Multiple Events

The purpose for development of the model in Table 6.1 was to allow engineers to simulate biochemical reactors in which all of the listed processes are occurring. Thus, it would be instructive to use it to investigate such a situation in a single CSTR. It is apparent, by now, however, that the conditions required for anoxic growth of heterotrophs and aerobic growth of autotrophs are mutually exclusive, since both are controlled by the dissolved oxygen concentration, but in the opposite manner. Consequently, if a CSTR is receiving a constant input and is operating at steady-state with a constant dissolved oxygen concentration, it is impossible for significant amounts of both nitrification and denitrification to occur. However, observations of treatment systems receiving diurnal variations in flow and concentration suggested that when the input was low, resulting in a high dissolved oxygen concentration, nitrification occurred, but when the input was high, driving the dissolved oxygen to very low concentrations, nitrification ceased and denitrification began, destroying part of the nitrate-N formed during the aerobic period. This made the daily average effluent total nitrogen concentration lower than in a system receiving adequate oxygen, suggesting that it might be possible to purposefully design a system in which the dissolved oxygen concentration varied sufficiently to allow both reactions to occur. In addition to reducing the amount of nitrogen discharged, this would reduce the amount of alkalinity destroyed and the amount of oxygen required, as discussed previously. Let us now consider each situation for a typical domestic wastewater.

6.5.1 Effects of Diurnal Variations in Loading

Figure 6.13 shows the simulated performance of a CSTR containing a biomass with the kinetic and stoichiometric coefficients in Table 6.3 and receiving an input with the variations shown in Figure 6.2. The average daily flow rate was 1000 m'/day and the reactor volume was 500 m\ giving an average HRT of 12 hr. The flow-weighted average concentrations of the various influent components were as shown in Table 6.6. The influent alkalinity, however, was assumed to have a constant concentration because it is influenced primarily by the characteristics of the carriage water. Two situations were considered. In one, the dissolved oxygen concentration was held constant at 2.0 mg/L throughout the entire 24 hr period. Given the value of K().„ in Table 6.3, this effectively eliminated denitrification. In the other, the mass transfer rate for oxygen was sufficient to maintain a dissolved oxygen concentration of 2.0 mg/L if the bioreactor received the daily average flow and concentrations. As shown in Figure 6.13a, however, because of the variable input, this resulted in excessive oxygen concentrations during periods of low loading, but inadequate concentrations during high loading. Under those conditions, significant denitrification occurred, as shown in Figure 6.13c. Comparison of the solid and dashed curves in parts b and d of the figure shows that while the lack of oxygen had only a minor

8 12 16 20 Time, hrs

Figure 6.13 Response of a CSTR to a diurnal input. The input patterns are like those in Figure 6.2, but the flow weighted average concentrations are those given in Table 6.6. The daily average flow rate was 1000 mVday. The solid curves are for a reactor in which the dissolved oxygen concentration was held constant at 2.0 mg/L. The dashed curves are for a reactor in which the mass transfer coefficient for oxygen was held constant at 3.83 hr~'. The average HRT = 12 hrs and the SRT = 240 hrs. The parameter values are given in Table 6.3.

8 12 16 20 Time, hrs

Figure 6.13 Response of a CSTR to a diurnal input. The input patterns are like those in Figure 6.2, but the flow weighted average concentrations are those given in Table 6.6. The daily average flow rate was 1000 mVday. The solid curves are for a reactor in which the dissolved oxygen concentration was held constant at 2.0 mg/L. The dashed curves are for a reactor in which the mass transfer coefficient for oxygen was held constant at 3.83 hr~'. The average HRT = 12 hrs and the SRT = 240 hrs. The parameter values are given in Table 6.3.

effect on the concentration of soluble organic matter in the effluent, it significantly lowered the nitrate-N concentration and raised the ammonia-N concentration. The raising of the ammonia-N concentration was due to retardation of nitrification, but the decrease in the nitrate-N concentration was caused by a combination of less production from nitrification and more consumption by denitrification. Nevertheless, it is apparent that less nitrogen was discharged from the system that had limited oxygen transfer capacity. Furthermore, less alkalinity was destroyed.

While it is not desirable to design a supposedly aerobic system with inadequate oxygen transfer capacity, the results from the simulation show clearly that all events can occur in a single biomass provided that the SRT of the system is sufficiently long to allow the nitrifying bacteria to grow during the aerobic period. This suggests that it should be possible to design a system in a way that maximizes nitrogen removal by controlling the periods with and without oxygenation. Such a system would have minimal power input as well as minimal alkalinity destruction.

6.5.2 Intermittent Aeration

Batchelor was among the first to use simulation to investigate the possibility of using intermittent aeration to achieve both nitrification and denitrification in a single CSTR receiving a constant influent. His study was conducted with a model that was conceptually similar to the one in Table 6.1, although it differed somewhat with respect to both the kinetics and stoichiometry employed. In addition, the values of the kinetic and stoichiometric parameters were slightly different from those in Table 6.3. Nevertheless, the results of his simulations illustrate some important concepts regarding the major variables influencing the performance of such systems.

The situation considered by Batchelors was of a wastewater containing 200 mg/L of biodegradable COD and 30 mg/L of ammonia-N, entering a CSTR with an SRT of 240 hr and an HRT of 4.3 hr. The bioreactor was operated with a cycle time (the time between initiations of aeration) of 0.5 hr while the aeration fraction (the fraction of time that the bioreactor was aerobic) was varied. In each case, the effluent nitrogen concentrations from the bioreactor achieved a stable oscillation and the average concentrations of ammonia-N, nitrate-N, and total nitrogen were calculated. Figure 6.14 shows clearly that at low aeration fractions, the average ammonia-N concentration rises whereas at high aeration fractions the average nitrate-N concentration rises. Furthermore, an optimum exists at which the discharge of nitrogen is minimized. This result is consistent with our previous discussions. At aeration fractions in excess of the optimum, nitrification is complete, but the anoxic period is insufficient to allow much reduction of nitrate-N. Conversely, at aeration fractions less than the optimum, the aerobic period becomes insufficient for growth of the nitrifying bacteria. If the aeration fraction is made so low that the product of the aeration fraction and the SRT is below the minimum SRT for the nitrifiers. they will wash out and no nitrogen removal will be achieved, other than that associated with incorporation of nitrogen into the heterotrophic biomass via synthesis. For the kinetic parameters used by Batchelor/ the minimum SRT for nitrification was 52.6 hr. Since the system SRT was 240 hr, this suggests that total process failure would occur at an aeration fraction of 0.22. The fraction of the SRT that is aerobic is called the aerobic SRT and it is now recognized as an important parameter in the design of systems in which both nitrification and denitrification are occurring.

Figure 6.14 Effect of aeration fraction on the concentrations of ammonia-N, nitrate-N. and total-N in a CSTR operated with intermittent aeration at a cycle time of 0.5 hr. The HRT = 4.3 hrs and the SRT = 240 hrs. The influent biodegradable COD = 200 mg/L and the influent nitrogen concentration = 30 mg/L. (Adapted from Batchelor ")

Figure 6.14 Effect of aeration fraction on the concentrations of ammonia-N, nitrate-N. and total-N in a CSTR operated with intermittent aeration at a cycle time of 0.5 hr. The HRT = 4.3 hrs and the SRT = 240 hrs. The influent biodegradable COD = 200 mg/L and the influent nitrogen concentration = 30 mg/L. (Adapted from Batchelor ")

The optimal aeration fraction in Figure 6.14 is 47%, but since the minimum allowable aeration fraction depends on the system SRT and the degree of denitrifi-cation also depends on the system SRT, we might expect the optimum aeration fraction to depend on the SRT as well. That this is the case is shown by the dark circles in Figure 6.15. The vertical bars in the figure indicate the range of aeration fractions at each SRT which results in total nitrogen concentrations within 1.0 mg/L of the minimum. That range is seen to broaden as the SRT is increased, suggesting that longer SRTs provide more latitude in operation of the bioreactor to achieve optimal nitrogen removal.

Both Figure 6.14 and Figure 6.15 were prepared from simulations conducted with a cycle time of 0.5 hr. However, we might also expect the cycle time to influence the degree of nitrogen removal. During aerobic periods, the ammonia concentration will fall and the nitrate-N concentration will rise as nitrification occurs. Conversely, during anoxic periods, denitrification will cause the nitrate-N concentration to fall while the continued influx of ammonia in the absence of nitrification will cause the ammonia concentration to rise. The longer the cycle time, the greater the amplitude of the cycles because a longer time will be available for each reaction. Since the effluent concentration is characterized by the daily average concentration, longer cycle times result in higher average nitrogen concentrations. On the other hand, as the cycle time is shortened, a point will be reached at which it is difficult for the biomass to shift rapidly enough from aerobic to anoxic metabolism, and vice versa. Consequently, there is also an optimum cycle time associated with each SRT and aeration fraction. At this time our understanding of the metabolic controls acting on the synthesis and activity of denitrifying enzymes is not sufficient to allow the optimum cycle time to be selected by simulation. Nevertheless, experimental studies have shown that cycle times on the order of 20 to 45 min work well.2"

0 0

Post a comment