Report into Sept. 19th Large Hadron Collider Incident Released
September 19th was a sad and frustrating day for accelerator physics.
After a long, long wait, the world was sitting on the edge of their seats, anticipating news about the first proton circulation on September 10th. Then, only a day later, a huge transformer broke down, stuttering LHC operations and forcing yet another delay. And then, only a week after the first successful circulation of particles, the worst possible news surfaces from CERN: A “quench” had occurred, stemming from a short circuit across two electromagnets, causing tonnes of liquid helium to leak into one of the tunnels. A huge amount of energy was dumped in a short period of time, heating the once-supercooled magnets by 100°C…
Now, a month after the quench, an official interim incident report has been released and it’s not good news. The report’s findings can be summed up by one of the engineers first on the scene of the damaged section of tunnel who said, “it wasn’t a pretty sight.”
According to Cosmic Variance, the LHC tunnel was too cold to enter until a few days ago. It would appear the liquid helium leak kept the underground chamber in deep freeze for a couple of weeks (after all, when chilled, the helium brought electromagnets down toward absolute zero). So now CERN engineers have compiled a report briefing us about what happened and how they intend to fix it. The summary of the report is a good opener:
On 19 September 2008, during powering tests of the main dipole circuit in sector 3-4 of the LHC, an electrical fault occurred resulting in mechanical damage and release of helium from the magnet cold mass. Proper safety procedures were in force, safety systems performed as expected, and no one was put at risk. An ad hoc task force was set up on 22 September 2008 to investigate the incident, establish the sequence of events, analyse and explain their development in relation with design assumptions and manufacturing and test data, and recommend preventive and corrective actions for further powering of the machine. Today a number of findings have been established, but inspections are not completed and investigations are continuing. Consequently this is an interim summary report of this task force as of 15 October 2008.
The main report can be summarized as follows:
- Serious mechanical damage to 24 long dipole magnets and 5 of the quadrupole magnets (the magnets which focus the beam of accelerated particles).
- The helium enclosure between two of the magnets was punctured by an electrical arc. This arc was caused by an overloaded circuit that failed during an electrical test.
- Helium leaked into the vacuum jacket surrounding the enclosure.
- An extreme overpressure built up in the vacuum jacket down the long chain of magnets.
- The magnet stands were ripped from their fixtures that were once bolted into the concrete floor by the rapid release of energy.
Two tonnes of liquid helium was initially lost as a result of the quench which instantly (as one might expect) clouded up the tunnel, raising the alarm. However, slower leaks after the incident resulted in a far greater leakage:
About 2 t of helium, corresponding to the volume contained in the magnet cold mass of subsectors 19-21, 23-25 and 27-29, were rapidly discharged and eventually released to the tunnel, producing a cloud which triggered the oxygen deficiency hazard detectors installed on the tunnel vault and tripped an emergency stop, thus switching off all electrical power and services from sector 3-4. In the subsequent leakage from the open circuits, and before restoration of electrical power enabled to actuate cryogenic valves, another 4 t of helium were lost, though at much lower flow rates. The total loss of inventory thus amounts to about 6 t, out of 15 t initially in the sector.
- No chance of any particle collisions in 2008.
- Scheduled winter maintenance would have halted experiments anyway, so repairs to the LHC will be carried out in parallel with routine tasks.
- Repairs will take months to be completed.
- It will take weeks to commission the repaired LHC some time in 2009.
- Initial particle collisions probably will not exceed 10TeV.
And what about questions of safety? Actually, all the safety precautions appeared to work as designed and no one was hurt during the quench. However, engineers are making some recommendations for future LHC operations including trying to seek out other related electrical faults that may resemble the circuit failure prior to the September 19th quench, and improved power-up procedures.
The report ends on this sobering note:
The technical parameters of the LHC are beyond precedent, and the energy stored in the superconducting magnets huge. Consequently, operation of this machine will always comprise a certain technical risk. We are however convinced that the repair actions under way and the improved protection systems to be implemented will ensure safe powering in the future.
Incidents such as these are, perhaps, inevitable. After all, we are dealing with the most powerful experiment ever built, faults and incidents are what goes with the territory. My appreciation goes to the scientists and engineers who ensure the safety of their team whilst trying to push physics into a new era of discovery. I’ll eagerly be awaiting (once again) for the first particle collisions in 2009.