Establishment and Application of Rapid Prediction Model for Total Bacteria Number in Water Distribution Network

Long time detection was made on the water quality in selected water distribution network. The correlation between several water quality parameters and total bacteria number was investigated. Temperature, free residual chlorine, total residual chlorine, turbidity and UV254 were analyzed and determined for their high correlation with total bacteria number and ease to be detected. Multivariate linear regression method was applied to found a rapid prediction model for total bacteria number in water distribution network. The prediction accuracy of the model is 64.55%. The model can be used in the rapid prediction of total bacteria number in water distribution network, and provide technical support for ensuring the microbiological indexes of water quality in water distribution network.