Fun With PUREX

Originally wrote this 16 years ago as I was getting my masters degree. As my adviser often asked “Don’t you want to learn normal things?”

While PUREX may save us from being buried in nuclear waste it is not without peril, though it’s more of the did-you-actually-think-this-through-before-you-started-pressing-buttons variety of peril.  It is vitally important to take a moment before you start playing with big machinery full of radioactive materials in the middle of simultaneous inorganic & organic chemistry.

For the record, I have personally performed the uranium half of the PUREX process…in a 20mL test tube.  It was damn impressive to observe this happening in my hands.

So, how it works:

PUREX is an acronym that stands for “Plutonium and Uranium Removal by EXtraction”. First, you bring the uranium up into solution as uranyl nitrate in concentrated nitric acid.  “Concentrated Nitric Acid” is one of those chemical phrases that lets you know that you are about to do something dangerous, just like Darkwing Duck.

Next, mix up the tributyl phosphate (TBP) in kerosene. Hearing that words phosphate and kerosene together make me start to think of Oklahoma City. Add the TBP/kerosene mixture to the uranyl nitrate aqueous solution. Now it really sounds like a truck bomb. However, the organic and aqueous phases are immiscible, with the kerosene floating on top.

What do you do next? That’s right, you shake the thing that sounds like an explosive VIGOROUSLY.

As you do this, greatly increasing the reacting surface area as you mix the bejeezus out of it, some of the uranium will slip out of the nitrate and pop into the middle of the TBP’s handy ring. You then decant the organic layer, add a fresh one, and repeat. Eventually, the vast majority of the uranium will be sucked out of the aqueous phase.

And that’s how PUREX works. You change the organic compounds slightly to grab the specific actinide element you want, but it’s more or less the same. What you do next with the organic phase depends on what you’re after.

Now, this is when it works correctly.


In the course of my research into radiation accidents, I came across one category called “Red Oil Fires” which had occurred at Savannah River, Hanford, and the Tomsk nuclear complex in Russia (a hot part of Siberia if there ever was one).  It was only in reading closely, that I realized that what the accident report was calling “red oil” was the same mix of chemicals used in the PUREX process.

There was a handy safety tip my radiochemistry teacher neglected to share. Below 130C, All Is Well Citizen! Above 130C…remember the Murrah Building? That’s right, if you get it too hot – oh, and the mixture is self-heating – the “red oil” mixture can detonate. Of course, since you are using this to separate plutonium and uranium, it is an explosion full of radioactive delights. I believe this would qualify as a radiologic dispersal device.

One would assume that these accidents could be prevented by keeping a cooling jacket of some sort on the system, or that there was a shut off for the mixers that would stop agitating the chemicals so that it could simmer down, or that there was some kind of ventilation thingee to take away the splody vapors. These would be good assumptions. Such things do exist. They do, however, need to be working and/or actually turned on by personnel. The good news is that there have been precious few of these accidents (4 total that I’ve found so far).  The bad news is that one of them, the one at Tomsk in 1993, gets compared with Chernobyl contamination-wise.

There is another hazard that comes with PUREX that is peculiar to uranium and plutonium: Criticality. The whole point of PUREX is to concentrate uranium and plutonium.  If you get too much fissile material in one place it, duh, achieves critical mass.  There’s some strange things that happen because this is all in a liquid phase (yes, liquids can go critical).  The key question becomes, “How enriched is the stuff in solution?” This was the question they failed to ask at Tokaimura, September 30th, 1999.

You see, they had developed some processes using PUREX to extract uranium to make reactor fuel. The one they used most often was the one for Low Enriched Uranium (LEU) with is less than 5% 235U. They had another one, that was rarely used, to process Highly Enriched Uranium (HEU), using the same system, which mainly involved a change in how large a batch they made. Both of these processes had been certified and approved by JAEA, Japan’s equivalent to the Department of Energy.

Over a couple of years operation they gotten a lot more business than they started with. Managers came up with a great idea to bump up the size of the LEU batch process that involved a large tub, a big paddle for stirring, and then climbing up a ladder to pour buckets of this solution into a separator tank. They did not consult with safety personnel or the governing body regarding this change.

Then they got their first order for HEU in years. They moved over to another building with an identical set up and got to work using the unapproved process.  When the worker added the fourth bucket of nitric acid with dissolved HEU to the separator within 90 minutes, there was the lethal blue Cherenkov flash that let everyone know a criticality had occurred.

Two died, one of them over a very long unpleasant month, several hundred were exposed, and roughly 300,000 were in harm’s way and “given instruction”.

There was another PUREX criticality that occurred at Hanford just because they turned the mixer speed up too high. Obviously, the fluid started creeping up the walls the faster it was spun. Not so obvious was the fact that as the shape of the fluid changed, so too did its critical mass. X kilograms of nuclear material in solution resting in a tank was fine; one tenth of that was enough to go critical in that nice spun up shape.

Chemistry and radiation is a sticky wicket. Really gotta think about things before you do things.