Chapter - 22 - Biological - Nutrient - Removal - Revised Edition
Chapter - 22 - Biological - Nutrient - Removal - Revised Edition
Chapter - 22 - Biological - Nutrient - Removal - Revised Edition
Biological Nutrient
Removal Processes
22-1
Copyright © 2007 Water Environment Federation.
22-2 Operation of Municipal Wastewater Treatment Plants
reactors with zones, cells, or baffles designed to promote the growth of the bacteria that
remove nutrients and the production of a good settling sludge.
Wastewater characteristics also affect both the ability to remove nutrients and the
rate of removal (Table 22.1; WEF, 2005). For example, if the influent’s inert (non-
biodegradable), solids change, operators may need to lengthen the solids retention
time (SRT) or mean cell residence time (MCRT) to maintain nitrification. If the waste-
water has a high biochemical oxygen demand (BOD) but the oxygen use rate is low,
then denitrifying or phosphorus-removing organisms may not be able to compete with
other organisms in metabolizing organic material.
Both BNR and BPR depend on sufficient amounts of readily biodegradable or-
ganic substrate [chemical oxygen demand (COD) or BOD]. Wastewater treatment pro-
fessionals have determined that 8.6 mg of COD is needed to remove 1 mg of nitrate-
nitrogen from wastewater, but only 0.7 to 1.9 mg of a more readily biodegradable
Growth Zones. Most wastewater treatment professionals are familiar with aerobic or
oxic activated sludge processes [Figure 22.2(a)], in which biological growth is managed
by controlling the oxygen concentration and recycling flows, such as return activated
sludge (RAS) and mixed-liquor recycle (MLR), to the reactor [Figure 22.2(b)]. The
wastewater’s oxygen concentration is kept near or above 2.0 mg/L, because nitrification
declines when dissolved oxygen concentrations drop below 0.5 mg/L. Also, the oxida-
tion–reduction potential (ORP) is kept near or above 100 mV.
To create an anoxic zone, a baffle or partial wall is installed in the reactor, and
the aerators in that area are shut off. Little dissolved oxygen is present (less than
0.5 mg/L) in this zone, but chemically bound oxygen (in nitrite and nitrate) may be
present in RAS or MLR flow. Also, ORPs should be between 100 and 200 mV (for
rapid denitrification).
Anaerobic zones contain neither dissolved oxygen nor chemically bound oxygen
and have ORPs below 300 mV [Figure 22.2(c)]. They typically are created by sending
MLR to denitrification selector cells rather than to the head of the anaerobic zone,
which would increase chemically bound oxygen levels too much. Sometimes a supple-
mental source of carbon is necessary to ensure that dissolved and chemically bound
oxygen are rapidly removed.
Selector Size and Equipment. Updating a treatment plant for filament control,
ammonia removal, denitrification, or BPR involves several changes (Figure 22.3). For
example, converting from conventional wastewater treatment to BNR may involve
changing gates, baffles, mixers, and recycle pumps; improving aeration equipment
and foam control; and adding clarification, RAS pumping, and instrumentation and
controls. Also, many BNR facilities operate at long MCRTs, so more aeration basin
volume may be required to maintain the necessary residence time to sustain nitrify-
ing bacteria.
The size and number of selector zones needed for BNR depends on wastewater
characteristics, treatment goals, and other factors. Typically, 20 to 30% of the total
basin volume will be dedicated to anoxic or anaerobic zones. Plants with stringent
nutrient limits may dedicate as much as 50% of the total basin volume to anoxic or
anaerobic zones. Bench or pilot testing may be needed to confirm selector sizing and
design criteria. [For details on sizing selectors, see the “Integrated Biological
Processes for Nutrient Control” chapter in Design of Municipal Wastewater Treatment
Plants (WEF, 1998b) or the 4th edition of Wastewater Engineering (Metcalf and Eddy,
Inc., 2003).]
The wastewater conveyance methods also may need to be changed. Wastewater
is often conveyed via pumping or air-entraining devices (e.g., screw pumps, flow-
splitting weirs, or launders), which inject air into the water. Entrained oxygen is detri-
mental to anoxic or anaerobic conditions, so air-entraining devices should be replaced
with submerged weirs, gates, launders, or centrifugal pumps.
The water surface of a BNR reactor and a conventional activated sludge aeration
basin look similar, but below the surface, the BNR reactor typically has a series of walls
or partitions (called baffles) that define discrete treatment areas. Baffles are typically
made of pressure-treated wood, fiberglass, reinforced concrete, or concrete block. They
often include blockouts for piping, tank drainage, main plant flow-through, and main-
tenance access. The tops of some baffles are below the water surface, while others in-
clude overflow weirs that direct water from one zone to the next. Because they are
elevated, overflow weirs can force foam into downstream aerated cells, rather than
allowing back mixing to occur. During peak flows, however, this elevated water sur-
face prevents the backflow of dissolved oxygen, which would allow readily biodegrad-
able BOD to be consumed. This will interfere with denitrification or phosphorus re-
moval by allowing the ORP to increase.
Numerous devices (e.g., diffused air, pontoon-mounted impellers, submersible
impellers, and bridge-mounted turbine impellers) have been used to mix the contents
of both anoxic and anaerobic selector zones. However, many wastewater treatment
professionals think that diffused air introduces too much entrained oxygen for the
zone to be truly anaerobic or anoxic. Some prefer the simplicity of pontoon-mounted
impellers because the electrical motors are above the water surface, but most use sub-
mersible impellers. Submersible impellers are typically mounted on a vertical rod;
they can be raised and lowered, and the horizontal angle can be adjusted for optimum
mixing.
Denitrification is an integral part of most BNR systems. It typically involves recy-
cling nitrified mixed liquor via low-head pumps to one of the anoxic cells. Low-head
pumps (0.3 to 0.9 m of head) are used because the MLR typically is moved a short dis-
tance, needs little or no valving, and enters a wastewater surface only a few inches
higher than the one it left. Mixed-liquor recycle pumps may include vertical turbine
centrifugal, propeller pumps, or axial flow pumps.
Upgrading to a BNR process often requires a change in aeration equipment.
However, more blowers may be unnecessary if the diffusers are very efficient or if the
upgrade includes denitrification, which can recover about 63% of the molecular oxy-
gen required to convert ammonia-nitrogen to nitrate-nitrogen. In a multiple-stage
BNR process, the first oxic stages typically contain more fine-bubble diffusers because
initial oxygen uptake rates are higher than in latter stages of treatment. Diffusers also
are installed in the last anoxic zone, which typically is a swing cell. Swing cells are
zones that have both mixers and diffusers so they can operate as either non-aerated or
aerated reactors.
Biological nutrient removal processes may need a long MCRT to allow nitrification
and denitrification to occur. However, long MCRTs typically lead to foam or filamen-
tous bacteria problems. To control foam, treatment plants can spray the reactor surface
with a chlorine wash or add a cationic polymer or foam suppressant. They also can
Mean Cell Residence Time. Mean cell residence time (also called sludge age or SRT) is
the most commonly used parameter when operating a conventional activated sludge
system. It measures the average length of time (in days) that microorganisms (sludge)
are held in the system, and is calculated as follows:
If the MCRT is too short, then the biological system will not have enough bacteria
to degrade the pollutants, resulting in poor effluent quality (Figure 22.4). The mean cell
retention time required depends on the wastewater constituent and the growth rate of
the microorganisms consuming it. For example, a simple carbon compound (e.g., ac-
etate) is metabolized by fast-growing heterotrophic organisms; it requires an MCRT of
less than 1 day to be synthesized. Conversely, ammonia is oxidized by slow-growing
bacteria, so its MCRT is much longer. Many treatment plants that remove CBOD from
wastewater may have been designed for 2 to 4 days of MCRT, but to remove ammonia,
the MCRT may need to be twice as long, depending on temperature and other factors.
If the basin size and MLSS are constant, then the MCRT will shorten as the sludge
wasting rate increases and lengthen as the wasting rate decreases.
Calculating MCRT for BNR systems is different because the aeration basin may in-
clude anaerobic or anoxic selectors. For example, a sequencing batch reactor (SBR) may
not be aerated 35 to 45% of the time. So, process control should include calculating both
the “aerobic” MCRT (for the aerated portion of the reactor) and the total “system” MCRT.
The anaerobic or anoxic selector zones will require separate MCRT calculations. In BPR
systems, experts recommend a 1-day anaerobic MCRT for proper growth of Acineto-
bacter organisms (Metcalf and Eddy, Inc., 2003). If the anaerobic contact is too long, the
Acinetobacter organisms will release their stored phosphorus and take up carbon (this is
called secondary phosphorus release).
If the treatment plant has primary clarifiers, then F is typically based on primary
effluent. If not, then F is based on the raw wastewater’s BOD load. The M for a conven-
tional activated sludge system is typically based on the entire aeration basin’s MLVSS.
Most conventional activated sludge systems are designed for F⬊M ratios ranging from
0.2 to 0.4 (WEF, 1998a). In BNR systems with several cells (Figure 22.5), the selector
may be configured to encourage the growth of nonfilamentous (floc-forming) organ-
isms. Or the selector may be configured to encourage the growth of Acinetobacter or-
ganisms to increase the denitrification rate. In either case, the F⬊M ratio of each selector
zone is important. When calculating the F⬊M ratio for a BNR system, F is based on the
secondary influent BOD. If the reactors are operated in series, then the first cell’s M is
based on its volume and MLVSS concentration. The second cell’s M is based on the
combined volume of the first and second cells, and so on.
One recommendation (Albertson, 1987) is that BNR systems have at least three
cells to create a high substrate concentration in the initial minutes of contact (Figure
22.5). Food-to-microorganism ratios of 6.0, 3.0, and 1.5 mg/d of BOD per 1.0 mg of
MLVSS for the first, second, and third cells, respectively, have been recommended for
a plug-flow reactor. However, if the F⬊M ratio is too high (more than 8.0 mg/d of BOD
per 1.0 mg of MLVSS), then a viscous, nonfilamentous organism could dominate the
reactor (this is called slime bulking).
One of the key characteristics is the amount of inert TSS in biological treatment in-
fluent (iTSS 2INF). A high concentration of inert TSS will increase the percentage of non-
biodegradable solids in the MLSS, so a longer MCRT will be needed to treat the waste-
water sufficiently. To calculate iTSS2INF,
Where
Operators also should better understand BOD’s constituents. The soluble BOD
measurement, for example, will indicate how much material is readily biodegradable.
Soluble BOD typically is measured based on a 5-day BOD test in which wastewater is
passed through a 0.45-m filter. A 1-day BOD test will provide an estimate of readily
biodegradable BOD—including the simple carbon compounds that are available for
rapid bioassimilation.
Operators can learn other valuable information by assessing the influent BOD and
waste activated sludge (WAS). In a conventional activated sludge system, for example,
5 mg of nitrogen and 1 mg of phosphorus will typically be used for every 100 mg of
BOD removed. In a BPR process, 3 to 5 mg of phosphorus may be used for every 100 mg
of BOD removed. If the BPR process is operated to remove total phosphorus, then a
secondary-influent-BOD-to-total-phosphorus ratio of 20⬊1 or more may be needed to
ensure that the treated effluent will contain less then 1.0 mg/L of phosphorus.
The required BOD-to-total-phosphorus ratio depends on process type and effluent
goals. Processes that do not nitrify and denitrify [e.g., the anaerobic/oxic (A/O) process]
may require a BOD-to-total-phosphorus ratio of 15⬊1, while those with full nitrification
and denitrification may require a BOD-to-total-phosphorus ratio of 25⬊1 or more.
Aeration Requirements. Aeration systems designed for CBOD removal may need to
be enlarged by 30 to 50% to provide enough oxygen to nitrify ammonia. Some plants
accommodate the increased demand by replacing air diffusers with fine-bubble sys-
tems or other more efficient aerators. Some incorporate denitrification (the reduction of
nitrate-nitrogen to nitrogen gas) into the BNR process because denitrification can theo-
retically reclaim 63% of the oxygen needed for nitrification. However, because denitri-
fication typically only reduces about half of the nitrate-nitrogen in the wastewater
(denitrification effluent typically contains 6 to 8 mg/L of nitrate-nitrogen), the actual
amount of oxygen reclaimed may be closer to 30 to 40%.
Sludge Settleability and Foam. Many BNR facilities operate at a high MCRTs (more
than 8 days) to fully nitrify and allow for denitrification. As a result, the facilities fre-
quently have sludge-settleability or foam problems. Wastewater constituents (e.g., soap,
oil, and grease) and streams recycled from solids handling processes can exacerbate
these problems. If so, the facility may need spray nozzles to spread antifoaming
chemicals.
Before RAS flows are raised to increase the basin MLSS concentration, plant staff
should evaluate the solids loading (g/m2d) on the secondary clarifiers to determine if
the capacity is adequate. Larger RAS pumps may be needed to obtain the higher MLSS
concentration often desired in a BNR system.
Return Flows. Conventional activated sludge systems typically have only one return
flow: RAS. Biological nutrient removal processes, on the other hand, may have more
return flows (e.g., MLR). The return activated sludge pumping rates are typically 30
to 100% of the influent flow (Q). The MLR pumping rates may be 100 to 400% of the
influent flow, depending on such factors as the target effluent nitrate concentration.
They also may be transferred from an aerobic or oxic zone (MLROX) or from an anoxic
zone (MLRAX). The source of the MLR—and which zone receives it—often distin-
guishes one BNR process from another.
Alkalinity and pH. Biological nutrient removal systems also may need pH control or
added chemicals (e.g., hydrated lime, soda ash, or caustic soda) to supplement the
available alkalinity. Because alkalinity is consumed during nitrification, the chemicals
can help maintain the minimum alkalinity level needed (typically, 60 to 100 mg/L of
alkalinity as calcium carbonate). Low alkalinity not only lowers pH but may limit the
growth of nitrifying organisms because they lack enough inorganic carbon to produce
new cells.
Hydraulic Retention Time. The hydraulic retention time (HRT) required to maintain
BNR depends on the size of the reactor, which in turn depends on the MCRT necessary
for growth. Once the reactor size is known, then the HRT can be found by dividing the
volume of the reactor (V) by the secondary influent flow (Q2INF):
The result is a nominal value used to relate basin size to plant flow. The nominal
value is typically used because internal recycle streams do not affect most of the reac-
tions that are important to plant operators. For some parameters, however, the actual
HRT may be affected by recycle streams, in which case, recycle flows (i.e., RAS and
MLR) should be added to the denominator in Equation 22.5.
Typical effluent from the primary treatment process contains 160 mg/L of BOD. Dur-
ing secondary treatment, the equivalent of about 80 mg/L of biological VSS is produced
(0.5 mg VSS/mg BOD). Ammonia is consumed during the production of VSS, as are
8 mg/L of nitrogen [(80 mg/L)(10/100)] and 1.6 mg/L of phosphorus [(80 mg/L)(2/100)].
The resulting biological sludge contains approximately 10% nitrogen and 2% phosphorus.
If the wastewater characteristics are not “typical”, then treatment plant staff will
have to adjust the target MCRT, F⬊M ratio, or other criteria.
The difference between conventional activated sludge and BNR systems is that a
conventional activated sludge system only removes nitrogen via sludge wasting. A bi-
ological nutrient removal facility can remove nitrogen via sludge wasting and a combined
biological reaction to oxidize the remaining ammonia (nitrification) and reduce oxi-
dized nitrogen (denitrification). If the treatment plant only has an ammonia limit, it
may only nitrify the wastewater. If the plant has both ammonia and oxidized nitrate
limits, then it will need to both nitrify and denitrify the wastewater.
Nitrosomonas
2NH
4-N 3O2 2NO
2 2H2O 4H
(22.7)
2NO
2 O2
Nitrobacter
2 NO3-N (22.8)
NH
4-N 2O2
Nitrifiers
NO
3-N 2H H2O (22.9)
Influences. Autotrophic bacteria typically grow two to three times more slowly than
heterotrophic bacteria, which are the predominant organisms in a biological treatment
system. Understanding the relationship between the bacteria’s “growth rate” and
MCRT (sludge age) can help treatment plant staff determine whether and how tem-
perature, dissolved oxygen, and other factors affect biological treatment. The biomass
growth rate () is calculated as follows:
The maximum growth rate (max) for nitrifying bacteria at 20 °C is typically between
0.14 and 0.23 kg/d (0.3 and 0.5 lb/d). To put this in terms more familiar to plant staff,
for a treatment system in equilibrium:
In other words, is the inverse of MCRT, which is used for process control:
MLSS X
MCRT = (22.12)
TSS 2EFF + TSS WAS
1
MCRTMIN = (22.13)
MAX
needed for nitrification (Figure 22.9). For example, the calculated MCRT (using max)
may be 10 days at 10 °C, but a design engineer would probably make the aeration basin
large enough to operate at a 15-day MCRT (factor of safety 1.5) or a 20-day MCRT
(factor of safety 2.0). So, operators may find that the process nitrifies reliably at a
shorter MCRT. (The actual “best” MCRT is plant-specific because of the number of
variables involved.)
Nitrification is also affected by pH. The optimum pH is typically about 7.5. As the
pH drops, so do the nitrifiers’ growth rate and activity. Nitrification may be inhibited
when the pH is less than 6.5, but some information sources have indicated that nitri-
fiers can acclimate to low pH.
Some metal, organic, and inorganic compounds can inhibit the growth of au-
totrophic bacteria. If plant personnel suspect that toxics are inhibiting bacteria growth,
they should conduct a bench-scale test assessing the nitrification rate. Such tests can be
conducted onsite, or samples may be sent to a contract laboratory for testing.
atmosphere naturally consists of more than 70% nitrogen, so the emissions do not harm
the environment.
NO
3-N carbon source facultative bacteria N2 CO2 H2O OH
The biochemical reactions associated with denitrification are key factors in operat-
ing wastewater treatment plants (Table 22.1). For example, reducing nitrate-nitrogen
eliminates some of the BOD demand, so the anoxic reactors help remove CBOD. If
CBOD is the only treatment consideration, then MCRT and F⬊M ratio calculations
could include the biomass in the denitrification process.
Theoretically, 2.86 mg of oxygen is recovered for every 1.0 mg of nitrate-nitrogen
reduced to nitrogen gas. This is more than 60% of the oxygen needed for nitrification,
so this recovered oxygen could be used to greatly reduce the amount of aeration equip-
ment needed in other areas of the treatment plant.
Denitrification also results in new bacterial cells. The cell yield depends on the car-
bon source. For example, if methanol is the carbon source, then the cell yield is about
0.5 mg VSS per 1.0 mg of nitrate-nitrogen removed. If BOD is the source, then the cell
yield is about 1.5 mg VSS per 1.0 mg of nitrate-nitrogen removed.
In addition, about 3.57 mg of alkalinity (as calcium carbonate) is produced for each
1.0 mg of nitrate-nitrogen removed. So, about 50% of the alkalinity lost during nitrifi-
cation can be recovered during denitrification.
Influences. Denitrifiers are less sensitive than nitrifiers, so if the treatment plant envi-
ronment does not inhibit the nitrifiers, then the denitrifiers should have no problem
functioning at optimal growth rates.
The rate at which denitrifiers remove nitrate is the specific denitrification rate
(SDNR):
This rate varies, primarily depending on the type of carbon source and amount of
carbon available (as measured by the F⬊M ratio).
Nitrification and dentirification can occur in one treatment unit [this is called
simultaneous nitrification/denitrification (SNDN)], or denitrification may occur separately
in either post- or pre-anoxic reactors. In the 1970s, BNR and denitrification typically
were done via post-anoxic reactors [Figure 22.10(a)] (Stensel, 2001). Methanol was
the carbon source for denitrification (methanol’s SDNR ranges from 0.1 to 0.3 mg
NO3-N/dmg VSS). A relatively small post-aeration reactor followed the denitrifica-
tion process to oxidize any remaining organics.
In the 1980s, many wastewater treatment plants were upgraded with selectors for
controlling filamentous organisms [Figure 22.10(b)]. Those that nitrified their waste-
water found that the anoxic zone was denitrifying RAS as well as controlling filaments.
Because methanol is expensive and a hazardous material, most plants took advantage
of this lesson. They added an MLR system and used a pre-anoxic selector for denitrifi-
cation [Figure 22.10(c)]. In this selector, either raw wastewater or primary effluent is
used as the carbon source for denitrification (wastewater’s SDNR typically ranges from
0.03 to 0.12 mg NO3-N/dmg VSS).
Overall, the denitrification rate drops as wastewater passes through various anoxic
cells because less readily biodegradable BOD is available in downstream anoxic zones than
Where
Mr MLR/Q and
Rr RAS/Q.
PNOx
Mr = − 1.0 − Rr (22.19)
NO X2EFF
So, given the treatment plant illustrated in Figure 22.11, an Rr of 0.5, and 25 mg/L
of produced nitrate, an Mr of 1.0 (MLR Q) will result in an effluent nitrate concen-
tration of about 11 mg/L. If the effluent nitrate limit is 5 mg/L, then Mr will need to be
3.5 (i.e., 350% of the treatment plant’s influent flow) to achieve this limit. When effluent
nitrate limits are less then 5 mg/L, a post-anoxic selector may be necessary to achieve
the limit without having a Mr that is excessively large.
limitations of nitrogen removal processes are listed in Table 22.4 (Metcalf and Eddy,
Inc., 2003). Descriptions of common nitrogen removal processes follow.
Ludzack–Ettinger Process. In the 1960s and 1970s, engineers made a number of mod-
ifications to the conventional activated sludge process to improve nitrogen removal.
For example, researchers Ludzack and Ettinger developed a version in which RAS and
secondary influent are combined in an anoxic zone that is followed by an aerobic zone
[Figure 22.12(a)]. The nitrate formed in the aerobic zone is returned to the anoxic zone
via RAS for denitrification.
Because the anoxic zone’s only source of nitrate is the RAS, denitrification is lim-
ited by the amount of RAS flow. If the influent NOX is 20 mg/L, the effluent nitrate
concentration will be 15 mg/L, 13 mg/L, or 10 mg/L depending on whether the RAS
recycle flow ratio is 0.3, 0.5, or 1.0, respectively. So, the process is only suitable for den-
itrification if nitrate limits are liberal or high RAS flows can be maintained.
HRT (hours)
recycles mixed liquor from the oxic zone to the anoxic zone [Figure 22.12(b)]. The
mixed-liquor flow may equal 100 to 300% of the secondary influent flow, depending on
how much denitrification is desired. Effluent nitrate concentrations typically range
from 4 to 7 mg/L.
The mixed-liquor pumps may be relatively small because short distances and
large transfer pipes typically result in pumping heads of less than 0.9 m (3 ft). They also
may resemble a fan rather than the conventional centrifugal pumps used at wastewater
treatment facilities.
This design [Figure 22.12(b)] is the cornerstone of many other BNR processes, in-
cluding those used for BPR, so the MLE process will be referred to in many subsequent
process descriptions. Although Figure 22.12 shows the anoxic (or anaerobic) selector as
one reactor, it probably consists of two or more cells, zones, or compartments operated
in series. If nitrified effluent is recycled for denitrification, the MLR is typically trans-
ferred to the second or third cell in the anoxic selector to minimize dissolved oxygen in
the first compartment, thereby better controlling filamentous bacteria and providing
optimal conditions for Acinetobacter organisms.
Advantages Limitations
General Saves energy; BOD is removed before
aerobic zone
Alkalinity is produced before nitrification
Design includes an SVI selector
MLE Very adaptable to existing activated- Nitrogen-removal capability is a
sludge processes function of internal recycle
5 to 8 mg/L TN is achievable Potential Nocardia growth
problem.
Dissolved oxygen control is
required before recycle
Step Feed Adaptable to existing step-feed activated Nitrogen-removal capability is a
sludge processes function of flow distribution
With internal recycle in last pass, nitrogen More complex operation than
concentrations less than 5 mg/L are mLE; requires flow split control
possible to optimize operation
5 to 8 mg/L TN is achievable Potential Nocardia growth
problem
Requires dissolved oxygen
control in each aeration zone
Sequencing Batch Reactor Process is flexible and easy to operate Redundant units are required for
operational reliability unless
aeration system can be
maintained without draining the
aeration tank
Mixed-liquor solids cannot be washed
out by hydraulic surges because flow
equalization is provided
Quiescent settling provides low effluent
TSS concentration
More complex process design
Effluent quality depends upon
reliable decanting facility
5 to 8 mg/L TN is achievable May need effluent equalization of
batch discharge before filtration
and disinfection
Batch Decant 5 to 8 mg/L TN is achievable Less flexible to operate than SBR
Mixed-liquor solids cannot be washed out Effluent quality depends on
by hydraulic surges reliable decanting facility
Advantages Limitations
Bio-denitro™ 5 to 8 mg/L TN is achievable Complex system to operate
Large reactor volume is resistant to Two oxidation ditch reactors are
shock loads required; increases construction
cost
Nitrox™ Large reactor volume is resistant to Nitrogen-removal capability is
shock loads limited by higher influent TKN
concentrations
Easy and economical to upgrade existing Process is susceptible to ammonia
oxidation ditch processes bleed-through
Provides SVI control Performance is affected by
influent variations
Bardenpho Capable of achieving effluent nitrogen Large reactor volumes required
(4-stage) levels less than 3 mg/L
Second anoxic tank has low
efficiency
Oxidation Ditch Large reactor volume is resistant to load Nitrogen-removal capability is
variations without affecting effluent related to skills of operating staff
quality significantly and control methods
Has good capacity for nitrogen removal;
less than 10 mg/L effluent TN is possible
Post-Anoxic with Carbon Capable of achieving effluent nitrogen Higher operating cost due to
Addition levels less than 3 mg/L purchase of methanol
May be combined with effluent filtration Methanol feed control required
Simultaneous Nitrification/ Low effluent nitrogen level possible Large reactor volume; skilled
Denitrification (3 mg/L lower limit) operation is required
Significant energy savings possible Process control system required
Process may be incorporated into existing
facilities without new construction
SVI control enhanced
Produces alkalinity
figured with denitrification (rather than phosphorus removal) as the primary goal. Both
processes can achieve effluent nitrate concentrations of less than 3 mg/L.
Air
Ax1 Ox1
2 CL
Aeration Tank
RAS
mLE
MLR
Air
Ax1 Ox
1
P 2 CL
Aeration Tank
RAS
Carbon
MLR
Air Air
Ox
1
Ax Ox1 2 CL
P 2
Ax 2
Aeration Tank
mLE
RAS
Ox Ox Ox 2 CL
Ax1 1 Ax2 2 Ax3 3
RAS
same size, a four-stage system should have an influent flow split of about 15⬊35⬊30⬊20%
so the F⬊M ratio will be the same in each step. Each stage should have its own influent
controls. The flow into the last step is critical because the nitrate produced there will
not be reduced. So, the anoxic step-feed process is best used when the effluent nitrate
limit is more than 6 to 8 mg/L.
Oxidation Ditches. Oxidation ditches are typically sized based on long aerobic MCRTs
(20 to 25 days), so excess volume is available for denitrification (Figure 22.14). This is
typically achieved by turning off one or more aeration rotors to create an anoxic zone
[Figure 22.14(a) and (b)]. Because the SDNRs are low, the zone must be large enough to
provide the needed anoxic time.
Another approach is to cycle the aeration by turning the aerators off at least twice
a day [Figure 22.14(c)]. Submerged mixers maintain recirculation in the ditch during
the anoxic phases. A variation of cyclic aeration, called the Nitrox™ process, involves
using ORP for control.
A third approach is phased ditch operation [Figure 22.14(d)]. In this process, two ox-
idation ditches are operated in series, and secondary influent is alternately pumped into
them. When the influent enters the ditch, its aeration equipment is turned off, and the re-
actor becomes anoxic. Submerged mixers maintain recirculation. After a period of time
(typically 1 to 2 hours), the influent is sent to the second ditch. At the time of the switch,
the aerators are turned on in the first ditch, making it oxic, and the aerators on the second
ditch are turned off. This cycle continues, allowing periods of aeration and anoxic fill to
occur. A variation of phased ditch operation, called the Bio-denitro™ process, uses four
phases to enhance oxidation and denitrification (Stensel and Coleman, 2000).
Dissolved Oxygen
Mixer
Oxic
Anoxic
Reactor 1
Ax
Ax
Cycle 1
Anoxic
Oxic
Ax
Oxic
Ax1
An MLR
Ax2 Ox2
Ox3 P
Ox1
Ox4
Anoxic MLR
Ax Zone
Oxic Zone Secondary
Clarifier
Ox1
P
Ox2
Anaerobic An
Zone
Secondary Influent RAS
One benefit of MBRs is that their effluent requires no further biological treatment
or filtration to meet water reuse standards.
Lagoons. Lagoons are probably one of the oldest nutrient removal technologies. How-
ever, their large size and loss of nitrification during cold weather have typically limited
their use.
A number of facultative or partially aerated lagoons have been upgraded to acti-
vated lagoons by converting the lined earthen basins to aeration basins and adding
secondary clarifiers and an RAS system. Conventional secondary clarifiers and RAS
pumps can be used, but several manufacturers supply in-basin clarifiers with travel-
ing-bridge sludge collectors and airlift RAS pumping.
Activated lagoons operate at 40 to 80 days of MCRT and 2000 to 3000 mg/L of
MLSS. They can nitrify wastewater reliably and do not have the algae problems of con-
ventional lagoons.
phosphorus is converted to cell mass per 100 mg of BOD removed (Table 22.1). So, tra-
ditional sludge wasting reduces phosphorus by about 1 to 2 mg/L.
Acinetobacter (BioP) microorganisms consist of up to 35% phosphorus, so mixed
liquor with a high percentage of Acinetobacter organisms may contain about 6% phospho-
rus. Systems designed to select for such organisms are called enhanced biological phosphorus
removal (EBPR) systems. They can reduce phosphorus concentrations by 3 to 6 mg/L.
Chemicals can also be used to precipitate phosphorus, but EBPR minimizes the
need for them, thereby reducing their side effects: alkalinity loss and extra sludge pro-
duction. Enhanced biological phosphorus removal systems also produce a better set-
tling sludge.
The disadvantages of EBPR include a higher capital cost (for baffles and mixers to
create selector zones), sensitivity to nitrate or oxygen toxicity, and more complex oper-
ations (WEF, 1998a).
should avoid nitrification, but if both nitrification and BPR are necessary, then plant
staff should take steps to eliminate interference from combined oxygen.
Settled or waste sludge from EBPR processes must be managed carefully to avoid
secondary release—phosphorus emitted from sludge when it is held under anaerobic
conditions. Such phosphorus can be inadvertently recycled to the process reactors.
INFLUENCES. Good BPR depends on the right environment (low ORP) and the
right type and amount of organic matter. Most wastewater will contain enough readily
biodegradable BOD to allow EBPR. Acinetobacter organisms prefer short-chain carbon
compounds, also called VFAs. If the treatment plant is in a warm climate or the collec-
tion system has an extended retention time so septic conditions can occur, then the
wastewater typically will have enough VFAs.
In areas with colder climates or wastewater diluted by infiltration and inflow (I/I),
staff may need to supplement naturally occurring VFAs. Some treatment plants gener-
ate VFAs by holding primary sludge (only 20 or 30% may need processing) in a grav-
ity-based thickener called a primary sludge fermenter. Others maintain a sludge blanket
in the primary clarifier. In either approach, staff maintain a sludge blanket with a 1- to
3-day MCRT to generate enough VFAs so the overflow can supplement the treatment
plant influent and allow EBPR to occur.
To optimize EBPR, either nitrate-nitrogen must be removed from the RAS, or the
RAS flow must be reduced (typically 20 to 30%) to minimize dissolved oxygen entrain-
ment. The ratio of BOD to total phosphorus is also important. A ratio of 25⬊1 or more is
considered necessary to achieve good phosphorus removal (the actual BOD-to-total
phosphorus ratio needed depends on the process involved).
Suspended solids at an EBPR facility may have a phosphorus concentration of 6%
or more, compared to 2% at a conventional biological treatment system. Likewise, a
secondary effluent with 30 mg/L of TSS will contain 1.8 mg/L of particulate phospho-
rus at an EBPR facility, but only 0.6 mg/L at a conventional plant. So, effluent filtration
may be necessary to meet low total phosphorus limits.
Anaerobic/Oxic Process. Developed in the 1970s and patented in the early 1980s, the
A/O process has a similar flow scheme to that of the LE process [Figure 22.12(a)] ex-
cept it uses an anaerobic zone rather then an anoxic one. The main difference between
the LE process and the A/O process is that A/O does not nitrify. Typically, its anaero-
bic zone’s HRT is between 30 and 60 minutes to select for Acinetobacter organisms, and
its oxic zone’s MCRT is between 2 to 4 days to discourage nitrification. The A/O
process typically is not used at treatment plants that need both nitrogen and phospho-
rus removal because other processes do both more effectively.
HRT (hours)
Typically, its anaerobic zone is about the same size as that in the A/O process, while the
anoxic zone has an HRT of 1 hour and the MLR is 100 to 400% of the secondary influent.
The A2/O process allows Acinetobacter organisms to be competitive in the anaero-
bic zone—even while nitrification is occurring—by lowering the nitrate content in the
RAS. It can achieve good denitrification via proper sizing of the anoxic zone.
Advantages Limitations
A/O Operation is relatively simple when Phosphorus removal declines if
compared to other processes nitrification occurs
Low BOD/P ratio possible Limited process control flexibility is
available
Relatively short hydraulic retention time
Produces good settling sludge
Good phosphorus removal
A2/O Removes both nitrogen and phosphorus RAS containing nitrate is recycled to
anaerobic zone, thus affecting
phosphorus-removal capability
Provides alkalinity for nitrification
Produces good settling sludge
Operation is relatively simple Nitrogen removal is limited by internal
recycle ratio
Saves energy Needs higher BOD/P ratio than the
A/O process
UCT Nitrate loading on anaerobic zone is More complex operation
reduced, thus increasing phosphorus-
removal capability
Requires additional recycle system
For weaker wastewater, process can
achieve improved phosphorus removal
Produces good settling sludge
Good nitrogen removal
VIP Nitrate loading on anaerobic zone is More complex operation
reduced, thus increasing phosphorus-
removal capability
Requires additional recycle system
Produces good settling sludge More equipment required for staged
operation
Requires lower BOD/P ratio than UCT
Bardenpho Can achieve 3 to 5 mg/L TN in Less efficient phosphorus removal
(5-stage) unfiltered effluent
Produces good settling sludge Requires larger tank volumes
SBR Both nitrogen and phosphorus removal More complex operation for N and
are possible P removal
Process is easy to operate Needs larger volume than SBR for
N removal only
Advantages Limitations
Mixed-liquor solids cannot be washed Effluent quality depends on reliable
out by hydraulic surges decanting facility
Quiescent settling may produce lower Design is more complex
effluent TSS concentration
Flexible operation Skilled maintenance is required
More suitable for smaller flowrates
PhoStrip Can be incorporated easily into existing Requires lime addition for phosphorus
activated sludge plants precipitation
Process is flexible; phosphorus-removal Requires higher mixed-liquor dissolved
performance is not controlled by BOD/ oxygen to prevent phosphorus release
phosphorus ratio in final clarifier
Additional tank capacity required for
stripping
Significantly less chemical usage than Lime scaling may be a maintenance
mainstream chemical precipitation problem
process
Can achieve reliable effluent
orthophosphate concentrations less
than 1 mg/L
Five-Stage Bardenpho Process. When an anaerobic zone precedes the four-stage Bar-
denpho process [Figure 22.12(c)], the resulting five-stage process can be used to encour-
age the growth of Acinetobacter organisms. The flow schematic of a five-stage (modified)
Bardenpho process resembles the A2/O process followed by a second anoxic zone and a
reaeration (oxic) zone [Figure 22.16(c)]. This process is typically designed to operate with
a total HRT of about 22 hours, although the HRTs vary for each zone: anaerobic (2 hours),
anoxic (3 hours), aerobic (12 hours), secondary anoxic (2 hours), and reaeration (1 hour).
Johannesburg Process. Originally used in Johannesburg, South Africa, this process is es-
sentially a simpler version of the modified UCT process. The Johannesburg (sidestream
denitrification) process minimizes the amount of nitrate fed to the anaerobic zone by in-
cluding an anoxic zone in the RAS flow pattern [Figure 22.16(d)]. It uses the bacteria’s en-
dogenous respiration to denitrify the RAS, while a second (mainstream) anoxic zone
denitrifies the MLR. Its anaerobic zone can operate at a higher MLSS concentration than
the modified UCT process, thereby reducing the HRT necessary for BPR.
Sequencing Batch Reactors. Typically used for denitrification and CBOD removal,
SBRs can also be modified to remove phosphorus (Figure 22.13). This is done by insert-
ing an anoxic period and adding carbon (typically methanol) after the aerated react
period. This ensures that little nitrate remains during the fill period, so the following
anaerobic period will encourage the growth of Acinetobacter organisms. Adding methanol
after the react period (while aeration is shut off) allows nitrate to be consumed. The fol-
lowing short aerated react period ensures that the methanol is totally consumed and
will not contribute to BOD in the final effluent. To ensure that enough VFAs are avail-
able, a carbon source is also added during the initial mix–react period.
These modifications typically cut the SBR’s cycles to three to four per day. The re-
duced cycles and extra complexity make BPR difficult to achieve, compared to conven-
tional, constant-flow BNR processes.
Oxidation Ditches. Oxidation ditches (Figure 22.14) typically are not considered a
BPR process because of the difficulty associated with achieving truly anaerobic condi-
tions when nitrates are being recycled (Stensel and Coleman, 2000). However, good
BPR can be achieved by constructing an external selector, which may incorporate flex-
ible cells allowing a mainstream BPR process to be used with the oxidation ditch [Fig-
ure 22.14(c)]. One oxidation ditch manufacturer promotes dissolved oxygen control in
the outer reactor ring as a means of accomplishing BPR [Figure 22.14(d)].
PhoStrip Process. The “PhoStrip” process, which removes phosphorus via a side-
stream process, was first introduced in the 1970s before any mainstream BPR processes
existed. Although seldom used today because of its chemical requirements and com-
plex operations, the PhoStrip method of removing phosphorus from RAS deserves de-
scription. It involves transferring about one-third of the RAS to an anaerobic stripper
tank (typically a gravity thickener) for a HRT of between 8 and 12 hours. A small
amount of primary effluent or raw wastewater may be added to the tank to enhance
anaerobic conditions. After fermentation, supernatant from the stripper tank is trans-
ferred to a separate reactor, where lime is added to raise the pH and precipitate soluble
phosphorus (Figure 22.15).
Although it uses lime, PhoStrip is considered a biological process because the RAS
in the stripper tank promotes the proliferation of Acinetobacter organisms.
PROCESS CONTROL
Operators of BNR facilities need more process-control knowledge than those of con-
ventional treatment facilities to keep them operating smoothly. The key operating pa-
rameters for a BNR facility typically include
• MCRT,
• F⬊M ratio,
• HRT,
• Oxygen levels,
• Alkalinity and pH control,
• Denitrification,
• Recycle flows, and
• Secondary clarification.
MEAN CELL RESIDENCE TIME. The mean cell residence time (defined in Equa-
tion 22.2) is the key to understanding whether the BNR process has enough time to
function effectively. When evaluating MCRT, operators should answer such questions as
Nitrifying facilities, such as conventional activated sludge, A2/O, and MLE pro-
cesses, typically need between 5 and 15 days of MCRT. Facilities with intricate recy-
cling patterns, such as the UCT process, need between 10 and 25 days of MCRT. Other
processes, such as extended aeration systems, SBRs, and oxidation ditches, need be-
tween 15 and 30 days of MCRT. In cold climates (less than 12 °C), nitrification systems
need at least 30 days of MCRT to function reliably.
MLSS
MCRT =
TSS 2EFF + TSS WAS
Where
Example 22.1: Determine Whether the MCRT is Sufficient for Nitrification. Given
the plant shown in Figure 22.17, what is its MCRT? Will it nitrify wastewater reliably at
12 °C? Assume that MLSS 2000 mg/L, TSSWAS 7000 mg/L, and QWAS 31 000 gpd.
8340 lb TSS
MCRT = = 4.05 days
2060 lb/d TSS
b. For the aerobic (oxic) treatment process:
6255 lb TSS
MCRTOx = = 3.04 days
2060 lb/d TSS
4. Evaluate MCRTOx for nitrification potential:
a. Figure 22.9 indicates that under optimal laboratory conditions, nitrification
could occur if the MCRT is between 2.5 and 8 days. At 12 °C, however, the
process needed an MCRT of about 11 days (using a safety factor 1.5).
b. Calculations showed that the MCRTOx is about 3 days.
c. Conclusion: Nitrification will NOT occur at 12 °C.
Example 22.2: Increase the MCRT. Again, given the plant shown in Figure 22.17, how
much should operators reduce the waste flow (QWAS) to increase the MCRTOx from
about 3 days to 5.5 days over a 5-day period?
Start by determining how many pounds of MLSS must be maintained to achieve
the desired MCRT:
Operators should evaluate the process and recheck their calculations before mak-
ing process changes.
Example 22.3: Determine the MCRT an SBR Needs to Achieve Nitrification. A se-
quencing batch reactor facility has the following characteristics:
The facility does not meet its ammonia limit when winter temperatures are 12 °C
or lower. What MCRT would ensure that nitrification occurs?
2. Determine MCRTOx and compare it to the target values suggested to maintain ni-
trification in Figure 22.9:
According to Figure 22.9, the target MCRTOx should be 12 days when the temper-
ature is 12 °C.
The F⬊M ratio is a good indicator of how well selector reactors will promote the
growth of floc-forming bacteria. When the F⬊M ratio is high, floc-forming bacteria
have a competitive advantage over filamentous bacteria.
Selector loading also helps ensure that nuisance bacteria will not cause operating
problems. The selector cells should be arranged so BOD is taken up rapidly.
Example 22.4: Determine the F⬊M Ratio of a BNR Facility. Given the BNR facility
described in Figure 22.17, determine its F⬊M ratio.
2752 lb BOD/d
2. Determine M for each treatment zone. If the MLSS is 80% volatile, then the cumu-
lative M is:
Zone M Cumulative M
MAx1 (0.035 mil. gal)(8.34)(2000 mg/L)(0.8) 467 467
MAn1 (0.035 mil. gal)(8.34)(2000 mg/L)(0.8) 467 934
MAx2 (0.055 mil. gal)(8.34)(2000 mg/L)(0.8) 734 1668
MOx (0.375 mil. gal)(8.34)(2000 mg/L)(0.8) 5004 6672
Total M 6672 lb
3. Calculate the F⬊M ratio in each selector zone and for the entire system:
With consecutive cells (e.g., An1, Ax1, Ax2, or Ox), F is the “applied” BOD in the
first compartment. M increases as greater portions of the reactor are considered in F/M
calculations.
Although not used in daily BNR operations, HRT indicates whether the plant is
operating within a normal contact time. Nitrifying facilities, such as conventional acti-
vated sludge, A2O, and MLE processes, typically have an HRT between 5 and 15 hours.
Facilities with more intricate recycle patterns, such as the UCT process, typically have
HRTs between 10 and 18 hours. Other processes, such as extended aeration, SBR, and
oxidation ditch systems, have HRTs between 18 and 36 hours.
Example 22.5: Determine the HRT of a BNR Facility. Given the BNR facility shown
in Figure 22.17, calculate the HRT of each zone. Assume that RAS 0.5 mgd and MLR
3.0 mgd.
Cumulative Incoming
Volume volume flow Q ⴙ Nominal HRT Actual HRT
Zone No. (mil. gal) (mil. gal) RAS ⴙ MLR (hr) (hr)
An1 0.035 0.035 2.0 0.56 0.42
Ax1 0.035 0.070 2.0 0.56 0.42
Example 22.6: Determine the Oxygen Demand for a BNR Facility. If a 2.3-mgd BNR
facility nitrifies completely, its influent ammonia 25 mg/L, and its BOD2INF 140 mg/L,
then:
2192 lb/d O2
5415 lb/d O2
3. Calculate how much aeration energy will be saved if the plant also adds denitrification.
a. Determine how much nitrate will be produced:
c. Determine how much money is saved each year if electricity costs $0.12/kWh
(assume the facility uses surface high-speed aerators with a field transfer rate
of 1.3 lb O2/hp/hr):
$650 $0.12
Savings = 30.3 hp × ×
hp ⋅ year ⋅ $0.10/kWh kWh
= $23, 630/yr
Example 22.7: Determine How Much More Air Blower Capacity a Treatment Plant
Needs to Nitrify Wastewater. Suppose an activated sludge plant must nitrify its
wastewater (its previous limits were for CBOD only), and its aeration basin has coarse-
bubble diffusers that are 14 ft below the water surface. Plant staff already know that
they need 1600 lb/d more oxygen to nitrify the wastewater. So, how much more horse-
power is needed?
[Note: For fine-bubble diffusers, the percent efficiency (0.66%/ft)(14 ft) 9.24%]
(100)(lb/d O 2 Required)
Airflow =
(0.0173 lb O 2/cu ft)(Percent efficiency)(1440 min/d)
(100)(1600 lb/d O 2 )
Airrflow =
(0.0173 lb O 2/cu ft)(5.25%)(1440 min/d)
1600 lb/d O 2
(0.2491)(5.25)
1220 cfm
3. Estimate the discharge pressure at the blower [assume 4-ft headlosses in air
piping and equipment (line losses)]:
14 ft 4 ft
Blower hp (cfm)(psi)(0.006)
57 hp
(Note: These calculations are strictly illustrative. See equipment curves, etc., for more
precise calculations.)
7.14 mg CaCO 3
Alkalinity depletion from N =
1 mg N oxidized
so the
7.14 mg CaCO 3
Alkalinity required = 22 mg/L NH 3 -N ×
1 mg NH 3 -N
= 157 mg/L CaCO 3
Example 22.9: Determine the Anoxic Selector Size Needed to Achieve a Specific
Nitrate Limit. Suppose a 2.3-mgd wastewater treatment plant plans to convert to the
MLE process. Its secondary influent has a BOD concentration of 160 mg/L. The waste-
water in the 0.6-mil. gal aeration basin is about 15 °C and contains 3000 mg/L of MLSS
(VSS 2400 mg/L). The plant needs to denitrify 18 mg/L of produced nitrate. What
size must the anoxic selector be to meet an effluent nitrate limit of 7 mg/L?
18 mg/L – 7 mg/L
NO X
SDNR =
lb MLVSS
211 lb/d NO X
=
(8.34)(2400 mg/L)(0.18 mil. gal)
211 lb/d NO X
=
3603 lb MLVSS
0.059 lb NO X
=
1 lb/d MLVSS
4. Compare the calculated SDNR with the range of SDNRs published in Metcalf and
Eddy, Inc. (2003) to confirm that too much nitrate will not be applied to the anoxic
zone.
RECYCLE FLOWS. For wastewater facilities with either ammonia and/or nitrate
limitations, it will be necessary to adjust recycle flows (typically RAS flow and/or
MLROx) to achieve operational goals. The following are examples of calculations and
that may be required.
Example 22.10: The Effect of BNR on RAS and MLR. Suppose that a 1.7-mgd acti-
vated sludge plant has a 0.4-mil. gal aeration basin, 32 mg/L of ammonia-nitrogen in
its secondary influent, 1500 mg/L of MLSS, and 8000 mg/L of TSS in its RAS. To nitrify
its wastewater, the plant will need to operate at an MLSS of 3000 mg/L. The plant will
also be converted to the MLE process, with MLR returned to the anoxic zone for deni-
trification. Its new nitrous oxide standard will be 12 mg/L. With this in mind, deter-
mine the RAS flow needed to maintain 3000 mg/L of MLSS, and calculate the MLR to
the anoxic zone (MLROx).
Because the TSS2EFF is small, assume it equals zero and rearrange the equation so
Q TSS RAS
= −1
Q RAS MLSS
Therefore
2. Calculate the MLROx (Equations 22.15 to 22.18). First, determine how much
nitrous oxide must be removed to achieve the limit (assume 1 mg/L of produced
nitrous oxide for each 1 mg/L of influent ammonia):
PNOx
Mr = − 1 − Rr (22.20)
NO X2EFF
Where
So
32 mg/L 1.02 mgd
Mr = − 1−
12 mg/L 1.7 mgd
MLR Ox MLR Ox
Mr = 1.07 = =
Q 1.7 mgd
Therefore
1.8 mgd
Example 22.11: Calculate Secondary Clarifier Capacity. Given the activated sludge
system in Example 22.10, assume that peak flows are three times the average. There-
fore, plant data are as follows:
Present Future
Flow (mgd)
Q average 1.7 1.7
Q peak 5.1 5.1
QRAS 0.4 1.02
MLSS (mg/L) 1500 3000
RAS (mg/L) 8000 8000
Suppose the plant has two 47-ft-diameter secondary clarifiers (total clarifier area
3470 ft2). Calculate the current and future clarifier loading.
When these HLRs are compared to those typically suggested for operations, the
average HLRs are within the range typically suggested for good performance (400 to
600 gpd/sq ft). Future peak HLRs, however, fall slightly above the recommended
range (800 to 1200 gpd/sq ft).
(Q + Q RAS )(8.34)(MLSS)
SLR =
Clarifier area
(1.7 mgd + 0.4 mgd)(8.34)(1500 mg/L)
Present SLR 1.7 =
3470 sq ft
= 7.6 lb TSS/d/sq ft
When these SLRs are compared to those typically suggested for operation, the
average SLRs are below that recommended for BNR plants (20 lb TSS/d/sq ft).
Future peak SLRs are slightly above that recommended for BNR plants (35 lb
TSS/d/sq ft).
So, the secondary clarifiers seem to be adequate for average loading but marginal for
peak loads. Success depends on the clarifier design and configuration. Sludge settleabil-
ity and characteristics will determine whether more clarifiers should be constructed.
TROUBLESHOOTING
Some of the most common operating problems at BNR facilities are foam control and
bulking (nonsettling) sludge (Table 22.7, 22.8, and 22.9).
Observations Solutions
BOD 25 mg/L, NH4-N 3 mg/L Ensure MCRT is sufficient for design temperature
Effluent NH4-N varies erratically, 3 mg/L and Waste sludge operation not stabilized, stop wasting
2 mg/L, in a short period of time until NH4-N is 2 mg/L for 3 consecutive days; restart
a careful sludge-wasting operation
Effluent NH4-N 3 mg/L, dark brown or Check aeration zone DO, should be 2 mg/L; check
black MLSS MLSS solids balance; MCRT may be too high
Brown clumps of sludge floating on final clarifier Nitrogen gas bubbles being produced in stagnant sludge
surface; effluent NH4-N 2 mg/L, BOD in clarifier; check return sludge recycle rate, increase
30 mg/L frequency of cleaning of clarifier walls and weirs
Effluent NH4-N 10 mg/L, BOD 25 mg/L Check MCRT, check mixed liquor DO, check MLSS; if all
appear satisfactory, plant may be receiving inhibitor of
of nitrification from industry; check pretreatment program
Effluent NO3-N 7 mg/L, NH4-N 2 mg/L DO in anoxic zone, should be less than 0.2 mg/L
FOAM CONTROL. Activated sludge processes typically have foam (froth) floating in the biolog-
ical reactor. In a well-operated process, a 50- to 80-mm (2- to 3-in.) layer of light tan foam will cover
between 10 and 25% of the reactor surface. Excessive foam, however, can degrade operations. Three
types of foam typically are problematic: stiff, white foam; brown foam (either greasy and dark tan or
thick, scummy, and dark brown); and very dark or black foam.
Observations Solutions
Effluent SP 1 mg/L, TP 2 mg/L, Ensure DO in anaerobic zone is 0.2 mg/L, NO3-N in
BOD 20 mg/L anaerobic zone is 5 mg/L
Effluent SP 1 mg/L, TP 2 mg/L, Check SS frequently, use polymer or metal salt to control SS;
BOD 25 mg/L, SS 30 mg/L find reason for high SS, such as high flow rates, high return
sludge recycle rate, or bulking
Effluent SP 0.2 mg/L, TP 1 mg/L, Plant may be organically overloaded; check F:M ratio
BOD 30 mg/L
Effluent SP 1 mg/L, TP 2 mg/L, May not be enough low molecular weight acids in anaerobic
BOD 15 mg/L zone; consider adding anaerobic digester supernatant to
this zone
Thin-looking MLSS, effluent TP 2 mg/L, Sludge may be forming a blanket in secondary clarifier, check
BOD 30 mg/L sludge blanket depth, adjust recycle rate if necessary
Effluent TP 2 mg/L, BOD 25 mg/L, Start or increase dose of metal salt to assist in precipitating TP
SS 25 mg/L
TABLE 22.9 Troubleshooting facilities that remove both nitrogen and phosphorus.
Observations Solutions
Effluent TP 2 mg/L, NH4-N 2 mg/L, Check recirculation rate between aerobic zone and anoxic
NO3-N 7 mg/L zone, increase rate if necessary; check DO in anoxic zone,
should be 0.2 mg/L
Effluent TP 2 mg/L, NH4-N 2 mg/L, Check feed rate of optional chemical dosing pump for
NO3-N 2 mg/L, BOD 30 mg/L adding organics to second anoxic zone
Effluent TP 2 mg/L, NH4-N 3 mg/L, Check DO in aerobic zone, should be 2 mg/L
NO3-N 5 mg/L, BOD 25 mg/L
Stiff, White Foam. If stiff, white foam builds up and the wind can blow it onto walk-
ways and other structures, where it looks unsightly and makes working conditions
hazardous. This foam also can be odorous and transport pathogens. If it overflows to
the secondary clarifiers, this foam tends to collect behind the influent baffles, creating
more cleaning problems, and can plug the scum-removal system.
This type of foam indicates a “young” sludge (low MCRT); it is typically found in
new or underloaded plants when the MLSS concentration is too low and the F⬊M ratio
is too high. The foam may consist of detergents or proteins that cannot be converted to
food by the type of bacteria that are predominant in the mixed liquor. Such foam may
occur when
Excessive Brown Foam. Brown foam occurs in plants operating at low loading ranges.
Nitrification plants operating in nitrifying mode, for example, typically have low to
moderate amounts of chocolate brown foam. If Nocardia (a filamentous organism) is
present, the foam will be strong (not easily collapsed), greasy, and dark tan. It also will
overflow onto the clarifier surface. Filamentous organisms containing scum should be
wasted from the system rather than returned to the biological reactors in order not to
concentrate foam-causing organisms.
Treatment plants that re-aerate sludge typically have heavy, greasy, dark brown
foam in their aeration stage. A thick, scummy, dark brown foam indicates an old
sludge (long MCRT). This foam can build up behind influent baffles in the clarifier, cre-
ating a scum disposal problem.
Overall, brown foam is likely to occur when
Very Dark or Black Foam. Very dark or black foam occurs when aeration is insuffi-
cient (anaerobic conditions) or industrial wastes, such as dyes and inks, are present. To
correct this problem, operators should
• Increase aeration;
• Investigate industrial waste sources for dyes or inks; and
• Reduce the MLSS concentration.
BULKING SLUDGE. If the supernatant above poorly settling sludge is clear, fila-
mentous microorganisms are hindering settling. To correct this problem, operators should
If the supernatant above poorly settling sludge is cloudy, the problem is dis-
persed-growth bulking caused by improper organic loading, overaeration, or toxics.
REFERENCES
Daigger, G. D.; Sedlak, R.; Eckenfelder, W.; Jenkins, D.; Stensel, D.; Banard, J. (1988)
Principles and Practice of Nutrient Removal from Municipal Wastewater. Presented
seminars sponsored by The Soap and Detergent Association, New York.
Harrison, J. R. (1999) Coupling TFs or RBCs with Activated Sludge. Paper presented at
the 4th Annual Central States Water Environment Educational Seminar, Madison,
Wisconsin,
Metcalf and Eddy, Inc. (2003) Wastewater Engineering, 4th ed.; McGraw-Hill, Inc.: New
York.
Randall, C. W. (1992) Design and Retrofit of Wastewater Treatment Plants for Biological
Nutrient Removal, Vol. 5; Technomic Publishing Company: Lancaster, Pennsylvania.
Stensel, H. D. (2001) Biological Nutrient Removal: Merging Engineering Innovation
and Science. Proceedings of the 74th Annual Water Environment Federation Technical Ex-
position and Conference [CD-ROM]; Atlanta, Georgia, Oct. 13–17; Water Environment
Federation: Alexandria, Virginia.
Stensel, H. D.; Coleman (2000) Technology Assessments: Nitrogen Removal Using Oxida-
tion Ditches; Project 96-CTS-1; Water Environment Research Foundation: Alexan-
dria, Virginia.
U.S. Environmental Protection Agency (1993) Nitrogen Control Manual; EPA-625/R-93-
010; Washington, DC.
Water Environment Federation (1996) Developing Source Control Programs for Commer-
cial and Industrial Wastewater; 2nd ed.; Manual of Practice No. OM-4; Water Environ-
ment Federation: Alexandria, Virginia.
Water Environment Federation (1998a) Biological and Chemical Systems for Nutrient
Removal; Special Publication; Water Environment Federation: Alexandria, Virginia.
Water Environment Federation (1998b) Design of Municipal Wastewater Treatment
Plants, 4th ed.; Manual of Practice No. 8, Water Environment Federation: Alexandria,
Virginia.
Water Environment Federation (2000) Aerobic Fixed-Growth Reactors; Special Publica-
tion; Water Environment Federation: Alexandria, Virginia.
Water Environment Federation (2001) Natural Systems for Wastewater Treatment, 2nd
ed.; Manual of Practice No. FD-16; Water Environment Federation: Alexandria,
Virginia.