Study On Drinking Water Quality Monitoring And Early Warning Technologies Based On Evaluation Of Spectroscopy,Total Organic Halogen And Toxicological Parameters

Drinking water quality is closely related to human health.But there are some limitations on the physicochemical and microbiological parameters currently used for water quality monitoring:(?)the complex measurement methods make them impossible to quickly detect the water quality;(?)the parameters are limited and unable to quantify all pollutants presented in drinking water;(?)the parameters cannot reflect the comprehensive toxicity produced by the interaction between numerous pollutants.Thus,the quality of drinking water cannot be monitored and evaluated in a timely and objective manner.As a result,the early warning of water pollution is hindered,and the sufficient scientific bases for improving the drinking water treatment processes and strengthening supervision of drinking water quality cannot be provided.Therefore,it is necessary to develope new water quality monitoring technologies to make up for the shortcomings of routine monitoring system.In this study,we collected long-term source water,finished water and tap water samples from two major drinking water treatment plants(DWTP)in urban area of W city in Hubei Province,and scanned ultraviolet-visible spectrum and three-dimension excitation-emission matrix(EEMs),detected the concentration of total organic halogens(TOX),performed Caenorhabditis elegans(C.elegans)-based bioassays on the water samples.Then the patterns of spatiotemporal variations of these parameters in drinking water treatment/distribution processes and long-term sampling were detected,and the correlations between these parameters and routine water quality monitoring parameters were analyzed.Based on above analyses,the performance of spectroscopic tecnologies,TOX and C.elegans-based bioassays in safety assessment and early warning of drinking water quality was evaluated.Part I Study on Drinking Water Quality Monitoring Technologies Based on Ultraviolet-visible Absorption Spectroscopy and Three-dimensional Fluorescence SpectroscopyObjectives:(?)To study the influences of drinking water treatment/distribution processes and seasonal factor on the spectral characteristics of source water(Yangtze river and Hanjiang river,respectively)and drinking water(including finished water and tap water)taken from A and B DWTPs located at urban area of W city.(?)To detect the surrogation of spectral parameters for routine parameters of source water and drinking water taken from A and B DWTPs.(?)To establish the normal range of spectral parameters of source water and drinking water taken from A and B DWTPs.Methods:The source water,finished water,tap water 1 at the midpoint of main distribution pipeline,tap water 2 at the terminal of branch distribution pipeline and tap water 3 at the terminal of main distribution pipeline(n=130)were collected monthly from A and B DWTPs at W city from June 2018 to September 2019.The ultraviolet-visible absorption spectrum and 3D-EEM were scanned for the filtered water sample.The absorption coefficient(a)represented the absorption intensity of a water sample at certain wavelength within ultraviolet-visible region(such as a₂₅₄ and a₂₁₀ measured the absorption intensity of light with a wavelength of 254 nm or 210nm,respectively).Parallel factor analysis was used to decompose the EEMs and extract the fluorescence components of the water samples.Then the spectral characteristics of water samples and their patterns of spatiotemporal variations were analyzed.Next,the Pearson correlation coefficient or Spearman correlation coefficient was calculated to analyze the relationship between the spectral parameters and routine parameters,and false discovery rate(FDR)was used to perform multiple test correction on the p value of above correlation coefficients.In addition,the partial least squares regression(PLSR)was further used to compare the prediction of a single spectral parameter and multiple spectral parameters to routine parameters.Finally,the normal range of spectral parameters intensity of surface water and drinking water taken from A and B DWTPs was established taking“Q1-1.5IQR”as lower limit and“Q3+1.5IQR”as upper limit(Q1 was lower quartile,Q3 was upper quartile and IQR was interquartile range).Results:(?)Three fluorescent components,component 1(C1,humic-like substance;excitation maximum and emission maximum?_ex/em=245/435 nm),component 2(C2,tryptophan-like substance;?_ex/em=230(275)/345 nm)and component 3(C3,tyrosine-like substance;?_ex/em=215(275)/290 nm)were extracted from source water,finished water and tap water samples;the removal efficiency of C1 by routine drinking water treatment processes was generally higher than that of C2 and C3;the fluorescence intensity of C2 of 74.36%of tap water samples and fluorescence intensity of C3 of 45.45%of tap water samples were higher than corresponding finished water after the distribution process in A DWTP;the mean value of a₂₅₄,C1and C3 intensity of surface water and drinking water taken in summer(from June to September)was 7.88 m^-1,2.76 R.U.and 1.60 R.U.,respectively,which was significantly higher(p<0.05)than its mean intensity in other months that was 6.51m^-1,1.41 R.U.and 0.40 R.U.,respectively.(?)For drinking water samples,strong correlation(p<0.05)was observed between a₂₁₀ and nitrate concentration,a₂₅₄ and TOC concentration,r_s was 0.88(n=52)and 0.76(n=52)respectively;compared with a₂₅₄,when all absorbance within the 230?400 nm wavelength range were included as independent variables,the explanation for TOC in drinking water sample was improved through PLSR analysis.The coefficient of determination R²(cross-validated coefficient of determination,Q²)increased from 0.70(0.67)to 0.85(0.76).(?)The normal range of a₂₅₄ for the source water and drinking water of A DWTP and source water and drinking water of B DWTP was 3.45?15.43 m^-1,3.22?6.91 m^-1,4.89?19.17 m^-1,4.76?9.30 m^-1,respectively.The normal range of a₂₁₀for the source water of both DWTPs,the drinking water of A DWTP and B DWTP was 141.62?302.86 m^-1,138.17?304.01 m^-1,132.13?277.23 m^-1,respectively.Conclusions:(?)Routine drinking water treatment processes had higher removal efficiency of organic matter represented by C1 than C2 and C3 in source water;in the distribution process,the intensity of C2 and C3 of drinking water showed an upward trend;organic matter components represented by C1 and C3 in surface water and drinking water increased significantly in summer.(?)a₂₁₀ and a₂₅₄ ofdrinking water could be used as a surrogator for monitoring nitrate and organic matter pollution,respectively;the prediction of organic matter concentration in drinking water could be improved by PLS-based regression model of multiple spectral parameters.(?)The normal range of a₂₁₀ and a₂₅₄ could be used to provide early warning of nitrate and organic pollution of drinking water taken from A and B DWTPs.Part II Study on Total Organic Halogens as a Surrogate for Monitoring DBPs in Drinking WaterObjectives:(?)To study the influences of routine drinking water treatment/distribution processes and seasonal factor on the TOX concentration.(?)To detect the correlations between TOX and routine water quality parameters,spectral parameters.(?)To establish normal range of TOX concentration of drinking water taken from A and B DWTPs.Methods:The TOX concentration of water samples was measured by microcoulometry.The THMs(trihalomethanes)and other routine water quality parameters were detected according to national standard methods(GB/T5750.6-2006).Person correlation coefficient or Spearman correlation coefficient was calculated to examine the relationship between parameters,and FDR was used to perform multiple test correction on the p value of above correlation cofficients.The upper limit of normal range of TOX concentration in drinking water was established using“mean+2.58standard deviation”.Results:(?)The TOX concentration of finished water taken from A and B DWTPs(72.7?134.4?g/L and 71.9?202.4?g/L,respectively)was significantly higher(p<0.05)than that of corresponding source water(7.0?70.3?g/L and 4.1?69.7?g/L,respectively),and significant correlation was observed between the increment of TOX concentration and DOC(dissolved organic carbon)concentration in source water,r_s=0.68(n=24,p<0.05);for A DWTP,the mean value of TOX concentration of tap water 2 and 3 was higher than that of tap water 1 and further higher than that of finished water,which was 123.8?g/L,110.3?g/L and 102.7?g/L,respectively;there was no significant difference in the TOX concentration of drinking water of A and B DWTP in different seasons(p>0.05),and the average value was 126.8?g/L in spring(April?May),131.7?g/L in summer(June?September),120.4?g/L in autumn(October?November)and 120.4?g/L in winter(December?March in next year),respectively.(?)For drinking water,the correlation between TTHMs(total trihalomethanes)and TOX concentration was weak,r_s=0.34(n=50);TTHMs could only explain 30%variations of TOX(R²=0.30);significant correlation was observed between TOX concentration and total organic carbon concentration and C2fluorescence intensity in drinking water(p<0.05),r_s was 0.64 and 0.34,respectively.(?)Under the conditions of this study,the upper limit of the TOX concentration of drinking water of A DWTP and B DWTP was 177.6?g/L and 222.1?g/L,respectively.Conclusions:(?)Decreasing the DOC concentration of source water could promote reduced production of TOX and/or DBPs in finished water;the TOX concentration of drinking water could show an upward trend with the increase of distribution distance;seasonal factor had unsignificant influence on TOX concentration of drinking water.(?)THMs could not sufficiently reflect the DBPs pollution level in drinking water;TOX was a potential parameter for monitoring DBPs pollution in drinking water.(?)The normal upper limit of TOX concentration could be used for early warning of DBPs pollution in drinking water of A and B DWTPs.Part III Study on C.elegans-based Bioassays for Monitoring Drinking Water QualityObjectives:(?)To optimize C.elegans-based oxidative stress assay for water quality monitoring.(?)To study the influences of drinking water treatment/distribution processes and seasonal factor on the C.elegans-based oxidative stress toxicity,lethal toxicity,reproductive toxicity and genotoxicity.(?)To analyze the correlations between toxicological parameter and routine water quality monitoring parameters,spectral parameters and TOX concentration.(?)To evaluate the consistency of results of drinking water toxicity tests based on C.elegans and Vibrio fischeri(V.fischeri).(?)To establish the normal range of C.elegans-based toxicity caused by drinking water of A and B DWTPs.Methods:In order to optimize the parameters of C.elegans-based oxidative stress assay and make it suitable for source water and drinking water quality monitoring,the dose-response relationship of the activation of Pgst-4::GFP in C.elegans induced by pro-oxidant juglone(5,10,20,40,60,80,100,and 120?M)was firstly observed under different exposure time(1,3,6,12,and 24 h);then the activation of Pgst-4::GFP were measured after 12 h-exposure to active chlorine(0.025,0.05,0.1,0.3,0.9,2,4,and 8 mg/L)and sodium sulfite(0.94,1.88,3.75,7,and 15 mg/L).According to established test parameters,the induction level of surface water and drinking water on C.elegans Pgst-4::GFP was measured.Water samples and ultrapure water(control)were concentrated with XAD-2 and XAD-8 resins.REF(relative enriched factor,REF)was used to indicate the exposure dose of water extracts.Set 5-6 doses for each water sample,L4-stage C.elegans were exposure to water extracts for 24 h,then the mortality was counted and the LC₅₀ was calculated.Smilarly,the V.fischeri luminescence inhibition rate and EC₅₀ was determined after15 min-exposure.L1-stage C.elegans were exposured to water samples with an REF of 100 for 96 h,the percentages of the increase of body length,fertility rate and offspring number of the experimental group relative to those of the control group were calculated.L3-stage C.elegans were exposured to water samples with an REF of 100 for 24 h,then the DNA of small numbers of C.elegans was extracted,the DNA lesions/10 kb(the frequencies of DNA lesions affecting PCR amplification per10 kb)was calculated after PCR amplification.Person correlation coefficient or Spearman correlation coefficient was calculated to analyze the relationship between toxicological parameters and routine water quality monitoring parameters,spectral parameters and TOX concentration,and FDR was used to perform multiple test correction on the p value of above correlation cofficients.For non-normally distributed toxicological parameter values,the normal range were established taking“Q1-1.5IQR”as lower limit and“Q3+1.5IQR”as upper limit;but for normally distributed toxicological parameter values,single-sided limit was established with“average+/-2.58standard deviation”.Results:(?)The activationof Pgst-4::GFP in C.elegans induced by juglone showed a bell-shaped concentration-response relationship;when the C.elegans was exposured to 40?M of juglone for 12 h,the highest activation of Pgst-4::GFP was observed;after12 h-exposure,2 mg/L of active chlorine could obviously increase the activation of Pgst-4::GFP,while 15 mg/L of sodium sulfite could not induce C.elegans oxidative stress.(?)Compared with the source water,the genotoxicity of the finished water of A and B DWTPs increased,the DNA lesion rate increased from0.01/10 kb to 0.49/10 kb(p<0.05),and from 0.18/10 kb to 0.42/10 kb,respectively;through distribution process,23.53%of tap water samples induced higher C.elegans DNA lesions rate than corresponding finished water(p<0.05).(?)For drinking water,there was no obvious correlation between toxicological parameters and routine parameters(including 19 parameters such as COD,nitrate and TTHMs),spectral parameters(including intensity of a₂₅₄,a₂₁₀,C1,C2 and C3)and TOX concentration.(?)The LC₅₀ of C.elegans mortality induced by source water and drinking water extracts were 102.40?1323.44 REF,EC₅₀ of V.fischeri luminescence inhibition rate were 0.29?29.65 REF,and no significant correlation was observed between LC₅₀ and EC₅₀(p>0.05,n=125).(?)Under the conditions of this study,the normal range of C.elegans Pgst-4::GFP activation induced by drinking water of A and B DWTP was 0.75?1.63 and 0.62?1.67,respectively;the percentage of descending C.elegans offspring number caused by 100 REF of drinking water was no more than 59.31%,and the C.elegans DNA lesion rate was no more than 1.21/10kb.Conclusions:(?)The C.elegans-based toxicity detection technologies for source water and drinking water of A and B DWTPs were successfully established.(?)Routine drinking water treatment/distribution processes and seasonal factor had effects on the biological toxicity of drinking water which could not be explained by routine monitoring parameters.(?)Biological toxicity test could promote the scientificity of water quality assessment based on routine parameters.(?)The results of water toxicity tests based on C.elegans and V.fischeri were different,and the strategy of using group biological toxicity tests might evaluate drinking water quality more objectively.(?)Under the conditions of the study,the normal range based on Pgst-4::GFP activation,offspring number and DNA lesion rate of C.elegans could be used for early warning of abnormal drinking water quality in A and B DWTPs.