Research On Reliability Assessment Of Petrochemical Safety Equipment Considering Different Maintenance Strategies

At the site of a petrochemical enterprise,the reliability of petrochemical safety equipment determines the safety production level of the installation.Once such equipment fails,the dynamic risk level of the installation will be affected.Therefore,it is of great significance to ensure the smooth operation of petrochemical plants to evaluate the reliability of the equipment in combination with the equipment failure data and accurately grasp the reliability of the equipment.However,the traditional reliability assessment method usually assumes that the life time of equipment is independent and distributed under the ideal condition of complete maintenance,and does not consider the impact of other maintenance activities on the reliability of the equipment.In addition,due to the limitation of external conditions such as monitoring cycle and funding,the failure time of the equipment is missing or the specific failure time can not be judged,which affects the accuracy of the equipment reliability assessment results.In view of the above problems,the main research contents of this paper are as follows:(1)For the censored life data of critical petrochemical safety equipment,traditional reliability assessment methods do not consider the effect of maintenance on time between failures.In order to avoid this shortcoming and the large error of median padding algorithm,an inverse transform interval censored data filling algorithm is proposed.Firstly,based on the minimum maintenance condition,the failure trend of the repairable system is described through the non-homogeneous poisson process,and the missing data in the monitoring interval is filled with this algorithm,which solves the difficulty of processing the interval censored data problem.Secondly,according to the Monte Carlo Expectation Maximization algorithm,the parameter estimates of the filled data are obtained,which solves the problem that the model is difficult to solve because of the unknown parameters.Finally,taking the emergency pressure relief valve on the wax oil hydrogenation unit as the case analysis object,by comparing the median filling method,the proposed algorithm can well integrate the existing data and the reliability index obtained is more accurate,and the shape parameter estimated value deviation reduced by more than 4% compared with the median padding algorithm.(2)Considering that incomplete maintenance is more in line with the actual situation of the project,the reliability assessment of petrochemical safety equipment is carried out under incomplete maintenance conditions.And a virtual service age model of Kijima I is established to quantify the restoration of the actual service age of equipment by maintenance activities effect.Through the equipment failure trend test and AIC and BIC information criteria,a suitable reliability assessment model is selected,and the non-linear constraint programming method is used to transform the estimated value of the distributed parameter under incomplete maintenance,which avoids the problem that the model parameters are difficult to solve.The model is applied to a ship unloading system of an LNG receiving station.The comparison results of the model parameters show that the Kijima I virtual service age model is better than the Weibull distribution and the NHPP model,which corresponds to the situation that incomplete maintenance activities are more actual project.(3)Based on the assessment of equipment reliability by considering information security failures,a petrochemical security equipment data collection framework that integrates functional safety and information security is proposed.Subsequently,in the reliability assessment of multi-category sample failure data,a hybrid Kijima I virtual age-of-service model is further proposed,which considers the influence of data differences under different failure causes on equipment reliability index assessment.Finally,based on the multi-category pressure transmitter failure data under this data collection framework,the reliability assessment is carried out.By comparing the reliability assessment results of the Kijima I virtual service age model with the other two types of samples,the reliability index of the hybrid Kijima I virtual service age model is more reasonable.To a certain extent,this model provides corresponding theoretical guidance for the differentiated maintenance of equipment,and avoids the transitional investment in enterprise safety costs,and helps to achieve a balance between differentiated maintenance and high equipment availability.