Update on the LHC

20 October 2008

The tunnel, sector 1-8 (archive)

Last Thursday, the office of the Director General released a report on the incident in sector 3-4, after the magnets had sufficiently warmed up and opened for a full investigation. Before the magnets had been opened, Lyn Evans gave a talk during the first day of the recent ATLAS week, October 6th. The information that follows is merged from these two sources.

At the time of LHC turn-on, all sectors had been commissioned to 5.5 TeV except one. Sector 3-4 had only been commissioned to 1 TeV, but that was enough to start lower-energy collisions.

“It started with a transformer failure at point 8, which took down the cryogenics in 7-8 and 8-1,” said Lyn Evans. The accelerator team realized that they had no spares, but CMS was able to lend a transformer. Because recovering the cryogenics would take time, they decided to put this time to greater use and commission that last sector to 5.5 TeV.

During this commissioning, on the morning of September 19th, “a resistive zone developed in the electrical bus”, according to the report. The zone was in between a dipole and a quadrupole magnet.

The resistive area created a potential difference of 300 mV, which was the first evidence of a problem, at 11:18 a.m. In a fraction of a second, that voltage had grown to 1 V, and within half a second, the power converter stopped supplying current. A second later, the current-dump resistors had engaged, resulting in the first magnet quenches. Approximately 100 magnets would quench in the next 20 seconds. These quenches worked perfectly.

However, during that first second, an electrical arc punctured the insulation vacuum surrounding the helium chamber. Helium rushed into what used to be a vacuum chamber. This chamber connects segments of magnets, but barriers exist at each quadrupole. The barriers can withstand pressures of 1.5 atm, and the valves open at 1.4, according to Lyn Evans.

But the valves couldn’t cope with the amount of helium escaping into the vacuum chamber, and the increased pressure pushed the magnets apart, breaking “jumper” connections between quadrupoles and dipoles and pulling anchors out of the tunnel floor. All told, two tonnes of helium leaked into the tunnel, triggering the oxygen deficiency system.

The preliminary report estimates that at most five quadrupoles and 24 dipoles will have to be replaced. The accelerator team has enough spares to make these replacements. In the plenary, Lyn Evans stressed that these magnets aren’t lost – the damage is reparable. The repairs are easier to make on the surface, however, so it’s simpler to replace the magnets and repair later. Also, a black “soot-like” substance, starting from the region of the electrical arc, has contaminated the beam pipe, and the team is considering how best to clean it up.

“Of course, the repair is one thing, but more essential to me is to be sure that it can never happen again,” said Dr. Evans. To that end, his team has been examining the behaviour of the system prior to the incident.

Up until the incident, the rise in temperature had been measured in thousandths of a degree Celsius, which is insignificant compared with normal fluctuations. However, they’ve found a “signature” in the Joule-Thomson valves that help regulate the cooling system. For an hour, “that valve was continuously opening, calling for more cooling in that cell,” said Lyn in the plenary.

The team identified a second suspicious splice in sector 1-2, during the commissioning to 5.5 TeV. During the winter shut-down, in addition to the repairs in sector 3-4, they will thoroughly test that splice and repair it if necessary.

The decision has not yet been made regarding the maximum energy at which the LHC will operate next year.


Katie McAlpine

ATLAS e-News