Predictive maintenance 

PhD Candidate Ewa Laskowska

Main Supervisor Jørn Vatn

Sponsor: Lundin Energy Norway AS

Modelling of degradation of safety valves (ESVs). The methodology is based on stochastic processes such as the Markov process or Wiener process. The idea is to enable prediction of Remaining Useful Lifetime (RUL) distribution, given the current degradation level of valve is known. Next, the obtained RUL can be used for optimization of inspection regimes. Degradation models assuming constant inspection intervals have been developed. The aim is to extend the modelling framework to treat the situation where inspection intervals depend on the current condition and the predicted RUL. Outcome of the project: Stochastic degradation models enabling RUL prediction and real-time optimization models taking condition and operational
constraints into account.

Project result: Optimization of testing strategy for Emergency Shutdown Valves 
A study case where condition-based testing allows to reduce the number of tests and repairs without increasing the probability of failure on demand