Fuzzy logic-based safety integrity level determination for multi-initiating-event scenarios: a case study of an ammonia primary reformer
The conventional Risk Graph method, widely used by technology licensors for Safety Integrity Level (SIL) allocation, often yields ambiguous or erroneous results when applied to scenarios involving multiple independent Initiating Events (IEs). This study validates a quantitative fuzzy Risk Graph framework through an industrial case study, demonstrating its ability to correct specific miscalculations inherent in traditional licensor assessments. Probabilistic input intervals are fuzzified and processed via a Mamdani inference system calibrated to reproduce conventional SIL assignments at category centers while explicitly aggregating multi-IE Demand Rates. Applied to 13 Safety Instrumented Functions (SIF) in a 3000-TPD ammonia plant’s synthesis-gas unit, the framework was benchmarked against the original licensor’s baseline. Results show that the conventional method misallocated SIL in four critical cases: underestimating SIL from 2 to 3 for two SIFs exposed to seven and eight IEs. These findings confirm that the proposed fuzzy approach provides a more accurate, discriminative, and uncertainty-aware basis for SIL determination in complex multi-IE scenarios where traditional qualitative methods fail.
Authors
- Mokhtar Shasadeghi (ORCID: https://orcid.org/0000-0001-5730-277X)
- Behrouz Safarinejadian (ORCID: https://orcid.org/0000-0001-5442-5600)
- Mohammad Sadegh Rezaei (ORCID: https://orcid.org/0000-0002-7619-613X)
Institutions
- Shiraz University of Technology (IR)
Publication Details
- Journal
- Scientific Reports
- Published
- 2026-11-19
- DOI
- https://doi.org/10.1038/s41598-026-80643-6
- Primary Topic
- Risk and Safety Analysis
- Type
- article
- Field-Weighted Citation Impact
- 0.00