Fault Tracking of the Superconducting Magnet System at the CERN Large Hadron Collider

The Large Hadron Collider (LHC) at CERN is one of the most complex machines ever built. It is used to explore the mysteries of the universe by reproducing conditions of the big bang. High energy particles are collide in particle detectors and as a result of the collision process secondary particles are created. New particles could be discovered during this process. The operation of such a machine is not straightforward and is subject to many different types of failures. A model of LHC operation needs to be defined in order to understand the impact of the various failures on availability. As an example a typical operational cycle is described: the beams are first injected, then accelerated, and finally brought into collisions. Under nominal conditions, beams should be in collision (so-called ‘stable beams’ period) for about 10 hours and then extracted onto a beam dump block. In case of a failure, the Machine Protection Systems ensure safe extraction of the beams. From the experience in LHC Run 1 (2009 - 2012) about 70 % of the beam fills were ended prematurely by a failure, thus reducing the average time of stable beams to less than 6 hours. To better understand the failures and the downtime, a fault-tracking project was started to develop strategies for coherent documentation of LHC failures. In this thesis failures of the LHC superconducting Magnet System are analysed, based on data from the intervention database of the years 2012 and 2015. The main focus is set on coherency improvements of the database entries and statistical results in form of histograms for failure analyses. Further goal is to identify key dependencies between failures to spot recurrent failure patterns and assess the impact of cumulative effects (e.g. radiation to electronics). Additionally, availability targets for a subsystem of the LHC Magnet Protection System are determined, based on financial budget assumptions.