A robust simulator of pressure-dependent consumption in Python

Abstract

Modeling of pressure-dependent users’ consumption is mandatory to simulate accurately the hydraulics of water distribution networks (WDNs). Several software solutions already exist for this purpose, but none of them actually permits the easy integration and testing of new physical processes. In this paper, we propose a new Python simulator that implements a state-of-the-art pressure-dependent model (PDM) of users’ consumptions based on the Wagner’s pressure–outflow relationship (POR). We tested our simulator on eight large and complex WDNs, for different levels of users’ demands. The results show similar precision and efficiency to the ones obtained by the authors of the original model with their MATLAB implementation. Moreover, in case of fully satisfied users’ demands, our simulator provides the same results as EPANET 2.0 in comparable computational times. Finally, our simulator is integrated into the open-source, collaborative, multi-platform, and Git versioned Python framework OOPNET (Object-Oriented Python framework for water distribution NETworks analyses); thus, it can be easily reused and/or extended by a large community of WDN modelers. All this work represents a preliminary step before the incorporation of new processes such as valves, pumps, and pressure-dependent background leakage outflows. HIGHLIGHTS A new simulator of pressure-dependent consumption in water distribution network.; Coded in Python, based on a state-of-the-art MATLAB model, easy to extend.; Numerical experiments on networks composed of up to 19,647 pipes and 17,986 nodes.; Instabilities handled through regularization and line search along the Newton descent.; Preliminary step for integration of new processes and first contribution to OOPNET.;

Keywords

• numerical optimization
• pressure-dependent model (pdm)
• python programing language
• user’s consumption
• water distribution network (wdn)