I want to discuss a few other reactor designs besides the typical US power reactors, the PWR and BWR types. Before I do that, it's important to understand a few things that force us down certain paths in designing reactor cores.
We need to understand the relative abundances of U-235 and U-238, why we use neutrons to split atoms, and also how a moderator works, in a physical sense.
First off, about Uranium. According to Wikipedia, naturally occuring Uranium consists of three major isotopes: U-239 (99.28%), U-235 (0.71%), and U-234 (0.0054%). The stuff we need for fission (U-235) is less than 1% of naturally occuring Uranium.
There is an entire industry based around the enrichment of Uranium. The process of enrichment is as fascinating as it is tedious. I might do a post on that process at some point, but for now it will suffice to point out that it is very, very difficult to create fission with naturally occuring Uranium. For this reason, in most reactors, the concentration of U-235 vs U-238 is increased, or enriched.
And we always accomplish large numbers of fissions by using neutrons. Why neutrons? A couple of reasons. Since neutrons have no charge, they can wander at will through any part of an atom, and through many materials other than Uranium, as if that material weren't there. Also, since fission produces more neutrons than it uses, they are ideal for initiating more fissions!
When a fission occurs, 2-3 neutrons are ejected, but they are moving at 2MeV, which is about 20% of light speed. At this speed, they CAN interact with another Uranium nucleus and cause a fission, but it is unlikely. In order to improve the likelihood of a fission, there are two things we can do: Increase the number of U-235 atoms in the core, or moderate (slow down) the neutrons.
In practice, both methods are usually used. Reactors using no moderator, and containing a core of highly enriched U-235 have been built and operated. Since there is no moderator slowing the neutrons down, these type of reactors are called "Fast Neutron Reactors".
The vast majority of reactors are "Thermal Neutron Reactors". The reasons most reactors use thermal neutrons are two-fold. Fast Reactors require fuel enriched above 20%, which can potentially be diverted and used for a nuclear weapon. Additionally Fast Reactors have poor safety characteristics from a reactor physics standpoint, and thus very poor safety record. Of the handful of Fast Reactors built, several have suffered meltdowns.
More on moderators though. The purpose is to slow neutrons down to the point where they are at thermal equillibrium with the surrounding material. The way to accomplish this is by allowing them to impact (or "scatter") against atoms with low mass. Think of the neutron as a ping-pong ball fired from a cannon. To remove the most energy by collision, ideally it should run into another ping-pong ball, which will recoil and remove some speed. If it collides with a heavy nucleus, like steel, it would be like our ping pong ball bounced off a boulder. It would not lose as much energy to the boulder.
Water is a decent, but not great, moderator. It has two hydrogen atoms for neutrons to scatter against, and it is dense. The oxygen molecule has the ability to scatter a neutron, but it due to its mass (16) it would take many more oxygen atom collisions to accomplish the task. Water is also plentiful and cheap. This is why in most applications, water is used as a moderator.
Water though, is not an ideal neutron moderator, and here is why: Both the hydrogen and oxygen atoms have a small, but noticable possibility of absorbing a neutron, thus removing it from the process and making it unavailable to cause a fission.
There are other moderators that are superior to light water from the standpoint of neutron absorption. Even though a Fast neutron may require a larger number of collisions to reach thermal equillibrium, the lower likelihood of being absorbed make other materials better moderators.
Moderator Number of collisions Likelihood of neutron absorption
Hydrogen 18 0.3326
Dueterium 25 0.000519
Beryllium 86 0.0076
Carbon 114 0.0035
Oxygen 150 0.00019
Uranium 2152 7.57
Dueterium (in the form of heavy water) is probably the best moderator known. This is water with hydrogen atoms that already have one neutron and one proton. Normal hydrogen atoms have just one proton, and occasionally capture one, taking it out of use for the reactor. Dueterium hydrogen atoms have already absorbed a neutron, and are unlikely to absorb a second one.
Dueterium is found in nature, but like U-235, is not abundant. It also needs to be enriched at great difficulty and expense, to be available in sufficient quantities to be used as a moderator.
Beryllium is another excellent moderator, with very little likelihood of absorbing a neutron. Its disadvantages are expense and high toxicity.
Carbon is an excellent moderator, evenwith the large number of collisions required, its low neutron absorption, and low cost make it a practical moderator, even if it is not a good material to remove heat.
The other two items on the list, Oxygen and Uranium are not moderators at all. They just show how increasing the mass of the target atom increases the number of collisions required to moderate a neutron. However on the right hand column, notice the likelihood of absorption for Naturally Occuring Uranium. It does love to vacuum up neutrons :)
In the next post, using what we have learned here, I will describe a few unusual reactors that have been built, either for testing purposes, or for breeding additional nuclear fuel, for nuclear weapons production, and how a nuclear weapon itself works.