v°/EPA
United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-82-081 August 1982
Project Summary
Effect of Aeration Basin
Configuration on Bulking at
Low Organic Loading
Sang-Eun Lee, Ben L. Koopman, and David Jenkins
Five series of experiments were
carried out in laboratory-scale activated
sludge systems (10.5 to 80 L) to
investigate the effect of aeration basin
configuration on possible causes and
cures of bulking at low food/micro-
organism ratios (F/M). Continuous-
flow, laboratory-scale activated sludge
units were operated on domestic
sewage from the City of Richmond,
California at low F/M. In continuous-
flow, stirred-tank reactors (CSTR) (at
the F/M range of 0.05 to 0.25g COD
removed/g of total mixed liquor
volatile suspended solids (TMLVSS)
per day), bulking did not occur with a
weak sewage feed (BOD5 = 139 mg/L)
and a total mixed liquor suspended
solids (TMLSS) of 1.5 g/L. Supple-
mentation of sewage by blending with
raw sludge produced a stronger
sewage (BOD5 = 315 mg/L), which
caused TMLSS to increase to 3.5 g/L.
Bulking occurred in CSTR units and in
2-, 4-, 8-, and 16-compartment units.
An aeration basin with an initial
compartment of 1/32 of aeration
basin volume prevented but did not
cure bulking. An aeration basin with
the initial compartments 1 /74 of the
total aeration basin volume prevented
and cured bulking. Anoxia in an
aeration basin with two initial com-
partments each 1 764 of total aeration
basin volume did not help to cure
bulking. The Sludge Volume Index
(SVI) of sludges at low F/M (0.3 to
0.35 g COD removed/g TMLVSS
per day) and high TMLSS (3.5 g/L) is
related to conditions in the initial
compartment rather than to those in
the remainder of the aeration basin.
Initial compartment COD, F/M, and
size are important; floe loading,
dispersion number, and total number
of aeration compartments are not.
This Project Summary was devel-
oped by EPA's Municipal Environ-
mental Research Laboratory. Cincin-
nati, Ohio, to announce key findings of
the research project which is fully
documented in a separate report of the
same title (see Project Report ordering.
information at back).
Introduction
Filamentous bulking in activated
sludge occurs under certain conditions
characterized by the presence of exces-
sive length — greater than 10 km/g
suspended solids (SS) — of various
types of filamentous organisms extend-
ing from the activated sludge floe Bulk-
ing has many causes, and they can often
be determined by the types of fila-
mentous organisms present For ex-
ample, Sphae-rotil-us natans and types
1701, 021N, and 1863 are character-
istic of plants with aeration basin dis-
solved oxygen (DO) too low for the
applied organic load; Microthr/x parvi-
cella and types 0041, 0092, 0581, and
1851 are associated with plants having
a low food/microorganism ratio or low
F/M. The remedy for low-DO sludge
bulking is to increase aeration basin
oxygenation capacity or lower the F/M.
Direct remedy of low F/M bulking
-------
usually is not possible because low F/M
may be required for nitrification, for de-
signer preference, or for lower than
anticipated wastewater flows or waste-
waters with significantly lower BODs
concentrations than expected.
Previous work suggests that the
growth of filamentous organisms at low
F/M is suppressed in systems that
either have a low degree of longitudinal
mixing ("plug flow") or are provided
with an initial compartment for a short
detention time during which the waste-
water feed and return activated sludge
are mixed.
The mechanism of filamentous or-
ganism suppression is not clear Sug-
gestions include lower filamentous
organism growth rate compared with
flock formers at high substrate levels,
inability of filamentous organisms to
store substrate for later use in growth,
and lower substrate utilization rate of
filamentous organisms compared with
flock formers. The term "biosorption"
has recently been used by Eikelboom to
lump all of these factors with physical
incorporation of particulate and soluble
substrates in the floe He proposes that
filamentous organisms growth in acti-
vated sludges with high biosorption
capacity can be suppressed by use of
high initial floc-loadmg values (mg
COD/g MLVSS)
Five series of experiments were
conducted in laboratory-scale activated
sludge systems ranging from 10 5 to 80
L to investigate the effect of aeration
basin configuration on possible causes
and cures of bulking at low F/M values.
Materials and Methods
The experiments were conducted in
laboratory-scale activated sludge sys-
tems with acrylic plastic aeration basins
that ranged m size from 105 L for the
continuous-flow, stirred tank reactor
(CSTR) to 80 L for various compart-
mentalized aeration basin units. Mixing
was by paddles with a mean velocity
gradient of 85 sec ~1. Mixing intensity
was increased only 10% when house air
was used for aeration Dissolved oxygen
(DO) in the CSTR control units and in up
to five of the initial compartments of the
other basin configurations was control-
led by a feedback system consisting of a
galvanic oxygen electrode, DO analyzer,
and a recorder-controller that operated
a solenoid valve on the air or Og supply
line. A control range of approximately -0
and +2 mg/L relative to the desired
minimum DO was achieved.
All wastewater used in the experi-
ments was domestic sewage from the
City of Richmond, California.
Secondary clanfiers were inverted
Erlenmeyer flasks with their bottoms cut
off, and in some cases with an acrylic
plastic cylinder attached. Bronze or
nichrome wires, bent to conform to the
conical portion of the inner clarifier
walls, related at 1 rpm to aid thickened
sludge flow to the recycle lines. Both
influent sewage and return sludge were
dosed into the aeration basin by
peristaltic pumps. The ratio of return
sludge (RAS) flow rate to influent flow
rate was 1 0. Sludge was wasted
directly from the aeration basin. The
systems were operated at ambient
temperature, which varied between 18
and 24°C.
The systems were characterized by (1)
the ratio of the total volume of the
aeration basin to that of the initial
aeration basin compartment (VT/Vi),
and (2) by the dispersion number, which
was determined from Rodamine B slug
addition tracer studies on the aeration
basin without the secondary clarifier or
return sludge stream. The F/M ratio
was calculated as
Total COD Soluble COD
F/M = in (g/day) - out (g/day)
TMLVSS (g)
where COD in = COD of feed
COD out = COD of effluent
TMLVSS = total mixed liquor
volatile suspended
solids, or aeration
basin volatile sus-
pended solids (VSS)
+ clarifier VSS
Results
In the first of the five series of
experiments, CSTR aeration basins
(Figure 1 a) were operated at several
steady states in the F/M range of 0.05
to 0.25 g COD removed/g TMLVSS per
day. Experiments commenced with
nonbulking sludge No bulking occurred
at any of the F/M ratios tested; neither
was a bulking sludge producd when the
settled sewage feed was made stale
(septic) by storage at room temperature
for 2 5 days. (Two other researchers had
suggested that stale sewage could
promote the growth of filamentous
organisms). In these experiments, the
rather weak Richmond settled sewage
(BOD5 = 139 mg/L, COD = 300 mg/L,
TSS = 75 mg/L) was the feedstock.
Because of the weakfeed,TMLSS levels
ranged from 1,0 to 1.7 g/L.
Previous work on bulking at low F/M
by other investigators has been conduc-
ted with stronger influent wastes, and
thus higher MLSS concentrations
occurred at an equivalent F/M. Another
researcher observed that a continuously
fed, one-compartment activated sludge
system produced bulking sludge at a
MLSS of 4 g/L, but not at 1 g/L.
Because of this, the strength of the
Richmond settled sewage feed was
increased by blending it with raw sludge
settled from the same sewage in the
primary clarifier. A comparative analysis
showed that the BOD5 of the supple-
mented feed increased to 315 mg/L and
the COD to 750 mg/L. But the BOD5/
COD ratio and the soluble percentages
of BOD5 and COD remained similar to
the previous feed All subsequent
experiments were conducted with
supplemented sewage
The second series of experiments was
conducted largely at an F/M of 0.15 to
0.2 g COD removed/g TMLVSS per day
and a TMLSS of 3.0 to 3.5 g/L. Two
parallel CSTR aeration basin systems
were operated (Figure 1 a). One received
stale supplemented sewage, and the
other was fed with fresh supplemented
sewage. When the experiment began
with a nonbulking sludge (SVI < 100
ml/g), bulking (SVI> 150 ml/g) occurred
after 35 days in both systems when
TMLSS concentration reached 3.0 g/L.
Thereafter, the two activated sludge
systems were operated on fresh, sup-
plemented sewage feed. System 1 was
designated at this time as the CSTR
control, and system 2 was operated at
steady-state, with the aeration basin
being divided progressively into 2, 4, 8,
and then 16 equal-sized compartments.
Sewage feed and RAS always entered
the first compartment Compartmental-
ization to 1 6 equal-sized compartments
did not improve sludge settleability over
the control CSTR
The third series of experiments
employed a CSTR control (Figure 1a)
and an aeration basin with 16 equal-
sized compartments (Figure 1 e). Both
units were operated at steady-state
F/M values in the ranges of 0.15 to
0.20, 0.20 to 0.25, 0.35 to 0.45, 0.50 to
0.60, 0.60 to 0.80, 0.90 to 1.15, and
1.15 to 1.40g COD removed/g TMLVSS
per day. For the F/M values of 0.15 to
0.20, 0.20 to 0.25, 0.35 to 0.45, and 0.50
to 0.60, bulking activated sludge was
initially used. For the F/M values 0.60
to 0.80, 0 90 to 1 15, and 1.15 to 1.40,
the control and 16-compartment unit
were started with nonbulking sludge
-------
b.
d.
e.
h.
Anoxia
Figure 1. Aeration basin configurations used. Values of V-^/V, are given in
parentheses.
a. CSTR system (1); for CSTR control, 10.5-L aeration basin; for 1-
compartment system, 40-L aeration basin.
b. 40-L aeration basin, 2 compartments (2).
c. 40-L aeration basin, 4 compartments (4).
d. 80-L aeration basin, 8 compartments (8).
e. 56-L aeration basin, 16 equal-sized compartments (16).
f. 58-L aeration basin, 16 compartments {32); first 8 are each equal to
1 /32 oftotal aeration basin volumes, and the next 8 are equal in size.
g. 58-L aeration basin, 16 compartments (74); first 2 are each equal to
1/74, the next 6 are each equal to 1/32 of total aeration basin
volume, and the last 8 are equal in size.
h. 42-L aeration basin, 3 compartments (64); first 2 are each equal to
1 /64 and the following compartment is equal to 62/64 of total
aeration basin volume.
i. 42-L aeration basin, 3 compartment (64); same as h, but the 2 initial
compartments are anoxic.
from a laboratory batch system. Though
sludge in the 16-compartment unit
bulked somewhat later than that in the
CSTR control, bulking was not prevented
at any of these F/M ranges by aeration
basins with 16 equal-sized compart-
ments. The conclusion was that bulking
could not be cured at the lower F/M
values by use of an aeration basin with
16 equal-sized compartments.
Because the 16 compartment aeration
basin did not prevent or cure bulking at
any of the F/M values tested in the third
series of experiments; it was decided to
return to an F/M value of 0.30 g COD
removed/g TMLVSS per day and to
examine the effect of using initial
compartments smaller than 1 /16 of the
total aeration basin volume. The first
eight compartments were reduced in
size to the desired fraction of the total
aeration basin volume; the last eight
compartments were sized so that the
total aeration basin volume was 56 to
58 L. This type of system is referred to as
a selector configuration. A CSTR control
was run in parallel at all times. All units
were operated at an F/M of 0.3 g COD
removed/g TMLVSS per day TMLSS
was kept at 3.0 to 3.5 g/L and the fresh,
supplemented sewage was used as the
feed
A selector configuration of eight
compartments, each 1/32 of the total
aeration basin volume (Figure 1f),
prevented bulking, whereas the CSTR
control bulked. At this point (day 26), the
test unit aeration basin was changed
back to 16 equal-sized compartments
(Figure 1e), and the control CSTR unit
was restarted with nonbulking sludge.
Again the CSTR bulked. The unit with 16
equal-sized compartments also bulked,
but less rapidly than the CSTR control.
The system with 16 equal-sized
aeration basin compartments (Figure
1e) was returned to a selector config-
uration, with the first eight compart-
ments each 1/32 of the total aeration
basin volume (Figure 1f). This selector
configuration, which had previously
prevented the bulking of a nonbulking
sludge, did not cure an already-bulking
sludge The SVI's of the selector and
control units fluctuated widely, but
these were similar to each other.
At this point, a plan was made to
change the selector configuration so
that the first two compartments were
1/64 of the total aeration basin volume,
the following six compartments were
1 /32 of the total aeration basin volume,
and the last eight compartments were
equal size at 1/1 Oof the total volume to
-------
give a total aeration basin volume of 58
L (Figure 1 g). But a measurement error
caused the first two compartments to be
1774 of the total aeration basin volume.
Starting with a bulking sludge, this
selector configuration rapidly reduced
SVI, and the CSTR continued to bulk (at
maximum SVI)
The unit with the selector configura-
tion containing nonbulking sludge was
converted to an aeration basin with 16
equal-sized compartments (Figure 1e)
The control CSTR was then restarted
with nonbulking sludge. Both units
bulked as they did previously at this
configuration. Again, the control CSTR
bulked somewhat more rapidly than did
the unit with the 16 equal-sized
compartments.
Because the 16 compartment unit
with two initial compartments equal to
1774 of the total aeration basin volume
was successful in both preventing and
curing bulking, it was decided to
determine whether the same results
could be obtained when the two initial
selectors were followed by a single
large CSTR rather than 14 more com-
partments. A decision was also made to
determine whether any advantage
existed in creating anoxia (no aeration)
in the first two compartments
To achieve these goals, three reactor
trains were set up int he fifth series of
experiments' A control CSTR (Figure 1 a)
and two three-compartment aeration
basins, each with two initial compart-
ments 1/64 of the total aeration basin
volume and a third that was 62/64 of
the total aeration basin volume (Figures
1h and 1i). In one of the three-
compartment units, the two initial
selector compartments were covered
and stirred. The freeboard volume was
purged with N2 gas to create anoxia.
The unit with two 1 /64-volume aerated
selectors reduced the SVI slowly and
erratically to final values of 173 ml/g(at
2.5 g TSS/L) and 150 ml/g (at 1 5 g
TSS/L). The final SVI value in the CSTR
control and the unit with two 1/64-
volume anoxic selectors were the
maximum values possible (i.e., 400
ml/g at 2.5 g TSS/L, and 667 ml/g at
1.5 g TSS/L). Though the final SVI value
in the aerated selector would still
classify this sludge as bulking, these
values were significantly lower than in
either the control or anoxic selector
systems. Table 1 summarizes average
operating data and initial and final SVI
values for all experiments that used
selector configurations.
Conclusions
1. Bulking at low F/M (0.05 to 0.25 g
COD removed/g TMLVSS per day)
did not occur when CSTR activated
sludge systems were fed fresh or
stale domestic wastewater and the
TMLSS was in the range of 1.0 to 1.7
g/L.
2. Bulking did occur when supple-
mented, settled wastewater was fed
to CSTR activated sludge systems at
an F/M in the range of 0.15 to 0.2 g
COD removed/g TMLVSS per day.
3 The conditions existing in the zone of
initial mixing of activated sludge and
wastewater are important in deter-
mining whether or not filamentous
bulking occurs at low F/M. The
degree of longitudinal mixing (as
measured by the dispersion number)
and the soluble COD gradient
throughout the reactor do not appear
to be important for bulking at low
F/M.
4 The F/M in the initial aeration basin
compartment was found to be con-
sistently related to the final SVI,
whereas the floe loading defined by
Eikelboom in 1981 (Eikelboom, D.H.,
Biosorption and Prevention of Bulking
Sludge by Means of High Floe
Loading. Paper 3, Water Research
Centre Conference, Cambridge,
England, 1981) did not show such a
relationship.
5 An aeration basin configuration
consisting of 16 compartments, the
first eight of which are 1/32 of the
total aeration basin volume, will
prevent low F/M filamentous bulking
from occurring in a nonbulking
sludge, but this configuration will
not cure low F/M filamentous
bulking in an already-bulking sludge.
6. An aeration basin configuration
consisting of 16 compartments
(compartments 1 and 2 equal to
1/74 of the total aeration basin
volume, compartments 3 through 8
equal to 1/32 of the total volume,
and compartments 9 through 16
equal to 1/10 of the total volume)
Table 1. Operating Data Summary for Experiments Employing Selector Configurations
Parameters
VM
Operation period
F/M
Units
day
g COD removed
g TMLVSS, day
Control
CSTR
1.0
216
0.31
1 6-Compartment Aeration Basins
A B C D
32
26
.31
16
33
0.30
32
42
0.30
74
23
0.31
Two Selectors x CSTR
E Aerobic Anaerobic
16
35
0.31
64
56
0.31
64
56
0.31
TMLSS
Average hydraulic
detention time
MCRT
Sewage strength
SVI Z5
Initial
Final
Initial
Compartment
F/M
Soluble COD
9/L
hr
day
mg COD/L
mL/g
mL/g
g COD removed
g MLVSS, day
mg/L
3.0
19
15
720
96
400*
0.32
33
3.2
17
14
620
98
70
9.0
63
3.4
18
15
660
78
373
5.0
43
3.4
19
14
720
373
376
8.0
53
3.3
20
12
745
389
50
22.0
87
3.4
21
13
790
46
387
5.0
48
3.2
21
12
830
400*
173
21.0
75
2.7
22
14
830
400*
392
20.0
183
* Maximum SVI value at 2.5 g SS/L
-------
will both prevent and cure low F/M
bulking. A system consisting of two
initial basins each 1/74 of the total
basin volume followed by a large
CSTR with the remaining 72/74 of
the volume should work as well. The
initial mixing zones for control of low
F/M bulking should apparently be
aerated; anoxic initial compartments
did not show any decrease in bulk-
ing.
7. Critical values for initial compartment
parameters to control low F/M
bulking for this waste are: soluble
COD >80mg/LandF/M in the first
reactor > 20 g COD removed/g
MLVSS per day. Further work is
needed to generalize these values
and to account for the effects of
variables such as waste character-
istics, recycle ratio, and MLSS
concentration.
The full report was submitted in
fulfillment of Grant No. R-806107 by
the University of California under the
sponsorship of the U.S Environmental
Protection Agency.
Sang-Eun Lee, Ben L. Koopman, and David Jenkins are with the Sanitary
Engineering and Environmental Health Research Laboratory, University of
California, Richmond, CA 94804.
Ronald F. Lewis is the EPA Project Officer (see below).
The complete report, entitled "Effect of Aeration Basin Configuration on Bulking
at Low Organic Loading." (Order No. PB 82-234 287; Cost: $7.50, subject to
change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati. OH 45268
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
RETURN POSTAGE GUARANTEED
PS 0000329
U S tNVIR PROTtCriUN A&ENCV
KtblGN 5 LIBRAHlT
230 S DEARBORN STREET
CHICAGO XL 60604
------- |