供水管网末端铁扩散实验及模拟

发布时间：2018-04-22 03:41

本文选题：水质模型 + 末端　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：城市供水安全是关系到国民的生命和健康的一个重要问题,经过多年的使用,金属管道会有不同程度的腐蚀从而形成水垢,铁释放是水质恶化的一个重要原因,会导致“红水”,影响饮用水观感,造成用户投诉,同时饮用水过量铁的摄入也会导致用户的健康问题。水质模型是实现水质监测和控制的重要工具,EPANET工具包是水力水质模型的重要方法,然而其水质推流模型并不考虑在供水管网末端水质的释放和铁的扩散效应,造成相应的模拟误差,导致对水质实时变化规律把握的不准确。为研究这一现象,修正EPANET并得到更为精确的水质模型,本研究设计实验膜池装置,使用已知物系KCL标定膜池反应器的膜池常数,将多次实验的平均值最为最终膜池常数。利用膜池常数以及络合滴定法对288.15K、293.15K、298.15K三个温度下的氯化亚铁积分扩散系数进行测定,以获得积分扩散系数为初值,通过最小二乘法得到微分扩散系数关于浓度的表达式,结果可知微分扩散系数随温度提高而增大,每度影响约为2%,随浓度增大而呈减小趋势。以上述实验测定的微分扩散系数为参数,建立CFD数值模型,以湍流场作为物质迁移场的耦合场。设计实验装置,在三个温度下分布设置七个不同工况对CFD数值模型的准确性进行标定,同时利用EPANET对同样情况进行模拟,比较三者的差距,结果显示传统水质模型由于物质迁移机制问题,难以模拟此情况下的水质变化,认为CFD数值模型能够一定程度的反映实际情况,EPANET需要进行一定的修正。利用标定后的CFD数值模型在288.15K、293.15K、298.15K下以末段长度、末端管径、末端连接管段管径、末端连接管段流速、末端存水浓度为自变量,研究其对铁扩散量的影响。为便于实际工程应用,本研究使用一种新型的多元非线性回归方法,先通过一元非线性回归确定单个自变量的最佳回归模型形式,并以多个自变量最佳模型的叠加作为多元非线性回归的初始值,继续进行迭代得到三个温度下的经验模型,所得到的多元非线性回归模型2可以达到99%,显著性远远小于0.05,在定义域区间内收敛速度快且拟合准确。本文工程应用以DG市某片区管网为例,综合利用该片区管网拓扑数据,建立该片区管网水力模型,选择测压点对管网水力模型的准确性进行标定。在符合相关标准的管网水力模型的基础上,分别使用改进前与改进后的EPANET对DG市该片区管网进行给水管网水质建模,由给水管网铁模型模拟结果可知,改进后的EPANET可以更为准确的描述铁在给水管网中的实时分布,具有较为广泛的应用价值。
[Abstract]:The safety of urban water supply is an important issue related to the life and health of the people. After years of use, metal pipes will corrode to varying degrees and thus form scale. Iron release is an important reason for the deterioration of water quality. It can lead to "red water", affect the perception of drinking water, cause user complaints, and drinking water excessive iron intake can also lead to health problems. Water quality model is an important tool to realize water quality monitoring and control. EPANET toolkit is an important method of hydraulic water quality model. However, its water quality pushing model does not consider the water quality release and iron diffusion effect at the end of water supply network. The corresponding simulation error results in inaccurate understanding of the real-time variation law of water quality. In order to study this phenomenon, correct EPANET and get a more accurate water quality model, an experimental membrane cell device was designed in this paper. The known KCL was used to calibrate the membrane cell constant of the membrane cell reactor, and the average value of many experiments was the final membrane cell constant. The integral diffusion coefficient of ferrous chloride at three temperatures 288.15K ~ 293.15K ~ 298.15K was determined by membrane cell constant and complexometric titration. The integral diffusion coefficient was obtained as the initial value, and the expression of differential diffusion coefficient about concentration was obtained by the least square method. The results show that the differential diffusion coefficient increases with the increase of temperature, the effect of each degree is about 2%, and decreases with the increase of concentration. With the differential diffusion coefficient measured in the above experiments as the parameter, the CFD numerical model is established, and the turbulent field is used as the coupling field of the mass transport field. An experimental device was designed to calibrate the accuracy of the CFD numerical model under seven different working conditions under three temperature distributions. At the same time, the accuracy of the CFD numerical model was simulated by EPANET, and the differences among the three models were compared. The results show that the traditional water quality model is difficult to simulate the change of water quality due to the problem of material migration mechanism. It is considered that the CFD numerical model can reflect the actual situation to a certain extent. By using the calibrated CFD numerical model at 288.15K 293.15K 298.15K, the influence of the end length, the end diameter, the end connecting pipe diameter, the terminal connecting pipe flow velocity and the terminal water concentration on the iron diffusion is studied. In order to facilitate practical engineering application, a new multivariate nonlinear regression method is used to determine the optimal regression model of a single independent variable by univariate nonlinear regression. Taking the superposition of the optimal model of multiple independent variables as the initial value of multivariate nonlinear regression, the empirical model at three temperatures is obtained by iteration. The multivariate nonlinear regression model 2 can reach 99th, the significance is far less than 0.05, the convergence speed is fast and the fitting is accurate in the domain of definition. Taking a district pipe network in DG city as an example, the hydraulic model of the pipe network in DG city is established by using the topological data of the district pipe network. The accuracy of the hydraulic model is calibrated by selecting the pressure measuring points. On the basis of the hydraulic model of the pipe network which accords with the relevant standards, the water quality model of the water distribution network in the DG region is established by using the improved EPANET before and after the improvement, and the simulation results of the iron model of the water supply network can be seen. The improved EPANET can more accurately describe the real-time distribution of iron in water distribution network, and it has more extensive application value.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TU991.33

【参考文献】