Abstract
| The Cryogenic system is one of the most critical component of CERN accelerators
and associated experiments. Any improvement in the maintenance plan leads to smoother
operation procedures and improves the reliability of the facility as a whole. To reduce the
recovery time after failure, a tool to quicken the identification of potential fault signatures has
been developed. It consists of dynamic models realized with EcosimPro™ and its associated
cryogenic library, CRYOLIB™, which are compared with process data. This comparison spots
potential failures by showing deviations on residues identified on critical variables. The
comparisons, that can be done both online and offline, will allow either the operation team to
take early mitigating actions ahead of the failure occurrence or to identify maintenance
consolidations to be implemented during the technical shutdowns. This contribution will
illustrate the method with several case studies, focusing on turbines, along with some examples
illustrating the actual limits of the tool and next steps for further development and
implementation. |