A new numerical modelling tool developed by researchers at the University of Toronto could help improve the design and operation of intermittent water distribution systems, which supply more than a billion people around the world.
While there are many commercial models designed for water distribution networks, they all make one important assumption: once the system is turned on, it stays on 24-7.
But for about 20 per cent of customers worldwide, that's simply not the case.
"In many places around the world, due to water scarcity or other factors, water is supplied for only a few hours per day," says David Meyer, an assistant professor in the department of civil and mineral engineering in the Faculty of Applied Science & Engineering.
"Over a billion people get their water this way. And yet these systems are almost universally designed and updated using hydraulic models that assume continuous water supply. That means they don't account for the pipes draining and refilling, and they don't account for the fact that customers store water in large tanks in their homes so they can have it for later use."
The new modelling tool developed by Meyer and PhD student Omar Abdelazeem combines current best practices to improve the design and operation of such systems. Their work is published in Water Resources Research .
"Water distribution networks are critical infrastructure, but they are also big, expensive and buried underground," says Abdelazeem, who is lead author of the study. "If you want to make a change to a system like that, you can't just try it out and see what happens. You need a computer model that can accurately predict how your changes might affect the system. That way, you can test out many different possible improvements before you start the difficult and expensive process of implementing them."
Meyer and his team have conducted numerous studies on intermittent water systems , including a comprehensive review of previous modelling attempts . That work earned them the 2025 Medal for Reproducible Research from the Journal of Water Resources Planning and Management.
Many of the models they reviewed ultimately derived from a single source: the Storm Water Management Model (SWMM), an open-source model created by the U.S. Environmental Protection Agency more than 50 years ago.
"That model was meant for stormwater, but with a little bit of work, you can adapt it for an intermittent water system that fills up and drains regularly," says Abdelazeem.
"The problem is that different people disagreed about which parts of the model to use for what purposes. And these differences matter - we found that sometimes values predicted by one model were thousands of times bigger or smaller than those from another.
"On top of that, many of the models were not reproduceable, meaning that even after reading the paper, you couldn't glean enough information to build your own version."
Building on this previous work, Abdelazeem and Meyer created a new, ready-to-use model, with a Python package that enables users to implement it automatically. Dubbed SWMMIN (SWMM for intermittent networks), the open-source model is available for free on GitHub .
"Our paper describes in detail how to use the model for various types of intermittent systems, and how to set up the numerical parameters to improve model performance," says Abdelazeem.
"In addition to synthesizing the best practices from all the previous models, we also found an ideal ratio of spatial and temporal resolutions that minimizes model error."
The team hopes the model will be used by researchers and water system operators around the world to test potential improvements to intermittent water systems all over the world.
"Ultimately, this is an attempt to model these water systems as they actually exist, rather than how we wish they existed," says Meyer.
"We hope that people will use it to find new design principles, and that those in turn improve service for all the people who depend on these intermittent systems every day."
