dpb wrote:
Dr. Hardcrab wrote:

"Stormin Mormon" wrote in message
...
Three Mile ISland...

I still consider that a near miss:
...

Actually, in the end it was a very good test and demonstration of the
adequacy of the system design to handle a LOCA (albeit an operator-error
induced one, but a LOCA nonetheless).

Probably ought to clarify that a bit--the LOCA (loss of coolant
accident) part wasn't actually an operator error; it occurred when the
PORVs (pilot-operated relief valves) did not automatically reseat after
they opened.

The problem occurred on the operator response to the incident wherein
they interceded w/ the HPI (High pressure injection) systems and later
the RCPs (reactor coolant pumps) on the basis of their misinterpretation
of the pressurizer level instrumentation indication and the fact they
did not recognize they had a leak (stuck PORV).

This came about because the PORV location was close enough to the
pressurizer outlet that the liquid-water interface became comingled
owing to the flow disturbance. This confused the level indication which
was a dP cell across the interface.

Consequently, they erroneously concluded that somehow they were in
danger of filling the pressurizer solid w/ water which is a no-no, hence
they turned off the HPI to avoid (they thought) doing that. Then, after
a while as they continued to lose cooling water, the RCPs began to
cavitate and they were turned off to prevent damage to them. At that
point they then had a core becoming uncovered and no longer had forced
circulation of what coolant they did have and things went downhill from
there.

As I noted previously, the shift which came on next recognized the
symptoms when they were going through the shift turnover meetings and
began recovery operations immediately.

As a very brief synopsis, this leaves out a lot of detail, of course,
but is a general description of what went wrong at TMI. There are a
couple of obvious things here, the most obvious of which was the
question of how, if the system were in danger of going solid, could
there have been such low coolant level as to have caused RCP cavitation?

What went right, of course, was that HPI was incorporated in the system
and once it was restarted to cover the core and the RCPS were again on
to circulate flow, the accident was on its way towards mitigation.

Of course, the sad part is that if the operators had simply left the
system alone, there would have been nothing but an operational incident
of the stuck PORV (for which there was already a pending corrective
action/notice in place) and after a short outage to repair/correct it,
the plant would have been back in operation.

--