An Engineer's Explanation of the Fukushima Incident | Hacker News Reader

An Engineer's Explanation of the Fukushima Incident

docs.google.com

Probably the clearest technical explanation of what occurred at Fukushima after the recent earthquake and resulting tsunami.

Open articleView on HN

although not submitted separately, this was discussed a bit yesterday on HN, when plux posted it:

http://news.ycombinator.com/item?id=2411141

it looks like the attribution is different for each presentation, though the content is identical

The attribution here is correct (Matthias Braun, Areva); the name and logos were stripped when it was released, apparently for some kind of legal reason. This is discussed at the end of this article:

http://www.nytimes.com/2011/04/03/science/03meltdown.html?pa...

Beat me to it :-)

It appears the Stanford presentation was given by an Areva executive, and the branding / sources on this copy were removed:

"On March 21, Stanford University presented an invitation-only panel discussion on the Japanese crisis that featured Alan Hanson, an executive vice president of Areva NC, a unit of the company focused on the nuclear fuel cycle."

Source here: http://www.nytimes.com/2011/04/03/science/03meltdown.html?pa...

i'm wondering about the main design principle of the reactor cores - by default they are "hot", ie. producing heat/energy, so one needs to make effort to keep them "cold", ie. non-producing heat/energy. This hot-by-default principle was the main reason of Chernobyl catastrophe and here as well.

Why not reverse the principle and design the cold-by-default core? For example, instead of inserting graphite rods into the core to calm it down, one would need to insert additional uranium rods (or pump molten uranium salt, or move the 6-8-12 pieces of core close enough, etc...) for the reactor to activate and become "hot".

There's a difference between sub/super critical (producing energy via fission) and "hot" (producing energy _at all_). Chernobyl became more reactive (= more fission) as the water boiled away, leading to a runaway criticality, and explosive disassembly of the core.

Fukushima, however, has a negative void coefficient. Moreover, the fission reaction was stopped immediately after the earthquake successfully by the automatic insertion of control rods. The heat that is being generated is from short-lived fission products (ie, nuclear waste) undergoing spontaneous decay events, thus producing heat (but only a small fraction of the heat of an active reactor).

It is physically impossible to design a reactor in which such spontaneous decay can be stopped. You can only remove the energy actively until these short-lived products decay to the point where air cooling is sufficient to keep them at a safe temperature. You can, however, design a reactor in which the passive containment structures can withstand the temperatures of fresh waste material with no active cooling; however such technology was not available at the time the Fukushima plants were built.

Do you know if new plants are generally built with containments that work without cooling?

No. The new reactors under construction are mostly very conservative and traditional Gen III+ PWR designs, with some added safety features. The reactors that can survive complete loss of cooling are all Gen IV designs, that are currently under discussion, not construction.

>You can only remove the energy actively until these short-lived products decay to the point where air cooling is sufficient to keep them at a safe temperature.

this is my point - necessity for active removal of energy.

The closest thing to what i was talking about seems to be molten salt reactors where energy producing reaction goes only when salt is pumped through the core/moderator. If something goes wrong - the valves open and the salt is dumped [gravitationally] into the tanks which can be large enough to allow for passive air cooling of the secondary decay energy.

Control rod insertion is designed to be an entirely passive operation, effectively making the default state one where the fission reaction is not active. Your proposals are all different ways to stop the fission reaction but do not address residual heat.

What you're really asking for is a system wherein the cooling of the fuel after fission has stopped (this is where the problems occurred at the Fukushima plants) is entirely passive.

I'm sure somebody could get clever with siphons to make this happen but you'll need an absolutely massive amount of water to do so and/or a way to do the cooling such that the water itself is not exposed to radiation. This would allow you to, for example, safely take it from one side of the dam to the other.

An active solution instead allows recirculation of the same water (often borated) and can generally be a closed-loop (and radiation apathetic) system.

>Control rod insertion is designed to be an entirely passive operation, effectively making the default state one where the fission reaction is not active.

well, the rods stuck at 1/3 while being dropped into Chernobyl core as the things had already started going wrong.

Convection driven cooling is also possible (no need for siphons).

> What you're really asking for is a system wherein the cooling of the fuel after fission has stopped (this is where the problems occurred at the Fukushima plants) is entirely passive.

This can be achieved by designing the safe core temperatures to be much higher, which makes cooling easier. The very interesting small molten salt breeder design FUJI MSR http://en.wikipedia.org/wiki/Fuji_MSR is designed to survive complete external cooling loss by shutting down fission (either under control, or by the heat physically destroying the neutron mirrors required for it's operation), and then dealing with the decay heat by conductive transfer to the environment and simply letting the reactor heat up.

The fact that the core is molten in normal operation, under no pressure and entirely chemically inert even at elevated temperatures makes passive safety somewhat easier.

And that was a fun way to spend an evening: new reactor designs are quite cool :). It seems like a lot of the innovation is occurring in parallel with both fuel types and containment approaches.

I too look forward to having completely passive shutdown of reactors, but from what I can tell (outside of a few small-scale test reactors) we're just not there yet. Someday...

The problem here isn't criticality but waste heat: heat generated by uncontrolled radioactive decay of waste products (fission products and products of neutron activation). There's no way to turn off this heat; it continues uncontrollably for a long time, decaying super-exponentially (here's a graph: [1]).

As far as the nuclear reaction goes, the Fukushima reactors were "off" within seconds of the initial earthquake, and remained so since then. (Though, tangentially, there's been speculation of some low-power criticality events since then, but the evidence is flimsy. It's not particularly important anyway). That's the polar opposite of the Chernobyl disaster, where it was an uncontrolled nuclear reaction which destroyed the reactor building and released vast amounts of nuclear waste in an explosion (and subsequent fire). In Fukushima (and Three Mile Island, and others) there was no uncontrolled reaction; but there was decay heat and its effects. The big explosions at Fukushima were an indirect effect of decay heat: explosions of hydrogen generated by when extremely hot zirconium "burned" in steam. And the large scale fuel failure ("meltdown") was also caused by decay heat -- too much decay heat, and not enough cooling.

[1] http://en.wikipedia.org/wiki/Decay_heat#Power_reactors_in_sh...

It seems that it might be a good idea to design future nuclear reactors with pneumatic-powered robots in mind; purpose-built shafts and ramps could allow robots with minimal electronics into areas unsafe for humans. I say pneumatic instead of hydraulic, as based on my very limited understanding of nuclear physics, lightweight gases are less likely to become neutron activated for extended periods of time.

I believe radiation messes with machines a bit too, though.

I could see neutrons activating the parts of a robot, but what physical mechanism would prevent a pneumatic motor from driving some wheels or moving an arm in the presence of radiation, other than the parts heating up and expanding?

A robot that can't see isn't very useful....

One problem I've read in these sorts of situations is the robot eventually or unexpectedly gets killed by the ambient radiation ... and now you have to get humans to dash in there to pull it out of the way.

Getting robots in wouldn't help with the Fukushima incident much; the real problem there is getting energy out, not getting robots in.

They could have been used for monitoring inaccessible locations earlier in the disaster, possibly by using fiber optic cameras and lighting.

I am certainty no expert on nuclear engineering or containment design but it seems odd to me that the spent fuel rods are stored in above ground pools. I wouldn't even let my kids play in one much less store radioactive materiel that requires active cooling in one.

I feel bad for your kids

I've read It's an issue with manipulating the fuel: to do so, you unbolt the containment cap on the top of the reactor pressure vessel and fill the entire path from on top of it to the initial spent fuel pool with water. That allows you to safely transfer hot (thermally as well as radiatively) rods and it's pretty hard to get around if you're using this sort of reactor geometry (I understand it holds for both PWRs as well as these BWRs).

All right. The Google URL's are different, so I'll assume this isn't just another copy/pasta karma endeavor.

http://news.ycombinator.com/item?id=2400790

I don't particularly care about my own karma (or lack thereof; and after all I was reformatting someone else's URL, albeit to avoid the PowerPoint dependency (and risk)), but I do care about the contribution of repetition to the increasing volume of noise on HN.

That aside, it is an interesting slide deck.

Reply on news.ycombinator.com