owned (9) out by central govern- equity ... ) International Trade good. Utilities with ment, environmental and Industry (MITI) nuclear investments have assessment by local developed substantial government. Final in-house technical authority rests with capabilities Prime Minister Sweden: State owned (50)%) National Government National budget (for Almost none. Financial health generally Privately final Iicensor after pro. State-owned utility), Electricity producers good, due to large share owned (35%) posals from utilities bonds for utilities not allowed to compete of inflation resistant Local owned by the state for large-scale hydro. Competition be- government (15°10) customers and tween utilities said to distributors hold rates down SOURCE. Mans Lonnroth and William Walker, The V~ab~l/fy of the CIvI/ Nuclear Industry, a working paper for the International Consultative Group on Nuclear Energy Publ}shed by the Rockefeller Foundation and the Royal Institute of International Affairs In 1979. tries, France and Japan have substantial domestic in West Germany, a group, called a “convoy,” markets for nuclear powerplants and are thus in of six similar nuclear powerplants was started stronger position to sustain nuclear industries. through licensing review in the spring of 1982 (7). The United Kingdom, West Germany, and Italy There are indications that political and public op- have nuclear powerplants underway, but these position may have peaked although there have been no changes in legal structure (24). Construc- could still be stopped by public opposition. Fur- tion stoppage is still a possibility, however, be- ther construction of LWRs in the United Kingdom cause citizens retain the legal right to sue to stop will depend on the outcome of the Sizewell B Inquiry. The case for new construction is greatly the plants and the courts are independent. weakened by the very low electricity demand In Italy, two nuclear plants are under construc- growth in the United Kingdom over the decade tion in addition to four in operation. Local oppo- of the 1970’s (32). sition to the two under construction caused ex- Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 197 tensive delays. The State electricity corporation, technical agencies (TUVs) for a series of power- EN EL, will begin a vigorous campaign of local plants to be ordered and licensed in groups of public education at each site of four plants pro- five or six or “convoys” over the next 10 to 12 posed to be built by 1990. Success of such ef- years (19). The basic process is modeled on the forts in avoiding site delays is still unknown; successful French program. Each series would similar site public relations efforts in France were have a standard design. Improvements on the targets of protestor bombings. design would be saved for a subsequent series. Of the countries with nuclear plants in exist- The various parties to the construction and li- ence and nearing completion, but with a hold censing of powerplants have agreed to several on further construction, Sweden has the most ex- changes in procedure designed to reduce the cost plicit moratorium. Switzerland also appears to and delays in plant construction. Documentation have halted further construction beyond plants requirements have been simplified. Technical scheduled to begin commercial operation in reviews of different aspects of the convoy plants 1985. Several countries have avoided nuclear have been assigned each to a separate technical power in developing a national energy policy. review agency. KWU will make maximum use Despite possessing some of the world’s richest of computer-assisted design and develop a con- uranium supplies, Australia is dedicated to bas- voy management system that controls the critical ing its electricity generation on coal. Austria and features of the design of each plant. Norway have decided against building nuclear The legal framework has not changed in West plants but are able to use abundant hydropower. Germany in order to facilitate the convoy con- New Zealand has surplus electricity from hydro- cept. As KWU concluded in its report on the con- power and coal; and Ireland will increase coal- cept, “without the broad consensus of all the par- fired electricity (12). ties involved (namely the customers, licensing authorities, authorized inspection agencies, and Implications for the U.S. manufacturers) the concept will remain nothing Nuclear Industry more than a collection of odds and ends, with every chance of real success denied it” (19). Implications for the U.S. industry can be drawn from the experience of other countries in devel- Although the institutions are different, the prob- oping nuclear power. Probably the most power- lems of getting a large number of different organi- ful lessons will come from countries more like zations to work together to streamline an increas- ourselves. ingly cumbersome process is similar to what would have to be accomplished in the United The West German “Convoy” Experiment.– States. Success of the West German effort would The German licensing system for nuclear power demonstrate that such an effort is possible. appears every bit as cumbersome as the U.S. system; in fact it involves even more regulatory man- Backfits.—ln both West Germany and France years per regulated megawatt (38). With seven there are policies that restrict backfits. In West State licensing authorities, assisted by seven dif- Germany, utilities are supposed to be compen- ferent independent engineering review organiza- sated for the costs of implementing any backfits tions (TUVs), the West German system adds State- after the State licensing authority has given its ap- to-State inconsistency to the several stages of proval (38). In practice, this provision has not hearings and the independent court reviews of been carried out very often and has not pre- the U.S. system. vented the escalation in required engineering man-hours described above. In an effort to halt the cycle of delays, requirements for rework, and increasing engineering In France, backfits are restricted once each manpower and paperwork, Kraftwork Union standardized design has been approved. Occa- (KWU), the chief German reactor vendor, has ne- sionally, certain backfits (e.g., several following gotiated a plan with State licensing authorities and the Three Mile Island accident) may be judged

25-450 0 - 84 - 14 : QL 3

198 ● Nuclear Power in an Age of Uncertainty important and then they are implemented on all plants nearby. Additional funds are used to re- plants of a certain design (36,6). duce electric bills of local residents (12,36,38). More funding is available if power is exported Standardized Training.–In West Germany from the local area. there is a single institution for certifying powerplant operators. This school, called the Kraftwerk- A similar approach is used in France where schule is owned by a joint organization of 116 electric bills in areas surrounding nuclear plants utility members in six countries. Operators com- are reduced by 15 percent, and funds to build plete a 3-year course including supervised opera- housing, schools, and other public facilities are tion of an actual powerplant. Such training is a lent by the utility to nearby towns which repay minimum requirement for a deputy shift super- the loan in utility property tax abatements. In visor. The shift supervisor must be an engineer Italy, there is an 18-month site review and ap- (33). proval process. Special public education centers are set up at each proposed site well in advance Siting of Nuclear Powerplants.–ln Japan and to help educate the public about the benefits and Italy, land is constrained, and finding sites for nu- risks of nuclear power (12). clear plants is difficult. In Japan, the most difficult part of the licensing process is the series of nego- Financial Risk.—There is far less financial risk tiations with local governments. MITI has in its to utilities investing in nuclear power in other budget about $60 million for public works grants countries than in the United States. In West Ger- to local governments that accept nuclear power- many, utilities set their own electric rates subject Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 199 to general Federal approval for household rates The public inquiry into the Sizewell B PWR and subject to antitrust provisions for industrial now underway in the United Kingdom has been rates. Financing is provided not only by the pri- one of the most extensive public debates on nu- vate sector market but also by regional govern- clear power ever held. Beginning in January 1983, ments, which are pat-t-owners of the largest util- it lasted most of 1983. Much of the argument fo- ities. In Sweden where utilities also use private cused on the economics of the proposed project. financing for investments, electric rates are gen- In all but one of the five electricity demand sce- erally not regulated, and utilities are in good narios proposed by the Central Electricity Gen- financial shape (see table 28) (26). Similarly, in erating Board (CEGB), electricity demand in the Japan, privately financed utilities are strong finan- United Kingdom is forecast to decline between cially. Electric rates in Japan are the highest in 1980 and 2000. The argument for building the the world and reflect the full costs of producing Sizewell B powerplant is that cheaper nuclear power, in order to encourage conservation (12). power will substitute for increasingly expensive In France, on the other hand, the Government oil and coal, but there is much official skepticism approves electric rates; in 1982 they were not in- about the economic plan from other government creased enough to prevent a deficit of about $1 agencies. The public inquiry has addressed ques- billion in the account of the electric utility (36). tions of the likelihood of cost overruns and of public pressure for expensive safety improve- Timing and Balance. —There are some indica- ments to match those being required in the tions that public opposition outside the United United States and West Germany (32). States has been intensified by very rapid development of nuclear power in Sweden and West Ger- The Sizewell B debate should provide a thor- many (26). At the same time, public acceptance ough exploration of many of the issues now fac- for nuclear power in many countries seems to ing the U.S. nuclear industry. Furthermore, it will be more solid if nuclear power is included as part provide one more example of a possible ap- of an overall national energy plan that includes proach to involving the public in decisionmak- a strong emphasis on energy conservation, re- ing on nuclear power. Conceivably such a full newables, and other sources of electricity. Such inquiry could precede the launching of a “con- plans have been formally announced in Japan, voy” of new advanced design LWRs when orders France, West Germany, and Italy, all countries pick up again. with potentially important roles for nuclear power (1 2). The U.S. Industry in an Shift to a New Reactor Technology.–As con- International Context sideration is given in the United States (see ch. Although the United States has by far the largest 4) to the desirability of shifting to a whole new number of nuclear plants of any country, future reactor technology—heavy water reactors (HWRs), prospects for the U.S. nuclear industry are re- gas-cooled reactors (GCRs), or “forgiving” LWRs, garded as dimmer than those of several other several lessons can be learned from foreign ex- countries, at least by some observers. Mans Lonn- perience. One is that it is quite possible to shift roth, a Swedish author of a book on the world- to a new technology. France shifted from GCRs wide nuclear industry (25), describes the com- to PWRs for plants ordered in the early 1970’s. ing decade as tough for all nuclear countries since The United Kingdom is now considering the shift the industry will have to demonstrate that it is a to PWRs in a formal public inquiry. In both cases, “safe, reliable and economic energy source. ” the shift has been towards a “standard” technol- Success will depend in part on the coherence of ogy now in use in the United States and else- each country’s response to this challenge (24). where, Clearly, this is a different situation from a shift to a relatively untested technology de- The French have been very successful at con- scribed as a possibility in chapter 4. structing nuclear plants quickly and cheaply. 200 ● Nuclear Power in an Age of Uncertainty

They have a large design, manufacturing, and Thus West Germany, France, Japan, and the construction capacity much of which has no United States–and perhaps Canada as well–will other alternative use. Among French industries, be competing for a nuclear export market in the the nuclear industry has been successful, and 1980’s and 1990’s. They will be competing for there will be considerable pressure to keep it a market that is far smaller than worldwide nucle- operating even in the face of slower growth in ar industrial capacity and far smaller than pre- electricity demand than anticipated. The French vious estimates had projected (see table 29 and nuclear industry, however, has had less success appendix table 7B). Several importing countries— in exporting sales than has West Germany or the Korea, Taiwan, Argentina, and Spain–are work- United States. This is the big challenge for the ing to develop domestic supply capabilities and next decade (24). will be curtailing imports. Many countries are importing far less equipment and engineering and By contrast, Lonnroth claims, in the United construction management services than they did States a “collapse of the nuclear industry would earlier in the decade. not seriously affect either the public authorities, the industry at large, or the American society” Given the softness of the export market, it will (24). The U.S. social and political system is more be difficult for suppliers of nuclear equipment and fragmented than either the French or West Ger- services to be sustained through the next two dec- man system and for this reason may have difficul- ades unless they have some domestic base. Al- ty developing a coherent long-range strategy in though U.S. firms are involved in 31 plants still the absence of full consensus. West German gov- under construction, all major components will ernmental processes are as complex and open have been delivered by the end of 1984. For as those in the United States, but the German those overseas plants engineered by U.S. AE com- nuclear industry acts far more as a single industry. panies, the basic design is now complete (37). It “views itself as one industry, with one collective If U.S. suppliers begin to get new domestic orders will and one identity” (24). For this reason Lonn- in about 5 years, they will provide a new basis roth suggests that the technical and economic for maintaining design teams and manufacturing coordination necessary for a long-range nuclear facilities capable of sales abroad. If new reactor strategy in West Germany is possible within a se- orders are delayed for 10 years or more, U.S. ries of “interlocking ownerships among the main firms may find themselves looking to the Japa- industrial actors” under the long-range guidance nese, West Germans, or French for joint ventures of key West German banks (24). or licensing arrangements in which the foreign company is an equal, or even dominant, partner. Japan has had its difficulties with nuclear This would be the reverse of the situation in each power, including prolonged siting processes and of these countries early in the history of the nu- construction delays. However, Japan has very clear industry. strong motivation to develop its nuclear industry because it has no indigenous fossil fuels. Japan has also demonstrated the ability to coordinate Nuclear Proliferation Considerations long-range industrial strategy in other areas and to develop an export-oriented strategy. Japan has U.S. and worldwide efforts to restrict prolifera- become the single most important exporter of tion of nuclear military technology are an impor- heavy electrical equipment and is expected to tant influence on the development of interna- provide about 25 percent of all exports in this sec- tional trade in civilian nuclear power (25). The tor from 1975-87 (24). This gives Japan a very perceived link between the civilian and military strong industrial base on which to develop a uses of nuclear power has stimulated much of nuclear exports business. For the moment the the opposition to nuclear power (see ch. 8). The nuclear industry still needs government support. reasoning is straightforward: commercial nuclear However, such support will not be available power requires the production of nuclear fuels, indefinitely. and some of these fuels and facilities that pro- Ch. 7—Survival of the Nuclear Industry in the United States and Abroad ● 201

Table 29.-Estimated Reactor Export Market (1983-87 inclusive) in Units

Hardware Software Previous suppliers Orders Industrial market market 1960-80 Low High capability Low High Low High (no. of units) Europe: 12 North ...... 3) Belgium ...... 0 1 ‘/2 o 0 0 0.5 U.S. (6), France (2) Finland ...... 0 1 2 0 0.5 0 0.75 Sweden (2), USSR (2) United Kingdom ...... 0 ‘/2 0 0.25 0 0.5 — South ...... (3 9) Greece/Turkey/Portugal. . . . . 0 1 3 0 1 01 – Italy ...... 2 4 1 0 0 0.5 1 Us.(l) Spain ...... 1 2 1 0 0 0.25 0.5 U.S. (12), Germany (2) Yugoslavia ...... 0 2 2 0 1 0 1.5 Us. (1) Latin America: o 5 Argentina ...... 0 1 2 0 0.5 0 0.75 Germany (2), Canada (1) Brazil ...... O 2 2 0 1 0 1.5 Germany (2), U.S. (1) Mexico ...... O 2 2/3 0 1.5 02 Us. (2) As/a and Pacific: 2 10 China ...... O 2 2/3 0 1.5 02 Indonesia ...... 0 1 3 0 1 01 Pakistan ...... 0 1 2/3 0 0.75 01 Canada (1) South Korea ...... 2 4 ‘/2 0.5 1 1 2 U.S. (6), France (2), Canada (1) Taiwan ...... O 2 2 0 1 0.75 1.5 U.S. (6) Africa and Mid-east: o 5 Egypt ...... O 2 3 0 2 02 Israel ...... O 1 3 0 1 01 South Africa ...... O 2 2/3 0 1.5 02 France (2) Total ...... 5 32 0.5 15.5 2.5 22.5 Average annual rate ...... 1.0 6.4 0.1 3.1 0.5 4.5 SOURCE: Mans Lonnroth, “Nuclear Energy in Western Europe,” based on research for Nuclear Power Struggles, Allen and Unwin, London, 1983. duce them can be used for nuclear weapons. The nuclear fuel. A few countries, most notably fundamental premise of U.S. and worldwide ef- France, are working with plutonium for use in forts to avoid proliferation to additional nations breeder reactors. has been to keep civil uses of nuclear energy dis- tinctly separated from military applications, and Current worldwide and U.S. concern about nuclear power and proliferation stems from these to try to erect barriers to prevent diversion or mis- use of civil nuclear materials and facilities. considerations about the potential use of civilian nuclear fuels. One fear is that a rapidly industrial- Technical and institutional aspects of nuclear pro- izing state with a nuclear power base in a trou- liferation were the subject of a previous OTA bled part of the world might be tempted to use study and a recent Congressional Research Serv- its civilian program as a base for developing nu- ice paper which is reprinted in full as an appen- clear weapons. The second concern is that some dix to this report (1 1,34). underdeveloped countries with nominal nuclear Although it is technically possible to make power programs might obtain enough technology crude nuclear weapons from the plutonium in and equipment on the world market to build fa- spent fuel from nuclear power reactors, it is more cilities to produce weapons-usable materials. The likely from the higher grade plutonium manufac- grave concern in the 1960’s that many nuclear tured in spent-fuel reprocessing, and in the opera- powerplants worldwide would give rise to the tion of breeder reactors (see appendix table 7C). wholesale spread of nuclear arsenals has given Economic prospects for commercial reprocess- way in the 1980’s to the concern that a few nu- ing have decreased worldwide as well as in the clear powerplants and related facilities scattered United States and no nation is currently pro- among some developing countries could bring ducing and using plutonium commercially for them much closer to an ability to make nuclear 202 ● Nuclear Power in an Age of Uncertainty weapons. It is this concern that drives proposals nonsignatory states include India, Pakistan, Brazil, for restrictions on international nuclear cooper- Argentina, and South Africa. ation and trade. After NPT took effect, other events stimulated Concern about the spread and use of nuclear further proliferation concerns. In 1974, India weapons has led to several international initia- tested a nuclear explosive that was derived from tives. The Non-Proliferation Treaty (NPT) pledges civilian facilities. Shortly thereafter, France and its members that do not have nuclear weapons West Germany agreed to supply enrichment and/ to forego future acquisition of them, and requires or reprocessing plants to nations (Pakistan and verification of the use of civilian facilities by in- Brazil) which had refused to sign the NPT. By the ternational inspection. It further stipulates that all late 1970’s, the nuclear supplier countries had the nuclear facilities in signatory states without become concerned enough to agree informally nuclear weapons will be safeguarded, even those to exercise additional restraint in nuclear coop- that are developed indigenously. This treaty rep- eration and trade, particularly in the area of the resents a significant departure from practices prior transfer of sensitive technology. The United States to 1970, and is an important element of the in- imposed even more severe restrictions than the ternational proliferation regime. However, it has other suppliers with the passage of the Nuclear not been as effective as it might have been be- Non-Proliferation Act of 1978. cause a number of nations have refused to par- U.S. policies and controls that guide nuclear ticipate in the treaty. As shown in table 30, the cooperation and exports include the following

Table 30.-No-Nuclear Weapons Pledges in Effect in 1981

Treaty and data of entry into force Treaty Prohibiting Antarctic Limited Test Nuclear Weapons Nuclear Non- Treaty, Ban Treaty, in Latin America, Proliferation Treaty, State 1961 1963 1 966a 1970 Argentina...... P s S b Australia ...... P P P P Belgium ...... P P P P Brazil ...... P P s c Canada ...... P P P P Cuba...... — Egypt ...... F.R.G, , ...... P P India ...... P — Iran ...... P — P Iraq ...... P P Israel ...... P Italy ...... P P P Japan ...... P P P Libya ...... P P Netherlands. . . . . P P P Pakistan...... s South Africa . . . . P P South Korea . . . . P P Spain ...... P P Sweden ...... P P Taiwan ...... P P Yugoslavia. . . . ., P P P = Party. S = Signatory. aNot yet h force for all signatories. bRatified Subject to preconditions not Yet met. cAdditional Protocol II applylng to Dutch territories in Latin America. dThere is some difference of opinion as to whether one small unsafeguarded laboratory should be considered a faCiiity. SOURCE: W. Donnelly and J. Pilat, “Nuclear Power and Nuclear Proliferation: A Review of Reciprocal Interactions,” Congres- sional Research Service for the Office of Technology Assessment, April 19S3. Ch. 7–Survival of the Nuclear Industry in the United States and Abroad ● 203 items: restrictive conditions for licensing exports enously or by finding suppliers who offer less de- of nuclear materials, equipment, and reactors; manding conditions. This could give rise to the restrictive conditions for providing technical as- emergence of a second tier of suppliers from the sistance and transferring technology; restrictions more industrialized developing nations, such as on export of dual-use items that can be applied Argentina and India, who might not comply with to weapons programs as well as to legitimate nu- the guidelines of the major suppliers. This could clear power programs; post-export controls, or significantly change the character of nonprolifera- prior rights over what may be done with or to tion control. U.S. nuclear exports, such as reactors or fuel; and In part because of fears of loss of U.S. nonpro- cutoff of nuclear cooperation and exports to liferation influence and trade, the Reagan admin- states which violate safeguards agreements with istration has shifted emphasis somewhat from the the United States. These restrictions are em- policies of the Carter administration. Rather than bedded in U.S. law, particularly in the Non-Prolif- emphasize across-the-board denial of nuclear eration Act of 1978 mentioned above, the Atomic supply, the Reagan administration has promoted Energy Act of 1954, and the Symington and the concept that the United States is a reliable Glenn amendments to the Foreign Assistance Act supplier of nuclear equipment to trusted coun- of 1961. tries who are not viewed as proliferation risks. Nuclear cooperation and trade has been cir- cumscribed in many aspects by the restrictive conditions and controls intended to prevent the Conclusion development of nuclear weapons. Specifically, For economic, political, and technical reasons, the supply of sensitive nuclear technology to the 1980’s will be a difficult decade for the nu- countries that have little visible economic need clear power industry in all countries. Those most for it is discouraged by the nuclear suppliers, and likely to survive are those with the political and particularly by the United States. The supply of industrial cohesion to develop a long-range strat- items that can be used for both civilian and mili- egy for demonstrating that nuclear power is a tary purposes has been little affected in the past, safe, reliable, and economic energy source. but is likely to be more restricted in the future France, Japan, and possibly West Germany and since export control lists are being made more Canada have a combination of national motiva- detailed and specific. Importing countries most tion and institutional coherence that makes it likely to feel the effects of such additional re- quite possible they will survive the decade with strictions include Argentina, Brazil, India, and a more viable nuclear industry than will the pakistan. United States. In the worst case, therefore, if the Pressures from both formal and informal non- U.S. industry emerges weakened from a long pe- proliferation regulation can be viewed as stimu- riod with no new orders, these countries may relating some nuclear customer countries to seek verse earlier roles and provide some of the exper- independence of the major suppliers, either by tise and hardware to U.S. companies during the building up their own nuclear industries indig- early years of a revival of the U.S. industry. 204 . Nuclear Power in an Age of Uncertainty

Appendix Table 7A.–Onsite and Offsite Nuclear-Related Job Vacancies at INPO Member Utilities Mar. 1, 1982

Vacancies Occupations Positions a Number Percent of positions Managers and supervisors ...... 5,765 432 7.5 Engineers: Chemical ...... 179 30 16.8 Civil ...... , ...... 872 40 4.6 Electrical ...... 1,518 239 15.7 Instrument and control ...... 506 91 18.0 Mechanical ...... 2,844 327 11.6 Nuclear and reactor ...... 1,427 287 20.1 Quality assurance/control ...... 791 147 18.6 Radiation protection ...... 140 30 21.4 All other engineers ...... 2,229 420 18.8 10,506 1,611 15.3 Scientists: Biologists ...... 144 6 4.2 Chemists ...... 269 37 13.8 Health physicists ...... 404 83 20.5 Other scientists ...... 235 28 14.6 1,052 154 14.6 Training personnel SRO/Relicensed/certified instructors ...... 405 109 26.9 Other technical/scientific instructors ...... 576 100 17.4 Other instructors ...... 188 52 27.7 Support staff...... 135 17 12.6 1,304 278 21.3 Operators: Shift technical advisors ...... 416 93 22.4 Shift supervisors ...... 735 119 16.2 Senior licensed operators (SRO) ...... 385 117 30.4 Licensed operators (RO)...... 1,094 230 21.0 2,214 466 21.1 Non-licensed operators assigned to shift ...... 2,286 242 10.6 Other non-licensed operators ...... 351 102 29.1 2,637 344 13.1 Individuals ingraining for SRO licenses...... 495 22 4.4 Individuals ingraining for Relicenses...... 878 66 7.5 Individuals ingraining for non-licensed positions ...... 838 246 29.4

7,478 1,237 16.6 Technical and maintenance personnel: Chemistry technicians...... 1,004 161 16.0 Draftsmen ...... 1,209 98 8.1 Electricians ...... 1,609 172 10.7 Instrument and control technicians ...... 2,463 320 13.0 Mechanics ...... 3,554 244 6.9 Quality assurance/control technicians ...... 793 88 11.1 Radiation protection technicians ...... 1,792 266 14.8 Welders with Nuclear Certification ...... 415 48 11.6 Other technical and maintenance personnel ...... 3,883 312 8.0 16,742 1,709 10.2 All other professional workers ...... 1,304 125 9.6 Other technical personnel ...... 1,061 114 10,7 Total ...... 45,212 5.660 12.5

%h{sincludes pereons employed by INPOrnember utilities, Including holdlngcornpany positions allocated tothe utilities, vecant positions, and contractor positions used in normal operations. Note: Fifty-five utilities providing offsite information; Onsite data comes from 82 plants representing 58 utilities, except onsite vacancy data, which was provided by only81 piants representing57 utilities. Ch. 7—Survival of the Nuclear Industry in the United States and Abroad Ž 205

...... - ...... 0...... 206 . Nuclear Power in an Age of Uncertainty

Appendix Table 7C.—Usability of Nuclear Materials for Nuclear Weapons or Explosives

Usability Material and form Direct Indirect Processing required Plutonium: Metal ...... Yes Oxide ...... Possibility Yes Chemical separation Oxide with uranium in nuclear fuels ..... No Yes Do Oxide or other forms in spent nuclear fuels. . . No Yes Reprocessing Thorium: Ore, metal, chemical forms ...... No No Uranium-235: Normal ore, metal, chemical forms...... No No Slightly enriched (3-6 percent) metal ...... No Yes Chemical processing to produce uranium hexafluoride for enrichment to weapons grade Oxide in nuclear fuels...... No Yes Do Oxide or other forms in spent nuclear fuels. . . No Yes Reprocessing, chemical processing and enrichment Moderately enriched (20 percent) metal ...... Unlikely Yes Chemical processing and enrichment Oxide ...... No Yes Do Oxide in nuclear fuels...... No Yes Do Oxide or other forms in spent nuclear fuels. . . No Yes Reprocessing, chemical processing and enrichment Highly enriched (90 percent): Metal ...... Yes Oxide ...... Yes Oxide in nuclear fuel...... No Yes Chemical separation Oxide or other forms in spent nuclear fuels. No Yes Reprocessing Uranium-233: Metal ...... Yes Oxide ...... Yes Oxide in nuclear fuels ...... No Yes Chemical processing and enrichment Oxide or other forms in spent nuclear fuels. . . No Yes Reprocessing, chemical processing and enrichment

SOURCE. W. Donnelly and J. Pilat, “Nuclear Power and Nuclear Proliferation: A Review of Reciprocal Interactions,” Congres- sional Research Service for the Office of Technology Assessment, April 1983. Ch. 7—Survival of the Nuclear Industry in the United States and Abroad “ 207

CHAPTER 7 REFERENCES

1. Atomic Industrial Forum, Nuc/ear Power Facts and Nuclear Power Industry, 1982-1991, September Figures, November 1982, 1982, ORAU-205. 2. Atomic Industrial Forum, “List of Nuclear Power- 19. Kraftwork Union, The KWU Convoy Concept, plants Outside the U.S.,” December 1981. KWU Report, May 1982, No. 36. 3. Atomic Industrial Forum, Nuc/ear Power P/ants in 20. Lanouette, William J., “The CANDU Reactor From the United States, Jan. 1, 1983. Canada—Can It Do the Job Down Here?” Nation- 4. Baker, Joe G., and Olsen, Kathryn, Occupation/ a/ Journa/, June 28, 1980; “Canadian Electricity Employment in Nuc/ear-Re/ated Activities, 1981, May Be Cheaper, But It Doesn’t Come Free of April 1982, ORAU-197. Problems,” Nationa/ Journa/, May 22, 1982. 5. Barkenbus, Jack N., “An Assessment of institution- 21. Lanouette, William j., “The Effects on the Nucle- al Alternatives for Nuclear Power Generation, ” ar Industry of No New Plant Orders 1985 to working paper for OTA, January 1983. 2000,” working paper for OTA, February 1983, 6. Benat, Jean, “The French Nuclear Program: Ben- to be published in vol. Il. efits of Standardization, ” Nuc/ear Europe, January 22. Lester, Richard K., Nuclear Power Plant Lead- 1983. Times, International Consultative Group on Nu- 7. Business Week, “West Germany: Can a Nuclear clear Energy, published by The Rockefeller Foun- ‘Convoy’ Run Over Its Opposition?” Aug. 9, 1982. dation and The Royal Institute of International 8. Curtis, Carol E., “Japan: The Nuclear Victor?” Affairs, November 1978. Forbes, Dec. 6, 1982. 23. Lester, Richard K., “U.S.-Japanese Nuclear Rela- 9. Department of Energy, Office of Research, A Study tions: Structural Change and Political Strain, ” of the Adequacy of Personnel for the U.S. Nuclear Asian Survey, vol. XXII, No. 5, May 1982. Program, DOE/ER-01 11, November 1981. 24. Lonnroth, Mans, “Nuclear Energy in Western Eur- 10. Doblin, Claire P., The Growth of Energy Con- ope,” paper prepared for a conference on Nuc/ear sumption and Prices in the USA, FRG, France, and Electric Power in the Asia–Pacific Region, Jan. the UK, 195(9- 7980 (Laxenburg, Austria: interna- 24-28, 1983, East-West Center, Honolulu, Hawaii. tional Institute for Applied Systems Analysis, May 25. Lonnroth, M%ns, and Walker, William, Nuc/ear 1982), RR-82-18. Power Strugg/es (New York: Allen and Unwin, 11. Donnelly, Warren H., and Pilat, Joseph F., Nuc/e- 1983). ar Power and Nuclear Proliferation: A Review of 26. Lonnroth, M~ns, and Walker, William, “The Via- Reciproca/ /interactions, report by the Congres- bility of the Civil Nuclear Industry,” a working sional Research Service for the Office of Technol- paper for the International Consultative Group on ogy Assessment, April 1983. Nuclear Energy and published by the Rockefeller 12. E. J. Bentz & Associates and Burns& Roe Industrial Foundation and the Royal Institute of International Services Group, Characteristics of Foreign E/ectric Affairs in 1979. Utility Experiences for Potential Useful Applica- 27. Nuclear Engineering International, Nuc/ear Station tions in the U. S., prepared for the Department of Achievement 7983, October 1983. Energy’s Electricity Policy Project, December 28, Nuc/ear News, November 1981. 1982. 29. Nucleonics Week, Sept. 23, 1982. 13, The Economist, “America Frets Over the Future 30. Nelkin, Dorothy, and Pollak, Michael, “Ideology of Its Technical Nous,” Mar. 12, 1983. as Strategy: The Discourse of the Anti-N ucle~’r 14. Electricitede France, “The French Nuclear Elec- Movement in France and Germany,” Science tricity Pr?gramme,” undated (included 1981 data). Technology and Human Values, vol. 5, No. 30, 15, Electricitede France, French 900 MWe Nuc/ear winter 1980. Powerp/ants, October 1981. 31 Nelkin, Dorothy, and Pollak, Michael, “Political 16. Institute of Nuclear Power Operations, 1982 Sur- Parties and the Nuclear Energy Debate in France vey of Nuc/ear-Related Occupational Employment and Germany,” Comparative Po/itics, vol. 12, in U.S. E/ectric Uti/ities, December 1982, INPO January 1980. Report 82-031. 32, NewScientist, “PWR Inquiry,” articles (a) by Gor- 17. )ennekens, J. H., Nuc/ear Regu/ation–the Cana- don Mackerron, “A Case Not Proven.” (b) Jack dian Approach, Atomic Energy Control Board, Edwards, “Lessons From Three Mile Island,” (c) September 1981. Remy Cade, “How France Went Nuclear.” 18. Ruth C. Johnson, Manpower Requirements in the 33. Office of Technology Assessment, Nuc/ear P/ant —

208 . Nuclear Power in an Age of Llncetiainty

Standatiization: Light Water Reactors, OTA- 37 S. M. Stoner Corp., Nuclear Supply Infrastructure: E-1 34, April 1981. Viability Study, prepared for the Argonne National

34< Office of Technology Assessment, Nuc/ear Prolif- Laboratory, contract #31-109-38-6749, November eration and Safeguards, OTA-E-48, June 1977. 1982. Summary, OTA-E-148, March 1982. 38 Stevenson, J. D., and Thomas, F. A., Se/ected Re- 35 Overseas Electrical Industry Survey Institute, Inc., view of Foreign Licensing Practices for Nuclear Electric Power Industry in Japan 1981, December Power P/ants, NUREG/CR-2664, April 1982. 1981. 39 Surrey, John, and Thomas, Steve, “ Worldwide 36, Procter, Mary, “Notes on the French Nuclear Pro- Nuclear Plant Performance: Lessons for Technol- gram,” OTA memo to the files, based qn conver- ogy Policy,” Futures, February 1980. sations with the French nuclear attachq a trip to the Tricastin nuclear plant and other sources, July 19, 1983.

Chapter 8 Public Attitudes Toward Nuclear Power

Photo credit: OTA staff Buttons and bumperstickers from the controversy about nuclear power Contents

Page Introduction: Public Opinion and Its Impact on Nuclear Power ...... 21 1 Actors in the Nuclear Power Debate ...... 214 The impact of Public Opinion on Nuclear Power ...... 215 The Experts’ View ...... 217 The Controversy Over Safety Studies ...... 218 The lmpact of Risk Assessments on Public Opinion ...... 220 Factors Influencing the Public’s View on Nuclear Safety ...... 221 Perceptions of Risk and Benefit ...... 221 Psychological Factors ...... 222 Values and Knowledge ...... 226 Which is More Important? ...... 226 Values ...... 227 The Role of the Media...... 231 The Extent of Media Coverage of Nuclear Power ...... 231 The Content of Media Coverage ...... 232 What Would lt Take To Increase Public Acceptance of Nuclear Power ...... 234 Enhance Nuclear Advantages...... 234 Reduce Concerns Over Nuclear Accidents...... 236 Minimize Linkage Between Nuclear Power and Weapons ...... 239 Case Study 1–Maine Yankee: Economics as the Key to Public Support ...... 240 Case Study 2–Diablo Canyon: Environmentalism and Industry Credibility ...... 242 Case Study 3–Public Perception of lmproved Management ...... 244 Chapter 8 References ...... 245

Tables

Table No. Page 31. History of Statewide Referendum Votes Dealing With Nuclear Powerplants ...... 212 32. Major National Groups Influencing Public Opinion For and Against Nuclear Power 215 33. State Laws and Regulations Restricting Construction of Nuclear Powerplants ...... 216

Figures

Figure No. Page 39. Trends in Public Opinion on Nuclear Power ...... 211 40. Relationship Between Judged Frequency and the Actual Number of Deaths per Year for 41 Causes of Death ...... 223 41. Expert and Lay Judgments of Risk Plotted Against Technical Estimates of Annual Fatalities ...... 224 Chapter 8 Public Attitudes Toward Nuclear Power

INTRODUCTION: PUBLIC OPINION AND ITS IMPACT ON NUCLEAR POWER Public attitudes toward nuclear power have Polls taken between 1976 and 1979 indicated become increasingly negative over the past two that slightly over half of the American public decades, with the most recent polls indicating favored continued construction of nuclear plants that a slight majority of Americans opposes fur- in the United States in general, while about 28 ther construction of reactors. During the 1950’s, percent were opposed and 18 percent unsure. nuclear power was still in the early states of de- The accident at Three Mile Island (TMI) in April velopment, and pollsters did not even bother to 1979 had a sudden and dramatic impact on these survey the public on the issue. In the early 1960’s, attitudes. As shown in figure 39, the percentage a few scattered protests against local plants gained of people who had been in favor of or uncertain national attention, but opinion polls indicated about continued construction of reactors de- that less than a quarter of the public opposed nu- creased immediately following the accident while clear power (41 ). From Earth Day in 1970 through the number opposed increased (57). In subse- the mid-1970’s, opposition levels averaged 25 to quent months, there was some return to previous- 30 percent, indicating that substantial majorities ly held opinions, but opposition levels remained of the public favored further nuclear develop- much higher than they had been. National polls ment. However, by 1976, anti-nuclear referen- taken since mid-1 982 indicate a continued slow da appeared on ballots in eight States. erosion in support for nuclear power. About a

Figure 39.—Trends in Public Opinion on Nuclear Power ,100

0 I 1 i I I t I I I I 0 1975 1976 1977 1978 1979 1980 1981 1982 1983 Year of survey

Question asked: “Do you favor or oppose the construction of more nuclear powerplants?” SOURCE Cambridge Reports, Inc.

211 212 ● Nuclear Power in an Age of Uncertainty third of the public now supports construction of looks to nuclear power as one solution to the Na- new plants in general, while over so percent are tion’s long-term energy problems. In a recent opposed (6, 10, 18). The accident at TMI appears survey, the majority of respondents believed that to have accelerated a trend of even greater op- most U.S. energy needs would be supplied pri- position to construction of new plants near to marily by nuclear” and solar over the next two where those polled live. By the end of 1981, a decades, and over a third of those polled ex- large majority of those polled opposed construc- pected nuclear power to provide most of the Na- tion of new plants in or near their communities. tion’s energy after the year 2000 (14). The ma- When compared with other energy options, in- jority of Americans favor neither a halt to all new cluding offshore oil drilling and coal plants, construction nor a permanent shutdown of all op- nuclear is now the least favored alternative. erating reactors. Opinion polls on this question have been verified by State ballot initiatives. As Despite the trend of declining support, the pub shown in table 31, most of the nuclear moratori- Iic’s overall current attitude toward nuclear um initiatives, and all referenda that would have power can best be described as ambivalent. For shut down operating plants were defeated in example, a 1983 poll indicates that about 40 per- 1976, 1980, and 1982. However, more of these cent of the public thinks currently operating reac- initiatives have been approved in recent years, tors are “mainly safe” while slightly over half and many restrictions on nuclear waste disposal think they are dangerous and 5 percent are “not have been passed, reflecting public doubts about sure.” There is some evidence that the public the technology.

Table 31.—History of Statewide Referendum Votes Dealing With Nuclear Powerplants

Year State Proposal Outcome Vote. split 1976 Arizona Would halt new construction Defeated 30-70 % California and reduce operations until Defeated 33-66% Colorado safety systems were found Defeated 29-71 % Montana effective, liability ceilings Defeated 42-58% Oregon lifted, and waste disposal was Defeated 42-58% Ohio demonstrated Defeated 32-68% Washington Defeated 33-67% 1978 Montana Same as ’76 referenda Approved 60-40% 1980 Maine Would shut down Maine Yankee Defeated 41 -59% Missouri Would prevent Callaway plants Defeated 39-61% from operating until safety systems were found effective, liability ceilings were lifted, and waste disposal was available Oregon Prohibits new construction until Approved 52-48% waste disposal is available and voters approve in a statewide referendum 1981 Washington Prohibits issuance of new bonds Approved 56-44% needed to complete WPPSS Unit 3 1982 Idaho Prohibits legislation limiting Approved 60-40% nuclear power unless approved by voters in a referendum Maine Would phase out Maine Yankee Defeated 44-56% over 5 years Massachusetts Prohibits new construction and Approved 66-33% waste disposal unless certain conditions, including voter approval in a referendum. are met Total restrictive referenda placed on ballots: 14 Total approved: 4 SOURCES: Atomic Industrial Forum, State Codes. Ch. 8—Public Attitudes Toward Nuclear Power ● 213

The public’s ambivalent attitude toward nucle- the safe shutdown of the reactor while downplay- ar power is due to a variety of factors including ing the failure of the Emergency Core Cooling Sys- the ongoing debate among experts over reactor tem— misrepresented the nearness of a disaster safety, individual perceptions of the likelihood of and ignored the lack of foresight which the acci- a catastrophic reactor accident, changing per- dent demonstrated. sonal values, and media coverage of the technology. Underlying all of these factors is increas- While these earlier accidents had an adverse ing doubt about the technical capabilities and the impact on popular support for nuclear technol- credibility of both the nuclear industry and its ogy, it was not until 1979 that a single accident governmental regulators. As discussed in chapter had a direct, measurable impact on public opin- 5, weak utility management has led to poor op- ion as reflected in national opinion polls. That erating performance at some reactors as well as spring, poor maintenance, faulty equipment, and skyrocketing costs and quality-assurance prob- operator errors led to a loss of coolant and par- lems at other plants under construction. These tial destruction of the core at the TMI Unit 2 reac- problems have led to accidents at operating re- tor located in Pennsylvania. Radioactive water actors, causing great public concern. spilled onto the floor of an auxiliary building, releasing a small amount of radioactivity to the As early as 1966, when large majorities of the environment, although the total radiation dose public supported nuclear power, a design error received by the population in the vicinity was far caused blockage of coolant, leading to melting less than their annual exposure to natural and of a small part of the core at Detroit Edison’s Fer- medical radiation (31 ). On March 30, Governor mi breeder reactor (3). Although no radioactivity Thornburgh advised pregnant women and pre- was released, and the event received relatively school children to leave the area within a 5-mile little publicity at that time, nuclear critics and radius of TMI. This advisory was not lifted until some members of the public became concerned. April 9. Conflicting statements from authorities They pointed to a University of Michigan study combined with obvious confusion at the reactor conducted prior to construction of the plant, site before and during the evacuation shook which indicated that if the plant had been larger public confidence in the nuclear industry and and had been operating at full design power for State and Federal officials. Following the accident, at least a year, a complete breach of containment majority support for nuclear power was lost, a combined with the worst possible weather condi- trend that continues today. Local opposition to tions might have led to as many as 60,000 deaths some reactors around the country also increased (26). Nearly a decade later, public discussion of after the accident, while local attitudes toward the accident increased in response to the 1975 other reactors remained favorable (see Case publication of the book, We Almost Lost Detroit Studies at the end of this chapter). (25). Opinion polls taken after the accident at TMI In 1975, a fire started by a worker using a can- indicated that at least half of those polled thought dle to test for air leaks spread through the elec- more such accidents were likely. Since that time, trical system of the Tennessee Valley Authority’s other incidents, such as the rupture of steam gen- (TVA’s) Browns Ferry plant in Alabama. The fire erator tubes at Rochester Gas & Electric’s Ginna caused some loss of core coolant in one unit of nuclear plant in January 1982, have occurred at the plant, and disabled the reactor’s safety sys- operating reactors. There is some evidence that tems (1 1). Because of confusion about how to put the public views these incidents, along with the it out, the fire burned out of control for 7 hours TMI accident, as precursors to a catastrophic ac- before being extinguished. Again there was no cident that might kill thousands (67). loss of life and no release of radiation, but the incident was reported in the national media, in- The handling of reactor safety issues by the creasing public fears of an accident. Critics felt nuclear industry and the NRC has led many peo- that the Nuclear Regulatory Commission’s (NRC’s) ple to the conclusion that both have seriously news release on the event—which emphasized underestimated safety problems. For example,

25-450 0 - 84 - 15 : QL 3 214 . Nuclear Power in an Age of Uncertainty opinion polls indicate that a majority of the public percentages of respondents were also extremely believed government officials understated the pro- or anti-nuclear. Thus, it appears that different dangers at TMI (57). In addition, since 1973 the groups among the public vary in the strength of nuclear industry has argued that the possibility their beliefs about nuclear power. of failure of reactor emergency shutdown systems During the 1970’s, critics of nuclear power and is negligible. Because of industry opposition, the their associated public interest groups became NRC delayed regulations requiring extra equip- increasingly well-organized at the national level. ment to avoid an accident in the event of such As shown in table 32, today all of the major na- a failure. However, it was exactly this type of fail- tional environmental groups are critical of at least ure that occurred not once, but twice within 4 some aspects of the U.S. nuclear program (33). days at Public Service Gas& Electric’s Salem, N. J., In addition to these environmental groups with plant in February 1983 (38). (See ch. 5.) broad agendas, several organizations, such as the The importance of public opinion to further de- Union of Concerned Scientists and the Critical velopment of nuclear power has been recognized Mass Energy Project of Ralph Nader’s Public Citi- by government and industry but is still little zen, Inc., focus primarily on nuclear power. understood (12,75). This chapter attempts to add Overall, about 1 million Americans belong to en- to the limited understanding of public percep- vironmental and energy groups critical of nuclear tions and to identify changes in the management power. Total annual expenditures for lobbying, of nuclear power that might make it more accept- public education, and other activities related able to the public. The analysis is limited to public directly to nuclear energy are estimated to be perceptions of operating nuclear reactors and about $4 million (33). In addition, these groups those under construction. An April 1983 OTA rely heavily on volunteer labor and donated re- stud y, Managing Commercial High-Level Radio- sources. active Waste, deals with public attitudes toward Partially in response to the publicity attracted transportation and disposal of spent fuel and nu- by nuclear critics, proponents of nuclear tech- clear waste in greater depth. nology have also formed advocacy groups, as shown in table 32. Most of the groups supporting the technology are trade and professional associa- Actors in the Nuclear Power Debate tions, although there are some broad-based pub- As discussed in chapter 1, there area number lic interest groups in this category as well. In total, of groups in the United States with sometimes about 300,000 individuals belong to professional conflicting interests in nuclear power. The ap- societies and public interest groups that directly parent contradictions in public attitudes toward or indirectly support nuclear energy develop- the technology are explained at least partially by ment. Some groups in this category, such as the the fact that there is not a single homogeneous American Nuclear Society and the Atomic indus- “general public” in this country. Opinion polls trial Forum, focus primarily on nuclear power, which survey the “general public” may fail to while others such as the Edison Electric Institute reveal the intensity of individual opinions. For are utility trade associations with broad agendas example, the phrasing of the question most fre- that include advocacy of nuclear power among quently asked to gage national public opinion— many other issues. In response to the accident “In general, do you favor or oppose the building at TMI, nuclear advocates stepped up their public of more nuclear power plants in the United education efforts through the creation of the States?” –leaves little room for people who are Committee for Energy Awareness (CEA). Current uncertain or have no opinion. When the ques- plans call for expenditure of about $27 million tion was rephrased in two surveys taken shortly in 1983 for CEA, a major increase over previous after the TMI accident, over a third of the re- expenditures of about $6.5 million by all groups spondents were uncertain or neutral (45). A na- combined (42). tional poll taken in 1978 indicated that about a Nuclear advocates and critics, including the third of respondents were neutral; however, large staffs of public interest groups, are knowledgeable Ch. 8—Public Attitudes Toward Nuclear Power . 215

Table 32.-Major National Groups Influencing Public Opinion For and Against Nuclear Power

Groups supporting nuclear power Groups opposing some aspects of nuclear power Category 1: Large organizations with a focus on nuclear Category 1: Groups with a focus on nuclear energy targeting a broad audience. energy and alternatives to it. — U.S. Committee for Energy Awareness — Union of Concerned Scientists — Atomic Industrial Forum — Critical Mass Energy Project of Public — American Nuclear Society Citizen, Inc. Category 2: Lobbying organizations with a primary or secondary — Nuclear Information and Resource focus on nuclear energy. Service — Americans for Nuclear Energy — Safe Energy Communications Council — American Nuclear Energy Council Category 2: Large environmental groups that — Americans for Energy Independence participate in lobbying and public criticism of Category 3: Trade and professional associations that support nuclear energy. commercial nuclear energy. — Sierra Club — Edison Electric Institute — National Audubon Society — American Public Power Association — Natural Resources Defense Council — National Rural Electric Cooperative Association — Friends of the Earth — Institute of Electrical and Electronics Engineers — Environmental Policy Center — American Association of Engineering Societies — Environmental Defense Fund — Health Physics Society — Environmental Action, Inc. — Scientists and Engineers for Secure Energy Category 4: Industry research organizations indirectly influencing public opinion. — Electric Power Research Institute — Institute for Nuclear Power Operations — Nuclear Safety Analysis Center

SOURCES Terry Lash, “Survey of Major National Groups Influencing Public Opinion Against Nuclear Power, ” Office of Technology Assessment contractor report, April 1983, M & D Mills, “Activities of GrOUpS Which Influence Public Opinion in Favor of Nuclear Power, ” Off Ice of Technology Assessment contractor report, May 1983 about nuclear power and much more committed The Impact of Public Opinion to their beliefs than the general public. They act on Nuclear Power on these beliefs both in seeking to influence Public concerns about reactor safety, nuclear nuclear power policies at the State and Federal waste disposal, and rising construction costs have level and in attempting to convince the public had a particularly notable impact on State policies of their point of view. affecting nuclear power. As discussed in chapter The nuclear establishment sometimes blames 6, State Public Utility Commissions (PUCs) must nuclear critics for the growth of public opposi- grant a license certifying the need for power prior tion to nuclear power. However, to some extent to construction of any type of new powerplant. these individuals simply are reflecting the con- Because PUCs have veto power over new plants, cern of the wider public which has grown in based on economic and financial criteria, State response to reactor accidents and the increasing laws essentially can halt further development of financial problems of the utility industry. In ad- nuclear power. While critics and advocates have dition, the success or failure of both advocates been involved in voter-initiated referenda restrict- and critics depends in part on public response ing further licensing of nuclear plants, it is ulti- to their arguments. A 1983 opinion poll indicates mately the voters of the State (the “general pub- that Ralph Nader, a leading environmentalist, is lic”) who decide whether or not to approve these considered very believable on energy matters (9). restrictions. Table 31 provides a history of State Electric utility trade associations are considered votes on nuclear energy referenda. Overall, the somewhat less believable, and nuclear industry trend appears to reflect accurately the trends associations have much lower credibility among shown in public opinion polls, declining from a poll respondents. Thus, it appears that the public large margin of suppt for nuclear power in 1976 may be more willing to listen to and accept the to an ambivalent position today. While all seven arguments of nuclear critics than those of advo- restrictive proposals were defeated in 1976, cates. voters in Oregon and Massachusetts approved 216 . Nuclear Power in an Age of Uncertainty

initiatives limiting new reactor construction in tion. A complete list of State laws and regulations 1980 and 1982. (including those enacted by voter referenda) affecting nuclear power is given in table 33. Be- In California, State legislators approved a law cause of these laws and regulations, utilities in restricting nuclear power development in 1976 10 States cannot obtain State licensing of pro- to head off a more stringent Statewide referen- posed nuclear reactors until certain conditions, dum with similar provisions that was then turned such as a clear demonstration of high-level waste down only a few months later. The law passed disposal, are met. by the legislature was upheld by the U.S. Su- preme Court in April 1983. Other State legisla- Even in those States where nuclear power de- tures and a few PUCs have limited further con- velopment is not limited by law or regulation, struction of nuclear plants by legislation or regula- State politics can influence utility decisions about

Table 33.—State Laws and Regulations Restricting Construction of Nuclear Powerplants

Year State Type of action approved Citation Provisions California Law-by 1976 Cal Pub Res Code, No licensing of new plants until Federal Govern- Iegislaturea Sees. 25524.1- ment approves a demonstrated high-level 25524.2 waste disposal technology and fuel rod reprocessing technology is available. Connect i cut Law-by 1979 H-5096, approved No licensing of a fifth plant until Federal legislature June 18 Government approves a demonstrated high- Ievel waste disposal technology. Law-by 1983 H.B. 5237 Limits construction costs of Millstone 3 to legislature rate-payers to $3.5 billion. Kentucky Resolution-by 1982 HR-85, Declares the State’s intention to prohibit legislature adopted March 26 construction of plants. Maine Law-by 1977 Me Rev Stat Ann, No licensing of new plants until Federal legislature Tit 10, Sees. 251- Government demonstrates high-level waste 256 (West 1980) disposal and a majority of voters approve in a referendum vote. Maryland Law-by 1981 Ann Code Md, No licensing of new generators of nonmedical legislature Health-Environ- low-level waste until Federal Government mental, Tit 8, demonstrates waste disposal or an interstate Sec. 402 compact is in effect. Massachusetts Law-by 1982 Question 3, No licensing of new plants or nonmedical referendum Approved Nov. 2 low-level radioactive waste disposal sites until (Chap. 503, Acts a Federally approved storage facility is of 1982) operating, and other conditions, including voter approval, have been met. Montana Law-by 1978 Mt Code Ann, No licensing of new plants until all liability referendum Tit 75, Sees. limits for an accident are waived, a bond is 20-201, 20-1203 posted against decommissioning costs, and other conditions, including voter approval, are met. Oregon Law-by 1980 Measure No. 7, No licensing of new plants until Federal referendum Approval Nov. 4 Government provides high-level waste disposal and a majority of voters approve in a referendum. Vermont Law-by 1975 Vt. Stat Ann, No licensing of new plants without General legislature 1970, V.8, Tit 30, Assembly approval. Sec. 248c Wisconsin Regulation-by 1978 Dkt No 05-EP-1, No licensing of new plants without progress Public Service Wis Pub Serv on waste disposal, fuel supply, decom- Commission Comm, Aug. 17 missioning, and other economic issues. Washington Law-by 1981 Chap. 80.52, Rev. No issuance of bonds for major new energy referendum Code of facilities (including nuclear plants) without Washington voter approval. aon Apr. ZI, 1983, the Us. Supreme Court upheld the COnSI iIUIiOditY Of this law SOURCES: Atomic Industrial Forum, State Codes, NRC Office of State Programs. Ch. 8—Public Attitudes Toward Nuclear Power ● 217 nuclear plants. Whether Public Utility Commis- control technology into new and existing plants. sioners are directly elected or appointed by an Negative public perceptions may also affect the elected Governor, they are sensitive to State pol- availability of financing for new nuclear plants. itics and broad public opinion. Public concerns The financial problems caused by the accident about nuclear power may lead Utility Commis- at TMI discouraged some investors and brokers sioners to disallow rate increases needed to from investing in utilities with nuclear plants finance completion of plants under construction underway, driving up the cost of capital for those or to simply deny a license entirely. Public op- utilities. Finally, negative public attitudes affect position at the local level, too, can discourage nuclear power’s future in less tangible ways: The utilities from implementing planned nuclear most gifted young engineers and technicians may plants. For example, Portland General Electric in choose other specializations, gradually reducing Oregon canceled its planned Pebble Springs reac- the quality of nuclear industry personnel. And, util- tor in 1982 following a lengthy siting controver- ities simply may not choose nuclear plants if they sy that made the project less economically attrac- perceive them as bad for overall public relations. tive. The approval of a State referendum in 1980 The future of nuclear power in the United banning licensing of new plants until waste dis- States is very uncertain due to a variety of eco- posal technology was available contributed to the nomic, financial, and regulatory factors outlined utility’s decision. in other chapters of this report. Both parties to Over the past few years, public concern about the nuclear debate are bringing these factors be- reactor safety in reaction to the accident at TMI fore the broader public. Some may argue that the has encouraged additional NRC safety studies and issues are too complicated for the general public new regulatory requirements, increasing nuclear to contend with. However, as Thomas Jefferson power costs and making it less attractive to util- said, “When the people are well informed, they ities. (A more detailed analysis of the costs of reg- can be trusted with their own government. ” ulatory requirements is included in ch. 6.) This None of the conditions seen by utilities as a re- trend is partially a continuation of increasing quirement for a revival of the nuclear industry– public concern about environmental quality that regulatory stability, rate restructuring, and began in the late 1960’s. Translated into laws and political support-can be met without greater regulations, those concerns drove up the price public acceptance. Thus, unless public opinion of both nuclear and coal-fired powerplants as util- toward nuclear power changes, the future pros- ities were required to incorporate more pollution pects for the nuclear industry will remain bleak.

THE EXPERTS’ VIEW

In contrast to the public, most “opinion particularly supportive of nuclear power. Seventy- leaders,” particularly energy experts, support fur- four percent of scientists viewed the benefits of ther development of nuclear power. This support nuclear power as greater than risks, compared is revealed both in opinion polls and in technical with only 55 percent of the rest of the public. studies of the risks of nuclear power. A March 1982 poll of Congress found 76 percent of mem- In a recent study, a random sample of scien- bers supported expanded use of nuclear power tists was asked about nuclear power (62). Of (50. In a survey conducted for Connecticut Mu- those polled, 53 percent said development tual Life Insurance Co. in 1980, leaders in religion, should proceed rapidly, 36 percent said develop- business, the military, government, science, edu- ment should proceed slowly, and 10 percent cation, and law perceived the benefits of nuclear would halt development or dismantle plants. power as greater than the risks (19). Among the When a second group of scientists with particular categories of leaders surveyed, scientists were expertise in energy issues was given the same 218 ● Nuclear Power in an Age of Uncertainty

choices, 70 percent favored proceeding rapidly The Controversy Over Safety Studies and 25 percent favored proceeding slowly with the technology. This second sample included ap- The Atomic Energy Commission (AEC) com- proximately equal numbers of scientists from 71 pleted its first major study of the consequences disciplines, ranging from air pollution to energy of a reactor accident involving release of radio- policy to thermodynamics. About 10 percent of activity in 1957. Commonly known as WASH- those polled in this group worked in disciplines 740, the study was based on a very small (by directly related to nuclear energy, so that the today’s standards) 165-megawatt (MW) hypo- results might be somewhat biased. Support thetical reactor. In the worst case, an accident among both groups of scientists was found to at such a plant was estimated to kill 3,400 peo- result from concern about the energy crisis and ple (5). While the study itself did not become a the belief that nuclear power can make a major source of public controversy, its findings contrib- contribution to national energy needs over the uted to concern about the impacts of an accident. next 20 years. Like scientists, a majority of engi- In 1964, AEC initiated a new study to update neers continued to support nuclear power after WASH-740 based on a larger, 1,000-MW reactor. the accident at Three Mile Island (69). The study team found that a worst-case accident Of course, not all opinion leaders are in favor could kill as many as 45,000 people but was un- of the current U.S. program of nuclear develop- able to quantify the probability of such an acci- ment. Leaders of the environmental movement dent. Rather than publish these disturbing find- have played a major role in the debate about ings, AEC chose to convey the results to Congress reactor safety and prominent scientists are found in a short letter. Nuclear critics were very dis- on both sides of the debate. A few critics of turbed by this action, which they viewed as an nuclear power have come from the NRC and the attempt to keep the facts away from the public nuclear industry, including three nuclear (22). In recent years, awareness of AEC’s handl- engineers who left General Electric in order to ing of this early safety study has added to public demonstrate their concerns about safety in 1976. skepticism about the credibility of both that agen- However, the majority of those with the greatest cy and its successor, the NRC. expertise in nuclear energy support its further In 1974, AEC published the first draft of the Re- development. actor Safety Study, also known as WASH-1400 Analysis of public opinion polls indicates that or the Rasmussen report. A panel of scientists people’s acceptance or rejection of nuclear organized by the American Physical Society (APS) power is more influenced by their view of reac- found much to criticize in this report. The panel tor safety than by any other issue (57). As dis- noted that AEC’s fatality estimates had considered cussed above, accidents and events at operating only deaths during the first 24 hours after an ac- plants have greatly increased public concern cident, although radioactive cesium released in about the possibility of a catastrophic accident. an accident would remain so for decades, expos- Partially in response to that concern, technical ing large populations to adverse effects. The most experts have conducted a number of studies of serious forms of illness resulting from a reactor the likelihood and consequences of such an ac- accident, the APS reviewers argued, would be cident. However, rather than reassuring the pub- forms of cancer that would not show up until lic about nuclear safety, these studies appear to years after the accident. Other APS reviewers have had the opposite effect. By painting a pic- found fault with the Rasmussen report’s methods ture of the possible consequences of an accident, used to predict the performance of emergency the studies have contributed to people’s view of cooling systems (23). the technology as exceptionally risky, and the On October 30, 1975, the NRC, which had as- debate within the scientific community about the sumed the regulatory functions of the former study methodologies and findings has increased AEC, released the final version of WASH-1400. public uncertainty. Again, there was an extensive, widely publicized Ch. 8—Public Attitudes Toward Nuclear Power ● 219 debate over the document. The Union of Con- troversy that continued for several weeks. This cerned Scientists released a 150-page report cri- analysis, known as the Sandia Siting Study, was tiquing the study, and in June 1976, the House initiated to determine the sensitivity of the con- Subcommittee on Energy and Environment held sequences of reactor accidents to local site char- hearings on the validity of the study’s findings acteristics (2). While the Sandia team did not (71). As a result of these hearings, NRC agreed study accident probabilities in depth, they esti- to have a review group examine the validity of mated the probability of a “Group 1“ or (worst- the study’s conclusions. case) accident involving a core meltdown, failure of all safety systems, and a large radioactive re- Three years later, in September 1978, the release, at 1 in 100,000 reactor years. The conse- view group concluded that although the Reac- quences of this and other less severe hypothetical tor Safety Study represented a substantial advance accidents were estimated for 91 U.S. reactor sites over previous studies and its methodology was using local weather and population data and as- basically sound, the actual accident probability suming a standard 1, 120-MW reactor. At the cur- estimates were more uncertain than had been rent site of the Salem, N. J., reactor on the Dela- assumed in the report (35). The panel also was ware River under the most adverse weather con- critical of the executive summary, which failed to ditions and assuming no evacuation of the local reflect all of the study findings. The following population, a Group 1 accident at the hypotheti- January, the NRC accepted the conclusions of the cal reactor was estimated to cause 102,000 review panel. In a carefully worded statement, “early” deaths within a year of the accident. If the agency withdrew its endorsement of the nu- the hypothetical reactor were located at Bu- merical estimates contained in the executive sum- chanan, N. Y., where the Indian Point plant now mary, said that the report’s previous peer review stands, a Group 1 accident under the worst-case within the scientific community had been “inad- weather conditions (the accident would be fol- equate,” and agreed with the panel that the dis- lowed by a rainout of the radioactive plume onto aster probabilities should not be used uncritical- a population center) might cause $314 billion in ly (47). property damage, according to the study esti- Two studies published in 1982 continued the mates. debate over the validity of accident probability estimates included in the Rasmussen report. The Although the Sandia report itself did not include first, conducted by Science Applications, Inc. estimates of the “worst-case” accident conse- (SAI) for the NRC, was based on the actual op- quences, background information containing the erating history of U.S. reactors during the 1969-79 estimates and a copy of the draft report were period. By examining the frequency of precur- leaked to the press on November 1, 1982. Media sors that could lead to an accident involving core accounts that day highlighted the high death and damage or meltdown, SAI estimated that the property damage estimates, while downplaying probability of such an accident during the pre- that part of the analysis which indicated that con- TMI decade was much greater than suggested by sequences of this severity had only a 0.0002- the Rasmussen report (43). In response, the In- percent chance of occurring before 2000 (51). stitute for Nuclear Power Operations (lNPO—a Some accounts suggested that the worst-case nuclear industry safety research group) published consequences had the same probability as the a report arguing that SAI’s probability estimates Group 1 or worst-case accident, which was esti- were about 30 times too high, and that the ac- mated to have a 2-percent chance of occurring tual probability of a core-damaging accident was before the end of the century. closer to the 1 in 20,000 reactor years estimated That same day, the NRC held a press confer- in the Rasmussen report (28). This controversy ence to clarify the purpose and findings of the has not yet been resolved. study, and on November 2, Sandia National Lab- While debate over the SAI report was limited oratory issued a statement saying that wire serv- to a small community of safety experts, a more ice accounts “seriously misinterpret the conse- recent study aroused a widespread public con- quences of nuclear power reactor accidents. The 220 . Nuclear Power in an Age of Uncertainty

probability of a very severe nuclear power reac- sequently, increasing polarization (41 ). Both crit- tor accident is many thousands of times lower ics and supporters of the study focused on the than stated in these accounts” (63). The nuclear methodology and quality of data. The debate industry took out full-page ads in major national over the study continues today, with critics argu- papers to try to counteract the story. At the same ing that NRC’s 1979 statement was a “rejection time, however, nuclear critics emphasized that of the report’s basic conclusion,” “repudiating the Sandia draft report itself had excluded the the central finding of the Rasmussen report” (23). worst-case consequence data and argued that Meanwhile, INPO challenges the methodology “the NRC is once again feeding selective data to and data of SAI’s more recent safety study, argu- the public on the theory that they know best what ing that the Rasmussen report’s probability esti- information the public should have” (73). While mates are still valid. nuclear advocates argued that the report’s findings on accident consequences had been great- Although the general public is uncertain about ly overstated by the press, critics charged that nuclear power, most people have more faith in data were used incorrectly in developing those scientific “experts” than in any other source on estimates. Examining the same information on ac- nuclear power questions (20,39,57). Because of cident probabilities at individual plants used by this faith, public disputes among scientists and the Sandia team, the Union of Concerned Scien- other energy experts, as in the case of the Ras- tists found that the likelihood of an accident in- mussen report, have a particularly negative im- volving a release of radioactivity might be much pact on public acceptance of nuclear power. greater than assumed in the Sandia report (65). Rather than attempting to follow the debate and This debate, too, has not been resolved. sort out the facts for themselves, many people simply conclude that nuclear technology has not The Impact of Risk Assessments yet been perfected. in other words, if the “ex- on Public Opinion perts” cannot agree on whether or not nuclear power is safe, the average citizen is likely to The release of the Rasmussen report raised par- assume it is probably unsafe. In Austria, the ticular concerns about nuclear power for some government attempted to resolve the growing people because of the public disagreements controversy over nuclear power by structuring among the “experts” that resulted. In June 1976 a series of public debates among scientists with hearings held by the House Interior Committee, opposing views. Rather than reassuring the pub- scientists from the Massachusetts Institute of lic, the debate led to increased public skepticism Technology and Princeton and Stanford Univer- and ultimately to a national referendum that sities, as well as a high-level official of the En- killed that country’s commercial nuclear vironmental Protection Agency, testified about program. the methodological weaknesses and limitations of the report. Thus, as Princeton physicist Frank If public debates about nuclear safety studies Von Hippel pointed out at the hearings, “Instead have only fueled the fires of controversy and of dampening the fires of controversy, the pub- added to public skepticism, what can be done lication of the Rasmussen report has had the ef- to make nuclear power more acceptable to the fect of adding fuel to them” (71). public? In order to answer that question, we need a better understanding of the public’s per- The controversy over the Rasmussen report, ceptions of nuclear power. In particular, it will like the rest of the nuclear debate, contains many be useful to compare the public’s view of the risks elements of “disputes among experts” as of nuclear energy with the risks estimated by most characterized by sociologist Alan Mazur: argu- nuclear experts. For example, if public percep- ing past one another instead of responding to tions of risk were based on misinformation, im- what the opposing expert has actually stated; re- proved public education programs might be an jecting data that develop the opponent’s case; appropriate response. However, this does not ap- interpreting ambiguous data differently; and, con- pear to be the case. Ch. 8—Public Attitudes Toward Nuclear Power ● 221

FACTORS INFLUENCING THE PUBLIC’S VIEW OF NUCLEAR SAFETY Perceptions of Risk and Benefit of activities in 1978 and 1979. Nuclear power was judged to be more risky than most of the other Studies of risk perception reveal a gap between activities and technologies, including drunk driv- lay people’s judgment of nuclear hazards and the ing, transporting chlorine by freight train, and risks estimated by technical risk assessments. In working as a big-city policeman (74). a 1979 study by Decision Research two small groups of informed people in Eugene, Oreg. (col- Several factors appear to enter into people’s lege students and members of the League of views of nuclear power as particularly risky. First, Women Voters) were asked to compare the ben- respondents in both the Netherlands and Oregon efits and risks of a variety of activities, ranging were concerned about the size of a potential ac- from smoking to vaccinations to swimming. The cident and the lack of individual control in pre- benefits of nuclear power were viewed as negli- venting an accident. In the Oregon study, nuclear gible, and the risks were judged to be almost as risks also were seen as “unknown to the public great as motor vehicle accidents, which claim and to science” and as particularly severe and about 50,000 lives each year (68). dreaded. Both the Oregon study and opinion surveys show that about 40 percent of the American Although the two groups estimated that the public believe that a nuclear plant can explode number of deaths from nuclear power in an av- like an atomic bomb, even though such an ex- erage year would be fewer than the number of plosion is physically impossible. Familiarity also deaths from any of the other activities or tech- played a role in people’s judgments. In the Ore- nologies, they used a very high multiplying fac- gon study, nuclear risks were perceived as greater tor to indicate how many deaths would occur in because they were unfamiliar. Another factor en- a “particularly disastrous” year. Almost 40 per- tering into risk perceptions was people’s difficulty cent of the respondents estimated more than in assessing the probability of a reactor accident. 10,000 fatalities would occur within 1 year, and more than 25 percent guessed there would be People’s opinions about nuclear power and 100,000 fatalities. Many of the respondents ex- other “hazardous” activities and technologies are pected such a disaster within their lifetimes, while not determined by perceptions of risk alone. The the Sandia study suggests that there is only one perceived benefits offered by a technology must U.S. reactor site–Salem, N.J.–at which an acci- be weighed against the perceived risk in deter- dent might cause as many as 100,000 “early” fa- mining how acceptable the technology is. Most talities and estimates that these consequences activities, including development of nuclear pow- have only a 10-6 or 0.000001 chance of occur- er, are undertaken initially in order to achieve ring at that site. That analysis suggests that the benefits, not avoid losses, and for many activities average American (who does not live near that the expected benefits far outweigh the potential site) has an even lower probability of being killed losses. within a year of a reactor accident. In general, in the case of nuclear power, perceptions of it appears that public perceptions of the possi- benefit may have played an important role in the ble consequences of an accident correspond trend of public opinion. During the 1950’s and somewhat with the findings of the most recent 1960’s, when electricity demand was growing technical studies, but that the probability of such rapidly, the development of nuclear energy was consequences is greatly overestimated. promoted as a means to meet future demand and Data from the Netherlands confirm the public’s there was little apparent opposition to the tech- perception of nuclear power as uniquely hazard- nology. As electricity demand slowed in the ous. Over 700 adults of varying ages, living at 1970’s and 1980’s some people may have seen varying distances from industrial activities were nuclear power as less vital to economic growth, asked to judge the “riskiness” of a wide range so that concerns about risk became more prom i- 222 . Nuclear Power in an Age of Uncertainty nent in their assessments of the technology. of nuclear power. instead, they argue, people re- Analysis of recent survey data indicates that act to nuclear energy on a purely emotional basis. judgments of “beneficiality” currently have a For example, psychiatrist Robert DuPont argues strong influence on Americans’ acceptance of that public concern about nuclear power is a nuclear power. After safety, the second most im- “phobia” resulting from irrational psychological portant factor in support for nuclear power ap- factors (1 7). Geographer Roger Kasperson cites pears to be the belief that nuclear powerplants frequently voiced concerns about genetic dam- are necessary to reduce American reliance on age and cancer as evidence of the “emotional foreign oil (57). roots” of opposition to nuclear power, and psychiatrists Philip Pahner and Roger Lifton have In both the Netherlands and the United States, suggested that fears of radiation from nuclear people living near to nuclear plants have been weapons have been “displaced” or “extended” more receptive to the technology. However, to nuclear power (30,36,55). while the Netherlands study appeared to indicate a resigned acceptance of the risks of nuclear Although there can be little doubt that emotion- power, some surveys in the United States indicate al factors enter into the public’s assessment of that those living nearby are more aware of the nuclear energy, further analysis of the public’s benefits. For example, a majority of people liv- view of risk indicates that the reasoning behind ing near Portland General Electric’s (PGE’s) Tro- these opinions is more rational than first appears. jan Nuclear Station continued to approve of the First, while people may be inaccurate in their as- plant following the accident at TMI in 1979, while sessments of the probability of a catastrophic customers throughout the entire PGE service ter- nuclear accident, they do not appear to overesti- ritory were ambivalent (7). The primary reason mate the seriousness of such a catastrophe. Both cited for local support of the plant was that it pro- proponents and opponents of the technology duced needed power. Similarly, residents of the have an equally negative view of the deaths, ill- town closest to Maine Yankee nuclear station, nesses and environmental damage that would re- who benefit from the jobs and taxes provided by sult from a reactor accident (66). People who are the plant, continued to support the reactor concerned about nuclear safety do not view a through two statewide referendum votes in 1980 radiation-induced death from a nuclear plant ac- and 1982 which would have shut the plant down. cident as significantly worse than a death from Defeat of the two referendum votes appears to other causes, and they do not perceive genetic be based primarily on the perception of Maine effects or other non-fatal consequences of such voters that Maine Yankee provides needed low- an accident as worse than death. The central area cost electricity (see Case Studies). of disagreement between the experts and the concerned public lies in the area of greatest Despite these favorable local attitudes toward uncertainty even among the experts: the prob- some nuclear plants, opinion polls at other plants ability and impacts of a major accident (21). show that local support shifted to majority opposition following the accident at Three Mile Island Secondly, lay people appear to rely on some- (24). Analysis of survey data at one host commu- what logical internal “rules of thumb” in assess- nity suggests that Federal safety standards are now ing the magnitude of various risks. As shown in seen as being too weak. It appears that national figures 40 and 41 people’s assessments of the risks events which increase perceived risk can offset associated with various diseases and technologies local perceptions of benefit. correlate fairly well with statistical estimates of the risks. While the relative riskiness of the vari- Psychological Factors ous activities was judged somewhat accurately, respondents in the Decision Research study tended The apparent gap between technical studies of to overestimate the risks of low-frequency events. nuclear power risks and people’s perceptions of According to the research team, this error results those risks has led some observers to suggest that from people’s assumption that an event is likely there is little thought involved in the public’s view to recur in the future if past instances are easy Ch. 8–Public Attitudes Toward Nuclear Power • 223

Figure 40.— Relationship Between Judged Frequency and the Actual Number of Deaths per Year for 41 Causes of Death

disease I I I I

I 1

1 u 1 10 100 1,000 10,000 100,000 1,000,000 Estimated number of deaths per year

NOTE Respondents were told that about 50,000 people per year die from motor vehicle accidents. If judged and actual frequencies were equal, the data would fall on the straight line. The points and the curve fitted to them represent the averaged responses of a large number of lay people. While people were approximately accurate, their judgments were systematically distorted. To give an idea of the degree of agreement among subjects, vertical bars are drawn to depict the 25th and 75th percentile of individual judgment for botulism, diabetes, and all accidents. Fifty percent of all judgments fall between these limits. The range of responses for the other 37 causes of death was similar.

SOURCE: SIovic, et at. (68). Reproduced by permission of P. Slovic.

to recall. Moreover, the “availability” of an event These findings help explain the increased pub- in people’s memories may be distorted by a re- lic opposition to nuclear power reported in opin- cent disaster or vivid film. Because life is too short ion polls over the past decade. Nuclear power’s to actually experience all the hazards shown in historic connections with the vivid, imaginable figures 40 and 41, people tend to focus on dangers of nuclear war lead people to associate dramatic and well-publicized risks and hence to the technology with catastrophe. Accidents such overestimate their probability of occurrence (68). as the fire at Browns Ferry and the near-meltdown 224 ● Nuclear Power in an Age of Uncertainty

Figure 41 .—Expert and Lay Judgments of Risk Plotted Against Technical Estimates of Annual Fatalities

I I I I I

Technical estimate of number of deaths per year NOTE: Each point represents the average responses of the participants. The broken lines are the straight lines that best fit the points. If judged and technically estimated frequencies were equal, the data would fall on the solid line. The experts’ risk judgments are seen to be more closely asociated with technical estimates of annual fatality rates than are the lay judgments. SOURCE: Slovic, et al., (68). Reproduced by permission of P. Slovic.

Ch. 8—Public Attitudes Toward Nuclear Power ● 225

Photo credit: William J. Lanouette In May 1979, 2 months following the Three Mile Island accident, there was a demonstration of about 200,000 people on the U.S. Capitol Grounds. Speakers urged the U.S. Congress to curtail further nuclear construction at TMI have added to the image of disaster in peo- ly calculated estimates of annual fatalities asso- ple’s minds. The publicity surrounding these ciated with various risky technologies and activ- events, movies and books such as “The China ities (21 ). However, while the experts knew more Syndrome” and We A/most Lost Detroit, and the facts, their risk assessments also were found to estimates of deaths in various safety studies have be greatly influenced by personal judgment. further enhanced the “availability” of nuclear Thus, expert studies also are subject to errors, in- power hazards in people’s minds, creating a false cluding overconfidence in results and failure to “memory” of a disaster that has never occurred consider the ways in which humans can affect at a commercial nuclear reactor. Public educa- technological systems. This latter problem was tion about nuclear safety systems, by identifying demonstrated clearly during the TMI accident, the various hazards those systems are designed which was caused in part by human error. to guard against, may only serve to increase the Because technical experts, like the general pub perceived risks of the technology. Iic, face limitations in evaluating the risks posed The Decision Research analysts also compared by nuclear power, it appears that there may be lay judgments of risks with the judgments of na- no single right approach to managing the tech- tionally known professionals in risk assessment nology. instead, it is most appropriate to involve (see fig. 41). The judgments of experts were much the public in order to bring more perspectives closer than those of the lay people to statistical- and knowledge to bear on the problem. There 226 ● Nuclear Power in an Age of Uncertainty are at least two other reasons for involving the important step in reducing public fears of nuclear public in decisions about nuclear energy. First, power is improved management of operating re- without public cooperation, in the form of politi- actors to eliminate or greatly reduce accidents cal support, observance of safety rules, and rea- and other operating difficulties. Even though ac- sonable use of the court system, nuclear power cidents at commercial reactors in the United cannot be managed effectively. Second, as a States have never caused a civilian death, the democracy, we cannot ignore the beliefs and public views both accidents and less serious desires of our society’s members (2 I). events at operating reactors as precursors to a catastrophe. An accident with disastrous conse- While public perspectives already are reflected quences already is viewed as being much more to some extent in NRC decisions, further efforts likely than technical studies and experts project, could be made to involve the public. More re- and any continuation of accidents or operating search could be conducted to define and quanti- problems will tend to confirm that perception. fy public opinion, and dialog with nuclear critics Approaches to increasing public acceptance are could be expanded with more attention paid to discussed in the conclusion of this chapter. the substance of their concerns. Perhaps the most

VALUES AND KNOWLEDGE Which is More Important? tionship between knowledge about nuclear power and support for the technology among the gen- Some analysts of public opinion have argued eral public. These findings supported a “selec- that basic values—those things that people view tive perception” hypothesis in which those as most morally desirable—play a relatively small strongly favoring or strongly opposing nuclear role in influencing people’s attitudes toward nu- energy selected and used information to bolster clear power. For example, Mazur has argued that their arguments. Attitudes toward nuclear power most of the general public, unlike energy activists, among all respondents were influenced heavily do not “embed their positions for or against a by preexisting political beliefs and values (58). technology in a larger ideological framework of These results could help to explain why the ac- social and political beliefs” (41). In addition, cident at TMI appeared to have little impact on nuclear advocates sometimes suggest that it is some people’s opinions about safety. For those primarily a lack of knowledge which leads peo- who already were firmly convinced that nuclear ple to oppose nuclear energy, and that better ed- power was safe, the accident confirmed the ef- ucation programs would increase public accept- fectiveness of safety systems. For those who were ance (1 3,42). However, as discussed below, the skeptical, it reinforced uncertainties. available evidence calls both arguments into question. A recent analysis of national survey data provides additional evidence that people’s values Although energy experts who are very knowl- and general orientations may be stronger deter- edgeable about nuclear power generally support minants of nuclear power attitudes than specific the technology, studies of the effects of slightly knowledge about energy or nuclear power issues. increased knowledge on attitudes among the In this study, sociologist Robert Mitchell tested broader public have yielded mixed conclusions. the strength of the correlation between various Two studies found greater support for nuclear “irrational” factors—such as belief that a nuclear power among more knowledgeable persons, but plant can explode–and people’s assessments of another found the opposite, and several studies reactor safety. While the analysis showed that the have found no significant relationship (46). For public was generally misinformed about energy example, a 1979 survey conducted just prior to issues, this lack of knowledge appeared to have the TM I accident revealed only a very weak rela- little effect on attitudes toward nuclear safety and Ch. 8—Public Attitudes Toward Nuclear Power Ž 227 hence on overall attitudes toward nuclear power. enthusiasm is reflected in the current computer When Mitchell went onto test the correlation be- boom and in the emergence of a flood of science tween values and attitudes toward reactor safe- magazines such as Omni, Discover, Science 83, ty, he found a much stronger relationship. Envi- and Technology as well as new television pro- ronmentalism was associated closely with con- grams including “Cosmos,” “Life on Earth,” and cern about nuclear safety among women, while “Nova,” and the reliance on high technology by skepticism about whether the future benefits of both major political parties. However, this sup- scientific research would outweigh the resulting port is tempered by a growing concern about the problems appeared to have a strong influence on unwanted byproducts of science, including accel- men’s concerns about reactor safety (46). Several erating social change, the threat of nuclear war, other studies also indicate that values have played and environmental pollution. The National Opin- an important role in both the growth of the anti- ion Research Center recently compared a nation- nuclear movement and continued support for al poll of adult attitudes toward science taken in nuclear power (8,27,32). 1979 with a similar survey conducted in 1957. They found that, over the 20-year period, an Values increasing number of survey respondents believed that “science makes life change too fast” A value has been defined as “an enduring or “breaks down people’s ideas of right and belief that a specific mode of conduct or end-state wrong.” The percentage of respondents who be- of existence is personally or socially preferred” lieved that the benefits of science outweigh the (61). While all people share the same values to harms declined from 88 percent in 1957 to 70 some degree, each individual places different pri- percent in 1979 (46). This curious duality of at- orities on different values. This ordering of the titudes may help to explain the public’s ambiva- absolute values we are taught as children leads lent attitude toward nuclear power. While ac- to the development of an integrated set of beliefs, ceptance of the technology has declined, the rate or value system. Because values have their origins of change has been slow, and votes on referenda in culture and society, changes in societal expec- have demonstrated that Americans are unwilling tations can, over time, lead to changes in an in- to forego the nuclear option entirely. dividual’s value system. One of the most undesired products of modern During the 1960’s and 1970’s, the emergence science is the threat of nuclear war. Because of of new social movements both reflected and en- the technological and institutional links between couraged changes in the priority some Americans nuclear power and nuclear weapons, opposition placed on different values. Values that appeared to buildup of nuclear weapons leads some peo- to become more prominent to many people in- ple also to oppose development of civilian nu- cluded equality of all people, environmental clear energy. AEC, which developed and tested beauty, and world peace. Critics of nuclear pow- weapons after World War II, was the original pro- er (and, to a lesser extent, proponents) were suc- moter of commercial nuclear power. In the late cessful in attracting broader public support by ap- 1950’s and early 1960’s, growing public concern pealing to these emerging values, and by linking about radioactive fallout from AEC’s atomic their organizing efforts with the related peace, bomb testing provided a context for increasing feminist, and environmental movements. Until fears about the possibility of radioactive releases a convincing case is made that nuclear power is from nuclear powerplants. Some prominent sci- at least as consistent with these values as other entists spoke out against both nuclear power and energy sources, it will have difficulty gaining ac- nuclear weapons, and links developed between ceptance with those who place a high priority on groups opposing the arms race and nuclear pow- these values. er (48). However, after the United States and the Overall, Americans are very supportive of sci- Soviet Union signed the Nuclear Test Ban Trea- ence and technology, viewing them as the best ty in 1963, concerns about both nuclear fallout routes to economic progress (39). The public’s and nuclear power temporarily subsided. 228 . Nuclear Power in an Age of Uncertainty

Since the mid-1970’s, rapid international de- most Americans continue to strongly support en- velopment of nuclear power and growing global vironmental laws despite recessions and an in- tensions have led to increasing concern about the creasing skepticism about the need for govern- possible proliferation of nuclear weapons from ment regulation of business (4). nuclear power technologies. Organizers in the The role of the environmental movement in co- peace movement and nuclear critics have built alescing and leading the criticism of nuclear on this concern in an attempt to renew the early power has been well-documented. The first na- linkages between the two movements. Case stud- tional anti-nuclear coalition (National Interveners, ies of the Maine Yankee and Diablo Canyon nu- formed in 1972) was composed of local environ- clear plants indicate that concern about nuclear mental action groups (40). By 1976, consumer weapons contributed to opposition to these advocate Ralph Nader, who later became allied plants during the late 1970’s and early 1980’s (see with the environmental movement “stood as the Case Studies). titular head of opposition to nuclear energy” (30). Today, a single Federal agency–the Depart- Today, all of the major national environmental ment of Energy—still is responsible for research groups are opposed to at least some aspects of and development of both nuclear weapons and the current path of nuclear power development commercial nuclear power. in addition, some pri- in the United States. While some of the groups vate firms are involved in both nuclear energy do not have an official policy opposing nuclear and nuclear weapons. These connections en- power, their staffs stay in close communication, courage a linkage in people’s minds between the and there is substantial cooperation and support peaceful and destructive uses of nuclear energy. on nuclear energy issues (33). A list of these Due to growing concern about the rapid buildup groups is shown in table 32. of nuclear weapons, some groups critical of nu- Both sides of the nuclear debate have em- clear energy are shifting resources toward weap- phasized environmental concerns to influence ons issues. However, most local groups and na- public opinion. in the late 1960’s and early tional organizations continue their efforts to im- 1 970’s, opponents of local nuclear plants most prove the safety of nuclear power as they expand frequently pointed to specific environmental im- their focus to include nuclear weapons. The pacts, such as thermal pollution, low-level radia- linkages between environmental and energy tion, or disruption of a rural lifestyle as reasons groups and anti-nuclear weapons groups may for their opposition. During that same period, strengthen the environmental groups and help nuclear proponents increasingly emphasized the them maintain their criticism of commercial air-quality benefits of nuclear power when com- nuclear power (33). pared with coal (41 ). Today, environmentalist op- On the pronuclear side of the debate, orga- ponents of nuclear power are more concerned nizers have emphasized the importance of nucle- about broad, generic issues such as waste dis- ar power to national energy independence which posal, plant safety, weapons proliferation, and “a is in turn linked with national security. Analysis set of troublesome value questions about high of 10 years of public opinion polls indicates that technology, growth, and civilization” (30). This a view of nuclear power as an abundant Ameri- evolution of concerns is demonstrated in two can resource which could reduce foreign oil de- case studies of local opposition to nuclear plants. pendence is a very important factor in favoring In these cases, environmentalists at first did not continued development of nuclear power (57). oppose the local nuclear plant because it offered environmental benefits when compared with a One of the most important values to affect coal plant. However, those positions were later opinions on nuclear power is environmentalism. reversed (see Case Studies). A 1972 poll by Louis Harris&Associates indicated that many Americans believed that the greatest Along with environmentalism, people’s general problem created by science and technology was orientations toward economic growth appear to pollution (46). Polls taken in 1981 indicate that influence their attitudes about nuclear power. In Ch. 8—Public Attitudes Toward Nuclear Power ● 229

1971, political scientist Ronald Inglehart identified However, even among those who most strongly a shift in the value systems of many Americans supported resource utilization, so percent in- away from a “materialist” emphasis on physical dicated that no more nuclear powerplants should sustenance and safety and toward “post-materi- be built. alist” priorities of belonging, self-expression, and Views about “appropriate technology” as de- the quality of life. At that time, he hypothesized fined by the British economist, E. F. Schumacher, that this shift could be attributed to the unprece- may also affect attitudes toward nuclear energy. dented levels of economic and physical security Most members of mainstream environmental that prevailed during the 1950’s and 1960’s. groups share this view, which endorses tech- Based on analysis of more recent surveys, lngle- nologies that are inexpensive, suitable for small- hart argued in 1981 that, despite economic un- scale application, and compatible with people’s certainty and deterioration of East-West detente, need for creativity (44,64). In 1976, Amory Lovins post-materialist values are still important to many brought nuclear energy into the middle of this Americans. And, he says, those who place a high technology debate with publication of his article, priority on post-materialist values “form the core “Energy Strategy: The Road Not Taken” in of the opposition to nuclear power” (27). Foreign Affairs magazine. In that article, Lovins argued that America’s energy needs could be met Like Inglehart, psychologist David Buss and his by the “soft energy path” of conservation, renew- colleagues have observed two conflicting value able energy and other appropriate technologies, systems or “worldviews” among Americans (71). and rejected nuclear energy as unneeded, cen- Using in-depth interviews with a random sample tralizing, and environmentally destructive (37). of adults from the San Francisco area, they iden- These concerns were found important in the local tified “Worldview A“ which favors development opposition to one case study plant (see Case of nuclear power as an important component of Studies). Residents of that rural area at first ob- a high-growth, high-technology, free enterprise jected to the plant on the basis that its electricity society, and “Worldview B“ which includes con- was not needed locally, and that the locality cern about the risks of nuclear power along with should not have to bear the impacts of plant con- an emphasis on a leveling off of material and tech- struction when it would not reap the benefits. nological growth, human self-realization, and par- Later objections were based on the contention ticipatory decisionmaking. that the electricity produced would not be needed anywhere in the surrounding three States. While different priorities within Americans’ value systems appear to influence attitudes toward Advocates of the appropriate technology phi- nuclear power, it is important to recognize that losophy fear that increased use of nuclear power the public is not completely polarized. lnglehart will lead to a loss of civil liberties and individual noted that “post-materialist” values can only be freedom, and decreased world stability due to given priority when basic human needs are met, weapons proliferation. The extent to which these making both priorities essential to individuals and views have been accepted by the American pub- to American society (27). An extensive national lic is difficult to ascertain. National opinion polls survey of attitudes toward growth conducted in showing that the majority of Americans prefer 1982 indicates that the public may be develop- solar energy to all other energy sources and view ing a new perspective that includes both a desire nuclear power as the least-favored energy option to ensure opportunities for development and con- would appear to reflect such values (57). A re- cerns about environmental quality (60). In this cent survey conducted in the State of Washington survey, few respondents could be classified as shows that large majorities there share Lovins’ totally favoring either resource preservation or view that it is possible to have both economic resource utilization, and the majority appeared growth and energy conservation (54). In addition, to be quite balanced in their views on economic many people, even those skeptical of renewable growth. Those who leaned toward resource pres- energy, share Lovins’ distrust of large centralized ervation were more opposed to nuclear power organizations (utilities and the government) that than those who favored resource utilization. promote nuclear energy.

25-450 0 - 84 - 16 : QL 3 230 Ž Nuclear Power in an Age of Uncertainty

Another shift in American value systems that This growing skepticism about industry and may help to account for increased opposition to government was reinforced by the accident at nuclear power is a growing distrust of institutions Three Mile Island. Post-TMl polls indicate that less and their leaders in both the public and private than half of the public were satisfied with the way sectors. The Vietnam War and the Watergate in- the accident was handled by Pennsylvania State vestigation contributed to growing public cyni- officials and the NRC, and Americans were even cism about the Federal Government during the less pleased with the utility (General Public Util- 1960’s and 1970’s. By 1980, an extensive survey ities) and the plant designer (57). One observer of Americans revealed a dramatic gap between has described public reaction to the accident as the public and leaders in both government and “essentially a crisis in confidence over institu- industry on questions of politics, morality, and tions” (30). A feeling that the nuclear utility was the family. Religious values were found to be of being dishonest helped spark the first referendum profound importance to the majority of Ameri- to shut down Maine Yankee, and events at Diablo cans, and respondents indicated that they placed Canyon led to nationwide doubts about the credi- a greater emphasis than before on the moral bility of the nuclear industry (see Case Studies). aspects of public issues and leadership (59). A final societal change that has been closely Some early supporters of nuclear power, in- intertwined with negative attitudes toward nucle- cluding prominent environmentalists, felt bear power is the growth of the women’s move- trayed by the nuclear establishment when new ment. Public opinion polls over the past 20 years information about the uncertainties of the tech- have shown a strong correlation between gender nology became known. The nuclear industry’s and attitudes toward nuclear power: Women early denials of the possibility of accidents, and are consistently more opposed (41). While the the Government’s handling of safety studies have strength of this correlation is well-known, the contributed to the critics’ and broader public’s reasons for it are not clear. Environmental values skepticism. Some critics have expanded their ac- and having young children have been linked with tivities from examination of technical safety issues women’s opposition to nuclear power (46). Soci- to include critiquing the nuclear regulatory proc- ologist Dorothy Nelkin argues that women’s dis- ess, and groups that formerly were concerned pri- trust of nuclear power cannot be attributed to a marily with “watch dogging’ Federal agencies greater aversion to risk in general. Instead, Nel- have entered the nuclear debate. These activities, kin’s analysis of women’s magazines and the fem- and Daniel Ford’s recent book, The Cult of the inist press suggests that women’s opposition be- Atom, which focuses primarily on regulatory gins with the specific risk of cancer in the event “misdeeds” in the early nuclear program, may of a major radioactive release from a reactor. per- contribute to the public’s disillusionment with sonal value priorities, including some women’s government in general and the NRC in particular. view of themselves as nurturers or “caretakers of life, ” also lead them to oppose what they view The American public’s growing concern with as a life-threatening technology (49). leadership applies to business as well as government. Americans increasingly are skeptical of the These connections have helped to bring nu- ability of both the public and private sectors to clear power as an issue into the mainstream of produce quality work. According to Loyola Uni- the women’s movement. Women’s magazines versity professor of business ethics Thomas Don- ranging from Redbook to Ladies Home Journal aldson, survey data indicate that, despite an im- to Ms. have questioned nuclear safety, and the proving corporate record, the public has become national Young Women’s Christian Association increasingly disappointed with corporate ethics (YWCA) took a public stand against nuclear pow- over the past 20 years. Corporations now are er in 1979. The League of Women Voters has de- viewed as “part of the overall social fabric that veloped a national policy favoring only limited relates to our quality of life,” not merely as pro- construction of new reactors, and the League’s viders of goods and services (1 5). local affiliates have taken even stronger anti- Ch. 8—Public Attitudes Toward Nuclear Power . 231

nuclear stands. In 1980, the National Organiza- orities could change again in the future. In addition for Women (NOW) recommended a resolution to values, knowledge about accidents at nu- tion opposing the “use of nuclear power in favor clear plants has played a major role in shaping of safer energy methods” (49). public attitudes. In the future, new information on improved management of nuclear power While changing value priorities appear to have could lead to a reversal of the current trend of contributed to increasing public concern about increasing opposition to the technology. nuclear power over the past 20 years, those pri-

THE ROLE OF THE MEDIA Both the amount and type of news coverage early 1960’s, most coverage was still positive, but have played an important role in shaping public a few protests against local plants—particularly attitudes toward nuclear power. As noted pre- the large demonstration at a proposed nuclear viously, people tend to overestimate the prob- plant site on Bodega Bay, Calif., in 1963–re- ability of certain hazards, including nuclear ceived national publicity. During the mid-1 960’s, powerplant accidents, in part because these there was a decrease in both the number of peri- hazards are discussed frequently in the media. odical articles on nuclear power and in public opposition as measured in opinion polls. This de- The Extent of Media Coverage cline reflected a shift in public concern and media of Nuclear Power attention away from nuclear issues and toward civil rights and other domestic issues. Beginning The most detailed analysis of print media cov- in 1968, magazine articles on nuclear power in- erage of nuclear power currently available is creased to cover local plant siting disputes. Print based on the number of articles on the subject media coverage rose even higher in 1969, and indexed in the yearly Readers’ Guide to Period- opinion polls showed a similar peak of opposi- ical Literature. In this study, sociologist Al Ian tion the following year. Mazur compared trends in media attention with From 1974 to 1976, anti-nuclear activism and trends in public opinion as revealed by national media coverage again increased, with a great deal opinion surveys, numbers of plant interventions, of national publicity given to the 1976 California and size of protests. This analysis suggested the referendum. After 1976, both negative public following three hypotheses (41): opinion and media coverage fell off, then rose 1. The greater the national concern over a ma- slightly in 1978 and early 1979 and finally rose jor issue that is complementary to a partic- massively following the accident at TMI in the ular protest movement, the more easily re- spring of 1979. Trends throughout 1979 appeared sources can be mobilized for the movement, to confirm the linkage between media coverage and therefore the greater the activity of and public opinion: Public opposition rose sharp- protesters. ly immediately following the accident, subsided 2. As the activity of protesters increases, mass within 2 months as media attention diminished, media coverage of the controversy increases. and then increased slightly during October and 3. As mass media coverage of the controversy November, coinciding with media coverage of increases, the general public’s opposition the final Kemeny Commission report (41). to the technology increases. Mazur argues that opposition to a technology At the time of the first citizen intervention such as nuclear power will snowball with in- against a nuclear plant in 1956, there was a great creased media coverage, whether that coverage deal of positive mass media coverage of president is positive or negative (40). The fluoridation con- Eisenhower’s “Atoms for Peace” program. In the troversy of the 1950’s and 1960’s, like the cur-

rent nuclear power debate, involved complex Analysis of the extent of television news scientific judgments and pitted the “established coverage of nuclear power has been much more order” against advocates of local self-control. limited than analysis of print media coverage. During this period of public debate, persons ex- Television nightly newscasts made relatively lit- posed to both positive and negative arguments tle mention of nuclear power over the decade about fluoridation were more likely to oppose the preceding the accident at TMI. Within the overall practice than persons who had heard neither ar- low level of reporting, the trends were somewhat gument, and communities where there had been similar to those in the print media: Coverage in- heated debate were most likely to defeat fluorida- creased in 1970, and then dropped off again un- tion in a referendum. The prominence given to til 1976, with greater coverage between 1976 and disputes between technical experts over the risks 1979 (70). of a technology appears to create uncertainty in people’s minds, which in turn raises concern and The Content of Media Coverage opposition, regardless of the facts under discussion. If this is true, the media play a key role in Just as there can be little doubt that media encouraging public opposition by giving exten- coverage influences public opinions toward nu- sive coverage to the experts’ disputes. clear power, there also is little doubt that jour- Ch. 8—Public Attitudes Toward Nuclear Power Ž 233 nalists, like most Americans, are ambivalent ing opposing views, these stories may have over- about this technology. In a 1980 survey, similar simplified complex issues and failed to convey percentages of media personnel and the public the prevailing consensus among scientists and (about 55 percent) viewed the benefits of nuclear energy experts. Psychiatrist Robert DuPont, after power as greater than the risks, while other “lead- viewing the same 10 years of television stories ership groups” were much more supportive of used in this analysis, suggested that fear, especial- nuclear energy (59). In another study, attitudes ly of nuclear accidents, was the underlying motif toward nuclear power were measured on a scale in all of the stories (16). Another study of 6 years ranging from -9 to +9, with a higher score in- of television news stories about various energy dicating greater support for nuclear power. While sources found that the risks and problems of nu- scientists were quite supportive of the technology clear power were emphasized, coal was given with an average score of 3.34, science journalists neutral treatment, and solar power was treated were much more skeptical, with an average score euphorically (56). of 1.30, and journalists reporting on general issues While these studies suggest that television cov- for major national newspapers were slightly less erage of nuclear power emphasizes the risks of supportive of nuclear power than science journal- the technology, there is no evidence that media ists, with an average score of 1.16 (62). personnel deliberately bias their coverage of Following the accident at TMI, the Kemeny nuclear power due to personal convictions. The Commission found that the public’s right to in- Kemeny Commission found that overall coverage formation had been poorly served. Confusion of the TMI accident was balanced although at and uncertainty among the sources of informa- times confused and inaccurate. One of the big- tion combined with a lack of technical under- gest factors in inaccurate reporting at TMI was standing by the media personnel were identified found to be the lack of reliable information avail- as contributing to the problem. Many of the re- able to the media. For example, national reports porters “did not have sufficient scientific and that the hydrogen bubble inside the reactor could technical background to understand thoroughly explode within 2 days were an accurate reflec- what they heard.” As a result of these difficulties tion of the views of NRC’s Washington office. in reporting on emergencies, the commission rec- Reporters, trusting these views and wanting to ommended that all major media outlets hire and “scoop” other reporters, tended to disregard the train nuclear energy specialists and that reporters onsite NRC officials who argued that the bubble educate themselves about the uncertainties and could not possibly explode. However, overall, probabilities expressed by various sources of in- the Commission found a larger proportion of re- formation (31). assuring than alarming statements in both television and newspaper reporting of the accident. The media’s need for balance in coverage of many issues, including nuclear power may lead Media coverage of nuclear power maybe influ- to understatement of the scientific consensus that enced by the fact that journalists are trained to the technology is acceptably safe. Media person- be skeptical of news sources, including the nu- nel are expected to bring various viewpoints be- clear establishment. Informed critics have been fore the public, and in the case of a controver- successful in publicizing many cases in which the sial technology such as nuclear power, this gen- nuclear industry, the Department of Energy, and erally means quoting both an advocate and a crit- the NRC have not been completely open about ic in any given story. One analysis of television safety problems. For example, during the first 2 news coverage showed that over the decade prior days of the accident at Three Mile Island, Metro- to Three Mile Island, most news stories dealing politan Edison withheld information on the situa- with nuclear power began and ended with “neu- tion from State and Federal officials as well as the tral” statements (70). However, among the “out- news media (72). According to the Kemeny Com- side experts” appearing most frequently in the mission, the utility’s handling of information dur- stories, 7 out of 10 were critics of nuclear power. ing this period “resulted in the loss of its credibili- Thus, while meeting the requirement of present- ty as an information source” (31). Experiences 234 ● Nuclear Power in an Age of Uncertainty such as this have led reporters to be particularly industry focuses on crashes. Because the public skeptical of nuclear industry sources and look to is unlikely to view a single plane crash as an in- the critics for the other side of any given story. dication that the entire airline industry is unsafe, the airline industry is confident that all airplanes Proponents of nuclear power are likely to view will not be grounded. With no such assurances media treatment of nuclear plant safety issues as for nuclear power, the nuclear industry may view biased because of the inherent complexity of coverage of accidents as a threat to its survival. those issues. It is important that problems such as construction errors, skyrocketing costs, and Finally, it is important to note that journalists operating difficulties be reported to the public. did not create the nuclear controversy. During However, since few people (including reporters) periods of greatest public concern, their cover- understand nuclear technology well, problems age of nuclear power has increased, which in turn may appear more threatening than they actually has contributed to still greater public uncertain- are. Considerable expertise is needed to sift the ty. If the media are more critical of nuclear power facts and accurately interpret them to the public. now than they were in the 1950’s, they may be By comparison, the media are not considered reflecting public opinion as well as influencing it. anti-airplane, even though most coverage of that

WHAT WOULD IT TAKE TO INCREASE PUBLIC ACCEPTANCE OF NUCLEAR POWER? It is unlikely that utility executives will order because of the present lack of trust in government any new reactors as long as they believe that a and industry. There are some extremists on both majority of their customers oppose nuclear sides of the nuclear controversy whose opinions energy. However, a societal consensus on the will not change, regardless of the evidence placed necessity for and benefits of the technology may before them. Even more moderate citizens, who be very difficult or impossible to attain. The pre- are willing to change their opinion on the basis vious analysis indicated that the general public of new evidence, are influenced strongly by pre- and the staffs of some public interest groups are existing attitudes and values so that they may concerned about the possibility of a catastrophic “filter out” or wrongly interpret new evidence. reactor accident. They perceive that the technol- Finally, for the majority of the public, new infor- ogy offers few or no benefits compared to these mation on improvements in utility management risks. In addition, many Americans’ personal val- of nuclear power will be viewed skeptically unless ues contribute to their skepticism of the tech- presented in a manner that arouses trust and in- nology and its managers. These value conflicts terest. However, while better communications may prevent a total resolution of the current con- are needed, the first and most important step is troversy. However, attitudes might change either to make concrete improvements responding to as a result of external events (e.g., another oil em- public concerns. bargo or new research findings on the environmental impacts of coal burning) or because of improvements made internally by government and Enhance Nuclear Advantages the nuclear industry. External events cannot be Research conducted in the United States and controlled, but it is up to the nuclear establish- the Netherlands suggests that people’s judgment ment to demonstrate the safety and economic at- of a technology or activity is influenced as much tractiveness of nuclear power. by their assessment of its potential benefits as by Assuming that major improvements were made their view of its risks. There are at least three in management of nuclear power, it would still potential benefits of nuclear power that could be be difficult to communicate them to the public perceived by the public: 1) its contribution to na-

Ch. 8—Public Attitudes Toward Nuclear Power ● 235

Photo credit: William J. Lanouette Protectors’ tent at a demonstration against the Seabrook nuclear plant in May 1977 tional energy and electricity supply, 2) its poten- Under some conditions, nuclear electricity can tial cost advantage relative to other energy op- be cheaper than its major competitor: electrici- tions, and 3) the fact that safely operated nuclear ty from coal combustion. Standardized plant de- plants produce no fossil air pollution. signs, increased predictability in the licensing It is difficult for many Americans to see a need process, and improved management of operating for nuclear powerplants at a time when electricity reactors all could help to realize the technology’s economic potential. New rate regulation systems demand has slowed. While this slow growth is expected to continue over the next several years, also could be used to reduce the initial costs of new powerplants of some kind still will be new nuclear and coal powerplants to the con- needed in the years ahead. Regions experienc- sumer. Assuming all of these changes took place and nuclear electricity did indeed offer long-term ing rapid economic and population growth will cost advantages, public opposition to new plants need new capacity sooner than others. Plans could be developed at a regional level to evaluate in hearings before State PUCs very likely would be reduced. the alternatives to meet demand growth. The planning process itself could become a vehicle However, coal is not the only alternative to for public participation, and any long-term cost nuclear power in meeting national energy needs. advantages of nuclear power could be most clear- Conservation, oil shale, and renewable energy ly demonstrated to the public this way. resources all can be used to match energy sup- 236 ● Nuclear Power in an Age of Uncertainty ply and demand but widespread application of mation, and a number of organizations are in- these technologies could be expensive. In Maine, volved in public awareness campaigns. public rejection of a 1982 referendum to shut down Maine Yankee was based in part on recog- Reduce Concerns Over nition that conservation and renewable energy Nuclear Accidents could not quickly make up for the inexpensive nuclear power lost by the shutdown (see Case While increased awareness of nuclear power’s Studies). benefits might decrease concerns about risk, one of the most favorable things that could happen Paradoxically, accelerated R&D on these alter- to the nuclear industry over the next 10 years natives might enhance the image of nuclear pow- would be an increasing output of nuclear elec- er and could confirm that they may never be tricity along with an absence of events causing widely competitive. As part of accelerated R&D bad publicity. Presently, both TMI-type accidents on alternative energy sources, the environmen- and incidents such as the failure of the safety con- tal costs and benefits of each source should be trol system at the Salem, N. J., plant are viewed examined. Environmental groups currently are by the public as precursors to a catastrophe. among the leading critics of nuclear power. These Given the slow rate at which public support for groups also are very concerned about the adverse nuclear power has declined, an extended period impacts of coal combustion and other energy of quiet, trouble-free operations could have very sources, and are monitoring research into those positive impacts on public attitudes. Chapter 5 impacts. If this research indicated that acid rain identifies a number of approaches to improved was a more serious problem than presently per- utility management of nuclear power, which, if ceived or that carbon dioxide buildup would re- implemented, could help to assure that neither sult in near-term climatic changes, some environ- major accidents nor precursors take place. mentalists might become less negative about nu- While a period of uneventful operation of nuclear relative to coal. This shift, in turn, could clear plants is necessary to restore public con- change attitudes among the broader public. fidence, it probably is not sufficient. Maine Yankee has had very high reliability but State Public relations or educational programs are voters have twice come close to shutting it down unlikely to increase public awareness of nuclear (see Case Studies). In addition, critics probably power’s potential benefits until those benefits are would remain skeptical. It would be important apparent. This might result either from events out- to demonstrate to them that the period of quiet side the industry’s control which decrease avail- operation was a result of real improvements and ability of alternative energy sources (e.g., an oil the beginning of a new trend, rather than just disruption) or from improvements in manage- luck. However, given the present level of distrust ment of the technology. Without such actions, between interveners, the NRC, and the nuclear public relations programs such as the current industry, it might be very difficult to do this. Committee for Energy Awareness campaign may have little impact, and possibly even a detrimen- Several steps could be taken to improve com- tal effect on public opinion. The response to this munications between the nuclear community campaign from critics may increase public uncer- and public interest groups critical of nuclear tainty and skepticism. Even programs viewed as power. An effort might be made to identify the unbiased by all sides, such as the League of concerns of particular groups and respond to the Women Voters Education Fund’s (LWVEF) “Nu- substance of those concerns. For example, some clear Energy Education Program” carried out in groups currently are concerned about insurance. 1980 and 1981, may do little to increase public A compromise on this issue might not decrease acceptance until the costs of new nuclear pow- the groups’ fundamental criticism of nuclear safe- erplants are better controlled (34). Nevertheless, ty, but it could improve the climate and allow the low level of public understanding of nuclear further negotiations to take place. If the current technology does indicate a need for more infor- heated debate could become a reasoned ongo- Ch. 8—Public Attitudes Toward Nuclear Power ● 237

ing dialog, the public might be less likely to view so unlikely as to be “impossible.” As a result, the technology as unsafe. As discussed previous- when the accident did happen, it greatly reduced ly, the prominence and stridency of the debate the credibility of the regulators and the industry. currently increases public uncertainty and en- The nuclear establishment should acknowledge courages opposition. that both operating events and more serious accidents can occur, attempt to educate the public As part of this effort, the Federal Government about the difference between the two, and dem- could actively encourage involvement of respon- onstrate its preparedness to deal with accidents. sible interveners in both regulatory proceedings For example, TVA immediately reports to the and long-range planning efforts through funding media any event that could be considered news- and other support. In Ontario, Canada, the inde- worthy. This very open approach increases the pendent Porter Commission funded knowledge- utility’s credibility with both the media and the able nuclear critics to conduct studies and par- public. Another positive example is offered by ticipate in extensive hearings as part of its long- a Midwestern utility that encountered quality-as- term electricity planning. In the Commission’s surance problems during construction. Once the interim report, health problems caused by im- company had greatly increased its construction proper disposal of uranium mine tailings were management capabilities, it launched a public re- identified, and environmental groups were ac- lations effort to educate the public about the knowledged for bringing the issue to the public’s problems and the steps it had taken to overcome attention. Similarly, based on testimony from them. These efforts appear to have increased leading critics, the commission found that the local trust in the utility. (See Case Studies.) probability of a loss of coolant accident causing a meltdown at Ontario Hydro’s heavy water reac- Two approaches to siting policy might help tors was much greater than the Canadian nuclear alleviate the public’s safety concerns. Both re- industry had claimed. Because the Commission spond to the public’s opposition to construction not only sought critics’ concerns but also ac- of new plants near where they live. First, as knowledged and responded to them, the process discussed earlier, some people living near nuclear had the effect of moderating some groups’ anti- plants tend to view them as less risky than peo- nuclear positions (see vol. II). ple who are less familiar with the technology. Previous U.S. efforts to involve government, in- While some polls show increasing opposition to dustry, and environmentalists in dialog or “en- nearby plants since the accident at Three Mile vironmental mediation” provide another model Island, support for other plants has remained for improved communication. Nonprofit organi- high. This fact has led Alvin Weinberg and others zations such as the Conservation Foundation in to promote a “confined siting” policy, under Washington, D. C., as well as several private firms which most new reactors would be added to ex- have brought all three parties together to discuss isting sites, rather than creating new sites. This topics such as radioactive waste disposal and approach has been used successfully in Canada chemical waste management. By careful staff (see vol. Ii) and is supported by some U.S. en- preparation and beginning the discussions with vironmental groups. It is most attractive in the a common objective (e.g., safe disposal of toxic East, where high population density makes re- wastes), these forums have succeeded in devel- mote siting infeasible. oping preliminary agreements on Federal and in- The second approach to dealing with local op- dustry policy. position to new construction is to site new reac- Nuclear regulators and the industry can in- tors at remote locations. This approach could increase their credibility with both interveners and corporate “confined siting, ” with new reactors the public by emphasizing candor in their public clustered at existing remote sites. Alternatively, information programs. Prior to the accident at new sites in remote areas could be identified. In Three Mile Island, the nuclear establishment either case, public opposition to such plants created the impression that such an accident was could be expected to be much less than opposi-

238 • Nuclear Power in an Age of Uncertainty

tion to new plants in densely populated areas. of the size of the plant construction program as Opinion polls show that the majority of the public a guarantee of Federal support for nuclear energy. favors remote siting of reactors, and the poten- However, if this definition of size were viewed tial impacts of a major accident would be re- as an absolute limit on the program, rather than duced greatly by this approach. However, the a target to be reached, the public might view it costs of transferring the power to load centers as an indication of Government skepticism of nu- would be much greater, and construction in re- clear power. A less drastic alternative would be mote areas might lead to adverse “boomtown” to limit the rate of growth in total nuclear capacity effects on nearby communities. by limiting the number of new construction permits granted in any one year. Current demand Public fears of a nuclear accident also might projections indicate that rapid growth of nuclear be reduced by controlling the rate of new plant power is unlikely for many years, but a limit might construction. Nuclear critics, fearing the impacts provide reassurance to those who feel the only of potential accidents and the possibility of a cen- choices are to eliminate the option now or for- tralized, undemocratic “nuclear state,” base their ever risk an uncontrolled resurgence. opposition in part on the rapid scaling up in size and number of reactors in the 1960’s and 1970’s After years of debate, Sweden passed a referen- and on the industry’s early projections of a “plu- dum in 1980 calling for completion of the 12 tonium economy.” These concerns might dimin- nuclear-generating units then under construction ish if the nuclear program were bounded. Some or planned, with a phaseout after 25 years (the within the nuclear industry also favor a definition expected lifetime of the plants). While this com- Ch. 8—Public Attitudes Toward Nuclear Power ● 239

promise might seem to offer little future for Minimize Linkage Between nuclear power, it did allow construction of six Nuclear Power and Weapons new nuclear plants, with the result that over half the country’s electricity is now nuclear. In the Another issue that should be addressed in pol- United States, a compromise under which regu- icy decisions about nuclear power is the connec- lators and nuclear critics agreed to encourage tion between weapons development and civilian completion and operation of units currently nuclear energy. Given the level of national con- under construction or planned might be prefer- cern over the arms race, public acceptance of able to the current impasse, especially in terms nuclear power cannot be expected to increase of financial return to investors. It has been sug- substantially until the two nuclear technologies gested that Americans might reach consensus on are separated in people’s minds. This report has a 150-gigawatt nuclear program (29). Similarly, not analyzed the impact of policies that might the advocacy arm of the League of Women Vot- minimize the linkages between nuclear weapons ers has adopted a national policy calling for a con- and power, but the effect on public opinion could tinuation of nuclear power in its current percent- be positive. For example, one action that might age of national energy supply. As energy and increase public acceptance by reducing the per- electricity demand grow in the future, this policy ceived linkages would be to remove nuclear would allow some growth in the nuclear pro- weapons development from the jurisdiction of gram. Any such compromise or cap would have the Department of Energy. Another step would to allow for adjustments as nuclear and com- be to legislate a ban on commercial fuel reproc- peting technologies are improved and economics essing. Many critics are more concerned about change. In addition, regional differences in the reprocessing than about reactors because plutoni- United States might make a State-by-State ap- um separated from the spent fuel might be stolen proach more feasible than a national referendum and used to construct a bomb or to threaten the as in Sweden. public. A legislated moratorium on reprocessing might have greater impact on these concerns While all of the approaches discussed above than the executive orders imposed by Presidents might decrease public concerns about current Ford and Carter that were later revoked by Presi- reactors, it is possible that public skepticism about dent Reagan. Such a ban might be especially ef- the technology is so great that these changes fective if imposed in conjunction with limits on would have little impact. In this case, other reac- the total growth of the program, as discussed tor concepts with inherent safety features might above. In addition, it might be best to keep in- be considered. Several alternatives, such as the dustry and military waste disposal strictly high temperature gas-cooled reactor (HTGR), the separate, although some public interest groups heavy water reactor (HWR), and an improved support joint disposal, because it encourages ac- light water reactor (the PIUS reactor) are dis- tion on military waste that has been allowed to cussed in chapter 4. A substantial federally accumulate for 40 years (53). funded R&D effort on one or more of these alternatives might meet with public acceptance, par- Policy makers also could take action to reduce ticularly if demand for power picks up over the the possibility of weapons proliferation through next two decades. The inherent safety features careful management of international nuclear of the chosen design might appeal to the general power development. A previous OTA analysis public and the choice of design could be used identified weaknesses in the existing international as a vehicle for much greater involvement of nu- nonproliferation regime (52). Recognizing the im- clear critics. By bringing critics into the R&D pro- pact these weaknesses have on public percep- gram and addressing their specific concerns, con- tions of nuclear power, Alvin Weinberg has sensus might be reached on an acceptable design argued, “We must strengthen the Nonprolifera- for future reactors. tion Treaty (NPT) regime and take the next steps, 240 ● Nuclear Power in an Age of l.uncertainty which involve both a reduction in nuclear arma- supportive of civilian nuclear power, reaches an ments and a strengthening of the sanctions that audience who might otherwise reject nuclear can be imposed on those who would violate the power along with weapons (48). NPT” (75). In conclusion, current public attitudes toward The link between nuclear weapons and nuclear nuclear power pose complex problems for the power also might be reduced in people’s minds nuclear industry and policy makers. However, if more proponents of nuclear power who op- technical and institutional steps could be taken pose the continued buildup of nuclear weapons that might lead the public to view nuclear power stated their beliefs publicly. For example, Hans as an important and attractive energy source in Bethe, a prominent nuclear physicist who has the years ahead. Constructive leadership and been active in the arms control movement and imagination will be required to start this process. Ch. 8—Public Attitudes Toward Nuclear Power ● 241

tefthe Referendum Committee, a business COW ~ erencjum proposed to shut down Maine Yankee sultant and Cl@ setup the Committee tO Saw .“ #t&r 5 yews, in order toabw tirnefor a smooth Maine Yankee {CSMY), raising over $800,W0.* !ramitkmtotisewationand renewable energy The funds were wed for advertisetnents, , .~*ces. CWY a~in launched a counter-carn- speakers, mdadirectmailc krnpaignemphasig- ~ p#@. ~W t~ economic advarlta~ of the fi~c~e.a. , - pbnLCSMYpr@ectedthatrf @acingMakwYa* . ~ !ii”bbvember 19$2, Maine voters again chose ‘ fgWep Maine Yankee operating, though by a kee with oil-fired &#k4ywouid add 30 pe~~ ‘ ~ cent to the ave~ residemia} customef”s bf~~+ , , .-nq~r -n {56 to 44 percent). Economics , . ..’ , -r@ the key factor in the vote. An inde- Th4? fU?f&@dlW’n C&&lit@ emphasized thm - .- -nt analysis by the Governor’s Energy Of- issueskl its@mJ?a”@n: thema@tude@apot#r& ; . ~:hadfound that ektricity prices would in- tkd. ac@(fqtK and its d@a&@tg effect on IW ~ ~ “:, i-by Igto’%i peroent if Maine Yankee were coast ofktaine; the impact ofradioactivew@S” - - ~ , #J&&J*, - a cmmtef-study by the ref- ori future generations; and the concept of k~” - ‘~.-*mWee, ‘ ~OSt VOtefs were con- franchi- in _ people have the ti@’t&ite@ensive cdl-fired generation would sibilityto make thfdrown value judgmen&a% : “ j :“ &f~~@@. M addition, the referendum com- ukar pQVVW ati OthW 155s that affect their ‘ : .: -Wtrcredibility on heakh and safety issues live. With \eSS than $IW,(X!(J tb ctim*@ ., ,, w+u~ M ~ U.S. Centers for Di$ease Control relied heavily cm the fairness doctrine to 43btqIq , ‘ the claim that leukemia rates media coverage of their views, . . *’”@@er dwt normal around Maine Yankee. . . ,. Fdowing a large Wter turnout, fheti@n- - “‘ .@!$ic at@udes towad the Maine Yankee @m c~mk k3st, by41 t~ S9 /J@ciXJt, Cl@- “ ~ ,’.; , . &!? $kMGr In many respects to national at- servers ori al! sid~ a#@d M ecOnOml C$ ~~ - .,. ~ ‘: . ~ ~ti~~ @f Wka$sett, which benefits been thq def%!ing fact~” ig $he vm. Wg Ulk ‘ r-nl= a~w ~avesW- end of~he a p#o@&r ‘ ; ,. Environmentalism, tOO, # the &lnivdmity qf”hdaine #t-0’O?W * W* ‘:. ‘ ‘ .~ *.*- hfkw!nce on concerns about the M IOWW? .* * 49@W@4!, :m ~ @@. ~w & the leaders of the referendum enW.& &@3@t* , ~mbI ‘-’; m~~ne- ~u= ~~e *@s th~ t~ *r@5e*~’r@ $T&kfi! ~flf~~ti”;””’ “ ‘ - ~; -, rqmtcharaa=.~mems about nuclear lost MaR’ieYa@eepo&@r. FroqMhenort,thqt@!$ $ : ~.= instrumental irl mobilizing opposition erendum fmrmnittee hadtrouble mfo@@n@@ ~ . ~@ee, and the nuclear weapons campai n q’ health and @@y issues &&f 4JtP, ~ : helped attract v~e~ mately Lt. : . . Another issue that M aGW@kiemxi ch-the debate was the The &@e turnout and VOt~ $@t in thi$ t!)t)ji “ - polls taken after the caRapai@n @ncol# that -e ~~e, I?eferw!ldl!m committee* @y u Ya* for eco~ic rea. &wvn because they disliked ula#’ 1962. - . .$ , - . . r, this same Yankee in- ‘~~~ to keeP the ~lant,

ik?f”~fi economic necessity, was@mfkkd * ttw m- ~. [page omitted] This page was originally printed on a dark gray background. The scanned version of the page was almost entirely black and not usable. Ch. 8—Public Attitudes Toward Nuclear Power Ž 243

cidence of a massive public protest based in part on concerns about earthquake-stance and an apparent industry failure to provide that resistance appeared to many people to vindicate the - Nuclear protectors. In addition, PG&E’s handling of the @@3!k {Mew Writ: Merrill event may have contributed to doubts about the R@wswwI at Diablo nuclear industry’s credibility. When an independent study confirmed the quality-assurance ~~;”;uv, ‘< ~ , - : ‘ “ ~ problems with the seismic restraints, a PG&E 6. ‘~bhfamh Fomi@ht- spokesman dismissed the finding as being"a *, 0ct*22*T%#?~ *;*: ~ ~ certain informality" in the utility’s “seismic sent- 7. Mmmtaih W4#-!l @#kfte, *ioeconorn- ice contracts.” Canyon ““pgi:qi$ Npcp@?@tifY-- Public reaction to Diablo Canyon has reflected .*$!@~:; . . k , national trends in several respects. Early COn- , 8. !!wQI%% Mkt and vonf.oewenfeldt, Paula, “The cerns about specific environmental impacts and :## at OiabkD Canyon,’V The Nation, Oct. 8, support among some environmentalists gave way to broader concerns about seismic risk, 9., U.S. brqpwss, House of Representatives, Sub- need for power, and, ultimately, the compe- ccMnMwe on Energy and Environment, Oversight tence and credibility of the utility. Conflicting #4b#9#g w O’abfo Canyon Nwkar Generating Mant @ne W, 1977. scientific studies of seismic risk appear to have 14). W. Qwwess, House of Representatives, Sub- increased public skepticism about the plant, and W-on Etwrgy ml Environment, Overs@ht the accident at Three Mile Island caused a sharp Hm@ng on NRC Operating Licensing Process, increase in local opposition. Mr. 9, 1981. II. ~;$.-tingress, Subcommittee on Energy ad Study References Case 2 Environment, Owersight Hearing on Quality 1. “A Seismological Shoot-out at Diablo Canyon,’t A.5&@t1i;+fitcleW-*wefplan~ C“struction, Science, vol. 214, Oct. 30, 1981. . t . 2. ‘(Diablo Canyon: Encore on Design QA,’r The Enetgy Daily, Apr. 25, 1983.

Photo credit: Pacific Gas & Electric Demonstrations and court actions called attention to issues of seismic design at Diablo Canyon [page omitted] This page was originally printed on a dark gray background. The scanned version of the page was almost entirely black and not usable. Ch. 8—Public Attitudes Toward Nuclear Power ● 245

CHAPTER 8 REFERENCES

1. “A Fresh Portrait of the Anti-Nuclear Movement,” 16. DuPont, Robert, “Nuclear Phobia–Phobic Think- Groundswe//, vol. 5, No. 4, July/August 1982. ing About Nuclear Power” (Washington, D. C.: 2. Aldrich, David, et al., Sandia National Labora- The Media Institute, 1980). tories, Technical Guidance for Siting Criteria De- 17. DuPont, Robert, “The Nuclear Power Phobia,” velopment, NUREG/CR-2239 (Washington, D. C.: Business Week, September 1981. U.S. Nuclear Regulatory Commission, December, 18. “Energy-guzzling: Most Consumers Are Cured,” 1982). Harris poll reported in Business Week, Apr. 4, 3. Alexanderson, E. P., Fermi-7: New Age for Nuc/ear 1983, Power (La Grange Park, Ill.: The American Nuclear 19. “Expert Knowledge,” Public Opinion, vol. 4., No. Society, 1979). 6, October/November 1981. 4. Anthony, Richard, “Trends in Public Opinion on 20 Firebaugh, Morris, “Public Education and Atti- the Environment, ” Environment, vol. 24, No. 4, tudes,” An Acceptable Future Nuclear Energy Sys- May 1982. tem, M. W. Firebaugh and M. J. Ohanian (eds.) 5. Atomic Energy Commission, Theoretic/ Possibil- (Oak Ridge, Term.: Institute for Energy Analysis, ities and Consequences of Major Accidents in 1980). Large Nuclear Power Plants, WASH-740 (Wash- 21. Fischhoff, Baruch, et al., “Lay Foibles and Expert ington, D. C,: U.S. Atomic Energy Commission, Fables in Judgments About Risk,” Progress in Re- 1 957). source Management and Environmental Planning, 6. Barbash, Fred, and Benjamin, Milton R., “States VOI. 3, 1981. Can Curb A-Plants,” Washington Post, Apr. 21, 22. Ford, Daniel, “The Cult of the Atom-l, ” The New 1983. Yorker, Oct. 25, 1982. 7. Beardsley & Haslacher, Inc., Nuc/ear Attitudes 23. Ford, Daniel, “The Cult of the Atom-l I,” The New After the Three Mi/e /s/and Accident (Portland, Yorker, Nov. 1, 1982. Oreg.: Pacific Gas & Electric, May 1979). 24. Freudenburg, William, and Baxter, Rodney, “Pub- 8. Buss, David M., et al., “Perceptions of Techno- lic Attitudes Toward Local Nuclear Power Plants: logical Risks and Their Management,” paper pre- A Reassessment,” paper presented at the 1983 an- sented at the Symposium on Environmental Haz- nual meetings of the American Sociological Asso- ard and People at Risk, 20th International Con- ciation, Detroit, Aug. 31 -Sept. 4, 1983. gress of Applied Psychology, Edinburgh, July 26, 25. Fuller, John G., We A/most Lost Detroit (New 1982. York: Reader’s Digest Press, 1975). 9. Cambridge Reports, Inc., American Attitudes 26, Gomberg, H. J., et al., “Report on the Possible Ef- Toward Energy Issues and the Electric Utility in- fects on the Surrounding Population of an As- dustry (Washington, D. C.: Edison Electric Institute, sumed Release of Fission Products From a 1983). 300-MW Nuclear Reactor Located at Lagoona 10. Cambridge Reports, Inc., 1982 Polls conducted for Beach,” APDA-120 (Atomic Power Development the U.S. Committee for Energy Awareness (unpub- Associates, 1957). lished). 27. inglehart, Ronald, “Post-Materialism in an En- 11. Comey, David, “The Incident at Brown’s Ferry,” vironment of Insecurity, ” American Po/itica/ Not Man Apart, September 1975. Science Review, vol. 75, No. 4, December 1981. 12. Committee on Nuclear and Alternative Energy Sys- 28, Institute of Nuclear Power Operations, Review of tems, National Research Council, Energy in Tran- NRC Report: “Precursors to Potential Severe Core sition 1985-2070 (Washington, D. C.: National Damage Accidents: 1969-1979: A Status Repott, ” Academy of Sciences, 1979). NUREG/CR-2497 (Atlanta, Ga.: INPO, 1982), p, 13. Delcoigne, Georges, “Education and Public Ac- 21. ceptance of Nuclear Power Plants, ” Nuc/ear Safe- 29. Kasperson, J. X., et ai., “Institutional Responses ty, vol. 20, No. 6, November/December 1979. to Three Mile Island, ” Bu//etin of the Atomic Sci- 14. “Despite Nuclear Power’s Problems, The Public entists, vol. 35, No. 10, December 1979. is Convinced,” Nucleonics Week, May 5, 1983. 30. Kasperson, Roger, et al., “Public Opposition to 15. Donaldson, Thomas, “Ethically, ‘Society Expects Nuclear Energy: Retrospect and Prospect,” More From a Corporation’, ” U.S. News & Wor/d Science, Technology, and Human Values, vol. 5, Report, Sept. 6, 1982. No. 31, spring 1980.

25-450 0 - 84 - 17 : QL 3 246 ● Nuclear Power in an Age of Uncertainty

31. Kemeny, John, et al., Report of the President’s 48. Nelkin, Dorothy, “Anti-Nuclear Connections: Commission on the Accident at Three Mile Island Power and Weapons,” Bu//etin of the Atomic (Washington, D. C., 1979). Scientists, vol. 37, No. 4, April 1981. 32. Kuklinski, James, Meday, Daniel, and Kay, W. D., 49. Nelkin, Dorothy, “Nuclear Power as a Feminist “Citizen Knowledge and Choices on the Complex Issue,” Environment, vol. 23, No. 1, January/ Issue of Nuclear Energy,” American )ourna/ of February 1981. Po/itica/ Science, November 1982. 504 “News Beat,” E/ectrica/ World, June 1982. 33. Lash, Terry, “Survey of Major National Groups in- 51, “Nuclear Study Raises Estimate of Accident Tolls,” fluencing Public Opinon Against Nuclear Power,” The Washington Post, Nov. 1, 1982. OTA contractor report, April 1983. 52, Office of Technology Assessment, Nuc/ear Prolif- 34. League of Women Voters Education Fund, “Edu- eration and Safeguards—Summary, OTA-E-1 48, cating the Public About Nuclear Power: The Final March 1982. Report to the U.S. Department of Energy,” Sep- 53, Office of Technology Assessment, Managing Com- tember 1981. mercial High-Level Radioactive Waste—Summary, 35. Lewis, H. W., et al., Risk Assessment Review OTA-O-1 72, April 1982 Group Report to the U.S. Nuc/ear Regulatory 54. Olsen, Marvin, et al., Washington Citizens View Commission (Washington, D. C.: U.S. NRC, 1978). Energy Issues (Pullman, Wash.: Washington State 36. Lifton, Robert J., The Broken Connection (New University, 1982). York: Simon & Schuster, 1979). 55, Pahner, 1975, 1976, cited in Kasperson, Roger, et 37. Lovins, Amory, “Energy Strategy: The Road Not al., “Public Opposition to Nuclear Energy: Retro- Taken?” Foreign Affairs, vol. 55, No. 1, 1976. spect and Prospect, ” Science, Technology, and 38. Marshall, Elliot, “The Salem Case: A Failure of Human Va/ues, vol. 5, No. 31, spring 1980. Nuclear Logic, ” Science, vol. 220, Apr. 15, 1983. 56 Rankin, William L., and Nealey, S. M., A Com- 39. Martin, L. John, “Science and the Successful Soci- parative Analysis of Network Television News ety,” Pub/ic Opinion, vol. 4, No. 3, June/July Coverage of Nuclear Power, Coal, and Solar 1981. Stories (Seattle, Wash.: Battelle Human Affairs 40, Mazur, AlIan, “Opposition to Technological in- Research Centers, 1979). novation,” Minerva, vol. X111, No. 1, spring 1975. 57. Ran kin, William L., et al., Nuc/ear Power and the 41, Mazur, AlIan, The Dynamics of Technical Con- Public: An Update of Collected Survey Research troversy (Washington, D. C,: Communications on Nuc/ear Power (Seattle, Wash.: Battelle Human Press, 1981). Affairs Research Centers, December 1981 ). 42, Mills, Mark & Donnamarie, “Activities of Groups 58. Reed, John, and Wilkes, John, “Nuclear Knowl- Which Influence Public Opinion in Favor of Nu- edge and Nuclear Attitudes: An Examination of in- clear Power,” OTA contractor report, April 1983. formed Opinion,” draft presented at AAAS Meet- 43, Minarick, M. W., and Kukielka, C. A., Precursors ing, San Francisco, Calif., 1980. to Potential Severe Core Damage Accidents: 1969- 59. Research and Forecasts, Inc., The Connecticut 1979 (Oak Ridge, Term.: Science Applications, Mutual Life Report on American Values in the Inc., 1982). 80’s: The Impact on Be/ief (Hartford, Corm.: Con- 44. Mitchell, Robert C., “How ‘Soft,’ ‘Deep,’ or ‘Left?’ necticut Mutual Life, 1981 ). Present Constituencies in the Environmental 60. Research and Forecasts, Inc., The Continent/ Movement for Certain World Views,” Natura/ Group Report: Toward Responsible Growth Resources Journa/, vol. 20, April 1980. (Stamford, Corm.: The Continental Group, Inc., 45, Mitchell, Robert C., “Polling on Nuclear Power: 1982). A Critique of the Polls After Three Mile Island,” 61. Rokeach, Milton, The Nature of Human Va/ues Po//ing on the Issues, Albert Cantril (cd.) (Cabin (New York: The Free Press, 1973). John, Md.: Seven Locks Press, 1980). 62. Rothman, Stanley, and Lichter, S. Robert, “The 46, Mitchell, Robert C., “Rationality and Irrationality Nuclear Energy Debate: Scientists, The Media and in the Public’s Perception of Nuclear Power, ” The Public,” Public Opinion, vol. 5, No. 4, discussion paper presented at AAAS Annual Meet- August/September 1982. ing, January 1982. 63. Sandia National Laboratories, statement of Nov. 47. “NRC Statement on Risk Assessment and the 2, 1982. Reactor Safety Study Report (WASH- I 400) in Light 64. Schumacher, Ernst F., Sma// is Beautifu/ (New of the Risk Assessment Review Group Report,” un- York: Harper & Row, 1973). published NRC Document, Jan. 18, 1983. 65. Sholly, Steven, “Consequences of a Nuclear Reac- Ch. 8—Public Attitudes Toward Nuclear Power ● 247

tor Accident,” F/uc/eus, vol. 4, No. 4, winter 1983, mittee on Interior and Insular Affairs, Reactor Safe- p. 4. ty Study (Rasmussen Report) (Washington, D. C.: 66. Slovic, Paul, et al., “Images of Disaster: Percep- U.S. Government Printing Office, June 1976). tion and Acceptance of Risks From Nuclear 72. U.S. Congress, House of Representatives, Com- Power,” Proceedings of the 15th Annual Meeting, mittee on Interior and Insular Affairs, 97th Cong., National Council on Radiation Protection and 1st sess., Reporting of Information Concerning the Measurements: Perceptions of Risk (Bethesda, Accident at Three Mile Island (Washington, D. C.: Md.: NCRP, 1980). U.S. Government Printing Office, 1981). 67. Slovic, Paul, et al., “Perceived Risk and Quanti- 73. Van Loon, Eric, “Executive Director’s Message, ” tative Safety Goals for Nuclear Power, ” Transac- Nuc/eus, vol. 4, No, 4, winter 1983. tions of the American Nuc/ear Society, No. 35, 74. Vlek, Charles, and Stallen, Pieter-Jan, “Judging 1980. Risks and Benefits in the Small and in the Large,” 68, Slovic, Paul, et al., “Rating the Risks: The Struc- Organizational Behavior and Human Perform- ture of Expert and Lay Perceptions,” Environment, ance, No, 28, 1981. vol. 21, No. 3, April 1979. 75. Weinberg, Alvin (cited in M. W. Firebaugh and 69. “Spectral Lines: Opinions on Nuclear Power, ” M. J. Ohanian (eds.), An Acceptable Futu;e Nu- IEEE Spectrum, vol. 16, No. 11, November 1979. c/ear Energy System (Oak Ridge, Term.: Institute 70. Theberge, Leonard, Television Evening News for Energy Analysis, 1980). Looks at Nuc/ear Energy (Washington, D. C.: The 76 Weinberg, Alvin, “Nuclear Safety and Public Ac- Media Institute, 1981). ceptance” Nuc/ear News, October 1982, 71, U.S. Congress, House of Representatives, Com-

Chapter 9 Policy Options

Photo credit: Ontario Hydro Pickering nuclear station has four identical CANDU units built in 1973 and four more scheduled for completion in 1985. Standardization helped reduce construction costs and improve operator training Contents

Page Introduction ...... , ...... , , ...... 251 Policy Goals and options ...... 252 Goal A: Reduce Capital Costs and Uncertainties ...... 252 Goal B: Improve Reactor Operations and Economics ...... ,...255 Goal C: Reduce the Risk of Accidents That Have Public Safetyor Utility Financial Impacts ...... 256 Goal D: Alleviate Public Concerns and Reduce Political Risks ...... 260 Major Federal Policy Strategies and the Likelihood of More Orders for Nuclear Plants. .263 Base Case: No Change in Federal Policies ...... 263 Base Case Variation: Let Nuclear Power Compete in a Free Market . ....,...... 270 Strategy One: Remove Obstacles to More Nuclear Orders...... 271 Strategy Two: Provide Moderate Stimulation to More Orders ...... 273 Strategy Two Variation: Support the U.S. Nuclear lndustry in Future World Trade ...275 Conclusions ...... 277 Chapter 9 References ...... ,...279

Tables

Table No. Page 34. Summary of Policy Options ...... 252 35. Major Policy Strategies ...... 264 36. Four Scenarios Affecting the Future of the Nuclear Industry ...... 265 37. Four Combinations of Economic and Management Scenarios Under the Base Case...... 267 38. Alternative Nuclear Futures Under Strategy One: Remove Obstacles to More Nuclear Orders ...... 272 39. Alternative Nuclear Futures Under Strategy Two: Provide a Moderate Stimulus to More Nuclear Orders ...... 273 Chapter 9 Policy Options

INTRODUCTION

The previous chapters have painted an un- source other than coal. The energy outlook for promising picture of the future of nuclear power the early 21st century, when oil and gas reserves in the United States. Projections for new central will become seriously depleted, is very uncertain. station generating capacity over the next 20 years If it is reasonably possible that nuclear power will are much lower than those of just a few years ago. be seen as very desirable or even indispensable The high financial and political risks involved with within 20 or 30 years, it probably is more effi- nuclear plants suggest that any central station cient to have a continuous learning curve than capacity that is added would be coal-fired. Under to try to put the industry back together when it existing conditions, there are few incentives for is needed. utilities to select nuclear plants and many reasons There are, of course, reasons for opposing these to avoid them. arguments. Even if it is conceded that it would It can be highly misleading, however, to fore- be in the national interest to have the nuclear op- cast future decisions on the basis of existing con- tion, that does not mean it is the responsibility ditions. Some of the problems that appear so for- of the Federal Government to ensure it. The eco- midable now will diminish. The plants under con- nomic penalty for not having more nuclear plants struction now were designed according to con- would not be crippling (though the total dollar cepts developed 10 to 15 years ago. Any future penalty could be quite high) (2,4), and as shown plants can be expected to incorporate major in chapters 3 and 5, unless nuclear plants are built changes that have been backfitted onto existing and operated well, they are not the most eco- plants and other changes that have been sug- nomic choices. If any more serious accidents oc- gested to improve operation. In addition, much cur, forcing long shutdowns and expensive back- has been learned about how to construct plants fits, the economics of nuclear power will be very more efficiently, While these and other changes hard to defend. Thus, it could be more produc- would go far toward eliminating the large cost tive economically to concentrate on the develop- overruns some utilities have experienced, they ment of alternative energy sources. probably are not sufficient to restore confidence Therefore, policy options presented here are in the financial viability of the technology. Other not intended to prop up a terminally ill industry, concerns exist that these changes will not ad- but to cure the problems of an industry that is dress. Therefore, it is probable that additional ini- salvageable and which the Nation decided was tiatives, including Federal actions, will be re- needed. I n addition, some of the options can be quired if the country concludes that nuclear en- of importance for ensuring that existing reactors ergy is to continue to grow past the plants now operate safely and economically regardless of the ordered. choice about the industry’s future. As recounted in chapters 1 and 3, there are rea- The next section presents a series of policy sons why the Nation could decide that it would goals and options that might be considered by be in the national interest to maintain a domestic Congress. For some of these, a lead congressional nuclear option. Nonfossil fuel energy sources role would be needed. For others, congressional may be urgently required for environmental rea- action may be no more than general policy set- sons within several decades, and nuclear energy ting and oversight because the main initiative could be the most economical source that can must arise elsewhere. be readily deployed. Even if such environmental conditions do not materialize, it could be One of the difficulties facing policy makers is economically prudent to retain a generating that few if any of these options will be very effec-

251 252 • Nuclear Power in an Age of Uncertainty tive by themselves. Actions need to be taken on to different levels of involvement to which pol- a broad front, but the responsibility for these ac- icymakers may want to commit the Government. tions is diffuse. The success of these strategies depends in turn The third section, therefore, groups the options on two other factors: the need for nuclear power according to three different strategies: first, no and how well the industry manages its present change in Government policy as it currently ap- reactors. Therefore this section also includes pears; second, remove obstacles in the way of economic and industry management scenarios further orders for nuclear plants; third, stimulate that are combined into four different futures to more nuclear orders. These strategies correspond help evaluate the strategies.

POLICY GOALS AND OPTIONS In order for nuclear power to become more Goal A: Reduce Capital Costs acceptable in general, progress must be made in and Uncertainties several different areas. Reactors must be more affordable, operations of existing nuclear plants At present, nuclear reactors pose too great a must be improved, concerns over potential ac- financial risk for most utilities to undertake. Few cidents must be alleviated, and public acceptance utilities can support such a great capital cost for must be improved. This section discusses the spe- the length of time required to build a nuclear cific policy initiatives that would contribute to plant, even if Iifecycle cost projections show that these goals. The goals and options are listed in it would be the cheapest power source over the table 34. Under each goal the options are listed lifetime of a powerplant. Not only are capital cost not by importance, but in order of ease of imple- estimates high, but the actual cost could be much mentation according to the strategies discussed higher if designs continue to change during con- later. struction. As discussed in chapters 3, 4, and 5,

Table 34.—Summary of Policy Options

Strategya Congressional role A. Reduce capital costs and uncertainties 1. Revise the regulatory process for predictable licensing ...... One Oversight, legislation 2. Develop a standardized, optimized design ...... One Moderate R&D funding (design) Two Major R&D funding (demonstration) 3. Promote the revision of rate regulation ...... Two Inquiry; FERC regulation B. Improve reactor operations and economics 1. R&D programs to improve economics of operations ...... Base Case Minor R&D funding 2. Improve utility management of nuclear operations ...... One NRC oversight 3. Resolve occupational exposure liability ...... Two Legislation c. Reduce the risk of accidents that have public safety or utility financial impacts 1. Improve confidence in safety Base Case NRC oversight; minor R&D funding 2. Certify utilities and contractors ...... Two Legislation 3. Develop alternative reactors ...... Two Major R&D funding 4 Revise institutional management of nuclear operations...... Two Inquiry, oversight D. Alleviate public concerns and reduce political risks 1. Accelerate studies of alternative energy sources ...... Base Case Minor R&D funding 2. Address the concerns of the critics ...... One Oversight, legislation 3. Control the rate of nuclear construction ., ...... One Legislation 4. Maintain nonproliferation policies ...... Two Oversight of legislation 5. Promote regional planning for electric growth ...... Two Legislation astrategies incorporating these policy options are described later in the chapter: Base Case, Strate9Y One, and Strate9Y TWO SOURCE: Office of Technology Assessment. Ch. 9—Policy Options ● 253 this situation should improve even without any Nuclear utilities are especially sensitive to back- policy changes, but probably not enough for util- fitting, which can be very costly and time-con- ities to be confident in their estimates. suming. The controversies surrounding backfits and the proposals for change have been de- Policy options intended primarily to support scribed in chapter 6. There are two related prob- this goal are discussed below. lems. One is that individual backfit orders do not always take into account the impact on other A 1. Revise the Regulatory Process parts of the plant. The second is that estimates Regulatory reform has many proponents in the of overall gains in safety are not made to weigh nuclear industry who argue that the licensing against the full cost. The prospect of ever greater process is unpredictable and unnecessarily time- costs associated with future backfits to completed consuming. Some revision may be necessary (if plants increases the uncertainty of investment in not sufficient) for a resurgence in nuclear plant nuclear power. Decisions on backfits generally orders. have been made implicitly and with little consistency. It is difficult to develop a universally ac- Efforts to change licensing will encounter dif- ceptable formula for these tradeoffs since safety ficulties, however, if they do not account for other is not easily quantifiable, and regulators are reluc- points of view. The primary purpose of nuclear tant to factor in costs if this could result in any regulation is to ensure safety. As discussed in decrease in safety. chapters 5 and 6, some utilities and contractors have not performed adequately. In such cases, Several proposals have been made to revise difficulties with regulation indicate that regula- NRC’s backfit rule and procedures. All proposed tion is working. In addition, nuclear critics ob- revisions have recommended changing NRC’s ject to any attempt to limit their participation in definition of a backfit to make it more explicit. the regulatory process, and suspect that changes In addition, it is generally suggested that threshold to enhance efficiency would reduce their effec- standards for invoking a backfit order be more tiveness in raising safety issues. Since critics have clearly identified, along with the procedure for considerable influence on public opinion, it will implementation. be difficult to achieve enough of a consensus on These changes could be accomplished through such revisions. Thus, a complete package of reg- administrative rulemaking, as proposed by NRC, ulatory change should improve the predictabili- or through legislation, as supported by the nu- ty and consistency of licensing nuclear plants clear industry and the Department of Energy while simultaneously ensuring their safety and (DOE). Legislation could make backfit decisions adequate public participation. more consistent but would have serious draw- Major proposals for legislative action concern backs if it attempted to be too precise. The tech- early approval of designs and sites, the hearing niques for quantifying safety improvements are process, combined licenses, and backfits. These still somewhat crude, and any cost-benefit anal- proposals are evaluated in chapter 6. It is likely ysis would be inherently uncertain and subject that efficiency and predictability could be en- to bias. Institutionalizing cost-benefit considera- hanced by banking designs and sites, and this tions through legislation also may reduce NRC’s change could be structured to allow adequate op- flexibility to improve the process later. Such portunity for public participation. It is less clear legislation also might be perceived by nuclear that revising the hearing process or combining critics as restricting safety improvements that construction and operating licenses would im- might be necessary even if they do not meet the prove efficiency or allow for adequate public in- cost-benefit criteria because of all the uncertain- volvement until the technology is more mature. ties in the technology. A productive Government Tighter management within the Nuclear Regu- role in this area might be to develop and refine latory Commission (NRC), perhaps with stricter risk assessment and cost-benefit analysis meth- congressional oversight, might make sufficient odologies so that they can be more confidently progress in these areas. applied in backfit decisions. 254 ● Nuclear Power in an Age of Uncertainty

A2. Develop a Standardized LWR Design, Opti- by incorporating improved construction tech- mized for Safety, Reliability, and Economy niques. It should be cheaper to operate because it would be designed to operate at a higher ca- For a variety of reasons discussed in chapters pacity factor, lower fuel cycle costs and lower 4, 5, and 6, the reactor plant designs currently operator exposure. Even if the technology were available could be significantly improved. An ef- similar to present reactors, this new design fort that rethinks the concepts by which reactors package might represent a major improvement have been designed could result in light water in the acceptability of nuclear power. reactors (LWRs) that are cheaper, safer, more operable, and perhaps smaller than the present There are also disadvantages, however. It generation. This effort would draw on all that has would be at least as expensive for the Govern- been learned about the characteristics of good, ment to sponsor such a design as it would for a safe reactors and integrate the best features into corporation—perhaps several hundred million a package that would represent the best of tech- dollars if a demonstration were required. In ad- nology. The design philosophy would emphasize dition, a Government lead in developing a new resiliency and passive safety features as well as design might imply to some groups a dissatisfac- affordability and economy. The system would be tion with present designs serious enough that ex- subjected to intensive analysis from every possi- isting reactors should be shut down. ble perspective to ensure that, insofar as possible, all contingencies had been covered. A3. Promote the Revision of Rate Regulation To a degree, the Westinghouse effort on the The process of rate regulation in most States advanced pressurized water reactor described in was designed for an era of relatively small capital chapter 4 meets these objectives. The rationale cost increments and declining costs per kilowatt- for a Government role is that a complete reac- hour. High interest rates and high capital costs tor and plant design is extremely expensive, and for new generating capacity have strained the sys- no corporation is likely to be able to finance it tem so seriously that changes may have to be unless it sees a major market, which is not now made before utilities resume ordering new cen- the case. In addition, there are several technical tral station capacity. The current overcapacity questions such as the unresolved safety issues re- gives utilities a welcome respite, but large con- quiring additional R&D that is best funded by the struction programs will be needed once again. Federal Government. A Government-initiated Regulatory changes that should be considered nonproprietary design could more easily draw on here are: 1) rate base treatment of utility assets the work of more than one vendor or architect- that takes inflation into account, 2) some con- engineer as well as a coordinated R&D program, struction work in progress (CWIP) to be included and be available to more producers. Therefore, in the rate base, and 3) real rates of return on a national design could have a better chance of equity appropriate to the actual investment risk. being truly optimized. The safety analysis also These changes and others are discussed in detail might be more convincing since it would be done in chapter 3. Their general intent is to even out in a more open atmosphere, with direct feedback rate increases and provide greater financial stabil- to the design to improve safety to the maximum ity for utilities and their customers. extent possible. There is also a growing feeling that current reactors have overshot the ideal size. A difficulty for Federal policy in this area is that U.S. vendors are unlikely to be in a position to rate regulation is the prerogative of the States. If redesign their new reactors to be smaller. Federal action is to be acceptable, it must be taken in a way that makes it in the interest of the There are several advantages to a standardized States. Federal encouragement of long-range redesign. The cost would be much more predict- gional planning and regulation (see option D5) able, since most of the regulatory and construc- may be useful since many States are finding that tion uncertainties could be cleared up before their regulatory programs are encountering in- construction started. Costs also could be lowered creasing difficulties in forging satisfactory com- Ch. 9—Policy Options ● 255 promises. To some extent, regulation of whole- ternational competitive position. A long-term sale power sales by the Federal Energy Regulatory R&D program could provide further benefits. Commission (FERC) influences State regulation. In the summer of 1983, there was extensive con- B2. Improve Utility Management of Nuclear gressional debate on legislation restricting FERC Power Operations allowance of CWIP. Consumer opposition to None of the policy options discussed in this CWIP has been intense because it allows pay- chapter will do as much to improve the attrac- ment for facilities before they are of any use to tiveness of nuclear power for all the parties to the the ratepayers. Some States, however, do have debate as improved utility management. Many partial CWIP allowances. utilities were unprepared for the complexities of nuclear power and the dedication required. This Goal B: Improve Reactor situation was perhaps unavoidable, given the Operations and Economics overenthusiasm gripping the nuclear supply industry and the Federal nuclear promoters. By Decisions on the desirability of future reactors now, however, we are in a period of operation, will be based not just on capital cost projections not expansion. Utilities now have the primary (as improved under goal A) but also on the per- responsibility, and all utilities responsible for formance of existing reactors. The low reliability nuclear plants must be adept at carrying it out. experienced at some plants negates their poten- it is important to recognize that much is being tial economic benefits and raises concerns over safety. Investors, critics, and the public will be done to improve the quality of operations as dis- opposed to more orders if some plants are notice- cussed in chapter 5. The Institute for Nuclear ably unreliable. Thus, it is in the interests not only Power Operations (IN PO) was set up for precisely of the specific utility involved but of the industry this purpose and has developed a large number as a whole to improve operations at all plants. of specific programs. The NRC has shifted some Other means for improving the economics of ex- of its scrutiny from the plants to the organizations running them. It is not yet clear whether these isting reactors could also improve the outlook for nuclear power as a whole. efforts will be sufficient. The specific options toward this goal follow. Specific areas for attention are training and organizational structure. Both previously had been left to the discretion of the utility but are now be- B 1. Support R&D Programs to Improve the Eco- ing addressed by both the NRC and IN PO. nomics of Operation DOE and most of the nuclear steam supply sys- Requirements for training show a remarkable tem (NSSS) vendors have modest programs for variation. Good training programs are expensive, developing extended burnup for fuel elements. and qualified instructors are in limited supply. It There would be a national benefit from expan- is important to set standards for training and establish reasonable programs for achieving them. sion of these programs. Fuel cycle costs could be reduced slightly, and the volume of spent fuel ac- INPO is beginning to do this by establishing a cumulation would be decreased considerably training accreditation program. It also may be necessary for NRC to impose these standards to (perhaps by 40 percent). This latter factor, by itself, could justify a significant Federal effort. Sav- achieve the optimum progress. The NRC probing 40 percent of the spent fuel would not reduce ably has the statutory authority to do this, but a the spent fuel problem proportionately, but it congressional directive would expedite NRC ac- would ease the total burden considerably in the tions. Careful observation of the results of INPO’s long term. The objection generally voiced to a efforts is important to see whether additional NRC Federal program is that private industry could action is necessary. handle it. While this is probably true, a Federal Criteria for organizational structure will be role would expedite matters and improve our in- harder to define. One factor that is apparent, 256 ● Nuclear Power in an Age of Uncertainty

however, is that the highest levels of the utility causes. Critics would argue that excluding such management must be involved with the plants cases would deprive a large number of potential and committed to their good operation. Again, victims of just compensation. the NRC probably has the authority to command Exposure levels during the weapons tests were attention at the utility headquarters, and is at- considerably higher than expected occupational tempting to do so. Still greater resolve seems to exposures at nuclear reactors. Some workers will, be in order at some utilities, however, and con- over their lifetimes, nevertheless accumulate a gressional encouragement of NRC to make this high enough dosage to qualify for awards if the a high priority item would help. floor is at the 10- to 20-percent level. Hospitals Both the NRC and INPO know which utilities also may find themselves liable for the exposure are most in need of upgrading their management. from X-ray machines and nuclear medicine. Com- All utilities are strongly influenced by the ex- pensation for test victims is an important social periences of these few. Strong measures may be issue. It also is important to recognize that it has required to get the operation of these plants up implications for the nuclear industry that could to minimally acceptable levels. Congressional ex- be serious if the awards are large. presson of the importance of a strong management commitment would be a significant incentive for the NRC and the utilities. Goal C: Reduce the Risk of Accidents That Have Public Safety or B3. Resolve the Financial Liability for Occupa- Utility Financial Impacts tional Exposure Nuclear reactor safety is a function of the de- The weapons testing program has focused at- sign of the plant, the standards by which it is built, tention on compensation for injuries arising from and the care with which it is maintained and op- exposure to radiation. New approaches are be- erated. If any of these are deficient, safety will ing developed for compensating test participants be compromised, perhaps seriously, and costs and downwind residents, and the industry is con- may well escalate unexpectedly. Option A2 has cerned that these plans will be applied arbitrari- discussed how to improve the designs of the next ly to commercial nuclear plants (and possibly the generation of LWRs, but this alone may not be medical industry). The proposals under consid- adequate. It would not affect existing plants, and eration for the weapons tests plaintiffs link radia- it may not go far enough in assuring safety in fu- tion exposure to the probability of contracting ture plants. Without a consensus that nuclear re- cancer, and then award compensation based on actors now are safe enough, there are unlikely that probability. With this approach, claimants to be any more. Therefore, ways to improve the who receive the most exposure also receive the safety of both present and future reactors are ex- greatest rewards. Recent legislation in Congress plored under this goal. proposes awarding $500,000 to a claimant if there is at least a 50-percent chance that the cancer The quality of the people involved appears to developed from the radiation exposure. At lower beat least as important as the design of the plant. levels of exposure that may only result in a 10- Option B2 discusses how to improve utility opera- to 20-percent chance of cancer, the claimant tion, but again this may be inadequate by itself. could receive $50,000. This proposal is controver- Some utilities simply may be unable to improve sial for two reasons. First, the nuclear industry their performance sufficiently. Others may think contests the relationship between low radiation they have done so but experience the same dif- doses and cancer since there is insufficient scien- ficulties in construction when they order another tific or technical basis to support it. In addition, plant. Two options discussed under this goal can many claimants who were exposed to low doses be considered if utility improvement is inade- would receive compensation for cancers that quate. Construction permits and operating li- were not produced by radiation but by other censes could be reserved for utilities and contrac- Ch. 9—Policy Options ● 257

tors that can demonstrate the commitment to safety. Resolving them expeditiously would elim- build and operate the plants to the exacting stand- inate some safety concerns, demonstrate a com- ards required. Second, different institutional mitment to maximum safety on the part of the arrangements might be considered to replace util- NRC, and permit more stable cost projections for ity management of reactors, This option also future plants. Resolution of some of the issues could be effective in stimulating further growth may call for modifications on existing plants. of nuclear power if utilities are reluctant to order While the utilities would not welcome such ex- more. penditures, the overall reduction of uncertainties and the gains in safety would be useful. Cl. Improve Confidence in the Safety of Existing and Future Reactors C2. Certify Utilities and Contractors As discussed in the options above, there has been a continual evolution in designs because It is readily apparent that some nuclear plants of frequent discoveries of inadequacies with re- are not being built and operated skillfully enough. spect to safety or operation. As our understand- As discussed earlier, all plants may be hostage ing of the technology has improved, formerly un- to the weakest because an accident, or even poor foreseen accident sequences or conditions are performance, reflects on all. If the persuasive ap- recognized. Unquestionably, the technology is proach of option B2 is insufficiently effective in maturing, but there is considerable dispute over improving nuclear plant management, more dras- how much farther it has to or can go. tic steps could be warranted. Part of the problem has been the partitioned For existing reactors, the NRC evidently already nature of the safety analysis both in the industry has the power to suspend an operating license and the NRC. Each system may be thoroughly if a utility is incapable of managing a reactor safe- scrutinized, but the entire plant is not viewed as ly. Few people expect the NRC to do this without a system, and responsibility for analyzing its over- the most compelling evidence of incompetence. all safety appears to be lacking. If higher standards are to be enforced, it probably will only be with congressional legislation. No amount of analysis will uncover all poten- Such improved standards would be in the best tial problems, but an intense analysis of each interests of the industry even though their imple- plant could identify design or operating flaws mentation could be traumatic. Even if this author- before they caused problems. These studies are ity were never invoked, it could be a strong inexpensive, but a few utilities already are under- centive to utilities to improve their performance. taking them in their own interests. The intent is The result would be greater confidence in the to discover weak points in the design and develop safety and operability of reactors. measures to address them, whether by changing plant equipment or modifying operations. Future reactors present a slightly different picture. Utilities have Iearned that building reactors Other efforts to improve safety could focus on is very difficult, and few, if any, would embark improving the analytical techniques. As has been on a new construction program unless they were stated above, probabilistic risk assessment is a confident they had the ability. Even then, how- useful tool that is still imprecise. Development ever, other parties of concern may not share that of this technique would be beneficial for both confidence. Certification of utilities as having the safety and economics. This will involve mainly necessary ability and commitment to build and improving the data base for failure rates and operate a reactor to high standards would ensure analyzing the human element, as is done in the that many of the expensive mistakes of the past aircraft industry. were not repeated. This would reassure many of The existence of unresolved safety issues, and the critics of nuclear power as well as investors, the probable introduction of more as new con- utility commissions, and the public. It also might cerns are developed, undermines confidence in be necessary to eliminate from contention con- 258 Ž Nuclear Power in an Age of Uncertainty tractors who had not demonstrated their capabili- the entire HTGR concept, including licensability of meeting the exacting standards required for ty, costs, operability, and acceptability would be nuclear construction. Presumably utilities would required. This would necessitate an increased de- know better than to select these contractors, but velopment program at DOE. some past experiences have been so poor that Even if the HTGR is not seen as a replacement making it official would increase confidence. for the LWR, there are still several reasons for sup- Even though this option is not likely to prevent porting an R&D program paced to make it avail- any plants from being built, it would be viewed able early in the next century. it would use urani- by the industry as another set of regulations to um more efficiently than LWRs, has relatively be- meet in what they consider to be an already over- nign environmental impacts, and could be used regulated enterprise. The utilities also may resent for industrial process heat. A small, modular form having a Federal agency judge utility manage- also has been proposed that could have major ment quality. An independent peer body analo- safety and financial advantages and be particular- gous to that being set up for review of medicare ly well suited to process heat applications. inpatient treatment might meet with better ac- It is harder to see a role for heavy water reac- ceptance. tors (CANDU) in this country. CANDUs are work- There are no clear criteria as to what constitutes ing extremely well in Canada. At least some of good management concerning construction of a that success, however, is due to the managerial nuclear powerplant. Nevertheless, as part of a environment in which the nuclear industry oper- strategy to rebuild confidence in the technology, ates in Canada. Transplanting it to this country this option clearly bears further examination. could lose these advantages, and would necessitate industry learning and investing in a quite C3. Develop Alternative Reactors different technology. While the technology can be mastered, a significant research program DOE has carried on a modest program for R&D would be necessary to adapt CANDUs to our reg- on the high temperature gas-cooled reactor ulatory requirements, or vice versa. It is not clear (HTGR). Given a higher commitment, the HTGR that this effort is warranted compared to other might develop into a superior reactor. In par- alternatives such as the HTCR or improved LWRs. ticular, it has inherent safety features that at least temporarily would shield it from some of the safe- The final alternative reactor discussed in chap- ty concerns of the LWR. Further, if problems deter 4 is the PIUS, which was conceived largely velop with the LWR that are too difficult to solve to meet safety objections to the LWR. While rad- economically, the HTGR probably would be the ically different from the LWR in some ways, it still next available concept in this country. An en- is an LWR. Therefore it has an element of famili- larged R&D program could prove vital in main- arity that the others do not. The concept, or at taining the nuclear option. least some features of it, appear promising, but only a significant research effort will confirm the On the negative side, it has to be noted that feasibility of the design since it is still a paper reac- gas reactors have not been a great success any- tor concept, There is great uncertainty over this where, and most countries have turned to the concept, but if the research program does prove U.S. developed LWR technology. Estimates of out the expectations of the developers, the reac- future costs and reliability are much more con- tor could be deployed rapidly. PIUS could be jectural than for the LWR. Many utilities would perceived as much safer by the public and critics. be reluctant to turn to a less familiar technology that might turn out to be subject to many unfore- Development of new technology will not by seen problems. Such uncertainties will only be itself solve the problems of the industry. resolved by a substantial R&D program. To a However, it will play a vital role in an overall greater degree than for the standardized LWR upgrading, whether the end result is an improved discussed above, a thorough demonstration of LWR or an alternative concept. Ch. 9—Policy Options ● 259

C4. Revise the Institutional Management of Goal D: Alleviate Public Concerns Nuclear Power if Necessary and Reduce Political Risks

If a utility has its license revoked as in option The issue of public acceptance has permeated C2, or such action seems likely, it might think of this report for good reason. If the long-term trend turning the plant over to a different operating in public opinion toward increasing opposition agent instead of just shutting it down. Utilities (described inch. 8) is not reversed, there will be already use a large number of consultants and few, if any, more orders for nuclear plants. service companies for specific tasks. Operating Many of the options discussed above are rele- service companies, discussed in chapter S, could vant to this goal. Nuclear energy will not be ac- bean extension of these, or they could be other ceptable so long as there are spectacular ex- utilities that have established good records and amples of out-of-control cost escalations and a are prepared to extend their expertise to other continuing series of alarming operating events. reactors. The NRC would be satisfied that the A major accident involving offsite loss of life plant was being given the management attention would almost certainly preclude future plants and required, the utility would have its plant operating quite likely close many operating reactors. There- again, probably at higher availability than before, fore, almost any action to improve operations and and the public would have greater assurance safety will pay dividends in public acceptance. about the commitment to safe operation. The options discussed here are intended to reduce the controversy or to confine a role for There are potentially serious liabilities to the nuclear power. idea, however. No utility would like to admit that it is incapable of operating its plant safely and would be reluctant to turn to another operating 01. Accelerate Studies of Alternative Energy Sources company except under extreme conditions. The contract between the two would have to be care- One of the major factors affecting public opin- fully worked out to determine who would pay ion against nuclear power is the feeling that the for modifications and maintenance. If a serious risks associated with it outweigh the benefits. As accident did occur, plant restoration costs and long as other energy sources are available that liability for offsite damages would have to be are perceived to be both more economical and spelled out. Premature plant closure due to unex- acceptable, there is little incentive to favor pected deterioration could be another problem. nuclear energy with its more controversial risks. Therefore, as more information is developed on There do not appear to be any legal impedi- the resource base, costs and impacts of these al- ments to the idea that would require legislation. ternatives, better decisions can be made on the However, Congress might want to encourage the relative merits of nuclear energy. NRC, and perhaps the Justice Department, to The major competitor of nuclear power for new undertake further analysis. central station plants is coal. Yet coal is arousing concerns (e.g., carbon dioxide and acid rain) that Alternative institutional arrangements also may exceed those of nuclear. Significant research couId be formed to encourage nuclear orders i n is going on in these areas, and the answers are the future. If individual utilities are unable to crucial for nuclear power. The sooner they be- undertake the risk, consortia of utilities, possibly come available, the easier it will be to make in- including vendors and architect-engineering firms formed decisions. etc., might be able to do so. Alternatively, Gov- ernment-owned power authorities might be the Some analysts feel that natural gas resources only way to maintain the nuclear option. These have been greatly underestimated. This cannot concepts are explored briefly in chapter 5. be confirmed for many years, but there is an im-

portant data-gathering role for the Government. a basis other than one’s attitude toward nuclear If gas remains plentiful and is permitted as a boiler power. The outcome, however, could be very fuel, it will reduce the competitiveness of nuclear important to the future of nuclear power. energy. From a different perspective, it also might be useful to expand R&Don the solar energy op- D2. Address the Concerns of the Critics tions that appear promising. Some of the euphor- ia about solar energy has withered under the hard Critics of nuclear power have long felt a deep light of costs, but some technologies such as distrust of the industry and the NRC. They feel photovoltaics are still candidates. Accelerating that their concerns have been ignored or down- these technologies actually could be beneficial played while the Nation plunged ahead to build to nuclear power. If they ultimately prove to be more reactors. The mistrust is mutual. The indus- not widely competitive with nuclear energy, we try feels that nothing would change the mind of would know that sooner. If they are reasonably the critics. competitive, then the Nation has another option. Bridging this distrust will be difficult at best. For None of these proposals is particularly contro- those critics who do not want nuclear power versial, though some might be expensive. In gen- under any conditions and for those in the nuclear eral, decisions on these options will be made on industry who refuse any concessions, resolution Ch. 9—Policy Options ● 261

probably is impossible. However, opposition to D3. Control the Rate of Nuclear Construction nuclear power is not monolithic. Many critics Many of the concerns over nuclear power origi- have specific concerns over the technology and nated during the late 1960’s and early 1970’s its implementation. These critics tend to be tech- when projections of very high growth seemed to nically knowledgeable and respected in the envi- be on the way to being realized. People became ronmental and anti-nuclear communities. Fur- alarmed over the thought of 1,000 reactors or ther, some in the industry are aware of these con- more around the country, many reprocessing cerns and appear to be willing to engage in a use- plants with spent fuel and plutonium shipments ful discussions. It is with these groups that a requiring security that disrupts normal transpor- bridge might be constructed. tation and threatens civil liberties, and the ever One important step is to resolve the safety con- present possibility of accidents. The industry itself cerns that have already been identified, as dis- found the rapid expansion more than it could cussed under option Cl. Further steps may be adequately manage. required to convince critics that every effort was Lower projections of nuclear growth have re- being made to identify previously unrecognized duced some of these concerns. However, a re- safety concerns and implement solutions at existing reactors. sumption of orders might rekindle the fears of another “too rapid” expansion. The most straightforward way of providing this Establishing a controlled growth rate may give assurance is to involve critics directly in the reg- assurance that the early. concerns about overex- ulatory process. This might be done through con- pansion would not recur. As discussed in chapter tracts to supply specific information or review ma- 8, the 1980 Swedish national referendum limiting terial (intervener funding), by including techni- the total number of reactors appeared to quell cally knowledgeable critics on the Advisory Panel the political controversy. for Reactor Safeguards, or by creating an office within the NRC that would serve as a liaison to Controlling the growth rate might realize this the critics. improvement in public acceptance without pre- cluding all future orders. The limit could be in Few proposals generate as much controversy the form of capacity that could be granted con- as this one. Industry sees it as opening the flood- struction permits in a year, or a sliding scale to gates to implacable opposition that would make any progress impossible. Utilities see it as presag- allow nuclear construction to remain at a roughly fixed fraction of total new capacity. If utilities’ in- ing a steady stream of new backfit orders and un- terest in new nuclear orders revives to the extent necessary regulations. Much of the NRC sees it that the growth rate could be exceeded, the NRC as an unwarranted infringement on its process. would allocate the permits using criteria of re- If it is to be implemented, congressional direc- gional need and ability to manage the technology tion will be needed. It may be worth the effort. as discussed under A3 above. There is no intrin- Nuclear technology is still imperfect, and the sic difficulty in the Government allocating limited sooner problems are discovered, the sooner the permits (e.g., airline routes were limited for many technology can be improved. Involving the critics years). is likely to speed this process. I n addition, improvements in public support is a prerequisite for This policy option would be controversial, at more orders; public support is unlikely to im- least at first. The industry would argue that it prove as long as the controversy over safety is would constitute unwarranted interference in the so bitter; and this controversy is unlikely to die marketplace and would distort economic deci- down until most of the concerns of the critics sions. In particular, if nuclear reactors turn out have been addressed. Given the current impasse to be the most economic form of electric power there may be little to lose by trying this approach. when managed carefully and if fully redesigned, 262 Ž Nuclear Power in an Age of Uncertainty controlled growth could result in a misallocation 1990’s. By then, many utilities may be finding of resources. However, until public acceptance their own capacity fully committed and bulk improves noticeably, no utility is going to order power purchases less available. Without major any reactors. If controlled growth were instru- changes in the way we generate and use elec- mental in improving public acceptance to the tricity-changes that are highly speculative now point that orders resumed, it would be of major —the Nation will need substantial new generating assistance to the industry. Furthermore, the bur- capacity to come online by 2000, and perhaps den of proof should be on the industry that it sooner. Some regions with high growth rates from could manage a rapid increase of orders, since population shifts and economic changes will ex- many of the present problems came from the last perience the need earlier. surge. The NRC has the authority to prevent such a surge by insisting on preapproved designs and Planning for electric growth can make clear evidence of utility capability, but a congressional- what the choices are and what the consequences ly imposed limit would be more convincing to might be. These plans could help build a consen- the public and the critics. sus on the necessary additional capacity and load management. At the least, plans would provide D4. Maintain a Strict Nonproliferation Stance a format for discussion. in conjunction with the controlled construction rates discussed above, Proliferation has been one of the major con- such plans could be quite effective. cerns of the critics and the public. All known reactors could be used in some fashion to facilitate National planning is probably too large a scale the production of nuclear weapons. If nuclear to be useful. State planning may be too small con- power is to regain public trust, this linkage must sidering the growing regional nature of power be minimized. wheeling. Regional planning appears best to cap- One step is to keep the U.S. weapons programs ture the commonality of interests, This might be and power programs sharply distinct, both tech- combined with regional rate-setting as mentioned in chapter 3. nically and institutionally. Separate waste disposal programs could be one step, even though the insofar as nuclear power is concerned, this pro- material is not much different. Consideration also posal might not make any difference. It would might be given to removing the weapons pro- not by itself eliminate any barriers to new reac- gram from DOE though that would be a compli- tors and might even raise an additional layer of cated decision beyond the scope of this study. regulation. On the whole, however, it should A related suggestion is to consider a ban or ex- allow utilities that decide they should build a tended moratorium on reprocessing. Reprocess- reactor to make a stronger case for it by showing is the focus of much of the opposition to ing how it will benefit the customers in the long nuclear power. A long-term legislated moratori- run. um, perhaps coupled with the controlled growth in option D3 and the extended burn up of option This policy option would be implemented by B1, would eliminate many of the major causes setting up regional planning authorities, possibly with ratemaking authority, that are agreed to by of concerns. the States. It is important that these authorities also have authority to determine power needs. D5. Promote Regional Planning for Electric Such responsibility should not go by default to Power Capacity Federal agencies such as the NRC, which are not One of the major reasons for the poor public well equipped to make such determinations. The opinion of nuclear power is the perceived lack concept of regional authorities appears promis- of need for it. As discussed in chapter 3, this need ing, but it has not been studied in detail in this is unlikely to be readily apparent before the late project. Ch. 9—Policy Options ● 263

MAJOR FEDERAL POLICY STRATEGIES AND THE LIKELIHOOD OF MORE ORDERS FOR NUCLEAR PLANTS It should be clear from the other chapters in to help evaluate how successful the strategies the study that none of the individual policy op- might be under different conditions. tions described earlier in the chapter is sufficient There are also two variations on these strategies by itself to improve significantly the prospects that are not analyzed in detail in this study but for more nuclear orders. There are too many dif- are worthy of consideration. One of these, a vari- ferent problems that have to be addressed before ation on the Base Case, would make several nuclear power can be again considered a viable changes in Federal policy to encourage more energy option for the future. market competition between nuclear power and If several of these policy options are coupled other generation (and load management) tech- in an overall strategy, however, they may be col- nologies. The other, a variation on Strategy Two, lectively much more effective. These strategies would consider nuclear power, not so much as should include options directed toward each of an important aspect of U.S. energy policy but as the four policy goals described earlier in the chap- a key element of U.S. industrial and world trade ter: reduce construction costs, improve reactor policy. economics, reduce the risks of accidents, and The Base Case and two strategies are outlined alleviate public concerns. Most of the policy opin table 35 with the policy options, discussed in tions will be controversial to some extent. For this the previous section, listed for each strategy. reason, it is likely to be necessary to take steps There is also a brief description of the two varia- that meet the concerns of several different groups tions with a general description of the probable at once: utility executives, critics, investors, reg- policies under each. ulators, and the public. The divergence among the views of these groups should be clear from the rest of the study. Base Case: No Change in In the face of the controversies surrounding nu- Federal Policies clear power and the uncertainties surrounding There are several reasons why policy makers its future, one obvious Federal course is to make might choose a strategy that avoids any major no changes in Federal policy. If such is the case, changes in the current Federal laws and regula- future nuclear orders will be heavily influenced tions affecting nuclear power. Some policy makers by two sets of conditions outside the direct con- may view nuclear energy as unimportant or un- trol of the Federal Government: economic condi- desirable, while others may feel that the Federal tions and improvement in nuclear industry man- Government already has done enough for the in- agement. In the section that follows, the pros- dustry, making further legislation unwarranted. pects for new nuclear orders in the absence of Still others may not wish to take any action right new Federal policies, the Base Case, are exam- now. At present there are many uncertainties ined for each of four nuclear futures that assume about future electricity demand, the environmen- different sets of economic conditions and industry tal impacts of coal combustion, and the poten- management success. tial of conservation and renewable resources The two other strategies described here assume which will affect the necessity for and attrac- various degrees of Federal intervention on several tiveness of nuclear power. Policy makers may fronts. The first of these, Strategy One, would prefer to wait 5 or 10 years to see how these merely remove obstacles to further nuclear or- uncertainties are resolved before revising current ders. A more active approach, Strategy Two, nuclear energy policies. Finally, policy makers would go further and attempt to stimulate more may feel that Federal legislation would have lit- nuclear orders. The four futures described under tle impact on the industry and that economic the Base Case also are examined for each strategy forces will ultimately determine its fate. 264 Ž Nuclear Power in an Age of Uncertainty

Table 35.–Major Policy Strategies (and the policy options included in each)

Strategy Policy Options Included Base Case: No change in Federal nuclear policy: Three noncontroversial policies that would be useful even in the absence of more orders Goals Policy Options Improve reactor economics ...... (B1) R&D to improve fuel burnup Reduce accident risk ...... (Cl) Improve analysis of reactor safety Alleviate public concern ...... (D1) Accelerate studies of alternative energy sources Variation: Sharpen market competition of nuclear power This strategy, not analyzed in detail, would include some or all of steps towards: reduction or removal of Federal subsidies for nuclear and alternatives; marginal cost pricing; deregulation; full costing of external impacts Strategy One: Remove obstacles to more nuclear orders: Three policies above plus five others Goals Policy Options Reduce capital cost barrier ...... (Al) Revise regulation (A2) Assist funding of standardized optimized LWR design Improve reactor economics ...... (B2) Improve utility management Alleviate public concerns ...... (D2) Address concerns of critics (D3) Control the rate of nuclear construction Strategy Two: Provide a moderate stimulus to more nuclear orders: Eight policies above plus eight others Goals Policy Options Reduce capital cost barrier ...... (A2) Assist funding of a demonstration of new LWR designs (A3) Promote the revision of rate regulation Improve reactor economics ...... (B3) Solve occupational exposure liability Reduce accident risk ...... (C2) Certify utilities and contractors (C3) Develop alternative reactors (C4) Revise institutional management of nuclear operations Alleviate public concern ...... (D4) Maintain nonproliferation policies (D5) Promote regional planning for electric growth Variation: Support the U.S. nuclear industry in future world trade This strategy, not analyzed in detail, would support industry and utility R&D and export financing policies aimed at obtaining a major share of the future world market in nuclear and other advanced electrotechnologies. SOURCE Office of Technology Assessment.

A “no change” strategy would continue the the act expires in 1987 and if it were not renewed, current Federal policy toward nuclear power. it could have a significant impact on the nuclear DOE could continue to fund R&D of both the power industry, although how much and what LWR and alternative reactor types at about cur- kind of impact has not been analyzed in this re- rent levels. Although current NRC efforts to port. A second assumption is that the Nuclear reduce backfit orders and streamline the licens- Waste Policy Act of 1982 is implemented success- ing process would continue, there would be no fully and the feasibility of safe waste disposal will major legislation and no fundamental changes in be demonstrated. present regulatory procedures. The strategy also assumes that three noncon- This strategy does assume continuation of two troversial policy options (actually expansions of fairly controversial Federal policies. One assump- existing efforts) discussed in the preceding section is that Congress will renew with no major tion could be implemented: (B I ) R&D for higher changes the Price-Anderson Act, limiting the lia- burnup and other improvements to reactor bility of plant owners and constructors in the economics would be funded; (Cl) Safety con- event of an accident (described in ch. 3). Part of cerns would be addressed more vigorously; and Ch. 9—Policy Options . 265

(D1) research into problems and opportunities ble combinations of events that could make nu- for alternative sources of electricity generation clear power more or less attractive to utilities over would be accelerated. the next 10 years. The likely outcome if Federal policy is not changed will depend on two major factors— Economic Conditions: Two Scenarios the success of industry efforts to eliminate cur- The major economic factors that will affect fu- rent problems, and the economy. Two alternative sets of external economic conditions are ture demand for nuclear power are the rate of growth in electricity demand, the price and considered here. One would result in a relative- availability of alternative energy and electricity ly high demand for new central station generating sources, and inflation and interest rates. All of plants and the other in low growth. Similarly, the potential range of results of current industry ef- these factors are discussed in greater detail in chapter 3 and summarized only briefly here. forts to improve the viability of nuclear energy are represented by two different outcomes: rela- Economic Scenario A: More Favorable to Nu- tively successful and only moderately successful. clear Orders.—As shown in table 36, Economic These outcomes, or scenarios, are summarized Scenario A includes a combination of those in table 36, and will affect the impact of each of economic factors that could be expected to make the other two strategies described in this report nuclear power more attractive. In this scenario, as well as the Base Case results. These scenarios rapid price increases for oil and gas might ac- are not predictions or projections of the future, celerate the shift to electricity helping create a but instead brief sketches of a few of the possi- moderately high increase in electricity demand

Table 36.—Four Scenarios Affecting the Future of the Nuclear Industry Economic Scenarios Affecting the Nuclear Industry Variable Scenario A: Favor more orders Scenario B: Hinder more orders Electricity demand (average annual growth rate 1983-93 ...... 3.5% 1 .5 ”/0 New capacity needed in 1995 (GW) would have to be ordered in late 1980’s a ...... 161 o Additional capacity needed between 1995 and 2000 at same demand growth rate (GW) would have to be ordered by 1992 ...... 218 84 Price of alternative fuels: Oil and gas ...... Real price increases faster than price Real price increases at same rate as of electricity electricity Renewable ...... Real price remains higher than price of Price becomes competitive with electricity electricity Inflation rates and interest ...... Low High

Environmental impacts ...... Concerns about acid rain, global CO2 No constraints on fossil increase

Industry Improvement Scenarios Variable Scenario A: Major improvements Scenario B: Modest improvements Average construction time ...... 7 years 12 years

Operation of existing reactors ...... 70% availability 60°/0 availability Safety risks ...... Currently operating reactors shown Little progress on unresolved safety much safer issues Reportable operating events...... Almost none over decade; management Continue at current rate; much media improvements obvious coverage see ch. 3 for a complete discussion of assumptions used in capacity projections; GW as used here means GWe. Possible factors In price increases: limited gas resemes; tight international oil market; increased environmental controls on coal burning. csteady resemes of oil and g=; continued consewation eases demand.

SOURCE: Office of Technology Assessment.

25-450 0 - 84 - 19 : QL 3 266 ● Nuclear Power in an Age of Uncertainty

(3.5 percent per year). This rate of growth in de- in Economic Scenario B, rapid inflation over mand, coupled with a moderate need to replace the next few years would cause continued price aging powerplants, is expected to create a need increases and continued high interest rates dur- for 161 gigawatts (GW)* of new central station ing completion of the 30 nuclear plants now un- generating capacity by 1995 and an additional 218 der construction. Utilities would be forced to re- GW by 2000. Given the time required to com- quest large rate increases from utility commissions plete new generating plants, utilities would be ex- as the plants are finished. These rate increases, pected to order this much capacity in the 1983-93 combined with the slow growth in electricity de- decade. mand, would cause consumer opposition and increased public skepticism about nuclear power. Under this scenario, increasingly stringent en- All of the assumptions included in Economic Sce- vironmental restrictions could make new coal nario B would be expected to make new nuclear plants very expensive. If this price increase oc- plants less attractive to utilities. curred at the same time as the projected growth in electricity demand, utilities would be faced Management Improvement Conditions: with the need for new capacity while their most Two Scenarios important fuel was becoming considerably harder to use. As a result, nuclear power would appear Industry and utility success or failure to make much more attractive to utilities placing power- substantial improvements in the management of plant orders. If the inflation rate and prevailing the nuclear enterprise will be reflected in several interest rates were relatively low, capital costs of indicators: Ieadtime to build nuclear plants; aver- nuclear plants would be more manageable and age plant availability; progress on unresolved predictable for utilities. Low inflation and the safety issues; and frequency of precursor events. decreasing construction costs over the next few These subjects were discussed in chapters 4 and years will stabilize rates to consumers, very like- 5. ly lessening hostility to utilities. In addition, the Management Scenario A: Major lmprove- benefits of nuclear power would grow in the eyes ment.—ln Management Scenario A, the nuclear of the public as electricity demand increases. industry would be very successful at overcom- Economic Scenario B: Less Favorable to ing some of its current difficulties without govern- Nuclear Orders. —If the economy follows this ment assistance. Utilities currently operating reac- path, nuclear power remains relatively less attrac- tors would overcome operating and safety prob- tive. In this scenario, moderate price increases lems, creating a steady improvement in reliabili- of gas and oil slow the shift to electricity. In ad- ty of operating reactors. Improved operation of dition, renewable energy sources become more existing plants and projections of reduced con- competitive with central station electricity. As a struction costs wouId make nuclear power more result, there is only slow growth (1.5 percent) in economically attractive to investors and public average annual electricity demand, and no new utility commissions as well as to consumers. generating capacity must be completed in the Management Scenario A assumes that operat- decade. However, even at this slow rate of ing plants and those completed over the next growth, if moderate numbers of existing plants decade would be shown to be much safer than are retired, about 84 GW of new generating ca- presently assumed because of improved opera- pacity would be needed by the year 2000, and tions and better understanding of the technology. this capacity would have to be ordered in the Improved analysis and information (e.g., the 1983-93 time period. With relatively small in- ongoing research into “source terms”) could creases in the price of coal, and high interest rates demonstrate other safety characteristics (as driving up the capital costs of nuclear plants, util- discussed in ch. 4). ities would be more likely to invest in coal-fired plants. Two consequences would follow from these safety improvements. First, the management *One gigawatt equals 1,000 MW (1 ,000,000 kW) or slightly less changes would greatly reduce major events, such than the typical large nuclear powerplant of 1,100 to 1,300 MW. as the failure of the automatic shutdown system Ch. 9—Policy Options ● 267 at the Salem, N. J., reactor, which are viewed by shut down for several months to a year for repairs. the public as precursors to a major accident. Sec- Without an adequate insurance pool, this would ond, the new information on the small amount cause a major rate hike to cover purchased pow- of radioactivity released in the event of an acci- er. The long construction periods, continuing dent would temper the reaction to the few oper- operating problems, and rate hikes due to out- ating events that did occur over the decade. ages would increase investors’ and consumers’ These safety gains should be helpful in reduc- skepticism of the technology. ing public opposition to the technology and fur- without Government intervention, this sce- ther increasing investor confidence. nario could be expected to have very negative In addition to the improvement in utility man- consequences for the industry regardless of ex- agement of operating reactors, the nuclear supply ternal economic events. industry would be offering improved standardized LWR designs such as the APWR currently Four Nuclear Futures Under the Base being developed by Westinghouse and Japan. Case: No Policy Change Under current regulatory policy, the NRC could give licensing approval for a complete design, The two sets of economic conditions described and, if a plant were built exactly to the design, above combine with the two management sce- there would be few regulatory changes during narios to form four futures that illustrate the range construction. Thus, the regulatory environment of possibilities for more nuclear orders. Future would become somewhat more predictable One is a combination of favorable economic con- without any major Federal legislation. ditions (Scenario A) and major improvements in management (Scenario A). Future Four com- Management Scenario B: Minor lmprove- bines the least favorable scenarios. Futures Two ments.—ln Management Scenario B, some of the and Three are intermediate. The discussion that weaker utilities would fail to improve their per- follows describes the factors under each future formance despite the efforts of INPO and the that would affect decisions on new nuclear NRC. Average availability for operating plants plants. The four futures and their likely outcomes would be only about 60 percent, and there would are summarized in table 37. be little progress in solving unresolved safety issues. poor management of operations would Future One.—Under the assumptions of Future continue to cause precursors to serious accidents, One, there is a clear need for more generating and the media wouId continue to give extensive capacity, the cost of other fuels is rising rapidly, coverage to these near-accidents and major con- operating plants are performing well, and industry struction problems. One operating event might offers improved, standardized designs. As dis- be so significant that a plant would have to be cussed in chapter 8, public acceptance of nuclear

Table 37.—Four Combinations of Economic and Management Scenarios Under the Base Case

Economic Scenario A: More favorable Economic Scenario B: Less favorable Fairly rapid growth in demand; utility Slow growth in demand; alternative alternatives costly; low interest/ energy available; high interest/inflation inflation. Management Scenario A: Major Future One improvement 7-year construction time; 70% per Some further orders possible, especial. availability; safer reactors; few precur- Iy if standardized preapproved designs sor events are available. Management Scenario B: Minor Future Three Future Four improvement 12-year construction time; 60°/0 A few well-managed utilities may order More orders very unlikely before 2000; availability; little progress on safety; plants over the next decade. a few possible after 2000. continued precursors. SOURCE: Office of Technology Assessment. 268 ● Nuclear Power in an Age of Uncertainty technology is influenced by perceived benefits As discussed earlier, Economic Scenario B en- as well as by perceived risks. Therefore, the in- visions that 84 GW of new electric-generating creased electricity demand in Economic Scenario capacity would have to be ordered only at the A, combined with the safety improvements envi- end of the 1983-93 decade. By then, the absence sioned in Management Scenario A could be ex- of orders and slowdown in construction would pected to increase support for nuclear energy. have eliminated many suppliers of nuclear plant Utilities would be more confident that plants components and services, increasing the cost of could be completed without unreasonable delays a new nuclear plant. This cost increase, com- due to changing regulations or intervention. bined with high inflation and interest rates, might offset the savings from reduced construction Under these circumstances, and if the 7-year times envisioned in Management Scenario A, fur- construction period for some recent plants ap- ther discouraging nuclear orders. Given the over- pears achievable for new plants as well, some all risks and uncertainties, it is unlikely that a util- utilities might be willing to order new nuclear ity would order a nuclear unit unless its projected plants even without changes in Federal policy. costs were much lower than a similar coal plant. However, most utilities would still be deterred This is unlikely, however, because poor business by uncertainties and risks, especially regulatory prospects would keep nuclear companies from delays and costs. investing heavily in the design and analysis Those utilities with a need for new capacity needed to reduce costs. might choose to share these risks by forming a At most, only one or two utilities could be ex- consortium. This consortium could order several pected to order a nuclear plant under Future plants based on the best current design and share Two, given no major change in Federal policy. the necessary startup costs with component man- Any orders that did occur probably would come ufacturers. Under current regulatoy procedures, from a very experienced nuclear utility and most the NRC could grant simultaneous construction likely would be in the form of initiating or com- permits for three or four new plants. The SNUPPS pleting construction at a currently inactive plant consortium in the early 1970’s is a prototype of site. However, it is more likely that there would such an effort (see ch. 5). If the plants were built be no orders at all over the next decade under in strict accordance with the complete design, these circumstances. Given some utilities’ cur- there would be only minimal requirements for rent problems with nuclear energy, no utility changes during construction. would want to be the only company venturing A problem-free licensing and construction into a new nuclear effort. process would demonstrate to other utilities the In the 1990’s, with few or no new orders, the benefits of standardization and show that at least U.S. nuclear industry would lose most of its ex- some utilities were committed to the technology. pertise in plant design and construction, and Given the need for additional power and the suppliers of key components would drop out. De- consortium’s expected success, nuclear orders spite these problems, some utilities could be in- might “snowball” without any major Federal terested in ordering reactors toward the end of action. the century, especially if demand growth starts Future Two.–The safety and management im- to increase. By then, the utilities would probably provements, reduced construction times, and ex- find a Japanese, French, or West German design pected increased public acceptance under Man- preferable to outdated U.S. plant designs. If U.S. agement Scenario A would make nuclear power companies wanted to offer the most current reac- a more attractive option. However, the assumed tor designs, they might have to license them from slow growth in electricity demand in Economic foreign companies, perhaps the very ones they Scenario B would make utility investment in any had licensed to build LWRs in the first place. In type of new powerplant unattractive throughout either case, as discussed in chapter 7, U.S. com- most of the decade. panies probably would still have the resources Ch. 9—Policy Options ● 269 to tailor the designs to American needs and to Future Four.–With the combination of both build the plants. Management Scenario B (Minor Improvement) and Economic Scenario B, there is little prospect Future Three.-The poor management condi- for any nuclear orders with no change in Federal tions of Management Scenario B, under a policy of no change in Federal policies, would create policies. Continued management and safety problems at plants currently operating and under serious tensions between the need for nuclear powerplants and continued opposition (by invest- construction, slow growth in electricity demand, and increasing competitiveness of other fuels, ors, critics, and the public) to their high costs and create a climate in which no utility could be ex- risk. As discussed previously, this scenario envi- pected to order a new nuclear plant. In the in- sions constraints on fossil fuel combustion, en- flationary environment expected under Economic couraging utilities to consider purchasing nuclear plants, Those utilities with successful experience Scenario B, utilities would be very reluctant to invest in capital-intensive plants of any type. In- in building and operating nuclear plants would not be deterred directly by the poor experiences stead, utilities could be expected to match supply and demand through load management, efficien- of others. Indirectly, however, all utilities are cy improvements to existing coal plants, and “tarred by the same brush” aimed at the weaker cogeneration, contributing to the slow overall rate utilities by the NRC actions, the public, investors, (1.5 percent) of growth in electricity demand. and others. Thus, the lack of major industry improvements envisioned in Management Scenario Near the end of the 1983-93 decade, some util- B would cause present uncertainties to continue. ity executives might order relatively inexpensive combustion turbines to supply the 84 GW of Moderate interest and inflation rates assumed new capacity needed by 2000. Although Eco- under economically favorable Economic Scenario nomic Scenario B expects moderate gas prices, A might help to offset the cost escalation resulting an increased reliance on gas and oil for electricity from the lengthy average construction period ex- generation might drive up prices, resulting in pected in Management Scenario B. This would much higher electricity prices. Despite high in- moderate the “rate shock” as new plants came terest and inflation rates, a few utilities might on line, reducing consumer opposition slightly. order coal- fired plants in the early 1990’s. The Public perception of increased electricity demand electricity from these plants would be rather ex- could be expected to offset concerns about safety pensive because of the high capital costs, further- risks slightly, perhaps returning public opinion to ing dampening growth. a 50-50 split on the technology. Most designers and equipment suppliers would The net effect of favorable economic condi- leave the business, leaving a much smaller U.S. tions combined with little change in the state industry. Because public opposition would be of the industry and no major Federal policy high under this combination of scenarios, utilities change would be that, at most, only a few would be unlikely to order new plants from utilities would order new nuclear plants over the abroad, and no new nuclear plants would be built next decade. Despite the constraints on fossil fuel in the United States before 2000. combustion, most new plants would be coal- fired, with environmental controls to meet cur- Even in Future Four, however, there is a rent regulations. With only a few new orders, possibility that new nuclear units could be built subsequent events would follow the path out- after 2000. As discussed in chapter 7, the U.S. lined for Future Two. In essence, the U.S. nuclear nuclear industry is very resilient. By maintaining industry wouId slowly decline, and any new nu- its expertise with fuel service and waste disposal clear plants ordered after 1995 might be of foreign business, the industry still should be capable of design. building a reactor (probably based on foreign 270 . Nuclear Power in an Age of Uncertainty

designs and components) if events after 2000 and regulated distribution companies. The the- created a renewed interest in nuclear power. oretical advantages and disadvantages of various kinds of deregulation and the many practical problems were analyzed in detail in a recent Base Case Variation: Let Nuclear comprehensive report (1 O). Power Compete in a Free Market Overall, the analysis concluded that the theo- Electric utility investment decisions are shaped retical benefits of more comprehensive forms of as much by regulatory practices as by the market. deregulation are sufficiently uncertain, and the This section notes some of the problems in in- practical problems sufficiently difficult, that any vestment decisionmaking which may have an im- move towards deregulation should proceed slow- pact on orders for nuclear powerplants and pro- ly and cautiously. The report also identified some posals intended to bring the discipline of the free limited steps toward deregulation that would en- market to generating capacity investments. courage the kind of competition among generating technologies that is most likely to lead to The first problem is that the regulated retail short- and long-term gains in efficiency. These price of electricity is based on the average cost steps include: 1 ) more Federal encouragement of all generating sources, but any incremental de- of power pooling and coordination, 2) rate struc- mand must be met with new generating capacity. tures (including experimental deregulation) for Under some conditions new generating capacity wholesale power sales regulated by FERC that en- costs considerably more than average. This is es- courage efficiency of operations, 3) encourage- pecially true if existing capacity includes a high ment of utilities to form generating and transmis- proportion of largely depreciated coal-burning sion companies within holding company struc- units or older nuclear units. Under these circum- tures, 4) mergers between small utilities partly to stances there is a mismatch between the price facilitate the contractual arrangements within signals consumers receive, and the decisions util- power pools, and 5) encouragement of retail rate ities have to make. structures that reflect the incremental cost of increased electricity demand. A second problem is that regulation combined with inflation can discourage investment in cap- Step 5 above includes “marginal cost pricing,” ital-intensive projects such as nuclear power- another category of proposed change. The rate plants even if they are the least expensive options structure could be adjusted so that customers pay in the long term. in times of inflation, regulators more for electricity at times of day or in seasons tend to delay increases in allowed return to equity when it costs more to provide. Alternatively, cus- investment, and the actual return lags behind tomers could be charged more for each incremental the allowed return. Furthermore, inflation block of electricity they purchase. A move toward combined with book value accounting tends to marginal cost pricing may be useful to encourage load the capital costs of a project into the early load management, especially in regions where years where it will be difficult to accommodate capacity utilization is poor. It is less obvious how them in the rate base. Some of these problems to use marginal cost pricing to improve the ac- have also been tackled in proposals for rate return curacy of the price signals with respect to nuclear (e.g., regional rate regulation, CWIP, trended power. Nuclear power is base load electricity original cost, etc. ) described elsewhere in this generation and would be affected only slightly chapter. by seasonal or time-of-day pricing. Further, rate regulation would have to be modified to account The most comprehensive proposals for changes accurately for the true marginal cost of nuclear fall under the general heading of the “deregula- generated electricity. Because of the peculiarities tion of electric utilities.” Such proposals range of current rate regulation described in chapter from selective deregulation of sales of wholesale 3 (and addressed in policy option A3), the cost power among utilities to a massive restructuring of the first year of nuclear power is more than of the electric utility industry into unregulated twice as much as the 20 or 30 year Ievelized cost generation and transmission (G&T) companies which is the true marginal cost. Ch. 9—Policy Options ● 271

Another problem with the current system is that clear power and the concerns of the public: (62) some of the costs of different sources of electricity improve utility management of nuclear plants, are not fully reflected in the private cost of such (D2) address the concerns of the critics, and (D3) sources. Nuclear power receives some direct or establish limits on future nuclear construction indirect Federal assistance of several kinds: 1 ) within the context of a balanced energy program. Federal limits on public liability following a nuclear accident (the Price-Anderson Act), 2) Fed- Of these policy options, the easiest to imple- eral subsidies for uranium enrichment, and 3) the ment is the involvement of the NRC in upgrading Federal cost of nuclear safety regulation and utility management. Such a program already ex- nuclear R&D. Coal-fired electricity also receives ists. The challenge is to make it motivate utilities Federal assistance from: 1) black lung payments to excellent performance (as is discussed in ch. (currently about $1 billion/year); 2) Federal coal 5) rather than merely to avoid getting in trouble mine regulation; and 3) no charge for air pollu- with the NRC. This is probably only possible if tion within regulatory limits. the NRC program is developed in close cooperation with IN PO. There are several possible ways investment comparisons for nuclear power and discussed earlier to involve interveners more competing generating technologies would be closely in monitoring and improving nuclear plant more accurate if these subsidies were eliminated. safety. This policy option is likely, however, to This study has not analyzed the consequences stimulate substantial opposition from utilities of reducing Government support, such as R&D unless it is clear that it is closely coupled with or the effects of eliminating the Price-Anderson licensing and backfit reform. liability limitations. It does seem clear, however, that such acts would be viewed by both the in- The effect of Strategy One on utilities’ percep- dustry and the public as signaling a lessening of tions of costs and schedules would be mixed. Li- the Government’s commitment to nuclear pow- censing and backfit reforms would help assure er. At this point, the industry can ill afford such utilities that they could build the plant as designed signals. Therefore, it should be recognized that once a construction permit were approved. How- taking these initiatives, whatever their overall ever, opening the process more to the critics in- merits, may well be tantamount to ending the nu- troduces another element of uncertainty, espe- clear option, at least for the foreseeable future. cially for the first few orders. Furthermore, these proposed reforms would not necessarily dramat- Strategy One: Remove Obstacles ically reduce capital costs, so electricity from to More Nuclear Orders average cost nuclear plants still could be perceived as more costly than electricity from The intent of this strategy (see table 35) for a average cost coal plants in most U.S. regions (see list of policy options) is to establish an environ- ch. 3). ment in which utilities would be more likely to consider nuclear reactors if demand does pick The impact of Strategy One on orders for new up over this decade and, at the same time, to es- nuclear plants also would vary sharply for each tablish policies that would win the support of nu- of the four nuclear futures described above for clear critics and the public. the Base Case. As can be seen in table 38, the The policy options included here work together impact of this strategy under each of the four to achieve these ends. Two of the options, (Al) futures does not differ greatly from the impact of revise regulation and (A2) develop standardized the Base Case. In light of the considerable dif- optimized LWR designs, are closely linked. These ficulty that would be involved in implementing options wouId eliminate some of the major con- the five policy options, this finding suggests that cerns utilities have over nuclear. It is more dif- Strategy One maybe only marginally effective. ficult to predict how to gain the necessary public support. Strategy One includes three policy op- Future One.—Under these conditions that are tions designed to reduce the controversy over nu- relatively favorable for more nuclear orders, Strat- 272 Ž Nuclear Power in an Age of Uncertainty

Table 38.-Alternative Nuclear Futures Under Strategy One: Remove Obstacles to More Nuclear Orders

Economic Scenario A: More favorable Economic Scenario B: Leas favorable Management Scenario A: Major Future One improvement A few orders likely if utilities are willing to be the first; a consortium and/or turnkey contracts could initiate ordering. Management Scenario B: Minor improvement

SOURCE: Office of Technology Assessment. egy One would increase the likelihood relative of good management on plant operations should to the Base Case. Utilities would have greater reassure nuclear critics and reduce the reasons confidence that a new reactor could be built for opposition to nuclear power in licensing pro- close to the projected cost and schedule. Most cedures and electricity rate hearings. The strong major design questions would have been worked need for more powerplants coupled with the rela- out before construction had started. The NRC tively high prices and environmental problems essentially would have approved the design and of coal in Economic Scenario A increase the rel- apparently would be ready to move an applica- ative advantage of nuclear and also reduce option through expeditiously. Critics would have position at State regulatory hearings. been given ample opportunity to critique the design. Controversy over nuclear power would be Future Two.– If economic and industry noticeably lower because the last of the present management conditions are less favorable than reactors under construction would be complete they are in Future One, the policies of Strategy without a continuation of the cost overruns they One will not succeed in stimulating very many are now experiencing, and operating reactors orders. In Future Two, industry management is would show considerably improved perform- assumed to improve substantially but electricity ance, demand grows slowly and inflation and interest rates are high, discouraging capital expenditures. Under such circumstances, vendors might be Because of fewer precursor incidents, public ac- willing to encourage the first orders by offering ceptance grows, but there is little obvious need a fixed price “turnkey” contract. The first few for nuclear powerplants. plants might not be much less expensive than present plants under construction, but simpli- With only 84 GW to be constructed by 2000, fied engineering and cumulative construction ex- there would be little pressure to diversify into perience would be expected to cut the cost of nuclear. With few prospective orders, vendors a typical plant 20 to 30 percent from current and architect-engineers would be unlikely to take levels (see chs. 3 and 5). Utilities might hesitate the risk of offering turnkey projects. A few util- to order the first few plants largely because of ities, seeking to preserve the option, might make doubts that cost and regulatory problems really a point of at least considering a few nuclear had been solved. Sharing the risk with the ven- plants. The new standardized designs, especialy dors would do much to alleviate these concerns. if smaller than current designs, and streamlined If the experiences of the first few orders were licensing could make reactors competitive with favorable, further orders could be expected in coal. Under Future Two, Strategy One has a line with demand growth. somewhat better chance than the Base Case of Intervener involvement in the licensing process leading to a few more orders by the end of the for standard designs combined with the effects century. Ch. 9—Policy Options . 273

Future Three. -Under these conditions, Strate- measures to reduce capital and operating cost of gy One actually could reduce the prospects for nuclear plants are combined with policy meas- new nuclear orders from what they are under ures to make nuclear power more acceptable to the Base Case. With continued poor plant nuclear critics and the public. operating performance and continued precursor events, nuclear critics will not be satisfied that The first policy option is a demonstration of the adequate progress is being made. With inter- reactor(s) developed under A2 as discussed in venors more closely involved in safety regulation, Strategy One. The main purposes would be to the lack of progress on resolving safety concerns show that the reactor could be built according could increase the time devoted to particular safe- to design and that the licensing process could ty issues. Even with high growth, the utilities are handle it as expected. It could be expected that likely to regard more nuclear orders as too risky, private industry would fund most of this project not only from technical problems raised by the since technological feasibiIity is not i n question, NRC and the critics, but also from the financial but significant Federal participation could be impact of public opposition on rate hearings and required. investor decisions. A second step is (A3) Stimulate improved rate Future Four.– Finally, the policy options of regulation treatment of capital-intensive projects. Strategy One could diminish still further the pros- No Federal budget is required for this policy op- pects for nuclear orders under the dismal condition but sensitive Federal leadership is needed tions of Future Four which combines both ad- since rate regulation traditionally has been left verse economic conditions and little industry im- to the states. provement. The combination of intervener in- Strategy Two adds a politically controversial volvement with poor industry improvement and step to improve operating economics with (64) little apparant need for nuclear power could Reduce uncertainty about occupational exposure create conditions of public opposition that liability. Clarifying occupational exposure con- would make orders unlikely even after 2000. ditions and ranges of possible payments to exposed workers with health problems might add Strategy Two: Provide Moderate expense to the operation of nuclear plants but Stimulation to More Orders would reduce the uncertainty which accompa- nies an unspecified liability (which could be the Strategy Two builds on Strategy One. It assumes subject of private lawsuits, such as is now hap- that efforts to remove obstacles to more nuclear pening with asbestos exposure liability). orders would be inadequate, largely because utilities still would be unconvinced that the prob- Strategy Two includes three additional options lems were resolved. As in Strategy One, policy to reduce the risk of a reactor accident. (C2) would

Table 39.—Alternative Nuclear Futures Under Strategy Two: Provide a Moderate Stimulus to More Nuclear Orders

Economic Scenario A: More favorable Economic Scenario B: Less favorable Management Scenario A: Major Future One improvement Some plants ordered by about 1990; more by the end of the century.

Management Scenario B: Minor Future Three Future Four improvement New orders Iikely only if Government Government actions to improve options to improve management have a management and R&D funding of new big impact on utilities and public opin- reactor types should improve pros- ion, resulting in Future One; new reac- pects for more orders after 2000. tor types would be useful. SOURCE: Office of Technology, 274 . Nuclear Power in an Age of Uncertainty

require the NRC to restrict construction licenses 1993, more coal fired capacity would have been for nuclear plants to qualified utilities and con- ordered over the previous 10 years than existed struction contractors. Substantial Federal funding, in 1983, but concerns over the environmental im- up to several billion dollars over 10 years or more, pacts would be rising sharply. Stringent new emis- would be required for a second step: (C3) Fund sion regulations would appear probable, but a major R&D program on alternative reactor would be very expensive. The costs of other fuels types. If this option is pursued at a high level, the also would be rising rapidly. Other industrial demonstration LWR probably would be deleted. countries, especially France, Japan, West Ger- For a third option, (C4) Revise institutional man- many, Great Britain, and Canada would have agement of utilities, there could be controversial maintained their nuclear programs, and their changes in Federal and State utility regulation, cheaper electricity could give them a competitive antitrust law and other long-standing institutional edge in certain areas of international trade. Util- guidelines. These steps should ensure the avail- ities would have raised their standards of nuclear ability of reactors and operators in which the pub- operation such that mishaps rarely would occur lic could have great confience. and reliability would be high. Americans in 1993 might well wonder what went wrong with our Substantial changes in public acceptance of nu- national decision making, clear power are needed to accumulate the political capital for either large Federal budget expen- What would Strategy Two have accomplished ditures or for Federal efforts to change utility rate in the same period? How would utility uncertain- regulation or utility institutional management. ty have been reduced? Under the conditions of Two options included in Strategy Two go further Strategy Two, utilities would have a clear dem- than the steps in Strategy One to alleviate the onstration of Federal government commitment public’s concerns about nuclear power and to resolving the problems with nuclear power and sharpen the basis for judgment about the long- good reason to expect that nuclear plants def- term need for nuclear power. (D4) reduces con- initely would be cheaper than coal plants in most cerns over the linkage between nuclear power regions of the country. and nuclear weapons by such steps as banning By 1990, the standardized design would be suf- reprocessing. (D5) encourages or requires the use ficiently well proven that interveners, NRC and of regional planning and perhaps rate regulation, nuclear designers would be satisfied that it was to improve the consensus on the long-term need very resilient to any accident sequences anyone for capital intensive technologies such as nucle- suggests. It would be clear to all that construc- ar power. tion of the first standardized plant was proceeding Future One.–Strategy Two would have a big smoothly on schedule and within budget. The impact on conditions such as those in Future One NRC would offer a construction permit based on which hypothesizes rapid demand growth and its previous review of the existing design, with successful industry improvement, but also util- only site specific characteristics to be approved. ity reluctance to place the first nuclear order after The financial burden on utilities would be less- a hiatus in orders and substantial public contro- ened because of reduced construction cost and versy. changes in rate regulation, and, perhaps, because smaller reactors would be available. Alternatively, Perhaps the best vantage point from which to the HTGR or one of the other reactors could be appreciate the possible need for Strategy Two is in an advanced stage of development. from a look forward to 1993 under the assumptions of Future One if there have been no nuclear Given the strong demand, low interest rates orders even though Strategy One has been and relative unattractiveness of other fuels in implemented. It is likely that the main ob- Economic Scenario A, there is little doubt that stacle to any nuclear orders would be utility un- utilities would order nuclear plants if they were certainty that all the problems had actually been convinced costs would come down and public resolved, despite the steps in Strategy One. By acceptance would be sufficient to avoid serious Ch. 9—Policy Options ● 275

economic risk to the plants. Nuclear critics and Three and Future Four. The consensus that would the public would have been reassured by a dec- be required for the large Federal budget expend- de of steadily improving nuclear plant perform- itures and the changes in Federal-State relations nce, the closer involvement of interveners in on utility rate regulation wouId not emerge. licensing, a nationally agreed limit on future For Future Three, with favorable economic nuclear construction and the several other meas- conditions for nuclear orders, Strategy Two might res of Strategies One and Two. In addition, the be implemented in stages beginning with options obvious need for generating capacity would lead to restrict construction licenses to qualified util- to considerably greater public acceptance for ities (C2)—and perhaps revoke operating licenses more nuclear plants by 1993. as appropriate,—and changes in utility institu- Under these conditions, some plants would tional structures (C4), allowing utility service com- be ordered in the early nineties and probably panies to take over poorly run plants. In effect, a larger number in the late nineties. Strategy Two these actions would change the situation to that thus makes probable what is only a possibility of Future One. under the same favorable conditions of Strategy If these actions are insufficiently successful, One. While it is difficult to be precise, this policy future orders probably would be contingent on package could lead to more nuclear orders even the development of inherently safe reactors under conditions somewhat less stimulating than which, by allaying many safety concerns, wouId in Economic Scenario A. A growth rate in elec- restore some degree of public acceptance. tricity demand averaging 2.5 percent might have the same probability of initiating nuclear orders Future Four.–Strategy Two would be very as 3.5 percent did under Strategy One. hard to justify for Future Four with both little industry improvement and unfavorable economic Future Two.– If utility management were im- conditions. The options to improve utility man- proved but economic conditions were much less agement, however, would help improve public favorable to nuclear orders, as in Future Two, acceptance for after 2000, which is the earliest much of the effort going into standardized design nuclear orders might be placed. It also is possi- could be wasted. When the time came for new ble that a modest level of investment in new reac- orders n-ear or past 2000, a foreign design, an tor types could lead to a more ambitious effort alternative reactor type, or possibly photovoltaics in the nineties when public acceptance would might appear more appropriate. Thus, this effort be improved as a result of NRC efforts to improve might be dropped if demand growth stays low. utility management. An effective program to develop alternative reactor technology could prove to be the crucial re- Strategy Two is clearly a high-risk, high-gain assurance to utilities contemplating new orders strategy. Under the circumstances of Future One after 2000. Utilities that might be unwilling to be it could assure the future of the U.S. nuclear in- the first or second to order a new reactor type dustry, although it might not even be necessary might be willing to be the third or fourth. to stimulate nuclear orders if the industry itself takes sufficient steps to improve the technology The other steps of Strategy Two (coupled with and if demand for electricity grows rapidly. Under two decades of good management) would be other circumstances, Federal efforts and funds useful in changing the climate of public opinion could produce little of value (Future Two) or be to be more receptive to nuclear power in the long impossible to carry out (Futures Three and Four). term. With demand growth of only 1.5 percent annually, such changes are likely to be impor- Strategy Two Variation: Support tant even when there has been good utility management. the U.S. Nuclear Industry in Future World Trade Future Three. -Strategy Two might not be feasible if utility management of nuclear reactors im- In this variation of Strategy Two, the Federal proved very little, as is assumed under Future Government also would play an activist role in 276 Ž Nuclear Power in an Age of Uncertainty

supporting the U.S. nuclear industry. The ration- on R&D and high technology industries may ale for this variation would come, however, from spend up to 10 to 15 percent. a completely different source. Rather than regard A major increase in R&D for the electric utility nuclear power as an element of U.S. energy pol- industry could be allocated among: icy, the strategy would treat the nuclear industry, and related electrotechnologies, as key elements ● elect rotechnologies for industry such as of an emerging U.S. industrial policy. plasma reduction of iron or microwave heating; Several assumptions about the current nature ● commercialization of photovoltaics and the of world competition in nuclear technologies solid-state control technology needed to in- underlie this approach. One is that the advanced tegrate them in the grid; and reactor designs now underway with joint U. S.- ● a more advanced nuclear reactor for beyond Japanese teams will establish a new standard 2000. LWR for the 1990’s that will make more modest improvements obsolete. These designs—probably It seems likely that a balance among support for to be licensed jointly by Japanese and U.S. different advanced generating technologies, in- vendors—may well account for any nuclear or- cluding photovoltaics as well as advanced nuclear ders in the 1990’s. These designs should give both reactors, will be necessary to get widespread sup- the United States and Japan a very strong posi- port. Utilities also are likely to favor diverse R&D tion in world nuclear trade. because of the financial risk inherent in relying on single technologies. The second assumption underlying this approach is that there will beat least one more gen- One approach to obtaining funding for such eration of non breeder reactors after these ad- an R&D effort is to make the treatment of R&D vanced LWRs become available. The slowdown in utility rate regulation more attractive. The in plant construction, potential for extended telephone industry traditionally has included burnup and new uranium discoveries all would R&D in the rate base in most States. make it likely that nonbreeder reactors will be Federal action in support of this policy strategy competitive for several more decades. The stand- would have several elements: ardized LWR discussed in option A2 or one of the advanced alternative reactors in C3 could be 1. Selection of these technologies as a signifi- the choice. cant element of U.S. long-term industrial pol- The question is: should the U.S. Government icy, and expansion of these R&D programs. encourage a strategy of R&D that leads to the next 2. Legislation providing incentives or requiring stages of reactor development beyond the joint States to allow R&D expenditures in the cost- Japanese-U.S. vendor projects? Several elements of-service. of a general industrial policy could be applied to 3. Increased Federal funding for some joint nuclear power: 1 ) support for R&D into product EPRI-DOE projects, development, 2) relaxation of antitrust prohibi- 4. Elimination of antitrust penalties for vendor tions on cooperation among businesses during cooperation on an advanced reactor design. the product development stage of R&D, and 3) 5. Consideration of long-term loans such as the financing assistance in export promotion through Japanese government made available to the the Export-Import Bank (7,8). semiconductor industry. 6. Attractive financing to foreign customers for Such a strategy might be especially productive nuclear technology exports-through the Ex- because of the historically low spending on R&D port-import Bank. by the electric utility industry. The $250 million budget of the Electric Power Research Institute Implementation of such a strategy would give (EPRI) represents only about 0.25 percent of elec- the U.S. a headstart in world competition just tric utility sales each year. General manufactur- when U.S. orders could be expected to pick up ing industries spend about 2 to 3 percent of sales because of load growth and replacement of ex- Ch. 9—Policy Options ● 277

isting plants. It could result in a healthy nuclear tric utilities. Utilities should derive a major part industry with brighter export prospects, and more of the benefit, aside from being able to buy the attractive options of both supply and demand for developed product. Another difficulty is that elec- the electric utility industry. Electricity consuming trical demand technologies and advanced gen- industries would also benefit, some quite signifi- erating technologies may not seem as important cantly. as other U.S. industries in claiming a role in a general high-technology industrial strategy. Cur- There are some obvious difficulties with this rent strategies tend to focus on computer, semi- strategy. Given the Federal system, it may be very conductor manufacture, and biotechnology rath- difficult to get State support for more favorable er than traditional manufacturing industries. The treatment of utility R&D expenditures. Although R&D effort would also have to be designed with some R&D will be carried out by manufacturers, care to ensure that the funds are spent produc- a comprehensive strategy will not work unless a tively. major part of the impetus comes from the elec-

CONCLUSIONS

The scenarios and discussions of the previous ● PUCs would have more confidence that a section underscore the conclusion that there is utility would not bankrupt itself with a new no simple key to restoring nuclear power as a na- reactor because costs would be predictable tional energy option. Nor is there any assurance at the start due to matured designs and reg- that even a rigorous set of policy initiatives would ulatory stability, and operation of existing do so. Uncertainties over the future growth rate plants had proven to be in the best interests of demand for electricity are at least as impor- of consumers. tant as the policy options analyzed here. At the ● Investors could expect more favorable reg- very least, a moderate demand growth rate and ulatory treatment from the PUCs as well as moderate improvements in management of exist- having more confidence in the economic ating plants are prerequisites for creating the con- tractiveness of nuclear power. ditions under which utilities could consider order- ● Critics would have far fewer specific con- ing additional nuclear plants. cerns with safety and the overall threat of a “nuclear economy. ” This would lower the The problems and uncertainties are great but intensity of the controversy even if few critics not insurmountable. If the technology and its changed their minds on the inherent desir- management improve sufficiently that confidence ability of nuclear power. in both safety and economics is much higher, if ● The public would be more supportive nuclear regulation shows a parallel improvement, because of the lowered controversy, the im- and if financial risks are shown to be less than proved operating records of existing plants, financial rewards, then nuclear power would be and the more visible benefits. a logical part of our energy future. To see how ● Finally, utilities, or generating entities set up this might be so, consider the seven sides to the to replace or supplement them, would see nuclear debate discussed in chapter 1: this improved environment for nuclear power, the predictability of costs and opera- ● The nuclear industry would have a product tions, and the affordability and com- that was thoroughly analyzed, demonstrably petitiveness of the new plants, and would be safe and economically competitive. much more receptive to proposals for new ● The NRC would have exhaustively examined nuclear plants. the design and be confident that few additional issues would be raised once a con- The purpose of the policy options discussed struction permit had been granted. above is to assist in the transition from the pres-

278 . Nuclear Power in an Age of Uncertainty

Photo credit: William J. Lanouette

Demonstration in front of the Capitol after the accident at Three Mile Island ent situation to this future. However, it is impossi- tors of Strategy Two would be more reassuring, ble to say with any certainty how much improve- but would introduce cost and operational uncer- ment in any one factor is necessary, or how much tainties. How far is it necessary to go to achieve each policy initiative would contribute. These a consensus that reactors are safe enough? It problems are interdependent: none can be seems reasonable to assume that greater safety solved in isolation, but the progress on each will would result in greater acceptance, but there is assist in the solution of others. little direct evidence to support that view or to quantify the relationship. The future listed above corresponds to Future One in the previous section, coupled with as Any policy strategy should be flexible enough many policy initiatives as it takes to start nuclear to try things that may not work or may be found orders. Strategy One might be adequate. Strategy inappropriate to changing conditions, and to dis- Two probably would be, but we cannot say for card the less successful initiatives in favor of the sure. The uncontrollable uncertainties are simply better ones. The strategy should include elements too great. For instance, future plants can be de- dealing with the technology, its management, nu- signed to be safer than existing plants. industry clear regulation, financial risk, and public accept- already is working on this. The national design ance. of Strategy One would improve on this and be more convincing to critics because of the open- Additional nuclear plants essentially have been ness of the design effort. The inherently safe reac- rejected by the American people because of per- Ch. 9—Policy Options ● 279

ceptions of the current technology and its man- proved significantly while knowledgeable critics agement. Major improvements will have to be still are voicing real concerns. made to restore their faith. To be successful, a The outlook for the nuclear supply industry is strategy must recognize the different concerns bleak but not hopeless. New policy initiatives can and try to balance the interests. In particular, the set the stage for a turnaround. Without appro- role of the critics in any nuclear resurgence will priate action, it is likely that the option of be crucial. Critics have been the messengers to domestically produced nuclear reactors gradually the public of many of the real problems with nu- will fade away. If such is to be the future, the clear power. They will continue to play that role decision should be made consciously, with the until they have been shown that the problems knowledge that even the nonnuclear futures have been solved or rendered inconsequential. available to the Nation contain risks and They also can be the messengers, even if only drawbacks. Nuclear power can be a significant by losing interest, when they are convinced that contributor to the Nation’s energy future, but nuclear power as a whole is a minor problem only if the efforts to restore it are undertaken with compared to other societal concerns. It is difficult wisdom, humility, and perseverance. to conceive of how public acceptance can be im-

CHAPTER 9 REFERENCES

1. Bauer, Douglas C., “Deregulation: Motivations, nology: Public Policies for the 980’s, ” Nationa Modalities and Prospects, ” paper presented at the Journal, March 1983. conference on E/ectr/c Power Generation orga- 9. Meyer, Michael B., “A Modest Proposal for the nized by the Energy Bureau, Inc. Partial Deregulation of Electric Utilities, Pub/ic 2. Committee on Nuclear and Alternative Energy Sys- Utilities Fortnjghtly, Apr. 14, 1983. tems, Energy in Transition 1985-2000 (Washing- 10. Massachusetts Institute of Technology, Deregu/a- ton, D. C.: National Academy of Sciences, 1979. ) tion o~Electrjc Power: A Framework for Analysis, 3. Fitzgibbons, Harold E., “A European Perspective Phase I report, DOE/N BB-0021, September 1982, on U.S. High Technology Competition, ” in “High published as: Paul Joskow and Richard Schmalen- Technology: public Policies for the 1980’ s,” ~a- see, Markets for Power: An Analysis of Electric tiona/ Journal, March 1983. W/ity Deregu/atjon (Cambridge, Mass.: MIT Press, 4. Ford Foundation, Nuclear Power Issues and 1983). Choices (Cambridge, Mass.: Ballinger, 1977). 11 Office of Technology Assessment, /ndustria/ and 5. Golub, Bennet W., and Hyman, Leonard S., “The Cornmercia/ Cogeneratjon, OTA-E-1 92, February Financial Difficulties and Consequences of Dereg- 1983. ulation Through Divestiture, ” Pub/ic Uti/ities Fort- 12. Oosterhuis, Paul, “High Technology Industries night/y, Feb. 17, 1983. and Tax Policy in the 1980’ s,” in “High Technol- 6. Graham, A. J., “Electricity Pricing: An Industrial ogy: Public Policies for the 1980’ s,” Natjona/ jour- View, ” printed in Energy Technology V. Deregu- nal, March 1983, lation. 13. Radford, Bruce C., “Marginal Cost Pricin~ Consid- 7. Krist, William R., “The U.S. Response to Foreign ered i n Recent Electric Rate Cases, ” Pub/ic Uti/ities Industrial Policies, “ in “High Technology: Public Fortnight/y, Oct. 9, 1980. policies for the 1980’ s,” Nat/ona/Journa/, March 14 Samuels, Warren J., “Problems of Marginal Cost 1983. Pricing in Public Utilities, ” Pub/ic Uti/ities Fort- 8. Merrifield, D. Bruce, “Forces of Change Affect- night/y, Jan. 31, 1980. ing High Technology Industries, ” in “High Tech- Appendix Appendix List of Acronyms and Glossary

List of Acronyms MITI –Ministry of International Trade and In- dustry (Japan) ACRS —Advisory Committee on Reactor NAS –National Academy of Sciences Safeguards NASA –National Aeronautics and Space AE —architect-engineers Administration AEC –Atomic Energy Commission NCRP –National Commission on Radiological AFUDC –allowance for funds used during Protection construction NEIL –Nuclear Electric Insurance Ltd. AIF –Atomic Industrial Forum NEPA –National Environmental Policy Act AGR —advanced gas reactor NERC –Northeast Electric Reliability Council ANS –American Nuclear Society NML –Nuclear Mutual Ltd. ASME –American Society of Mechanical NPCC –Northeast Power Coordinating Council Engineers NPDES —National Pollutant Discharge Elimination BWR —boiling water reactor System permit CAA –Clean Air Act NPRDS –Nuclear Plant Reliability Data System CEGB –Central Electricity Generating Board NPT –Non-Proliferation Treaty COL —construction and operating license NRC —Nuclear Regulatory Commission CP —construction permit NSAC —Nuclear Safety Analysis Center CPCN –Certificate of Public Convenience and NSSS —nuclear steam supply system Necessity NTOL —near-term operating licenses CRBR –Clinch River Breeder Reactor NUTAC –Nuclear Utility Task Action Committee CRGR –Committee for Review of Generic OECD –Organization for Economic Cooperation Requirements and Development CRS –Congressional Research Service OL —operating license CWA –Clean Water Act OTA –Office of Technology Assessment CWIP —construction work in progress PCRV –prestressed concrete reactor vessel DOE –Department of Energy PG&E –Pacific Gas & Electric Co DRI –Data Resources, Inc. PHWR –pressurized heavy water reactor EIS –Environmental Impact Statement Plus –process inherent ultimately safe reactor EPRI –Electric Power Research Institute PRA –probabilistic risk assessment ERDA —Energy Research and Development PSAR –Preliminary Safety Analysis Report Administration PUC —public utility commission ER –Environmental Report PURPA —Public Utility Regulatory Policies Act FEMA –Federal Emergency Management Agency PWR –pressurized water reactor FERC —Federal Energy Regulatory Commission RNPA –regional nuclear power authority FSAR –Final Safety Analysis Report RNPC –regional nuclear power company GAP –Government Accountability Project SAI –Science Applications, Inc. GCR –gas cooled reactor SEE-IN –Significant Events Evaluation and Infor- GNP –gross national product mation Network GPU –General Public Utilities SER –Safety Evaluation Report HTGR —high temperature gas-cooled reactor SNUPPS–Standardized Nuclear Unit Power Plant HWR –heavy water reactor System IDCOR –Industry Degraded Core Rulemaking SPDS –Safety Parameter Display Systems Program SPP –Southwest Power Pool IN PO —Institute of Nuclear Power Operations SRP –Standard Review Plan JCAE –Joint Committee on Atomic Energy TM I –Three Mile Island KWU —Kraftwork Union TVA –Tennessee Valley Authority LMFBR –liquid metal fast breeder reactor WPPSS –Washington Public Power Supply LWR –light water reactors System

283 284 ● Nuclear Power in an Age of Uncertainty

Glossary core of a reactor is known as the primary coolant. It absorbs heat in the core and then transfers it to absorption, neutron: Any reaction in which a free a secondary coolant system. neutron is absorbed by a nucleus, including cap- core: The region of a reactor in which the nuclear ture and fission. chain reaction is initiated, maintained, and con- Allowance for Funds Used During Construction trolled. Coolant is constantly circulated through the (AFUDC): An account in the income statement of core to remove heat produced by the fission a utility in which interest is accumulated on the con- process. struction expenditures for construction work in decay heat: The heat produced by radioactive decay progress that has not been entered into the utility’s of materials that are primarily the remnants of the rate base and is therefore not yet earning a cash chain reaction. return on investment. The accumulated interest is deplete: To reduce the fissile content of an isotopic then added to the actual construction expenditures mixture, particularly uranium. when the plant enters the rate base. elasticity: The ratio of change in demand for a prod- base loaded: Keeping a power station continuously uct (in this case electricity) to change in a category loaded at the maximum load because it is one of of prices, or to change in income. the lowest cost power producers on the system. emergency core cooling system: Any engineered boiling water reactor: A reactor cooled by water that system for cooling the core in the event of failure is allowed to boil as it passes through the core. This of the basic cooling system, such as core sprays or coolant is used directly to produce the steam which injectors. generates electricity. enrichment: The process of increasing the concen- capacity factor: Ratio of average plant electrical tration of one isotope of a given element. energy output to rated output. fabrication: The final step in preparing nuclear fuel chain reaction: The continuing process of nuclearfis- for use in a reactor. sioning in which the neutrons released from a fis- fast breeder reactor (FBR): A reactor cooled by liquid sion trigger at least one other nuclear fission. sodium rather than waste. In this type of reactor, cladding: The term used to describe any material that the transformation of uranium-238 to plutonium oc- encloses nuclear fuel. In a water-cooled power reac- curs readily. Since plutonium fissions easily, it can tor this is the fuel rod tube. be recycled and used as fuel for a breeder reactor. Construction Work in Progress (CWIP): An account The conversion of uranium to plutonium is so effi- on the asset side of the utility’s balance sheet that cient in an FBR that this reactor creates more fuel includes all construction expenditures for plant and than it consumes. equipment on plant that has not yet been placed feedwater: Water, usually from a condenser, supplied i n service. to replenish the water inventory of components construction Ieadtime: The time required to complete such as boilers or steam generators. construction of an electric generating plant, usual- fertile: Material composed of atoms which readily ably defined from either date of reactor order or con- sorb neutrons to produce fissionable materials. One struction permit to commercial operation. such element is uranium-238, which becomes plu- containment building: A thick concrete structure sur- tonium-239 after it absorbs a neutron. Fertile rounding the pressure vessel and other reactor com- material alone cannot sustain a a chain reaction. ponents. It is designed to prevent radioactive fissile: Material composed of atoms which readily fis- material from being released to the atmosphere in sion when struck by a neutron. Uranium-235 and the unlikely event that it should escape from the plutonium-239 are examples of fissile materials. pressure vessel. fission: The process by which a neutron strikes a control rods: Long thin rods that are positioned nucleus and splits it into fragments. During the proc- among fuel rods to regulate the nuclear chain reac- ess of nuclear fission, several neutrons are emitted tion. Control rods are composed of material that ab- at high speed, and heat and radiation are released. sorbs neutrons readily. They interrupt or slow down fission products: The smaller atoms created when a a chain reaction by capturing neutrons that would nucleus fissions. The mass of the fission products otherwise trigger more fissions. is less than that of the original nucleus. The dif- coolant: Fluid that is circulated through the core of ference in mass is released as energy. a reactor to remove the heat generated by the fis- fossil plant: A powerplant fueled by coal, oil, or gas. sion process. In reactors that have more than one fuel: Basic chain-reacting material, including both coolant system, the fluid which passes through the fissile and fertile materials. Appendix: List of Acronyms and Glossary Ž 285

fuel cycle: The set of chemical and physical opera- curring during that period. Also used as a tions needed to prepare nuclear material for use in synonym for capacity factor. reactors and to dispose of or recycle the material load eye/e pattern is the variation in demand over after its removal from the reactor. Existing fuel cycles a specified period of time. begin with uranium as the natural resource and loss-of-coolant accident (LOCA): A reactor accident create plutonium as a byproduct. Some future in which coolant is lost from the primary system. cycles may rely on thorium and produce the fissile market potential: The number of instances in which isotope uranium-233. a technology will be sufficiently attractive-all things fuel rod: An assembly consisting of a capped zircalloy considered—that the investment is likely to be or stainless steel tube filled with fuel pellets. made. half life: The period required for an unstable radio- market-to-book ratio: The ratio of the market price active element to decay to one-half of its initial of a firm’s stock to its book value. mass. MWe: Megawatts of electrical energy. heat rate: A measure of the amount of fuel used to MWt: Megawatts of thermal energy. produce electric and/or thermal energy. moderator: A component (usually water, heavy water, total heat rate refers to the amount of fuel (in Btu) or graphite) of some nuclear reactors that slows neu- required to produce 1 kilowatt-hour of electrons, thereby increasing their chances of being ab- tricity with no credit given for waste heat use. sorbed by a fissile nucleus. incremental heat rate is calculated as the addi- neutron: A basic atomic particle that has no electrical tional (or saved) Btu to produce (or not pro- charge. Neutrons and protons, which are positive- duce) the next kilowatt-hour of electricity. ly charged particles, form the central portion of the net heat rate (also measured in Btu/kWh) credits atom known as the nucleus. Negatively charged the thermal output and denotes the energy electrons orbit the nucleus at various distances. The required to produce electricity, beyond what chemical and nuclear properties of an atom are wouId be needed to produce a given quanti- determined by the number of its neutrons, protons, ty of thermal energy in a separate facility (e.g., and electrons. a boiler). neutron poison: The general name given to materials interest coverage ratio: The ratio of a firm’s earnings that absorb neutrons. These materials either in- to its current interest obligations. terfere with the fissioning process or are used to isotopes: Atoms having the same number of protons, control it. but a different number of neutrons. Two isotopes nuclear island: The buildings and equipment that of the same atom are very similar and difficult to comprise the reactor and all its emergency and aux- separate by ordinary chemical means, Isotopes can iliary systems. have very different nuclear properties, however. For nuclear steam supply system (NSSS): The basic reac- example, one isotope may fission readily, while tor and support equipment, plus any associated another isotope of the same atom may not fission equipment necessary to produce the steam that at all. drives the turbines. light water reactor: A general term that refers to all once-through fuel cycle: A nuclear system wherein nuclear reactors which use ordinary water as a cool- nuclear materials are introduced into a reactor only ant. This includes pressurized water reactors and once; they are not recycled. boiling water reactors, which are the predominant plutonium: An element that is not found in nature, reactors in the United States. but can be produced from uranium in a nuclear load: The demand for electric or thermal energy at reactor. Plutonium fissions easily, and can be used any particular time. as a nuclear fuel. base load is the normal, relatively constant de- power density: The power generated per unit volume mand for energy on a given system. of the core. peakload is the highest demand for energy from pressure vessel: A heavy steel enclosure around the a supplying system, measured either daily, core of a reactor. It is designed to withstand high seasonally, or annually. pressures and temperatures to prevent radioactive intermediate load falls between the base and material from escaping from the core. peak. pressurized water reactor: A reactor cooled by water load factor is the ratio of the average load over that is kept at high pressure to prevent it from boil- a designated time period to the peak load oc- ing. Primary coolant passes through the core of a 286 ● Nuclear Power in an Age of Uncertainty

PWR, and then transfers its heat to a secondary water in which spent fuel elements are stored pend- coolant system. Steam is produced from the heated ing their shipment from the facility. water in the secondary system. spent nuclear fuel: Material that is removed from a primary coolant: The fluid used to cool the fuel reactor after it can no longer sustain a chain reac- elements. It may be liquid or gas. tion. Spent fuel from a light water reactor is com- qualifying facility: A cogenerator or small power proposed primarily of uranium and contains some ducer that meets the requirements specified in the radioactive materials, such as fission products. Spent Public Utility Regulatory Policies Act of 1978–in the fuel also contains some valuable nuclear materials, case of a cogenerator, one that produces electrici- such as uranium-235 and plutonium. ty and useful thermal energy for industrial, commer- steam generator: The main heat exchangers in a cial, heating, or cooling purposes; that meets the pressurized water or gas-cooled reactor powerplant operating requirements specified by the Federal En- that generates the steam that drives the turbine gen- ergy Regulatory Commission with respect to such erator. factors as size, fuel use, and fuel efficiency); and thermal efficiency: In a powerplant, the ratio of net that is owned by a person not primarily engaged electrical energy produced to total thermal energy in the generation or sale of electric power (other released in the reactor or boiler. than cogenerated power). thermal load: The stresses imposed on a component radioactive decay: The process by which a nucleus due to restriction of thermal growth caused by tem- of one type transforms into another, accompanied perature changes. by emission of radiation. thermal neutron: A neutron whose energy level has radioactive waste: Waste materials, solid, liquid, or been lowered sufficiently so that upon collision with gas, that are produced in any type of nuclear facility. another atom it will cause the atom to split and rate base: The net valuation of utility property i n serv- release energy. Neutron energy levels can be ice, consisting of the gross valuation minus accrued lowered by recoil off moderating atoms. depreciation. thorium-232 (Th232): A fetile, naturally occurring iso- reactor: A facility that contains a controlled nuclear tope from which the fissile isotope uranium-233 can fission chain reaction, It may be used to generate be bred. electrical power, to conduct research, or exclusively turbine generator: The assembled steam turbine to produce plutonium for nuclear explosives. coupled to an electric generator that produces the reactor containment boundary: The pressure enve- electric power in a powerplant. lope in which a reactor and its primary cooling sys- uranium: A metallic element found in nature that is tem are located. commonly used as a fuel in nuclear reactors. As reactor vessel: The container of the nuclear core or found in nature, it contains two isotopes— critical assembly; may be a steel pressure vessel, a uranium-235 and uranium-238. prestressed concrete reactor vessel (PCRV), or a uranium-233 (U233): A fissile isotope bred by fertile low-pressure vessel (e.g., a calandria or sodium pot). thorium-232. It is similar in weapons quality to plu- reprocessing: Chemical treatment of spent reactor fuel tonium-239. to separate the plutonium and uranium from the fis- uranium-235 (U235): The less abundant uranium sion products and (under present plans) from each isotope, accounting for less than one percent of nat- other. ural uranium. Uranium-235 splits, or fissions, when safeguards: Sets of regulations, procedures, and struck by a neutron. When uranium is used as a fuel equipment designed to prevent and detect the in a nuclear reactor, the concentration of urani - diversion of nuclear materials from authorized um-235 is often increased to enhance the fission channels. process. For example, the fuel for light water reac- safety system: A mechanical, electrical, or instrumen- tors contains about 30/0 uranium-235. tation system or any combination of these, whose uranium-238 (U238): The more abundant uranium purpose is the safety of the reactor or of the public. isotope, accounting for more than 99 percent of nat- scram: The rapid shutdown, via introduction of ural uranium. Uranium-238 tends to absorb neu- neutron absorbers, of the chain reaction. trons rather than fission. When it absorbs a neutron, seismic load: The stresses imposed on a component the uranium atom changes to form a new element- by a seismic shock. plutonium, shutdown: The act of stopping plant operation for an y water hammer: The shock load imposed on a flow- reason. ing pipeline by the rapid closure of a shutoff valve. spent fuel storage pool: The pool of demineralized Index Index

Advisory Council on Historic Preservation, 149 Certificates of Public Convenience and Necessity Advisory Committee on Reactor safeguards (ACRS), 145, (CPCN), 151 164, 165 C. F. Braun, Co., 136 Alternative reactor systems, 83-109, 258 (see also Cincinnati Gas & Electric Co., 116, 157 nuclear powerplant technology) Colorado Public Service Co., 101 advanced light water reactor design concepts, 94-96 combined construction and operating license (COL), basics in nuclear powerplant design, 84 165, 168, 169 heavy water reactors, 96-99 Combustion Engineering, Inc., 87 high temperature gas-cooled reactor (HTGR), 99-102, Committee for Energy Awareness, 214, 236 103 Committee for Review of Generic Requirements inherently safe reactor concepts, 102-105 (CRGR), 130, 157, 159 light water reactors, safety and reliability of, 87-94, Commonwealth Edison, 59, 66, 67, 126, 148, 168 102 Connecticut Mutual Life Insurance Co., 217 comparison of fossil units to all nuclear units, 89 Connecticut Yankee, 136 examples of specific concerns, 90 Congress: overview of U.S. reactors, 87 Congressional Research Service (CRS), 201 reliability concerns, 88 House Committee on Science and Technology, 8 safety concerns, 87 House Interior Committee, 220 probabilistic risk assessment, 88 House Subcommittee on Energy and Environment, unresolved safety issues, 91 219 small reactor, 105-107 Joint Committee on Atomic Energy, 144 standardized reactor, 107 Senate Committee on Energy and Natural Resources, American Electric Power Co., Inc., 136 8 American Nuclear Society, 88, 214 Conservation Foundation, 237 American Physical Society, 218 construction permit (CP), 145, 148, 151, 157, 158, 160, American Society of Mechanical Engineers (ASME), 182 161, 163, 165, 168, 170 architect-engineers (AEs), 15, 22, 87, 107, 114, 135, Consumer Product Safety Commission, 164 179, 183 Council on Environmental Quality, 149 Argentina, 200, 202, 203 Critical Mass Energy Project, 214 Arizona Public Service Co., 115, 125, 127 ASEA-ATOM, 103, 105 Data Resources, Inc. (DRI), 33, 34 Atomic Energy Commission, 101, 144, 218, 227 Decision Research, 222, 225 WASH-740 study, 218 Department of Agriculture, 149 Atomic Industry Forum, 214 Department of Defense, 136, 149 Audubon, 32 Department of Energy (DOE), 33, 38, 44, 46, 60, 61, Australia, 197 65, 102, 136, 149, 154, 157/ 158, 159, 161, 162, 165, 168, 169, 170, 172, 179, 185, 228, 233, 239 Babcock & Wilcox Co., 87 Office of Policy Planning Analysis, 33 Baltimore Gas & Electric Co., 5 policy options, 253 Bechtel Corp., 127 Department of Housing and Urban Development, 149 Blue field Water Works and Improvement Co. v. West 701 Comprehensive Planning Assistance Program, 151 Virginia Public Service Commission, 51 Department of the Interior, 149 Bonneville Power Administration (BPA), 138 Department of Justice, 132, 171, 260 Boston Edison Co., 118 Attorney General, 145 Brazil, 202, 203 Department of Transportation, 149 Brown & Root, Inc., 127 Detroit Edison’s Fermi Breeder reactor, 213 Browns Ferry, 6, 87, 122, 123, 153, 213 Diablo Canyon, 116, 153, 228, 230 Buss, David, 228 case study, 242 Donaldson, Thomas, 230 California, 136, 151, 216, 231 Duke Power Co., 5, 60, 66, 67, 136 Calvert Cliffs plant, 5, 122 DuPont, Robert, 222, 233 Canada, 17, 18, 23, 71, 96, 98, 99, 109, 138, 179, 182, 191, 200, 203, 237 Ebasco Services, Inc., 127 Carolina Power & Light Co., 118 Edison Electric Institute, 32, 214 Carter administration, 203 Eisenhower administration, 144 case studies, 240-244 Electric Power Research Institute (EPRS), 46, 62, 63, 73, Central Electricity Generating Board (CEGB), 96 88, 115, 118, 127, 182

289 290 . Nuclear Power in an Age of uncertainty

Energy Information Agency, 32 Florida, 59, 72, 92 Environmental Impact Statement (EIS), 145, 148, 168 Florida Power & Light Co., 59, 72, 115, 136 Environmental Protection Agency, 149, 220 Ford, Daniel, 230 Eugene, Oreg., 221 Font St. Vrain, Colo., 9, 101, 102, 128 France, 22, 23, 67, 189, 191, 196, 199, 200, 202, 203 Federal Aviation Agency, 149 backfits, 197 Federal Energy Regulatory Commission (FERC), 138, 255 siting of plants, 198 Federal Power Commission v. Hope Natural Gas Co., Fuel Supply Service, 136 51 Federal Register, 166, 167 Garrett, Pat, 240 Federal Trade Commission, 163 General Accounting Office (GAO), 171 financial and economic future, 13-15, 29-76 General Atomic Co., 101 cost of building and operating nuclear powerplants, General Electric Co., 87, 94, 180, 218 57-71 General Public Utilities (GPU), 68, 136 cost of electricity from coal and nuclear plants, 64 Georgia Power Co., 40 financial risk of operating a nuclear powerplant, 68 Great Britain, 22 risk of public liability, 69 Great Lakes Basin Commission, 149 future construction costs of nuclear powerplants, 66 gross national product (GNP), 29, 32, 33, 34, 36, 40, 41 impact of delay on cost, 63 Gulf Power Co., 40 impact of risk on the cost of capital, 70 increase in nuclear construction Ieadtimes, 62 Hodel, Donald, Secretary of Energy, 162 rapid increase in cost, 58 Houston Lighting & Power Co., 127 reasons for increased construction costs, 60 Hutchinson, Fla., 115 electricity demand, 13, 29, 31-42 sources of uncertainty, 33-41 Illinois, 56, 136 estimated impact of industrial electrotechnologies in impasse, 5 the year 2000, 38 India, 202, 203 nuclear power in the context of utiltity strategies, Indian Point Station, 150 71-76 Indiana Public Utility Commission, 56 alternative utility construction plans, 73 Industry Degraded Core Rulemaking Program, 88 estimated cost of nuclear plants under construction, Inglehart, Ronald, 229 76 Institute for Nuclear Power Operations, 9, 19, 20, 25, implications for Federal policies, 74 128, 130, 131, 133, 134, 137, 185, 219, 255 plant construction, 14, 29 near-term operating licenses (NTOL), 132 rate regulation and powerplant finance, 13, 46-57 Significant Events Evaluation and Information Network accounting for capital costs of powerplants under (SEE-IN), 129, 130 construction, 50, 64 Systematic Assessments of License Performance allowance for funds used during construction (SALP), 131 (AFUDC), 47, 50, 51, 52, 57 changes in rate regulation, 52 Japan, 22, 23, 102, 191, 194, 199, 200, 203 construction work in progress (CWIP), 50, 52, 56, Ministry of International Trade and Industry (MITI), 57, 71 195, 198 history of the deterioration in the financial health of siting of plants, 198 electric utilities, 1960-82, 48 impact of changes in rate regulation in electricity Kasperson, Roger, 222 prices, 57 Kemeny Commission, 231, 233 implications of utilities’ financial situation, 50 Komanoff, Charles, 58, 65 market-to-book ratios and dilution of stock value, Korea, 194, 200 47 obstacles to a long-term commission perspective, 56 League of Women Voters, 230, 236, 239 pay-as-you-go inflation schemes, 56 legislation: phased-in rate requirements, 52 Administrative Procedures Act (APA), 164 public utility commissions (PUC), 50, 51, 52, 56, 57 Atomic Energy Act, 21, 22, 143, 144, 151, 154, 160, utilities current financial situation, 46 161, 164, 172, 203 utility accounting, 53 Clean Air Act, 151 recent past, 29 Clean Water Act (CAA), 151 reserve margins and retirement, 42-46 Economic Recovery Tax Act of 1981, 50, 51 economic obsolescence, 43 Energy Reorganization Act of 1974, 144 loss of availability of generating capacity, 44 Foreign Assistance Act of 1961, 203 need for new powerplants, 45 Fuel Use Act, 72 retirements due to age, 43 National Environmental Policy Act (NEPA), 151 Index . 291

Nuclear Licensing and Regulatory Reform Act of 1983, external factors, 121 154 historical labor requirements, 123 Nuclear Non-Proliferation Act of 1978, 202, 203 internal factors, 124 Nuclear Powerplant Licensing Reform Act of 1983, technological factors, 118 154 new institutional arrangements, 134 Nuclear Waste Policy Act of 1982, 264 certification of utilities, 137 price-Anderson Act, 69, 144, 264 Government-owned regional nuclear power Public Utility Regulatory Policies Act (PURPA), 138 authority, 138 Lifton, Roger, 222 larger role for vendors in construction, 135 Lloyd’s of London, 69 privately owned regional nuclear power company, Long Island Lighting Co., 60 138 Lonnroth, Mans, 190, 199, 200 service companies, 136 Louis Harris & Associates, 228 variations in quality of construction and operation, Lovins, Amory, 229 114-118 performance of selected U.S. LWRs, 117 Maine Yankee, 136, 222, 228, 230, 236 Nuclear Mutual Ltd. (NML), 68 case study, 240 nuclear powerplant technology, 15-18 management of nuclear powerplants, 18-20 boiling water reactor, 84, 85, 86, 90, 92, 94 Massachusetts, 136, 215 heavy water reactor, 17, 84, 96-99, 175 Massachusetts Institute of Technology, 220 high temperature gas-cooled reactor (HTGR), 16, Mazur, AlIan, 231 17, 84, 87, 99, 100-102, 128, 175 Mexico, 194 light water reactors (LWRs), 15, 16, 18, 83, 84, 85, Michigan, 45, 136 88, 93, 94, 96, 97, 98, 102, 113, 128, 156, 174 Middle South Utilities, Inc., 136 safety and reliability of, 87-94 Mitchell, Robert, 226 construction records, 115 process inherent ultimately safe reactor, 17, 18, 103, Nader, Ralph, 214, 215, 228 104, 105, 175 National Aeronautics and Space Administration, 136, pressurized water reactor, 84, 85, 86, 90, 94, 95, 96 182 prestressed concrete reactor vessel (PCRV), 99, 100, National Interveners, 228 103 National Oceanic and Atmospheric Administration, 149, Nuclear Regulatory Commission (NRC), 4, 5, 6, 16, 18, 164 19, 20, 23, 25, 26, 63, 69, 87, 90, 92, 93, 115, 118, National Opinion Research Center, 227 120, 122, 123, 128, 133, 134, 137, 145, 150, 154, National Organization for Women (NOW), 231 156, 157, 158, 159, 164, 165, 166, 168, 180, 213, National Pollutant Discharge Elimination System 233 (NPDES), permit, 151 approval of site suitability, 170 National Technical Information Service, 10 Committee for Review of Generic Requirements, 130, Nelkin, Dorothy, 230 157, 159 Netherlands, 231, 234 Office of Nuclear Reactor Regulation, 145, 148, 149, New England Electric System, 72 155 New Hampshire public Service Co., 136 Performance Assessment Team (PAT), 131, 132 New Jersey, 68, 118 policy options, 253 New York, 56, 135, 150 PSAR, 150 New Zealand, 197 Regional Administrators, 167 Non-Proliferation Treaty (NPT), 202 Regulatory Reform Task Force, 158 Northeast Electric Reliability Council (NERC), 44, 45 responsibilities in licensing, 146 Northeast Power Coordinating Council (NPCC), 44 Safety Evaluation Report (SER), 145, 148 nuclear disincentives, 4 Standard Review Plan (SRP), 145, 166 Nuclear Electric Insurance Ltd. (NEIL), 68 Nuclear Safety Analysis Center (NSAC), 129 nuclear enterprise, management of, 113-139 nuclear steam supply system (NSSS), 135, 145, 179, assessments of efforts to improve quality, 132-134 180, 182, 188 improving the quality of nuclear powerplant manage- Nuclear Test Ban Treaty, 227 ment, 127-132 Nucleonics Week, 181 classifications for INPO evaluations, 132 NUS Corp., 118 creating the right environment, 129 detecting and improving poor performance, 130 institutional approach, 128 Ontario Hydro, 138 technical approach, 127 operating license (OL), 148, 149, 151, 160, 161, 163, management challenges, 118-127 165, 168 comparison of manpower requirements, 122 Oregon, 151, 215, 217, 221 292 ● Nuclear Power in an Age of Uncertainty

Organization for Economic Cooperation and Develop- regulation of nuclear power, 21-22, 143-175 ment, 189 backfitting, 154-160 definition, 158 Pacific Gas & Electric Co., 116 legislation, 160 Pahner, Philip, 222 specific concerns, 155 Pakistan, 202, 203 Federal regulation, 143-150 Palo Verde plants, 125 historical overviews of nuclear regulation, 144 Pennsylvania, 68, 72, 213, 230 NRC responsibilities in licensing, 146 Philadelphia Electric Co, 101 role of emergency planning, 150 Philippines, 135 utility responsibilities in licensing and construction, policy options, 3, 25-26, 251-278 147 major Federal policy strategies, 263-277 hearings and other NRC procedures, 160-167 base case: no change in Federal policies, 263 changes in role of NRC staff, 164 base case variation: let nuclear power compete in a efficiency improvements, 162 free market, 270 enforcement, 167 economic conditions: two scenarios, 265 improvements in public participation, 162 four nuclear futures under the base case, 267 management control, 165 management improvement conditions: two other NRC procedural issues, 164 scenarios, 266 proposals for change, 161 policy goals and options, 252-262 use of rulemaking, 166 alleviate public concerns and reduce political risks, issues surrounding nuclear plant regulation, 153-154 260 licensing for alternative reactor types, 174 improve reactor operations and economics, 255 other NRC responsibilities, 171 reduce capital costs and uncertainties, 252 safety goals, 172 reduce the risk of accidents that have public safety State and local regulation, 150-153 or utility financial impacts, 256 environmental policy, 151 strategy one: remove obstacles to more nuclear need for power, 151 orders, 271 State siting activities, 153 strategy two: provide moderate stimulation to more State siting laws, 152 orders, 273 two-step licensing process, 168-170 Porter Commission, 237 Richland, Wash., 126 Portland General Electric, 217, 222 River Basin Commissions, 149 Preliminary Safety Analysis Report (PSAR), 145 Rochester Gas & Electric, 213 President’s Commission on the Accident at Three Mile Rogovin Report, 164, 166 Island, 166 Rumania, 194 Princeton University, 220 public attitudes toward nuclear power, 23-25, 211-244 Salem, N. J., 93, 214, 219, 221, 236 actors in the nuclear power debate, 214 Sandia Siting Study, 219 expert’s view, 217-220 Schumacher, E. F., 229 factors influencing the public’s view on nuclear safety, Science Applications, Inc., 219 221 selected U.S. light water reactors, comparison of, 19 psychological factors, 222 Shadis, Ray, 240 history of statewide referendum votes dealing with Shoreham plant, N. Y., 5 nuclear powerplants, 212 S. M. Stoner Corp., 179 impact of public opinion on nuclear power, 215 Solar Energy Research Institute, 32 impact of risk assessments on public opinion, 220 South Africa, 202 public acceptance of nuclear power, 234-244 Southern Co., 136 role of the media, 24, 231-234 Southwest Power Pool (SPP), 44 State laws and regulations restricting construction, 216 Spain, 200 values and knowledge, 226 Standardized Nuclear Unit Power Plant System Public Citizen, Inc., 214 (SNUPPS), 96 Public Service Electric & Gas Co., New Jersey, 118, 214 Stanford University, 220 purpose of study, 8 Strauss, Lewis, 144 survival of the nuclear industry, 179-206 Quadrex Corp., 136 effects in the U.S. nuclear industry of no, few, or delayed new plant orders, 179 Rasmussen Report, 69, 218, 220 architect-engineering firms, 183 Reactor Safety Study, 218, 219 engineering manpower, 181 Reagan administration, 203 impact on future new construction of nuclear regional nuclear power company (RNPC), 138 plants, 186 Index ● 293

impact on nuclear plant operation, 183 Union of Concerned Scientists, 214, 220 nuclear component suppliers, 182 United Kingdom, 189, 194, 196 reactor vendors, 180 Central Electricity Generating Board (CEGB), 199 prospects for nuclear power outside the United States, University of Michigan, 213 189-206 U.S. Army Corps of Engineers, 149 economic context for nuclear power, 189 U.S. Geological Survey, 149 foreign technical experience, 191 U.S. Supreme Court, 151, 216 implications for U.S. nuclear industry, 197 licensing and quality control, 191 Vermont, 151 no-nuclear weapons pledges in effect, 202 Vermont Yankee, 136 nuclear-generating capacity outside the United Von Hippel, Frank, 220 States, 205 nuclear proliferation considerations, 200 Walker, William, 190 onsite and offsite nuclear-related job vacancies, 204 Washington Public Power Supply System (WPPSS), 4, overall nuclear development, 194 126 public acceptance, 190 We Almost Lost Detroit, 213, 225 sample construction times for nuclear plants in Weinberg, Alvin, 239 various countries, 192 West Germany, 23, 56, 102, 189, 191, 194, 196, 200, U.S. industry in an international context, 199 202, 203 Sweden, 196, 199, 238 “Convoy” experiment, 197 Switzerland, 197 Kraftwork Union (KWU), 197 backfits, 197 Taiwan, 194, 200 standardized training, 198

Teledyne, Inc., 136 Westinghouse Electric Corp., 87, 94, 95, 135{ 180 Tennessee Valley Authority, 105, 138, 213, 237 Wintersburg, Ariz., 115 “The China Syndrome, ” 225 Wisconsin, 151 Three Mile Island, 6, 16, 62, 68, 70, 72, 87, 90, 92, 93, 102, 120, 121, 122, 125, 129, 148, 150, 153, 155, Yankee Atomic Electric Co., 136 156, 213, 233

U . S . GOVERNMENT PRINTING OFFICE : 1984 0 - 25-450 : QL 3