Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2 Principles of Nuclear Safety Research During its study the committee was provided with evidence suggesting that members of the public, of the nuclear industry, of the Congress, and of the administration have little faith in the soundness of the federal government's program of nuclear safety research. One source of this lack of confidence is the failure of the government to organize and implement a program according to fundamental principlesâprinciples of nuclear safety research. The purpose of this chapter is twofold: first, to develop gen- eral principles of nuclear safety research; second, to examine some of the implications of these principles for decisionmaking in gov- ernment and in industry. Chapter 3 illustrates how these general principles can be applied to a set of important research topics that represent elements of a future program of nuclear safety research for the nation. The reader should recognize that hi discussions of research that principally benefits utilities or the NRC these organi- zations are being used as proxies for larger groups (viz., ratepayers and stockholders in the former case, taxpayers and the public in the latter case). THREE CENTRAL QUESTIONS There are three major questions from which fundamental prin- ciples of nuclear safety research can be derived: ⢠Who should fund nuclear safety research? ⢠Who should set the research agenda? ⢠Who should conduct the research? 11
12 "Who should fund the research?" asks from which organization's budget should the research be funded. "Who should set the re- search agenda?" asks who should decide in what areas research is needed, what questions should be investigated, what research projects should be done, and how should meaningful projects be designed. And "Who should conduct the research?" asks which organizations or which researchers should actually carry out the research. WHO SHOULD FUND AND WHO SHOULD SET THE AGENDA FOR NUCLEAR SAFETY RESEARCH? The economics and science policy literature cite three basic reasons why industry usually underinvests in R&D relative to the level that would be socially optimal: (1) the uncertainties as to whether R&D investment will pay off make such investment highly risky; (2) most knowledge, once obtained, cannot be kept to oneself for very long, implying that the firm that pays for the RfcD cannot appropriate to itself all the benefits from it; and (3) the cost of R&D projects are often large compared to what an individual firm can afford to invest. These reasons apply to the nuclear industry no less than to others. Indeed, the current state of the nuclear industry and the historical role of the federal government in nuclear R&D suggest that the nuclear industry might underinvest in R&D more than many other industries. In many respects nuclear power is at a standstill in the United States; there have been no nuclear plants ordered for almost a decade, and dozens previously on order have been canceled. As a result of the industry's diminished prospects, nuclear reactor and equipment vendors cannot have a high expectation of being able to benefit from investment in safety research. The utility industryâ the primary owners and operators of commercial nuclear power plantsâis itself fragmented, so that sizable research investments are possible for only a handful of the larger utilities. With a few notable exceptions (such as in New York and California), util- ities have traditionally relied upon reactor vendors for research on power generation technology. State public utility commissions, which regulate utility rates of return, have significant influence upon, and an incentive to limit, the size of utility budgets. In fact, there have been instances in which PUCs have restricted
13 the amount of money raised from utility ratepayers to fund EPRI research. There are thus both internal and external cost-cutting pressures that are likely to push the industry in the direction of avoiding expenditures whose direct contribution to operating ef- ficiency is unclear. Finally, the federal government has a legacy of large investments in all facets of nuclear research and devel- opment, and this legacy has created a general expectation within the industry that the federal government should continue to bear a large measure of the responsibility for safety research. These facts, combined with the more general arguments cited above, im- ply that the nuclear industry may voluntarily fund less nuclear safety research than is in the public interest. These factors need to be kept in mind in considering who should fund and who should set the agenda for nuclear safety research. In what follows the two questions are treated simulta- neously, not only because they are intimately connected to one another but because there is no general answer to either one of them. Principles for Determining Who Should Fund and Who Should Set the Agenda for Research Logically, either government or industry or a combination of the two could be involved in funding research. If the government wants research, it has several means of getting it done. It can direct industry to do it, it can pay to have it done by others, or it can do it itself. If industry wants research done, it can either do it itself or pay to have it done by others. Agendas, on the other hand, may be set not only by industry and government but also by other performers of research and independent technical experts. In cases where both government and industry benefit from the research, they will have to work out the details of their respective funding and agenda-setting responsibilities on a case-by-case basis. This discussion of who should fund and who should set the agenda for research is guided by the following general principles. They represent the committee's best judgment of what principles are applicable, and they derive not only from its study but from its views of the appropriate roles of government and industry in the nuclear safety research enterprise. 1. Where the proximate beneficiary of the research is industry, the presumption is that industry should pay for it. However, there
14 are countervailing considerations that can override that presump- tion. 2. Where the expected societal benefit of the research is greater than the expected return to industry, government should con- tribute to paying for it. 3. Industry should pay for much of the research necessary for the performance of the regulatory functions of the NRC, but the NEC should pay for research when it is necessary to protect its independence, assure the timely availability of results, or explore issues involving great uncertainty. 4. Anyone who pays for a piece of safety research should contribute to setting the agenda for that research. 5. Performers of research and independent technical experts should participate in setting the research agenda. It is not always going to be easy to decide, a priori, who the principal beneficiary of some research may be. In the case of basic research, it is not always possible to say even if there is a beneficiary, much less who that beneficiary may be. Yet basic research is an essential feature of a sound program of nuclear safety research. Implications for Decisiomnaking The application of these general principles has implications for decisionmaking in government and industry. Below we make a distinction between regulatory and nonregulatory research and between research on current and future reactors in order to explore these implications. It should be noted that as a practical matter it is frequently not possible to make a sharp distinction between research to improve regulations and research to improve the design, performance, and safety of reactors. There are, of course, examples of research that is quite clearly directed at achieving some economic benefit but is of only marginal benefit either to safety or to regulation. Similarly, there are examples of research where the purpose is clearly to support improvement of regulation or the independence of the regulatory agency in the light of some vested industry position. In many cases, however, the public and private benefits of the research are mixed, and the overlap between regulatory and utility requirements is extensive, so that it is difficult to allocate neatly the responsibility for the research.
15 Category 1: Research on Current Reactors That is Aimed Di- rectly at Improving Regulatory Decisions The proximate beneficiary of research in this category is by definition the NRC, so it must have at least partial authority to set the research agenda. Yet the NRC does not and need not fund all research in this category since it has the authority to direct industry to fund it. It exercised such authority, for example, with respect to research on relief valve reliability; and it also exercised this authority when it initiated the reassessment of Babcock & Wilcox (B&W) reactors now being conducted by the B<kW Owners Group. The practice of directing industry to do research is more widely used by the nuclear regulatory authorities of other countries (no- tably Canada) and by other U.S. regulatory agencies (notably the Food and Drug Administration). It could also be used more widely by the NRC. Of course, the agency would need to employ safeguards to protect its independence. Such safeguards might include insisting that agency personnel have access to facilities and to the raw data from such work; requiring that the results be peer reviewed; and retaining within the agency, through the use of consultants or staff of the national laboratories, the ability to interpret the data and evaluate claims made by industry. With the employment of these kinds of safeguards, such an approach to nuclear safety research can and should be more widely used by NRC, and doing so would not jeopardize the NRC's independence. When the NRC directs industry to do work applicable to a particular reactor, the work should be funded by the owner of that reactor; more generic research could be funded by licensees cooperatively through, for example, Owners Groups or through EPRI. In such cases, agenda-setting ought to be shared between industry and the NRC. The NRC should decide in what areas research is needed to address regulatory issues. The industry should decide what research projects would be responsive to the NRC, and how to design them, although these issues might be better decided after consultation with the NRC. Industry cannot be directed to fund all research in this cat- egory. On the contrary, some of it will have to be funded by the NRC, with the agency also setting all aspects of the research agenda. This includes research where NRC's independence cannot otherwise be assured (e.g., research on emergency planning where
16 the industry and NRC disagree on how much emergency pre- paredness is needed; research where not enough is known about the problem for the industry to be able to decide what to do or for the NRC to be able to specify exactly what research it should require industry to do; and research where the results are required more quickly than could be anticipated if industry alone were re- quired to fund the work (i.e., where the NRC is more motivated to do the research rapidly than industry). Since it is often difficult to determine whether industry or government is the proximate beneficiary for a piece of research, it will often be appropriate to adopt an intermediate approach, namely, "cooperative research." Cooperative research is research in which the NRC jointly funds and jointly sets the research agenda with either the utilities or the vendors or both. The record of such projects involving industry and the federal government has been uneven. One complaint about cooperative research is the difficulty of overcoming the adversarial nature of the regulatory system within which it is conducted. On technical problems that are linked to licensing, industry often tends to offer the NRC no more than it feels it has to for fear of triggering a time-consuming regulatory response. The NRC often believes that industry is dragging its feet, particularly on any problem where the two dis- agree fundamentally on the safety implications. The net result is for cooperative projects to get bogged down in seemingly endless meetings, at great cost in both time and money. Yet despite the potential obstacles, there are enough recent cases of cooperative research that are viewed as having been a success to suggest that more cooperative efforts should be encouraged. The NRC and industry need to work to reduce the atmosphere of mistrust that pervades their dealings with one another. Within the NRC, most research should be funded by the Office of Research. Other program offices in the NRC have funds to perform short-term research that they may need to conduct regulatory activities. These funds are commonly referred to as "technical assistance," and they should continue to be available. Of course, in order to ensure the integrity of the research program, it is important for agency management to ensure that what is properly "research" and what is properly "technical assistance" come under the managerial authority of the appropriate NRC offices.
17 In the past, some research in support of the NRC's ability to regulate has been funded through international cooperation. Such cooperation is to be encouraged whenever possible as a way of spreading the cost of safety research among all potential benefi- ciaries. Category 2: Research on Current Reactors Where the Research has Safety Implications but is not Directly in Support of Regula- tory Decisions Safety research on current reactors where the objective of the research is not directly aimed at supporting NRC's ability to regulate is conducted by both industry and government, in the case of the government, principally by DOE. The government has a role to play in funding and setting the agenda for such research. When the work is developmental, the government is fulfilling its responsibility to ensure that the nation has secure and safe future sources of energy and thus is benefitting the general public. Therefore the general public has a responsibility to pay for a portion of the cost of the research. Much of the research in this category is and should continue to be funded by industry for its own purposes, including the following: ⢠Research to improve the product. ⢠Research to support claims about safety in proceedings before the NRC. ⢠Research in anticipation of the NRC's raising issues (so- called "defensive research"). ⢠Research by entrepreneurs who perceive potential business opportunities in supplying utilities with new safety-related equipment or services. In all these instances, industry should set the agenda for the research. Ordinarily, government funding of any research in this cate- gory should occur in the context of government-industry coopera- tive research programs, and the research agenda should be jointly set by government and industry. However, there are instances in which research of this type should be fully funded by the government. These are cases in which industry is unwilling to fund or contribute to research that, although primarily of benefit to industry, is nonetheless expected to have significant societal benefits. For example, research to
18 develop specialized equipment or to test equipment used in other industries for possible use in the nuclear field may be constrained because manufacturers and suppliers have insufficient incentive to fund such work due to the limited scale of the nuclear market for such products. Research of this type should usually be funded by DOE because of its mission, although other agencies may also fund or conduct research with applications in the nuclear industry. In such cases, government should have the authority to set the research agenda, although this might be shared, either de facto or de jure, with industry and with the contract research community through appropriate advisory committees. Although research that is not aimed directly at improving regulatory decisions is not of immediate interest to the NRC, it will often have implications for reactor safety that regulators must care about. For this reason, NRC participation in the funding of such research will, on occasion, be fully justified. In the past, some research not in support of regulation has been funded through international cooperative arrangements. Such cooperation should be encouraged whenever possible as long as the original research objective is maintained. Category 3: Research on Future Reactors Research on future commercial reactors is conducted by DOE, which has no regulatory responsibilities in the commercial reactor sector. However, if future reactors ever become commercially vi- able, they will have to be regulated by the NRC and meet NRC safety regulations. Research on future reactors entails both safety and economic considerations. The research should be funded and the agenda set primarily by reactor vendors and suppliers as part of their development programs, particularly if a sufficiently early commercial payback can be anticipated. However, DOE has statu- tory responsibility for ensuring adequate future energy supplies and conducts an advanced reactor R&D program. DOE's decision on whether to allocate funds to advanced reactor research should be made on the usual grounds, i.e., whether the expected benefit of the research to society exceeds the expected return for the private sector, implying insufficient incentive for private investment in the research. If DOE contributes to such research, the agenda will neces- sarily be set cooperatively. Even when DOE totally funds the
19 research, the potential users of the technology (i.e., reactor ven- dors and utilities) should participate in setting the research agenda in order to ensure that it takes account of market considerations. The NRC needs to conduct regulatory research on advanced reactors in order to establish what safety standards these new designs must meet. Such work is a normal function of regulatory policy development, and in any case is not an important cost relative to other safety research activities. It is not practicable for industry to fund NRC's advanced reactor research since no specific applicant for an advanced reactor license yet exists. However, once the agency begins receiving applications for the review of advanced reactor designs, it could begin recovering from each applicant a portion of the costs incurred. Here again, some research of this kind has been funded through international cooperation, and use of this mechanism of sharing the costs of research should be encouraged. General Requirements for the Design of Government Research Programs All government research programs, including the nuclear safety research programs of the NRC and DOE, have certain gen- eral requirements not considered in the foregoing discussion. These requirements are as follows: ⢠To support basic or exploratory research ⢠To ensure agency competence ⢠To ensure an adequate national pool of trained scientists and engineers. Support for Basic Research All government agencies that conduct research should sup- port some research that is exploratory or basic in nature. The fundamental aim of this research is to continue the development of new knowledge. In theory, basic research could be funded either by industry or by government, but the usual reasons why indus- try tends to underinvest in basic research apply to the nuclear industry no less than to others. Therefore, government agencies, primarily DOE but also the NRC, should reserve some portion of their budgets for mission-oriented basic and exploratory re- search. The agenda for basic research will necessarily have to
20 be set at least partly by the government, but both industry and the university research community can and should participate in agenda-setting through their participation in independent research advisory groups. Ensuring the Competence of Government Research Managers and the Availability of Trained Scientists and Engineers In principle, if the government has a vigorous program of nuclear safety research under way, no additional research will be necessary to fulfill either of these objectives. Furthermore, there are ways for the government to promote these two kinds of ob- jectives other than by performing research. Nevertheless, some special research may be needed to ensure that the government is technically competent and aware of new developments in the rele- vant fields of science and engineering, and some may be required to ensure that researchers having specialized training and experience are available to perform research when needed. At a time when the industry's fortunes are in serious decline, it is imperative that the government take a strong interest in the long-term availability of scientific and engineering talentâ talent that will be needed to maintain agency staffs at a high level of capability and to conduct future research on safety-related problems. If NRC and DOE are to have effective safety research programs in the future, they must begin now to cultivate top-rate researchers and research managers, involve them over a period of years in the regulatory process, and support them in research on safety-related topics. WHO SHOULD CONDUCT NUCLEAR SAFETY RESEARCH When it comes to determining who should perform the re- search, there are several possibilities. Research might be con- ducted by the following: ⢠the government (e.g., NRC, DOE's national laboratories, U.S. Geological Survey, National Bureau of Standards); ⢠industry (including reactor vendors, equipment suppliers, utilities, including EPRI and consortia); ⢠the contract research community, including universities;
21 ⢠an international cooperative arrangement involving the U.S. government or U.S. industry and one or more for- eign governments or firms; and ⢠all combinations of the above. All these combinations can and do exist because prime contractors of nuclear safety research often let subcontracts for a portion of the research they have been engaged to perform. Principles for Determining Who Should Conduct Nuclear Safety Research Despite the complexity of these arrangements for performing research, there are a number of principles that should be applied: 1. Nuclear safety research, like all research, should be con- ducted in institutions and by researchers who can do a good job. This principle encompasses both the competence of the researcher and his or her access to suitable facilities and equipment. 2. In deciding who should conduct research, where the princi- pal objective of the research is aimed directly at improving regu- latory decisions, care must be taken to protect the independence of the NRC from the regulated industry. 3. There must be institutional mechanisms to ensure that research results can be made available in response to short-term requirements without disrupting longer-term research. 4. Industry is a source of knowledge and expertise important to the conduct of nuclear safety research. When designing research programs, the problem of transferring the results of the research must be considered at the time agreements are reached concerning who should conduct the research and how the agenda should be set. 5. Universities have important roles to play as training grounds for future researchers, as centers of high-quality basic and ex- ploratory research, and as sources of independent ideas. 6. National laboratories have unique experimental and com- putational resources as well as highly competent, experienced re- search teams of potential usefulness to a program of nuclear safety research.
22 Implications for Decisionmaking The above principles have implications for determining who should conduct nuclear safety research. The process of selecting researchers is far from straightforward. The following discussion illustrates how these principles apply to research conducted in industry, in the university, in government laboratories, through international cooperative arrangements, or in the small business community. Research Conducted in Industry If the proximate beneficiary of a nuclear safety research project is industry, the research should involve industry to the maximum extent possible regardless of who pays for it. This is accomplished automatically, of course, when industry itself funds the research, when the NRC directs a licensee or applicant to conduct the re- search, when industry conducts research on contract either for the NRC or DOE, or when either of those agencies engages in government-industry cooperative research. Even NRC-funded re- search should often involve industry, except that where research results can be expected to directly influence regulatory decisions, care must be taken, in ways previously identified, to preserve the independence of the NRC. Where the facilities that are required for nuclear safety research exist only in the national laboratories, arrangements should be made to enable and encourage industry researchers to utilize these facilities; or the national laboratories should themselves conduct the research and help industry gain access to the results. In either case, industry should help set the research agenda. Research Conducted in the Universities University research is important for several reasons. It is a way of engaging some of the country's best theoretical and ex- perimental scientists and engineers; it provides training for future government and industry researchers and research managers; and it fosters and encourages independent thinking about scientific and engineering problems. For all these reasons, the federal gov- ernment has a fundamental responsibility to preserve and protect university research. For research related to current reactor types, the government's responsibility should be discharged primarily by
23 the NRC; and for research related to more advanced reactor types, by DOE. Research Conducted by the Government The NRC and DOE do virtually no research of their own. When either of them wants research, they contract for it. The greatest part of the research supported by the NRC is conducted in DOE's government-owned, contractor-operated national labo- ratories. The national laboratories, in turn, distribute varying amounts of the work to other sectors through subcontracts. Of course, the NRC's enabling legislation directs the agency to make use of the national laboratories, and there are identifiable benefits to be gained by contracting with them. The laboratories provide ready access to experienced researchers from a wide variety of dis- ciplines; they have instruments, equipment, and facilities that are sometimes not readily available elsewhere; their services can be ob- tained quickly through contracting procedures that are relatively easy for federal agencies to use; and they are accustomed to form- ing and maintaining interdisciplinary research teams. However, while the national laboratories have made identifiable contribu- tions to nuclear safety research, and can continue to do so, they should be considered as only one among several alternative sources of quality research. Research Conducted Through International Cooperation Some nuclear safety research of interest to the United States is conducted jointly with other countries, both here and abroad. U.S. participation in these international arrangements for research is partially funded either by U.S. government agencies or by Ameri- can industry. Examples include an international program of exper- imental research being conducted at the NRU reactor in Canada, a similar program at the Halden facility in Norway, and another at the Marviken reactor in Sweden. Foreign researchers sometimes participate in nuclear safety research in the United States as well. Such international cooperation in nuclear safety research should be encouraged whenever it is mutually beneficial. Research in the Small Business Community The United States has a long tradition of entrepreneurial cre- ativity. The nation's small businesses provide major sources of
24 invention and innovation and are quick to respond to the needs of industry. The contribution that small businesses have made to the country's nuclear safety research effort has been considerable and can continue. The DOE and the NRC should maintain their small business programs at the level necessary to exploit small business capabilities, whether or not the size of their research programs happens to be above the level at which a program of support for small business research and development is mandated by existing law. Attaining Balance in Allocating Research Dollars Determining who is best suited to perform a particular piece of research is one of the most challenging responsibilities facing directors of research. The challenge is particularly acute for the federal government's directors of research because several factors must be weighed in reaching a decision. These factors include not only finding sufficient talent and resources to do the job but also monitoring the impact of the selection process on the research community as a whole. To undernourish or overutilize any segment of the research community is to improperly manage the nation's limited resources for research. The committee is concerned that the nation's limited resources for conducting nuclear safety re- search have been poorly allocated in the past, particularly by the NRC. In Chapter 4 this issue is stressed, with the focus on the specific responsibilities of the NRC in contracting for research, and a reevaluation of the way the NRC allocates its research dollars is recommended. A Final Note It is important to recognize that in this chapter when the committee has focused on who should fund research it has been concerned with who should control the dollars being spent. A separate issue not addressed here is where those dollars come from. For example, even when the NRC provides funding, the dollars may flow from general revenues, license fees charged to particular nuclear utility licensees, or charges to reactor vendors for the review of standardized plant designs.
