In this paper:
http://wwwphy.princeton.edu/mumu/target/King/king_WEBR6_pac99.pdf
"Potential Hazards from Neutrino Radiation at MUON Colliders"
Abstract
High energy muon colliders, such as the TeV-scale conceptual designs now being considered, are found to produce enough high energy neutrinos to constitute a potentially serious off-site radiation hazard in the neighbourhood of the accelerator site. A general characterization of this radiation hazard is given, followed by an order-of-magnitude calculation for the off-site annual radiation dose and a discussion of accelerator design and site selection strategies to minimize the radiation hazard.
It suggests "muon colliders" could produce neutrino beams powerful enough to pose a potentially dangerous radiation hazard, but I don't recall it going into the details of why the danger existed. How is that possible? Neutrinos are not usually very interactive, and the only other time I heard of "deadly neutrinos" was in regard to a supernova explosion, where if you were within 1 AU, it would be enough to be deadly, but if you were within 1 AU, you would be inside the stellar envelope anyways, and thus the neutrinos would be the last thing you'd be concerned about. I can't possibly imagine that a man-made source could come anywhere close to the power of a supernova, much less in emitted neutrinos.
The only thing I can think of is that, apparently, as the neutrino energy increases, they become more interactive, and the muon collider would be generating neutrinos with per-particle energies far in excess of those produced by a supernova. Thus while the total neutrino areal energy density would be nothing compared to that in a supernova, the much higher per-particle energy would dramatically boost the interactivity. Is this the correct explanation?
Best Answer
Something not mentioned in the other answers, but very important to understanding this is that the neutrino-other stuff cross-section grows with neutrino energy. In the regime between solar energies (~1 MeV) and current accelerator energies (a few to a few tens of GeV) the growth is roughly linear a trend which continues some ways further up the energy scale.
The oft mentioned notion of a neutrino going through a light-year of lead with only a 50% chance of interacting refers to solar neutrinos. At the current accelerator scale this is down to around 1/1000 light-year, and at the TeV scale down to one millionth a light-year.
And of course, every muon in a storage ring results in both a $\nu_\mu$ and a $\bar{\nu}_e$ both of which will have an appreciable fraction of the muon's kinetic energy.
Then to get an appreciable interaction rate the muon current in the ring will have to be prodigious.
Then when those (anti-)neutrino interact with matter most of the products continue at high enough boost to be ionizing radiation on their own. Thus the concern about "dirt events" generating a measurable amount of conventional ionizing flux in the plane of the ring.