Comparison of Pressures Simulated Using Transient Analysis with Field Data from a Full-Scale Distribution System

Abstract

Negative pressures were measured in a full-scale distribution system following low pressure events at a water treatment plant. Transient analysis was used to model three downsurge events and compare the simulated pressure profiles with field measurement data. The objective of this work is to assess the source of uncertainty and variability associated with the estimation of intrusion volumes calculated by a transient analysis model. This assessment was conducted by comparing actual field pressure measurements and model outputs under various model settings (e.g., cavitation head, wave speed). For the three downsurge events, the modeled pressure profiles matched reasonably well with the measured pressures, as long as the pressures remained positive at a site. When the pressures reached negative values, the amplitude of the modeled pressures was larger than that of the recorded pressures. The difference between measured and modeled pressure is strongly related to a greater energy dissipation in the real distribution system, which is affected by the uncertain presence of air in pipes, the level of network skeletonization, and the allocation of demand. The estimation of intrusion volumes and risk for public health is directly affected by the pressure results obtained using transient analysis. Comparison to field data is therefore important to evaluate the accuracy of such a process.