You are on page 1of 5

J. Agric. Engng Res.

(1999) 73, 59}63


Article No. jaer.1998.0391, available online at http://www.idealibrary.com on

Treatment of Dairy Wastewater Using an Up#ow Anaerobic Sludge Blanket


Reactor
H. N. Gavala; H. Kopsinis; I. V. Skiadas; K. Stamatelatou; G. Lyberatos
Department of Chemical Engineering, University of Patras and Institute of Chemical Engineering and High Temperature Chemical Processes,
GR-26500 Patras, Greece; e-mail: gavala@chemeng.upatras.gr

(Received 16 March 1998; accepted in revised form 5 November 1998)

Wastewater coming from cheese-producing industries in Greece is high in organic matter (about 40}60 g/l
Chemical Oxygen Demand, COD) since it generally contains discarded cheese-whey as well. This wastewater is
rich in easily biodegradable carbohydrates and has a relatively low content in suspended solids (1}5 g/l).
Because of the high organic content of dairy wastewater, anaerobic digestion is essentially the only viable
treatment method.
An Up#ow Anaerobic Sludge Blanket (UASB) reactor is a high rate treatment system, especially feasible
for treating soluble (containing low solids) wastewaters. A well-performing UASB reactor is characterized by
highly #occulated, well-settling, compact methanogenic sludge granules, resulting in a very high biomass
content.
The aim of this work is the study of a UASB reactor treating dairy wastewater. A UASB reactor of 10 l
useful volume was constructed and inoculated with anaerobic mixed liquor from dairy wastewater and glucose
fed digesters. The digester e$ciency of treating dairy wastewater at various organic loading rates was studied
and its performance was assessed by monitoring pH, dissolved chemical oxygen demand (COD), biogas
production and composition. Operation at an organic loading rate of 6)2 g COD/l d was found to be safe
and could be increased to a maximum of 7)5 g COD/l d . A draw and "ll digester fed with non-diluted
wastewater was also studied in parallel for comparison purposes. The advantages of using a UASB
reactor versus a conventional digester become questionable for the high COD ('42 g/l), non-diluted dairy
wastewater case.
 1999 Silsoe Research Institute

1. Introduction a method for reducing the instability and low e$ciency


problems caused by its high organic content, espe-
Dairy wastewater comes either from milk or cheese- cially for high-rate anaerobic systems, such as Up#ow
producing industries; in the "rst case the wastewater has Anaerobic Sludge Blanket (UASB) reactors2 or Up#ow
a low chemical oxygen demand (COD) of 3000 mg/l, Anaerobic Filters.4 The two-stage anaerobic digestion
whereas in case of cheese-producing industries the of cheese whey has been studied also by Cohen et al.5
generated wastewater generally contains discarded using two pilot scale UASB reactors. The combination
cheese whey as well, with a COD value of 50 000 mg/l.1 of two treatment methods, biological and chemical, is
Aerobic processes are most suitable for the treatment another possible option for treating cheese whey and
of milk-producing industries wastewater, although dairy wastewater.6,7 Co-digestion of cheese whey with
there are several studies on anaerobic methods for animal wastes8,9 or domestic wastewater10 has also been
treating this kind of wastes as well.2,3 On the other studied.
hand, anaerobic digestion is essentially the only viable The aim of this work is to study the behaviour of
method for treating high organic content wastewater a high rate system, a UASB reactor, when treating dairy
coming from cheese-producing plants. The dilution (including cheese whey) wastewater and to compare it
of cheese whey by mixing with other wastewater is with the use of a conventional digester.
0021-8634/99/050059#05 $30.00/0 59  1999 Silsoe Research Institute
60 H . N . G AV AL A E A .

2. Materials and methods was changed so as to allow a satisfactory COD removal.


A recirculation rate of approximately 52 ml/min was
Determinations of dissolved (after "ltration) chem- used throughout. Although long-term population shifts
ical oxygen demand (COD), total Kjeldahl nitrogen could be observed upon changing the HRT or the OLR,
(TKN) and total suspended solids (TSS) were carried out changes were e!ected whenever it was felt that a quasi-
according to standard methods.11 For total P determina- steady state was reached (typically two residence times).
tion the persulphate digestion method and the ascorbic For comparison purposes, an 8 l (useful volume) draw
acid method (standard methods11) have been employed. and "ll digester was inoculated with anaerobic mixed
liquor from dairy wastewater and glucose fed digesters
and was fed with non-diluted dairy wastewater (COD of
2.1. =astewater characterization 60 g/l) at progressively lower hydraulic retention time.
The two reactors were operated at the mesophilic
Characterization of a cheese-producing industry temperature range (353C) throughout the period of the
wastewater gave a pH of 4)7, TSS of 2)5 g/l, dissolved experiments. Wastewater was continuously fed into the
COD of 60 g/l, TKN of 830 mg/l and total P of 280 mg/l. UASB reactor using a variable-speed peristaltic pump,
Di-ammonium phosphate (NH ) HPO was added to whereas the draw and "ll digester was fed manually on
 
make up for N and P de"ciencies; NaOH was used for a daily basis.
pH adjustment of the feed.

3. Results and discussion


2.2. Reactor design and operation
3.1. ;p-ow anaerobic sludge blanket experiment
A 10 l (useful volume) UASB reactor was constructed with dairy wastewater
of Plexiglas (total height of 113 cm with inner diameter of
11 cm), inoculated with anaerobic mixed liquor from Figure 1 shows the digester in#uent and e%uent COD
dairy wastewater and glucose fed digesters and fed with concentration as well as the system e$ciency (based on
dairy wastewater. The objective was to determine the %COD removal) versus time while Fig. 2 shows the
maximum possible organic loading rate (OLR) as well as hydraulic retention time (HRT) changes, the organic
the maximum in#uent COD concentration that allows loading rate (OLR) changes and pH values over time.
satisfactory COD removal. The digester was initially fed The maximum digester COD removal e$ciency of
with diluted dairy wastewater (2500 mg/l COD); sub- 98% was reached at an HRT of 6 d with an in#uent COD
sequently the in#uent COD concentration was increased concentration of 37 g/l (OLR"6)2 g COD/l d). When
gradually (by reducing the dilution) while at the same increasing the in#uent COD concentration to 42 g/l
time the hydraulic retention time (HRT) of the system (OLR"7)5 g COD/l d), the COD removal e$ciency

Fig. 1. Inyuent and e{uent chemical oxygen demand (COD) concentration and system ezciency (based on COD removal) versus time
for an Upyow Anaerobic Sludge Blanket reactor treating dairy wastewater
TRE AT M EN T O F DA IRY W AS T EW A TE R 61

Fig. 2. (a) For an Upyow Anaerobic Sludge Blanket reactor treating dairy wastewater: hydraulic retention time (HRT) and organic
loading rate (OLR) changes versus time; (b) biogas production and organic loading rate (OLR) changes versus time; (c) pH values
versus time

was reduced to 85}90% with a mean COD e%uent 3.2. Suspended growth experiment
concentration of 5 g/l. After this point, the increase of
in#uent COD resulted in even lower e$ciencies accom- For comparison purposes, an 8 l (useful volume) draw
panied by a sharp decrease in biogas production and pH and "ll digester was fed with non-diluted dairy waste-
values as well. Consequently, the total experiment may water (COD of 60 g/l) at progressively lower hydraulic
be divided into three periods (Fig. 1): period I, with al- retention times as shown in Fig. 3. The highest possible
most complete COD removal; period II, with constant OLR was found to be 2)3 g COD/l d at a hydraulic
in#uent COD concentration but with a progressive in- retention time of 26 d.
crease in HRT from 6 d to approximately 20 d in order to Table 1 presents typical values for the performance
maintain satisfactory COD removal (80}90%); and of each reactor during the di!erent phases of the
period III during which the in#uent COD concentration experiments.
was increased to the undiluted value. In this period, the In general, it is di$cult and risky to compare systems
HRT had to increase above 30 d in order to sustain operated in di!erent laboratories, meaning that the an-
reasonable reactor performance. aerobic sludge history and characteristics might be quite
Thus, an organic loading rate for a UASB reactor di!erent. However, in Yan et al.12 the removal e$ciency
treating dairy wastewater of 6)2 g COD/l d (diluted to of a UASB reactor has been reported to be about
37 g COD/l, with an HRT to 6 d) may be safely used and 81}86% at HRT 5}6 days with an in#uent COD concen-
could be increased up to 7)5 g COD/l d. Above that tration of 41 g/l (OLR"7)9}8)2 g COD/l d) and 97}99%
OLR, reduced performance is observed; while for non- at an HRT of 5 d with an in#uent COD concentration of
diluted dairy wastewater, an HRT in excess of 30 d is 5}29 g/l (OLR"0)91}6 g COD/l d). In Yan et al.,13 it
required. was stated also that the UASB reactor could treat cheese
62 H . N . G AV AL A E A .

Fig. 3. E{uent chemical oxygen demand (COD) concentration, percentage COD removal and hydraulic retention time (HRT) versus
time for a draw and xll digester treating dairy wastewater

whey of strength up to 28)8 g COD/l. From this study, it UASB or suspended growth reactors (mostly with sys-
was concluded that the maximum digester COD removal tems based on attached microorganism growth). Indeed,
e$ciency of 98% was reached at an HRT of 6 d with this fact is not surprising. The present paper focused on
an in#uent COD concentration of 37 g/l (OLR" the comparison of a UASB and a draw and "ll reactor
6)2 g COD/l d). When increasing the in#uent COD that were inoculated with similar anaerobic sludge and
concentration to 42 g/l (OLR"7)5 g COD/l d), the treated the same wastewater.
COD removal e$ciency was reduced to 85}90%. Addi-
tionally, the removal e$ciency of a continuous stirred
tank reactor was reported12 to be about 18}58% for 4. Conclusions
HRT values of 14}70 d with an in#uent COD concentra-
tion of 69 g/l. The present experiments showed that the An Up#ow Anaerobic Sludge Blanket (UASB) re-
removal e$ciency of the draw and "ll digester lay be- actor was used successfully for treating dairy waste-
tween 83 and 92% for HRT values of 26}40 d with an water containing large amounts of cheese whey, while
in#uent COD concentration of 60 g/l. In each case, the a draw and "ll digester fed with non-diluted waste-
treatment e$ciencies were similar and/or better than water was also studied in parallel for comparison pur-
those reported by Yan et al.12,13 The organic loadings poses. Operation of the UASB reactor at an organic
were reported12 to be much higher (up to 27 g COD/l d) loading rate of 6)2 g COD/l d was found to be safe
than those in this paper only with systems other than and could be increased to a maximum of 7)5 g COD/l d.

Table 1
Typical values for the performance of each reactor during the di4erent phases of the experiments

Duration, pH Mean inyuent Hydraulic Organic Removal, %


d COD, g/l retention loading rate, (mean vlaue)
time, d g COD/l d

UASB experiment
1st period 76 7)0}7)3 12}44 6 2)0}7)3 85}99
2nd period 111 6)6}7)2 45 10}20 2)3}4)5 79}91
3rd period 77 6)6 57}60 30}40 1)5}1)9 81

Draw & Fill experiment


1st period 41 7)0 60 35}40 1)5 94
2nd period 12 7)0 60 29}32 2)1}1)9 94
3rd period 12 (6)5 60 26 2)3 (83

COD, chemical oxygen demand.


TRE AT M EN T O F DA IRY W AS T EW A TE R 63

5
The high retention times required for non-diluted waste- Cohen A; Thiele J H; Zeikus J G Pilot-scale anaerobic
water, however, do not justify the use of a UASB reactor, treatment of cheese whey by the substrate shuttle
as a less expensive conventional reactor, such as a Con- process. Water Science and Technology, 1994, 30(12),
433}442
tinuous Stirred Tank reactor could equally well be em- 6
Barford J P; Cail R G; Callander I J; Floyd E J Anaerobic
ployed for such a high COD wastewater ('40 g/l). digestion of high-strength cheese whey utilizing semicon-
tinuous digesters and chemical #occulant addition. Bio-
technology and Bioengineering, 1986, 28(11), 1601}1607
7
Acknowledgement Fang H H P Treatment of wastewater from a whey process-
ing plant using activated sludge and anaerobic processes.
Journal of Dairy Science, 1991, 74, 2015}2019
The authors wish to thank the Commission of the 8
Lo K V; Liao P H; Chiu C Mesophilic anaerobic digestion of
European Communities for the "nancial support of this a mixture of cheese whey and dairy manure. Biomass,
work under grant No BRE2-CT92-0355. 1988, 15(1), 45}50
9
Gavala H N; Skiadas I V; Bozinis N A; Lyberatos G Anaer-
obic codigestion of agricultural industries wastewaters.
Water Science and Technology, 1996, 34(11), 67}75
References 10
Guiot S R; Sa5 B; Frigon J C; Mercier P; Mulligan C;
Tremblay R; Samson R Performances of a full-scale novel
1
Nemerow N L Dairy wastes. In: Industrial Water Pollution multiplate anaerobic reactor treating cheese whey e%uent.
Malabar, FL: R E Krieger, 1987, 378}391 Biotechnology and Bioengineering, 1995, 45(5), 398}405
2 11
Oztu3 rk I; Eroglu V; Ubay G; Demir I Hybrid up#ow American Public Health Association, American Water Works
anaerobic sludge blanket reactor (HUASBR) treatment of Association, Water Pollution Control Federation Standard
dairy e%uents. Water Science and Technology, 1993, Methods for the Examination of Water and Wastewater.
28(2), 77}85 17th edition, Washington, 1989
3 12
Viraraghavan T; Kikkeri S R Dairy wastewater treatment Yan J Q; Liao P H; Lo K V Methane production from cheese
using anaerobic "lters. Canadian Agricultural Engineer- whey. Biomass, 1988, 17(3), 185}202
13
ing, 1990, 33, 143}149 Yan J Q; Lo K V; Liao P H Anaerobic digestion of cheese
4
Mendez R; Blazquez R; Lorenzo F; Lema J M Anaerobic whey using an Up#ow Anaerobic Sludge Blanket reactor:
treatment of cheese whey: start-up and operation. Water III. Sludge and substrate pro"les. Biomass, 1990, 21(4),
Science and Technology, 1989, 21(12), 1857}1860 257}271

You might also like