Re: TMI root causes

[Andy Hull <hull@mail.sep.bnl.gov>]

Perhaps it's time to reread the Kemeny report. It spoke critically about
"mind set".
 Andy


At 10:35 AM 3/6/98 -0600, Brian Rees wrote:
>Sandy wrote:
>"The Root Causes of the TMI accident were:
>
>1. Refusing to believe their instrumentation
>2. Lack of or inadequate system's training, as well as general 
>    overall training and competency reviews
>3. Inattention to details
>4. No "real" Emergency Plans or Procedures, nor, Emergency Plan 
>    training.
>
>All of these issues could have been mitigated with the necessary 
>management attention. Of course in the 1980's, the edict was to keep 
>the plants running and maximize profits. Overhead $$ were to be 
>minimized. Training, while provided, was not detailed at all, and, 
>Emergency Exercises, for all practical purposes, were a joke!"
>
>
>The indication for the motor operated relief valve, (the one that stuck 
>open, and directly lead to the rest of the accident) was actually an 
>indication that power was supplied to the motor.  When the motor  was 
>denergized, the control panel indication was that it had closed.  This 
>indication was (initially, and for quite a while) believed. 
>
>The TMI control room had something like 2,000 alarms that sounded during a 
>"routine" scram, sorting out the details is not a simple task.  There were 
>other indications that, in retrospect, should have had more attention paid 
>to them.   Many indications were disregarded later because they conflicted 
>with what was expected,  this can be attributed to lack of 
>training/understanding.  
>
>There are of course other problems that surfaced during the time we've had 
>to look back, but until you're the person who's trying to respond to a 
>myriad of alarms, and make the proper decision in VERY short time period, 
>it's difficult to really appreciate how fast (literally) hundreds of 
>considerations are made. Yes, that's what the operators are paid for, but 
>keep their  operating environment in mind along with the events.  I still 
>believe that it was the ($$$/"management") operating environment that lead 
>to some of the *faulty* decisions that were made (like turning off MCPs 
>because they were cavitating). 
>
>We must keep these things in mind, there are times that we (HPs) could be 
>susceptible to some of the same traps.
>
>Brian Rees
>brees@lanl.gov
>
>