August 20, 2005
Optimizing Denitrification in Anoxic Zones
What first comes to mind when you hear the word denitrification? For many operators, the answer may be a somewhat negative "floating sludge on the surface of the secondary clarifiers." This is understandable, as floating or denitrified sludge can reduce effluent quality by increasing secondary-effluent suspended solids as it breaks up.
Despite this negative connotation, however, denitrification is being incorporated into many wastewater treatment plants (WWTPs) because it helps the biological treatment process, improves operational efficiency, and benefits the environment. Specifically, denitrification does this by* removing nitrogen from wastewater and returning it to the atmosphere, thus completing the nitrogen cycle;
* helping control filamentous organisms and improving sludge settling;
* using nitrate to oxidize carbonaceous biochemical oxygen demand (CBOD), thus reucing oxygen requirements and energy use; and
* recovering alkalinity (approximately half of the alkalinity lost during nitrification is recovered during denitrification).
Because denitrification is being incorporated into many secondary treatment process designs, it is important for operators to know how to implement effective denitrification at their facilities.
Necessary Conditions for Denitrification
Denitrification is the biological reduction of nitrate (NO^sub 3^^sup -^) to nitrogen gas (N^sub 2^). Whereas nitrification - the oxidation of ammonia (NH^sub 3^) to nitrate - converts nitrogen from one form to another, denitrification removes nitrogen from the wastewater.
For effective denitrification process control, one must first understand the four environmental conditions required for denitrification to occur:
1. Denitrifying organisms must be present. This is not a limiting factor for denitrification, because such organisms are common in wastewater.
2. Nitrate or nitrite must be present. Nitrates are seldom found in raw wastewater or primary effluent due to the denitrification that occurs in the collection system and primary clarifiers. For most WWTPs, the nitrates produced during nitrification serve as the source for denitrification processes.
3. There must be little or no dissolved oxygen (DO) present. Most denitrifying organisms are facultative and prefer oxygen over nitrate. If too much DO is present, denitrification will not happen.
Table 1. Denitrification Simulation
Aerator influent and effluent samples were collected from a biological nutrient removal facility using a process mode shown in the following schematic. The process mode achieved complete nitrification (NH^sub 4^^sup +^-N less than 1 mg/L), and partial denitrification (NO^sub 3^-N less than 5 mg/L).
The samples were transferred to settleometers and the environmental conditions of each sample adjusted as shown below to promote denitrification. Denitrification is indicated when the settled sludge floats to the surface of the settleometer and is time- dependent based on where the sample was collected.
4. Organic carbon must be present. Denitrifying organisms are heterotrophic. Thus, they use organic carbon (CBOD substances) as an energy source. Some WWTPs use a separate-stage denitrification process that follows nitrification. In these plants, a carbon source, usually methanol, is added.
Tables 1 and 2 (below and p. 52) provide a graphic simulation of denitrification using activated sludge in settleometers. Table 1 illustrates how the above-listed environmental conditions are varied, while Table 2 lists the order in which the sludge is denitrified in each of the settleometers. This simulation has been performed many times at many treatment plants using site-specific activated sludge samples. With few exceptions, only the order of settleometers 3 and 4 has varied from the order shown in Table 2.
With the increased emphasis on nutrient removal in general and nitrogen removal in particular, many WWTPs have incorporated anoxic zones into the biological treatment process to promote denitrification.
Anoxic zones usually are placed at the influent end of the activated sludge tank to take advantage of the organic carbon and low DO associated with primary effluent. Since nitrates typically are not present in primary effluent, a mixed liquor recycle pipe brings nitrified mixed liquor - hence nitrates - to the anoxic zone. With all the environmental conditions necessary for denitrification present, denitrifying organisms oxidize organic carbon and reduce nitrate to nitrogen gas. Mechanical mixing promotes contact and releases the nitrogen gas into the atmosphere.
Minimizing DO in the anoxic zone. Since denitrifying organisms are facultative and prefer oxygen over nitrate, one element of optimization should focus on minimizing DO in the anoxic zone. Some common ways that oxygen enters the anoxic zone are shown in Figure 1 (p. 53) and discussed below.
* Primary effluent (Figure 1a). Oxygen can enter primary effluent in several ways. For example, aeration occurring over primary clarifier weirs, through channel-aeration mixers, and through flow- splitting structures can entrain oxygen.
Table 2. Denitrification Simulation Results
* Cascading aerator influent (Figure 1b). When primary effluent or return activated sludge (RAS) freefalls into the anoxic zone, each naturally entrains oxygen from the air. The amount of oxygen entrained depends on the wastewater temperature and the freefall height.
* Mixed liquor recycle. The mixed liquor recycle pipe is designed to return nitrates to the anoxic zone. However, the place where the recycle is withdrawn - the end of the aeration basin - also tends to have the highest DO concentration. The aerobic zone should be optimized to have enough DO for complete nitrification (including diurnal variations) but not be overaerated. Excessive DO in the mixed liquor recycle will reduce the efficiency of the anoxic zone.
* Anoxic zone mixer speed (Figure 1c). Mechanical mixers in the anoxic zone can increase oxygen transfer from the atmosphere to the wastewater. The mixer speed should be sufficient to keep solids in suspension, rapidly distribute primary effluent, and prevent "clearwater" trails across the anoxic zone but also should prevent surface vortices that entrain oxygen.
* Backmixing from the aerobic zone (Figure 1d). Air from the activated sludge tank's aerobic zone can backmix into the anoxic zone if there is not enough baffling between the two zones. In addition to providing adequate baffling, backmixing can be overcome by inducing headloss between the two zones.
* Surface discharges into the anoxie zone (Figure 1e). RAS and mixed liquor recycle discharge pipes should be introduced below the surface of the anoxic zone. Discharges onto the surface entrain oxygen and reduce efficiency. (Note: The pipes shown in Figure 1e once discharged above the anoxic zone surface, but they have since been extended below the surface.)
Optimizing nitrate reduction. Unless an industrial user discharges a significant amount of nitrate into the collection system, it is uncommon to see nitrate in the raw wastewater or primary effluent. Consequently, in a biological treatment system, anoxic zone performance depends on the nitrates produced during nitrification in the aerobic zone. These nitrates can then be returned to the anoxic zone through the RAS or mixed liquor recycle. Figure 2 (p. 54) shows the theoretical nitrogen removal rate that can be achieved through recycle. As can be seen, RAS alone (assuming a rate of approximately 50%) can achieve between 30% and 40% nitrogen removal. In a typical design, mixed liquor recycle can achieve 80% removal based on a 4:1 recycle ratio.
Figure 1. Sources of Dissolved Oxygen in Anoxic Zones
Figure 2. Theoretical Maximum Nitrogen Removal Based on Recycle Ratio
Some WWTPs can achieve even higher nitrogen removal through simultaneous nitrification-denitrification. If the DO concentration in the aerobic zone is not high enough to maintain aerobic conditions in the center of the biological floe, denitrification can occur in the center while nitrification happens elsewhere in the floe. This is most common in aeration systems with poor oxygen transfer, such as surface aerators, and in higher-temperature wastewater.
Size of the anoxic zone. To account for variable wastewater characteristics, many WWTPs can vary the size of their system's anoxic zone. Assuming the zone is truly anoxic, its optimum size is a function of temperature, concentration, nitrates, and readily biodegradable organic matter. For example, higher wastewater temperatures and the presence of short-chain volatile fatty acids (a type of readily biodegradable organic matter) permit the use of smaller anoxic zones.
Performance, however, is determined by the anoxic zone's effluent nitrate concentration. Thus, the zone should be operated to reduce its nitrate level completely, and the size of the zone used to accomplish this should be only as large as necessary. If the zone is too small, nitrates will pass into the effluent, and the full benefits of denitrification will not be realized. If the zone is too large, some portion of it will be anaerobic (no nitrates or oxygen) and therefore may suboptimize oxidation potential.
When deciding how large the anoxic zone should be, diurnal peaks also must be considered. Diurna\l peak flows may require a larger anoxic zone to account for reduced detention time, but this need might be offset by higher concentrations of organic material. High flows due to infiltration and inflow may bring DO into the anoxic zone; if this happens, a larger anoxic zone may be needed to deplete the oxygen so that denitrification can occur.
Putting It All Together
Ultimately, anoxic zone performance can be optimized by applying the knowledge of the environmental conditions required for denitrification. These include minimizing DO in the anoxic zone, optimizing nitrate return within the treatment plant's capabilities, and adjusting the size of the anoxic zone based on nitrate reduction. Optimal anoxic zone performance improves operational performance and minimizes the need for capital improvements as flows and loads change.
Woodie Muirhead is a vice president and operations specialist in the Honolulu office of Brown and Caldwell (Walnut Creek, Calif.). The author gratefully acknowledges the support of Rob Baur of Clean Water Services (Hillsboro, Ore.) for his help in setting up and performing the denitrification simulation.
Copyright Water Environment Federation Aug 2005