Thank you. It's a pleasure to be here and to be invited back to the PSA conference. I see many familiar faces from last year's meeting in New York.

As most of you are aware, the past year has been marked by positive change in the way federal regulators approach nuclear plant oversight.

The Nuclear Regulatory Commission has undertaken the historic task of reforming both its oversight procedures and its regulations. This new course will allow the agency to focus resources more effectively on safety.

Probabilistic safety assessments are playing a key role in this process.

This is a remarkable achievement, considering that until now federal regulators have relied on policies that date to the birth of the atomic age—policies that lacked insights from plant operating experience. Today, the NRC recognizes the value of more than 2,500 reactor years of U.S. operating experience. Worldwide, the industry's collective experience amounts to 6,000 reactor years. This experience ... combined with the development of analytical tools that are familiar to all of you ... provides a sound basis for fundamental change.

In June, nine nuclear plants in every NRC region began testing the agency's new regulatory oversight process. The lessons learned from this pilot study will help bolster a full industry rollout next April.

The pace at which the NRC is moving toward an objective oversight process … informed by risk analysis … didn't seem possible several years ago. Now, there is no reason that a shift toward risk-informed regulations cannot proceed at a comparable rate. Anything less would be bad policy.

As is often the case with reform, however, there is a tendency to preserve the old rather than pursue the new.

Fortunately, NRC Chairman Dicus has demonstrated strong support of reform and is keenly aware of its challenges. Last March, at the NRC's regulatory information conference, she noted that NRC managers must guide staff through the uncertainties of implementing risk-informed oversight and regulation to ensure the transition's success.

In following the commission's leadership, NRC staff has made great strides in developing a framework for applying risk insights to licensing decisions at operating nuclear plants …and to regulatory activities.

However, the commission's policies have not always been applied consistently. Today, I'd like to address one bump in the road, namely, the application of risk insights to proposed regulations for plants that are being closed, or decommissioned.

For some time, the NRC has treated the public health risk of operating plants and decommissioning plants the same way. In other words, certain regulations for operating plants also apply to decommissioning plants.

This practice fails to recognize the obvious difference created by decommissioning status and thus, the difference in the potential risk of accidents. Even today, decommissioning plants must seek exemptions to regulations … such as emergency planning and liability insurance.

Although the NRC has the opportunity to incorporate risk insights in decommissioning plant regulations, it is having difficulty doing so.

Let me give you an example. Earlier this year, NRC staff announced the results of a risk analysis of spent fuel pool accidents at decommissioned facilities that would serve as the basis for new regulations.

The results stunned the industry. The report concluded that a large radioactive release was as likely to occur at a decommissioning plant as it was at an operating plant. Based on that miscalculation, NRC staff recommended that such requirements as emergency planning and financial protection continue for years after a plant closes.

Something was clearly wrong! Radiation release from a spent fuel pool … which is considered insignificant as an operating plant … could not possibly carry the same magnitude of the overall risk as at an operating plant.

The industry wanted to unearth the underlying problem. To that end, NEI commissioned a peer review study of the draft by industry experts who have performed and reviewed dozens of operating plant PSAs.

Our study found that the NRC staff's quantitative analysis was not "probabilistic." Nor was it attuned to "risk." In fact, the NRC draft violated the tenets of a PSA policy statement adopted in 1995 by the commission.

A cynic might say that the draft appeared to support a pre-ordained conclusion.

Let me quote the PSA policy statement:

PSA evaluations in support of regulatory decisions should be as realistic as practicable.

Ironically, the drafters of the NRC report vowed to "reduce the unnecessary conservatisms associated with current regulatory requirements and staff practices." Yet the report falls short of that goal. It quantifies risk based on hypothetical accidents and worst-case scenarios that give new meaning to Pascal's theory of probability.

For example, one worst-case scenario assigns an unsuitably high probability that operators of a decommissioning plant would fail to notice a problem with the spent fuel pool over the course of five days. That's 15 shift changes!

This audience knows that a PSA evaluation is of little value without realistic, best estimates for such things as initiating events … equipment and human performance data … sequence development … and success criteria. Otherwise, the fundamental purpose of assessing relative risk is lost. Application of the realistic analysis principle espoused in the NRC's policy statement on PSA was nowhere to be found in the draft report.

I mentioned earlier that the NRC study didn't represent risk effectively. The study combined "worst case" probabilities with deterministic analysis to gauge radiation release.

This approach does not actually reflect risk. It yields misleading conclusions.

Let me cite an example involving the potential zirconium fire in the spent fuel pool. NRC staff calculated the probability of the fire based on whether water in the spent fuel pool would drain to the top of the fuel assemblies. But, the deterministic assumptions were based on complete loss of pool water and obvious uncoverage all of the fuel. These two endpoints in the event sequence are significant, because they can occur days apart from each other.

Had the NRC staff relied solely on probabilistic analysis, they would have assumed realistic conditions. An approach that is used for operative nuclear plants that are in service.

With a proper risk analysis, the NRC could have assessed more objectively the continued need for emergency planning and other existing regulations.

All of this said, I am not suggesting that PSAs should be the sole instrument used in developing decommissioning regulations, but rather that the PSA itself must be performed properly and incorporate realistic event sequences.

There is still time to reassess the NRC's risk model so that it properly clarifying the risk posed by decommissioning plants. The industry's review of NRC's probabilistic framework provides a good starting point.

The industry report isn't alone in contradicting the findings of the NRC draft. In fact, the draft's conclusions clash with NRC's previous assessments of the scenarios in question. A comparison of the overall risk from nonseismic events in the draft study reveal that they are higher by nearly two orders of magnitude than those reached in NRC's 1989 regulation on design basis accidents and spent fuel pools.

This NRC analysis will is simply result in continuation of the status quo for the decommissioning plants. It fails to suggest a reduction in the level of regulatory burden that should accompany the decommissioning of a plant. Simply put, the draft study provides no relevant comparison of these risks and cannot lend context to developing risk-informing regulations for decommissioning plants.

To address this disparity, the industry's detailed report provides concrete examples of how the NRC can fine-tune its PSA model for decommissioning plants. At the same time, the industry's recalculations preserve NRC's overarching goal of adopting safety-focused regulations.

The industry will submit its comments well in advance of a final draft that NRC staff will prepare for commissioners. The final draft is scheduled for release in March.

In light of our report and in view of the commission's demonstrated commitment to real reform, we trust that the commission will ensure that risk insights are applied properly—and in compliance with the agency's 1995 PSA statement.

We look forward to working with the commission as it moves toward a consistent application of safety-focused regulation that includes decommissioning plants.

Thank you.