Dissertation submitted in compliance with the requirements for the Master's Degree in Technology: Biotechnology, Durban Institute of Technology, 2004. / M

HT biomass sonicated at ES inputs < 56KJ/gTS decreased floc size by 41% and enhanced its metabolic activity by 50-250% compared to control. ES inputs >118 KJ/gTS caused HT biomass solubilization and irreversible loss of its metabolic activity and reflocculation ability. During continuous activated sludge processing (ASP) of real primary effluent the observed yield (Yobs) decreased by 20% compared to control ASP at SF (stress factor) of 1 (biomass exchanged without USPT). At SF of 0.5, 1 and 1.5 (biomass exchanged with USPT) the Yobs further decreased by 33, 25 and 44% respectively as compared to control. This indicated that combining biomass anoxic exposure with USPT enhanced sludge reduction by increasing microbial maintenance metabolism likely in combination with microbial flora shift in the ASP depending on SF.

The objective of this investigation was to evaluate Holmberg's method for simultaneous dissolved oxygen control and parameter estimation in a completely mixed diffused air activated sludge reactor.

An investigation into the removal of phosphorus in the activated sludge process showed that: (1) Luxury uptake of phosphorus may be induced by the presence of an anoxic zone (2) If a primary and secondary anoxic reactor are present in the system, the actual retention time of the primary anoxic reactor will have the dominant effect on the degree of phosphorus removal. (3) In a system where only a primary anoxic reactor is present, there is an optimum actual anoxic retention time for obtaining maximum phosphorus removal. This optimum value appears to coincide with the condition where there is no nett release of phosphorus in the anoxic zone. Decreasing or increasing the actual anoxic retention time either side of the optimum value has a detrimental effect on phosphorus removal. (4) Increasing the influent ammonia-nitrogen and hence the reactor nitrate concentration, enhances phosphorus removal. (5) An increase in aeration reactor pH to between pH 7-8 enhances phosphorus removal. (6) The mechanism of removal does not appear to be precipitation. Although an increase in pH through the anoxic reactor was observed due to denitrification, phosphorus was released into solution. (7) The degree of phosphorus removal is independent of the influent phosphorus concentration. (8) Anoxic stripping of phosphorus is not a pre-requisite for luxury uptake of phosphorus to occur. (9) The concentration of phosphorus removed is a function of the COD utilized. The ratio of phosphorus removed to the COD utilized is constant at a particular sludge age, so that as the concentration of COD utilized increases, the concentration of phosphorus removed increases. (10) The concentration of phosphorus removed is a function of the sludge age. For a fixed biodegradable influent COD, the longer the sludge age, the less phosphorus removed.

M.S.

This research involves design and implementation of computer program for simulation of a wastewater treatment plant. The program has the capability to be interfaced with an existing analog wastewater plant process flow training board which is equipped with meter readouts of key process variables and adjustable control valves. It is planned that the total system simulator could be used to train wastewater treatment plant operators to afford them hands-on dynamic experience in plant operations. The wastewater treatment process modeled is the activated sludge process. Beginning with the known plant design stead-state equations for this particular process an algorithm was developed to simulate the treatment process through probable system dynamics. All assumptions are presented in a logical manner and used to develop the necessary transient equations. The success of this project demonstrates that a simulation program which emulates a waste treatment process is possible; however, it is suggested that further research is needed to provide deeper insight into variable changes during system transients.

Utilization of the activated sludge process is widespread although many of the mechanisms that make it work are still relatively misunderstood. Recent studies have indicated that dual substrate limitations may occur in the process. Several misconceptions in the basic fundamentals regarding the rates and mechanisms involved in oxygen transfer to wastewater systems also exist. This research investigation examined the effects of the mean cell residence time and wastewater stoichiometry on the operation of the completely mixed activated sludge process under a dual substrate limitation. At low mean cell residence times (θ_c) the system was growth C limited with respect to carbon and at high mean cell residence times the system was oxygen limited. Oxygen transfer studies were conducted to ascertain the relationship between the steady state oxygen transfer coefficient (K_La) and the oxygen uptake rate of the mixed liquor (R). The objectives of this research were accomplished by operating two continuous flow bench scale activated sludge units at COD:TKN ratios of 6.07:1 and 0.65:1. Reactor-1 was operated at a COD:TKN = 6.07:1 and was always growth limited with respect to organic carbon while Reactor-2 was operated at a COD:TKN = 0.65:1 and was carbon limited at low mean cell residence times and oxygen limited at high θ_c values. Mean cell residence time served as the primary control parameter during the laboratory studies and was varied form approximately 2.5 to 21.0 days. Theoretical studies were also conducted in which biokinetic and stoichiometric equations were used to develop a model to simulate the process operating under carbon and oxygen limitations. The model was found to yield results that were similar to the actual experimental data collected. Further refinement of the model by including inhibition functions would result in a model with better predictability. Examination of the experimental data collected during the laboratory study revealed several interesting conclusions. Operation of the activated sludge process at a low COD:TKN ratio (0.65:1) and under an oxygen limitation at high mean cell residence times can result in high levels of free ammonia and nitrite that will lead to a deterioration in effluent quality. Increased removal efficiencies for COD, TKN and NH₃-N can be achieved by operating the process at a high COD:TKN ratio (6.07:1). Steady state oxygen transfer coefficients determined in the mixed liquor of the reactors indicated there was a direct relationship to the oxygen uptake rate of the activated sludge (R). This observation is quite significant since standard aeration theory states that K_La is constant for a given aeration device. Nonsteady state K_La values determined on the effluent from each reactor indicated that K_La was a constant. Alpha and beta coefficients determined from nonsteady state tests on wastewater effluent from each reactor showed no trend with the mean cell residence time. / Ph. D.

This study compared the sludge characteristics of a lab-scale Air Products (A/O) and a conventional activated sludge system. The sludges were analyzed for operational properties and the exocellular biopolymers of each sludge was characterized. Operational analysis mainly consisted of measurements of settling (SVI) and dewatering (specific resistance, CST) parameters at each sludge age. Biopolymers were measured using pH extraction followed by gel filtration and subsequent analysis for carbohydrate and protein concentrations. The results showed a high degree of similarity between the two systems both operationally and with respect to biopolymer characteristics. The A/0 system did produce a lower effluent soluble COD at sludge ages below 5 days. Also, the A/O system showed some consistent differences in the distribution of bound and unbound fractions of protein and carbohydrate ECP but these had no discernible· effect on sludge operational characteristics. In addition relationships between unbound ECP and effluent BOD, bound HMW ECP and SVI, and phosphorus and Mg uptake were observed, with varying degrees of consistency, in this study. / Master of Science

The dewatering characteristics of sludges produced by water and wastewater treatment plants bear heavily on the methods chosen to treat and dispose of the sludge, as well as on the costs associated with handling the large volumes of sludge produced at these facilities. This study investigated why different sludges dewater to different dry solids concentrations, how sludge structure affects dewatering, and how sludge structure changes during gravity thickening and during different types of mechanical dewatering. It is generally thought that sludge can be described as having a three-tiered structure: (1) primary particles, (2) floc particles, and (3) aggregate particles. To investigate sludge structure in relation to sludge dewatering, this study has defined sludge structure using several sludge particle micro properties. A laboratory technique incorporating isopycnic centrifugation in gradients of Percoll^R media was developed to measure one micro property known as floc particle density. Six field and laboratory sludges were subjected to a series of dewatering tests: gravity thickening, centrifugation, and vacuum filtration. Each sludge was analyzed for macro- and micro properties through each stage of dewatering. It was concluded that improvements in thickening and dewatering characteristics were heavily dependent upon increases in sludge floc density and decreases in aggregate water content. / Master of Science

The purpose of this study was to investigate the interaction between colloidal substances and activated sludge and attempt to relate this information to the performance of activated sludge processes, particularly the contact stabilization process. Protein and carbohydrate organic colloids and a combination substrate were separately studied to determine if the type or classification of substrate colloids is a factor in the colloid-sludge interaction. Albumin, potato starch, and jack bean meal were the substrates chosen. Two series of studies were performed. The first involved monitoring metabolic uptake of the colloidal substrates alone and then with glucose supplementation. The second series used mercury poisoned activated sludge to investigate the physical-chemical removal of different loadings of the colloidal substrates. During these experiments, conditions of pH, ionic strength, and cationic valence were varied to determine their role in the physical interactions between the colloids and the sludge. Data obtained from the metabolic studies were monitored for unusual responses in substrate and oxygen utilization. Other factors determined included yield and oxygen utilization constants, substrate removal rates, and initial removal capacities of the sludge. Physical studies on the metabolically inactive sludge yielded data on the facility of organic colloid removal by activated sludge under a variety of different environmental conditions. Removal capacities were recorded as well as any particle size variations of the activated sludge after colloid addition. Further information was obtained on the settleability of the colloid-sludge mixtures. An adsorption and release phenomenon was observed for the two carbohydrate containing substrates while a two step oxygen utilization was observed for albumin and starch. Glucose addition was hypothesized to repress extracellular enzyme production thereby decreasing the sorption capacities of the activated sludge. Studies with the metabolically inactive sludges indicated that variations in pH, ionic strength, and cation valence play important roles in the physical removal of organic colloids by activated sludge. Sorption capacities of sludge varied with MLSS concentration. Unit sorption capacities for all three substrates decreased as MLSS levels increased. Total sorption capacities increased for albumin and jack bean meal with an increase in MLSS; the starch removal capacity however still decreased. The type of sorption occurring (adsorption or enmeshment) was believed to be important for these variances. Better quality supernatants were noted after substrate colloid addition. It was hypothesized that dispersed bacteria in the supernatant were coagulated by the organic colloids. Finally, carbohydrate colloids responded in a manner consistent with contact stabilization theory. Specifically the carbohydrates exhibited better sorption characteristics as .wel1 as an adsorption and release phenomenon. / Ph. D.