Bioresource Technology 98 (2007) 2131–2136

pH regulation of alkaline wastewater with carbon dioxide:A case study of treatment of brewery wastewater in UASB reactor

coupled with absorber

A. Gangagni Rao ¤, T. Sasi Kanth Reddy, S. Surya Prakash,J. Vanajakshi, Johny Joseph, P.N. Sarma

Bioengineering and Environmental Centre, Indian Institute of Chemical Technology, Tarnaka, Hyderabad 500 007, Andhra Pradesh, India

Received 2 June 2006; received in revised form 22 August 2006; accepted 23 August 2006Available online 12 October 2006

Abstract

Studies were carried out with carbon dioxide absorber (CA) to evaluate the usage of carbon dioxide (CO2) in the biogas as an acidify-ing agent by Up-Xow Anaerobic Sludge Blanket (UASB) reactor. Investigation on the 5 l absorber revealed that ratio of brewery waste-water (BW) Xow rate to biogas Xow rate of 4.6–5.2 was optimum for minimum consumption of CO2 for acidiWcation. The acidiWed BWafter the absorber was treated in UASB reactor with optimum organic loading rate (OLR) of 23.1 kg COD/m3/day and hydraulic reten-tion time (HRT) of 2 h. UASB reactor exhibited good performance with respect to reduction of chemical oxygen demand (COD) andmethane yield. The implications of the present study on the full scale anaerobic reactor of medium scale brewery revealed that suYcientcost savings could be made if CO2 in the biogas or CO2 that was being wasted (let out to the atmosphere) can be used instead of sulfuricacid (H2SO4) for pH control.© 2006 Elsevier Ltd. All rights reserved.

Keywords: Acidifying agent; Carbon dioxide; Brewery wastewater; Biogas; HRT; OLR; UASB; Absorber

1. Introduction

The brewing process employs a number of batch-typeoperations in processing raw materials to the Wnal beerproduct and produces large quantities of wastewater, typi-cally 3–12 l of wastewater for every litre of beer produced(Luc Fillaudeau et al., 2006). The quality of brewery eZuentdepends on various processes that take place within thebrewery (Driessen and Vereijken, 2003). In general, breweryeZuents are easily biodegradable (BOD/COD ratio is in therange of 0.6–0.7) and amenable for anaerobic treatment(Leal et al., 1998; Parawira et al., 2005; Driessen and Vereij-ken, 2003; Hanqing Yul and Guowei Gu, 1996; Etheridgeand LeroV, 1994). Anaerobic treatment is applied success-

* Corresponding author. Tel.: +91 40 27160123x2664; fax: +91 4027193159.

E-mail address: gangagnirao@yahoo.com (A.G. Rao).

0960-8524/$ - see front matter © 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.biortech.2006.08.011

fully for the treatment of brewery eZuent in laboratory,pilot and full scale with a variety of reactor conWgurations(Parawira et al., 2005; Speece et al., 2001; Hanqing Yul andGuowei Gu, 1996; Etheridge and LeroV, 1994). The pH ofbrewery eZuent varies in the wide range of 4.5–12 (KarlOckert, 2002; Charles, 2001). The wastewater obtained frombrewing operation is acidic whereas the wastewaterobtained from caustic operation is alkaline (Briggs et al.,2004; Karl Ockert, 2002). Most of the breweries providebuVer tank (pH Equalization tank) before the anaerobicreactor so that wastewater generated in diVerent operationsof the brewery can be made to uniform pH. Even after mix-ing, pH of the combined wastewater in the buVer is alkalinesince two thirds of the wastewater from the brewery is alka-line (Luc Fillaudeau et al., 2006). Therefore, acids (sulfuricacid or hydrochloric acid) are used to maintain the pHin the range of 7–7.5 for feeding to the anaerobic reactor(Cronin and Lo, 1998). The addition of acids led to the

2132 A.G. Rao et al. / Bioresource Technology 98 (2007) 2131–2136

formation of sulWde and added to the cost of the eZuenttreatment operation. The reactor performance deteriorates,if the sulWde concentration is in the range of 200–250 mg/l(Gangagni Rao et al., 2003; Sawyer and McCarty, 2000) dueto the inhibition of the mixed cultures of anaerobic micro-bial consortia. Carbon dioxide (CO2) is abundantly pro-duced in breweries during fermentation and it is used in theWnal stage of beer production for Xavour and excess isvented to the atmosphere (David Willison, 2006). If this canbe utilized in the buVer tank to reduce the pH of wastewaterprior to the anaerobic reactor, the cost of the acids and neg-ative aVects of sulWde can be avoided. Alternatively, if thebrewery is having anaerobic eZuent treatment plant (ETP),biogas generated in the anaerobic reactor can also be usedfor pH reduction. In one study CO2 (Lom, 1977) is used forneutralization of alkaline brewery wastewater. However,there are no reports in the literature regarding the usage ofCO2 as an acidifying agent in the buVer tank prior to theanaerobic step. Hence, in the present study, the pH of thebrewery wastewater was regulated with CO2 and anaerobictreatment of the same was studied in the UASB reactor.

2. Methods

2.1. Wastewater

The brewery wastewater (BW) was obtained (approxi-mately 200–300 l per week) from United Breweries (UB),Hyderabad, India. The BW was collected from buVer tankin which pH was in the range of 9–12. The physicochemicalcharacteristics of the BW were determined and shown inTable 1. All the chemicals used during the experiments wereof A R grade.

2.2. Inoculum

Sludge obtained from an anaerobic lagoon was used asthe inoculum for seeding the reactor. Volatile suspended

solids (VSS) of the sludge were 27.66 kg/m3 and the metha-nogenic activity was 0.21 kg CH4-COD/kg VSS/day. Thesludge was acclimatized with BW for three weeks underanaerobic conditions as a result of which, the VSS of thesludge increased to 33.63 kg/m3 and the methanogenicactivity rose to 0.26 kg CH4-COD/kg VSS/day.

2.3. Analysis

The physicochemical characteristics of the wastewaterwere determined as per standard methods (APHA, 1998).The characteristics of the inlet and outlet viz., COD, pH,alkalinity and volatile fatty acids were determined dailyduring the reactor operation. The volume of the biogas pro-duced and its composition were estimated using Orsatapparatus (APHA, 1998). The washout of the microbialcells in the reactor outlet was estimated following the pro-cedure of Gangagni Rao and Bapat (2006).

2.4. Experimental set up

The experimental set up (Fig. 1) consisted of peristalticpump, centrifugal pump, feed tank, absorber, buVer tank,UASB reactor and biogas holder. Absorber was made of

Table 1Characteristics of brewery wastewater

a All parameters except pH and temperature are in mg/l.

Parametera Brewery wastewater

pH 3–12Temperature 18–40 °CCOD 2000–6000BOD 1200–3600VFA 1000–2500Phosphates as PO4 10–50TKN 25–80TS 5100–8750TSS 2901–3000TDS 2020–5940

Fig. 1. Experimental set-up: (1) Biogas holder, (2) absorber, (3) feed tank, (4) centrifugal pump (5) methane gas holder (6) buVer tank, (7) peristaltic pumpand (8) UASB reactor.

6

43

2

1

5

78

Sludge Blanket

Sludge Bed

GLS Separator

A.G. Rao et al. / Bioresource Technology 98 (2007) 2131–2136 2133

5-l Plexiglas (Perspex glass) column with arrangements forcountercurrent Xow of biogas and BW. Centrifugal pumpwas used for spraying the BW from the top of the absorberand biogas was passed from the bottom. The UASB reactorwas designed with a total volume of 25 l with a liquidretaining capacity of 22.5 l. The reactor was made of Plexi-glas and was insulated to maintain the temperature. Thefeed to the UASB was pumped by means of a peristalticpump and its temperature was always maintained at37§ 2 °C. Provision was made for measuring liquid temper-ature at the reactor outlet. This was found to be always inthe range of 33–37 °C. The wastewater was introduced atthe bottom of the reactor via an inlet distribution networkand the gas, solid and liquid phases were separated at thetop by means of a three-phase separator. The gas producedwas measured using a wet gas Xow meter.

2.5. Experiments

2.5.1. AbsorberThe BW from the feed tank having pH in the range of 9–

12 was sprayed from the top of the absorber using centrifu-gal pump. Biogas containing CO2 in the range of 30–40%from the biogas holder was counter currently passed fromthe bottom of the absorber. Enriched biogas (in terms ofmethane) after the absorber was separately collected in aholder. The absorber was operated at diVerent biogas Xowrates (G) from 100 to 160 l/h. At each biogas Xow rate (G),BW Xow rate (L) was varied in the range of 100–1000 l/h.

2.5.2. UASB coupled with absorberThe BW from the absorber, the pH of which was in the

range of 7–7.5 was collected in the buVer tank. The BWfrom the buVer tank was pumped from the bottom of the

UASB reactor after inoculating the reactor with the accli-matized sludge. The UASB reactor was started as per thestart-up protocols (Gangagni Rao and Bapat, 2006) andthe OLR was increased in a stepwise manner from 1 to46.2 kg COD/m3/day to determine the maximum OLRachievable at optimum HRT.

3. Results and discussion

3.1. Optimization of absorber

The variation of CO2 consumption and outlet pH withrespect to the BW Xow rate at each biogas Xow rate wereplotted and shown in Fig. 2. It could be observed that at allbiogas Xow rates under study, outlet pH of the absorberincreases and CO2 consumption decreases gradually withincrease in BW Xow rate. Fig. 2 also shows that during thevariation of biogas Xow rate from 100 to 160 l/h, absorberoutlet pH increased in the range of 6.5–9.2 and CO2 con-sumption decreased in the range of 32.8–13 l during theincrease of BW Xow rate from 100 to 1000 l/h. Fig. 2 alsorevealed that CO2 consumption and increase in outlet pHpattern were proportionate to the BW and biogas Xowrates. In order to operate the absorber eYciently at a partic-ular BW Xow rate, it was essential to determine the opti-mum biogas Xow rate and CO2 consumption at desiredoutlet pH of absorber. It is known that UASB reactor needsto be fed at inlet pH in the range of 7–7.5 (Gangagni Raoand Bapat, 2006) to get the optimal performance. Therefore,an optimal pH of 7.3 was selected as inlet pH of the BW tothe UASB reactor. Accordingly, absorber outlet pH of 7.3was selected to optimize the CO2 consumption and BW Xowrate at each biogas Xow rate under study. Using the polyno-mial regression analysis, pH and CO2 consumption curves

Fig. 2. Variation of pH and CO2 consumption with BW Xow rate.

2134 A.G. Rao et al. / Bioresource Technology 98 (2007) 2131–2136

in Fig. 2 were Wtted and these equations were as shown inTable 2. Using the pH polynomial equation, at absorberoutlet pH of 7.3, optimum BW Xow rate was determinedand at this optimum Xow rate of BW, corresponding opti-mum CO2 consumption was determined using the polyno-mial equation of CO2 consumption. The optimum BW Xowrate and CO2 consumption were determined for all biogasXow rates under study using the aforesaid procedure andwere tabulated in Table 2. It was observed that at biogasXow rates of 100, 110, 120, 130, 140, 150 and 160 l/h, opti-mum BW Xow rates were 506, 578, 620, 640, 663, 675 and734 l/h and CO2 consumption rates were 20.0, 20.3, 20.8,20.9, 21.0, 21.9 and 22.2 l, respectively. The results revealedthat minimum CO2 consumption could be obtained whenthe absorber operated in the optimum L/G ratio of 4.5–5.3.

3.2. Performance of UASB reactor

As stated earlier, UASB reactor was started as per proto-col. Excellent granular sludge was formed during the opera-tion of the reactor due to the inoculum quality, systematicand stepwise increase of organic and hydraulic loading ofthe reactor at optimal pH and temperature. Previous studies

(Wen-Tso Liua et al., 2002) also showed that granularsludge could be formed with brewery wastewater. Theabsorber was coupled to the UASB reactor and it was oper-ated at optimum L/G ratio in the range of 4.5–5.3 as deter-mined previously. COD reduction and methane yield of theUASB reactor at diVerent organic loading rates (OLR) wereplotted (Fig. 3). Evidently COD reduction was in the rangeof 92–96% and methane yield was in the range of 0.27–0.32 m3CH4/(kg COD reduced) during the variation of OLRin the range of 1.9–23.1 kg COD/m3/day. However, the CODreduction was observed to be 65–55% and methane yieldwas 0.15–0.12 m3CH4/(kg COD reduced), when the OLRwas increased in the range of 29.6–46.2 kg COD/m3/day.

Inlet pH, outlet pH and hydraulic residence time (HRT)of the UASB reactor at diVerent OLR are shown in Fig. 4.Outlet pH was maintained in the range of 7.7–7.4 and HRTwas in the range of 24–2 h, when OLR steadily increasedfrom 1.9 to 23.1 kg COD/m3/day. However, when OLR wasincreased from 29.6 to 46.2 kg COD/m3/day, HRT was inthe range of 1.5–1 h and pH came down from the previousrange of 7.7–7.4 to a range of 6.6–6.4. This was observedeven though inlet pH was maintained constantly in therange of 7.1–7.5 during the entire course of operation of the

Table 2Absorber optimization data

L D BW Xow rate, G D Biogas Xow rate, pH D Outlet pH of the absorber, C D CO2 consumption.a L was determined at outlet pH of 7.3 by using pH polynomial equation.

G (l/h) pH polynomial equation La (l/h) CO2 consumption polynomial equation CO2 consumption(l/h)

L/Gratio

100 pH D 0.0212L2 + 0.0576L + 6.4667 506 C D¡0.0045L3 + 0.1078L2 ¡ 2.2665L + 29.267 20.0 5.1110 pH D 0.0193L2 + 0.0354L + 6.4517 578 C D¡0.0076L4 + 0.1736L3 ¡ 1.261L2 + 1.7078L + 27.475 20.3 5.3120 pH D 0.0223L2 ¡ 0.024L + 6.5917 620 C D 0.0007L3 + 0.0385L2 ¡ 1.835L + 30.48 20.8 5.2130 pH D 0.0117L2 + 0.0484L + 6.5117 640 C D 0.0055L5 ¡ 0.1623L4 + 1.7759L3 ¡ 8.6539L2 + 16.121L + 19.92 20.9 4.9140 pH D 0.0212L2 ¡ 0.0333L + 6.5667 663 C D¡0.0041L3 + 0.1575L2 ¡ 2.8082L + 33.927 21.0 4.7150 pH D 0.0148L2 + 0.012L + 6.545 675 C D¡0.0112L4 + 0.2576L3 ¡ 1.9301L2 + 3.5875L + 29.692 21.9 4.5160 pH D 0.0197L2 ¡ 0.0664L + 6.7267 734 C D 0.0054L3 ¡ 0.0382L2 ¡ 1.777L + 35.12 22.2 4.6130 Average values 631 Average values 21 4.9

Fig. 3. Performance of the UASB reactor.

A.G. Rao et al. / Bioresource Technology 98 (2007) 2131–2136 2135

reactor (from 1.9 to 46.2 kg COD/m3/day OLR). The aboveresults revealed that beyond the OLR of 23.1 kg COD/m3/day and HRT of 2 h, the reactor performance deterioratedin terms of COD reduction, pH and methane yield (Figs. 3and 4). The reactor could be operated at this optimizedhigh OLR of 23.1 kg COD/m3/day and low HRT of 2 h dueto the formation of excellent granular sludge during thestart up period. The reactor up Xow velocity was main-tained in the range of 0.04–0.5 m/h during the operation ofthe reactor at the OLR and HRT in the range of 1.9–23.1 kgCOD/m3/day and 24–2 h, respectively. It was in the range of0.65–1.0 m/h when the reactor was operated at OLR andHRT in the range of 29.5–46.2 kg COD/m3/day and 1.5–1 h,respectively. In both cases washout of granular biomasswas not observed due to the good settling characteristics ofthe granular sludge, even though performance of the reac-tor deteriorated beyond the OLR of 23.1 kg COD/m3/day.In previous studies on opaque beer wastewater with UASB,57% COD reduction was obtained at an HRT of 24 h(Parawira et al., 2005). Similarly, studies of Cronin and Lo(1998) and Driessen and Vereijken (2003) on UASB withbrewery wastewater showed that COD reduction of 91%with the HRT in the range of 12–36 h and COD reductionin the range of 75–80% with OLR in the range of 5–15 kgCOD/m3/day was possible. In the present study UASBreactor could be optimally operated at an OLR of 23.1 kgCOD/m3/day and HRT of 2 h with COD reduction of 96%and methane yield of 0.32 m3CH4/(kg COD reduced).

3.3. Application of present results to full-scale plant and cost implications

During the fermentation to produce 1080 kL of beerapproximately 1100 m3/day of CO2 was generated and outof this approximately 300–500 m3/day of the CO2 was

absorbed in beer storage kettle in order to maintain theCO2 in the range of 2.8–3 v/v to enhance the Xavour ofbeer. The balance amount of CO2 was let oV to the atmo-sphere. The brewery produces 420 m3 of wastewater perday. In this, approximately 280 m3/day of wastewater wasalkaline having pH in the range of 9–12. The brewery hadeZuent treatment plant based on anaerobic (anaerobichybrid reactor) followed by aerobic (activated sludge pro-cess). The full-scale anaerobic reactor at the factory wasgenerating 350 m3 of biogas having 35% CO2. The brew-ery was using approximately 3000 l of 98% commercialgrade H2SO4 per month. It was determined in theabsorber optimization studies that average optimum L/Gratio was 4.9 (Table 2). Therefore, approximately 60 m3 /day of biogas was required for neutralizing the BW. Inour experiments, it was observed that the CO2 in the inletbiogas to the absorber was approximately 35% and 20%in the outlet. Therefore, 15% of the CO2 was consumed.Accordingly, for the case of factory, minimum require-ment of CO2 was approximately 9 m3/day. Since CO2 wasavailable at no cost in the brewery plant, the direct savingwas the cost of H2SO4. In some cases where the CO2 mightnot be available within the plant, biogas generated in theanaerobic process could be used and if it was insuYcient,feasibility of buying extra CO2 should be analyzed sepa-rately.

4. Conclusions

The studies showed that CO2 could be used as a cheapand acidifying agent for decreasing the pH of the alkalinewastewaters before the anaerobic reactor, replacing theconventionally used acids. The studies also revealed that itwas economical to implement this idea at full-scale plant ifthe CO2 was available at source with no cost, which was the

Fig. 4. Optimization HRT of the UASB reactor.

2136 A.G. Rao et al. / Bioresource Technology 98 (2007) 2131–2136

case with brewery. CO2 absorbed was economized by opti-mizing the L/G ratio.

Acknowledgements

The authors wish to thank the management of M/sUnited Breweries at Hyderabad for the supply of wastewa-ter and giving data regarding their existing eZuent treat-ment plant. The authors also wish thank Director, IICT forhis encouragement during the course of work.

References

APHA, 1998. Standard Methods for the Examination of Water andWastewater, 20th ed. APHA, AWWA, WEF, Washington, DC.

Briggs, D.E., Boulton, C.A., Brookes, P.A., Stevens, R., 2004. Brewing: Sci-ence and Practice. Woodhead Publishing/CRC Press, Cambridge, UK/Boca Raton, Florida, USA. p. 70.

Charles Bam forth, W., 2001. pH in brewing: an overview. MBAA TQ 38(1), 1–9.

Cronin, C., Lo, K.V., 1998. Anaerobic treatment of brewery wastewaterusing UASB reactors seeded with activated sludge. Bioresour. Technol.64, 33–38.

David Willison, 2006. Cost EYciencies in the Brewing Business. http://www.oceta.on.ca/documents/sleeman_fnl.pdf.

Driessen, W., Vereijken, T., 2003. Recent developments in biological treat-ment of brewery eZuent. Paques March 2–7. http://www.environmen-tal-expert.com/resultarticlept.asp?idcategory D 299&idmainpage D 70&idproducttype D 6&levelD 2.

Etheridge, S.P., LeroV, U.E.A., 1994. Anaerobic digestion – a viable optionfor industrial eZuent treatment. MBAA TQ 31 (4), 138–141.

Gangagni Rao, A., Bapat, A.N., 2006. Anaerobic treatment of pre-hydro-lysate liquor (PHL) from a rayon grade pulp mill: pilot and full-scaleexperience with UASB reactors. Bioresour. Technol. 97, 2311–2320.

Gangagni Rao, A., Krishna Prasad, K., Venkata Naidu, G., Chandrashe-kar Rao, N., Sarma, P.N., 2003. Removal of sulWde in integrated anaer-obic–aerobic wastewater treatment system. Clean Technol. Environ.Policy 6, 66–71.

Hanqing Yul, Guowei Gu, 1996. Biomethanation of brewery wastewaterusing an anaerobic upXow blanket Wlter. J. Clean. Prod. 4 (3–4), 219–223.

Karl Ockert, 2002. Practical Wastewater Pretreatment Strategies for SmallBreweries. MBAA TQ 39 (1), 39–46.

Leal, K., Chacin, E., Behling, E., Gutierez, E., Fernandeza, N., Forstef,C.F., 1998. A mesophilic digestion of brewery wastewater in anunheated anaerobic Wlter. Bioresour. Technol. 65, 51–55.

Lom, T., 1977. A new trend in the treatment of alkaline brewery eZuents.MBAA TQ 14 (1), 50–58.

Luc, Fillaudeau, Pascal, Blanpain-Avet, Georges, DauWn, 2006. Water,wastewater and waste management in brewing industries. J. Clean.Prod. 14, 463–471.

Parawira, W., Kudita, I., Nyandoroh, M.G., Zvauya, R., 2005. A study ofindustrial anaerobic treatment of opaque beer brewery wastewater in atropical climate using a full-scale UASB reactor seeded with activatedsludge. Process Biochem. 40, 593–599.

Sawyer, N.C., McCarty, L.P., 2000. Chemistry for Environmental Engi-neering, fourth ed. TATA McGraw-Hill, New Delhi.

Speece, R.E., Ahn, Y.H., Min, K.S., 2001. Full scale UASB reactor perfor-mance in the brewery industry. Environ. Technol. 22 (4), 463–476.

Wen-Tso, Liua, On-Chim, Chanc, Fang, H.P. Herbert, 2002. Characteriza-tion of microbial community in granular sludge treating brewerywastewater. Water Res. 36, 1767–1775.