In challenging times, it’s reassuring to see CERN’s accelerator complex fully up and running again, with physics being delivered to the experiments at ISOLDE and HIE-ISOLDE, n_TOF, AD-ELENA, the East Area, the North Area, AWAKE, HiRadMat, CLEAR and, of course, the LHC – the current temporary unscheduled stop notwithstanding – and great work being done with test beams and at the irradiation facilities.
On the LHC side, following extensive recommissioning with beam, the first collisions with the detectors on were produced the day after we celebrated the 10th anniversary of the discovery of the Higgs boson. The first stable beams were followed by a period of interleaved commissioning and intensity ramp-up. Every year, the number of bunches per beam is carefully increased in stages, with sign-off by the Machine Protection Panel after a designated length of time/number of fills at a given configuration. This year, the LHC ramped up from 72 to 315, 603, 987, 1227, 1551, 1935, 2173 and then 2413 bunches per beam in the space of five and a half weeks, with the first 1227-bunch fill taking place on 29 July, a few days ahead of schedule. Healthy progress was made, despite a familiar mix of issues along the way, and 2440 bunches were achieved by 12 August.
Experience tells us that the first year of operation with beam after a three-year shutdown has the potential to be a little rocky. The challenges foreseen included additional main dipole training quenches due to the machine now operating at 6.8 TeV, electron cloud, and unidentified falling objects (UFOs).
The vacuum team had anticipated fully deconditioned beam screens and the need to restart from scratch with an electron cloud reduction campaign. A full scrubbing programme successfully brought the initially very high electron cloud to acceptable levels, with further conditioning foreseen during the long, high-intensity physics runs. Here, the key issue is the e-cloud heat load to the cryogenics system – a real operational limit on the maximum intensity that can be handled by the LHC.
UFOs, a real bugbear in 2015, were also expected to reappear in number after LS2. This did indeed prove to be the case but, fortunately, they have conditioned down quickly and are now occurring less often. Although still a cause of occasional premature dumps, thanks to careful management of beam loss thresholds, they haven't been debilitating.
In parallel, there has been the necessary re-bedding in and debugging of extensive, complex accelerator systems. Recent availability has been moderate compared with the impressive levels achieved at the end of Run 2.
Luminosity performance has been stunning. On the back of the improvements made during the injector upgrade programme (LIU), the injectors have been delivering high-quality beam, with low transverse beam size. Well established procedures and excellent parameter control in the LHC have enabled the full potential of the beams to be exploited. For the moment, the Operations team is still working with around nominal bunch intensity, with the possibility to go significantly higher yet to be exercised. The excellent performance is testament to the continued investment in understanding, tools, machine development, accelerator physics, accelerator systems such as instrumentation and transverse feedback, as well as a lot of hard work.
Although the LHC has the potential to go significantly higher, the peak luminosity for Run 3 is limited to around 2e34 cm-2 s-1 due to the heat load from the luminosity debris, which impacts the superconducting inner triplet magnets. The luminosity is limited through transverse displacement or by varying the beam size at the interaction point. Sophisticated new operation tools have been deployed to gently reduce the beam size in stable beams (beta* levelling) in order to keep the luminosity level at its maximum value for as long as possible.
With reasonable availability and some long fills, production rates have been good, and 11 fb-1 were delivered to ATLAS by 23 August. However, when the luminosity curve points high, never extrapolate – you will anger the accelerator gods. We’d foreseen training quenches, UFOs, electron-cloud heat load and system debugging and, indeed, got caught by a big one on 23 August.
A cooling tower control problem temporarily knocked out the cryogenics at Point 4. Here, the cryogenics system cools not only the magnets but also the superconducting RF cavities. Following the incident, the liquid helium in the RF cryomodules warmed and vaporised, increasing the pressure inside the modules. This situation is foreseen and release valves are in place should the pressure rise above a certain level, carefully set to avoid damage to the RF cavities. The release valves are backed up by thin graphite “burst discs”, which are designed to open at a higher pressure than that which triggers the opening of the release valves.
On 23 August, the release valves opened as designed. Unfortunately, in the minutes that followed 3 burst discs (out of 16) opened at below their design value. A task force was already in place and had performed detailed investigations following a similar incident earlier in the year; mitigation measures had already been planned for the coming year-end technical stop.
A blown burst disc opens the modules to air, necessitating a ten-day warm-up to flush any moisture off the cavities, followed by cool-down and cavity reconditioning. The tail end of the recovery period overlaps with a planned five-day technical stop and we hope to be back in action with beam in the second half of September.
The cryogenics team has developed an energy economy mode for the LHC and is able to switch within a day to a configuration with fewer active units, saving around 9 MW. This mode is used during the beam commissioning period and ion runs, when the full cooling capacity of the system is not required. This mode was deployed immediately for the duration of the RF recovery.
Despite the RF incident, the performance of the LHC and, indeed, the whole accelerator complex is very encouraging and bodes well for a productive Run 3. That these decades-old machines (the PS is 63 this year!) and the associated facilities continue to deliver their incredible spectrum of physics at the limits of their capabilities is testament to the continuing dedication, commitment and ingenuity of everyone involved.