Heuristic-Based Pipe Dimensioning Model for Water Distribution Networks

Abstract

An economic design of water distribution system components plays an important role in the water-supplying agency because of the requirement of a large financial investment on construction, operation, and maintenance. Population-based stochastic search algorithms have become a futuristic method for optimal design of water distribution networks. However, the effectiveness of population-based stochastic search algorithms purely depends on several factors, such as population size, degree of randomness involved, penalty function, algorithm operators, and search space size. An analysis of published research shows that a large number of evaluations is required to arrive at the least cost combination of pipe diameters. This paper proposes a new approach, in which flow velocity in a pipe is considered as implicit information for pipe sizing. Three well-known benchmark networks were used to illustrate the applicability of the proposed approach for least-cost pipe sizing of water distribution networks. The result of the model application reveals that the computational effort required to arrive at an economical design is very minimal.