| Chlorine disinfection always produces disinfection by-products in thedistribution network. People pay more and more attention to the health risk ofdisinfection by-products. This study selected the water distribution network in TianjinTD campus as the research object. To ensure drinking water quality and human health,the network in this research was monitored for a year with chloroform anddichloroacetic acid as the representative indicators. Human health risk of disinfectionby-products risk in the target network was assessed comprehensively and accurately.The influences of water quality and operation conditions on disinfection by-productsgeneration were analyzed. The reference concentration distribution provides thetheoretical basis and decision support for network operation.Firstly, through principal component analysis and logistic regression model, riskprobability simulation model was established with chlorine consumption, total organiccarbon, water temperature, pH, SUVA (SUVA=UV254/TOC) and ammonia asindependent variables. According to the chloroform and dichloroacetic acidconcentration distribution of the different risk probabilities, three levels of risk weredefined.Secondly, according to independent variable distributions of risk probabilitymodel, the values of six independent variables were simulated with the Monte Carlomethod. Combined with formulas, water quality risk probabilities under threescenarios were calculated, getting cumulative probability distribution curves anddistribution. Then with the above risk levels, risk probability distributions weresimulated and compared under different scenarios.Thirdly, on the basis of health risk assessment model recommended by US EPA,some exposure parameters were adjusted by consulting the literature survey andcharacteristics of the target network such as average time, body weight, etc.According to the network practical situation, probability distributions of cancer risksand hazard indexes under different scenarios (residential area facilities, businessfacilities and public facilities) were simulated with the Monte Carlo method.Finally, according to the model independent variables distributions of three risklevels, change rules between disinfection by-products and independent variables were explored and impact factors of disinfection by-products were analyzed. Independentvariables distributions of three risk levels were compared and analyzed, and then thereference concentration ranges were determined, which would have a certain guidancesignificance in selecting the treatment processes of water plant and network operation. |
