Environmental Protection Technology Series
CARBON, NITROGEN, AND
PHOSPHORUS REMOVAL IN STAGED
NITRIFICATION-DENITRIFICATION TREATMENT
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-670/2-75-052
June 1975
CARBON, NITROGEN, AND PHOSPHORUS REMOVAL IN STAGED
NITRIFICATION-DENITRIFICATION TREATMENT
By
James A. Haidman, Dolloff F. Bishop,
and
John B. Stamberg
EPA-DC Pilot Plant
Washington, D.C. 20032
Program Element No. 1BB043
Project Officer
Dolloff F. Bishop
Advanced Waste Treatment Research Laboratory
National Environmental Research Center
Cincinnati, Ohio 45268
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
REVIEW NOTICE
The National Environmental Research. Center—Cincinnati has reviewed
this report and approved its publication. Approval does not signify that the
contents necessarily reflect the view and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
11
-------
FOREWORD
Man and his environment must be protected from the adverse effects
of pesticides, radiation, noise and other forms of pollution, and the
unwise management of solid waste. Efforts to protect the environment
require a focus that recognizes the interplay between the components
of our physical environment—air, water, and land. The National Environ-
mental Research Centers provide this multidisplinary focus through
programs engaged in
o studies on the effects of environmental contaminants on man
and the biosphere, and
o a search for ways to prevent contamination and to recycle
valuable resources.
This work describes an advanced biological treatment system to remove
organic, nitrogen and phosphorus pollutants from the aqueous environment.
The complex approach provides the maximum possible biological treatment
reliability for treatment of municipal wastewaters and may be used in
future wastewater reuse systems for recycle of our water resources.
A. W. Breidenbach, Ph.D.
Director
National Environmental
Research Center, Cincinnati
111
-------
ABSTRACT
A three-stage activated sludge system with mineral addition for nutrient
removal was operated with District of Columbia primary effluent. Influent
flow followed a programmed diurnal cycle and averaged 205 nrVday (54,000
gpd). The first biological reactor was operated as a modified aeration
system with ferric chloride addition for supplemental phosphorus removal.
The clarified effluent then flowed to the second reactor for the biological
nitrification of ammonia and organic nitrogen. Dry lime was used for pH
control. Methanol was added to the nitrified effluent, and biological
denitrification occurred in the final activated sludge system. Prior to
clarification, the denitrification effluent was briefly aerated for nitrogen
gas removal and for consumption of any excess methanol. The clarified
effluent was then split into two equal streams for comparison of filtration
performance of a dual-media coal and sand filter with that of a multi-media
coal, sand, and ilmenite filter. Effluent quality consistently met the
proposed D.C. discharge standards of BOD < 4.5 mg/1; total N 5 2.5 mg/1;
and P < 0.22 mg/1.
This report was submitted in partial fulfillment of Contract No, 68-01-0162
by the Department of Environmental Services, Government of the District of
Columbia, under the sponsorship of the Environmental Protection Agency.
Work was completed as of September 1973.
IV
-------
CONTENTS
Page
Abstract iv
List of Figures vi
List of Tables vii
Acknowledgement viii
Sections
I Conclusions 1
II Recommendations 3
III Introduction 4
IV Experimental Plan 5
V Methods and Procedures 10
VI Results and Discussions 12
VII References 36
VIII Publications 37
-------
FIGURES
No. Page
1 Flow Diagram for Three-Stage Activated Sludge 6
Treatment
2 Diurnal Flow Pattern 7
3 Diurnal Flow Pattern and Relative Recycle on Chemi- 8
cal Feedrate
4 BOD Removal and MLVSS in the Modified Aeration 15
System
5 Phosphorus and Nitrogen Removals in the Modified 17
Aeration System
6 Selected Process Parameters for the Nitrification 20
System
7 Changes in Nitrogen Concentrations in the 23
Nitrification Process
8 Summary of Nitrification Kinetic Data from 24
January 1972 to September 1973
9 Changes in Nitrogen Concentrations in the 27
Denitrification Process
10 Selected Process Parameters for the Denitrification 29
System
11 Summary of Denitrification Kinetic Data from 34
January 1972 to September 1973
12 Changes in BOD, Nitrogen and Phosphorus Resulting 35
from Filtration
VI
-------
TABLES
No. Page
1 Average Monthly Characteristics of District of 13
Columbia Primary Effluent
2 Primary Clarifier Loadings and Average Process Flow 14
3 Ferric Chloride Dosages and Phosphorus Removal for 16
Modified Aeration
4 Sludge Volume Index and Settling Characteristics 18
of the Modified Aeration Activated Sludge
5 Average Monthly Characteristics of Modified Aeration 19
Clarified Effluent
6 Average Monthly Characteristics of Nitrification 22
Clarified Effluent
7 Sludge Volume Index and Settling Characteristics 26
of the Nitrification Activated Sludge
8 Average Monthly Characteristics of Denitrification 28
Clarified Effluent
9 Methanol Dosages and Nitrate Removal for 30
Denitrification
10 Alum Dosages and Phosphorus Removal for 31
Denitrification
11 Sludge Volume Index and Settling Characteristics 33
of the Denitrification Activated Sludge
VII
-------
ACKNOWLEDGMENT
The assistance of the District of Columbia operators, technicians and
laboratory staff at the EPA-DC Pilot Plant is gratefully acknowledged.
viii
-------
SECTION I
CONCLUSIONS
The three-stage activated sludge system with final filtration is entirely
satisfactory for meeting the proposed District of Columbia discharge '
standards. This system was the one selected for the 812 m-^/min (309 mgd)
Blue Plains Sewage Treatment Plant.
Specifically, the modified aeration operated at an SRT of approximately
1 day with mineral addition exhibited excellent stability and produced a
satisfactory effluent for the subsequent processes in the three stage
system. With FeCl-j dosage equal to a 1:1 mole ratio Fe/P, modified aeration
removed approximately 81% of the BODij, 72% of the phosphorus, and about
31% of the total nitrogen.
The subsequent nitrification process with the pH controlled to 7.0-7.2 by
an average addition of 60 mg/1 of dry CaO produced essentially complete
nitrification (average residual TKN of 1.2 mg/1, excluding an upset from
mechanical failure) and produced essentially complete removal of carbona-
ceous BODr (nitrifier inhibited BODt- residual of approximately 3 mg/1).
Batch nitrification kinetics on the process mixed liquor provided a strong
correlation (correlation coefficient r = 0.837) between the nitrification
rate constant-and temperature. The relationship is:
where KJJH N = gm NH3~N/day/gm MLVSS
t = Temperature, C
The denitrification process with methanol addition and with alum addition
removed an average of 94% of the nitrate nitrogen with an annual average
of 0.72 mg/1 of residual NO^-N. A dosage of four units of methanol (by
weight) per unit of M^-N produced essentially complete denitrification.
Batch denitrification kinetics tests on the process mixed liquor produced
a moderately weak correlation (r = 0.629) between the denitrification rate
constant and temperature. The relationship is:
^N = 0.0212 t - 0.1657
where ^0_^ = gm NO-j-N/day/gm MLVSS
t = temperature, C
The alum addition in the denitrification process at an A1:P mole dosage
ratio between 3:1 and 5:1 reduced the_influent phosphorus by about 40%
from about 3.4 mg/1 to 2.1 mg/1 as P07. The real impact of the alum
addition in denitrification was to insure good phosphorus and solids
-------
removal by the final filtration process.
Dual- or multi-media filtration of the denitrified effluent produced a
final effluent that consistently exceeded the discharge standards for
the proposed new plant in Washington, D.C. The residual BOD5 averaged
2 mg/1; the total nitrogen, 1.6 mg/1; and the total phosphorus, 0.52 mg/1
as P0|.
-------
SECTION II
RECOMMENDATIONS
Considering the length of time over which the three-stage activated sludge
process was evaluated, there is little doubt that it will produce a high
quality effluent throughout the year on District of Columbia wastewater.
Therefore, further work is not needed to establish the basic reliability
of the process.
There are, however, several areas that need to be explored so that the
process operation can be optimized. The minimum methanol dosage needs
to be ascertained more carefully. The possibility of varying the ferric
chloride and alum feed in direct response to the incoming phosphate also
requires study. Finally, it may not be necessary to add lime to the
nitrification process, especially in the summer, and this could result in
additional cost savings. Once the main District of Columbia treatment
plant is constructed, all of these areas could be fully evaluated under
full-scale operating conditions.
-------
SECTION III
INTRODUCTION
The three-stage activated sludge system was examined on a pilot-plant scale
as one of several process alternatives that could potentially meet the
proposed discharge standards for the District of Columbia Blue Plains
Wastewater Treatment Plant. These standards, which apply year-around, call
for a maximum effluent BOD of 4.5 mg/1; total nitrogen not to exceed 2.5
mg/1; and total phosphate of less than 0.67 mg/1 (0.22 mg/1 of P). The
three-stage system was put on stream in 1970 and was operated through
September of 1973. This report summarizes the results obtained from the
last year of operation, i.e., October 1972 through September 1973.
The three-stage biological system-'- consisted of modified aeration with
mineral (FeC^) addition for removal of organic carbon and phosphorus;
nitrification with lime addition for oxidation of ammonia to nitrate under
controlled pH; denitrification with methanol as an external carbon source
for removal of nitrate and with alum addition for residual phosphorus
removal; and finally, filtration for removal of residual solids (C, P and N)
-------
SECTION IV
EXPERIMENTAL PLAN
A schematic diagram of the three-stage activated sludge process is presented
in Figure 1. The influent process flow, consisting of a small portion of
the primary effluent from the District of Columbia Blue Plains Wastewater
Treatment Plant operating at 762 nrVmin (290 mgd), was pumped to the modi-
fied aeration reactor on the diurnal flow pattern shown in Figure 2.
Average process flow was approximately 205 m^/day (54,000 gpd). The modi-
fied aeration reactor consisted of three completely mixed passes of equal
size in series with an effective total volume of 22.0 m^ (5,820 gal). The
reactor provided a detention time of 2.6 hours at average flow. Compressed
air was supplied through perforated PVC pipe diffusers and the dissolved
oxygen levels in each stage were maintained between 0.5 and 4.0 mg/1.
Ferric chloride was added to the third pass of the reactor. The chemical
dosage rate was manually changed three times per day to correspond to the
diurnal flow pattern.
The modified aeration reactor effluent discharged to a circular peripheral
feed clarifier with an effective surface area of 8.9 m^ (96 ft^). The area
provided an average overflow rate of 21.2 in/day (520 gpd/ft^). Recycle
solids were returned at a reasonably constant percentage of influent flow
with manual adjustment of the recycle pumping rate at the times indicated
in Figure 2. A typical relationship between process flow and recycle flow
is shown in Figure 3. Except for brief periods, the three time per day
manual adjustment of the recycle flow provided a reasonably constant rate
between influent and recycle flow. The various chemical feed rates (ferric
chloride, methanol, and alum) were also manually changed three times per
day. The relative relationship between process flow and chemical feed
rate is also presented in Figure 3.
The effluent from the modified aeration clarifier was pumped to the
second biological system for nitrification. The nitrification reactor
consisted of four complete mix passes operated in series. Total effective
volume was 29.3 m (7,740 gal), which provided for an average detention
time of 3.4 hours. Air was supplied independently to each pass through
perforated PVC pipe diffusers and the D.O. was maintained between 0.5 and
4.0 mg/1. A dry lime feeder was located above the first pass and lime was
automatically fed to maintain the desired effluent pH.
The nitrified effluent flowed to a circular center-feed clarifier with a
n n
surface area of 8.9 m (96 ft ). Recycle solids were returned from the
clarifier to the reactor at a constant rate that was not varied in response
to the diurnal flow pattern. However, the recycle rate was changed from
time to time as operating conditions warranted.
Methanol was added to the effluent from the nitrification clarifier and
process flow was sent to the denitrification reactor. The denitrification
reactor consisted of four, covered, mechanically stirred tanks of equal
size in series. The tanks were covered to exclude oxygen transfer from
-------
RAW
u r i u
FERRIC CHLORIDE
/
f
t
T
*
Tl 1
r
==5:
MODIFIED AERATION
ALUM METHANOL
LIME
I.
AIR
DENITRIFICATION
AIR
1—
W
NITRIFICATION
T V
FILTRATION
Figure 1. Flow diagram for three-stage activated
sludge treatment
-------
200-i
160
.E 120-
E
0)
O
80-
40-
RECYCLE
CHANGED
RECYCLE
CHANGED/
RECYCLE
CHANGED,
0000 0400
Figure 2. Diurnal flow pattern
0800 NOON
TIME
1600
2000 2400
-------
00
160-
PROCESS FLOW
RECYCLE OR CHEMICAL FEED
0000
+
0400
+
+
0800
1200
TIME
1600
r50
-40
-30
-20
-10
2000
2400
Figure 3. Diurnal flow pattern and relative recycle on chemical feedrate
0
fe?
to
<
Q
Ul
LU
U
X
u
a
u
u
LU
-------
the air. The total effective volume of the reactor was 27.3 m-^ (7,210 gal),
with a corresponding detention time of approximately 3.2 hours.
Denitrified effluent flowed directly into a single, aerated, completely
mixed chamber with an effective volume of 4.35 m^ (1,150 gal). The purpose
of the chamber was to strip nitrogen gas from the water and to oxidize any
excess methanol that was not consumed in the denitrification reactor. Alum
was also added to the chamber at the particular dosage rate desired.
Following aeration, the denitrified effluent flowed to a circular center-
feed clarifier with a 5.85 m^ (63 ft^) surface area. The corresponding
average overflow rate was 32.2 m/day (790 gpd/ft^). Recycle solids were
returned from the clarifier to the denitrification reactor at a constant
rate of flow.
Effluent from the denitrification clarifier flowed to a splitter box, where
it was equally divided before flowing to a dual-media and multi-media
filter. The dual-media filter consisted of 0.30 m (12 inches) of sand with
an effective size of 0.6-0.7 mm overlain by 0.61 m (24 inches) of coal with
an effective size of 1.2-1.4 mm. The multi-media filter consisted of 0.08 m
(3 inches) of ilmenite with an effective size of 0.2-0.35 mm; overlain by
0.23 m (9 inches) of sand of effective size 0.4-0.5 mm; overlain by 0.20 m
(8 inches) of coaj. of effective size 1.0-1.1 mm; overlain by 0.41 m
(16 inches) of coal with an effective size of 1.5-1.6 mm. The uniformity
coefficient of all materials in the multi-media filter was 1.8. Filter
loading varied from a low of 111 ra/hr (1.9 gpm/ft^) at low flow to a high
of 235 m/hr (4.0 gpm/ft^) at high flow. The average loading was 176 m/hr
(3 gpm/ft^). Differential pressure readings were taken at various bed
depths, and the filters were backwashed either when the total pressure
drop reached 3.0 m (120 inches) or after 24 hours even though the 3.0 m
(120 inch) pressure drop had not yet been attained.
-------
SECTION V
METHODS AND PROCEDURES
The process was operated on a 24-hour a day, 7-day a week schedule for the
entire year of operation covered in this report. The only interruptions
in the normal operating sequence resulted from mechanical malfunctions
and these were of short duration.
Grab samples of influent, effluent, mixed liquor, etc., were taken every
4 hours. The samples collected for laboratory analysis were composited
over a 24-hour period on Tuesday, Wednesday, and Thursday; samples collected
on Friday-Saturday and on Sunday-Monday were composited over the 48-hour
period. The single exception to this was that the samples for BOD analysis
were just 24-hour composites and the analysis was always started within a
few hours (4-10 hours) after the last sample had been collected for the 24-
hour composite. All samples were refrigerated at 2°C. In addition, all
samples except those taken for BOD or suspended solids analysis were pre-
served with one drop of I^SO^ per 30 ml of sample while they were being
held in storage. All laboratory analyses (except BOD) were performed on a
Monday through Friday schedule.
The following analyses were performed in the EPA-DC Pilot Plant laboratories
according to the procedures specified in Standard Methods^: suspended
solids, volatile suspended solids, BOD, COD, and TKN. BOD analyses in which
N03 production was inhibited by the addition of o.5 mg/1 of l-allyl-2-
thiourea were also performed. The procedures specified in the EPA Manual^
were used for the following: TOC with a Beckman analyzer; total solids; and
NH3, N03, N02 with a Technicon autoanalyzer. The method of Gales et al.,^
was used for the determination of total phosphorus.
In addition to collecting samples for laboratory analysis every four hours,
the operating personnel checked the dissolved oxygen levels in the appro-
priate reactors with a portable field probe and adjusted the air flow rates
as necessary; obtained solids samples for 30-minute sludge volume deter-
minations in one-liter cylinders; measured temperature, pH and alkalinity
of selected samples; measured the depth of the sludge blankets in the
three clarifiers; measured and adjusted chemical feed rates as needed; and
obtained differential pressure readings on the parallel filters and
backwashed them when required.
Sludge wasting on the modified aeration system was accomplished automatic-
ally by diverting the recycle flow to a drum with a level control probe; a
timer was used to control the frequency of diversion to the drum; and the
level control probe switched the recycle flow back to the process after
74.9 liters (19.8 gal) had been added to the drum. Sludge wasting on the
other two systems was done manually. Whenever the waste rate was 0.19 m^/day
(50 gpd) or less, the wasting was done once per day. For waste rates in
excess of 0.19 m-Vday (50 gpd), wasting was done twice per day in roughly
equal amounts.
10
-------
Throughout the year, samples of mixed liquor were removed periodically
for settling tests in 2.3 m x 0.15 m (7.5 ft x 6 inches) diameter stirred
columns. The stirring mechanism consisted of two 0.64 cm (1/4 inch)
diameter rods that extended the length of the column and rotated around
the vertical axis at a rate of 10-14 rph.
On numerous occasions, batch kinetic studies were undertaken in the
laboratory to establish the process kinetic rates for nitrification and
denitrification. The nitrification studies were performed by mixing a
sample (1-2 liters) of the nitrification recycle solids with the effluent
from the modified aeration clarifier in a ratio similar to that existing
in the process at the time. A water bath was employed to insure that the
temperature of the kinetic analysis remained the same as that which existed
in the process. The mixture was aerated and the decrease in NH3 was moni-
tored by Technicon analyses. In all cases, the NH^ removal followed zero
order kinetics and the rate of removal was determined per unit of mixed-
liquor volatile suspended-solids. The change in mixed-liquor solids
concentration during the course of the kinetic study was insignificant.
The denitrification kinetic studies were performed by mixing 1-2 liters
of denitrification recycle solids with a sample of the nitrification
effluent in a ratio similar to that which existed in the process. Methanol
was also added. A flexible, plastic screw top container was used to hold
the mixture and the container was "squeezed" to exclude all air prior to
putting on the plastic cap. A large magnet and magnetic stirrer was used
to keep the contents thoroughly mixed and a sample was continuously
withdrawn for (N02 + NC^-N analysis by a Technicon autoanalyzer. A
water bath was employed to insure that the temperature remained the same
as that which prevailed in the actual denitrification process at the time.
The decrease in (N0£ + N03)-N followed zero order kinetics in all cases
and the rate was expressed per unit of volatile suspended solids.
11
-------
SECTION VI
RESULTS AND DISCUSSIONS
The average monthly characteristics of the District of Columbia primary
effluent are summarized in Table 1. The wastewater is largely of domestic
origin with very little industrial discharge. The presence of a combined
sewer system plays a significant part in the monthly variation indicated.
In spite of the overloaded primary clarifiers (Table 2), the primary
wastewater is relatively weak with an average BOD of 103 mg/1 and a COD of
236 mg/1. The TKN averaged 23.7 mg/1 and 15.6 mg/1 of this was present
as NH3-N. The influent (N02 + N03)-N content was negligible. Influent P04
averaged 20.8 mg/1.
As indicated in Figure 4, the modified aeration reactor solids were varied
throughout the year to produce the desired effluent quality. The large
variation in necessary reactor solids results from winter wastewater
temperatures of about 15°C versus summer wastewater temperatures of 26°C.
To insure the development of a sufficient quantity of heterotrophic orga-
nisms for good floe formation and settleability in the subsequent
nitrification process, the residual effluent BOD was maintained within
the range indicated. (The relatively low effluent BOD residual entering
nitrification during the last four months of the study, however, did not
adversely affect the settling performance of the nitrification system.)
The volatile solids concentration ranged between 61-66% of the MLSS. This
low volatile solids content reflects the inert solids buildup resulting
from ferric chloride addition. As indicated in Table 3, the ferric
chloride dosage was uniform throughout most of the project period. The
average dosage was 36 mg/1, which produced a ferric to influent phosphorus
mole ratio of 1:1. The combination of chemical precipitation and biological
uptake resulted in an overall average phosphorus removal of 72.3%.
Although a considerable portion of the influent carbonaceous material and
phosphate was removed in the modified aeration process, the average
reduction in total nitrogen was only 7.4 mg/1 (Figure 5). Of this, organic
nitrogen removal accounted for the major decrease in TKN. Since the process
was operated at a sludge retention time (SRT) that varied from 0.65 days to
1.3 days during the year's operation, there was no opportunity for a
nitrifying population to develop. Consequently, the effluent (N03 4- N02>-N
concentration was negligible throughout the year.
The results presented in Table 4 summarize the settling rates obtained
from the 0.15 m (6 inch) column studies. The combination of relatively
low reactor solids and the addition of ferric chloride produced a sludge
with excellent settling characteristics. The monthly variation in effluent
suspended solids is summarized in Table 5.
The effluent BOD and phosphate residuals from the nitrification system and
the variations in mixed liquor volatile solids throughout the year's
operation are presented in Figure 6. The large decrease in reactor solids
12
-------
Table 1. AVERAGE MONTHLY CHARACTERISTICS OF DISTRICT OF COLUMBIA
PRIMARY EFFLUENT
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
TOC
73
61
61
69
69
79
68
79
77
66
63
__
COD
251
239
238
246
234
263
230
234
238
221
207
225
BOD
115
106
100
97
99
110
99
115
111
96
88
99
TKN
26.2
24.5
26.8
25.7
23.5
24.8
21.7
23.3
23.0
21.6
20.7
22.6
NH3-N
17.8
16.9
16.4
16.7
15.5
14.1
13.3
14.9
14.7
15.5
15.0
16.4
P04
23.5
21.9
20.7
22.2
21.8
21.7
19.1
21.2
21.8
18.2
16.9
21.1
SS
95
112
115
106
108
128
120
109
111
101
102
110
VSS
70
81
78
77
82
97
85
80
83
75
79
85
All concentrations in mg/1.
-------
Table 2. PRIMARY CLARIFIER LOADINGS
AND AVERAGE PROCESS FLOW
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
Average Loading On
D.C. Primary Clarifier
m/day
83
90
87
88
88
91
88
87
88
88
Average Process
Flow
m3/day
175
240*
210
205
200
200
200
205
200
215
210
215
* Includes 4 days of 380 m3/day operation to simulate rain peaks.
14
-------
120
no
100
90
- 30
20
O)
E
in
10
Q
LLJ
< 1200
u
Q
? 1100
1000
900
600
u
O
INFLUENT BOD
EFFLUENT BOD
>
O
u
LU
Q
Z CO
•
<
O
Q.
LU
to
Figure 4. BOD removal and MLVSS in the modified aeration
system
15
-------
Table 3. FERRIC CHLORIDE DOSAGES AND PHOSPHORUS REMOVAL
FOR MODIFIED AERATION
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
FeCl3 Dose
mg/1
47
37
35
37
36
35
34
37
35
32
35
35
P04> Inf.
mg/1
23.5
21.9
20.7
22.2
21.8
21.7
19.1
21.2
21.8
18.2
16.9
21.1
P04> Eff.
mg/1
5.6
5.0
6.2
6.3
7.1
7.4
6.5
6.2
4.3
5.0
4.9
4.8
P04
% Removal
76.2
77.2
70.0
71.6
67.4
65.9
66.0
70.8
80.3
72.5
71.0
77.3
Fe:P
Mole Ratio
1.17:1
0.99:1
0.99:1
0.98:1
0.97:1
0.94:1
1.04:1
1.02:1
. 0.94:1
1,03:1
1.21:1
0.97:1
-------
35
30
15
10
o Influent PC>4
Effluent PO4
^m
J
"m— ""
o Influent TKN
•Effluent TKN
> u Z ca **•
o 3 < s <
a Influent NH3
» Effluent NH3_N
1 1 | i |
— •
a
u
u
Figure 5. Phosphorus and nitrogen removals in the
modified aeration system
17
-------
Table 4. SLUDGE VOLUME INDEX AND SETTLING
CHARACTERISTICS OF THE MODIFIED
AERATION ACTIVATED SLUDGE
Process
Settling Test Results
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
SVI
ml/gm
89
80
67
69
73
71
68
80
103
96
94
106
m/hr
8.4
6.6
7.0
6.8
10.7
8.0
7.3
12.6
7.4
7.4
13.7
12.0
10.7
16.6
10.2
9.8
7.3
13.3
6.4
9.3
4.2
8.7
11.4
4.3
5.9
11.9
5.5
6.1
°C
25
22.2
22.2
21.7
17.8
16,7
16.7
16.4
13.5
14.0
15
15
14.5
15.5
16
18
18
21
23.5
19.5
25.5
24.5
24.7
25.5
27
23.5
22.5
24
MLSS
rng/1
1300
1250
1200
1500
1850
1200
1350
1050
1850
2100
1100
1800
2000
1650
1600
1350
1650
1800
1800
1700
1850
1150
1300
1300
1450
1500
1450
1550
18
-------
Table 5. AVERAGE MONTHLY CHARACTERISTICS OF MODIFIED AERATION
CLARIFIED EFFLUENT
TOC
COD
BOD
TKN
NH3-N
P04
SS
VSS
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
28
26
29
26
28
28
24
23
19
19
19
1
61.5
75.7
84.7
74.5
70.6
78.3
65.3
55.9
42.3
51.3
46.7
43.9
22.3
22.5
25.8
28.0
26.8
27.6
21.9
18.8
11.7
18.1
13.9
13.2
17.7
17.4
18.2
18.3
16.2
17.1
15.5
15.8
14.9
15.4
13.9
14.9
14.8
15.3
14.1
14.4
13.5
12.7
12.3
13.6
13.6
13.3
13.6
13.8
5.58
5.04
6.19
6.25
7.10
7.44
6.50
6.23
4.30
5.00
4.88
4.76
21
32
33
33
31
38
31
22
13
20
17
18
15
22
23
20
23
27
20
16
10
14
13
13
All concentrations in mg/1.
-------
44
2
o
LU
u
Z 2200-
2000-
1800-
1600-
1400
1200H
1000
EFFLUENT PC>4
EFFLUENT BOD
EFFLUENT BOD (NOs INHIBITED)
MLVSS
o i
u Z
LU <
Q ^
tfl «
ffi <
? <
z
Z)
>- 1*1 I—
^ °"
^ < to
Figure 6. Selected process parameters for the nitrification system
20
-------
indicated for February and March was the result of a pump failure in
February and the concomitant loss of a substantial portion of the solids
stored in the clarifier. This loss resulted in the lower-than-desired
solids levels for winter operation. The mixed-liquor volatile suspended-
solids varied from a low of 48% to a high of 64% of the mixed liquor
suspended solids.
As previously indicated, dry lime (CaO) was added to the first pass of
the nitrification reactor in sufficient quantity to provide an effluent
pH of 7.0-7.2. Since the D.C. wastewater is of low alkalinity (^ 100 mg/1
as CaCC>3 after modified aeration), the natural buffer capacity is not
sufficient to prevent pH depression resulting from the nitric acid produced
during nitrification. The average lime dose for the year was 60 mg/1.
Inert material present in the lime was responsible for the relatively low
percentage of mixed-liquor volatile suspended-solids observed. No attempt
was made to evaluate the nitrification process under natural pH conditions.
However, it is quite likely that one could reduce the lime dosage and
operate at somewhat lower pH values without a significant decrease in
nitrification kinetics.
The inhibited BOD values indicated in Figure 6 represent the average of a
maximum of eight samples for any given month. Whereas the inhibited BOD
analysis was performed approximately twice per week, the normal effluent
BOD values represent the average of daily analyses taken throughout the
month. This difference in the data base was responsible for much of the
variation in the relative differences between the inhibited and normal
BOD values. The nitrified effluent was essentially free of carbonaceous
BOD. The effluent phosphate concentration was very steady throughout
the year and averaged 3.40 mg/1. Other effluent parameters are summarized
in Table 6.
The large solids loss in February was responsible for the relatively high
TKN and NH3~N levels in the nitrification effluent during part of February
and March. Except for this temporary effluent deterioration resulting from
mechanical failure, the nitrification process functioned very well through-
out the year (Figure 7). Excluding the February and March data, the
average effluent TKN and NI^-N values were 1.21 mg/1 and 0.63 mg/1,
respectively. For the entire year's operation, the effluent TKN averaged
1.72 mg/1; the NH3-N, 1.15 mg/1.
A summary of the nitrification kinetic rates determined from the laboratory
batch studies is presented in Figure 8. The results include all kinetic
constants obtained from the system and cover nearly two years of operation.
The kinetic rate was significantly influenced by process temperature. In
fact, treating the kinetic rate as a function of just the temperature
produces a linear correlation coefficient of 0.837. Other factors, such
as variation in the BOD to TKN ratio and operation at different SRT's also
influenced the kinetic rate and account for some of the variation shown by
examining the rate as a simple function of temperature. Nonetheless, the
temperature dependency is quite obvious.
21
-------
Table 6. AVERAGE MONTHLY CHARACTERISTICS OF NITRIFICATION
CLARIFIED EFFLUENT
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
TOC
7.1
9.1
10.5
8.6
11.9
9.0
6.5
8.2
7.5
6.9
6.9
___
COD
20.0
29.6
26.4
23.4
23.8
24.4
19.1
16.9
17.6
17.2
16.4
17.3
BOD
11.3
H.8
13.3
16.6
17.4
12.3
9.2
10.2
10.9
12.3
9.7
10.1
TKN
0.69
2.80
1.82
1.56
4.74
3.75
1.05
0.68
0.76
0.91
0.70
1.14
NH3-N
0.38
1.42
1.01
0.65
4.51
2.98
0.66
0.24
0.16
0.77
0.32
0.64
N03-N
14.4
11.8
10.4
13.1
9.0
10.6
12.4
12.7
13.4
12.5
12.1
12.2
P04
3.50
2.28
3.24
3.90
3.38
3.79
3.40
3.83
3.42
3,43
3.41
3.24
SS
5
9
11
12
9
8
7
7
8
7
7
8
VSS
4
5
6
6
5
5
4
4
5
4
5
5
All concentrations in mg/1.
-------
12-
8-
4-
16-
12-
10-
o Influent TKN
n Influent NH3-N
n Effluent (NOs)-N
o Effluent TKN
A Effluent NH 3-N
to
Figure 7. Changes in nitrogen concentrations in the nitrification
process
23
-------
KNH_= 0.0171 t°
15 16 17 18 19 20 21 22 23 24 25 26 27
TEMPERATURE (°C)
Figure 8. Summary of nitrification kinetic data from January 1972
to September 1973
-------
The results of the 0.15 m (6 inch) column settling studies are summarized in
Table 7. During winter operation, the process SVI was higher and the
settling velocities lower than during summertime operation. This is just
the reverse of the modified aeration process, where the increased filamentous
growth during the summer produced somewhat higher SVl's and lower settling
velocities than in the winter months.
The nitrogen transformations occurring in the denitrification activated
sludge are indicated in Figure 9. There was a very small decline in the
NH^-N concentration. A slight increase in effluent TKN resulting from a
greater solids concentration in the effluent than in the influent (Table 8)
also occurred. For all practical purposes, however, there was no signifi-
cant change in the TKN or NH3 levels in the effluent when compared with the
denitrification influent. Denitrification reduced the influent NOg-N (NC>2
analyses were performed three times per week but no more than trace
quantities were ever detected) from an average of 12.1 mg/1 to 0.72 mg/1
and effected an overall average removal of 94%. To obtain these nitrogen
removals, the denitrification process was operated at MLVSS levels that
varied between 1350 and 2400 mg/1 (Figure 10). The volatile suspended
solids comprised between 53 to 60% of the mixed liquor solids. These
concentrations resulted from process SRT's that varied from 10 to 19 days.
In order to obtain denitrification, it was necessary to add a degradable
carbon source to support heterotrophic growth. Once the dissolved oxygen
present in the influent wastewater was removed, the oxygen bound in the
NOo molecule served as the terminal hydrogen acceptor and the nitrate
nitrogen was converted to nitrogen gas.
The average monthly methanol dosages, NO? removals, and methanol dose
per unit of NO^-N removed are summarized in Table 9. No sustained attempt
was made to reduce the methanol dosage to the minimum amount needed. Rather,
the difficulty of accurately controlling the feed rate on a small scale
and other problems resulted in a tendency to feed more methanol than
actually needed. For this reason the methanol dosages for some months
were higher than actually required. In spite of this, there were several
occasions when excess nitrate was present in the effluent because of short
periods of insufficient methanol. A dose of four units of methanol (by
weight) per unit of NO-j-N removed seemed to be the upper limit of the
amount actually required.
The effluent PO^ concentration and effluent BOD's are also presented in
Figure 10. As was the case for the nitrification process, the inhibited
BOD values are based on about 25% of the number of samples comprising the
uninhibited BOD average. This difference in the data base accounts for
some of the deviation shown. Since the denitrification process included
a terminal aerated chamber, the difference between influent and effluent
BOD is not an indication of possible methanol overdosing.
The influent and effluent phosphate concentration and the average monthly
alum dosages are summarized in Table 10. The A1:P mole ratio varied
between 3:1 and 5:1 with no direct correlation between alum dosage and
25
-------
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
Table 7 . SLUDGE VOLUME INDEX AND SETTLING
CHARACTERISTICS OF THE NITRIFICATION
ACTIVATED SLUDGE
Process Settling Test Results
SVI
ml/gm
104
124
150
182
165
151
189
138
81
84
76
76
m/hr
5.6
5.0
4.7
4.3
2.1
1.7
1.7
2.7
2.4
2.6
2.5
5.3
3.3
3.8
4.8
5.0
2.1
1.8
2.0
2.2
2.8
8.6
4.8
4.7
4.2
5.7
4.8
5.2
4.8
5.7
5.7
4.7
°C
21.7
22.2
21.7
21.4
17.8
17.8
17.5
16.1
17.2
16.7
14.0
14.0
14.5
17.5
15.5
15.3
15.5
15.5
17.5
17.0
19.0
19.5
22.0
23.5
23.5
26.5
25.0
23.5
26.0
26.0
24.0
22.0
MLSS
mg/1
2600
2100
2300
2600
3200
2400
3200
3800
3400
4000
3000
1600
2000
1900
1800
1800
3200
3500
3900
3800
3000
2900
2900
3600
2800
2600
3900
2800
3400
3400
2500
2900
26
-------
a Influent TKN
Effluent TKN
a Influent
A Effluent NH3-N
a Influent NOs —N
Effluent NO3-N
Figure 9. Changes in nitrogen concentrations in the
denitrification process
27
-------
ho
CO
Table 8. AVERAGE MONTHLY CHARACTERISTICS OF DENITRIFICATION
CLARIFIED EFFLUENT
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
/PRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
TOC
9.9
10.3
12.9
10.4
15.9
11.0
8.9
8.8
8.8
8.0
7.8
___
COD
28.4
30.3
37.2
27.7
33.3
27.9
24.2
19.0
19.1
20.4
19.3
21.5
BOD
9,5
8.4
11.0
12.5
13.9
10.1
7.4
7.4
8.5
10.2
6.8
7.6
TKN
1.79
1.66
2.50
2.13
5.19
3.69
1.62
1.03
1.11
1.38
1.10
1.53
NH3-N
0.41
0.53
0,62
0.38
3.80
2.33
0.26
0.08
0.17
0.39
0.35
0.48
N03-N
0.77
0.80
0.11
0.98
0.25
0.61
0.61
1.17
0.88
1.38
0.30
0.77
P04
2.43
1.10
1.85
2.10
2.54
2.48
2.26
2.27
1.98
2.22
1.82
2.25
SS
17
18
21
19
25
19
21
16
•14
19
16
18
VSS
9
10
13
9
15
11
10
9
7
9
10
11
All concentrations in mg/1.
-------
O)
E
20-
15-
(/)
5 10
5-
<
Q
14004
EFFLUENT PO4
EFFLUENT BOD
EFFLUENT
^^—^X
BOD (NO3 INHIBITED)
1200
o I
Figure 10
UJ
Q
oa
UJ
Of
a.
>-
<
§
Q.
UJ
in
Selected process parameters for the
denitrification system
29
-------
Table 9. METHANOL DOSAGES AND NITRATE REMOVAL
FOR DENITRIFICATION
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
Methanol
Dose
mg/1
67
55
49
46
48
49
46
47
46
46
56
54
N03-N, Inf.
mg/1
14.4
11.8
10.4
13.1
9.0
10.6
12.4
12.7
13.4
12.5
12.1
12.2
N03-N, Eff.
mg/1
0.8
0.8
0.1
1.0
0.3
0.6
0.6
1.2
0.9
1.4
0.3
0.8
N03-N
Removed
mg/1
13.6
11.0
10.3
12.1
8.7
10.0
11.8
11.5
12.5
11.1
11.8
11.4
Methanol
N03-N Re
4.9
5.0
4.8
3.8
5.5
4.9
3.9
4.1
3.7
4.1
4.7
4.7
-------
Table 10. ALUM DOSAGES AND PHOSPHORUS REMOVAL
FOR DENITRIFICATION
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
Alum Dose
mg/1
41
35
33
35
37
35
37
38
42
39
44
I
47
P04, Inf.
mg/1
3.50
2.28
3.24
3.90
3.38
3.79
3.40
3.83
3.42
3.43
3.41
3.24
P04, Eff.
mg/1
2.43
1.10
1.85
2.10
2.54
2.48
2.26
2.27
1.98
2.22
1.82
2.25
"°4
% Removal
30.6
51.8
42.9
46.2
24.9
38.8
33.5
40.7
42.1
35.3
46.6
30.6
A1:P
Mole Ratio
3.7:1
4.9:1
3.3:1
2.9:1
3.5:1
3.0:1
3.5:1
3.2:1
3.9:1
3.6:1
4.1:1
4.6:1
-------
effluent PO^ concentrations. This lack of correlation largely resulted
from the variations in clarification efficiency during the year. The
real impact of the alum dosage was in insuring good phosphate removal on
the filters receiving the effluent from the denitrification clarifier.
The settling characteristics of the denitrification system were entirely
satisfactory throughout the year (Table 11). The lowest rates obtained
were nearly 3 m/hr.
The denitrification kinetic rate constants determined in laboratory batch
studies are presented in Figure 11. In contrast to the strong correlation
between kinetic rate and temperature in the nitrification kinetic studies,
the denitrification kinetic constants exhibit a relatively weaker correla-
tion when considered as just a function of the single variable temperature.
The data points shown between 14 and 18.5°C and grouped below the regression
line (a total of 11 values) are kinetic rates obtained in earlier operations
during the winter of 1972 when there was considerable methanol overdosing
(2 to 3 times the required dose) and high SRT operation. Both of these
factors contributed to these low kinetic rates.
The changes in BOD, total nitrogen, and PO^ as a result of filtration are
presented in Figure 12. These values represent the average effluent
quality from the dual- and multi-media filters. The filters were not
operated during February. Although the denitrification effluent meets the
proposed discharge standards for total nitrogen, it does not meet the
proposed standards for BOD or PO/^. The addition of the filtration step,
however, produced an overall average BOD of 2.0 mg/1, total nitrogen of
1.6 mg/1, and total phosphate of 0.52 mg/1. The phosphate concentration
could be decreased further by feeding more alum ahead of filtration.
Generally the multi-media filters produced a slightly improved effluent
quality, with typical improvements usually not exceeding 5%. Filter
runs of 24 hours were normally obtained with either media. The filtration
results have been presented in detail elsewhere.5
32
-------
Table 11. SLUDGE VOLUME INDEX AND SETTLING
CHARACTERISTICS OF THE DENITRIFICATION
ACTIVATED SLUDGE
Process Settling Test Results
OCTOBER
NOVEMBER
DECEMBER
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
SVI
rnl/grn
85
82
71
76
72
79
79
81
61
54
55
56
m/hr
4.5
3.7
3.3
4.3
3.9
3.3
3.6
2.7
3.0
3.3
2.9
3.0
3.4
2.9
3.0
4.5
4.4
4.0
5.2
5.3
4.7
6.3
6.5
7.1
7.1
3.7
5.4
4.8
5.7
°C
22.2
23.3
21.4
17.8
16.7
17.2
17.0
13.5
14.5
15
15.5
16
14.5
15.5
16
16
18
20
17
20.5
24.5
23.5
26.5
26
25.5
25.5
27.7
24
24
MLSS
mg/1
2400
2800
1600
3300
3800
3600
3900
4100
3900
3300
4400
4300
4000
4100
4100
2800
2500
2500
2300
2800
3000
3500
3000
2600
2800
3800
3100
3600
4100
33
-------
co
CO
0.6 r
0.5--
E
O)
^0.4
i 0.3
Z n o--
Z
O
u
0.1--
o
O
CD
8
OD
KNO3-N = 0.0212 t°-0.1657
r = 0.629
^—,—i
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
TEMPERATURE (°C)
Figure 11. Summary of electrification kinetic data from
January 1972 to September 1973
-------
20
16
12
8
Influent BOD
Effluent BOD
Influent Total Nitrogen
Effluent Total Nitrogen
Figure 12. Changes in BOD, nitrogen and phosphorus
resulting from filtration.
35
-------
SECTION VII
REFERENCES
1. Earth, E. F., Brenner, R. C., and Lewis, R. F., "Chemical-Biological
Control of Nitrogen and Phosphorus in Wastewater Effluent," Jour.
Water Poll. Control Fed., _44, 2040 (1968).
2. "Standard Methods for the Examination of Water and Wastewater."
13th ed., American Public Health Association, New York (1971).
3. "Methods for Chemical Analysis of Water and Waste." Report No.
16020-07/71, U.S. EPA, Cincinnati, Ohio (1971).
4. Gales, M., Julian, E. C., and Kroner, R. C., "Method for Quantitative
Determination of Total Phosphorus in Water," Jour, of Am. Water Wks.
Assoc., 58, 1363 (1966).
5. O'Farrell, T. P., and Bishop, S. L., "Filtration of Effluent from
Staged Nitrification-Denitrification Treatment," presented at the
76th National AIChE Meeting, March 10-13, 1974, in Tulsa, Oklahoma.
36
-------
SECTION VIII
PUBLICATIONS
Heidman, J. A., Bishop, D. F., and Stamberg, J. B,, "Carbon, Nitrogen,
and Phosphorus Removal in Staged Nitrification Activated Sludge Treatment,1
AIChE Symposium Series 145, Water 1974, 71 , 264 (1975).
37
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-670/2-75-052
2.
3. RECIPIENT'S ACCESSIOP+NO.
4. TITLE AND SUBTITLE
CARBON, NITROGEN, AND PHOSPHORUS REMOVAL IN STAGED
NITRIFICATION-DENITRIFICATION TREATMENT
5. REPORT DATE
June 1975; Issuing Date
6. PERFORMING ORGANIZATION CODE
7. AUTHORS)
> Heidman
Dolloff F. Bishop
John B. Statnherg
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Government of the District of Columbia, Department of
Environmental Services, EPA-DC Pilot Plant, 5000
Overlook Avenue, S.W., Washington, DC 20032
10. PROGRAM ELEMENT NO. 1BB043
ROAP 21-ASO Task 017
11.
68-01-0162
12. SPONSORING AGENCY NAME AND ADDRESS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final Report - 10/72 to 9/73
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A three-stage activated sludge system with mineral addition for nutrient removal was
operated with District of Columbia primary effluent. Influent flow followed a pro-
grammed diurnal cycle and averaged 205 m-Vday (54,000 gpd). The first biological
reactor was operated as a modified aeration system with ferric chloride addition for
supplemental phosphorus removal. The clarified effluent then flowed to the second
reactor for the biological nitrification of ammonia and organic nitrogen. Dry lime
was used for pH control. Methanol was added to the nitrified effluent, and biological
denitrification occurred in the final activated sludge system. Prior to clarification,
the denitrification effluent was briefly aerated for nitrogen gas removal and for
consumption of any excess methanol. The clarified effluent was then split into two
equal streams for comparison of filtration performance of a dual-media coal and sand
filter with that of a multi-media coal, sand, and ilmenite filter. Effluent quality
consistently met the proposed D.C. discharge standards of BOD ^4.5 mg/1; total N ^
2.5 mg/1; and P ^ 0.22 mg/1.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
*Nitrification
Aeration
^Activated Sludge
Process
^Anaerobic Processes
Methyl Alcohol
^Filtration
n i
------- |