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.
8 The Regulatory Environment The nuclear industry is one of the most heavily regulated in- dustries in the United States. The activities of nuclear utilities are constrained by a large set of external organizations and legal institutions that includes federal regulatory agencies, state public utility commissions, siting agencies, and environmental agencies. In addition, nuclear utilities are subject to constraints imposed by self-regulating agencies such as INPO and NUMARC, interested third parties such as independent standard-setting agencies and accreditation boards, and legal systems that define financial lia- bility (e.g., the Price-Anderson Provisions of the Atomic Energy Act of 1954 (P.~.8~703) and establish administrative procedures and public rights. Although the regulatory environment is seldom considered as a topic for human factors research, the pane} believes it is integral to decisions made within the nuclear utilities and to human factors in nuclear safety. At a general level, the regulatory organizations and institutions affect the incentives and the legal considerations of operation. Regulatory requirements constrain the choices of utilities in hardware, management, and personnel areas; and they impose costs on utilities, which may be either effective or coun- terproductive from a safety standpoint. These costs influence the resources available to the utilities for safety. Regulation, in ad- dition, constrains the profit-making opportunities in ways that may interact with their safety activities (Joskow and Noll, 1981; 89
go Burness, Montgomery, and Quirk, 1980; Breyer, 1981; Joskow, 1974). At the very least a regulatory body should ensure that its own regulations meet accepted human factors standards with respect to lack of ambiguity, comprehensibility, and conformance to human performance requirements. All regulations should be reviewed with this in mind. The pane] identified three research areas that are of particular importance to current policies for nuclear regulation, each of which includes a set of research topics. The first concerns the possibil- ity of assigning regulatory activities to different organizations and devolving some regulatory functions from the regulatory bodies to the utilities themselves. This topic is motivated by current propos- als for reform, particularly those advocating more self-regulation by the industry and consequent greater financial liability. The second research area of importance is a closer examination of the effects of regulation. Specific topics of importance include assessing costs imposed by regulation on utilities and incentives created for (or against) innovation by utilities. Diverse regula- tion by state public utility commissions has raised concerns that regulations tying profits to certain performance measures may de- tract from safety. Related research topics include exarn~ning the extent to which this is true and the possibility of formulating con- tracts that provide incentives for safety as well as short-term cost · . · . m~n~m~zat~on. The~third set of topics concerns human factors within the NRC itself such as staffing requirements and management issues. Because the NRC's activities are so pervasive, its own internal operations define the types of regulations it is able to engender and enforce. MODELS OF R1DGU[ATION Rationale and Background A critical factor in nuclear safety is the incentives facing nu- clear utilities to ensure the safety of their operations. Tradi- tionally, the government has relied on a combination of financial responsibility and direct regulation; however, federal regulation of the organization and management of utilities is limited. Many of
91 the human factors aspects of nuclear safety, including specifically organization and management, has been left to individual utilities. Since the Three Mile Island accident, attention has focused on the role of institutions and organizational factors in nuclear safety. The NRC conducted and comrn~ssioned research in this area in the early 1980s and considered more stringent regulation. At the same tune, the nuclear industry organized industry groups (most notably, INPO and NUMARC) that have undertaken a self-regulatory effort in operator training and other organization and management assessment activities. Many human factors areas have been proposed as appropriate for self-regulation, and utili- ties have taken the initiative in regulating training programs and related activities. Third parties have not received as much at- tention, but their roles have been proposed in certain areas (e.g., nuclear experts to run utility-owned power plants). Finally, atten- tion has focused on reforming the Price-Anderson Provisions of the Atomic Energy Act of 1954 (P.~.83-703), with its implicit finan- cial incentives, and the potential safety impacts of other financial constraints facing nuclear utilities (Wood, 1983~. The nuclear industry is a special case with regard to non- governmental regulation. Catastrophic accidents, which are of spe- cial concern, are associated with small probabilities of occurrence (SIovic, Fischhoff, and Lichtenstein, 1985~. They are associated with consequences that are necessarily beyond the financial liabil- ity of the utility. Furthermore, beliefs about nuclear safety vary substantially among the public, the industry, and third parties, and divergence of opinion about nuclear safety has been viewed as a major constraint on industry expansion (Wood, 1983~. It should be emphasized that, although different parties may have differ- ent beliefs regarding accident risks, it does not follow necessarily that industry incentives for safety are less than those of the public (Starr, 1980~. Nevertheless, utilities cannot (because of limited financial exposure) and may not (because of different assessments- of-accident probabilities) view the risks of operations in the same way as do other interested parties, such as the government or the public, and their attitudes to regulation will vary accordingly. The NRC has the statutory responsibility for regulating nu- clear power plants so as to ensure public health and safety. In theory, there is ample room for variation in how the NRC actu- ally exerts its responsibility. At one extreme, the NRC could give extremely detailed specification for all aspects of nuclear power
92 plant construction, operation, maintenance, and management- for example, imposing educational and experience qualifications for managers, delineating specific training curricula, decreeing con- tro} room design retrofits, specifying surveillance and maintenance schedules for equipment, and so forth. In a sense, this represents regulation of inputs to safety, i.e., those aspects of plant opera- tion that are believed to lead to safe production of electricity. At the other extreme, the NRC could fulfill its obligations by giv- ing utilities wide latitude in how they operate their plants, i.e., self-regulation. The role of the NRC in this instance would be to monitor plants to ensure that they are being operated safely, w~th- out regard to the specific measures the individual utilities adopt to achieve safety. In reality, the mode} of regulation adopted by the NRC varies over time and with respect to domain of interest. For example, in the area of operator training, the NRC responded to congressional mandates about training contained in the Nuclear Waste Policy Act of 1982 (P.~.97-425) by allowing INPO to take the lead in developing training criteria and accrediting the training programs of individual utilities. The NRC's role in this case has been to review INPO's accreditation program. This example might be contrasted to more specific NRC requirements with respect to the minimum numbers and qualifications of control room and senior control room operators on each shift. Research Recommendations The wide latitude that exists in how the NRC carries out its statutory responsibilities begs the question of whether one ap- proach results in greater safety than another. Arguments can be developed to defend aLrnost any regulatory approach, from detailed specifications to self-regulation. Over the last several years there has been increased reliance on industry initiatives in the human factors area including such areas as training, management, and performance indicators. And there has been a concomitant de- crease in NRC initiatives in these areas. Is this wise? Will it lead to improved safety in the short or long run? What types of NRC sanctions and incentives are most effective in a self-regulatory model? How is public input and participation protected in self- regulation? These are all questions amenable to an avenue of
93 research known as policy analysis, incorporating such methods as case studies and comparative studies of organizations. While much has been written about regulatory reform in gen- eral (Joskow and Noll, 1981; Noll, 1985; Nail and Owen, 1983; Friedlaender, 1978; Abolafia, 1985), to our knowledge research specific to the NRC has not been carried out. Given the unique aspects of nuclear power production and the potential for endan- gering public health and safety, this represents a serious omission. Several types of analysis can be used to address the issue of preferred regulatory models. Case studies and historical analyses of other regulatory agencies and of nuclear regulatory bodies in other countries may yield data on the expected effects of different regulatory models. Similarly, within the specific NRC context, case studies of specific instances of newly unplemented regulations and instances of specific regulations being deferred to industry initiatives could yield important information as to the effective- ness of the different approaches. In addition, the effectiveness of new initiatives could be treated empirically using performance indicators. Ideally, an empirical basis for selecting models of regulation that will yield greatest safety would be the outcome of such re- search. In reality, we recognize that the empirical basis would represent only one of several bases for selecting modes of regula- tion. Other bases would include political and economic pressures, hortatory argument, and the personal biases of key decision mak- ers. Studies of several years' duration would be required to allow sufficient time to fully assess the impacts of different regulatory approaches. How can the public and its representative, the NRC, be as- sured that regulations and regulatory actions result in the net enhancements to safety that are intended? Regulations and reg- ulatory requirements have increaser} over the years, as have the concomitant demands on regulatory and utility resources. Is there a point at which these demands on resources reduce safety? In addition, given the number of different regulatory bodies affecting utility behavior, is it possible that the actions of one regulatory body have results that run counter to the goals of another? The answers to these questions bear directly on the issues of overall en- hancement of safety at nuclear power plants and have implications for regulatory strategies. NRC regulations and requirements place substantial costs on
94 utilities. Obviously, such regulations are intended to ensure safety; nevertheless, some are confusing, repetitive, internally inconsis- tent, and require an onerous quantity of paperwork to comply with accounting and reporting requirements (see Advisory Committee on Reactor Safety, 1980; Cohen, 1980; Wood, 1983; Committee on Nuclear Safety Research, 1986~. The msue arises as to whether some regulations are detrimental to safety, because utilities divert excessive resources in particular, staff time and energy toward formal compliance with regulations. To the extent that utility staff believes that safety regulations are counterproductive, the regula- tions degrade utility-NRC relations and hence may be detrimental to nuclear safety overall. This problem is hardly new. Recommendations to rationalize and simplify NRC regulations have cropped up frequently (e.g., SECY-85-129, 1985k). That a recommendation ~ not novel does not detract from its importance. Research on effective organiza- tion and behavior must consider the time and efforts of staff in complying with regulations. If the regulations can be improved, benefits accrue both in Reproving utility staff efforts and in im- proving relations between utilities and the regulators. PLANT PERFORMANCE INDICATORS Rationale and Background To know whether regulations and utility actions are increasing safety at nuclear power plants, one must first be able to measure plant safety. The fundamental purpose of performance indicators is to be able to readily monitor and assess individual plant perfor- mance and take action when appropriate. The major driving goal is to be able to track plant performance and to understand how changes in plant operation and maintenance, whether utility- or NRC-initiated, affect that performance. Ideally, we should have valid indicators that temporally are as far removed from actual threats to safety as possible. Knowing that a plant is heading for trouble 10 months in advance is certainly more valuable in- formation than having that knowledge 10 minutes in advance. Of course, there is typically a trade-off in terms of the accuracy versus timeliness of such information. Both INPO and the NRC have embarked on well-publicized programs of performance indicators (see SECY 8~317, 19861;
as NUREG/CR-437B, 1986c; NUREG/CR-461t, 1986e). However, the current set of performance indicators is limited to information that is publicly available. The data that utilities currently must report to the NRC (e.g., licensee event reports) do not cover all the valid indicators, especially those having early diagnostic value. Research Recommendations We believe that an ongoing research program is required to support these efforts and should include several types of analysis. A research program that examines a wide variety of measures, not limited to those currently publicly available, should be initiated. These measures should include a search for "inputs" (e.g., mean time between maintenance of critical equipment, thoroughness of training progam), Through puts" (e.g., unplanned down time due to equipment failure, retention rate of personnel throughout the training program), and ~outputs" (e.g., radiation exposure due to equipment failure, events due to human error). In addition, the search for indicators should include those at different levels of the nuclear power plant system, including individual, team, and plant performance. Finally, the research should involve the con- tinuous update and validation of these indicators. This is essential because, as technology changes, plants age, and new regulatory requirements are Implemented or rescinded over time; ad of which may affect the validity of an indicator. - The result of this research will be both a regulatory too! with the ability to differentiate plants in terms of their likely safety performance and an important research too} against which all other changes can be assessed. The search for candidate indicators beyond those already col- lected by the NRC and their initial validation will require a two- to-three-year effortdepending on the difficulty of obtaining the measures from utilities. Beyond that, a modest level of effort will be required to continually update and validate the measures. Increasing discontent with traditional rate-of-return regula- tion has led a number of states to institute or experiment with al- ternate incentive financial regulation for utilities (Seagraves, 1984; Block et al., 1985~. Details of these contracts differ, but the general point is to create incentives for utilities to operate their systems in a cost elective manner, generally by sharing extra profits or im- posing penalties should costs exceed some predeterminer! amount.
96 Although this trend in regulation raises concerns, it simultane- ously presents opportunities to enhance nuclear safety. The concerns raised are that, in attempting to minimize short- run operating costs, utilities may incur increased nuclear safety risks by cutting maintenance and other safety-related expenses. Descriptive, but not statistically significant, evidence suggests that in fact long-run electricity generation is greater and costs are lower at plants that are well-managed, weD-staffed, and have lower incidences of events that are associated with safety problems (e.g., Wall Street flours al, July 28, 1987, p.2~. Nevertheless, presented with some of the financial incentives proposed by state public utility commissions, it is possible that utilities may have incentives to cut some short-term costs in ways that are detrimental to safety. The opportunity presented by these changes in state regu- lation is that the concept of incentives regulation can include incentives for long-run maintenance, investment, innovation, and, in general, safety. While only some of these issues have been con- sidered, current research in economics has focused on such mech- anisms, using both experimental and theoretical analyses (Smith, 1974; Loeb and Magat, 1979; Cox and Isaac, 1987~. A natural extension of this work is to consider mechanisms that incorporate long-run safety incentives. Because of the potential relevance of the results, and because the situation is currently in flux, this is a particularly important and timely research issue. There have been several instances ot public utility companies' instituting incentive programs for utilities whereby the given util- ity's rate is dependent on meeting certain financial or performance standards. The effects of these programs on safety, not to mention performance, have not been fully analyzed. Regulatory actions such as these, which have clear implications for productivity and therefore have safety implications as well, should be subjected to analysis and research. To begin, those incentive programs that are already in place could form a body of case studies, whereby characteristics of the specific incentive plan, utility and plant management responses to this plan, and actual plant performance would be carefully delineated and analyzed in terms of both safety and production. The interactions among these three sets of variables incentive plan characteristics, utility responses, and plant performance- would serve as the basis for structuring regulatory mechanisms to improve both safety and performance. _ . . .. . ...
97 CONCLUSION Having reviewed the course and nature of human factors re- search at the NRC over the last ten years, the pane} is encouraged by the recent initiative shown at the NRC to develop and funs] a new human factors research program. If this plan is implemented in 1988, receives the strong support of the NRC and of the indus- try, is managed by a qualified human factors specialist, is staffed by a team of multidisciplinary scientists, and ~ organized as a branch rather than as a subdivision of the reliability branch, then the initial steps of leadership required of the NRC in this critical area will have been taken. In this report, the pane} has emphasized higher priority items as a point of departure. The pane] envisions, however, that in the next few years, the NRC would develop a fuB program cov- ering all topics listed in the report, and that, over time, the full program would be implemented. It is anticipated that additional research will be required as the NRC and the industry review and implement the recommendations set forth by the pane! in this report.