Fake Banner
    Accident Forces Two-Month Delay At The Large Hadron Collider (LHC)
    By Alan Gillis | September 20th 2008 01:07 PM | 4 comments | Print | E-mail | Track Comments

    No collisions and no beams either next week at the Large Hadron Collider (LHC). The BBC reported a quench of about 100 superconducting magnets yesterday that heated up as much as 100 Celsius. A ton of liquid helium spilled into the tunnel and the CERN fire brigade went in. Cause of the quenching has not been announced, nor have any injuries been reported.

    The fault does not pose any longer-term threat to the LHC. The quench occurred during final testing of the last of the LHC's electrical circuits. Liquid helium leaks vaporize back to a gas almost instantly and would freeze or choke personnel present but no workers were at risk, according to CERN, the European Organization for Nuclear Research, which runs the LHC.

    Ordinarily magnet quenches occur when part of the superconducting coil enters a 'normal' state, either the field inside the magnet is too great, eddy currents occur or, more rarely, because of a defect in the magnet. CERN's press statement states "the most likely cause of the problem was a faulty electrical connection between two magnets, which probably melted at high current leading to mechanical failure."

    If beams had been running, they would have been at their lowest power, 0.45 TeV per beam, though beams 11 times more powerful at 5 TeV were scheduled before October 12th. Vacuum conditions were also lost said the BBC, citing LHC's online logbook.

    This is on top of news that a thunderstorm burned out a giant 30-ton 12 Million Volt Amperes surface transformer that powers some of the helium cryogenics system, a problem CERN admitted on September 18th, without mentioning the thunderstorm, and said the transformer was replaced last weekend. The failure of the transformer caused an initial warming in the helium coolant in 2 of the colliders 8 sectors, some magnets as late as September 17th at near 7 K.

    The TimesOnline initial report on damage said, "One of the beams had been captured by Friday, but work was then interrupted by the loss of electrical transformers that power the cryogenic cooling system . . ." A beam was apparently running during the thunderstorm, though it seems the beam was stored before the giant transformer failed, but perhaps more than one transformer lost in this early account. CERN's lastest progress report of Sept 18th was sketchy.

    The damaged Sector3-4, an eighth of the 17 mile collider has still not been stabilized since yesterday morning's accident. About half of it or a mile length was well above normal 1.9 K design temperature, about a quarter of the sector and its magnets yesterday afternoon at 15:46 PM were about 82 K to 110 K. Currently all magnets have warmed. The warmest show a slight recovery from 110 K down to 99 K. and Sector 3-4 is still in crisis as of the 20th September 0:418 AM, with no accurate readout for about half the magnets, and the rest showing more warming, with a few outside the spike zone climbing abruptly in temperature since yesterday afternoon.

    Given the nature of the accident, repairs could take a week or longer. At Fermilab's Tevatron, a quench took 2 weeks to repair and get the collider up to speed again. The damaged LHC ring Sector 3-4 is considered to be equivalent to the Tevatron's ring size, overall like 8 Tevatrons.Near Geneva, the 17 mile Large Hadron Collider ring underground in red. Courtesy cnet/CERN2008 

    A later update Sept 20th by Associated Press adds that a shutdown for repairs would take about 2 months according to CERN. Their preliminary finding shows a faulty electrical connection melted.

    For a comparison of Cooldown Status Graphs mentioned above from the accident yesterday, see Accident Forces Two-Month Delay At The Large Hadron Collider (LHC).


    Higgins, Alexander G. "CERN: Damage to new collider forces 2-month halt", Sept 20, 2008, AP News on NewsTimes, http://hosted.ap.org/dynamic/stories/E/EU_SWITZERLAND_PARTICLE_COLLIDER?SITE=CTDAN&SECTION=HOME&TEMPLATE=DEFAULT

    BBC. "Hadron Collider forced to halt", Sept 19, 2008, BBC News, http://news.bbc.co.uk/1/hi/sci/tech/7626256.stm

    CERN. "LHC progress report, week 1", Sept 18, 2008, LHC First Beam, http://lhc-first-beam.web.cern.ch/lhc-first-beam/News/lhc_080918.html

    Henderson, Mark. "'Big Bang Machine' back on collision course after its glitches are fixed", Sept 18, 2008, TimesOnline, http://www.timesonline.co.uk/tol/news/uk/science/article4774817.ece


    Perhaps CERN could take this gift of time to begin to properly address the credible concerns of scientists who challenge aspects of CERN's safety assumptions.

    I would also like to see CERN agree to at a minimum follow senior astrophysicist Dr. Habil. Rainer Plaga's feasible risk mitigation proceedures that include "SLOW DOWN!"[1]

    [1] arxiv.org/PS_cache/arxiv/pdf/0808/0808.1415v1.pdf On the potential catastrophic risk from metastable quantum-black holes produced at particle colliders - Rainer Plaga Rebuttal (2008)

    We were making fun of their 'hurry up' concept more than a year ago, though not for the paranoid 'let's-make-up-disaster-scenarios' you do, but because that's not how it works with big machines:

    Is the Large Hadron Collider Ready For Prime Time?

    “The low-energy run at the end of this year was extremely tight due to a number of small delays, but the inner triplet problem now makes it impossible,” said LHC Project Leader Lyn Evans. “We’ll be starting up for physics in May 2008, as always foreseen, and will commission the machine to full energy in one go.”

    That was bordering on ridiculous because, as the Tevatron showed, quenches happen all of the time in a new machine. Of course, this startup was delayed by a long time (years) so they have a certain pressure on them. As the saying goes, "there may not be time to do it right but there will always be time to do it over."

    If anything, a delay of two months will make the hype die down and disaster scenario types can go back to worrying about Mayans.

    Hatice Cullingford
    Going back to the concept of risk assessment of a large engineered system like the LHC, each phase of such a project would entail its own risks. I would ask simply, was this "accident" identified in the LHC risk assessment study?? As I wrote here in "LHC Is Not a Fast Reactor (But Is It Also Safe?)", there are three questions to be answered: 1. What can go wrong with LHC? 2. How likely is that to happen at LHC? 3. What are the consequences? Hence, my invitation to Lyn Evans stays for us to do an independent risk asessment of the LHC. Why? Because we want the scientists to succeed!

    points out that Plaga applied a formula for BH evaporation rate to a physical situation where it is not valid, and got an answer about 20 orders of magnitude too large.
    Indeed, any physicist can make a mistake, but he should also acknowledge it and respond to criticisms. So far no reply from Plaga. At least he is modest and careful enough not to be giving interviews all over the world yet.

    Indeed one should make some very careful checks before claiming that a calculation is good enough that we should use it to determine the fate of a multi-billion dollar experiment, let alone the Earth... it looks like Plaga did not ask anyone to look over the equations before releasing the paper.

    It is also disappointing that people are only recently starting to dream up their favourite disaster scenarios... LHC has been planned for decades, LHC-black hole papers have been written since 2000-2001, somehow it only becomes worthwhile for LHC critics to go on the record a few months or weeks before the startup?

    Unfortunately there are an infinite number of wrong equations that one can write to 'show' that LHC is possibly dangerous. One cannot refute all of them in time and still start the LHC without an infinitely long wait.

    Still, it is in principle a good thing that people are critically examining the Ellis and Giddings papers on LHC safety and willing to put meaningful arguments and equations down on paper.

    What we had last week was a 'quench' which is a widely anticipated technical problem. Of course it is not welcome but none of the components are irreparably damaged and there are no injuries.