Slashdot Mirror
Antineutrino Device Tackles Nuclear Proliferation
KentuckyFC writes "One of the biggest problems in nuclear proliferation is verifying that countries are not secretly transferring fissile material by taking it out of reactors and selling it. Now a group of US scientists say they've developed a machine that can remotely detect whether a reactor has been switched on and off by detecting the antineutrinos produced by nuclear reactions. The detector is about the size of a car engine and is designed to be left near a reactor to record data. The group has been testing a prototype at the San Onofre Nuclear Generating Station in Southern California and says it works well (abstract). Now it's up to the International Atomic Energy Authority in Vienna to decide whether to deploy the new machine."
They are extremely difficult to detect, but not impossible. Moreover, nuclear reactors produce quite a lot of them.
As a recent example, the KamLAND neutrino experiment (http://kamland.lbl.gov/) used a 1000 ton detector in Japan to study the flux of neutrinos emitted from dozens of reactors in Japan and Korea, some hundreds of miles away. KamLAND performed precision studies of the propagation of neutrinos over distance, and was also able to detect the rising and falling neutrino fluxes as various reactors powered up and down.
The detection device described in the article is much smaller, but it's located much closer to the reactor. I've heard talks on this, and it seems quite reasonable.
However the proposal is to place the device rather near to a nuclear reactor. A reactor generates enough flux that big detectors can measure the neutrino flux from miles away. So a small (and probably less shielded) detector that is much closer (<100 m) will receive enough flux to get statistically significant data.
The preprint says: For example, cubic meter scale hydrogenous scintillator detectors, containing 10^28 target protons Np, will register thousands of interactions per day at standoff distances of 10-50 meters from typical commercial nuclear reactors. The proposal thus requires the detector to be installed on-site at the nuclear reactor. You may wonder "Is this actually useful for monitoring, then? If someone wants to lie about their reactor usage and burn rate, they will just falsify the detector records, too." Well, the idea is that this local detector is completely independent from the reactor (it doesn't rely on sensors hooked into the reactor, for instance). Thus it can be locked and sealed off completely; installed and managed by a completely independent oversight organization.
First, it's very very heavy. You need a thick lead shield to block cosmic rays, but that's not the biggest problem.
Neutrino (antineutrinos work the same, but give off a different signal) collisions are very rare. The device needs to be close to the source, so that many neutrinos pass through the device. The density of neutrinos fall off as the square of the range (inverse square law). If you double the range, you get one quarter the number of collisions. Even close to the reactor, they are rare enough that it's hard to tell collisions from the local reactor versus background noise (the sun and other reactors are probably the biggest sources). This why it takes them hours to figure out when the reactor has turned off. The whole thing relies on the statistics of a very low probability event. From space, the background noise would easily swamp the signal from a specific reactor.
You might be able to tell which part of the earth is releasing more neutrinos if your orbit was low enough, but it would be very tough to pinpoint a source to a specific country. It would be very hard to get both decent spacial resolution and time resolution.
Unlike this detector, then neutrino detector at South Pole, Amanda, can tell which direction a neutrino came from. It uses the Earth for shielding and many kilometers of ice as the collision target. I assume they detect neutrinos from reactors, but probably not often enough to tell exactly when they are on (just a guess).
The problem is one of mass, not of interference. It takes a long time (days) to start a reactor back up after shutting it down, so a satellite in a typical 90-minute polar orbit could check every reactor on Earth twice a day, at a distance close enough to tell it from all other reactors. Since the atmosphere provides almost no shielding against high-energy cosmic rays, both a ground-based detector and an orbital detector require some other way of filtering those out.
However, such a satellite would be too massive to launch. A detector with a 50-meter range weighs about 1000kg, and neutrinos obey the inverse-square law. A satellite at an altitude of 500km would weigh over 1,000,000,000 kg.
"They redundantly repeated themselves over and over again incessantly without end ad infinitum" -- ibid.
One of the things that makes reactor operation so hard is that they don't like being run at low power. If you run a reactor at 50% power for more than a few minutes, the buildup of fission byproducts will cause the reactor to shut down, requiring a very expensive restart process.
"They redundantly repeated themselves over and over again incessantly without end ad infinitum" -- ibid.
Disclaimer: I am NOT a particle physicist (but I did stay at a Holiday Inn Express last night). We seem to be drawing an overall conclusion that it is antineutrinos that are being sought by these detectors. However, since the neutrino and antineutrino are both neutral (read: they have no charge) particles, is it not possible that they are the same particle, or perhaps simply two variants of a core type of particle?
If I recall correctly, the only differing property between the neutrino and antineutrino is helicity; the neutrino has left-handed helicity, the antineutrino right-handed. Perhaps someone can clarify this a bit better?
During the http://en.wikipedia.org/wiki/Neutrino_experiment they got only 3 neutrons per hour from a reactor just 11 meters from a detector.
Neutron flux obeys the inverse square law, so this detector should detect only few neutrons per _day_ at the distance if 100 meters.
If you read the article, it mentions that they can monitor the on/off status of the reactor with a time resolution of 5 hours, and power output over month long intervals... if they were getting, say, 10 anti-neutrinos an hour those time resolutions would be much more granular... I would guess they're getting 3-8 detections a day based on those figures - which would fit nicely with your estimate.
They do mention that they're using some sophisticated filtering techniques (sounds like a combination of hardware and software) to try to up the detection rate and reject the falses...
Even though it's sort of a not-so-well thought-out question, I'll bite.
When sometimes the discussion goes to pollution of nucear power plants, the discussion is not that it pollutes on a daily basis it's immediate surroundings. Indeed, if all things go well it should be a remarkably clean operation.
However:
- there is the small question of what to do with the fuel rods, some other consumables, and even the power plant itself, after they are "spent"
- in case of an incident, pollution does become relevant (this does not have to be a big bang, but can be leaks into the ground)
- the production of the fuel rods is by the way also a pretty dirty job
It is also important to note the fact that incidents very hard to clean up. So in a theoretical function of "toxicity", "chance it might happen" and "cost to clean up" we have -to offset the others- done a heck of a lot to reduce one of those coefficients, but *if* it happens it is really bad news.
Would this prevent the use of such detectors in satellites? I realize that this first design requires that it be essentially next to the reactor, and that its time resolution is far too short for common orbits. But would there be too much interference in space for this to ever be of practical use?
According to the article, the device has filter materials meant to reject cosmic rays before they reach the detector. They're also looking for a very characteristic double-detection event; the anti-neutrino reacts with protons in the tank in a two-stage reaction that produces two gamma rays seperated by a very small amount of time. (The gamma rays are then converted to visible light through the use of a scintillating material, and detected with photomultiplier tubes)
By all sounds, they've done a very good job both in hardware and software to eliminate false detection events. Of course, take that with a grain of salt; the person who most has a vested interest in the device eliminating false detection events is also the only person telling you how good that is. If it proves out in peer-review, it'll be an impressive accomplishment...
So it might work in space. One of the size from the article would be next to worthless - it'd be good enough to tell you if a nuclear powered spacecraft was moments away from capturing or destroying your satellite, and that'd be about it... The Japanese built a 1000-ton detector that got events from most of their island nation a few years ago... maybe with these guys' techniques, you could scale that guy down to about 100-tons... but then, that's just the weight of the scintillator - the Japanese detector's scintillator was in the middle of a much larger facility...
http://en.wikipedia.org/wiki/Kamioka_Liquid_Scintillator_Antineutrino_Detector
Using that detector's sensitivity as a guideline, it has a published mean distance to the reactors it was monitoring of 180km - just barely enough for an LEO satellite to graze the surface.
Really, putting one of these in space makes no sense until more than one nation has nuclear-powered space ships up there...
There are a ton of fission-powered man made objects up there, sure... but they all use a miniscule amount of material in a low power output configuration that is likely undetectable from any distance... right now there's just nothing to detect up there...
That isn't a problem because the difficulty is to detect when the reactor is off thou they claim it isn't. During normal operations a reactor will run close to full power for up to a year at a time, but if you want to produce nuclear weapons you must keep shutting it down to refuel since leaving the plutonium in the reactor for too long severely degrades its suitability in a bomb. Thus all the detector needs to do in order to blow the whistle is to show that the reactor operated in a very unusual pattern rather than the continuous high power mode that is more common.
Furthermore, for baseload plants like large nuclear power stations you can't just change the power output as you like because the type of turbines used do not spin up and down very easily. Special load-leveling plants ( typically natural gas or hydroelectric ) are used for this purpose, so if somebody is running a large baseload nuclear plant at 50% power it would almost certainly attract suspicion.
People usually focus on charge reversal for particle-anti particle opposition; charge reversal is a consequence of anti-particle formation, not a cause. If a given particle has, say, 3 specific quarks, its anti-particle is the one with the 3 anti-quarks of them. If the sum of the quarks yield a charge, then by nature, the anti-particle will have the opposite charge, but if the total charge is 0, then the anti-particle is also neutral - but not equivalent.