Agenda - 11-17-1998 - 5a - Laserfiche WebLink

Browse Search

22 -s ,y,~ a., ~; ~~. ~::; •~' E'. Chapter Eight • Spent Fuel Risks fir, public H•elfare are posed by loss of spent fuel pool cooling, radiation overexpu- ?== sure to spent fuel, and spent foe! handling. Near misses have been logged over the years, including numerous repeti- lions of the some events. Spent fuel risk assessments conclude that the probabil- -' ities of the ~•orst case accidents occurring are reasonably lo~v as to provide ade- _. quate assurance of public safety However, these risk assessments are nonconser- . ~ -- vative because several of their key assumptions are invalid. For example, the risk ~.; assessments assume that a relatively long period is available following the initi- ating event to implement actions to prevent or mitigate resulting fuel damage. Yet "' the spent fuel pool temperature and level instrumentation at most nuclear po~+~er (' plants is nonemergency equipment with a history of frequently being inoperable or unavailable follo++ring initiating events. The initiating event may go undetect- __ _• ed for several hours, thus redudng or eliminating the available response time. The risk assessments nonconservatively assume that the operating staff has ' unlimited access to the systems that can be used to mitigate a spent fuel pool acci- ' ' ; dent. The spent fuel pool, along ~~ith most of the equipment that cools and pro- i -•'w vides makeup water to the pool, are located inside the secondary containment `: structure at most BWR plants. Following a reactor accident involving fuel dam- ~~ age, secondary containment can be rendered inaccessible due to extremely high radiation levels. The consequences of such a reactor accident can mechanistically cause a loss of spent fuel pool cooling event. Hence, the risk assessments are wrong to assume unlimited access for restoration and mitigation efforts. - The risk assessments also nonconservatively assume that the probabilit~• of ~` a fuel cladding fire propagating to adjacent fuel assemblies is significantly lower '~ .for BWRs than for PW1Zs because BWR fuel channels act as fire barriers. Many BWR spent fuel assemblies are stored with the fuel channels in place; not for risk t`~=: reduction, but for convenience. There are no requirements or even recommends- ,; lions that BWR spent fuel assemblies be stored with their fuel channels. Hence, it ~: ~;: ^;~: a is wrong for the risk assessments to assume that the fuel channels are ahvavs pre- ~~F;; sent to prevent a fuel cladding fire in a BWR spent fuel pool from propagating. The predominant invalid assumption in the spent fuel storage risk assess- ~,::::. ,•~ ~' menu involves the extent of fuel damage resulting from a spent fuel pool acci- ~~.':.~ dent. From the initial NRC risk assessment in 1975 through the NRC's updated 'f~~;~ risk assessments in the 1980s, it has been assumed that only one-third of a reac- ~{ ~~ for core's inventory ~•ould suffer damage following a spent fuel pool accident. :~E;-- ~~?'; Nuclear power plants currently store the equivalent of several reactor cores in '_sY~ closely packed arrays within the spent fuel pools (see Table r-1-}- Stvdies have demonstrated that a spent fuel assembly discharged within the past three to four _~; years still produces sufficient decay heat to cause its foe! cladding to burn. ~'. Studies have also demonstrated that the resulting cladding fire can spread tc~ involve every irradiated fuel assembly in the spent fuel pool, e~•en those that ,,,~= were discharged many years ago. Therefore, limiting the postulated rele~~e~ from ,... .1Z1 ,.~x ..