8/13/2019 010025223

1/14

._ . .

War. Sci .Tech. Vol. 25 No. I.pp.223-235. 1992.Rioted in GreatBritain.Au rightsreserved. 0273-1223EJ2 15QOCopyrightO 1992IAWPRC

ANAEROBIC TREATMENT OFAPETROCHEMICAL WASTEWATER FROMA TEREPHTHALIC ACID PLANTH.Macarie , A. Noyola** and J.P. Guyot****Universidad Autonoma Metropolitana, Departamento de Biotecnologia,A. P. 55-535, Iztapalapa, 09340 Mkxico D.F,, Mkxico. Present address: Laboratoirede microbiologie ORSTOM Universitk de Provence, 3 Place Victor Hugo,13331 Marseille, Cedex 3, France**Corresponding author: Instituto de Ingenieria UNAM, A. P. 70-472, Coyoacan,04510 Mtkico D.F., Mdxico***Mission ORSTOM, Homero 1804-1002, losMorales 11510, M hic o D. F., Mkxico

AAnaerobic treatment of te reph thalic acid plant wastewater was tested using twoUASB reactors (T and U) nd adownflow tubular fixed film reactor. UASB T was inoculated with sludge sampled from an anaerobicstabilization pond receiving waste activated sludge from a petrochemical industry treatment plant. UASB U andthe f i. @ film reactor were inoculated with anaerobically adapted activated sludge from a municipal plant. Raweffluent had to be se ttled and neutralized before reactor feeding. Sedimentation resulted in 70% TSS and 37%COD removal. UASB digeste rs presented comparable treatment efficiencies with rather low COD removals: thebest results were 46.4 for UASB T at 2.6 kg COD/m3.d and a hydraulic retention time O) of 2.7 days and43.9% for UASB U at 2.2 k g COD/m3.d and e of 3.2 days. The perform ance of the tubular reactor was muchhigher, 74.5% COD removal at 1.89 kdm3.d and eof 3.4 days. The better efficiencies of this last digester areexpla ined mainly by a higher VSS content and a better resistance to toxicity caused by the arom atics present inthe wastewater. A prim ary settling-anaerobic-aerobic process is proposed as an alte rnative to the conventionalaerobic process for treating terephthalic wastewater, but disposal of solids from primary & men ta tio n and costof neutralization have to be considered before application.

KEYWORDSAromatic compounds, Terephthalic acid; p-Toluic acid; UASB; Fixed film reactor, Anaerobic inoculum.

INTRODUCTIONPolymer-grade terephthalic acid (1,4-benzenedicarbxylic acid) is among the top 50chemicals manufactured inthe United States (C&EN, 19 88; Webber, 1984). In 1990, its roduction in the world was estimated at 6.52producers are Japan, Taiwan, South Korea and the EEC (Savostianoff, 1990). In Latin America, purifiedterephthalic acid (PTA) is made in Mexico and Brazil, and represents around 6 of world production. Thisproportion will increase due to the augmentation of plant capacities and the construction of a new fac tory inColombia. Polyester textile fibers are he principal outlet for PTA. Other important end products are polyesterresins used to make bottles for carbonated drinks (PET bottles), and polyester films which have manyapplications in the audio-visual, photographic, computer, packaging and decoratives fields. Smaller amountsareused to make technica l plastics (Savostianoff, 1990).The A merican Amo co process is the established technology for polymer-grade terephthalic acid manufacture(75%of world production). It consists of a liquid-phase air oxidation of p-xylene at 173-230 OC and 1.5-3.0

106 ton/year and should increase in the future to about 9 10gton/year. A part from the USA, the principal

223

- -

8/13/2019 010025223

2/14

8/13/2019 010025223

3/14

Anaerobicpetrochemical wastewater Ireatmerit 225with an internal diameter of 9.6 cm was packed using67 cm high, 1.27 cm diameter PVC tubes. The tubesprovided a spec ific area of 221 m2/m3. The digesters were placed in a controlled temperature m m t 33f OCand were con tinuously fed with peristaltic pumps, in an upflow mode fo r the UASB and a downflow mode forthe fured film reactor. Biogas was evacuated at the top of the colum ns and directed to a gasmeter device filledwith an acidified brine (pHc4) to avoid CO2 dissolution.

GAS HETERCOLUMH

Ul SB REACTOR

IHFLUEHT

GIIS NETER

TUBULAR F I X E D F I M REACTOR

Fig. 1. Schematic diagram of the labo ratory reactors.Inoculum,2 sludg es from different sources were used as seed in the case of the U ASB diges ters in order to assess anylimitations owing o the kind of inocuhm . The seed for UASB T was obtained from an anaerobic stabilizationpond receiv ing waste sludge from the aerobic wastewater treatment syste m of a PTA plant. The seed for reactorU was an anaerobically adapted activated sludge from the conventional wastewater treatment plant of t4eNational University campus at Mexico City. The fixed film reactor had been used during a previous experiment(Monroy et al., 1988) and had already developed a biofilm on its tubular packing from the same kind ofinoculum used for UASB U.Wastewater Characteristics.The effluent employed in th is study was sampled on a weekly basis at the entranceof the first aeration basin ofa PTA production wastew ater treatment plant. The effluen t had already passed through an equalization basinwith seve ral days hyd raulic retention time and had been supplem ented with nitrogen and phosphorus. Duringthe experim ent, the w astewater was kept at 6OC. Raw efflue nt analys is is given in Tab le 1 ogetherwt themain arom atic compounds found in the mother liquor of the purification unit, which is o ne of the majorwastew ater constituents. The raw wastew ater had a high con tent of fast settling suspended solids, so it wasdecided to feed the reactors with primary settled wastewater. Sedimentation was performed directly in theconta iners used fo r effluent transport at the temperature of storage. The pH of decanted wastew ater was thenadju sted to 6.15 (standard deviation (s)=0.22); an average of 3.15 g ( s= l) in dustrial N a m 3 was necessaryfor one liter of effluent. A characterization of the settled, neutralized influent is presented in Tab le 2. The valuescorrespond to an ave rage calculated over the 7 nd half months of the experiment.

8/13/2019 010025223

4/14

226 H.MACARIEet aLTABLE Characteristics of Raw Wastewateq.

Parameter ConcentrationRaw wastewaterPH 4.5

OD^COD 9500 mgn5500m a2200 mgflSSAcetic acid 500-loo0mg/lTerephthalic acid 2670 mg/lp-toluic acid 480 mgnBenzoic acid 354 mgA

Mother liquor (*)

4fo rmy lbenzo ic acid "g/l(*) Supernatant after 10min sedimentation.

TABLE2 Ch-tcs of Neutra lized. Sett ed In-Parameter Concentxition

TotalCOD 6477 75 941Soluble COD 5958 5 836TS 6644 31 1674TVS 3459 31 1551TSS 704 32 413vss 406 30 287Alkalinity (as caC O3 ) 1777 62 619N-NH4+ 93.2 27 39.3

(*) n = number of samples; s= standard deviation.

s

Start-UD and Omration.Each UASB received 1 litre of wastewater, litre of tap water and litre of sludge. Feeding with rawwastewater resulted in pump c loggin g problems, socontinuous operation could only begin one month later,when primary settled w astewater was used. Then, 3 hydraulic retention times e) were testcl for UASBdigesters: 7 days (day 0-108), 3 days (day 109-134), and 2 days (day 135-194). Th e fixed film reactor wasdrained of its previous contents and filled with settled wastewater and tap water in equal proportions.Continuous feeding started imm ediately, with a e of 10 days in orde r to initiate sludg e acclimatization. Tw oother e were applied, 5.8 days from day 21 to 87 and 3.4 days from day 88 today 164. An additionale of 2.9days was maintained or the two ast weeks.

Total and solubleCOD; Biochemica l Oxygen Demand (BODS); Total, Fixed and Volatile Solids TS, TFS,TVS); Total, Fixed and V olatile Suspended Solids (TSS, FSS, VSS); p H alkalinity topH 4.3 and ammoniumnitrogen were determined as indica ted in StandardMethods (APHA, 1980).TotalCOD, pH and alkalinity weremeasured twice a week. Soluble COD, TS, TFS, TVS, TSS, FSS, VSS, N-NH4+ were measured once aweek. BOD5 was analysed 3 times during steady state periods. The atomic absorption spectophotometer

. .... ccLc

8/13/2019 010025223

5/14

Anaerobic petrochemical wastewater treatment 227(Varian spect AA-20 ) was used to determ ine heavy metals and Na, K, Mg and Ca. The samples for atomicabsorption spectxophotometry were prepared according to Standard Methods (APHA, 1980). The SludgeVolume Index (SVI) was adapted from the usual technique (APHA, 1980): in a 100 m est tube, 25 m freactor sludge were diluted with 7 5 m f settled effluent; after mixing, the 30 min settled sludge volume wasdetermined and the SVI was ob tained dividing this value by the TSS content of the 2 5 ml sludge. A maximumsettling velocity was calculated with the steepest slope of the settling curve obtained during SV I determination.With the experimental conditions employed thisvelocity corresponds to a hindered sett ling velocity (Viii). Thesize of granules was estimated with the method described by Mahoney et al. (1987). Counts of anaerobicbacteria were performed using the Most Probable Number (MPN) technique with 5 ubes per dilution (Guyot eral. 1990 a). Cultivation media and the inoculation technique s were that of Hungate (1969) and Balch et al.(1979). Gas production was obtained by w ater displacement and methane content was measured by g aschromatography using a thermal conductivity detector (Noyola et al. 1988).

RESULTS AND DISCUSSIONEffect of Decantation on Effluent Characteristics.Unde r the primary settling conditions (pH 4.5,6"C, batch mode), TSS and COD removal efficiencies were70% and 37% respectively (Tables 1and 2). The solubility of terephthalic acid in water (19 mgil, 25OC) and itsoperation. Xu et al. (1988) show ed that terephthalic acid (pK1 = 3.51; pK2 = 4.46) can be separated from aPTA alkaline wastewater by acidification-flocculation -sedimentation . ctually, precipitation of PTA starts atpH 5.1, is almost complete at pH 4.5 and is entirely finished at pH 3.8. When applied to a PTA wastewatercontaining an amoun t of tereph thalic acid similar to our case, the acid ification @H

8/13/2019 010025223

6/14

228 H. MACARIE er aLIn order to aIlow fo r sludge adaptation, the sam e 6 was maintained till day 108,but noimprovem ent could beobtained u a b le 3). Another explanation for the low COD removal efficienc ies could be a deficiency in sometrace metals necessary fo r methanogenic bacteria as it is often the case with industrial effluents (S peece, 1983).To test this hypothesis, from day 82 to day 122, a mixture of metals (Table 4) was added to the digesterinfluent, but no amelioration was observed. T he period of test, correspo nding to 8 e values, was consideredlong enough. As a consequence, the metal addition was stopped. Analysis of trace nutrients was performed ona sample of settled wastewater. The results presented in Table 5 confirmed that there was nolimitation due tothese compounds. The particular ly high conten t of cobalt and manganese is exp lained by the use of Co and Mnacetate ascatalysts in the process of terephthalic acid production.

TABLE 4 Commsition of Trace Metal Solution Added to the Wastewater During Davs 82 to 122 (*).compound concentrationin the cocktail

gfl1111000.20.0510

concentration in thewastewatermgfl0.48 (as Zn)0.28 (as h4n %0.25 (as Co)11.96 (as Mg)0.05 (as Ni)

0.019 (as Cu)5:4 (as Mo)*) Each com pound was dissolved in a separate solutionand 1m f each solution was added to 1 itre of influent.

TABLE 5 Inorganic Composition of Settled Terephthalic Acid Plant Wastewater.Compound Na Mn K Co Fe(* ) Ca Pb Zn Mo Cu Cd Mg Ni

Concentration 430 75.6 60 '56.6 13.3 12.4 11.3 4.76 4.66 3.33 1 0.86 0.26(*) Includes all oxidation states.mgfl

Over this first period of operation (day 30 to 108), UASB T presented a bette r behaviour than UASB U higherCOD emoval, biogas production and stability. The superiority of reactor T d uring the first months of operationmay be explained by the origin of the inoculum , a higher quantity of seed sludge and better microbial andsedimentation characteristics (Table 6). At lower 6 (3 and 2 days), UASB reactors achieved better CODremoval efficiencies than at 7 days. However the improvement was limited on ly to 46% in the best case forUASB T at a e of 2.7 days. It seems that highe r upflow velocities and gas production rates led to better sludgebed mixing and improved mass tranfer, sobetter COD removal efficiencies were obtained at lower 0or higherorganic loads. T he differ ence in UASB perform ance, observed during the firs t period, tended to disappear assimilar removal e fficienc ies were achieved for the rest of the study. This behaviour was the result of sludgeevolution, as may be noticed in Table 6. At the end of the experim ental work, s ludges U and T presented goodsimilar settling characteristics (SVI, Viii), but their MPN counts were different: sludge U maintained itsremained at the sam e level, and propionate and butyrate users decreased. In addition, granule size was verysmall for both reactors indicating that PTA wastewater does not favour granu le formation. Results in Table 6show that at the end of the study, sludge U had better microbial chara cteristics than sludge T, which wassupposed adapted to some extent to the w astewater. As a consequence, aerobic sludge from municipaltreatment plants, once it ha s been adapted anaerobica lly, may be considered as a good inoculum alternative

'hydrogenoph ilic methanogen counts wh ile they decreased in sludge T; acetoc lastic methanogenic bacteria L

8/13/2019 010025223

7/14

Anaerobic petrochemical wastewater treatment 229when granular sludge or anaerobic digested sludge are not available. A sim ilar observation was made by Wu etal. (1987) with different industrial wastewaters.

80 -2 60-8A 40 -n8 20-O

r

m UASBTUASBUA Fixed film

TABLE 6. Characteristics of UASB Sludpes at the Beclinning and at the End of the Study *IUASB T UASB T UASB U UASB U

Day of operation O 194 O 194TSS gA 42.3 (100%) 57.4 (100%) 26 (100%) 56.3 (100%)FSS gJ 16.1 (38%) 26.9 (46.9%) 8.6 (33% ) 29.9 (53%)vss gA 26.2 (62%) 30.5 (53.1%) 17.4 (67%) 26.4 (47% )g VSS/reactor 26.2 26.2 17.4 11.4SVI mllg 33.3 29.9 86.8 26.2Granule size mm n.d. 0.31 (S = 0.39) n.d. 0.22 (s = 0.23)Viii m/h 1.8 3.2 0.84 3.47Bacterial Counts ( ba ct ed g VSS)Hydrogenophilic methanogens ,7.7 E l0 4.9 E7 3.7 E10 9.4 E10

9 Acetoclastic methanogens 3.1 E8 4.9 E8 5.0 E8 4.7 E8Propionate users 2.8 E9 < 3.9 E6 5.8 E8 8.4 E7Butyrate users 1.5 E9

8/13/2019 010025223

8/14

23 H.MACARIEet al.Peasons for SuperiOritv of Fixed Film R eactor over UASB Reactors When Treahne IT A Plant EffluenLSeveral facts can explain this result.VSS content and mass trans er, At the end of experimentation, he VSS content of the fixed iilm reactor (100.2g ) was much higher than that of the UASB (Table 6). Fixed film reactor VSS content was estimated bysubtracting its drained volum e (2.5 1) from its void volume (4.75 1) and applying a film density of 0.116 kgVSS/m2 for a 2.6 mm biofilm th ickness (Van den Berg and Lentz, 1980). Thus, when similar loading rates(kg COD/m3.d) were applied to each digester, the organic specific loads (kg COD/kg V SS-d) were lower in thecase of the fixed film reactor (T able 3). However, even when the same organ ic specific load was applied, thetubular reactor had a better behaviour. For instance, at 0.09kg CODkg VSS.4 its COD removal efficiencywas 74.5 (day 145-164) while at 0.1 kg COD& VSS.d (day 30-108), the COD removal efficiency ofUASB T was only 33.8% (Table 3). This indicates that mass transfer phenomena might have been ofimportance. In the downflow fixed film digester, the tubular packing favors an homogeneous distribution ofwastewater and practically a l l the biofiim surface is in contact with the substrate. Moreover, the biogas bubbleson their way up along the tubes create perturbations which may increase the liquid-solid transfer. On thecontrary, the biogas production of UASB reactors was not su sc ie nt to create by itself an adequate biomass-substrate contact in t he sludge and so,only a fraction of the bed was really active and could contribute to CODremoval.Biomass distribution and resistance to toxic compounds. As discussed previously, residual terephthalic acid isstill present after primary settling in the form of near colloida l VSS. In order to assess its biodegradability, ananaerobic test was run in serum bottles during 70days. The inoculum used for the test was sludge T, diluted to20 % (v/v) with a mineral medium (Balch eral., 1979). The experiment showed that settled wastewater (0.56gCOD/g VSS) as well as terephthalic acid (0.183 glgVSS), nhibited biogas production with respect to a controlcontaining the same amount of sludge but without substrate (Fig. 3).

O 20 40 60 80DaysFig. 3. Cumulative methane production of batch cultures in serum bottles containingsludge (0.6 g VSS) (controt A ), sludge and settled wastewater(0.56 g COD/g VSS) (O), sludge and terephthalic acid (0.183 g/g VSS) n.

In the fii ed film reactor design, the biomass is distributed all aldng the column , whereas for the UASB, thebiomass is located at the bottom. The presence of toxic compounds in the wastewater under the form ofsuspended solids would affect more the UASB reactors because the TSS could accumulate into the sludge bedand inhibit it, while the upper part of the tubular reactor biofilm would be untouched and active. In addition,the downflow operation of the fixed film digester would allow a continuous wash out of the settled TSS.Onthis subject, it has been shown in batch conditions that dimethyl terephthalate plant wastewater is toxic formethanogenic bacteria even at a m oderate concentration. Only 10% waste or less could be tolerated for growth

8/13/2019 010025223

9/14

Anaerobic petrochemical wastewater treatment 231and m ethane production (DiTommaso and Elkan, 1973). In the Grontm ij report (1980 ), this inhibition wasexplained by the presence of formaldehyde n the effluent.Scanning electron microscopic observ ations of sludge sampled on day 169 from UASB T and U showedprecipitates entrqpped i n the granules (Fig. 4)which supports the hypothesisof a poss ible UASB inhibition bysettled VSS. From the TSS and VSS removal efficiencies of UASB reactors (Table 3), it can be estim ated that30 g TSS and 7 gVSS were eliminated by UASB U and that 20gTSS and 7 g VSS were eliminated by UASBT during the entire operation period. Part of these TSS and VSS could have been removed by sim ple physicalmeans, the upflow velocities in the UASB reactors being very sm all (0.25-0.9 cm/h). T he methane yields ofthe 2 ast steady states were well under the theoretical value (0.35 m3 CH&g COD) (Table 3), indicating thatnot all the COD removed was actually biodegraded.

Fig. 4. Scanning electron microg raphs of UASB reactor g r a d & sludge. (a) Entire granule from UASB T. (b)Same for UASB U. (c)Detail show ing precipitates entrapped in a granule from UASB T. (d) Same forUASB U.Moreover, it must be considered that resistance to aromatic toxicity may be greater when bacteria are fixed onasupport. Dwyer et al. (1986) showed that a consortium composed of a phenol oxidizing bacteria, aMe fhno fhrir-lik e bacteria and an H2-utilizing methanogen could tolerate higher concentrations ofphenol whenimmobilized in agar than in the fo rm of suspended cells.Process ntemation for PTA Waste water Treatment.Anaerobic treatment of industrial wastewater is a very effective pretreatment process but, in order to complywith dis charge standards, a secon dary wastewater treatment is usually needed.

8/13/2019 010025223

10/14

232 H.MACARIE et aLIn previo us work (Noyola et al., 1990) a system consisting of primary settling, anaerobic fixed film reactorand aerob ic treatment was proposed for PTAwastewater treatment. This approach was also taken by a researchteam in China, from laboratory work to a real scale plant, which i s now under start-up (Yang, 1991); the plantwill treat 8400 m3/d discharged by a PTA factory. The treatment system , primary settling, hybrid anaerobicreactor and fixed media activated sludge achieves 95% COD removal with a e of40 hours. The anaerobicreactor reaches 70% COD removal, 0.4-0.6 m3 biogaskg COD removed with an organic load of 8 kgCOD/m3.d and 37C (Yang, 1991).The higher performance obtained in that work supportsour observation that a higher organic load could havebeen treated by the downflow fixed fim reactor (Fig. 2). Nevertheless, our results clearly showed that theUASB reactor was inhibited, which was not the case in the C hinese work. In fact, that study compared UASB,anaero bic filter and hybrid reactors, and the conclusion w as that all performed in a similar manner (Yang,1991). At the moment, we do not know the detail s of that research work, and an explanation of the differencesencou ntered cannot be proposed . Howe ver, the sludge blanket of the hybrid reactor may eventua lly beinhibited by the toxic VSS that are not retained in the primary settler. The process proposed by the Chinesegroup has been recently protected by a patent (Yang, 1990).Apparen tly, Ely and Olsen (1989) investigated a different approach for Amoco Corporation. Like in ourstudy, they used an anaerob ic downflow filter followe d by an optional aerobic treatment, but instead of aprimary decantation to avoid the precipitation of fast settling, poorly soluble aromatic acids in the filter, theyincreased the so lubility of these aromatics by the addition of NaOH to the reactor influent. Caustic soda raisesthe wastewa ter pH, moving the acid -base equilibriumof the aromatics towards the more soluble deprotonatedforms. The solubility of terephthalic acid in water is 19mgfl at 25 OC against 140fl t 25 C for its disodiumsalt.In he integrated process suggested in the presen t paper, so me points should also be considered:PisDosal of Drimarv sett ed solids The solids are mainly organic (92% W/W nd an analysis performed by aprivate laboratory showed that at least 40 was tereph thalic acid. It is not the purpose of this work to discussthe disposal of that solid waste, but their characteristics may suggest that either PTA shou ld be recovered ordisposed of directly in an indusmal landfill, due to the expected low biodegradability. Nevertheless, PTAprecipitation under acidic pH seems to be used in Asia fo r its recovery from PTA production wastewater o rpolyester fiber wastewaters (Motojima et al., 1986; Tan et al. 1986; Yang et al.. 1989).Consumption of chemicals for settled wastewater neut alization, The study was realized in controlledconditions of wastewater pH and alkalinity, but the quantity of NaHCO3 needed was high, 3.15 g / l which isnot economically feasible at real scale. The capacity of anaerob ic sludges to generate their own alkalinity and toadapt to treat unneutralized or partially neutralized acidic effluents is a vel1 known phenomena (Brune et l.,1982; Nel and Britz, 1986; Moreno et al. 1990). In our case, at the beginning of the study, an attempt to start-up the UA SB reactors with unneutralized settled wastewater failed due to sludge acidification. Moreover, asindicated by F igure 5, when fed with neutralized influent, the fixed film reactor which presented the bestbehaviou r, produced little alkalinity, 1174 mg CaC03A du ring the f i s t steady state (day 73-87), and 1136"mgCaC03A during the second (day 145-164).

,

lkalinityinW i i t y o u t42008 7006 3200

2700*E 2200,x*g 17004

1200700

O 20 40 60 8 100 lu 140 160 180 200Operating days

Fig.5 Evolution of influent and effluent alkalinity during the operation of the fied film reactor.

l

8/13/2019 010025223

11/14

Anaerobic perrochemid wastewater treatment 233The use of oth er neutralization agents, Cao NaOH, Na2C03, may lower the cost of neutralization. Othertechnical alternatives to reduce the requiremen t for neutralization chemica ls consist of applying high effluentrecycle to the reactor in order to take advantage of the effluent buffer capacity (Witt et al . , 1980;Fergusonetal., 1984). Moreover, gas stripping to reduce CO2 in the recycle stream can also decrease base addition andrecycle ratios (Ferguson et al., 1984). All these altematives would have to be studied at lab oratory scale todetermine if neutralization of PTA wastewater for anaerobic treatment can be realized at an economical leveLEffluent recycle and C O2 stripping were also retained in Amoco patent (Ely and Olsen. 1989) to decreasecaustic soda requirements for aromatic acids dissolution. In contrast, primary settling, which elim inatesinsoluble aromatics, reduces ch emicals consumption for reactor influent neutralization and should be moreeconomical on thispoint.

CONCLUSIONA lab scale arrangement consisting of a static primary sedimentation of PTA raw wastewater, followed by ananaerobic treatment of the settled wastewater with a downflow fixed film reactor gave an overall reduction of84 in COD and 80% in TSS for an hy drau lic retention time of 3.4 days.

uThe use of UASB reactors for the biological step was not successful. This failure is mostly explained bysludge inhibition caused by toxic characteristics of the wastewater. The tubular reactor was not affectedbecause of its different biomass distrib ution and its higher VSS content. Each reac tor presented good resistanceto shock loads and periods without feeding.Anaerobica lly adapted aerobic activated sludge was a good inoculum for seeding the anaerobic reactors andmay be recommended when granular or anaerobic digester sludges are not available.These results indicate that th e arran gement proposed is a good alternative to the co nventional aerobic processused to treat PTA effluent. Nevertheless, two major problems have to be studied before applying this system:disposal of primary settled solids and reduction in chemical consumption for e ffluen t neutralization.

AKNOWLEDGEMEhVSThis work is part of a project supported by the Organization of American States (OAS) and the EuropeanEconomic Community (EEC).We want to thank Eduardo Cas tillo for excellent technical assistance and weare grateful to the company "DescontaminAccion S. A. de C. V." and its manager Miguel del Villar for havingbrought its support to the study. W e grea tly appreciated the help of the PTA fac tory in the form of w astewatersupply. We are indebted toDrG. Lettinga and DrP. Hulshoff for sendingusthe Grontmij report and to DrY.Yang for providing information ab out PTA w astewater treatment in China. We finally want to thank DrK.Ilangovan for advice abou t the determination of traceelements using spectrophotometric atomic absorptionand U. Kropf for translating the W. Reule article.

REFERENCESAftring, R.P., Chalker, B.E. and Taylor, B.F. (1981). Degradation of phthalic acids by denitrifying, mixedcultures of bacteria. Appl. En viron. Microbiol.,APHA, AWWA and WPCF. (1980). Standard Methods for the Examination of Water and W astewater, 15thedition,Washington D. C.Balch, W. E,, Fox , G. E., M a g " L. J., Woese C. R. and Wolfe R. S. (1979). Methano gens: Reevaluationof a unique bio logica l group. Microbiol. R ev.,4,260-296.Battersby N. S. and Wilson V. (1989). Survey of the anaerobic biodegradation potential of organic chemicalsin digesting sludge. Appl. E nviron. Microbiol., 55 (2), 433-439.Bemis, A . G., Dindorf, J. A., Horwood, B. and Samans, C. (1982). Phthalic acids and otherbenzenepolycarboxylicacids. In: Kirk Othmer Encyclopedia of Chemical Technology. H. F.Mark, D. F.Othmer, C. G. Overberg, G. T. Seaborg, M. Grayson, and D. Eckroth, (eds)., Vol. U ohn Wiley andSons, New Y ork, pp. 732-777.

1177-1183.

8/13/2019 010025223

12/14

234 H.MACARE etaLBerry , D. F., Francis, A. J. and Bollag, J. M. (1987). Microbial metabolism of homocyclic and heterocyclicarom atic compounds under anae robic conditions. Microbiol.Rev.,3 l) , 43-59.Brugnaro, S. and Polo, B. (1985). Effect of some w astes on the proliferation of filamentous bacteria inactivated sludge.Inquinamento.,22(4), 65-68. (in Italian).Brune, G. choberth, S. M. and Sahm, H. (1982). Anaerobic treatment of an industrial wastewatercontaining acetic acid, furfura l and sulphite. Process. Biochem.,u 0-35.C&EN. F acts and figures for the che mical industry. (1988). Chem. Eng.News., 612 une 20,34-44.Chou, W. L., Speece, R. E. and Siddiqi, R. H. 1978). Acclimation and degradation of petrochemicalwastew ater components by meth ane fermentation . Biotech.Bioeng. Symp., 8,391-414.DiTommaso, A. and Elkan, G . H. (1973). Role of bacteria in decomposition of injected liquid waste atWilm ington, North Carolina . In: Int.Symp. Undergroun d Waste Management and A rtificial Recha rge, J.Brauns tein (ed)., Vol. 1.Amer. Ass. Petrol. Geol., Tulsa, Okla, pp. 585-599.Dwyer, D. .,Krumme, M. L., Boyd, S. A. and Tiedje, J.M. (1986). Kinetics of phen ol biodegradation byan mmobilized methanogenic consortium .App l. Environ. Mic robio l.,2 2). 345-351.Ely, S. R. and Olsen, G .P. (1989). Process for treatm ent of wastewater. US Patent. 4,826,600. pp 10 .Evans,W. . and Fuchs, G . (1988). Anaerobic degradation of aromatic compounds. Ann. Rev. Micro biol.,Ferguson, J. F., Eis, B. J. and Benjam in, M. M. (1984). Neutralization in anaerob ic treatme nt of an acidicwaste. Wut. Res.,a 3, 73-580.Frankin, R. J. and Koevoets, W. A. A. (1991). Application of the biobed upflow fluidized bed (UFB) systemfor Hoechst DMT waste water. In: Poster A bstracts, 6th IntSymp. Anaerobic Digestion.pp 156.Grontmij n. v. (1980). Anaerobic treatment of DMT w aste water. Final report. O. N . 97I35. CT: II7OlJans.39 pages.Gu yot, J. P., Noyola, A. and Monroy, O. (1990 a) Evolution of microbial activities and population in g ranularsludge from an UASB reactor. Biofechnol.Lett., u ( 2 ) , 155-160.Guyot, J. P., Macarie, H. and Noy ola, A. (1990 b). Anaerob ic digestionof a petrochemical wastewater usingthe UASB process. Appl. Biochem. Biotech.,a 79-589.Holliger, C., Stams, A. J. M. and Zeh nder, A. J. B. (1988). Anaerob ic degradation of recalcitrantcampounds In: Anaerobic Digestion 1988. Adv . Wut. Pollut. Control no 5)E. R. Ball and P. N. .Hobson (eds.)., Pergamon Press, PP. 211-224.Horowitz, A. Shelton,D. ., Comell, C. P. and Tiedje, J. M. (1982). Anaerobic degradation of aro maticcompounds in sediments and dige sted sludge. Dev. Ind.Microbiol.,a,35-444.Hu ngate , R. E. (1969). A roll tube method fo r cultivation of strict anaerobes. In: Methods in Microbiology,J. R. Noms and D. W. Ribbons (Eds)., Vol.s cadem ic Press, New York, pp. 117-132.Lau, C. M. (1978). Staging aeration for high ef ficien cy treatment of aroma tic acids plant wastewater. Proc.Ind. Waste.Conf. Pwdue University.,s 3-74.Leenh eer, J. A., Malcolm, R. L. and White, W. R. (19 76). Investigation of the reactivity and fate of certainorga nic components of an industiai waste after deep-w ell injection. Env. Sci. Tech.,Lund, F. A. and Rodriguez, D. S. (1984). Acc limation of activa ted sludge to mono -substituted derivatives ofphenol and benzoic acid. J . Gen. Appl. Microbiol.,3 , 5 3 - 6 1 .Mahoney, E. M., Varangu, L. K., Caims W. L., Ko saric, N. and Mu rray, R. G . E. (1987) The effect ofcalcium onmicrobial aggregation dur ing UASB reactor start-up. Wut.Sci. Tech., 3(1 /2) , 249-260.Monroy, O., Noyola, A., Ramirez,F. and Guyot, J. P. 1988). Anaerobic digestion and water hyacinth as ahighly efficient treatment process for deve loping countries. In: Poster Papers, 5th Int. Symp. AnaerobicDigestion, A. Tilche and A. Rozzi (eds)., M onduzzi Editore, Bologna, pp. 747-751.Moreno, G. odriguez, F. J., Jimenez, C. and Noyola, A. (1990). Treatm ent of sugar cane alcohol vinasseswith U ASB reactors: Process start-up and reduction of its inhibition. Proc. VII National Congress ofEnvironmental Engineering and H ealth, Oaxaca, pp. C172-C178. (in Spanish).Motojima, K., Sugii, N. and Takeda, H. (1986). Recovery of highly-pure terephthalic acid from itswastewater. Japanese Parent. 61 43,140. pp 2. (in Japanese).Nel, L. H. and Britz, T. J. (1986). The influence of different substrate pH values on the performance of adownflow anaerobic fixed bed reac tor treating a petrochemical effluent. Biotechnol. Lett., 8 (4), 293-298.Noyola, A., Capdeville,B. and Roques , H. (1988). Anae robic treatment of domestic sewage with a rota ting-stationary fixed-film reactor. Wut. Res.,2 12), 1585-1592.Noyola, A., Macarie, H. and Guyot, J. P. (1990). Treatment of terephthalic acid plant wastewater with ananaerob ic fixed film reactor. Environ. Tech nol.,1 ,239-248 .Nozawa, T. and Maruyama, Y. (1988). Denitrification by a soil bacterium with phtha late and other aroma ticcompounds as substrates.J . Bacteriol.,170 6) , 501-2505.

&& 289-317.

(5),445-451.

8/13/2019 010025223

13/14

Anaerobic petrochemical wastewater treatment 235Reule, W, 1990). Methane from chemical industry wastewater. Chem. Ind. Duesseldorf . 113 (9),Savostianoff, D. (1990). TPA-DMT, Confidence is coming back. Informa tions. Chimie.,u ay, 191-202.Shelton, D. R. and Tiedje, J.M. (1984). General method for determining Anaerobic biodegradation potential.Speece, R. E. (1983). Anaerobic biotechnology for induseial wastewater treatment. Environ. Sci. Tech.,nTan, K., Hm na ka, T., Yano,H.1986). Recove ring terephthalic acid. Japanese Patent. 61 43,139. pp 9. (inJapanese)Van den Berg L. and Lentz C. P. (1980). Comparison between up- and dow n-flow anaerobic fixed filmreactors of vary ing surface to volume ratios fo r the treatment of bean blanching waste. Proc. Ind, WasreConf. Purdue University.,3,19-325.Webber D.1984). Basic chemical output retums to growth. Chem. Ehg. News. 42 19),8-10.Weissermel,K. nd Arpe, H. . (1981). Industrial Organic C he mi so .Masson, New York. (in French).Witt, E. R., Humphrey, W. . nd Roberts, T. E. (1980). Full-scale anaerobic filter treats high strengthWu, W., Hu, J. Gu, X., Zhao, Y., Zhang, H. nd Gu, G. (1987). Cultivation of anaerobic granular sludgeXu,K., Xu, .,Dong, H., Sun, J. nd Zhang, Q. (1988). Treatmentof alkaline terephathalic acid wastewaterYang, G., Teng, D., Han, X. er al. (1989). Recovery of para-benzene dicarboxy lic acid from wastewater w ithYang, Y. (1990). Treatment of wastewaters from production of pure p-phthalic acid. Chinese Patent.Yang, Y. (1991) personal communication.

20,22,24. (in German).Appl. Environ. Microbiof.,a(4), 850-85741 A-427A.

wastes. Proc. Ind. Wasre. Conf. Purdue University.,3,229-234.inUASB reactors with aerobic activated sludge as seed. Wut.Res.,U (7), 789-799.by floccula tion sedimentation. Water. Treat.,3,481-490.ferric salts. Chinese Parent. 1,033,176. pp 5. in Chinese).1,039,784. pp 7. (in Chinese).

. c

8/13/2019 010025223

14/14