Research Reactor Procedure Writing and Root Cause Analysis Advice
The following was provided by Mike Whaley, Associate Director at the UT-Austin TRIGA Reactor.
I don’t claim to be an expert in procedures or root cause analysis, but I have found the following useful:
- I don’t know much about the Professional Procedure Writer’s Association but the manual is pretty good and captures most of what I understand about procedures. You should not feel you have to follow it the PPA manual verbatim if you want to do something different than suggested, but you should make your own facility specific writer’s guide. IMHO the key isn’t necessarily to use a specific system, the key is to not have traps the to be consistent in format.
PPA AP-907-005 Procedure Writer’s Manual
Here’s a link to how to develop a process description, I haven’t used it but it might be helpful.
- Personnel errors are virtually never a root cause; 99% of the time if you have an event that starts with a personnel error the root cause will be some aspect of management. INPO developed a Management Risk Oversight Tree (I think, not sure, from NTSB). DOE is an INPO member, and frequently adopts INPO guidance for their use. As a government agency DOE information is usually available to the public. DOE took the INPO guidance developed for nuclear plants and used it as a DOE guide:
DOE G 231.2 – OCCURRENCE REPORTING CAUSAL ANALYSIS GUIDE
DOE G 231.2 is an extremely useful tool because it lays out Management Oversight and Risk Tree analysis applied directly to nuclear operations. The guide describes the method, defines the terms, and provides application guidance. It’s a complicated system, but once you try to go through the process make sense.
The only cautions I have:
First I always look at the definitions of cause codes before considering the cause code final – I don’t always understand what the short phase is intended to mean at first glance.
Second you should look through all related codes to make sure everything get captured.
- A more general look root cause analysis that includes less rigorous systems is available here.