Wat. Sci Tech. Vol. 40, No. I, pp. 191-198.1999 <01999
Published by Elsevier ScienceLtdon behalfof the IAWQ Printedin GreatBritain. All rightsreserved 0273-1223/99 $20.00 + 0.00
RECIRCULATION OF POWDERED ACTIVATED CARBON FOR THE ADSORPTION OF DYES IN MUNICIPAL WASTEWATER TREATMENT PLANTS L. Nicolet and U. Rott Institute/or SanitaryEngineering, WaterQualityand Waste Management, University ofStuttgart, Bandtdle 2. 70569 Stuttgart. Germany
ABSTRACT The usc and recirculation of powdered activated carbon (PAC) as an advanced treatment for colour removal in municipal wastewater treatment plants is presented. Studied wastewaters consist of domestic effluents with a high portion of dyehouse residual waters. The particulanty of the treatment is that PAC is not disposed of before being recirculated several times. Therefore, it enables the usc of a great part of the total adsorption capacity of the PAC. A positive side effect is that halogenated and refractory organic compounds, which arc not degraded by micro-organisrns in a conventional municipal wastewater treatment plant, arc removed too. This paper describes results which were obtained in batch experiments and in a pilot plant during two years of observation, and concludes with advantages and drawbacks of this technology. @ 1999 Published by Elsevier Science Ltd on behalfof the IAWQ. All rights reserved.
KEYWORDS Adsorption; flocculation (coagulation); dye removal; powdered activated carbon; wastewater. INTRODUCTION Textile residual waters are often discharged into the municipal sewerage and are then treated together with domestic wastewaters in municipal wastewater treatment plants. These plants are typically composed of a primary settling tank, an activated sludge tank and a secondary settling tank. However, dyes are not biologically degraded through this conventional treatment so that the effluent from the municipal treatment plant remains coloured and can show high levels of halogenated organic compounds. This is why advanced treatment should take place. Powdered activated carbon is well known to adsorb dyes and other refractory organic compounds with high efficiency and speed. However, PAC can not be regenerated and has to be disposed of before its adsorption capacity is totally depleted. Since in Germany environmental standards concerning wastes are quite high, the use of PAC for advanced treatment may only be taken into consideration if PAC is recirculated, in order to profit as much as possible from its adsorption capacity and to achieve a reduction of costs in the purchase and disposal of powdered activated carbon. In this respect, a research project was conducted at the Institute for Sanitary Engineering, Water Quality and Waste Management at the University of Stuttgart over a two year period. The aim of the research project was to determine the conditions under which it is feasible and cost efficient to recirculate PAC for the decolorization of wastewater in municipal treatment plants (Rott et al., 1998). 191
L. NICOLET andU. ROTT
ANALYSIS PARAMETERS AND MATERIALS Analysis parameters Parameters used for the analysis of coloured wastewater were pH, conductivity, COD (Chemical Oxygen Demand), adsorbable organic halogenated compounds (AOX), suspended solids (SS) and absorption coefficients at different wavelengths, for example, at 436 run, at 525 run and at 620 run (Abs. 436, Abs. 525, Abs. 620) or at the wavelength where absorption is maximum. The amount of suspended solids is determined by means of filtration with a membrane filter, which has a pore diameter of 0,45 urn. It gives an indication on the amount of PAC in solution. The absorption coefficient is measured in a photometer according to the absorption through a sample of wastewater (A) and the length of the sample dish (L): Abs. =
According to the Lambert-Beer law, it is possible to derive how much of a compound is dissolved in wastewater from the absorption coefficient. That is why the determination of the dye concentration is not necessary to assess colour contamination in wastewater. Instead the measurement of the absorption coefficient can be very simply carried out. This method is described in the European Standard EN ISO 7887 (1994). Colour is the most important parameter of this research work to assess the efficiency of adsorption and recirculation. Dyehouse residual waters Dyehouse residual waters used in the project contained reactive dyes, salts and auxiliary chemicals. They had been pumped from dye processing machines before the rinsing step and were therefore highly concentrated. Reactive dyes are known for their fastness, that is to say their ability to stay on materials so that the colour does not come out. However reactive dyes have a very low reaction yield of about 55% to 95% with natural textile materials, like cotton or wool, so that they are found in large amounts in dyehouse residual waters (Haelters, 1980; Fiebig et al., 1985; Minke and Rott, 1995, Menzel, 1997). Four types of residual waters from different dyehouses were used in the project. Their main characteristics are shown in Table 1. Table 1. pH-value, conductivity, COD, AOX and absorption at 436, 525 and 620 run of four different dyehouse residual waters which were used in the research project Content
Name A B
054% Cibacron yellow CROI 4 J0% Cibecron red C2G o 12% Cibacron red CR ISO gil Sodium carbonate 1.22% Cibecron yellow F3R 2.32%C.bacron red FB 0.12% Cibaeron blue FB 70 gil Salt 0095% Levafix yellow KR 3.95% Levafix red KGR 70 gil Salt 20 gil Sodium carbonate o8% Remazol red 45% Remazol black 80 gil Sail
Abs. m (m'l)
Abs. 620 (m'l)
Dyehouse residual waters A, B and C were mixed with the effluent from a domestic wastewater treatment plant in order to obtain wastewaters A, B and C with a constant absorption coefficient of 3 mol at the wavelength 525 run.
Recirculation of PAC for the adsorption of dyes
Powdered activated carbon Powdered activated carbon was the main chemical used in this project. The employed PAC, "SA Super", produced by the Norit GmbH company, was known to be very appropriate to remove reactive dyes. Coagulant A solution of 8% aluminium sulphate was used as a coagulant in order to flocculate the powdered activated carbon, so that it can be removed easier from wastewater. The PAC and the coagulant form together a PAC. sludge that can settle rapidly in a settling tank and be recirculated or disposed of. METHODS Batch experiments were performed in order to determine which conditions are favourable for the contact of PAC with wastewater. Parameters like contact time, concentration of PAC, COD, temperature and stirring speed were examined. Moreover, the role of the coagulant was studied with regard to additional decolorization effects and the increase of the PAC-sludge concentration. The batch experiments were conducted with a stirring device, with which it was possible to examine six different samples of I liter at once. A half scale pilot plant (HSPP) was constructed and run. Wastewater A, B and C were used as input with a flow of about 250 l/h. The PAC-sludge recirculation rate was varied in order to obtain various concentrations of PAC in the contact tank. The purpose of the treatment was to obtain an effluent with an Abs.52S of about 0.3 m", that is to say a decolorization rate 0(90%. The principles of the HSPP is as follows (cf. Figure 1): •
Wastewater which has been treated in a conventional domestic treatment plant flows through a contact tank, a flocculation tank, a settling tank and a filter.
PAC is added to the wastewater in the contact tank, where enough turbulence is provided by a stirring device. The hydraulic retention time in the contact tank is 20 min.
A coagulant is added to the water in the flocculation tank, where the flocculation of PAC to PAC-sludge takes place.
The PAC-sludge settles in the settling tank.
The PAC-sludge is recirculated from the bottom of the settling tank back to the contact tank by means of a pump.
Finally, the wastewater flows through a filter in order to eliminate the remaining small PAC particles. RESULTS AND DISCUSSION
Contact time. As expected, the longer the contact time of PAC with wastewater, the greater the colour removal. In wastewater B for instance there was a decolorization of about 57% after 20 min and a decolorization of 70 % after 60 minutes. This shows that the plot of decolorization vs. time is not linear and that a greater part of the decolorization takes place during the first few minutes . Consequently long contact times are not economical for a plant because of the huge volume of the corresponding contact tank. That is why contact tanks are generally designed for about 20 minutes retention time (Sontheimer, 1985). However this experiment proves that the adsorption capacity of the PAC is not totally depleted when PAC is disposed of after 20 minutes. That means that PAC is only partially loaded with dyes. The recirculation of partially loaded PAC in order to fully use its adsorption capacity is hence justified.
L. NICOLET and U. ROTT
D om estic
D ych ouse.---,
Coa gulan t
F IR. L1 P L.
Flow Indicatio.n and Re gistra lion " Level Indication . Pre ssure lndi cat ion
~ r 1 I
I PAC·S uspension
Floccula tion Tan k
PAC -R ecirculation
Air / Water Filter
- O utput
Figure I. The wastewater pilot plant for the removal of dyes with powdered activated carbon half scale pilot plant IISpP).
Concentration of PAC The greater the concentration of PAC in wastewater, the greater the colour removal. However, even with equal PAC concentration levels there were great differences in the decolorization rates for wastewater A, B, and C. Since the process of adsorption of activated carbon largely depends on the characteristics of the dyes, e.g. molecule structure and weight, polarity, pH and ionisation, this result is not surprising. COD in wastewater. Batch experiments were carried out in order to study the cornpentive adsorption between dyes and other compounds. At equal Abs. m values it was discovered that the higher the COD was in wastewater, the lower the decolorization was. These results prove that not only dyes are adsorbed with powdered activated carbon, but also other compounds which contribute to COD. These compounds may be humic acids, proteins and detergents, which arc generally found in effluents of municipal wastewater treatment plants (Nowack et al., 1995). This experiment shows that in order to keep the PAC consumption as low as possible, PAC should be only added to wastewater after a biological treatment. Temperature. Adsorption is an exothermic process (Sonthcimer, 1985). This means that the higher the temperature is, the lower the adsorption capacity due to desorption mechanisms will he. However if the temperature of the wastewater is increased, the diffusion and thus the kinetics of thc adsorption is faster because of the acceleration of the movement of the molecules in the solution. This theory was veri tied in a batch experiment. One sample of wastewater A was heated to 70°C, the other sample not. PAC was added with the same concentration in both samples. After 10 min the first sample had already reached the same decolorization rate as the second sample did after 40 minutes. This means that with a temperature elevation it is therefore possible to design smaller contact tanks. However, this method demands a huge energy consumption, which does not cover the cost saving of the contact tank volume. That is why methods which use temperature elevations to enhance adsorption are not economical.
Recirculation of PAC for the adsorption of dyes
Stirring speed. By increasing the speed of the stirring device from 100 to 600 rotations/min the adsorption kinetics are accelerated and thus the decolorization rate is higher. This has the same reasons as when increasing the temperature. However the energy consumption for the increased stirring is very costly and does not pay for the saving of contact tank volume or PAC. Effect of the coagulant. It is well known that dyes can be removed from wastewater through precipitation/coagulation processes (Koprivanac et al., 1992; Hovelmann et al., 1993; Glockler, 1995). In the HSPP a coagulant was added to the wastewater in order to speed up the settling of PAC, so it was not clear if the coagulant would contribute as well to decolorization. That is why batch experiments were performed, where a PAC and an aluminium sulphate solution were each added to the wastewater either separately or at the same time. The results are represented in Figure 2. 100
z I !i
_ _ 30 mglL PAC
- ..- 30 mglL PAC and 11 mg/L Aluminium
......- - -~-
-+- 5 ,4 mg/L Aluminium
- - 15 mg/L PAC
- ' - 15 mg /L PAC and 5,4 mglL Aluminium _ _
20 Time (min)
Figure 2. Influence of PAC and aluminium sulphate on colour removal in wastewater C (a) and D (b).
Figure 2(a) illustrates that the addition of 30 mg/l PAC in wastewater C enabled a better decolorization than the dosage of II mg/I aluminium. When both products were combined, an even better decolorization was achieved, which was slightly lower than the arithmetical addition of both. This means that the action of each product is only slightly affected by each other. The same batch experiment with wastewater A resulted in a better decolorization due to the aluminium sulphate than due to PAC. An even better colour removal was achieved due to the combined action of both of them. However, a further experiment with wastewater 0, Figure 2(b), had another outcome: the dosage of 5.4 mg/l aluminium resulted in a very low decolorization of about 6% at the wavelength 608 nm. Moreover the combined action of 15 mg/I PAC and 5.4 mg/l aluminium showed a lower decolorization compared to the action of PAC alone. These batch experiments have proved that in the main cases the coagulant used to settle PAC can contribute a great extent to colour removal. However, the effect of adsorption of PAC may be lowered due to the fact that the coagulant may plug the pores of the activated carbon. That is why it is necessary in the HSPP to add the coagulant separately to the PAC. This effect may also have an influence upon recirculation of PAC. Half scale pilot plant (HSPP) The half scale pilot plant described in Figure I was run continuously with wastewater B which had an Abs. S25 of 3 m· l . Every day samples were taken in the feed, in the contact tank, in the flocculation tank, in the settling tank and in the effluent of the HSPP where Abs. m , SS, COD, pH and conductivity were measured (Table 2).
Phase J. In Phase I the recirculation pump of the PAC sludge was not switched on in order to obtain a value as reference for the PAC consumption. The PAC pump was manually adjusted, so that the value of 0.3 m'l was obtained in the effluent of the HSPP. Figure 3 shows the plots of the Abs. m values at different places of the HSPP vs. time. In the contact tank about 54% of colour removal occurred with a PAC concentration of 27 mg/1. As the above batch experiments have already showed, a great part of decolorization (about 29%)
1.. NICOLE T and U. Run
occurred as we ll in the flocculation tan k. whe re 27 mg/l alum inium was necessary to settle approximatel y 27 mg/l PAC.
Figure 3. Abs.", in feed. conta ct tank. settling tank and effluent of the HSPP without recir culation of PAC (Phase I).
Phase 2. In phas e 2 the pump for the recircul ation of the PAC -sludge was sw itched on with a const ant recirculati on rate of 25% compared to the feed flow . That is to say of abo ut 60 IIh. The SS of the PACsludge were varied in order to obtain di fferent co ncen trat ions of partially loaded PAC in the contact tank. This was achiev ed by havin g the PAC-slud gc acc umulate at the bottom of the settling tank and becom e thicker and thicker . In orde r to calc ulate which port ion o f suspended soli ds in the PAC-slud ge was to be attributed to PAC and which port ion was to be attributed to the coa gulant. an iterati ve meth od was developed . Il is based on a mass balance in the co ntact tank. flocculation tank and sett ling tank. A "recircula tion ratio" R was defined as the ratio of reci rcula ted PAC mass to new added PAC mass in the conta ct tank . T he recirculated PAC mass was ealculated with the porti on of PAC in PAC-slud ge. mult iplied by the S5 and the flow of the recircul ated PAC- sludge. Th e HSPP was run at different R varyin g approxi mately from I to 6. Param eters and results of phase 2 compared to phase I arc presented in Table 2. The higher the recirculation ratio R was. the lower the PAC consumption was. PAC sav ing rates of about 50 % were ach ieved. Howe ver. by increasing the level of susp ended so lids in the PAC-sludge wh ich was recircul ated. the sedimentation and filtration steps were overloaded. so that the removal of S5 from the wastew ater did not take plac e effe ctively. As a co nseq uence the filter had to he back was hed twice as often as wi thout recirculation. Furth ermore. the highe r the reci rculation ratio R was. the lower the port ion o f PAC in the PAC-sludge was. This means that the amo unt of PAC-sl udge due to the dosage of a coa gulant made up a great part of the sludge produ ction and was not decreased as PAC co nsumption was decreased. Furth er experiments were conducted witn wastewater A and C. and simi lar results were achieved . In all case s the pH value was slightly redu ced in the effluent compared to the feed and the conductivity increased. mainl y due to the addition of aluminium sulphate. The AOX-elimination reach ed approximately 5()'X•• with concentrations in the feed of about 30 f.I g/1and in the effluent of about 15 f.I g/1. A cost analysis was performed in ord er to determine if the rec ircul ation of partially loaded PAC is economical. It was found that the main source of cos t comes from the purchase of the recirculation pump . It depends in which country the treatm ent sho uld take place. on how exp ensive electricity . powdered act ivated carbon, sludge disposal and pumps are. In Germa ny. for instance, with the sam e co nditions as dep icted for the H5 PP in a treatm ent plant with a daily wastew ater discharge o f abo ut 100 m J or more. it was calcu lated that this technology reduces costs. The recirc ulation o f PAC from the bottom of the sedi mentation tank to the cont act tank is therefore a very simple techni cal so lution that brings adva ntages from an environm ent al and an eco nomic al poi nt of view .
Recirculation of PAC for the adsorption of dyes
Table 2. Parameters and results of the HSPP without (phase 1) and with (phase 2) recirculation of PAC Wastewater B Recirculation ratio R New added PAC PAC saving rate Total Decolorization Decolorization in contact tank Decolorization in flocculation tank Decolorization in sedimentation tank Decolorization in filter COD in Feed COD elimination in contact tank Total COD elimination SS in contact tank SS in flocculation tank SS in Sedimentation SS in effiuent HSPP SS in recirculation flow Portion of PAC in SS of contact tank PH feed HSPP PH effiuent HSPP Conductivity feed HSPP Conductivity effluent HSPP
Without recirculation 27.2 mg/l 90.5% 52.6% 30.7% 3.6% 3.5% 23.4 rng/l 18.7% 51.9% 30 mg/l 163 mg/I 49mg/l 8rng/l 90.7% 7.56 6.53 97611S/cm 1030 /lS/cm
2.1 20.8 mg/l 23.5% 92.3% 78.5% 11.3% 0% 2.5%
2046 mg/l 2172 mg/l 149 mg/l 30 mg/l 9757 mg!1 13.4%
Recirculation 4.2 18.4 rngIJ 32.5% 92% 80.7% 7.1% 0% 4.2% 18.7 mg/l 40.8% 55.8% 2558 mg/l 2877 mg!1 136mg/1 4mg/l 15963 mgIJ 10.2% 7.76 6.81 872I1S/cm 92111S/cm
5.9 41.9% 93.6% 82.9% 8.6% 0% 2.1%
3625 mg/l 3763 mg!1 213 mg/l 25 mg/l 20763 mgll 8.9%
CONCLUSIONS The studies conducted with batch experiments and continuous experiments in a half scale pilot plant confirmed that powdered activated carbon is very appropriate to remove colour and other refractory organic compounds from wastewater in municipal treatment plants. Colour removal was not only due to the adsorption of dyes with activated carbon, but also due to the reactions of dyes with the coagulant, which was added to wastewater in order to flocculate and settle the activated carbon rapidly. The combined effect of adsorption and reaction of dyes with the coagulant mostly increased the decolorization rate in wastewater compared to the single action of adsorption. The recirculation of partially loaded PAC from the bottom of the settling tank into the contact tank contributed to a decrease of the PAC consumption of about 50%. The production of sludge was barely reduced, since the amount of coagulant added to wastewater could not be lowered. The removal of suspended solids would otherwise have been affected due to the high SS level in the recirculation flow. A cost analysis showed that the recirculation of partially loaded powdered activated carbon is feasible and economical. ACKNOWLEDGEMENT This work was funded by the German Federal Government Ministry of Education and Research BMBF. REFERENCES EN ISO 7887 (1994). Water Quality - Examination and Determination ofColour. European Committee for Standardisation. Fiebig, D., Schulz, G. and Herlinger, H. (I~85). Die Bedeutung von hydrolisierten Reaktivfarbstoffanteilen flir Ftirbung und Abwasser. Textil praxis international, August 1985, pp. 855-861. Gleckler, R. (1995). Optimiertes Textilabwasserreinigungverfahren. Melliand Textilberichte, pp. 1020-1021. Haelters, M. (1980). Sichere Reaktivfarberei, ein immer wieder neues Ziel. Melliand Textilberichte 61, 1016-1026. Hovelmann, A.• Bidinger, S. and Linder, A. (1993). Kostengimstige EntflJrbung von Texti/fiJrberei-Abwllssem mittels Ftillung> Ergebnisse von Versuchen. Textil praxis international, pp. 507-509. Koprivanac, N., Jovanovic-Kolar, J., Bosanac, G. and Meixner, J. (1992). Studies on wastewater decolonzation by precipitation/flocculation process. Microchemical joumal46, 379-384. Menzel, U. (1997). Optimierter Einsatz von Pulveraktivkohle zur Elimination organischer Reststoffe aus Klilranlagenablllufen. Stuttgarter Berichte zur Siedlungswasserwirtschaft, Band 143. Minke, R. and Rott, U. (1995). Verfahren der innerbetrieblichen Behandlung von Abwllssem der Textilvcredelungsindustrie. AWT Abwassertechnik.A, 15.20.
L. NICOLET and U. ROTT
Nowack, G. and Ueberbach, O. (1995). Die kontinuierliche SAK-Messung. Aussagekraft, Statistik und Anwendungen. Korrespondenz Abwasser, 11/95,2020-2030. Rott, U. and Nicolet, L. (1998). Entwicklung von Wiederbeladungsverfahren zum optimierten Einsatzvon Pulveraktivkohle in der weitergehenden Abwasserreinigung. AbschluBbericht zurn BMBF-Forschungsvorhaben 02 WA 9547/6. Institut fUr Siedlungswasserbau, Wassergllte und Abfallwirtschaft der Universitlit Stuttgart. Sontheimer, H. (1985). Adsorptionsverfahren zur Wasse"einigung. DVGW-ForschungsstelJe am Engler-Bunte-Institut der Universitllt Karlsruhe (TH).