Since the restart of the LHC for its third operational run in 2022 (Run 3), the machine has been routinely operated with higher-brightness beams following the successful deployment of the LHC Injector Upgrade Project during Long Shutdown 2 (LS2). At SPS extraction, the Run 3 parameters required by the LHC have been fully realised, while further work and improvement is still required during the remainder of the run to achieve the nominal HL-LHC brightness at LHC injection. The successful deployment of the first HL-LHC deliverables, namely the upgraded injection kickers (MKI cool) and new injection absorbers (TDIS), have played a major role in safely injecting these increased-brightness beams into the LHC, gathering valuable first experience before completion of their series deployments in the year-end technical stop (YETS) 2024/25 and Long Shutdown 3 (LS3).

At the end of 2023, the LHC had delivered 32 fb^-1 out of the 75 fb^-1 originally forecasted for the operational year, as the pp run was cut short by several long faults. These included a helium leak between the cold mass and the insulation vacuum in the inner triplet left of the LHCb experiment, issues with the overpressure burst disks of the RF accelerating cavities, an RF finger issue in the warm vacuum module on the left side of ATLAS (which led to the decision to limit the bunch intensity to 1.6x10¹¹ ppb for the rest of the year instead of the targeted 1.8x10¹¹ppb), and two leaks in the TDIS IR8 bellow, which led to the cancellation of the pp reference run.

Despite these major faults, machine performance during stable periods was excellent, achieving stable beams 50% of the time and reaching rates of 0.8 fb^-1 accumulated in 24h, with a record of 1.2 fb^-1/24h. In addition, very advanced levelling schemes were developed and tested, and were actually even more complex than what is presently planned for the HL-LHC baseline! Offsets at the collision points, ß* levelling, and changes of gaps and centres of the local collimators around ATLAS and CMS were orchestrated together to provide the desired pile-up at these experiments. This new scheme was commissioned and deployed smoothly.

Run 3 also marks a new era for the LHC as an ion collider since all foreseen HL-LHC upgrades for ion operation are now implemented, including upgraded ion beams and slip-stacking in the injectors and the full WP5 ion collimation hardware upgrades. The targeted stored beam energy for lead ion beams is 20.5 MJ, i.e. more than a factor of 5 larger than the LHC design target and almost a factor of 2 larger than what was achieved in Run 2. This poses important challenges for the IR7 betatron cleaning system that have been addressed by the deployment of crystal collimation. In addition, an important upgrade of the ALICE experiment — the LHC’s heavy ion-dedicated experiment — took place in LS2, reaching peak luminosity 6 times larger. A machine limitation that must be overcome to achieve this important goal comes from collisional losses generated by the bound-free pair production mechanism, when colliding ions capture an electron. Their rigidity is thus unmatched to the beam’s rigidity and these ions are then lost in the downstream dispersion suppressor (DS) where the induced power load can quench the impacted magnet.

Safely disposing of this beamlet downstream of ALICE, whilst keeping beam losses below the quench limit of superconducting magnets, is only possible by adding new collimators for local protection in the cold region (see figure 1). The integration of this collimator, called TCLD, in the cold region involves specific issues related to the tight space constraints and the need to ensure a continuous cold interconnection. The TCLD design was elaborated within WP5, and the required units were successfully produced during LS2 and installed in time for the ALICE in Run 3. Figure 2 demonstrates that the new TCLDs can efficiently intercept the DS losses. The closure of the TCLD jaws to the required setting, together with a special orbit bump in this region (bottom graph), efficiently suppresses the collisional losses on the downstream cold magnets. This was demonstrated already in 2022, in a dedicated validation test where only the TCLD on the left side of IP2 was closed. This collimation solution was successfully deployed for the whole Pb run in 2023. It was demonstrated that this upgrade enables ALICE to operate at the upgrade target luminosity, a factor of 6 beyond what has previously been achieved!

The restart for proton physics in 2024 has seen several important machine configuration changes, most notably the decision of performing a triplet polarity inversion in IP1, which allows the redistribution of the radiation peaks in the most exposed triplet and warm-dipole magnets and, as such, extends their lifetime in view of soon surpassing the 300 fb^-1 integrated design luminosity of the LHC. The triplet polarity inversion, however, requires that the individual quadrupoles of Q4 remain unpowered, which was shown to increase the muon flux to the FASER forward physics experiment. This requires additional radiation studies/simulations for the LHC and HL-LHC optics to unambiguously identify, and eventually mitigate, the origin of these muons.

A rebalancing of the refrigerators between Sector 78 and 81 was also adopted, aiming to enable operation with 5x36b trains of up to 1.6e¹¹ ppb without the need to mix with 8b4e beams in the 2024 run. This was, however, not fully successful and operation is today still limited in using 3x36b trains, and the electron cloud and available cryogenic capacity will very likely remain the main performance limitation until the end of Run 3 in 2025. The adoption of the Beam Screen Treatment project, which consists of applying an in-situ aC coating to around 20% of the main dipole beam screens of the LHC, has been integrated as a baseline activity during Long Shutdown 3 to allow – in the absence of further major degradation of non-treated parts of the machine – reaching the target beam intensities of 2.3e¹¹ ppb at injection during the HL-LHC operational era.

In pushing machine configuration and performance further, new operational challenges have emerged in 2024, for which additional work will be required as some of these might otherwise impact the HL-LHC performance reach. Increased losses at the start of the ramp, likely created by the capture inefficiency between the SPS and the LHC and the creation of off-momentum particles, are being investigated but will most certainly require an upgrade of the RF power injection through the installation of high-efficiency klystrons during LS3 to ensure a sufficient operational margin to efficiently capture the beams with nominal HL-LHC beam parameters. Increased levels of losses have also been observed when bringing the beams into collision and at the start of physics, the origin of which is still to be precisely determined but is hinting today towards resonances near the present collision tunes. Lastly, a loss of the collimation hierarchy has been observed when pushing ß* below 30cm. The origin of this hierarchy breakage has been confirmed as off-momentum halo particles. For the time being, sufficient operational margin for reliable and safe operation at ß* at 30 cm can be reestablished through a tuning of operational chromaticity and dispersion knobs, while the root cause and the generation process of these off-momentum particles at flat top still remains to be fully understood and mitigated.

The 2024 run has also seen a performance comparison between the standard and BCMS schemes, demonstrating an 8% performance gain of the latter thanks to the injection of higher brightness that can be partially preserved all the way into collisions. BCMS beams will therefore be retained as the baseline for the remainder of the run. With the machine just going into its first 1-week-long technical stop of the year, close to 30 fb^-1 have already been collected in this year’s proton run (see figure 3), in line with the predicted target of collecting 120 fb^-1 by the end of this year. The remainder of the operational year will also still see a multitude of machine development studies being performed to validate the full HL-LHC baseline including, amongst others, quench tests, the validation of beam instrumentation technologies, and the final HL-LHC optics, halo, and collimation studies.