Study On Risk-Based Intelligent Maintenance Decision-Making For Petrochemical Rotating Machinery

The petrochemical industry plays an important role in national economic development. However, its process media are inflammable, explosive and toxic. Therefore, the failure of equipment may cause great economic losses. Furthermore, it may lead to serious personnel injury even death and environment pollution. And social disaster will be a nightmare. So it is of great significance for petrochemical industry to take effective measurements to prevent failures and improve the reliability of equipment. With the production trend of large-scale, complication and automation, dangerous media keep increasing both in species and quantity. Hence, it requires stringent regulations on equipment safety and reliability. Maintenance is extremely essential for ensuring the equipment to keep running safely, improving product quality and increasing economic benefits of plants. Nowadays, the maintenance cost takes a great proportion of the total production cost. How to reduce it and how to effectively make decisions and optimize maintenance have caused great concerns. It is an urgent problem to be solved.Compared with static equipment, the failure rate of rotating machinery is generally higher. In order to ensure process safety, stability and reliability, modern industry requires effective maintenance management in rotating machinery so that the safe and long-life running of the petrochemical plant can be achieved.Based on the kinds of advanced equipment maintenance management concepts, our particular focus is on maintenance decision-making for petrochemical rotating machinery. The studies in this dissertation include following parts.(1) Establishment of a dynamic risk-based intelligent maintenance systemA dynamic risk-based intelligent maintenance system (RBIMS) was established by using the advanced information science technology, and systems engineering methods. The maintenance decision-making is taken as a guidance of condition monitoring / inspection and monitoring / inspection has been integrated with dynamic risk assessment. Based on the results of condition monitoring and fault diagnosis, the equipment dynamics risk level is determined, and the risk early-warning is carried out. With the changes in equipment condition, the plans are keeping updating and adjusting. In this way, a dynamic closed-loop maintenance system is achieved. (2) Research on criticality evaluation method for petrochemical rotating machineryEquipment criticality evaluation is an important foundation for reasonable maintenance resources allocation. Aiming at the uncertainty, fuzzy and multivariable of the evaluation factors, an equipment criticality evaluation model based on fuzzy comprehensive evaluation is set up. Benefit from the strong learning ability and simple structure of BP neural network, a fuzzy comprehensive evaluation model with BP neural network is established. The result of case study shows that this model not only truly represents the equipment importance in the whole production process, but also realizes intelligent evaluation of the equipment criticality.(3) On Condition Maintenance (OCM) optimization study of petrochemical rotating machineryPresently, OCM is an effective predictive method for maintaining petrochemical rotating machinery. Its cost is an important part in the equipment life cycle cost. Therefore, to reduce OCM cost is an important mission in maintenance management. In this study, an OCM cost optimization model is established. First, the judgment conditions are considered as a necessary criterion of the OCM feasibility. Moreover, in order to minimize maintenance costs per unit time, the concept that there must be an interval between the potential failure point (P) and the functional failure point (F) was introduced into the optimization model of OCM. Considering the randomness, a Monte Carlo simulation program for optimizing monitoring period was developed. Finally, an optimization study case on centrifugal pump shows that compared with the existing monitoring plan, the maintenance costs saved 14,327 RMB/year. The economic benefit is obvious.(4) Study on risk-based maintenance decision-making for rotating machinery, development of analysis software and engineering applicationWith the introduction of risk assessment theory, the maintenance decision-making process of rotating machinery was studied. The risk assessment criteria are set up, which is suitable for petrochemical rotating machinery in China. The method of risk-based maintenance decision-making is established. On the basis of failure root cause analysis, radical maintenance has been proposed.Based on the research results of risk-based maintenance decision-making for rotating machinery, computer-aided analysis software has been developed. Compared with traditional manual decision-making methods, the results show that the normalization and ordering analysis of maintenance decision-making process can be achieved and this software takes advantages in efficiency and effectiveness.The risk-based maintenance decision-making method for rotating machinery has been applied in a SINOPEC ethylene plant. Failure mode analysis, risk assessment and root cause analysis had been carried out for 141 rotating machinery. 549 failure modes were obtained, including 67 high risk models (12%). Then, according to the process of maintenance decision-making, the radical maintenance and condition monitoring / inspection recommendations were proposed. The results of engineering application shows that this method could realize reasonable allocation of maintenance resources, highly consider the safety and economy, improve maintenance benefits.