-------
4728
CHAPTER 1
FACTORS AFFECTING NITRIFICATION- -KINETICS* —:
Harry E. Wild, Jr., Clalr N. Sawyer, and Thomas C. McMahon
The nitrification phenomenon has been studied inten-
sively by soil scientists for the past century. With the advent.
of biological wastewater treatment systems, chemists and engi-
neers were impressed by the fact that the same phenomenon oc-
curred in their treatment plants. Originally, in the absence
of biological methods of assessing degrees of purification,
chemical analyses served as the major means of evaluation. Ex-
perience soon taught the lesson that highly nitrified effluents
were immune to putrefaction. As a result, waste treatment
plants prior to 1930 were designed as standard or conventional
•
plants intended to accomplish a relatively high degree of ni-
trification, at least during the summer months.
With the development and widespread application of the
BOD test, it became apparent to many designing engineers that
high degrees of waste treatment, in terms of BOD removal, could
i
be accomplished, at marked savings in capital and operating
costs, by designing to avoid nitrification. Thus, from 19*10
until the late 1960's the main objective in the United States
was to design to minimize nitrification.
*Paper presented at the Boston Meeting of the WPCF,
October 1970. Scheduled for publication in Jour. WPC^. Per-
mission to reproduce for distribution at this EPA Symposium
granted by WPCF.
METCALF ft EDDY
-------
1-2
Many of the newly designed high-rate or modified plants,
and some older plants suffering from overloads, were plagued
with denitrification and resultant "rising sludge" problems in
the final clarifiers. These problems stimulated numerous
studies on how to control nitrification, since it was a physical
impossibility to accomplish high degrees of nitrification, in
most cases, without expansion of the plant facilities.
Although the NOD of unnitrified effluents was well under-
stood, sanitary engineers generally dismissed this matter from
their minds on the basis of three premises:
1. Nitrification is caused by special organisms,
the population of which is minimal in surface
waters.
2.• The reaction constant for nitrogenous oxidation
is small in relation to the constant for carbo-
naceous matter.
3- Oxidation of ammonia to nitrates simply converts
dissolved oxygen to a form from which it is still
available to prevent formation of anaerobic
conditions.
The philosophy that unnitrified effluents are not dam-
aging to receiving streams has been undermined by biologists
and conservationists who point out that nitrates will not
satisfy the oxygen requirements of fish and many other aquatic
organisms and by the river and stream investigations of Gannon(
and the Michigan Water Resources Commission, as reported by
Courchaine^2).
MCTCALF a EDDY
-------
1-3
As a result of the studies conducted In Michigan, manv
states are now requiring that NOD be considered as well as BOD
in any analysis of pollutional loads that streams can bear.
This will undoubtedly mean that many plants of the future will
be designed to accomplish extensive nitrification, at least
during the warmer months of the year when oxidation rates are
highest and stream flows are apt to be minimal.
With regard to eutrophication of surface waters, nitrogen
in the fixed forms of ammonium and nitrate ions is considered
to be one of the major nutrients supporting blooms of green
and nonnitrogen fixing blue-green algae. Its removal from waste-
waters is being requested in some areas and considered in many
others. Where discharge is to lakes or reservoirs with signif-
icant detention times, seasonal removal will not suffice and
365-day per year performance will be expected. Removal through
nitrification followed by denitrification represents the most
promising method at this time. It has the advantage of return-
ing nitrogen to the atmosphere in its natural form.
Nitrification and Population Dynamics
It seems certain at this time that nitrification will
play a greater and greater role in wastewater treatment in the
future because of anticipated increased NOD removal requirements
and possible use of systems employing nitrification-denitrifica-
tion for nitrogen removal.
METCALF a EDDY
-------
1-1
It is conceivable that NOD removal will be a seasonal
requirement in most locations and will occur during the warm
months of the year. If so, then conventional designs of bio-
logical systems, similar to those used prior to 1930 or any
which are capable of maintaining conditions so that the recip-
rocal growth rate of the nitrifying bacteria is less than the
mean cell residence time or sludge retention time as described
by Jenkins and Garrison(3), will be required. In simple terms,
this means that nitrification in plants can be maintained only
when the rate of growth of nitrifying bacteria is rapid enough
to replace organisms lost through sludge wasting. When they
can no longer keep pace, the ability to nitrify decreases and
may become extinct.
It h'as been quite well established that no treatment
plants, including those of the extended aeration type, are ca-
pable of accomplishing complete nitrification on a year-round
basis in our northern states. In situations where nitrogen re-
moval is required and the nitrification-denitrification route is
preferred, it will be mandatory to accomplish nitrification in
a separate biological system where the reciprocal growth rate
can be kept less than the mean cell residence time at all times.
This will mean that a large part of the normal BOD will have to
be removed before the wastewater enters the nitrification unit.
Such a system is shown as Figure 1. It is believed that a BOD
of 40 or 50 mg/L can be tolerated in the feed to the nitrification
METCALF ft EDDY
-------
1-5
unit; consequently either high-rate activated sludge or trick-
ling filter systems should be acceptable for the first stage
of treatment.
Nitrification Kinetics
The response of both nitrite and nitrate forming bac-
teria in pure culture to various environmental conditions has
been extensively studied. The effect of pH upon the respira-
tion rate of Nltrosomonas as reported by Meyerhof(^) and Enge]
and Alexander(5), is shown on Figure 2 and the effect upon
Nitrobacter, as reported by Meyerhof(6) is shown as Figure 3-
Early studies on nitrification in wastewater treatment
were mainly related to its control to prevent "rising sludge"
problems in the activated sludge process. These brought dis-
•
solved oxygen under close scrutiny, since it was the only envi-
ronmental condition that could be considered readily controllable
under normal operating conditions. Bragstad and Bradney(?) re-
ported that dissolved oxygen must be kept below 0.5 mg/L to
control nitrification. Recently, Downing et al(8) ana Jenkins
and Garrison^) have reported on other aspects affecting nitri-
fication and Zanoni(9) investigated the effect of temperature on
the velocity constant for nitrification in treated effluents.
Our study was prompted by three major considerations:
1. A paper by Borchardtdu) which indicated that
temperature had little effect on nitrification in
the range of 15 to 35 deg C (see Figure lJ), in
opposition to published data(H)-
METCALF ft EDDY
-------
1-6
2. A lack of information on sludges in systems
receiving feed stock containing relatively
low BOD.
3- A considered need to establish a quantitative
basis for evaluating the ability of nitrifying
sludges to convert ammonia to nitrate, under
various temperature and pH conditions.
Method of Study
The investigations to be described were conducted at A
Marlborough, Massachusetts, where a 10-gpm pilot nitrification
unit, receiving settled high-rate trickling filter effluent,
was operated, open to the weather from October 1969 through
April 1970. All observations on the effects of dissolved
oxygen were made in the pilot plant. The studies on the in-
fluence of temperature and pH were made in the laboratory
using return sludge from the nitrification unit and settled
trickling filter effluent in the apparatus shown as Figure 5-
The batch studies on pH and temperature were conducted with
dissolved oxygen levels above 2 mg/L to ensure that it would
not be an inhibiting factor.
Experimental Results
Pilot Plant
The results of measuring the dissolved oxygen in the
aeration tank of the pilot plant and the resulting effluent
quality are indicated on Figure 6. The dissolved oxygen
MCTCALF ft EDDY
-------
1-7
concentration was measured twice daily and found not to have a
significant variance during any given day. The effluent ammonia
nitrogen concentration was taken from a 24-hour composited
sample.
The wide range of dissolved oxygen concentration resulted
from breakdowns in one of the two available compressors and a
varying demand for air at other locations.
Figure 6 indicates that there was apparently no inhibi-
tion of nitrification occurring at dissolved oxygen levels
exceeding 1.0 mg/L.
Laboratory
The laboratory studies were concerned with determining
the effect of temperature and pH under carefully controlled
conditions. The procedure used involved collection of samples
of return sludge from the nitrification pilot plant and of
settled trickling filter effluent, determination of suspended
and volatile suspended solids in the return sludge, and the
adjustment of portions of each to definite pH and temperatures
before making the desired mixtures in the aeration units. In
most instances, the trickling filter effluent was supplemented
with a dilute aqueous solution of ammonium chloride in order
to give runs of sufficient duration to obtain three or more
experimental values.
The rate of nitrification was determined by measuring
residual ammonia nitrogen on grab samples of mixed liquor which
were filtered immediately after collection. Dissolved oxygen,
METCALF a EDDY
-------
1-8
pH, and temperature were monitored continuously during the course
of each study. Dilute sodium hydroxide was added to control oH
as needed. The system of study may have involved some slight
loss of ammonia at pH levels above 8.5 but such losses were too
insignificant to be detected from a plotting of the data.
It was assumed that the relative population of nitrifiers
in the total MLVSS concentration for the duration of the study
remained constant. It is felt that this assumption was justi-
fied due to the long duration of the pilot studies run under
the same conditions, employing settled trickling filter efflu-
ent as feed stock.
Mixed Liquor Volatile Suspended Solids. The nitrifica-
tion studies were conducted with MLVSS concentrations within
the range of 800 to 6,000 mg/L.
A sample of two of the experiments run at the same pH
and temperature conditions but with two different mixed liquor
volatile suspended solids is shown on Figure 7. It was observed
that the time to completely nitrify the same amount of ammonia
nitrogen per gram of MLVSS was constant given the same environ-
mental conditions. This allows direct comparisons to be made
for different MLVSS concentrations in the study and permits
subsequent data to be expressed in terms of mg of ammonia nitro-
gen per mg of MLVSS.
Ammonia. The augmented ammonia nitrogen concentrations
for the studies varied from 6 to 60 mg/L. The ammonia nitrogen
concentration had to be augmented on many occasions because the
MCTCALF a EDDY
-------
1-9
time required for complete nitrification of low levels was so
short that only one or two samples could be analyzed prior to
attaining the zero level. Two sample results are shown on
Figure 8. Both of the experiments were conducted at the same
pH and temperature conditions. As can be seen from the figure,
the slopes of the lines are parallel and constant for all re-
sidual concentrations of ammonia nitrogen regardless of the
initial concentration. This would indicate that nitrification
is not inhibited at concentrations normally found in a domestic
wastewater system. This also allows adjustment of other
data for different ammonia nitrogen concentrations by con-
structing a line parallel to the experimental line at the de-
sired concentration.
BOD. A special study was made to determine the effect
of variable BOD upon the rate of nitrification. This was ac-
complished by nitrifying three different samples. The temper-
ature and pH for all three units were the same. The wastewater
in the first unit was primary effluent with a BOD of 110 mg/L,
the second unit contained settled trickling filter effluent with
a BOD of **5 mg/L, and the third unit contained nitrification
effluent from the pilot plant with a BOD of 5 mg/L. All samples
were augmented with enough ammonium chloride to give a reason-
able duration for the test.
Figure 9 shows the results of this special study. Within
the limits of the study, there was no apparent inhibition of
MCTCALF ft EDDY
-------
1-10
nitrification for the various BOD concentrations. It should be
\
realized that this study was undertaken to determine the reac-
tion of the nitrifiers to a shock loading of BOD and that any
sustained high BOD loading would eventually cause nitrification
to cease, due to the washing out effect wasting of sludge would
have on the nitrifiers. This is concluded because of the low
growth rate exhibited by the nitrifiers as compared to those
organisms utilizing carbonaceous BOD and the established fact
that an increased BOD loading in a conventional system leads
to greater sludge production.
pH. The pH range investigated in these studies was from
6.0 to 10.5. The samples were adjusted to the desired pH level
and maintained at that level for the duration of the experiment.
The ammonia nitrogen weight per MLVSS weight ratio of the grab
samples was plotted against time and all the other variables
were noted. The time plot allowed calculation of the exact time
of complete nitrification, i.e., complete oxidation of ammonia.
A sample graph is presented on Figure 10. This graph shows two
sample results both of which were obtained at a temperature of
20 deg C. The pH of one sample was 8.5 and the other 6.5. The
figure also shows an adjusted line to compensate for different
initial concentrations of ammonia.
Three factors are immediately evident from the preceding
figure.
1. There was no apparent initial uptake of ammonia
nitrogen by the nitrifiers.
MCTCALF A EDOY
-------
1-11
2. There was no lag time Involved in the rate
of nitrification.
3. The rate was uniform and constant for the entire
length of the experiment. This indicates that
the nitrifiers work at maximum efficiency at
all times independent of the residual concen-
tration of ammonia nitrogen.
Our studies indicate an optimum pH for nitrification to
be 8.4. Figure 11 indicates that 90 percent of the maximum
o
rate occurs in the range of 7.8 to 8.9 and that outside the
ranges of 7-0 to 9.8 less than 50 percent of the optimum rate
occurs.
Temperature. The temperature studies covered the range
from 5 deg C to 30 deg C, and nitrification occurred at all
temperatures investigated. The rate of nitrification increased
with temperature throughout the full range. Figure 12 indicates
the straightline relationships for two sample experiments run at
different temperatures but the same pH. One adjusted line is
shown to offset the initial ammonia concentration difference.
There was no lag period observed nor any decrease in the rate of
nitrification as the residual ammonia concentration decreased.
The relationship of the rate of nitrification at all
temperatures studied to the rate at 30 deg C is indicated on
Figure 13. Since 30 deg C is a very high wastewater temperature
for all but the most southerly states in the United States, a
METCALF & EDDY
-------
1-12
summary of relative rates in terms of other maximum temperatures
is as shown in Table 1, based upon the data of Figure 13.
Table 1. Relative Rates of Nitrification
at Various Temperatures
Temperature deg
^0
100
25
80
100
20
60
75
100
15
48
60
80
C
10
27*
34
45
5
12*
16*
21
"Abnormal temperatures for maximums stated.
These data indicate, on the basis of temperature alone
and the most adverse conditions considered possible, that up to
five times the detention time may be needed to accomplish com-
plete nitrification in the winter as is needed in the summer.
However, temperature effects can be overcome to a considerable
degree by increasing mixed liquor suspended solids and adjustment
of pH to more favorable levels. Optimum design for complete
nitrification will depend upon the best combination of aeration
tank capacity, mixed liquor, suspended solids, and pH for winter
operating conditions. Under summer conditions, operation will
be possible at less favorable pH levels and lower mixed liquor
solids.
Discussion
When all of the above information is evaluated, rates of
nitrification can be computed. The rate of nitrification has
been defined as the weight ratio of ammonia nitrogen oxidized
per day to the mixed liquor volatile suspended solids.
MCTCALF ft EODY
-------
1-13
The rates for any pH within the range of 6.0 to 10.5 are
shown on Figure 1^. All of the rates are for a temperature of
20 deg C. As can be seen from the figure, the rate varies from
a maximum of 0.185 g NHg-N nitrified per day per g MLVSS at a
pH of 8.4 to a minimum of 0.020 g NI^-N nitrified per day per
g MLVSS at a pH of 6.0.
The results obtained in this study with respect to pH
show good correlation with the results indicated in the pre-
vious section of the paper on work performed by others.
Our results on temperature effects are opposed to what
was observed at Ann Arbor, Michigan, and reported by Borchardt(1
as shown on Figure 4. Borchardt's low temperature observations
were made in extended aeration studies by measuring the ammonia
and nitrate nitrogen in the effluent of the units. It is felt
by the authors that the apparent effects of temperature were
observed because the units were not being stressed to their
limit of nitrification at the higher temperatures and complete
nitrification was being obtained in less time than the detention
time of the units. As the temperature decreased, the time re-
quired to obtain complete nitrification approached the detention
time of the units and when the temperatures dropped low enough,
the time required for complete nitrification exceeded the deten-
tion time and this lesser percent of nitrification was noted.
Our results are from units which were stressed to their capacity
at all times, and indicated an immediate drop in efficiency as
the temperatures decreased, which is in agreement with results
reported by Sawyer and
METCALF a EDDY
-------
Figure 15 gives the expected rate of nitrification com-
pared to temperature for various selected pH conditions. This
figure allows the computation of the time required for complete
nitrification at any MLVSS concentration, ammonia nitrogen con-
centration, temperature, and pH.
The curve at the optimum pH of 8.4 was determined from
information gathered during the course of the study. The curves
for 75 percent and 50 percent of the optimum rates were com-
puted from the rates experienced at pH 8.4 The pH values for
the 75 percent and 50 percent curves were obtained from
Figure 11.
From a practical standpoint, Figure 15 indicates that if
the nitrification system were run at 50 percent of the optimum
conditions, the time required to completely oxidize the ammonia
nitrogen would double or the MLVSS would have to be carried at
twice the level necessary for complete nitrification under op-
timum conditions.
Summary
The preceding may be briefly summarized as follows:
1. The ammonia nitrogen concentration did not inhibit
nitrification in concentrations of less than
60 mg/L.
2. pH did affect the rate of nitrification. Optimum
pH was found to be 8.4.
METCALF a EDDY
-------
1-15
3. Temperature did affect the rate of nitrification.
The rate increased through the range of 5 deg C to
30 deg C, in reasonable agreement with the van't
Hoff-Arrhenius law.
4. The time required for nitrification is directly
proportional to the amount of nitrifiers present
in the system.
5. Instantaneous increases or decreases in BOD concen-
tration from 50 mg/L to 5 or to 110 mg/L did not
affect the rate of nitrification. However, it
would be expected that a change in the average BOD
concentration of the feed would affect that per-
centage of MLVSS which is composed of nitrifiers,
and as a result would affect the time to achieve
complete nitrification.
Acknowledgments
This research was supported by the Commonwealth of
Massachusetts, Water Resources Commission, Division of Water
Pollution Control, Grant 68-1. Appreciation is acknowledged
for the interest shown and the cooperation extended by Messrs.
John Elwood and Alfred Ferullo of the Division of Water Pol-
lution Control, and Mr. Harry P. Loftus, Commissioner of
Marlborough Public Works Department.
METCALF a EDDY
-------
1-16
Special appreciation is extended to Mr. John Hartley,
Superintendent, and the staff of the Marlborough Easterly
Treatment Plant. Their work on the installation and help in
operation of the nitrification pilot plant contributed im-
measurably to the success of the project.
METCALF a EDDY
-------
1-17
REFERENCES
1. Gannon, J. J. "River BOD Abnormalities". Bull. 05168-1-F,
The University of Michigan, Office of Research Administration,
Ann Arbor, Michigan.
2. Courchaine, Robert J., "Significance of Nitrification in
Stream Analysis - Effects on the Oxygen Balance." This Jour.,
40, 835 (1968).
3. Jenkins, D. and Garrison, W. E., "Control of Activated Sludge
by Mean Cell Residence Time". This Jour. 40, 1905 (1968).
4. Meyerhof - Arch. f. die ges Physiologic, 166, 255 (1917).
5. Engel, M. S. and Alexander, M., "Growth and Autotrophic
Metabolism of Nitrosomonas Europaea". J. Bact., 76, 217
(1958). —
6. Meyerhof - Arch. f. die ges Physiologie, 164, 416 (1916).
7. Bragstad, R. E. and Bradney, L., "Packing House Waste and
Sewage Treatment at Sioux Falls, South Dakota". Sewage
Works J., 9, 959 (1937).
8. Downing, A. L., Painter, H. A., and Knowles, G., "Nitrifica-
tion in the Activated Sludge Process". Jour. Inst. Sewage
Purif. (Brit.) Part 2, 130 (1964).
9- Zanoni, A. E., "Secondary Effluent Deoxygenation at Different
Temperatures", This Jour., 4^, 640 (1969).
10. Borchardt, J. A., "Nitrification in the Activated Sludge
Process", Bull. "The Activated Sludge Process". The
University of Michigan, Division of Sanitary and Water
Resources Engineering", Ann Arbor, Mich.
11. Sawyer, C. N. and Rohlich, G. A., "The Influence of Temper-
ature upon the Rate of Oxygen Utilization by Activated
Sludge". Sewage Works J., in, 94* (1939).
METCALF & EOOY
-------
NITRIFICATION
BOD
REMOVAL
r
WASTE
SLUDGE RECYCLE
SLUDGE RECYCLE
WASTE
FIG. I TWO STAGE BIOLOGICAL SYSTEM REQUIRED
TO GUARANTEE COMPLETE NITRIFICATION
-------
<
o
u.
o
X
o
u.
o
UJ
90
80
70
60
50
40
<
30
20
10
0
;
/
/
'
L
Al
c
l-jj
fAFTER ENGLE
AND ALEXANDER
\
.
/
/
f
1
i
f
r
/
?
i
/
r
^.^s
AFTER MYERHOF
FSHr-.'?'**
•
"> f
\
fc%
•
\
1
«
1
i
^
6.0
7.0 8.0 9.0
pH
FIG. 2 THE EFFECT OF pH ON OXIDATION OF
AMMONIA BY NITROSOMONAS
100
90
UJ
t 80
te.
\-
z 70
u.
° 60
0
g 50
o
§ 40
u.
0 30
UJ
§ 20
10
0
/
/
/
/
/
/
/
it
/
f
/AFTER
f
/
/
MYI
:RHO
\
F
\
\
\
\
\
5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 105
pH
FIG. 3 RATE OF OXIDATION OF NITRITE
BY NITROBACTER
10.0
METCALF A EDDY
-------
o
o
O
ro
<
o:
UJ
o
tr
100
90
80
70
60
50
40
30
20
10
1
1
t
t
t
1
1
1
1
1
1
1
1
1
1
/'
1
I
1
1
1
1
1
1
1
1
1
1
1
^^*
bx
\
\
V
<
\
\
\
\
\
\
\
\
\
\
\
\
\
»
\
\
\
\
\
I
t
t
I
I
t
I
t
t
\
\
\
\
\
v
FIG. 4
10 20 30 40 50 60
TEMPERATURE,°C
EFFECT OF TEMPERATURE ON
NITRIFICATION AS REPORTED BY
BORCHARDT (10)
70
METCALF a EC*DY
-------
D.O. PROBE
pH PROBE
THERMOMETER
AIR STONE
FIG. 5 LABORATORY AERATION UNIT
METCALF a EDDY
-------
10
8
o»
£
l±J
O
O
o:
02468
NITRIFICATION UNIT AVE. AERATION TANK
10 12
D.O., mg/L
FIG. 6 RELATIONSHIP OF RESIDUAL
AMMONIA TO DISSOLVED OXYGEN
-------
s
II
n
CONDITIONS
pH 8.5
TEMPERATURE 2I°C
TIME.HRS
FIG. 7 EFFECT OF VARIATION
IN MIXED LIQUOR VOLATILE
SUSPENDED SOLIDS
CO
o>
E
tr
UJ
a.
.025
.020
.015
.010
UJ
tr
i
o»
E
.005
CONDITIONS
pH 8.5
TEMPERATURE 20° C
INITIAL NH3-N = 46.5mg/L
-INITIAL NH3-N=26.4mg/L
I 2
TIME, MRS
FIG. 8 EFFECT OF VARIATION
IN AMMONIA CONCENTRATION
-------
.020
CONDITIONS
pH 8.5
TEMPERATURE 20° C
.015
en
X
tr
ui
a.
z
z
.010
CONDITION:
TEMPERATURE 20°C
.005
10
FIG. 9
TIME.HRS
EFFECT OF VARIATION
IN BOD
23456
TIME,HOURS
FIG. 10 EFFECT OF VARIATION IN pH
-------
Ul
100
90
80
70
<
ft
UJ
o
tr.
UJ
a.
50
40
30
20
tv/
,
10
6.0
FIG. II
AT 20° C
X
7.0
8.0
PH
9.0
10.0
PERCENT OF MAXIMUM RATE OF NITRIFICATION
AT CONSTANT TEMPERATURE n pH
.015!
a>
UJ
a.
i 010
UJ
u>
o
OL
005
<
u.
o
CONDITION
pH 8.5
2 3
TIME ,HRS
FIG. 12 EFFECT OF VARIATION IN TEMPERATURE
METCALF a EDDY
-------
too
10
15 20 25
TEMPERATURE,°C
35
FIG. 13 RATE OF NITRIFICATION
AT ALL TEMPERATURES
COMPARED TO THE RATE
AT 30 DEGREES C
VFTCALF A EOI3Y
-------
£
o
cr
UJ
Q.
O
UJ
z
'ro
o»
en
E
o>
RATE OF NITRIFICATION vs pH AT
CONSTANT TEMPERATURE
10.0
i
UJ
0.
0
UJ
H
ro
X
o>
H A_
W OR.
/
^
**^
.
x/
TIMUM
7.5,
>TIMUM
7.0,
pH 8.'
RATE
9.3
RATE
9.8
\
iH
H
"6 5 10 15 20 25 30 35 40 45 5
TEMPERATURE °C
FIG. 15 RATE OF NITRIFICATION vs TEMPERATURE
AT VARIOUS pH LEVELS
METCALF a EDDY
-------
2-1
CHAPTER 2
NITRIFICATION AND DENITRIFICATION
Nitrification and denitrification have been well recog-
nized phenomena in wastewater treatment for many years. The
former occurred to the greatest degree during the warmer months
of the year and was considered highly beneficial in most in-
stances because of the oxygen resource that the nitrates pro-
vided. Because of additional capital and operating costs re-
quired to produce nitrates, American engineers, in general,
attempted to design or use processes which minimized
nitrification.
The problems of "rising sludge" in conventional activated
sludge and standard rate trickling filter plants were shown to
be due to denitrification. The common way of controlling the
problem was to limit nitrification.
The Michigan studies on the significance of nitrogenous
oxidation (NOD) in creating oxygen sag in receiving streams and
other studies showing the role of ammonia and nitrate nitrogen
in stimulating algal blooms have demonstrated the need for in-
formation on how wastewater treatment plants can be designed
to accomplish nitrification and denitrification. Figure 5-1
shows the facilities required to accomplish both in a controlled
manner.
A three-stage biological system is considered necessary
in northern climates where wastewater temperatures drop below
METCALF a EDDY
-------
2-2
65 deg F (18 deg C). The first stage is necessary to remove
carbonaceous BOD5 to levels of about 50 mg/L. The second stage
is needed to accomplish nitrification and should be designed
to employ the plug-flow principal as closely as possible. The
third stage accomplishes denitrification. A source of carbona-
ceous BOD must be added to reduce the nitrates to nitrogen gas
in a reasonable period of time.
METCALF a EDDY
-------
CARBONACEOUS
BOD
NITRIFICATION
DENITRIFICATION
FIG. 5-1 MODEL SYSTEM FOR NITRIFICATION AND DENITRIFICATION
m
O
O
-------
3-1
CHAPTER 3
DESIGN CRITERIA OF NITRIFICATION SYSTEMS
Discussed below are the design criteria which appear to
be reasonable at this time (July, 1971). It must be emphasized
that these criteria are based on very limited experience. A
large amount of work is in progress and it is expected that
many of the criteria will be refined as experience is gained.
NITRIFICATION TANKS
Tank Layout
Because the rate of oxidation of ammonia is essentially
linear (zero order reaction), short circuiting must be prevented,
The tank configuration should ensure that flow through the tank
follows the plug-flow mixing model as closely as possible. This
can be accomplished by dividing the tank into a series of com-
partments with ports between them. Three compartments is a
minimum number. Tanks can be designed for either diffused air
or mechanical aeration systems as shown on Figure 6-1.
Since the oxidation rate of the process varies widely
with temperature, special provisions may be necessary to in-
corporate the necessary flexibility in the oxygen supoly
system, as discussed hereinafter.
pH Control
Nitrification tanks should be sized to permit complete
nitrification under the most adverse combination of ammonia
load and temperature expected and at the optimum pH (8.2-8.6).
METCALF a EDDY
-------
3-2
The nitrification process destroys alkalinity and the pH may fall
to levels which will inhibit nitrification
+ 402 - > 2HN03 + lJH20 + 2C02
_ 2HN03 + Ca(HCOo)? - » Ca(N03)p + 2COg + 2H?0
2NH/|HC03 + 402 + Ca(HC03)2 > CatN03)2 + ^C02 + 6H20
unless excess alkalinity is present in the wastewater or lime is
added to maintain favorable pH levels. Theoretically, 7.2 pounds
of total alkalinity are destroyed per pound of ammonia nitrogen
oxidized to nitrate. One-half of this is due to loss of alkalinity
caused by ammonia and the remainder is due to destruction of
natural alkalinity, as shown in the equations above.
Whether or not lime additions will be required depends
upon the alkalinity of the wastewater and the desired pH of opera-
tion. For operation under the most adverse temperature conditions
and at optimum pH, sufficient lime must be added initially to
raise the pH into the desired range and then 5-^ pounds of hydrated
lime per pound of ammonia nitrogen will be required to maintain
the pH. An actual titration test should be conducted to obtain
design criteria. In Boston sewage, about 250 pounds of hydrated
lime are needed per million gallons to raise the pH initially to
optimum pH range and an additional 700 pounds to hold it there
during the course of oxidation of the ammonia. The total hy-
drated lime requirements are estimated to be about 115 mg/L.
Additional amounts of lime may be required if chemicals, such as
alum, have been added previously for phosphorus removal.
METC ALF ft EDDY
-------
3-3
Marked reductions in lime requirements will result in
any system that can be designed to operate at pH levels of 7.8
or less because carbon dioxide resulting from destruction of
alkalinity and organic matter will be washed out of the liquid
phase by air contact. The pH of such systems will vary somewhat
with the rate of aeration (ventilation).
The type and sensitivity of the pH control system will
depend on the character of the wastewater and the variations
in the ammonia load fed to the system. A proposed system for
pH control under the most demanding situation is shown as
Figure 6-2. In many situations, a lesser degree of control will
be feasible, in some none will be needed.
MLSS and MLVSS Concentrations
Designs based upon MLSS concentration alone should be
avoided since MLSS will not truly reflect the biological mass
in the system. The ratio of MLVSS to MLSS may vary depending
on the nonvolatile suspended solids (including residual chem-
ical precipitates) in the feed. The fraction of the MLVSS
attributable to nitrifying organisms is as yet unknown. How-
ever, for nitrification systems receiving normal secondary
effluents, MLVSS concentrations of 1,500 to 2,000 mg/L appear
to be safe for design.
Tank Capacity
The choice of the "design peak" load depends upon the
circumstances of the specific project, and need not necessarily
METCALF & EDDY
-------
3-4
be the absolute maximum expected load. For many projects, the
use of a peak load factor of 1.5 represents a reasonable peak
at low temperature conditions.
Figure 6-3 shows the permissible volumetric loading of
the nitrification tanks at a pH of 8.4 and at various temperatures
and mixed liquor volatile suspended solids concentrations,
based upon the nitrification kinetics studies at Marlborough,
Massachusetts'-1-'.
Figure 6-4 shows the corrections that must be applied to
the permissible loadings when the pH is different from 8.4. In
plants with well buffered wastewater, it may be more economical
to provide the additional tankage to permit operation at a lower
pH, rather than to add an alkaline material. The following is
a sample calculation for computing the tank size:
Sample Calculation for Tank Volume
Given: Design Flow - 10 mgd.
Average NH^-N concentration to nitrification
tanks - 20 mg/L.
Minimum temperature - 10 deg C.
Operating pH - 7.8.
MLVSS concentration = 1,500 mg/L.
Computed:
1. NH3 load
a. Average - 10 x 8.34 x 20 = 1,670 Ib/day.
b. Maximum - 1,670 x 1.5 = 2,500 Ib/day.
METC ALF ft EDDY
-------
3-5
2. Tank volume at 10 deg C, MLVSS = 1,500 mg/L.
a. From Figure 6-3, volumetric loading
=8.2 lb/1,000 cu ft.
b. Tank volume = 2,500 -\ r>3
8.2
= 305,000 cu ft.
3. Tank volume adjusted to pH 7.8 (See Fig. 6-4)
= 347,000 cu ft.
4. Check detention period
3*17,000 x 24 x 7.^
10 x = 6'2 nr'
Oxygen Requirements
Stoichmetrically, each pound of ammonia nitrogen that is
nitrified requires 4.6 pounds of oxygen. (The amount of ammonia
nitrified is usually slightly more than the amount of nitrate
measured because some denitrification occurs.) Usually, it is
assumed that all of the ammonia fed will be nitrified. An
additional oxygen allowance must be made for carbonaceous BOD
that escapes from the secondary treatment process.
Nitrification appears to be uninhibited at dissolved
oxygen concentrations of 1.0 mg/L or more. Design based on
maintaining 3.0 mg/L of dissolved oxygen in the mixed liquor
under average loading conditions includes a reasonable factor
of safety. Under peak loading the dissolved oxygen concentra-
tion may be permitted to fall somewhat but not below 1.0 mg/L'.
METCALF 8r EDDY
-------
3-6
Sample Calculation for Oxygen Requirements
Given: Design Plow - 10 mgd.
Average NH^-N concentration - 20 mg/L.
Average BOD - 30 mg/L.
Computed:
1. NH3 load
a. Average = 1,670 Ib/day.
b. Maximum = 2,500 Ib/day.
2. BOD load = 2,500 Ib/day.
3. Oxygen requirement:
a. NHo oxidation -
2,500 x 4.6 = 11,500
b. BOD requirements
2,500 x 1.5 3,750
c. Total 15,250 Ib/day.
To design the aeration system, the total oxygen require-
ment must be corrected to actual operating conditions by the
use of well-known equations incorporating such factors as:
1. Critical wastewater temperature.
2. Minimum dissolved oxygen concentration.
3. Coefficient of wastewater oxygen uptake rate (alpha).
l|. Coefficient of wastewater dissolved oxygen satura-
tion (beta).
5. Altitude of plant.
The rate of nitrification will vary significantly with
temperature and pH, and compensation for this must be made
METC ALF a EDOV
-------
3-7
In the design of the plant. During the summer, the following
methods can be used to match the oxygen demand rate to the
oxygen supply rate:
1. Reduce MLSS concentration.
2. Reduce pH by reducing chemical supply.
3. Reduce tankage in service while increasing
oxygen supply to the tanks remaining in service.
Miscellaneous
Although the nitrification process will handle the normal
variations in ammonia load found in raw wastewater, experience
at the Washington, D.C. pilot plant indicates that nitrification
in the secondary treatment process must be carefully controlled
to ensure stable operation. Experience at South Lake Tahoe,
California, indicates that the addition of 2-8 mg/L of chlorine
to the effluent of the carbonaceous aeration tank will effectively
prevent nitrification.. In addition, excessive amounts, of car-
bonaceous BOD and suspended solids that escape from the secondary
treatment process, such as those associated with "bulking"
sludge caused by filamentous growths, must not be so great that
sludge wasting from the nitrification process causes a washout
of the nitrifying organisms. Carbonaceous BOD concentrations
higher than 50 mg/L in the nitrification influent may interfere
with winter operation.
Foam spray systems have not been found to be necessary
where the MLSS concentration was greater than 2,000 mg/L.
1ETCALF A EDOY
-------
3-8
The substances listed below have been shownto have an
inhibiting effect on the nitrification process in concentrations
greater than those shown:
Halogen substituted phenolic compounds - 0.0 mg/L.
Thiourea and thiourea derivatives - 0.0 mg/L.
Halogenated solvents - 0.0 mg/L.
Heavy metals - 10 to 20 mg/L.
Cyanides and all compounds from which hydrocyanic
acid is liberated on acidification - 20 mg/L.
Phenol and cresol - 20 mg/L.
SETTLING TANKS
Design information on settling tanks serving nitrification
systems is generally limited to pilot plant research studies.
The criteria given herein represent what has been determined
to date.
Surface Loadings. The maximum permissible hydraulic
surface loading appears to be approximately 1,000 gpd/sq ft.
It may be necessary to reduce this loading somewhat if the MLSS
concentration is greater than 2,000 mg/L, because of limiting
sedimentation tank solids loadings.
Mulbarger^ noted at the Manassas, Virginia, pilot plant
that settling improved in the nitrification settling tanks
when alum was added to upstream processes, probably due to
carryover of alum floe. It has also been noted that the periodic
addition of waste sludge from the secondary treatment process
M ETC ALF » EDDY
-------
3-9
improves settling. In cases where nitrification units follow,
addition of waste sludge from them may facilitate a more rapid
buildup of nitrifying organisms.
Number of Tanks. Because of the relatively slow growth
and settling rates of nitrifying sludges, it is desirable to
provide more than two settling tanks to ensure that the sludge
is kept within the system when a tank is down for maintainence
and repair. Pour tanks is a desirable minimum number.
Depth. A depth of 12 feet is recommended.
Sludge Collection Equipment. Denitrification may occur
in settled nitrified sludge and create problems of rising sludge.
Suction-type sludge collection equipment may be used to remove
the sludge as quickly as possible. However, because the food
supply will be very limited, little problems with floating sludge
can be expected. The settling tanks should be equipped with
skimmers and provision should be made to use the scum system to
pump floating sludge to the nitrification tank influent.
Sludge. It is recommended that capacity be provided for
a return sludge rate of 50 to 100 percent of flow since the
nitrification sludge is light and does not compact well.
Sludge wasting was not necessary at the pilot plants
at Washington, D.C. , and Marlborough, Massachusetts, but pro-
visions should be made to dispose of waste nitrification sludge
with the waste sludge from the secondary treatment process.
METCALF a EDDY
-------
3-10
References
1. See Chapter 1.
2. Drew, E.A., Chief Engineer, Middle Regional Drainage
Scheme, England.
3- Mulbarger, private communication.
METCALF a EDDY
-------
MECHANICAL AERATION
SEC.
EFF.
T
i
i
RETURN SLUDGE
SEC.
DIFFUSED AIR
EFF
err.
RETURN SLUDGE
MODEL NITRIFICATION SYSTEMS
FIG. 6-1
METCALF a EDDY
-------
FLOW
SIGNAL
ANALYZER - CONTROLLERS
SLURRY
PH
•c
PI
H
n
>
r
•n
FIG. 6-2 pH CONTROL FOR NITRIFICATION SYSTEM
PLAN VIEW
-------
30
25
u 20
o
o
o
to
10
I I I
BASED UPON NITRIFICATION RATES
OBSERVED AT MARLBORO, MASS
TEMPERATURE, °C
FIG. 6-3 PERMISSIBLE NITRIFICATION TANK LOADINGS
METCALF & EDDY
-------
UJ
x
<
u.
o
UJ
o
az
UJ
a.
10.0
2
PI
H
O
FIG. 6-4 PERCENT OF MAXIMUM RATE OF NITRIFICATION
AT CONSTANT TEMPERATURE vs pH
-------
CHAPTER 1| /|_1
DENITRIFICATION BY SUSPENDED GROWTH SYSTEMS
Very little information is available at the present time
to serve as the basis of suspended growth denitrification sys-
tems. In our opinion, the most valid information which can
serve as a basis of rational design comes from the oilot plant
studies at Manassas, Virginia, as reported by Mulbarger (^ .
Figure 7-1 shows the kinetics of the denitrification reaction
in relation to temperature for a given pH range, as reported
by him. The data from which the figure was developed were ob-
tained in laboratory studies in a manner comparable to those
shown on Figure 15 of Chapter 1 and are considered to be fully
as reliable.
Denitrification Tank
The tank layout should assure that the plug-flow mixing
model is followed as closely as possible, because nitrates are
not adsorbed by biological growths and detention periods may be
quite short. Whether covered tanks are required to minimize
absorption of oxygen from the atmosphere is a matter of conjec-
ture. There is some evidence to indicate that properly designed
denitrification units can be made to seal themselves by forma-
tion of a floating scum. In any event, airtight covers are
to be avoided.
Studies by Mulbarger^1) have indicated that optimum pH
for the denitrifying organisms is in the range of 6.5-7-5, the
METCALF & EDDY
-------
1-2
same as for most saprophytic bacteria. The effect of pH as
reported by Mulbarger is as follows:
Reaction rate as minimum
pH percent of optimum rate
6.5 - 7.5 Optimum
6.3 - 7-7 90$
5.6 - 8.6 50%
Although the pH of the effluent from the nitrifying
units may exceed 7-5 at some time during a year, this is no
particular problem because carbon dioxide generated from
oxidation of carbonaceous matter in the denitrification unit
reduces the pH into the favorable range below 7-5 very quickly.
There is no need for addition of chemicals to control pH.
MLSS and MLVSS
The limited experience available has shown that denitri-
fying sludges have settling properties comparable to good acti-
vated sludges. It seems reasonable to assume, therefore, that
mixed liquor solids in the range of 2,000 to 3,000 mg/L can be
maintained without excessive rates of returning sludge. The
volatile matter in the denitrifying sludges at Manassas and
Washington, D.C. is about 75 percent.
Size
Reference to Figure 7-1 will show that the minimum tem-
perature to be allowed for will play a great role in deter-
mining the size of the denitrification tanks, as well as the
MLVSS that can be carried in the system. Figure 7-1A may be used
to compute the size of the denitrification tanks as follows:
METCALF A EDDY
-------
4-3
Sample Calculation for Denitrlfication Tank Volume
Given: Design Flow = 10 mgd.
Average NOo-N + NOg-N concentration = 20 mg/L.
Minimum temperature = 10 deg C.
Expected operating pH = 7-7-
MLVSS = 2,000 mg/L.
Computed:
1. N03-N plus N02-N loading
a. Average = 10 x 8.35 x 20
= 1,670 Ib/day.
b. Peak = 1,670 x 1.5
= 2,500 Ib/day.
2. Tank loading at 10 deg C, optimum pH (from Fig 7-1A)
=22.5 lb/1,000 cu ft.
3. Tank volume at MLVSS = 2,000, optimum pH
= 2,500 x 103
22.5
= 111,000 cu ft.
4. Tank volume at pH 7.7
_ 111,000
.90
= 123,000 cu ft.
5. Check detention period
= 123,000 x 7.48 x 24
10 mgd
= 2.22 hours.
METCALF ft EDDY
-------
Such a system would have over twice the tankage needed
at 20 deg C. For this reason good design will allow for idle
operation of part of the capacity during the warm months of
the year. A design similar to that shown for the diffused
air nitrification system in Figure 6-1 is recommended.
Carbonaceous Matter
Effluents from nitrifying units are exceptionally free
of oxidizable carbonaceous matter (BODC). For this reason
denitrification is very slow unless a readily oxidizable source
of carbonaceous matter is added. Methyl alcohol (Methanol) is
the cheapest commercial source of carbonaceous matter at this
time. Glucose (corn sugar) is the next cheapest source. Methanol
is preferable because it is more completely oxidized than glucose
and, consequently, produces less sludge for disposal.
In some areas, nitrogen deficient industrial wastes, such
as brewery wastes, might be available and suitable for use. All
such waste materials should be employed before considering
methanol, because it is produced from natural gas which is not an
unlimited resource.
When methanol is used for denitrification the basic
reaction involved is:
5CH3OH + 6H+1 + 6NO^ > 5C02 + 3N2 + 13H20
(5 x 32) = 160 (6 x HO = 84.
From the above equation and weight relationships, it might be
concluded that each pound of nitrate nitrogen would require
1ETCALF & EDDY
-------
about 2 pounds of methanol for its reduction. This is true
but some of the methanol is used to produce new cell growth
(sludge) as follows:
(CH..OH) > CO. + (CH.O) + H?0
J A C. C. X e-
Also, nitrified effluents normally carry some dissolved oxygen
into the denitrification tank and some may enter the mixture
as a result of agitation. This increases the amount of methanol
required. An equation commonly used to estimate methanol
requirements is:
Methanol, Ib/day = 2.^7 NO -N + 1.53N02-N + 0.87D.O.
Reports indicate that from 3.0 to 3-5 pounds of methanol per
per pound of nitrate nitrogen are required to consume dissolved
oxygen and leave sufficient to reduce the nitrate to nitrogen gas.
The amount of methanol fed must be very closely con-
trolled by a system such as shown on Figure 7-2 to ensure that
enough is fed to reduce the nitrates and to avoid an excess.
Any excess is not only a waste of chemical but it creates an
undesirable residual BOD.
Equipment
The contents of the denitrification tanks are mixed with
underwater mixers comparable to those used in flocculation tanks
in water treatment plants. The energy provided must be suf-
ficient to keep the MLSS in suspension but controlled to prevent
pickup of atmospheric oxygen as much as possible, unless the
tanks are covered or some other scheme is used to exclude contact
with the air.
METCALF a EDDY
-------
Power requirements of 1/4 to 1/2 hp per thousand cubic
feet have been found to be adequate.
Nitrogen Release
The denitrification reaction results in the formation of
carbon dioxide and nitrogen gas. Both have limited solubility
in water, especially the latter. Because of the gentle mixing
used in the denitrification tanks, the mixed liquor leaving the
tanks is supersaturated with nitrogen, and possibly carbon
dioxide. As a result, gas bubbles tend to form and adhere to
the MLSS and inhibit settling in the final clarifier. Super-
saturated conditions can be relieved by employing aerated open
channel conduits between the denitrification and final settling
tanks. It is recommended that from 5 to 10 minutes detention
be provided at peak flow.
Settling Tanks
The limited experience available indicates that the
settling properties of denitrification sludge, following relief
of supersaturation, are very similar to conventional activated
sludge.
A minimum tank depth of 12 feet is recommended and surface
overflow rates should not exceed 1,200 gal/sq ft/day at peak flows
MLSS concentrations greater than 2,500 mg/L may require larger
tanks due to the higher settling tank solids loadings.
A suction type sludge collector is recommended for
circular tanks. Long rectangular tanks should be equipped with
mid-tank sludge drawoff systems.
METCALF a EDDY
-------
1-7
Skimming facilities should be provided on the settling
tanks and provisions should be made for returning the scum to
the denitrification tank when desired.
Sludge
Return
Capability of returning sludge to the denitrification
tank of up to at least 50 percent and preferably of up to 100
of average flow is recommended.
Waste
Provision should be made for periodic wasting of sludge
from the denitrification systems. Normally, the sludge should
be wasted to mix with primary and/or waste activated sludge and
be disposed of with them. However, the waste sludge line should
be designed to transport sludge to the nitrification tank when
desired. In the event nitrifying sludge is lost from the nitri-
fying system, it is normally captured by the denitrifying system.
It can be returned to its normal home, at least in part, by using
denitrifying sludge to reseed the nitrification tank.
Quantity of Waste Sludge
It is reported that about 0.2 pounds of sludge will be
generated for each pound of methanol fed. This would correspond
to about 0.7 lb/lb of nitrate nitrogen reduced.
Effluent Quality
Based upon pilot plant studies operating under steady
state conditions the following effluent quality is predicted
from a nitrification-denitrification system.
METCALF a EDDY
-------
4-8
mg/L
Suspended Solids 20
BOD5 10
Organic-N 1.5
NH3-N 0.5
N03-N l.Q
Total -N 3.0
Thus, it appears that 90 percent removals of nitrogen
may be difficult to achieve in actual practice.
Reference
1. Mulbarger, M.C., Private Communication.
METCALF a EDDY
-------
0.7
o
UJ
O 0.6
UJ
o:
o 0.5
or
a
UJ
N
g
CO
.a
i
UJ
1
<
o
0.4
0.3
0.2
5 O.I
l-
ui
o
pH 7.2-7.8
REF. MULBARGER
5 10 15
TEMPERATURE, °C
20
FIG. 7-1 EFFECT OF TEMPERATURE UPON
RATE OF DENITRIFICATION
METCALF a EDDY
-------
w
O
O
O
<
O
180-,
160-
140-
120-
100-
80-
m
o
3 60
40-
20-
BASED UPON DENITRIFICATION RATES
OBSERVED AT MANASSA, VA.
(REF: MULBARGER)
3000 mg/L MLVSS
2500 mg/L MLVSS
2000 mg/L MLVSS
1500 mg/L MLVSS
1000 mg/L MLVSS
1
5
I
10
I
15
I
20
i
25
TEMPERATURE °C
PERMISSIBLE DENITRIFICATION TANK LOADINGS
FIG. 7-1A
METCALF ft CODY
-------
CH3 OH (METHYL ALCOHOL)
NITRATE
ANALYZER
NITRIFIED WASTEWATER
WASTE SLUDGE
$ *
I—4
1
AIR
t;
(s~^\~
-( CLAR. V^-
I RETURN SLUDGE
MODEL SYSTEM FOR FEEDING METHYL ALCOHOL
TO DENITRIFICATION TANK
FIG. 7-2
------- |