Japan Earthquake: Nuclear Plants at Fukushima Daiichi

[MadderDoc]

50000 ≈ 37^3 - a bit oversize I'd say but close enough. Hydrogen can still explode even when 50% by volume though.

Don't forget the explosion energy will heat & expand the gases significantly (well, pressurize them first, then expand them). Hydrogen burning in air is up to about 2300K ~ 8x volume at atmospheric, less the decrease from oxygen consumption, -> ~7x, by my rough calculation.

Fair enough, but I don't think we can assume the temperature of a dark non-glowing cloud to be 2300K or anywhere near it.

 

[MadderDoc]

In a funny way the video of the Unit 3 explosion indicates the possible behaviour of combustion gases from a hydrogen explosion by its kind exposure to our view of the fire phenomenon in the SE corner -- and the ensuing development of a knob of the condensed steam from that combustion, at the downwind side of the stem of the mushroom cloud.

Unfortunately Internet has been thoroughly cleaned of the best videos of the explosion, but if you can get your hands on one still, and it has more than the first dozen of seconds after the blast, this knob of steam from the explosion in the SE corner can be followed, as it travels downwind along with the mushroom. It stays low, and appears to have little tendency to rise, rather it just slowly grows and thins out by entraining air, and gradually disperses, much like the behaviour of the clouds we saw going with the wind from unit 1.

Image above is the last frame from this Unit 3 explosion animation

 

[r-j]

nuclear plant question

What is the source of the hydrogen that is believed to have caused the explosions?

 

[Most Curious]

What is the source of the hydrogen that is believed to have caused the explosions?

Short answer; Zircaloy cladding of the fuel rods in a steam or water atmosphere and very high temperatures releases a lot of hydrogen.

 

[Joffan]

Fair enough, but I don't think we can assume the temperature of a dark non-glowing cloud to be 2300K or anywhere near it.

No, well, that's full of dust and rubble too by the time we see it, don't forget. The 2300K burn temperature is (1) a maximum and (2) quickly diluted and reduced as the gas expands. It won't have transmitted more than a small fraction of that heat into the debris.

And anyone who knows more about the physics of explosions should jump in and correct me, incidentally, since I'm definitely no expert.

 

[zapperzero]

OK, thanks, I'll give it another go, at least now I know there's supposed to be a method.

I did not mean to be chiding you, sorry. It is simply that you know the size of the building, so you can get a fair estimate of how fast the ball grows because of also knowing at what intervals the frames are taken.

 

[zapperzero]

More excitingly, one could make a swift twist to the lid and jump back. Whoosh!

'Zackly. I don't know if you can get the same effect when you blow the roof off a depressurized equipment hall.

 

[zapperzero]

The initial 14400 cubic meters of hydrogen would have come mixed with some air, the nitrogen of which would add to the final volume of combustion products.

Dunno about that... I just stated the volume of steam at normal temp. That figure implies a rather low partial pressure (iow the air is pre-mixed :).

Also, initially there would have been, oh, however many meters of hydrogen-air mix the building can fit :).

 

[MadderDoc]

No, well, that's full of dust and rubble too by the time we see it, don't forget. The 2300K burn temperature is (1) a maximum and (2) quickly diluted and reduced as the gas expands. It won't have transmitted more than a small fraction of that heat into the debris.

Re - by the time we see it - 'it' being the mushroom cloud, when that would be that we see it. In the initial development of the vertically projected cloud, its upper edge would represent the front of ejected debris. The material to produce the mushroom cloud is of course there too. As the cloud progresses upwards, the debris slows down to eventually, except for the fine dust, start falling back to the ground. It is at this stage the mushroom can be most clearly seen, as it is emerging out of the top of the eruption cloud, steadfastly continuing its buoyancy driven travel upwards.

To illustrate, here's an animation of 17 frames, one for each of the first 17 seconds of the explosion. Images are heavily color enhanced to allow better distinction between the different cloud formations.

 

[SteveElbows]

I see the reactor 3 TIP room investigation was a bit of a failure, due to door etc debris.

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_120524_06-e.pdf

Given that a human was able to visually inspect one part of the room, I am less than impressed about a lack of photo of this area. It makes me curious about the nature of any debris inside the room.

 

[MadderDoc]

'Zackly. I don't know if you can get the same effect when you blow the roof off a depressurized equipment hall.

I think you are speaking in the context of a shortlived pressure drop below ambient, that would be present in connection with a detonation. But it takes time to nucleate water, that's why it works well to jump back from the radiator in the case of the car. I think to get some steam out of the effect in a pool in depressurised hall, you'd need more than a brief underpressure pulse, something like the permanent decrease of the pressure above the water in the car radiator. Also, you must lower the pressure, such that the new saturation temperature is lower than the temperature of the water. In the case of the pool we are probably looking at a body of water at about 50 deg C, so that is quite some pressure drop that must be maintained.

The theoretical mass of steam that can be produced by the depressuring can be fairly easily estimated. It is directly proportional to the difference between the initial temperature and the boiling temperature at the new lower pressure, and directly proportional to the amount of water present. The proportionality factor is about 0.002K^-1.

Mass(steam)=0.002*(Twater-Tsat)*Mass(water)

Example: You have a pressurised PCV filled with saturated steam and 4000 m3 of liquid water at 150 deg C. Swiftly release the containment lid and jump back a mile. Don't try this at home.

Mass(steam)=0.002*(150-100)*4000 = 400 tons

 

[MadderDoc]

I see the reactor 3 TIP room investigation was a bit of a failure, due to door etc debris.

http://www.tepco.co.jp/en/nu/fukushima-np/images/handouts_120524_06-e.pdf

Given that a human was able to visually inspect one part of the room, I am less than impressed about a lack of photo of this area. It makes me curious about the nature of any debris inside the room.

Weird to see that the door and apparently the door frame has not been blown away, but has been blown inward, into what they call the labyrinth. The implied pressure differential that should have existed to produce that effect is intriguing.

 

[zapperzero]

I think you are speaking in the context of a shortlived pressure drop below ambient, that would be present in connection with a detonation.

Yes.

But it takes time to nucleate water, that's why it works well to jump back from the radiator in the case of the car.

How much time? We have about a second, second and a half to work with, no?

I think to get some steam out of the effect in a pool in depressurised hall, you'd need more than a brief underpressure pulse, something like the permanent decrease of the pressure above the water in the car radiator. Also, you must lower the pressure, such that the new saturation temperature is lower than the temperature of the water. In the case of the pool we are probably looking at a body of water at about 50 deg C, so that is quite some pressure drop that must be maintained.

That's what I meant, when I said this should be experimented upon. But are you sure about the pool water temp? Could easily have been more.

The theoretical mass of steam that can be produced by the depressuring can be fairly easily estimated. It is directly proportional to the difference between the initial temperature and the boiling temperature at the new lower pressure, and directly proportional to the amount of water present. The proportionality factor is about 0.002K^-1.

Mass(steam)=0.002*(Twater-Tsat)*Mass(water)

Example: You have a pressurised PCV filled with saturated steam and 4000 m3 of liquid water at 150 deg C. Swiftly release the containment lid and jump back a mile. Don't try this at home.

Mass(steam)=0.002*(150-100)*4000 = 400 tons

I'm reasonably sure that the RPV didn't have that much water in it, in the event :). Anyways, http://www.spiraxsarco.com/esc/SS_Properties.aspxgives me a saturation temp of 60.something degrees celsius at 0.2 atm absolute. At 7.5 m^3/kg that's... a lot of steam, should the pool boil over. I am not sure how fast it nucleates, though. It should still be reasonably clean at this point, unless crud was thrown in by the earthquake.

 

[zapperzero]

Weird to see that the door and apparently the door frame has not been blown away, but has been blown inward, into what they call the labyrinth. The implied pressure differential that should have existed to produce that effect is intriguing.

The explosion could have done that, because the hydrogen-air mix occupies much more volume than the resulting steam. I can't think of anything else.

 

[MadderDoc]

Dunno about that... I just stated the volume of steam at normal temp. That figure implies a rather low partial pressure (iow the air is pre-mixed :).

Then I am not sure how you arrived at that figure, and how it landed in the right ball-park range for the volume of steam that can be produced from 1 ton of hydrogen. Are you a beginner? :-)

Also, initially there would have been, oh, however many meters of hydrogen-air mix the building can fit :).

Yes, I think we've arrived at something like 50000 cubic meters on that one. It could be less of course, but not significantly more. Once exploded, the combustion gases would come out as clouds from the building, and likely somewhat expanded in volume (due initially to elevated temperature, later due to air entraining). The hydrogen explosionists will necessarily have to claim responsibility for the clouds encircled here under, the mushroomists will have nothing to do with them. The 50000 cubic meters of hydrogen-air mix would seem to suffice to produce those clouds.

http://gyldengrisgaard.dk/fuku_docs/Unit3_hydrogenexplosioncloud.jpg

 

[MadderDoc]

The explosion could have done that, because the hydrogen-air mix occupies much more volume than the resulting steam. I can't think of anything else.

That would hold true even more for a steam explosion.

 

[MadderDoc]

How much time? We have about a second, second and a half to work with, no?

I am not sure. Nucleation is a function of diffusion rate hence a function of temperature. At relatively low temperatures it is not easy to make water flash to steam.

That's what I meant, when I said this should be experimented upon. But are you sure about the pool water temp? Could easily have been more.

I can't see how, we are only about 70 hours after cooling of the pool was halted. Mind Occam.

I'm reasonably sure that the RPV didn't have that much water in it, in the event :).

No, surely not. But it was just an example, and anyway it was about water in the primary containment, not the RPV. Under normal circumstances the primary containment holds about 3000 cubic meters of water. Not sure how much they lost by vents and how much that was countered by injection of external water. On the face of the pressure data, it might have lost only a few hundred cubic meters of its liquid water as steam in connection with the explosion in unit 3.

 

[zapperzero]

I can't see how, we are only about 70 hours after cooling of the pool was halted. Mind Occam.

Hm. Okay.

On the face of the pressure data, it lost only a few hundred cubic meters of its liquid water as steam in connection with the explosion in unit 3.

And where do you think that went? Up in a mushroom cloud?

 

[zapperzero]

Then I am not sure how you arrived at that figure, and how it landed in the right ball-park range for the volume of steam that can be produced from 1 ton of hydrogen.

PV=nRT. Checked with a steam calculator I found on the web.

Are you a beginner? :-)

In all things, I should hope.

 

[MadderDoc]

Hm. Okay.

And where do you think that went? Up in a mushroom cloud?

Let's see if there is room for more after the hydrogen explosionists have made their claims.

 

[r-j]

Is there some sort of mystery about what blew up building three?

 

[MadderDoc]

Is there some sort of mystery about what blew up building three?

:-) Now mystery is such a strong expression, r-j, but if the keyword is 'up' there may be a thing or two which are not yet fully understood: http://gyldengrisgaard.eu/fuku_docs/unit3cloud45/

 

[Yamanote]

In all things, I should hope.

That's the right way zz, but hard to go...

 

[MadderDoc]

<..>you know the size of the building, so you can get a fair estimate of how fast the ball grows because of also knowing at what intervals the frames are taken.

OK, so I made an overlay of frame 1 with the ball visible, and the next frame color inverted. I assume width of building profile (SE to NW corner) is 55 meter. Pixel distance there is 280 pixels, yielding a pixel resolution of 0.2 m. I assume the photo is not significantly out of aspect vertically/horizontally. The blue color in the overlay represents the apparent growth of the ball in frame 2 relative to frame 1. I measure this to be 30 pixels, or 6 m. Frame rate is 30 fps, so the apparent growth speed of the ball is 180 m/s from frame 1 to frame 2. Seeing limited precision, better to say 2E2 m/s.

In the overlay one can also measure the extension of the ball from the building in frame 1. This yields an apparent growth rate, relative to frame 0, of about 3E2 m/s which is about sonic. It could be this you think of as seeing the (just) supersonic speed of a combined flame/shock front of a detonation. Can you think of it in other ways, too?

 

[Rive]

:-) Now mystery is such a strong expression, r-j, but if the keyword is 'up' there may be a thing or two which are not yet fully understood: http://gyldengrisgaard.eu/fuku_docs/unit3cloud45/

As I recall, the previous 2-3 frames are also parts of the process. The explosion starts with the slight expansion of the (visible parts of) the building, the fire phenomenon comes later.

(When the video were first released I also made some pixel hunting on them (substracting the first frames from each other) but the result was not considered worthwhile or important to be posted. )

 

[zapperzero]

In the overlay one can also measure the extension of the ball from the building in frame 1. This yields an apparent growth rate, relative to frame 0, of about 3E2 m/s which is about sonic. It could be this you think of as seeing the (just) supersonic speed of a combined flame/shock front of a detonation. Can you think of it in other ways, too?

We are in the same ballpark, measurements-wise. There is one more measurement you can take - time and space from the leftmost edge of the blast (as seen in frame 0) to where/when the wall on the opposite side of the building bulges out, nearly reaching the apparent edge of reactor building 2 (frame 2, I believe).

 

[MadderDoc]

As I recall, the previous 2-3 frames are also parts of the process. The explosion starts with the slight expansion of the (visible parts of) the building, the fire phenomenon comes later.

Your reference is to an animation of 1 frame per second of thehttp://gyldengrisgaard.eu/fuku_docs/unit3cloud45/ , counting from the first visible fire phenomenon. Naturally seeing the source is 30 fps that leaves out 29/30th of the frames. The animation could have been made starting one frame earlier, when the first signs of something untoward happening with the building can be seen. But then the animation wouldn't have included any of the frames with the fire phenomenon.

Looking at the earliest initial frames, I don't think we can detect any expansion, as part of the first visible effect on the building. I'd say the strong impression is one of building collapse, i.e. the roof appears to come down, so that the building seems to be shrinking in the vertical direction.

 

[MadderDoc]

We are in the same ballpark, measurements-wise.

I assume we are also in the same ballpark, method-wise. The measurements are made from the thought that they represent the movements of a detonation shock/flame front. Can you think of them as perhaps representing something else?

There is one more measurement you can take - time and space from the leftmost edge of the blast (as seen in frame 0) to where/when the wall on the opposite side of the building bulges out, nearly reaching the apparent edge of reactor building 2 (frame 2, I believe).

That would imply measuring to ill-defined points in the shadow of the building and the developing plumes, it doesn't seem immediately promising to me. What would that measurement be good for?

 

[Rive]

Looking at the earliest initial frames, I don't think we can detect any expansion, as part of the first visible effect on the building. I'd say the strong impression is one of building collapse, i.e. the roof appears to come down, so that the building seems to be shrinking in the vertical direction.

- thanks for the link
- between frames -3 and -2 I think the shape of U3 starts to behave differently than U2's and U4's.

 

[MadderDoc]

- between frames -3 and -2 I think the shape of U3 starts to behave differently than U2's and U4's.

You too? I've had the same thought, Rive, but the signal is so weak, feels like dancing on the thin edge of the DL to base something on it. If other evidence likewise points to the possibility of damage preceding shortly the visible blast -. Hm. The peculiarities of the roof in the SE corner, a big flame in the middle of it, but how did it start, how did it progress.
http://www.gyldengrisgaard.eu/fuku_docs/20120311193727exp2.jpg