Municipal wastewater treatment with anodizing solid waste

A. Correia*, T. Chambino, L. Goncalves, A. Franco, R. Goncalves,A. Goncalves, V. Limpo, F. Delmas, C. Nogueira, F. Bartolomeu

Department of Materials and Production Technologies, National Institute of Engineering, Technology andInnovation, Estrada do Paco do Lumiar, 1649–038 Lisboa, Portugal

Tel. þ351 21 716 51 41x2327; Fax þ351 21 716 65 68; e-mail: anabela.correia@ineti.pt

Received 21 March 2005; accepted 10 April 2005

Abstract

In this study we have investigated the feasibility of the use of aluminium anodizing waste as coagulant for

the treatment of municipal wastewaters instead of conventional inorganic materials. We have selected three

different anodizing aluminium facilities for this study. The fresh mud was collected in the industrial facilities and

prepared for tests. The anodizing muds were tested in two different forms: as powder and as mud suspension.

For the powder form the fresh muds were submitted to a stabilisation process, homogenised and grinded. For

the mud suspension form, water was added to the fresh mud in a way that the content of solids was known. A

series of coagulation/ flocculation tests were made, being controlled important parameters as pH, coagulant

dose, and impeller type, stirring conditions (time and speed) and settling time for the different sludges. The

effective coagulant capacity of the anodizing sludges was verified based on the final turbidity and on the

pollutant removal (COD) of the supernatant samples as a function of the coagulant dose and pH.

Keywords: Coagulation; Anodizing waste; Municipal wastewater; Aluminium; Flocculation

1. Introduction

In aluminium anodizing processes, the classi-cal surface treatment of this metal, large quan-tities of waste (anodic mud) are generated. In theindustrial facilities there are normally two kindsof effluents clearly differentiated: concentratedsolutions from the chemical baths and washing

waters. In most installations there is only a set-tling tank where all wastes are blended, neutra-lized, flocculated and settled. The final productis the anodizing mud, which has a variable com-position but that contains three main constitu-ents: aluminium hydroxide, oxy-hydroxides andbasic sulphates, namely of aluminium. This ano-dic sludge is a problem for many countriesbecause it is relatively difficult to manage due

Presented at the Conference on Desalination and the Environment, Santa Margherita, Italy, 22–26 May 2005.

European Desalination Society.

0011-9164/05/$– See front matter � 2005 Elsevier B.V. All rights reserved

*Corresponding author.

Desalination 185 (2005) 341–350

SM 154

to its complex nature. At the present time thedisposal of this sludge on land is a commonpractice and the cost of this operation is gettingincreasingly higher. Today this situation repre-sents a really great environmental problemwhose resolution is necessary and urgent.

The use of aluminium mineral salts in was-tewater treatment is much spread. The chemicalcoagulation and flocculation followed by neu-tralization is a process well known and mostcommon in water and wastewater treatment.For a long time we have been studying thefeasibility of the aluminium anodizing wasteas coagulant [1,2]. In previous works we havetested anodizing sludge from one facility in thetreatment of municipal wastewaters [3] and inthe treatment of paint industry wastewater [4]instead of the conventional inorganic coagu-lants. In this work three anodizing sludgeswere selected due to their chemical compositionin aluminium and two forms of the coagulantwere tested: powder and mud suspension.

2. Coagulation basic principles

Coagulation is a well-known process whichpurpose, combined with a solid-liquid separa-tion process, is the removal of turbidity, colouror micro-organisms that are present in the was-tewaters as colloidal suspensions. These suspen-sions are a heterogeneous mixture of particleswith different size, shape and chemical composi-tion. A colloid has been defined as a dispersionof distinguishable particles in the size range of0.01–10 mm in a medium that may be regardedas a structure less continuum [5]. Colloidal sys-tems will usually scatter light, that is, they exhi-bit turbidity, which is related to the sizes of theparticles involved. Colloidal suspensions in aqu-eous media appear cloudy, and the observedturbidity depends on both the particle size dis-tribution and the mass concentration present.This type of particles tends to remain in

suspension for a long period of time and dueto its great stability colloids do not formaggregates.

The most important interactions affectingsuspension stability are electrostatic repulsionand Van der Waals attraction. These twointeractions are assumed to be additive andtogether establishing the total energy of inter-action between particles as a function ofseparation distance. Attraction predominatesat short distances and repulsion is more effec-tive at greater distances. To eliminate theseparticles the electrostatic forces of the suspen-sion must be destabilized. Then if there isenough kinetic energy available a separationdistance can be reached where attractionbecomes more effective and particle collisionand aggregation can occur.

Coagulation can be described as the agent-induced aggregation of particles suspended inliquid media into larger particles. The coagu-lation favours, with the help of slow stirring,the contacts between the destabilized parti-cles. The particles aggregate to form flocsthat are more easily removed. According toO’Melia [6] four mechanisms of coagulationare recognised: compression of the diffuselayer, adsorption to produce charge neutrali-zation, enmeshment in a precipitate andadsorption to permit interparticle bridging.The destabilisation of colloids in water andwastewater is probably accomplished byadsorption of oppositely charged solubleand insoluble coagulant hydrolysis specieson the colloid and subsequent destabilisation,enmeshment of colloid within hydroxide orcarbonate precipitates, or both.

The use of aluminium in the clarificationof water is common practice in wastewatertreatment. When aluminium salts are dis-solved in water, dissociation into the consti-tuent ions occurs. These ions are submitted tohydration reactions. The products of the

342 A. Correia et al. / Desalination 185 (2005) 341–350

hydrolyse reactions of the aluminium salts arethe effective coagulant agents.

Several hydrolysis species can be formeddepending on the pH, temperature and the con-centration of mineral salts. These hydroximetalcomplexes adsorbon the particle surfaces and thecharges they carry may cause charge reversals ofthe surfaces they adsorb on, contributing to thedestabilisation of the suspension. The hydrolysisreactions have a great tendency to release Hþ,lowering the pH.A different but important effectof the coagulation characteristics of aluminium isthe formation in the alkaline range of a hydro-xide precipitate that appears as a fine colloidaldispersion. These particles tend to aggregateforming hydroxide flocs and then enmesh thecolloidal particles present in the wastewater.Which possibility will occur will depend on theconcentrationof aluminium, the final pHand thewastewater particle concentration as observed byGregory [7]. It can be seen that in acid range itpredominates the coagulation mechanism ofadsorption and charge neutralization and inalkaline range it happens the mechanism ofsweep floc with formation of a precipitate thatinvolves and drags the suspended particles.

3. Characterization of the sludge and the muni-

cipal wastewater

3.1. Anodizing sludge characterization

The anodizing muds were collected inthree Portuguese different facilities havingthe operations of aluminium anodizing. Thesludges came from the wastewaters treatment

of the anodizing plants and were collected atthe press-filter discharge. A representativecomposite sample was collected, well mixedand homogenised. The sludges were charac-terised and the results obtained are presentedin Tables 1 and 2.

Like it was expected aluminium is one ofthe major constituents of the composition ofthe solid fraction along with the total sulphurexpressed in sulphates.

When using the sludges as mud suspen-sions, water was added to the fresh muds ina way that the solids percentage was known.The total solids, in the mud suspensionstested, were between 5 and 12.5% for thethree different sludges.

When the sludges were used in the powderform they were first stabilized. They weresubmitted to a process of drying at roomtemperature, for several days until constantweight loss. The time needed for the stabiliza-tion process was dependent on particle/cakesize, room temperature, air humidity andsludge moisture. After stabilization the driedsludges were grinded in a hammer mill with ascreen of 0.50 mm so that a homogeneouspowder of constant characteristics was pro-duced for the use in the wastewater treatmentexperiments. The stabilized muds have moist-ure values that are presented in Table 3.

The particle size distribution of the threedried sludges tested as flocculants in thiswork was characterised and we have verifiedthat 90% of the particles in S1 sample have adiameter smaller than 286.96 mm, in S2 sam-ple have a diameter smaller than 50.18 mmand in S3 sample they have a diameter smal-ler than 35.10 mm (Fig. 1).

3.2. Municipal wastewater characterization

The municipal wastewaters used in thisstudy were collected in a large municipal was-tewater treatment plant, near Lisbon. This

Table 1

Moisture and pH of the fresh aluminium anodizing

sludges

Parameter Results

S1 S2 S3

Moisture (%) 81 69 70pH 7.4 6.9 6.7

A. Correia et al. / Desalination 185 (2005) 341–350 343

plant recovers the domestic and industrialwastewaters corresponding to 709,000 popu-lation equivalent, being 374,000 industrialequivalents. Representative samples of efflu-ent were collected at the entrance of the treat-ment plant, in the Parshall channel. Sampleswere collected with an automated sampler bycombining small samples collected at succes-sive times previously established. The knowl-edge of the constituents of the municipalwastewaters is important for the quantifica-tion of the pollutants removal after the treat-ment. For this work, two different samples ofwastewater were collected in different dates.Relevant chemical parameters of the waste-waters were determined and are presented inTable 4.

The two samples presented concentrationsof the individual constituents that allow itsclassification accordingly to Metacalf &Eddy [8] between medium and strong.

0

20

40

60

80

100

0.01 0.1 1 10 100 1000diameter (um)

Cu

mu

lati

ve v

alu

es (

%)

S1 S2 S3

Fig. 1. Size distribution of the stabilized aluminium

anodizing sludges.

Table 2

Chemical analysis of the aluminium anodizing sludges after drying at 105�C

Parameter Results

S1 S2 S3

Weight loss 600�C (%) 28 21 7.9Weight loss 1000�C (%) 37 40 9.8Sulphur, total (SO4) (%) 14 32 13Aluminium, total (Al) (%) 22 18 24Sodium (Na) (%) 0.17 1.0 3.3Calcium (Ca) (%) 0.062 10 0.18Iron (Fe) (%) 0.34 0.13 0.023Tin (Sn) (%) 5 <0.1 (q.l) <0.1 (q.l)Manganese (Mn) (mg/kg) 124 30 8.8Chromium total (Cr) (mg/kg) 110 <100 (q.l) 0.50*103Chromium VI (Cr) (mg/kg) <5.3 3.2 4.5Nickel (Ni) (mg/kg) 35 13 <8 (q.l)Cooper (Cu) (mg/kg) 77 48 4.4Zinc (Zn) (mg/kg) 68 41 18Lead (Pb) (mg/kg) 17 5.1 <6.2 (q.l)Strontium (Sr) (mg/kg) <20 (q.l) 33 <20 (q.l)Cyanides (CN) (mg/kg) <0.047 (q.l) <0.029 (q.l) 8.4

q.l. – quantification limit

Table 3

Moisture content of the stabilized and grinded

aluminium anodizing sludges

Results

S1 S2 S3

Moisture (%) 29.60 11.45 11.65

344 A. Correia et al. / Desalination 185 (2005) 341–350

4. Use of anodizing mud as coagulant agent for

municipal wastewater treatment

4.1. Coagulation tests conditions

A coagulation-flocculation process con-sists on three steps: coagulation of the sus-pended solids, growing of the microflocs andelimination of the floc aggregates formed.Besides the wastewater composition the pro-cess is strongly influenced by kinetics processparameters such as rapid and slow mixingsteps.

The initial phase of the coagulation pro-cess is the rapid mixing. The coagulant spe-cies causing destabilisation are transported byturbulent eddies which interact with the par-ticles in the fluid by collisions. The rapidmixing step is then followed by a period ofless intense agitation where floc growth takesplace up to sizes suitable for removal.

A series of tests was programmed, beingcontrolled important parameters such as

coagulant dose and mixing and flocculationconditions (stirring time and speed and settlingtime). The laboratorial tests were performedusing samples of 500 ml of wastewater, twotypes of impellers and using supernatant tur-bidity as control parameter (turbidity mea-surements have 2% accuracy). Turbiditymeasurements represent a convenient experi-mental procedure for the determination ofthe stability of colloidal suspensions. As aggre-gation occurs and the colloids settle out ofsolution, turbidity decreases. The mixing andflocculation conditions used in the tests areindicated in Table 5.

The coagulation process was effective forthe three tested muds and the two differentforms, powder and mud suspension. The for-mation of flocs that settle rapidly was visible.The tests performed have demonstrated theefficiency of the anodizing muds as coagulantbeing the optimal dose dependent on the stir-ring conditions.

Table 4

Chemical analysis of the municipal wastewaters

Parameter Results Limita value

1st sample 2nd sample

pH at 20�C 7.9 7.7 -COD (mg/ l O2) 7.3*102 8.9*102 125BOD5 (mg/ l O2) 3.8*102 3.0*102 25TDS (mg/l) 4.9*102 5.4*102 -TSS (mg/l) 4.0*102 3.6*102 35Aluminium (mg/l Al) 3 4.8 -Iron (mg/l Fe) 1 2.2 -Lead (mg/l Pb) <0.25 (q.l) <0.2 (q.l) -Chromium (mg/l Cr) <0.25 (q.l) <0.2 (q.l) -Chromium VI (mg/l Cr (VI)) <0.1 (q.l) <0.1 (q.l) -Cooper (mg/l Cu) <0.13 (q.l) 0.1 -Nickel (mg/l Ni) <0.25 (q.l) <0.1 (q.l) -Zinc (mg/l Zn) 0.32 0.7 -Turbidity (NTU) 215 270 -

q.l. – quantification limitaEmission limit value, accordingly to the Portuguese Legislation of discharge of Urban Wastewater TreatmentPlants, D.L. 152/97 September 1997.

A. Correia et al. / Desalination 185 (2005) 341–350 345

4.2. Coagulant dose

The degree of clarification obtained whenchemical aids are used in wastewater treat-ment depends on the chemicals quantity andin the control and monitoring of the process.The use of aluminium salts in the treatmentof wastewaters is common practice and thequantity added usually ranges between 0.05and 0.3 g/l [8]. In this work the coagulantdose varies between 0.05 and 1 g/l in orderto detect the effect of the used dose in thecoagulant performance. The muds doses usedwere greater than the recommended for alu-minium salts, but this is due to the fact thatthe muds have aluminium in its compositionbut aluminium salts are not the onlyconstituents.

The final turbidity of the supernatant sam-ples was chosen as indicative parameter forverifying the efficiency of the coagulationprocess. Some results from these experimentsare displayed in Fig. 2.

The results have showed that as the dosedecreases, for the same conditions of stirring,the supernatant turbidity increases. Neverthe-less in a dose range between 0.3 and 1.0 g/lthe turbidity reduction was almost alwaysgreater than 80% and for a coagulant dose

Table 5

Coagulation conditions

Stirring ConditionsType of impeller

N.° 1 N.° 2

Rapid Stirring Speed 100 or 300 rpm 300 or 800 rpmRapid Stirring time 5 or 15 min 5 or 15 minSlow Stirring Speed 50 rpm 50 rpmSlow Stirring time 30 min 30 min

Sedimentation time 30 min, 1 h and 2 h 30 min, 1 h and 2 hCoagulant dose 1 to 0.05 g/l 1 to 0.05 g/l

(100 rpm-5 min) and (50 rpm-30 min)settling-1 h

0%

20%

40%

60%

80%

100%

1 0,5 0,3

1 0,5 0,3

Mud suspension dose (g/l)

Tu

rbid

ity

red

uct

ion

(100 rpm-5 min) and (50 rpm-30 min)settling -1 h

02468

101214

Mud suspension dose (g/l)

Su

per

nat

ant

pH

S1 S2 S3

S1 S2 S3

Fig. 2. Turbidity reduction and pH obtained with

different mud suspension doses for the same stirring

conditions and using impeller n�1.

346 A. Correia et al. / Desalination 185 (2005) 341–350

of 1 g/l the turbidity reduction reached the90%, for the three muds (Fig. 2).

In all performed tests the supernatant pHwas greater than 7.8 for all the coagulantdoses and for all the muds tested. So, sincethe final pH is in alkaline range we haveconcluded that the predominant coagulationmechanism was the mechanism of sweep flocwith formation of a precipitate that involvesand drags the suspended particles. During thewastewater treatment process hydroxide pre-cipitates are formed. The suspended contami-nants are removed by heterocoagulation orare enmeshed by the precipitates, destabilisa-tion mechanisms also mentioned by Dempsey[9]. The soluble contaminants are adsorbedon the precipitates.

4.3. Rapid stirring

Accordingly to Francois [10] the coagula-tion/flocculation process is strongly influencedby kinetic process parameters: duration ofrapid mixing, slow mixing steps and the energyinput during the different phases.

Experiments were done in order to estab-lish the influence of the rapid mixing period,speed and time, for different combinations ofcoagulant dose as powder or suspension. Theslow mixing phase was always long enough toform fully grown flocs. Typical experimentalresults from these experiments are shown inFig. 3, with 1 hour of settling time. Theresults presented in Fig. 3(a) were obtainedwith dry mud and impeller n�1 and inFig. 3(b) also with dry mud but using impel-ler n�2. In all the experiments the slow stir-ring time and speed was the same.

The results obtained for the three sludgestested and for the different rapid mixing con-ditions are very similar, being the turbidityreduction always greater than 78%, valueobtained for S1 (300 rpm-15 min; 50 rpm-30 min; impeller n�2).

The difference between the minimum andthe maximum value of turbidity reductionobtained for the dose of 1.0 g/l was for S1of 8%, for S2 of 5% and for S3 of 4%. Thesesmall differences in the turbidity reductionvalues for the different mixing conditionsagree with Amirtharajah [11], which saysthat in the processes where sweep coagulationdominates there is little difference in coagula-tion results obtained with different intensitiesof rapid mixing.

Other particularity of the results is the factthat in some tests the increase of rapid mixingtime results in the increasing of the superna-tant turbidity on the contrary of the expected.We believe that in these cases extended rapidmixing has disturbed the growth of the flocs.A critical time exists for rapid mixing, andmixing for a period longer than that criticaltime leads to a disturbance of floc growthwith disadvantageous consequences for thefloc characteristics [10]. The energy fromeach rapid mixing allows a certain maximumfloc diameter and a rapid mixing time greaterthat the critical time yields smaller flocsbecause larger flocs will be formed and rup-tured during the rapid mixing period.

The best results obtained, for the Fig. 3conditions, for the turbidity reduction was86.9% for S1 sludge for 800 rpm-15 min;50 rpm-30 min and impeller n�2, 86.0% forS2 sludge for 300 rpm-15 min; 50 rpm-30 min and impeller n�1 and 93.2% for S3sludge for 100 rpm-15 min; 50 rpm-30 minand impeller n�1. These results show that forthe dose of 1.0 g/l of dry mud the 15 minrapid stirring time gives the greater turbidityreduction for the three muds tested.

4.4. Coagulant characteristics

In this work we have tested the use of theanodizing muds as coagulant in two differentforms. The coagulant was added to the

A. Correia et al. / Desalination 185 (2005) 341–350 347

wastewater in the form of a powder or asmud suspension. For the S2 sludge, examplein Fig. 4, the solids percentage in the mudsuspension was 12.5. We have verified thatthe coagulant as suspension gives betterresults in the turbidity reduction of the super-natant and also in the COD removal (Fig. 4).

In the two conditions presented in Fig. 4for S2, the COD reduction obtained for thepowder form range between 79 and 90% andfor the mud suspension the COD reductionrange between 84 and 93%. For the condi-tions presented in Fig. 4 for the powder formthe selected dose should be 1 g/l that gives a

COD reduction of 90.1% (Fig. 4(a)) and forthe suspension form we could work with adose of 0.2 g/l and still obtain a COD reduc-tion of 92.9% (Fig. 4(b)).

5. Discussion

The municipal wastewater tested has aCOD, close to 800 mg/l and a content insuspended matter around 380 mg/l. The ano-dizing muds tested have, effectively, throughcoagulation process, contributed for theremoval of the wastewater COD, reaching,in some cases, values higher than 90%.

(a) - Impeller n°1

(300 rpm-5 min;50 rpm-30 min;1 h)

(300 rpm-15 min;50 rpm-30 min;1 h)

(800 rpm-15 min;50 rpm-30 min;1 h)

(800 rpm-5 min;50 rpm-30 min;1 h)

Dry muds dose -1,0 g/l

010

2030

4050

6070

8090

100

Tu

rbid

ity

red

uct

ion

(%

)

S1 S2 S3

(b) - Impeller n°2

Dry muds dose -1,0 g/l

010

2030

4050

6070

8090

100

(300 rpm-5 min;50 rpm-30 min;1 h)

(300 rpm-15 min;50 rpm-30 min;1 h)

(100 rpm-15 min;50 rpm-30 min;1 h)

(100 rpm-5 min;50 rpm-30 min;1 h)

Tu

rbid

ity

red

uct

ion

(%

)

S1 S2 S3

Fig. 3. Turbidity reduction obtained with different stirring conditions for the same dry muds dose.

348 A. Correia et al. / Desalination 185 (2005) 341–350

The coagulation process was effective andit was visible the formation of flocs thatsettled rapidly. The size of the flocs formeddepended on the agitation conditions andmud dose used. The coagulation processobserved for all the three muds occurred inpH zone between 7.8 and 8.5, what accordingto the coagulation diagram for aluminiumsalts proposed by Johnson [12], is the pHzone were there is a predominance of themechanisms of enmeshment in a precipitate.

The precipitate mechanism will act in thefollowing way: precipitation of the metalhydroxide and contaminants removal byenmeshment in this precipitate. Soluble con-taminants are adsorbed on the hydroxide andparticles are removed by heterocoagulation[9]. Also according to Amirtharajah [11] thehydroxide is positively charged when the pHis less than 8, so, in the range of the tests theprecipitate is positively charged and theenmeshment mechanism is enhanced by het-erocoagulation between the precipitate andthe negatively charged suspension. Accord-ingly to Duan [13] the sweep flocculation

mechanism gives improved particle removalcomparatively with the mechanism of chargeneutralization. Hydroxide precipitates have arather open structure having a higher prob-ability of capturing other particles. This canexplain the high removal rates achieved in theflocculation experiments with the anodizingsludge.

The results have showed, as expected, thatfor the same conditions of stirring, the super-natant turbidity decreases with an increase ofthe mud dose. Nevertheless in the dose rangetested, for the three muds, 0.05–1.0 g/l theturbidity reduction was almost always greaterthan 70% for the dry muds samples andgreater than 80% for the muds suspensionform.

We have verified for all the tested sludges,S1, S2 e S3 that the mud suspension gavebetter results in turbidity and COD reductionthan the use of dried mud. We believe thatthis is due to the fact that when dealing witha powder, the dissolution process is veryimportant to ensure that the aluminium pre-sent acts as coagulant, and to make the

Impeller n° 1; (300 rpm - 5 min) and (50 rpm - 30 min) settling - 1 h

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 0,4 0,2 0,1

S2 dose (g/l)

CO

D r

edu

ctio

n (

%)

Dry mud Mud suspension 12,5%(w/w)

Impeller n° 1; (300 rpm-15 min) and (50 rpm-30 min) settling -1 h

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

S2 dose (g/l)

CO

D r

edu

ctio

n (

%)

(a) (b)

1 0,4 0,2 0,1 1 0,4 0,2 0,1

Dry mud Mud suspension 12,5%(w/w)

Fig. 4. COD reduction for different doses of S2, powder and mud suspension and for two different stirring

conditions (time and speed).

A. Correia et al. / Desalination 185 (2005) 341–350 349

maximum use of the coagulation substancespresent. The presence of a considerable per-centage of aluminium in the chemical compo-sition of the sludge solid fraction is the mainfactor that contributes to the sludges coagu-lation capacity.

6. Conclusions

The coagulation process is a versatilemethod used in wastewater treatment. This pro-cess provides a high removal efficiency of dif-ferent parameters, including COD, BOD, SS,and microorganisms. The coagulation and floc-culation processes remove SS at an efficiency of80–90%, CBO at an efficiency of 50–80% andmicroorganisms at an efficiency of 80–90% [8].

It was verified through the experimentalwork that the three anodizing sludges testedhave an effective coagulant capacity since wehave recorded efficiencies in COD and turbidityreduction of 90%. The aluminium in the sludgesolid fraction is the fundamental element for thesludge coagulation action. Floc formation isobserved and flocs are easily separated fromthe supernatant by settling. Also, for all thethree sludges tested the coagulant in suspensionform presents better results in contaminantsreduction for smaller doses.

The final pH of the treated wastewateragrees with environmental regulations notbeing needed a neutralization phase in opposi-tion to the classical treatment with aluminiumsulphate. So the use of the anodizing sludge ascoagulant in urban wastewater treatmentseems promising due to its efficiency.

References

[1] F. Bartolomeu, T. Chambino, L. Sota and F.

Delmas, Use of Aluminium Anodizing Sludge

in Wastewater Treatment, in: Congress on Char-

acterisation and Treatment of Clean-up Sludge

from Dredging, Sewage Sludge, Drinking Water

Sludge and CIPS, CATS III, G. de Schutter and

R. Vanbrabant, eds., Belgium, KVIV - Technolo-

gisch Instituut Oostende, 1996, 389–398.

[2] F. Delmas, F. Bartolomeu, T. Chambino, L.

Goncalves and L. Sota, Reciclagem de Lamas de

Anodizacao de Alumınio. in: V Conferencia Nacio-

nal sobre a Qualidade do Ambiente, C. Borrego, C.

Coelho, L. Arroja, C. Boia and E. Figueiredo eds.,

Comissao de Coordencao da Regiao Centro

(CCRC), Averio, 2 (1996) 1725–1734.

[3] T. Chambino, A. Correia, A. Goncalves and

F. Bartolomeu, Reuse of Industrial Waste

Sludges, Wastewater Treatment International

Water Association (IWA) 2� World Water Con-

gress. Proceedings in CD-Rom, Berlin 2001.

[4] A. Correia, T. Chambino, A. Goncalves, C.

Ribeiro and F. Bartolomeu, Use of Solid Waste

from Surface Treatment of Aluminium as Coagu-

lant. 7th Conference on Environmental Science

and Technology, Ermoupolis, Greece, University

of the Aegean, T. D. Lekkas eds., 2001, 129–138.

[5] M.B. Hocking, K.A. Klimchuk and S. Lowen,

Polymeric Flocculants and Flocculation. J M S -

RevMacromol Chem Phys, C39(2) (1999) 177–203.

[6] C.R. O’Melia, Coagulation and Flocculation,

Physicochemical Processes for Water Quality

Control, Wiley Interscience, New York, 1972.

[7] J. Gregory, Flocculation by Inorganic Salts. in: The

Scientific Basis of Flocculation, K.J. Ives, eds.,

Sijthoff & Noordhoff, Alphen aan den Rijn, 1978.

[8] Metcalf and Eddy, Wastewater Engineering –

Treatment, Disposal and Reuse, 3rd ed., McGraw-

Hill 1995, Tata Mcgraw-Hill, New Delhi.

[9] B.A. Dempsey and C.A. O’Melia, Removal of

Naturally Occurring Compounds by Coagula-

tion and Sedimentation. Critical Reviews in

Environmental Control, 14(4) 1984.

[10] R.J. Francois and A.A. Van Haute, The Role of

Rapid Mixing Time on a Flocculation Process.

Journal Water Science and Technology, 17

(1984) 1091–1101.

[11] A. Amirtharajah and K.J. Mills, Rapid-Mix

Mechanisms of Alum Coagulation. J AWWA,

74(4) (1982) 210–216.

[12] P.N. Johnson and A. Amirtharajah, Ferric

Chloride and Alum as Single and Dual Coagu-

lants. J AWWA, 75(5) (1983) 232–239.

[13] J. Duan and J. Gregory, Coagulation of Hydrolys-

ing Metal Salts. Adv Colloid Interface Sci, (100–

102) (2003) 475–502.

350 A. Correia et al. / Desalination 185 (2005) 341–350