PhD Oral Exam - Fatemeh Boloukasli Ahmadgourabi, Civil Engineering

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

Drinking water utilities must maintain disinfectant residuals throughout distribution networks while limiting the loss of treated water. This trade-off is most evident in dead-end branches, where low and intermittent flows increase residence time and utilities may use continuous discharge to improve turnover. This thesis develops a decision-support framework for dead-end water quality management. Evidence from seventeen Canadian municipalities and interviews with eight utilities is synthesized into a process for selecting appropriate practices. Decay coefficients are derived from a hold study and full-scale calibration and used to compare an advection-reaction model with aggregated demands against a dead-end model combining advection-dispersion-reaction transport with stochastic demands and spatial-aggregation corrections. The advection-reaction model predicted higher residuals, changing which locations were identified for intervention. A hybrid optimization approach is developed, representing blow-off operation as a binary decision and combining efficient screening with detailed validation. It identified configurations allowing 68-73% of operational blow-offs to be closed while satisfying chlorine and pressure criteria and reducing annual treatment costs by 64-70%. A surrogate modeling approach is developed using 9.6 million detailed-model node-hour records. It was more accurate than the conventional-model baseline, reduced prediction time from two hours to 0.29 seconds, and generalized to configurations and demand scenarios withheld from training. Finally, a branch-level reliability and risk framework is developed and applied across 19,200 evaluations under empirical demand uncertainty, showing that node-hour assessment overstated reliability in every scenario. The framework identifies where continuous discharge remains necessary, where treated-water loss can be reduced, and where monitoring and intervention should be concentrated.