Research On Prediction Method Of Corrosion And Water Side Corrosion Of Heat Exchanger Based On Big Data Technology

Since the reform and opening up,China's economy has entered a rising stage of sustained and high-speed development,people's material living standards have been changed dramatically,and the demand for oil and gas resources is increasing day by day.In order to ease the tension of China's oil reserves,the import of crude oil has continued to grow in recent years,among which there are a large number of poor crude oil.Poor crude oil has increased the burden of heat exchange equipment in petrochemical refining enterprises,resulting in more and more serious corrosion problems,among which the cold heat exchange equipment is the most affected by corrosion and is not easy to be found.Therefore,it is urgent to study the corrosion law of heat exchange equipment.At the same time,the advent of the Internet and intelligent AI era has made a qualitative leap in the informatization level of petrochemical enterprises,which has greatly changed the situation of unclear update of design status and untimely discovery of problems for a long time,and also accumulated a large number of equipment monitoring data in the production process.How to use these data reasonably and effectively to discover the potential corrosion laws It is a very valuable research topic to predict corrosion of heat exchange equipment in petrochemical and oil refining enterprises and to realize intelligent monitoring and problem warning of heat exchange system in oil refining.In this context,based on the corrosion data of multi-source heat exchanger collected on site,this paper studies the occurrence and development of corrosion of atmospheric and vacuum heat exchanger widely used in petrochemical and oil refining enterprises from three aspects of statistics,simulation analysis and corrosion prediction.The specific research contents are as follows:In the first place,the corrosion statistical analysis of atmospheric and vacuum heat exchanger is carried out by using the long-term accumulated corrosion detection data of heat exchanger in petrochemical plant.The heat exchanger can transfer part of the heat of the fluid to the cold fluid,so as to reduce the temperature of the fluid and provide an important guarantee for the working temperature of each equipment.Based on literature review and years of overhaul corrosion inspection,this paper analyzes the problems of the heat exchanger of atmospheric and vacuum distillation unit,classifies and analyzes the common corrosion morphology of the heat exchanger with statistical analysis method,obtains the corrosion parts and morphology of the heat exchanger in three aspects:low temperature corrosion,high temperature corrosion and water side corrosion,analyzes its corrosion mechanism,and puts forward corresponding suggestions and measures.Secondly,in order to study the exact location of corrosion distribution inside the tube bundle,the corrosion prone tube box and tube bundle in the heat exchanger are simulated and analyzed by CFD software.The flow pattern of the medium in the tube of the heat exchanger is simulated.The flow field of the tubular heat exchanger at different speeds is simulated by FLUENT software.The erosion cloud diagram,velocity cloud diagram and erosion particle trajectory cloud diagram of the tube box and tube bundle are obtained.The results show that the flow pattern will change greatly when the fluid flows through the pipeline components,which will lead to the erosion corrosion of the pipeline.Finally,a corrosion prediction model based on neural network and multi-source analysis data of circulating water is established to predict the corrosion rate of the circulating water pipeline in the refinery.Eight kinds of routine monitoring data are selected as the sample standard database,and KPCA is used to preprocess the original data to extract the main factors affecting the external corrosion of the pipeline,and GRNN is applied A mathematical model for the prediction of corrosion rate of pipeline is established.The effect of kpca-grnn model on the prediction of corrosion rate is verified by the corrosion monitoring data of the hanging piece of circulating water field and the test pipe,and the prediction effect is compared with that of BP neural network model.The results show that the combination of KPCA and GRNN algorithm can predict the corrosion rate of circulating water,and the predicted value is more consistent with the actual value than the BP algorithm.