Comparison of laboratory-scale thermophilic biofilmand activated sludge processes integrated with a mesophilic

activated sludge process

J. Suvilampi *, A. Lehtom€aaki, J. Rintala

Department of Biological and Environmental Science, University of Jyv€aaskyl€aa, P.O. Box 35, FIN-40351 Jyv€aaskyl€aa, Finland

Received 30 October 2002; received in revised form 28 November 2002; accepted 4 December 2002

Abstract

A combined thermophilic–mesophilic wastewater treatment was studied using a laboratory-scale thermophilic activated sludge

process (ASP) followed by mesophilic ASP or a thermophilic suspended carrier biofilm process (SCBP) followed by mesophilic ASP,

both systems treating diluted molasses (dilution factor 1:500 corresponding GF/A-filtered COD (CODfilt) of 1900� 190 mg l�1).

With hydraulic retention times (HRTs) of 12–18 h the thermophilic ASP and thermophilic SCBP removed 60� 13% and 62� 7% of

CODfilt, respectively, with HRT of 8 h the removals were 48� 1% and 69� 4%. The sludge volume index (SVI) was notably lower in

the thermophilic SCBP (measured from suspended sludge) than in the thermophilic ASP. Under the lowest HRT the mesophilic ASP

gave better performance (as SVI, CODfilt, and CODtot removals) after the thermophilic SCBP than after the thermophilic ASP.

Measured sludge yields were low (less than 0.1 kg suspended solids (SS) kg COD�1filt removed) in all processes. Both thermophilic

treatments removed 80–85% of soluble COD (CODsol) whereas suspended COD (CODsusp) and colloidal COD (CODcol) were in-

creased. Both mesophilic post-treatments removed all CODcol and most of the CODsusp from the thermophilic effluents. In con-

clusion, combined thermophilic–mesophilic treatment appeared to be easily operable and produced high effluent quality.

� 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Thermophilic; Mesophilic; Combined treatment; Activated sludge process; Suspended carrier biofilm process

1. Introduction

Industries, such as the pulp and paper industry, have

a growing opportunity to apply thermophilic biological

treatment processes for wastewater and process water

management. This is due to increasing concentrations oforganic and inorganic pollutants along with higher

temperature of wastewaters related with reduced nomi-

nal water consumption. If external treatment units such

as the activated sludge process (ASP) could be operated

under thermophilic conditions, the need and investment

for increasing cooling capacity (and subsequent waste of

energy) of the wastewater before treatment could be

minimized.Thermophilic aerobic wastewater treatment appears

to have the advantage over mesophilic treatment of high

removal rates along with low net sludge yield (reviewed

e.g., by LaPara and Alleman, 1999). In thermophilic

aerobic treatment of industrial wastewaters high and

stable COD removals have been reported in laboratory-

and pilot-scale (e.g., Barr et al., 1996; Malmqvist et al.,

1999; Jahren et al., 2002; Suvilampi and Rintala, 2002;Vogelaar et al., 2002a; Suvilampi et al., submitted for

publication a), but also controversial results have been

reported (e.g., Tardif and Hall, 1996; LaPara et al.,

2001). Comparative studies with aerobic processes under

different temperatures have reported that as temperature

increases, COD removals either increase (Couillard and

Zhu, 1993; Barr et al., 1996; Lim et al., 2001) or decrease

(Tripathi and Allen, 1999; LaPara et al., 2001; Vogelaaret al., 2002a). Increased operating temperature has been

suggested to improve the removal of specific compounds

like, e.g., methanol, di-2-ethylhexyl phthalate, and lipids

(Banat et al., 1999; Becker et al., 1999; B�eerub�ee and Hall,2000). Poorer removals have been explained by, e.g.,

reduced diversity of the thermophilic microbial com-

munity and increased part of recalcitrant COD (e.g.,

*Corresponding author. Present address: Satafood Development

Association, Risto Rytin Katu 70 C, FIN-32700, Finland. Tel.: +358-

400-801416; fax: +358-2-6026335.

E-mail address: juhani.suvilampi@bh.spt.fi (J. Suvilampi).

0960-8524/03/$ - see front matter � 2002 Elsevier Science Ltd. All rights reserved.

doi:10.1016/S0960-8524(03)00006-3

Bioresource Technology 88 (2003) 207–214

Tripathi and Allen, 1999). Our recent study on acti-

vated sludge treatment indicated that mesophilic (35 �C)and thermophilic (55 �C) processes removed the same

amount of 0.45 lm––filtered, soluble COD (CODsol,)

with similar volumetric loading rates (VLRs) and hy-

draulic retention times (HRTs) (Suvilampi and Rintala,

2002). We have also noticed that thermophilic effluents

have had higher turbidity and also higher GF/A-filteredCOD than mesophilic effluents, probably consisting

mainly of free-swimming bacteria (Suvilampi and

Rintala, 2002; Suvilampi et al., submitted for publica-

tion b). Subsequently we suggested mesophilic aerobic

post-treatment to reduce the turbidity in the effluent

from the thermophilic ASP.

Aerobic biofilm processes, especially suspended car-

rier biofilm process (SCBP) with moving carriers inside aclosed reactor tank, are claimed to have the advantage

over conventional aerobic treatment processes based on

suspended sludge. The suggested benefits over ASPs

include, e.g., higher volumetric degradation rate and

loading capacity, lower sludge yield, higher biomass

retention, better sludge settling properties, and better

tolerance of varying process conditions (Lazarova and

Manem, 1994; Ødegaard et al., 1994; Jahren, 1999;Kaindl et al., 1999). Thermophilic SCBPs have been

reported to operate with high COD removal rates, short

HRTs, and low sludge yields, compared to conventional

mesophilic ASPs (Jahren, 1999; Malmqvist et al., 1999;

Suvilampi et al., submitted for publication a). Biofilm

processes have also been applied to increase the treat-

ment capacity or to improve the COD removals of ex-

isting activated sludge processes, either by installing apre-treatment unit of suspended carrier process (Hansen

et al., 1999), or by introducing carrier material into the

aeration tank (Gebara, 1999). While thermophilic pro-

cesses should also have higher loading capacity and

lower sludge yields than mesophilic processes, thermo-

philic SCBP should be highly competitive in, e.g.,

loading and degradation rates in comparison with other

thermophilic aerobic processes, like ASPs, sequencingbatch reactors (SBRs), and membrane bioreactors

(MBRs).

The objective of this study was to evaluate continu-

ously operated laboratory-scale combined aerobic

thermophilic–mesophilic wastewater treatment pro-

cesses using two treatment lines. SCBP and ASP were

compared as thermophilic processes whereas ASP was

used as a final mesophilic process in both treatmentlines.

2. Methods

Experimental setup. Two separate laboratory-scale

wastewater treatment lines (referred to as A- and B-

treatments) consisting of thermophilic (55 �C) ASP

(A1), mesophilic (35 �C) ASP (A2), thermophilic (55 �C)SCBP (B1), and mesophilic (35 �C) ASP (B2), each PVC

reactor with liquid volume of 1.5 l, was placed into

temperature controlled water baths. All reactors were

covered with aluminum foil to prevent evaporation. In

B1 the KMT carriers (Kaldnes Miljøteknologi) were

filled up to 50% of reactor volume. Rena 200 aquarium

aerators provided aeration to exceed dissolved oxygen(DO) 2 mg l�1 in all reactors. In ASPs (A1, A2, and B2)

the sludge recirculation ratio from settling units to ae-

ration units was 3:1 to feed flow, in B1 the sludge was

not recirculated. Sludge age in treatment lines was

maintained within 13� 5 d in A-treatment and 12� 4 d

in B-treatment by decanting a measured volume of

mixed liquor from the aeration units. The schematic

presentation of the experimental setup is shown inFig. 1.

Feed. Molasses from a sugar beet factory (Sucros

Ltd., Finland) was used as a synthetic substrate. The

molasses was diluted with tap water in the ratio 1:500

corresponding to total COD (CODtot) of 2100� 150

mg l�1 and GF/A-filtered COD (CODfilt) of 1900� 190

mg l�1 (measured from samples taken daily from the

feed line before reactors). A few samples were alsoanalysed for different fractioned CODs, which were,

total (CODtot), suspended COD (CODsusp), colloidal

COD (CODcol), and soluble COD (CODsol), with values

of 2000� 150, 400, 100 and 1500� 90 mg l�1, respec-

tively. CODsusp and CODcol are calculated values and

therefore have no standard deviation values. Nutrients,

NH4Cl and K2HPO4, were added into the feed container

to produce CODfilt:N:P; 200:5:1. The feed was prepared2–3 times per week in a 30 l feed container, kept at 4 �Cunder nitrogen atmosphere. The fresh feed pH was ad-

justed to 7.5 in the feed container with 2 M NaOH; pH

was adjusted high because the diluted molasses acidified

rapidly in the feed lines. Feed pH varied between 6 and

7. Feed lines were cleaned daily.

Fig. 1. Experimental setup. (1) feed pump; (2) aeration pump; (3)

sludge return line; (4) aerating stone; (5) settling unit; (6) refrigerator.

208 J. Suvilampi et al. / Bioresource Technology 88 (2003) 207–214

Seed. Mixed liquor from a full-scale mesophilic (30–

35 �C) activated sludge plant treating thermomechanical

pulp and paper mill wastewater (UPM Kymmene Ltd,

Kaipola, Finland) was used as seed sludge. The initial

volume of seed sludge in reactors was 50% of total re-

actor volume.

Analyses. CODtot and CODfilt were analysed accord-

ing to SFS 5504 (closed tube method; Finnish StandardsAssociation, 1988). For selected samples the COD was

fractionated to CODtot, CODsusp, CODcol, and CODsol

by different pore size filters. Schleicher and Schuell

GF50-filter (until day 40 and Whatman GF/A after-

wards) was used to measure CODsusp. A 0.45 lmSchleicher and Schuell membrane filter was used to

measure CODsol, CODcol was calculated as the difference

of CODsusp and CODsol. Suspended solids (SS), volatilesuspended solids (VSS), mixed liquor SS (MLSS), and

MLVSS were measured according to Standard Methods

(APHA, 1998) using Schleicher and Schuell GF50 and

Whatman GF/A filters (pore size 1.6 lm). Total solids(TS-fix) as attached biomass in B1 was measured by

weighing 50 dried (4 h at 105 �C) carriers from the re-

actor and 50 unused carriers, measuring the difference

and calculating the whole amount of reactor TS-fix. DOwas measured with a YSI Jenway 9300 DO meter, and

pH and temperature were measured with a Hanna In-

strument 6028 pH meter. The sludge volume index (SVI)

was measured in a 250 ml graduated glass by settling

diluted samples (dilution ratio 1:1) for 30 min. Sludge

yield was calculated according to Metcalf and Eddy

(1991).

3. Results

The reactors were inoculated at 35 �C and wastewaterfeeding, with HRT of 18 h, was started immediately.

Subsequently the temperature of A1 and B1 was in-

creased to 55 �C within 18 h. After 24-h operation at 55

�C the CODfilt removals were 24% and 60% for A1 and

B1, whereas after two days operation CODfilt removal

was 60% in both reactors. The combined processes re-

moved 91% CODfilt already after two-day operation.

Both thermophilic–mesophilic treatment lines were

operated under similar conditions, including HRT,

SRT, VLR, and temperature. The HRT was gradually

reduced in all reactors in three stages, from 18 h (days 0–

23) to 12 h (days 24–45) to 8 h (days 46–59), exceptduring the final period when B1 had HRT of 7 h (Tables

1 and 2, Fig. 2). Throughout the runs A-treatment (A1

followed by A2) gave CODtot and CODfilt removals of

81� 7% and 85� 5%, respectively, and B-treatment (B1

followed by B2) gave CODtot and CODfilt removals of

82� 8% and 87� 3%, respectively. A1 and B1 CODfilt

removals remained at the same level (60� 13% and

62� 7%) until HRT was reduced to 8 h in A1 and to 7 hin B1, resulting in CODfilt removals of 51� 9% and

69� 3% in A1 and B1, respectively (Table 1). Significant

difference was measured in A2 and B2 CODtot and

CODfilt removals with 8 h HRT; A2 CODtot and CODfilt

removals were 79� 3% and 75� 5%, respectively: for B2

both removals were 86� 1% (Table 2).

With HRTs of 12 and 8–7 h B1 had low SVI in all

tested samples and averaged 26� 8 ml g�1 (Table 1). InA1 the SVI values were notably higher and the average

was 184� 86 ml g�1: at the end of the trial the sludge did

not settle at all but floated on the surface of the mea-

surement vessel. In the mesophilic ASPs SVI was

115� 61 ml g�1 in A2 and 64� 40 ml g�1 in B2 (Table 2,

Fig. 2). HRT and VLR did not appear to influence the

reactor MLSS values. During most of the runs meso-

philic reactors had higher MLSS values than thermo-philic ones (Fig. 3). Typically effluents after treatment

lines had lower SS values than untreated wastewater.

The sludge yield in all processes, both mesophilic and

thermophilic, was less than 0.1 kg SS kgCODfilt removed

during most of the experiment (Table 1). Attached

growth in B1 carriers was detected after day 30 and was

included to B1 MLSS values from that day onwards.

After the first measurement, biofilm growth appeared to

Table 1

Performance of thermophilic aerobic ASP (A1) and SCBP (B1) processes under different loading rates and HRTs

Unit Period I days 1–23 Period II days 24–45 Period III days 46–59

A1 B1 A1 B1 A1 B1

HRT h 18� 0 18� 0 12� 0 12� 0 8� 0 7� 0

VLR kgCODfilt m�3 d�1 2.7� 0.1 2.7� 0.1 3.3� 0.4 3.3� 0.4 5.3� 0.4 6.0� 0.3

SLR kgCODfilt

kgMLVSSd�17.9� 6.4 7.1� 2.7 3.4� 1.1 1.8� 0.2 1.8� 0.6 2.2� 0.3

CODtot removal % 44� 13 45� 8 44� 5 48� 4 39� 2 44� 2

CODfilt removal % 58� 14 63� 7 66� 5 61� 6 51� 9 69� 3

CODsol removal % 84a 88a 81� 12 85� 4 nd. nd.

SVI ml g�1 nd. nd. 190� 05 26� 11 167a 25� 1

Sludge yield kg SSkgCOD�1filt rem 0.05� 0.03 )0.01� 0.00 0.02� 0.01 0.09� 0.01 0.11� 0.04 0.09� 0.02

nd.¼ not determined.aN ¼ 1, no standard deviation.

J. Suvilampi et al. / Bioresource Technology 88 (2003) 207–214 209

be rapid and reached steady values in a few weeks of

operation. Most of the biomass in B1 was attached to

carriers. In B1 the calculated SVI values accounted for

only suspended sludge and attached TS was excluded.

The COD in untreated wastewater had only a small

portion of suspended and colloidal COD; CODsol was

75–90% from CODtot. The thermophilic effluents were

turbid (visual perception) indicating the presence ofsuspended and colloidal matter. Selected samples (for

days 14, 34, 40, 44, and 45) indicated that thermophilic

treatment efficiently reduced CODsol (80–85%) and also

CODtot (55–60%), whereas CODsusp and CODcol remo-

vals were markedly poorer or even negative (Fig. 4).

Mesophilic post-treatments removed all CODcol from

thermophilic effluents and also efficiently eliminated

CODtot and CODsusp, whereas CODsol removals werenegligible (Fig. 4).

4. Discussion

The present results showed that the combined

thermophilic–mesophilic treatment of diluted molasses

wastewater produced good effluent quality. Thermo-

philic pre-treatment efficiently removed all biodegrad-

able CODsol. Mesophilic post-treatment removed the

colloidal COD (difference between CODfilt and CODsol)

present in the thermophilic effluents. Similar phenomenawere previously observed in a study in which thermo-

philic effluent CODfilt was notably reduced (40%) by 24-

h post-aeration at 35 �C, but not at 55 �C (Suvilampi

and Rintala, 2002). The removal under mesophilic

conditions probably took place because the colloidal

material, such as free bacteria, apparently present in

thermophilic effluent, aggregated and formed flocs in the

post-treatment process under mesophilic conditions.LaPara et al. (2001) also reported mesophilic post-

treatment to reduce COD from thermophilic effluent.

On the other hand, they suggested a mesophilic–meso-

philic combination to remove more COD than therm-

ophilic–mesophilic treatment. Good sludge settling

properties have been reported in a few studies (Barr

et al., 1996; Vogelaar et al., 2002a), in the latter study

Ca2þ had apparent effect in bridging the flocs. Thermo-philic bacteria have been shown to form weak and small

flocs, which cause poor sludge settling as well as higher

effluent COD values (Tripathi andAllen, 1999; Suvilampi

et al., submitted for publication b). The poor aggrega-

tion under thermophilic conditions may be because high

temperature reduces liquid surface tension, which in-

fluences the hydrophobicity of the bacteria, subse-

quently weakening their floc forming capabilities (Zitaand Hermansson, 1997). Another possible explanation

suggested by Vogelaar et al. (2002b) is that colloidal

particles originate from feed wastewater; this howeverTable

2

Perform

ance

ofmesophilic(A2andB2)andcombined

thermophilic–mesophilic(A12andB12)aerobicprocesses

under

differentloadingratesandHRTs

Unit

PeriodIdays1–23

PeriodIIdays24–45

PeriodIIIdays46–59

A2

B2

A12

B12

A2

B2

A12

B12

A2

B2

A12

B12

HRT

h18�0

18�0

36�0

36�0

12�0

12�0

24�0

24�0

8�0

8�0

16�0

15�0

VLR

kgCOD

filt

m�3d�1

1.1

1.0

1.4

1.4

1.2

1.3

1.6

1.6

2.6

1.9

2.7

2.8

SLR

kgCOD

filt

kgMLVSSd�1

0.8

0.6

1.5

1.2

0.9

1.0

1.2

1.0

1.8

0.5

1.1

1.2

COD

tot

removal

%74

63

86

79

63

69

79

84

60

73

75

86

COD

filt

removal

%64

64

86

87

60

68

86

88

58

52

79

86

COD

sol

removal

%4a

3a

84a

88a

)16�45

16�22

81�5

87�3

nd.

nd.

nd.

nd.

SVI

mlg

�1

nd.

nd.

nd.

nd.

76�38

69�51

nd.

nd.

172�32

56�33

nd.

nd.

Effluent

SS

mgl�

1–

–18�19

80�53

––

55�38

40�42

––

160�68

120�125

Sludge

yield

kgSSkg

COD

�1

filtrem

0.06�0.12

)0.07�-

0.04

0.04�0.03

)0.03�-

0.02

0.08�0.07

0.03�0.07

0.05�0.03

0.07�0.02

0.07�0.05

0.11�03

0.10�

0.05

0.10�0.01

nd.¼

notdetermined.

aN

¼1,nostandard

deviation.

210 J. Suvilampi et al. / Bioresource Technology 88 (2003) 207–214

was apparently not the case here since CODcol was lowin untreated wastewater.

The results suggest that combined thermophilic–me-

sophilic treatment could result in significantly lower

excess sludge production than present conventional

mesophilic ASPs in, e.g., pulp and paper industry

wastewater treatment where sludge yields are between

0.3 and 0.5 kg SS kgCODremoved. The measured low

sludge yields of both thermophilic SCBP and ASPsupport the generally proposed low excess sludge pro-

duction of thermophilic aerobic processes as reviewed

by LaPara and Alleman (1999). However, mesophilic

treatments produced similar yields as thermophilic ones.

Low yields in mesophilic processes might have been due

to markedly lower loading rates, as after thermophilic

treatments 80–85% of readily biodegradable CODsol was

used and only CODsusp and CODcol were availablefor mesophilic microorganisms. Improved flocculation

and sludge settling properties under mesophilic condi-

tions did not reduce CODtot, but only transferred

CODcol to CODsusp or to CODtot (i.e. increase CODsusp

and CODtot). It was likely that both fractioned CODs

hydrolysed partly in mesophilic treatment and in thatway made CODsol available to microorganisms (which

subsequently decreased CODtot).

The application of combined thermophilic–meso-

philic wastewater treatment for industrial wastewaters

and wastewater partial recirculation might be a cost-

and energy-effective process option. Since a high-loaded

thermophilic unit cuts off most of the CODsol, a part of

the thermophilic effluent can be circulated back tomanufacturing process (via e.g. ultrafiltration) while the

rest may be treated by an external polishing treatment.

Thermophilic effluent for reuse in manufacturing plant

must be treated only for removal of suspended particles

while readily biodegradable substances, causing, e.g.,

slime formation in pipes and odour problems, are effi-

ciently removed in thermophilic treatment. If the rest of

a high-temperature wastewater from thermophilictreatment still needs cooling before subjecting it to me-

sophilic treatment, the low amount of CODsol in the

wastewater benefits the operation of, e.g., heat ex-

changers, due to lower slime formation and subse-

quently lower maintenance costs. The low sludge yield is

Fig. 2. (a) ðMÞ VLR A1, ð�Þ VLR B1, ð}Þ HRT A1, ð�Þ HRT B1; (b) ðMÞ VLR A2, ð�Þ VLR B2, ð�Þ HRT A2 and B2; (c) ðMÞ A1 CODfilt

removal, ð�Þ B1 CODfilt removal, ð}Þ A12 CODfilt removal, ð�Þ B12 CODfilt removal, (r) A1 SVI, (d) B1 SVI; (d) ðMÞ A2 CODfilt removal, ð�Þ B2CODfilt removal, ð}Þ A12 CODfilt removal, ð�Þ B12 CODfilt removal, (r) A2 SVI, (d) B2 SVI.

J. Suvilampi et al. / Bioresource Technology 88 (2003) 207–214 211

also of great importance for industries with high

wastewater production, such as pulp and paper industry

with typical daily flow rates of 10 000–30 000 m3, and

where daily amounts of excess biosludge may vary be-

Fig. 3. (a) ð–Þ A1 and A2 HRT, ð�Þ feed wastewater SS, ð}Þ A1 MLSS, ðMÞ A2 MLSS, ð�Þ effluent from A2; (b) ð–Þ B1 and B2 HRT, ð�Þ feedwastewater SS, ð}Þ B1 MLSS (include both TS-fix and suspended SS, ðMÞ B2 MLSS, ð�Þ effluent from B2.

Fig. 4. Relative changes of different (CODtot, CODsusp, CODcol, CODsol) CODs. (a) (1) thermophilic ASP (A1), (2) mesophilic ASP (A2) treating

thermophilic effluent from A1, and (3) combined thermophilic–mesophilic treatment line (A12). (b) (1) thermophilic SCBP (B1), (2) mesophilic ASP

(B2) treating thermophilic effluent from B1, and (3) combined thermophilic–mesophilic treatment line (B12). Positive bars indicate reduced COD,

negative bars indicate increased COD.

212 J. Suvilampi et al. / Bioresource Technology 88 (2003) 207–214

tween 3 and 10 tons of TS. However, it is obvious that

laboratory-scale experiments suffer from small reactor

and flow volumes and more precise sludge yield can be

confirmed only after long-term pilot- or full-scale stud-

ies.

To our knowledge this is the first reported study on

comparison of SCBP and ASP performance under

thermophilic conditions and with similar operation pa-rameters, such as VLR, SRT, and HRT. Also compar-

ative studies of SCBP and ASP under mesophilic

conditions are rare if any exist. In this study SCBP

CODfilt removal increased with increasing VLR,

whereas ASP removal decreased, which suggests that

SCBP might require a certain, and higher than ASP,

loading rate to maintain high COD removal. This is

apparently not related to temperature. Several authorshave reported SCBPs (or moving bed bioreactors

(MBBRs)) as well as ASPs and SBRs treating industrial

wastewaters, to operate under large variation of differ-

ent loading rates and COD removals (Couillard and

Zhu, 1993; Barr et al., 1996; Malmqvist et al., 1996;

Tardif and Hall, 1996; Broch-Due et al., 1997; Becker

et al., 1999; Jahren and Ødegaard, 1999a,b; Malmqvist

et al., 1999; Rusten et al., 1999; Tripathi and Allen,1999; Jahren et al., 2002; Suvilampi and Rintala, 2002;

Vogelaar et al., 2002b). It appears that biofilm reactors

can operate under markedly higher loading rates than

processes based on suspended sludge, however, a

thermophilic process with no specific sludge circulation

or biofilm fixation was reported to have extremely high

(80–150 kg COD m�3 d�1) VLR (Becker et al., 1999).

One of the reasons why biofilm processes have beenclaimed to have higher loading capacity is due to their

higher biomass concentrations (Jahren, 1999). In this

study the average MLSS (when in a biofilm reactor the

MLSS includes both suspended sludge and attached

sludge) in both thermophilic reactors was approximately

the same. Subsequently the substrate utilization rates (as

kg CODsol removed by kg MLSS within 24 h) were al-

most the same for both thermophilic processes. Thissuggests that biofilm processes have an advantage at

higher loading rates over suspended sludge process by

some other means, such as higher microbial activity and

better mass transfer capabilities (Lazarova and Manem,

1994), which might be enhanced under thermophilic

conditions. One possible advantage is also that the

sheltered areas inside carriers provide environments to

totally different bacteria than the flocs (Tiirola et al., inpress).

5. Conclusions

The present study showed that thermophilic–meso-

philic treatment of diluted molasses provides high COD

removals; while the thermophilic stage removes almost

all CODsol, and the mesophilic stage improves effluent

quality by decreasing CODtot and CODcol values.

Thermophilic SCBP CODfilt removal increased with in-

creasing VLR, whereas ASP removal decreased. All of

the processes had low sludge yields. Two-stage processes

maintained stable CODfilt removals throughout the tri-

als.

Acknowledgements

The Foundation for Research of Natural Resources

in Finland is acknowledged for financial support (grant

no. 1581/00). Sucros Ltd., Finland is acknowledged for

supplying the molasses. UPM Kymmene Ltd., Kaipola,

Finland, is acknowledged for supplying the seed sludge.

References

APHA, 1998. Standard Methods for Examination of Water and

Wastewater, 20th edition. American Public Health Association,

Washington, DC.

Banat, F., Prechtl, S., Bischof, F., 1999. Experimental assessment of

bio-reduction of di-2-ethylhexyl phthalate (DEHP) under thermo-

philic conditions. Chemosphere 39, 2097–2106.

Barr, T., Taylor, J., Duff, S., 1996. Effect of HRT, SRT and

temperature on the performance of activated sludge reactors

treating bleached kraft mill effluent. Wat. Res. 30, 799–810.

Becker, P., K€ooster, D., Popov, M.N., Markossian, S., Antranikian, G.,

M€aarkl, H., 1999. The biodegradation of olive oil and the treatmentof lipid-rich wool scouring wastewater under aerobic thermophilic

conditions. Wat. Res. 33, 653–660.

B�eerub�ee, P., Hall, E., 2000. Effects of elevated operating temperatures

on methanol removal kinetics from synthetic kraft pulp mill

condensate using a membrane bioreactor. Wat. Res. 34, 4359–4366.

Couillard, D., Zhu, S., 1993. Thermophilic aerobic process for the

treatment of slaughterhouse effluents with protein recovery. Env.

Pollution 79, 121–126.

Finnish Standards Association, 1988. SFS 5504, Determination of

chemical oxygen demand (CODCr) in water with closed tube

method, oxidation with dichromate. Finnish Standard Association,

Helsinki, Finland.

Gebara, F., 1999. Activated sludge biofilm wastewater treatment

system. Wat. Res. 33, 230–238.

Hansen, E., Zadura, L., Frankowski, S., Wachowicz, M., 1999.

Upgrading of an activated sludge plant with floating biofilm

carriers at Frantscach Swiecie S.A. to meet the new demands of

year 2000. Wat. Sci. Technol. 40 (11–12), 207–214.

Kaindl, N., Tillman, U., M€oobius, C., 1999. Enhancement of capacity

and efficiency of a biological wastewater treatment plant. Wat. Sci.

Tech. 40 (11–12), 231–239.

Jahren, S., 1999. Thermophilic treatment of concentrated wastewater

using biofilm carriers. Doctoral Thesis, Norwegian University of

Science and Technology, Norway.

Jahren, S., Ødegaard, H., 1999a. Thermophilic aerobic treatment of

high strength wastewater using the moving bed biofilm reactor.

African International Environmental Symposium �99. Pieterma-ritzburg South Africa. 8 pp.

Jahren, S., Ødegaard, H., 1999b. Treatment of thermomechanical

pulping (TMP) whitewater in thermophilic (55 �C) anaerobic–

aerobic moving bed biofilm reactors. Wat. Sci. Tech. 40 (11–12),

81–89.

J. Suvilampi et al. / Bioresource Technology 88 (2003) 207–214 213

Jahren, S., Rintala, J., Ødegaard, H., 2002. Aerobic moving bed

biofilm reactor treating thermomechanical pulping (TMP) white-

water under thermophilic conditions. Wat. Res. 36, 1067–1075.

LaPara, T., Alleman, J., 1999. Review paper; thermophilic aerobic

biological wastewater treatment. Wat. Res. 33, 895–908.

LaPara, T., Nakatsu, C., Pantea, L., Alleman, J., 2001. Aerobic

biological treatment of a pharmaceutical wastewater: effect of

temperature on COD removal and bacterial community develop-

ment. Wat. Res. 35, 4417–4425.

Lazarova, V., Manem, J., 1994. Advances in biofilm aerobic reactors

ensuring effective biofilm activity control. Wat. Sci. Tech. 29 (10–

11), 319–327.

Lim, B., Huang, X., Hu, H-Y., Goto, N., Fujie, K., 2001. Effects of

temperature on biodegradation characteristics of organic pollu-

tants and microbial community in a solid phase aerobic bioreactor

treating high strength organic wastewater. Wat. Sci. Tech. 43 (1),

131–137.

Malmqvist, �AA., Welander, T., Gunnarson, L., 1996. Suspended-carrier

biofilm technology for treatment of pulp and paper industry

effluents. Symposium Pre-print, the Fifth IAWQ Symposium

on Forest Industry Wastewaters. Vancouver, Canada, pp. 189–

194.

Malmqvist, �AA., Ternstr€oom, A., Welander, T., 1999. In-mill biological

treatment for paper mill closure. Wat. Sci. Technol. 40 (11–12), 43–

50.

Metcalf and Eddy, Inc, 1991. Wastewater engineering. Treatment,

disposal, and reuse, third ed. McGraw-Hill, New York, USA.

Ødegaard, H., Rusten, B., Westrum, T., 1994. A new moving bed

biofilm reactor––applications and results. Wat. Sci. Technol. 29

(10–11), 157–165.

Rusten, B., Johnson, C., Devall, S., Davoren, D., Cashion, B., 1999.

Biological pre-treatment of a chemical plant wastewater in high-

rate moving bed biofilm reactor. Wat. Sci. Technol. 39 (10–11),

257–264.

Suvilampi, J., Rintala, J., Nuortila-Jokinen, J., submitted for publi-

cation a. On-site aerobic suspended carrier biofilm treatment for

pulp and paper mill process water under high and varying

temperatures.

Suvilampi, J., Rintala, J., 2002. Comparison of activated sludge

processes at different temperatures: 35 �C, 27–55 �C, and 55 �C.Env. Technol. 23 (10), 1127–1134.

Suvilampi, J., Lehtom€aaki, A., Rintala, J., submitted for publication b.

Thermophilic–mesophilic wastewater treatment and biomass char-

acterization.

Tardif, O., Hall, E.R., 1996. Alternatives for treating recirculated

newsprint whitewater at high temperatures. Symposium Pre-print,

the Fifth IAWQ Symposium on Forest Industry Wastewaters 1996.

Vancouver, Canada.

Tiirola, M., Suvilampi, J., Kulomaa, M., Rintala, J., in press.

Microbial diversity in thermophilic aerobic biofilm process; Anal-

ysis by Length Heterogeneity PCR (LH-PCR) Water Research.

Tripathi, C., Allen, D., 1999. Comparison of mesophilic and thermo-

philic aerobic biological treatment in sequencing batch reactors

treating bleached kraft pulp mill effluent. Wat. Res. 33, 836–846.

Vogelaar, J., Bouwhuis, E., Klapwijk, A., Spanjers, H., van Lier, J.,

2002a. Mesophilic and thermophilic activated sludge post-treat-

ment of paper mill process water. Wat. Res. 36, 1869–1879.

Vogelaar, J., van Lier, J., Klapwijk, A., de Vries, M., Lettinga, G.,

2002b. Assessment of effluent turbidity in mesophilic and thermo-

philic activated sludge reactors––origin of effluent colloidal mate-

rial. Appl. Microbiol. Biotechnol. 59, 105–111.

Zita, A., Hermansson, A., 1997. Effects of bacterial cell surface

structures and hydrophobicity on attachment to activated sludge

flocs. Appl. Env. Microbiology 63, 1168–1170.

214 J. Suvilampi et al. / Bioresource Technology 88 (2003) 207–214