Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.enzyme.chem.msu.ru/ekbio/article/1997_hybrid.pdf Äàòà èçìåíåíèÿ: Sun Jan 5 13:02:24 2003 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 22:26:26 2012 Êîäèðîâêà:

Copyright@ 1997 by Humana Press Inc. All rights of any nature whatsoever reserved. 0273-2289/97/6603-0291$10.75

~

.

,,;:,,;\

The Anaerobic Treatment of Soft Drink Wastewater in UASB arid Hybrid Reactors
S. V. KAlYUZHNYI,*,l,2 VAlADEZ SAUCEDo,2 J. AND J. RODRIGUEZ MARTINEZ2
1Department of Chemical Enzymology, Chemistry Faculty, Moscow State University, 119899 Moscow, Russia; and 2Biotechnology Department, Chemistry Faculty, Autonomous University of Coahuila, 25000 Saltillo, Mexico Received March22, 1996;Accepted November 1996 11,

ABSTRACT
The anaerobictreatment of soft drink wastewater (SOW)was studied in two laboratory reactors-a 1.8-T~ UASB reactor and a 3-L hybrid reactor-sludge bed containing a layer of polyurethane in the upper part, at 35°C.The~ghest organic loading rates(OLR) achievedwere 13and 16.5 g COO/L. 0 for hybrid and UASB reactors,respectively, with the treatment efficiency of about 80%for both reactors. Despite the higher treatment productivity achieved for the UASB reactor, its lower ability to generate sufficient level of alkalinity led to difficulties in maintaining a a stable operation performance.Therefore,the hybrid reactor seemsto be indicated for OLR higher than 10 g COO/L. d and HRT lower than 1 0, from the point of view of reliability of thesetwo systems. Both reactorscan treat the SOWwith pH influent up to 11.0.ll1e feeding of reactorswith higher pH influent values led to their quick failure because alkali shock. The duration of the recovery period after alkali of shockwas about 1.5-2 mo. Index Entries: UASB reactor; hybrid reactor; soft dru1k waste\vater; performance;alkali shock. INTRODUCTION Typical wastewaters from soft drink industries are mainly composed of \vashing waters from production lines. For this reason,they ha\'e a mod*Author to whom all correspondenceand reprint requests should be addressed.

"

AppliedBiochemistty Biotechnology and

291

Vol.66, '997


."'::'",'C,c'f

5
;"~)
Table 1

:

~

..7

Kalyuzhnyi, Saucedo, and Martinez

.

Comparison of OLRs and Treatment Efficiencies Achieved in Various Anaerobic Treatments of Soft Drink Wastewaters :tor
~p.,

type
DC)

OLR.
9COD/I"day

Treatment
efficiency, %

Reference
-

[continuos

batch

with

)ended support
Irobic es with filter UASB in

(0.66

*

)

78

(3)

0.88

75

(4)

in

series

with

:c~

;; ~
~
" " :

robic

filter:
,

1.22
2.78

87
73

(5)
(5)

'..I -c '1

robic

filter

0.88

87

(5)
'c

,i o~, ~ "*f ~~
-,}. "',

; J

I

;

7.5

78

(6)

~ ~.
",,'"

Lng

Anaerobic

6-8

80

(7)

!:~ "j

?f; ,., !;}'

.'

LIter

Systems

6-10

85-95

(7)

16.5

80

this

study

d parent (calculated from

13.0 feed data).

82

this

study

:oncentration of pollutants (usually not more than 10 g COO/L), th~ )sition of which is derived from the ingredients used in the final prodd high pH (up to 13.0). Owing to the nature of the main ll1gredients, llit syrups, sugars, flavorings, colorants, and so forth, the organic pol5 are soluble and easily biodegradable (1-3). b1 practice, many soft wastewaters (SOW) are frequently treated together with beer waste;, but little research has been reported regarding the application of "n high-rate an~erobic digestion technologies for the treatment of
>nly.

Some features of recent studies (3-7) are summarized. in Table 1.

lent productivities and organic loading rates (OLR) achieved in these ; were relatively low in comparison with those usually observed for confaining wastewaters (8-10). The possible explanation can be

iochemislty and Biotechnology

Vol. 66, 1997

;.

*
;~i:


~::'

~;:;k~;~::;:;;.. " ~,'ii;".,'" related to low sludge activities caused, in turn, by nutrient deficiency, for
',c

:;t"

Anaerobic Treatment opSoftOrink Wastewpters f

~~
,';

:

:
..

example~ nitrogen, soforth. by and

:
'...

\'"

This article de~withthetreatment of SDWin two modern high-rate anaerobic reactors, the upflow anaerobic sludge blanket (UASB) reactor (8, 11,12) and the so-calledhybrid reactor-sludge bed combined with anaero" bic filter (13), which we consider to be the most appropriate reactor systems for treatment ofSDW in Mexican conditions from the economic point of view. Th~ aim of this work was to (1) compare treatment efficiency (TE) of these two reactors under elevated OLR and (2) investigate an effect of high pH value of i:nfluent SOW on performance of both reactors as well as their recovery after alkali shock.

.

METHODS Reactors The UASB reactor of 2.1 L was made from glass (internal diarneter6 cm, total working volume-1.B L). The temperature (35°C) in the reactor was thermostatically: controlled by pumping water from the thermostat through the jacket surrounding the reactor. A relatively small (60-mL) conical gas separator was installed in the upper part of the reactor. The hybrid reactor (UASB in the down part and anaerobic'filter in the upper part with a 5.5-cm layer of polyurethane) of 4.24 L was made from plastics (internal diameter-11.4 cm, total working volume-4 L). During the experiments the reactor was placed in a thermostat (35°C). No recycling or mixing facilities were provided for both reactors, which were fed by pumps. Wastewater The SOW used was obtained from a continuous soft drink processing factory located in Saltillo, Coahuila (Mexico). The daily stream of wastewaters from this factory was on the average about 400 m3. The main characteristi'cs of SOW were very variable because of different production regimes of the factory. The range of variation of these characteristics during the period of execution of our experiments (January-September 1995) is shown in Table 2. Becauseof a very low proportion of nitrogen/COD and low buffer capacity, the SOW was supplemented by 0.5-1 g/L NH4Cl and 1-6 g/L sodium bicarbonate(when previous neutralization of SOW was necessary). Seed Sludge and Start-up of Reactors The hybrid reactor was seeded with 1 L of sludge originating from an anaerobic lagoon (Toluca, Mexico) treating beer industry wastewater. The concentration and specific methanogenic activity of this sludge ~re 12 g
AppliedBiochemislty Biotechnology and Vol,66, 1997

;


i

)? .

and carbon dioxide concentrations ~asured using gas chromatography as previously described (15). lIne of methane producedwasdetern1ined by the liquid displacethoc!after removing CO2by adsorption into NaOH solution (8). All uyseswere performed according to standard methods (8, 16).
atile fatty acids (VFA), methane,

; AND DISCUSSION Period generalized results of running period for both reactors are shown -3. It is seenthat the influent COD concentration applied was subcarp\Tariation (Fig. 2), becausewe tried to work in the conditions the teal conditions in the factory. Concerning the applied OLR the Qeneralstrategy consisted of its stepwise increasing (without riation) for both reactors. For this reason,applied HRT also varied


r*"'l!ci
~.;:;: ' t,;l.:: ,,' ... i"' ' ~cc;;;

':

I"

c

::

ii'!

10

Time, days

Fig. 1. Applied HRT and OLRfor the VASB(A) and hybrid (B) reactors treating SOW 35°Cduringrunningperiod. at

The results obtained show that during gradual increaseof the OLR up to 8 g COD IL . d (first 2 moof the experiment), both reactors demonstrated good operational performance with the TE on the total COD usually higher than 80% (Figs. 1,2). In this,time, U1e formation of good settling,granules was detected visually as well as by light microscopic examination of samples taken from the sludge bed zone of both reactors. On further gradual increaseof the OLR up to 10-12 g COD II . d, the hybrid reactor continued to maintain the TE, on the average, higher than 80%(Fig. 2B), but the TE of the UASB reactor decreasedto 60-70% (Fjg. 2A). The effluent COD concentrations were, on the average, 1 (Fig. 2B) and 2 g COD/L (Fig. 2A), respectively. The relatively worse performance of ~e UASB reactor at elevated OLR and shorter HRT «1 d) can be mainly attributed to increased biomass washout, which was observed on this stage of the experiment (effluent COD insoluble varied from 0.3-0.5 g CaDIt). On the contrary, the biomass washout from the hybrid reactor was not as noticeable (effluAppliedBiochemislty Biotechnology and Vol.66, 1997

,-.;~I

.:;"":1


1__11 :-

r
I

~

Kalyuzhnyi,Saucedo,and Martinez

~ncentration. COD/I 9

TE,"
1100

"r25/ 20r 15 !
10

l
I

CO~
I

I.

I I 1-.

-1""\
TE

:80

l
. ".."-"'"1
40 50 80

'so
40 f

1

5
0 ".
0 10

J~OD'fr
'...' '
20 30

I Uc__,20
70 80 0

.,"

;':

-'. 0.":,:,,, :~""'.'

iB

Concentration,9 COD/I r---" '

TE, % ---,
1100

:,~, c..
" ..

:;f~lA-l;-~r
~
:}
lol
1 ',

j

80

::

I G~
0

-n
I
10

CODln

,

CODe!!
.., 40 days ""."'..'.'.'-.." 50 so 30 Time, 70

20 ..,...' 0

oC

",=:=::=:.:: 20

:. Influent and effluent COD concentrationsand treatment efficiency for the l) and hybrid (B) reactorstreating SDW at 3SoC during ruru1ingperiod.

,

) insoluble was 0.1-0.2 g COD/L) becauseof more effective reten)iomass aggregatesby polyurethane layer. e biogas content was sufficiently stable (CH4-60-65%; COzJr both reactors as well as methane yield (0.32-0.33nL/ g COD conDuring the last week of running period, the sludge taken from the. e of hybrid reactor had the following characteristics: concentra\-20 g VSS/L of bed zone, speci~ic methanogenic activity)

~

~ I

j
'
;

g CH4-COD g VSS. d. The analdgous / parameters the sludge, of

:

>m the bed zone of UASB reactor during the last week of the run'iod were slightly higher: concentration-22-24 g VSS/L of bed ecific methanogenic activity-o.61-o.63 g CH~-COD/g VSS . d ile, determina tion of sludge profile iI1 the both reactors performed me time sho",ed that the mean sludge concentration (pel total reactor volume) ,,,,,asslightly lower for the UASB reactor (8-9 g
:

total reactor volume) than for the hybrid rea~tor (10-11 gVSS/L

;
i

lemislry and Biotechnology

Vol. 66,

/.99;/

I j i ,

i


AnaerobicT rootment SoftDrinkWastewaters of
;..i

297
70 80

pH
11 10 8 8 7 8 5

A

0

10

20

30

40

&0

80

pH
11

B

.
.':;~: .;

10

.. & 0 10 20 30 40 50 60 70

Time,

days

Fig. 3. Influent and effluent pH for the UASB (A) and hybrid (B) reactors treating SOWat 35°Cduring running period.

of total reactor volume). Thus, the mean sludge loading rates (SLR) during the last week of the running perio.d were 1.29 and 0.95 g COO/gVSS. d for the UASB and hybrid reactors, respectively. Taking into account low alkalinity of the fresh SOW (Table 2), special attention was paid to prevent acidification of the reactor media by addition of bicarbonate to the influent. In spite of the fact that the quantity of bicarbonate added was usually higher for the VASE reactor than for the hybrid reactor, the effluent pH was usually lower in the first case (compare Figs. 3A and B) because of the higher concentration of YFA (C2-C4) generated (on the average, the effluent VFA concentrations were 1 and 0.5 g COD/L for the UASB and hybrid reactors, respectively!). Though the UASB reactor was able to cope even with the medium pH of about 5.5 (Fig. 3A) without noticeable decrease in the TE (Fig. 2A), the observed low medium alkalinity was a subject of continuous \\rorries owing to the danger of an acidif~cation -collapse of the system. The above-mentioned problems were less
Applied Biochemistry and Biotechnology Vol. 66, '997


,

;

'-


,';::J;~~:;': ,

'0.

:c

.,

Anaerobic Treatmen(ofSoft Drink Wastewaters

.

299
100
0

Concentration, 9 COD/I
;
;'

16

A

.

'

.
.'

,,'..
0

""...

~

"

""'>'1

0

0

0

10

20

30

40

60

60

TO

80

gO

0 100 100 i 0

Concentration, 9 COD/I
15

TE, %

B

1

0

40

20

0

10

20

30

40

50

60

TO

80

90

0 100

Time, days

Fig. 5. Influent and effluent COO concentrationsand treatment efficiency for the UASB (A) and hybrid (B) reactorstreating SOW at 35°C during alkali shock recovery period.

restored at day 10 after the shock (Fig. 5). Further on, the slow stepwise increase of the OLR up to 2.5 g COD /L . d led to the increase of the TE higher than 80% at days 45-50. Thus, the recovery period of 1.5 mo seems to be necessary for restoring the sludge bacterial activity after the alkali shock. This is comparable with the duration of the start-up procedure usually observed for anaerobic reactors inoculated by disperse sludge (8,17). After termination of the recovery period (day 50), the OLR were increased more rapidly and reached 13 and 16.5 g COD/L . d for hybrid and VASE reactors, respectively (Fig. 4). The average TE was about 80% for both reactors (Fig. 5). Tl1e formation of new granules was observed in both reactors at this stage of the experiment. Tl1e higher OLR achieved for the VASE reactor after the shock were a surprise, and can be explained by higher specific methanogenic activity (0.7-0.73 g CH4-COD/g VSS . d) detected in this reactor (day 99) in comparison with that (0.55-0.61 g CH4COD / g VSS . d) for the hybrid reactor. The reactor medium pH oscillated
Applied Biochemistry and Biotechnology Vo'- 66 1QO7


6 0 10 20 30 40 60 80 70 80 80 100

pH
10

B

I

9

-L---'---

,
SO 40

.
60

,
60

,
70 80 80 100

0

10

20

Time, days

:nfluentand effluentpH for theUASB(A) and hybrid (B) reactors treating ;oC during alkali shock recovery period.

I and 6.0 for hybrid and UASB reactors, respectively (Fig. 6). The :anpotentially lead to disturbing the operation performance of the actor. Further increase of the OLR led to decreasing the TE for ~tors(more sharply for the UASB reactor). Therefore, the experi~restopped. .mparison of the OLR and the TE achieved with those reported for lerobic treatment systemsfor the SOW (Table 1) indicates that our 'e comparable with the best ones from the point of view of the TE ~d at least two times the reported data on the OLR applied. JSIONS .tudy showed that, on the whole, the UASB and hybrid reactors demonstrated satisfactory TE and operation performance r anaerobic treatment of SDW. The highest OLR achieved were
Jemistry and Biotechnology Vol. 66, /997

.
.:~: 0. :;"'~: ~

"

c::j,:.",..

;.;,,1'.~,;;;:;,1!:g

:~.


,~

Anaerobi2Tr~atment SoftDrink Wastewaters 6j
-,

301

13 and 16.5g COD /L . d for hybrid and UASB reactors,respectively, with the TE of about 80%for both reactors. . 2. In spite of a higher OLR achieved for the UASB reactor, its lower abilitY to generatea sufficient level of alkalinity can potentially lead to difficulties in maintaining stable reactor operation. Therefore, the hybrid reactor seemsto be preferential with the OLR higher than 10 g COD /L . d and the HRT shorter than 1 d from the point of view of reliability of thesetwo systems. 3. Both reactors can treat the SDW with pH influent up to 11.0. The feeding of reactors with higher pH influent values led to their failure becauseof alkali shock. 4. The duration of recoveryperiod after alkali shock was about 1.5-2 mo.

.
REFERENCES
1. O'Shaughnessy, J. C., Blanc, F. C., Corr, S. H., and Toro, A. (1987),in Proc. Ind. Waste ConI, USA, pp. 607-617. 2. Borup, M. B. and Ashcroft, C. T. (1991),J. Water Pollut. Control Fed.60,445--448. 3. Botja, R. and Banks, C. J. (1994), Chern.Tech.Biotechnol. J. 60,327-334. 4. Vicenta, M., Pacheco,G., and Anglo, P. G. (1989),in Alternative EnergySources, Proc.8th, Miami Int. Conf., Veziroglu, T. N., ed., Hemisphere, New York, pp. 865-875. 5. Silverio, C. M., Anglo, P. G., Luis, V. S. J., Avacena, V. P., and Ana, A. S. (1989), in Alternative EnergySources, Proc.8th, Miami Int. Coni. Veziroglu, T. N., ed., Hemisphere, New York, pp. 843-853. 6. Stronach, S. M., Rudd, T., and Lester,J. N. (1987),Biomass 173-197. 13, 7. Vogel, P. and Nagatani, I (1994), PosterPaperPreprintsof Seventh International Symposium of AnaerobicDigestionGanuary23-27,1994), Cape Town, South Africa, pp. 252-255. 8. Lettinga, G. and Hulshoff Pol, L. W., eds. (1992),International Course AnaerobicWaste on WaterTreatment,Wageningen,Agricultural University, the Netherlands. 9. Kalyuzhnyi, S. V., Sklyar, V. I., Davlyatshina, M. A., Parshina, S. N., Simankova, M. V., Kostrikina, N. A., and Nozhevnikova A. N. (1996),Bioresource Technol.55,47-54. 10. vanStarkenburg, W. (1996),in Proc. EERO Workshop "Methanogenesis Sustainable for EnvironmentalProtection,"St. Petersburg,Russia Gune 19-21, 1996),p. 16. 11. Lettinga, G., van Velsen, A. F. M., Hobma, S. W., de Zeeuw, W. J., and Klapwijk, A. (1980),Biotechnol.Bioeng. 22,699-734. 12. Lettinga, G., Hulshoff Pol, L. W., Jansen,A., Fi~d, J., van Lier, J. and Rebac, S. (1996), in Proc EERO Workshop"Mefhmlogcncsis SustainableEnvironlrlental Protection," St. for Petersburg, Russia Gune 19-21, 1996),pp. 6, 7. 13. Guiot, S. R., and van den Berg, L. (1985),Biotechnol.Bioeng.27,800--806. 14. Kalyuzhnyi, S. V., Estradade los Santos, L., and Rodriguez Martinez, J. (1995), in Proc. VI Mexican Congress Biotechnol. Bioeng., lxtapa, p. 106.(BIOAMB CO8). 15. Varfolomeyev, S. D. and Kalyuzru1yi,S. V. (1989),Appl. Biochem. Biotechnol. 22,331-350. 16. StandardMethods the E.talllinationof Waferand Wastervater, for 14th ed. American Public Health Association, Washington, 1975. 17. Kalyuzhnyi, S. V., Danilovich, D. A., and Nozhevnikova, A. N. (1991),Anaerobicbiological treatment of rvastervalers, VINITI Press (Itogi nauki i tehniki, ser. Biotechnology, 29), Moscow.

AppliedBiochemislry Biotechnology and

. Vol.66, 1997