Search

The different kinds of aerated lagoons, which exclude anaerobic pre-treatment ponds, are described and the design approach for aerated lagoons is explained. This hinges around ensuring that the 1st lagoon is suspension mixed and the second and any additional are facultative. Selection of the retention time for the 1st lagoon is important to ensure complete utilization of the influent biodegradable organics. Minimum retention times to achieve this at 14 degreesC and 22 degreesC were determined with the general activated sludge kinetic simulation model for (i) readily biodegradable soluble organics (BSO) only, (ii) slowly biodegradable particulate organics (BPO) only, (iii) real municipal wastewater (20% BSO and 80% BPO) and (iv) real municipal wastewater with 5% OHO active VSS mass seed. The minimum hydraulic retention times for these four cases are: at 14 degreesC 1.3, 3.0, 2.0 and 1.5 d, respectively, and at 22 degreesC 0.3, 2.0, 1.2 and 1.0 d, respectively. From a comparison of the simulation results with the steady-state model calculations, washout of OHOs takes place at about 75% of these retention times. Approximate equations to estimate the power requirements for aeration by mechanical surface aerators and mixing are given. These equations are combined with those of the steady-state activated sludge lagoon model for calculating the oxygen requirements and the aeration power density (W/m3) in each lagoon. With these equations, it is shown that influent COD concentration needs to be between an upper and lower limit band to ensure that the 1st lagoon is suspension mixed and the second lagoon is facultative. This COD concentration band decreases as the influent flow increases. The important conclusion arising from this is that if the aerated lagoon system is applied for small rural communities, where land for these large systems is likely to be available, then additional mixing energy above that for aeration will need to be provided to ensure that the 1st lagoon is suspension mixed - this additional aeration cost makes it unlikely that aerated lagoons will be applied for municipal wastewater treatment. Matching mixing and aeration power requirements for industrial organic wastewaters is easier because these usually are significantly stronger than municipal wastewaters.

The characterization of wastewater and biomass in relation to the Activated Sludge Model No. 2 is described. A new fraction of organic fermentable matter is needed. Phosphate accumulating organisms and their structural compounds polyphosphate and polyhydroxyalkanoate have to be included in the biomass characterization. There is still a need for development of analytical methods for characterization of the various components.

The Activated Sludge Model No. 2 is introduced as a further development of Activated Sludge Model No. 1. Model No. 2 introduces phosphorus accumulating organisms (PAO) and allows us to simulate the behaviour of biological nutrient removal activated sludge systems. Typical wastewater composition and a set of stoichiometric and kinetic parameters are provided in order to make reasonable predictions. The model has not yet been calibrated or verified in any full-scale application. This report contains a simplified version of the full Model No. 2; the full model is published in the final report of the Task Group (Henze et al., 1994).

Successful application of upflow anaerobic sludge bed (UASB) technology hinges on the generation of sludge aggregating into bio-conglomerates, to facilitate retention of the sludge in the reactor. In this paper an hypothesis for formation of one type of bio-conglomerate, the biopellet, is presented. From the hypothesis, characteristics that a waste must possess to induce pelletization, and process design and operational procedures to optimize system performance, are identified.

Experiments have been performed to investigate the hypothesis that the alternation of anoxic and aerobic conditions is the major factor influencing filamentous bulking in low F/M systems. The results provided strong supporting evidence for the hypothesis; nitrite, rather than nitrate, appears to play a dominant role in causing the bulking.

From laboratory research and a literature review of the biochemical pathways of aerobic-facultative heterotrophic organisms, an hypothesis is proposed for the proliferation of anoxic-aerobic (AA) filamentous organisms in nitrification-denitrification (ND) and nitrification-denitrification biological excess phosphorus removal (NDBEPR) systems. In activated sludge, under anoxic conditions floc-forming organisms execute the denitrification of nitrate (NO3−) through each of the denitrification intermediates to dinitrogen (N2), in the process of which the intermediate nitric oxide (NO) is accumulated intracellularly. Intracellular NO is inhibitory to the utilization of oxygen in the subsequent aerobic zone. In contrast, the filamentous organisms execute only part of the denitrification pathway, i.e. the reduction of NO3− to NO2−; they do not accumulate NO and hence are not inhibited in the subsequent aerobic zone. Thus in anoxic-aerobic systems, floc-formers are placed at a disadvantage in the aerobic zone giving an advantage to the filaments in the competition for substrate. Experimental evidence to support this hypothesis is presented and a tentative proposal of a strategy for control of AA filament proliferation is described and tested experimentally.

A suite of semi-empirical models has been developed to quantify the transport of phosphorus in the Berg River. This paper describes the results of the model simulations to quantify (1) the input of phosphorus from point and nonpoint sources, and (2) the transport of phosphorus in the water prism along the main river channel. Export of phosphorus from nonpoint sources is influenced by the drainage flow during storm events. Mass export under the rising limb of the drainage hydrograph is higher than under the falling limb giving rise to a hysteresis relationship between concentration and flow. Implementation of the 1 mg P/l effluent standard for point sources is shown to have a minor influence on the phosphorus budget of the river - nonpoint sources contribute over 80 percent of the annual load. In the Berg River, 80 percent of the phosphorus export takes place during storm events which occupy less than 3 percent of the annual hydrological cycle.

A 5 pH point acid titration method is proposed to measure the short chain fatty acids, total carbonate species concentration and H2CO3*alkalinity in anaerobic systems. To test the method, aqueous solutions containing NaHCO3 (1990; 2488 mg/ℓ as CaCO3) and the short chain fatty acid (SCFA) acetic acid with concentrations ranging from 100 to 1000 mg/ℓ were titrated using the method; the H2CO3*alkalinity and SCFA estimates ranged around the expected values with an average standard deviation of 5 and 8 percent respectively. Estimates of the SCFA by the method and by Montgomery et al 's wet chemical method on the effluents of UASB reactors treating brewery and wine distillery waste correlated closely (r = 0.98). Other tests demonstrated as predicted that the presence of ammonium has a negligible effect on the derived results. However, phosphates can have a significant effect on the H2CO3*alkalinity but not on the SCFA estimates. The phosphate effect is eliminated in the derived estimates if its concentration is known. The method also identifies errors in pH due to the residual liquid junction effect or poor calibration and corrects the estimates accordingly; deliberately imposed pH calibration errors were identified and their error effects on the H2CO3*alkalinity and SCFA estimates largely corrected.

Grape wine distillery waste developed a pelletised sludge bed in a UASB system. Product formation along the line of flow in the pelletised bed was similar to that when treating a pure carbohydrate, apple juice waste. Pelletised sludge production was about 0.14 mgVSS/mgCOD removed (as against 0.42 mgVSS/mgCOD removed for apple juice waste), indicating a low influent COD carbohydrate fraction. The pellets were not as compact as with apple juice waste and were smaller (< 2 mm). The distillery waste COD ranged from 20 000 to 30 000 mg/ℓ. An appreciable amount of H2CO3*alkalinity was generated internally due to deamination of proteins and removal of organic salts. Provided the system was operated with a recycle from the effluent to influent at a recycle ratio sufficiently high to dilute the base influent COD to an effective influent COD (CODe) < 2000 mg/ℓ, sufficient of the H2CO3*alkalinity generated internally was recycled to maintain a minimum sludge bed pH > 6.6. Recycle ratios as high as 33:1, reducing the base influent COD of 27 000 mg/ℓ to a CODe of 790 mg/ℓ, did not adversely affect COD removal. No nitrogen, phosphate or trace element supplementation was required. COD removal was greater than 94 percent for COD loading rates up to the maximum of 15 kg/(m3 sludge bed.d); the maximum COD loading rate was fixed by gas lifting pellets in to the settling section, not by process failure.

This paper reviews developments in modelling the kinetics of activated sludge systems: Completely aerobic nitrification, anoxic/aerobic nitrification denitrification (ND), and anaerobic/anoxic/aerobic nitrification denitrification biological excess phosphorus removal (NDBEPR) systems. The paper highlights the progress in developing a general NDBEPR activated sludge kinetic model – development of polyP organism enhanced cultures, their kinetics, simplification of the kinetics for enhanced cultures under constant flow and load conditions, extension of the simplified model to mixed culture NDBEPR systems under constant flow and load conditions, integration of the polyP organism enhanced culture kinetics with the ND kinetics to give a general NDBEPR kinetic model for cyclic flow and load which incorporates the increased specific denitrification rates observed in NDBEPR systems compared to ND systems. Areas of research that require attention to complete the development of the general NDBEPR kinetic model are identified – denitrification by polyP organisms, calibration and verification of the model for cyclic flow and load, etc.

A survey of long sludge age activated sludge plants in South Africa showed that bulking was widespread in nitrogen and nitrogen and phosphorus removal plants with the dominant causative filamentous organisms being the low F/M types 0092, 0675, 0041, Microthrix parvicella, 0914 and 1851. The literature suggests that in aerobic systems proliferation of the low F/M filaments can be controlled by aerobic selector reactors. From an extensive investigation into the selector reactors, and the effect these stimulate, it was concluded that in long sludge age systems, irrespective of whether or not a selector effect is present in the sludge, (1) under completely aerobic conditions low F/M filaments do not proliferate, (2) sequential or alternating anoxic-aerobic periods like intermittent aeration are necessary, though not sufficient, conditions for the proliferation of low F/M filaments, and (3) an aerobic selector installed on an intermittent aeration system did not control low F/M filament proliferation.

This paper reviews developments in the kinetics of biological excess phosphorus removal (BEPR) in nitrification denitrification BEPR (NDBEPR) activated sludge systems – development of polyp organism enhanced cultures, their kinetics, simplification of the kinetics for enhanced cultures under constant flow and load conditions, extention of the simplified model to mixed culture NDBEPR systems under constant flow and load conditions. The next stage will be the development of a general NDBEPR kinetic model valid for cyclic flow and load. Before such a model can be established, an important problem needs to be resolved – the kinetics of denitrification in NDBEPR systems; recent studies show that the specific denitrification rate due to slowly biodegradable COD is significantly higher in NDBEPR than in nitrification denitrification (ND) systems.

Using plugflow anoxic reactors and batch tests, an experimental investigation was undertaken into the kinetics of denitrification in nitrification denitrification biological excess phosphorus removal (NDBEPR) systems. It was found that (1) in the primary and secondary anoxic reactors, the specific rate constant for denitrification associated with the utilization of slowly biodegradable COD (SBCOD) was respectively 2½ and 1½ times higher than in these reactors in nitrification denitrification (ND) systems and (2) in the primary anoxic reactor, the rapid rate of denitrification attributable to readily biodegradable COD was absent. The increased rate is hypothesized to be due to a stimulation in the active sludge mass of an increased rate of hydrolysis of SBCOD in these anoxic reactors of the NDBEPR system apparently induced by the presence of the anaerobic reactor in these systems.

Three biochemical models for biological excess phosphorus removal are critically analysed: the Comeau/Wentzel, Mino and modified Mino models. There is agreement between the models except in one respect, the generation of reducing equivalents (NADH2) required to convert acetate to poly-β-hydroxybutyrate under anaerobic conditions. In this regard a procedure is suggested to determine which of the models' premises are correct.

Laboratory research has indicated that a possible cause of low F/M filament bulking in ND (nitrification-denitrification) and NDBEPR (nitrification-denitrification biological excess phosphorus removal) systems occurs as a result of competition for substrate between filamentous and floc-forming organisms which have different denitrification pathways. In ND and NDBEPR systems alternating anoxic-aerobic conditions prevail and continuous utilization of particulate slowly biodegradable COD (SBCOD) by floc-forming organisms in these systems leads to accumulation of the denitrification intermediates nitrite (NO2−) and nitric oxide (NO) under anoxic conditions. It is proposed that a cause for low F/M filament bulking is that the intermediate NO inhibits the utilization of SBCOD by floc-formers under subsequent aerobic conditions, with high concentrations of NO2− exacerbating this effect, thereby allowing filamentous organisms, which do not accumulate NO, to dominate. Some experimental evidence to support this proposal is presented.

Optimal design of nitrification-denitrification (N-D) systems reduces to obtaining maximum denitrification subject to the constraint of efficient nitrification. Under constant flow and load conditions simple approximate highly accurate formulae are developed to describe the N-D system. Selection of the N-D configuration is found to be crucially influenced by the TKN/COD (influent) ratio, µnm, maximum and minimum temperature and sludge age. In terms of these factors design charts are presented which rapidly allow the best configuration and sizes of the anoxic and aerobic reactors to be selected.

Experimental observations indicate that the magnitude of biological excess P uptake is linked strongly to the magnitude of P release in the anaerobic reactor. A theory describing the kinetics of P release is presented in terms of the readily biodegradable COD in the influent, the non-polyP heterotrophic mass, the anaerobic mass fraction and the reactor flow regime. Observed P release conforms well to that predicted over a wide range of wastewater characteristics and process conditions.

The average influent wastewater characteristics - (i) the COD concentration, (ii) the TKN/COD concentration ratio, (iii) the rapidly biodegradable COD concentration, (iv) the maximum specific growth rate of the nitrifiers at 20°C attainable in the wastewater, (v) the maximum and minimum temperatures, and (vi) the P/COD concentration ratio - are shown to govern the design of, and effluent quality from single sludge activated sludge processes for both biological nitrogen and phosphorus removal. The TKN/COD ratio governs the selection of the process type: For the Phoredox process, complete denitrification is essential to obtain excess P removal, and this is shown to be feasible only for TKN/COD ratios less than 0,07 to 0,08 mgN/mgCOD; as the TKN/COD ratio increases above 0,08, complete denitrification becomes increasingly unlikely, and the UCT or Modified UCT processes are appropriate because in these processes complete denitrification is not essential to achieve excess P removal - in these processes N and P removal can be traded off against each other depending on the critical nutrient to be removed. Primary sedimentation significantly reduces the biological nutrient removal potential of activated sludge process because it increases the TKN/COD and P/COD ratios and reduces the COD load; however it significantly reduces the process volume and total oxygen demand.

The influent COD of municipal wastewaters has been categorised into fractions: readily (soluble) and slowly (particulate) biodegradable, and soluble and particulate unbiodegradable. Procedures are presented for determining the division into these four categories, as are experimental methods for determining the maximum specific growth rate of the heterotrophs within the IAWPRC Task Group model structure. Some of the procedures are dependant on knowledge of the heterotroph yield and endogenous mass loss rate constants; comment is made on suggested values.

The IAWPRC hypothesis, that particulate biodegradable COD hydrolyzes to a readily biodegradable (soluble) form, introduces an additional source of readily biodegradable COD for which the floc-formers and filaments compete. Acceptance of this hypothesis requires a re-evaluation of the selection criterion for the growth of filaments or floc-formers proposed by Chudoba et al. (1973). An alternative selection criterion, and some experimental observations supporting it, are presented.

The new generation of anaerobic digestion systems provide a high level of performance under optimal conditions (~ 37°C). However, in practice many influents will be at lower temperatures - heating of digesters presents problems, particularly in developing countries where technical backup is limited. This paper reports on a study of the treatment of a low/medium strength apple juicing waste in an upflow anaerobic sludge blanket reactor (UASB) at temperatures Less than optimal (25°C and 30°C). Maximum loading rates of approximately 12 and 16 kg COD m−3 d−1 were attained at 25°C and 30°C, respectively, for Influent concentrations in the range 2500 to 5000 rag COD ℓ−1. The comparative treatment capacity is In accord with the reported temperature sensitivity of mesophilic anaerobic processes. Formation of pelletised (granular) sludge enabled high upflow velocities and low hydraulic retention times.

Biological excess phosphorus removal in nitrification-denitrification single sludge activated sludge processes is shown to be stimulated by having a concentration of rapidly biodegradable COD (Sbsa) ≧25 mg/ℓ in the anaerobic reactor; the magnitude of the P removal is determined by a P removal propensity factor (Pf) defined by the product of (Sbsa−25) and the fractional mass of sludge in the anaerobic reactor. Sbsa is rapidly depleted by nitrate entering the anaerobic reactor; in the Phoredox process treating municipal waste flows if the TKN/COD ratio of the influent is greater than about 0,08 mgN/mgCOD the process, if designed to ensure efficient nitrification, is unlikely to remove all the nitrate and nitrate is recycled to the anaerobic reactor whereupon P removal declines. A new process is proposed that protects the anaerobic reactor from the nitrate in the effluent; tests indicate that this process can give excess P removal for TKN/COD ratios up to 0,14 mgN/mgCOD.

Fluctuations in hydraulic and organic loads cause wastewater treatment plant operating problems, necessitating some form of process control to ensure attainment of design objectives. A comparison of in–plant control with equalization control indicates that the latter is more appropriate for the case of long sludge age nutrient removal activated sludge processes which include anaerobic, anoxic and aerobic zones. A microprocessor-based control strategy is presented for the operation of an equalization tank upstream of the process to reduce, optimally, diurnal fluctuations in both flow and organic load rates. Results from implementation on a 150 Mℓ.d−1 plant with an in-line equalization tank (4,5 hour mean retention time) demonstrate the successful performance of the control strategy.

This paper reviews the antecedents to the general activated sludge model proposed by the IAWPRC Task Group on modelling of the activated sludge system. Modifications to the Group model are proposed and sets of experimental data from a wide range of single sludge systems are presented to validate the model.

This paper discusses the application of the general activated sludge model as set out by Dold et al. ( Prog. Wat. Technol. 12, 47–77, 1980) and extended by Van Haandel et al. ( Wat. Res. 15, 1135–1152, 1981), to anoxic-aerobic digestion of waste activated sludge. The laboratory scale experimental investigation comprised a 6 day sludge age activated sludge process, the waste sludge from which was fed to a number of digesters operated as follows: single reactor flow-through digesters at 4 or 10 days sludge age (retention times) under aerobic or anoxic-aerobic conditions (with 1.5 and 4 h cycle times) and 3-in-series flow-through aerobic digesters each with 4 days sludge age; all digesters were fed draw-and-fill wise once per day. The general kinetic model simulated accurately all the experimental data without the need to change the values of the kinetic constants. Both theoretical simulations and experimental data indicate that (i) the rate of volatile solids destruction is not affected by the incorporation of anoxic cycles and (ii) the specific denitrification rate constant in a digester is about two-thirds of that in the secondary anoxic reactor of the single sludge activated sludge system; this allows definition of a fourth denitrification rate constant K 4 for the anoxic-aerobic digester with K 4 T = 0.046(1.029) ( T -20) mg(NO 3 -N) (mgAVSS d) −1 , a constant independent of sludge age. An important consequence of (i) and (ii) above is that the denitrification can be integrated readily into the steady state digester model of Marais and Ekama ( Wat. SA 2, 163–200, 1976) and used for design purposes.

In nitrifying activated sludge systems several practical advantages accrue from inclusion of unaerated (anoxlc) zones. Accepting the general activated sludge model proposed by the IAWPRC Task Group, as modified by Dold and Marals (1985), simulation studies on nitrifying and nltrlfylng-denltrlfylng systems demonstrate the advantages of the latter with respect to (1) reduction in dally oxygen requirements, (2) reduction in effluent nitrate load, (3) elimination of pH control, (4) alleviation of the stringency in D.O. control, and (5) reduction in the required peak aeration capacity.

In aerobic digestion of waste activated sludge reduction in pH due to nitrification is commonly observed. pH reduction in turn affects nitrification efficiency, settleability and other properties of the digested sludge. Consequently for optimal performance it is necessary to monitor the pH and to control its value to near 7 by lime addition. An alternative control procedure, that eliminates both the need for pH monitoring and its control by lime addition, is to operate the process under alternating aerated/unaerated periods with 50 percent aerated, 50 percent unaerated. With this ratio all the nitrate produced during the aerobic period is denitrified during the unaerated (anoxic) period; the Alkalinity loss during nitrification is completely regained by denitrification and ammonification arising from protein degradation during both periods. A further advantage in anoxic-aerobic digestion is that by incorporating denitrification the oxygen requirement is approximately 20 percent less than that for purely aerobic digestion. Solids destruction is not affected.

Bulking in activated sludge systems due to proliferation of Sphaerotilus natans is very common in laboratory-scale but rare in full-scale systems. From two laboratory-scale studies it is concluded that a cause for proliferation was attached growth of S.natans on the walls of the feed lines and reactor surfaces continuously seeding the mixed liquor. It is suggested that S.natans bulking in laboratory-scale systems is common compared to full-scale systems because the surface area/volume ratio of the former is orders of magnitude higher than that of the latter so that the potential for seeding from attached growths in laboratory-scale plants is correspondingly higher. It would appear that in laboratory-scale activated sludge systems regular cleaning of the feed lines and daily scrubbing of the reactor and other wetted surfaces will eliminate S.natans bulking due to seeding of the mixed liquor from these surfaces.