Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Civil and Environmental Engineering

Major Professor

Jeffrey A. Cunningham, Ph.D.

Committee Member

Sarina J. Ergas, Ph.D.

Committee Member

Aydin Sunol, Ph.D.

Committee Member

Kathleen Scott , Ph.D.

Committee Member

Luke Mulford, Ph.D.


BioWin, energy, microbial fuel cell, nutrient removal, sidestream enhanced biological phosphorus removal


A variety of process configurations are employed at municipal water reclamation facilities (WRFs), such as 5-stage Bardenpho, oxidation ditch, modified Ludzack-Ettinger (MLE) process, and sidestream recycle. However, all of these configurations face certain challenges in achieving optimum treatment of nitrogen and phosphorus. In this dissertation, the driving objectives were to: (1) quantitatively assess the fate of nitrogen and phosphorus (mass fluxes) at a treatment facility that employs biological nutrient removal, aerobic digestion, and sidestream recycle, (2) evaluate the performance of a microbial fuel cell (MFC) as a technology to remove nitrogen from sidestreams at a treatment plant that employs that employs biological nutrient removal, aerobic digestion, and sidestream recycle, (3) determine the fate of phosphorus at a treatment facility that employs an oxidation ditch and is characterized by simultaneous nitrification and denitrification, and (4) assess the potential of implementing enhanced biological phosphorus removal (EBPR) to MLE process via the addition of recycled activated sludge (RAS) fermentation, using both a lab-scale continuous treatment MLE system and a BioWin model of a treatment facility that employs MLE.

To achieve objective 1 of this dissertation, mass fluxes of nitrogen and phosphorus were quantified through the treatment trains at the Northwest Regional Water Reclamation Facility (NWRWRF) and the adjoining Biosolids Management Facility (BMF) in Hillsborough County, Florida. I determined that nearly half of the overall phosphorus flux into NWRWRF was recycled from the BMF sidestream. This leads to an increased cost of treatment, e.g., for alum used in phosphorus removal at NWRWRF. In contrast to phosphorus, the flux of nitrogen from BMF to NWRWRF is small (~3%) compared to the flux of nitrogen entering NWRWRF in raw wastewater. However, nitrogen in the sidestream is mostly in the form of nitrate, which prevents anaerobic conditions from developing in the fermentation basin at NWRWRF, and thereby interferes with the EBPR process. Some measurements suggest that fermentation and release of phosphorus may occur in the RAS line, which supports EBPR and may partially compensate for anoxic (denitrifying) conditions in the fermentation basin. Therefore, overall, NWRWRF was able to meet its permit limits for phosphorus through a combination of EBPR and alum addition. The general trends observed are likely to apply to many similar facilities that employ biological nutrient removal, aerobic digestion, and sidestream recycle. I recommend that such facilities consider (1) removal or recovery of phosphorus from their sidestreams and (2) returning sidestreams downstream of fermentation basins to avoid inhibition of EBPR processes.

I constructed a bench-scale MFC and operated it for a period of 125 days to remove nitrogen (nitrate) from an aerobic digester sidestream from BMF. The average removal rate of nitrogen was 14 mg/(L∙d), the average power production was 0.38 mW/m2 of electrode surface area, and the apparent efficiency of electron transfer from anode to cathode was 41%. The nitrogen removal rate and apparent electron transfer efficiency are similar to those observed in previous MFC studies treating other nitrate-containing streams via cathodic denitrification. The low power generation may be due partly to the two-chamber configuration of the MFC employed, which was appropriate for the goals of this study but is not the most advanced MFC configuration. Therefore, I conclude that the MFC remains a promising candidate for nitrate removal from aerobic digester sidestreams, but that more advanced MFC configurations than the one employed here will be required for the technology to be viable at a larger scale.

I analyzed data collected by a previous student from Falkenburg Advanced Wastewater Treatment Plant (FAWTP) in Hillsborough County, Florida. Samples from six locations at FAWTP were collected and tested for total phosphorus, orthophosphate, ammonium, nitrate, nitrite, alkalinity, and pH. Results indicated the occurrence of simultaneous nitrification and denitrification (SND) and EBPR in the oxidation ditch at FAWTP. Ammonium average removal efficiency was 99.5 %. The net production of nitrite and nitrate was insignificant, indicating the conversion of ammonium to nitrogen gas. EBPR was responsible for 90% of phosphorus removal and chemical precipitation was responsible for the remainder.

EBPR via RAS fermentation was evaluated at a MLE process using: (1) a laboratory-scale continuous flow system with influent collected from NWRWRF and (2) a BioWin model of Pinellas County’s South Cross Bayou (SCB) Water Reclamation Facility that employs MLE configuration for wastewater treatment. Despite, phosphorus release in the RAS fermentation reactor, both the laboratory-scale system and the BioWin model were not successful in achieving removal of orthophosphate. In the laboratory-scale system and the BioWin model, orthophosphate in the fermentation reactor reached 24.7 mg/L as P and 10.2 mg/L as P, respectively. In the lab-scale system, average orthophosphate concentration in the effluent was 5.3 mg/L compared to 4.3 mg/L in the influent. In the BioWin model, average orthophosphate concentration in the effluent was 2.7 mg/L compared to 2 mg/L in the influent. RAS fermentation might have been limited by low supply of readily biodegradable chemical oxygen demand (rbCOD), in which case fermenters would have been relying on endogenous decay, which slowed down the growth rate and VFA production. One way to test this hypothesis is to analyze rbCOD in the RAS. A possible solution to this problem could be adding raw wastewater or methanol to the RAS fermentation reactor for the necessary supply of rbCOD