Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.enzyme.chem.msu.ru/ekbio/article/abb_2003_253263.pdf Дата изменения: Sat Jul 19 13:31:06 2003 Дата индексирования: Mon Oct 1 22:52:35 2012 Кодировка:

"a ;n 5, 'r , .,
~

Copyright @2003 HumanaPressInc. by All rights of any nature whatsoeverreserved. 0273-2289/03/109/0253/$20.00

-

CombinedBiologic (Anaerobic-Aerobic) and ChemicalTreatment
of Starch Industry Wastewater
VLADIMIR SKLYAR,1 ANDREY DANILOVICH,2 EpOV,1 AND MARINA GLADCHENKO,1

DMITRII
1

SERGEY KALYUZHNYI*,1

Department of Chemical Enzymology, Chemistry Faculty,

Moscow State University, 119899 Moscow, Russia, E-mail: svk@enz.chem.msu.ru;and 2MOSVODOKANAL, Pleteshkovskiypereulok 2, 107005 Moscow, Russia Abstract
A combined biologic and chemical treatment of high-strength (total chemical oxygen demand [CODtoJup to 20 g/L), strong nitrogenous (totalNup to 1 g/L), and phosphoric (total P up to 0.4 g/L) starch industry wastewater was investigated at laboratory-scale level. As a principal step for COD elimination, upflow anaerobic sludge bed reactor performance was investigated at 30°C. Under hydraulic retention times (HRTs) of about 1 d, when the organic loading rates were higher than 15g of COD / (L.d), the CODtot removal varied between 77 and 93%, giving effluents with a COD/N ratio of 4-5:1, approaching the requirements of subsequent denitrification. The activated sludge reactor operating in aerobic-anoxic regime (HRT of about 4 d, duration of aerobic and anoxic phases of 30 min each) was able to remove up to 90% of total nitrogen and up to 64% of COD tot from the anaerobic effluents
under 17-20°C. The coagulation experiments with Fe (III) showed that 1.4mg of resting hardly biodegradable COD and 0.5 mg of phosphate (as P) could be removed from the aerobic effluents by each milligram of iron added. Index Entries: Activated sludge reactor; coagulation; nutrient starch wastewater; upflow anaerobic sludge bed reactor. removal;

Introduction The current Russian food industry is demonstrating impressive growth and is one of the biggest starch consumers. The volume of the starch market was 110,000t in 1999and continues to grow (1). Although the starch industry compared to the other branches of food industry is potentially
,.Author to whom all correspondence and reprint requests should be addressed. Applied Biochemistry Biotechnology and

253

Vol. 109,2003

..


254

Sklyaret al.

~
S ~ ..
"<

~
.., {/)

Coagulant addition

Final Air emuent

Air Influent /ntennediate reservoir

Fig. 1. Flow sheetof laboratory installation for treatment of starch industry wastewater. able to operate with a closed water cycle, up to 40% of its wastewater in Russia is discharged on wastewater treatment plants of neighboring towns or directly on so-called filtration fields. 5uch a practice not only increases expenses (i.e., payments to the wastewater treatment plants for treatment and increased concentrations of chemical oxygen demand (COD) and nutrients, penalties to the ecologic authorities) but also severely damages the environments and leads to nonrational usage of arable land. The internal starch production in Russia accounts for 45,000 t with 1.5 million cubic meters of wastewater to be treated (1). The wastewater content is highly variable, but the most typical concentrations of main pollutants in this wastewater are the following: 10 g/L of total COD (CODtoJ, 2.5 g/L of suspended solids (55), 0.6 g/L of total nitrogen, 0.15 g/L of total phosphorus (1). A possible solution for utilization/treatment of starch industry liquid refuses would include separation into a solid and a liquid fraction. The solid fraction, mainly consisting of gluten, can be reused in a main technological process while the liquid fraction should be treated before discharge. The objective of this study of starch industry wastewater ants. As a first step, the upflow investigated for the elimination
Applied Biochemistry and Biotechnology .~

was to develop a technology for treatment targeting the aforementioned main pollutanaerobic sludge bed (UA5B) reactor was of the major part of COD and 55. This is a
Vol. 109,2003


Treatment of Starch Industry Wastewater Table1 Rangeof Variation of SomeCharacteristics of Wastewaterfrom Starch-Producing Factory !bred (Ryazan'province) Parameter pH CODtot (g/L) CODss (g/L) CODcol (g/L) CODsol (g/L) SS(g/L) Ash contentof SS(%) Total protein (g/L) Solubleprotein (g/L) Total nitrogen (g NIL) NN1i3 NIL) (g Total phosphorus(g PIL) Solublesugars(g/L) Variation 4.19-7.03 2.98-26.22 0.35-9.83 0.12-0.95 2.47-15.44 0.23-5.48 18.8-33.7 1.02-5.21 0.42-2.58 0.204-1.042 0.049-0.115 0.050-0.385 0.18-6.69

255

usual treatment step for starch-containing wastewater having a high strength (2-8). In a second step, the conventional activated sludge process operating in an aerobic-anoxic regime was used for removal of remaining biochemical oxygen demand and nitrogen. Finally, iron coagulation was applied for effluent clarification and phosphate precipitation. The flow sheet of the integrated laboratory installation used in this study is shown in Fig. 1. Materials and Methods Wastewater Wastewater was directly taken from gluten decanters of factory !bred (Ryazan' province) producing starch from maize. The range of variation of some characteristics of this wastewater is given in Table 1. VASa Reactor A laboratory UASB reactor (rectangular cross-section: 38 cm2;height: 85 cm; total working volume: 2.68 L) made from transparent plastic and equipped with six sampling ports along the reactor height was used. An operating temperature of 30 ::!: 1°C was maintained by placing the reactor into a thermostat TS-80 (Mashzavod, Odessa, USSR).The reactor was seededwith 1 Lofsludge (55gofSS/L;ashcontent: 30%;specific aceticlastic activity: 0.55 g of COD/[g of VSS.d] at 30°C) originating from a submesophilic UASB reactor treating pig manure wastewater (9). Anaerobic batch biodegradability and activity assayswere conducted as described by Lettinga and Hulshoff Pol (7).
Applied Biochemistry Biotechnology and .' Vol. 709,2003


256

Sk/yar et a/.

Activated Sludge Reactor The continuous-flow stirred tank type of reactor (Fig. 1) was made from transparent plastics and equipped with a mechanical stirrer. It had a working volume of 0.8 1 and functioned at ambient laboratory temperatures (17-20°C) in alternating aerobic/ anoxic regime for treatment of the anaerobic effluents. During the aerobic phase, the feeding was stopped, while air at a flow rate of 1 L/min was pumped through the reactor. Aeration was switched off during the anoxic phase, while the continuous feeding was restored. The effluent passed through the external settler. The sludge return was organized via air lift with an additional air separator (Fig. 1). A programmable multichannel timer controlled all pumps used. Additionally, in the reactor, an electronic sensor (Datchik, Russia) was inserted for online monitoring of soluble oxygen. The electric signal from this sensor was transferred to a programmable data logger system. Data were recorded every 30 s and were averaged (when necessary) over 3-min intervals. A personal computer programmed to function as a terminal emulator was used to communicate with the data logger. Secondary sludge from Kur'yanovskaya wastewater treatment plant (Moscow) was used as a seed sludge (initial concentration: 1.5 g of mixed liquor suspended solids [MLSS]/L). Excess sludge was periodically withdrawn from the settler. Nitrification and oxygen uptake rates (OUR) were determined as described by Klapwijk and Rensink (10). Coagulation Assays The assayswere performed with 200 mL of aerobic effluent ratory glass under continuous stirring and pH control. Addition lant (FeCI3.6H2O) was carried out at 200 rpm. Then the intensity was reduced to 40 rpm to complete a flocculation process during was maintained at 7.2-7.5 by the addition of NaOH. in a laboof coaguof stirring which pH

Analyses COD was analyzed spectrophotometrically using Hach tubes. Raw samples of influents or effluents were used to determine CODtol' 4.4-~mfolded-paper-filtered (Schleicher & Schuell 5951/t' Germany) samples to determine filtrated COD (CODfil)' and 0.45-~m-membrane-filtered(Schleicher & Schuell ME 25, Germany) samples to determine soluble COD (CODsoJ. The suspended solids COD (CODss)and colloidal COD (CODcoJwere cal-

culatedby the differences betweenCODtot and COD fill' and between COD fill
and COD sol'respectively. All other analyses were performed three to five times per week or asdescribed previously (8,9).All gasmeasurements were recalculated to standard conditions (1 atm, O°C).Statistical analysis of data was performed using Microsoft Excel. Results and Discussion Performance of VASa Reactor Since the preliminary batch tests showed a quite high (-95%) anaerobic biodegradability of starch industry wastewater and the seedsludge was
Applied Biochemistry Biotechnology and
.~

Vol. 709,2003

, I...c;~

I :, I,';f~


,I.
e
a ,-

Treatment of Starch Industry Wastewater

257
-Int:-

A
6

-HRT-OLR

B
20
15 ~ 0

-. -Eft:-Rem
100
90
g

25
0
~ 20

e

~

~

"

~ ~

.

4

10 u

§

.

15

80

~
-'ii.

2
0 0 20 40
TJIre, days

5 0 60 80

~

00

c5

-

>

u 10 ~5 0 0 20 40 60 80

70 4) 60 ~ 50

~

:
.
;

TJIre, days

C
-Int: 8 7 :a,6 5 4 0 20 40 60 TJIre, days 80 .v,. ..
~

Eft: -CH4
"

D
4 3

-Int:
6

Prot . - . . Eft: Prot
Eft:N-NID
.:

-Int:N-NID

"'u'...

-~"\..~'

'~I

.

0,6 0,4 0,2 0

~

~4
.g ~2 0 0 20 40 60 TJIre,days 80

~
~ I Z

2 -=~ ~ lU 0

Fig. 2. Performance of UASB reactor. (A) HRT and OLR; (B) Influent and effluent total COD concentrations and removal efficiency; (C) Influent and effluent pH and specific methaneproduction; (D) Influent and effluent protein and ammonia concentrations.

sufficiently active (specific aceticlastic activity of 0.55 g of COD I [g of volatileSS.d]), the UASB reactor reached 90% of COD
tot

rem oval in a week (Fig. 2B).

Then the hydraulic retention time (HRT) was reduced to -2 d; however, the organic loading rate (OLR) decreased (d 15-46, Fig. 2A) owing to reduced strength of incoming wastewater (d 15-46, Fig. 2B). As a result, the COD tot removal was quite high (>90%) in this period (Fig. 2B). After a subsequent decreasein HRT to -1 d, the OLR reached values> 15 g of COD I (L.d) (d 47 onward, Fig. 2A), while the COD totremoval varied between 77 and 93% (d 47 onward, Fig. 2B). In spite of acidic effluents fed to the UASB reactor, the effluent pH was fairly stable, slightly oscillating around 7.9 throughout the entire study (Fig. 2C). This was owing to consumption of volatile fatty acids (data not shown) and production of ammonia as a result of protein degradation (Fig. 2D). The methane production rate was subject to significant fluctuations (Fig. 2C). This was related to entrapment and subsequent slow hydrolysis ofSS from incoming wastewater. In spite of the fact that the entrapped yellow aggregates were sometimes seen inside the sludge bed, no difficulties in the UASB reactor performance (e.g., sludge lifting as reported in refs. 7 and 8) were observed. The effluent COD IN ratio varied
Applied Biochemistry and Biotechnology -' Vol. 109,2003


.
j
.,

258

Sk/yar et a/.

between 4 and 5, i.e., lower than practically established values (usually around 6) to fulfill the requirements for subsequent denitrification (11). To avoid a deficiency of COD for nitrogen removal under planned implementation of the proposed technology, the COD removal efficiency of the UASB reactorcanbe decreased decreasingHRT or working temperature(9). by Performance of Activated Sludge Reactor An activated sludge reactor was started up in a nitrifying mode using the anaerobic effluents containing low concentrations of biodegradable COD (run Nl, Table 2). When the nitrification efficiency reached values around 80%, the reactor was switched on alternating (aerobic-anoxic) operation using anaerobic effluents having higher COD (run DNl). During this run, when aerobic and anoxic phaseslasted 40 and 20 min, respectively, the average HRT was 1.24 d while the average OLR was 2.28 g of caDI (L.d) (Table 2). The average CODtotremoval accounted for 83% with the CODtoteffluent concentrations oscillating at about 0.39 g of COD IL (run DNl, Table 2). However, becauseof the high ammonia loading rate (ALR)
applied (443 mg of N I [L.d], on average), the average efficiency of nitrifica-

tion was low (49%) (run DNl, Table 2). Together with a relatively low denitrification efficiency observed (60%, on average), this resulted in an average total nitrogen removal of 26% (run DNl, Table 7). To improve nitrogen removal, the HRT was increasedto 2.33d (on average) while duration of aerobic and anoxic phases was set as 30 min each during run DN2 (Table 2). Such an operational regime immediately led to almost complete denitrification (98%, on average); however, the average nitrification efficiency remained at a relatively low level (54%) (run DN2, Table 2). This can be related to the inhibiting pH values for nitrification observed in the reactor; they oscillated around 9.0 (runDN2, Table 2) owing to generation of alkalinity by denitrification and stripping of car In spite of these unfavorable conditions for nitrification, the average total nitrogen removal increased to 53% while the average CODtot removal slightly dropped (to 72%) during run DN2 compared with run DNI (Table 2). For balancing nitrification-denitrification processes,the HRT during run DN3 was further increased to 4.18 d (on average) while keeping the sameratio between aerobic and anoxic phases as in run DN2 (Table 2). This resulted in a significant improvement in nitrification efficiency (87%, on average) with almost no losses in denitrification efficiency (92%, on average), giving an average total nitrogen removal of 81% (run DN3, Table 2). Theeffluent CODtot concentrations oscillated at about 1 g/L, resulting in a further drop in COD tot removal to an averagevalue of 64%(run DN3, Table 2). It is likely that this remaining COD, which comprises about 5% of the initial strength of starch industry wastewater, is hardly biodegradable in either anaerobic or aerobic conditions. The observed efficiencies of nitrification, denitrification, and total nitrogen removal are plotted in Fig. 3 vs the imposed ALR. For the investigated activated sludge system (1.5-2.5 g of MLSS/L), a total nitrogen
Applied Biochemistry Biotechnology and
.'

,

Vol. 109,2003

I


Treatment Starch Industry Wastewater of
000\1r)t N Ir)\OO\ 0\0\-00 ~ONO-~~Ir)Ct) ~-_OOCt)O\t \0 oot Ct)Ct)O-OOO~Ir) Ct)Ct)~~~~Ir)~~ 1r) --Ct) N"" t
Ct) 00 0 OO 00 0\ 00t N t Ir) 1r)Ct)

259
OON"""""" 0\ 00--00 t 00 \ONOOIr)OO\O

--

"J Z 0

~~~1j~~11~~~~~11 ,., 00 Ct)
'!Ii 00 ~, ~ O\lr)lr)o\\oOONCt)Ct) Ct)ONO . . .. t

.. t

~
~

$ ~ u
..0 0
~ '"'~,
N ~ 00

Ct) N \0 00 ~ Ct)0\01r) NONO-~~Ir)\o\o- O\o\Ot
,..
~ L'

--0\ Ir)-Ct)

o
\O-Ir)--OOIr) ON-O\""~Ct)

---~Ct) t

---Ir) Ir) N

< 00
t 000\NIr)Ct)

ca ~

Z ~

~
L...

Oot Ct)~t Ct)Ct)~~~~OO~~Ir)t

oCt)t

~
NON
N Ir) N

NONI I I 0\ 0\ 0 000

~J
I
~ N Ir) N 00 Ir) \0

I O\~~N
00

I

I

I

1
I I I

°OO\\O"""" II

.op ca

0

\OOOOOIr)\oO~

0

0

~ 00

~ '"' E.-. '"'

0

u 2 ca

~00Ct)0\

~r

\0
N

00 \0 0\

NNNCt)

O\Ir)Ct)O\O
Ct) --~
Ct)
. .

""' "tj

~~ ~
~NNO-~OO~Ct)\O_Ct)--Ct)t

N"tj~
~

~
~
. . . . L'"

...

z
oo

~N
~~-b~J6-b~c1>J~J~~~~ Ct)N900~0\Ir)\OCt)\ONO\Ct)
O\\Ot
o~~o ~00Ct)Ct)

0\ ~ ~ N
Ir)
""

00 Ct) N -Ct) Ir) 00
0

---~-\ON-Ct) v' ,., ~ t 00 Ir) Ir) ~
0
.~

Ir)

- -- -\0 Ct) OO\o
""'

0\

~~
..o~ ca >

Nt

t

O\t

Ct)

N

a

.

E.-.".p u

<
Ct)Ct)1r)00 O\NNO 0 0 0 t 0 \0 0 ot
00 --~t N -.0

0

G
oo~-\OOO N \0 ~
'

-\0 __N
~ Ir)

g) \0 \0 -

~

x

Z . :r:o X
""

'""'i". ~o

u

;.+=

rTl

Z
0 0 Ct) 00

N 0\ 00
-

Ct) N N
\0 '" ,0 .

Ct) \0
.\oIr)NOO\

""'
~ t Ct)

_Ct)N
I

N -

-t 00

Ir) ~

-

~ ~~~ ""' ~ .

(/) '-. Qi Z .-';

Qi...~,z

]'"
(/)

~~~~~~~J~~1~ \0 \0 0\ 00 00
00 N N 0 0000 ~Iri""""""

~J~ N

N

'"'

J!; ~ S
ca

' '~ Z '""'i" ~ ~ Qi~..J" >,;I..i
'ti-Z
bO'""?"

~~~

QiOZ
"1j
.c

'"'
ca

~Z~

I

+

p.,

-

oZ~'ti~

~

ca ~ 0
'op

~
0

ca
'"'

O (l.

~

-

S '~ M

,S M ~

-"""""""""""0

-11
",~~

# -G
"""
.

~
:1 :1 >, :1

~""
Qi'

"tj""u~~"

10

Z :10000...00.""0"" ~ ~ u-o:::OOOO " S S.~ S;.+= M -~ ~ ""
t::. ~ 1--"t;i

00

00-0:::

~-

-- ~-""'"00> 22ca

~-

-:I :I

~

c,;

Z

...;>---

~~

° .-. ° :I: Z
~
ca ~
p.,

1
;13;

;'+='" ~ ca ~

~ M

~"O"" v'

0

"'000

>

~
"""",t1S-

(/)

c,;
'"iU .~

~~

~ u"tj ~~ M ..0.""- -00 ~ ~

o..~ u--

§OOS:I::I:Z

UUU~(l.(l.oozz".pze~p.,p., ca """",", ~
~

8 ~~ :I~$O~$..J~$ >6 E.-. ~ ;j ;j 0 ;j
~

2~

~

~

~

;j

Applied Biochemistry

,~§~..gS~~
M

~]]
bO$$

~

~

~U ~;.+=

~
~

;.+=

~
«~O~~U~~<~~Z~OE.-.~~
and Biotechnology

.b
;j ;j]

Z
;j.2"; $ ~o~$

~
Qi

""

~
~

t1S;j;j

~
;j;j

u
>'-"-'

U

Q'"

J0
Vol. .' 709, 2003


"

~~

~.
260 100 ~ 80 ~ g 60
Q)

Sk/yar et a/.

~
~

40 20 0 0 100 200 300 400 500 600 ALR, mg N-NH3/l/d -0Denitrification

Fig. 3. Efficiencies of nitrification, denitrification, and total nitrogen removal vs imposed ALR. 1 200
150

~ 0
8

0,8

~

06 ' 0,4 0,2

. .
0

100 S .; TotalCOD P-PO4

~
50
~

0 50 100
AddedFe (III), mg/l

0 150

Fig. 4. COD and phosphate concentrations aerobiceffluentsafter coagulation in with differentconcentrations iron (III). of removal >90% could be obtained only under relatively low ALR (about 100mg ofN-NH4/[L.d]). Under such conditions, the effluent total nitrogen concentrations were about 50 mg of N /L (run DN3, Table 2), Le., in the range for direct discharge into municipal wastewater treatment plants without additional charges for excessive nitrogen concentrations. The concentrations of phosphates slightly dropped during all nitrification-denitrification runs (Table 2), probably owing to partial precipitation on the activated sludge, taking into account that the pH values in the reactor were about or higher than 8.0 (Le.,favorable for hydroxyapatite and struvite precipitation) (9). Performance of Coagulation Step

Sincethe concentrations CODtot and phosphate remained relatively of high in the aerobic effluents (Table 2), efficiency of their coagulation with iron (III) was investigated (Fig. 4). The remaining COD and phosphate concentrations were adversely proportional to the iron concentration used. From the data in Fig. 4, one can calculate that 1.4 mg of COD and 0.5 mg of phosphate (as P) were removed by each milligram of iron added, respecApplied Biochemistry Biotechnology and .J Vol. 109,2003

II


Treatment

of Starch Industry 100 ~. 80
§ '.. ..

Wastewater 8CODtot 8Ntot DPtot

261

~ 60
11

~ 40
~ ] 20

~

0 Raw UASB ASR CoaglWoon

Fig. 5. Relative decrease in total COD, N, and P concentrations after each treatment step. tively. In addition, about 1.2 mg of NaOHI mg of iron should be added for pH correction to the optimal for the coagulation range (7.2-7.5). Conclusion Figure 5 summarizes the average data showing decreases in CODtot' nitrogen, and phosphorous concentrations for the raw starch industry wastewater after each treatment step in laboratory-scale investigations. About 85% of CODtot could be removed in a UASB reactor; however, it is essential to avoid too deep exhaustion of biodegradable COD in order to fulfill subsequent denitrification requirements. Up to 90% of total nitrogen could be removed from the anaerobic effluents using an activated sludge reactor operating in an aerobic-anoxic regime. Since the generated effluents still contained significant concentrations of hardly biodegradable COD (about 1 g/L) and phosphates (about 150 mg of P IL), a coagulation step with iron (III) could be applied for effluent clarification and P removal. Generally, the application of the three step treatment that we investigated could produce effluents approaching the limits established for discharge into a sewage treatment system, which is often an option for the starchproducing industry in Russia. If more stringent limits for effluent quality should be fulfilled, some posttreatment steps such as constructed wetlands can be recommended. Acknowledgment We gratefully acknowledge financial ing factory Ibred (Ryzan' province). References 1. Yevstigneyeva, Y. (2001), www.vedomosti.ru/stories/2001/03/20-44-06.html. 2. Landine, R. c., Brown, G. J., Cocci, A. A., and Viraraghavan, T. (1983),Agric. Wastes 7,111-123. 3. Christensen, D. R., Gerick, J. A., and Eblen, J. E. (1984), WPCF 56, 1059-1062. 4. Koster, I. W. and Lettinga, G. (1985), Biotechnol.Bioeng.27,1411-1417. 5. Nanninga, H. J. and Gottschal, J. C. (1986), Water Res.20, 97-103. Applied Biochemistry Biotechnology and .~ Vol. 709,2003 support from the starch-produc-


262

Sklyaret al.

6. van Wambeke,M., Grusenmeyer, Verstraete,W., and Longry, R. (1990), S., Process
Biochem.25, 181-186. 7. Lettinga, G. and Hulshoff Pol, L. W. (1992),AnaerobicReactorTechnology, Wageningen Agricultural University, Wageningen, The Netherlands. 8. Kalyuzhnyi, S. V., Estrada de los Santos, L., and Rodriguez-Martinez, J. (1998), Bioresour.Technol.66, 195-199. 9. Kalyuzhnyi, S., Sklyar, V., Epov, A., Archipchenko, I., Barboulina, I., Orlova, 0., and Klapwijk, A. (2001), Wat. Sci. Technol.45(12),79-87. 10. Klapwijk, A. and Rensink, J. H. (1996), Aerobic WastewaterTreatment, Lecture Notes, Department of Environmental Technology, Wageningen Agricultural University, The Netherlands. 11. Henze, M., Harremoes, P., la Cour Jansen,J., and Arvin, E. (1997), WastewaterTreatment: Biological and ChemicalProcesses, ed., Springer, Berlin. 2nd

Applied Biochemistry Biotechnology and .' I

Vol. 709,2003