Water Asset Management International 3.3 (2007) 14-18

Modelling the structural degradation in water distribution systems using the artificial neural networks (ANN)

Raed Jafar, Isam Shahrour

Mechanical Laboratory of Lille UMR, University of Science and Technology of Lille, Villeneuve d'Ascq
France

Ilam Juran

Urban Utility Centre, Polytechnic University, Brooklyn
USA

This paper presents data analysis of the degradation rate of water mains using the Artificial Neural Networks (ANN). This method is used for the analysis of data collected in Wattrelos city in the North of France. The first part of the paper presents the case study. Subsequently, the paper presents the ANN method and its calibration on the data collected from 1991 to 1999. Six artificial neural network models are elaborated. The validation of these models on data collected from 1999 to 2004 illustrated good performance, as defined by the Coefficient of Determination through comparison of forecasted and observed number of failures. The spatial repartition of the risk of degradation is illustrated using the geographic information system, which constitutes an effective tool for the elaboration of strategies of rehabilitation of water distribution systems.

Full article (PDF Format)