Common menu bar links
CONSTRUCTION INNOVATION, Summer 1998
[ Table of contents | Subscribe ]
Selecting and scheduling restoration measures for water-distribution networks
Aging watermains, subject to continuous operational and environmental stresses, have led to significant deterioration of water-supply systems throughout North America. This deterioration manifests itself in increased operating and maintenance costs, water losses, frequent service disruptions and a reduction in the quality of water supplied.
The most expensive component of a water-supply system is the distribution network. Its restoration - to ensure adequate flow, pressure, water quality, and uninterrupted service - may require significant investment in pipe replacement or renovation.
Water utility engineers are constantly faced with the dilemma of which restoration measure to select for each pipe in the system, and when to implement it - the actual costs associated with any structural restoration depend on the timing of its implementation (see figure). Delays in pipe replacement (or structural relining) decrease the present value of the renewal investment but increase maintenance costs, primarily because the older the pipe, the more repairs it requires.
The figure depicts the optimal renewal timing of a single water main. Although relatively simple to determine for a single pipe, this task becomes extremely difficult when the deterioration of the hydraulic capacity of an entire distribution system has to be considered.
OPTIMAL TIMING FOR RENEWAL OF A SINGLE WATER MAIN
A method of facilitating this decision-making process has been developed at the University of Toronto, and is currently being enhanced at IRC. It involves the selection of a rehabilitation/renewal alternative and the time of its implementation for each pipe in an existing network, so that the investment in rehabilitation and all maintenance costs are minimized for the entire distribution system. This method explicitly takes into account the diminishing structural integrity and hydraulic capacity of every pipe in the system, allowing consistent and unbiased consideration of the various rehabilitation/replacement alternatives. The method, based on a dynamic programming approach, provides accurate results but is, at present, suitable only for small distribution systems.
IRC is now initiating a research consortium project to extend this approach to large water-distribution systems. Suitable approaches will be identified and validated, and a prototype software package developed. It will likely be possible to incorporate this software into a utility's existing asset management system, providing engineers and decision-makers with a practical planning tool for the long-term rehabilitation of water-distribution networks. If you are interested in joining the consortium, please contact Dr. Yehuda Kleiner at (613) 993-3805, fax (613) 954-5984, or e-mail yehuda.kleiner@nrc-cnrc.gc.ca
