Monday 14 March 2011

Japan

I'm sure everyone is aware of recent events in Japan, looking around the internet it's fantastic to see the support offered to the people effected by the earthquake. Whether that support comes from the emergency and aid services, or from the regular folk who are wishing the Japanese well.

There has been a lot of discussion on the nuclear power plant which was only 40 miles from the epicentre. (40!!!)
I cannot claim to understand the current predicament they're facing, but I am told that this facility is 40 years old, and the safety features installed by the nuclear engineers are fantastic. I read a very interesting post on reddit by user rynvndrp.

"After scram (complete shutdown), about 5% is still there and that level reduces in half every 15 minutes. So for a 3000MW thermal reactor (like this one), 150MW is still there after immediate shutdown, 75MW 15 minutes later, 37.5MW 30 minutes later, and so forth.
In a Boiling water reactor of that model, there are 3 cooling pumps in series (one for the cold feed and two after each hot feed loop). During a scram, those pumps must be shut off so that you don't get vibrational feedback (which could break a pipe). However, in the core is designed such that there are jet pumps (16 in parallel on the standard design) on the wet stream loop. This jet pump has no moving parts but can circulate the some water regardless of what the pumps are doing. Further, there is an single emergency main pump on the main loop that can push the full coolant load. This pump can be driven by any one of several diesel systems or a battery backup.
If all of this fails, you then start to depend on the cold water back up re-condenser. In a BWR, underneath the first reactor containment, there is a large torus shaped area that is filled with a large amount of water. You can drop the steam from the reactor into this ring. The steam recondenses and reduces pressure. With the jet pumps, you can push the cold water into the reactor. This system can prevent first containment failure but doesn't prevent core damage.
Then there is a second level containment. The design based accident for this reactor is assume a main coolant pipe dissappears (double guillotine break), all of the coolant flushes out and there is no way to cool down the reactor. In this case, 600MW post scram level will melt the reactor and possibly damage first containment, but second containment can hold the total heat produced post scram. You also have a coreium (term for molten core) catcher that catch the core, mix in a huge amount of radiation shield, and reduce the temperature.
What is happening: Earthquake happens, core is scrammed because something might break. When the scram happens, main pumps turned off, emergency main pump turned on. However, the diesel generator doesn't work, maybe damage in earthquake, maybe something else. They turn on the battery system. It can last 4 hours, which is enough to reduce core power to 10kW. However, it is hot in there and 10kW is still enough to continue to make steam. So the water level (which tells you how much steam has been made) is dropping. They can go to the recondenser if they need to, but the recondenser doesn't rule out core damage (not melting, but material warping and oxidation) which would be expensive or impossible to repair. So for now, they are bringing in new coolant so they can prevent expensive core damage.
They are NOWHERE close to the design based accident, and there is very little worry about containment failures. They are doing the evacuation because there isn't the safety factor they want, not because they are anywhere close to a radiation leak."

I am not up to date with the recent developments at the facility, but it's a fascinating read about the intricacies of the safety features.

3 comments:

  1. Good read. I've been following this stuff closely, but no one explains it well.

    ReplyDelete
  2. There is always far too much panic over these things. Nobody hears about it until there is a problem, which gives the wrong impression due to how rare problems really are.

    ReplyDelete
  3. I agree with both posts. I try to follow it up closely, but the different media never give the same info

    ReplyDelete