Science Inventory

Better Water Demand and Pipe Description Improve the Distribution Network Modeling Results

Citation:

Zhao, Y., J. Yang, Y. Lee, S. Panguluri, AND Y. Shao. Better Water Demand and Pipe Description Improve the Distribution Network Modeling Results. In Proceedings, World Environmental and Water Resources Congress 2017, Sacramento, CA, May 21 - 25, 2017. American Society of Civil Engineers (ASCE), Reston, VA, 593-601, (2017). https://doi.org/10.1061/9780784480625.055

Impact/Purpose:

Generalized water demand patterns can be used to simulate water flow and quality for distribution system monitoring, however this approach has not been examined for its reliability. The purpose of this project was to analyze data and water demand patterns to determine model accuracy. This research can be used by other researchers, communities, water utilities, and regulators who are performing network modeling and water quality simulations.

Description:

Distribution system modeling simplifies pipe network in skeletonization and simulates the flow and water quality by using generalized water demand patterns. While widely used, the approach has not been examined fully on how it impacts the modeling fidelity. This study intends to evaluate these consequences in a typical water distribution system. We first analyzed tracer test data at 13 locations using a calibrated EPANET model with generalized water demand patterns. Modeling results agree with actual measurements in six locations with high water demand and flow velocity near the storage tank, while fail to replicate the variations in network abundant of dead-ends and Tee/Cross junctions. Then we further investigated the causes of model inaccuracy using real-time water demand data covering ~35% water users. This paper compares several levels of pipe skeletonization and water demand aggregation in water age simulation, and shows significant impacts on water age results because of discrete water demand at pipe nodes. The analysis reveals varying flow properties and the movement of highly aged water parcels in two types of pipe configurations. These results have the implications on network modeling and further studies of water quality simulations.