Research On Proof Testing And Its Effect On The Reliability Calculation Of Safety Instrumented System

The trend of enlargement of modern petrochemical industry demands more andmore severe safety requirement for operation and controlling of process plant. Since mostmaterials in petrochemical are flammable and explosive, severe loss (health, economicand environment) was occurred occasionally due to maloperation of plants. As one ofmain protection measures for process safety, the reliability of safety instrumented systemis of great importance.In this thesis, the technique of the calculation of the safety levels of the safetyinstrumented system is put forward with the concept of the safe failure. The program iscompiled to achieve the results according to the process of the theory calculation, themethod of markov process is used to calculate the reliability index of the repairable safetyinstrumented system. The reliability designs of the petrochemical plant and the safetyinstrumented system, especially the design of the proof test interval are the mostimportant aspect in the design of the safety of the process to ensure that the systemreliability index reaches the requirement of the international standard IEC61511 as wellas the benefit of the economic target.In this paper, the method of markov process is mainly discussed to evaluate thesafety integrity of the safety instrumented system. The failures are classified by thestructure of the instruments in the safety instrumented system in order to calculate thefailure rates which are the basis of the calculation of the reliability index of the safety instrumented system. The proof test interval is determined considering the economicproblem and the safety integrity which is based on the calculations of the reliability indexin different proof test interval. In this thesis, the method of failure modes and effectanalysis (FMEA) is used to analyze the potential failure modes of the product and theeffects to the product. Every potential failure is classified by the severity it caused and thefrequency it happened. Then, the questionnaire about the product is designed toinvestigate the customers to get the data of failure modes and the failure rates of everyfailure modes are calculated by the data. In this paper, evaluation method of invalid ratesfor the exponential distributions are proposed, based on the exact Bayes estimatorsconsidering the combination of the literature data and test data. The literature data and thetest data are fully combined by the proposed method, which improves the accuracy of theestimation of the invalid rates. The results show that the proposed method is of goodpracticability and guidance in practice. At last, the safety integrity of the safetyinstrumented system of the hydrocracker is acquired by the method ofmarkov process.