Research On Incomplete Mixing Model Based On Simulation Of Mixing Flow Pattern At Pipe Junctions Of Water Supply Network

Due to the rapid development of the economy and the acceleration of urbanization in China,drinking water quality safety has drawn increasing attention.In the urban water supply and distribution system,there is a common issue that the quality of effluent water meet standard,while,the water quality of the tap water at the user end is not.The pipe network,as an important infrastructure that connects water factories and users,its internal water quality directly affects the water quality of tap water due to retention time of drinking water.The current water quality simulation of urban water supply and distribution network is based on the hypothesis of instantaneous complete mixing at pipe junctions,resulting in a large difference between water quality model predictions and actual observations.The purpose of this study is to research the water quality mixing mechanism at pipe junctions(cross junctions,double-T junctions,and H junctions)in the water distribution network,and to build an incomplete mixing empirical model of the pipe network,improve the EPANET water quality model,and provide data support for improving the accuracy of water quality model of water supply network.In this study,CFD numerical simulation and orthogonal experiment design are used,the pipe diameter,total Reynolds number,Reynolds number ratio,and distance between junctions that affect the solute mixing at pipe junctions,is taken as orthogonal experiment factors.The orthogonal experimental conditions are designed according to the factors and the corresponding levels,three types of mixing water quality models at pipe junctions are constructed respectively,and the boundary condition parameters of the simulated working conditions are set,to perform the orthogonality analysis.The results show that both the inflow Reynolds number ratio and the outflow Reynolds number ratio have significant effects on the mixing degree of the three types of pipe junctions,while the pipe diameter has no significant effect on the mixing degree of the cross junctions and H junctions.The distance between double-T junctions has the most significant effect on its mixing degree,while the distance between the two inflows of H junction has no significant effect.The same factors have different impacts on mixing degrees of different types of pipe junctions.Build a water distribution network pipe junctions mixing experimental platform and use it as a basis,according to the orthogonal experiment conclusions,the important factors that affect the mixing degree of pipe junctions are selected as key experimental factors,and some factors are used as auxiliary experimental factors to design experimental schemes and conduct experiments.In the laboratory simulation experiment,data analysis is carried out by measuring the pressure,flow,and conductivity obtained from the online remote transmission equipment.The effects of inlet and outlet Reynolds number ratio,pipe diameter,inflow Reynolds number,and distance between junctions(inflow spacing)on the mixing behavior of pipe junctions are compared and analyzed under the circumstance of the laboratory experiment,existing research,EPANET water quality simulation and CFD numerical simulation.The results show that both the inflow and outflow Reynolds numbers ratios have significant effects on the mixing behavior,which verifies the numerical simulation results;the pipe diameter and total inflow Reynolds number have no significant influence on the mixing behavior of the three kinds of pipe junctions;the laboratory experimental results of cross junctions and H junctions are consistent with the numerical simulation results,while the experimental results of double-T junctions are significantly different from the numerical simulation results.Using the response surface method(RSM),taking the ratio of inlet Reynolds number and outlet Reynolds number as the design factors of central composite test,adopting the full factor central composite sequential design to design the numerical simulation scheme.High significance incomplete solute mixing water quality model of cross junctions,double-T junctions(L=D,L=5D,L=10D)and H junctions(L=D)at water distribution network are established according to the results obtained through CFD numerical simulation,which can reflect the interaction of the inlet Reynolds number ratio and the outlet Reynolds number ratio on the solute mixing at pipe junctions,providing a basis for the improvement of the water quality model of water distribution network.Based on the complete mixing water quality model at water distribution pipe junctions of EPANET,combined with the mixing law of pipe junctions,the complete mixing water quality model is modified to incomplete mixing model,according to the characteristics of pipe network structure and operation parameters.A small-scale 3×3pipe network is constructed to compare the difference between the two models under static and transient solute transport conditions.The results show that the modified model is more close to the actual operation of the pipeline network under the condition of static solute transport,and there is a big difference between the two models under the condition of transient solute transportation,which proves the necessity of improving the water quality model of pipe network.Build a numerical simulation model and an experimental platform of pipe junctions of a small-scale 4x4 water distribution network respectively,set up two conditions of the same direction inflow and opposite direction inflow,take the sodium chloride concentration at the outlet of the pipe network as the evaluation index,to compare the results of CFD numerical simulation,EPANET model simulation and experimental platform under the same conditions.The results show that the CFD numerical simulation results and the experimental platform results tend to be consistent,but EPANET model simulation results are significantly different from the two,which further proves the mixing phenomenon at pipe junctions of the water distribution network is incomplete mixing.The modified EPANET mixing water quality model was applied to the simulation of residual chlorine concentration in the water distribution network of EQ City.The difference in residual chlorine contour map distribution between the complete mixing model and the modified incomplete mixing model at each time were compared.The results show that the existence of various pipe junctions is the reason for the difference of simulation results.Selecting the same pipe junction to analyze the changes of residual chlorine concentration at a different time under the two models,the results show that the residual chlorine concentration of the modified model is lower than that of the complete mixing model,which provides a method for accurate calculation of the residual chlorine concentration in a water distribution network.