It's been a while since I did a nerdy nuke post. I've had a subject in mind, and been thinking a bit about how to describe some of that nerdy stuff in an understandable way... I guess it's time to give that a shot.
Today's post is about delayed neutrons, and how they allow a reactor to be operated without power taking off exponentially within microseconds. I brushed over delayed neutrons in another post, but this will be just a little more in depth discussion of the how and why.
As I mentioned in another post, the Strong Nuclear Force holds the nucleus (neutrons and protons) of an atom together. This force is strong enough to even overcome the powerful electrostatic repulsion of positively charged protons at very close range, and holds them together in spite of their desire to fly apart. Think of this force as the kind of force that would be able to hold two powerful magnets together, even with the poles aligned North-North.
As an aside, there are four fundamental forces in the universe that drive everything in and around us:
The Strong Nuclear Force keeps the atoms of our bodies (and everything else in the universe) from falling apart into simple protons and neutrons. Happily, the Strong Force is a very short-range force, about 2.5 times the diameter of a proton, so it also doesn't pull all the atoms of the universe together into a single massive black hole. It's interesting how balanced the universe is with these repulsive and attractive forces...
Back to the Strong Force though. There's a lot of complex stuff going on in the nucleus of an atom, and I don't pretend to understand it all. I'm a nerd, but too much of a slacker to follow the calculus.
What I do understand is that when you have a lot of positively charged protons repelling one another in a nucleus, it takes more and more neutrons in that nucleus to hold it together. Neutrons help bind the nucleus together, because they attract via the strong force, but don't repel each other like the protons do. As a result, the neutron-proton ratio changes as we work our way up the periodic table.
Below, the red line indicates a 1:1 ratio of neutrons and protons. The blue dots are a plot of actual stable elements. As you can see, more and more neutrons are required to bind the nucleus together as the number of repelling protons increases.
The "Band of Stability" in the above graph indicates elements that are stable, i.e. not radioactive.
At this point let's talk about an old friend, Uranium 235. Uranium 235 has 92 protons and 143 neutrons in the nucleus, for a proton-neutron ratio of 1 to 1.55. Uranium 235 would not be on the above graph because it is not stable. It is *almost* stable, however, with a half life of 704 million years. That means that after 704 million years, half of it would have decayed (eventually to Lead).
When an additional neutron is added to a U-235 nucleus, it (very briefly) becomes U-236, but fissions in 10^-14 seconds. that's 0.000000000000001 seconds. As the U-235 fissions, it also emits 2.43 neutrons on average. These can go on to cause more fissions, and within .001 seconds, your entire reactor is converted into an expanding cloud of glowing plasma, before the control rod can even move to slow down the reaction. That's because the very brief time scale of each generation of neutrons goes by so quickly.
Except that's *not* what happens in a nuclear reactor. That's what happens in a nuclear weapon. Here's why there is a difference: Delayed neutrons. The neutrons I was discussing earlier are called "prompt neutrons". Those are the extremely high energy neutrons ejected from the fissioning nucleus within 1 x 10^-13 seconds of the fission, ones that were bound by the strong nuclear force, but are now free to roam about the reactor core.
The fission fragments - those two new lower-mass nuclei that result from the split Uranium nucleus, suddenly have an excess number of neutrons for the now lower number of protons. The proton/neutron ratio is not correct for these new nuclei (too many neutrons), and they are unstable (radioactive). Because they are neutron-rich, they typically decay by converting a neutron to a proton, and ejecting a beta ray (electron), so that the electrical charge is conserved. They also eject an anti-neutrino to conserve angular momentum, or "spin". This type of decay usually happens several times before the neutron-rich daughter nuclei are down to a stable (non-radioactive) state.
HOWEVER, six of these neutron-rich fission products decay by simply ejecting the excess neutrons. They are known as "delayed neutron precursors". The half-life of these fission products range from 0.23 seconds to 57 seconds. These "delayed" neutrons are what allow us to have fine control over reactor power. Only 0.0064% of the neutrons in a U-235 reactor are delayed neutrons, but this is enough for control.
The delayed neutrons, being generated by radioactive decay, rather than directly from fission, have considerably less energy (speed) than prompt neutrons. Because of this lower energy, they require fewer collisions to thermalize, and are less likely to escape the reactor than a prompt neutron. So the percentage that survive to be thermalized and create a new generation is greater for delayed neutrons than for prompt neutrons. This is important, because it means the population of thermal neutrons is quite a bit higher than the miniscule 0.0064% neutrons that are born delayed.
When a reactor is "critical", that means it has a constant neutron population. The chain reaction is creating new neutrons just as fast as they leak out of the core or are lost to absorption in non-fuel materials. The reactor is critical due to the combination of prompt + delayed neutrons. The key is to always keep it that way. The delayed neutrons act to slow down the rate at which the reactor changes neutron population (power level).
One of the many keys to reactor safety is to always stay below the point where the reactor could be critical on prompt neutrons alone. That condition is called "prompt critical", and once that happens, nothing mechanical can respond quickly enough to prevent a runaway nuclear reaction. You really do get the runaway reaction scenario I first described above in a prompt critical situation.
There have been a few prompt criticality accidents. Chernobyl is a notable accident. SL-1 is a less well-known, but still lethal accident. Soviet submarine K-431 is another prompt criticality accident that was shrouded in the secrecy of the Cold War. Numerous prompt criticalities have occurred in nuclear fuel and weapons manufacturing settings. Plutonium 239 fissions produce fewer delayed neutrons than Uranium 235, making reactor control more challenging. The same situation applies to fast reactors, as the fraction of delayed neutrons is smaller than in a thermal reactor.
Unless a reactor is designed to withstand prompt criticality, it's best to avoid the situation :)