How did they shut down the TMI-2 reactor?

[Drakkith]

I believe that a cooling valve at the TMI NPP malfunctioned, closed, and water was unable to flow to the core. And this valve had malfunctioned at least a dozen times at other plants.

I'd have to check, but I thought the valve was stuck open, which allowed massive amounts of coolant to leak out, dropping the level in the core and eventually causing the meltdown from lack of coolant/coolant flow.

 

[Morbius]

I believe that a cooling valve at the TMI NPP malfunctioned, closed, and water was unable to flow to the core. And this valve had malfunctioned at least a dozen times at other plants.

Fisherman,

There are no valves in the TMI cooling system that when closed prevent coolant flow.

The valve in question failed in the open position after it relieved a pressure spike. The continued outflow from that valve dropped the pressure in the TMI primary system. As the pressure went down, so did the boiling point of the water. It got to a point at which the temperature in the core which is not high enough to boil the water at nominal pressure was able to boil the coolant at the reduced pressure. The core was being cooled by a two-phase mixture of steam and water.

However, the core was undamaged at that point, and the TMI accident was totally reversible up to 90 minutes into the accident. It was at that time that the operators, who were unaware of the fact that the coolant was boiling because they never checked their steam tables, decided that the primary coolant pumps should be stopped because they were making strange noises.

Those strange noises were a badly needed clue that the operators ignored. The reason the pumps were complaining is because they were pumping a two-phase steam and water mixture. The operators missed that.

They decided to "save" the pumps, and in so doing "lost" the reactor.

Dr. Gregory Greenman

 

[QuantumPion]

Fisherman,

There are no valves in the TMI cooling system that when closed prevent coolant flow.

The valve in question failed in the open position after it relieved a pressure spike. The continued outflow from that valve dropped the pressure in the TMI primary system. As the pressure went down, so did the boiling point of the water. It got to a point at which the temperature in the core which is not high enough to boil the water at nominal pressure was able to boil the coolant at the reduced pressure. The core was being cooled by a two-phase mixture of steam and water.

However, the core was undamaged at that point, and the TMI accident was totally reversible up to 90 minutes into the accident. It was at that time that the operators, who were unaware of the fact that the coolant was boiling because they never checked their steam tables, decided that the primary coolant pumps should be stopped because they were making strange noises.

Those strange noises were a badly needed clue that the operators ignored. The reason the pumps were complaining is because they were pumping a two-phase steam and water mixture. The operators missed that.

They decided to "save" the pumps, and in so doing "lost" the reactor.

Dr. Gregory Greenman

Not only did they turn off the reactor coolant pumps due to cavitation, but they also turned off the automatically-actuated safety injection pumps because they thought the system was TOO full of water (due to the pressurizer level indication).

 

[Morbius]

Not only did they turn off the reactor coolant pumps due to cavitation, but they also turned off the automatically-actuated safety injection pumps because they thought the system was TOO full of water (due to the pressurizer level indication).

Quantum Pion,

Yes - that is also true. The emergency coolant injection system started and the operators turned it off.

They didn't realize that they had an accident underway. They thought they had "stabilized" the plant because the temperature stopped changing. The reason the temperature stopped changing is that they were on the "saturation line" - they were boiling.

After he completed his inquiry, Prof. Kemeny came to MIT where I was a student to give a seminar on the inquiry. When he was in the TMI Unit 1 control room, he asked the operators to get him a "steam table" - the Equation of State for water. It took the operators about 45 minutes to find a copy of Keenan and Keyes "steam tables".

Like the Eastern Airlines pilot that didn't know he was losing altitude while he fiddled trying to change a bulb for the landing gear locked indicator; the operators didn't know they had an accident in progress.

Dr. Gregory Greenman

 

[texpat]

>After he completed his inquiry, Prof. Kemeny came to MIT where I was a student to give a seminar on the inquiry. When he was in the TMI Unit 1 control room, he asked the operators to get him a "steam table" - the Equation of State for water. It took the operators about 45 minutes to find a copy of Keenan and Keyes "steam tables".

This is highly reminiscent of Feynman's demonstration of the shuttle SRB o-ring behavior in cold temperatures.

Theatrical, but damning demonstration, of a giant failing. I'm sure many hearts sank at the conclusion of both demonstrations.