Voir en


LHC Report: Another run is over and LS2 has just begun…


Screen shot of the LHC page 1 showing the dump of the last beam of the LHC Run 2 on Monday 3 December 2018 (Image: CERN)

The CERN accelerator complex is being switched off step-by-step in order to prepare for Long Shutdown 2 (LS2). On 12 November, all protons were stopped in the LHC and throughout the accelerator complex. On Monday, 3 December, the LHC lead-ion run ended and, finally, on Monday, 10 December, the lead ions for fixed target physics were also stopped, signalling the start of LS2.

In view of the restart of the LHC in 2021 possibly at 7 TeV instead of 6.5 TeV per beam, a magnet test and training campaign was scheduled last week. The aim of this test campaign was to get a better idea of how long it takes and how many training quenches are required to train all the superconducting magnets in the LHC so that they can sustain the magnetic field required to collide the beam at 7 TeV per beam. For the dipole magnets, this field corresponds to 8.33 Tesla. Arc 1-2, approximately 3 km of the machine connecting Point 1 (ATLAS) with Point 2 (ALICE), was chosen for this campaign. Unfortunately, a thunderstorm in the evening of Monday, 3 December caused a major power cut at CERN, delaying this magnet test campaign by two days. To make up for this lost time the test campaign has been extended from Monday morning 10 December at 6 am to Wednesday 12 December at 6 am.

The LHC cryogenic status display, clearly showing the magnet test/training campaign. The upper part indicates the readiness of the different parts of the machine in terms of cryogenics. The lower part shows the temperature evolution of the different parts of the machine as a function of time. For example, the blue curve indicates the temperature in arc 1-2. The sudden rise indicates a magnet quench, while the slow descent shows the recovery time. (Image: CERN)

Looking back on 2018, one can only conclude that it was a successful year. The target integrated luminosity of 60 fb-1 for the ATLAS and CMS proton run was reached, and even exceeded by 10%, resulting in a total integrated luminosity during Run 2 (2015 – 2018) of 160 fb-1, and of 189 fb-1 since the start of LHC physics. The integrated luminosities for the proton run of LHCb and ALICE, for which luminosity levelling is applied, were 2.5 fb-1 and 27.3 pb-1 respectively. The fourth lead-ion run since the start of the LHC was challenging, but, here again, the goals were reached: 1.8 nb-1 was integrated for each of ATLAS and CMS, 0.9 nb-1 for ALICE and 0.24 nb-1 for LHCb.

In addition, important steps have been made towards the High-Luminosity LHC (HL-LHC): during re-commissioning in April, the new Achromatic Telescopic Squeezing (ATS) optics, developed for the HL-LHC, were deployed, allowing for smaller β* and hence higher peak luminosities. The value for β* at ATLAS and CMS in the LHC design report was 80 cm, while in 2018 the LHC ran with a β* of 30 cm and at the end of each fill even 25 cm. For the HL-LHC, a β* of 15 cm or even 10 cm is planned.

In addition to the techniques of levelling through beam separation, used for ALICE and LHCb, and the change of crossing angle, which were both deployed in previous years, levelling through β* was deployed operationally in 2018. The purpose of this type of levelling is that, at the start of collisions, when the peak luminosity is too high for the experiments, the beam size is increased while, later, when the beam density or brightness decreases due to the collisions, the beam size is reduced. All these techniques aim at reducing, in a controlled manner, the cross section of the beam encounters and limiting the pile-up of physics events in the detectors.

The re-commissioning of the machine and beam were remarkable in 2018. The initial schedule, based on past experience, allowed five weeks from first beam in the machine to physics with 1200 bunches per beam, which is when luminosity production starts to become significant. Thanks to the high machine availability during this period, 1200 bunches per beam were reached on 28 April. Collisions with the full machine – 2556 bunches per beam – were established on 5 May, thirteen days ahead of schedule.

In 2017, the LHC performance was hampered by the so-called 16L2 issue: frozen air in an interconnection between magnets in the arc connecting Point 1 (ATLAS) with Point 2 (ALICE). Despite the work carried out during the Year End Technical Stop, when the arc was warmed up to 100 K, not all the gas had been eliminated and, in 2018, the beam was dumped several times due to losses induced by the remaining, probably minuscule, amount of ice in 16L2. Fortunately, this allowed running with Bunch Compression, Merging and Splitting (BCMS) beam instead of the 8b4e (8 bunches and 4 empty buckets) beam that was used in 2017. However, to avoid these 16L2-induced dumps, the bunch intensity was not pushed beyond 1.15 × 1011 protons, the design value for the LHC bunch intensity.

Preliminary analysis of the 2018 availability statistics shows that 49% of the time, beams were colliding and luminosity was being produced. Equipment failures and other faults account for 24% of the time, while the remainder, 26%, was spent on “operation”, i.e. recovering from beam dumps, preparing the machine, injecting, accelerating, squeezing and adjusting the beams.

The entire accelerator complex is now in shutdown. The majority of the LHC helium inventory will be moved to the surface before the Christmas break and, as of January, people rather than particles will be running around the machine to perform all the planned maintenance and upgrade activities to prepare the machine for Run 3.