I have a pretty lengthy interview with SRNL’s John Marra in the August issue of the Bulletin of the American Ceramic Society regarding materials implications of the nuclear situation in Japan, and I will include a link as soon as the electronic version is online, but I did want to note that the NRC’s Near-Term Task Force has issued its recommendations (PDF).
I just looked over the document quickly. There are some recommendations about hardened vent designs in boiling water reactors and steps for more detailed studies related to hydrogen-producing reactions, hydrogen mitigation and control strategies, and spent fuel pool capability/instrumentation.
Of interest, also, is that the Task Force didn’t wait for the discovery of specific “smoking guns” from the Fukushima Dai-ichi events before making the recommendations, and rightly, I think, notes that this isn’t the first time it has advocated for preemptive changes:
“The Task Force notes that, after the attacks of Sept. 11, 2001, the Commission established new security requirements on the basis of adequate protection. These new requirements did not result from any immediate or imminent threat to NRC-licensed facilities, but rather from new insights regarding potential security events. The Task Force concluded that the Fukushima Dai-ichi accident similarly provides new insights regarding low-likelihood, high-consequence events that warrant enhancements to defense-in-depth on the basis of redefining the level of protection that is regarded as adequate. The Task Force recommendation for an enhanced regulatory framework is intended to establish a coherent and transparent basis for treatment of the Fukushima insights. It is also intended to provide lasting direction to the staff regarding a consistent decision-making framework for future issues.” [emphasis added]
In other words, the task force is acknowledging that there is a class of somewhat rare events that individually carry enormous consequences because these low-likelihood events may carry unanticipated correlations with each other (sometimes referred to as fat-tail events in the statistics and financial fields, the “tail” being the far end of the probability distribution curve). In a broad sense, it amounts to acknowledging, for example, that in some locations the odds of having separate rare evens, such as an earthquake, a tsunami and an extended power outages occuring simultaneously are much higher that than sum of the individual probabilities.
My example oversimplifies the complexities, and nuclear regulators and planners might have given some low-level consideration to these correlations. However, to do it right requires a lot of modeling, computing power and development of rigorous stress tests that use hypothetical and historical data to tease out the correlated risk events. Moreover, the tests and analysis have to be done repeatedly using the latest data from engineers, materials researchers, geoscientists, weather experts plus onsite data from plant operators. The art is to keep looking for correlated risk events, and then delve into why the correlations exist.
This opens up a great opportunity for cooperative work among the international nuclear energy community. This contrasts with the financial and banking communities around the world, which did a lot of pioneering work in fat-tail/long-tail modeling and analysis in the early 2000s, but since have worked to keep their “library” of correlations data proprietary (e.g., each investment or hedge fund attempted to use their models to attain competitive edges).
The nuclear community, however, has every incentive to develop a transparent open-access set of probability and correlations data libraries. This would be relatively easy to put into place, and it is hoped those concepts are included in long-term NRC and IAEA plans.