CFD Simulations of Incomplete Mixing at Pipe Junctions: Junction Angle Influence and Predictive Modelling

<p dir="ltr">Accurate modelling of solute transport in water distribution networks (WDNs) is critical for ensuring water safety, yet conventional approaches often rely on the oversimplified assumption of complete mixing at pipe junctions—an idealization that fails under realistic flow conditions and can lead to substantial errors in predicting contaminant spread and disinfectant efficacy. This study advances the understanding of incomplete mixing by employing high-resolution computational fluid dynamics (CFD) simulations to systematically analyse how junction geometry, specifically the pipe connection angle, and the distribution of inflow and outflow rates govern solute dispersion at cross-junctions. The results demonstrate that these hydraulic and geometric factors dominate mixing behaviour, while diffusion coefficients and turbulent Schmidt numbers exert negligible influence. To translate these findings into practical applications, a nonlinear regression model is developed and validated against experimental data, offering fast and accurate predictions of solute concentrations at junction outflows. The model is readily integrable into existing WDN solvers such as EPANET, enabling scalable and physically realistic simulation of water quality dynamics. By incorporating incomplete mixing effects, this approach significantly enhances the fidelity of disinfection strategies and contaminant risk assessments in complex urban water systems.</p><p dir="ltr">This paper was presented at the 21st Computing and Control in the Water Industry Conference (CCWI 2025) at the University of Sheffield (1st - 3rd September 2025).</p>