The Chesapeake Bay watershed has seen an increase in population, nutrient loading, and stringent effluent limits; therefore, cost-effective technologies must be explored and implemented to intensify the treatment of regional wastewater. This work describes the bioaugmentation and retention of anammox (AMX) granules in a continuous adsorption/bio-oxidation (A/B) mainstream deammonification pilot-scale process treating domestic wastewater. The AMX granules were collected from the underflow of a sidestream DEMON® process. The bioaugmentation rate was based on several factors including full-scale sidestream DEMON® wasting rate and sidestream vs mainstream AMX activity. The retention of bioaugmented AMX granules required a novel settling column at the end of the deammonification step. The settling column was designed to provide a surface overflow rate (SOR) that allowed dense AMX granules to settle into the underflow and less dense floccular biomass to outselect into the overflow. B-Stage was operated to out-select nitrite oxidizing bacteria (NOB) by maintaining an ammonia residual (>2 mg NH4-N/L), a relatively high dissolved oxygen (DO) (>1.5 mg O2/L) concentration, an aggressive solids retention time (SRT) for NOB washout, and intermittent aeration for transient anoxia. AMX activity was not detected in the mainstream at any time. The settling column AMX retention quantification suggested but did not confirm AMX were maintained in the mainstream. NOB were not suppressed during this study and no nitrite accumulation was present in the mainstream process. It was theorized that AMX granules were successfully settled into the settling column underflow and accumulated in the intermittently mixed sidestream biological phosphorus reactor (SBPR) where they disintegrated. This work also describes optimization of carbon addition to an anoxic partial denitrification anammox (PdN/A) moving bed biofilm reactor (MBBR) testing glycerol, acetate, and methanol as carbon sources to maximize total inorganic nitrogen (TIN) removal through the anammox pathway and to minimize effluent TIN. A carbon feeding strategy was developed and was evaluated by the extent of partial denitrification vs full denitrification (partial denitrification efficiency, PdN efficiency). All three carbon sources were capable of high TIN removal, low effluent TIN, and moderate to high PdN efficiency. Average TIN removal for glycerol was 10.0 ± 3.6 mg TIN/L, for acetate it was 8.7 ± 2.9 mg TIN/L, and for methanol it was 11.5 ± 5.6 mg TIN/L. Average effluent TIN for glycerol was 6.0 ± 4.0 mg TIN/L, for acetate it was 5.0 ± 1.1 mg TIN/L, and for methanol it was 4.3 ± 1.5 mg TIN/L. Average PdN efficiency for glycerol was 91.0 ± 9.0%, for acetate it was 88.0 ± 7.7%, and for methanol it was 74.0 ± 8.5%. When PdN efficiency was factored into the cost of each carbon source, methanol was 5.83% cheaper than glycerol per mass TIN removed and 59.0% cheaper than acetate per mass TIN-N removed. / Master of Science / The Chesapeake Bay watershed has seen an increase in population, nutrient loading, and stringent effluent limits; therefore, cost-effective technologies must be explored and implemented to intensify the treatment of regional wastewater. This work involves removing nitrogen from wastewater in a pilot sized modeled from a real wastewater treatment plant. The removal of nitrogen from wastewater can become costly. This cost is due to aeration and chemical demands to remove the nitrogen. This masters work uses a type of microorganism that can remove nitrogen without the need for aeration or chemicals through anaerobic ammonia oxidation (AMX bacteria). A specific environment has been created for AMX bacteria during this study to ensure they perform nitrogen removal optimally. Often times, communities of bacteria can help remove nitrogen more effectively when they work together. Therefore, communities of bacteria were encouraged to grow during this study. We were able to see that nitrogen removal was indeed occurring at high rates and producing high effluent water quality. We used several different metrics to prove this nitrogen removal technology worked well. This research was important because it showed the capabilities of a highly intensified process of successful nitrogen removal at a pilot-scale facility. It is the hope that these findings can be improved upon and implemented at full-scale facilities. These full-scale facilities would be able to achieve low levels of nitrogen in their effluent while saving millions of dollars on operational costs.

Research was performed at a pilot-scale wastewater treatment plant operating an adsorption/bio-oxidation (A/B) process at 20C. The study compared B-Stage performance under DO Control, Ammonia Based Aeration Control (ABAC), and Ammonia vs. NOx (AvN) control. AvN in 1) fully-intermittent and 2) intermittently-aerated MLE configurations was compared to DO Control and ABAC, each with continuous aeration, in an MLE configuration. The study also examined operation of each aeration strategy with two different feed types: A-Stage effluent (ASE) and primary clarifier effluent (PCE). Operating modes were compared on the basis of nitrogen removal performance, COD utilization efficiency for denitrification, and alkalinity consumption. AvN was found to provide comparable nitrogen removal performance to DO Control and ABAC. The highest nitrogen removal performance was seen when operating DO Control (81.4 ± 1.2%) and ABAC (81.1 ± 1.2%) with PCE. High nitrogen removal efficiency (77.5 ± 6.1%) was seen when fully-intermittent AvN operation was fed ASE containing a high particulate COD fraction. A high effluent nitrite accumulation ratio (NAR = NO2-/(NO2-+NO3-)) was seen during this period (46 ± 15%) accompanied by the out-selection of Nitrospira. Feeding effluent from AvN control to an Anammox MBBR improved removal efficiency. Increased soluble COD loading resulted in greater nitrogen removal with strategies operating in an MLE configuration while particulate COD was found to be important for processes where removal was designed to occur in downstream reactors. Efficiency of COD for denitrification was found to vary based on the amount and type of influent COD; however AvN in an MLE configuration was found to use COD more efficiently than fully-intermittent AvN. In either configuration, AvN required less alkalinity addition than DO Control or ABAC. High sCOD concentrations in PCE led to increased nutrient removal as compared to ASE but increased heterotrophic growth and mixed liquor concentrations in the B-Stage making the A-Stage an attractive option for its ability to control the C/N ratio fed to BNR processes. / Master of Science