Water and Uranium-235 Reactivity in W88 Secondary: Safety Concerns Explained

[lowellite]

(I apologize in advance if this is in the wrong section of the forum)

I found an article online describing the safety concerns of the W88 thermonuclear warhead. Apparently, water can turn the uranium-235 coating of the W88's secondary into a "runaway boiling hot water reactor" (not sure what that means...), causing the secondary to go critical. Since my knowledge of nuclear physics is practically nonexistent, I can't really understand what the article is talking about. Can someone explain to me why/if this process would occur?

From http://www.fas.org/sgp/eprint/morland.html

This use of uranium-235 in the outer shell, or pusher, of the W-88 secondary raised a curious environmental problem that was first noted with the Hiroshima bomb, Little Boy, in 1945. If uranium-235 is used as reactor fuel, the moderator can be ordinary salt or fresh water (not heavy water, or graphite as required with uranium-238), and the reactor can be quite small. Most research reactors use highly enriched fuel for this reason, including the one on the campus of the Massachusetts Institute of Technology in the heart of greater Boston, but that's another story.

Uranium-235 plus water equals a nuclear reactor, especially if the geometry of the metal allows the water to get between metal surfaces where neutrons can be moderated, or slowed down, as they travel from one surface to another. In theory, the hollow uranium-235 projectile of Little Boy could become a runaway boiling water reactor if it fell into the ocean. (Until John Coster-Mullen's book Atomic Bombs was self-published in 2002, all descriptions and graphic depictions of Little Boy had the large hollow uranium piece as the stationary target and a smaller cylindrical piece as the projectile. Coster-Mullen's contrary account comes from years of interviewing men of the 509th Composite Group who assembled and dropped the bomb.)

The problem with Little Boy was ignored. However, by the time of the first Polaris Submarine, the standard primary design employed a hollow plutonium shell which could theoretically fill with water and become a reactor. Because the Polaris W-47 warhead was deployed at sea, a test was done on May 11, 1960, code-named the Stardust Program. A hollow plutonium pit was filled with water in an underground tunnel at the Nevada Test Site. The pit did not go critical.

But the secondary is much more massive than the primary. When the decision was made with the W-88 to replace the uranium-238 pusher shell in the secondary with uranium-235, the boiling water reactor problem was undeniable. If the submarine sank, or a warhead was damaged and dropped into the ocean, the secondary would go critical as soon as water got inside it. Considering the damage a nuclear war is designed to cause, such safety considerations are of course trivial, but the problem was studied, anyway.

 

[theCandyman]

A nuclear weapon's core is similar to the fuel in a nuclear reactor in that it will fission and produce energy, the difference being the rate at which this occurs. If this core gets dropped into the ocean as suggested, the moderation of the water will slow down neutrons and allow the core to fission at a slower rate than if the bomb were detonated. As the core heats up, it will heat up the water around it - like a mini-reactor.

All this is based on guesses of the bomb structure and hypothetical scenarios though. You could do a similar check yourself and see how 90% or so enriched uranium would act in water (e.g. find how critical a homogeneous mix of U-235 and water is).

 

[lowellite]

Do you know where these neutrons come from? I thought the neutrons come from the fission reaction of the primary, but dropping the warhead in water wouldn't set off the primary, right?

 

[Bob S]

Do you know where these neutrons come from? I thought the neutrons come from the fission reaction of the primary, but dropping the warhead in water wouldn't set off the primary, right?

Read about the nuclear reactor at the Oklo site in Gabon, Africa, when it was flooded with water. See
http://www.ocrwm.doe.gov/factsheets/doeymp0010.shtml
Bob S

 

[theCandyman]

There is a neutron source in a warhead to get it started, I think it is Beryllium. Also, spontaneous fission and (alpha, n) will contribute to neutron sources as well.

 

[Bob S]

There is a neutron source in a warhead to get it started, I think it is Beryllium. Also, spontaneous fission and (alpha, n) will contribute to neutron sources as well.

The PuBe (plutonium-beryllium) source neutron yield is given in this abstract:
http://prola.aps.org/abstract/PR/v98/i3/p740_1
Bob S

 

[Hologram0110]

Some nuclear warheads us an (alpha,n) reaction (commonly a mixture of Pu and beryllium) to supply the initial neutrons. However, these two materials are separated (alphas are easy to block) until the core is imploded.

There are other sources of neutrons such as spontaneous decay of heavier Pu isotopes (greater than Pu-239) which are found in significant quantities which could provide starting neutrons, although this would significantly reduce yield.

I can see three ways that it could go:
1) The core remains sub critical (depending on geometry, pressure, temperature, purity)
2) The core becomes a self regulating reaction. In this case the core would get hot, expand and boil off some water, maybe deform. This could make the core switch between sub and super critical without ever resulting in an explosion. The core would just slowly release its energy
3) The core becomes super critical and fissile out. Even if it went super critical, it wouldn't result in a big explosion because it wouldn't hold itself together long enough to generate that much energy. It would make a small boom and blow itself apart (or melt) without releasing most of its energy.

 

[lowellite]

Thanks so much for the explanation, everyone.

So basically, the water is slowing down the fast neutrons released by the radioactive decay of uranium-235, thus exponentially increasing the speed of the fission reaction? (And beryllium can also supply neutrons when hit by alpha particles released from radioactive decay?)

Also, can someone tell me how spontaneous fission occurs? Is it the same thing as radioactive decay?

 

[theCandyman]

It is only my guess where the neutrons come from, but those cover all the internal sources it could be. The rate of fission will only increase exponentially if the material and geometry allow it. As for (alpha, n), the energy of the neutron will depend on the alpha particle, which comes from alpha decay of Uranium. That article Bob linked will explain more.
Yes, spontaneous fission is one route for decay. I don't think it is very common, except for one isotope of Plutonium.

 

[Hologram0110]

Spontaneous fission is basically a type of radioactive decay. The isotope just breaks apart ejecting a neutron.

Water moderates (slows down fast neutrons) so that more of them get absorbed in the U-235. More neutrons then cause fission of U235 which results in more neutrons. This is the basic premiss of 'thermal' reactors (reactors which slow down fast neutrons). Without the water, the neutrons would be moving too fast and most of them would not get absorbed and simply leak out of the u235.

While there are many sources of neutrons (even radiation from space, spontaneous fission, neutron sources like PuBe ect) the number of neutrons provided by these sources is relatively low but ensure that there is always some neutrons in the core. When the core goes super critical, these small number of neutrons start the chain reaction that produces many many more neutrons.

When a nuclear weapon is trigger the weapon becomes highly super critical and is held together by the force of the implosion long enough to release lots of energy before blowing itself apart. Since the core in this case doesn't have that, I suspect that it will fissile out without a large explosive yield. It could still release a lot of radiation and radioactive waste however.

 

[Astronuc]

With respect to criticality, the term for a nuclear detonation is 'prompt supercritical'.

 

[vanesch]

Normally you cannot have something like a "nuclear explosion" with moderated thermal neutrons. The reason is that the inter-generation time is too long.

You need something of the order of a hundred "neutron generations" to reach "nuclear explosion" energy. Under a "neutron generation" we understand the succession of a neutron that is liberated from a (neutron-induced) fission reaction until it gets absorbed itself in U-235, causing a (next-generation) fission reaction. At each generation, the number of neutrons is multiplied with a number k, called the Fermi k-factor. If k > 1, this means that each generation has more neutrons in it than the preceding one, and the system is "critical". If k < 1, then the number of neutrons dies out, and we have a sub-critical system.
In order to reach, starting from a handful neutrons, to "nuclear bomb" energy (order of kilotons), you grossly have:

bomb energy = k ^ (number of generations) x (few initial neutrons) x 160 MeV
(160 MeV is about the energy liberated by a single fission).

During this phase, the system dissipates more and more energy, and in fact, from a certain point on, it will destroy itself. If things go "slowly" then the energy needed to destroy the bomb will be way lower than a "nuclear bomb" energy. A small explosive will be sufficient to blow the pieces apart. It is only if the chain reaction goes very fast, that enough energy is released to reach "bomb energy" before the thing flies apart. The brake is essentially inertia.

It turns out that you need something like 10^25 or so for k ^ (number of generations) in order to start being able to speak about a nuclear explosion. Now, it is difficult to get k larger than 2, and that means that you need something of the order of 100 generations to reach 10^25.

The inertia that can hold your bomb together only works on the few (tens of) microsecond level.

This means that the inter-generation time of neutrons in a bomb must be smaller than something of the order of 100 nanoseconds.

Well, in a thermal reactor, the moderation process already takes several microseconds to tens of microseconds. You only have time for a few generations. That's by far not enough to have a nuclear explosion worthy of the name.

So you can't make a nuke with thermal neutrons. The inter-generation time is too long.
(this is btw also why a thermal reactor can never turn into a nuclear bomb).

You CAN release a lot of energy, but it will be way way way below what a bomb is supposed to release. If the energy is not evacuated, it will destroy the structure of the bomb, but it won't cause a nuclear explosion.

Note: the numbers above are very rough, I typed this from memory. Can be a factor 10 wrong on some of them.

 

[lowellite]

Thanks for so patiently explaining this to me. It's a great help, but sometimes I feel like the more information I get, the more questions I have! ;)

So if fissile materials tend to blow themselves apart if they undergo fission for too long, how are nuclear bombs engineered to release all of the energy at once? I've read that the fission reaction is catalyzed in the primary (essentially a fission bomb) by detonating the explosives placed around it, but how does that prevent the fissile material from blowing itself apart? Why doesn't the fissile material (or the whole bomb) just blow apart to pieces when the explosives go off?

And the next part is also a bit confusing. So once the primary starts its fission reaction, the energy is transferred to the secondary (lithuim-6 deuteride surrounded by uranium 235, Wikipedia has a picture http://en.wikipedia.org/wiki/File:W-88_warhead_detail.png). The energy will start a fusion reaction in the secondary, using lithium-6 deuteride as fuel. According to http://en.wikipedia.org/wiki/Nuclear_fusion#Inertial_confinement , fusion is made possible when the fusion fuel is "imploded" by X-rays from the fission reaction of the secondary. How is this possible? I thought an X-ray is just high frequency wave of light...

 

[Astronuc]

So if fissile materials tend to blow themselves apart if they undergo fission for too long, how are nuclear bombs engineered to release all of the energy at once? I've read that the fission reaction is catalyzed in the primary (essentially a fission bomb) by detonating the explosives placed around it, but how does that prevent the fissile material from blowing itself apart? Why doesn't the fissile material (or the whole bomb) just blow apart to pieces when the explosives go off?

Most weapons are an implosion type. Explosive compress a sphere of fissile material long enough to allow a sufficient number of generations of 'prompt' neutrons to be generated in a few microseconds. The fissile material does blow itself apart - which is the nuclear detonation.

And the next part is also a bit confusing. So once the primary starts its fission reaction, the energy is transferred to the secondary (lithuim-6 deuteride surrounded by uranium 235, Wikipedia has a picture http://en.wikipedia.org/wiki/File:W-88_warhead_detail.png). The energy will start a fusion reaction in the secondary, using lithium-6 deuteride as fuel. According to http://en.wikipedia.org/wiki/Nuclear_fusion#Inertial_confinement , fusion is made possible when the fusion fuel is "imploded" by X-rays from the fission reaction of the secondary. How is this possible? I thought an X-ray is just high frequency wave of light...

The nuclear detonation produces an intense burst of gamma (from the nuclei) and X-rays (from atomic electrons). The X-rays are focussed by a special lens and a directed to a volume of material such as LiD which is heated to temperatures sufficient for fusion. If the volume of fusile material is surrounded by a fissile material, then the primary fissile detonator would cause that fissile material to fission, which has a great effect than a chemical explosive, and the fusile material is compressed (imploded) and heated by gamma/X-rays such that fusion occurs.

The process must occur in a matter of microseconds before the system is disrupted.

 

[vanesch]

fusion is made possible when the fusion fuel is "imploded" by X-rays from the fission reaction of the secondary. How is this possible? I thought an X-ray is just high frequency wave of light...

Yes, this was in fact the strange property that made thermonuclear weapons possible: the way to "ignite" the fusion was not, as one would have thought, some mechanical/fluid shock wave (which turns out to be too slow), but rather radiation pressure. It is the radiation pressure from X-rays that makes the ignition possible and fast enough. In fact, this discovery was Ulam's, some say, against Teller's opinion, after which Teller claimed credit (and others say the opposite) but nevertheless it is called http://en.wikipedia.org/wiki/Teller%E2%80%93Ulam_design" .

 

[theCandyman]

To give you something to thing about, light has momentum equal to hv (Plank's constant times the frequency of the light). Many of the X-rays might escape, but there's still a lot of them in a small space - those will contribute to confinement/heating of fusion material.