Evaluation of Robustness of Water Distribution Networks by Varying Peak Factors in Hydraulic Models: Case Study of Mombasa North Mainland

Abstract

Most old water distribution networks are extended overtime to accommodate new connections which make them heavily looped. These networks also lack well-documented mapping incorporating all alterations. Consequently, there are difficulties in the monitoring of water flow in pipelines and in the management of water losses. The result is poorly sized and managed water networks that do not meet flow velocities and pressure requirements leading to inadequate and erratic water supply. Upgrading of the networks involves testing the robustness of the network to handle varied flows and construction of capacity augmentation pipelines running parallel to the old pipelines. In this study, the existing alignments of the Mombasa North Mainland distribution network were mapped and modelled using KYPipe Software. Three models; namely, existing network model for year 2020, and future network models design years 2030 and 2040, were developed. Design flow is the projected peak flow in the distribution network and usually has a design peak factor of between 1.0 and 1.2. Flows greater than the design flow were simulated in the distribution networks by applying peak factors of 1.5 and 2.0 in the models. The adequacy of the network was determined based on the achievement of acceptable design pressures and pipeline flow velocities. Results showed that the existing and future Mombasa North Mainland hydraulic model analyzed under a peak factor of 1.2 met the water supply design criteria for pressures range 100 - 600 kPa and flow velocity range 0.01 - 2.0 m/s and pressures of 80 to 524 kPa. The 2030 model simulated using a peak factor of 1.5 was found robust to transmit the increased flow. However, under 1.5 peak factor, 2040 model was found insufficient. Under a peak factor of 2.0, the 2030 and 2040 models were found insufficient to transmit the increased flows with both models experiencing negative flows and zero flow pipelines in some parts. The study optimized the networks for possible increased flows by proposing augmented new pipelines for short distances along the main transmission mains to increase the robustness of the ultimate model for a 1.5 and 2.0 peak factor. The study recommended testing the robustness of water distribution networks to handle unexpected flows before implementation of the networks. Moreover, water utilities should develop, maintain, and update models for quick and informed decisions making on their network expansions.