Speech-97-17: Chairman Jackson to Nuclear Power Reactor Safety Course at MIT
OFFICE OF PUBLIC AFFAIRS
|Office of Public Affairs||Telephone: 301/415-8200|
|Washington, DC 20555-0001||E-mail: firstname.lastname@example.org|
|Web Site: Public Affairs Web Site|
July 29, 1997
Current Regulatory Issues
Dr. Shirley Ann Jackson, Chairman
U.S. Nuclear Regulatory Commission
Nuclear Power Reactor Safety Course
Massachusetts Institute of Technology
July 21, 1997
Good afternoon. I am pleased to be able to participate once again in the MIT Nuclear Power Reactor Safety Course.
Today I will summarize for you the NRC mission and safety philosophy; discuss several areas of current NRC focus, including (1) the incorporation of risk insights into NRC regulation, (2) digital instrumentation and controls, and (3) achieving international consensus and cooperation on matters of nuclear safety regulation. I will close with a discussion of the changing industry environment occasioned by electric utility deregulation.
II. NRC MISSION AND SAFETY PHILOSOPHY
Based on the Atomic Energy Act of 1954, as amended, the mission of the NRC is to regulate the Nation's civilian use of byproduct, source, and special nuclear materials to ensure adequate protection of the public health and safety, to promote the common defense and security, and to protect the environment. Some of the principal terms of the NRC regulatory mandate, however--such as "protect health and safety," or "provide adequate protection"--are not defined in the Act, nor are they self-explanatory. The process of interpreting and applying these terms and provisions is a continuing effort that has evolved over several decades of Commission regulation, Congressional oversight, and judicial review of specific NRC actions. The result has been the creation of a body of regulations, decisions, and practices through which the NRC safety philosophy is expressed.
This philosophy comprises several closely interrelated elements, which were articulated earlier this month in the consultation draft Strategic Plan the Commission submitted to the Congress. The elements are: defense-in-depth, licensee responsibility, safety culture, regulatory effectiveness, and accountability to the public.
"Defense-in-Depth" ensures that successive measures are incorporated into the design and operating procedures for nuclear installations to compensate for potential failures in protection or safety measures, wherever such failures could lead to serious public or national security consequences.
"Licensee Responsibility" embodies the principle that, although the NRC is responsible for developing and enforcing the standards governing the use of nuclear installations and materials, it is the licensee who bears the primary responsibility for conducting those activities safely.
"Safety Culture" recognizes each licensee's responsibility to establish and maintain a set of attitudes and operational principles to ensure that safety issues get the attention they warrant. A safety culture encourages a questioning and learning attitude toward safety issues and discourages complacency.
"Regulatory Effectiveness" emphasizes the approach that, because safety is paramount in the Commission's regulatory program, certain standards and practices to ensure adequate protection will be required, whatever the cost. Over and above that baseline, additional safety upgrades will be required only if their benefits justify the added cost. Regulatory Effectiveness also involves the ongoing examination of NRC regulations, internal procedures, and oversight activities, to ensure consistency, fairness, ease of implementation, and compatibility with the overall NRC mission and program.
"Accountability to the Public" dictates that, just as licensees are accountable to the NRC, the NRC is accountable to the American people and to their elected representatives. This accountability entails being candid about NRC activities and their results, acknowledging the public interest and right-to-know about safety issues, and ensuring that the public has sound, complete, up-to-date information on which to base their judgments.
III. NRC FOCUS AREA: INCORPORATING RISK INSIGHTS INTO NRC REGULATION
I now would like to highlight several areas of current NRC focus, beginning with a discussion of risk assessment activities.
A. Risk Assessment Activities
The NRC has established its regulatory requirements to ensure that a licensed facility is designed, constructed, operated, maintained, and modified in a manner that will not result in undue risk to public health and safety. NRC requirements have been based largely on deterministic engineering criteria, establishing safety margins through the use of multiple barriers and the "defense-in-depth" philosophy I spoke of earlier. Probabilistic Risk Assessment (PRA) methods offer the potential to sharpen the focus and to improve the effectiveness of these requirements, allowing better decision-making by concentrating on those aspects of a facility most important to safety, thereby achieving better utilization of resources and reducing unnecessary burdens. PRA insights and information have been applied successfully in numerous regulatory activities, and have proven to be a valuable complement to deterministic engineering approaches. Accordingly, the Commission continues to focus on expanding the application of risk assessment methodologies to improve the overall regulatory process. NRC risk assessment activities have included the preparation of draft Regulatory Guides and Standard Review Plans, development of the Commission Safety Goals, initiation of pilot applications in specific areas, and review of licensee Individual Plant Examinations.
1. Regulatory Guides and Standard Review Plans
In the area of reactor regulation, the Commission recently released for public comment a set of draft Regulatory Guides (RGs) and Standard Review Plans (SRPs) that will support implementation of the August 1995 Commission Policy Statement on Probabilistic Risk Assessment, by providing guidance on how to use PRA information to support and evaluate plant-specific changes to the licensing basis. By "licensing basis," I am referring generally to that set of regulations, license conditions, technical specifications, and commitments that define the design and operating envelope within which a licensee must maintain and operate its facility. These RGs and SRPs describe acceptable approaches to decision-making for any area in which risk assessment can be used. They contain specific guidelines applicable to the areas of technical specifications, in-service testing, and graded quality assurance.
Five fundamental safety principles form the foundation for the approach described in these guidance documents. When using risk assessment insights to support plant-specific changes to the licensing basis, a licensee must adhere to these principles.
The licensee must meet existing regulations or propose an appropriate change or exemption.
The licensee must maintain defense-in-depth.
The licensee must maintain sufficient safety margins--which means that, where limits have been established through regulation, or through commitment to codes, standards, or Regulatory Guides, those limits should be observed, unless a change is proposed and approved.
The licensee must ensure that any increases in risk (including the cumulative effect of all proposed increases) should be small, and should not cause the NRC Safety Goals to be exceeded.
The licensee must use performance-based implementation and monitoring strategies that address uncertainties in the analysis models and data, and provide for timely feedback and corrective action. Performance-based implementation and monitoring is a method of checking key PRA analysis assumptions by observing actual equipment or system performance. If performance is not consistent with the assumptions in the analysis, then feedback and corrective action should be taken to revise the proposed change or to improve equipment or system performance. In effect, it is a way to reduce uncertainty and ensure PRA validity.
These five principles are intended to ensure that the essential elements of traditional NRC approaches to safety regulation are maintained, and that the insights from risk assessment are integrated into the safety review process in a way that complements the existing review process by focusing the reviewers on the most important issues. Both licensees and the NRC staff are expected to employ an integrated decision-making process that assesses a proposed change against these five principles.
2. Commission Safety Goals
While defense-in-depth and safety margins have been part of the NRC regulatory process for a long time, NRC guidelines associated with risk have not. In 1986 the Commission issued its Safety Goal Policy Statement which, for the first time, expressed quantitatively the Commission's expectation on the safety of nuclear power plant operation. Through the Safety Goal Policy Statement, the Commission propagated its philosophy that the risk from the operation of a nuclear power plant should be no more than 0.1 percent of the risk to which people are exposed from other sources. This statement of risk translates into objectives on individual risk of 2x10-6/yr of a latent fatality and 5x10-7/yr of an early fatality. These risk objectives were then used to develop a design-related subsidiary objective of 1x10-4/reactor year (RY) core damage frequency (CDF) for accident prevention, and a standard (characterized as an "implementation guideline") of 0.1 conditional containment failure probability for accident mitigation.
These subsidiary objectives have been applied in a generic fashion to the review of standard plant designs and new generic requirements. However, for evaluating risk-assessment-based changes to a plant licensing basis, risk goals were needed that could be applied to individual plants. The draft RGs and SRPs propose the following guidelines for this purpose:
Accident Prevention Guidelines
10-4/RY CDF, as the point beyond which increases in risk would not be allowed
10-5/RY CDF, as the maximum increase in risk from an individual change to the licensing basis
When any change is within a factor of 10 of the CDF guideline, additional evaluations will be focused on factors such as uncertainty, plant performance, and other qualitative considerations before a change to the licensing basis is approved.
Accident Mitigation Guidelines
10-5/RY Large Early Release Frequency (LERF), as the point beyond which increases in risk would not be allowed
10-6/RY LERF, as the maximum increase in risk from an individual change to the licensing basis
When within a factor of 10 of either guideline, additional evaluations will be performed similar to those for CDF.
These goals are intended for comparison with a full-scope PRA (i.e., one which includes internal and external events, full power, low power, and shutdown). However, a less than full-scope PRA can be used, provided that the proposed change is not affected by issues outside the PRA scope, or that qualitative factors can be used to compensate for missing quantitative information.
3. Pilot Applications
To test the process, approach, and guidelines laid out in the RGs and SRPs, the Commission has begun reviewing proposed plant-specific changes to the licensing basis through several licensee pilot applications. These pilots are in the areas of technical specifications, in-service testing, and graded quality assurance, with two or more licensees participating in each technical area. Examples of the types of licensing basis changes being reviewed in the pilot applications include:
changes to allowable equipment outage times;
changes to equipment testing intervals; and
reduced quality assurance measures on specified equipment.
In addition, the NRC staff plans to initiate a pilot application in the area of in-service inspection later this year. For each area, once the pilot applications have been completed, the staff intends to complete and issue application-specific guidance for that area.
4. Individual Plant Examination (IPE) Reviews
In November of 1988, the Commission requested that each nuclear power plant licensee perform Individual Plant Examinations (IPEs) to search for plant-specific vulnerabilities to severe accidents. These IPEs were intended to increase licensee awareness of severe accident behavior, to improve understanding of the most likely accident sequences at each plant, to gain a more quantitative understanding of the overall probabilities of core damage and fission product releases, and to form the basis for reducing these probabilities, where necessary, by modifying plant design or procedures.
The NRC is completing its reviews of these IPEs and, based on the results to date, has issued a report, NUREG-1560, summarizing the resultant insights. Overall, the Commission views the IPE program as having accomplished its goals. All licensees chose to perform a Level 1 (and most a Level 2) PRA, conducted either entirely by utility personnel or with the support of contractors and substantial utility involvement. As a result, licensees have an enhanced understanding of the behavior of their plants, as well as a more quantitative understanding of risk. Licensees have used the IPE results to make improvements in hardware and procedures, as well to collate information that will be useful in developing accident management programs.
The NRC staff also plans to use IPE information in certain follow-up activities, such as assessing the need for regulatory action on remaining risk-significant plant-specific accident sequences, evaluating the need for generic regulatory action where the IPEs revealed common safety issues, and identifying areas in which additional research could improve risk assessment methods. In addition, the staff has used the IPE review insights to develop a draft report, NUREG-1602, which describes the attributes of a quality PRA, and was used to support the development of the RGs and SRPs discussed earlier. Many of the IPEs, as performed, do not conform in all aspects to the quality PRA described in this NUREG. As a result, licensee proposals for changes based on these IPEs will require additional staff review, and/or, in certain cases, additional licensee PRA analysis.
5. Other Risk Assessment Activities
To support the expanded use of risk assessment in the regulatory process, efforts are underway to improve or develop risk assessment methodology in certain key areas, including human reliability analysis, plant aging, fire protection, and digital instrumentation and control. We also plan to expand our international cooperation in the risk assessment arena, sharing with other regulators our methods, our tools, and our experience with various applications.
In addition, we are using risk assessment in our accident sequence precursor program, to analyze better the risk significance of operating events. We also are continuing to develop more structured ways of addressing uncertainties in risk-informed decision-making, considering factors such as the uncertainty in data and models used, as well as the uncertainty resulting from what is not modeled.
B. Risk Assessment Topical Areas
As you can tell by this point, the integration of risk assessment insights with traditional approaches to regulation continues to be a complex endeavor. I would like to illustrate this complexity further by briefly discussing two areas of continuing NRC staff effort: first, how to address PRA uncertainties when evaluating conformance with decision criteria, and second, how to incorporate degraded equipment into PRA modeling.
1. Addressing PRA Uncertainties in Evaluating Conformance With Decision Criteria
In considering uncertainty, we must understand two issues: First, what do PRA results really tell us about whether a given reactor meets a specified decision criterion? Second, in what defensible manner can PRA results be used to set and analyze decision criteria? As I said earlier, the NRC staff has proposed a core damage frequency (CDF) value of 10-4/RY as the guideline where further increases in CDF would not be acceptable. For plants with CDFs <10-4/RY, the proposed guidelines would permit small changes in CDF (no more than 10-5/RY CDF).
These decision criteria do not account for uncertainty in the PRA results, but use point values to determine if the criteria are met. I will now describe briefly an approach that allows the uncertainty of PRA results to be factored into the decision-making process. However, I will use median values rather than mean values as required by the guidance. This, of course, can be remedied easily, but it is not my point of emphasis.
Remember that the actual CDF is always unknown;PRAs only provide an estimate of this unknown CDF, with the goal that it not exceed 10-4. Because we can never be absolutely certain of the actual CDF value, we can never be absolutely certain that the criterion has or has not been met. Further, because of the uncertainty in the PRA estimate, we cannot conclude positively that the criterion has been met simply because the PRA estimate does not exceed 10-4. Recognizing the presence of these uncertainties, however, we can compute the probability that the goal has been met after the PRA estimate becomes available. Let me indicate how this may be done.
We use a Bayesian model to calculate the desired probabilities. Our parameter of interest is the CDF for a given nuclear plant. A log-normal probability distribution is used to quantify the uncertainty, and to express the best state-of-knowledge about the actual CDF prior to the PRA results. After the PRA results become available, a log-normal distribution is also used to characterize the uncertainty in the PRA estimates.
Let me digress for just a moment to remind you of some of the properties of a log-normal probability distribution. First of all, it is related to the normal distribution through a logarithmic transformation. It is a two-parameter distribution, and can be completely characterized by its median and an error factor. The distribution is positively skewed (i.e., the long right tail), and frequently is used in PRAs to model uncertainty in the failure rate of components.
For ease of communication, let denote the unknown (actual) CDF, -hat denote the PRA best estimate of , and * denote the decision criterion of 10-4. Using Bayes' theorem, we can then write Equation 1. This expression provides the probability that the criterion has been met, given the uncertain PRA results, and given Equation 2, where is the posterior median of (i.e., after the PRA results are known), and is the posterior standard deviation of n.
Note that there are two primary sources of uncertainty which can be accounted for explicitly in the analysis: (1) uncertainty in our state-of-knowledge about the actual (unknown) CDF prior to the PRA results, and (2) uncertainty in the PRA estimate of CDF. We will use the probability to measure (in the sense of a figure-of-merit) the compatibility of the reactor with the decision criterion based on the PRA-produced estimate. Let me illustrate these notions.
For a given hypothetical reactor, suppose that we ultimately want to assess the compatibility of a PRA core damage estimate having an error factor of 10 with the proposed decision criterion for CDF of 10-4/RY. Using a log normal prior distribution (having a median of 1.2 X 10-4 and an error factor of 12), and a PRA median estimate of 7X10-5, the probability that the actual core damage frequency is less than 10-4 is computed to be 0.54. If changes were proposed to the plant licensing basis that would raise this estimate to 7.2X10-5, the assurance probability would remain basically unchanged. However, if the estimate increased to 9X10-5, the assurance probability decreases to .49.
Figure 1 gives a plot of the probability that the criterion of
10-4 is met as a function of the PRA estimate, with error factors of 5, 10, 30, and 100 with the same prior distribution as considered above. As the CDF increases, the probability of meeting the criteria decreases. Note also that the assurance probability increases as the error factor decreases for estimates smaller than 10-4, and the assurance probability decreases as the error factor decreases for estimates above 10-4.
Obviously, such analyses are useful in setting and analyzing decision criteria. Remember, however, that this procedure depends upon several probability modeling assumptions whose validity is open to question. In addition, some factors exist that are not quantifiable (e.g., management and organization), or that have been explicitly omitted from the analysis (e.g., external events, for a PRA of less than full scope). The sensitivity to these omissions and assumptions may be examined by considering alternative models, by providing allowances for the missing analyses, and by considering qualitative factors such as recent plant performance. In addition, biases in the PRA results may be accounted for in the Bayesian model through the introduction of location and scale parameters. The location biases are appropriate when known components of the PRA are omitted from the analysis (e.g., low power and shutdown, external events, etc.). The scale biases are applicable when the uncertainty in the PRA estimates is known to be understated. From these considerations a range of solution values can be obtained that better reflect the uncertainties.
I continue to believe that it is important for the NRC to develop approaches and methods for adequately treating uncertainty in the decision-making process. The question that remains is how can we better use PRA results in decision-making despite these uncertainties, while reducing the remaining uncertainties where practical.
2. Incorporating Degraded Equipment Into PRA Modeling
Typically, PRAs model components as being in either a failed state or an operating state, and do not consider degraded equipment as a separate category. This practice can result in either non-conservative assumptions that would treat degraded equipment as operable for PRA purposes (despite marked differences in actual reliability that would undermine PRA validity), or, conversely, conservative assumptions when degraded equipment that is still basically functional is considered failed for PRA purposes. To illustrate the effect of such assumptions on PRA modeling, consider two examples of events that involved degraded equipment.
The first event, which is well known, occurred in April 1991 at the Shearon Harris plant. During testing while in a refueling outage, common cause failures were identified that would have affected both trains of high pressure safety injection. The failures involved alternate mini-flow lines for both Charging/Safety Injection Pumps (CSIPs). These alternate mini-flow lines are designed to protect the CSIPs for accidents in which the reactor coolant system (RCS) repressurizes after safety injection is actuated. Water hammer in these alternate mini-flow lines had damaged the relief valves and test connections such that a significant portion of the safety injection flow would be diverted from the RCS. The conditional core damage probability associated with this event was estimated initially to be 6.3X10-3, assuming that the degraded system would provide no safety injection flow. Later evaluations showed, however, that two charging pumps possibly could have provided adequate injection flow even with the failed relief valves. If this were the case, the conditional probability estimated for the event would be about 1.3 X 10-4 without the use of Steam Generator depressurization, and 2.3 X 10-5 if Steam Generator depressurization and Low Pressure Injection were effective in providing core cooling.
A more recent example of degraded equipment occurred at the Crystal River plant earlier this month. A portion of the emergency diesel generator (EDG) radiator exhaust air was recirculating back to the radiator inlet and caused the average outside air temperature being supplied to the EDG components to increase. This affected the supply air temperature going to the radiator, the engine room (including the EDG and other components), control room ventilation air, and engine room combustion air for both EDGs. The increased supply air temperature could allow the design basis temperatures to be exceeded for all of these areas and components. Information received from the EDG vendor indicated that the maximum load that could be carried by each EDG, given these supply air temperatures, was 3,000 kilowatts--which is below the maximum expected loading required to mitigate various accidents. The licensee is developing a modification to eliminate this air recirculation.
These two events provide some perspective on the importance of degraded equipment as a factor in performing risk analysis. I believe that as we progress further toward risk-informed regulation, the consideration of degraded conditions will become even more important. A plethora of data exists that describes degraded equipment and/or conditions. One source of this data is, of course, plant maintenance logs. The NRC inspection program accounts for the fact that many degraded conditions will be self-revealing, and places high emphasis on post-maintenance testing and substantive surveillance testing at appropriate intervals. For areas not covered, the NRC and the industry need to determine how components and systems can be ranked more appropriately in accordance with their risk significance. The omission of degraded equipment from PRAs could undermine seriously the validity of the results.
IV. NRC FOCUS AREA: DIGITAL INSTRUMENTATION AND CONTROL (I&C)
In November 1995, I asked the NRC staff to develop an updated regulatory framework that would support timely reviews of nuclear power plant digital I&C systems while ensuring that safety margins were not compromised. The staff has recently completed this effort. The primary product is a revision to the Standard Review Plan, Chapter 7, entitled "Instrumentation and Control," which incorporates six new Regulatory Guides endorsing IEEE standards on software quality, new Branch Technical Positions addressing review aspects of digital systems, and new sections in Chapter 7 on diverse I&C systems and data communication systems. The major revision to this chapter was the addition of detailed review guidance covering the regulatory basis, information to be reviewed, and detailed acceptance criteria. The NRC also has endorsed key Electric Power Research Institute (EPRI) topical reports dealing with digital I&C topics including dedication of commercial-off-the-shelf software and electromagnetic interference protection. In addition, twenty publicly available NRC technical reports provide the background for this guidance, and cover such digital I&C issues as diversity analysis, software reliability and safety, environmental effects, programming languages, human-system interface, and software verification and validation. The overall framework provided by the Standard Review Plan revisions and supporting documents sets out an acceptable approach to implementation of digital I&C systems, but is not a requirement. Licensees may provide alternative approaches with appropriate justification.
On a related topic: early in 1995, the NRC asked the National Research Council of the National Academy of Sciences to conduct a study on the application of digital I&C technology to commercial nuclear power plant operations. Their recently issued final report defined and analyzed technical and strategic issues, along with conclusions and recommendations. The NRC staff has reviewed the report and responded to all 39 recommendations. Many of the recommendations were addressed in the update to Chapter 7 of the Standard Review Plan. Several others concerned research areas still under consideration, such as (1) how to measure digital hardware and software reliability, and (2) development of techniques to ensure proper integration and completeness of design specifications so as to address the "total system requirements" (i.e., hardware, software, human, and system functions).
V. NRC FOCUS AREA: INTERNATIONAL COORDINATION AND COOPERATION
As you know, the safety of nuclear power is an issue that transcends national boundaries. Safety problems, concerns, and improvements in one nation are very likely to be matters of interest for others. Many questions related to nuclear regulation--for example, the causes of the nuclear accidents at Three Mile Island and Chernobyl, the effects of exposure to low-level radiation, or new approaches to reactor safety--command attention throughout the world. To enhance international communication and cooperation, representatives of eight nations formally created the International Nuclear Regulators Association (INRA) during (or, in the case of Germany, following) a meeting in Paris, France in May of this year.
The Association was established so that the most senior officials of well-established independent national nuclear regulatory organizations could discuss and make recommendations on issues of mutual concern. The senior regulatory officials who signed the terms of reference for INRA were from Canada, France, Japan, Germany, Spain, Sweden, the United Kingdom, and the United States.
The purpose of INRA is to influence and enhance international nuclear safety from a regulatory perspective. The organization seeks to carry out this goal in several ways:
by establishing a forum in which senior regulators can exchange views;
by building a global nuclear safety culture;
by encouraging the most efficient use of resources;
by seeking international consensus on nuclear safety issues and facilitating international cooperation;
by cooperating with other international and national organizations involved in nuclear safety; and
by identifying nuclear regulatory challenges.
The INRA will act by consensus to fulfill its objectives, and will make recommendations to international and national agencies on nuclear safety issues. At the constituting meeting in Paris, the founding members of INRA elected me to serve a 2-year term as the Association's first chairman. I am pleased and honored by the challenges and opportunities that come with this responsibility, and I will continue to make INRA issues an area of personal focus.
VI. CHANGES TO THE REGULATORY ENVIRONMENT: ELECTRIC UTILITY RESTRUCTURING
The final topic I would like to cover today relates to a significant change occurring in the NRC regulatory environment, related to electric utility restructuring. As many of you are aware, the Energy Policy Act of 1992 included provisions that enabled wholesale competition in electricity generation. As the transition to a competitive market began to take shape, several areas of NRC concern began to emerge. You know from my earlier description of the NRC mission that the NRC is not an economic or rate regulator. However, as utilities restructure internally, as ownership changes, as mergers occur, and as licensees work to control and reduce costs, the NRC must understand and respond appropriately to the effects of the changing business environment on nuclear safety. NRC challenges related to electric utility restructuring fall under three general headings: (1) the availability of funds for decommissioning; (2) electrical grid reliability; and (3) the impact of cost-competitiveness on safe nuclear operations.
A. Decommissioning Funding Assurance
Under Section 161 of the Atomic Energy Act, the NRC has general authority to regulate the decommissioning of the nuclear facilities and materials that it licenses. Existing NRC decommissioning regulations require power reactor licensees to set aside funds periodically in external trust fund accounts (or to provide third party guarantees for estimated decommissioning costs). As such, by the time a licensee permanently ceases operations, the total amount of funds estimated as needed to complete decommissioning is expected to be available.
In the emerging environment of electric utility restructuring, the NRC has had to re-evaluate certain aspects of these provisions for decommissioning funding assurance, including the NRC definition of "electric utility," the potential impact of new ownership arrangements, and the problem of above-market or "stranded" costs. In February 1996, the NRC issued a comprehensive action plan to provide a framework for evaluating this overall area. This action plan has resulted in a range of NRC actions, which I will discuss briefly.
1. Commission Policy Statement
In September 1996, the Commission issued for public comment a draft policy statement on electric utility restructuring and deregulation. The Commission recently approved the final version of the policy statement, and it should be issued shortly. The policy statement indicates that the NRC:
will continue to conduct its financial qualifications, decommissioning funding, and antitrust reviews;
will identify all direct and indirect owners of nuclear power plants;
will establish and maintain working relationships with rate regulators (including the Federal Energy Regulatory Commission (FERC) and the State Public Utility Commissions (PUCs); and
will re-evaluate the adequacy of its regulations in this area.
2. Rulemaking Activities
The NRC is also about to issue for public comment a Proposed Rule on decommissioning funding, based on the results of an April 1996 Advance Notice of Proposed Rulemaking and the continuing analysis of emerging industry developments. The proposed rule would modify NRC decommissioning regulations in four areas:
First, it revises the NRC definition of "electric utility," to ensure that decommissioning funding assurance requirements are clarified for all responsible licensee entities.
Second, it allows credit on the earnings from decommissioning trust funds.
Third, to keep the NRC informed of licensee decommissioning fund status, it requires periodic licensee reports on the status of such funds and any changes to licensees' external trust agreements.
Fourth, to ensure adequate licensee accumulation of decommissioning funds, the NRC would take additional action as needed on a case-by-case basis, either independently or in cooperation with the FERC and the State PUCs, including the modification of a licensee schedule for accumulation of decommissioning funds.
3. Related Actions
Several other significant NRC actions have taken place in this area. Two draft Standard Review Plans were issued for public comment, one in the area of licensee financial qualifications and decommissioning plans, and the other related to antitrust reviews. The NRC staff expects to issue the final Standard Review Plans later this summer. In addition, through a variety of measures, NRC has upgraded its knowledge of licensee ownership and antitrust conditions. Numerous meetings have been held with industry representatives, State and Federal rate regulators, the financial community, and other stakeholders, and staff-level liaisons have been established where appropriate. The overall effect of these measures has been to improve NRC, licensee, and public awareness on issues of concern related to electric utility restructuring.
B. Electrical Grid Reliability
An equally important area of NRC focus has been electrical grid reliability, or security. In recent years, NRC probabilistic risk assessments have made it clear that a Station Blackout at a nuclear power station is a major contributor to core damage frequency. While Station Blackouts have been extremely rare to date, the possibility of a Station Blackout continues to be an area of NRC focus.
In 1996, within a 5-week period, two electrical disturbances on the United States' Western Grid caused 190 power generating plants to trip off-line, including several nuclear units. In reviewing the electrical disturbances, the Western Systems Coordinating Council listed the following contributing factors: high Northwest transmission loads; equipment out of service; inadequate maintenance of right-of-way; operation in a condition in which a single failure would overload parallel lines, triggering cascading outages; communication failures to neighboring utilities, prior to the disturbances; and the lack of response to earlier events.
These events and studies tell us that, while nuclear generating stations are robust in design and operational standards, they also are vulnerable to grid disturbances, and especially to Loss-of-Offsite-Power events. Grid reliability governance structures must reflect this vulnerability. Standards of performance, operational criteria, and training of personnel are critical oversight issues that must all be factored in and properly addressed as deregulation goes forward. To address concerns in this area, the Department of Energy (DOE) has created a working advisory committee on the reliability of the U.S. electric system. This committee is considering whether current efforts to maintain reliability are sufficient to ensure future reliability, and whether a need exists for increased Federal authority over reliability in the future. The NRC has been coordinating with DOE, and will continue to monitor closely the impact of electric utility restructuring on grid reliability.
C. Cost-Competitiveness and Safe Nuclear Operation
The NRC also continues to focus on the possible impact of cost-competitiveness pressures on safe nuclear operations. While the overall safety of the U.S. nuclear power industry continues to improve, NRC safety assessments at several reactor facilities have identified deficiencies that may stem from the economic pressure on a licensee to be a low-cost energy producer, which in turn may limit the resources available for corrective actions and plant improvements. The NRC is developing measures that would identify plants where economic stress may be adversely impacting safety.
In closing, I hope that this discussion has given you a better understanding of the NRC regulatory perspective on nuclear power plant safety, and has enhanced your appreciation of some of the technical challenges the NRC faces. Here I have covered only a sampling of current issues, but in each case, a sound regulatory approach requires both theoretical knowledge and practical experience, combined in the unique NRC mixture of science, technology, law, and public policy. The overall regulatory agenda must then be evaluated in relation to the three questions that derive from my overarching vision for the NRC: (1) as an agency, are we fulfilling our primary mission of protecting public health and safety, promoting the common defense and security, and protecting the environment? (2) as regulators, are we effective? and (3) have we anticipated and readied ourselves for change?
Thank you for the invitation to speak again at this annual safety course, and thank you for your attention. I will be happy to address any of your questions.
NRC speeches are available through a free list serve subscription at the following Web address: http://www.nrc.gov/public-involve/listserver.html. The NRC homepage at www.nrc.gov also offers a SUBSCRIBE link. E-mail notifications are sent to subscribers when speeches are posted to NRC's Web site.