Application of the technology neutral framework to sodium cooled fast reactors

Abstract

Sodium cooled fast reactors (SFRs) are considered as a novel example to exercise the Technology Neutral Framework (TNF) proposed in NUREG- 1860. One reason for considering SFRs is that they have historically had a licensing problem due to postulated core disruptive accidents. Two SFR designs are considered, and both meet the goals of the TNF that LWRs typically would not. Considering these goals have been met, a method for improving economics is proposed where systems of low risk-importance are identified as candidates for removal, simplification, or removal from safety grade. Seismic risk dominates these designs and is found to be a limiting factor when applying the TNF. The contributions of this thesis are the following: 1. Functional event trees are developed as a tool to allow different designs to be compared on an equal basis. Functional event trees are useful within the TNF as a method for the selection of Licensing Basis Events (LBEs) which take the place of traditional Design Basis Accidents. 2. A new importance measure, Limit Exceedance Factor (LEF), is introduced that measures the margin in system failure probability. It can be used directly with the TNF where standard importance measures cannot. It also reveals that some systems that appear to be of high risk-importance with standard importance measures may have significant margin. 3. The seismic risk dominates these designs. It is shown that even under optimistic assumptions, the goals of the TNF cannot be met by a typical reactor. The effect of seismic isolation to reduce the frequency of seismically initiated large releases is also analyzed and found to be insufficient to reach the goals of the TNF. A limit on initiating event frequency that is consistent with current practices is proposed.