(Hat tip: Brown and Caldwell California Water News)
Iran has unveiled a heavy water plant, capable of producing 8 tons of heavy water each year. In five months, it's expected to double its output.
Heavy water is water with heavy hydrogen (deuterium) instead of normal run-of-the-mill hydrogen. Light water is H2O; heavy water is light water enriched with additional DHO (one hydrogen atom, one deuterium atom) and D2O.
Deuterium weighs very close to twice what hydrogen weighs, and each deuterium atom in an average molecule of heavy water increases its mass by about one atomic mass unit. (Up to a maximum of two, if every molecule contains two deuterium atoms to one oxygen atom.)
A small fraction of all water has at least one deuterium atom in place of one of its hydrogen atoms. These can be concentrated by various methods. Indeed, certain bodies of water, like the Dead Sea, have naturally high concentrations of heavy water, because heavy water evaporates a little more slowly than light water does. Over the centuries, the heavy water has noplace else to go, so it piles up. (Actually, it reaches an equilibrium concentration because it does evaporate.)
The increase of mass doesn't seem like a whole lot. DHO has an atomic mass of 19 5.5% heavier than H2O's atomic mass of 18. D2O is 11% heavier than H2O.
5.5% and 11% heavier doesn't seem like that big a deal, except that increase in weight happens to be right where all the action is.
In a nuclear reactor, you have two essential components: nuclear fuel, and a moderator. Other components that are extremely nice include control rods and a cooling system.
In order to get a nuclear reactor going, you need some of the atoms in your fuel to split and emit neutrons, and you need enough of those neutrons to be captured by other atoms of fuel, causing further splitting and generating further neutrons, to keep the reaction going.
That brings up the notion of the "interaction cross section". Neutrons moving through fuel are like bullets flying through a collection of targets. If you blindfold a marksman (marks-person?) and let him shoot in random directions in a room full of targets, there's a small chance he'll hit a target. If the targets are made bigger, the odds go up. Being bigger, each target has a larger interaction cross section.
The thing about nuclear reactions is the interaction cross section varies with the speed of the oncoming particle. It's as if the marksman's targets grew or shrank depending on the muzzle velocity of his gun.
Natural uranium is mostly U-238 with a small fraction of U-235. U-235 has a very large cross section for slow neutrons, and it can be burned in a light water reactor very easily. In a light water reactor, the fuel is surrounded by light water. Neutrons speeding out into the water are pretty likely to run into a molecule of the water. If they hit a hydrogen atom, they will be slowed down quite a lot, and may even come to a screeching halt. Since neutrons are only slightly heavier than the nucleus of hydrogen atoms, they bounce off each other like pool balls on a pool table. If one ball hits another dead on, the first ball comes to a stop and the other ball goes speeding off with all the energy. The water acts as a "moderator", slowing fast neutrons down so they can be absorbed by the U-235 and cause them to split.
U-238 has a relatively small cross section for slow neutrons. Its cross section is much higher for fast neutrons. Therefore, you don't want to slow neutrons down if you want to burn U-238. You want some material that will reflect neutrons back into the fuel without slowing them down.
Going back to the analogy of balls crashing into each other, let's look at what happens when the masses are unequal. Suppose we replace our cue ball with one that's half the mass. It turns out that even if we hit another ball dead on, we can never get it to come to a dead stop. It will transfer, at most, two thirds of its momentum to the other ball, and thus keep a lot of its own speed. The result is, neutrons speeding through heavy water take a longer time to slow down, and thus spend more time at a speed where they are likely to hit the U-238 targets.
Heavy water thus makes it possible to burn natural uranium in a reactor without going through the work of enriching it. (And enriching water to make heavy water is much safer to deal with than enriching uranium. Uranium is radioactive, and in order to process it, you have to react it with some very reactive chemicals.
The main problem with an Iranian heavy water program is that a nuclear reactor that burns natural uranium can also produce other byproducts. Not all neutrons will cause uranium atoms to split. Some of them will be absorbed by uranium atoms without causing a fission. If the resulting U-239 declines to split, it will decay (with a half-life of less than 24 minutes) into Neptunium 239. Neptunium then decays (with a half-life of some two and a half days) into Plutonium 239.
Pu-239 has a half-life of 24,390 years, so it will build up in a reactor for as long as you care to run it. It will capture neutrons and split very readily, so you have to balance various factors to keep the stuff around, but it's certainly doable.
In order to get large quantities of "weapons grade" uranium, you need to do a lot of work to separate isotopes. In order to get "weapons grade" plutonium out of used nuclear fuel, you only need to do a chemical separation. That's a lot easier.
That's why people are worried about Iranian heavy water production.
Now, the Iranian government says it will use its nuclear technology for peaceful ends. Given their track record, I think it's reasonable to have doubts about that.