Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Environmental Engineering and Science

Committee Member

Tanju Karanfil Ph.D., (Committee Chair)

Committee Member

Brian A. Powell, Ph.D.

Committee Member

Cindy Lee, Ph.D.

Committee Member

David A. Ladner, Ph.D.


The seasonal (summer, fall, winter and spring) and local weather related patterns of N-nitrosodimethylamine (NDMA) formation potentials (FPs) were examined with water samples collected monthly for a two year period in 12 surface waters. This long term study allowed monitoring the patterns of NDMA FPs under dynamic weather conditions (e.g., high/low rainfall periods) covering several seasons. Anthropogenically impacted source waters (SWs) which were determined by relatively high sucralose levels (>100 ng/L) had higher NDMA FPs than limited impacted SWs (<100 ng/L). In some sources, NDMA FP showed more variability in spring months. However, seasonal mean values were in general relatively consistent in most sources. These results showed that watershed characteristics played an important role in NDMA FP levels. For one of the sampled surface waters, a large reservoir on a river examined in this study appeared to serve as an equalization basin for NDMA precursors. In contrast, in a river without an upstream reservoir, the NDMA FP levels were influenced by the ratio of an upstream wastewater treatment plant (WWTP) effluent discharge to the river discharge rate. The impact of WWTP effluent decreased during the high river flow periods due to rain events. Linear regression with independent variables consisting of dissolved organic carbon (DOC), dissolved organic nitrogen (DON), and sucralose yielded poor correlations with NDMA FP (R2<0.27). However, multiple linear regression analysis using DOC and log (sucralose) yielded a better correlation with NDMA FP (R2=0.53). This study also examined the removal of NDMA FPs for various operational conditions (e.g., alum clarification, powdered activated carbon [PAC] application, use of polymers, pre- and post-oxidation with chlorine [Cl2] and/or chlorine dioxide [ClO2], and Ct [concentration x contact time] changes) at full-scale water treatment plants (WTPs). The impacts of different seasons and dynamic local weather conditions (e.g., high/low rainfall periods) on the removal efficiency of NDMA FP at nine WTPs were investigated for a two year period. NDMA FP removal by alum clarification process remained between 12 to 30% for different seasons and temporal weather conditions. PAC addition (>4 mg/L) increased significantly the NDMA FP removal and PAC doses showed a good correlation (R2=0.71) with the NDMA FP removal. The contribution of polymers to NDMA FP depended on the polymer type used and concentration. The simultaneous application of Cl2 and ClO2 for pre-oxidation and post-oxidation were beneficial for the removal of additional NDMA FPs. The average NDMA FP removals for reverse osmosis (RO) and microfiltration (MF) units were 81% and 7%, respectively. The effect of ClO2 oxidation on the control of NDMA precursors was investigated for different background waters (e.g., low/non-impacted vs. wastewater [WW]-impacted) under various oxidation conditions (e.g., pH, oxidant dose, and Ct). The removal of NDMA FP from all water types (low/non- or WW-impacted) was ≤25% at pH 6.0 with ClO2 oxidation. However, under the similar oxidation conditions, NDMA FP removals increased up to ~80% with increasing influence (i.e., 10%, 25% and 50%) of WW effluents at pH 7.8. This indicates that the majority of WW-derived NDMA precursors can be deactivated with ClO2 oxidation at higher pH (≥7.8). This was due to the better oxidative reaction of ClO2 with amines that have lone pair electrons to be shared at higher oxidation pH conditions. Similarly, NDMA formation levels under uniform formation condition (UFC) also significantly decreased in WW-impacted waters with ClO2 oxidation at pH 7.8. Furthermore, natural attenuation simulation experiments were conducted in this study, and the results showed that ClO2 oxidation can be more effective to deactivate NDMA precursors from relatively "freshly" impacted waters. Finally, the effects of different oxidation scenarios (individual [ClO2 or Cl2 only] and integrated [simultaneous or sequential application of ClO2 and Cl2]) on the removal of NDMA FP from different waters (e.g., non-impacted vs. either 20% wastewater- or polymer-impacted waters) were investigated. The removal efficiency of NDMA FP in non-impacted waters for all the oxidation scenarios was ≤25%. In 20% WW-impacted waters, NDMA FP removals improved about 50% by only ClO2 oxidation at pH 7.8 (~75%) compared to pH 6.0 (≤25%). However, the increase was limited for only Cl2 oxidation case under same oxidation conditions. For integrated oxidations, NDMA FP removals increased (20-45%), especially, at pH 6.0 compared to individual application of oxidants (Cl2 or ClO2). This indicates that ClO2 and Cl2 can react with different amine precursors at lower pH (i.e., 6.0). However, increasing oxidation pH from 6.0 to 7.8 transformed amine precursors to more reactive forms for both oxidants. Furthermore, integrated use of Cl2 and ClO2 also increased (10-40%) the removal of polymer-derived (poly epichlorohydrin dimethylamine [polyamine], poly diallyldimethylammonium chloride [polyDADMAC] and polyacrylamide [Sedifloc 400C]) NDMA precursors independent of oxidation time (10 vs. 60 min) and pH (6.0 vs. 7.8).