2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
Post on 02-Jun-2018
214 Views
Preview:
Transcript
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
1/12
Water Research 37 (2003) 35903601
Effects of temperature transient conditions on aerobic
biological treatment of wastewater
Fernando Morgan-Sagastume, D. Grant Allen*
Department of Chemical Engineering and Applied Chemistry, Pulp & Paper Centre, University of Toronto, 200 College Street,
Toronto, Ont., Canada M5S 3E5
Received 17 May 2002; accepted 10 April 2003
Abstract
The effects of temperature variations on aerobic biological wastewater treatment were evaluated with respect to
treatment efficiency, solids discharges, sludge physicochemical properties and microbiology. The effects of controlled
temperature shifts (from 35 to 45C; from 45 to 35C) and periodic temperature oscillations (from 31.5C to 40C, 6-
day period, for 30 days) were assessed in 4 parallel, lab-scale sequencing batch reactors (SBRs) that treated pulp and
paper mill effluent.
Overall, the temperature shifts caused higher effluent suspended solids (ESS) levels (25100 mg/L) and a decrease (up
to 20%) in the removal efficiencies of soluble chemical oxygen demand (SCOD). Lower ESS levels were triggered by a
slow (2C/day) versus a fast (10C/12 h) temperature shift from 35 to 45C, but the SCOD removal efficiencies
decreased similarly in both cases (from 6673% and 6572% to 4973% and 5173%). Temperature oscillations caused
an increased deterioration of the sludge settleability [high sludge volume indices (SVI); low zone settling velocities
(ZSV)], high ESS levels and lower SCOD removals.The temperature transients were associated with poor sludge settleability (SVI>100 mL/g MLSS, ZSVo1 cm/min),
more negatively charged sludge (up to 0.3570.03 meq/g MLSS), increased filament abundance (B4 to 4.5, subjective
scale equivalent to very common), and decreased concentrations of protozoa and metazoa (25,00050,000
microorganisms/mL sludge). The controlled, periodic temperature oscillations had a slight impact on SCOD removal
efficiency (5% decrease), and did not seem to select for robust microorganisms that withstood the temperature shift.
Sludge deflocculation and filament proliferation caused by these temperature transients may explain the higher ESS
levels.
r 2003 Elsevier Ltd. All rights reserved.
Keywords: Activated sludge; Temperature transients; Temperature oscillations; SBR; Settleability; Pulp and paper mill effluent
1. Introduction
Transient, non-steady state conditions in biological
wastewater treatment are common, and can be caused
by changes in substrate and nutrient characteristics or
concentration, and by changes in the environmental
conditions to which the biomass is exposed [e.g.,
dissolved oxygen (DO), pH, temperature]. The effects
of substrate concentration transients on internal poly-
mer storage, growth rate, and substrate accumulation
have been widely investigated and are better understood
than other types of transients [17]. Environmental
transients have been associated with system instability
and/or perturbations, but have been less studied until
recently. DO transients, specifically anaerobic condi-
tions, have been related to sludge deflocculation[8], and
toxic transients (e.g., phenol spikes) have induced sludge
ARTICLE IN PRESS
*Corresponding author. Tel.: +1-416-978-8517; fax: +1-
416-971-2106.
E-mail address: allendg@chem-eng.utoronto.ca
(D.G. Allen).
0043-1354/03/$ - see front matter r 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0043-1354(03)00270-7
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
2/12
deflocculation and decreased oxygen uptake rates
(OURs)[9,10].
Temperature transients in biological wastewater
treatment can result from seasonal variations, and from
the operation of batch units and shutdowns/start-ups in
upstream industrial processes. Industrial treatment
systems may be subjected to frequent and drastictemperature transients that affect treatment perfor-
mance. On the contrary, sewage treatment systems
may experience mainly seasonal transients of which
winters may represent the most challenging due to
reduced microbial activity. Pulp and paper mill effluents,
similar to those from several food processing industries,
are characterised by high temperatures (above 2035C)
[11]. A better understanding of mesophilic and thermo-
philic aerobic treatment of pulp and paper mill effluent
has been achieved, especially, in relation to steady-state
operation at different temperatures [1218]. Neverthe-
less, the effects of temperature transitions on sludgemetabolism, microbial community structure, settling
characteristics, and bioflocculation are not well under-
stood.
Temperature shifts have been related to decreased
treatment performance and system instability [e.g.,
lower activity, poor settling, high effluent suspended
solids (ESS)], as in full-scale biological plants treating
pulp and paper mill effluent over 38C during the
summer[19,20]. There are a few reports of the effects of
temperature shifts on aerobic biological wastewater
treatment, most of which come from temperature
adjustments in steady-state studies. The effects have
been dependent on the magnitude of the shift and on the
temperature range studied, and have been linked to
decreased sludge metabolic activity and/or poor sludge
settling [15,21,22]. Observations of deteriorated sludge
settling due to temperature shifts have been anecdotal,
but not from systematic studies, and have been reported
as biomass washout, increments in ESS levels, and
variability in sludge settling parameters [12,15,23].
Effluent turbidity increase has been related to defloccu-
lation and weak flocculation due to a temperature
decrease from 20C to 4C[24].
Biological treatment plants of high-temperature ef-
fluent traditionally operate within the mesophilic tem-perature range of 2535C. In aerobic treatment systems
that operate at the limit of the mesophilic range (35
40C), operating at higher temperatures (e.g., 45C)
during the summer and back down during the fall-winter
may represent a way of cutting down on costs
of cooling equipment and limited cooling through
direct effluent dilution. Treating industrial effluent
at higher temperatures (e.g., 4045C) may be feasible,
as demonstrated for pulp and paper mill effluent by
Tripathi and Allen [16]. However, the transition
from 3035C to 4045C in the summer, and back to
3035
C in the fall-winter, may represent a challenge
due to system instability. Destabilisation due to
transients is becoming of greater concern as treatment
systems are pushed to work at their treatment limits
to meet more stringent regulations and/or increased
loads.
The goal of this study was to investigate the effects of
controlled temperature transients on the performance ofan activated-sludge-type system. The impacts of a 10C
temperature upshift and a 10C temperature downshift
on the sludge metabolic activity, settling and biofloccu-
lation characteristics were assessed at the upper limit of
mesophilic treatment (3045C). In addition, the poten-
tial to enhance the robustness of the sludge to handle
temperature shifts through adaptation to temperature
oscillations was evaluated.
2. Experimental procedures
2.1. Experimental apparatus
Bleached hardwood kraft pulp and paper mill effluent
was used in this study. Approximately 2200 L of mill
effluent were collected from the outlet of the primary
clarifier during a period of 1.5 h, and immediately
refrigerated at 4C. The mill produces approximately
300 t/day of Elemental Chlorine Free (ECF) bleached
hardwood kraft pulp, 120 t/day of recycled bleached
corrugated pulp, and 700 t/day of fine paper. The
treatment plant handles about 128,000 m3/day of waste-
water.
The biomass used as inoculum (approximately 0.35 L/
reactor) was return-activated-sludge mixed liquor ob-
tained from the same mill wastewater treatment plant,
and was refrigerated until inoculation. The sludge was
aerated for 1 day at room temperature before inoculat-
ing and starting up the reactors. This sludge suspension
had a total suspended solids (TSS) concentration of
12,5707230 mg/L and a volatile suspended solids (VSS)
concentration of 98307130 mg/L.
The effluent from the mill was transported to our
research laboratory in a refrigerated truck, and then
frozen at 20C. The wastewater was thawed as
required (about 84 L/week). Nitrogen (N) and phos-phorus (P) were added to the thawed, raw mill effluent
as ammonium chloride (NH4Cl; Mallinckrodt Inc.,
Paris, Kentucky) and di-ammonium hydrogen ortho-
phosphate [(NH4)2HPO4; BDH Inc., Toronto, Ontario],
in a soluble-COD:N:P ratio of 200:5:1. The pH of the
feed (conditioned mill effluent) was decreased to 6 by
adding 20% v/v sulphuric acid (H2SO4, Reagent A.C.S.
Fischer Scientific, Nepean, Ontario), which maintained
the mixed liquor pH between 7 and 8 (pH=7.670.3).
The prepared feed was then stored in 4 separate 9-L
containers (High-density polyethylene carboy, Nalgene,
VWR Scientific, Mississauga, Ontario) at 4
C in a 153-L
ARTICLE IN PRESS
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 35903601 3591
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
3/12
refrigerator (W.C. Wood Co., Ottawa, Ohio), part of the
reactors setup.
An hour and a half before each feeding cycle, 1 L of
feed from each refrigerated container was pumped into a
2-L glass holding tank, where the temperature reached
28C by means of a water bath. The pre-warmed feed at
28C caused the temperature of the sludge in thereactors at 35 or 45C to decrease to a minimum of
33.5C (from 35C) and to a minimum of 40C (from
45C) during feeding, respectively. The initial tempera-
ture of 35 or 45C recovered within less than an hour
since the beginning of feeding. These batch-feeding
temperature transients, similar to the feast-starvation
transients inherent to the batch reactors operation, had
no observable disturbing effect on the treatment
performance of the reactors since relatively constant
operating conditions were achieved over time. The
temperature cooling during feeding, therefore, did not
represent a significant shock to the system.Four parallel sequencing batch reactors (SBRs) were
operated to mimic the processes taking place in an
activated sludge system, and were connected to the feed-
storage refrigerator, preheating tank, and water baths,
as described elsewhere [16,23]. The 4 SBRs were
operated in three 8-h cycles per day. Each 8-h cycle
consisted of a 25-min anoxic filling phase with mixing, a
reaction phase with continuous mixing and aeration
(385min), a 60-min settling phase, and a 10-min
discharge phase. The DO levels were above 23 mg/L
during the reaction phase in the 4 reactors, except for the
initial 20 min, after anoxic filling, when the DO levels
were below 1 mg/L. A sludge retention time (SRT) of
approximately 25 days in the 4 SBRs was maintained by
the amounts of mixed liquor wasted every 2 days (B8%
of mixed liquor), taking into account the wastage due to
suspended solids in the effluent.
2.2. Temperature transients
Two main temperature variations in the 4 SBRs were
conducted on days 117 and 146, as illustrated in Fig. 1.
Before the first shift (Day 117), the 4 reactors were
acclimated at 35C, and the sludge was mixed and
redistributed among the 4 SBRs (Day 111). One SBR(SBR1) was subjected to a fast temperature increase
(10C/12 h) from 35 to 45C. A second SBR (SBR2)
was subjected to a slow, 2C/day temperature increase
during 5 days to achieve a net increase from 35 to 45C.
The temperature in a third SBR (SBR3) was initially
increased from 35 to 40C, and after 3 days at 40C,
decreased to approximately 32.5C to begin periodic
temperature oscillations from 31.5 to 40C with a 6-
day period. The fourth reactor (SBR4) acted as control
at 35C for the temperature shifts, and provided for a
paired experiment showing that the shift effects were not
due to random operating characteristics.
Twenty-nine days after the first temperature shift, the
mixed liquors from SBR1 and SBR2 were mixed and
redistributed between the 2 reactors, and a second
temperature transient was conducted (Day 146). SBR1
remained at 45C, the temperature in SBR2 was
decreased from 45 to 35C, SBR3 was subjected to
an increase from 31.5 to 45C, and the temperature in
SBR4 was increased from 35 to 45C. In this case, the
reactor performance after 4535C decrease (SBR2) was
compared to that under constant temperature at 45C
(SBR1), and the effects of a 31.545C increase after
temperature oscillations (SBR3) were compared to those
of a 3545C increase after constant temperature
operation at 35C (SBR4).
The reactors temperature was monitored with 76 mm
mercury thermometers. Deep-chamber water baths
(3 SBRs: 33 L, 1295 PC, VWR Scientific, Mississauga,
Ontario; 1 SBR: MagniWhirl Constant Temperature
Bath, Blue M Electric Company, Blue Island, Illinois)
connected to the reactors water jackets were used tocontrol the reactors temperatures. The accuracy of the
temperature readings was 71C.
2.3. Microbial activity
Organic carbon biodegradation was monitored by
measuring soluble chemical oxygen demand (SCOD)
from the inlet of the reactors and the treated effluent
discharged and collected at the end of a cycle. The
samples were filtered through 1.5-mm-pore-size glass
microfibre filters (934-AH, Whatman Inc., Clifton, New
Jersey), and stored at 4
C before digestion. COD
ARTICLE IN PRESS
30
35
40
45
50
100 110 120 130 140 150 160 170
Time of operation (days)
T
emperature(C)
SBR1
SBR2
30
35
40
45
50
100 110 120 130 140 150 160 170
Time of operation (days)
Temperature(C)
SBR3
SBR4
(B)
(A)
Fig. 1. Temperature profiles in the 4 SBRs during the 3545C
temperature upshift, the 4535C temperature downshift, and
during temperature oscillations. (A) Profiles of SBRs 1 and 2
and (B) profiles of SBRs 3 and 4.
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 359036013592
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
4/12
measurement in the filtrates was conducted following
Standard Methods[25].
The total COD of the fresh raw pulp mill effluent
when collected was about 800 mg/L. The average
SCOD of the effluent was 578720mg/L before
storage at 20C. A reduction in the SCOD levels
of the original effluent took place after freezing,thawing and refrigerating the effluent. Therefore, the
feed to 4 SBRs had a lower average SCOD of
371753 mg/L (average from the 4 SBRs in Table 1)
during the 165 days of operation than the original
effluent; however, the 4 SBRs were fed with the same
prepared batches of mill effluent.
Specific oxygen uptake rates (OURs) were calculated
from DO measurements (YSI Model 57, YSI 5750 BOD
Bottle Probe, YSI Inc., Yellow Springs Instrument Co.
Inc., Yellow Springs, Ohio) taken within the reactors at
different points during the reaction phase of an
operating cycle.Mixed liquor total suspended solids (MLSS), MLVSS,
and effluent total suspended solids (ESS) were measured
based on Standard Methods[25].
2.4. Floc characterisation
Sludge surface charge was determined by
cationic-anionic titration [26,27]. A 0.002-N hexadi-
methrine-bromide (Polybrene) solution and a 0.001-N
sodium-salt-polyanetholesulphonic-acid solution were
used as the cationic and the anionic standards,
respectively.
2.5. Sludge settling characteristics
Sludge volume indices (SVIs) [25] and zone settling
velocities (ZSVs) [27] were used for assessing
sludge compressibility and settleability, respectively.
Measurements were conducted within the SBRs
during the settling phase of a cycle. No statistically
significant differences between SVIs measured within the
reactors and in a 1000 mL graduated cylinder were
obtained.
2.6. Microbiology
The abundance of filamentous bacteria was recorded
9 times during 165 days of operation based on Jenkins
et al.s [29]subjective scoring system: none (0), few (1),
some (2), common (3), very common (4), abundant (5),
excessive (6). Filament identification was conductedbased on Eikelboom types [28].
Protozoa and metazoa were enumerated in
an undiluted mixed liquor sample using a
corpuscle counting chamber (Improved Neubauer
Levy Chamber, Hausser Scientific, Blue Bell, Philadel-
phia) under phase contrast microscopy at 400
magnification.
2.7. Statistical analyses
Mean values are reported with 71 standarddeviation, except for the SCOD removals before (Days
1116) and after (Days 117146) the shift from 35 to
45C (Day 117; as in Fig. 2) and SCOD from days
147165 (as in Fig. 2) for which 95% confidence limits
are used.
The statistical significance of differences between
means from the same SBR before and after a
temperature shift was assessed by paired Students t
hypothesis tests, and the statistical significance of
differences between means from two different SBRs
was assessed by Students t hypothesis tests for
independent samples and unequal variances, both at
the 95% confidence level. The levels of significance of
the tests (p) are reported for those cases where a
significant difference was found.
The relationships between sludge settling parameters
(SVI and ZSV), filament abundance, temperature and
ESS were evaluated with the Pearsons product moment
linear coefficient at the 99% or 95% confidence
level. Whenever a linear correlation was not
statistically significant at the 95% confidence level, the
relationship was evaluated with the Spearmans rank-
order correlation coefficient at the 99% or 95%
confidence level.
ARTICLE IN PRESS
Table 1
Operating conditions in the 4 SBRs as averages71 standard deviation (from at least 25 observations) calculated from data collected
from day 1 to 165, unless indicated
SBR Actual SRTa (d) Average MLVSSa (mg/L) Influent SCOD (mg/L) pH DOb (mg/L)
1 2678 29007390 374749 7.770.3 2.871.0
2 26710 29007590 368757 7.670.3 3.371.2
3 22711 26007400 372750 7.670.3 4.071.3
4 2577 28007390 372756 7.670.3 4.971.4
aData from stable conditions after initial acclimation and before transient conditions (days 40120).b
Values from DO levels during the react phase from stable and transient conditions (days 40165).
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 35903601 3593
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
5/12
3. Results and discussion
3.1. Sludge acclimation and 3035C shift
To isolate the impacts of the temperature transients
on system performance, the hydraulic retention time
(HRT), the SRT, pH, macronutrients availability, and
aeration rate were maintained approximately constant
among the 4 reactors (Table 1). The HRT was set at 12 h
and the SRT was approximately 25 days, which was
achieved for most of the operating period except when
biomass washout occurred due to filamentous bulking
(SRTB22 d in SBR 3) or due to temperature transients
after day 140. Although lower DO levels were main-
tained in the reactors where the temperature was
increased (SBR1, 2 and 3), the DO concentrations were
on average higher than 2 mg/L in all the SBRs.
The biomass within the 4 SBRs was allowed to
acclimate (B3 SRTs) before shifting the temperature.
Acclimation was considered complete when the MLVSS
concentrations, the SCOD removals, the sludge settling
curves, and the DO uptake profiles within a cycle were
relatively constant among the reactors operating at the
same temperature.SBRs 1 and 4 were started up at 30C and SBRs 2 and
3 at 35C. Steady performance was reached in approxi-
mately 30 days and the temperature in SBRs 1 and 4 was
increased to 35C on day 60. The increase in tempera-
ture from 30 to 35C was conducted in duplicate, in
SBRs 1 and 4 at the same time.
During the 165 days of operation, the reaction-phase
SOUR profiles were consistent among reactors, both
before and after temperature shifts. This indicated that
the microorganisms had similar oxygen require-
ments across the 4 reactors, even under temperature
transients.
The performance of SBRs 1 and 4 was not affected by
the 3035C temperature increase (Day 60); therefore,
this mesophilic temperature shift seems insignificant. In
the 4 reactors, during the first 116 days of operation, the
SCOD removals were not significantly different (Days
1116: SBR1=6572%; SBR2=6673%; SBR3=
6972%; SBR4=6772%), the filament abundance
remained within common to very common (3.3
4), the dominant filaments were Haliscomenobacter-
hydrossis-like type and type 021N, the effluent sus-
pended solids (ESS) concentrations were below 30 mg/L,
and the sludge volume indices (SVIs) and the zone
settling velocities (ZSVs) were similar before and after
the shift (Table 2). In addition, there was no significant
difference in performance among the 4 reactors during
the early acclimation period (Days 360): between SBRs
1 and 4 at 30C, between SBRs 2 and 3 at 35C, and
between the SBRs at 30C and those at 35C.
The performance of the 4 reactors at 35C (Days 60
116) was reproducible since no significant difference in
performance was found among reactors with respect to
all of the parameters measured (SCOD removals, SVI,
ZSV, sludge surface charge, and ESS). The reproduci-
bility in performance obtained by operating the 4 SBRsin parallel at the same temperature of 35C for
approximately 60 days ensured that the response of
the reactors under the temperature transients was not
random; this reproducibility is shown by similar
operating values7standard deviation among reactors
(Table 2).
Variable ESS levels, SVIs and ZSVs were observed in
SBR1 and especially in SBR3 during the first 104 days of
operation, but these were due to filamentous bulking
(Table 2). The incidents of poor sludge compressibility
(SVIs>100 mL/g MLSS) and settleability (ZSVs
o1 cm/min) in SBRs 1 and 3 correlated to filamentous
ARTICLE IN PRESS
Fig. 2. Soluble chemical oxygen demand (SCOD) average removals in the 4 SBRs before and after the 3545C temperature upshifts
(SBRs 1 and 2; SBR4), the 4535C temperature downshift (SBR2), during 31.540C temperature oscillations (SBR3), and the period
from day 147 to 165. The error bars represent 95% confidence levels.
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 359036013594
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
6/12
proliferation (>3.5, common) and high ESS levels
(>30mg/L, up to 900 mg/L in SBR3), but not to
temperature. These incidents of filamentous bulking in
SBR 1 at 30C and SBR 3 at 35C remained without
apparent explanation and point out the potential forfilamentous bacteria to bloom unpredictably.
3.2. Temperature upshift from 35C to 45C
The temperature shift from 35 to 45C was
conducted relatively rapidly in SBR1 (10C/12 h), and
more slowly (2C/day) in SBR2. SBR4 was operated as
a control at 35C. At the same time, SBR3 was operated
under oscillating conditions.
The reactors were further monitored for 29 days, after
which time the sludge in SBRs 1 and 2 was mixed
together in preparation for a temperature downshiftfrom 45C t o 3 5C in SBR2 (Day 146). For this
temperature shift, SBR1 remained as a reference at
45C.
3.2.1. Soluble chemical oxygen demand (SCOD)
removal
Increasing the temperature (Day 117) from 35C to
45C quickly (SBR1) and slowly (SBR2) reduced the
SCOD removals up to 1820% with respect to those of
the control reactor at 35C (SBR4) (Fig. 2). The SCOD
removals in the constant-temperature reactor (SBR4 at
35
C) were the same before and after the temperatureshift: 6772% (Days 1116) and 6373% (Days 117
146). The SCOD removals in the fast-shift reactor
(SBR1: 5174%) and the slow-shift reactor (SBR2:
4974%) were statistically significantly lower than
those from the control SBR4 (pSBR1 1:9 105
;
pSBR2 4:7 105) and oscillating SBR3 (pSBR1
0:002; pSBR2 0:0003), which shows the reproducibility
of the effect of increasing the temperature up to 45C.
No statistically significant difference in SCOD removals
was observed between conducting the 3545C tem-
perature shift quickly and slowly. The SCOD-concen-
tration profiles over a cycle time were similar among
reactors, and showed that the treated effluent SCOD
concentrations from the fast-shift (SBR1) and slow-shift
(SBR2) reactors were approximately 50 mg/L higher
than those from the oscillating (SBR3) and control
(SBR4) reactors (data not shown).Although the temperature in SBR2 was later de-
creased from 45C to 35C (Day 147), no statistically
significant increase in SCOD removal was observed
between operating at 35C (SBR2: 5875%) and 45C
(SBR1: 5872%). The SCOD removal in SBR1 after the
3545C shift (Days 117146: 5174%) increased
significantly (p 0:017) after operating SBR1 for 19
more days at 45C (Days 147165: 5872%), thereby
indicating a gradual acclimation. It is possible that the
biomass in SBRs 1 and 2 would have acclimated further,
with higher removal efficiencies, if the reactors would
have been operated at 45
C for longer than 29 days(Days 117146), as reported by Tripathi and Allen[16].
A possible explanation for the decrease in SCOD
removal is microbial activity reduction due to the
readjustment of microbial enzymatic activity. Microbial
metabolic deterioration and microbial death and lysis
could have also led to reduced SCOD removals.
However, the reduction in SCOD removal efficiencies
due to the temperature upshift was not correlated with
any change in SOUR profiles. Therefore, the release of
soluble products from the sludge flocs due to defloccula-
tion and lysis, as reported by Barker and Stuckey[30], is
also a plausible cause of increased effluent SCOD levels.
3.2.2. Sludge settling characteristics
The temperature shifts, both 3545C upshift and 45
35C downshift, deteriorated the sludge settling char-
acteristics. Before the 3545C upshift (Day 117), the
sludge in the 4 SBRs was settling slowly (ZSVp1.5cm/
min) and had a moderate compressibility (SVI=75
150 mL/g MLSS)(Fig. 3), compared to previous values
of highly settleable (ZSV>2 cm/min) and compressible
(SVI o75 mL/g MLSS) sludge. The 3545C upshift
(Day 117) caused the sludge compressibility to decrease
further (SBRs 1 and 2: SVI=120210 mL/g MLSS),
ARTICLE IN PRESS
Table 2
Operating parameters of the 4 SBRs as averages71 standard deviation before and after the 3035C shift in SBRs 1 and 4 (day 60)
SBR SCOD removal (%) Filament abundance SVI (mL/g MLSS) ZSV (cm/min) ESS (mg/L)
Before After Before After Before After Before After Before After
1 6675 6578 3.5 3.53.7 98742 87727 1.871.9b 3.271.8 1974 19711
2a 66711 6775 33.5 3.53.7 61719 115756 3.571.6 2.371.8 23717 1875
3a,b 6975 6977 2.83.8 3.74 115756 177771 2.371.9 0.770.5 2477 1697303
4 6775 6776 2.83.5 3.5 55711 5979 4.071.1 4.171.1 2073 2176
Before=days 1559; After=days 60104, except for the SCOD removals where Before=days 360 and After=days 60112.aReactors continuously at 35C.bHigher variability and/or different after values reflect filamentous-bulking incidents.
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 35903601 3595
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
7/12
especially in the fast-shift reactor (SBR1), and the sludge
continued to settle poorly (SBRs 1 and 2: ZSV o1cm/
min) during the 20 days after the shift. On the contrary,
the compressibility and settleability in the control SBR4
at 35C improved due to stable operation (SVI
o100 mg/L; ZSV=1.53.1 cm/min).
The poorer sludge compressibility and settleability
after the temperature upshift (in SBRs 1 and 2) were
accompanied by higher ESS levels, due to an increase in
filament abundance (sludge bulking). One cause of high
ESS levels (>100 mg/L), as in the cases of SBRs 1 and 3shown inFig. 5, was poor sludge settling characteristics
since the sludge blanket rose above the outlet port where
the treated effluent was discharged, and biomass was
washed out. Filament abundance may have been
promoted by the temperature shifts. A slight increase
in filament abundance (up to B4) with respect to that in
the control reactor (SBR4 at 35C; filament abundance
B3.53.7) occurred in the fast-shift (SBR1) and slow-
shift (SBR2) reactors after the 3545C temperature
shift. The filament abundance in SBR2 remained similar
to that of SBR1 (at 45C) after decreasing the
temperature from 45
C to 35
C in SBR2 (Day 147).
These results agree with observations of poor sludge
settling at higher temperatures under steady-state
conditions. Settleability reduction (SVI increase) at high
temperatures under steady-state was reported by Car-
penter et al.[19] in continuous stirred-tank reactors that
treated pulp and paper mill effluent and operated with
acclimated activated sludge at 37C, 42C, 47C, and
52C. Krishna and Van Loosdrecht [6] reported a
continuous decrease in sludge settleability (SVI increase)
with increasing temperature (15C, 20C, 25C, 30C,
and 35
C) in aerobic SBRs treating an acetate mediumunder steady state. Some authors report similar sludge
settling characteristics at temperatures between 35C
and 45C [12,16], but at steady state. Tripathi [23]
reported higher variability in SVIs during a transient
between 35C and 45C than at a constant temperature
of 45C.
About 25 days after the 3545C upshift (Days 144
153), the sludge that had been operating at 45C (SBRs
1 and 2) changed significantly: a firmer, more compact,
less negatively charged sludge with common abundance
of filamentous organisms and improved settleability and
compressibility was observed. This could be explained
ARTICLE IN PRESS
0
50
100
150
200
250
300
350
105 110 115 120 125 130 135 140 145 150 155 160 165
Time of operation (Days)
SVI(mL/g
MLSS)
SBR1
SBR2
SBR3
SBR4
T=35C(4 SBRs) T=45C(SBR1&2); T=35C(SBR4); T=30-40C(SBR3) T=45C(SBR1,3&4); T=35C(SBR2)
0
1
2
3
4
5
6
7
105 110 115 120 125 130 135 140 145 150 155 160 165
Time of operation (Days)
Zonesettlingvelocity
(cm/min)
SBR1
SBR2
SBR3
SBR4
T=35C(4 SBRs) T=45C(SBR1&2); T=35C(SBR4); T=30-40C(SBR3) T=45C(SBR1,3&4); T=35C(SBR2)
Fig. 3. Sludge volume index (SVI) and zone settling velocity (ZSV) measured in the 4 SBRs before and after the 3545C temperature
upshifts, the 4535C temperature downshift, and during temperature oscillations (31.540C). The SBRs conditions are labelled on
top of each graph.
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 359036013596
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
8/12
by acclimation at 45C in SBRs 1 and 2. The decrease in
temperature in SBR2 from 45C to 35C (Day 147) did
not improve or maintain the good sludge compressibility
and settleability in this reactor; similar behaviour and
values of SVI and ZSV were registered in both SBR1
(constant at 45C) and SBR2 (4535C downshift).
These observations suggest that both temperatureupshifts and downshifts cause poor sludge compressi-
bility and settleability, and that this condition is
irreversible with respect to temperature downshifts
within the time frame studied.
3.2.3. Sludge deflocculation
The temperature increase from 35C to 45C (Day
117; SBRs 1 and 2) caused a net decrease in the sludge
surface charge (Fig. 4). Similar sludge surface charge
values, hence similar sludge physico-chemical character-
istics, among the SBRs were measured before the 35
45
C temperature shift (SBRs 1 and 2). A slowtemperature shift (SBR2) in comparison to a faster
one (SBR1) delayed the sludge becoming more nega-
tively charged, but a similar sludge surface charge was
ultimately reached in both reactors after 8 days of the
shift (SBR1: 0.2470.03meq/g MLSS; SBR2:
0.2470.01 meq/g MLSS; SBR4: 0.1570.01 meq/g
MLSS). This suggests that the factors determining the
sludge charge characteristics are a function of
the temperature rather than of the mode in which the
temperature is attained. The later decrease in tempera-
ture from 45C to 35C in SBR2 (Day 147) did not
change the sludge charge in SBR2, which remained
similar to that of SBR1 at 45C. The sludge charge
among SBR1 (at 45C), SBR2 (at 35C since day 147),
and SBR4 (at 45C since day 147) were similar, ranging
from 0.11 to 0.22 meq/g MLSS (Fig. 4).
Sludge surface charge is believed to influence sludge
floc stability and floc formation due to the interaction of
electrostatic forces at the solidliquid interface of sludge
particles, as indicated by Zita and Hermansson[31]and
Mikkelsen et al. [32].A more negatively charged sludge
(i.e., less hydrophobic) has been correlated[27]with pin-
point flocs and probably deflocculation. The increase in
negative surface charge may have been due to the
presence of more soluble compounds with anionicfunctional groups released by floc fragmentation and/
or sludge lysis, as a consequence of the temperature
shift. Negative sludge surface charge under neutral
conditions has been attributed to the presence of anionic
functional groups (e.g., carboxyl, hydroxyl, phosphate
groups) on the sludge floc surface[33,34]. Deflocculation
may have also increased the sludge surface area per g of
sludge due to the presence of smaller floc fragments with
increased surface, thereby increasing the sludge surface
charge. Lower sludge hydrophobicity (and presumably
more negatively charged sludge) has been associated
with deflocculating sludge under phenol disturbancesand has been partially explained by the effect of cellular
components released by lysis [10]. Lysis may be
explained by cell death, at least in some bacteria and
microfauna, due to irreversible damage as a result of
transient conditions; the cause of deflocculation,
although associated with sludge physico-chemical prop-
erties, is not known. Further research in this area is
being conducted in our laboratory.
Further evidence of deflocculation as a result of the
3545C shift (Day 117) came from the high ESS levels
(SBRs 1 and 2). High ESS levels (25100 mg/L) were due
to pin-point flocs in suspension in the treated effluent,
which may have come from stressed flocs that became
structurally weak and deflocculated. Before the tem-
perature shift, the ESS concentrations amongst the 4
reactors were similar and below 25mg/L (Fig. 5).
However, after one day of the temperature shift (Day
118), the ESS concentration increased above 25 mg/L in
ARTICLE IN PRESS
-0.40
-0.35
-0.30
-0.25
-0.20
-0.15
-0.10
-0.05
0.00105 110 115 120 125 130 135 140 145 150 155 160 165
Time of operation (Days)
Sludgecharge(meq/gM
LSS)
SBR1
SBR2
SBR3
SBR4
T=35C(4 SBRs) T=45C(SBR1&2); T=35C(SBR4); T=30-40C(SBR3) T=45C(SBR1,3&4); T=35C(SBR2)
Fig. 4. Sludge surface charge in the 4 SBRs before and after the 3545C temperature upshifts, the 4535C temperature downshift,
and during temperature oscillations (31.540
C). The error bars represent standard deviations of surface charge measurements.
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 35903601 3597
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
9/12
the fast-shift reactor (SBR1). In this fast-shift reactor
(SBR1), high ESS concentrations of 45 mg/L and up to
187 mg/L were registered within 15 days after the shift
(Fig. 5). In contrast, in the slow-shift reactor (SBR2),
only a slight increase in ESS (up to 32mg/L) was
registered compared with the control at 35C (SBR4:
ESS o 25 mg/L) and to the period before the shift.
These results show that a 3545C temperature upshift
produces higher ESS levels, and suggests that the higher
the magnitude of the temperature upshift, the higher the
ESS levels.
There are few reports on ESS levels under tempera-
ture transients. Lee et al. [12]mention slight settleability
deterioration as a result of a 3545C shift in a lab-scale
aerated lagoon treating bleached kraft mill effluent.
Although contradictory results have been reported on
settling at different temperatures when treating pulp mill
effluent, as pointed out by Barr et al. [15], the results
from the present work are consistent with those from
some researchers[16, 35]in that high ESS levels occur at
temperatures higher than 40C under steady state.
Reproducibility of the effects of the shift from 35C to
45C on high ESS levels are demonstrated by the
increase in ESS concentrations in the 4 SBRs when
subjected to temperature transients within this range(Fig. 5).
After the 4535C decrease in temperature in SBR2
(Day 147), the ESS levels returned to the normal levels
below 25 mg/L, which supports the idea that while
increasing temperature causes higher ESS levels, de-
creasing temperature counteracts this effect.
3.2.4. Sludge microorganisms
A shift in the Eikelboom types of filaments may have
occurred due to the 3545C shift. Before the shift in
SBRs 1 and 2, the prevalent types were Haliscomeno-
bacter-hydrossis -like,Thiothrixspp., and types 0041 and
021N; however, after the shift, the dominant types were
type 021N and Thiothrix spp. This is in agreement with
the prevalence ofThiothrix spp. and Type 021N in the
full-scale pulp and paper mill wastewater treatment
plant during the summer, from where the inoculum and
the effluent were collected. In the control reactor at
35C (SBR4), Haliscomenobacter-hydrossis-like and type
0041 were the dominant types. No further changes in
filament type dominance were recorded. The types of
filaments identified in the reactors have been identified
as some of the dominant types in bulking sludge of
activated sludge plants treating pulp and paper mill
effluent [36].
Protozoa and metazoa concentrations decreased
significantly with the 3545C temperature shift (SBRs
1 and 2). After the 3545C shift, the protozoan/
metazoan concentrations in the fast-shift reactor
(SBR1=46,000726,000 microorganisms/mL sludge)
and slow-shift reactor (SBR2=47,000725,000 micro-
organisms/mL sludge) were significantly lower than
those in the control at 35C (SBR4=130,000759,000
microorganisms/mL sludge). Whereas the low concen-
trations of higher life forms prevailed at 45C (SBRs 1
and 2), the protozoan/metazoan concentrations in-
creased to about 150,000 microorganisms/mL sludge 2days after the temperature was decreased from 45C to
35C (SBR2, day 147). A diverse microfauna was
observed at 35C (stalked ciliates, small free-swimming
ciliates, flagellates, rotifers, rotifer cysts, and nema-
todes), but mostly small free-swimming ciliates/flagel-
lates and inactive/dead rotifers, inactive/dead stalked
ciliates and inactive/dead nematodes were observed after
the shift (SBRs 1 and 2 at 45C). These observations
give additional evidence of floc fragmentation since
thriving of small flagellates and free-swimming ciliates
may be indicative of deflocculation[29]and the presence
of soluble organic matter.
ARTICLE IN PRESS
0
25
50
75
100125
150
175
200
225
105 110 115 120 125 130 135 140 145 150 155 160 165
Time of operation (Days)
ESS(m
g/L)
SBR1SBR2
SBR3
SBR4
T=35C (4 SBRs) T=45C (SBR1&2); T=35C (SBR4); T=30-40C (SBR3) T=45C (SBR1,3&4); T=35C (SBR2)
Fig. 5. Effluent suspended solids (ESS) in the 4 SBRs before and after the 3545C temperature upshifts, the 4535C temperature
downshift, and during temperature oscillations (31.540C). The error bars represent standard deviations of ESS measurements.
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 359036013598
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
10/12
3.3. Temperature oscillations (31.540C)
At the same time of the 3545C temperature shift
(SBRs 1 and 2 on day 117), another SBR (SBR3) was
subjected to a fast temperature increase from 35C to
40C, and to periodic temperature oscillating condi-
tions. The oscillations consisted of variations in tem-perature from 40C to 31.5C with a period of 6 days,
operating consecutively for 3 days at 40C and 3 days at
31.5C (Fig. 1). After 29 days under oscillations (Day
146), a temperature upshift from 31.5C to 45C was
conducted in the oscillating reactor (SBR3) to test the
hypothesis that temperature oscillations select for a
microbial community that handles temperature varia-
tions. Also on day 146, the control reactor at 35
(SBR4) was subjected to a 3545C temperature shift,
after operating for 86 days at a constant temperature of
35C.
Temperature oscillations (Days 117146) had nosignificant effect on the SCOD removal efficiency
(SBR3: SCOD removal=6174%) compared to that of
the control at 35C (SBR4: SCOD removal=6373%).
Nevertheless, a slight reduction of 5% in SCOD
removals was observed during oscillations (Days 117
146) in the SBR3 compared to the previous removals
before oscillations in the same reactor (Days 1116);
there was a statistically significant difference (p 0:002)
between the SCOD removals before (6972%) and after
(6174%) oscillations in the SBR3 (Fig. 2). No
statistically significant differences in SCOD removals
were observed between the oscillating (SBR3) and
control (SBR4) reactors before or after the temperature
upshift to 45C (Day 146; Fig. 2). Nevertheless, the
sludge settling characteristics, the physico-chemical
sludge characteristics, and ESS concentrations changed
under temperature oscillating conditions.
Under temperature oscillations (SBR3: days 117
146), the sludge settling characteristics deteriorated
steadily with time (Fig. 3). In the oscillating reactor
(SBR3), both the sludge compressibility and settleability
were poor (SVI>100 mL/g MLSS; ZSVo1 cm/min)
compared to those of the control at 35C (SBR4:
SVIo75 mL/g MLSS and ZSV=13.1 cm/min). After
the temperature increase to 45
C in SBRs 3 and 4 (Day147), the sludge settling characteristics in the previously
oscillating reactor (SBR3) further deteriorated, and the
SVIs reached values as high as 345 mL/g MLSS
after 6 days of the shift. In the control (SBR4), the
SVI also increased as a result of the temperature upshift
to 45C and reached similar values as in SBR3 after 13
days of the upshift. This agrees with the previous
observations from the 3545C shift (SBRs 1 and 2: day
117), and supports the reproducibility of the experi-
ments and the conclusions on the negative impacts of
temperature upshifts on sludge compressibility and
settleability.
This poor sludge settling may have also been a
consequence of filament proliferation. In the oscillating
reactor (SBR3), a change in filament abundance was
scored from B3.7 up to B4 (very common) after
shifting to oscillating conditions, and from B3.8
up to B4.5 after the temperature upshift to 45C (Day
147). Similarly, in the control (SBR4) an increase infilament abundance from B3.53.7 to B3.84.1 was
scored after increasing the temperature in this reactor
to 45C.
The sludge surface became more negatively charged
under temperature oscillations, decreasing steadily (Fig.
4) until it reached a value of0.3570.03 meq/g MLSS,
the most negative charge measured in the 4 SBRs. This
was significantly lower (po0:05) than the sludge charge
of the control (SBR4: 0.1570.01 meq/g MLSS). The
sludge surface in the oscillating reactor (SBR3) was
more negatively charged than in the control (SBR4)
before and after the temperature increase to 45
C inboth reactors (Day 147). The sludge surface charge in
SBR4 did not change significantly after the 3545C
temperature upshift.
The ESS levels during temperature oscillations (SBR3:
1632 mg/L) were higher than those of the control at
35C (SBR4: >15 mg/L), and they increased steadily
after 25 days under oscillating conditions (up to 49 mg/
L) and also due to the subsequent upshift to 45C (Day
47). Although the temperature increase to 45C in SBR4
was followed by an increase in the ESS concentrations
(above 15mg/L), this was not as drastic as in the
previous temperature shifts (SBRs 1 and 2); this suggests
that the long run of 86 days under a constant
temperature of 35C (SBR4) improved the stability
during the transition.
Similar to the 3545C shift (SBRs 1 and 2), the
temperature increment to 45C in the oscillating (SBR3)
and control (SBR4) reactors (Day 147) caused a
decrease in protozoan/metazoan concentrations (from
approximately 150,000 to 25,00050,000 microorgan-
isms/mL sludge). Conversely, the temperature oscilla-
tions did not cause any change in the protozoan/
metazoan diversity and concentration (B150,000 micro-
organisms/mL sludge) with respect to those of the
control at 35
C (SBR4). Overall, the changes inprotozoan/metazoan diversity and concentration due
to the temperature shifts suggest that from 35C to 41C
a similar, diverse, and active protozoan-metazoan
community exists, and that some microorganisms
survive up to 4145C.
In conclusion, these types of periodic temperature
oscillations (31.540C, 6-day period, for 30 days) did
not select for a microbial community that handled
temperature variations (up to 45C) more robustly. On
the contrary, operating the reactor at a constant
temperature for a long period seemed to have helped
buffer the effect of the 45
C temperature upshift.
ARTICLE IN PRESS
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 35903601 3599
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
11/12
3.4. Overall correlation of sludge settling characteristics
and temperature
Statistically significant correlations (Table 3) amongsludge filament abundance, settling characteristics, and
temperature suggest that filament proliferation was
promoted by the temperature upshifts studied here
(filament abundance vs. temperature), and that sludge
settling deteriorated due to filament proliferation (SVI
and ZSV versus temperature and filament abundance).
In addition, an overall correlation of ESS versus
temperature confirmed that the temperature upshifts
within the temperature range examined here lead to
higher ESS concentrations.
4. Conclusions
The temperature upshifts (from 35C to 45C) had 2
major effects: a reduction (up to 20%) in SCOD removal
efficiency and an increase in effluent suspended solids
(ESS) levels.
Temperature upshifts (from 35C to 45C) and
periodic temperature oscillations (from 31.5C to
40C, 6-day period, for 30 days) deteriorated the sludge
settling characteristics [poorer sludge compressibility
(high SVIs) and settleability (low ZSVs)] by promoting
filament proliferation.
Poor sludge compressibility and settleability, and highESS levels due to the 3545C shift were attenuated by
a gradual temperature increase (2C/day), compared to
a faster temperature increase (10C/12h) in these
experiments. The SCOD removals, however, decreased
in a similar fashion under fast and slow temperature
upshifts.
Periodic temperature oscillations (from 31.5C to
40C, 6-day period, for 30 days) did not select for a
microbial community that handled temperature varia-
tions more robustly from the ESS and sludge settling
perspective. These periodic oscillations slightly de-
creased the SCOD removal efficiency in 5%.
Sludge deflocculation and poor sludge settling due to
the temperature shifts were the origin of high ESS levels.
Sludge deflocculation could have decreased the SCOD
removal efficiency by increasing the effluent SCODconcentrations.
Temperature upshifts (from 35C to 45C) and
periodic temperature oscillations (from 31.5C to
40C, 6-day period, for 30 days) caused a more
negatively charged sludge, a shift in filamentous organ-
isms, and a reduction in protozoan/metazoan concen-
trations and diversity.
Acknowledgements
The authors acknowledge the financial support fromthe members of the Consortium Minimizing the Impact
of Pulp and Paper Mill Discharges at the Pulp and
Paper Centre, University of Toronto: Aracruz Celulose,
Carter Holt Harvey Tasman, S.A., Domtar Inc., EKA
Chemicals Inc., Georgia-Pacific Corporation, Irving
Pulp and Paper Ltd., Japan Carlit Co. Ltd., ERCO
worldwide (formerly sterling Pulp Chemicals Ltd.), and
Tembec Inc. In addition, the financial support from the
Government of Ontario/DuPont Canada Graduate
Scholarship in Science and Technology is gratefully
acknowledged, as well as the partial support from the
Natural Sciences and Engineering Research Council
(NSERC) of Canada. The authors thank Amy Lo at
Domtar Inc. for facilitating wastewater samples.
References
[1] Barford JP, Pamment NB, Hall RJ. Lag phases and
transients. In: Bazin MJ, editor. Microbial population
dynamics. CRC Series in Mathematical Models in Micro-
biology. Boca Raton, FL: CRC Press; 1982. p. 5589.
[2] Daigger GT, Grady Jr CPL. The dynamics of microbial
growth on soluble substrate. A unifying theory. Wat Res
1982;16:36582.
ARTICLE IN PRESS
Table 3
Pearsons product moment linear and Spearmans rank-order correlation (Italics) coefficients among sludge settling parameters,
filament abundance, temperature and ESS for an extended period of operation (174 days) of the 4 SBRs within the temperature range
from 35C to 50C
Correlation Overall N SBR 1 SBR 2 SBR 3 SBR 4 n/SBR
SVI vs. filament abundance 0.61a
56 0.47 0.22 0.55b
0.82b
14ZSV vs. filament abundance 0.61a 56 0.38 0.51 0.79b 0.83b 14
Filament abundance vs. temperature 0.46a 56 0.23 0.67b 0.61b 0.56b 14
SVI vs. temperature 0.43a 152 0.58b 0.14 0.46b 0.66b 38
ZSV vs. temperature 0.28a 152 0.31 0.06 0.44b 0.45b 38
ESS vs. temperature 0.59a 280 0.47b 0.61b 0.43b 0.15 70
aStatistically significant at the 99% confidence level.bStatistically significant at the 95% confidence level. N=total pair of observations. n/SBR=pair of observations per reactor.
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 359036013600
8/10/2019 2003_F.Morgan-Sagastume_Effects of temperature transient conditions on aerobic biological treatment of wastewater.pdf
12/12
[3] Majone M, Massanisso P, Ramadori R. Comparison of
carbon storage under aerobic and anoxic conditions. Wat
Sci Technol 1998;38(89):7784.
[4] Majone M, Dircks V, Beun JJ. Aerobic storage under
dynamic conditions in an activated sludge process. The
state of the art. Wat Sci Technol 1999;39(1):6173.
[5] Krishna C, Van Loosdrecht MCM. Effect of temperatureon storage polymers and settleability of activated sludge.
Wat Res 1999;33(10):237482.
[6] Van Loosdrecht MCM, Henze M. Maintenance, endoge-
neous respiration, lysis, decay and predation. Wat Sci
Technol 1999;39(1):10717.
[7] Lishman LA, Legge RL, Farquhar GJ. Temperature
effects on wastewater treatment under aerobic and anoxic
conditions. Wat Res 2000;34(8):226376.
[8] Wil!en BM, Keiding K, Nielsen PH. Anaerobic defloccula-
tion and aerobic reflocculation of activated sludge. Wat
Res 2000;34(16):393342.
[9] Galil NI, Schwartz-Mittelman A, Saroussi-Zohar O.
Biomass deflocculation and process disturbances exerted
by phenol induced transient load conditions. Wat SciTechnol 1998;38(89):10512.
[10] Schwartz-Mittelmann A, Galil NI. Biological mechanisms
involved in bioflocculation disturbances caused by phenol.
Wat Sci Technol 2000;42(12):10510.
[11] Patrick KL, Young J, Ferguson KH, Harrison A. Closing
the loopThe effluent free pulp and paper mill. Pulp Pap
Pap Making Additives 1994;68(3):S17.
[12] Lee EG-H, Mueller JC, Walden CC. Effect of temperature
and sludge loading on BOD5 removal and sludge settle-
ability in activated sludge systems treating bleached kraft
effluent. TAPPI J 1975;58(6):1003.
[13] Graczyk M. Purification of pulp industry effluents- A
modification in activated sludge method in a thermophilic
system. Gaz Woda Tech Sanit 1991;58:1427.
[14] Johnson RM. The performance of a sequencing batch
reactor for the treatment of whitewater at high tempera-
tures. MASc Thesis, University of British Columbia,
Vancouver, 1995.
[15] Barr TA, Taylor JM, Duff SJB. Effect of HRT, SRT and
temperature on the performance of activated sludge
reactors treating bleached kraft mill effluent. Wat Res
1996;30(4):799810.
[16] Tripathi C, Allen DG. Comparison of mesophilic and
thermophilic aerobic biological treatment in sequencing
batch reactors treating bleached kraft pulp mill effluent.
Wat Res 1999;33(3):83646.
[17] LaPara TM, Alleman JE. Thermophilic aerobic biologicalwastewater treatment. Wat Res 1999;33(4):895908.
[18] LaPara TM, Konopka A, Nakatus CH, Alleman JE.
Effects of elevated temperature on bacterial community
structure and function in bioreactors treating a synthetic
wastewater. J Ind Microbiol Biotechnol 2000;24:1405.
[19] Carpenter WL, Vamvakias JG, Gellman I. Temperature
relationships in aerobic treatment and disposal of pulp and
paper wastes. J WPCF 1968;40(5):73341.
[20] Cocci AA, McCarthy PJ. Sequencing batch reactors in the
pulp and paper industry. A bench-marking study. Proceed-
ings of the TAPPI International Environmental Confer-
ence and Exhibit, Vancouver, 1998. p. 12038.
[21] Prensner DS, Muchmore CB, Gilmore RA, Qazi AN.
Wastewater treatment by heated rotating biological discs.
Biotechnol Bioeng 1976;28:161521.
[22] Kriebitzsch K, Adamietz E, Schiegl C, Helmreich B,
Wilderer PA, Wuertz S. Assessment of biological activityduring temperature changes in a bench-scale sequencing
batch reactor fed with synthetic medium containing lignin.
Wat Sci Technol 1998;37(45):2514.
[23] Tripathi C. Thermophilic aerobic biological treatment of
bleached kraft pulp mill effluent and its effects on floc
formation and settleability. Ph.D. Thesis, Graduate
Department of Chemical Engineering and Applied Chem-
istry, University of Toronto, Toronto, 1999.
[24] Wil!en BM, Nielsen JL, Keiding K, Nielsen PH. Influence
of microbial activity on the stability of activated sludge
flocs. Colloids and Surfaces B 2000;18:14556.
[25] APHA. Standard methods for the examination of water
and wastewater. 18th ed. Washington, DC: American
Public Health Association, 1992.[26] Morgan JW, Forster CF, Evison L. A comparative study
of the nature of biopolymers extracted from anaerobic and
activated sludge. Wat Res 1990;24(6):74350.
[27] Liao BQ, Allen DG, Droppo IG, Leppard GG, Liss SN.
Surface properties of sludge and their role in biofloccula-
tion and settleability. Wat Res 2001;35(2):33950.
[28] Ekama GA, Barnard JL, Gunther FW, Krebs P, McCor-
quodale JA, Parker DS, Wahlberg EJ, Secondary settling
tanks: theory, modelling and operation. Scientific and
Technical Report Series, Report No 6., IWA Publishing
London, 1997.
[29] Jenkins D, Richard MG, Daigger GT. Manual on the
causes and control of activated sludge bulking and
foaming. Boca Raton: Lewis Publishers, 1993. p. 193.
[30] Barker DJ, Stuckey DC. A review of soluble microbial
products (SMP) in wastewater treatment systems. Wat Res
1999;33(14):306382.
[31] Zita A, Hermansson M. Effects of ionic strength on
bacterial adhesion and stability of flocs in a wastewater
activated sludge system. Appl Environ Microbiol
1994;60(9):30418.
[32] Mikkelsen LH, Gotfredsen AK, Agerbk ML, Nielsen
PH, Keiding K. Effect of colloidal stability on clarification
and dewatering of activated sludge. Wat Sci Technol
1996;34(34):44957.
[33] Forster CF. The surface of activated sludge particles in
relation to their compression characteristics. Wat Res1968;2:76776.
[34] Forster CF. Activated sludge surfaces in relation to the
sludge volume index. Wat Res 1971;5:86170.
[35] Flippin TH, Eckenfelder WW. Effects of elevated tem-
perature on the activated sludge process. Proceedings of
the TAPPI International Environmental Conference and
Exhibit, Portland, Oregon, 1994. p. 94749.
[36] Thompson G, Swain J, Kay M, Forster CF. The treatment
of pulp and paper mill effluent: a review. Biores Technol
2001;77:27586.
ARTICLE IN PRESS
F. Morgan-Sagastume, D.G. Allen / Water Research 37 (2003) 35903601 3601
top related