Safety Integrity Level Design Method And Application With Data Uncertainty

Safety Instrumented Systems(SIS)is an important category of safety-related systems and is widely used in petrochemical and other process industries.Historical failure data may be weak and parameters may be subject to uncertainty,since SIS are high reliable systems.Those facts may lead to an inaccure SIL level.Finding quality reliability data is a critical step of evaluation.The traditional simplified model introduces a variety of uncertain factors because of parameter hypothesis.Inaccurate even erroneous SIS reliability and usability evaluation results will have a great impact on the actual safety of SIS and have an important impact on the accuracy of the risk decision results.However,the influence of uncertainty factors is often neglected,which reveals lack of minimizing the impact of uncertainty.Based on the existing methods of safety instrumentation system evaluation,this paper takes typical safety instrument control system of process industry as the research object,proposes safety integrity level determination and verification models and supporting software toolkits under uncertainties.Methods of reducing the influence of uncertainty are supporsed by summarizing the uncertain factors,and the influence of different parameters on the average demand failure probability is analyzed.Based on the method of IEC 61508 risk graph method and protective layer analysis method,combined with the scene,safety integrity model of the improved risk graph method is established under uncertainty,while results of this method is compared with the traditional method.The importance measure method of failure rate and diagnostic coverage is approcimated by finite difference.Combining the probability model and the fuzzy mathematics theory,the methods of different types of data uncertainty of small sample or non-data corresponding to safety integrity evaluation methods are studied.Based on the ISA-TR 84.00.02 STR calculation method,the STR calculation model is established under uncertainty.Assistance software is established under the Matlab environment based on IEC 61508 and IEC 61511,using classical fuzzy method and Monte Carlo simulation as the mechanism.The results show that the proposed methods can complete the verification of the safety integrity under uncertainty.The assistance function evaluation software of the safety instrument system can complete the calculation of the average demand failure probability and the STR,which provides technical support for safety integrity assessment under uncertainty.