On “Death Rays”

One bad thing about SteelyKid’s preschool graduation yesterday was that it drained my phone battery, causing me to miss an interview request from a local TV station looking for somebody to talk about a a couple of local guys arrested for a plot to build a “death ray” from X-ray components. This is pretty far from my area of expertise, but, hey, I’d be willing to go on tv and talk about just about anything physics-related. And, as always, I promise to be at least 90% less wrong than Michio Kaku.

Of course, on another level, I’m kind of glad that I didn’t get the message in time, because when I decided to write something about it this morning for the blog, I was reminded that I absolutely hate dealing with medical physics. I work in the world of lasers and optics in the visible-ish region of the spectrum, where the language we use to talk about optical phenomena is pretty straightforward: we talk about energy in terms of either the total energy delivered to a given area or the number of photons delivered to a given area. Medical physics, on the other hand, is a sucking swamp of cryptic units for exposure, dose, absorbed dose, effective dose, dese and dose, and all kinds of other crap. this page is about the best I’ve found for connecting these things together, and it’s still more unit conversion than I want to worry about. On top of that, it’s damnably difficult to get straightforward information about the numbers you would need to talk sensibly about any of this stuff– everything is either proprietary or custom, and thus specs aren’t available on the Web, or it’s a resource aimed at panicky laypeople afraid that they’re going to sprout a second head because their dentist x-rayed their teeth, and God knows, including actual numbers would only freak those people out.

But let’s see what we can come up with based on the little information given in that news story, and see what can be done to assess the plausibility of this plot. The information given in the article suggests that they were planning to use a “powerful industrial X-ray machine that would be powered by a makeshift, 2,000-watt battery,” that would fit in a van, and aim a “high-energy lethal beam of radioactivity at human targets.” The object was to have victims “mysteriously die from radiation poisoning within days.”

Now, given that these guys were apparently trying to shop this to both Jewish groups (for “killing Israel’s enemies while they slept”) and the Ku Klux Klan, they’re fairly clearly a few fries short of a Happy Meal ™. But that aside, is it physically possible to do this?

In order to kill someone “within days,” the Wikipedia page on radiation sickness suggests you would need to deliver a dose of around 30 Gray. This is a measure of the amount of radiation delivered to a body, which is a function of both time and intensity. You don’t necessarily need a “high-energy lethal beam” to do this– enough time spent in front of a more modest source, like a medical scanner, would do the trick.

So, how long would this take? Well, the data provided by a radiology association suggest that the largest effective dose of any of the procedures they list is around 20 milli-Sieverts, another of the weird units used in medical physics. The actual conversion from one to another is potentially messy, but a passing mention on this page is going to serve as justification for claiming that a Gray is a Sievert and vice versa. We’re doing order-of-magnitude here, anyway. So, the highest medical scan dose would clock in at 0.02 Gy, where you need 30Gy to kill somebody quickly. That suggests you’d need to spend (appropriately enough) 666 times the duration of a typical scan inside the medical x-ray machine in order to kill yourself with radiation poisoning. Not having had such a scan, I can’t really convert that to a time with any great accuracy, but 10 seconds seems like a reasonable order-of-magnitude for the time they run such a scanner, in which case it would take a bit more than 100 minutes to kill somebody with a CT scanner.

That seems kind of improbable, but they’re talking about a “powerful industrial x-ray machine,” so maybe the flux is higher. You’d need it to be 100 times more powerful to get down to a minute of exposure to kill, which still seems like a long time to keep your target in the beam, but if you’re killing Israel’s enemies in their sleep, maybe that works. I’m not sure I believe you can run that off 2kW of batteries in the back of a van, though.

But there’s another problem, here, which has to do with the distance. The x-rays from a typical source spread out as they travel, which means that the same beam covers a larger area as you move away from the device. This area increases as the square of the size of the beam, so doubling the working distance reduces the intensity, and thus the dose, to a quarter of its initial value. If you’re getting a medical x-ray, you’re probably on the order of 0.1m away from the source, but if you’re trying to poison sleeping enemies, I’d imagine your van would need to be at least 10m away. That’s a factor of 100 in distance, or 10,000 in the dose. If it took one minute to cook somebody at 0.1 m, you would need 10,000 minutes to kill them at a distance of 10m, or about a week.

Ah, but what about the “beam” aspect of this? They were planning to focus the x-rays on target, weren’t they? Doesn’t that take care of the expanding beam issue? Well, maybe, but I doubt it. You see, it’s damnably difficult to focus x-rays. If you were trying to cook somebody with visible light, you could just make a big curved mirror to focus the beams at a particular distance, and take care of that. X-rays aren’t like visible light, though– they either pass through or are absorbed by most of the materials you would normally make a mirror out of.

Focusing of x-rays can be done, but it’s a very daunting project. Something like the Chandra X-ray observatory uses an array of metal “lenses” that reflect x-rays at a glancing angle (for reasons that don’t really matter here, it’s really easy to get light of any frequency to bounce off things at a glancing angle– this is part of the reason why asphalt, not ordinarily noted for its high reflectivity, produces enough glare to justify polarized sunglasses for driving). The resulting array is pretty big– a meter or more across– and requires quite a bit of precision machining. And then there’s a significant issue of targeting– getting the tightly focused beam on the target (who presumably is not sleeping out in the open) is a non-trivial problem, even before you factor in that they planned to trigger this thing remotely, from half a mile away.

So, given the limited specs available, the “Hiroshima on a light switch” idea that these guys were pitching to a selection of mutually incompatible groups seems pretty far-fetched on the basis of physics. I don’t think there was very much danger of this succeeding.

(Of course, you don’t even really need to invoke physics to see the implausibility of this scheme: if it were possible to make a van-mounted death ray that ran off a 2kW battery with parts that you could scavenge from an industrial machine shop, every military in the world would be arming themselves with these. The fact that we don’t see these in general use suggests that these guys were just whack jobs.)