Award Date

May 2018

Degree Type


Degree Name

Master of Science in Engineering (MSE)


Civil and Environmental Engineering and Construction

First Committee Member

Daniel Gerrity

Second Committee Member

Jacimaria Batista

Third Committee Member

Erica Marti

Fourth Committee Member

Boo Tseng

Number of Pages



Potable reuse has been growing as a strategy to augment water supplies, especially in highly populated and water-scarce regions. Ozone and chloramines have emerged as important disinfectants and oxidants in potable reuse applications, but reactions with wastewater-derived constituents can lead to the formation of potentially carcinogenic disinfection byproducts (DBPs). One DBP that has received considerable attention is the nitrogenous DBP N-nitrosodimethylamine (NDMA). NDMA is a potential human carcinogen and mutagen at trace concentrations — even at the sub-ng/L level. Several studies have reported successful attenuation of NDMA in biofiltration systems at wastewater treatment plants, but the associated mechanisms and design criteria are not well understood.

In the current study, a pilot-scale ozone-biofiltration system was used to treat membrane bioreactor (MBR) filtrate from a full-scale water reclamation plant to assess the role of various operational parameters, including ozone dose and empty bed contact time (EBCT), on NDMA removal. In the ozonated biological activated carbon (BAC) and anthracite columns, longer EBCTs (e.g., 10-20 minutes) achieved >90% NDMA removal, while shorter EBCTs (e.g., 2 min) achieved only 30-40% NDMA removal. In the non-ozonated BAC column, longer EBCTs were more important, with NDMA attenuation exhibiting a relatively steady increase toward ~45% for an EBCT of 20 min.

Pre-oxygenation of the MBR filtrate (i.e., instead of ozonation) also achieved ~90% removal in the BAC column, thereby suggesting that biodegradable dissolved organic carbon (BDOC) availability did not impact NDMA removal. Interestingly, when receiving ambient MBR filtrate (no pre-oxygenation or pre-ozonation), the typically ozonated column still achieved >90% NDMA removal, while the typically non-ozonated column only achieved 50% NDMA removal. In other words, NDMA removal was dependent on EBCT but did not necessarily require high concentrations of BDOC or dissolved oxygen. Instead, long-term exposure to ozonated MBR filtrate may have been critical in promoting the development of microbial taxa that were better adapted to NDMA biodegradation. The presence of monooxygenase genes responsible for NDMA biodegradation was confirmed by quantitative polymerase chain reaction (qPCR), although possible DNA extraction limitations for the BAC media prevented a reliable comparison by media type. Finally, this study confirmed the efficacy of ozonebiofiltration (but not biofiltration alone) for attenuating chloramine-reactive NDMA precursors. An overall reduction of 96% was observed, with a majority of that attenuation achieved by ozonation because of its ability to transform primary and secondary amines into nitrated intermediates and tertiary amines into N-oxides.

These data suggest that ozone-biofiltration is effective in achieving net reductions in NDMA in some potable reuse systems, particularly when chloramines are expected to be used as a final disinfectant. However, UV photolysis might still be necessary as a final polishing step to ensure compliance with relevant guidelines and

regulations (e.g., 10-ng/L notification level in California). Also, additional studies are needed to better characterize microbial community structure and function in potable reuse systems.


Biofiltration; NDMA; Ozone; Potable Reuse; qPCR; TOC


Environmental Engineering