Info

'All components and coefficients are expressed as COD. "Coefficient must be multiplied by - i to express as oxygen.

consideration and limit the reactions to those associated with heterotrophs, as was done in Chapter 5. The influent contained sufficient ammonia-N for heterotrophic growth, but no nitrate-N, thereby eliminating nitrate-N, or reactions associated with it, from consideration as well. These selections simplified the model to include only processes 1, 4, and 7 and components 2, 3, 5, 7, and 8, as shown in Table 6.5. Finally, the dissolved oxygen concentration was held constant at 4.0 mg/L, which together with the Knn value in Table 6.3, made the oxygen term in the rate expressions approach 1.0.

6.2.1 Steady-State Performance

Figure 6.1 shows the effect of SRT on the mixed liquor suspended solids (MLSS) concentration, the active fraction, and the oxygen requirement for bioreactors receiving the two types of substrate at a constant flow and concentration. The MLSS in the bioreactor receiving soluble substrate is composed of heterotrophic biomass and biomass debris, whereas the MLSS in the bioreactor receiving particulate substrate contains heterotrophic biomass, biomass debris, and unreacted particulate substrate, with the relative quantities depending on the SRT. The curves for the soluble feed are similar to those in Chapter 5 and can serve as a reference point with which to see the effect of particulate feed. The most obvious effect is that a longer SRT is required to get biomass growth on particulate substrate. As indicated by the oxygen requirement, washout occurs at an SRT of about 4.5 hr when the substrate is soluble, but at an SRT of about 26 hr when it is particulate. This reflects the fact that hydrolysis reactions are slow. At SRTs below 26 hr, nothing happens to the particulate substrate and it acts like inert material, giving a MLSS concentration equal to the influent concentration times the SRT/HRT ratio (500 X 20 = 10,000) and an active fraction of zero, i.e., the MLSS is composed entirely of unreacted particulate substrate. As the SRT is increased past 26 hr, degradation of the particulate substrate begins, causing the MLSS concentration to drop and the active fraction and oxygen utilization to increase. In this region, the MLSS is composed of unreacted particulate substrate, active heterotrophic biomass, and some biomass debris. Eventually, at longer SRTs, bioreactor performance becomes independent of the feed type and is essentially the same for each. This occurs when both substrates are almost completely degraded so that system performance is governed primarily by biomass death and lysis, and the MLSS in each bioreactor is composed primarily of heterotrophic biomass and biomass debris.

Two significant points arise from Figure 6.1. The first is that long SRTs are required to achieve substantial degradation of particulate substrates. The second is that use of the process loading factor with a particulate substrate can be confusing. It will be recalled from Eq. 5.39 that the active fraction must be known before the process loading factor can be related to the specific growth rate of the biomass. Figure 6.1b, however, shows that the active fraction varies in a complex manner as the SRT is changed when the substrate is particulate. This is because the active fraction is the active biomass concentration divided by the MLSS concentration, which includes the undegraded particulate substrate. The SRT, on the other hand, is still related to the biomass specific growth rate by Eq. 5.12, and thus is more descriptive of process performance. Because most wastewaters contain some particulate i—("ci 111-1-

Soluble Feed

' Particulate Feed l

Soluble Feed

Soluble Feed

SRT, hrs

Figure 6.1 Effect of the nature of the organic matter in the feed on the performance of a CSTR with 0,/t = 20. Intluent biodegradable COD = 500 mg/L in each case. Flow = 1000 ml/day. Particulate substrate was assumed to be removed by the biomass separator and retained in the reactor. Parameter values are listed in Table 6.3. (1A was set equal to zero.

SRT, hrs

Figure 6.1 Effect of the nature of the organic matter in the feed on the performance of a CSTR with 0,/t = 20. Intluent biodegradable COD = 500 mg/L in each case. Flow = 1000 ml/day. Particulate substrate was assumed to be removed by the biomass separator and retained in the reactor. Parameter values are listed in Table 6.3. (1A was set equal to zero.

substrate, SRT is preferable to process loading factor as a basic design and operational parameter for suspended growth bioreactors.

6.2.2 Dynamic Performance

So far we have only considered the steady-state performance of a CSTR, i.e., the performance that results when a bioreactor receives a constant influent flow at a constant concentration. Most wastewaters are subject to time dependent variability, however, and thus it would be beneficial to investigate the impact of the nature of the substrate under those conditions.

Because of variations in human activities, municipal wastewater treatment systems experience diurnal variations in the flow and concentration of the wastewater entering them. Figure 6.2 shows typical variations experienced over a 24 hr period, beginning at midnight as time 0. The patterns correspond to those observed at a large municipal plant in South Africa over a period of one week," but the values have been normalized to a daily average flow of 100 m'/day and a flow-weighted average concentration of 100 mg/L. The patterns are also typical of those experienced in the United States," and will be adopted herein for demonstration purposes. As with kinetic and stoichiometric parameters, however, the necessity for determining the actual variations associated with a given wastewater cannot be overemphasized.

To determine the effect of the type of substrate on the dynamic response of a CSTR, the bioreactor in Figure 5.1 was subjected to the variations in flow and concentration shown in Figure 6.2. The flow values were adjusted to give a daily average flow of 1000 m'/day and the biodegradable COD was adjusted to give a flow-weighted daily average concentration of 265 mg/L, a value commonly seen in United States domestic wastewater."* The SRT was set at 240 hr, a value sufficient to make the steady-state performance for the two substrate types essentially the same, as shown in Figure 6.1, while the HRT based on the daily average flow rate was set at 6 hr. As with the steady-state response, two situations were considered, one in which the organic matter was entirely soluble and one in which it was entirely particulate. All other constituents were assumed to be constant at concentrations sufficient to not limit the reactions. As before, the matrix in Table 6.5 described the system.

The response of the bioreactor is shown in Figure 6.3. Before considering the effect of the type of substrate on the bioreactor performance, we will examine only the soluble substrate case in order to understand why the bioreactor behaves as it does. Examination of the figure shows that the soluble substrate concentration has

Figure 6.2 Typical diurnal patterns of wastewater flow and concentration for a community with little night time activity (after Dold and Marais'"). The flow has been normalized to an average of 100 mVday. The concentration has been normalized to give a flow weighted average of 100 mg/L.

Time, hrs

Figure 6.3 Effect of the nature of the organic matter in the feed on the response of a CSTR to the diurnal flow pattern in Figure 6.2. Flow weighted average influent biodegradable COD = 265 mg/L in each case. Daily average flow = 1000 mVday; SRT = 240 hrs; average HRT = 6 hrs. Parameter values are listed in Table 6.3. |i.N was set equal to zero.

Time, hrs

Figure 6.3 Effect of the nature of the organic matter in the feed on the response of a CSTR to the diurnal flow pattern in Figure 6.2. Flow weighted average influent biodegradable COD = 265 mg/L in each case. Daily average flow = 1000 mVday; SRT = 240 hrs; average HRT = 6 hrs. Parameter values are listed in Table 6.3. |i.N was set equal to zero.

greater relative variations throughout the day than does the MLSS concentration. This is a direct result of the fact that the residence time of the MLSS in the bioreactor (the SRT) is much longer than the residence time of the soluble substrate (the HRT). As a result, variations in the MLSS concentration are dampened. In fact, the mass of MLSS in the bioreactor is almost constant throughout the day.

If we consider the mass of MLSS to be approximately constant throughout the day, we can then see why the soluble substrate concentration varies. Examination of Figure 6.2 reveals that the mass of substrate entering the bioreactor per unit time (flow X concentration) varies throughout the day. This means that the mass of substrate available to a unit mass of microorganisms also varies throughout the day. However, the rate at which a microorganism can remove substrate is controlled by the concentration of substrate surrounding it. Thus, as the mass flow rate of substrate into the bioreactor increases, the substrate concentration must rise to allow the microorganisms to remove substrate faster. Conversely, as the mass flow rate of substrate decreases, the microorganisms will drive the substrate concentration lower until the rate of substrate removal is decreased to be consistent with the rate of input. Thus, the variation in substrate concentration is a direct consequence of the necessity for the microorganisms to vary their activity in response to the changing input rate of substrate. The variation in the oxygen requirement directly reflects that variation in activity.

The soluble substrate curves in Figure 6.3a also demonstrate an important point about the growth characteristics of the biomass in a CSTR receiving a time varying input; the specific growth rate is not constant over time. It will be recalled that the specific growth rate is controlled by the soluble substrate concentration, as expressed by the Monod equation (Eq. 3.36). Since the soluble substrate concentration is varying over time, so is the specific growth rate. This means that the bacteria are in a continually changing state. For the reactor configuration in Figure 5.1, the SRT is determined solely by the reactor volume and the wastage flow rate. Consequently, the SRT can be held constant, even though the specific growth rate of the bacteria is varying. In other words, the specific growth rate of the microorganisms in a CSTR that is not at steady-state is not fixed by the SRT. This can also be seen by performing a mass balance on heterotrophic biomass in the reactor. Such an exercise using the kinetics and stoichiometry in Table 6.5 reveals:

1 1 dXllH

Since the heterotrophic biomass concentration varies over time in response to the changing input, so will the specific growth rate. It is important to recognize that the constant relationship between SRT and specific growth rate depicted in Eq. 5.12 is only valid for steady-state conditions.

Even though the specific growth rate is not determined solely by the SRT for a non-steady-state CSTR, the SRT is still a good indicator of average performance, with longer SRTs giving lower average substrate concentrations. Nevertheless, the average substrate concentration leaving a CSTR receiving a dynamic input will always be greater than the concentration leaving a steady-state CSTR. For the conditions imposed in Figure 6.3, the flow-weighted average output concentration for the soluble substrate case in 2.11 mg/L as opposed to 1.80 mg/L for a steady-state CSTR with the same SRT. The higher average substrate concentration results from the nonlinear nature of the Monod equation (Eq. 3.36) describing the relationship between substrate concentration and microbial activity. This is one reason why it is advantageous to practice equalization prior to a biochemical reactor.

Now consider the impact of the type of substrate on the dynamic behavior of the CSTR. Figure 6.3a shows the effluent soluble substrate concentrations resulting from the two feed types. When the influent feed is all particulate, soluble substrate arises from hydrolysis of the particulate substrate, which is a slow reaction. Thus, the response is dampened and the flow-weighted average concentration is lower (1.92 mg/L vs. 2.11 mg/L). Figure 6.3b shows that there will be little difference in the MLSS concentration or its variation for the two substrate types. The slightly higher concentration in the bioreactor receiving particulate substrate is because of a slight build up of that substrate in the system caused by the slow hydrolysis reactions, but the effect is small. Generally, one would not expect to be able to distinguish much difference in the amount of MLSS in the two systems.

The major impact of particulate substrate on the dynamic response of a CSTR is in the utilization of oxygen, as shown in Figure 6.3c. As might be expected from the previous discussion of the slow nature of hydrolysis, the impact of the presence of particulate substrate is to dampen the system response, thereby reducing the peak oxygen requirement. In addition, the need for hydrolysis to make substrate available causes a time lag in the occurrence of the maximum and minimum oxygen consumption rates. Examination of Figure 6.2 reveals that the minimum mass input rate occurs at about 6 hr and the maximum at about 13 hr, which correspond closely to the times of the minimum and maximum oxygen requirements in the bioreactor receiving soluble substrate, thereby demonstrating the rapidity with which biomass can respond to soluble substrate. In contrast, the need for hydrolysis of particulate substrate, in combination with the fact that its concentration does not vary rapidly because its residence time in the system is the SRT, delays the minimum response by about 3 hr and the maximum by about 5 hr. Since both systems use about the same amount of oxygen in a 24 hr period, this suggests that consideration must be given to the physical state of the substrate during design of the oxygen transfer system for a suspended growth bioreactor.

0 0

Post a comment