The complexity behind the date
For the waiting world, and indeed for most of us here at CERN, ‘the LHC schedule’ simply means the date that the LHC will restart - and we only take notice when that end-date changes. But in fact the schedule is a constantly evolving intricate document coordinating all the repairs, consolidation and commissioning in every part of the machine. So, what actually goes on behind the scenes in timing and planning all the work on one of the most complex scientific instruments ever built?
Each week the nitty-gritty details are discussed at the LHC Machine Committee (LMC) meeting, where the four experiments and every technical group from the accelerator sector are represented. "Nearly every week the schedule changes and, although the end-date nearly always stays the same, inside it looks completely different," explains Steve Myers, Director for Accelerators, who chairs the LMC meeting.
Due to the huge amount of inter-dependency between different areas of work in the LHC, even a small change can necessitate a complete overhaul of the schedule. For example, something as simple as cleaning a water cooling tower - required regularly by Swiss law to prevent Legionella - has a huge impact on the planning: "When you clean the water tanks it means we don’t have water-cooling for the compressors, that means we can’t run the cryogenics, so the temperature starts to go up," explains Myers. "If a sector gets above 100 K, then the expansion effects of heating can cause problems, and we could have to replace parts."
A bigger task, such as warming a sector, causes even more of a headache in planning, and must at times seem like trying to solve the world’s most complex Rubik’s Cube. For instance, despite having one of the world’s largest liquid helium storage facilities, CERN’s tanks hold only just over half of the total helium in the LHC – the rest must remain in the machine itself. "To warm up a sector, you have to shift all the helium to another sector. And if that one already has helium, then you need to shift that too. You end up playing musical chairs with helium!"
The initial restart schedule was decided upon during the Chamonix meeting in February this year. "In fact we’ve done much more work than we planned in Chamonix," says Myers. Originally the schedule focused mainly on the repairs in Sector 3-4 - repairing, reinstalling and interconnecting the magnets - but since then a lot of extra consolidation work has been done all around the ring: most importantly, the new pressure release valves, and a whole new system for ‘symmetric quench’ protection. "All of these will make the machine that much safer to operate," Myers adds.
"At Chamonix we agreed on a success-orientated schedule, but we knew from last year that in all likelihood there would be unexpected things that could delay us", says Myers. "Even with all the additional work that we added, and we added a lot, we were still on schedule up to the beginning of May," he continues. "We managed to do that by putting a huge amount of more people on the critical operations and by redoing the schedule every week."
While most of the extra work has not postponed the schedule, two problems have meant it has been necessary to warm sectors, causing unavoidable delay. "At the end of May we found a problem with the copper stabilized busbars, and there was no way round that, and we had to warm Sector 4-5". At present, all the eight sectors have undergone resistance measurements to check for any imperfection. These crucial measurements will also determine the energy that it is safe to run at (see press release). More recently, two leaks were found in the insulating vacuum that required partial warming up of two other sectors.
Even with the repairs completed, planning the restart schedule will continue to be a very complex task, especially when you consider that a powering test in just one of the LHC’s eight sectors actually prevents work on half of the entire ring. "Obviously we want to restart as early as possible," says Myers. "However, we can’t afford to take any shortcuts with a machine like this, or the price will be very high."