Least-Cost Design of Water Distribution Networks Including Fire Damages

Abstract

The implicit goal of water network design is to provide acceptable service during normal and peak demands and to limit damages to property and people during fires. Despite this, fire damages are seldom explicitly included in network design. This paper presents a single-objective particle swarm optimization program that incorporates a new measure of expected conditional fire damages to size local water distribution mains for fire flow protection in residential service areas. The optimization approach generates trade-off information to help utilities determine the cost effectiveness of adding new pipe capacity to reduce the risk of fire damages in water networks. The optimization program was applied to an 8-pipe network and a 34-pipe network to generate trade-off curves for pipe cost and fire damages. A sensitivity analysis indicated that the level of uncertainty in fire flow had a little impact on pipe sizing and cost in the eight-pipe network. Optimization results in the 34-pipe network supported the industry practice of using a minimum 150-mm distribution main sizing to provide fire flow protection.