Talk:Natrium reactor

Safety of Sodium-cooled reactors

Statement: by Robert Steinhaus, former physicist at Lawrence Livermore National Laboratory in the FaceBook forum Molten Salt Reactors
Sodium proponents make claims such as "walk away safe" (but in my opinion, make such claims are irresponsible and without good engineering evidence). I respect evidence and the factual accumulated record of multiple decades of reactor operating history. SFRs (Sodium-cooled Fast Reactors) have been built for now 70 years, and anyone having a serious interest should take the time to examine the operating safety record for SFRs as a reactor class.

Conclusion - approximately half of the SFRs constructed over seven decades had their operating lives shortened by a safety-related accident or incident. This is a vastly inferior safety record to any other current reactor class. My specific safety concerns for sodium-cooled reactors are in The case for not building large numbers of Sodium Cooled Fast Reactors

Response: from Captain Roger Blomquist, United States Navy (retired) email 19-Feb-2924
Robert Steinhaus, a Liquid Fluoride Thorium (molten salt) Reactor (LFTR) advocate, has argued that these tests do little to demonstrate passive safety because they were conducted with VIPs present and under ideal operational conditions. In other words, they were public relation stunts, not realistic accident scenarios. In my experience, all reactors are operated under conditions that are the starting points for accident scenario safety analyses. And for many reactors, operating outside of the safety case operating envelope results in an automatic shutdown. Of course, all experimentalists aspire to provide repeatability and precision – it’s good experiment design. I expect an LFTR prototype will also do so in similar full-system tests to support licensing. And I would be very surprised if an LFTR safety demonstration test were conducted without VIPs and journalists present. Perhaps Dr. Steinhaus can suggest some additional specific operational constraints for future tests.