On 4/16/2011 12:02 PM, Alain wrote:

>> On 4/15/2011 5:26 AM, Bill Pragnell wrote:
>>> Invisible<voi### [at] devnull> wrote:
>>>> Nuclear power works in theory. In practise, if you make even the
>>>> tiniest
>>>> mistake, just once, everything is ruined forever. (Or at least, for
>>>> several centuries.) And there's nothing you can do to fix it.
>>>
>>> That's not really true. Chernobyl was caused by a very long chain of
>>> mistakes,
>>> all committed with a reactor design which was already itself a long
>>> chain of
>>> mistakes. It should be noted that most other countries have never
>>> built a
>>> reactor that could fail as catastrophically as this, even through wilful
>>> sabotage.
>>>
>>> TMI was also long chain of mistakes, which resulted in only the
>>> reactor being
>>> ruined, and they did fix it.
>>>
>>> I suppose you could say that Fukushima was really only one very big
>>> mistake,
>>> i.e. how big a tsunami was ever likely to be. However, they show every
>>> sign of
>>> being able to fix it eventually.
>>>
>> No, the "big" mistakes where:
>>
>> 1. Having no way to cool it, or certainty that the power systems would
>> still work, to do so, if enough failures happened. And, no, battery
>> backup doesn't work, if it lasts less than 24 hours.
>>
>> 2. Placing the old, spent, fuel in something that was ever *less*
>> effectively cooled.
>>
>> And, I would add 3. Presuming that a *big* reactor, which produces
>> massive amounts of power, but where it would be nearly impossible to
>> either make it less hazardous, or run battery backup long enough, or
>> otherwise create a system that *could* compensate for major problems,
>> remains bloody stupid.
>>
>
> Some more gross mistakes at Fukushima:
>
> a. Need *EXTERIOR* power source to run the cooling system when you
> produce that power localy.
> The pumps failed when the power lines TO the reactor got damaged by the
> tsunami. That's a humongously huge mistake!
>
> b. NO passive shutdown mechanism. Sanity *demands* that there are
> several controll rods suspended by electro magnets powered by the
> reactor itself over the core. If the cooling system fails, the turbines
> stop, they no longer produce current, shutting down the magnets whitch
> let the controll rods fall into the core, stoping the nuclear reaction
> and thus the heat generation. Those rods are usualy made of cadmium
> because that metal can absorbs huge amounts of neutrons.
> Those rods must be set and designed so that gravity alone will make them
> fall completely into position.
> There where obviously none! Totaly insane!
>
> c. NO passive cooling mesures. A passive cooling mesure should be enough
> to evecuate the residual heat from the shut down reactor.
>
> d. Severly skipping on maintenance for over 10 years.
> The director of the station said so himself...
>
>
>
> Alain
Yeah. Those would have been a damn good idea too.

a) makes no damn sense to me at all (even if it was some minimal system, 
enough to just keep pumps going), unless the tsunami took out 
lines/systems in the actual turbine part of the system. I would have 
thought those would be internal to the reactor, but.. Then again, their 
systems reuse a lot of waste heat/energy, so its hard to say how complex 
the whole thing was, and thus "where" those systems where, and thus 
whether it was even feasible to have the turbines in the reactor 
buildings. Its a definite WTF for me, but may go towards the whole, "If 
you make the thing so damn big you can't do *basic* shit to keep it 
working, you may be building them too big." If I where to guess... They 
probably have "internal systems that funnel water to and from the 
reactor, then a "heat exchange" point, with and entirely separate 
system, which is external to the reactors themselves. This would then 
run out to their actual turbine systems, where the heated water produces 
the actual power, is cooled, then pumped back into the exchanger. You 
don't want to run radioactive water through the turbines, or use a 
system where water from one might get into the other. This means that, 
for practical purposes, power isn't generated "in" the reactor itself, 
at all. So...

b) not sure any of them have this. Its not enough to just drop in a few, 
in most cases, and.. well, if you can bend the damn things to shit in a 
system that runs them in slowly, you can imagine what sort of mess you 
end up with if some of them jam being "gravity dropped" into place. 
Again, the whole design seems a problem, and most of it due to scale, 
not just this one issue.

c) Hmm.. This is an interesting problem. How do you make a passive 
cooling system which doesn't move contaminated air, or water, or 
something else, and exchange it often/fast/effective enough to do that? 
We are talking about something heating to temperatures on massive 
scales, pretty damn fast, with a *major* exchange requirement. Again, 
size here matters. Think of it like trying to cool a high end CPU, using 
only a bit of foil taped to the top, because someone had to build the 
case so an actual radiator grill and electric fan wouldn't fit in there. 
You can get by with it if you are using low power, a lower end chip, 
etc., but... scale up to a PC, or, in the case of power plants, a full 
size nuclear reactor... and you "passive" systems are just not going to 
cut it in any practical sense.

So.. Again, seems to me, once you get past the basics of, "make sure the 
damn thing can run the active systems long enough to matter", everything 
else is, "scale, scale, scale", and the impracticality of doing jack to 
solve such problems at the scales being talked about. And, short of 
building the damn things in Antarctica, and somehow getting the power 
from there to every place else, I don't see "passive" being too viable...

Post a reply to this message