Voir en


LS1 report: Full-size security

Last week several CERN teams were involved in creating a life-size mock helium leak in Sector 4-5


Last week several CERN teams were involved in creating a life-size mock helium leak in Sector 4-5. The goals of the exercise were to measure how quickly the helium cloud spread, the change in temperature and the level of oxygen in various parts of the tunnel at different points in time, as well as to check the effects of the helium on the machine and the infrastructures and, above all, to re-evaluate the safety standards in force in the tunnel.

"The safety valves designed to release helium in the event of excess pressure are currently located within a 6-metre "no-stopping area" (3 metres on either side of the valves)," explains Johan Bremer of the TE-CRG group, who was in charge of the exercise. "People working in the tunnel can move through this area but are not authorised to stop. The idea is to avoid anyone being too close to the valves in the event of a leak." However, the results of recent simulations of this safety area appear to be somewhat at odds with those of the first simulations performed a few years ago. "That's why we decided, with the backing of the Management, to perform a real test to allow us to define the optimum safety rules," explains Bremer.

As it was obviously out of the question to produce a helium leak in the machine itself, the TE-CRG team performed the test by connecting two 500-litre liquid helium containers to a specially made safety valve on the accelerator. The two containers were emptied through the valve at a rate of around 1 kg per second, which corresponds to the maximum expected escape rate in the event of an accident.

With the ventilation system operating as normal to create an airflow, several sensors placed at different points of the tunnel - a maximum of 100 metres upstream and 200 metres downstream of the leak - ensured that nothing was missed. "We were able to follow the operation live from one of the CMS caverns using cameras placed inside the tunnel," explains Bremer. "We worked closely with the EN-CV and GS-ASE groups on the test, as well as with representatives of the Wroclaw University of Technology (Poland), who helped us to set up part of the measuring system."

The tests, which were carried out on the nights of Tuesday 4, Thursday 6 and Friday 7 February, were also closely monitored by the CERN Fire Brigade, which took measurements of the oxygen level in the tunnel at various stages of the leak. "For the last test, the firemen remained inside the tunnel, 200 metres from the leak," adds Bremer. "That enabled them to take measurements right at the start of the leak, which will give us valuable input for our study."

The data collected will help to provide a better understanding of the risks associated with a large-scale liquid helium leak and thus to re-evaluate the safety standards.  The results will be available within the next few weeks. Stay tuned.

Meanwhile, elsewhere…

At the LHC, the SMACC project is progressing well and the delay has been reduced from three to two weeks since last October. 80% of the interconnections have already been consolidated, and 85% of the 27,000 shunts have been installed.

The electrical feedbox (DFBA) at Point 6 (right), which was undergoing repairs above ground until last week, has been reinstalled and reconnected in the tunnel. The operation, although highly delicate, was a complete success. Meanwhile, the DFBA at Point 8 has just been lowered back into the tunnel. It will be reconnected within the next few days.

In Sector 6-7, the first pressure tests have been followed by the ELQA tests at ambient temperature, which have proven very positive. Only two slight non-conformities have been identified on the instrumentation of the warm part of the accelerator, which should not pose major problems.

The tension is mounting as the date of the final closure of access to the pre-injectors at the end of March approaches. The teams are currently working flat out to complete their work on time for the restart.