Choose timezone
Your profile timezone:
Powered by Indico
v3.3.3-pre
Most analytical formulas developed for PFD and PFH calculation assume a constant failure rate. This assumption does not necessarily hold for system components that are affected by wear. For the analysis of systems where constant failure rate of components is assumed, the preferred approach by industry is the use of fault tree analysis or analytical formulas developed on basis of reliability block diagrams. Examples of such analytical formulas of PFD and PFH calculation for M-out-of-N redundancy architecture are found in IEC 61508 and IEC 62061. Unfortunately, these analytical formulas assume constant failure rates, and the non-existence of similar analytical formulas for components with non-constant failure rates often results in constant failure rate being assumed also for components in the wear-out period. In this presentation we present modifications, adaptation and application of existent methods of reliability/availability assessment to PFD/PFH calculations of mixed systems (electronic/electrical/mechanical) and systems with non-constant failure rates. Such mixed systems may be used, for instance, in old mechanical systems when instead of full replacement of safety system we can apply redundancy by using electrical/electronic equipment. The first method we present is a method of analytical calculations of PFD for systems with non-constant failure rates and M-out-of-N redundancy architecture. It is based on ratio between cumulative distribution functions, and Weibull distribution for a characteristic of the system behavior. The second method demonstrates application of perturbed semi-Markov processes to PFD calculation of mixed M-out-of-N systems. And the third method is Markov method with iterations.
ATS Seminars Organisers: H. Burkhardt (BE), M. Modena (TE), T. Stora (EN)
Coffee / tea will be served after the seminar in room 30/7-012