Abstract
Nuclear safety analysis and licensing criteria are based upon the concept that plant situations that are expected to have a high frequency of occurrence must not pose a danger to the public, and that plant situations that could pose a danger to the public must be limited to situations that have a very low expected frequency of occurrence. This concept is implemented by grouping postulated plant situations (or events) into categories that are defined according to their expected frequencies of occurrence (i.e., high-frequency, low-consequence events, and low-frequency, high consequence events). In plant licensing basis analyses, events in each category must be shown to yield consequences that remain within the limits that are specified for that category. To protect the integrity of this categorization scheme, events must not be allowed to develop into the more serious events that belong in other, higher-consequence categories. In other words, nuclear plant designs must not allow high-frequency, low-consequence events to degrade into high-frequency, high-consequence events. The development of this system of frequency-based categorization is discussed, followed by an evaluation of various methods that could, and could not be used to demonstrate, for licensing purposes, that benign events are prevented from becoming serious accidents.