An Enhanced Vesely Approach for Common Cause Failure Analysis in Non-Identical M-out-of-N Voting Systems

Abstract

Traditional methods for common cause failure (CCF) analysis, such as the Vesely approach, assume uniform failure probabilities across all channels, limiting their applicability to safety instrumented systems (SIS) with diverse channel designs. This study develops an enhanced Vesely method, extending the classical Vesely approach to incorporate per-channel beta factors (βi) for non-identical M-out-of-N voting systems. A generalized formula for the effective beta factor (βeff) is derived and validated through case studies, achieving reductions in βeff ranging from 4% to 28% across diverse system configurations, with the method’s benefits scaling with system complexity. Robustness is assessed through sensitivity and uncertainty analyses, while log-transformations ensure computational efficiency for large systems. The method aligns with IEC 61508-6, IEC 61511, and ISO/TR 12489, offering a practical tool for reliability engineers to improve safety assessments and reduce costs in SIS design. Validated through three comprehensive case studies spanning baseline (N=3, M=2), high-redundancy (N=4, M=3), and large-scale (N=12, M=4) configurations.