05-13-2008, 07:09 AM
Ah, confusion about passive.
You've pointed out a mistake I've made: I have confused the term "passive" with "no human necessary". The "no human" can mean "automated". When I read the articles' "no human" I thought they meant the "passive" controls.
I am paid to be a software engineer (I call myself programmer tho). I think for many things automated is better than manual, but I also know firsthand that as the saying goes, it takes a computer (automation) to really eff things up. Complex systems that are automated using software written by people cannot be treated as 100.0% reliable. Maybe 99.9999% or so, sure. It may also be true that my qualifier "complex" may slide as technology progresses. ABS in cars for instance would be considered complex in the 1960's but not so by the 1990's.
My point is, I don't care that much about whether a plant is manual or automated. The automated would most likely be better, but I still wouldn't totally trust it.
The difference in plant designs that is important to me is what happens when either operators or automated processes fail to do the right thing.
Let me use the Browns Ferry atomic plant fire as an example. (My source is a reference book I saw in the library last week that was transcripts of the hearings on Capitol Hill in 1975. I haven't read another source yet.)
Engineers were testing the wall of "the containment". They were looking for an airflow that would indicate that a hole in the containment had not been sealed. They did this by lighting a candle next to the wall. The flame went horizontal, indicating an airflow. So they sprayed some sealing foam into the hole. They did the candle test again. The flame went horizontal again. The engineer with the candle noticed that the flame had caught the foam on fire. Another worker got a fire extinguisher. The hole was sprayed with the extinguisher. The engineer noted that the spray had pushed the foam into the containment and the foam was still on fire.
"What are the odds on that?" one of the senators asked at the hearings. (I paraphrase -- it was more of "I wonder if mr. soandso could tell me the odds of that happening...")
Anyway, the fire started spreading out of control. The human operator of the reactor unit did the right thing, cooling the reactor and slowing down its pace. Yay human operator. Note that an automated system may have done the same. Yay automated system.
However, after four hours, the fire had damaged the electronics used to control the cooling and the core temperature started going up again. Darn.
I had to stop reading and go back to my schoolwork, so I didn't find out what came after that.
So would it be much issue if it were a pebble reactor or some other reactor design that just falls apart when things get too intense? No. No problem, except a reduction of power output for a day or so.
But what about a "landmark" reactor? Certainly a big problem, if there's any chance of the "containment" not containing the mess. Wait, wasn't there a hole in the containment?? Can we really trust a container that was just worked on by Larry, Moe, and Curly?
And that is only if the runaway process does not run away too far. Well, supposedly we don't have to worry about that. See, though, that this is where I stop worrying about it: when you have a pebble reactor or some "passive" design that cannot "run away". The big early reactors don't give me that warm fuzzy.
I don't trust people nor their automated processes to keep it from happening.
Even the container is unacceptable to me. How many thousands of years does it need to contain? How much damage will it do when nature or misanthropes break it open early? Unacceptable.
To compare this to my riding a bus is not a valid comparison. I can accept the risk that the driver will do something stupid that will kill me. However, if every time a bus crashed we had hundreds or thousands of square miles of toxic dead zone for ten thousand years I'd say no buses please.
Comparing it to the burning of coal is valid. The coal damage is not as intense, yet is more definite and more widespread. And then there is the likely advent of global warming, which if it ever reaches runaway status, would be worse than thousands of nuclear catastrophes. Anyway, I feel that nuclear design has improved to the point where it is definitely better than coal. I don't have much reason to be optimistic about nuclear waste, yet I am -- I think we'll figure out what to do with it.
I need to crosspost my last thought to the other thread. That is: kandrathe started that thread by lamenting the lack of nuclear research because of anti-nuclear dummies. Yet if you read the original article on the thorium reactor it says "the biggest obstacle to (the idea) is a mature nuclear industry." What? How did it get to be mature if the know-nothings closed it down???
-Van
You've pointed out a mistake I've made: I have confused the term "passive" with "no human necessary". The "no human" can mean "automated". When I read the articles' "no human" I thought they meant the "passive" controls.
I am paid to be a software engineer (I call myself programmer tho). I think for many things automated is better than manual, but I also know firsthand that as the saying goes, it takes a computer (automation) to really eff things up. Complex systems that are automated using software written by people cannot be treated as 100.0% reliable. Maybe 99.9999% or so, sure. It may also be true that my qualifier "complex" may slide as technology progresses. ABS in cars for instance would be considered complex in the 1960's but not so by the 1990's.
My point is, I don't care that much about whether a plant is manual or automated. The automated would most likely be better, but I still wouldn't totally trust it.
The difference in plant designs that is important to me is what happens when either operators or automated processes fail to do the right thing.
Let me use the Browns Ferry atomic plant fire as an example. (My source is a reference book I saw in the library last week that was transcripts of the hearings on Capitol Hill in 1975. I haven't read another source yet.)
Engineers were testing the wall of "the containment". They were looking for an airflow that would indicate that a hole in the containment had not been sealed. They did this by lighting a candle next to the wall. The flame went horizontal, indicating an airflow. So they sprayed some sealing foam into the hole. They did the candle test again. The flame went horizontal again. The engineer with the candle noticed that the flame had caught the foam on fire. Another worker got a fire extinguisher. The hole was sprayed with the extinguisher. The engineer noted that the spray had pushed the foam into the containment and the foam was still on fire.
"What are the odds on that?" one of the senators asked at the hearings. (I paraphrase -- it was more of "I wonder if mr. soandso could tell me the odds of that happening...")
Anyway, the fire started spreading out of control. The human operator of the reactor unit did the right thing, cooling the reactor and slowing down its pace. Yay human operator. Note that an automated system may have done the same. Yay automated system.
However, after four hours, the fire had damaged the electronics used to control the cooling and the core temperature started going up again. Darn.
I had to stop reading and go back to my schoolwork, so I didn't find out what came after that.
So would it be much issue if it were a pebble reactor or some other reactor design that just falls apart when things get too intense? No. No problem, except a reduction of power output for a day or so.
But what about a "landmark" reactor? Certainly a big problem, if there's any chance of the "containment" not containing the mess. Wait, wasn't there a hole in the containment?? Can we really trust a container that was just worked on by Larry, Moe, and Curly?
And that is only if the runaway process does not run away too far. Well, supposedly we don't have to worry about that. See, though, that this is where I stop worrying about it: when you have a pebble reactor or some "passive" design that cannot "run away". The big early reactors don't give me that warm fuzzy.
I don't trust people nor their automated processes to keep it from happening.
Even the container is unacceptable to me. How many thousands of years does it need to contain? How much damage will it do when nature or misanthropes break it open early? Unacceptable.
To compare this to my riding a bus is not a valid comparison. I can accept the risk that the driver will do something stupid that will kill me. However, if every time a bus crashed we had hundreds or thousands of square miles of toxic dead zone for ten thousand years I'd say no buses please.
Comparing it to the burning of coal is valid. The coal damage is not as intense, yet is more definite and more widespread. And then there is the likely advent of global warming, which if it ever reaches runaway status, would be worse than thousands of nuclear catastrophes. Anyway, I feel that nuclear design has improved to the point where it is definitely better than coal. I don't have much reason to be optimistic about nuclear waste, yet I am -- I think we'll figure out what to do with it.
I need to crosspost my last thought to the other thread. That is: kandrathe started that thread by lamenting the lack of nuclear research because of anti-nuclear dummies. Yet if you read the original article on the thorium reactor it says "the biggest obstacle to (the idea) is a mature nuclear industry." What? How did it get to be mature if the know-nothings closed it down???
-Van