RELIABILITY DESIGN OF INDUSTRIAL PLANTS USING PETRI NETS

Purpose – The paper aims to develop a new methodology for reliability modelling able to use the data collected in an FMECA analysis to simulate the reliability behaviour of a complex system and the effects of different maintenance policies. Design/methodology/approach – A new methodology has been developed, that combines the FMECA analysis with the Petri nets, oriented graphs able to simulate the behaviour of complex systems with concurrent events. This approach was tested on a case study in API oil refinery in Falconara Marittima (AN, Italy). Findings – By using data collected during an FMECA analysis as input for a Petri net, it is possible to obtain important reliability parameters that describe the behaviour of the system, such as MTTR, MTBF, availability of the plant/machine. It is possible to perform what-if analysis observing how the reliability parameters change depending on changes in preventive maintenance policies. Research limitations/implications – In the case study proposed it was not necessary to simulate on condition maintenance. Even if the approach here described is able to do that, it would be interesting to evaluate the methodology presented also in other case studies. The approach described can be used as a starting point to develop techniques of optimization of preventive and on condition maintenance. Practical implications – The application of the methodology of a case study demonstrated its practical effectiveness, obtaining data very close to experimental ones. This approach is able to provide useful data that help maintenance managers in scheduling maintenance activities and assessing their effect on the behaviour of the whole system. Originality/value – The methodology presented shows a new solution for reliability modelling of complex industrial systems and represents the first combination of Petri nets with FMECA technique.