Although the A-stage high-rate activated sludge (HRAS) process destroys some of the chemical energy present in municipal wastewater, this process has been gaining attention as a viable technology for achieving energy neutrality at water resource recovery facilities. In addition to carbon capture for energy recovery, A-stages are also being utilized upstream of shortcut biological nitrogen removal (BNR) processes as these BNR processes often require a controlled influent carbon to nitrogen ratio that is lower than required for conventional BNR processes. While there is extensive knowledge on conventional activated sludge processes, including process controllers and activated sludge models, there has been little detailed research on the carbon removal mechanisms of A-stage processes operated at solids retention times (SRT) less than about one day. The overall objective of this study was to elucidate the chemical oxygen demand (COD) removal mechanisms of short SRT activated sludge processes with a specific focus on the removal of the different COD fractions under varying operating conditions including dissolved oxygen, hydraulic retention time, temperature, and SRT. Once understood, automatic process control logic was developed with the purpose of producing the influent characteristics required for emerging shortcut BNR processes and capturing the remaining COD with the intent of redirecting it to an energy recovery process. To investigate the removal and assimilation of readily biodegradable substrate (SS), this study evaluated a respirometric method to estimate the SS and active heterotrophic biomass (XH) fractions of the raw wastewater influent and effluent of an A-stage pilot process. The influent SS values were comparable to the SS values determined using a physical-chemical method, but the effluent values did not correlate well. This led to the measurement of the heterotrophic aerobic yield coefficient and decay rate of the pilot process. The yield coefficient was estimated to be 0.79±0.02 gCOD/gCOD, which was higher than the accepted value of 0.67 g/g. It was speculated that the batch respirometry tests resulted in the aerobic storage of SS and this likely contributed to the error associated with the determination of SS and XH. Therefore, physical-chemical fractionation methods were used to determine the removal of the individual COD fractions. It was concluded that the SRT was the primary control parameter and below a 0.5 day SRT the dominate COD removal mechanisms were assimilation and oxidation of readily degradable substrate and sedimentation of particulate matter. At SRTs between 0.5-1 days, COD removal became a function of hydrolysis, as adsorption of particulate and colloidal matter was maximized but not complete because of limited adsorption sites. Once adequate adsorption sites were available, effluent quality became dependent on the efficiency of bioflocculation and solids separation. While the SRT of the pilot process could not be directly controlled because of severe biofouling issues when using in situ sensors, a MLSS-based SRT controller was successfully implemented instead. The controller was able to reduce total COD removal variation in the A-stage by 90%. This controller aslo provided the capability to provide a consistent carbon to nitrogen ratio to the downstream B-stage pilot process. To ascertain the settling, dewaterability, and digestibility of the sludge produced by the pilot A-stage process, several standardized and recently developed methods were conducted. The results from these tests indicated that the A-stage had similar dewaterability and digestibility characteristics to primary sludge with average achievable cake solids of 34.3±0.4% and average volatile solids reduction (VSR) of 82±4%. The A-stage sludge also had an average specific methane yield of 0.45±0.06 m3CH4/kgVS. These results were attributed to low extracellular polymeric substance (EPS) content. However, further research is needed to better quantify EPS and determine the effect of HRAS operating parameters on EPS production. Overall the A/B pilot study was able to capture 47% of the influent COD as waste sludge while only oxidizing 45% of the influent COD. Of the COD captured, the A-stage contributed over 70% as dry solids. Coupled with high sludge production, VSR, and methane yield the A/B process was able to generate 10-20% more biogas and 10-20% less dry solids after anaerobic digestion than a comparable single-sludge BNR process. / Ph. D.

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