Graduation Year


Document Type




Degree Granting Department

Civil and Environmental Engineering

Major Professor

Andrés E. Tejada-Martínez, Ph.D.

Co-Major Professor

Qiong Zhang, Ph.D.

Committee Member

James R. Mihelcic, Ph.D.

Committee Member

Muhammad M. Rahman, Ph.D.

Committee Member

Hongxia Lei, Ph.D.


computational fluid dynamics, ozone disinfection, reaction kinetics, turbulent flow, water treatment


In the last two decades, Computational Fluid Dynamics (CFD) has shown great potential as a powerful and cost-efficient tool to troubleshoot existing disinfection contactors and improve future designs for the water and wastewater treatment utilities.

In the first part of this dissertation two CFD simulation methodologies or strategies for computing turbulent flow are evaluated in terms of the predicted hydraulic performance of contactors. In the LES (large eddy simulation) methodology, the more energetic, larger scales of the turbulence are explicitly computed or resolved by the grid. In the less computationally intensive RANS (Reynolds-averaged Navier-Stokes) methodology, only the mean component of the flow is resolved and the effect of the unresolved turbulent scales is accounted for through a turbulence model. For baffled contactors, RANS performs on par with the LES in predicting hydraulic performance indices. In this type of contactors, hydraulic performance is primarily determined by quasi-steady recirculating (dead) zones within the contactor chambers which are well-resolved in both RANS and LES. Testing of the RANS methodology is also performed for a wastewater stabilization pond leading to prediction of hydraulic performance indices in good agreement with field measurements. However, for column contactors, LES performs better than RANS due to the ability of the LES to resolve unsteady or unstable flow structure associated with spatial transition to turbulence which is important in the determination of the hydraulic performance of the contactor.

In the second part of this dissertation the RANS methodology is adapted in order to develop a novel modeling framework for ozone disinfection of drinking water. This framework is unique as it combines CFD with kinetics-based reaction modeling to predict disinfection performance and bromate formation for the first time. Bromate, a human health hazard, is an undesired by-product of the disinfection of drinking water via ozonation. The modeling framework is validated via application to a full-scale ozone contactor. Predictions of ozone and bromate concentrations are consistent with data from physical experiments.