AWWA WQTC62511

AWWA WQTC62511 Predicting Source Water Quality Using Neural Network
Conference Proceeding by American Water Works Association, 11/01/2005

Carriere, Annie; Barbeau, Benoit; Tremblay, Genevieve; Gauthier, Vincent

More info

Full Description

Rapid fluctuations of source water quality can upset routine water treatment plant operations.Several waterborne outbreaks have been caused by the co-occurrence of source watercontamination and treatment upsets (Woo and Vicente, 2003; Fox and Lytle, 1996).Understanding source water fluctuations according to watershed activities increases therobustness of WTP operation. The main objectives of this project were to: identify theorigins of source water turbidity fluctuations at the inlet of the Montreal water treatment plant(WTP); and, use this information to forecast turbidity peaks 24 hours in advance using anartificial neural network (ANN) methodology.The first step of this project was to become familiar with the phenomenon of interest,turbidity variations. For this purpose, daily turbidity data for a period of 40 months werecollected and observed to characterize the major events and define any existing patterns. Forthe same period, data were collected for 43 variables possibly related to turbidity variationsbased on a literature review. From this list of variables, those presenting significant seasonalvariation were conserved as potential independent variables. The main causes of turbidityvariations were identified by superposing graphs of turbidity and potential indicators. Thisexercise also allowed observing time lags between the parameters. As a complement to thegraphical method, a correlation matrix was produced between the indicators and the turbidityvalues for different time lags.Once it was felt that the main causes of turbidity fluctuations had been identified, artificialneural networks were selected as a modeling tool to forecast them. The methodologyemployed was elaborated using the approaches proposed by research teams working in theenvironmental and water resources fields (Baxter et al, 2002; Maier and Dandy, 2000). Itincludes six main steps: the identification of the needs; the choice of the performance criteria;the development and organization of the database; the construction of neural network models;and, the final model choice. Includes 7 references, tables, figures.