Although the LHC, the SPS North Area and part of the PS East Area have transitioned to lead-ion operation, proton and antiproton beams remain essential for several facilities, including ISOLDE, n_TOF, part of the PS East Area and the AD complex. As in every end-of-run period, the scientists want to make the most of the little beam time remaining, and every proton, antiproton and lead ion seem to count.
Maintaining high beam availability during these final days of the run is therefore crucial, but this is no mean feat.
One set of limitations is intrinsic to the beam production itself. The filling of the LHC with lead ions is a demanding process for the injector chain and has a direct impact on the beam availability for fixed-target physics experiments. The SPS cycle required for LHC lead-ion production is 57.6 s long and requires 14 injections from the PS. Unlike proton operation, where LHC filling can be interleaved with beam delivery to other SPS users, lead-ion filling is performed in dedicated mode, meaning that no beam can be delivered to the SPS North Area during the LHC filling period.
At the PS, preparing the beam for the SPS requires 14 consecutive cycles of 2.4 s, resulting in more than half of the PS supercycle being allocated to LHC lead-ion production. Consequently, less than half of the supercycle remains available for other facilities downstream of the PS such as n_TOF, the PS East Area and the AD.
In addition, filling the LHC with lead ions typically takes around one hour, and the stable-beam duration (the time the lead-ion beam is in collision) is about six hours, compared to roughly ten hours for proton operation. This results in more frequent LHC refills, each of which is longer than during proton running.
The second limiting factor is the availability and reliability of the individual accelerator systems and the technical infrastructure, which have generally been very good and comparable to last year. However, the end of last week was marked by several uncorrelated failures that caused significant downtime across multiple accelerators in the complex, resulting in a rather low LHC stable-beam ratio and extended periods without beam for the fixed-target experiments.
Beam availability from the injectors remained good until 21 November, when a series of unrelated incidents led to substantial downtime. All beam production in the PS complex was halted around noon following a blackout of the PS complex access control and safety systems, triggered by a PLC (programmable logic controller) that failed. By design, this safety system places all the machines in a safe state in the event of a failure, meaning no beam production, all safety elements activated and no access to the machines. The experts were able to repair the system and re-establish conditions for beam production only at 7.20 p.m.
In the meantime, at around 3.30 p.m., a major power cut occurred on the Meyrin site, caused by the opening of a circuit breaker in the 66 kV supply of a Meyrin substation feeding the accelerators. In addition to generating a large number of equipment alarms in the injector chain, the power cut also stopped the cooling of the ATLAS toroid magnet, triggering a slow abort and requiring almost 20 hours of cryogenic recovery. After intensive work by the experts, power to the substation was restored around 5.20 p.m.
Although the access-system issue had not yet been resolved at that time, the Operations teams were able to restart all machine equipment affected by the power cut. Uncontrolled power cuts always cause collateral effects, considerably increasing the time required to bring the machines back to an operational state.
The PS complex managed to resume beam operation at around 10.20 p.m., but only for proton beams. In the meantime, the LEIR injection system had developed a fault that required an intervention by the on-call expert, preventing the production of lead-ion beams.
In the early morning of 22 November, a fire alarm in the SPS brought beam production to a halt. Fortunately, this was traced to a faulty sensor rather than an actual fire. Later the same day, at around 4.30 p.m., the Technical Infrastructure team in the CERN Control Centre received an ODH (oxygen deficiency hazard) alarm in the LHC arc 1-2, which prompted an intervention by the CERN Fire and Rescue Service. They identified a defective sensor and, thankfully, confirmed that no helium leak was present.
Finally, all the accelerators in the chain were fully operational again, and beam was restored in the LHC. The lead-ion beam, with 1240 bunches, was brought back into collisions at 3.00 a.m. on 23 November. The total LHC downtime had been approximately 38 hours.
Despite the low stable-beam ratio of 34% during week 47, lead-ion luminosity production remains on track to meet the targets set for the 2025 lead-ion run. This is largely thanks to the injector chain delivering higher-than-nominal bunch intensities.
If beam availability and performance remain good during the final week of the 2025 run, we will be able to declare not only the proton run but also the lead-ion run a success. But it is still too early to celebrate, so let’s keep the champagne on ice until the targets are fully reached.