The air flow rale required to keep the biomass in suspension is larger than that required to deliver the needed oxygen, and therefore it controls. The tank volume cannot be decreased to reduce this amount because then it would be impossible to deliver the needed oxygen to the first equivalent tank. This commonly occurs in CAS systems. The design can be considered to be acceptable.
Design of High-Purity Oxygen Activated Sludge Systems. Design of a HPOAS system is essentially the same as design of a CAS system, with two significant differences. First, the system is actually staged, so that the number of tanks-in-series is chosen by the designer. Thus, there is no need to iterate when distributing the oxygen requirement from tank to tank. Second, because high-purity oxygen is introduced into the first stage and the gas moves sequentially from stage to stage from the inlet to the outlet, the partial pressure of oxygen in the gas phase decreases as it moves through the system. As a result, the gas phase is most heavily enriched in oxygen in that part of the system where the required oxygen transfer rate is highest, allowing high oxygen transfer rates to be achieved with lower power inputs than would be required if the gas phase were air. This makes the power inputs more uniform throughout the system and alleviates some of the problems associated with balancing the power inputs into CAS systems. Of course, mixing must still be checked to ensure that sufficient energy is expended to maintain the MLSS in suspension.
Design of Selector Activated Sludge Systems. SAS systems are used when either the wastewater or the bioreactor configuration of the chosen activated sludge variation have a tendency to favor the growth of filamentous bacteria. A selector counteracts that tendency. As described in Section 10.1.2, a selector is a highly loaded section at the inlet end of an activated sludge bioreactor. The high loading condition creates an environment that favors the growth of floc-forming rather than filamentous bacteria, as discussed in Section 10.2.1. This results in improved solids settling characteristics, allowing the use of higher MLSS concentrations and/or higher secondary clarifier solids loading rates.
The selector must be properly sized if kinetic selection of floc-forming bacteria is to occur. First, it must be large enough to remove the majority of the readily biodegradable organic matter applied. If it is not, some of that organic matter will pass through the selector into the remaining portion of the bioreactor where envi ronmental conditions favor the growth of filamentous bacteria. Second, it must be small enough to maintain the readily biodegradable substrate concentration on the right side of the cross-over point on a plot of specific growth rate versus substrate concentration for the two bacterial types, as shown in Figure 10.12. If it isn't, environmental conditions favoring the growth of floc-forming bacteria will not be established in the selector, resulting in the preferential growth of filamentous bacteria in it.
Several factors must be considered when choosing the size and configuration of a selector. Because it is difficult to generalize about the kinetic parameters describing growth of floc-forming and filamentous bacteria, it is impossible to identify the readily biodegradable substrate concentration associated with the crossover point on the |x:Ss curve. However, empirical evidence has suggested the minimum initial process loading factor that will lead to good settling biomass, as shown in Figure 10.14 and discussed in Section 10.2.1. Since the process loading factor is proportional to the specific growth rate, specification of the process loading factor for the selector is analogous to specifying the crossover point on the |x:Ss curve. Consequently, selector design is typically based on specification of a process loading factor for it. However, because influent loadings vary, it is difficult to specify a single selector volume that is large enough to remove all of the readily biodegradable organic matter while also providing a sufficiently high process loading factor under all loading conditions. This problem can be solved by using a staged selector, as illustrated in Figure 10.6. During periods of low loading, the process loading factor will be sufficiently high in the first stage to favor the floc-forming bacteria, yet the bulk of the readily biodegradable substrate will be removed. During periods of high loading, on the other hand, the additional selector stages ensure that the readily biodegradable organic matter is removed prior to the main bioreactor.
Our understanding of the organism selection and organic substrate removal mechanisms occurring within SAS systems is evolving and, consequently, so are the approaches used to design selector systems. Based on current knowledge, however, the following general approach has been successful:"'
1. A minimum of three stages should be used. The first two stages should each contain 25% of the total selector volume, while the third stage should contain the remaining 50%.
2. The total selector volume is chosen to give an overall process loading factor of 3 kg total biodegradable COD/(kg MLSSday), which is analogous to about 1.75 kg BODs/(kg MLSS day). Note that even though the purpose of the selector is to remove readily biodegradable organic matter, the process loading factor is based on the total biodegradable COD. This is because the empirical evidence upon which it is based, i.e., Figure 10.14, did not distinguish between the readily and slowly biodegradable organic matter in the wastewater.
This relatively low overall process loading factor results in good removal of readily biodegradable organic matter, even during periods when the organic loading is temporarily higher. Furthermore, use of the selected overall process loading factor will result in a process loading factor in the initial selector stage of 12 kg COD/(kg MLSS-day). Reference to Figure 10.14 indicates that this high initial stage process loading factor provides a factor of safety to ensure that sufficiently high process loading factors are achieved regardless of the influent load.
Based on the above considerations, the selector volume is chosen from rearrangement of the definition of the process loading factor as given by Eq. 5.37:
Us " Avi.i where the influent COD includes both readily and slowly biodegradable substrate, and Vs and Us are the volume and process loading factor for the selector, respectively. The MLSS concentration is the same as in the rest of the activated sludge system, and is determined from (XM., • V)SvML„, as calculated with Eq. 9.11, where V is the selected total system volume. The volume of the rest of the system is just V — Vs.
Determination of the required oxygen transfer rate within a selector is not straightforward. Experience with some selector installations indicates that respiration rates in the first stage may be as high as 40 to 60 g 0:/(kg MLSS-hr).M v' These values are high, but are less than would be expected if the heterotrophic bacteria were growing at their maximum specific growth rate. It appears that, in some instances, the specific growth rates in the selector are sufficiently high to trigger a substrate storage response called the selector effect.In those cases, selector oxygen requirements corresponding to the oxidation of only about 20% of the COD removed have been observed."' It has been hypothesized that the remainder of the removed substrate is stored as intracellular carbon storage polymers such as glycogen and/or PHB. Because COD is conserved when storage polymer formation occurs, it is not necessary to supply sufficient oxygen in the selector to oxidize all of the removed substrate. Rather, the stored substrate is oxidized in the main bioreactor, and the oxygen requirement associated with it must be met there. Further research is needed to refine our understanding of the conditions under which this phenomenon occurs. In the meantime, it is prudent to design aerobic selector systems with significant flexibility and capacity in terms of the oxygen that can be transferred.
The determination of the oxygen requirement in the selector can be made by the same techniques used to spatially distribute the oxygen requirement in multitank systems, as discussed earlier in this section. From a conservative perspective, the oxygen requirement in the selector can be calculated by assuming that all of the readily biodegradable organic matter is removed and used for biomass synthesis in the selector. In other words, the selector effect is assumed to not occur. The oxygen requirement associated with the removal of slowly biodegradable organic matter can be distributed in proportion to the selector volume as a fraction of the system volume in which biomass synthesis occurs on slowly biodegradable substrate as determined with Eq. 10.27. The oxygen requirement for decay is given by Eq. 10.26 in which V, is the selector volume, Vs. Finally, nitrification will be occurring at its maximal rate in the selector, as given by Eq. 10.15. Because the selector is highly loaded, the volumetric oxygen transfer rate is likely to be quite high, particularly during periods of peak loading. As a consequence, special care should be given to the design of the oxygen transfer system in it to ensure that the needed transfer rate can be achieved.
The oxygen transfer system for the main bioreactor should be designed as if the selector were not present. There are two reasons for doing this. First, it will be necessary to bypass the selector periodically for maintenance and other purposes, placing the entire oxygen requirement into the main bioreactor. Second, the extent to which the selector effect will occur is usually unknown. However, the larger it is, the more the oxygen requirement is shifted to the main reactor. Thus, designing the main bioreactor to handle all of the oxygen requirement ensures that any situation can be handled.
Consider the CMAS system that was the subject of the examples in Section 10.3.3. In Example 10.3.3.6 the range of acceptable bioreactor volumes and associated MLSS concentrations was determined for the unequalized case, and was found to be small. It was also found that it would be preferable to design the system for the maximum feasible volume. Thus, assume that a volume of 7,300 ml was chosen, giving a MLSS concentration of 2,350 mg/L. Size a three-compartment selector for the system and determine the maximum potential oxygen requirement under average loading conditions.
a. What is the size of the selector?
As seen above, aerobic selectors are typically sized with an overall process loading factor of 3.0 kg total biodegradable COD/(kg MLSS - day). Adopting that value, and recognizing that the MLSS concentration in the selector will be the same as the MLSS concentration in the main bioreactor, i.e., 2.350 mg/L = 2.35 kg/m\ the selector volume can be calculated with Eq. 10.39. Recalling that the average flow rate is 40,000 mVday, and that the average readily and slowly biodegradable substrate concentrations are 115 mg COD L (0.115 kg/trT) and 150 mg COD/L (0.150 kg/m'), respectively, gives:
b. How is the selector configured?
It should be configured as three tanks-in-series, with the first two each being 25% of the total volume, and the third 50%. Thus, the first two selectors each have a volume of 375 m' and the third a volume of 750 m\
c. What is the volume of the main CMAS bioreactor?
The total system volume is not changed by the addition of the selector, so the main reactor volume is reduced by the volume of the selector. Therefore:
d. What is the oxygen requirement in the selector under average load conditions in the summer?
The oxygen requirement associated with biomass synthesis from readily biodegradable substrate was determined in Example 10.3.4.1 to be 1,840 kg 0:/day by using Eq. 10.23. All of it will occur in the selector.
The oxygen requirement for biomass synthesis from slowly biodegradable substrate was determined in Example 10.3.4.1 to be 2,400 kg 0,/day by using Eq. 10.24. Because the rest of the system is not compartmentalized, the utilization of slowly biodegradable substrate will occur uniformlv throughout it. Therefore, the utilization of slowly biodegradable substrate in the selector can just be assumed to be in proportion to its volume as a fraction of the total system volume. (If the remainder of the system was like a CAS system, rather than being a CMAS system, the oxygen requirement would have to be distributed in the same way as in Example 10.3.4.1.) Therefore.
The oxygen requirement associated with decay of heterotrophic biomass was determined in Example 10.3.4.1 to be 2.020 kg O /day by using Eq. 10.25. The amount of oxygen required in the selector lor heterotrophic decay will be in proportion to its volume:
The oxygen requirement associated with autotrophic biomass synthesis was determined in Example 10.3.4.1 to be 5,280 kg OVday by using Eq. 10.32. If the DO concentration in the selector is maintained at 1.5 mg I., then nitrification will occur at the maximum rate, as calculated in that example. Using the logic used in Example 10.3.4.1 to determine the traction of the nitrification oxygen requirement occurring in each tank of a CAS system, the fraction of the nitrification oxygen requirement occurring in the selector is (1,500/7,300) 0.49 = 0.42. Therefore,
The total oxygen requirement in the selector is the sum of all the above components:
RO, = 1,840 + 490 + 415 + 2,210 + 13 = 4,968 kg O day = 207 kg O/hr e. What is the required volumetric oxygen transfer rate?
The required volumetric oxygen transfer rate is determined bv dividing RO, by V„ giving a value of 138 g O/^m'hr). This can be achieved by giv ing special attention to the design of the oxygen transfer system.
Techniques for designing aerobic selectors will continue to evolve as our understanding of microbial competition and metabolic selection increase. Consequently, the approach presented here can be expected to be modified or replaced in the future.
10.3.5 Step-Feed Activated Sludge and Contact
Stabilization Activated Sludge—Systems with Nonuniform MLSS Concentrations
Examination of Figures 7.15 and 7.26 reveals that SFAS and CSAS have two characteristics in common: (1) the MLSS concentration is not uniform throughout the system, and (2) the concentration in the tank that discharges to the final settler is the
Finally, the oxygen requirement associated with decay of autotrophic biomass was determined in Example 10.3.4.1 to be 62 kg O /day by using F.q 10.34. The amount of oxygen required in the selector for autotrophic decay will be in proportion to its volume:
lowest of all of the tanks in the system. These are a direct result of the influent and recycle flow distributions as discussed in Sections 7.3.4 and 7.4.4. Furthermore, they provide the justification for the choice of these activated sludge variations for a particular installation. Guiding principle No. 4a in Table 9.1 states that all activated sludge variants with the same SRT contain about the same mass of biomass. As a consequence, when a SFAS or CSAS process is designed to have the same MLSS concentration entering the final settler as one of the activated sludge processes with uniform biomass concentration, the SFAS or CSAS process will always have a smaller total volume. Furthermore, the savings in system volume will be greater the longer the system SRT and the easier the organic substrate is to remove. Thus, SFAS and CSAS are often chosen for situations where space is limited. Another reason for designing a system so that it can be operated as SFAS was discussed in Section 7.3.4, and the same reason also applies to CSAS. If a CAS or CMAS system receives extremely high flows periodically, the high flow rate may cause the solids loading on the final settler to exceed allowable values, leading to loss of MLSS, poor effluent quality, and process failure. However, if the operating configuration can be switched to SFAS or CSAS, the MLSS concentration entering the final settler will be reduced, thereby keeping the solids loading on the clarifier within an acceptable range." Thus, another reason wastewater treatment plants are designed with the ability to operate in the SFAS or CSAS mode is for operational flexibility.
One consequence of the hydraulic characteristics of a CSAS system is that its performance is much more dependent on the recycle ratio (and on the RAS flow rate) than other activated sludge variations, as illustrated in Figures 7.24 and 7.25. This is because of the effect that the recycle ratio has on the MLSS concentration gradient through the system, with higher recycle ratios diminishing the gradient. For the same reason, the performance of a SFAS system is also somewhat sensitive to the recycle ratio, but much less so, as illustrated in Figures 7.17 and 7.18. Nevertheless, for both systems, more consideration must be given to the impacts of the RAS flow rate during system design, particularly when its impact on the final settler size is taken into consideration. As far as effluent quality is concerned, it is often acceptable to consider only the minimum anticipated recycle ratio because it is the critical one, causing the maximum MLSS gradient and minimizing the amount of biomass in the last bioreactor. If the system produces an acceptable effluent soluble substrate concentration at that recycle ratio, it will at all higher ones, as shown in Figures 7.17 and 7.24. On the other hand, the oxygen requirement in the contact tank of a CSAS system can increase significantly as the recycle ratio is increased, particularly when nitrification is occurring, as shown in Figure 7.25. Therefore, the distribution of oxygen requirements should be examined at both the upper and lower anticipated recycle ratios.
For an existing system with uniform MLSS concentration, the impact of switching to a SFAS or CSAS operational configuration can be calculated by using mass balance techniques based on a selected RAS flow rate, the anticipated distribution of influent and RAS flows to the various tanks, and the mass of MLSS in the system as calculated with Eq. 9.11. For the design of a new system in either SFAS or CSAS mode, the situation is more complicated because of the large number of choices involved, particularly for CSAS as illustrated in Section 7.4.4. Nevertheless, in all situations the overriding criterion is that the specific growth rate of the biomass in the tank discharging to the final settler must be low enough to allow effluent quality objectives to be met. Furthermore, all of the criteria discussed earlier about oxygen transfer and mixing energy input must also be met. This requires distribution of the oxygen requirement among the various tanks. Because it is the simpler of the two, we will first investigate SFAS. Then we will consider CSAS.
Design of Step-Feed Activated Sludge Systems. As shown in Figure 10.2 and illustrated schematically in Figure 7.10, SFAS systems are usually configured so that they behave as equally sized tanks-in-series with flow distributed equally to each tank. Other configurations can, and often are, used. For example, the equivalent tanks may not all be of equal size or influent may not be distributed to all, particularly the last. Furthermore, it some cases it may be advantageous to distribute the influent in unequal portions to the various tanks. Because it would be impossible to quantitatively describe all possible configurations, this presentation is limited to the case of equally sized tanks with equal distribution of flows to all of them. The concepts presented can easily be extended to other configurations by the reader should the need arise. The design of a SFAS system follows the same general approach as the design of the other alternatives presented earlier, with some additional steps. The emphasis here will be on the additional steps.
The first task in the design is to choose the bioreactor configuration, fixing the number of equivalent tanks in series, N, and the SRT, choices which are made by considering the factors presented earlier. Once those choices have been made, the mass of biomass in the system, (XM , ■ V)s>Ml.„„ can be calculated by using F.q. 9.11 for a CMAS system.
The next task is to ensure that the selected configuration and SRT will result in the desired effluent quality. For that to occur, the specific growth rate in the last tank, p.N, must be equal to or smaller than the specific growth rate as calculated by the Monod equation, Eq. 3.36. The approach used to calculate that specific growth rate is similar to the approach for determining the specific growth rate in the fictitious bioreactor used in the distribution of the oxygen requirement in CAS, HPOAS, and SAS systems, as discussed in Section 10.3.4. Thus, Eq. 10.28 may be used in which SSi is set equal to the desired effluent substrate concentration. The chosen effluent concentration can be either the readily biodegradable substrate concentration or the ammonia-N concentration, depending on the type of standard that must be met. A lower specific growth rate will be required to meet an ammonia-N standard than a soluble biodegradable COD standard. Nevertheless, for the development that follows, organic substrate removal will be assumed to be the objective, making the heterotrophic specific growth rate controlling. The logic is similar to that used to derive Eqs. 10.29 and 10.30, except that in this case the production of soluble substrate by hydrolysis of slowly biodegradable substrate will make a significant contribution to biomass growth in the last tank. If we assume that utilization of both readily and slowly biodegradable substrate is important, it can be shown that the specific growth rate in the last tank is given by:
in which VN is the volume of the last tank (which is the same as Vr/N since all tanks have the same volume) and XMTN is the MLSS concentration in the last tank. The derivation of this equation is dependent on the assumption that the mass flow rate of substrate into the last tank from the preceding tank is much less than the mass flow rate from the influent, which will generally be true. Since the value of |xM K as calculated with Eq. 10.40 must be less than or equal to the value of p,H N associated with the desired effluent substrate concentration as calculated with Eq. 10.28, then:
where p,M N has been calculated with Eq. 10.28. In other words, as long as the fraction of MLSS in the Nth tank, fxm.n, is greater than or equal to the right side of Eq. 10.41, the effluent quality will be acceptable.
The fraction of MLSS in any given tank of a SFAS system is determined totally by the hydraulics of the system. If both growth and wastage are neglected, then the fraction of MLSS in any tank i of an SFAS system in which the influent flow is split equally among N equal size tanks is given by:
(■^-Ml' VJs,Mtm ^ 1.0 + a where a is the biomass recycle ratio, which is the RAS flow rate, Fr, divided by the influent flow rate to the system, F. Furthermore, since the last tank is tank N, the fraction of MLSS in the last tank is:
Was this article helpful?