Disinfection By-products Prediction Model And Its Precursor Characteristic Analysis In Wastewater Reclamation Process

Urban sewage recycling is one of the important measures to solve the problem of water shortage and urban water shortage in the world.Disinfection is a necessary measure to ensure the safety of reclaimed water.Due to the advantages of strong sterilization ability,long duration and low price,chlorination is still widely used in many domestic sewage treatment and recycling plants.The conventional pollutants in the effluent of municipal wastewater are reduced significantly after the traditional two stage biological treatment,but there are still some residual dissolved organic matters which are difficult to biodegradation and the new generation in two stage biological treatment process.The content of these materials are not only higher than the surface water,but also the composition are more complex,which will lead to the formation of disinfection by-products in reclaimed water.These DBPs does not act on the human body directly,but will enrich in the environmental,and eventually harm to human health through the food chain.Based on the study of the structure of precursor product and the prediction model of by-products,we can provide a theoretical basis for the effective control of by-products in the process of wastewater reclamation.Four different chemical techniques,including the elemental analysis,Fourier transform infrared?FT-IR?spectrometry,solid-state CP TOSS MAS 13 C nuclear magnetic resonance?13C-NMR?spectroscopy and three dimensional fluorescence excitation-emission-matrix?3DEEM?spectra were used to characterize the chemical composition and structural characteristic of humic acid?HA?in the excess sludge from four different wastewater treatment plants?WWTP?.In order to obtain the prediction model of the reclaimed water disinfection by-products suitable for the specific conditions,which could provide the basis for the study of the disinfection by-products,the test was carried out with coagulation-sedimentation-filtration reactor.The results were as follows:?1?The HAs presented almost the same chemical composition,however,the structure of the HAs was significantly different.The humic acids in the excess sludge of the sewage treatment plant were generated from biological source,and there were big differences between the chemical composition and structure compared with soil humic acid,which were characterized by more fat components and relatively weak aromaticity.?2?Due to the low concentration of Br in different humic acid,the content of disinfection by-products containing Cl were the largest,and the disinfection by-products containing Br were very rare.The formation potential of chlorine disinfection by-products in different humic acid solutions were different,and the order of the formation potential of THMs was C>D>A>B,but the order of the formation potential of HAAs was B>D>A>C.Under the different chlorination conditions,the concentration range of THMs was from 35.74 to 122.57 ug/L,and the chloroform was in the range of 31.81 to 108.54ug/L,which accounted for the main part of the whole THMs.?3?Multiple linear regression model and nonlinear model can be used to predict the THMs generated after the disinfection in the reclaimed water:THMs=-502.50+0.03T+60.87pH+7.00DOC+4.96Cl₂THMs=1.86×10^-5·T-0.01·pH 6.09·DOC0.62·Cl₂0.63But the prediction accuracy was different,the R² of multivariate linear model of THMs was 0.72,and the R² of nonlinear model was 0.69.The R² of multivariate linear model validation analysis of THMs was 0.56 and the nonlinear model's R² was 0.52,indicating that the predictive accuracy of the linear model was better than that of nonlinear model in terms of THMs.?4?Chloroform accounted for the main part of THMs,linear model and nonlinear model were:CHCl₃=-368.90-0.14T+44.35pH+5.16DOC+4.25Cl₂CHCl₃=1.62×10^-5·T0.06·pH 5.91·DOC0.64·Cl₂0.77In terms of chloroform,the R² of multiple linear prediction model was 0.78,and nonlinear prediction model's R² was 0.74.And the R² of multivariate linear model validation analysis was 0.70,and the nonlinear model's R² was 0.68,which also showing that predictive accuracy of the linear model was better than that of nonlinear model.And on the whole,it was found that the correlation coefficient of chloroform was greater,that is,the prediction was more accurate.