Discussion of the Problem-----------------Return to Risk Evaluation Forum Home Page

Recent developments in physics suggest the possibility that an experiment at the CERN research facility in Europe will destroy the Earth. CERN is installing a new high-energy particle collider, the Large Hadron Collider (LHC), that is scheduled to begin experiments late in 2007. Some of those experiments could be dangerous.

The LHC will use a ring of super-conducing magnets, over five miles in diameter, to accelerate two beams of particles. The beams will collide with tremendous energy. When they collide they will produce many new particles, including, it is expected, particles never before seen by scientists.

Two of the expected new particles, black holes and strangelets, could be dangerous, so dangerous that they could destroy the entire Earth.

Black holes have gravity so strong that even light cannot escape. That is why they are black. In the worst case, that immense gravity could suck in and swallow the entire Earth. Some physicists have published papers predicting that CERN will produce a black hole every second.

Strange matter has a different arrangement of quarks than normal matter. Some physicists think it can catalyze the conversion of normal matter into more strange matter. If this reaction gets going, it will convert the entire earth into a small ball of strange matter.

CERN has published a paper asserting several safety factors: [Blaizot et al, see our References section]

1) Black holes are supposed to dissipate via Hawking radiation. However, a new study [Helfner] questions whether Hawking radiation really exists. (Hawking radiation is theoretical, it has never been seen.)

2) A collection of strangelets is supposed to be electrically positive on its surface, and therefore not attract other matter. However, a new study [Peng et al] finds that a collection of strangelets can be electrically negative on its surface.

3) Cosmic rays, which can have energy comparable to the LHC and which have been hitting the Earth for billions of years without catastrophic effect, are claimed to demonstrate that there is no problem. However, as CERN itself says, "the kinematic conditions are not identical." [Blaizot et al, p. 4]

Cosmic ray collisions are not the same as collider collisions. The difference is in the speed of collider products. The products of cosmic ray collisions will usually be moving very rapidly. The products of collider collisions can be moving slowly.

In both cases there are caveats.

The products of a cosmic ray collision sum the momentum of a very fast cosmic ray particle and a slow earth particle. The center of mass of collision products will be moving rapidly. However, a particle exploded backward in the frame of the products just might be moving slowly in our rest frame. CERN cites two assumptions about the rapidity distribution of cosmic-ray-created strangelets, one assumption that permits an occasional slow strangelet, the other that does not. [Blaizot, p. 4] By the precautionary principle, we should consider that the conservative assumption might be true.

In the case of black holes, it seems reasonable to assume that the event horizon captures all products of two colliding partons. Since one of the two partons has a large momentum, the black hole will be moving rapidly. While parton distribution functions permit some partons that contain a low portion of cosmic ray momentum, we can assume that they would not have the energy to form black holes. If these assumptions are true, cosmic rays cannot produce a slow black hole.

On the other hand, the products of collider collisions can sometimes be moving slowly. The momentum of the colliding particles more or less cancels. The collision of consequence is that of the partons, which carry a random fraction of proton momentum. Since each has a random fraction, they do not cancel exactly, and the surplus momentum is usually enough to impart substantial velocity. But not always. In an email exchange Greg Landsberg calculated that 158 black holes per year made by colliders would be moving at less than escape velocity from earth.

A difference in velocity means a difference in results. If they are moving rapidly, there is reason to think that neither strangelets nor black holes can harm us. In the case of a strangelet, CERN predicts that a strangelet cannot survive high speed collisions with normal nuclei. In the case of a black hole, there is reason to think that its initial accretion rate would be low, similar that of a neutrino. If it acts like a neutrino, it would pass right through the earth. It would need to accrete thousands of particles to slow below escape velocity from earth. The probability of accreting enough particles in one pass through earth, or even in several passes through the sun and other stars, is so low that the probability of even one case in trillions of trials over billions of years is essentially zero. On the other hand, if strangelets or black holes are moving slowly, they have time to cause problems. Black holes would drop into Earth and begin to accrete at a rate that has not been studied carefully, a rate that might or might not be fast enough to harm us, but a rate that in all models increases exponentially. Thus the analogy between colliders and cosmic rays is not definitive. We can not count on experience with cosmic rays to prove that colliders are safe.

Physicists have offered a plethora of other reasons why nothing can go wrong. See our section, "Help us find a limit to this model." Some of these reasons may prove definitive. At the moment we think they need work. We encourage more work in this area.

It is useful to think of reasons why we may be safe. However, we ask people who do so, to please think out these reasons carefully before they are used to assure the public that nothing can possibly go wrong. The precautionary principle asserts that proponents of risky experiments should prove they are safe. It is unethical to do so with reasons that do not hold water. We are talking about destruction of Earth here. We do not want to make a mistake.

One physicist published a popular article assuring the public that nothing can go wrong. His main reason, not mentioned in the article, was that he believes that there is no such thing as a black hole. His reason for this belief: "When equations go to infinity, that is a sign that there is something wrong with the equation." He may be right. However, we should not base the safety of the entire earth on a theory that has a reasonable chance of being false.

Reporters and editors, when considering this story, often ask a physicist. That physicist often repeats one of the rather questionable reasons why nothing can go wrong. Reporters have an obligation to follow up.

Brian Cox was recently quoted by Reuters as estimating the probability that the LHC will destroy earth as 1 E -40. When we followed up via email, he wrote back, "I don't estimate the probability as that - I probably said, in answer to a question from the audience, "less than a very large number such as 10E-40" ! i.e. not going to happen." He did not stand by his own offhand estimate, for which he claimed no basis. We cannot expect him to be perfectly prepared for every question, but his answer should not have been given to the world by reporters as a reason not to worry without a bit more checking. Good follow-up would have produced a more accurate article.

Return to Risk Evaluation Forum Home Page