Influent Water Quality Characteristics Analysis And Effluent TN、TP Prediction From A Municipal Wastewater Treatment Plant

During the 13th five-year Plan period,a large number of municipal sewage treatment plants have been built and put into operation throughout the country,and the control requirements of their effluent indicators have become increasingly stringent in various regions.The 14th Five-Year Plan has proposed that more stringent pollutant discharge requirements for municipal sewage treatment plants can be put forward in areas such as the Yangtze River Delta,Beijing-Tianjin-Hebei,and areas along the South-to-North Water transfer Project.As one of the reference indicators to evaluate water pollution,the excessive discharge of TN and TP will seriously affect the water ecological environment.The excess of TN and TP in effluent of sewage treatment plant is often related to influent quality and treatment process.Therefore,based on the analysis of influent quality characteristics and key process node parameters,the prediction and control of effluent TN and TP of sewage treatment plant is of great significance to process operation management and next standard raising work.A stable prediction model of effluent TN and TP reaching the standard is established,which can timely determine the effluent reaching the standard in case of sudden abnormality in the sewage treatment process,avoid excessive discharge,and provide guidance for the efficient and stable operation of sewage treatment.This paper analyzes the influent water quality characteristics of the first phase project（phase I）and the expansion project of Handan East sewage treatment Plant in2020.The results show that the influent index concentration distribution of the two places accords with the natural logarithmic normal distribution;the influent quality change is consistent with the seasonal variation;the influent NH₃-N,TN and TP index concentration of the expansion project is higher than that of the phase I project;the correlation between the influent indexes of the phase I project is lower than that of the expansion project.The designed influent index of the phase I project CODcr,BOD5,SS,TP and the expansion project CODcr,BOD₅,SS,TN has been lower than the actual influent guarantee rate water quality index of 85%in 2020,especially the influent BOD₅ far exceeds the designed influent quality index,which needs to be considered in the next step of upgrading.Based on the improved BP neural network,the prediction models of TN and TP of Handan East sewage effluent are established.The phase I project is 12-9-2,and the expansion project is 15-13-2 network structure.The remove abnormal water quality and operation data of East sewage treatment Plant in 2020 are used for training and testing.By comparison,the prediction result of neural network using LM algorithm is more accurate.The simulation of the operation data of the phase I project from June to July in2020 shows that when the influent CODcr reaches 95%of the actual influent frequency value in 2020,the effluent TN will exceed the standard;the actual operation data of the expansion project from July to August in 2020 are simulated,and when the influent TN reaches the actual influent frequency value of 95%in 2020,the effluent TN has the risk of exceeding the standard.When the influent SS reaches 95%of the actual influent cumulative frequency in 2020,or when the dosage of phosphorus removal agent is low（4 t/d）,it is predicted that the effluent TP will exceed the standard.