This thesis discusses the development of an improved microbial inactivation model for analyzing continuous flow UV-LED air treatment systems and use of the model to evaluate the impact of several treatment system design parameters on inactivation. Model development includes three submodels: a radiation submodel, a fluid flow submodel, and an inactivation kinetics submodel. Radiation modeling defines the UV irradiance throughout the system. Fluid flow modeling provides the residence times that microbes spend exposed to the UV irradiation while passing through the system. Inactivation modeling combines irradiance and residence times with inactivation kinetics to calculate species-specific inactivation in a treatment system. The most significant development focuses on the radiation submodel as it is key to linking the UV intensity emissions to treatment system properties and inactivation rates. Various radiation transfer models previously developed by other researchers are evaluated for computational efficiency and effectiveness in modeling non-uniform LED emission and diffuse and specular wall reflections. The Discrete Ordinates Method (DOM) with Legendre-Chebyshev quadrature sets is selected for use in this research due to its ability to represent both non-uniform LED emission profiles and combined specular and diffuse surface reflection. The DOM and associated quadrature schemes are reviewed in detail and limitations in representing LED emissions discussed. Sensitivity to spatial and directional discretization is evaluated. The radiation submodel is combined with a well-accepted inactivation kinetics correlation and two simple fluid flow models: a uniform flow model and a fully-developed flow model. The use and validity of these submodels is explained and their limitations discussed. Predicted microbial inactivation from the overall model is shown to compare well with limited data from a test system. Model flexibility in evaluating several system operating and design parameters is illustrated. These analyses show that for a similar number of LEDs, highly reflective surfaces (diffuse or specular) produce higher inactivation. Other parameters are shown to impact inactivation but to a lesser degree. Square ducts result in higher inactivation than non-square ducts, a fully-developed flow profile slightly increases inactivation over a uniform flow profile, positioning LEDs on all four duct walls slightly increases inactivation when surfaces are non-reflective or diffuse, and positioning LEDs closer together results in slightly higher inactivation.



College and Department

Mechanical Engineering



Date Submitted


Document Type





UV LED; microbial inactivation; UV radiation; discrete ordinates method; combined radiation reflection; air treatment system design



Included in

Engineering Commons