Integrated Gis-based Management of Water Distribution Networks

The need to more intelligently and efficiently manage water distribution systems is increasingly more important to agencies managing such networks seeking a way to increase the reliability of their systems, the uninterrupted quality service of their “customers” and the cost-efficient operations and maintenance of the aging distribution networks. Repair and/or replacement of aging water mains, especially in urban environments, impose major expenditures on already financially strained municipalities and state governments, and the need to more actively engage in the monitoring and management of such networks is progressively increasing as existing distribution networks continue to age and therefore deteriorate. The work included herein presents a framework for the management of urban water distribution networks based on both analytical and numerical modeling techniques and coupled with geographical distribution systems (GIS) for improved visualization and dissemination of relevant information to the management teams. The condition of elements within the water distribution network is assessed by means of mathematical modeling and artificial intelligence techniques, patterns in the underlying historical data are examined by means of artificial neural networks, knowledge is assembled and assessed by use of database management systems and fuzzy logic, and finally results are mapped on GIS for improved visualization and graphical querying of the implicit and explicit knowledge gained during maintenance and management of the water distribution network. The framework and processes developed and presented in this work were based on data collected in New York City and subsequently applied in Freeport (Long Island, NY), and the cities of Limassol and Larnaca (Cyprus). This report is based upon work supported partly by the National Science Foundation (under Grant No. CMS-0118376) and partly by the Cyprus Research Foundation (under Grant No. AKGEN/1204/01).