炼油厂常压系统腐蚀因素关系模型及评价研究

发布时间：2018-03-05 00:35

本文选题：常压系统　切入点：腐蚀因素　出处：《西安石油大学》2015年硕士论文　论文类型：学位论文

【摘要】：炼油厂常压塔装置是石油加工的基本装置也是重要装置,其安全运行决定着整个炼油厂的正常作业与安全。腐蚀是常压装置的最普遍的问题,常压系统的腐蚀影响着炼化装置的正常运行及安全,腐蚀问题的存在,给炼油厂带来的安全隐患,造成装置的非计划停工,给企业带来了巨大的经济问题,因此,为了炼油厂的安全有效长期的运行,研究炼油厂常压系统的腐蚀问题极为重要。通过对炼油厂常压系统腐蚀的调研,得出了最常见的腐蚀类型是电化学腐蚀,常压系统中换热器的腐蚀类型有:点腐蚀、应力腐蚀、垢下腐蚀、均匀腐蚀等。对换热器循环冷却水进行分析,得出其主要的腐蚀类型为点腐蚀,影响点腐蚀的因素主要是循环冷却水的温度、p H值、Cl-浓度、NH4+浓度和S2-浓度。评价点腐蚀的指标有点蚀电位Eb、标准偏差SD、孔蚀指标PI和噪声电阻Rn等,本文选用的腐蚀评价指标是孔蚀指标PI和噪声电阻Rn。选用非线性回归方法建立腐蚀模型,应用软件SPSS及1st Opt软件对循环冷却水的数据进行数据分析。软件SPSS对五种影响因素进行主成分分析,提取了4个成份,并得到了每一个主成份的表达式,成份1至成份4包含了所有的因素(温度、p H、S2-浓度、N4+浓度、Cl-浓度),也就是需要这五种因素来综合评价炼油厂常压塔顶的腐蚀情况。建立常压系统的腐蚀模型,应用回归分析方法,首先建立了线性回归模型,得出评价指标与各个因素之间的模型的相关系数最高为0.595,模型明显不理想,因此选用非线性回归分析模型,第一种非线性模型就是非线性转线性模型的建立方法,得出模型的相关系数最高为0.496,此模型不理想。第二种非线性模型就是分别建立每个因素与评价指标之间的最佳一元回归模型,最后人工合成多元非线性模型,得出每个模型的相关性系数(RPI=0.876、RRn=0.984),并进行回归方程显著性检验,检验结果为每个腐蚀评价模型拟合均理想。最后对腐蚀非线性模型进行分析,应用模型对炼油厂常压系统的腐蚀进行模拟预测,预测结果与理论分析的结果相一致,故本文建立的腐蚀评价指标与各个因素之间的非线性模型可用于腐蚀的预测分析。
[Abstract]:The atmospheric pressure tower unit is the basic and important unit of petroleum processing, and its safe operation determines the normal operation and safety of the whole refinery. Corrosion is the most common problem in the atmospheric pressure unit. The corrosion of atmospheric pressure system affects the normal operation and safety of refinery plant, the existence of corrosion problem, the safety hidden danger to refinery, the unplanned shutdown of unit, and the huge economic problem to the enterprise. In order to run the refinery safely and effectively for a long time, it is very important to study the corrosion problem of the atmospheric pressure system in the refinery. Through the investigation of the corrosion of the atmospheric pressure system in the refinery, it is concluded that the most common corrosion type is electrochemical corrosion. The corrosion types of heat exchangers in atmospheric pressure system are pitting corrosion, stress corrosion, scale corrosion, uniform corrosion, etc. By analyzing the circulating cooling water of heat exchanger, it is concluded that the main corrosion type is point corrosion. The main factors affecting the pitting corrosion are the temperature of circulating cooling water and the concentration of NH4 and S2-. The evaluation indexes of pitting corrosion are pitting potential Eb, standard deviation SDs, pitting corrosion index Pi and noise resistance rn, etc. The corrosion evaluation indexes used in this paper are pitting corrosion index Pi and noise resistance Rn.The corrosion model is established by using nonlinear regression method. The data of circulating cooling water were analyzed by software SPSS and 1st Opt. The software SPSS analyzed the five influencing factors, extracted four components, and got the expression of each principal component. Components 1 to 4 contain all the factors (temperature, temperature, pH, S2-, N4, Cl-Concentration), that is, these five factors are required to comprehensively evaluate the corrosion of the top of the atmospheric tower in a refinery. A corrosion model of the atmospheric pressure system is established. Based on the regression analysis method, the linear regression model is established, and the correlation coefficient between the evaluation index and each factor is the highest 0.595.The model is obviously not ideal, so the nonlinear regression model is chosen. The first kind of nonlinear model is the method of establishing nonlinear translinear model. The highest correlation coefficient of the model is 0.496, this model is not ideal. The second kind of nonlinear model is to establish the best univariate regression model between each factor and the evaluation index, and finally to synthesize the multivariate nonlinear model. The correlation coefficient of each model is obtained, and the regression equation significance test is carried out. The test results show that each corrosion evaluation model fits well. Finally, the nonlinear corrosion model is analyzed. The model is used to simulate and predict the corrosion of atmospheric pressure system in refineries. The predicted results are consistent with those of theoretical analysis. Therefore, the corrosion evaluation index established in this paper and the nonlinear model between various factors can be used to predict and analyze corrosion.
【学位授予单位】：西安石油大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TE986

【参考文献】