For this scenario, count your blessings that they bothered to call you at all. SURPRISE ACCELERATOR is the worst kind of accelerator. I wish I could say this has never happened in my career. But you already knew this one existed. It’s the modification that’s the problem.

[The fourth in an ongoing series of my compiled explainers for my CHOOSE YOUR OWN RADIATION ADVENTURE quizzes. There’s never really a right answer but some might work out better under the constraints of the scenario. It’s like poetry, really.]

At the most basic level, an accelerator is a machine that makes charged particles go much faster. Everything beyond that is just getting fancy and going faster in highly esoteric ways. The machine doesn’t particularly care what particles you put in it; it just makes them go fast. Because your client isn’t technically inept, they were wise enough to make the changes that reversed the polarity to turn an electron accelerating machine into a positron accelerating one. Otherwise, it would’ve just thrown the positrons right back at the ion source. But it VERY MUCH MATTERS what’s at the business end of the accelerator. Because your beam line must come to an end and your fast particles are going to have to smack into something. Hopefully your intended target, but that’s why you build a backstop for the ones that miss. 
And because you’re working with positrons here, that means it’s time to worry about matter-antimatter annihilation radiation reactions. Your positron is going to go away and in its wake it’s going to give you two 511keV gamma photons moving in equal and opposite directions. But unless your research is on accelerator technology development, the whole point of having an accelerator is to make charged particles go fast to get nuclear reactions. The resulting gammas from the VERY short lived things in your target tend to be a lot more energetic than 511keV. If you’ve built your backstop and target cave right, they should take care of all those pesky annihilation gammas. The positron interactions in the ion sources and the accelerator are going to be a pittance of dose contribution compared to the x-rays from the accelerator itself. 
Of course, those are your intended reactions. If your accelerator is operating at a high enough energy, you can start causing incidental activation of materials, like those stainless steel screws slowly growing more and more Co-60 over time. If your accelerator was built with this in mind for the interactions from using electrons and now you’ve swapped to positrons, all your dose and activation calcs that you submitted for registration & permitting of the accelerator go out the window. Luckily, the different activations don’t deviate too badly from a safety point of view, but all the documentation is now wrongwrongWRONG. If your accelerator is licensed for isotope production, say radiopharmaceuticals, you just invalidated your permit to operate. That is, from a business point of view, extremely bad. But an accelerator that’s shut down until they can put it back the way it was so that it matches the paperwork again (they better be able to do that) or you can get new paperwork approved (may take months) is very safe indeed. 

Which brings us to the most important question several of you identified that you’ll be asking. “Where did you get a positron ion source large enough for this? Howwwwww???” If you’re lucky, it’s one you know about that they’ve installed into the accelerator. To have enough positrons to make a decent current ion source, you’re going to need an isotope with a half-life long enough to build with. Because there aren’t all that many positron emitters like that, this means your ion source is probably a big pile of Na-22. While Na-22 is a positron emitter with a ~2.5yr half-life that you can utilize, it is also a VERY potent gamma emitter, especially when you get enough of it together to think of it as a positron source instead. Usually, ion sources for accelerators just make a lot of soft (low energy) x-rays as you generate plasmas to throw down the line. Soft enough that the ion sources are typically self-shielding. This is *NOT* the case for a GBq Na-22 source. You need some lead, stat. Lots of it.

Also, as you may have noted Na-22 is sodium. Working with sodium is *messy*. Once you’ve finished building a lead cave for the ion source, it’s time to survey everything and everyone to determine exactly how much Na-22 they spread all over the place while building this. 
In the events that partially inspired this scenario, a researcher had a large Na-22 source as part of their lab’s inventory. But they were retiring and put out the word that they’d like to give this source and a lot of vacuum system gear away to a good home. This is the classic “If someone wants it, then it isn’t waste” gambit to avoid decon and disposal fees, but that’s a rant for another time. And so, a very large Na-22 source, in NaCl chemical form, showed up and got put in the extra beefy source safe. Even with the tungsten walls, counting experiments everywhere in the entire building began to show the fingerprint Na-22 gamma line. 

I’m happy to say that they called before trying to reconfigure the accelerator. This was because they were stymied trying to figure out a way to build the ion source safely. Their first attempt resulted in a contamination incident and everything got put away with a harrumph. But it’s cool, they’ve got a decade and change to figure it out before that source dies away too much to be useful. Also, they added a lot more lead around the safe to be better neighbors.