The Research And Application For Failure Probabaility Of Risk-based Inspection Method For Atmospheric Crude Oil Tanks

With the continuous development of petroehemical enterprises,the demand for crude oil keep rising,and the number of atmospheric crude oil storage tanks in China is also increasing exponentially.At the same time,the physical characteristics of crude oil,tank structure,operating conditions and other factors make the storage tank inherently dangerous,and the failure probability of atmospheric crude oil storage tank is greatly increased due to insufficient experience of personnel,loose management and other factors.Risk Based Inspection(RBI)is used to assess the risk of atmospheric crude oil storage tanks,which can effectively realize the rational allocation of inspection resources and reduce the risk level.However,in risk-based inspection methods,there are still some problems in the calculation of failure probability,such as the large gap between failure mechanism and foreign countries,the lack of failure data and the less consideration of original defects.Therefore,in view of the above problems and in accordance with the development trend of large data,dynamic risk assessment of failure mechanism of atmospheric crude oil storage tank is carried out,and the general equipment failure probability factor and equipment correction coefficient factor in failure probability are modified to make the analysis result of failure probability more accurate.Firstly,the risk-based testing method was introduced.The research status at home and abroad was summarized and analyzed,and the key problems existing in the application of RBI in atmospheric crude oil tanks in China were summarized.The basic concepts,theoretical support,evaluation process and so on were understood.Then,the structural characteristics of several types of storage tanks were analyzed,and their failure forms and failure mechanisms were sorted out.By comparing the advantages and disadvantages of the accident tree,event tree and bow model,a dynamic analysis model combining bow model with Bayesian network was proposed to analyze the leakage events of a storage tank,effectively identified the high-risk factors and the most likely paths of accidents,and realized dynamic risk assessment by modifying Bayesian network model for the future,laid the foundation for the future application of big data.Secondly,in view of the large gap between China and foreign countries in design and manufacturing,and the inaccuracy of failure data obtained directly from foreign standards,this paper put forward the treatment of failure data in China.By determining the content of failure data collection in China,and comparing the advantages and disadvantages of the statistical model,we chose Weibull analysis method to do failure data statistics.After listing the detailed steps of the statistical model,we determine the sample space.By comparing the statistical results of tank accidents and failure data in recent years in China,it was found that the statistical methods were in line with Chinese national conditions and scientific.Finally,a multi-level grey comprehensive evaluation method combining analytic hierarchy process(AHP)and grey comprehensive evaluation model was proposed.Firstly,a mathematical model of failure probability for atmospheric crude oil storage tanks was established based on risk-based inspection(RBI)failure probability evaluation process.Secondly,on the basis of coefficient correction method,equipment correction factors were divided into first-level target layer A and second-level factor layer B.At the third index level C,the weight of each evaluation index was calculated by AHP method.Then,the grey comprehensive evaluation model was introduced to evaluate each index.Finally,the multi-level grey comprehensive evaluation method was eapplied to practical cases.The results showed that when dealing with multi-factor and multi-level comprehensive evaluation,this method can avoid the risk masking problem which may be caused by traditional scoring method.At the same time,it can get the rarking of the influence of each factor on the correction coefficient of the required equipment,and understand the importance of the in:fluence factors according to the results,so as to take corresponding measures.