Experience with SAMG implementation at various nuclear power plants-problems, solutions

Many nuclear power plants (NPPs) today have Severe Accident Management Guidelines (SAMG) in place to mitigate severe accidents, in the unlikely case such an accident occurs. As such, these plants have increased their level of safety. Yet, it appears that there is a wide scatter in the level of implementation and in the extent and quality of the implementation. NSC Netherlands has executed a number of reviews at NPPs, both in such review missions by the IAEA and on a bilateral basis. Plants reviewed include PWRs, BWRs, and a CANDU PHWR. NSC Netherlands has also reviewed generic SAMG programs, including those by the PWR Owners Group and the BWR Owners Group and the SAMG program for all reactors in Sweden. The reviews usually lasted about one week and involved various staff with international SAMG experience and reputation. Basis of the review were applicable IAEA safety standards and technical reports, such as the Safety Guide for Severe Accident Management NS-G-2.15 and the report SRS 32, Implementation of Accident Management Programmes. Of course, national regulations - where existing - were also considered. Where possible, SAMG exercises were observed, and reviewed for their robustness and latitude, as well as their organisational set-up. Although many positive experiences could be gathered, also a number of weak points have been found. Some plants even had quite a weak implementation, as many basic features of SAMG program were lacking. Also some generic programs were found to have weak spots, notably from the human engineering point of view. Some exercises went well, others were too simply stylized to train staff sufficiently for the immense load that a severe accident poses on the plant staff. Some plants have responded seriously to the findings of the reviews, others lacked interest or budget to implement improvements. Regulator attitudes - where involved - also varied widely. The probable reason being that severe accidents are highly unlikely, and not considered credible events by many power plants. The conclusion is that in-depth external peer review of both generic and plant specific SAMG programs as well as a strong commitment to improve shortcomings are essential to arrive at a credible risk-reducing set of SAMG. And that such review is highly relevant as no plant is immune for severe accidents. (author)