Modeling for evaluation of safety instrumented systems with heterogeneous components

Abstract

A novel methodology for evaluation of safety instrumented systems (SISs) with heterogeneous components based on multi-stage dynamic Bayesian networks (MDBNs) is proposed. The unified evaluation model for M-out-of-N architectures based on MDBNs is constructed. An automatic modeling method of conditional probability tables for time-slice and inter-slice transitions is established. Average probability of failure on demand (PFDavg) and probability of failing safety (PFS) of the structure can be calculated by the established model. A method for calculating the life-cycle cost of SISs is presented by using the results of system evaluation and specific cost details. A high-integrity pressure protection system (HIPPS) is adopted to demonstrate the application of the proposed approach. A MDBNs-based evaluation and cost analysis software of heterogeneous redundant structure is developed by using MATLAB graphical user interface. The analysis of system evaluation results and life-cycle cost can help the designer determine the configuration of the heterogeneous redundant structure.