After almost two years of shutdown, all planned upgrades of CMS have been performed. The endcap muon system now has a complete forth layer with additional Cathode Strip Chambers (CSC) and Restive Plate Chambers (RPC), enclosed by two new 120t shielding disks made as a sandwich of steel and a special heavy concrete containing borate and hematite for efficient neutron absorption.

The entire muon system in the barrel and the endcaps underwent a major overhaul, in which parts of the trigger system of the Drift Tube Chambers in the barrel were moved to the service cavern for easier maintenance. The RPCs in the barrel underwent a laborious campaign to fix gas leaks, helping reduce emission of climate active gases and costs. To recover the efficiency as close as possible to 100%, electronics of many chambers in all muon systems were repaired or exchanged. The first layer of the endcap CSCs closest to the IP were removed, brought to the surface for an upgrade of the frontend electronic, and re-installed. In the outer part of the hadronic calorimeter, the photo detectors were replaced with Silicon Photomultiplier because the original detectors develop lifetime problems in the magnetic stray field of the solenoid.

In the center of CMS, the central beam pipe was exchanged against a new one with a smaller outer diameter of 45mm, in preparation for the installation of a new 4 layer Silicon Pixel Tracker during the year-end technical stop 2016/17. In the summer of 2014, after the beam pipe was installed [Fig.1], baked out, and found leak tight, the remaining programme looked quite clear and straightforward.

The current Pixel Tracker, which was removed for the beam pipe replacement and minor overhaul, was planned to be re-installed in early October, followed by the final closure of CMS. As last step a crucial magnet test was foreseen to ensure that the enormous effort put into the thermal insulation of the Silicon Trackers allowing cold operation without condensation had been successful. Instead, in early August, a severe fault in the barrel pixel system was reported, with 25% of the modules in one quadrant (~50) being found not responding. By the end of August, the half-shell containing the faulty quadrant was brought to PSI for detailed investigation. In the beginning of September, the diagnostics revealed that electro-migration between adjacent bond pads of the High Density Interconnect was the reason for the failure.

On September 18, the repair strategy of repairing some faulty modules but replacing the bulk of them with new ones was presented. The schedule agreed foresaw the detector being back at CERN and ready for installation by the end of November, which is almost exactly 2 months later than originally planned.

At this stage, it was decided to radically change the end-game sequence of the shutdown, not waiting for the barrel pixel repair, but rather closing CMS immediately and conducting a short, pre-emptive magnet test. The idea was to clear risk, which otherwise would not have appeared until the final closure for beam operation. After finishing the remaining work on the bulkhead seal, necessary for the tracker being operated cold, this closing, magnet test, and re-opening sequence became the critical path for two months, with remaining activity on the endcaps and the HF not essential for the magnet test being postponed.

The new sequence implied unexpected tight deadlines for several teams, particularly those responsible for the magnet operation and the forward region maintenance, and a massive additional workload for our heavy movement team. The additional closing and opening sequence required 36 single movements of heavy disks and 12 heavy raiser platform insertions or removals.

A concerted and exceptional effort resulted in a full closure of both forward regions and CMS ready for the magnet test by November 6. On November 10, the magnet was at full field and kept there until November 12. The test proved that the main objective of LS1, being able to operate the tracker at about -15^oC, has been reached.

At the same day the magnet test ended, the opening of CMS began with the objective of installing the pixel tracker before the annual closure of CERN. Again, with an exceptional effort of the heavy movement team, the detector was fully open on both ends and all necessary platforms were installed by December 3. After some careful beam pipe adjustments and some mockup tests for the installation of the new pixel tracker, on December 5 the installation of the barrel pixel started, followed by connecting cables and cooling lines. On December 11, the first test showed that the detector works, giving the go ahead for the installation of the forward pixel. During the last working week of the year, the forward pixel will be connected and a fast check out programme will be carried out. In parallel, the tracker region will be sealed again. During the closure of CERN, the detector will be kept at a temperature just above the ambient dew point to be safe in case of service failures or a possibly delayed piquet response.

Installation of the new CMS central beam with a smaller diameter around the IP – Image Credits: M. Hoch

From January 5 onward, the tracker will be run cold for about two weeks allowing a careful checkout and calibration. Assuming no serious issue is found, the detector will be brought back to room temperature and the seal will be opened for the installation of the upgrade Beam Condition Monitor (BCM) and the Pixel Luminosity Tracker (PLT). Around January 26, the installation and checkout of these components will be finished, giving the start signal for the final closure, which is still in time for allowing intensive tests with and without magnetic field before the start of beam operation.

The barrel part of the CMS Pixel Tracker during installation – Image Credit: M. Hoch

To keep this report short, the major upgrade programmes on the DAQ, the trigger, and the infrastructure at PT5, which are all coming to a successful end, cannot be described in any detail.

With all these upgrades and improvements, CMS will be perfectly prepared for the exciting times ahead of us, with the LHC providing collisions of unprecedented energy and luminosity.

Though, for CMS, LS1 is not over yet, there is clearly light at the end of the tunnel.