## P 1000 Q

The necessity to indicate that the minimum feasible volume calculated with Eq. 10.4 comes from the floe shear criterion stems from the fact that there is a maximum volumetric rate at which oxygen can be transferred in activated sludge systems, and it also imposes a lower limit on the bioreactor volume. This maximum rate is device specific, and the manufacturer of the particular equipment of interest must be contacted to determine the appropriate maximum value for a given application. For example, for a floor coverage diffused air system, the limitation may be caused by the maximum number of diffusers that can be placed in the bioreactor per unit of floor area. Nevertheless, for the types of oxygen transfer systems typically used today, the maximum volumetric oxygen transfer rate that can be achieved economically on a sustainable basis is around 100 g 0./(rTT-hr), which is equivalent to 0.10 kg 0-/(m' • hr). During short-term transients this rate can sometimes be pushed to 150 g 0:/(rrv ■ hr), but typical mechanical aeration equipment should not be counted on to deliver oxygen at such a high rate on a sustainable basis because of excessive wear. Thus, if such high transfer rates are needed, specialized high-efficiency transfer systems must be used. As a result of this constraint, the lower limit on bioreactor volume based on oxygen transfer, V, m, should also be calculated:

where RO is expressed as kg 0:/hr and the V, <>, is in m'. The smallest allowable reactor volume is given by the larger of V, ,s and V, <,,.

Although floe shear has been correlated with the volumetric power input, a more fundamental parameter describing flocculation in general is the root mean square velocity gradient, G, and thus it is often used when examining flocculation in activated sludge systems. For diffused air systems, G (sec ') can be calculated as:

where Q is the airflow rate in m'/min, y is the liquid specific weight in N/m". h is the liquid depth above the diffuser in m. V is the bioreactor volume in m". p.„ is the absolute viscosity in N-sec/m:, and 60 is the conversion from minutes to seconds. Note the direct relationship between the volumetric air flow rate and G. For mechanical aerators, G can be calculated as:

where P is the aerator power input in kW ([W] = N • m/sec) and V and have the same units as above. Again, note the direct relationship between the volumetric power input and G.

For diffused air systems, effluent suspended solids concentrations have been correlated with G, with G values in excess of 125 sec 1 causing values to rise. 4 A G value of 125 sec 1 corresponds closely to a volumetric air input rate of around 20 m7(min- 1000 m') and a volumetric power input of around 14 kW/1000 m', which are the minimum considered necessary to keep MLSS in suspension, i.e., II,. At the other extreme, the G values associated with IIr are on the order of 270 sec ', and above that value, excessive floe destruction occurs. Between those extremes, there is a continual rise in effluent suspended solids concentration with increasing G, 4 and a designer can use the calculated G value to get an idea about likely effluent suspended solids concentrations. A prudent designer will anticipate having effluent suspended solids concentrations above the minimum attainable unless provisions are made for reflocculation prior to clarification.

For mechanically aerated facilities, another factor that must be considered is the location of the aerator relative to the discharge to the final clarifier."4 This is because mechanical aerators have very high localized velocity gradients. Consequently, in such systems the type and layout of the aerators has a stronger effect on effluent suspended solids concentrations than does the average G based on the overall volumetric power input.

Sheared floe can be reflocculated7". Thus, if the value of G exceeds 125 sec which will be true for most facilities, effluent quality can be improved by passing the activated sludge through a reflocculation zone prior to entering the final clarifier. A reflocculation time of 20 minutes at a G value of about 15 sec 1 may be appropriate. h,M The same thing can be accomplished in CAS systems by using low mixing energy in the latter stages where the oxygen requirement is low, but care must be exercised to keep all solids in suspension.

### 10.2.6 Nutrients

As discussed is Section 3.8.2, adequate nutrients are required to allow balanced growth of biomass in biochemical operations. Failure to provide them can have several consequences. For example, low nutrient concentrations can favor the growth of filamentous bacteria over floc-formers, as discussed in Section 10.2.1, resulting in a poor-settling activated sludge. More severe nutrient deficiencies can result in unbalanced growth of all bacteria, leading to the production of exocellular slime. In severe cases, the slime gives the activated sludge a jelly-like consistency, resulting in a sludge that settles slowly and compacts poorly."' s Virtually no liquid-solids separation will occur in such cases.

Procedures to calculate nutrient requirements are described in Sections 5.1.4 and 9.4.1, and Tables 3.3 and 9.2 provide guidance as to the quantities needed. Experience suggests that such calculations should consider only the inorganic nitrogen and phosphorus available in the influent wastewater. " Organic nitrogen and phosphorus will be released into solution and become available to the biomass as organic matter is biodegraded. However, the rate of biodégradation of some of these materials can be relatively slow, making the associated nutrients unavailable to heterotrophic bacteria metabolizing readily biodegradable organic matter. Thus, limiting nutrient concentrations can occur within the process, even though the total mass of nutrients may be adequate. Consistent maintenance of residual inorganic nitrogen and phosphorus concentrations throughout the process of approximately 1 mg/1 should be adequate.

### 10.2.7 Temperature

Temperature affects the performance of activated sludge systems as a result of its impact on the rates of biological reactions. Procedures for estimating the magnitudes of its effects are presented in Section 3.9. Two additional factors must be considered: the maximum acceptable operating temperature, and the factors that affect heat loss and gain by the process.

The maximum acceptable operating temperature for typical activated sludge systems is limited to about 35° to 40°C, which corresponds to the maximum temperature for the growth of mesophilic organisms. Even short-term temperature var iations above this range must be avoided since thermal inactivation of mesophilic bacteria occurs quickly. Successful operation can also be obtained if temperatures are reliably maintained above about 45° to 50°C, since a thermophilic population will develop, provided that thermophilic bacteria exist with the capability to degrade the wastewater constituents. Unacceptable performance will result for temperatures between about 40° and 45°C due to the limited number of microorganisms that can grow within this range. These considerations are particularly important for the treatment of high temperature industrial wastewaters.

One of the factors that affects heat gains in biological processes is the production of heat as a result of biological oxidation. As discussed in Section 2.4.1, the growth of bacteria requires that a portion of the electron donor be oxidized to provide the energy needed for biomass synthesis. Energy is also needed for cell maintenance. This oxidation and subsequent use of the energy results in the conversion of that energy into heat. Although this may seem surprising at first, it is directly analogous to the release of energy that occurs when material is burned; the only difference is the oxidation mechanism. The amount of heat released in the biooxidation of carbonaceous and nitrogenous material is directly related to the oxygen utilized by the process. For each gram of oxygen used, 3.5 kcal of energy are released.4"" Since 1 kcal is sufficient energy to raise the temperature of one liter of water 1°C, the impact of this heat release depends on the wastewater strength. For example, a typical domestic wastewater requires only one gram of oxygen for each 10 liters treated, therefore the temperature rise would be only 0.35°C, a negligible amount. On the other hand, it is not unusual for an industrial wastewater to require one gram of oxygen for each liter treated, in which case the temperature rise would be 3.5°C. This could be quite significant, particularly if the wastewater itself is warm.

Other heat gains and losses occur in biological systems. Heat inputs to the system include the heat of the influent wastewater, solar inputs, and mechanical inputs from the oxygen transfer and mixing equipment. Heat outputs include conduction and convection, evaporation, and atmospheric radiation. Models for accurately calculating heat balances across suspended growth bioreactors have been developed/'1""" They are discussed in Section 14.2.5.

If experience or a heat balance suggest the likelihood of unsatisfactorily high or unstable temperatures, the bioreactor should be configured to maximize heat losses. Measures to accomplish this include the use of relatively large basins to increase the HRT, shallow sidewater depths to increase basin surface area, above ground construction to maximize conductive and convective heat loss, and the selection of an oxygen transfer device, such as mechanical surface aeration, which maximizes heat loss. Another solution is to provide mechanical cooling of the process influent or the bioreactor contents. Designs such as HPOAS or facilities using deep bioreactors with diffused aeration will have minimal heat loss and should be avoided in this situation. In fact, they may require mechanical cooling even when large heat inputs are not expected.

Although heat gain is not generally a concern during treatment of municipal wastewaters, heat loss can be, depending on the type of oxygen transfer systems used and the bioreactor HRT." For example, submerged oxygen transfer systems, such as diffused aeration, have low heat losses, whereas mechanical surface aerators have high losses. This difference may affect the geographic region in which a par ticular type of oxygen transfer device can be used. When needed, heat loss can be minimized through proper facility design.

### 10.3 PROCESS DESIGN 10.3.1 Overview

The basic approach to the design of suspended growth biochemical operations is presented in Chapter 9. In this chapter, we focus that approach on activated sludge systems and examine the types of decisions that are required in their design. As discussed in Chapter 9, design is an iterative procedure and can take place at several levels of sophistication, depending on the information available to the designer. At the simplest level, in which little information is available, a preliminary design can be accomplished by applying the guiding principles articulated in Table 9.1. This approach is illustrated in Section 9.4.1, and the steps involved are summarized at the end of that section.

The next level, stoichiometric-based design, uses the simple model of Chapter 5, as extended in Section 9.4.2, and requires that specific information be available about the nature of the wastewater and the parameter values describing its biodégradation. That information sometimes can be obtained from historical records at a facility that is to be expanded, or from facilities treating similar wastewaters when a new system is being designed. In this case, it will usually be necessary to convert the information from traditional measurements, such as BOD5, into the more descriptive measurements, such as biodegradable COD, in use today. The procedures for doing this are presented in Sections 8.6 and 8.7. In other cases, treatability studies will be required to provide the necessary information. The procedures for conducting them are presented in Sections 8.2 and 8.3. Stoichiometric-based design provides quantitative information about the mass of MLSS to be contained in the activated sludge process, the steady-state oxygen requirement, and the mass of excess solids to be disposed of daily. The equations are for a single completely-mixed bioreactor, such as in CMAS, but as guiding principle No. 4 in Table 9.1 states, the calculated values are applicable to any of the activated sludge variants. Thus, they can be used as the basis for decisions about bioreactor configuration and the distribution of MLSS and oxygen supply within the bioreactors. These decisions require heuristic approaches, which are presented in the material that follows. It should be recognized, however, that all of the calculated values are based on the daily average flow and substrate concentration entering the facility, even though wastewater treatment facilities are subject to diurnally variable inputs, as illustrated in Figure 6.2. Thus, unless the activated sludge process is to be preceded by equalization, the impact of those variations on the design must be considered. This also requires the application of heuristically derived approaches.

As discussed in Section 9.4.3, the third level of design is simulation-based design. It is the most precise way to consider the impact of dynamic loadings on activated sludge systems and requires the use of a suitable dynamic model, such as activated sludge model (ASM) No. 1, and a computer code that implements it. Several such codes are listed in Table 6.4; all are simulation programs, not design programs. This means that the designer must choose a particular activated sludge process and provide the sizes of the component bioreactors as input to the programs. Sim ulations are then run to examine the performance of the process and the output is used to assess its acceptability. Should the design be unacceptable, the process must be modified and another simulation run. This procedure must be repeated in a logical manner until an acceptable design is arrived at, with the output providing needed information about the distribution of oxygen, MLSS, etc. Thus, the designer must have already accomplished a basic process design before beginning the simulations. This can be done with either of the first two approaches. However, characterization of the wastewater constituents and the parameters describing treatment is much more complex than that for the other design levels. The techniques for performing that characterization are described in Section 8.5.

The primary focus of this section is on stoichiometrically-based design. There are several reasons for doing this. First, preliminary design based on the guiding principles is discussed in sufficient detail in Section 9.4.1 to allow its application. It need not be expanded upon here. Second, as discussed above, simulation-based design requires the designer to provide a basic process flow diagram as input to the simulation program. The most effective way of doing this is by stoichiometrically-based design. Third, execution of a stoichiometrically-based design requires the designer to understand the most important aspects of activated sludge design. Thus, it provides an excellent framework within which to present them.

As discussed in Sections 8.6 and 8.7, many types of measurements have been used to express the concentrations of wastewater and activated sludge constituents. However, to allow us to focus on the decisions to be made during design and not distract the reader with multiple unit conversions, we use only biodegradable COD as the measure of organic substrate and TSS as the measure of MLSS. We have chosen the former because of its fundamental importance as a measure of available electrons, and the latter because of its widespread use in the profession. Sections 8.6 and 8.7 provide the information needed to convert between unit systems. In addition, to provide continuity in the examples, we use a standard wastewater throughout that is typical of domestic wastewater after primary treatment. It is the one in Table E8.4, and the translation between the conventional characterization in the top of the table and the more complete characterization in the bottom is explained in Example 8.6.1. Finally, we must emphasize that the calculations presented in this book are meant only to illustrate the procedures and decisions the process designer must make. The numerical results should not be considered to be typical of the application of biochemical operations to any particular real wastewater.

### 10.3.2 Factors to Be Considered During Design

Selection of the Appropriate Process Option. The selection of a particular activated sludge process is based on many considerations, including the wastewater characteristics, effluent quality goals, facility capital and operating costs, facility operational objectives, other processes at the facility, and the desires of the owner. Consequently, a full discussion of the selection of the process option is beyond the scope of this book. Nevertheless, a few generalizations are possible. CAS is popular for treatment of domestic wastewater because of its proven reliability and ability to achieve high effluent quality, including complete nitrification. However, for treatment of industrial wastewaters containing inhibitory organic compounds, CMAS has advantages, although special consideration must be given to the settling properties of the resulting biomass. If a wastewater contains a high percentage of readily biodegradable organic matter and no inhibitory materials, SAS may be required to control filamentous sludge bulking. On the other hand, if the wastewater contains a high percentage of colloidal organic matter that can be removed by entrapment in the biofloc, and the readily biodegradable substrate can be removed at an SRT shorter than that associated with good bioflocculation, then CSAS and SFAS have distinct advantages relative to system volume, although full nitrification may be difficult to achieve. For small communities wishing to minimize the number of operational personnel and types of unit operations on site, EAAS is popular. Finally, when little space is available and the emission of volatile organic compounds must be minimized, situations commonly associated with industrial facilities, HPOAS is often used. For additional information on the selection of the process option the reader should consult design manuals, such as MOP No. 8. "

Selection of the Solids Retention Time. The effects of SRT on activated sludge performance are discussed in Section 10.2.2, while the factors that must be considered during its selection are covered in Section 9.3.2. Consideration of the information in those sections makes it clear that selection of the SRT is a multifaceted decision requiring input from a number of sources. An important consideration of course, is the SRT needed to meet the required effluent quality. If a single completely mixed bioreactor is to be used, such as in CMAS or EAAS, the SRT required to produce a particular effluent COD is given by Eq. 9.7:

As discussed in Section 9.4.2, the parameters associated with autotrophs can be substituted for the heterotrophic parameters and SNM can be substituted for Ss to determine the SRT required to achieve a required ammonia-N concentration through nitrification. In either case, the SRT calculated from Eq. 9.7 would not necessarily produce the required effluent quality. There are several reasons for this. One is uncertainty in the kinetic parameters, influent characteristics, natural variability in the microbial community, and other factors. Such factors cause statistical variability in the effluent quality as illustrated in Figure 9.9.™ Thus, the calculated SRT must be multiplied by an appropriate safety factor, to account for that uncertainty, or the value of Ss (or SNH) must be chosen with the uncertainty in mind.

Because Eq. 9.7 represents only steady-state performance, another factor that must be considered is the impact of loading variations on process performance. Loading variations take two forms, day to day variations caused by seasonal and other events, and diurnal loading variations within a day, such as those illustrated in Figure 6.2. Discussion of the factors that go into decisions about design loadings and the use of equalization to dampen them is beyond the scope of this book. However, it is important to recognize that seasonal loading variations must be considered by the designer in selecting the SRT. With regard to typical diurnal loading variations, examination of the curves labeled continuous stirred tank reactor (CSTR) in Figure 7.4 reveals that COD removal is much less subject to their effects than is nitrification. Consequently, the safety factor for uncertainty is often sufficient to guard against unsatisfactory effluent organic substrate concentrations as a result of diurnal loading variations, but not against high effluent ammonia-N concentrations. Rather, an ad ditional safety factor is required to account for the effects of ammonia-N loading variations on nitrification. It is the peak load safety factor, g,,,:

The choice of the period over which (F- SN|,()),.tak is defined, i.e., peak diurnal within the average day, maximum month, maximum week, etc., depends on the nature of the discharge standard that must be met, the plant process flow diagram, etc., and as such, is also beyond the scope of this book. Consequently, the reader should consult other sources, such as the U. S. EPA Nitrogen Control Manual1 for more information.

Finally, nitrification is very sensitive to DO concentration, as discussed in Section 6.3.1 and illustrated in Figure 6.7. That sensitivity can be expressed by a double Monod expression, such as Eq. 3.46, which is used in ASM No. 1 (Table 6.1). Since DO was not included in Eq. 9.7, it can be included by defining another safety factor, ;,„„ that is the reciprocal of the DO term in Eq. 3.46:

A similar safety factor does not need to be applied to organic substrate removal because the half-saturation coefficient for DO for heterotrophic bacteria is sufficiently small to make Eq. 10.9 approach a value of 1.0 at typical bioreactor DO levels.

Considering all of these safety factors, the required value of the SRT, (M)„ r, can be determined from the computed SRT, (M),, as:

For organic substrate removal, gIM and 5,,,, are normally set at 1.0, as discussed above, simplifying the expression. However, the full expression is typically used for nitrification. Application of Eq. 10.10 results in a very conservative design for nitrification because the safety factors are multiplicative. Consequently, application of simulation-based design techniques, which can explicitly account for dynamic conditions, often allows reductions in the SRT required to achieve adequate nitrification performance from a CMAS system.

Application of Eq. 10.10 to nitrification in a CAS system, which behaves like a plug-flow or tanks-in-series system, would result in selection of a longer SRT than is actually necessary. This can be seen in Figure 10.17 where the effluent ammonia-N concentration is plotted as a function of SRT for activated sludge systems with various numbers of CSTRs in series. The curves in the figure were developed by simulation with ASM No. 1, using the parameter values in Table 6.3. The DO concentration was set at 4.0 mg/L so that it would not be a factor. To illustrate the impact of bioreactor configuration, consider a situation in which the desired ammonia-N concentration is 1.0 mg/L as N. Examination of the figure reveals that the calculated SRT for a CMAS system (N = 1) would be 4.4 days, whereas the SRT required for a CAS system (N = 9) would be only 2.36 days. Thus, if the SRT calculated with Eq. 9.7 (which is for a CMAS system) were substituted into Eq. 10.10 to determine the required SRT for a CAS system, it is clear that the CAS system would be badly overdesigned. Furthermore, the overdesign becomes worse

SRT, days