A critical element to the successful operation of activated sludge systems is efficient solid liquid separation achieved by bioflocculation. Bioflocculation refers to the process of microbial aggregation to form activated sludge flocs, dependent on the interaction of exocellular polymeric substances (EPS) to form the matrix that holds microbes, other organics and inorganic particles in a flocculent mass. Numerous factors affect bioflocculation; two key parameters are the Solid Retention Time (SRT) and the substrate loading rate. The latter is related to the two basic designs in activated sludge bioreactor configurations: the Plug Flow Reactor (PFR) and the Completely Stirred Tank Reactor (CSTR). PFR systems have a high substrate loading rate, whereas CSTRs have a low substrate loading rate. Research has shown that the PFR configurations produce better sludge quality, in terms of settleability and dewaterability, and subsequently better effluent quality than CSTR systems. In this experiment, the effect of SRT and substrate loading rate on activated sludge was investigated using bench scale SBRs. PFR and CSTR configurations were simulated by adjusting the fill period to be shorter or longer respectively. A series of SBRs were operated, each with an operating volume of 6L, to obtain data for PFR (fast feed) versus CSTR (slow feed) configurations at 10 Day, 5 Day and 2 Day SRTs. Effluent quality was monitored by measuring effluent TSS, VSS, total and soluble COD and soluble biopolymers. Sludge quality was monitored for the aerobic phase by measuring total and suspended solids, total and suspended volatile solids, Sludge Volume Index (SVI), Capillary Suction Time (CST) and Zeta Potential. Anaerobic digestibility was measured for the sludge produced in these systems by measuring gas production, similar to estimating biogenic methane potential (BMP) and determining short term odor productions, specifically Total Volatile Organic Sulfur Compounds (TVOSCs). As expected the change in feeding pattern and SRTs affected the effluent and sludge quality during the aerobic operation phase. Effluent quality was found to be better for the fast feed system at all SRTs, with all monitored parameters being of similar or significantly lower concentration than for the slow feed system. In terms of sludge quality, the fast feed system was found to retain more of its biomass in solution, indicating better flocculation and settleability in this system. COD was given a lower rank as an effluent quality indicator, since the 5 Day and 2 Day SRT datasets did not correlate well with other datasets, specifically effluent TSS and biopolymers. The data was included because it is believed that the trends were accurate representations of fast versus slow feed system behavior. The trends were comparable to those of effluent TSS and solution biopolymer datasets. In terms of anaerobic digestion potential, the fast feed sludge exhibited greater volumetric gas production per gram of solid at the 5 and 2 Day SRTs. Gas production was similar for both systems at the 10 Day SRT. Total and Volatile Solid reduction were however found to be higher for the slow feed sludge than for the fast feed. This may indicate higher gas and potential odor production per gram of solid degraded for the fast feed sludge. This theory is supported by the odor analyses, which revealed that the fast feed sludge had a higher TVOSC production at each SRT. This was related to the higher protein content of the sludge, indicated by the effluent biopolymers being much higher in protein content than carbohydrates. Shearing, which is part of the solids handling process at most plants, releases these proteins and makes them bioavailable, allowing them to be oxidized to produce TVOSCs and hence higher odors. In conclusion it was found that the fast feed effluent and sludge quality appeared to be overall better at each SRT simulated; the higher TVOSC content may indicate a problem with solids handling, but research has shown that these can be overcome with the addition of iron. Additionally, both systems, the fast and slow feed systems operated better at longer SRTs, with the fast feed system performing better in all cases. The difference was not completely significant in all cases and this is attributed to being a by-product of operating at the optimal M:D salt ratio. This project has strength in terms of its potential for large scale applications. SRT is the considered the most important design parameter and one of the more complicated parameters to manipulate due to its widespread effect on reactor behavior, specifically sludge and effluent quality. Additionally, the fast feed versus slow feed concept is one that has been gaining significant interest, since bioreactor configuration impacts the effluent and sludge quality. Feed configurations have been investigated more frequently within the past decade. The novel approach taken by this project is that it combines these two parameters, both of which are important to large scale plants, both industrial and municipal. / Master of Science

An activated sludge system that incorporates a sidestream anaerobic bioreactor, called the Cannibal process, was the focus of this study. A prior study of this process (Novak et al., 2007) found that this system generated about 60% less solids than conventional activated sludge without any negative effects on the effluent quality. Although that study showed substantial solids reduction, questions remain concerning the specific mechanism(s) that account for the solids loss. In this study, the mechanisms that account for the loss of biological solids was the focus of the investigation. The first part of this study was conducted to evaluate those effects in terms of the role of iron in the influent wastewater and feeding patterns on the performance of the Cannibal system. It was found that the Cannibal system with high iron in the influent produced less biological solids than the system receiving low iron. The data also showed that the Cannibal system operated under fast feed (high substrate pressure) produced much less solids than the system with slow feed (low substrate pressure). The high substrate pressure was achieved by feeding the influent wastewater to the Cannibal system over a short time period so that the substrate concentration would initially peak and then decline as degradation occurred. This is called "fast feed." For low substrate pressure, the influent was added slowly so the substrate concentration remained low at all times. This is called "slow feed." Later, an attempt to increase substrate pressure in the slow feed Cannibal system was conducted by either manipulating the aeration patterns or adding a small reactor in front of the main reactor (selector). It was found that either interrupting aeration in the aerobic reactor or providing a small aerobic reactor in front of the main reactor resulted in an increase in solids reduction. The second part of this study was to investigate the mechanisms of floc destruction in the fast and the slow feed Cannnibal systems. It was found that higher accumulation of biopolymers (proteins and polysaccharides) occurred in the fast feed system and this was associated with a greater solids reduction in the fast than the slow feed system. In addition, more protein hydrolysis and more Fe(III)-reducing microorganism activity in the fast feed environment were found to be factors in higher solids reduction. The last part of this study was to investigate the structure of the Cannibal sludge flocs generated under the fast and the slow feed conditions. It was found that the readily biodegradable (1 kDa.) protein is larger in the flocs from the fast feed than the slow feed Cannibal system. This resulted in higher floc destruction in the fast feed condition. / Ph. D.