#3 is the one with the MOX fuel rods. They contain Plutonium and other nasty stuff.
But don’t worry. Haven’t you heard? Radiation is good for you.
MOX fuel is a bogeyman. First of all, it is only 7% Pu when fabricated, and almost all of that is fissioned in the course of one fuel cycle, so you end up with about as much Pu as you would with ordinary low-enrichment uranium fuel, which breeds Pu as a natural consequence of neutron capture in 238U.
Second, when analyzing a reactor accident, of primary importance is what is called the source term. That is what contains information on what is in the material released, and how much. Guess what? The source terms for MOX and ordinary uranium fuel are about the same. Why? Because in a LWR accident involving release of fission products, fuel isn't important, fission product spectrum is. And you get essentially the same fission products in the same amounts: noble gases like 133Xe and 85Kr, and volatile, reactive forms like 131I, 90Sr, and 137Cs. Very, very little of the fuel itself is contained in the source term? Why? Because in a LWR accident, the fuel particles generally lack the stored energy necessary to drive them out of the fuel pellets (even if damaged), then drive them out of the cladding (even if damaged), then drive them out of the pressure vessel (unlikely to be damaged), then drive them out of the containment (unlikely to be damaged), and out into the environment. And if they happen to get out, they aren't terribly mobile, at least compared to things like cesium and iodine, which you are going to get whether you use MOX or not.
So, please, lets save the bogeyman for the kiddie stories.
Fission product yield from MOX (which include some fissionable Plutonium) is virtually indistinguishable from “regular” reactor fuel of ~5% U-235 enrichment.
The fission products that are measured in the environment “get out” based on their physical characteristics - gases and iodine highly mobile, and to a much lesser extent, particulates like Cs or Sr/Y or Pm. The Uranium Oxide or Plutonium oxide in fuel is in ceramic form, and is not mobile – it may turn to rubble as fuel cladding “melts” (proved at TMI-2) but is NOT mobile and is NOT going to be released from the plant systems.
Bottom line, don’t buy into ANY hysteria about MOX. NO sample taken in the environment will detect Uranium or Plutonium. You don’t have to believe “radiation is good” to understand how these materials behave physically.