6SOR80009C States Municipal Environmental Research September 1980
mental Protection Laboratory
Cincinnati OH 45268
Reiearch and Development
International Seminar on
Control of Nutrients in
Municipal Wastewater
Effluents
Proceedings
Volume III: Nitrogen and
Phosphorus
Hotel del Coronado
(San Diego)Coronado, California 92118
September 9, 10, and 11, 1980
-------
CONTROL OF NUTRIENTS
IN MUNICIPAL WASTEWATER EFFLUENTS
VOLUME III: NITROGEN AND PHOSPHORUS
Proceedings of the International Seminar
San Diego, California
September 9-11, 1980
Seminar Convener:
E. F. Barth
Wastewater Research Division
Municipal Environmental Research Laboratory
Speakers:
Dr. H. D. Stensel, Salt Lake City, Utah
Mr. M. C. Goronszy, Sidney, Australia
Dr. R. L. Irvine, Notre Dame, Indiana
Mr. H. R. Kohl, Lake Buena Vista, Florida
Dr. T. E. Wilson, Chicago, Illinois
Mr. L. G. Shur, Corvallis, Oregon
Mr. D. E. Eckmann, Chicago, Illinois
Mr. W. A. Peterson, Boston, Massachusetts
Municipal Environmental Research Laboratory
and
Center for Environmental Research Information
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
AGENDA FOR
INTERNATIONAL SEMINAR
ON
CONTROL OF NUTRIENTS IN WASTEWATER EFFLUENTS
SEPTEMBER 8, 1980
7:30 to 9:00 p.m. RECEPTION/EARLY REGISTRATION
VOLUME I Page
SEPTEMBER 9, 1980 PHOSPHORUS CONTROL TECHNOLOGY
7:30 to 9:00 a.m. REGISTRATION
9:00 to 9:15 WELCOME AND INTRODUCTION TO PROGRAM
Mr. Edwin Barth
Chief, Biological Treatment Section
Wastewater Treatment Division
U.S. EPA/MERL
9:15 to 10:05 NUTRIENT REMOVAL TECHNOLOGY - THE CANADIAN 1
CONNECTION
A presentation of the rationale for nutrient
control; the development of an R&D, legislative,
and technology transfer program; implementation
of low cost technology at existing municipal
plants; and impact and current status of control
technology.
Speaker: Dr. Norbert W. Schmidtke, Director
Wastewater Technology Centre
Environmental Protection Service
Environment Canada
Burlington, Ontario, Canada
10:05 to 10:20 COFFEE BREAK
10:20 to 11:10 PHOSPHORUS REMOVAL IN LOWER GREAT-LAKES MUNICIPAL 39
TREATMENT PLANTS
A survey of phosphorus removal processes of
various types, with statistical summary of lower
lakes facilities, histograms of performance and
loadings, and a discussion on phosphorus
availability in relation to treatment processes.
Speaker: Dr. Joseph DePinto
Department of Civil and Environmental
Engi neeri ng
Clarkson College, Potsdam, New York
iii
-------
VOLUME I (Continued) Page
11:10 to 12:00 EXPERIENCES AT GLADSTONE, MICHIGAN UTILIZING 91
ROTATING BIOLOGICAL CONTACTORS FOR BOD,
PHOSPHORUS AND AMMONIA CONTROL
A rotating biological contactor facility with
summary data on effluent residuals, key daily
operational points, actual cost data, and
recommendations on future facility design from an
operational standpoint.
Speaker: Mr. Willard Lee Morley, Superintendent
Water and Wastewater Treatment
City of Gladstone, Michigan
12:00 to 1:00 LUNCH
1:00 to 1:50 CONTROL TECHNOLOGY FOR NUTRIENTS IN MUNICIPAL 113
WASTEWATER TREATMENT IN SWEDEN
Necessity for nutrient control in Sweden,
techniques to translate basic nutrient research
into full-scale facilities, and extent of
implementation of nutrient control in municipal
faci1i ti es i n Sweden.
Speaker: Dr. Bengt Gunnar Hultman
Swedish Water and Wastewater
Works Association
Stockholm, Sweden
1:50 to 2:40 RESEARCH ON PHOSPHORUS CONTROL IN JAPAN (separate
The type of research on phosphorus control being manuscript)
conducted in Japan, the reasons why phosphorus
control is necessary, and views of operating
facilities that utilize phosphorus removal
processes.
Speaker: Mr. T. Annaka
Department of Sewage and Serage
Purification
Ministry of Construction
Japan
2:40 to 3:30 ECONOMICAL AND EFFICIENT PHOSPHORUS REMOVAL AT A 139
DOMESTIC-INDUSTRIAL WASTEWATER PLANT
The combination of industrial and domestic waste
characteristics considered in the design of the
facility, a summary of several years of plant
efficiency, and the low-cost experience of
phosphorus control.
Speaker: Mrs. Doris Van Dam, Superintendent
Wastewater Treatment Plant
Grand Haven, Michigan
iv
-------
VOLUME I (Continued) Page
3:30 to 3:45 COFFEE BREAK
3:45 to 4:35 THE PHOSTRIP PROCESS FOR PHOSPHORUS REMOVAL 159
The PhoStrip process is discussed with emphasis
on efficiency, cost, and reliability in relation
to original design approaches.
Speaker: Mr. Carl J. Heim
Assistant Staff Engineer
Union Carbide Corporation
Linde Division
Tonawanda, New York
4:35 to 5:00 DISCUSSION ON PHOSPHORUS CONTROL TECHNOLOGY
-------
VOLUME II Page
SEPTEMBER 10, 1980 NITROGEN CONTROL TECHNOLOGY
8:00 to 8:50 EMERGING STRATEGY FOR NITROGEN CONTROL BASED ON 1
RECEIVING WATER QUALITY CONSIDERATIONS
Emerging nitrogen strategies and the need for
nitrification will be discussed, along with
research needs for nitrification to suit European
situations.
Speaker: Dr. Willi Gujer
Swiss Federal Institute for Water
Pollution Control
Dubendorf, Switzerland
8:50 to 9:40 FULL-SCALE CARBON OXIDATION/NITRIFICATION STUDIES 43
AT THE METROPOLITAN SANITARY DISTRICT OF GREATER
CHICAGO
Large-scale plant manipulations to accomplish
single-stage nitrification, with operational
control techniques related to nitrification
kinetics and to implications of control and costs
for a 1,300 MGD facility.
Speaker: Dr. Cecil Lue-Hing, Laboratory Director
Metropolitan Sanitary District of
Greater Chicago
Chicago, Illinois
9:40 to 10:30 PHOSPHORUS REMOVAL WITH IRON SALTS AT BLUE PLAINS 98
Data from the world's largest nutrient control
plant on mineral addition for phosphorus control.
Discussion of costs, alternate chemical
selection, and sludge production, plus what it
takes to put a plant of this size on-line.
Speaker: Mr. Ed Jones, Chief Process Engineer
Wastewater Treatment Plant
Washington, D.C.
10:30 to 10:45 COFFEE BREAK
10:45 to 11:45 NITRIFICATION AT LIMA, OHIO 129
Design of second-stage plastic media for
nitrification, summarizing several years of
efficiency data, operational control, and costs,
and relating these to design changes of second
generation designs.
Speaker: Mr. Felix Sampayo
Jones and Henry Engineers, Ltd.
Toledo, Ohio
VI
-------
VOLUME II (Continued) Page
11:45 to 1:30 LUNCH
Speaker: Dr. Henry Heimilch, Professor of
Advanced Clinical Studies
Xavier University
Cincinnati, Ohio
Author of the Heimlich Maneuver
1:30 to 2:20 OPERATING EXPERIENCE WITH A 30 MGD TWO-STAGE 153
BIOLOGICAL NITRIFICATION PLANT
A summary of efficiency data, control-loops,
operational modifications, and costs for the John
Eagan Plant.
Speaker: Mr. Earl W. Knight
Assistant Chief Engineer
Metropolitan Sanitary District
of Greater Chicago
Chicago, Illinois
2:20 to 3:10 NITRIFICATION-DENITRIFICATION IN FULL-SCALE 170
TREATMENT PLANTS IN AUSTRIA
Single stage nitrification/denitrification, plus
status of nitrification control in Austria and
the need for this technology.
Speaker: Dr. Norbert F. Matsche
Assistant Professor
Technical University, Vienna, Austria
3:10 to 3:25 COFFEE BREAK
3:25 to 4:15 SINGLE STAGE NITRIFICATION-DENITRIFICATION AT 194
OWEGO, NEW YORK
Second generation design for single-stage
nitrification/denitrifi cation systems, and a
real-world perspective on reliability, efficiency
demands, cost, and operation.
Speaker: Mr. Donald E. Schwinn, P.E.
Stearns and Wheler
Civil and Sanitary Engineers
Cazenovia, New York
4:15 to 5:00 DISCUSSION ON NITROGEN CONTROL TECHNOLOGY
vii
-------
VOLUME III Page
SEPTEMBER 11, 1980 COMBINED PHOSPHORUS AND NITROGEN CONTROL
TECHNOLOGY
8:15 to 9:05 DESIGN AND OPERATION OF NITROGEN CONTROL 1
FACILITIES AT TAMPA AND THE NSSD
Three-step nitrogen control at Tampa and two-step
nitrogen control at the North Shore Sanitary
District in Illinois. A summary of the design,
operation, and use of the unusual flexibility
buillt into these plants.
Speaker: Mr. Thomas E. Wilson
Principal Engineer
Greely and Hansen
Chicago, Illinois
9:05 to 9:55 PERFORMANCE OF FIRST U.S. FULL-SCALE BARDENPHO 34
FACILITY
A managed biological system for nitrogen and
phosphorus control.
Speaker: Dr. H. David Stensel, Manager
Sanitary Engineering Technology
Development, EIMCO PMD
Salt Lake City, Utah
9:55 to 10:10 COFFEE BREAK
10:10 to 11:00 DENITRIFICATION IN CONTINUOUS-FLOW SEQUENTIALLY 74
AERATED ACTIVATED SLUDGE SYSTEMS AND BATCH
PROCESSES
Present developments on batch systems controlled
by time-clocked valves and evolution into a
microprocessor-controlled municipal facility.
Speakers: Dr. Mervyn C. Goronszy
Senior Investigating Engineer
State Pollution Control Commission
Sidney, Australia
and
Dr. Robert L. Irvine, P.E.
Deptartment of Civil Engineering
University of Notre Dame
Notre Dame, Indiana
viii
-------
VOLUME III (Continued) Page
11:00 to 12:00 NITROGEN AND PHOSPHORUS REDUCTION FROM LAND 118
APPLICATIONS AT THE DISNEY WORLD RESORT COMPLEX
Several approaches to attaining defined effluent
residuals and accumulating large amounts of
analytical data for this entertainment complex,
with data on phosphorus control in the activated
sludge system, overland flow, spray, and
perculation basins.
Speaker: Mr. Robert Kohl, Director
Reedy Creek Utilities Company, Inc.
Walt Disney World
Lake Buena Vista, Florida
12:00 to 1:00 LUNCH
1:00 to 1:50 EXPERIENCE WITH AMMONIA REMOVAL BY SELECTIVE ION 137
EXCHANGE AND CLOSED-CYCLE AIR STRIPPING
REGENERANT RENEWAL
A discussion of the Tahoe-Truckee Sanitation
Agency and the Upper Occoquan facility in
Virginia, covering a closed-cycle stripping
process in relation to efficiency and effluent
residuals, operational considerations, and cost
data.
Speaker: Mr. L. Gene Shur
Vice President and Director
CH2M-Hill Consultants
Corvallis, Oregon
1:50 to 2:40 NITRIFICATION AND PHOSPHORUS REMOVAL IN A 35 MGD 185
ADVANCED WASTE TREATMENT PLANT AT ROANOKE, VA
Design parameters related to operational results
for control of nitrification and phosphorus
residuals.
Speaker: Mr. Donald E. Eckmann
Alvord, Burdick, and Howson Engineers
Chicago, Illinois
and
Mr. Harold S. Zimmerman, Plant Manager
Waste Treatment Plant
Roanoke, Virginia
2:40 to 3:30 FULL-SCALE EXPERIENCE WITH TWO-STAGE 214
NITRIFICATION AND PHOSPHORUS REMOVAL
Accumulated efficiency and cost data from two
facilities, a summary of efficiency data, as
frequency distribution, overall costs, and
operational modifications necessary for enhanced
second generation design.
Speaker: Mr. Winfield A. Peterson, Chief
Plant Operating Group N.E.
Metcalf and Eddy, Inc.
Boston, Massachusetts
ix
-------
DESIGN AND OPERATION
OF NITROGEN CONTROL FACILITIES
AT TAMPA AND THE NSSD
Thomas E. Wilson Eugene D. Lukasik
Associate Director of Operations
Greeley and Hansen North Shore Sanitary Districl
David W. Pickard
Plant Administrator
Department of Sanitary Sewers
City of Tampa
INTRODUCTION
The most recent additions to the Hookers Point plant, in Tampa,
Florida, have included full nutrient removal facilities to protect
Hillsborough Bay from eutrophication. The recent additions to the
Waukegan plant and the new Gurnee plant, of the North Shore Sani-
tary District (NSSD) , include ammonia removal facilities to pro-
tect the fish in the Des Plaines River. Each plant has operated
its nutrient control facilities for about two years. This paper
discusses the most recent year of operation for each planr, and
shows how the somewhat unique flexibility built into each plant has
been used.
DESCRIPTION OF PLANTS
The Tampa plant and the two NSSD plants have essentially the same
liquid stream process flow sheet, as shown on Figure 1. The
principal differences between the NSSD plants and Tampa is that the
former are air-activated sludge plants, providing only nitrification,
and the latter is an oxygen-activated sludge plant, with ancxic
denitrification. Simplified site plans are shown on Figures 2-4
and simplified Bases of Design in Tables 1-3.
-------
ac
o
00
ce
UJ
oo
-
00
OJ
C
^
D
O
S T3
us d
VJ (0
en
to ••
•H c
Q to
cn
S (1)
O -V
•H 3
cn
w ^
-------
—
X
03
a
03
EH
c
.Tl
i—I
0<
-------
C
(0
3
£
(1)
oo
(U
tj
D
-------
-------
Table 1
Key Design Parameters - Tampa Plant
Design Year
Average Flow, mgd
Maximum Flow, mgd
1985
60
140
Wastewater Characteristics
Design Temperature, °C
BODs, mg/1
SS, mg/1
N, mg/1
P, mg/1
20
224
221
32
12
Preliminary Treatment
Pre-aeration, coarse screens, grit tanks
Partial primary sedimentation
Activated Sludge
Flow, MGD
Aeration Tanks
Type
Number
Retention Time, hours
Average Sludge Recycle, %
Maximum Normal Bypass, %
Sedimentation Tanks
Number
Depth, ft
Retention Time, hours
Surface Overflow Rate, gpd/ft2
First Step
60*
Second Step
60
Covered, oxygen, mechanical aerators
2 4
1.0 2.0
50 100
25
4
12
2.4
893
8
12
4.8
447
Denitrification Filters
Average Flow, mgd
Maximum Rate
Number
Media (effective) size, mm
Media depth, ft
Empty bed displacement time,
Hydraulic Loading,
mm.
60
120
20
4
5.5
20.6
2.0
Chlorination - Post Aeration Tanks
Number
Depth, ft.
Chlorine Contact (Total Tank) min.
Post Aeration (first portion of tank) Zone,
nan.
2
10
38.2
8.6
*9 MGD Bypass Normally.
All parameters given in terms of average flow (60 mgd)
-------
Table 2
Key Design Parameters - Waukegan Plant
Annual Average Flow, mgd
Maximum Flow, mgd
19.9
39.8
Wastewater Characteristics
Design Temperature, °C
BODs, mg/1
SS, mg/1
N, mg/1
11
166
174
32
Preliminary Treatment
Coarse Screens, Storm Water Retention and Sedimentation Basins, Prechlorination,
Aerated Grit Chambers
Primary Sedimentation Tanks
Number
Displacement time, hours
Surface Overflow Kate, gpd/ft2
6
2.0
4 @ 580(1>
2 @ 905
(1) Existing Imhoff Tanks.
Activated Sludge
Aeration Tanks
Type
Number
Displacement, hours
Maximum Sludge Recycle, %
Maximum Bypass, %
Sedimentation Tanks
Number
Displacement Time, hours
Surface Overflow Rate, gpd/ft2
First Step
Air; Coarse Bubble
Diffusers
6
2.5
50
25
5
2.5
780
gecond Step
Air; Coarse Bubble
Diffusers
9
3.8
100
25
6
3.2
620
Effluent Filters
Number (includes one standby)
Media (effective) size, mo
Media, depth, ft.
Hydraulic Loading, gpd/ft2
6
2.1
5
2.6
Chlorine Contact
Maximum Flow, mgd
Displacement at max. flow, min.
39.8
13.7
All parameters calculated on basis of average flow (19.9 mgd)
-------
TaDle 3
Key Design Parameters - Gurnee, Initial Years
Annual Average Flow, mgd
Maximum Flow, mgd
12.0
27.9
Wastewater Characteristics
Design Temperature, °C
BODs» rag/1
SS, mg/1
N, mg/1
12
235
222
32
Primary Clarifiers
Retention Time, hours
Surface Overflow Rate, gpd/ft^
2.7
695
Activated Sludge
Aeration Tanks
Type Aeration
Number
Retention Time, hours
Maximum Sludge Recycle, %
Clarifiers
Number
Retention Time, hours
Surface Overflow Rate, gpd/ft^
Carbonaceous
Step
Nitrification
Step
Air, Coarse Bubble Diffusers -
5 7
3.0 4.2
70 100
3
1.8
1,080
5
3.1
645
Effluent Filters
Media (effective) size, mm
Media Depth, ft
Hydraulic Loading,
2.1-2.4
5
1.7
Chlorine Contact/Post Aeration Tanks
Depth, ft
Contact Time, min.
Type Aeration
8
36
Surface Aerators
-------
In addition to the processes shown, the Waukegan and Gurnee plants
have storm water retention basins. These are located on-site at
Waukegan and remotely, at the NSSD's North Chicago pretreatment
plant, for Gurnee.
Discussion of sludge handling and disposal facilities are beyond
the scope of this paper. However, at the NSSD, all of the sludge
dewatering facilities for the District's three AWT plants (each
currently treating about 15 + mgd) are located at the Waukegan
plant. The recycle streams from this operation have a signifi-
cant impact on the treatment at Waukegan. Further details con-
cerning these plants are available elsewhere .
Each of these plants is extremely flexible. This flexibility,
called MID-STEP nitrification, has been discussed in detail else-
where and is illustrated on Figure 5. Briefly, each plant:
o Can operate a single-step plant, a two-step plant, or
modes intermediate to these
o Allows, in the two-step mode, the operator to vary the
fraction of aeration and clarifier tankage used in each
step, without taking any process unit out of service
o Has internal bypasses of each step of activated sludge
Furthermore, the NSSD plants both have full step aeration (i.e.
step-feed of wastewater) capability which enables each step
(carbonaceous or nitrification) to be operated in modes ranging
from plug flow (all wastewater added to first pass) through contact
stabilization (all wastewater added to last pass, with remaining
tankage used to reaerate sludges).
-------
Q
i
3
£
rt
v\\>\\\\
\\xv\\v
v\\\\\\\
\\\N?\\\\
X\\X\\V
v\\\\\\\
tr
O
o
<
LU
c
o
2
o
to
| 300 GPSFPD
\\\NK\\\
x\\x\\v
v\\\^\\X
\\\>\\\\
x\\x\\v
v\\\\\\\
\\\\\\\>
V\\\X\\V
\\\Vv\\\
\\\\\\\>
C\\\X\\V
\\\\\\\\
Q
LU
LU
Q
O
Cfl
i
Q
a
3
Q
o
s
-------
PERFORMANCE
The performance of these plants for the most recent year (May 1979
through April 1980) is summarized in Tables 4, 5 and 6. The design
requirements are also included in these tables. The period reported
includes the second full winter of operation, in the nitrification
mode, for each plant.
TAMPA
As can be seen from Table 4, the plant always produced an effluent
with significantly less BOD5, SS and Total N than required by the
NPDES permit, averaging, respectively, 5.6, 3.1 and 2.7 mg/1. Only
February of 1980 even exceeded the design total nitrogen level of
3.0.
The Tampa plant operated at an average flow rate of 45.8 mgd which
is about 76 percent of design flow. There were months (August and
September at 55.6 and 58.6 mgd, respectively), however, where the
flow approached the design flow, on a monthly average basis, and
exceeded it on a daily basis.
At Tampa, the 8005 and SS loadings to the first-step reactor have
typically exceeded design loadings:
Parameter
BOD5, Ib/d
SS, Ib/d
Design
81,700
63,000
March-April
1980
120,100
86,700
This has necessitated bringing on 50 percent more first-step reactors
(from 2 to 3 reactors) to keep the MLSS near the design level of
4,000 mg/1. This is discussed in more detail elsewhere in this paper.
11
-------
(0
a
EH
•=*• 1
0) 0)
rH U
.Q C
(0 «J
I
O
M-"
S-i
OJ
OH
C
•H
Ul
2
*
en
<
O
CO
en
iH
Cft
r»
eft
rH
rH
•H
SH
a
jC
0
flj
«s
,
^^
a)
DL,
.
C
a
^
.
o
2
>
0
z
,
4-1
o
o
4J
a
0)
en
•
cr>
3
•<
>i
rH
3
1-5
(1)
^
>
(d
meter
f r-
co r-
• •
oo r-
ro f*»
rO 0**
Cfi CN*
ro CO
CN en
CN 'SJ1
T CO
•* in
• •
O \Q
m oo
V0 CN
oo* r^
m oo
vfi \s>
m r»
in co
^* eft
• •
TJ« m
•ff CO
cr> m
TT H
^r co
vo vo
• •
m r*
in r^
^^_
b*
0
-o ^
? Qj
0 g
rH (I)
*5j* ro
CM rH
CM
in vo
CTt •
CM in
CM in
^*
^
rH vD
eft
en
CO CO
r-
cn
CO 5T
P^
d
CN VO
fO
CM rH
VO rH
m
in vo
r^
m
O •*•
00
rH
vO ^*
CO
rH
ro CM
rH
CM
vo n
CM
^r in
CN
CM
^^
rH -P +J
\ C C
CP
vo in
eft
CN
rH CO
O
CO
o m
eft
CN
O CN
CO
CN
CO CM
CN
m CN
CN
vO CN
^*
CN
CN ro
rH
i-H
CN TT
V0
rH
00 CM
f^
rH
CN cn
O
CM
r~ ^«
eft
rH
— 4J 4J
rH C C
\ (0 0)
CP 3 3
J3 rH rH
C <4H
CO H H
cn
1
rH
O
o
*
o
CN
«
O
vO
d
o
d
rH
d
o
d
o
d
o
d
o
d
o
o
rH
d
o
d
4.^
rH
\ -P
II
rH
z il
ff) y
Z
CN
>^
I rn
rH Cft
rH •
CM
vo in
• •
O CM
VO C*^
• •
O CM
rH 00
• •
rH m
CO CN
• •
O CN
ro H
rH fO
o m
rH CM
o in
• •
rH CN
r^ CM
• •
O CM
Cft ^*
d CM"
0 T
rH CN
o ^
rH CM
o m
• •
H CN
^Ht,
*-* rH
rH \
CT* ^J g -P
•^ Q) (0
3 Z 3
Z H rH
| VM r-H ttH
UH «J UH
ro P3 j_i PT*|
O O
Z H
.
rl
3
a)
rl
0)
g
*
rl
0)
4J
3
g
3
g
•H
C
•rl
G
•O H
C \
(0 W
1 o
•H in
X
§0)
(I)
rl
H1 '3
5 S1
C rl
§ -U
•rl
*W &
o n
0)
0) O4
CCJ CO
H Cd
vV Q
^ C^
< z
rH CN
*•
0)
4J
0
12
-------
First-step SRTs have been, typically (March-April 1980), at about
1.8 days, versus the design value of about 1.4 days. However, the
F/M ratio during March-April 1980 has been about 1.2 Ibs BOD5/d/
Ib MLSS, versus the design value of 1.11, suggesting less sludge
production per pound of 8005 applied than assumed in the original
design.
First-step clarifier surface overflow rates and solids loadings
may be compared:
March-April
Parameter Design 1980
Overflow, gpd/ft2 760 400
Solids Loading, lb/d/ft2 41.5 19.2
The light loadings were required due, in part, to accommodate the
poorer mixed liquor settling/thickening properties encountered:
March-April
Parameter Design 1980
SVI 100 150
RSS, mg/1- 13,000 13,000
% Return Sludge 50 45
The second-step SRT generally has been kept higher than designed:
March-April
Parameter Design 1980
SRT, days 6 10
The method of SRT calculation used, however, may not account for
all of the solids wasted. At Tampa about 40-120 pounds of second-
step solids per million gallons of wastewater flow are removed by
13
-------
the scum collection equipment and are not included in the SRT cal-
culation. If the calculation were corrected for this, the value
of SRT would be only about 6-8 days.
While the second-step clarifier surface overflow rate has been
lower than design, the solids loading has been higher:
March-April
Parameter Design 1980
Overflow Rate, gpd/ft2 450 400
Solids Loading, lb/d/ft2 5.6 11.2
The higher solids loading is due to the need for keeping a higher
MLSS, since only 2 or 3 reactors are in use instead of the design
4 reactors. This has resulted in keeping a MLSS of about 2,500
mg/1 instead of the 1,000 mg/1 used in the design. However, since
the design of these was not solids loading limited, there have been
no problems.
During March and April of 1980, there were 19 of the 20 denitrify-
ing filters on-line:
Parameter
Empty Bed Detention Time, min.
Hydraulic Loading Rate, gpm/ft2
N03~N in Filter Influent, mg/1
Methanol Dosage Ratio, mg/mg NO3~N
"Bump" Frequency, times/day/filter
Full Backwash Frequency, times/week/
filter
Design
20.64
1.98
19
3.0
4-9
<2
March
32
1.36
9.0
5.9
6.9
5.6
April
24
1.93
9.5
3.85
6.9
5.6
From the above it can be seen that, compared to design:
14
-------
o The influent NC>3-N is lower, being only about half of
design levels. This is due to greater uptake of N in
the sludge wasted from the first step.
o The methanol dosage ratio was higher in March, due to
unintentional overdosing. April is more typical. One
possible reason for this higher dosage ratio is that the
automatic methanol feed system was not yet functional.
The methanol feeding was done by manually measuring
NC>3-N levels and periodically changing the methanol feed
rate-
o Full backwashes are needed more frequently than originally
expected. This appears to be chiefly due to differences'
in the media used in the pilot study (upon which the
design was based) and the media supplied.
As shown in Table 4, the performance of these filters has been excel-
lent, with effluent N03-N averaging 1.1 mg/1. At higher methanol
dosages, and with closer control, it is expected that concentrations
approaching zero nitrate could be approached. Generally, the level
is intentionally kept at about 1.0 mg/1, which keeps the total
effluent N below 3.0 mg/1, but minimizes the use of methanol.
WAUKEGAN
The performance of the Waukegan plant was more marginal than that
of the Tampa or Gurnee plants. The monthly BOD§ standard of 10 mg/1
was exceeded once (11 mg/1 in March of 1980). The 12 mg/1 SS
standard was exceeded twice (April and June of 1979 at 22 and 18
mg/1, respectively). This was due, at least in part, to the fact
that this plant was operating with no effluent filters during the
entire period reported and was often at, as discussed below, load-
ings far in excess of design loadings. Similarly, the NPDES 3.0
mg/1 NH3~N standard was exceeded/ on a monthly average basis, during
the first three months reported (5.6, 7.1 and 3.5 mg/1, respectively)
This was due to several factors, including excessive recycled solids,
15
-------
exceptionally high flows at low temperatures, and lack of a good
SRT control strategy. In spite of this, for the period reported,
the plant met the effluent daily3.0mg/l standard for ammonia over
87 percent of the time.
The annual average flow of 17.5 mgd represents 88 percent of de-
sign flow. The maximum month, however, exceeded the design flow by
a factor of about one-third. The winter of 1978-79 was one of the
worst winters in the history of the NSSD and the combination of
high flows, low wastewater temperature, inaccessibility of tankage
due to snow and low air temperature, and inexperience with winter
nitrification, caused nitrification to be seriously impaired until
late June 1979. However, even though the water temperature dropped
to design levels during the winter of 1979-80, the nitrogen standards
were consistently met (Table 5).
As at Tampa, the BOD5 and SS loadings on the first-step activated
sludge typically exceeded design loadings. In some months, the
loadings (particularly the SS loadings) have been extraordinarily
high, due to recycle streams:
Parameter
BOD 5, Ib/d
SS, Ib/d
Design
20,200
12,995
February ,
1979
30,308
50,600
June,
1979
25,109
64,720
While the plant could handle the excessive BODs loadings, the exces-
sive SS loadings made it difficult to maintain adequate second-step
SRTs. Data to accurately calculate SRTs for these months are not
16
-------
c
r
^
<
§
0^
r-l
ji
o
fy
^
r*
fa
§
(0
IT)
rH
3
*"3
•
35
. I
•H
^1
Oi
W
rt)
g
2
2
°1
en
rH
in
r-
rH
vO
vO
0
t
in
rH
en
vo
rH
VO
r-
rH
rH
rH
en
rH
CO
r-
rH
rH
tH
r* co
H
r*
VO CO
H
r- co
rH
\Q
^ft vO
rH
VO
o in
CN
VO
m ^r
CN
en
o r-
cn
rH VO
CN
in co
in H
in
r»
co en
en
rj> CN
Cn CN
en
^ 4J 4J
H C C
\ 0) 0)
» 3 3
S H rH
CO H »
CO
m
•
rH
CN
*
CN
CO
00
•
rH
O
H
•
rH
cn
vO
t
CN
CO
rH
CN
VO
t
o
r-
CN
d
o
o
r-
CN
•
0
m
en
•
rH
VO
^t
•
en
•*
rH
t
r-
m
vo
•
m
r7
X. 4J
Q^ CN
H
J2 4^
1 ^W
i
i
CN
CN
•
O
O
•
rH
rH
m
CN
rH
CN
CN
fji
O
rH
rH
O
t
H
VO
rH
VO
H
in
•
o
rH
rH
•
CN
rH
CD
in
•
00
m
rH
•
CN
VO
in
•
H
rH*
\ 4J
I §
rH
53 ^^
Jo a
S
CN
en 1
CN r^
vo ^
CN rH
rH in
rH ^*
CN H
O O
in vo
CN rH
CT* rH
co in
H H
rH ^P
CN en
CN rH
rH CN
in vo
CN H
in en
t *
in co
CN rH
•v in
CN H
rH H
• t
CN T
CN rH
CO rH
• •
^ V0
CN rH
in co
* •
r^* n
<*") rH
n ro
» •
rH CN
rn rH
vo r-
• t
in rH
en H
4J 4J
c c
ft) CD
233
rH rH
rH UH VM
«J C MH
4J H a
£
*
M
OJ
4J
c
•rH
?
OJ
4-1
C
•H
H
S
o
•
i1
in
si
4J
C
e
5
?
OJ
S
•H
Q
§
S
•H
X
(8
rH
•H
(0
T3
CT
B
O
m
(0
iH
0
VM
CQ
rH
H
(Tj
0
4J
•rH
S
r«|
0)
Cu
01
a
H
«•
CU
4-1
0
z
17
-------
available. However, it is possible to calculate the average sus-
pended solids sludge age (S.A.):*
Parameter
Second Step Sludge Age,
Wastewater Temperature,
Theoretical Minimum SRT
(rounded) , days
days
oF
Required
February,
1979
0.7
49
10
June,
1979
0.6
62
5
Even if one assumes that the solids production was equal to only
half the influent solids load (i.e. a net destruction of 50 per-
cent of the total solids) these sludge ages would correspond to
SRTs of less than 1.5 days. As shown above, these are considerably
less than the SRTs required for nitrification at the wastewater
temperatures encountered.
First-step SRTs generally were kept low (3-4 days) until July of
1979. After some initial problems, an SRT control strategy was
devised ' which allowed the first-step SRT to be kept above
9 days, ranging from 9.6 to 14.1 days, on a monthly average basis,
during the period October, 1979 through March, 1980. This higher
SRT was achieved by using a higher than design fraction of carbon-
aceous tanks, coupled with a contact stabilization mode (see next
section of the paper) and careful wasting. More stable operation,
and typically, partial nitrification (1-7 mg/1 N03~N formed) re-
sulted. F/Ms, not SRTs, were used to design this plant. Comparing
design and operating first-step F/Ms and MLVSSs:
*S.A. «MLSS x Aeration Tank Volume/(Influent SS x Flow)
18
-------
Parameter
Design
April, 1979-
March, 1980
Oct., 1979-
March, 1980
F/M Ib BOD5/lb MLVSS-d
MLVSS (average), mg/1
0.77
1,500
0.45
4,900
0.32
4,900
The higher average MLVSSs and lower F/Ms resulted from the use of
the contact stabilization mode and 50 percent more first-step tank-
age than originally designed (9 vs 6).
First-step clarifier surface overflow rate, solids loading, return
sludge fractions and return sludge concentrations may be compared:
Parameter
Overflow, gpd/ft^
Solids Loading, lb/d/ft2
Return Sludge, %
RSS, mg/1
Design
778
16.2
25
10,000
April, 1979-
March, 1980
717
15.6
14.5
8,900
All of these, except return sludge rate, appear to be close to de-
sign values. The lower return rate resulted from the use of the
contact mode.
The second-step SRTs, as mentioned before, cannot be accurately
calculated. The design parameter used was NH3~N loading:
Parameter
Design
April, 1979-
March, 1980
Loading, Ib NH3-N/lb MLVSS-d
MLVSS, mg/1
0.079
1,800
0.039
2,233
The lower loading is primarily due to lower ammonia concentrations
reaching the second step. This is due to a combination of lower
19
-------
raw sewage nitrogen, higher BODss (causing higher uptake, as sludge
solids, of N in carbonaceous step) and partial nitrification in the
first step. It should be noted that the volume of the second-step
aeration tanks had been reduced to two-thirds of original design
value (6 vs 9) .
Second-step clarifier surface overflow rates and return sludge
solids concentrations have been near the design levels:
Parameter
Overflow Rate,
Solids Loading,
Return Sludge,
RSS, mg/1
gpd/ft2
, lb/d/ft2
%
Design
622
18.6
50
7,200
April, 1979-
March, 1980
574
Not Available
Not Available
6,200 (Oct.-
March only)
GURNEE
Table 6 shows that, for the period April, 1979 through March, 1980,
the Gurnee plant, on a monthly average, met all its effluent
requirements, except for effluent suspended solids during April.
That month it averaged 14 mg/1 instead of the required 12 mg/1.
High flows (about 43 percent above design) and an industrial waste
induced filamentous bulking problem were the chief problems .that
month. (For details of the filamentous bulking problem, see
reference 12.)
The Gurnee plant flow averaged 12.9 mgd, which is 108 percent of
design flow. The flow exceeded the design level of 12 mgd six
of the 12 months reported.
20
-------
0)
(U
c
1 ,
M
3
O
vo 1
0) 01
rH U
& C
(0 (0
£n g
U
o
M-l
^1
V
f*
M-l
P
•^
V
cJ
c
>
<
o
CO
cn
,j
r^
•S
^
(T
s
x
d
r
W
c
I-)
r\
rH
sx
Q
2
4J
U
O
*
4->
a
0)
CO
ff
cr
3
<
>•
rH
3
h-i
1 j
0)
2
h™i
1 j
>
S
rH
•H
M
a
<
rl
OJ
4-1
2
in
A
n o
•
CM
rH
cn
9
CD CD
— 533
•^ rH rH
• I4H 14H
a in c vw
S Q H W
0) Q
& &
CN CN
CN H
CN
-H CN
O I1
CM •
VO
in
o r-
CN
o
CN ro
CN
oo r-
^ rH
CM
T
CN cn
CM
rH ro
CN
ro
CO ro
rH
O
r^ CN
H
ro
O CN
CN
m
ro ro
CN
in
rH tp
CM
r- o
CO rH
rH
cn ^r
cn H
— 4J -P
H C C
\ 0)
§
rH
4-4
MH
W
CN
ro 1
O CO
CM in
ro H
rH "31
» t
m H
ro H
CO CO
• •
rH CN
ro rH
ro ro
CO ^
CM rH
m m
rH Tl"
rO rH
cn ^
r- r~
CM H
00 CO
co in
CN H
^r ro
HI rt
vf
ro rH
vo ^
vO CO
CN rH
^" CN
O CM
TJ> CN
r~ co
m co
ro rH
CN cn
in in
ro H
CM ro
• •
ro CM
ro rH
H
cn
S v 4->
"- C C
-------
While the flows were above design loadings, the 3005 and SS load-
ings were actually somewhat less than design levels. This can be
seen in Table 6, by noting that the influent BOD5 and SS concen-
trations were at somewhat less than design levels.
First-step SRTs were erratic, ranging, on a monthly average basis,
from 3.8 to 16.3 days. Generally, no sludge was intentionally
wasted from this step and its effluent SS were quite high due to
the bulking problem mentioned above. F/M was used as the design
parameter:
Parameter
F/M, Ib BOD5/lb MLVSS-d
MLVSS, mg/1
Design
0.95
1,500
April, 1979-
March, 1980
0.45
1,800
It should be noted that the first-step tankage used during this
period was 20 percent greater than in the design (6 vs 5 tanks).
Also, two-thirds of these aeration tanks were used for reaeration
and only one-third for contact. The resulting light loadings
allowed significant nitrification to occur in the first step.
Typically, about 7 mg/1 of N03-N (about half of the nitrate formed)
were produced in this step, considerably reducing the ammonia load-
ing to the second step.
Second-step SRTs often actually were lower than the first-step SRTs,
typically being in the 5-8 day range. Ammonia loadings were used
as the design parameter:
22
-------
Parameter
Design
April, 1979-
March, 1980
Loading, Ib NH3-N/lb MLVSS-d
MLVSS, mg/1
0.070
1,800
0.019
1,800
The aeration tank volume used was one-seventh less than design
(6 vs 7 tanks). The low ammonia loading was primarily due to the
high degree of nitrification occurring in the first step.
The surface overflow rates in both steps were essentially the same
as design levels. The return sludge metering systems were not
functioning, so no accurate values of solids loading could be cal-
culated. Return sludge concentrations were, however, quite low
due to the bulking problem:
Parameter
First-Step RSS,
Second-Step RSS
mg/1
, mg/1
Design
10,000
12,000
April, 1979-
March, 1980
2,600
4,200
FLEXIBILITY
All three plants were designed and expected to be normally operateu;
o As two-step (separate step) nitrification plants
o Using most of the tankage in the second (nitrifying) step
In actuality, these plants have often been operated quite differently
than this.
23
-------
SINGLE-STEP NITRIFICATION
The NSSD plants have only been successfully operated as two-step
nitrification plants. An attempt was made to start the Gurnee
plant as a single-step plant, but it was unsuccessful. At Tampa,
however, start-up limitations forced the use of single-step nitri-
fication during the period from August, 1978 through January, 1979.
During this period:
o No primary tanks were in service. Raw sewage was fed
directly to oxygen reactors until October, 1978. At
that time, 4 primaries were gradually brought on-line
o The sludge drying beds were not complete, and there was
no place to put all of the sludge solids
o All 6 reactors were used as first-step reactors
o All 12 clarifiers were used for single-step nitrification
o The filters were first being brought on-line, gradually
from October 1978 through December 1978
At the 41.8 mgd flow encountered during this period, the final
clarifier surface overflow rate dropped from about 420 gpd/ft^
when using 6 clarifiers in each step of a two-step plant to about
210 gpd/ft2 by using 12 clarifiers in a single-step mode. At the
same time, the final clarifier solids loading dropped from about
43 lb/d/ft2 (assuming 7,000 mg/1 mixed liquor solids, a 75 percent
return sludge rate and a 42 mgd flow) to 21.5 Ib/d/ft2. This made
the difference between being able to hold the biosolids in the
plant and not.
Of significant note is the higher effluent NH3-N concentration (1.2
mg/1) observed during this mode of operation. During virtually all
two-step modes, the NH3~N was less than 0.1 mg/1
24
-------
TANKAGE DISTRIBUTION
At Tampa, the 6 oxygen reactors and 12 clarifiers have been dis-
tributed as in Table 7. As shown on this table, the reactors were
typically equally divided between the first and second steps. Most
recently, 60 percent of the reactors (3 of 5) have been used in the
first step instead of the 2/4 division anticipated in design. The
final clarifiers have essentially always been equally divided rather
than split 4/8, as designed.
The major reasons for the deviations from the design split of tank-
age include:
o Higher first-step 8005 loadings than expected. Even
though the plant has averaged somewhat less than design
flow, higher than design strengths 8005 have led to
loadings greater than design levels
o Poorer settling/thickening mixed liquors. This has been
discussed in the previous section
o Need for using reactors and clarifiers to store sludge
o Mechanical problems in reactors
At the Waukegan plant the 15 aeration tanks and 11 final clarifiers
have been distributed as follows:
Aeration Tanks
Period
Design
Pre July, 1978
July, 1978-Present
First
Step
6
6
9
Second
Step
9
9
6
Clarifiers
First
Step
5
5
5
Second
Step
6
6
6
Until July, 1978 the nitrification was erratic. Studies indicated
that the first step was loaded at BODs and suspended solids load-
ings several times design levels. It was impossible to consistently
25
-------
to
EH
fO
H
(0
C
o
•r-l
3
.0
•H
M
4J
CO
P
JJ
OJ
•r-l
S-l
(C
rH
O
to
M
O
O
tO
ca
O
I
•a
c ftl
o a>
O 4J
CU CO
CO
to ft
Vl G
-rH 4-1
CM CO
•a
C ft
O CU
O 4J
co
CO
co a
Vj CU
H 4->
fo CO
04
beds incomplete,
&
G
•H
^
•a
^
CO
0)
•H
M
flJ
-H
^j
ft
o
ft
3
£
Tj
4J
to
«,
CO
3;
H
M-4
U
o
H
O
4J
CO
1
rH
•§
ft
4J
C
V
g
ft
•H
3
V C
CU §
o
M -O
O
•P CO
O 4J
<0 -H
0) C
reasing
0
c
•rl
j?
O
G
CU
C-
•H
to
T)
-H
rH
O
(0
ft
4J
CO
4J
CO
M
•H
X
-H
.C
,G
•H
£
M
O
4J
C
^
G
-rl
CO
•a
•H
i— I
o
to
•0
id
o
Q
8
n to hold solids
s go on-line,
e, filters on-line
0 CU C ~
•rl -r| -H O
•a w H C
4J •> Id -H
CO 4J rH Vl
C 3
CU (d CU T)
H rH tyi
(7> ft T3 rH
C 3 rH
CO -H CO >—
ft
CU
4J
to
O
0
4J
0
H
0
VI
4J
C
O
O
VI
T3
rj
to
•O
-H
rH
0
to
£j
O
id
0)
&
tors to maintain
o
fl
0)
J^
o
2
4-1
^
r-H
G
o
T3
0)
rrj
(1)
Q)
G
EH
Pi
CO
ft
4-1
01
4-1
Cfl
^1
•H
M-l
caused solids inven-
CQ
g
CU
H
O
ft
•O
0)
XI
G
•H
^i
Vl
•o
iH
Q
0
Vl
ft
^1
^
o
4J
S
ft
H
id
o
•H
•g
s
CU
3
•o
§
Q
r0
VI
0
4J
O
CU
6
CU 0)
C rH
0 X)
00
\O
VO
vO
VO
vO
VO
CM 01
rn
G
Cf<
•H
CO
CU
Q
00
CO
CO
(N
CO
CO !>•
cn CN
(N \
^ I
I CO
CO l-»
\ en
H
CM
I
CO
CO
cn
(N
Cft
r»
CN
CN
VO
vO
ro
a\
o
H
rH
I
r-
(N
H
vO
to
O
CO
o
ro
ro
I
O
-------
provide enough air and keep an adequate SRT to assure good mixed
liquor microbiology. By increasing the first-step tankage by 50
percent and converting two-thirds of the first-step tankage to re-
aeration, these problems were solved and the second step started
to nitrify consistently.
Even with this modification, the recycle streams became so exces-
sive that, in January, 1979, the plant started to not fully nitrify.
Complete nitrification was not obtained again until June, 1979, when
the recycle loads were brought under control. Figure 6 shows that
when the influent SS were increased by recycle loads to a weekly
average of 500 mg/1 (about 2-1/2 times design level) that nitrifi-
cation started to be lost.
The 12 aeration tanks and 8 final clarifiers at the Gurnee Plant
have been distributed as follows:
Aeration Tanks
Period
Design
Pre May, 1978
May- July, 1978
July, 1978-Present
First
Step
5
12
5
6
Second
Step
7
0
7
6
Clarifiers
First
Step
3
8
3
3
Second
Step
5
0
5
5
The single-step mode used prior to May, 1978 did not nitrify satis-
factorily, but did provide adequate BODs and SS removal. In May,
1978 the plant was switched to a two-step mode. The BODs and SS
immediately deteriorated. The cause was found to be an industrial
waste-induced filamentous bulking problem. By switching the first-
step mode from feeding all the primary effluent and return sludge
at influent of the aeration tanks (i.e. plug flow mode) to using
27
-------
N-SHN
28
(0
o
<=>
co
oa
00
o
OJ
(0
c
o
o
GO
C
o
a
z
LU
GO
DI
C
ju
-U
cn
!
0)
oe
LU
cn
LU
00
4-1
O
-U
o
-------
all but two tanks for reaeration of sludge (i.e. a contact stabil-
ization mode) the plant was quickly stabilized and nitrification
was consistently obtained by August, 1978. The details of this
(12)
have been discussed elsewhere
The Gurnee plant has operated consistently well since this point
in time, in spite of a continuing battle with the sludge bulking
problem, except for a few weeks in early 1979 when flows in excess
of the plant metering capacity (more than three times design flows)
temporarily washed the mixed liquor out of the system.
COSTS
The capital costs for these three plants are summarized in Table 8.
Note that these costs include more than just nitrification facili-
ties and may not be directly comparable.
The operating and maintenance costs for each of the three plants
are summarized in Table 9. Since bookkeeping procedures are slightly
different at Tampa than at the NSSD, the breakdown of costs may not
be directly comparable. The chemical costs at Tampa are for de-
nitrifying, sludge conditioning and disinfection chemicals. The
Waukegan chemical costs are for sludge conditioning and disinfec-
tion only. Gurnee chemical costs also include the cost of odor
control chemicals.
29
-------
cn
4J
CO
o
u
1 — 1
03
4-1
CO -H
O,
CD (fl
rH CJ
rts 4-1
EH 0
-
1^1
in
(0
E
g
cn
cn
(U
•H 'O
4-1 ,
4-1
•H
rH
•H
CJ
fl
fa
G
O
•rH
4J
fl
^1
0>
fl
c
0
•rH
4-1
fl
U
•H
4H
•H
iH
4J
•H
G
%
U)
J-t
0)
•rH
4H
•H
M
fl
rH
O
>1
iH
fl
Q
•H
iH
a
^
4-1
C
C1J
e
4J
fl
0)
rl
4J
0)
^4
d
O
O
O
<»
in
en
00
^*
•w-
^
r**
i
o
["•*
^
rH
O»
•
^
rH
rH
H
H
H
0)
•H
4H
•H
H
fl
rH
O
T3
fl
Ul
C
4-1
T3
c
fl
T3
c
fl
tn
j>4
(U
G
0)
O
•H
43
4-1
0)
Cn
T3
3
rH
cn
0,
S
CD
4-1
01
>,
C
o
•H
4->
fl
•*H
M
O
rH
43
O
cn
iH
0)
4->
rH
•H
<4H
Q
cn
cn
z
(!)
iH
•H
4-1
C
i
cn
Cn
c
•rH
)H
0)
4-1
fl
3
0)
•0
0)
T)
3
rH
01
<4H
0
cn
Q,
0)
4J
cn
0
S
4J
^-^
rH
^.^
*
CO
iH
•r(
UH
•rH
1-1
fl
CJ
>»,
^
fl
g
•H
^(
a
r7
•^-»
4J
C
X
^
Q)
•H
14_|
•H
^
(^
rH
O
qQ
c
in
•S
fl
4-1
C
O
•H
4-1
(C
a
fl
'^-' tjl
* 'O
tn 3
(H rH
cu oi
•M
fH «*^
•-HH
4H-'
4-1 ••
c tn
(U iH
3
•rH 3
3 H
43 01
cn
0)
•H
4-1
•H
rH
•rH
CJ
fl
UH
i-H
o
4-1
c
0
CJ
rl
0
T3
o
0
MH
0
CO
CO
0
3
4-1
4-1
C
4J" g
G d)
fl iH
rH
a
CU v rH
Cn cn -H
G^
0) 43 G
Cn 0
X! C 4J
O -H fl
>i O
O iH -rH
•H T3 «*H
G -H
(U 0) SH
Cn Cn 4-)
0 T3 -H
>< 3 C
iH rH 0)
O 01 T3
fl
* ^ rrj
4J fl -H
G ft iH
•H g O
O fl rH
CU EH fa
4J
G
fl
i-H
ft
4J
iH
fl
•O
O
o
0)
Ul
Q
C3
m
(N
01
fl
01
N
30
-------
en
0)
H
X!
(0
en
o
u
a
i3
O
M-l
O
(0
5
3
C/5
4J *
tn «
0 Q
U C
•p \
•H
W C
W D
O
H
tO M
3 >
C \
C
5
JJ *
tn *
0 Q
u u
2!
-p \
2 -H
<; c
O D
D
1
•H
s
.p
tn *
0 Q
U O
s
•M \
•H W-
C
D
^«»
**<
E-i
i-l
03 ^
3 >i
C
rt!
g
5
p
H
!-)
(^
1-1
O
O
O
O
cn
o
•*
H
00
H
CN
O
O
o
o
en
o
w
1-1
CTi
•^
i-(
0
O
O
^
O
m
rr
CN
M
CD
9
c
(0
s
T3
fl
C
0
•H
•P
fl
M
p
tn
•H
c
•H
CM
«c
iH
CD
O
O
O
a
CN
in
vD
m
o
o
o
o
fn
•'T
00
m
o
0
o
^
o
m
1
Oi
u
CO
c
u
r- &
en >i)
•5T
0 0
o o
0 0
0 0
CN CO
vo en
CN
m ^r
CN O
H in
0 0
0 0
0 0
0 0
r- o
en en
m
CN VO
00 "—I
•*r
o o
o o
o o
^ *
0 0
m fi
m co
-H VO
M H
CO <0
J3 -P
-P O
o ^
co
0)
M
3
O
O
u-i
Q
in
t)
C
,
i
Q
O
in
cn
•H
cn
a
CN
C
•H
cn
D
31
-------
REFERENCES
1. Wilson, T. E. "Nitrogen Control by Means of the MID-Step
Biological Nitrification Process." IAWPR Conference on
Nitrogen as a Water Pollutant, Copenhagen, Denmark (August
1975).
2. Wilson, T. E. "Process Designs for Nitrogen Control: NSSD,
Peoria, Tampa." Presented at USEPA Technology Transfer
National Conference on Nitrogen Control, Chicago, Illinois
(July 1976).
3. Newton, D. and Wilson, T. E. "Advanced Waste Treatment at
Tampa, Florida." EPA Technology Transfer Seminar, Atlanta
(October 1971).
4. Newton, D. and Wilson, T. E. "Oxygen Nitrification Process
at Tampa." Applications of Commercial Oxygen to Water and
Wastewater Systems, University of Texas at Austin (November
1972) .
5. Wilson, T. E., Newton, D. and Kapoor, S. K. "Pilot Studies
for Advanced Waste Treatment at Tampa, Florida." 46th
Annual Conference, Water Pollution Control Federation,
Cleveland (October 1973).
6. Bizzarri, R. E., Langdon, P. E. Jr. and Newton, D. "Design
of a Two-Stage Oxygen Activated Sludge System for Tampa,
Florida." Symposium on New Trends in Water and Sewage
Treatment Using Pure Oxygen and Ozone, Denver (October 1974)
7. Bizzarri, R. E., Langdon, P. E. Jr. "60 MGD Advanced Waste
Treatment for Tampa, Florida." 48th Annual Conference,
Water Pollution Control Federation, Miami Beach (October
1975) .
8. Wilson, T. E. and Newton, D. "Brewery Wastes as a Carbon
Source for Denitrification at Tampa, Florida." Preceedings
of 28th Annual Purdue Industrial Waste Conference (May
1973), p. 138.
9. Kapoor, S. K. and Wilson, T. E. "Biological Denitrification
on Deep Bed Filters at Tampa, Florida." Presented at
Second World Congress on Water Resources, New Delhi, India
(December 1975).
10. Wilson, T. E. and Riddell, M. D. R. "Studies and Design for
Advanced Waste Treatment at the North Shore Sanitary
District." Central States Water Pollution Control Asso-
ciation 47th Annual Meeting, St. Paul, Minnesota (May 1974).
32
-------
11. Denniston, W. Jr., Snyder, B. and Lukasik, E. D. "Improve
Plant Performance Through Better Forecasting" presented at
52nd Annual Conference,, CSWPCA, St. Charles, Illinois (May
1979}.
12. Wilson, T. E., Lukasik E. D., and Ogle, D. "Treatment of a
Filamentous Industrial Waste in a Municipal Step Aeration
Plant" presented at the International Association of Water
Pollution Research Workshop on Treatment of Domestic and
Industrial Wastewater in Large Plants, Vienna, Austria
(September 1979).
13. Wilson, T. E., Lukasik, E. D., and Koespsel, W. D. "AWT
Operating Procedures, NSSD" presented at 53rd Annual Meeting,
CSWPCA, Delavan; Wisconsin (May 1980).
33
-------
PERFORMANCE OF FIRST U. S. FULL SCALE BARDENPHO FACILITY
H. David Stensel, Manager Wastewater Technology Development,
Eimco PMD, Envirotech Corporation, Salt Lake City, Utah.
Naohiko Sakakibara, Project Engineer, Sumitomo Jukikai,
Envirotech Incorporated, Tokyo Japan.
David R. Refling, Project Engineer, Glace & Radcliffe, Inc.
Consulting Engineers Winter Park, Florida.
Chuck R. Burdick, Manager Water and Wastes, Reedy Creek
Utilities Company, Inc. Walt Disney World Lake Buena Vista,
Florida.
INTRODUCTION
The first full scale Bardenpho wastewater treatment facility
in the United States has been operating at Palmetto, Florida
since October, 1979. The Bardenpho system, first developed
and applied in South Africa, is an advanced biological treat-
ment system capable of removing BOD, suspended solids,
nitrogen and significant levels of phosphorus without chemi-
cal additions.
The City of Palmetto discharges its treated effluent into
Terra Ceia Bay, which is a small embayment on Tampa Bay on
the west Florida cost. Tampa Bay is located in a large re-
creational area and the bay has limited flushing action and
a relatively warm water temperature. The presence of nitro-
gen and phosphorus can be conducive to the formation of algae
blooms, which has been a problem in the past in the Tampa
Bay area.
In the early 1970's the State of Florida passed the Wilson-
34
-------
Grizzle bill which required advanced treatment for municipal-
ities discharging sewage effluent in the Tampa Bay area.
This has been interpreted as requiring an effluent BOD_ and
suspended solids of 5 mg/£, total nitrogen of 3 mg/£ and
total phosphorus of 1 mg/£. A number of wastewater treatment
alternatives were studied for Palmetto, including spray ir-
rigation, the Bardenpho system with sand filtration, con-
ventional advanced wastewater treatment systems with rotating
biological contactors and conventional advanced wastewater
treatment using a two sludge process for nitrification and
denitrification (1). The Bardenpho system was chosen as the
most cost-effective alternative with the economics signifi-
cantly affected by the chemical cost savings for nitrogen and
phosphorus removal.
BARDENPHO SYSTEM DESCRIPTION
As shown in Figure 1 the Bardenpho process consists of a
fermentation zone followed by four complete mix activated
sludge zones in series and a final clarifier. Settled sludge
from the clarifier is returned to the fermentation zone where
it is mixed with the influent sewage to create an anaerobic
condition. It has been found that an anaerobic stress condi-
tion cause the bacteria to release stored phosphorus and then
assimilate and store greater than normal quantities of phos-
phorus under subsequent aerobic conditions (2). Mixed liquor
flow from the fermentation basin is then directed to the first
35
-------
CC.
X
o
-------
anoxic zone where it is mixed with recycled mixed liquor
from the following nitrification zone. This internal recycle
is normally about four times the influent flow rate. Nitrates
produced in the nitrification zone are reduced to nitrogen
gas as the nitrate is used in the biological oxidation of the
organic material in the sewage. In the nitrification zone
additional BOD removal, ammonia oxidation, phosphorus removal
and sludge stabilization occurs. The second anoxic zone pro-
vides additional denitrification via mixed liquor solids
respiration which decreases nitrate levels further. The mixed
liquor in the reaeration zone will take up phosphorus released
or oxidize ammonia released in the second anoxic zone.
The nitrification zone is designed on the basis of providing
a sufficient solids retention time as a function of temper-
ature for nitrification and sludge stabilization if desired.
The anoxic zones are designed on the basis of using specific
denitrification rates as a function of the total system sludge
age or organic loading, design temperature, and influent nitro-
gen concentration (3). In many cases the nitrification and
anoxic zone design detention times are increased slightly to
provide a total solids retention time that will result in an
aerobically digested sludge for disposal.
Barnard first reported on a four stage system without an
initial fermentation zone to achieve over 90 percent nitrogen
removal in the early 1970's (4). In the course of evaluating
37
-------
the performance of the system he found that phosphorus re-
moval was also occurring in the system. He felt that this
was due to the occurrence of an anaerobic stress condition
on the biological sludge in the system. Barnard later hy-
pothesized that high levels of phosphorus removal could be
expected in a modified Bardenpho system as shown in Figure 1,
by contacting the recycled sludge with the influent sewage
(5). Nicholls further verified that the biological phosphorus
removal mechanism is stimulated by anaerobic stress conditions
by creating an anaerobic zone in front of the full scale
Johannesburg Alexandria extended aeration facility in South
Africa (6). Modification of the Olifantsvlei extended aera-
tion plant in Johannesburg to create an anaerobic zone ahead
of the aeration, also showed that 50 to 80 percent phosphorus
removal could occur (7). In view of these developments a
number of pilot plant tests have occurred and many Bardenpho
facilities have been constructed in South Africa. The most
notable of these is a 40 MGD facility at the Johannesburg
Goudkoppies sewage works which incorporates the initial fer-
mentation zone.
PALMETTO BARDENPHO PLANT DESCRIPTION
Figure 2 shows a schematic of the 1.4 MGD two train Palmetto
Bardenpho plant. Prior to converting to this design, the
Palmetto plant consisted of primary treatment, anaerobic
digestion and trickling filters. The trickling filters were
eliminated and the primary clarifier effluent was directed
38
-------
OL
O
O
u
05
c
01
03
O
-P
-U
OJ
rH
fO
CN
01
^
3
(n
•H
fa
39
-------
to the fermentation zone of the Bardenpho system. In order
to insure that sufficient BOD removal would be available to
create proper anaerobic conditions in the fermentation zone,
design steps were taken to allow the feeding of primary
clarifier underflow sludge to the fermentation zone. This
was done by connecting the primary sludge underflow line to
the return sludge pump pit. Another precaution taken in
design was to by pass a portion of the primary effluent flow
to the first anoxic zone during periods of high flow to the
plant. In this way the high excess flow would not decrease
the anaerobic fermentation time and thus interfere with the
phosphorus removal mechanism. By-pass was designed hydraulic-
ally so that any flow in excess of the average flow of 1.4
MGD would flow into the primary anoxic zone.
The secondary clarifier effluent is directed to two automatic
backwash low head sand filters for polishing prior to chlorine
contacting and discharge. The design return sludge recycle
rate was 100 percent. This is controlled by telescopic
valves in the secondary clarifier. Excess sludge is wasted
from the secondary clarifier underflow directly to drying
beds. Sludge is hauled to a landfill after sufficient drying.
In some cases sludge has been wasted directly to a truck
used for land spreading of the sludge. The drying bed under-
drain flow is directed to the headworks of the plant.
40
-------
A low alkalinity was expected for the wastewater so lime
addition was provided at the front of the plant to maintain
the desired pH level for nitrification.
Tables 1 and 2 summarize the Bardenpho zones detention
times and sizing for the secondary clarifiers and polishing
filters.
Submerged turbine mixers were used to mix the fermentation
and anoxic zones. Because of the small size tanks used for
each of the trains, 5 HP mixers were provided even though
the mixing horsepower requirements were less. It was found
in operation that it was possible to turn off the mixers for
the fermentation zone since the influent flow provided enough
energy to cause circulation of the solids. Two high volume,
low head pumps were provided in each of the trains to recycle
mixed liquor from the nitrification zone to the anoxic zones.
Each pump was rated at four times the influent flow. Normal
operation is with one pump operating and one as a spare. Two
20 HP submerged turbine aerators were used to provide aeration
and mixing in each of the nitrification tanks. Two speed
motors were used to allow a wider range of dissolved oxygen
control during low load conditions. A five HP submerged tur-
bine aerator was also used in each of the reaeration zones.
Conventional center feed, peripheral discharge, center sludge
withdrawal, secondary clarifier units were used with conven-
tional hydraulic loadings. The filter backwash is directed
to the head of the plant.
41
-------
0
-P
4->
(U
C
rH
(T3
a,
C
01
•H
cn
0)
Q
e
0
.p
CO
>1
CO
o
.C
a
a
0)
CQ
2
O
EH
§ 2
W O
-rH rH
C 4J EH
•H CO
X rH rH
•rH CD i
-P 0
G
-------
Table 2
Palmetto, Florida Clarification and Filtration Design
Unit Operation Design Actual Operation
Secondary Clarifiers 2 - 12.2 m (40 ft)
(Diameter)
°'94 °'82
Polishing Filters
43
-------
ACTUAL PLANT OPERATING CONDITIONS
Table 3 compares the actual operating conditions after start-
up to the design conditions. The plant was at 80 percent of
design flow but the influent BOD5 concentration was consider-
ably lower than expected. The influent total nitrogen was
also slightly weaker than expected in design. The diurnal
flow rate varied between 60 and 130 percent of the average
daily flow rate.
In spite of the longer than design detention time, the plant
was still operated to control mixed liquor suspended solids
level on the basis of maintaining a 14 to 20 day solid re-
tention time. Sludge wasting was done on a daily basis by
wasting a certain fraction of the recycled sludge underflow
to the drying beds every morning during the plant operation,
according to Walker's solids retention time control method
(8). The mixed liquor suspended solids averaged around
3500 mg/£ with this solids retention time control.
Excellent sludge settling characteristics were observed as
the average sludge volume index from October to May was 65
mjl/gr. Due to this SVI, a low solids blanket of 0.5 to 1.5 ft
was easily maintained in the secondary clarifier. The sludge
was also found to dewater rapidly on the sludge drying beds
and sludge cracking has been observed in some cases within
18 hours of application.
44
-------
Taole 3
Design Conditions Versus Actual Operating Conditions
(Average Values)
Parameter
Daily Flow (MGD)
Total Detention Time (Hrs)
Influent BOD5 (mg/Ji)
Influent Suspended
Solids (mg/£)
Influent TKN (mg/£)
Influent Phosphorus (mg/£)
MLSS (mgA)
Percent Volatile Suspended
Solids
SVI (m£/gr)
Temperature
pH (Nitrification Zone)
Design
1.4
11.6
270
250
43
14
3500
—
-
18-25°
Actual
1.22
13.3
164
143
32
8.4
3346
70.0
65
19-23° C
6.8
Actual Fraction
Of Design
0.87
1.15
0.60
0.57
0.75
0.60
0.98
—
45
-------
After the plant start-up in late October, operation in
November and December consisted of building the mixed liquor
in the system, and training the plant operators relative
to operational procedures, mixed liquor suspended solids and
dissolved oxygen control methods, and sampling and analytical
procedures. The plant was operated by the City of Palmetto
personnel, who also carried out most of the sampling and
analytical work.
Table 4 shows the operational periods when routine data was
collected. During the early part of January (Phase A) the
internal recycle pumps were off due to mechanical problems.
During this period, the denitrification capability was de-
creased, since the nitrate was not being recycled to the
first anoxic zone. From the middle of January to the end of
the first week in February (Phase B), the plant was operated
according to design. In early February, the lime addition
was terminated as there appeared to be sufficient alkalinity
in the plant effluent.
During Phase B the phosphorus removal was not as high as
desired. This was attributed to the lower than expected BOD^
in the influent. Because of the weaker wastewater, it was a
continuous practice to feed the primary underflow solids
routinely to the fermentation zone via the sludge pump pit.
46
-------
Table 4
Palmetto Plant Operational Periods
Phase Dates Description
A 1/1 to 1/10 Internal recycle
pumps off
B 1/11 to 2/7 Normal operation
per design
C 2/7 to 3/20 Increased anaerobic
contacting - Plugging
problems
D 3/20 to 4/20 Same as Phase C
No plugging problems
47
-------
In Phase C, the fermentation contacting time for the return
sludge was increased by modifying the operation. Due to the
plant piping arrangement, primary underflow and return sludge
from the sludge pump pit could be directed to the head of the
plant and thus to the primary clarifiers. On-off valves
located on the primary clarifier underflow lines were con-
trolled by timers to control the detention time of the sludge
in the primary clarifiers to allow an increase in fermenta-
tion time for the sludge. One clarifier was filling with
sludge, while the other clarifier underflow line was open to
return sludge to the sludge pump pit. During the initial
operation, the time selected for the sludge in the primary
clarifiers was too long so that the increased sludge thicken-
ing resulted in plugging of the primary clarifier underflow
lines. The plugging upset the plant operation by not allow-
ing sufficient mixed liquor solids to get to the Bardenpho
system for phosphorus uptake, nitrogen removal and BOD
removal.
The timer period was then set at three hours which allowed
the sludge to build-up to a depth of about 4-5 feet before
switching the clarifier operations. Thus, the average in-
crease in fermentation time in the primary clarifiers was
1 to 1.5 hours. It was also decided to flush the primary
underflow lines out with a high pressure water hose every
2-3 days to prevent build-up and clogging in these under-
flow lines. This was successful and no plugging problems
occurred during Phase D.
48
-------
BARDENPHO SYSTEM PERFORMANCE
DATA COLLECTION PROCEDURES
Wastewater influent, secondary clarifier effluent and sand
filter effluent samples were composited on an hourly basis
from 7:00 A.M. until 11:00 P.M. on the days selected for
sampling. COD, total suspended solids, pH, ammonia nitrogen,
ortho-phosphorus, and nitrate nitrogen analyses were done on
a daily basis on the influent and secondary clarifier efflu-
ent samples. These analyses were performed on the sand filter
effluent samples two or three days per week. BOD5, total
kjeldhal nitrogen, alkalinity and total phosphorus analyses
were performed on the influent and secondary clarifier effluent
and sand filter effluent samples 2-3 days per week. Mixed
liquor suspended solids, temperature, pH and sludge volume
index analyses, were performed on a daily basis on samples
from the aeration basin. Mixed liquor volatile suspended
solids analyses were performed once per week. Polarographic
dissolved oxygen measurements were made daily on the nitrifi-
cation zone mixed liquor. Grab samples were taken occasion-
ally on the fermentation zone to observe the phosphorus re-
lease to determine if the biological phosphorus removal
mechanism was occurring. All analyses were performed in
accordance with Standard Methods for the Examination of
Water and Wastewater (9).
49
-------
BOD5 AND SUSPENDED SOLIDS REMOVAL
Figure 3 shows that high levels of BOD_ and suspended solids
removal were obtained on a daily basis for the full scale
plant operation. The BOD- concentration was always less than
5 mg/£ for the average daily composite samples, and the sus-
pended solids concentration was usually less than 10 mg/£ ,
and in many cases, less than 5 mg/£ in the secondary clarifier
effluent. Table 5 shows that over 98 percent removal of
suspended solids and BOD were achieved. Because of the high
quality of the secondary clarifier effluent, the sand filters
only removed a small amount of BOD- and suspended solids.
The secondary clarifier performance is significant consider-
ing the average overflow rates of 484 gal/day/sq ft. The
relatively long solids retention time used for the system
was an important factor in providing a low effluent BOD5 .
NITROGEN REMOVAL
Table 6 shows that total effluent nitrogen concentrations
averaged less than 2.5
Figure 4 shows that complete nitrification occurred in the
system during the entire operating period. In most cases,
the effluent ammonia nitrogen concentration was less than
0.5 mg/&. This occurred with an effective solids retention
time of 6 to 8 days for the nitrification zone which is well
above the nitrification washout solids retention time for the
19-23° C operating temperature.
50
-------
2/5UJ) SQI10S Q3QN3dSnS
O
e
5
ss
in
O
33
in
O
'.fl
C
OJ
Oj
to
D
CO
CO
OJ
^1
D
CP
51
-------
Table 5
Bardenpho Average Operating Performance
BOD5 and Suspended Solids
January 1 - April 20
Influent
Secondary
Clarifier Effluent
Sand Filter
Effluent
BOD5
mg/£
164.0
2.5
1.7
% Removal
98.5
98.9
Suspended Solids
mg/£
143.0
4.8
2.0
% Removal
96.6
98.6
52
-------
Table 6
Bardenpho Average Operating Performance
Nitrogen Removal
January 11 - April 20
Influent
Secondary
Clarifier Effl.
Sand Filter
Effluent
Total Nitrogen
(TKN + N03-N)
mg/£ % Removal
32.0
2.3 92.8
2.1 93.4
TKN
mg/£ % Removal
32.0
1.0 96.9
0.8 97.5
NO3-N
mg/£
1.3
1.3
53
-------
o
CM
03
>
O
e
c
a;
a>
o
S-i
-u
3
CT>
NM1
N-CON
54
-------
The average nitrate nitrogen effluent concentration averaged
7.1 mg/& during the first two weeks of January due to the
fact that the internal recycle pumps were not operating.
During normal operation the nitrate nitrogen effluent con-
centration was usually below 2 mg/£. The effluent nitrate
nitrogen did increase to slightly above 2 mg/£ on occasion
during late March and early April when the nitrogen loading
increased. When the primary sludge underflow line was
clogged, the effluent nitrate nitrogen concentration in-
creased to 2-4 mg/£. This data is not shown on the graph
due to the unusual nature of the clogging operation.
On April 15 and April 17 significant rainfall occurred and
the average plant influent flow increased by 50 percent.
This large change in influent flow rate did not significantly
affect the nitrogen removal performance for the system.
The detention time and denitrification rate in the first
anoxic zone was sufficient to accomplish complete reduction
of the nitrate recycled from the nitrification zone. The
second anoxic zone was capable of removing 5-6 mg/& nitrate
nitrogen. Grab samples showed that there was about a 0.3
rag/A nitrate nitrogen concentration increase in the final
aeration zone due to further ammonia oxidation as a result
of ammonia release in the second anoxic zone.
55
-------
PHOSPHORUS REMOVAL
Table 7 compares the phosphorus removal capabilities for this
system during the different phases of operation. Due to the
low solids concentration in the effluent, most of the phos-
phorus in the effluent from the final clarifier or sand
filter was in the form of ortho-phosphorus.
During Phase A the nitrate nitrogen in the plant effluent
and recycled solids averaged around 7.1 mg/5, due to the mal-
functioning of the internal recycle pump. During this period
phosphorus removal averaged only 50 percent. This compares
well with a previous Bardenpho system pilot plant study
reported by Simpson and McLaren (10) which showed that 90
percent phosphorus removal could be achieved only after ob-
taining high levels of nitrate removal. With an effluent
nitrate nitrogen concentration of 6.7 mg/&, only around 55
percent phosphorus removal was obtained. The higher nitrate
concentration also existed in the return sludge, which then
decreased the anaerobic stress conditions in the fermentation
zone. The lower stress condition would then impair the bio-
logical phosphorus removal.
Phase B results indicate that as the nitrate nitrogen con-
centration was decreased in the effluent and return sludge,
the phosphorus removal increased to an average of 59 percent
phosphorus removal.
56
-------
(0
(U
o
C
a
g
o
0)
Oi >
C O
-P OJ
0) W
CX 3
o u
o
(!) x:
en a
(0 u}
M O
0) X!
^ p-l
ex
C
(1)
'O
tO
ffl
(0
CM
O
CN
n
n
(N
u
(M
CQ
dP
0
in
^
00
(N
o>
oo
4J
C
(U
3
rH
MH
C
H
CTi
in
co
O
in
m
0)
•H
a
i-i
o
•P
>i C
M (U
«J 0
t> H
C «H
0 MH
O W
0)
to
CO
CM
(N
CM
CTi
m
rl
0) -P
-P C
rH 0)
•H 3
m
TJ HH
C H
ti
to
57
-------
When the return sludge was recycled through the primary
clarifier tanks in Phase C, the average effluent phosphorus
concentration was 2.2 mg/Si and the percent removal increased
to 72 percent. However, during this operation, clogging of
the primary clarifier underflow line resulted in high effluent
phosphorus concentrations, as well as phosphorus concentra-
tions below 1 mg/£, as shown in Figure 5. During the clogging
period, it was possible for a portion of the Bardenpho system
sludge to be held in the primary clarifier for a period of up
to 6-8 hours. The released phosphorus from this long an-
aerobic period could enter the Bardenpho system via the
primary clarifier overflow. However, due to the clogging,
there were insufficient biological solids in the Bardenpho
system to take up this high level of released phosphorus,
resulting in high effluent phosphorus concentrations.
Effluent phosphorus concentrations were extremely low (less
than 1 mg/&) following the clogging period. This could be
attributed to two possible reasons. The first is that the
sludge had less stored phosphorus than before the clogging
problem, and thus could take up more phosphorus to reach its
previous sludge phosphorus content. The second is that the
longer anaerobic stress period resulted in the sludge develop-
ing a capacity to store a higher percent of phosphorus.
The average influent BOD^ concentrations during Phases C and
D were similar (155 versus 152 mg/£) and the average amount
58
-------
D>
g
03
§
o
a
o
K
12. Ot-
11.0 -
10.0 -
0 CLOGGING
PROBLEM
10 20
FEBRUARY
4.0 -
3.0 -
2.0-
1.0-
Figure 5. Phosphorus Removal
59
-------
of phosphorus removal was similar (5.7 versus 5.6 mg/S,) .
Since Phases C and D showed the same average amount of
phosphorus removal, it appears that the clogging may not
have resulted in a higher phosphorus storage capacity in
the sludge. However, since the 6-8 hour stress condition
was not maintained continuously, no final conclusions can
be drawn concerning the maximum phosphorus content in the
sludge due to extensive anaerobic stress conditions.
Comparison of Phases C and D and Phase B indicates that
returning sludge through the primary clarifier to increase
the anaerobic contact time resulted in greater phosphorus
removal. The average influent wastewater strength in Phases
C and D (153 mg/fc BOD) were also weaker than in Phases A and
B (170 mg/£ BOD).
On April 15 and 17, rain resulted in about a 50 percent in-
crease in the average plant flow. At this time, the influent
phosphorus concentrations decreased,but the effluent concen-
tration decreased only slightly. The phosphorus removal was
only around 3 mg/£ versus around 5.6 mg/£ normally. It
appears that the rainy period resulted in a sufficient dilu-
tion of the influent BOD (122 mg/fc) and reduction of the
anaerobic fermentation time to decrease the biological phos-
phorus removal.
Evidence of high levels of phosphorus uptake was observed by
60
-------
measurements of the phosphorus content of the biological
solids produced in the Bardenpho system. Phosphorus content
of the biological solids was determined on a bi-weekly basis.
The phophorus content averaged between 4-4.5 percent phosphor-
us on a dry weight solids basis.
Figure 5 also shows the importance of dissolved oxygen con-
trol for phosphorus removal. From March 24 to March 29 D.O.
control problems existed and the dissolved oxygen level was
not maintained above 0.5 mg/£. This apparently interfered
with the uptake of phosphorus.
It should be noted that when biological phosphorus was occur-
ring, fermentation zone grab samples showed ortho phosphorus
concentrations in the range of 14-20 mg/Ji. This was well in
excess of the influent phosphorus concentrations. The phos-
phorus release is typical of conditions during biological
phosphorus uptake.
Bardenpho Phosphorus Removal Compared to Conventional Acti-
vated Sludge Operation
During the last six weeks of the study, a small laboratory
activated sludge system was set up and seeded with the mixed
liquor from the Bardenpho system. The laboratory unit was
operated under aerobic conditions only, with a dissolved
oxygen level greater than 2 mg/S,. It was batch fed about
every 2 hours with influent wastewater. Figure 6 shows that
for the first 3 days of operation, phosphorus removal was
61
-------
J
M
«
U
K
JJ
03
>
•H
-4->
o
O (0
•H >
£1 O
o e
^ a;
0) «
tn 3
3 iu
en o
Vi JS
<1J a
> w
C CU
OJ tn
TD T3
SJ 3
03 rH
03 CO
OJ
M
D
cn
'snaOHdSOHd 1V1OJ,
62
-------
similar for the aerobic system and the Bardenpho full scale
facility. However, the effluent phosphorus concentration
increased for the aerobic system and the removal averaged 29
percent. This compared to about 67-70 percent removal for
the full scale facility. This side by side comparison shows
the importance of the fermentation zone on the biological
phosphorus removal.
During the rainy period, the activated sludge aerobic unit
effluent was identical to that for the Bardenpho plant, thus
indicating that no biological phosphorus storage was occurring
in the Bardenpho system during this period.
ALKALINITY AND pH
The pH of the influent wastewater ranged between 7 and 7.4.
When lime was added to the system the pH in the aeration tank
averaged around 7.0. After lime addition was discontinued on
February 7, the pH in the aeration tank averaged around 6.8
for the remainder of the plant operation. Alkalinity measure-
ments taken during early February showed an average influent
alkalinity of around 180 mg/£ as calcium carbonate. The aver-
age effluent alkalinity was around 110 mg/£ as calcium carbon-
ate. The alkalinity is decreased by about 178 mg/£ as calcium
carbonate due to the biological oxidation of about 25 mg/£
total nitrogen. With denitrification returning 3.46 mg/£ per
mg/£ of nitrate nitrogen reduced, about 86 mg/£ of alkalinity
as calcium carbonate is returned. Thus the denitrification
63
-------
helps to maintain an acceptable pH level for nitrification
without lime addition.
SLUDGE DISPOSAL
Operation of biological phosphorus removal systems require
that sludge be disposed of under aerobic conditions to pre-
vent the release of phosphorus back into the system. At the
Palmetto operation, sludge is disposed of daily on sand dry-
ing beds. The sludge properties were such that the sludge
drained rapidly. A small amount of phosphorus release from
the drying beds was observed with phosphorus concentrations
in the range of 5 to 10 mg/i . However, the amount of flow
from the drying beds was such a small amount of the total
plant flow that the amount of phosphorus released did not
affect the overall plant performance.
OPERATION CONSIDERATIONS
The Palmetto wastewater facility has a plant superintendent,
a chemist, four operators and two maintenance men. The
superintendent and some of the employees were also involved
in other projects for the City. Seven people were normally
at the plant between 7:00 A.M. and 4:00 P.M. on Monday
through Friday. Between 3:00 P.M. and 11:00 P.M. one oper-
ator was present at the plant. An operator was also present
at the plant on Saturdays and Sundays between 7:00 A.M and
4:00 P.M. A training period was required to familiarize
plant personnel with activated sludge plant procedures and
64
-------
new laboratory analytical methods.
The operating requirements for the Bardenpho system are very
similar to operating a conventional activiated sludge system.
The important operating parameters are listed as follows:
1. Solids retention time control.
2. Clarifier sludge blanket control.
3. Dissolved oxygen control in the nitrification basin.
4. Internal recycle pump operation.
The solids retention time control is important to maintain the
proper biological solids level in the Bardenpho system to
accomplish the necessary denitrification and nitrification,
as well as BOD removal. This was done at Palmetto by wasting
a certain percent of the return sludge flow to the drying
beds every day.
The clarifier sludge blanket level was observed by using a
tubular device that would sample the clarifier contents and
indicate the depth at which the sludge blanket occurred.
Sludge blanket control was important to minimize the solids
time in the clarifier to minimize the possibility of anaerobic
sludge conditions and subsequent phosphorus release in the
plant effluent.
The dissolved oxygen control in the nitrification basin was
one of the most important control parameters observed in
this study. If the level is too high then too much oxygen
65
-------
is recycled back to the denitrification zone and thus could
affect nitrogen removal. It was common practice at Palmetto
to turn off one of the submerged turbine aerators in each
train during the evening. As the study showed, if the dis-
solved oxygen level was too low it could also affect the
ability of the sludge to take up the phosphorus in the
aeration zone. A dissolved oxygen level of 1-2 mg/£ was
recommended for the Palmetto operation. Dissolved oxygen
control was done manually with measurements taken in the
aeration basin 2-3 times per day.
COST OF PALMETTO BARDENPHO SYSTEM
The bid price for the 1.9 MGD Palmetto Bardenpho system was
1.95 million dollars. Since this plant was already operating
as a trickling filter plant this cost does not include the
cost of the master pump station, the grit removal system,
a small laboratory building and the primary clarifier tanks.
The cost, however, is comparable to that for secondary treat-
ment with a polishing filter. The operating cost for the
facility is estimated to be around $122,000/year, including
salaries and benefits. The chemical costs for alum and
methanol to achieve the same levels of nutrient removal
would be an additional $60,000/year. This does not include
the additional sludge disposal costs.
66
-------
SUMMARY AND CONCLUSIONS
1. High Quality Effluent Achieved
This paper has reported on the first few months of operation
and performance of the Bardenpho facility at Palmetto, Florida,
The operating data showed that a high quality effluent could
be obtained from the secondary clarifier. The BOD^, suspended
solids, total nitrogen and phosphorus averaged less than
3.0 mg/£, 5.0 mg/£, 2.5 mg/£ and 3.0 mg/£ respectively
without any chemical addition. This resulted in over 98
percent nitrogen and 65 percent phosphorus removal. Due to
the low suspended solids concentration from the secondary
clarifier, the polishing filter increased the percent
removals by only 1-2 percent.
The sludge consistently exhibited excellent settling proper-
ties with the SVI averaging 65 mVgr. This helped to ease
operational requirements relative to solids retention time
control and sludge blanket control in the secondary
clarifier.
2. Phosphorus Removal Affected by Wastewater Strength,
Fermentation Time and Nitrate Removal
Phosphorus removal was affected by the ability to develop an
anaerobic stress condition in the fermentation zone. A
separate bench scale aerobic system treating Palmetto
wastewater showed the importance of the fermentation zone
to accomplish high levels of biological phosphorus removal.
67
-------
The full scale plant results showed that the anaerobic stress
condition and phosphorus removal could be diminished by signi-
ficant nitrate nitrogen levels in the return sludge flow.
Results from this study and South African reports (10) support
the need to have maximum nitrate removal in Bardenpho systems
to maximize the phosphorus removal.
The anaerobic stress condition and phosphorus removal can be
affected by the influent BOD_ concentration. Lower BOD con-
centrations experienced at night time could also reduce the
phosphorus removal capability of the system. Increasing the
fermentation zone detention time can improve phosphorus removal
for weaker wastewater conditions. Bardenpho systems should be
designed to handle peak flow conditions without decreasing the
fermentation zone detention time by by-passing the increased
flow to the first anoxic zone.
3. Percent Phosphorus Concentration in Waste Sludge Important
Design Parameter
The percent phosphorus in the waste activated sludge was 4.0 to
4.5 percent on a dry solids basis. Thus the amount of phosphorus
that can be removed in a Bardenpho system should be based on the
influent BOD5 strength, the net solids yield (Ib solids/lb BOD5
removed) and the percent phosphorus in the waste sludge. For a
20-30 day SRT, about 1 mg of phosphorus can be removed per 30 mg
BOD,., at 4.5 percent phosphorus in the sludge. It is not known
at this time how the percent phosphorus in the sludge may be
affected by wastewater strength or further increases in anaerobic
contact time. As the SRT is varied, the net solids production
68
-------
will change. Lower SRT systems should remove more phosphorus
due to a high Sludge production. However, the practicality of
lower SRT systems will be affected by the detention time needed
for nitrification and denitrification, the need for stable
sludge and sludge disposal methods.
4. Phosphorus Removal Similar to South African Pilot Plant
Study
The phosphorus removal results at the Palmetto facility were
very similar to the results observed in the Simpkins and
McLaren Bardenpho pilot plant (10). Weak wastewaters were
present in both cases/ the fermentation zone detention times
were similar, and the BOD removal to phosphorus removal
ratio was about 31:1 in both cases. This was based on the
assumption that the change in COD divided by 1.5 equaled the
change in BOD as only COD data was reported for the pilot
plant. The amount of phosphorus removal was 5.5-5.6 mg/,2, for
this study and 4.9 to 5.6 mg/Jl for the pilot plant work. The
lower removals occurred in the pilot plant when the influent
phosphorus concentration was lower, as the effluent phos-
phorus concentration did not decrease below 0.5 to 0.7 mg/£.
5. Dissolved Oxygen and Solids Retention Time Control Most
Important Operational Parameters
An operating objective was to maintain the dissolved oxygen
level between 1-2 mg/& in the nitrification zone. This
resulted in complete nitrification. Too high a dissolved
oxygen level would not only waste energy, but result in
69
-------
feeding oxygen to the first and second anoxic zones and
interfere with denitrification. A design modification sug-
gested for future systems is to use an oxidation ditch system
(Carrousel System) design for the nitrification. In this
design the withdrawal of the internal recycle and second
anoxic stage feed would be at a point of low or zero dissolved
oxygen concentration. Dissolved oxygen monitoring in the
nitrification zone is important, as too low a dissolved
oxygen level in the nitrification zone was found to prevent
biological phosphorus uptake.
Operation at a selected solids retention time level was
found to be a simple operational technique and resulted in
adequate mixed liquor levels in the system to provide effi-
cient nitrification, denitrification and phosphorus removal.
6. Sludge Disposal of Phosphorus Rich Sludge on Drying Beds
Possible
Sludge disposal on the drying beds did not result in exces-
sive phosphorus release back to the system to deteriorate the
effluent phosphorus concentration. This was likely due to
the good draining characteristics of the sludge which mini-
mized anaerobic conditions and phosphorus release. In some
other cases dissolved air flotation thickening may be de-
sired prior to mechanical dewatering. Gravity thickening
would likely result in excessive phosphorus release from the
sludge.
70
-------
7. Advanced Wastewater Treatment Achieved at Cost Similar
to Secondary Treatment Cost
The cost for the 1.4 MGD Bardenpho facility was similar to
that expested for secondary treatment, yet nutrient removal
was obtained in addition to low effluent BOD and suspended
solids concentrations.
8. Future Design Considerations
The Bardenpho System must be designed specifically for each
application as a function of the wastewater influent concen-
trations of BOD , suspended solids, nitrogen and phosphorus
and wastewater temperature. Low effluent total nitrogen
levels can be achieved, provided sufficient BOD exist in the
influent and proper design nitrification and denitrification
rates are used. The maximum phosphorus removal obtained will
be dependent on the influent BOD concentration, fermentation
zone design and solids retention time. For weak wastewaters
or wastewaters with relatively high phosphorus concentrations,
some use of chemicals for effluent polishing may be necessary
to reduce phosphorus concentrations to less than 1 mg/£.
Other designs for the fermentation and nitrification zones
could improve operational reliability and performance.
Acknowledgements
The results of this study could not have been possible with-
out the personal dedication and quality effort of the plant
superintendent, Don White, and the plant personnel. At the
time of this study David R. Refling was a Senior Research
71
-------
Engineer with Eimco PMD, Envirotech Corporation, Salt Lake
City, Utah, and Chuck Burdick was a Project Engineer with
Glace and Radcliffe, Inc., Consulting Engineers, Winter Park,
Florida. Dr. Stensel and Mr. Sakakibara's afflilations
were as shown on the paper heading.
72
-------
References Consulted
1. Burdick, C. R. and Dallaire, G., "Florida Sewage Plant
First to Remove Nutrients with Bacteria Alone - No Need
For Costly Chemicals" Civil Engineering - ASCE, p. 51,
October 1978.
2. Nicholls, H. A. and Osborn, D. W., "Bacterial Stress:
Prerequisite For Biological Removal of Phosphorus"
Journal Water Pollution Control Federation, p. 557,
Vol. 51, No. 3, March 1979.
3. Refling, D. R. and Stensel, H. D., "A Rational Approach
to Biological Nitrogen Removal" Presented at the Annual
Water Pollution Control Federation Conference, Anaheim,
California, October 1978.
4. Barnard, J. C., "Biological Denitrification" Journal
International Water Pollution Control Federation,
72, 6, 1973.
5. Barnard, J. C., "Cut P and N Without Chemicals" Water
and Water Engineering, 11, 7, 1974.
6. Nicholls, H. A., "Full Scale Experimentation on the New
New Johannesburg Extended Aeration Plants" Presented
at the workshop on New Aspects of Biological Treatment,
Pretoria, South Africa, February 24-26, 1975.
7. Venter, S. L. V, Halliday, J. and Pitman, A. R.,
"Optimization of the Johannesburg Olifantsvlei Extended
Aeration Plant For Phosphorus Removal" Progress in Water
Technology, 10, 279 (1978).
8. Walker, L. F., "Hydraulically Controlling Solids
Retention Time in the Activated Sludge Process" Journal
Water Pollution Control Federation, 43, 30, 1971.
9. Standard Methods For the Examination of Water and
Wastewater,13th Edition, American Public Health
Association, Washington, D. C. (1971).
10. Simpkins, M. J. and McLaren, A. R., "Consistent
Biological Phosphorus and Nitrate Removal in an
Activated Sludge Plant" Progress in Water Technology,
10, 33 (1978).
73
-------
DENITRIFICATION IN CONTINUOUS-FLOW SEQUENTIALLY AERATED
ACTIVATED SLUDGE SYSTEM, AND BATCH PROCESSES
by
Mervyn C. Goronszy
Senior Investigation Engineer
State Pollution Control Commission
Robert L. Irvine
Professor
University of Notre Dame
APPROACH
This paper is divided into two distinct but interrelating parts.
The first part deals with sequentially aerated activated sludge
systems in Australia; the second, with a full scale United States
Environmental Protection Agency demonstration study on batch pro-
cesses. In both systems discharge is interrupted while treatment
and sedimentation take place. The paper describes the general
operation of each system along with strategies employed in each
for nutrient removal.
CONTINUOUS FLOW SEQUENTIALLY AERATED ACTIVATED SLUDGE SYSTEMS
Reasons traditionally offered for controlling the content of
nitrogen in treated wastewater discharges include the high oxygen
demand on receiving waters exerted by ammonia, the increase in
chlorine break point requirements and contact time for adequate
disinfection in the presence of ammonia, the toxicity of ammonia
to fish, the serious health hazards of high concentrations of
74
-------
nitrate nitrogen in drinking water which can lead to infantile
methaemoglobineaemia and the" problem of eutrophicration which is
associated with receiving water offering limited dilution to
wastewater discharges containing high concentrations of nitrogen.
A further reason relates to the biological treatment of waste-
waters having a low reserve alkalinity in which case denitrifi-
cation procedures may be used to enhance the production of an
activated sludge having favourable settlement properties.
The biological nitrification-denitrification process is perhaps
the most common scheme for removing nitrogen from municipal
wastewaters in use today. In this method of treatment, following
the release of organically bound nitrogen as ammonia by hetero-
trophic bacteria during the organic removal stage, ammonia is
oxidized primarily to nitrate by autotrophic bacteria. Nitrate
conversion may then take place through either of two mechanisms,
assimilatory or dissimilatory denitrification. In assimilatory
denitrification, nitrate is reduced to ammonia which is used in
cell synthesis. In dissimilatory denitrification, nitrate serves
as the hydrogen acceptor in the oxidation reduction reactions of
the carbon substrate to provide energy for cell growth and is
converted to gaseous end products, principally nitrogen, by
heterotrophic bacteria. Biological denitrification is achieved
under anaerobic or near anaerobic (anoxic) conditions; a wide
variety of common facultative bacteris such as Pseudomonas sp,
Micrococcus sp, Achromobacter sp, Denitrobacillus sp, Spirillum sp,
and Baccillus sp have been reported to accomplish denitrification
(1).
75
-------
The principle factors which affect nitrification in the activated
sludge process include sludge age or solids retention time, pH
value, alkalinity, temperature, dissolved oxygen concentration,
and the presence of toxic substances including heavy metals (2,
3,4,5,6). These factors, together with the carbonaceous energy
source, also affect denitrification. It is generally believed
that assimilative nitrate reduction is uninhibited by oxygen
concentration (7,8). Reports of denitrification under aerobic
conditions may be explained in terms of a dissolved oxygen gradient
across bacterial floes with active denitrification the product of
bacteria not directly exposed to dissolved oxygen (7). This is
further supported by the work of Mueller et. al. (9) who devel-
oped relationships between floe size and the dissolved oxygen
concentration necessary to render the floe completely aerobic.
For a floe size of 115 urn their results showed that at an oxygen
uptake rate of 80 mg/l/hr, a dissolved oxygen concentration of
6 mg/l was required to maintain an aerobic floe, while an uptake
rate of 4 mg/l/hr required a dissolved oxygen concentration of
0.6 mg/l.
Present day biological nitrigication-denitrification process are
differentiated as combined or separate sludge systems or in terms
of the carbon sources, provided for denitrification. Internal
carbon sources, from raw sewage and endogeneous respiration, and
external carbon sources, such as methanol, may be identified. The
early bench scale experiments of Ludzack and Ettinger (10) have
76
-------
seen the development of a number of combined sludge systems, as
distinct from separate and attached sludge, 'capable of removing
80-90% of influent nitrogen through biological nitrification-de-
nitrification without the use of supplemental chemicals. Pasveer
(11) was one of the first to suggest the possibility of combined
sludge carbon and nitrogen removal in early work on oxiditation
ditches. Since then variations to this combined sludge approach
for nitrification and denitrification have received considerable
attention (12, 13,14,15,16,17).
The achievement of denitrification in activated sludge systems
without the addition of chemicals requires the use of zones of
low oxygen tension, with sufficient overall oxygenation capacity
within the system for full nitrification. Some early attempts
which used raw sewage as a carbon source suffered loss of effi-
ciency because of degradation of effluent quality due to uncon-
verted organic matter and ammonia nitrogen. Separate sludge
systems incorporating supplemental carbon addition, such as
methanol (18,19) have been exhaustively researched over recent
years. These systems appear to be disadvantaged by current
circumstances, because of high capital and operating costs.
This paper includes details of sequentially aerated activated
sludge systems and reports on operational procedures which
maximize nitrogen removal eithout the use of supplemental chemi-
cals. Plant performance and associated kinetic data are shown.
77
-------
Capital cost of these plants are compared with conventional
continuous systems sized on similar criteria.
SYSTEM DESCRIPTION
Continuous flow sequentially aerated activated sludge systems
have been used in Australia since 1965, initial development was
for small community municipal wastewater treatment (17,20).
Early plants were designed as shallow endless channel units
(1.9 m deep) and were mixed and aerated by floating brush aera-
tors. Current plants are designed as deep (up to 4 m deep) rec-
tangular vessels and employ either submerged jet or floating
vertical shaft mechanical aeration. These plants marry contin-
uous activated sludge technology with intermittent system operation
to yield a total 'continuous-intermittent' process. System oper-
ation is 'on-off to enable carbonaceous oxidation, nitrification,
and denitrification and to effect settlement of mixed liquor
suspended solids and removal of treated effluent, without inter-
ruption to the inflow of wastewater, all of which takes place
within a single vessel. Wastewaters are received and treated
without separate equalization.
The operating sequence for "continuous intermittent treatment" is
shown schematically in Figure 1. The spacial arrangement of unit
operations and processes of conventional continuous activated
sludge treatment is incorporated as a cycle of timed sequences
(tfl -t1, t.. -t_, t2 ~t3). Each sequence is initiated by a
78
-------
u ni H
^ t U_ LU LU UJ UJ
h- ^ LLJ < c/> o cr
II II II II II II II
^ M UJ < CO Q CC
a
3
w
I
H
Oi
o
H
D
O
79
-------
variable time selector process controller, to enable simultaneous
load balancing, biological processes and settlement to occur
within the single vessel. Following settlement (t.. - t_) the
final effluent is discharged by means of a moving weir arrangement
during the period (t_ - t,), aeration is provided from t_ termin-
ating at t.,, settlement follows during (t.. — t_) . The total cycle
of sequences is repeated. Operational sequences have been dev-
eloped to maximize either nitrification or denitrification by
incorporating different cycle times. Denitrification processes
require that there is sufficient aeration to provide for total
carbonaceous and nitrogeneous treatment within a modified (t- -
t.) period, but extend the period (t, - t_) to effect the reduction
of nitrate ions. The determination of a time scale for each of
the sequences requires optimization of a number of independent
factors:
•Hydraulic considerations determine the duration and
frequency of the decant period, t2 - to, and therefore
dominate the duration of the total cycle, tQ - t-.
•The volumetric discharge rate associated with the decant
period, t-2 - t3 is critical and must not cause settled
solids to be entrained in the effluent.
•The aeration period, tg - t]_, must be of sufficient
duration to accomodate the requisite process oxygen
at an economic rate of power usage.
•Vessel geometry must be compatible with the process
requirements for all of the sequences, tn - t1, t. - t,,
+•—•+• Oil*
t2 t3.
• The mass of biological solids must be sufficient to achieve
the biological processes and have an adequate solids flux
at maximum solids concentration to permit efficient solids-
liquor separation for design decant rate.
80
-------
••System sizing requires that biological solids activity
be such that settled solids do not rise during non-
aeration periods.
Sequence durations used in the treatment of municipal wastes to
achieve nitrification or denitrification are shown in Table 1.
Basic design criteria and sizing of rectangular configurations
are presented in Table 2. In practice systems are sized to
accomodate a flow variation of about seven times dry weather
flow (peak wet weather flow) without appreciable loss of effluent
quality. Accommodation of wet weather flow is a most important
self regulating feature of this form of sequential treatment
whereby these high flows are accepted, for extended periods,
without washing out the mixed liquor solids, a factor which
would otherwise lead to a subsequent loss of process efficiency.
In operation the high flow sequence provides 14.4 decants, 12
hours of aeration and 7.2 hours of decant, in every 24 hours.
SYSTEM PERFORMANCE
Many bacteria in sewage and sewage sludge are capable of respi-
ratory nitrate reduction. The continuous feed sequential mode
of operation of the activated sludge process can provide the
prerequisites for denitrification, i.e., the presence of nitrate,
an electron donor (carbon source) and anaerobiosis. Sequential
operations, related to the duration of the aeration period (t. -
t.) may be selected to favor either nitrification or denitrification,
Nitrogen loss is favoured by long anoxic periods (relatively) or
81
-------
TABLE 1. TYPICAL SEQUENCE DURATION FOR TREATMENT
OF MUNICIPAL WASTES
- DRY WEATHER FLOW CONDITIONS -
CONFIGURATION AERATION SETTLEMENT DECANT
MIN MIN MEN
ENDLESS OffiNNEL
. NlTKii'lCATICN 1 270 60 30
. DENITRIFICATION 1 180 150 30
. NITRIFICATION 2 150 60 30
. DENITRIFICATION 2 120 90 30
RECTANGULAR, NON-BAFFLED INLET
. DENITRIFICATION 90 60 30
RECTANGULAR, BAFFLED INLET
. DENITRIFICATION 1 270 60 30
. DENITRIFICATION 2 150 60 30
. DENITRIFICATION 3 150 45 45
- WET WEATHER FLOW CONDITIONS -
ALL CONFIGURATIONS 50 20 30
82
-------
• > 0, -H
• \ Qa CU Q) \
Q) a a> cu \ a»
o ^ > > * 5
o t^ o in o
o o o vo o o
- o n« co o • o
T r» (N -«3« T— o m
_ £ o in oo
O O Cn CN (N
vo m o m i-
rn r- ro T- »- ro
«- oo
in ' ' g
in • in
«- VO CN
CM
CO VO
(N
Cn
o
Z
H
N
H
co
a
z
a
w
a
u
CO
z
o
EL<
Z
o
u
u
to
to
a, -o
0)
in
o
o
o
in
s, f |
I
a aT
OJ \
in ^«
o o vo o
in o m •
(N m T- o
.. sr s? .as
• >i (0 (0 'O
£X (0 t3 TS \
0) v. ?£ "^ S*-
_ v. co
E S > e
in S
CN fM O
oj • in in
m vo en rr
I
i
CO
8 ^
Cb ^
T- VO
a,
a>
o
o
m
8-
o
in
co
u
a
u
H
CO
<
03
W
a
Q)
S
O
VO
>, fl1 fr
_ig ^O ^D
a:
S)
D> Orrofl SSoocn
oocooo'Do omr*
o^oovo •ov.cooooo •
r^(N^"«— oinoor— cp»ror~»von
CO
2 •*• ' ' £
O • vo
-------
continued operation at low dissolved oxygen concentrations during
the aeration sequence in order to maximize anaerobic conditions
within floes. The effectiveness of the latter is governed by
floe size and the oxygen uptake rate of the biological solids.
Denitrification can therefore be effective over the duration of
the whole sequence (tQ - t^) by manipulation of the (-t. - t-)
period in relation to the (tQ - t..) period to effect the minimal
dissolved oxygen concentrations brought about by a temporary
imbalance of oxygen demand and oxygen supply. It is important,
however, to exceed the oxygen demand at some stage, preferably
at the completion of the aeration sequence, to render the floe
completely aerobic. Continued operation with a significant
anaerobic floe fraction may lead to sludge bulking, poor solids
settlement and subsequent process failure. The extent to which
anoxic conditions exist in relation to partial oxic conditions
has been found to be important to the maintenance of a biological
floe having good settlement properties. Inflow during non-aeration
(t, - t3) serves to increase the relative substrate to biomass
concentration of the biomass in the inlet contact zone. This
phenomena is more pronounced in baffled inlet systems, but is
also apparent in non-baffled systems, but to a lesser degree.
Biomass oxygen utilization rate of this floe fraction is signi-
ficantly higher than the remainder and is indicative of a high
rate assimilation mechanism. The longer anoxic sequences, com-
bined with the baffled inlet sections all serve to decrease the
degree of longitudinal mixing in these systems. Such an obser-
84
-------
vation is in accord with separate investigations which indicated
that activated sludge units with a low dispersion number (plug
flow) exhibit better solids settlement characteristics than units
with a higher dispersion number (completely mixed) and that anoxic
zones can improve the settlement characteristics than units with
a higher dispersion number (completely mixed) and that anoxic
zones can improve the settlement characteristics of activated
sludges (21). The presence of anoxic sequences tends to increase
the concentration of carbon dioxide in the system. Evidence has
been reported to suggest that some micro-organisms which exhibit
poor settlement properties can be favoured by a low carbon dioxide
concentration (22).
Dissolved oxygen profiles associated with nitrification and de-
nitrification using a six hour sequence are shown in Figure 2.
Several other sequences of various duration have been investi-
gated (23) for which similar performances were observed in those
cases of equivalent mass transfer of oxygen. In practical terms
shorter aeration sequences require a greater rate of oxygen
transfer to effect equivalent levels of treatment at the same
organic loading rate. Figure 3 shows typical biological solids
oxygen uptake rates associated with inlet and aeration sections
in baffled and non-baffled deep rectangular tanks at the onset
of and during aeration sequences. A denitrifying dissolved
oxygen profile is applied to the main aeration sections of both
systems.
85
-------
o
u
I
Q
LU
1 2
NITRIFYING CYCLE
TIME. HOURS
1 2
DENITRIFYING CYCLE
i. 5
TIME. HOURS
AERATION SEQUENCE kV\\M ANOXIC/SETTLEMENT SEQUENCE
DECANT SEQUENCE
FIGURE 2: DISSOLVED OXYGEN CONCENTRATION DURING NITRIFYING AND
DENITRIFYING CYCLES FOR CONTINUOUS FEED SEQUENTIALLY
OPERATED TREATMENT SYSTEMS
86
-------
JZ
(ft
to
_J
2
o>
0>
Ul
cr
LU
0.
X
o
30
20
10
© INFLUENT END
x EFFLUENT END
AERATION
NON AERATION
INLET ZONE
0 50 100
TIME MINS
OXYGEN UPTAKE RATE FOR 4000 ep NON-BAFFLED
DEEP TANK I23°C, OVERALL F/M 0-11 75 MINS AERATION SEQUENCE)
8 30
o
o>
u
*
OL
LU
Z
LU
O
X
o
20
10
o STAGE 1 INLET ZONE
x STAGE 2 AERATION SECTION
e
0 100 200
TIME MINS
OXYGEN UPTAKE RATE FOR 800 ep BAFFLED
DEEP TANK (220C, OVERALL F/M 0-024 270 MINS AERATION SEQUENCE)
FIGURE 3: OXYGEN UPTAKE RATES FOR SEQUENTIALLY
OPERATED ACTIVATED SLUDGE
87
-------
In the two examples shown the rate of oxygen uptake for the non-
baffled deep rectangular system during aeration was approximately
13 mg O./g MLSS hr (23°C). However, in the partially mixed inlet
zone during non-aeration the measured rate of oxygen uptake reached
a 2.3 times greater rate than the mean aerobic value. The rate
of 30 mg 0_/g MLSS hr can be assumed to represent the activity
associated with an adsorption oxidation process under conditions
of low oxygen tension. Similar and more definitive results were
observed in the inlet sections of the baffled sequentially oper-
ated deep tank system. The initial oxygen uptake rate of 21 mg
0-/g MLSS hr, at 22°C, decreased with time and after 2-3 hours
(the approximate retention time in the first stage) approached
the rate observed in the second stage of the plant, 8 mg 0 /g
MLSS hr. Overall mean P/M ratios were 0.11 and 0.024 for the
non-baffled and baffled systems, respectively. Denitrification
was effective in both instances and was in excess of 80% removal.
The results of nitrogen removal associated with various nitrifying
and denitrifying dissolved oxygen profiles, are presented in
Table 3. The data applies to six hour sequential operation al-
though similar removals have been obtained for four and three
hour sequential operations (20). Measured rates of denitrifi-
cation using raw municipal wastewater as the only added carbon
source are shown in Table 4. Associated biological sludge ac-
tivity, measured as specific oxygen uptake rate, is also shown.
Denitrification rates compare favourably with several tabulated
88
-------
TABLE 3.
INFLUENT
SUMMARY OF DENITRIFYING SEQUENTIAL OPERATION
STATISTICS - 6 HR CYCLE INCLUDING 60 MINS
SETTLEMENT AND 30 MINS DECANT
TOTAL NITROGEN
X - 51.9
S » 16.5
n = 92
EFFLUENT
AERATION SEQUENCE TOTAL NITROGEN
OXIDISED NITROGEN MEAN NITROGEN
REDUCTION PER CENT
120 MINS
154 MINS
101 MINS
172 MINS
X
S
n
X
S
n
X
S
n
X
S
n
- 8.6
= 2.6
- 12
= 5.2
= 1.7
- 35
= 4.2
= 1.7
= 11
- 18.4
= 4.5
• 14
X
S
n
X
S
n
X
S
n
X
S
n
» 5.8
= 2.4
- 12
- 2.5
- -1.3
= 36
» 1.6
= 0.8
» 11
= 16.8
= 4.9
= 14
83.4
90.0
91.9
64.5
X = mean
S = standard deviation
n = number of samples, 2 samples per week
89
-------
TABLE 4. DENITRIFICATION RATES IN SEQUENTIALLY
OPERATED ACTIVATED SLUDGE
MLSS
ng/1
3360
3410
3660
6210
6380
5810
2890
4480
7650
8640
6800
4070
2145
7870
% SEWAGE
10
10
25
10
10
25
10
10
10
10
10
10
10
10
TEMPERATURE
17
17
20
23
23
20
-
22
23
27
23
27
25
30
27
25
28
BCD /N
(initial)
_
_
_
_
_
-
1.0
1.62
1.13
1.07
1.73
1.48
1.65
CUR
mgO-Xg MI
(init
—
—
-
—
-
—
6.7
11.4
8.9
10.2
5.6
18.0
9.3
DENTERIFICATICN
0.51
0.46
0.25
0.31
0.23
0.35
0.83
0.98
0.50
0.68
1.66
1.91
0.45
0.61
0.74
0.27
0.30
90
-------
by Christensen and Harremoes (24) for a number of suspended
growth, combined sludge internal carbon source systems. The
rate is influenced by temperature but is independent of initial
oxidised nitrogen concentration. Maximum and minimum values
quoted by these authors, where raw sewage was used as the carbon
source, were 2-3 and 0.2-0.3 mg N/g MLSS hr (20°C) respectively
Barnard (13) has reported denitrification rates of 1.3-1.6 mg
N03-N/gm MLSS hr at 20°C. Typical effluent quality for sequen-
tially operated denitrifying systems is shown in Figure 4.
In operation nitrification/denitrification efficiency is not
markedly affected by temperature even at a low vessel temperature
of around 7°C. According to Barnard (13) the minimum sludge age
required to complete nitrification at this temperature is of the
order of seventeen days, a figure which is well and truly exceeded
in practice. Combined with the fact that the rate of denitrifi-
cation exceeds the rate of nitrification, such a performance is
not therefore totally unexpected. Temperatures below 10°C general-
ly reduced total nitrogen removal efficiencies by less than ten
percent.
IMPLICATIONS OF DENITRIFICATIQN PERFORMANCE
The capacity of continuous feed sequentially operated activated
sludge systems to achieve efficient and substantial denitrification
is an important feature of the system of treatment. Although the
rate of denitrification/mass of mixed liquor is slow for sequentially
91
-------
50
20
- 10
o>
E
u.
u.
01
BOO
20
FILTRABLE
RESIDUE
10 50 90
PERCENT EQUAL TO OR LESS THAN
ITAL
B NITROGEN
99
FIGURE 4: EFFLUENT QUALITY FOR SEQUENTIALLY OPERATED
4000 ep PLANT
92
-------
operated systems relative to methanol fed separate sludge systems
(by a factor of 10) the time required for effective denitrification
is comparable for the two -systems. This is feasible because of
the large mass of biological solids held in the system and the
available hydraulic capacity of the system. In a conventional
separate sludge system using methanol the biological solids con-
centration must not exceed 3000 mg/1 to meet the requirements of
effluent clarification. Also it is usual to include a safety
factor to account for the effects of diurnal load variation, the
safety factor can be taken as the ratio of peak to average loads
and a factor of 2 is often appropriate. For a sludge with a rate
of denitrification of 12 g NO -N/g MLSS d at 25°C an anoxic reten-
tion time of about 2 hours is necessary to denitrify 25 mg/1 of
nitrate.
In continuous feed sequential systems the total sludge mass is
available for denitrification. The solids flux is substantially
lower than the equivalent loading on a separate clarifier and
therefore MLSS concentrations in excess of 5000 mg/1 can be
retained within the plant. For a municipal wastewater with a
BOD_ of 70 g/capita d a flow of 240/1 capita d and 6 g of nitrogen/
capita d (25 mg/1), to be treated in a continuous feed sequentially
aerated plant designed with an F/M ration of 0.05, the mass of
mixed liquor per head is 1400 g. For a denitrification rate of
0.02 kg NO^-N/kg MLSS d a total of 28 g of oxidised nitrogen can
be denitrified every 24 hours. Therefore a plant operating with
93
-------
four 6 hour cycles and containing a non-aeration period of 6 hours
per day can denitrify the required mass of oxidised nitrogen (6 g
per capita). A plant with a lower applied organic loading will
have an excess capacity to remove nitrogen.
SYSTEM ECONOMICS
Continuous feed sequentially aerated systems require about five
hours of semi-skilled operational involvement per week. Power
usage for operation on a six hour nitrifying cycle is typically
0.1- 0.13 kw/capita d for systems treating the wastes from 10,000
persons. Marginally lower costs can be obtained as a result of
reduced aeration arising from denitrification, some 10-25% power
savings are possible. Figure 5 shows the relative capital cost
of continuous flow sequentially aerated systems when compared with
continuous conventional systems sized on equivalent organic load-
ing parameters (25), In applying this data it should be remembered
that local factors and exact process design have a significant
effect on the relative costs of an installation. An approximate
breakdown of costs apportions 53% to aeration tank, 28-32% to
aeration, 15-18% to effluent withdrawal mechanism and controls
for plants around 240 m /day capacity. For plants around 5300 m /
day capacity the cost proportions are around 68,18,13 percent
for aeration tank, aeration and effluent withdrawal mechanism
and controls, respectively.
94
-------
1-0.
0-5 -
I I
1000 _
PLANT CAPACITY M^/DAY (ADWF)
10000
FIGURE 5: FRACTIONAL COST OF SEQUENTIAL/CONTINUOUS
CONVENTIONAL ACTIVATED SLUDGE TREATMENT OF
MUNICIPAL WASTES
95
-------
PROCESS CONTROL
Mixed liquor suspended solids determinations are needed from time
to time to check on the mass and volume of sludae held in the
system. Samples are taken at bottom water level under fully
mixed conditions. This test also serves to check on the sludge
wasting programme, which is automatically initiated within each
aeration sequence. A relatively small volume of mixed liquor
is removed, using a positive displacement pump, equal to the
daily accumulation of biomass in order to maintain the required
mass of solids under aeration. Removed mixed liquor is pumped
to sludge lagoons and displaced supernatant liquors returned to
the aeration vessel. Comprehensive analyses are needed from
time to time to check on overall efficiency and to determine
the need for increasing or decreasing the aeration sequence this
being dependent on the residual ammonia nitrogen concentration.
Sufficient process oxygen is provided when the ammonia nitrogen
concentration is determined at or near nil. Regulation of the
duration of the aeration sequence offers a simple and efficient
means for providing sufficient process aeration for loadings
less than design. Under these circumstances start-up procedures
can invariably be simplified and operating economies achieved.
Nitrification of wastes with a low associated alkalinity often
results in the production of a sludge of high settled volume and
hence high sludge volume index brought about by filamentous
96
-------
growth, due mostly to a reduction of the system pH (26). Control
to effect denitrification can reverse this situation due to the
production of hydroxyl ions and the maintenance of pH. Table 5
shows the case of a poor effluent quality caused by a sludge of
poor settling characteristics leading to solids entrainment during
effluent decant. Introduction of denitrification procedures led
to improved solids settlement and cessation of solids entrain-
ment after only three weeks. Similar nitrogen removal is achieved
with both high and low alkalinity wastes.
CONCLUSIONS
•Continuous flow sequentially aerated activated sludge
systems present a simple method for maintaining aerobic-
anoxic zones under varying plant flow and loading without
the need for sophisticated process controls.
•Anoxic conditions generated between aeration sequences
contribute to mean total nitrogen removals in excess of
eighty five percent for widely varying operating conditions
normally associated with municipal wastewater treatment.
•Current sizing of these systems enable high nitrogen removal
to be maintained at temperatures below 10°C. At these
temperatures stated efficiency is only reduced by about
ten per cent.
• Operation of continuous feed sequentially aerated activated
sludge systems for effective denitrification maintains
a balance of pH where wastewaters contain a low reserve of
alkalinity and therefore provides an effective method for
maintaining an activated sludge of an acceptable settle-
ability.
•Continuous flow sequentially operated activated sludge
systems are simple and straightforward to operate re-
quiring only semi-technical operational skills.
•The level of nitrogen removal in systems sized to accept
volumes up to about 10,000 m^/day may be achieved at
significantly lower capital and operating cost when
97
-------
TABLE 5. EFFECT OF DENITRIFICATION SEQUENCE
ON SLUDGE SETTLED VOLUME
DAYS
MLSS itg/1
EFFLUENT
NFR mg/1
NH.-N mg/1
A
Org-N mg/1
NOx-N mg/1
ALKALINITY
as Ca CO, mg/1
PH
1
3630
950
0.4
52.2
33.8
56
6.2
7
3980
32
Nil
3.5
3.1
145
7.2
13
3700
9
0.8
4.1
2.9
164
7.2
20
3050
7
Nil
2.5
5.3
167
7.4
MIXED LIQUOR
(1 HOUR SETTLEMENT)
PER CENT SETTLED
VOLUME
SVT
74
204
68
171
57
254
33
108
(k HOUR SETTLEMENT)
PER CENT .SKTI-I/RD
VOLUME
SVI
88
242
86
261
75
203
42
138
98
-------
compared to other nitrogen removal processes. The system
can also be constructed in modules in accordance with de-
mand and is therefore a viable alternative to other conven-
tional methods of treatment.
THE BATCH PROCESSES: SBRs
Because of the need to operate minimal maintenance facilities in
many locations in Australia, experience with the continuous-flow
sequentially aerated activated system has been restricted pri-
marily to low loadings and reasonably infrequent sludge wastings.
In these systems, biological nitrogen removal is conveniently
promoted by alternating aerobic and anoxic conditions. On the
other hand, biological phosphorous removal has not been reported.
In contrast, the Sequencing Batch Reactors (SBRs) developed in
the United States are designed to operate over a wider range of
loadings and sludge wasting (27,28,29,30). As a results, both
biological nitrogen and phosphorous removal are possible (31,32,
33,34). In addition, because SBR operation calls for the inter-
ruption of raw waste flow into each reactor once each cycle,
discharge and inflow do not normally overlap. This also results
in the final major difference between the two systems: in the
SBR, after the raw waste inflow is diverted, aerobic, anoxic and
anaerobic reactions can be regulated as needed in a mixed batch
reactor with sedimentation delayed until the desired reactions
are completed; in the Australian mode of operation, sedimentation
commences as soon as the reactor liquid volume reaches a predeter-
99
-------
mined level with the liquid level continuing to increase through
sedimentation until discharge.
Figure 6 illustrates the five major periods (i.e., FILL, REACT,
SETTLE, DRAW and IDLE) in one SBR cycle (27) . In way of compari-
son with continuously operated reactors, the intermittent systems
retain the sludge until wasting is desired while conventional
suspended growth systems with external clarification continuously
discharges solids only to return that which is required for proper
operation.
Virtually all suspended growth biological waste treatment facilities
in the United States are of the continuous flow variety. Because
the SBR offers the potential for control of municipal discharges
in an efficient and cost effective manner, the United States Envi-
ronmental Protection Agency funded a two year demonstration project
to investigate SBRs in terms of operation, design and control.
The project is directed by the University of Notre Dame and gen-
erally involves the conversion of a conventional municipal acti-
vated sludge plant in Culver, Indiana to a two tank SBR system.
The details of the conversion and the strategies developed for
the microprocessor control systems are described below.
PRECONVERSION OPERATION AT CULVER
After passing through a comminutor and grit chamber, the raw
waste flow enters two of the three primary tanks before being
100
-------
O
o
10
3iAiniOA amoin
o
IO
o'
IO
CVJ
d
o
o
10
Is-;
d
o
LU
_J
>
U_
O
CJ
<
cc
10
CD
u
H
o
04
OS
O
6*
U
a;
w
z
o
OS
o
fa
w
en
o
Q
H
0
H
i-3l
vo
Pi
D
U
H
jo
101
-------
aerated in one of two existing activated sludge basins. Final
sedimentation (two tanks) and chlorination are provided before
discharge. Principal design data (e.g. flow, tank volume, etc.)
are given in Table 6. A summary of preconversion operating data
are presented in Table 7. As can be seen from the data, the raw
waste is relatively weak and the final effluent limitations of
3 3
10 g/m 5-day Biological Oxygen Demand (BOD ), 10 'g/m suspended
solid (SS) and 1 g/m phosphorous were met consistently. Ferric
chloride is added for phosphorous removal. There are no effluent
limitation on nitrogen. The raw waste, however, contains an un-
usually high concentration of oxidized nitrogen (approximately
9 g/m3 as N).
The Culver treatment facility was designed for an average flow of
2592 m /d. The actual average flow is less than half that value.
As a result, only one of the two aeration tanks were used. Com-
pressed air was supplied through coarse bubble diffusers with
one 45 KW blower dedicated to each tank. The air flow rate was
throttled to the minimum (approximately 0.3 m /s) in an attempt
to limit the air supply. Even so, excess air was wasted through
the aeration tank not receiving raw waste. As a result of the
low loaded -conditions, the settling characteristics of the mixed
liquor suspended solids (MLSS) were poor. The Sludge Volume Index
(SVI) in ml/g MLSS averaged approximately 200 between 1977 and
1979 with the monthly average exceeding 300 on six occasions.
A polymer was added to help settling. Nevertheless, attempts to
102
-------
TABLE 6.
PRINCIPAL DESIGN DATA
YEARLY AVERAGE FLOWS:
1977:
1978;
1979:
1173 nr/d
1124 m3/d
1173 m3/d
TANK VOLUMES:
THREE PRIMARY CLARIFIERS: 64 m3 EACH
TWO AERATION TANKS: 460 m3 EACH
TWO SECONDARY CLARIFIERS: 127 m3 EACH
CHLORINE CONTACT CHAMBERS: 65 m3
TWO AEROBIC DIGESTERS: 569 m3 EACH
APPROXIMATE LOADING:
0.2 Kg BOD5/Kg MLSS-d
TYPE OF AERATION:
TWO 45 KW CENTRIFUGAL BLOWERS
CAPACITY OF BLOWER:
0.6 m3/s
103
-------
TABLE 7. SUMMARY OF PRECONVERSION OPERATING DATA -
YEARLY AVERAGE CONCENTRATIONS
RAW WASTE PRIMARY EFFLUENT FINAL EFFLUENT
NH3
Jj\j\j
SS
plus NO -N
H
- (g/m3)
1977
1978
1979
(g/m3)
1977
1978
1979
*
31.3
140
154
148
137
150
105
85
109
88
62
96
62
25.4*
6.3
7.4
6.5
4.8
5.4
5.1
PHOSPHOROUS (g/m3-P)
1977
1978
1979
3.5
6.4
7.4
-
-
-
1.0
1.1
1.0
AVERAGE FOR NOVEMBER AND DECEMBER 1979 ONLY
104
-------
raise the MLSS above approximately 1500 g/m resulted in poor
operation.
In spite of the problem caused by the overdesign, the conventional
activated sludge system located at Culver, Indiana is a well op-
erated facility which meets consistently the imposed effluent
limitations (see Table 2).
The Town of Culver agreed to participate in the demonstration
project in mid-1978. The project was funded in April 1979. Plant
modifications began in August 1979.
SBR CONVERSION
Virtually all modifications were directed at the aeration tanks.
A summary of the modifications and approximate equipment costs
are provided in Table 8. The costs are for both aeration tanks
and include structural steel components, piping, hardware, concrete
modifications, valves, fittings, etc,
A schematic illustrating the changes is given in Figure 7. Ex-
cept for most of the electrical work, the installation was carried
out by personnel from Culver and Notre Dame.
Original plans called for the replacement of the coarse bubble
diffuser systems in one of the two aeration basins with a jet
aeration system. Because of the more efficient use of oxygen by
105
-------
TABLE 8.
SUMMARY OF MODIFICATIONS AND APPROXIMATE COSTS
ITEM
DESCRIPTION
PINCH VALVES
STILLING CHAMBERS
CENTER WALL
COMPRESSORS
DIRECTIONAL JETS
AIR LINES AND VALVES
DECANT SYSTEM
DECANTER PLATFORM
LEVEL DETECTORS
MOTOR CONTROL CENTER
MICROPROCESSOR
REGULATE INLET FLOW
ALLOW AUSTRALIAN MODE
IMPROVE OXYGEN TRANSFER
FOUR AT 7.5 KW EACH
FIVE PER TANK AT 1.6 KW EACH
AIR SUPPLY TO JETS
FLOATING SUBMERSIBLE PUMP
LIMIT DECANTER DISCHARGE
MONITOR LIQUID LEVEL
CONTROL INSTALLED EQUIPMENT
CONTROL SYSTEM
TOTAL
APPROX. COST
$3,500
1,300
3,000
6,000
23,500
9,500
15,000
500
3,300
6,000
12,000
$83,600
106
-------
EFFLUENT
•^
PUMP
CONTROLS
CONTROL-
9*-
CENTER
1
VALVE
CONTROLS
L
INFLUENT ^
i^.
W
r -
> r
LEVE
SENSC
i
1
_L J
PRIMARY
TANK
— — =y ^
*>
• ••< o o o )_
:L
DRSA
^
A
V
Dl
n-z-j
V A
o
- -4
FLOATING
""PUMPS
RECT1ONA
U-JET
ERATORS
&f^
>
^(00 0)
A
L
A
V
y
o
L AERATION A
TANKS
^i \
£
^^ALTERNATING CYCLES^
Fll 1 AMD ORA\A/
a AIR
COMPR
7 (TO E
JET)
FIGUEE 7:
SCHEMATIC ILLUSTRATING CHANGES
AT CULVEH (PLAN VIEW)
107
-------
the jets, a decision was made to replace the existing 45 KW
blowers with four 7,4 KW units. Two compressors, one operating
and one operating-standby, were dedicated to each tank. After
some debate, the coarse bubble diffuser system in the other basin
was also eliminated in favor of the jets. As can be seen from
Figure 7, both aeration basins were modified identically.
The decant system consists of a floating submersible pump coupled
to a suction draw-off pipe. This device was designed only after
attempts to secure the Australian discharge system were frustrated.
In order to insure -against pumping the entire contents of the tank
in the event of a failure of a level controller, a platform was
installed to prevent the glecanter from reaching the bottom of
the tank.
All modifications were completed by April 1980. Usual delays
associated with delivery and installation prevented an early
start-up.
MICROPROCESSOR CONTROL SYSTEM
There are two versions of the program residing in the microproc-
essor. One set operates in real time and normally operates the
plant. The other set is a modified version of the first set and
operates 60-times as fast. This second version is used for train-
ing and demonstrations.
108
-------
Each program set consists of the following:
•Tank One controls sewage treatment in the south tank.
It monitors the liquid level and elapsed time constantly,
and decides which devices to switch on and off at the
appropriate time or liquid level,
•Tank Two is logically identical to Tank One, but controls
the north tank. It should be noted that the two tank
programs operate independently. In this.way, the program
set can easily be adapted to run a larger treatment plant
by simply duplicating the same tank program for each re-
acting tank.
•SCRAM monitors the levels and process-states of both tanks.
Since the two tanks cycle independently, it is possible
for dangerous conditions to occur; e.g., the total shut-
off of influent. This program continuously checks for
the occurrence of such conditions, and instantly remedies
the situation. In the case of a total disaster, such as
the failure of both decanter pumps, this program will put
the treatment plant into a continuous-flow state and sound
an alarm.
•Main controls the time-sharing of the programs resident in
the microprocessor. This routine makes the computer
switch back and forth between the various programs being
executed, spending a brief time executing each one. Due
to the extreme speed of the computer, it appears to the
human observor that all of the programs are running simul-
taneously. This routine implements the same time-sharing
concept used in most large computer systems.
•Teletype monitors the process-states of the tanks. Upon
sensing a change in the state of either tank, this routine
turns on a typewriter, prints the current time, lists the
state and level of the tanks, then turns off the typewriter
to avoid wear.
Many different field programs will be used to operate the treat-
ment plant. Feasible field programs were developed by operating
a computer program supplied hourly flow data for three separate
seven consecutive day periods in 1978, The periods selected
represented minimum, average and maximum flow conditions at Culver.
109
-------
Bench scale reactors were operated in accordance with the computer
simulations in order to determine the treatment efficiency of
the proposed operating modes.
The program developed for SBR operation during the first month
is based on time clock and liquid level control only and is in-
tended to develop organisms with the needed settlina characteris-
tics and increase the population of nitrifiers. Table 9 summarizes
this initial program for normal volume control only for Tank 1.
As can be seen from the table, the events in Tank 1 depend upon
the liquid levels in both Tanks 1 and 2. After nitrification is
established, a program designed to alternate aerobic and anoxic
periods will be implemented.
Program modification is a simple matter, requiring the replacement
of one microprocessor chip with another. Direct modification of
program logic through the teletypewriter will not be allowed.
The microprocessor system is designed to handle up to thirty
individual inputs (such as dissolved oxygen concentration) for
feedback control.
POSTCONVERSION OPERATION AT CULVER
The transition from the continuous flow to the sequencing batch
mode took place on May 12, 1980, On May 15, 1980 the level de-
tectors were reset to be more compatible with actual operating
conditions. Data collected between May 15 and June 10. io°n
110
-------
TABLE 9. INITIAL MICROPROCESSOR PROGRAM FOR
TANK 1 - NORMAL VOLUME CONTROL ONLY
NAME
DESCRIPTION
FILL 1
FILL 2
FILL 3
REACT
SETTLE
DRAW
IDLE
-SIMULTANEOUS FILL AND DISCHARGE
•LEVEL -3, TANK 1 EXCEEDED DURING
DRAW
•AIR FED TO JETS
•ACTIVATES WHEN LEVEL 4, TANK 1 REACHED
•ANOXIC FILL - JETS Oil WITHOUT AIR
•ACTIVATES WHEN LEVEL 5, TANK 2
REACHED
•INLET VALVE CLOSED
•ACTIVATES WHEN LEVEL 5, TANK 1 REACHED
•AIR AND JETS DEACTIVATED
•ACTIVATES WHEN LEVEL 2, TANK 2 REACHED
•DISCHARGE PUMP ACTIVATED
•BEGINS AFTER PREDETERMINED TIME FOR SETTLE
•TANK 1 AWAITS RAW WASTE FLOW
•BEGINS WHEN LEVEL 1, TANK 1 REACHED
111
-------
summarized in Table 10. Although additional data must be collected
before proper comparisons can be made with the previous continuous
flow operation, a preliminary check with the data shown in Table 7
would suggest that the SBR changeover has been quite successful.
BOD_ and suspended solids reductions are essentially the same.
The SVI is markedly improved. Phosphorous removal (with ferric
chloride addition) has increased somewhat. While the aeration
policy implemented between May 15 and June 10 was not selected
for nitrogen removal, there was a slight increase in the net
fraction of inorganic nitrogen removed. This increase resulted
primarily from the denitrification of the raw waste oxidized
nitrogen during the anoxic fill period.
Also shown in Table 10 are data on decant volume and typical
times for each period for one day. In contrast to the systems
described in the first portion of this paper for continuous
flow sequentially aerated systems, the decant depths is roughly
50 percent of the total tank depth, or 1.68 m. In addition,
because of the anoxic fill period (FILL 2), the organisms are
under aeration approximately ten hours during each twenty-four
period. This plus the low dissolved oxygen concentration during
FILL 3 (usually less than 0.5 g/m ) has resulted in little nitri-
fication.
System operation during the remainder of the summer of 1980 will
be modified such that biological nitrogen removal occurs in one
aeration tank (and, possibly, biological phosphorous removal in
112
-------
TABLE 10. SUMMARY OF AVERAGE SBR OPERATING DATA
FROM MAY 15 to JUNE 10, 1980
RAW WASTE
NORTH
TANK
SOUTH
TANK
FINAL
EFFLUENT
FLOW (m /d)
BOD5 (g/m3)
SS (g/m3)
PHOSPHOROUS (g/ra3-P)
NH_ PLUS NO -N•(g/m3)
O Ji
MLSS (g/m3)
SVI (ml/g)
DECANT VOLUME (m3)
DECANT DEPTH (m)
TYPICAL TIMES PER DAY
FILL 1 (Hours)
FILL 2 (Hours)
FILL 3 (Hours)
REACT (Hours)
SETTLE (Hours)
DRAW (Hours)
IDLE (Hours)
AERATION TIME (Hours)
APPROX. NO. CYCLES
1300
150
112
5.4
22.8
6.0 7.1
6.0 7.6
0.82 0.52
16.4 15.8
1880 1770
144 147
230 230
1.68 1.68
0.1
5.5
6.3
3.8
2.8
3.3
2.1
10.1
4
0
5.2
6.8
3.4
2.8
3.2
2.6
10.2
4
1.6
9.0
0.76
16.9
BASED ON RESULTS FROM NINE DAYS
113
-------
the other). Modifications planned include elevating the dissolved
oxygen concentration during FILL 3 to greater than 1 g/m , de-
creasing time for FILL 2 in favor of time for FILL 3 and increasing
the MLSS. After nitrification is achieved, an alternating pattern
of anoxic and aerobic conditions will be instituted for denitrifi-
cation. A similar strategy was employed successfully in the bench
scale reactors. Results from these^ efforts will be presented
during the conference.
DISCLAIMER
The views presented in this paper are those of the author (s).
The paper is not intended to reflect any matters of policy of
the State Pollution Control Commission, nor its approval for
particular plants or processes.
REFERENCES
1. Painter, H.A. , "A Review of Literature of Inorganic
Metabolims in Microorganism." Water Res., 4_, 393, (1970).
2. Downing, A.L., et. al. "Nitrification in the Activated
Sludge Process>'ir~Jour. Proc. Inst. Sew. Purif. 130, 537 (1964)
3. Wild, H.E., et. al. "Factors Affecting Nitrification
Kinetics." Jour. Water Poll. Control Fed..43, 1845 (1971).
4. Haug, R.T., and MaCarty, P.L,, "Nitrification with Submerged
Filters." Jour. Water Poll. Control Fed., 44, 2987 (1972).
5. Kiff, R.J., "The Ecology of Nitrification/Denitrification
Systems in Activated Sludge." Water Poll. Control, 71,
475 (1972).
6. Sutton, P.M., et. al., "Continuous Biological Denitrification
of Wastewater." Report No EPS4-WP-76-6, Environment Canada,
August, 1974.
114
-------
7. Skerman, V.B.D., and MacRae, I.C., "The Influence of Oxygen
on the Degree of Nitrate Reduction by Pseudomonas Denitri-
fication." Can. Jour. Microbiol., _3, 505 (1975).
8. Wuhrmann, K., "Effects of Oxygen Tension on Biochemical
Reactions, in Sewage Purification Plants." Biological
Waste Treatment, Reinhold, Manhattan College, (1960) .
9. Mueller, J.A., et. al., "Nominal Diameter of Floe Related
to Oxygen Transfer." Jour. San. Eng. Div., Amer. Soc.
Civil Eng. , £2, 9 (1966).
10. Ludzack, F.J., and Ettinger, M.B., "Controlling Operation
to Minimize Activated Sludge Effluent Nitrogen." Jour.
Water Poll. Control Fed. , 34., 920 (1962).
11, Pasveer, A., "A Contribution to the Development in Activated
Sludge Treatment. " Jour. Proc. Ins_t. Sew. Purif. , 4_, 436
(1959).
12. Matsche, N.F., "The Elimination of Nitrogen in the Treat-
ment Plant of Vienna-Blumental. " Water Res. , 6_, 485 (1972).
13. Barnard, J.L., "Cut P and N Without Chemicals." Water Wastes
Eng., 11, 33 (1974).
14. Christensn, N.H., "Denitrification of Sewage by Alternating
Process Operation." 7th IAWPR Conference, Paris, Sept.,
(1974) .
15. Bishop, D.F., et. al., "Single Stage Nitrification-Denitri-
fication." Jour. Water Poll. Control Fed., 48, 520 (1976).
16. Drews, R.J.L.C., and Greef, A.M., "Nitrogen Elimination by
Rapid Alternation of Aerobic Conditions in Orbal Activated
Sewage Plants." Water Res. , 1_, 1183 (1973).
17, Batty, J.A., et. al., "Development of the Pasveer Extended
Aeration System." The Shire and Municipal Record, 67, 608
(1974).
18. Barth, E.F., "Design of Treatment Facilities for the Control
of Nitrogeneous Materials." Water Res. , 6., 481 (1972).
19. Sutton, P.M., et. al., "Nitrogen Control a Basis for Design
with Activated Sludge Systems." Prog. Water Tech., 8_, 467
(1977).
20. Goronszy, M.C., "Intermittent Operation of the Extended
Aeration Process for Small Systems." Jour. Water Poll.
Fed., 51, 274 (1979).
115
-------
21. Tomlinson, E.J., and Chamber, B., "The Use of Anoxic
Mixing Zones to Control the Settleability of Activated
Sludge." TR 116, Water Research Centre, Stevenage (1979).
22. Water Research Centre and Wessex Water Authority, "Palmers-
ford Sewage Treatment Works, Investigation of the Oxygen
Activated Sludge Process." Water Research Centre, Stevenage
(1979).
23, Goronszy, M.C., "Single Vessel Intermittently Operated
Activated Sludge for Nitrification - Denitrification."
51st Water Poll. Control Fed, Conf., Anaheim (1978).
24. Christensen, M.H., and Harramoes, P., "Biological Denitri-
fication of Sewage: a Literature Review." Prog. Water
Tech., 8., 509 (1977).
25. Goronszy, M.C,, and Barnes, D., "Intermittent Single Vessel
or Conventional Continuous Activated Sludge - Economic
Considerations," 52nd Water Poll. Control Fed. Conf,,
Houston (1979).
26. Pasveer, A., "A Case of Filamentous Activated Sludge."
Jour. Water. Poll. Control Fed., 41, 1340 (1969).
27. Irvine, R.L., and Busch, A.W., "Sequencing Batch Biological
Reactors - An Overview." Jour. Water Poll. Control Fed.,
51, 235 (1979).
28. Irvine, R.L., et. al., "Investigation of Fill and Batch
Periods of Sequencing Batch Biological Reactors." Water
Res., 11, 713 (1977).
29. Irvine, R.L., and Richter, R.O., "A Comparative Evaluation
of Sequencing Batch Biological Reactors." Jour. Environ.
Eng. Div., Amer. Soc. Civil Eng., 104, 503 (1978).
30. Dennis, R.W., and Irvine, R.L,, "Effect of Fill:React
Ratio on Sequencing Batch Biological Reactors." Jour.
Water Poll. Control Fed., 51, 255 (1979).
31. Irvine, R.L., et. al., "Sequencing Batch Treatment of
Wastewaters in Rural Areas." Jour. Water Poll. Control
Fed. , 51, 244 U979),
32. Alleman, J.E., and Irvine', R.L,, "Nitrogen Removal from
Wastewater Using Sequencing Batch Reactor Design." Amer.
Inst. Chem. Eng, Res. Symp. Ser. - Water 1978, 75, 181
~~
116
-------
33. Alleman, J.E., and Irvine, R.L,, "Nitrification in the
Sequencing Batch Reactor." Jour. Water Poll. Control Fed.,
(in press).
34. Alleman, J.E., and Irvine. R.L., "Storage-Induced Denitri-
fication Using Sequencing Batch Reactor Operation." Water
Res. (in press).
117
-------
"NITROGEN AND PHOSPHORUS REDUCTION
FROM LAND APPLICATION SYSTEMS
at the
WALT DISNEY WORLD RESORT COMPLEX"
H. Robert Kohl
Director
and
Ted McKim
Sanitary Engineer
Reedy Creek Utilities Co., Inc.
Walt Disney World Co.
INTRODUCTION
The Reedy Creek Improvement District provides wastewater
collection and treatment for the twenty-seven thousand
four hundred (27,400) acres which make up the WALT DISNEY
WORLD Resort Complex. The Reedy Creek Utilities Co., Inc.
is a private utility company which plans, designs, con-
structs, and operates the facilities owned by the Reedy
Creek Improvement District.
Prior to the construction of the WALT DISNEY WORLD Resort
Complex, the area was part of a larger swamp and the high-
land area was limited to agricultural development.
The entire volume of wastewater originates on the property
with the majority produced in the Theme Park, hotels and
the WALT DISNEY WORLD Shopping Village at Lake Buena Vista,
totalling approximately three thousand (3,000) acres.
The wastewater treatment plant was constructed for a
nominal capacity of 3.3 MGD in 1970 with design requirements
of ninety percent (90%) B.O.D. and suspended solids removal.
118
-------
Subsequent improvements upgraded the capacity to 4.0 MGD.
The secondary process train consists of a standard activated
sludge process with comminution, grit removal, primary treat-
ment, travelling screens, aeration tanks, final clarifiers,
and chlorine contact tank. Sludge treatment includes air
flotation for waste activated sludge thickening, aerobic
digestion, drying beds, and a tank truck for land spreading
of digested sludge.
Initial construction of the system allowed for the direct
discharge of chlorinated secondary effluent to Reedy Creek.
However, with the attempt by the Environmental Protection
Agency (EPA), in 1975, to promote non-point source discharge
to waterways, the Reedy Creek Utilities Co., Inc. was inter-
ested in complying with this concept even though the State
of Florida Department of Environmental Regulation had no
requirements of this nature at the time. From 1976 until
1979, four land application systems were constructed to
minimize the impact of direct discharge of secondary effluent
and to provide an advanced wastewater treatment system capable
of reducing the long range environmental impact on the seven
thousand five hundred (7,500) acre conservation area down-
stream of the treatment system at the WALT DISNEY WORLD Resort
Complex.
SYSTEMS DESCRIPTION
The four land application systems presently in operation are
shown schematically in Figure 1 and described as follows:
119
-------
\\ BERM
WASTEWATER TREATMENT
PLANT SITE
r1
flUHf WATER HYACINTH
SYSTEM
DIGESTED
SLUDGE
APPLICATION
AREA
WETLANDS
OVERLAND
FLOW
SYSTEM
HOLDING POND
NORTH
POLISHING
POND
410 WEIR
'AT RC-18
APPROX. SCALE
I" = 1,000'
PERCOLATION PONDS/OVERLAND FLOW SYSTEM
UNDEFINED CHANNEL-J
Figure 1. Wastewater Treatment Facilities
At Walt Disney World
120
-------
Percolation Pond - Overland Flow System. Placed in opera-
tion on January 1, 1977, the percolation pond - overland
flow system consists of a 5.2 acre polishing pond (7.5 M.G.
capacity) and three two-acre percolation ponds (7.8 M.G.
total capacity) with a maximum operating depth of four feet.
The percolation ponds are constructed in virgin sandy soil,
U-shaped, with an underdrain system at a point 150 feet
distant from the water surface with an operating head of
9.0 feet to the invert of the underdrains. The underdrains
outlet through a 30" pipe to an open channel weir, 200 feet
in length, thus creating a sheet flow effect across a wooded
area and into a wetlands prior to discharging over a 2.0
foot rectangular weir into Reedy Creek. The total area of
overland flow and ponding is 14.5 acres.
The average capacity of the system is one million gallons
per day.
Most comparisons used in this paper for the percolation pond
system will relate primarily to the overland flow portion,
the area between the discharge of the underdrains and the
overland flow effluent weir.
Wetlands - Overland Flow System. The wetlands - overland flow
system, constructed in July 1977, consists of 102 acres of
wetlands immediately adjacent to the wastewater treatment
plant with a vegetative makeup of pines, cypress, and dense
understory. The soil characteristics are generally muck.
121
-------
Chlorinated secondary effluent is allowed to flow into the
area by gravity. The entire area is bermed and isolated so
that the flow travels a total distance of 5,000 feet to one
discharge point over a 20-foot rectangular weir with end
contractions. The gradient of the overland flow is approxi-
mately 0.10 percent. The average flow to this system is 2.0
MGD.
This system consists of a combination of overland flow and
ponding or wetlands. Ponding in some areas is two to three
feet and velocities are nearly non-detectable.
Fixed Irrigation. Fixed irrigation has been installed on
approximately 92 acres utilizing secondary effluent on the
WALT DISNEY WORLD Tree Farm.
Secondary effluent is pumped from polishing ponds to the
sprinkler heads. Fixed irrigation is underdrained on 20
acres with the drainage flowing into a wetwell and returned
to the polishing pond. The remaining irrigation filters
through the soil into the shallow ground water table. This
system has a capacity of a 1.0 MGD rate, however, average
daily use is approximately 200,000 gallons.
Water Hyacinths. A fourth system utilizing water hyacinths
to treat wastewater was completed in May 1979. The system
consists of three one-quarter acre channels 29 ft. x 360 ft.
which can receive either primary effluent or secondary
122
-------
effluent from the main wastewater treatment plant. The
depth of flow in the channels is 14 inches which allows for
a detention time of 5.4 days at 50,000 gallons per day flow.
This is a demonstration project which has a duration of four
years and is funded by the U. S. Environmental Protection
Agency, NASA, the Gas Research Institute, Aquamarine Corp.,
and WED Enterprises.
The objectives of the project are to produce an energy effi-
cient secondary and advanced wastewater effluent and to
ultimately optimize hyacinth growth to produce methane gas
through anaerobic digestion.
SYSTEMS OPERATION
The primary goal in early 1977 for instituting land appli-
cation processes at the WALT DISNEY WORLD Resort Complex was
to reduce nutrient loading on Reedy Creek and attain a non-
point discharge from the wastewater treatment facility.
It was anticipated that the original facility, the percolation
pond system, would provide a capacity of 3.5 MGD; however,
due to algae plugging the system has provided a consistent
1.0 MGD since its initial start-up.
The percolation ponds are drained every seven to nine months
and allowed to dry for a period of two weeks. At the end
of the drying period, a one-inch crust of sand penetrated
with algae remains which is then disced to a depth of twenty
123
-------
inches. Placing the system on-line after this procedure
results in restoring the capacity to 1.0 MGD. During the
drying period, the system is operated at 1.0 MGD with the
percolation ponds being by-passed and the water from the
polishing pond draining directly to the overland flow
portion of the system. No degradation in effluent water
quality occurs during these periods.
The 92-acre fixed irrigation system is operated on an as
needed basis by the WALT DISNEY WORLD Horticulture Department.
All irrigation on the Tree Farm is provided from secondary
effluent.
The wetlands - overland flow system receives all flow over
and above that pumped to the percolation ponds and spray
irrigation. In addition to the 102 acres which is used
as the primary treatment area, approximately 298 acres of
wooded and undeveloped area drains into the facility as
storm water runoff. This adds to the nutrient loading of
the system. The only maintenance required on this system
is that of clearing the biological growth from the discharge
weir two or three times per year.
The water hyacinth system was placed on line in July 1979
receiving 50,000 gallons per day of primary settled waste-
water from the main plant. Two channels utilize the water
hyacinths while the third channel is a control channel and
acts as a facultative lagoon. The channels are harvested
124
-------
from one to two times per month (depending upon the season),
chopped and placed on a compost pad. The compost piles are
turned twice weekly and records are kept on temperature and
moisture content of the compost. Basic observations of the
plants are as follows:
o Unusual growth of dollar weed
o Blind mosquitoes
o Frost burn during winter weather
o Composted material stringy and difficult to distri-
bute - not friable
WATER QUALITY AND TREATMENT EFFICIENCIES
The parameters tested and their frequency for the various
systems are tabulated in Table 1.
Table 2 outlines the biochemical oxygen demand and suspended
solids removal throughout the system.
Due to high amounts of rainfall in Central Florida during
certain seasons of the year, the average flow discharging
from the systems exceeds the influent on a yearly average.
These increases amount to 119% and 153% of the input to the
percolation ponds and overland flow respectively.
Nitrogen removal efficiencies, as outlined in Tables 3 and
4 , in both percolation pond and overland flow systems, are
excellent. Much research has been accomplished in the area
of nitrogen reduction through nitrification - denitrification
125
-------
Table 1
Parameters Tested and Frequency Per Month
Reedy Creek Land Application Systems
PERCOLATION
PARAMETER PONDS
B.O.D.
S.S.
TKN
NH3
N03
0-P
T-P
Fe
Cl
Conductivity
TC/TOC
PH
Temperature
D.O.
4
4
4
-
4
4
4
4
4
4
4
30
30
30
OVERLAND
FLOW
4
4
4
-
4
4
4
4
4
4
4
30
30
30
FIXED
IRRIGATION
4
4
4
-
4
4
4
4
4
4
4
WATER
HYACINTHS
8
8
4
8
4
8
8
0
1
0
2
30
30
30
126
-------
Table 2
B.O.D. and Suspended Solids Summaries
Reedy Creek Land Application Systems
FLOW B.O.D. S.S. B.O.D. S.S.
LOCATION MGD Mg/1 Mg/1 Ibs/day Ibs/day
Influent to WWTP
Discharged from
WWTP
To Percolation
Ponds
From Percolation
Pond Underdrains
Discharged from
Percolation Ponds
Overland Flow
To Wetlands -
Overland Flow
Discharged from
3
3
1
1
1
2
3
.296
.296
.051
.051
.254
.012
.077
293.
11.
11.
4.
1.
11.
1.
4
00
00
12
52
0
36
201
18.
18.
10.
1.
18.
1.
80
80
10
69
80
50
8,065
302
96.40
35.90
15.9
184.60
34.90
5,525
518
164.
88.
1?.
315.
38.
80
50
7
50
50
Wetlands - Over-
land Flow
To Spray 0.233 11.00 18.80 21.40 36.50
Irrigation
127
-------
Table 3
Phosphorus and Nitrogen Summaries
Reedy Creek Application Systems
October 1978 to April 1980
FLOW T-N T-P T-N T-P
LOCATION MGD Mg/1 Mg/1 Ibs/day Ibs/day
Influent to WWTP 3.296 28.2 6.00 775 165
Discharged from 3.296 12.11 2.90 333 80
WWTP
To Percolation 1.051 12.11 2.90 160 25.4
Ponds
From Percolation 1.051 3.54 1.08 31.00 9.50
Pond Underdrains
Discharged from 1.254 1.39 1.21 14.60 12.70
Percolation Ponds
Overland Flow
To Wetlands - 2.012 12.11 2.90 203 48.7
Overland Flow
Discharged from 3.077 1.44 3.22 37.00 82.60
Wetlands - Over-
land Flow
To Spray 0.233 12.11 2.90 24.00 5.90
Irrigation
128
-------
Table 4
Nutrient Removal Summaries
Reedy Creek Land Application Systems
LOCATION
Percolation
Pond Overland
Flow
Wetlands -
Overland Flow
REMOVAL
EFFICIENCIES
T-N T-P
% %
REMOVAL RATES APPLICATION
T-N T-P RATES
#/ac/day #/ac/day inches/week
53
81.8
-34
-69.5
1.13
1.62
-.22
-.33
22.2
5.1
129
-------
on the soil surface which substantiates the success of this
process. Application rates as shown in Table 4 to date
have indicated that 22.2 inches per week and 5.1 inches per
week can be applied to the percolation pond - overland flow
and wetland - overland flow systems respectively with excellent
efficiencies achieved.
Phosphorus removals, as noted in Table 3 and 4, are poor.
Sandy soils in general have very little capability for the
removal of phosphorus. It is assumed that the addition of
storm water runoff in both systems accounts for the actual
increase in phosphorus loading exiting the systems.
Figures 2 through 5 depict monthly average plottings of the
pounds of nitrogen and phosphorus discharged from each of
the overland flow systems as they relate to the influent
loadings. For the past nineteen months, the nitrogen removals
have been very consistent and show no increased trends. The
higher increase of all nutrients being discharged from the
systems in September 1979 is attributed to Hurricane David
which produced 3.5 inches of rainfall on the area. In general,
the discharge of nitrogen remains quite constant compared to
the influent loadings.
The plot of phosphorus discharge from the percolation pond
overland flow discharge noted on Plate 111 is very consistent
with only small increases attributed to rainfall due to the
smaller drainage area.
130
-------
LJ
LJ
t-
co
>
V)
o
Q
z
LI
>
O
i
Q
Z
O
0.
o
a
LJ
O.
t- h-
z z
UJUJ
b_ LJ_
Z U_
HLJ
Cfl
>i
cn
C
(C
o •
i a>
T3 >
Ci
n3
•Q
y\
PJ*
Cd
C^2
I
cs
0)
OUJ
131
-------
cs«
<9
C9
(0
Q
(M
— 00
— 00
<9CO H
— 00 H
Cvl® u
— 00 f—
«i 1
^ CL.
0)0)
000
Is*
0) r**
. r^-
(DO)
0)LO
0)CO
. Is*
(M 09
r>.
00 —
000
<9
O
_1
a
CO
a
z
u
3
'Z.Z
UJ LJ
\-~l-3
1
en
B
•-I
fc,
'O
(0
iH
iJ-—
OJ
o >
I <
co
tJ >i
c m
(CO
-u=<*
0)
52
QH
l_J ^X
o
tf
ro
0)
H
P
Cr>
132
-------
-:>f-
1—00
\
u,*3
-S"5
CM®
00
\
O0>
1^
0>0)
o>r^
.r^
(00)
r-
O)\t)
r^
N
0)CO
h»
\
<\IO>
r^
00^
r^
ooo
Q
^~
CM
Q
Q
H
H
H
LJ
h-
(f)
>
Q
Z
<
_l
a
LJ
>
O
I
LJ
OOZ Lull.
C^ 2lJ-
\ HLJ
g
~
O •
I 0)
Ti >
C<
O
Ol >1
(0
•Q
(a
U
i
•H
OLU
133
-------
E
Q)
4J
Cfi
TD
(0
rH
S-f-
OJ •
>
I i
C (0
(OD
4-1*
0)
Sfr
U
m
134
-------
Significant phosphorus discharges in excess of the influent
from the wetlands - overland flow system as noted on Plate
IV are attributed to the large drainage area contributing
storm water. A small portion of the wastewater sludge
spreading area is included in the wetlands - overland flow
drainage area, probably resulting in higher levels of phos-
phorus discharge.
SUMMARY
The overall land application systems at the WALT DISNEY
WORLD Resort Complex are quite complex; from underdrainage
in the spray irrigation system to a combination of percolation
ponds with underdrains leading to an overland flow system to
a wetlands - overland flow system.
All overland flow systems have demonstrated an extremely high
capability for removing nitrogen. The application rates range
from 5.1 to 22.2 inches per week with varying influent con-
centration. The actual removal rates have ranged from
averages of 1.13 Ibs/ac/day to 1.62 Ibs/ac/day with resultant
discharge concentrations remaining well within the Florida
Department of Environmental Regulation standards of 3.0 mg/1.
Phosphorus removal cannot be considered successful in overland
flow systems in Central Florida. As a result, if phosphorus
removal is required, chemical treatment will most probably be
the mechanism.
135
-------
CONSIDERATIONS FOR FURTHER STUDY
Much data has been accumulated on the various land application
systems at the WALT DISNEY WORLD Resort Complex since January
1976. It has been the intent of this paper to obtain indi-
cations of any trends occurring on the systems for the past
nineteen months.
The data described in this paper will continue to be obtained
plus the following areas will be considered for future publi-
cations :
o Biological changes throughout the overland flow systems
o Effects upon groundwater quality surrounding these
systems
o Effects of storm water runoff to the systems
o Incremental water quality testing throughout the
systems
o Water budgets for the systems
ACKNOWLE DGEMENTS
Acknowledgement is given to the assistance provided by the
Reedy Creek Improvement District Environmental Laboratory for
providing all the water quality data on the systems described.
136
-------
EXPERIENCE WITH AMMONIA REMOVAL BY
SELECTIVE ION EXCHANGE AND
CLOSED-CYCLE AIR STRIPPING REGENERANT RENEWAL
By
L. Gene Suhr
INTRODUCTION
Second only to carbon in amount, nitrogen is a major con-
stituent of municipal wastewaters. Various compounds of
nitrogen, which may occur in raw or treated wastewater, are:
o Organic nitrogen
o Ammonia nitrogen—present either as the ammonium
ion (NH.+) or as dissolved ammonia gas (NH3)
o Nitrate ion (N0_-)
o Nitrite ion (N02~)
Table 1 lists typical ranges of nitrogen concentrations
found in domestic wastewaters as reported by five separate
references.1'2'3'4'5
TABLE 1
TYPICAL RANGES OF NITROGEN CONCENTRATIONS
FOUND IN RAW DOMESTIC WASTEWATER
CONCENTRATION IN mg/l AS N
NITROGEN
FORM
ORGANIC N
AMMONIA N
NITRATE N
NITRITE N
TOTAL N
1
8 to 20
12 to 35
0 to 0.5
0 to 2.0
20 to 55
REFERENCES
234
8 to 35
12 to 50
0
0
20 to 85
20
35
NR
NR
55
10 to 15
25 to 35
0
0
35 to 55
5
7 to 22
4 to 35
NR
NR
11 to 57
REF. 1 — Public Works, May 1973
REF. 2 — Wastewater Engineering, Metcalf and Eddy, McGraw Hill Pub. Co. 1972
REF. 3 - MOP No. 8, WPCF, 1977
REF. 4 — Manual of Treatment Processes, Environmental Services Corp., 1968
REF. 5 — The Eftect of Industrial Wastes on Sewage Treatment, New England Interstate WPCC, 1965
137
-------
Nitrogen, depending upon its form and concentration, can be
deleterious to the aquatic environment in a variety of ways.
For example, the role of nitrogen as a primary algal nutrient
is well documented as is the causative relationship between
nitrate and infant morbidity through methemoglobinemia.
Evidence also strongly links high nitrate concentrations to
lowered birth and survival rates for livestock calves. In
addition, nitrites are linked to formation of nitrosamines,
long known to be carcinogens in laboratory animals . Ammonia,
aside from its propensity to deplete dissolved oxygen content
of receiving waters, can, depending on circumstances, be
acutely toxic to fish fry particularly trout and salmonoid
species. The unionized portion of ammonia in water is
especially toxic to aquatic life.
The U.S. EPA on January 3, 1980, promulgated notice of its
intention to add ammonia to the list of toxic pollutants
under the clean water act . While inclusion of ammonia on
the toxic pollutant list has not (as of May 1980) yet
occurred, this action by the U.S. EPA underscores the
national concern for the control of nitrogen in wastewater
effluents.
Nitrogen control technology can range from removal of nitro-
gen to simply changing the form of nitrogen present. Control,
either removal or change of form, can be accomplished by a
variety of means including biological and physical-chemical
processes. This paper discusses nitrogen removal, specifically
ammonia removal, by use of a combined physical-chemical
process. The unit processes discussed will be ammonia
removal by ion exchange, regeneration of spent ion exchange
beds, and renewal of spent regenerant using a closed cycle
air stripping process. Design and operating data discussed
138
-------
herein are specifically drawn from the Tahoe-Truckee Sanita-
tion Agency Advanced Wastewater Reclamation Plant, which is
now in its third year of operation. This paper highlights a
nitrogen removal process specifically designed to meet a
very strict effluent standard.
FUNDAMENTALS OF AMMONIA REMOVAL BY ION EXCHANGE
Among the prerequisites for successful ammonia removal by
selective,_ion exchange is the necessity for the ammonia to
be present in the form of ammonium (NH.+) ions. Ammonia in
wastewater is a "changeling." It is not only relatively
easily oxidized from the ammonia form to nitrate, but may
also take the form of a dissolved gas or an ammonium ion in
true solution or both, depending on pH and temperature. At
100
• FIGURE 1
*100
14
EFFECTS OF pH AND TEMPERATURE ON
DISTRIBUTION OF AMMONIA AND
AMMONIUM ION IN WATER
139
-------
a pH of 7 or less, only ammonium ions in true solution are
present; at a pH of 12, only dissolved ammonia gas is present;
at pH values between 7 and 12, both forms may be present.
In addition to the pH, temperature also plays a major role
in determining the ratio of ammonia and ammonium species
present. The effects of pH and temperature on the distri-
bution of ammonia and ammonium ion in water are shown in
Figure 1.
Selective Ion Exchange
The selective ion exchange process derives its name from the
use of zeolites which are selective for the ammonium ion
relative to most other cations normally present in municipal
sewage. The zeolite currently favored for this use is
clinoptilolite, which occurs naturally in several extensive
deposits in the Western United States. The order of exchange
Q
preference for the various cations by clinoptilolite is :
Cs+>Rb+>K+>NH4+>Ba++>Sr++>Na+>Ca++>Fe+3>Al+3>Mg++>Li+
The clinoptilolite is crushed and sieved to obtain a 20 by
50 mesh size. Ammonium ion is removed by passing the waste-
water through a bed of clinoptilolite at rates between 6 to
10 bed volumes per hour. Ammonia removals of as high as 96
percent may be expected with influent ammonia nitrogen con-
centrations of about 20 mg/1.
The ammonia capacity of clinoptilolite is nearly constant
over a pH range of from 4 to 8, but diminishes rapidly
outside of this range. Wastewater composition affects the
ammonia exchange capacity. For relatively constant influent
ammonia concentrations, the ammonia exchange capacity
decreases sharply with increasing concentrations of competing
cations (such as calcium and magnesium). Ammonia removal to
140
-------
residual levels below 0.5 mg/1 ammonia nitrogen is techni-
cally feasible but requires very short service cycles and
greatly increases the complexity of regeneration requirements.
Flow rates in the range of 7.5 to 15 bed volumes per hour
through the ion exchange beds do not appear to have measure-
able effects on ammonia effluent values.
After about 150 to 200 bed volumes of normal strength municipal
waste have been passed through an ion exchange bed, the
capacity of the clinoptilolite will have been used to the
point that ammonia will begin to "leak" through the bed. At
this point, the clinoptilolite must be regenerated so that
its capacity to remove ammonia is restored.
The key to the applicability of this process is the method
of handling spent regenerant. The exchange media is regen-
erated by passing concentrated salt solutions through the
exchange bed when the effluent ammonia concentration has
reached the maximum desirable level. Following regeneration,
the ammonia-laden spent regenerant volume is about 2-1/2 to
25 percent of the throughput previously passed through the
bed. This will depend on the type and concentration of the
regenerant used. Obviously this regenerant cannot be
disposed of directly to the aquatic environment and there-
fore various means have been employed to allow reuse of the
regenerant solution.
One of the earliest solutions attempted was the use of lime
slurry as a regenerant so that the ammonium stripped from
the bed during regeneration would be converted to gaseous
ammonia because of the pH of the regenerant solution. This
was then followed by an air stripping process which could
9
remove the gaseous ammonia from the lime slurry . Unfor-
tunately regeneration with lime alone was found to be a
rather slow process; therefore, the ionic strength of the
141
-------
regenerant solution was increased by the addition of sodium
chloride. The increased ionic strength of the regenerant
plus the presence of sodium ion accelerates the removal of
ammonia from the zeolite. Although most of the sodium
chloride added to the regenerant is converted to calcium
chloride by continuous recycle of the regenerant, sufficient
sodium ion remains under steady state conditions to promote
the elution of ammonium ions from spent clinoptilolite beds.
The use of the high pH regenerant system is accompanied by
extreme operational problems. Plugging of the clinoptilolite
beds with magnesium hydroxide and calcium carbonate occurs
when the high pH regenerant is used. Attrition of the
zeolite is aggravated by the violent backwashing necessary
to remove the solids.
In an alternate approach, ammonia in the regenerant solution
is converted to nitrogen gas by reaction with chlorine which
is generated electrolytically from the chlorides present in
the regenerant solution. This process can be carried out
with a regenerant of neutral pH so that the problem of
precipitation of magnesium hydroxide and calcium carbonate
within the bed during regeneration is eliminated. Also cold
weather does not affect this regenerant recovery process.
The regenerant solutions used are rich in sodium chloride
and calcium chloride which provide the chlorine produced at
the anode of the electrolysis cell. The reactions for the
destruction of ammonia by chlorine in this process are the
same as for breakpoint chlorination. During regeneration of
the ion exchange bed, a large amount of calcium is eluted
from the zeolite along with the ammonia. This calcium may
be removed from the spent regenerant solution by a softening
process prior to passing the spent regenerant through the
electrolytic cells. The softening step lowers the calcium
concentration below the level that would cause calcium
142
-------
hydroxide formation in the electrolytic cells. Unfor-
tunately, return of the solids in the recycle resulting from
this softening step to the treatment plant unavoidably
results in an increase in effluent total dissolved solids.
High flow velocities through the electrolytic cells are
required in addition to a low concentration of calcium to
minimize scaling of the cathode by calcium hydroxide and
calcium carbonate. Acid flushing of the cells is necessary
to remove the scale when the cell resistance becomes too
high for economical operation.
A third alternative regeneration scheme is that of steam
stripping. Steam stripping of spent regenerant was demon-
strated at the 0.6 mgd Rosemount, Minnesota Physical Chemical
Plant . This process is economically feasible only with
high pH regenerant. At Rosemount, ammonia was recovered
from the spent ion exchange regenerant in an ammonia stripper,
Steam was injected into a distillation column countercurrent
with the regenerant solution to strip off the ammonia. An
air-cooled plate and tube condenser was used to condense the
vapor for collection and storage in a covered tank as a
1-percent aqueous ammonia solution which could be sold as a
commercial fertilizer. The steam stripping process is based
on the use of high pH regenerant which has the disadvantages
noted earlier. Battelle Northwest's evaluation of steam
stripping indicates that it is economically feasible if the
regenerant volume is held to four bed volumes per cycle. A
stripping tower depth of 24 feet and a loading of 7 gallons
per minute per square foot were used at Rosemount. Ceramic
saddles were used in the stripping tower rather than wooden
slat packing to minimize delignification which would other-
wise occur at the high pH values attendant to the process.
A new process for removal of ammonia and recovery of the
regenerant was developed by CH2M HILL to eliminate many of
143
-------
the shortcomings of other methods of handling or recovery of
spent regenerant from the clinoptilolite ion exchange process.
This process was labeled ARRP (ammonia removal and recovery
11 12 13
process). ' ' The regenerant solution used with ARRP
is sodium chloride. In the regeneration process, sodium
exchanges with ammonia on the clinoptilolite resin. The
regeneration cycle is a countercurrent process to minimize
the makeup brine required, and uses four tanks of regenerant
having different degrees of activity ranging from virgin to
spent.
The freshly renewed regenerant solution at equilibrium
conditions contains approximately 3 percent sodium chloride,
1 percent calcium chloride, plus about 600 mg/1 potassium
chloride. Initially the regenerant will be a 3-percent
sodium chloride solution and this is the active regenerant.
Calcium and potassium ions are allowed to build up in the
regenerant solution until their concentration and that of
the influent wastewater to the ion exchange system are in
balance. No further significant removal of these ions
occurs following such equilibrium. A small amount of
magnesium will be present in the influent to clinoptilolite
beds. This will be captured in the beds and subsequently
removed from the beds during regeneration. Final removal of
magnesium in the regenerant occurs by chemical precipitation
following addition of sodium hydroxide.
Regeneration involves pumping from one tank to another
through the ion exchange bed. Towards the end of the cycle,
freshly renewed regenerant is recycled through the bed.
This has the effect of neutralizing the alkaline pH carry-
over from the ARRP process. ARRP effluent is normally at a
pH of between 10.7 and 11.0. This is reduced to a pH of
about 9.5 by recycling. In this manner, the pH is con-
trolled without excessive use of acid and wasting of alkaline
chemicals.
144
-------
When spent regenerant is accumulated to a predetermined
amount, the recovery portion of the process is activated.
This system operates at a flow rate of approximately 7.7
percent of the average clinoptilolite bed throughput, which
means that the concentration of ammonia in the spent regen-
erant solution will be about 13 times that of the influent
to the clinoptilolite beds. Initially, sodium hydroxide is
added to the regenerant to achieve a pH of about 11. The
amount added is stoichiometric for ammonium and magnesium
ions removed. Sodium chloride is also added because of some
salt losses from the regenerant solution during clinoptilolite
bed rinsing and from magnesium hydroxide sludge removal.
Following pH adjustment, the regenerant is clarified and
magnesium hydroxide is removed by sedimentation. Effluent
is then pumped to stripper adsorber columns for ammonia
removal and recovery.
The ARRP process is a closed cycle air stripping and adsorp-
tion system. The ammonia gas is first stripped into an air
stream in the stripping tower. The air stream containing
the ammonia is then directed into an absorption tower where
the ammonia is absorbed. By maintaining the absorbent
liquid in an acid condition, the absorbed ammonia is
immediately converted to ammonium ions which are thus effec-
tively trapped as the ammonium salt of the acid used.
TAHOE TRUCKEE SANITATION AGENCY WATER RECLAMATION PLANT
Plant Description
The 4.83-mgd Tahoe-Truckee Sanitation Agency (T-TSA) Water
Reclamation Plant became operational in February 1978. The
plant provides extremely high-level tertiary treatment to
protect the pristine quality of the Truckee River, which is
the primary water supply for the City of Reno and eventually
145
-------
flows into Pyramid Lake, Nevada. Discharge requirements
established by the regulatory agency to protect receiving
water quality are shown in Table 2.
TABLE 2
T-TSA EFFLUENT LIMITATIONS
MEAN MAXIMUM
CONSTITUENT UNITS CONCENTRATION CONCENTRATION
COD mg/l 15 40
SUSPENDED SOLIDS mg/l 2 4
TURBIDITY NTU 2 8
TOTAL NITROGEN mg/l 2 4
TOTAL PHOSPHORUS mg/l 0.15 0.4
MBAS mg/l 0.15 0.4
TDS mg/l 440 —
CHLORIDE mg/l 110 —
TOTAL COLIFORM ORGANISMS MPN/100 ml — 23
Processes installed to meet the discharge requirements
include the following:
o Primary treatment
o Pure oxygen activated sludge
o Lime treatment with two-stage recarbonation
o Dual media filtration
o Activated carbon adsorption with onsite regeneration
o Selective ion exchange using clinoptilolite for
ammonia removal
Bilogical sludges are anaerobically digested. Both biological
and chemical sludges are dewatered in a plate and frame
filter press and landfilled.
The process diagram and plant layout are shown on Figures 2
and 3, respectively. Flow is by gravity through the primary,
secondary, and chemical treatment facilties and is then
pumped in a single-stage through the pressure filters,
carbon columns, and ion exchange vessels on the way to the
land effluent disposal system.
146
-------
I 1 NO9UV3
'93d
<
cc
o
5
0)
u
o
o
2z
o
o
> m < 5
S fn tft w
_ S52£
z i2
— LU LU — ^
cc u u O
3cc§§± S
I CC _| _| CC <
U < 00 W O O
CM
D 0)
O h"
IL h-
cc
LU
IS
itii
u z S °-
uj o K ,_
o: 5So w
Z
g
i-
CL
CC
O
D t/3
gS|3sli
E LU uj < r < _i
LL CC (0 CO LL U O
CO CT) O «-
-------
in oo
CM CO
s
SzOcoSe!
o: O _i o cc<
o. O i o < O i
O <(/) o £
- < < - -
,,000:0:
w E E £ H
^ i- t- w w
J o o uj u
-J ui uj a u
< -i -i s =
CD UI UJ O Q
148
-------
The ammonia removal system is a unique feature of the plant.
It consists of four 40-foot long by 10-foot-diameter clinop-
tilolite ion exchange beds that reduce the ammonia in the
plant flow stream from 20 to 30 mg/1 down to 2 to 3 mg/1.
When the capacity of a bed is exhausted, it is regenerated
with a 3-percent salt solution with a pH in the range of 8
to 10. The ammonia is then removed from the brine by using
an ammonia removal and recovery system (ARRP). The ion
exchange and salt regeneration system serve to concentrate
the ammonia so that the ARRP process operates on a flow
stream only about 8 percent of that of the main plant.
The ARRP system includes a constant head box where sodium
hydroxide is added to raise the pH to 11.1 prior to flowing
to two 20-foot-diameter spent regenerant clarifiers. The
high pH clarified spent regenerant is then pumped to three
ARRP modules that each consist of 12-foot-diameter stripping
and absorber towers, interconnecting ductwork, and a 40-hp
fan. Ammonia is stripped from the spent regenerant solution
in the stripping tower and absorbed in the absorber tower
where the pH is maintined at 2 by addition of sulfuric
acid. The end result is that ammonia is removed from the
regenerant solution and recovered as concentrated ammonium
sulfate. The salt regenerant solution is used repeatedly
for regeneration of ion exchange beds and all recovered
ammonium sulfate is used as commercial fertilizer in the
Sacramento Valley of California.
The 2 to 3 mg/1 of ammonia remaining in the plant effluent
is removed by breakpoint chlorination so that the
combination of organic nitrogen, nitrates, and ammonia do
not exceed the 2-mg/l effluent limitation for total
nitrogen. Ion exchange used in conjunction with the ARRP is
believed to be the best process available to provide the
149
-------
ammonia removal required at this location. The very low
winter temperatures prohibit effective use of biological
nitrogen removal or conventional air stripping.
Wastewater Characteristics
The treatment plant receives wastewaters strictly from
domestic and commercial sources; there is no industrial
contribution. The wastewater characteristics are fairly
typical of a moderate-strength domestic sewage except during
spring snowmelt conditions when infiltration and inflow
dilute the wastewater.
Maximum plant loadings occur during August when waste strengths
are at a maximum due to the large number of tourists and
seasonal residents at that time. The plant was designed to
treat a maximum 7-day average flow of 4.83 mgd during this
period. Assumed August waste strengths used in design as
well as actual measured waste strengths are listed in Table 3.
TABLE 3
T-TSA RAW SEWAGE CHARACTERISTICS
ASSUMED AUGUST
DESIGN ACTUAL AUGUST
PARAMETER CONCENTRATION CONCENTRATION
BOD, mg/l 300 250
COD, mg/l 450 460
SUSPENDED SOLIDS, mg/l 220 220
TOTAL NITROGEN, mg/l 40 43
PHOSPHORUS, mg/l 15 11
150
-------
General Plant Performance
The plant has produced a high quality effluent since startup.
Table 4 shows average effluent quality for the first 12-month
period. The data in Table 4 indicates compliance with the
very stringent discharge requirements, except for limitations
on TDS and chlorides. The limitations were established for
these two parameters after the plant was designed, and there
are no processess included in the design to reduce these
dissolved inorganic constituents. The TDS and chloride con-
centrations in the raw sewage, coupled with increases through
the plant inherent with the chemical treatment and sludge
handling processes, essentially preclude the plant's ability
to control the concentrations of these consitutents in the
effluent. In recognition of these factors, T-TSA staff has
requested and hopes to receive a relaxation in the discharge
requirements for TDS and chloride.
TABLE 4
T-TSA PLANT PERFORMANCE
JUNE 1978 THROUGH MAY 1979
AVERAGE
EFFLUENT EFFLUENT
CONSTITUENT UNITS CONCENTRATION LIMITATIONS
COD mg/l 10 15
SUSPENDED SOLIDS mg/l 0.8 2
TURBIDITY NTU 0.6 2.0
TOTAL NITROGEN mg/l 2.6 2.0
TOTAL PHOSPHORUS mg/l 0.17 0.15
MBAS mg/l 0.06 0.15
TDS mg/l 584 440
CHLORIDE mg/l 206 110
TOTAL COLIFORM ORGANISMS MPN/100 ml <2 23
151
-------
In addition to the limitation on constituents in the effluent,
the discharge requirements have a year-round flow limitation
of 4.83 mgd for any maximum 7-day period. Since the plant
was designed for 4.83 mgd at maximum August sewage loadings,
the Agency contended that the plant could treat higher flows
during the winter when the sewage was diluted with infiltration/
inflow (I/I). They demonstrated this capacity by performing
a simulated test during August 1979. For 30 days the Agency
added river water to the sewage flow to produce a simulated
wet weather sewage flow diluted with I/I. The ratio was
one part of river water to two parts sewage. The plant
successfully treated these flows at a maximum 7-day average
flow of 5.64 mgd.
The plant is presently operating under a temporary waiver of
discharge limitations on COD and phosphorus. The Agency has
obtained approval to discharge higher concentrations of
these two constituents to evaluate their removal in the
effluent disposal system. Effluent is discharged to a
subsurface leach field and then percolates to the Truckee
River approximately 1 mile away. The removal of phosphorus
and COD is being evaluated by sampling monitoring wells
between the disposal field and the river. The objective is
to determine removal efficiency of the soil system to even-
tually reduce the level of treatment required in the plant.
The quality of effluent entering the Truckee River from the
disposal field under the land treatment system being evaluated
must, at all times, equal the quality specified in the plant
effluent limitations. If the test demonstrates the soil can
provide a high removal for a long period, there can be
substantial savings in chemical and energy cost in operating
the tertiary treatment facility. During this waiver period,
the plant is providing secondary treatment followed by
filtration and ammonia removal. The lime treatment system
and carbon adsorption are normally bypassed. To date,
removals measured in the soil system have been encouraging.
152
-------
Plant Clinoptilolite-ARRP Operation
Ammonia removal at T-TSA is accomplished by passing effluent
from the upstream carbon contacters through the ion exchange
system. As previously described, four individual ion exchange
beds are provided; these are operated in parallel and the
carbon column effluent passes through the system without
recirculation at a rate of from 6 to 10 bed volumes per
hour. Since only a finite number of ion exchange sites are
available and the contact time is limited by system geometry,
an unavoidable small amount of ammonium ion will escape even
a freshly regenerated bed. This, in the T-TSA system, is
approximately 1 mg/1 of ammonium ion. During use as the
number of available exchange sites is diminished, increasing
amounts of ammonium ion will pass the bed. Typically at
T-TSA, a bed is removed from service and regenerated when
effluent concentration of ammonium ion is between 2 and 3
mg/1. Given the influent conditions prevalent at T-TSA
(average 24 mg/1 NH .), ammonium breakthrough to the
2-mg/l level occurs after passage of approximately 129 bed
volumes. Figure 4 is a graphical representation of effluent
ammonium ion concentration in the system as a function of
total flow through a bed subsequent to regeneration.
At T-TSA recovery/reuse of the regenerant solution employed
to regenerate exhausted clinoptilolite beds is accomplished
by the ARRP system. In this system a number of processes
occur simultaneously. The first of these processes, that of
clinoptilolite regeneration, simply elutes ammonium ions
from the clinoptilolite by exchanging them for sodium which
is contained in the regenerant solution. This is accom-
plished by passing a sodium chloride brine solution through
the beds. At T-TSA, counterflow technology is employed to
achieve the goal of maximizing ammonia concentration in
spent regenerant. Towards this end, the regenerant solution
153
-------
1OA 039 6£l = 1VD 000'ZZ9'l
154
-------
is contained in four basins. Each of the four basins contains
(during operation of the ARRP system) regenerant solution
having varying concentrations of ammonia ranging from 250 to
50 mg/1. In operation, the regeneration sequence of an
exhausted clinoptilolite bed is as follows:
o Step 1—The exhausted clinoptilolite bed is back-
washed using carbon column effluent as the backwash
water source.
o Step 2—The exhausted clinoptilolite bed is purged
of backwash water by slowly pumping nearly exhausted
regenerant from basin No. 3 in an upflow mode
through the bed. Displaced liquid from the bed is
discharged to the backwash storage tank and step 2
is terminated by timers, calibrated by conductivity
probes.
o Step 3—Nearly exhausted regenerant from tank
No. 3 in the amount of 10 bed volumes is pumped
through the exhausted clinoptilolite bed. The now
totally spent regenerant is collected and stored
in regenerant basin No. 4.
o Step 4—Partially spent regenerant from basin
No. 2 in the amount of 10 bed volumes is pumped
through the clinoptilolite bed. This regenerant
is collected and stored in basin No. 3 and is
subsequently used during a succeeding regeneration
cycle in step 3.
o Step 5—Fresh regenerant from basin No. 1 in the
amount of 10 bed volumes is pumped through the
clinoptilolite bed. Discharge is collected and
stored in basin No. 2, where it is used in sub-
sequent regenerations in step 4.
155
-------
o Step 6—An additional 10 bed volumes of fresh
regenerant from basin No. 1 is recirculated through
the clinoptilolite bed and discharged back to
basin No. 1. This has the effect of lowering the
pH of freshly recovered regenerant from the stripping
process, which is simultaneously discharged into
basin No. 1.
o Step 7—The clinoptilolite bed is purged of the 1
bed volume of fresh regenerant remaining therein
after step 6 by pumping carbon column effluent
into the bed. Discharge is to basin No. 1.
o Step 8—The clinoptilolite bed is rinsed with
approximately 2 to 3 bed volumes using carbon
column effluent. Discharge is to the backwash
equalization tank. This has the effect of re-
moving residual brine and ammonia from the bed
thus reducing "spike" concentrations of ammonia
and TDS in the initial few bed volumes of dis-
charge. The "spike" which would otherwise occur
is shown on Figure 4 previously described.
Figure 5 is a graphical representation of the ammonia elutria-
tion curve which occurs during operation at T-TSA. Examination
of this figure indicates the two major advantages attributable
to the four-basin regeneration scheme: first, concentration
to the highest possible extent of the ammonium ion in basin
No. 4, and second, the ability to "polish" the bed in the
final stage of regeneration through the use of highly active
regenerant contained in basin No. 1. Figure 6 is a graphical
representation of the entire regeneration and regenerant
recovery system utilized at Tahoe-Truckee.
156
-------
ELUTRIATION
PHASE I (STEP 3)
BASIN NO. 3
PUMPED TO
BASIN NO. 4
PHASE 2 (STEP 4)
BASIN NO. 2
PUMPED TO
BASIN NO. 3
PHASE 3 (STEP 5) I PHASE 4 (STEP 6)
BASIN NO. 1
PUMPED TO
BASIN NO. 2
500
BASIN NO. 1
PUMPED TO
BASIN NO. 1
(RECYCLE)
2olBV 30 !BV
116,700 GAL 233,400 GAL 350,100 GAL
VOLUME OF REGENERANT SOLUTION PASSING THROUGH THE BED
CONCENTRATION OF NH4 -N IN
REGENERANT SOLUTION COMING
OFF OF THE BED
CONCENTRATION OF NH4 -N IN
REGENERANT SOLUTION GOING
INTO BED
NH4-N ELUTRIATED
CURRENT CYCLE
NH4-N ELUTRIATED IN
PREVIOUS CYCLES
40 BV
466,800 GAL
• FIGURES
TYPICAL AMMONIA ELUTION CURVE
157
-------
SPENT
SPENT REGENERANT ——x
REGENERANT —
HEADER.
CLINO
BED
REGENERANT
HEADER
REGENERANT-^
REGENERANT
RECOVERY
SYSTEM
*
NO.4
LOW '
VWEIR
-w-1 „
Hf'
INTERMEDIATE
WEIR
NO.3
NO.2
HIGH
WEIR'
NO.1
-REGENERANT
BASIN
TYP
-CONSTANT ^CLARIFI- "-WET ^ARRP
HEAD BOX CATION WELL
• FIGURE 6
T-TSA PLANT REGENERATION AND
REGENERANT RECOVERY SYSTEM
158
-------
A second concurrent process in the overall system is that of
regenerant clarification. In this process, spent regenerant
from basin No. 4 is pumped into a constant head box where
its pH is elevated to approximately 11.0 with sodium hydroxide.
This converts the ammonium ion to dissolved ammonia gas and
also precipitates magnesium (exchanged by the clinoptilolite
along with ammonia although to a lesser extent) from the
spent regenerant as magnesium hydroxide. The pH-adjusted
spent regenerant then flows into regenerant clarifiers where
the magnesium hydroxide sludge is removed by settling.
Supernatant flows over clarifier weirs and is stored in a
wet well for subsequent pumping into the ARRP modules.
The third major concurrent process is that of ammonia removal
and recovery. This function is accomplished in a closed
cycle air stripping system which strips ammonia gas from the
clarified regenerant solution and subsequently reabsorbs the
ammonia gas in a sulfuric acid solution. This process is
graphically depicted in Figure 7. In operation, the clarified
regenerant is sprayed onto and passes vertically downward
through the stripper tower media. Simultaneously a counter-
current closed-loop air stream flows upwards through the
media. The now ammonia gas-laden airstream exits the top of
the stripper and is directed into the bottom of a similar
absorbing tower where it passes upward, countercurrent to a
downward sprayed sulfuric acid solution (pH 2.2 to 2.6).
This scrubs the ammonia gas from the airstream and produces
a residual of ammonium sulfate which is collected in the
bottom of the absorber tower. The cleansed airstream is
continuously returned to the stripper by fans. Liquid
effluent from the stripper section is continuously returned
to regenerant storage basin No. 1 previously described.
The ammonium sulfate produced in the absorber tower eventually
reaches a saturation value of approximately 40 percent. It
159
-------
LLJ < CO
cc cc z
O C5
LU LLJ O
OC QC (-
160
-------
is blown down to a storage tank where a 50-percent sodium
hydroxide solution is added to neutralize pH. The ammonium
sulfate is subsequently sold as a commercial fertilizer.
OPERATION RESULTS14
The T-TSA clinoptilolite-ARRP system began initial operation
in February 1978. Construction was not then complete nor
had testing and shakedown operation been undertaken. Initial
operation was not without problems and effluent ammonia
concentrations ranged between 1 and 11 mg/1 during this
time. Performance of the nitrogen removal system improved
through the spring and summer of 1978 as final construction
was completed and the system was optimized. During the
period August through December, 1978, final plant effluent
was consistently under the 2-mg/l total nitrogen discharge
limitation.
Beginning in the late fall of 1978, the clinoptilolite beds
appeared to be loosing ammonia removal capacity. An exten-
sive investigation was made to identify the problems.
Operational and equipment modifications were made to the
nitrogen removal system during the winter of 1979. These
were completed by April 1979 and, for the balance of 1979,
final effluent was again maintained at less than 2 mg/1
total nitrogen. Figures 8 and 9 graphically portray both
ammonia concentration and TKN concentration in effluent as
well as other process points for the period February 1978
through September 1979.
Figure 10 presents daily data for a typical 30-day operation
period (April 1979). This figure shows the ammonia/nitrogen
concentrations in and out of the clinoptilolite beds and
that of the final effluent after chlorination. During this
161
-------
LU
o
SECpNDARY
EFFLUENT
20
15
10
JFMAMJJA
1979
• FIGURES
T-TSA PLANT
MONTHLY AVERAGE NH3-N
1978-1979
162
-------
45
40 •
35
CARBON COLUMN
A M J
1979
J A S
• FIGURE 9
T-TSA PLANT
MONTHLY AVERAGE TKN
1978-1979
163
-------
40
30
CARBON COLUMN-
EFFLUENT
= 2q
Ul
U
O
cc
z
O
1 10
CLINO BED
EFFLUEN
^FINAL
EFFLUENT
I
10
1 5
FIGURE 10
10 15
TIME IN DAYS
20 '
25
30
T-TSA PLANT
EFFLUENT NH3-N BY UNIT PROCESS APRIL 1979
164
-------
month, clinoptilolite bed influent ammonia-nitrogen concen-
tration averaged 27.4 mg/1 and the effluent averaged 2.5
mg/1 ammonia nitrogen. Average ammonia nitrogen removal was
91 percent. Final effluent contained 0.3 mg/1 ammonia
nitrogen and total ammonia nitrogen removal efficiency was
99 percent. Data are presented in Figure 11 for the same
time frame showing the relationship of the various forms of
nitrogen in the final effluent. This figure indicates that
during April 1979, ammonia nitrogen comprised 43 percent of
the total Kjeldahl nitrogen in the final effluent and organic
nitrogen averaged 57 percent (0.4 mg/1). Total Kjeldahl
nitrogen averaged 0.7 mg/1 nitrogen.
Data for April 1979 presented in Figure 12 show the effi-
ciency of the ammonia removal and recovery process. Average
ammonia nitrogen concentration in the spent regenerant was
248 mg/1 and the renewed regenerant from the closed cycle
air stripping system averaged 30 mg/1. This is an 88-percent
removal and recovery of ammonia. Later operation during the
summer of 1979, as the stripping system continued to be
optimized, achieved removal efficiencies as high as 95
percent.
OPERATION COSTS
Total capital cost of the T-TSA plant, constructed during
the period 1976 to 1978, was $19.98 million. Of this total,
$3.5 was for the nitrogen removal system. Operation and
maintenance cost for the T-TSA clinoptilolite-ARRP system
was characterized by a special study for a 3-month period
from February through April 1979. The results are as
follows:
o Chemicals, 9.16C per thousand gallons
o Electrical power, 2.20C per thousand gallons
165
-------
1 5 10
• FIGURE 11
15 20
TIME IN DAYS
25
30
T-TSA PLANT
FINAL EFFLUENT NITROGEN
APRIL 1979
166
-------
300
200
z
UJ
O
DC
Z
<
100
o l
TO
STRIPPERS
FROM
STRIPPERS
10 15
TIME IN DAYS
20
25
30
• FIGURE 12
T-TSA PLANT
AMMONIA REMOVAL FROM SPENT REGENERANT
BY CLOSED CYCLE AIR STRIPPING PROCESS
APRIL 1979
167
-------
o Direct labor, 6.21C per thousand gallons
o Indirect labor, 1.55C per thousand gallons
o Miscellaneous overhead and other expenses, 3.82C
per thousand gallons
This total of 22.94C per thousand gallons compares to a
total operation and maintenance cost for the plant during
the same time frame of $1.70 per thousand gallons.
Figure 13 presents anticipated capital cost curves for
similar installations of varying capacity. ' It is
based on an Engineering News Record Construction Cost Index
of 3,000. Figure 14 presents projected operation and main-
tenance cost for similar installations of varying capacities.
Basis of costs used are as indicated on Figure 14.
ENERGY REQUIREMENTS
The fact that approximately 10 percent of overall operation
and maintenance cost of this system is for power is indicative
of the amount of energy consumed when nitrogen removal must
be practiced. Figures 15 and 16 present estimated primary
energy requirements respectively for ammonia removal by ion
exchange including regenerant renewal by closed cycle air
stripping and primary energy requirements for ammonia re-
18
moval by ion exchange, excluding regenerant renewal.
Primary energy utilization is, however, only one aspect of
total energy consumption for a given system. For example,
when one includes the energy necessary to produce consumable
chemicals used in various nitrogen removal processes, a much
more complete and accurate evaluation results. For instance,
168
-------
100Q,
'8fOO
2 » 15678
"TOGO
FLOW MGD (U.S.)
COST BASIS
ENR INDEX = 3,000
EQUATION FOR CURVES
CH2M HILL COST $ U.S. = 925,500 x
(CAPACITY MGD) °-72
CWC
COST SU.S. = 422,160 x
(CAPACITY MGD) °-85
BECHTEL COST SU.S. = 418,660 x
(CAPACITY MGD) 0-93
• FIGURE 13
CONSTRUCTION COST FOR AMMONIA
REMOVAL BY SELECTIVE ION EXCHANGE
WITH REGENERANT RENEWAL BY CLOSED
CYCLE STRIPPING (ref. 15,16)
PLANT A = UPPER OCCOQUAN S.A.
B = TAHOE-TRUCKEE S.A.
C = DENVER WATER BOARD
(in design)
169
-------
10Q
|llBwiituMii«»ymBW^ stnwa****?™? ny»
f ~_ ~* „!" . ii~l
i " ". ~ ! * "" I , IT I
•0.01
cc
LU
g
O
a.
V)
cc.
in
co
_l
<
O
HI
I
CJ
1.0
' J " '" 10
FLOW MGD (U.S.
TOO
—J 0.001
1000
COST BASIS
POWER $0.02 (U.S.)/kWh
LABOR $7.50 (U.S.)/man hour
SULPHURIC ACID. .66° Be' $50 (U.S.)/ton
REGENERATION. . .40 bed Volumes/Cycle
THROUGHPUT . . . .100-150 bed Volumes/Cycle
REGENERANT . . . .2% NaCI Solution
SALT USE 0.1 lb/1000 gals (U.S.) throughput
CAUSTIC SODA . . .0.9 mg NaOH/mg NH4-N removed
• FIGURE 14
OPERATION AND MAINTENANCE COSTS
FOR AMMONIA REMOVAL BY SELECTIVE ION
EXCHANGE WITH REGENERANT RENEWAL BY
CLOSED CYCLE STRIPPING (ref. 1 7)
_
-------
10,000,000 '"
1,000,000
Si
s
c
LU
cc
D
a
LU
cc
o
x
LU
2
LU
<
O
O
LU
100,OOQ
w 10,000
1000
0.1
DESIGN ASSUMPTIONS:
"TOO
1,000
PLANT CAPACITY, mgd
EQUATION FOR CURVES:
kWh/year(wjth recovery) = 120,000 x
(CAPACITY MGD) 1-°
kWh/Vear(without recovery) = 6-667 x
(CAPACITY MGD) 1-°
REGENERANT SOFTENED w/NaOH AND CLARI-
FIED AT 800 gal/ft2/day
INFLUENT TO CLINO NH3-N 15 mg/l
THROUGHPUT 150 Bed
Volumes/Regeneration
REGENERATION 40 Bed Volumes
Regeneration/Cycle
REGENERANT STRIPPER 760 gal/ft2
AIR FLOW 4.23 m3 air/I
STRIPPER HEIGHT 32 ft
NH3 RECOVERY BY ABSORBER TOWER WITH
H2SO4
• FIGURE 15
PRIMARY ENERGY REQUIREMENTS FOR AMMONIA
REMOVAL BY ION EXCHANGE WITH REGENERANT
RENEWAL BY CLOSED CYCLE AIR STRIPPING
(ref. 18)
_
-------
10,000,OOQ
ID
0>
i
JC
a
LU
EC
§
HI
oc
o
oc
UJ
z
UJ
o
E
o
1,000,009
100,009
UJ 10,009
1,000*
10
2 ,45
? J 4 » " ?¥,ooo i J 4 s 6/s)B,ooo 2 3 4 s 6
CLINOPTILOLITE BED (4 ft depth), SQ FT
7ftib,ooo
DESIGN ASSUMPTIONS:
INFLUENT NH3-N 15 mg/l
INFLUENT SS < 5.0
THROUGHPUT 150 Bed Volumes
LOADING RATE 6 Bed Volumes/hour
HEAD AVAIL.TO GRAVITY 7.25 ft
PUMP HEAD FOR PRESSURE ... .10 ft
REGENERATION 40 Bed Volumes/Cycle
REGENERATION TIMING 1 Regeneration/day
REGENERATION PUMP HEAD . . .10ft
EQUATION FOR CURVES:
kWh/year(gravjty beds) = 9.0 x
(Bed Area - ft2 @ 4 ft deep) 1-°°
+ Regeneration
kWh/year(pressure beds) = 125 x
(Bed Area - ft? = 4 ft deep) ^ °°
+ Regeneration
kw"/Vear(regeneration) = 16-6? x
(Bed Area - ft2 @ 4 ft deep) 1 -00
• FIGURE 16
PRIMARY ENERGY REQUIREMENTS FOR AMMONIA
REMOVAL BY ION EXCHANGE EXCLUDING
REGENERANT RENEWAL (ref. 18)
172
-------
the expenditure of approximately 42,000,000 Btu's plus
4,000 kWh of electricity is required to produce 1 ton of
17 19
chlorine and 1.13 tons of sodium hydroxide. ' Using
conventional conversion efficiencies, this equates to the
expenditure of about 4,350 kWh of equivalent electrical
energy per ton of chlorine produced and about 3,850 kWh of
equivalent energy per ton of sodium hydroxide produced.
Similarly sulfuric acid production results in the utili-
zation of approximately 350 kWh of equivalent electrical
energy per ton produced. Sodium chloride, on the other
hand, is produced in many instances through solar evapora-
tion, hence its production energy requirements do not
necessarily result in utilization of scarce fossil fuels.
Taking both primary energy requirements and the secondary
energy requirement for consumable chemicals into account, a
nitrogen removal system serving a 10-mgd average flow plant
would utilize approximately 5.2 million kWh of equivalent
electrical energy per year if the ion exchange system
described herein were used, approximately 3.35 times as much
18
if a breakpoint chlorination system were used, and approxi-
mately 4.81 times as much if a reverse osmosis system were
18
used. On the other hand, if climatic conditions allow and
one disregards the energy requirements of high pH lime
coagulation (attributing these to phosphorus removal), an
air stripping system for the same plant flow and ammonia
concentration would require approximately 1.15 times as much
18
energy expenditure. Viewed in the light of these comments,
ammonia removal utilizing ion exchange and and closed cycle
air stripping for regenerant renewal is certainly among the
most energy-efficient systems available for ammonia-nitrogen
removal. In addition, its ability to produce (at far less
energy requirement than conventional means and without
dependence on natural gas,) ammonium sulfate fertilizer is
certainly a plus from an energy standpoint.
173
-------
OPERATIONAL PROBLEMS ASSOCIATED WITH AMMONIA REMOVAL BY
ION EXCHANGE AND REGENERANT RECOVERY BY CLOSED-CYCLE
AIR STRIPPING
Operational problems associated with ammonia removal by the
process described herein largely fall into two categories.
The first of these is the problem of scaling. Scale forma-
tion can be prevalent both in the clinoptilolite bed and
associated piping, as well as the regenerant renewal
stripping system.
At T-TSA a portion of the ammonia removal capacity of the
clinoptilolite media was temporarily lost due to scaling of
the media. Scaling of the media was a result of extremely
high pH conditions in the regenerant solution during the
initial operation of the system. During the pre- and post-
regeneration purge cycles, this high pH reacted with the
carbonate alkalinity in the carbon column effluent to form
calcium carbonate scale which effectively coated the clinop-
14
tilolite media, thus reducing its exchange capacity.
The media was subsequently washed in place with hydrochloric
acid which removed most of the calcium carbonate scale.
This restored essentially all of the exchange capacity of
the media but may have contributed to a loss of the media by
dissolving a slight amount of clinoptilolite. At T-TSA, a
permanent solution to the scaling problem of media in the
beds has been to operate the regenerant system at a lower pH
through the bed of the final 10 bed volumes of regenerant
14
solution as previously described.
Scaling of the media in the ARRP stripping towers has, on
occasion, significantly lowered capacity of the regenerant
14
recovery system. In order to meet the stringent discharge
limitation at T-TSA, the regeneration sequence must terminate
174
-------
with "polishing" the clinoptilolite beds with as fresh a
regenerant solution as possible. To achieve the low ammonia
concentration necessary for this last 10-bed-volume step,
the ammonia stripping efficiency in the ARRP system must be
maintained above 85 percent. Scaling of the stripping media
has, at times, reduced the stripping efficiency to 70 percent,
thus allowing the ammonia concentration in the renewed
regenerant to increase to a value as high as 200 mg/1. This
quite naturally results in lessened ammonia removal capacities
in the clinoptilolite beds.
Investigations into scale formation in the ARRP system
stripping towers revealed that the scale consists mainly of
calcium carbonate with lesser amounts of silica and magnesium.
Initially, scale was removed from the stripping tower teller-
ettes by acid cleaning. After a second acid cleaning, it
was noted that only approximately the top 14 inches of
tellerettes were clean and that below this 14-inch level, a
gray, acid-insoluble, gelatinous material was present on the
tellerette surfaces. This gelatinous material could be
hydraulically removed from tellerettes by washing; however,
the placement of tellerettes within the stripping tower made
obtaining sufficient velocity of washwater to remove the
material in place impossible. Eventually all tellerettes
had to be removed from the stripping tower to enable cleaning
of the entire depth. The procedure is laborious (approxi-
mately 80 man-hours are required to clean the three stripping
towers provided at T-TSA). The gelatinous residue consists
primarily of silica; its origin is quite likely from the
clinoptilolite media dissolved by earlier periods of operation
utilizing high pH regeneration and, possibly, from the acid
washing of clinoptilolite referred to earlier. Lowering the
pH of the regeneration solution has markedly decreased the
rapidity of deposition of the gelatinous material; however,
at full-scale operation, stripping tower cleaning is still
14
anticipated to be required about twice per year.
175
-------
Since the T-TSA clinoptilolite-ARRP system was placed in
operation in February 1978, approximately 60 tons of makeup
clinoptilolite have been added to the beds. This amounts to
a loss of about 20 percent per year. Besides the loss of
media during acid washing, other losses have been attributed
to attrition caused by the high regenerant solution pH and
loss during clinoptilolite bed backwashing. The rate of
clinoptilolite loss has been markedly reduced but it is
still somewhat higher than expected. The cause and solution
to this problem is still being investigated.
Potential Process Improvements
At the T-TSA installation, the clinoptilolite beds are
contained in horizontal cylindrical steel pressure vessels.
Inspection of the media in these vessels has indicated that
"dead" areas occur in the beds. These areas are the seg-
ments bounded by the curved sides of the tanks and a vertical
chord from the top side to the bottom side of the media in
the tank. Media in these side areas appeared septic and
contained large amounts of carbon fines which indicates that
these "dead" areas are not being backwashed and regenerated
properly. These dead spaces amount to approximately 10
percent of the total bed volume. With current clinoptilolite
prices over $300 per ton, it does not appear cost effective
to fill these dead areas with clinoptilolite. Potential
solutions include the use of vertical cylinders (which may
increase pumping head requirements); conventional filter
design with straight sidewalls (for larger plants); or,
where economics dictate the use of horizontal cylinders, the
blanking off of unavoidable dead space through the use of
internal baffles or filling with lightweight concrete appro-
priately secured to the cylinder shell.
176
-------
The regenerant renewal air stripping system at T-TSA does
not provide the opportunity to interchange the stripping and
absorbing sections. This was done because the absorbing
tower section can, in fact, be designed somewhat smaller
than the stripping section, thus resulting in economy of
construction. The two sections of the system could be
designed to be interchangeable. This feature would allow
alternating use of each section as the absorbing section.
This would mean that the sections would alternately be
exposed to the acid conditions prevalent in the absorbing
section. This would completely eliminate calcium carbonate
scale and may well reduce the buildup of gelatinous silica
scale.
Installation of hard-piped acid washing systems is another
simple improvement which has now been made at Tahoe-Truckee
to reduce labor requirements for cleaning.
Fan placement, nozzle selection, and mist eliminator designs
have been improved since the original prototype construction.
These improvements will reduce maintenance costs and will
result in slightly less power requirements for the operation
of the ARRP system.
As previously noted, the clinoptilolite beds at T-TSA were
designed to remove approximately 920 pounds per day of
ammonia nitrogen. The actual ammonia loading to the clinop-
tilolite beds at this plant has exceeded the maximum design
loadings for a significant portion of the time since initial
operation. The reasons for these higher than design loadings
have been that the raw sewage has a higher than expected
total nitrogen and ammoniacal nitrogen content, and supernatant
returned from the anaerobic digestion process as well as
filtrate from the dewatering of digested sludge has resulted
177
-------
in a higher than expected ammonia recycle to the clinoptilo-
lite system. At T-TSA, digested sludge is dewatered by
filter press after conditioning with lime and ferric chloride.
Filtrate from the press has a pH between 11.5 and 12 and a
temperature ranging between 80°F and 85°F. These are optimum
conditions for stripping ammonia nitrogen. The press filtrate
contains ammonia nitrogen concentrations in the range of
from 700 to 1000 mg/1. In July 1979, two small temporary
stripping units were constructed to treat this filtrate.
Effluent from these two pilot strippers is reduced in
ammonia concentration to approximately 150 mg/1. The two
strippers have been operating to remove from 100 to 150
pounds of ammonia nitrogen per day from the filtrate, thus
reducing by that amount the load which would otherwise
report to the entire clinoptilolite system. This is ex-
tremely significant since the removal of this significant
amount of ammonia without the use of clinoptilolite is
extremely cost effective. Figure 17 is a simplified nitro-
gen balance for the T-TSA system based on operation at
g
design flows. As can be seen from Figure 17, the permanent
installation of a filtrate stripping system could relieve
the entire clinoptilolite system of the necessity of removing
an additional 400 pounds per day (equivalent to one-third of
the raw sewage ammonia content).
The use of the ARRP stripping system for ammonia removal
from high pH filtrates may well hold promise for use at
other plants where complete ammonia removal is not required
or where oxygen capacity for nitrification is not available
to handle such return flows.
PROCESS ADVANTAGES AND DISADVANTAGES
The advantages of selective ion exchange for ammonia nitro-
gen removal are: efficient ammonia removal in all seasons,
178
-------
TKN = 1730 Ib/day TKN
/NH3 = 1200 Ib/day /NH3
I|NO2+NO3 > 0 Org N
II NO2+NO3
I DAl/U _„,_„„ _ .
IWASTE- 1 1 \
VWA™ J PRIMARY + [V
M W • __ -_____-._ \J ™
1 TREATMENT
^-REMAINING
PRIMARY <• RECYCLE
AND NH-, 40 Ib/day
WASTE
ACTIVATED i '
1 ANAEROBIC 1
DEWATERED DEWATERING IFILTRATE
SLUDGE |NH3 =
TKN = 310 Ib/day
NITROGEN OUT
WITH DEWATERED SLUDGE = 310 Ib/day
TO CLINO REGEN SYSTEM * 920 Ib/day
TO BREAKPOINT CI2 * 200 Ib/day
TO AIRSTRIP *= 400 Ib/day
WITH EFFLUENT ^ 80 Ib/day
SUBTOTAL 1710 Ib/day
UNACCOUNTED FOR 20 Ib/day
• FIGURE 17
= 1180 Ib/day TKN
= 1140 Ib/day /NH3
40 Ib/day NO2+NO3
20 Ib/day lOrgN
\
^™""^™1V I
CLINOPTILOLITE! \
AND IV .J
REGENERANT P^^^^l
STRIPPING 1
I
I STRIPPING
I
I I
TO ATMOSPHERE
AS N
= 260 Ib/day
= 220 Ib/day
= 20 Ib/day ,_ N = 2QQ |b/d
= 40 Ib/day 2
1
BREAKPOINT I rv
CHLORIIMATION P1111^"^!^^
FT
TKN =60 Ib/day'
NH3 - 20 Ib/day
NO2+NO3 = 20 Ib/day
OrgN = 40 Ib/day
T * 1 M ^ QA 11^ /ft n\*
Total N - oO Ib/day
r
TO (NH4)2SO4 TO (NH4)2SO4
NH3 = 400 Ib/day NH3 =
920 Ib/day
SIMPLIFIED NITROGEN MASS BALANCE
FOR T-TSA AWT PLANT AT 4.83 MGD
(1 8,286 m3/DAY)
FLOW RATE
179
-------
and reliable efficiency of ammonia removal regardless of
variation in influent ammonia concentration. Using the ARRP
process for regenerant recovery provides a means of mini-
mizing the amount of waste for final disposition. Furthermore,
a commercially valuable byproduct fertilizer is produced.
Process disadvantages include the following: the selective
ion exchange process is accompanied by a complex regener-
ation and regenerant recovery problem. Many unit processes
are of necessity included, and these must be operated in
concert with one another. The calcium concentration of
influent to the clinoptilolite beds can be a significant
consideration. Above approximately 40 mg/1 (as calcium)
influent calcium concentration will reduce bed volumes to
exhaustion by a ratio of approximately 2 bed volumes per
milligram per liter of calcium. In some cases, this may
require softening of the influent to maintain service cycle
of the clinoptilolite beds between regenerations. Clinopti-
lolite resin is a soft material and attrition of the resin
due to backwashing and rinsing of the beds is significant.
Scaling of the media is also a potential disadvantage that
must be carefully considered in design. Since ammonia is
exchanged for sodium in the process, there is a net unavoid-
able increase in sodium, chloride, and TDS concentrations
which may be of concern to water supplies downstream of the
wastewater discharge. At the T-TSA plant, experience to
date has shown an increase in chloride and total dissolved
solids of approximately 140 mg/1 and 250 mg/1, respectively.
This is due primarily to the clinoptilolite-ARRP process.
SUMMARY
The selective ion exchange and closed cycle air stripping
regenerant renewal system at the T-TSA Sanitation Agency's
Wastewater Reclamation Plant was initially placed into
180
-------
operation in February 1978. Operating problems which have
at times occurred within the system have been eliminated or
reduced to the extent that the system was shown to be capable
of consistently producing an effluent meeting a discharge
requirement of 2 mg/1 per liter total nitrogen.
Operating costs for nitrogen removal are high, regardless of
whether physical-chemical or biological processes are employed,
The operating costs for the clinoptilolite-ARRP system are,
by comparison with other available systems, reasonably
attractive considering the strict effluent nitrogen limita-
tions imposed. In terms of constant dollars, these costs
can and will undoubtedly be reduced somewhat as additional
experience is gained with the system and as flows approach
design capacity.
The air stripping system for regenerant renewal has been
shown to be effective as a direct means of nitrogen removal
from certain recycle streams such as those emanating from
the dewatering of lime conditioned anaerobically digested
sludge. The overall system including the clinoptilolite ion
exchange merits consideration at other locations where
extremely low effluent ammonia/nitrogen concentrations may
be required.
/RPT65A
181
-------
BIBLIOGRAPHY
1. Gonzales, John G. and Gulp, Russell L. "New Developments
in Ammonia Stripping." Public Works. May 1973.
2. Metcalf and Eddy, Inc. Wastewater Engineering. McGraw-Hill
1972.
3. Water Pollution Control Federation. Wastewater Treatment
Plant Design (MPO-8). 1977.
4. Eckenfelder, W.W., Manual of Treatment Processes.
Vol. 1. Environmental Science Services Corporation.
1968.
5. Masselli, J.W., et al. "The Effect of Industrial
Wastes on Sewage Treatment." New England Interstate
Water Pollution Control Commission. June 1965.
6. National Academy of Sciences. Drinking Water and
Health, Washington, D.C. 1977.
7. U.S. Environmental Protection Agency. Federal Register.
45FR803, January 3, 1980.
8. Suhr, L. Gene and Hamann, Carl L. "Fundamentals of
Physical Chemical Processes for the Removal of Nitrogen
Compounds from Wastewater", Nutrient Control Technology
Seminar, Calgary, Alberta, Canada. February 1980.
9. Battelle-Northwest, "Ammonia Removal from Agricultural
Runoff and Secondary Effluents by Selective Ion Exchange."
Robert A. Taft Water Research Center Report No. TWRC-5.
March 1969.
182
-------
10. U.S. Environmental Protection Agency. Technology
Transfer. Process Design Manual for Nitrogen Control.
October 1975.
11. Suhr, L. Gene, and Kepple, Larry G. "AWT Plant Gets
Tough With Ammonia," Water and Wastes Engineering,
March 1975.
12. Suhr, L. Gene and Kepple, Larry G. "Design of a
Selective Ion Exchange System for Ammonia Removal."
Unpublished.
13. Suhr, L. Gene and Evans, David R. "Ammonia Removal
from Wastewaters by Selective Ion Exchange," Nutrient
Control Seminar, Kelowna, British Columbia, Canada.
October 1975.
14. Prettyman, Raymond D., et al. "Ammonia Removal at
Tahoe Truckee Advanced Wastewater Treatment Plant—First
Year in Review." Unpublished.
15. CWC Consulting Engineers. "Estimating the Costs of
Wastewater Treatment Facilities." Virginia State Water
Control Board Report. March 1974.
16. Bechtel, Inc. A Guide to the Selection of Cost Effective
Wastewater Treatment Systems. U.S. Environmental
Protection Agency Report. No. 430/9-75-002. July 1975.
17. EPA (Burns and Roe Consulting Engineers). Innovative
and Alternative Technology Assessment Manual. U.S.
Environmental Protection Agency Report No. 430/9-78-009.
1978.
183
-------
18. CWC Consulting Engineers. Energy Conservation in
Municipal Wastewater Treatment. U.S. Environmental
Protection Agency Report No. 430/9-77-011. March 1977,
19. Bardie/ D.W.F. Electrolytic Manufacture of Chemicals
From Salt. The Chlorine Institute, Inc., New York.
1975.
/RPT65A
184
-------
NITRIFICATION AND PHOSPHORUS REMOVAL IN A
35 MGD ADVANCED WASTE TREATMENT PLANT AT ROANOKE, VIRGINIA
Donald E. Eckmann, Partner
Alvord, Burdick & Howson-Chicago
Harold S. Zimmerman, Plant Manager
Waste Treatment Plant, Roanoke, Va
GENERAL DESCRIPTION
The Wastewater Treatment Plant described in this paper is
located in Roanoke, Virginia, United States of America. Roanoke
is a city in the State of Virginia and is near the east coast of
the North American Continent, about 37° N. latitude, 80° W.
longitude. The plant is about 900 ft. above mean sea level and
the average monthly temperature in the area varies from a low of
39° F in December to a high of 76° F in July. The annual
rainfall is about 43", varying from 2.9 to 4.9" per month during
the year.
HISTORY
The Roanoke Waste Treatment Plant is owned by the city and
operated with city personnel. It serves the cities of Roanoke,
Salem, Roanoke County, and the Town of Vinton, which has a
combined population of approximately 200,000. The plant was
originally constructed as a 14 MGD activated sludge type plant
in 1950 and by' 1970 had been increased to 21 MGD with
disinfection facilities. During the 1960's the Appalachian
Power Company constructed a dam downstream of the plant outfall
on the Roanoke River, creating one of the finest recreational
lakes in southwest Virginia. The current National Pollutant
185
-------
Elimination System permit requires the plant effluent meet the
following average monthly requirements:
NPDES PERMIT REQUIREMENTS
5-day Biochemical Oxygen Demand (BOD) - 5 mg/1
Suspended Solids (SS) -2.5 mg/1
Phosphorus (P) - 0.2 mg/1
Total Kjeldahl Nitrogen (TKN) - 2.0 mg/1
Fecal coliform (FC) - 200/100 ml
pH - 6 to 9
Chlorine residual - 1.5 to 2.5 mg/1
The plant with an additional 14 MGD of secondary and 35 MGD of
tertiary treatment facilities was placed in operation in January
1977 and operating data for its last full calendar year of
operation, 1979, was selected for presentation in this paper.
SAMPLING
All of the data presented in this paper was taken from operating
reports which are based on test results taken on samples tested
in the plant laboratory. The laboratory has been verified by
the State of Virginia and is used for testing samples from other
waste treatment plants. The samples of the raw sewage and final
effluent are taken hourly 7 days a week and proportioned to
flow. Samples of intermediate treatment processes are taken in
the same way but only 5 days a'week.
PLANT PERFORMANCE
The Roanoke plant has a design capacity of 35 MGD. During 1979
it treated an annual average day of 30 MGD which is
186
-------
approximately 86% of its design capacity and produced the
following effluent:
Annual Average Day Final Effluent
BOD -1.6 mg/1
SS - 1.1 mg/1
TKN - 0.4 mg/1
Phosphorus - 0.18 mg/1
Fecal Coliform - 4.3/100 ml*
*Geometric mean
(Average monthly flows are shown in Table 1.)
Figures 1 through 5 were prepared to show the relation of daily
to annual average values for sewage flows, suspended solids,
BOD, phosphorus and TKN. These graphs were prepared by
organizing the data for each item from its maximum to minimum
daily amount. It can be seen from Figure 1 that the daily flows
during the year were typical of what would be expected at a
Waste Treatment Plant. The maximum day flow was 63 MGD which
was about twice the annual average day flow and the minimum day
flow was 19 MGD which is about two-thirds of the annual average
day flow. Figure 1 shows the design capacity of the plant of 35
MGD was exceeded about 20% of the time or 73 days during 1979.
Figures 2 through 4 show the daily quality of the final effluent
from the plant. From Figure 2 it can be seen that the suspended
solids in the final effluent exceeded 2.5 mg/1, about 10% of the
time and 1 mg/1 about 30% of the days during the year. There
were several days during the year in which the final effluent
187
-------
(0
EH
4J
C
rc
rH
cu
4J
Q)
4J
(1)
0} rC
rO -H
S C
•H
O -H
C >
fO
> -
T3 (!)
O
I C
I (C
CO OH
-p
cn
c
•H
4J
(U
CTl
188
-------
0)
0)
CO
0)
en
to
S-i
O
OJ CM
o .
-P C O 03
03 oS C -P
>
0) TD
DH
-------
U)
TD
o
CO
0>
d
0)
m
CO
0)
en
03
0)
> -P
< d
03
iH i-H
(0 DJ
D
d -P
d d
< 0)
o -P
-P (0
0)
o
a
0)
en
d
•H
-P
(C
•H
(0 OJ
Q -U
cn
M-l (0
o s
C T3
o d
m m
•H >
a> T3
CN
a>
jj
a
en
•H
O
-r-t
d <^
•H r-
Jj M
•H
> S
O
- i~i
0) fci
,v
O (0
d -P
03 03
o a
190
-------
m
Q
O
CQ
0)
tn
(0
>j
0)
>
C
(0
rH <-\
(0 GJ
3
C -U
C C
g
o
C ^ ^ r-l
O CU D &J
•H u ,y
-u c o fC
(0 tO C -P
rH > <0 (0
CD TD O Q
S-i
3
cn
•H
191
-------
r
i
4-
•
—
— — «
"
• 1 **•
^
f §
1
I
en
3
O
^
O
s:
o
m
j-i
0)
> -P
S
O
* L>
0) CD
^:
o
-------
0)
Cn
m
xj
-u
< c
(0 Oi
3
C -H
c c
< OJ
g
O 4J
^j n3
QJ
tn
4J
!u
O
Q4
0)
C
-H
4-1
(C
S-l
OJ
a
O
•r-l
(C 0
Q -U
CO
U-l (C
O S
O <1>
•H U
-P C
0 TD
in
a;
i~i
3
en
-H
C
•H
> e
O
«• SJ
0) CD
^
O (0
c -u
(0 ns
193
-------
contained high suspended solids which was the result of
malfunctioning equipment so the graph depicts the effluent
results which can be realistically obtained with an operating
plant.
Figure 3 is a platting of the daily BOD in the final effluent.
It never exceeded 8 mg/1 during the year and 3 mg/1 was exceeded
only about 8% of the time.
Figure 4 shows the daily phosphorus in the final effluent of the
plant during the year 1979. It can be seen from the figure the
maximum day phosphorus in the final effluent did not exceed 2
mg/1 and about 25% of the time the plant produced an effluent
with less than 0.2 mg/1 of phosphorus. The maximum day TKN was
7.3 mg/1 and 1 mg/1 was exceeded on about 40 days during 1979.
PLANT PERFORMANCE DURING PERIODS OF HIGH FLOW
During the 12-day period from April 8 through April 19, 1980,
the plant treated an average daily flow of 52.3 MGD which was
about 150% of its design capacity. The minimum daily flow
during that period was 44.6 MGD and the maximum was 70.2 MGD.
The concentration of the raw sewage and the summary of the plant
performance is shown in Table 2. It can be seen from that table
that during this 12-day period of high flows the plant produced
the following effluent:
BOD - 1 mg/1
SS - 2 mg/1
Phosphorus - 0.1 mg/1
TKN - 0 mg/1
194
-------
Table 2
Summary of Plant Performance
During High Flows*
Roanoke, Virginia
(April 8 to 19, 1980)
Residuals in Effluent-mg/1
BOD J3S P TKN
Raw Sewage 144 173 4.4 6.6
Primary Basin Effluent 57 64
Activated Sludge Process Effluent 27 64 1.0 1.0
Nitrification Process Effluent 3 17 0.5 0
Flocculation - Coagulation
Process Effluent 3 12 0.3
Filtration - Disinfection
Process Effluent** 1 2 0.1 0
*Daily Flows varied from 44.6 to 70.2 mgd and averaged 52.3 mgd
(149% of Design Flow)
**Final Effluent
195
-------
DESCRIPTION OF FACILITIES
The plant consists of primary settling and the activated sludge
processes followed by the suspended growth nitrification
process, the flocculation-coagulation process, filtration, and
disinfection. Figure 6 is a schematic diagram of the plant.
Comminuting devices reduce the size of solids in the raw sewage
prior to the raw sewage pumps lifting the sewage from the wet
well to the grit basins in which mostly inorganic materials are
removed. From the grit basins flow proceeds through the primary
settling basins which during 1979 had an average hydraulic
loading of 1000 gpd/sf of surface area. The activated sludge
process consists of aeration basins which have a hydraulic
detention time of 6 hours at design flow, excluding return
sludge. However, during 1979 aeration equipment was being
replaced in two of the basins and therefore the average
detention time of the aeration basins was about 5.6 hours. The
activated sludge settling basins had a hydraulic loading of 600
gpd/sf. The nitrification process consists of nitrification
aeration basins and nitrification settling basins. Based on the
annual average day flow of 30 MGD during 1979, the nitrification
aeration basins had a detention time of 4.2 hours and the
nitrification settling basins had a hydraulic loading of 514
gpd/sf.
For reducing phosphorus from 11.9 mg/1 in the raw sewage through
the activated sludge process to 2.4 mg/1, 1.07 gallons of pickle
liquor were utilized per pound of phosphorus removed. This
196
-------
Figure 6. Basic Flow Diagram
Advanced Waste Treatment Plant
Roanoke, Virginia
197
-------
amounted to 85 gallons of pickle liquor per million gallons of
sewage treated.
Chemical costs for the pickle liquor and ferric chloride were
$10,535 during 1979 which resulted in a unit cost of $12.14 per
1000 pounds of phosphorus removed, or $0.96 per million gallons
of sewage treated. For these dollars 80% of the incoming
phosphorus was removed. There is an unexpected benefit in the
nitrification process which removes about 1 mg/1 of phosphorus.
The next process in which chemicals are added for phosphorus
reduction is the flocculation-coagulation process. Aluminum
sulfate, commonly referred to as alum, is added there. Based
upon the reduction of 1.3 to 0.39 mg/1 of phosphorus and the
amount of alum used during the year, 47.8 Ib. of alum were used
per pound of phosphorus removed. This is considerably more than
the theoretical 9.5 Ib. of alum required to remove a pound of
phosphorus and demonstrates the efficiency of chemicals to
precipitate phosphorus at low levels is not good.
Overall chemical costs in 1979 for pickle liquor, ferric
chloride and alum amounted to $26,118. Based on reduction of
phosphorus from 11.9 mg/1 in the raw sewage to 0.18 mg/1 in the
final effluent, chemical costs amounted to $24 per 1000 pounds
of phosphorus removed.
The next major process in the treatment chain consists of a
rapid mix basin with about 1 min. detention time, flocculation
basins with 35 minutes and coagulation basins which had a
hydraulic loading on the annual average of 685 gpd/sf.
198
-------
Following the coagulation basins there are dual media, down flow
gravity filters which were designed for 3 gpm/sf and were
operated at an average hydraulic rate of 2.6 gpm/sf during
1979. The average chlorination detention time was 35 minutes
prior to discharge of the final effluent to the Roanoke River.
PHOSPHORUS REMOVAL
Table 3 shows the phosphorus in the raw sewage and effluent of
the different plant processes. Through the activated sludge
process, phosphorus is reduced 80% on the annual average from
11.9 to 2.4 mg/1. This is accomplished primarily in the
aeration basins by the addition of pickle liquor to the effluent
of the primary settling basins which is the influent to the
activated sludge aeration basins. The pickle liquor used is a
ferrous chloride type which contains about 10% iron by weight.
The pickle liquor is delivered to the site by railroad and
during 1979 there were some problems with delivery so commercial
liquid ferric chloride was used intermittently during the year.
For convenience, the gallons of ferric chloride were converted
to equivalent gallons of pickle liquor by multiplying the
gallons of FeCl3 used by a factor of 1.62. However, costs
reflect the actual expenditures for ferric chloride and pickle
liquor.
The addition of chemicals has some effect on the pH of the
sewage. The following is a pH profile through the treatment
plant.
199
-------
Table 3
Annual Average Day Phosphorus Reduction
Advanced Waste Treatment Plant
Roanoke, Virginia
(1979 Data)
% Reduction from
mg/1 Raw Sewage Chemical Costs
Raw Sewage 11.9
Activated Sludge Effluent 2.4 80% $ 10,535
Nitrification Process
Effluent 1.3 89%
Coagulation Basin Effluent 0.39 97% 15,583
Final Effluent 0.18 98%
TOTAL $ 26,118
200
-------
Location pH
Raw sewage 7.5
Primary effluent 7.5
Activated sludge process effluent 7.3
Nitrification process effluent 7.3
Coagulation basin effluent 7.2
Final effluent 7.1
Figure 7 has been prepared which shows the average monthly
phosphorus reduction through each process in the plant for which
data are available.
Pickle liquor, which is a waste product from the steel process-
ing industries, is hauled about 200 miles by rail to the Roanoke
plant. Facilities for adding chemicals to the activated sludge
plant cost $310,000 in 1972 but included a spur track and switch
off the main Norfolk and Western Railroad track, a bridge across
the Roanoke River and two 50,000-gallon lined steel pressure
vessels. The system was designed to feed alum, ferric chloride
or pickle liquor. Only ferric chloride and pickle liquor have
been used to date. The pickle liquor is delivered to the rail
siding in 9500 gallon tank cars and transferred in about 2 hours
from each car across the river to the storage tanks utilizing an
air pad on the tank cars. An air pad is also used on the
storage tanks for transporting the pickle liquor which has a pH
of less than 1-, to a rotodip type of chemical feeder. The
system eliminates the need for the pickle liquor to come in
contact with any moving parts except a corrosion-resistant
dipper in the chemical feed machine. The system has operated
satisfactorily now for 8 years.
201
-------
4-)
c
(0
0)
£.
4J
C
c
o
o
D
# c
(0
C/3 ,-1
3 01
o
Lj j_J
O C
CX £
10 4->
0 fO
jC 0)
EH -H
iH d) -H
x; 4-i cn
4J CO S-l
C 03 -H
O £ >
s
(D OJ QJ
CT> U 4^
fO C O
0) > OJ
> T3 O
202
-------
At the point the pickle liquor is introduced into the primary
settling basin effluent, there are two mixing units which have a
total velocity gradient, G = 500 ft/sec/ft.
NITRIFICATION
During 1979 the annual average day Total Kjeldahl Nitrogen in
the raw sewage was 13.8 mg/1 and it was reduced 97% to 0.4 mg/1
in the final effluent as shown on the following table.
Annual Average Day Total Kjeldahl Nitrogen Reduction
Advanced Waste Treatment Plant
Roanoke, Virginia
% Reduction
from
mg/1 Raw Sewage
Raw sewage 13.8
Activated sludge process effluent 3.5 75%
Nitrification process effluent 0.6 96%
Final effluent 0.4 97%
The temperature of the raw sewage averaged 68° F and for the
three coldest months of January, February, and March averaged
59° F during 1979. Figure 8 shows the daily temperature varied
from a low of 50° F to a high of 83° F. There were 9 days under
55° F.
The monthly average TKN in the final effluent is generally less
than 1 mg/1. However, during April it averaged 1.7 mg/1 and the
effluent from the nitrogen process was 2.2 mg/1, which was
higher than the 2 mg/1 design of the system. We believed it
would be of interest to compare the operating conditions for the
months of January, February, and March with the operating
203
-------
3
-U
0)
a
en
OJ
OJ
en
-P
< c
(C
•H ,-H
(D CU
D
C -P
C C
< 0)
O -P
-u) (0
en
0)
en
c
•r-l
4J
0>
•v r nift
>i X-i "3 O
•-I E-< -H
•H c O>
03 ^J rH
m (d -H
o s > e
o
C 73 - SJ
O CD (U Cu
US
00
0)
Vj
3
Cn
•H
C O 03
03 C -U
> tO (0
'nod
204
-------
conditions for the first week in April when the effluent
averaged 3.5 mg/1 TKN. During the months of January through
March/ the influent to the nitrification aeration basins had an
average BOD that ranged from 33 to 55 mg/1 and a TKN that ranged
from 6.6 to 8.0 mg/1. The loadings on the aeration basin varied
in BOD from 13 to 25 Ib. per 1000 cu.ft. and TKN ranged from 2.6
to 3.6 Ib. per 1000 cu.ft. The detention period in the aeration
basins based on the flow of raw sewage only varied from 3.3 to
4.0 hours. The actual detention time because of the 43 to 50%
return sludge is less. Dissolved oxygen was maintained between
9.2 and 10.8 mg/1. This D.O. required between 0.51 and 0.74
cu.ft. of air per gallon of sewage treated. Using the nitrogen-
oxygen demand of 4.6 TKN, the calculations show between 963 and
987 cu.ft. of air was used per pound of BOD and TKN applied.
Settling basin overflow rates ranged from 542 and 641 gallons
per day per square foot based on raw sewage flow. Under these
conditions the TKN in effluent from the nitrification process
ranged from 0.8 to 1.1 mg/1.
During the first week of April the BOD and TKN loading and the
aeration detention time were all within the range experienced
during the months of January through March. The pH was slightly
lower, 7.0 vs. 7.1 to 7.2, however, the temperature was higher
67° F vs. 59 to 60° F. The cubic feet of air used per pound of
BOD and TKN (using an NOD of 4.6 TKN) was 545 which resulted in
a DO of 3.7 mg/1. The settling basin hydraulic loading was
within the range experienced during the months of January
through March.
205
-------
From the data it is apparent that maintaining the dissolved
oxygen in the nitrification aeration basin of over 3 mg/1 is not
sufficient to produce a properly nitrified effluent. It is
necessary to use at least 963 cu.ft. of air per pound of BOD and
TKN applied to maintain a concentration of around 1 mg/1 TKN in
the effluent from the process. Detention periods can be as
short as 3.5 hours with a sewage temperature of 60° F and pH of
7.1.
Operating data for 1979 (Table 4) shows that the plant utilizes
1.7 cu.ft. of air per gallon of sewage treated, or about 1600
cu.ft. per pound of BOD and TKN removed (using an NOD of 4.6
TKN). The plant continues to utilize about the same cubic feet
of air per gallon of sewage in the activated sludge system since
the tertiary treatment plant was built (see design criteria,
Table 5). Of the total amount of air blown, about 40% is used
in the nitrification system which, based on the total cost of
energy for the blowers in 1977 amounted to $122,000 for the
nitrification system. On the annual average, there is no
decrease in the 7.3 pH through the nitrification system. The
alkalinity decreases from 133 mg/1 in the influent to 106 in the
effluent which is about 9 mg/1 per 1 mg/1 of TKN removed.
It can be seen from Figure 5 the plant consistently produces TKN
of less than the standard of 2 mg/1 more than 95% of the time
during the year and the days on which the limit is exceeded are
largely the result of insufficient air blown to the nitrifi-
cation aeration basins.
206
-------
-p
c
fO
P-H
^r
•H
4->
•H W
< • cO
H £
fu CO (D
o o w
CD
Q
0)
M cu co
T) W) >
3 0
H r-l
-P *Z
C «
0) EH
5 0
-
VD O^vCM
K^itAr^
[>>£> t^-
C^-OJ >-
VDvD^O
rOH
rH rH VO
rH CO O
O^iH O~N
H
O CTvCT\
VD LHlTv
Settling Basins
Aeration Basins
H
rH
-P 2
£EH
H
^
o cd
°.^o
OOiJD
J-
X) C
•H
00
X)
CTi
rH
in
d
H
O^v
rO
H
rH
CM
CT.
XI
in
o
H
p-
s
VO
>o
d
rH
Pi
P-
VD
ro\
H
in
V£)
J-
o
in
-H-
m
>x>
rA
O
O
f^
00
CM
CO
H
in
rA
H
P-
(A
[^ O K> P~
rA m
•
CM
K\
H
£)
D
CJv
CM
K-\
^3
CM
00
rH
00
^O
H
cA
^t
vO
l<^
in
CO
S
o.
-------
Table 5
Design Criteria
Advanced Waste Treatment Plant
Roanoke, Virginia
Avg. flow
Max. flow-'-
Item
Comminutors
Raw Sewage Pumps
Grit Basins
Primary Setting
Basins
Flow Equalization
Basins
Aeration Basins
Return Activated
Sludge
Secondary Settling
Basins
Nitrification
Aeration
Return Nitrified
Sludge Pumps
Nitrification
Settling Basins
Blowers
Rapid Mix
35 mgd
60 mgd
Design Data2
72 mgd capacity
62 mgd capacity
47 mgd based on 95%
removal of 100 mesh grit
Surface loading
1200 gpd/SF
Detention 1.5 hrs
30 mg storage
Detention 6 hrs
23 mgd capacity
65% return
Surface loading-700gpd/SF
Detention-2.5 hrs
Detention 3.6 hrs
30 mgd capacity
85% Return
Surface loading-600 gpd/SF
Detention 3.7 hrs
85,000 cfm capacity
3.5 cf air/gal sewage
Detention 1 min. vel. gradient-
G = 500 ft/sec/ft
iDesign Capacity of Interceptors
2Surface Loading & Detention based on 35 mgd
208
-------
Table 5
Design Criteria
Advanced Waste Treatment Plant
Roanoke, Virginia
(Continued)
Item
Flocculation
Coagulation
Filters
Wash Water Pumps
Chlorine Contact
Basins
Design Data2
Detention 30 min vel. Gradient
1st stage-100 ft/sec/ft
2nd stage- 60 ft/sec/ft
3rd stage- 20 ft/sec/ft
Surface loading-800 gpd/sf
Detention-2.7 hrs
3 gpm/SF
Media
20" anthr. es. 0 . 9 mm
uc 1.8
10" sand es. 0.6 mm
uc 1.4
14 gpm/sf, 22" rise
Detention-30 min
^Design Capacity of Interceptors
2Surface Loading & Detention based on 35 mgd
209
-------
SLUDGE
There are four points in the treatment process from which sludge
is removed: the primary settling basins, the activated sludge
process, the nitrification process, and the flocculation-
coagulation basins. The waste wash water from the filters is
recycled through an equalizing basin to the raw sewage. The
sludge from the primary settling basins and coagulation-
flocculation process is commingled in a gravity thickener which
thickens the sludge to about 5% solids before it is discharged
to digesters. The waste activated sludge which averages about
0.8% solids and the waste nitrified sludge which averages about
0.6% solids is thickened to about 4% in a flotation type
thickener and also discharged to digesters. Sludge was wasted
from the nitrification system a total of 150 days during 1979 at
an average rate of 102,000 gallons per day. The total amount of
sludge from the digesters amounted to 1837 dry pounds of solids
per million gallons of sewage treated.
ENERGY REQUIREMENTS
The raw sewage pumps are driven with engines which can be run on
either digester gas or natural gas. The air blowers which are
used solely for the activated sludge process and the nitrifi-
cation process are also engine driven. Those engines can be
operated on oil or digester gas or natural gas. The heat is
captured from both the pump and blower engines and is used to
heat the digesters, which produce the digester gas, and all the
buildings on the site, including all hot water. There are no
separate heating furnaces in the plant. All of the rest of the
210
-------
equipment is operated with electricity. The total energy
requirements amounted to 3921 kw-hr per million gallons of
sewage treated. This is an increase of about 1000 kw-hr per
million gallons from when the plant was only an activated sludge
type plant.
A breakdown of the energy uses is as follows:
Energy for Operation of
Advanced Waste Treatment Plant
Roanoke, Virginia
Description % of Total
Electricity 12%
Natural Gas 22%
Digester Gas 22%
Oil 44%
The total cost for energy amounted to $524,880, or about 5$zf
per 1000 gallons of sewage treated. This did not include the
cost of digester gas. If an equivalent amount of natural gas
had been purchased in place of the digester gas, the cost of
energy would have increased by approximately $90,000.
COSTS
OPERATION AND MAINTENANCE
During 1979 the cost of Operating and Maintaining the Roanoke
Plant amounted to $2,088,848. Based upon the actual cost for
chemicals, energy, labor and materials, the following unit costs
were calculated using the total volume of sewage treated.
211
-------
1979 Operating & Maintenance Expenses
Chemicals 3^/1000 gallons
Total Energy 5^/1000 gallons
Labor and Materials ,11/zJ/lOOO gallons
Total 19jzi/1000 gallons Sewage Treated
CAPITAL COSTS
The total plant now in operation was constructed over about a
25-year period, starting in 1950. The city's investment
amounts to approximately $30 million. Using the Engineering
News-Record Construction Cost Index, which is a commonly
accepted index of rising construction cost, the actual
construction costs were trended to an ENR Index of 3200. On
that basis the value of the plant is about $60 million, or
approximately $1.7 million per MGD of capacity. Assuming the
plant would be reconstructed new in 1980, the fixed charges
necessary to amortize the interest and debt would be about 47£
per 1000 gallons of sewage treated. Therefore the total unit
cost to operate, maintain, and amortize the debt of the plant
new would be as follows:
Fixed charges $0.47 per 1000 gallons
O&M 0.19 per 1000 gallons
Total $0.66 per 1000 gallons
SUMMARY AND CONCLUSIONS
The plant consistently produces an effluent low in phosphorus
and Total Kjeldahl Nitrogen. The effluent is clear and looks
better than the river water into which it flows. Assuming a
212
-------
typical home owner discharges about 7500 gallons per month to
the sanitary sewers, the proportionate share of Operation and
Maintenance costs amounted to $17.10 during 1979.
213
-------
FULL-SCALE EXPERIENCE
WITH TWO-STAGE
NITRIFICATION AND PHOSPHORUS REMOVAL
By: W. Peterson
Operations Specialist
Metcalf & Eddy, Inc.
Boston, Mass.
INTRODUCTION
In the last 10 years Metcalf & Eddy has designed a number of
Advanced Wastewater Treatment (AWT) plants based upon two-
stage nitrification, several of which employed phosphorus removal
as well. This paper describes the operation of two of these
plants: the first, in Marlborough, Massachusetts, operating since
1974, and the second, in Princeton, New Jersey, operating since
1978. Both plants have been well operated, with excellent
laboratories and staff. Operating data can be viewed with a
relatively high degree of confidence. Operating experience at
these and other plants has, for the most part, confirmed early
pilot work, provided valuable feedback for current designs, and
has produced process control techniques for fine-tuning the
operation of such plants.
THE MARLBOROUGH EASTERLY AWT PLANT
This plant was designed for an average flow of 5.5 mgd and treats
domestic and commercial wastewaters, septic wastes and landfill
leachate. The plant was built at a cost of $6.2 million and went
on line in 1974. The plant design data and flow schematic are
presented in Table 1 and Figure 1.
The Marlborough Plant provides comminution, grit removal, primary
treatment, and two-stage nitrification with phosphorus removal by
214
-------
e
•4->
C
OS
-U EH
Q 0
io, Ib/lb .27 Overflow rate, avg. GPDSF 274
ors 8 peak GPDSF 820
40. Return Sludge cap.
•P 4-1
id id jr,
HMO
0 id
to < 0
to
J 0 -
Son
-x id 0
Cu I*H 3
C 3 ft
tntO 0
•H in
in • 14
SO 0
•z w
o
o
o
o
rH
^
id
•o
\_
in
a
rH
*
10
en
c •
•H ^l
•o >
id f£
o
rH I/
Q
• O
tn m
0
a
% influent 200
• 2
90 No. Chlorine Contact Tanks 2
w .
£ +J
S1"
4J .
10
• 0
0 -
to H
0
• 4->
O 0
£§
•rl
• Q
0
z
O O O
o o in
O O ro
O CN CT*
CN rH rH
pX • rX
id Cn id
CD > 0
ft id ft
to
CO
EH
12
g. , GPDSF 434 Final Treatment Cascade
id
J 0
id
0 3
O
'rH
ft 0
0 £
D o
o o
0 0
CN *»
rH CN
* -^
D^ (TJ
> 0
10 ft
».
ft
i-H
id
4->
O
EH
No. vacuum filters 2
:. system) 3 conditioning: Lime
0
— '
n
|
ft
|
<
0
2
CN in
CN
to
^1
0
4J
3 D
c o
•rl S
g .
O IB
U 0
Vj_| ,
o a
id
• o
o
194. Ferric Chloride
io, Ib/lb 2.0 Sludge Disposal Landfill
a ra
rH
O *x
rH
10
0 C
en
• -rl
Cu 10
id 0
0 Q
CN
.
in
H
g
Lt~]
CJ
4J
•rl
H
U
•a
0
4J
id
^i
0
<:
.
o
z
2
actor 4
>H
>H
O "
S «
. id
z
. d
z
k£> cn
CO CN
c c
•H 'H
e e
* *
• r^
en id
> 0
id ft
0
E
•rH
4-1
,
4-1
0
O
•a-
r-H
• id
0 0
CN 0
p^
tn
C
id
EH
.
•0
0 •
tO 4J
>1
rl •
(0 10
e 0
•H
VI •
ft id
•H
• o
o
r-
O IN
O 0
in • .
rH in CO
a
rH
ja en
rH QJ
•a
o"u
•rl
-U - tn
IB ft lH
\ 04-1
tntn 4J id
g en ^
> en 0
tO 21 'r!
> z"« o
E B 0 >2
ft H
C C O 3
en Cn en
•H -rl .
0 0 -rl 0
CN
CN rt O
rH rH •<)•
t^ rH
CN
tl
• tn
4-1 0 O PM
MH 4-1 Q, tO
10 U D
* OH pi
id - o
0 3 •
-rH > X
£1 "H 10 10
4J l-l 0
ft 0 ft
0 >
Q O
in
CN
.C
O
id
0
^
tu
1
0
in
h
O
215
-------
o
•H
-U
§
1
rH
0)
w
03
D
o
OJ
M
D
216
-------
chemical addition to the first stage. Nitrified effluent is
chlorinated and reaerated by a series of cascade steps. Sludge
from the secondary and nitrification systems is pumped to the
primary sedimentation tank(s) for cosedimentation with plant in-
fluent. The resulting "primary" sludge is conditioned with lime
and ferric chloride, vacuum filtered and landfilled on a site
adjacent to the plant.
In the following pages overall plant performance will be reviewed;
the performance of each unit treatment operation will be examined;
plant research and optimization studies will be briefly
summarized; and plant 0 & M costs will be presented. Metcalf &
Eddy has benefitted from a considerable amount of "design
feedback" from the Marlborough plant. A complete discussion of
this design feedback would require many more pages than can be
devoted here. However, selected examples of design considerations
have been included.
Annual average plant influent and effluent characteristics for a
four-year period are displayed in Table 2. Fiscal 1976 was the
first complete year of operation and fiscal 1979 was the most
recent year of operation. Plant flows and loadings have been
roughly half of design flows and loadings over the four-year
period. There has been little change, on an annual average basis
in plant influent flow but influent loadings, with the exception
of phosphorus, have steadily grown.
217
-------
TABLE 2
POUR YEAR SUMMARY
MARLBOROUGH EASTERLY AWT PLANT
Annual Averages
1976 1977 1978
Flow, mgd
Influent, Average
BOD,-, mg/1
TSS, mg/1
NH.-N, mg/1
Total P, mg/1
Effluent, Average
BOD5, mg/1
TSS, mg/1
NH -N, mg/1
Total P, mg/1
2.50
123.
181.
15.5
7.7
5.7
11.4
.25
.86
2.75 2.70
145. 152.
225. 317.
16.4 13.8
8.4 6.8
3.5 3.8
9.0 10.6
.74 .22
.61 .82
1979
2.50
159.
306.
18.0
6.8
2.6
7.9
.12
.61
1. By fiscal years, (July-June) plant started up In 1975.
Plant operating data for the most recent fiscal year, 1979, have
been analyzed with the aid of a computer. Complete probability
and calendar plots of concentrations and mass loadings are
included in Appendix A. One of these plots is included here as
Figure 2 displaying variability of ammonia concentrations in raw
influent, primary effluent, secondary effluent and final effluent.
Primary effluent ammonia concentrations generally exceed influent
ammonia concentrations due to the addition of septic wastes and
plant recycle loads between the raw influent and primary effluent
sampling points. The variability of ammonia concentration in
these two streams is seen to be comparable. Secondary effluent
ammonia concentrations are considerably more variable because
218
-------
0
*** PERCENT EQUAL OR LESS ***
Figure 2. NH3-N Influent/Effluent Variability
Marlborough, Mass. - Fiscal Year 1979
219
-------
secondary effluent ammonia concentrations are prone to drop in
warmer weather due to partial nitrification in the first stage.
Final effluent ammonia concentrations are consistently low.
All of the concentrations used to develop Figure 2 and the similar
plots included in Appendix A are based upon daily composite
samples. The ratio of the maximum value for the year to the
annual average is the max day peaking factor. Similarly, the
ratio of the 95th percentile value* to the average is the 95th
percentile peaking factor. Again, using 1979 data, peaking
factors for mass loadings (not concentrations) are displayed in
Table 3. The ratios for median to average are shown as well.
TABLE 3
INFLUENT/EFFLUENT VARIABILITY
MARLBOROUGH EASTERLY, 1979
Mass Loadings
Ratio To Average*
Location
Parameter
Flow
Influent
BODR
TSS5
NH--N
Total P
Primary Effluent
BODc-
TSS5
NH--N
Total P
Secondary Effluent
BOD.
TSS5
NH-.-N
Total P
Final Effluent
BOD,-
TSS5
NH_-N
Total P
Median
.89
.92
.81
.96
.98
.86
.78
.98
.87
.76
.90
.95
.80
.86
.88
.31
.87
90$
1.4
1.4
1.8
1.3
1.2
1.7
2.1
1.3
1.7
1.6
2.0
1.9
1.7
1.6
1.5
3.0
1.4
95$
1.6
1.6
2.4
1.4
1.4
2.1
2.5
1.5
1.9
2.0
2.7
2.0
2.4
1.8
2.0
3.9
1.7
Max
day
3.3
2.5
4.8
2.3
1.8
2.9
3.0
2.2
3.7
4.0
13.2
3.3
4.9
8.9
11.2
16.4
4.2
Average
mass
loading
Ibs/day
3,299
5,082
373
145
3,069
3,941
409
207
299
369
206
25
58
182
3
13
* Peaking factors for 90, 95 percentiles and maximum daily values.
220
-------
Note that in every case median values are less than average
values, (ratio <1.0) but particularly so for effluent ammonia.
When effluent ammonia values are high, they tend to be very high
relative to the average or median values. In 1979 the maximum
effluent ammonia loading was 16 times higher than the average
loading. The average loading represented an ammonia concentration
of only 0.14 mg/1. It is the nature of nitrification systems,
that with a given inventory of nitrifiers and a given set of
environmental conditions, there is a fixed system capacity for
ammonia. Loadings in excess of capacity will bleed through by an
amount equal to the increment in excess. For instance, this
excess might only be 10 percent over capacity, but all 10 percent
passes or "bleeds" through the system and winds up in plant
effluent, where it may be 10 times the normal effluent load.
The plants discharge permit calls for a maximum day value which is
no more than twice the maximum monthly value. Ratios of maximum
daily values to average (not maximum) monthly values are well over
2 (Table 3) but the plant is well operated and rarely exceeds
permit requirements, the reason for this is clear after examining
Table 4 which contrasts plant performance with permit require-
ments. Note that annual average values are well below the permit
specified maximum monthly values. Ammonia-nitrogen, a parameter
requiring special attention, for the reasons described earlier, is
the only major parameter to display noticeable deviation from the
standards. The effluent standards for ammonia-nitrogen shown in
•Table 4 were not met during one 9-day period in 1979. During that
month, the maximum week and maximum month values were exceeded
also.
221
-------
TABLE 4
PERFORMANCE RELATIVE TO PERMIT
LEVELS*
MARLBOROUGH EASTERLY, EFFLUENT
Permit
max.
Parameter month
BOD5, mg/1 7.
TSS, mg/1 15.
NH3-N, mg/1 0.
Total P, mg/1 1.
0
0
5
0
1979
annual
average
2
7
0
0
.6
.9
.12
.61
1979
max
month
4
9
0
0
.5
.7
.71
.87
Permit
max.
week
10
22
0
1
.5
.5
.75
.50
1979
max.
week
5
16
2
0
.1
.4
.2
.94
Permit
max.
day
14.0
30.0
1.0
2.0
1979
max.
day
9.1
47.0
3.2
1.0
*Effluent concentrations specified for major parameters. The
plant's NPDES permit includes many other requirements as well.
The plant's major wastewater treatment systems are the primary,
secondary (including phosphorus removal) and nitrification
systems. To illustrate the performance of each of these three
systems, influent, intermediate, and final effluent concentrations
are presented in Table 5. Overall reductions in BOD,- and TSS were
98 and 97 percent in 1979. Overall reductions in ammonia and
phosphorus were 99 and 91 percent, respectively.
TABLE 5
PLANT PROFILE
MARLBOROUGH EASTERLY, 1979
Average Concentrations, mg/1
Parameter
BODC
0
TSS
NH -N
Total P
Raw Primary
influent effluent
159. 147.
238. 191.
17.9 19.5
6.8 9.8
Secondary
effluent
13.8
15.9
9.3
1.2
Final
effluent
2.6
7.9
0.14
0.6l
222
-------
Influent values do not include plant recycle loads (primarily
vacuum filtrate and waste activated sludge nor do they include
septic wastes. As a result, performance of the primary treat-
ment system implied in Table 5 is understated considerably.
Performance of the secondary system has generally been quite high
even when operating at F/M ratios of 1.0 Ib BOD/lb MLSS and above.
This high performance is due largely to the addition of ferrous
sulfate for phosphorus removal. Operators have found it difficult
to operate the secondary system poorly.
The nitrification system has generally achieved a high degree of
nitrification,as expected, but the system has also routinely
achieved substantial reductions in secondary effluent BOD^, TSS
and total phosphorus. At the Marlborough plant there is only one
dosing point for phosphorus removal within the secondary system.
Reductions in phosphorus by the nitrification system parallel
reductions in suspended solids by the nitrification system, and
are due largely to entrapment of fine suspended solids by the
well-flocculated nitrification mixed liquor.
The addition of chemical to the first activated sludge stage tends
to overpower the "biological sensitivity" of the system. One
would expect to observe a wide variation in mixed liquor
characteristics and system performance with Sludge Retention Times
(SRT's) varying from 1 day to 15 days. Long-term trials during
the first few years of operation were undertaken to study plant
performance and quantify costs as a function of first-stage SRT.
SRT variations spanned four operating modes for the secondary
(first stage) system. Modified aeration, conventional activated
223
-------
sludge, high-rate activated sludge and single-stage nitrification
were studied. Wide variations in mixed liquor characteristics
were not observed but there were variations in sludge production
and energy use. Plant final effluent quality was best when the
first stage was operated at SRT's of 1 to 2 days (modified
aeration mode). Sludge production was quite high but aeration
power costs were quite low. On the other hand, overall
performance was worst when operating in a single-stage
nitrification mode (with 10-25 day SRT's). Sludge production was
quite low but power costs were quite high. Examining these
trade-offs plant staff has tended to operate with 7 to 8 day SRT's
in the cooler months and 4 to 6 day SRT's in the warmer months
(based upon inventory under aeration).
During the first four years of operation, the Marlborough plant
used aluminum sulfate (Alum) for phosphorus removal. In the 9-
months after start-up aluminum to phosphorus (Al/P) ratios
averaged 1.9 Ib/lb. Nearly two years were spent attempting to
bring automatic alum control on-line using a feed-forward control
loop which integrated flow and primary effluent phosphorus con-
centration. The weak link was the phosphate analyzer which never
performed consistently in spite of frequent and intense efforts to
make it work. Operated in the manual mode, the system performed
well in the plant's first full year (FY 76) and Al/P ratios
averaged 1.5 Ib/lb in that year. Midway through the next year the
phosphorus analyzer was abandoned as a control element and the
system was modified slightly to run in the flow proportional
(automatic) mode. Al/P ratios dropped still further, averaging
224
-------
1.25 Ib/lb. Alum dosages in the third year may have approached a
practical minimum averaging 1.1 Ib/lb for Al/P.
Having minimized alum dosage plant staff turned their efforts to
finding a cheaper alternative to alum for phosphorus removal. In
late 1978, a three month study examined sodium aluminate and
ferrous sulfate for phosphorus removal. Sodium aluminate is
alkaline and was found to entirely eliminate the need for lime (pH
control) in the nitrification system. Ferrous sulfate was
obtained cheaply as a manufacturing by-product and was
considerably less acidic than alum. During the trial all three
chemicals performed equally well. Dosages for alum and sodium
aluminate were 1.1 Ib/lb (Al/P) and for ferrous sulfate 1.8 Ib/lb
(Fe/P). The monthly chemical cost including lime for
nitfification pH control was $6,000, $4,650, and $3,850 for alum,
sodium aluminate and ferrous sulfate. The plant has been using
ferrous sulfate ever since the trial, duplicating and confirming
trial dosage and cost figures during the 1979 year.
During the first year of operation additional testing was con-
ducted at the Marlborough plant to confirm or modify the
conclusions of earlier pilot work(l) and examine on a full-scale
basis the kinetics of the nitrification process. Nitrification
rate was measured in batch laboratory tests and by continuous flow
modeling (2). Most of these rate measurements were made at mixed
liquor pH levels between 6.5 and 7.5 and with D.O. concentrations
above 2 mg/1. When ammonia concentrations were above 3 mg/1
nitrification rates corresponded to those identified as optimum
225
-------
in earlier pilot work (figure 3). In this pH range, there
appeared to be no visible pH effect. However, there was a marked
correlation between mixed liquor ammonia concentration and
nitrification rate (figure 4). Given the dilution by return
sludge of influent ammonia concentrations, this concentration
effect can be important in designing systems to meet very
stringent permit requirements. One last factor to be mentioned is
the BOD^/NH ratio for nitrification influent which has generally
been between 1 and 2 Ib/lb at the Marlborough plant (1.4 Ib/lb in
1979). Those relatively low ratios bear on interpretation of rate
data from the plant.
Table 6 summarizes nitfification operating parameters during 1979.
Secondary system operating data reflects operation with two of 4
aeration tanks on line (.29 mg each) and one of two secondary
sedimentation tanks on line. The nitrification system consists of
two four-staged reactors. To date, only one reactor has been
needed at the Marlborough plant, with two-stages on line during
the summer and four stages on line in the winter. Each stage has
a volume of 0.14 MG. Switchover refers to the addition or
dropping of a stage. The stress at switchover is the ratio of
system loading to maximum nitrification rate at the time of
switchover. Given a secondary effluent ammonia peaking factor of
2.0 the stress should be no more than 50 percent to allow enough
slack to handle peak loads without bleed through. Greater stress
risks permit violation.
226
-------
LU
U.
CC
H
Z
ro
I
Z
CO
m
Q
cc.
LU
Q-
CO
CO
CO
CO
.35
.30
.25-
.20
.15
.10
.05'
TWO-STAGE SYSTEMS (BOD/TKN <3/1)
SPECIFIC (MAXIMUM)
NITRIFICATION RATES
10 15 20
MIXED LIQUOR TEMPERATURE,
25
30
MARLBOROUGH EASTERLY
Figure 3. Optimum Nitrification Rate Versus Temperature
227
-------
<
Q
>
O)
.E
o>
Ul
cc
2
o
LL
OC
z
.02
.01
.5 1.0 1.5 2.0
NH3 CONCENTRATION, mg/L AS N
ENVIRONMENTAL CONDITIONS: pH: 6.5 TO 8.0
TEMP: 6° C to 9.5° C
2.5
Figure 4. Nitrification Rate Versus Ammonia Concentration
228
-------
TABLE 6
TWO-STAGE NITRIFICATION
OPERATING PARAMETERS
MARLBOROUGH EASTERLY, 1979
Parameter Average
Flow, MGD 2.5
Secondary System
MLSS, mg/1 2216
MLVSS, mg/1 1519
RSSS, mg/1 439^
F/M, Ib BOD,-/lb MLVSS 0.44
Return Sludge Rate, % 50
Mixed liquor settleability
SSV5, ml/1 330
SSV30, ml/1 170
Nitrification System
MLSS, mg/1 3180
MLVSS, mg/1 I960
RSSS, mg/1 4490
BODR/NH_-N, Ib/lb 1.7
NH.-N/MEVSS, Ib/lb 0.03
Stages on line (.142 MG each) 2 to 4
Switchover temperatures
Two - three - Two stages, C deg. 13
Three - Four - Three stages, C deg. 8
Stress at switchover, % 40
Mixed liquor pH 6.9
Influent Alkalinity, mg/1 109
Effluent Alkalinity, mg/1 50
Mixed liquor settleability
SSV5, ml/1 560
SSV30, ml/1 310
Return Rate, % 95
Since switching from alum to ferrous sulfate secondary effluent
alkalinity has increased from 50 mg/1 to just over 100 mg/1. It
is quite difficult to accurately assess lime usage for pH control
at the Marlborough plant because a by-product lime is used which
is highly variable and generally poor in quality. It is estimated
that in 1979 lime dosage averaged 60 to 70 mg/1 as Ca(OH)2.
Effluent alkalinity values of 20 to 30 mg/1 have been observed
with effluent pH values of 6.5 to 6.8, and there appeared to
229
-------
be no adverse impact upon plant performance. Since start-up, lime
has normally been fed at the influent end of the nitrification
system. Although the plant was provided with automatic pH
controls for each nitrification stage these have not been needed.
The pH drop from the first to last stage is generally less than
0.2.
The Marlborough plant has operated well at nitrification MLVSS
concentrations as high as 3200 mg/1 and final sedimentation tank
solids loadings as high as 35 Ib/sf-day. Performance deteriorates
above these values. Mixed liquor tends to settle rapidly leaving
"straggler floe," with Sludge Volume Index's typically 50 to 100
ml/g. A key objective in nitrification fine -tuning has been to
balance mixed liquor microorganism diversity to produce optimum
settling characteristics, but not at the expense of nitrification
rates. To do this, operators occasionally waste sludge out of the
nitrification system (SRT's of 75 to 150 days) and occasionally
pump small amounts of sludge back into the nitrification system
from the first activated sludge stage. Typically, for every 100
Ibs. of nitrification sludge wasted out, some 20 to 50 Ibs of
secondary sludge has been pumped back in.
Table 7 is a nitrogen profile for the Marlborough plant based upon
a sampling and analysis series extending over several months.
Note the substantial reductions in total nitrogen across the
secondary and nitrification systems. Biological assimilation may
account for much of the secondary system's nitrogen reduction.
However, low sludge wasting rates for the nitrification system
230
-------
TABLE 7
NITROGEN PROFILE
MARLBOROUGH EASTERLY*
Nitrogen Forms
Organic
Ammonia
Nitrite
Nitrate
Total -
* 1975
- N
- N
- N
- N
N
Study.
Raw
Influent
8.7
14.8
.03
.09
23.6
Primary Secondary
Effluent Effluent
8.2
17.3
.09
.11
25.7
3.6
15.8
.04
.29
19.7
Final
Effluent
.56
.18
12.3
.00
13.0
rule out assimilation as a major factor for total nitrogen
reduction by the nitrification system. There may be a great deal
of "incidental" denitrification by facultative microorganisms
present in the nitrification system in spite of aerobic
conditions.
Plant staffing is presented in Figure 5 and plant operation and
maintenance costs are summarized in Table 8. The plant is manned
around the clock, seven days per week. Plant staffing and
budgeting were lean in 1979. The plant staff was shy one man, and
maintenance costs have not been fully accounted for in the 1979
budget summary (Table 8). In previous years maintenance costs
have run six percent and miscellaneous costs two percent of total
O&M costs. The 1979 O&M cost summary does not include
amortization of capital but does include the cost of sludge
treatment.
231
-------
^
bx
LU •
X">
U CO
.(-
UPERINTENDEN
RADE7
GO O
CC
1-
LU
tr.
o
LU ~
(/J —
2
O
2
OQ O
^ LU
-J 1-
b
LU
^< §
Q S m
buT CC _J
5 ^ -^
lol P2
FOF
iTENANCE
_ —
I <
u S
CC
O
t-
UJ CC
— LU
5fe
°t* ^i
JIPMENT
•RATOR"
O Q- S
LU O —
«f 5 cc x 55
§ w D LU O
i iii n
Z CM M S *
r-
CO
1- LL
Z U.
< <
Q 1-
"~"^^^^~ 1- Z »'
I LU _l
oh_ <
i1
OJ
-P
W
(C
w
3
O
(0
S
x:
u
c
o
•H
-P
03
N
•H
C
fO
01
^
o
d)
Vi
3
01
•H
232
-------
TABLE 8
OPERATING AND MAINTENANCE COSTS
MARLBOROUGH ESTERLY, 1979
Labor
Chemicals
Energy
Misc.
Total
$157,300
104,300
99,200
7,600
368,400
43#
28%
21%
2%
100$
Unit Cost 37.5*/l,000 gal treated
In 1979 sludge production averaged 4,400 Ibs/mg of wastewater
flow. Some 48 percent of the sludge can be attributed to primary
treatment, 50 percent to secondary (biological and chemical)
treatment and 2 percent to nitrification. Average lime and ferric
chloride sludge conditioning dosages were 68. and 5.0 percent of
dry sludge solids, respectively. The lime purchased by the plant
is a manufacturing by-product with an activity level (measured by
titration with HC1) which is generally 40 to 50 percent of the
advertised level. This is the major reason for high lime dosages.
Attempts to condition with polymer instead of lime and ferric
chloride have not been successful to date, due primarily to sludge
cake release problems. Vacuum filter cake averaged 23.6 percent
total solids in 1979 while vacuum filter yield averaged 4.0
Ibs/sf-hr on a dry sludge feed solids basis.
THE STONY BROOK AWT PLANT
The Stony Brook Regional Sewerage Authority operates their River
Road AWT Plant in Princeton, New Jersey. This plant is designed
for an average flow of 10.0 MGD and treats primarily domestic,
233
-------
commercial and septic wastes. The plant was built at a cost of
$20.6 million and went on line in January of 1978. Selected design
data are presented in Table 9 and a flow schematic in Figure 6.
The Stony Brook plant is a two-stage plant like the Marlborough
plant, and like Marlborough is designed for phosphorus removal as
well. The plant does not provide primary treatment. Plant
influent passes through aerated grit chambers and on to the
modified aeration system for the first stage of biological
treatment together with chemical addition to remove phosphorus.
The nitrification reactors are very similar to Marlborough's with
two reactors, each with four stages, followed by, in this case,
three nitrification sedimentation tanks. Nitrification is
followed by a second chemical addition point for phosphorus
removal and multimedia filtration. Plant effluent is chlorinated,
dechlorinated and reaerated before final discharge to the
sensitive Millstone River. Data is based upon daily composite
samples taken at plant influent and modified aeration,
nitrification, and final effluents. Sludge treatment is by
gravity thickening, vacuum filtration (polymer conditioning) and
incineration.
In the following pages overall plant performance will be reviewed;
the performance of each unit treatment system will be examined;
and plant O&M costs will be presented.
The first full year of operation for the Stony Brook plant is also
its most recent year of operation, 1979. Plant performance rela-
tive to design and permit standards is shown in Table 10. The
234
-------
(T3 -P
H-> c
fC OS
Q rH
Oi
c
CT>EH
tn -H 2
in -i
j-> CQ
U
0) >H
H C
(U O
C/3 -P
U)
co
en
^
C
CO
EH
C
o
H
-P
menta
• H
•O
CD
en
rH
cd
C
•H
h
O
Z
0
rH
P
O
S
I/I
rH CD
fU W
crt
esign
Aver;
O
o
rH
rH
.
•P
4H
.C
O
Diameter, ea
IT,
CN
A;
CO
CD
CM
CN
rH
.P
IIH
.C
U
CO
0)
.C
4->
CM
CD
P
r- r- 1
H r-- o inoinr-r-jo^r CN co rH-H CN CN o
in r~ r— i co rH CN co
m CO rH rH
s
CM
o
.c
u
-P rO
C CD
CD
3
rH h >1 rH
En in W 4J \
en c g -H 01
Q -H CM D g
CM fa CO O co
CJ CO dP rH . a K
Q H . O ft £ Oi OJ
-CM- 0) .yZOrH -^01 ")5
IDO>,O cO XS10-^^
01 4J CD - -.n ™ n c \
CO--H rl CMCneOrH EHcD COpQ
Mi!U CD ftCM >EHrH
(DcOco-P -Sg-'Pco en
>(DO,H tn330 " C-^ .c
COCMCO -H..CCMCM-H ™ » OrH C fj
D &M-P-P-H 4J -^{D-HCO CO ft
- mmTJggco 5oiH-ip EH fD
CD CD co (03305 ,9 co COT3
4-) cji-H**OrHrH e-'co O--H Ctn^
CO 13 'O.C.CrHetfrtCCM O Cw OrHCM
« sajou -x.eDQ.ri m -HOW
rH S CO CO 0 >1 >1rH C y « 4J JJ
5 in ICDCD-HrlM<< 'PCD O COrrjtfl
O -H rHCOCO oC'rl^'^-1-1
rH c -p^-STj-ac °-r( J5r^ S^o
«H rl rH4J£,rJCCCJ, HSH UO JfS-P
IH 3 3tJi+JUOO-H fl O OJ CD-H>iCDCDCD X! O ooj
O « -rJgKencnp ^u ^en ^ZeC
Z Z g Z
U3
o o o in o in ^5
CTlrH^ CNir-rO CNrHOCOkOLn CNLDCNVI31/1
--- in r- rH.-IOrH
O CO rH ^ rH ^D in
CN rH H
44
c
EH
^ fc
c c m
•rl .0 P
>-H ^ X. S S ° Q
Wcug5 -Qcut; »rM
CO CDrrj^EH CM^ 01
T3 J3^iX- -eoc co-
\ gcDcog oin i.rix
M CO>CD 5 -HlHCN "+J CDCO
a XCOCM'H. HJJO— .^XH >CD
rH U +J o \< 4J m cB D,
psoiKcox; „, -•
CD -PCD J^ g vicj «Jj5jJ-
WCD 01 -HE 5- wcDco ruiHCD
OitnQjcO tH-rl ^CD-cflccjrD ™cO 4-)
CcociM UEH ^geflJ 2 CD-CO
•rllHCOCD ^SeOSCDK ^j .CH
T3CDt(> T3c (BrHiJXUCD S-U
cO>Q)cO 0)O ;H, OSCuCOS gricOS
oco> -P-H l*1> mo ^CDOO
ij co- "S4J -H ee-ia i". -P rH
Z ric ^rHDiOi3CD tlcDXJIH
am-i CDOJ Oco-H-ritnen i^g-Pri
oiQcnro
mCOEnZ 'P "EHPPSffi JpPO
CD ° O 0
P Z B s
CN
m
rl
ckene
•H
s;
EH
CD
Oi
•a
p
rH
en
o
Z
in
CO
s
^
MH
e
• H
00
fc
4J
rn sludge capaci
3
1
ro (N
in
rl
CD in
4J rl
No. Vacuum Fil
No. Incinerate
o
CO o CN
^i*
CN
cn
CM
g
3
. rH
• rl \
O SB A
•Z \rH
rH
- CO -
g a o
CD -H
cn £ co
tH U rl
en co
I^S
< -P
•H C
>i U Oi
M CO -rl
ca CM en
g cO CD
•rl U P
rl
PH
rH
rH
• H
tl
e
CO
J
Ash Disposal
CN **
en
1H
O
4J
.trification Reac
es per reactor
ScT
JJ
d"
^
CO
CO
'
e volume, MG
CO
•P
cn
i£>
CN
1 vo lume , MG
CO
•P
e
CO
o
o * co o
O "^
in
rH
rH T3
\ OJ
£1 flJ
3 CU
W
^ |
o
-------
3 U «
.5 I . 0 § S g
o
•H
-P
I
0)
X!
O
CO
O
iH
fc,
4J
C
o
u
en
c
o
4J
0)
M
D
cn
•H
236
-------
State of New Jersey substantially relaxed plant discharge criteria
just before plant start-up. Although the plant was designed for
stringent phosphorus removal standards, none were set In the
initial operating permit. Although the plant was designed for
year round nitrification, the State only required seasonal nitrifi-
cation. As a result, the phosphorus removal systems have only
been operated for brief periods of time, and then for the purpose
of equipment and instrumentation checkout. Although the nitrifi-
cation system is operated year-round, winter operations are
relaxed, with only two or three stages on-line to minimize O&M
costs. As a result, occasionally high effluent ammonia
concentrations have been observed in the winter. Summer
concentrations are more consistent (Table 10).
TABLE 10
PERFORMANCE RELATIVE TO
DESIGN/PERMIT LEVELS
STONY BROOK, EFFLUENT
Design
Parameter average
BOD5, mg/1 8.
TSS, mg/1 7.
NH3-N, mg/1 0.5
Total P, mg/1 0.5
1979
annual
average
1.8
1.8
0.92
3.5
1979
max
month
3.0
2.3
4.4
6.2
Permit
max
week
8.0
10.0
-
-
1979 Permit
max max
week day
5.7
3.9
7.9** 2.0*
8.0*** -
1979
max
day
6.7
8.6
26.9**
8.0
*June 1 to Sept. 30
**Max week and max day values during June 1 to Sept. 30 were
.76 mg/1 and 2.4 mg/1, respectively.
***0nly measured once per week.
Plant operating data for 1979 have been analyzed with the aid of a
computer. Complete probability and calendar plots of concentra-
tions and mass loadings are included in Appendix B. One of these
237
-------
plots, displaying effluent ammonia concentrations for 1979, is
included here as Figure 7. Since the plant maintains an
inadequate inventory of nitrifying microorganisms in the cooler
months the phenomenon of ammonia bleed through is graphically
illustrated.
At the Stony Brook plant the potential for ammonia bleed through
is further aggravated by the schedule for sludge treatment
operations. There have been delays in completing sewer connections
within the new Stony Brook Regional Sewerage Authority and as a
result, initial year flows and loads have been well below design
flows and loads (Table 11) and sludge processing has had to be
scheduled on an every other week basis. This concentrates two
weeks of sludge processing recycle loads into two days of sludge
processing operations. Potential for bleed through of recycle
ammonia loads is increased as a result.
TABLE 11
ACTUAL VERSUS DESIGN LOADS
STONY BROOK, INFLUENT
Parameter
BOD5!
TSS,
NH3,
, Ib/day
Ib/day
Ib/day
1979
Average
1,887
1,998
542
Design
Average
20,900
18,100
1,670
1979
% Design
9
11
33
The performance of the plant's major wastewater treatment systems
(modified aeration, nitrification and multimedia filtration
.systems) is illustrated in Table 12, which profiles pollutants
across the plant. Substantial reductions in BOD,- and TSS are
238
-------
PERIOD I/ 1/79 TO 12/31/79
AVERAGE -14.786 RANGE - 3.0000 TO H3.000
*** PERCENT EQUAL OR LESS ***
Figure 7. NH3-N Influent/Effluent Variability
Stony Brook AWT Plant, 1979
239
-------
achieved by each system. Reduction in TKN and NH--N occurs
primarily in the nitrification system. Total Nitrogen and Total
Phosphorus are reduced by one-fourth overall - the latter without
any chemical addition for phosphorus removal.
TABLE 12
PLANT PROFILE
STONY BROOK, 1979
Average Concentration, mg/1
Parameter
BOD
COD5
TSS
NH,-N
TKN
NO^-N
Total N
P°4.
Total P
Raw
influent
51.7
124.6
54.8
14.7
17.5
0.6
18.1
3.5
5.1
Mod. Aeration
effluent
15.1
49.3
11.4
12.4
15.5
1.5
17.0
2.8
3.3
Nitrification
effluent
5.5
6.5
1.3
2.6
11.5
14.1
3.1
3.8
Final
effluent
1.8
22.
1.8
0.92
2.2
11.0
13.2
3.0
3.7
Influent and effluent peaking factors are displayed in Table 13.
Flow variability is limited by the maximum pump station capacity
(8 mgd) of the plant's major contributor, the Town of Princeton.
The variation in ammonia loading is substantially greater than the
variability of other parameters due to the seasonal nitrification
requirement.
240
-------
TABLE 13
INFLUENT/EFFLUENT VARIABILITY
STONY BROOK AWT PLANT, 1979
Mass Loadings
Relative to Average*
Parameter
Flow
Influent
BODR
TSS5
NH.,-N
Total P
Mod. Aer. Effluent
BOD,.
TSSD
NH,-N
Total P
Nitrified Effluent
BODc
TSS3
NH,-N
Total P
Final Effluent
BOD5
TSS
NH3-N
Total P
median
0.94
.84
.80
.97
.93
.88
.85
.95
.99
.88
.85
.21
.96
.86
.82
.14
.94
9 Of,
1.57
1.5
2.0
1.4
1.4
1.8
1.8
1.5
1.4
2.0
2.3
2.6
1.3
1.9
1.8
2.0
1.3
95$
1.68
1.7
2.3
1.6
1.6
2.5
2.6
1.8
1.6
2.6
2.5
5.6
1.4
2.6
2.8
2.9
1.7
max
day
1.77
5.3
3.9
3.7
3.1
6.2
9.7
3.4
2.3
5.2
3.6
14.5
1.9
3.1
4.0
29.
2.1
Average
mass
loading*
Ibs/day
4.69
1887.
1998.
542.
185.
573.
450.
451.
139.
228.
264.
49.
143.
72.
70.
35.
138.
Except Flow, in MGD.
Operating parameters for the modified aeration and nitrification
systems are displayed in Table 14. Mixed liquor suspended solids
are maintained at a very low concentration, averaging 514 mg/1 in
1979, but due to the dilute influent, the first-stage F/MLVSS
ratio was only 0.73 Ib/lb-day. Sludge wasting is on an SRT basis.
Ignoring effluent TSS the net SRT for the first stage averaged 1.1
days. Mixed liquor settles rapidly and leaves a turbid
supernatant. Mixed liquor activity is high with an average
respiration rate of 26. mg/g-hr.
241
-------
TABLE 14
TWO-STAGE NITRIFICATION
OPERATING PARAMETERS
STONY BROOK, 1979
Parameter
Flow, MGD
Average
4.69
Modified Aeration System
MLSS, mg/1
MLVSS, mg/1
RSSS, mg/1
F/M, Ib BODR/lb MLVSS
Return Sludge Rate, %
Mixed Liquor Data
Settleability, SSV5, ml/1
Settleability, SSV30, ml/1
Respiration Rate, mg/g-hr
SRT, Days*
Nitrification System
MLSS, mg/1
MLVSS, mg/1
RSSS, mg/1
BOD /TKN, Ib/lb
TKN/MLVSS, Ib/lb-day
Stages on line (.33 mg each)
Mixed liquor data
Settleability, SSV5, ml/1
Settleability, SSV30, ml/1
Respiration Rate, mg/g-hr
First Stage pH
Last stage pH
First Stage D.O., mg/1
Last stage D.O., mg/1
Return Sludge Rate, %
Sludge Wasting
Out of System, Ib/day
In to system, Ib/day
Effluent TSS, Ib/day
514.
395.
1465.
0.73
41.
47.
42.
26.
1.1
1540.
1039.
2695.
0.97
0.07
2 to 3
159.
90.
17.
7.1
7.1
2.4
4.6
115.
227.
118.
264.
*Based upon MLSS under aeration
The second stage mixed liquor is substantially more concentrated
with MLSS averaging 1540 mg/1. The mixed liquor Settleability is
rapid with a sludge volume index of 58 ml/g and quite active with
a respiration rate of 17 mg/g-hr. The BOD,-/TKN ratio is just
242
-------
under 1. To counteract the very "old", rapidly settling sludge,
which tends to be developed in the nitrification system, plant
staff routinely pump small amounts of first-stage mixed liquor
into the nitrification system, just as the Marlborough plant does,
but on a more formalized, systematic basis. This sludge
"recharge" averaged 50 percent during the nitrification season at
Stony Brook. Ignoring effluent suspended solids, the net
nitrification system SRT averaged 100 days. When effluent solids
are included, the nitrification SRT averaged 31 days.
Table 15 briefly summarizes multimedia filter system operation
data. During 1979 generally 2 or 3 filter cells (out of 6
available) were placed on line. An annoying problem during summer
months has been filter flys (midges) which develop from larvae
attached to the inner walls of buried concrete pipes carrying
effluent between the nitrification and MMF Systems. Prechlorina-
tion has helped but not eliminated this problem. Shock applica-
tions of hypochlorite at the nitrification effluent structure
produced better results, but still did not cure the problem.
TABLE 15
MULTIMEDIA FILTER OPERATING DATA
STONY BROOK, 1979
Parameter Average Value
Influent TSS, mg/1 6.5
Effluent TSS, mg/1 1.8
Percent Removal, % 12.
Hydraulic loading, average, gpm/sf 3.1
max day, gpm/sf 6.2
Solids loading, average, Ibs/day-sf 0.25
max day, Ibs/day-sf 0.70
Backwash frequency, average, #/cell-month 36.5
Backwash duration, average, minutes 9.2
Backwash rate, average, gpm/sf 17.2
Backwash Recycle, % Influent Flow 8.2
243
-------
In 1979 sludge production averaged 620 Ibs/mg of wastewater flow.
Of this, some 7 percent can be attributed to the nitrification
system. All the rest is pumped to the sludge thickeners from the
modified aeration system. As described earlier waste activated
sludge is gravity thickened, vacuum filtered and incinerated.
Selected operating data for sludge treatment systems is included
as Table 16. The plant has had little success in using polymer to
condition sludge for vacuum filtration. The major problem with
polymer sludge conditioning has been in obtaining cake release
from the vacuum filter media. The plant will be converting from
vacuum filters to belt presses in 1980 based upon a relatively
short payback in fuel savings to offset belt press costs.
TABLE 16
SLUDGE TREATMENT SYSTEMS
OPERATING DATA
STONY BROOK, 1979
Parameter Average value
Sludge Thickening
Solids loading, average, Ib/day-sf 1.2
Thickened sludge, TS, % 5.4
Thickener overflow, TSS, mg/1 28
Vacuum Filtration
Filter yield, dry Ib/sf-hr 2.2
Cake, TS, % 21.3
VS, % 53.2
Filtrate, TSS, mg/1 1076
Lime, Dosage, % Sludge solids 67
Incineration
Hearth temperatures, 1 607
2 994
5 1136
7 359
Fuel oil usage, gal/DT 198
Average Operating Time, hrs/month 30
244
-------
Plant staffing is presented in Figure 8. Plant operation and
maintenance costs are summarized in Table 17. The plant is manned
around the clock 7 days per week. Shift schedules have been
staggered to meet sludge processing staffing needs. The plant has
gone through substantial personnel turnover during its first two
years of operation, with an average turnover rate estimated to be
nearly 50 percent per year.
TABLE 17
OPERATIONS AND MAINTENANCE COSTS
STONY BROOK, 1979
Category of expense Cost
Labor* $256,345
Power 171,421
Fuel Oil, Propane 87,719
Water 3,171
Operating Chemicals 44,963
Maintenance Tools and Supplies 20,297
Vehicle purchases and maintenance 2,027
Miscellaneous O&M Expenses 7»019
Percentage
43
29
15
0.5
7.5
3.5
.3
1.2
Total $592,962 100
Unit Treatment Cost 34.6<|;/1,000 gal.
"Including benefits.
SUMMARY AND OVERVIEW
The Marlborough Easterly and Stony Brook AWT Plants represent
state of the art operation of two-stage nitrification plants for
the 1970 decade. The years of operating data and experience which
have been obtained from these two plants stand as ready resources
for designers and operators of similar plants. The two-stage
nitrification process is capable of producing consistently high
degrees of nitrification even in climates where mixed liquor
245
-------
STONY BROOK REGIONAL SEWERAGE AUTHORITY 7
EXECUTIVE DIRECTOR I
EXECUTIVE SECRETARY] i
CLERK /BOOKKEEPER
1
NOTE:
FOR INITIAL YEARS
OF OPERATION
PLANT SUPERINTENDENT
I
CHIEF OF OPERATIONS
OPERATIONS
ASSISTANT
1
CHIEF OF FACILITIES
CHEMIST
LAB TECHNICIAN 2
FOREMEN
OPERATORS
10
MAINTENANCE-
MECHANICAL
ELECTRICAL
UTILITY MAN| l]
Figure 8. Staffing - Stony Brook AWT Plant
246
-------
temperatures fall below 5 deg. C. Extremely low effluent
suspended solids and BOD concentrations can be consistently
attained as well. The process is well suited to phosphorus
removal by chemical addition to the first-stage, and 1.0 molar
dosages can be approached with a l.mg/1 total phosphorus standard.
Annual budgets for these plants, if well operated, can be quite
reasonable.
Staged reactors have been effective as a means of controlling
mixed liquor inventory and minimizing energy costs. Nitrification
pH control requirements can be met with relatively simple control
systems, and close to optimum nitrification rates achieved over a
wide range of mixed liquor pH. Nitrification mixed liquor
settling characteristics tend to be those of an "old" activated
sludge. With attentive process control mixed liquor settleability
can be adjusted to achieve optimum results with respect to
supernatant (effluent) suspended solids and turbidity. If
experience at Stony Brook is any indication, effluent from a two-
stage nitrification system is readily filterable. Future designs
should seek to provide the operator with a good deal of
flexibility, as earlier designs have done; should provide simple,
easily maintainable pH control systems; should allow for the use
of a variety of potential chemicals for phosphorus removal; and
should take into account the observations of nitrification
kinetics which have been made at plants like Marlborough and Stony
Brook. Of course, the best design in the world will fail if
staffing and training are not adequate. The designer is well
advised to guide his client wisely in the areas of staffing and
247
-------
budgeting, and should play an active role in the comprehensive
training of new personnel.
248
-------
APPENDIX A
MARLBOROUGH, MASSACHUSETTS
STATISTICAL PLOTS
Figures
Al Flow Variability
A2 BOD - Influent/Effluent Variability
A3 TSS - Influent/Effluent Variability
A4 NH3-N - Influent/Effluent Variability
A5 Total P - Influent/Effluent Variability
A6 BOD - Effluent Variability, Weekly Averages
A? TSS - Effluent Variability, Weekly Averages
A8 NHg-N - Effluent Variability, Weekly Averages
A9 Total P - Effluent Variability, Weekly Averages
A10 Secondary Effluent - BOD, TSS
All Secondary Effluent - NH^-N, Total P
A12 Final Effluent - BOD, TSS
A13 Final Effluent - NH3~N, Total P
249
-------
9.0
PERIOD 7/ 1/78 TO 6/30/79
AVERAGE -5.6868 RANGE -1.1000 TO 8.9000
8.0
7.0
6.0
-o
I
. 5.0
O
4.0
3.0
a.o
II I I'
9.5 99. E
0.1 0.2
075
95 98
.9
*** PERCENT EQUAL OR LESS ***
Figure Al.
Flo/? Variability, Marlborough, Massachusetts
Fiscal Year 1979
250
-------
500
400
*** PERCENT EQUAL OR LESS ***
Figure A2.
BOD Influent/Effluent Variability
Marlborough, Mass., Fiscal Year 1979
251
-------
1000
900
*** PERCENT EQUAL OR LESS ***
Figure A3.
TSS Influent/Effluent Variability
Marlborough, Mass., Fiscal Year 1979
252
-------
*** PERCENT EQUAL OR LESS ***
Figure A4.
NH3-N Influent/Effluent Variability
Marlborough, Mass., Fiscal Year 1979
253
-------
*** PERCENT EQUAL OR LESS ***
Figure A5. Total P Influent/Effluent Variability
Marlborough, Mass., Fiscal Year 1979
254
-------
PERIOD 7/ 1/78 TO 6/30/79
AVERAGE -2.6531 RANGE - 1.3000 TO 5.1000
MAX WEEK PER MIT
10.5
5.0
4.0
3.0
2.0
1.0
LL
0.1 0.2 DT5 I 2 :
50 70 80 SO 95 98 99 99.5 99.8 99.9
*** PERCENT EQUAL OR LESS ***
Figure A6.
BOD Effluent Variability, Weekly Averages
Marlborough, Mass., Fiscal Year 1979
255
-------
PERIOD II 1/78 TO 6/30/79
AVERAGE -7.9151 RANGE - 2.9333 TO 14.067
1H
MAX WEEK PERMIT
22.5
BO 9) 95 S 1 99 99.3 99.B99.9
11 d 31 4 I DU 6 I /U
*** PERCENT EQUAL OR LESS ***
Figure A7.
TSS Effluent Variability, Weekly Averages
Marlborough, Mass., Fiscal Year 1979
256
-------
1.8
1.6
PERIOD 7/ 1/78 TO 6/30/79
AVERAGE -0.1405 RANGE - 0.0800 TO 1.7100
MAX WEEK PERMIT
0.75
0.8
O.T
*** PERCENT EQUAL OR LESS ***
Figure A8. NH3-N Effluent Variability, Weekly Averages
Marlborough, Mass., Fiscal Year 1979
257
-------
PERIOD 7/ 1/78 TO 6/30/79
AVERAGE '0.6110 RANGE » 0.3500 TO 1.0000
i.OO
0.90
0.80
0.70
<
O
0.60
0.50
0.40
TO 20 3D 40 30 60 70 BO 90 95 SB 99 99.5 997899.9
*** PERCENT EQUAL OR LESS ***
Figure A9.
Total P Effluent Variability, Weekly Averages
Marlborough, Mass., Fiscal Year 1979
258
-------
CD
t^
CO
z
LU
3-
tg
UJ £
> O"
o: O
< m
o
z
3 S
1/DW 'OO8
§
1/3M 'SSI
- = 5
CD
[^
CO
LU 3
> 5
518
a i-
8
CD
r^
en
r-
(0
cu
0
Cfl
•r-l
6-1
W
-P
-U
0)
o
(0
en
to
(0
a
en
a
o
(0
S
tJ>
•r-l
CM
259
-------
* «£
li
8
00
IN
en
U)
c~-
(J)
33
It *
LLI Q-
> ^
1^
o
00
c^
CT)
CTl
cu
03
O
W
•H
Cu
•U
0)
tn
3
x:
u
(0
tn
en
D
O
m
<
CU
3
•H
Cu
1/SW 'd nVJ.00.
260
-------
en
c^
en
1 s
• D
o
00
fe
3 -J
sis
U. 2
< t-
z
00
t>
en
r-
ON
03
(0
o
CO
•H
03
CO
D
x:
o
03
CO
CO
03
o
M
o
OS
s
0)
S-i
D
CJi
-H
Cu
I/OH 'ooa
261
I/ON 'SSI
-------
h
CO
t^
cn
CTi
(U
u
cn
•H
fci
-U
-P
0)
as
cn
cn
03
s
01
3
o
u
o
CO
0)
M
3
cn
•H
'd 1VI01
262
-------
APPENDIX B
STONY BROOK AWT PLANT
STATISTICAL PLOTS
Figures
Bl Flow Variability
B2 BOD - Influent/Effluent Variability
B3 TSS - Influent/Effluent Variability
Bi| NH3-N - Influent/Effluent Variability
B5 BOD - Effluent Variability, Weekly Averages
B6 TSS - Effluent Variability, Weekly Averages
B7 NHo-N - Effluent Variability, Weekly Averages
B8 Modified Aeration Effluent - BOD, TSS
B9 Modified Aeration Effluent - NH3~N
BIO Nitrified Effluent - BOD, TSS
Bll Nitrified Effluent - NH3-N
B12 Final Effluent - BOD, TSS
B13 Final Effluent - NH3-N
263
-------
PERIOD I/ 1/79 TO 12/31/79
AVERAGE -H.6904 RANGE - 1.8000 TO 8.3000
8.0
7.0
6.0
5.0
1.0
3.0
2.0
0.1 0.2075I?
10 SO 3D HO 50 BO To 80 90 95 96 99 99.5 99.399.9
*** PERCENT EQUAL OR LESS ***
Figure Bl.
Flow Variability
Stony Brook AWT Plant, 1979
264
-------
800
180
160
39.9
*** PERCENT EQUAL OR LESS ***
Figure B2.
BOD Influent/Effluent Variability
Stony Brook AWT Plant, 1979
265
-------
200
180
160
140
120
100
to
80
60
20
OT
*** PERCENT EQUAL OR LESS ***
Figure B3.
TSS Influent/Effluent Variability
Stony Brook AWT Plant, 1979
266
-------
PERIOD I/ 1/79 TO 13/31/79
AVERAGE -14.726 RANGE - 3.0000 TO 43.000
*** PERCENT EQUAL OR LESS ***
Figure B4.
NH3-N Influent/Effluent Variability
Stony Brook AWT Plant, 1979
267
-------
6.0
PERIOD I/ 1/79 TO 12/31/79
AVERAGE -1.B190 RANGE - 0.7667 TO 5.6500
5.0
4.0
Q
O
m
3.0
2.0
1.0
0.1 0.2 075 I s
TO STJ3D 40 SO 60 70 80 SO S5 98 99 99.5 99.899.9
*** PERCENT EQUAL OR LESS ***
Figure B5.
BOD Effluent Variability, Weekly Averages
Stony Brook AWT Plant, 1979
268
-------
3.2
3.0
a.s
2.6
2.4
2.2
PERIOD I/ 1/79 TO 12/31/79
AVERAGE -1.7831 RANGE - 0.8200 TO 3.0000
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.I 0.2
2 30 40 50
70 80 90 95 98 99 99.5 99.899.9
*** PERCENT EQUAL OR LESS ***
Figure B6. TSS Effluent Variability, Weekly Averages
Stony Brook AWT Plant, 1979
269
-------
12.0
11.0
10.0
9.0
8.0
- 7.0
6.0
5.0
4.0
3.0
2.0
1.0
PERIOD I/ 1/79 TO 12/31/79
AVERAGE -0.9191 RANGE ' 0.0350 TO 11.770 4 JUN 1980
0.1 0.2
*** PERCENT EQUAL OR LESS ***
Figure B7. NH3-n Effluent Variability, Weekly Averages
Stony Brook AWT Plant, 1979
270
-------
cs
~7
o Q"
£ m
cn "S
£ 0S
^ PS
s s
'aoa
S ?
i/ow 'ssi
t.
ri
CD
t^
"CD
UJ
r>
CD
CTl
r--
EH
S
-------
en
r^
'CD
CTi
4J
C.
"3
iH
Oi
z
h
I
z
O
0
>J
CQ
C
O
-P
U)
t^
'cn
CQ
0)
M
d
CT>
•r-l
Cu
272
-------
\
(J)
O)
S §1
CD -58
|i
i/owaoa
S S S
1/9W'SS1
^^
JUL
UENT
is
4J
c
m
rH
Ol
EH
S
O
O
Sj
CQ
C
O
-P
cn
o
rH
CQ
cr>
•H
273
-------
CD
N
CO
II
en
i^
en
en
r~
CTl
4J
c
O
O
ij
CQ
C
O
4-)
CO
CQ
N-EHN
274
-------
J
"
ss
CD
I/DW 'aoa
s?
I/DW 'ssi
J.
"\
C
03
iH
Oi
O
O
C
O
-l-l
CO
CN
r-l
CQ
01
M
D
cn
•H
275
-------
en
t~-
en
4J
c
i
o
o
^J
CQ
C
O
JJ
s-
Pl
en
t^
CT)
ro
i—i
CQ
OJ
!LJ
3
Oi
•H
276
«US GOVERNMENT PRINTING OFFICE 1980-657-165/0079
------- |