Numerical and experimental investigation of UV LED water treatment reactor

Shah, Jainil and Aggidis, George and Zidonis, Audrius (2025) Numerical and experimental investigation of UV LED water treatment reactor. PhD thesis, Lancaster University.

Text (2025jainilphd)
2025jainilphd.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (4MB)

Abstract

UV LED water treatment systems represent an emerging alternative to conventional mercury lamp reactors, offering energy efficiency, compact design, and environmental benefits. However, achieving effective pathogen inactivation in commercial-scale LED reactors requires optimized flow structures to ensure uniform UV dose distribution throughout the treated water. This thesis presents the first comprehensive investigation of swirl-enhanced flow dynamics in commercial-scale UV LED reactors through integrated computational and experimental approaches. A validated CFD model was developed using ANSYS CFX with SST turbulence modeling, coupled with ray-tracing optical calculations to predict UV dose distribution. The model was rigorously validated through dual approaches: biodosimetry experiments using MS2 bacteriophage across nine operating conditions (three flow rates: 80-250 m³/h; three UV transmittance levels: 90-98%) achieved excellent agreement with the numerical model, well within experimental uncertainty (±29.4%). Particle Image Velocimetry measurements in a scaled reactor (DN200) validated flow field predictions for both tangential and axial velocity components. The validated model quantified key performance drivers: stationary vanes contribute >97% of swirl generation (swirl number S = 0.37), providing 30% residence time enhancement leading to an improved mixing uniformity. Comprehensive geometric analysis established design guidelines: LED ring spacing <40 mm optimizes efficiency for UV transmittance <85%; upstream piping configurations (U-bends, T-junctions, reducers) produce <5% dose variation, eliminating straight-pipe installation requirements. This research provides the first validated predictive framework for commercial UV LED reactor design, enabling rational optimization of geometric configurations, LED arrangements, and operating conditions. The findings have been implemented in commercial product development, reducing prototype testing requirements.