Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

LHC meltdown before first collision

Europe's largest particle accelerator might not produce data until 2009.

The shutdown may be due to a welding failure. Credit: M. BRICE/CERN

A mechanical failure has shut down the world's largest particle accelerator for at least two months.

The Large Hadron Collider (LHC), a US$4.1-billion machine located at CERN, Europe's particle-physics laboratory near Geneva, Switzerland, will be out of commission possibly until early next year to allow repairs to be made to a section of superconducting cable that is thought to have melted during a test on 19 September, just nine days after the collider circulated its first beams.

"We've had better days," says James Gillies, CERN's chief spokesperson.

The LHC is designed to accelerate protons to energies of 7 teraelectronvolts before smashing them together. To keep the beam focused and fast, the machine depends on around 9,600 superconducting niobium–titanium magnets. The magnets must be kept at temperatures of just 1.9 kelvin while carrying roughly 10,000 amps. The magnets carry this enormous load without generating heat, but if a section of the niobium–titanium rises above its operating temperature it can lead to a catastrophic failure known as a 'quench'.

That is exactly what seems to have happened as physicists passed 8,000 amps into a sector of the LHC's 27-kilometre underground ring. A cable feeding current between two of the LHC's beam-focusing quadrupole magnets suddenly heated to above superconducting temperatures and melted. The failure seems to have happened at a joint where two sections of cable were spliced together. Tens of thousands of joints run around the LHC and many of them had already been tested without incident.

The failure caused the liquid helium that was being used to cool the magnets to boil off, apparently rupturing the machine and releasing as much as a tonne of the gas into the LHC tunnels. During testing the tunnels are evacuated and no injuries were reported.

Such failures are not uncommon during the early commissioning of an accelerator, Gillies says. "With a normally conducting machine you could fix it in a couple of days." But the LHC's superconducting status also makes it difficult to service. To fix the broken sector, physicists must heat thousands of tonnes of magnets from near-absolute-zero to room temperature, make the necessary repairs, and then slowly cool the system back down. Just warming and cooling will take at least two months, Gillies says.

The setback has dashed the hopes of thousands of researchers working on the four main detectors that will use the LHC. They had hoped to get their first data from proton–proton collisions as early as this week, but it now seems unlikely that collisions will occur this year. CERN normally shuts down during the winter months to save on electricity costs, says Tommaso Dorigo, a physicist at the CERN's Compact Muon Solenoid detector. He and others think the repairs will run into that winter shutdown. "I'm not sure collisions are going to happen in 2008," Dorigo says.

That doesn't mean physicists will be sitting on their hands in the coming months. Much must still be done to prepare the detectors for first collisions, according to Pauline Gagnon, a senior scientist on CERN's ATLAS detector. Physicists must still calibrate the hundreds of thousands of sub-detectors in each of the larger experiments, and that can be done with high-energy cosmic rays from outer space rather than beam from the machine. "It would be easier with the beam, but we can still do it," Gagnon says.

But for graduate students trying to complete their theses, there is no substitution for the real thing. Fresh data are imperative, especially for the American contingent, according to Seth Zenz, a graduate student at the University of California, Berkeley, who is currently working on ATLAS. Without new data, Zenz explains, he can't complete his thesis, and he is now planning on spending seven years in graduate school instead of six. Still, Zenz says, he's willing to put up with the delays: "I can wait a few more months before getting back to affordable sushi and burritos in California," he says.


Additional information

UPDATE: CERN has announced that the LHC will not restart until the spring of 2009, after the winter maintenance period.

Related links

Related links

Related links in Nature Research

LHC Special

Related external links

CERN's statement

LHC hardware commissioning

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Brumfiel, G. LHC meltdown before first collision. Nature 455, 436 (2008).

Download citation

  • Published:

  • Issue Date:

  • DOI:


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing