Time-dependent unavailability assessment of final element of safety instrumented systems- an application of multiphase markov process - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2020

Time-dependent unavailability assessment of final element of safety instrumented systems- an application of multiphase markov process

Résumé

Safety Instrumented Systems (SISs) are commonly used as a preventive barrier to avoid unwanted accidents caused by hazardous events. Simplest structure of a SIS consists of a sensor to detect anomaly of a given process parameter, a logic solver to decide appropriate actions given the nature of detected anomaly and a final element to execute necessary actions to prevent an imminent failure. As a critical safety barrier, unavailability assessment of SIS receives a great deal of attention and its safety requirement is regulated mainly by IEC 61508 standard. It demands a systematic and quantitative evaluations of SIS unavailability by some suggested methods such as Fault Tree Analysis (FTA), Reliability Block Diagrams (RBD), Markov Chains, Petri Nets, etc. which must then be compliant with the SIL (Safety Integrity Level) required for that particular application. In general, quantification of SIS unavailability adopts a convenient assumption that failure rates of the SIS elements are constant over the operational lifetime. Although such an assumption is usually justified for sensor and logic solver subsystems of a SIS, it may not be equally valid for final elements. These final elements are often located and operated in harsh environment-making them subject to degradation. Furthermore partial tests and proof tests for maintenance purpose are believed to have some negative impact on the condition of the final element. Several challenges exist when considering time-dependent failure in unavailability quantification, for instance, difficulties in detecting failures due to passive nature of low demand SISs, lack of relevant data for parameter estimation when modeling with stochastic processes, imperfectness of tests, non-periodicity in testing practice, etc. In this paper, a multiphase Markov model is proposed to model a SIS final element subject to both partial and full proof test by considering the negative impact due to tests, non-detection and non-negligible repair time. The analytical model is validated by Monte Carlo simulation and the impact of negative tests and non-detection is discussed in the light of sensitivity analysis. Finally, number of partial tests over a given mission time is optimized to show the trade off between PFDavg and number of tests.
Fichier non déposé

Dates et versions

hal-04317734 , version 1 (01-12-2023)

Identifiants

Citer

Islam Abu Md Ariful, Srivastav Himanshu, Vatn Jørn, Barros Anne, Lundteigen Mary Ann. Time-dependent unavailability assessment of final element of safety instrumented systems- an application of multiphase markov process. 29th European Safety and Reliability Conference, ESREL 2019, 2019, Hannover, France. pp.2583-2590, ⟨10.3850/978-981-11-2724-3_0515-cd⟩. ⟨hal-04317734⟩
10 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More