Gulation settling of natural organic matter from soft tropical water


Published on

Published in: Technology, Business
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Gulation settling of natural organic matter from soft tropical water

  1. 1. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME118COAGULATION-SETTLING OF NATURAL ORGANIC MATTERFROM SOFT TROPICAL WATER USING ALUMINIUM ANDIRON(III) SULPHATEJ.M. Sieliechi1,*, R. Kamga1, G. Joseph Kayem21. Department of Applied Chemistry, ENSAI, P.O. Box 455, University of Ngaoundere,Cameroon2.Department of Process Engineering, Water Treatment and Filtration Group, ENSAI, P.O.Box 455, University of Ngaoundere, CameroonABSTRACTHigh natural organic matter (NOM) containing tropical forest river waters are verydifficult to treat for drinking because of their low values of alkalinity, mineralization andturbidity. In addition, owing to cost of chemicals, coagulation of these waters is performedunder variable pH conditions. We report studies on such high NOM water, using syntheticwater containing NOM extract from a tropical forest river and natural raw water from thesame river, by Jar test and column settling experiments. Two coagulants were tested,aluminium and iron (III) sulphate, in order to determine, firstly, the pH range at which theinitial water pH should be set for best coagulation and minimisation of residual NOM, andsecondly, the influence of each coagulant on NOM removal and floc sedimentation. Jar testresults showed that minimum coagulant consumption for maximum NOM removal attendedby minimum NOM residuals obtained for initial pH in the range 5.5 – 6.5 for bothcoagulants. The amount of Al sulphate required was at least 2 times that of Fe (III) sulphate.In column settling of metal-NOM flocs, Al was generally better than Fe but both gavecomparable results at hydraulic loading rates ≤ 1.0 m/hr.Keywords: Coagulation, humic substances, sedimentation, soft water, water treatment.1. INTRODUCTIONIn the hot humid tropical regions, surface waters, especially river waters constitute theprincipal source of raw water for drinking water production for cities. The river waters aresoft with slightly acidic to neutral pH, have low levels of alkalinity and mineralization. In theINTERNATIONAL JOURNAL OF ADVANCED RESEARCH INENGINEERING AND TECHNOLOGY (IJARET)ISSN 0976 - 6480 (Print)ISSN 0976 - 6499 (Online)Volume 4, Issue 4, May – June 2013, pp. 118-133© IAEME: Impact Factor (2013): 5.8376 (Calculated by GISI)www.jifactor.comIJARET© I A E M E
  2. 2. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME119dense luxuriant rainforest zones, river waters have low turbidity and medium to high valuesof dissolved natural organic matter (NOM), generally called humic substances [1, 2, 3].Aquatic humic substances represent at least 50% of the dissolved organic on matter(DOM) in natural surface waters, and in the tropical rainforest, they are essentially, the waterextractable fraction of soil humus that derives from biogeochemical processes such as thedecay of plant residues [1,4]. The yellow to brown colour of medium to high NOMcontaining waters can at least in part be attributed to the presence of dissolved colouredhumic substances. At typical natural surface water pH, the humic substances present are amixture of fulvic and humic acids, which account for about 80 % of DOM; with fulvic acidspredominating in most waters especially non-coloured waters and humic acids can be ofgreater proportion in highly coloured waters [4,5]. Fulvic and humic acids possess phenolicand carboxylic groups whose dissociation increase with pH so that in the pH conditions ofnatural surface waters, they have negative charges making them more soluble and stable, andenabling them to contribute to the stabilisation of aquatic inorganic colloids [2].Several water quality problems are posed by the presence of humics, necessitatingtheir complete removal or at least minimization. First, aesthetically, colour, taste and odourmake these waters unfit for drinking. The presence of humic substances in natural watersrenders these waters difficult to treat, causing higher coagulant demand [6, 7]. Indeed,membrane fouling, trihalomethanes formation during chlorine desinfection, or biologicalregrowth in the distribution network, have all been linked to the presence of residual humicsubstances in clarified water [8,9,10]. Humics substances form complex with hazardousinorganic and organic contaminants such as trace (heavy) metals and pesticides, thusprobably facilitating their transport through treatment systems [11,12].Chemical coagulation by hydrolysing metal salts is the major technique used aroundthe world for removal of NOM from water, with Al salts most widely used and iron (III) saltsto a lesser extent [13,14, 15, 16]. For the purpose of improving dissolved NOM removal,considerable research efforts have been made in recent years towards investigating themechanism of destabilisation of aquatic NOM by Al and Fe salts [17, 18, 19, 20]. Three mainmechanisms are generally invoked to explain the removal of humic substances bycoagulation: charge neutralization/complexation preferentially applies at acidic pH and findsexperimental support from stoichiometric relationships between coagulant demand anddissolved organic matter concentration, and from suspension restabilization upon overdosing[21, 22, 23]. On the other hand, under conditions favoring metal hydroxide precipitation,physical ensmeshment and/or adsorption onto the freshly formed precipitate are assumed toplay a major role in humic substances elimination [24].The studies undertaken so far on NOM removal from natural waters have beenessentially concerned with hard, high alkalinity waters. However, Eikebrokk [25, 26], hasinvestigated NOM removal by Al salts from low alkalinity soft Norwegian waters at constantpH and found that minimisation of NOM residuals occurred at pH 5 – 6 and Al metalresiduals at pH 5 - 7. The case of tropical soft humic waters with low alkalinity,mineralization and turbidity, has hardly received any attention. Removal of NOM from thesewaters by water suppliers notably in tropical Africa is based on turbidity and colour reductionwith Al sulphate. Insufficient importance is accorded to process improvement with respect tominimisation of DOM residuals, despite the risks associated with the presence of humics inwater as indicated earlier above.In addition, although it has been established that pH is one of the most crucial factorsregulating NOM removal and metal ion speciation [19, 27, 28, 29, 30], coagulation of
  3. 3. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME120tropical soft humic waters is carried out under conditions of variable pH. Coagulant is addedto the water at its natural pH or after raising pH to a desired value by liming, depending onin-plant experience. The pH is then allowed to drift downwards, towards, it is expected, theoptimum pH for NOM destabilisation. Unfortunately, the coagulation process often fails inthis case, causing coloured water to arrive at the tap. Even though the high cost of treatmentchemicals is at the root of this practice in developing tropical Africa of allowing pH driftduring coagulation. The operative mechanisms have been inferred from the discussions byDuan and Gregory[31].There are no studies available to indicate the optimum initial pH rangefor coagulation of these soft waters under conditions of variable pH, where pH is not adjustedafter adding coagulant.Al is preferred over Fe III sulphate as coagulant for treating humic surface waters inthe tropics because of the corrosiveness of Fe III salts and the risk of staining of clothes bythese salts. However, research is giving greater consideration to the use of Fe III salts becausethe aluminium is non essential metal. Also, it is claimed that Fe III forms stronger and denserflocs with NOM than does Al and it should thus be expected that Fe-NOM flocs should settlefaster than Al-NOM flocs. Yet, no pilot scale study has been performed to test the settlingproperties of the two types of metal-NOM flocs. Deductions are made based on small scaleJar tests [32].The settling characteristics of flocs can be evaluated by column settling pilot scaletests in the laboratory [33,34]. Zanoni et al. [35] have shown that for columns of height equalto or greater than 2.0 meters and internal diameter of 100 mm or above, the results of thecolumn settling tests were similar. Most studies by column settling of aqueous flocculentsuspensions have been performed on inorganic colloids and waste water systems.This study was therefore directed at the investigation of the removal of NOM withfreshly prepared Al and Fe III sulphate from soft, low alkalinity and mineralization tropicalriver water in a treatment situation involving pH drop from a preset initial value. It had twoaims, namely, determining the initial pH range for optimum coagulation with minimum NOMresiduals and evaluating on pilot scale the settling of metal-NOM flocs formed by the twocoagulants and hence appreciate their relative effects on clarification. The study wasperformed by Jar test and column settling tests and involved synthetic waters and naturalriver water.2. MATERIALS AND METHODS2.1. Water sourcesFeedstock waters used in this study were from two sources, distilled water and rawriver water. Distilled water from a borosilicate glass still (JENCONS, UK), was used forextraction of natural organic matter (NOM) and preparation of synthetic waters containingNOM for Jar test experiments and column settling tests. Raw river water was from RiverNyong in the Cameroon rainforest south of Yaounde (Latitude 4°N), on which is located thewater treatment plant supplying Yaounde, Cameroon’s capital. The river water was used forJar test and column settling experiments.2.2. ChemicalsAll chemicals were analytical grade reagents. Aluminium sulphate (Al2(SO4)3.18H2O)coagulant, sulphuric acid (H2SO4) and sodium hydroxide (NaOH) used for pH adjustment,were obtained from PROLABO France, now MERCK-Eurolab. Ferric sulphate, Fe2 (SO4)3xH2O coagulant was from RIEDEL DEN HAEN, Germany. Since the number of molecules of
  4. 4. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME121water in the ferric sulphate was unknown, this chemical was dried to constant weight at 105°Cand cooled in a desiccators before being used to prepare solutions based on the formula weight of399.88 g/mol.2.3. GlasswareAll glassware used was of borosilicate glass.2.4. ApparatusThe pH of solutions and suspensions was measured by means of a HANNA HI 931000pH-meter (Hanna Instruments, Portugal) with a precision of 0.05 pH units.A SPECTRONIC GENESYS 2PC model 336003 single beam UV-Visiblespectrophotometer (Spectronic Instruments Co, USA) equipped with a quartz cuvette (10 mmoptical path) was used to determine the absorption peak of dissolved NOM and also forestablishing the calibration graph for determining NOM concentration in waters tested. Theabsorption peak was confirmed on a double beam UV-visible spectrophometer.Coagulation - flocculation of NOM in water was performed on a VITTADINI Jar testapparatus (PROLABO, France) equipped with 6 paddle stirrers.Column settling experiments to determine removal rates of coagulated NOM were carriedout with the system shown in Figure 1. It consists of a rapid mixing tank, a pump and transparentacrylic polymer (Polymethylmethacrylate) Column of height 2m and internal diameter 100mm.Both the mixing tank and the column are equipped with paddle stirrers as shown. Sampling portson the column were fitted with short silicone tubing tightly sealed with screw valves at 0.5m,1.0m and 1.5m depths. A syringe fitted with a small hypodermic needle was used to withdrawsamples the column were fitted with short silicone tubing tightly sealed with screw valves at0.5m, 1.0m, and at each port.Figure 1 Schematic diagram of gravity sedimentation apparatus0500 mm1000 mm1500 mm2000 mmMotorSedimentationcolumnPumpMixing tankMotorValve
  5. 5. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME1222.5. ProceduresNOM extract for coagulation-flocculation studies was obtained by alkaline extractionfrom river sediment (soil) knowing that aquatic NOM and soil humus have the same chemicalbehaviour (Stevenson 1994). The extraction was done by the Jackson [36] method as follows.Fresh sediment was dug out during the rainy season from Nyong river bed upstream from thewater treatment plant. The sediment was allowed to dry in air at room temperature (24±2°C)inside a vacuum cupboard. The dried sediment, hereafter called soil, was then ground to finepowder in an agate mortar. Aliquots (100g) of powder soil were placed in glass beakers (2L)to each of which was then added 1L of NaOH solution (0.25M). The mixture in each beakerwas agitated for 20 minutes with a HB502 magnetic stirrer (Bibby Sterilin, UK). Afterstopping agitation, the beakers were covered with parafilm and the suspensions formed wereallowed to settle for 12 hours. The supernatant (pH 12) was carefully poured into other 2Lglass beakers, pH adjusted to 7 by adding drops of H2SO4 (0.5M) and the resultantsuspension allowed to stand from 12 h. The supernatant solutions containing NOM extractwere then carefully transferred to 1L screw capped glass bottles.The aqueous NOM extract so obtained was filtered by means of a CARBOSEP (TECH-SEP,France) tubular membrane module having a membrane pore size of 0.14 µm. The clearfiltrate, also called permeate, containing dissolved NOM was stored in 1L screw capped glassbottles at 4°C until required for experiments.Before use, the concentration of the stock NOM extract was determinedgravimetrically after drying the extract solution. A sample of the NOM extract solution wasplaced in a Spectra/Por membrane dialysis tube (Spectrum Medical Industries, USA) anddialysed exhaustively against distilled water (conductivity 3 µS/cm) under agitation in a glasstank. Water was changed every 12 hours until the conductivity of the dialysing water wasconstant and close to that of distilled water. Conductivity was measured with a TACUSSELCD 60 Resistimeter (TACUSSEL Electronique, France). The concentration of the dialysedstock NOM extract solution was then determined gravimetrically by drying a sample (200mL) to constant weight at 105°C. Triplicate determinations were made and the average takento calculate the concentration of the stock NOM solution.The residual concentration of NOM after Jar test coagulation-flocculation and settlingexperiments was determined as follows. The UV-Visible absorption peak of dialysed stockNOM solution at several pH values was found by measuring absorbance versus wavelengthafter diluting the NOM solution to give an absorbance value in the range 0.1 to 0.9 [37]. Theabsorption peak was at 280 nm was quite independent of pH as can be seen in Figure 2Korshin et al. [38] in their work show that the presence of coagulant species doesn’t modifythe spectrum of NOM. Knowing the concentration of the dialysed stock NOM solution andthe dilution factor used, a calibration graph of NOM concentration, C (mg/L) versusabsorbance at 280 nm was established and shows a linear graph. This graph of equation,Absorbance = 0.0230*C + 0.0364, has a correlation coefficient of 0.994.
  6. 6. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME123Figure 2 UV-Visible absorption spectra of diluted dialysed NOM extract solution at variousvalues pH 4 (●), 5 (○), 6 (▼), 7 (∆)Coagulation-flocculation experiments were performed on synthetic waters and NyongRiver water by means of a Jar test bench with 6 glass beakers of 1L each filled with 800 mLof water. Nyong River water was tested as such, the average dissolved NOM concentrationbeing 20±0.6 mg/L and pH 5.9±0.2 during the period of investigation. The synthetic waterswere obtained by diluting stock NOM extract solutions with distilled water down to a desiredconcentration (10, 25 or 40 mg/L), and the pH was then adjusted under agitation to apredetermined value with sulphuric acid or NaOH (0.05M). The Jar test procedure involvedadding coagulant from a freshly prepared stock solution of 10g/L not more than 6 hours oldto the NOM containing water under rapid agitation at 260 rpm, after which, stirring wascontinued for 2 minutes at 260 rpm followed by 10 minutes at 50 rpm. Agitation was thenstopped and the resultant suspension was allowed to settle for 20 minutes. The final pH wasmeasured, the ccc ascertained and residual NOM at ccc determined by absorbance at 280 nmon a small sample prefiltered on a 0.45 µm porosity disc membrane. The sample analysed forresidual NOM was withdrawn each time at the same level (5cm) below the water surface inthe beaker using a syringe fitted with a small hypodermic needle. The values of ccc, final pH,as well as residual NOM concentrations, were established on the basis of triplicatecoagulation-flocculation tests.Column settling studies were carried out with the set-up shown in Figure 1 onsynthetic waters (25mg/L NOM) at initial pH 5.0±0.05 and 6.0±0.05 and on Nyong riverWavelength (nm)0 150 200 250 300 350 400 450 500 550 600 650 700 750Absorbance0123
  7. 7. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME124water (20.0±0.6mg/L NOM) at pH 5.9±0.2. For this purpose, the predetermined volume ofcoagulant solution to attain the ccc was added to water in the mixing tank under high speedagitation and stirring continued for 2 minutes at the same speed then stopped. The pump wasthen used to rapidly transfer enough suspension (15.7L) to fill the settling column to the zeromark. The suspension in the column was agitated for 15 minutes at slow speed (velocitygradient 40/s). Agitation was stopped and samples were withdrawn from the indicated ports(0.5m, 1.0m, 1.5m) at predetermined times. Sample withdrawal was staggered and alternatedbetween column sides in order to avoid channelling effects. Removal rate of flocculent NOMparticles was determined by measuring turbidity of withdrawn samples with a HACHRatioXR turbidimeter (HACH Co., Loveland, USA). The column settling tests wereconducted in duplicate in each case studied. At the end of each run, the percentage removal atthe three selected depths for each settling time was calculated [39]. Then, following themethod of Krishnan [33] the average percentage removal in the column and thecorresponding settling velocity also called hydraulic loading rate (HLR) was evaluated foreach of the chosen settling times. HLR was given by the height of the column divided by thesettling time. In literature, sometimes surface overflow rate (SOR) is also used but that issimply obtained by dividing the HLR values by the cross-sectional area of the settling tank.3. RESULTS AND DISCUSSION3.1. Coagulant Demand and NOM ResidualsThe influence of initial pH of water (pHi) on the critical coagulation concentration(ccc) and thus on coagulant demand on the one hand and on NOM residuals on the other handafter coagulation and settling in Jar tests are shown in Figure 3 for synthetic water containing25mg/L NOM. Table 1 presents the results for experiments carried out on Nyong River watercontaining 20mg/L NOM.pH0,04,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0CCC(mg/L)050100150200250300350pH0.04.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0024681012ResidualNOM(mg/L)Figure 3. Variation of the ccc of coagulant (a) and residual NOM (b) with initial solution pHfor coagulation (25±2 °C) of 25 mg/L NOM water with Al sulphate (●); Fe (III) sulphate (○).(Added coagulant concentration is that at CCC)(b)(a)
  8. 8. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME125Table 1: Coagulation of Nyong River water, CameroonRaw watercharacteristics(in rainy season)pHinitialAl sulphate coagulation Fe sulphate coagulationCCC(mg/L)Residual NOM(mg/L)CCC(mg/L)Residual NOM(mg/L)pH 5.9±0.2Alkalinity 9.1±0.5(mg/L)NOM 20.0±0.6(mg/L)Turbidity 3.8±0.4(NTU)Colour 180±20Pt/Co (mg/L)Temperature24±0.5°C5.0±0.15.5±0.16.0±0.16.5±0.17.0±0.144.5959.3159.8257.8089.194.822.692.682.742.3629.9838.7337.6739.0559.903.402.102.201.951.70It can be seen that the graph of figure 3a, ccc vs pHi, has three segments defined bypHi for both Al and Fe(III) sulphate coagulation. First, when pHi is in the range 4.5 – 5.5,coagulant demand rises gently with increase in pH, 54.33mg/L per pH unit for Al and29.63mg/L per pH unit for Fe(III). Then for pHi between 5.5 and 6.5, coagulant demandroughly stabilizes, with a difference per pH unit of only 11.38 mg/L for Al and 5.88 mg/L forFe (III). Finally, for pHi in the range 6.5 – 7.5, coagulant demand rises rapidly with increasein pH, being per pH unit, 128.00 mg/L for Al and 78.45mg/L for Fe (III). The trends seen infigure 3a were also observed in coagulation experiments on other synthetic waters containingNOM at concentrations of 15mg/L and 40mg/L except that the values of ccc were lower for15mg/L NOM and higher for 40mg/L NOM. The results shown in table 1 for raw river waterconfirm the stabilization of ccc for pHi in the range 5.5 – 6.5. Throughout the pHi rangeinvestigated in figure 3a, coagulant demand is much higher for Al than for Fe (III), and onaverage 100% higher; the ccc value for Al sulphate is at least two times that of Fe (III)sulphate. The higher coagulant demand involved in using Al sulphate is confirmed in table 1even though the difference between the two coagulants is around 50% only for thecoagulation of raw Nyong River water. Figure 3b also presents the variation of NOMresiduals against pHi for synthetic water containing 25mg/L NOM. The curves shows threesegments for Al sulphate coagulation whereas Fe (III) sulphate coagulation has only twosegments for NOM residuals vs pHi. Both coagulants show the same trend for NOM residualsvs pHi for pHi below 5.5. Above this value of pHi, the curve for Al has two segments, onewith almost constant (4.0 + 0.5 mg/L) but slightly increasing NOM residuals for pHi between5.5 and 6.5, and one with falling NOM residuals at pHi above 6.5; whereas for Fe (III), at pHiabove 5.5, there is only one segment, with roughly constant NOM residuals generally lessthan 1mg/L. In general, Al sulphate coagulation leaves higher NOM residuals than does Fe(III) sulphate coagulation. This is corroborated by the results shown in table 1 on raw riverwater but the difference in NOM residuals arising from the use of the two coagulants is muchsmaller. Furthermore, table 1 also confirms the stabilization of NOM residuals for the twocoagulants for the condition, 5.5 < pHi > 6.5.The trends shown by the results displayed in Figure 3 for synthetic water as well asthose in table 1 for raw Nyong river water, when considered together, enable establishment of
  9. 9. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME126an optimum pH region for minimization of both coagulant demand and NOM residualsduring coagulation of soft, low alkalinity, low mineralization and low turbidity water undervariable pH conditions where, pH is preset but allowed to vary by dropping after coagulantaddition. The optimum value of pHi is between 5.5 and 6.5. Both coagulant demand (ccc) andNOM residuals are roughly constant in this pHi range. This corresponds to a final pH insolution (pHf) of 4.5 – 5.0 for Al sulphate and 4.5 – 5.5 for Fe (III) sulphate. For pHi below5.5, NOM residuals are high and for pHi above 6.5, NOM residuals are minimized but at theexpense of steeply increasing coagulant demand. The optimum pH range is not affected bythe presence of particulate turbidity as shown by the results on raw river water (table 1). Thisis in agreement with Edwards and Amirtharajah [40], who studied removal of colour causedby humic acids using alum and found that added turbidity had no effect on the optimumcoagulation domain for medium to high humic acid water. Eikebrokk[25], Cheng et al, [41]studied the removal of humic substances from soft, low alkalinity and low turbidity watersby Al coagulants and stated that Al was minimized at pH 5 – 7 while NOM residuals wereminimized at pH 5 – 6.The results of ccc vs pH and NOM residuals vs pH as seen above show that Fe (III) isa better coagulant than Al both in terms of coagulant demand and NOM residuals. This resultis in accord with Black et al. [42] who pointed out that Fe (III) sulphate was more than twotimes as effective as Al sulphate in the coagulation of colour causing organic compounds inwater. In the optimum pHi domain (5.5 – 6.5) where ccc is stabilized in our studies,consideration of coagulant demand in terms of metal ion to NOM mass ratio (mg metal ion /mg NOM) gives for Al, 0.27± 0.01, and for Fe (III), 0.46 ± 0.02 for synthetic water; but forraw river water, 0.24 mg Al / mg NOM and 0.50 mg Fe (III) / mg NOM. Outside theoptimum pH domain for ccc, it is found that NOM residuals increase below 0.22 mg Al/ mgNOM and 0.4 mg Fe (III) / mg NOM; also, at pHi = 6.5 – 7.5, mg Al / mg NOM rises from0.27 to 0.49 and mg Fe (III) / mg NOM rises from 0.46 to 0.91 but NOM is still minimized.Thus, NOM residuals are minimized at mg Al /mg NOM values of 0.27 – 0.49 and at mg Fe(III) / mg NOM values of 0.46 – 0.91. These results are in general agreement with those inthe literature. In the case of Al, Eikebrokk [25] for soft, low alkalinity Norwegian watersreported dosage requirements of 0.29 – 0.56 mg Al / mg NOM to minimize NOM and metalresiduals and Jekel [43] suggest a minimum dosage 0.4 – 0.5 mg Al/mg NOM for low Alresiduals and best removal of NOM. As regards Fe (III), the results of Cheng [44],recalculated along these lines give mg Fe / mg NOM values of 0.34 – 0.57 for good NOMremoval with minimization of residuals.In fact in the optimum range of initial pH of water, pHi 5.5 – 6.5, determination ofNOM residuals after Jar Test coagulation experiments show that NOM removal was as highas 97.58% for Fe (III) as against 86.69% for Al. The difference in behaviour of Al and Fe(III) sulphate with respect to pH depression during coagulation, coagulant demand and NOMresiduals, points to the possibility that in at least part of the pH range explored in thesestudies, some of the types of species involved in NOM coagulation by the two metals aredifferent.3.2. Column Settling of Metal-NOM FlocsFigure 4 represent the column settling results obtained on the humic synthetic watertested and Figure 5 shows those obtained on raw natural humic water (Nyong River). Thevalues of pHi at which experiments were performed on synthetic waters were selected on the
  10. 10. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME127basis of the results obtained for NOM residuals, with pH 5 (Fig.3b) being in the zone of highNOM residuals and pH 6 (Fig. 3b) in the zone of minimum NOM residuals.It can be seen from Figure 4 for synthetic waters that, for both coagulants, the graphsshow the same trends generally. Metal - NOM floc removal is low (< 20%) and independentof HLR at HLR higher than 5 m/hr. For HLR less than 2 m/hr, NOM removal increasesrapidly with decreasing HLR. Comparison of Figure 4a and b shows Metal – NOM flocremoval is generally higher for Al coagulated NOM than for Fe (III) coagulated NOM. Thisimplies that Al – NOM flocs have higher settling rates compared to Fe (III) – NOM flocs. Atall values of HLR investigated, Fe (III) – NOM flocs seem to have the same settling rates atpH 5 and pH 6, as indicated by results of NOM removal. The absence of a pH effect on flocsettling points to the possibility that in the conditions prevailing in these studies, thestructuring of aggregates, Fe (III) – NOM flocs, deriving from Fe (III) coagulation of NOMin these waters occurs by the same mechanism at both pH values. This situation is similar tothat applicable in the mechanism of coagulation which is complexation – chargeneutralization at both pH values, as suggested earlier. However, Vilgé – Ritter et al. [45]stated that floc structuring in Fe (III) coagulation of NOM in natural waters (Lake Ribou andSeine River) was pH dependent as indicated by the fractal dimensions of aggregates and wascontrolled by the nature of the organic matter. It is suggested that the nature of the tropicalriver organic matter may be responsible for the absence of a pH effect in our case. RegardingAl –NOM flocs, settling rates are similar at both pH values for HLR less than 2 m/hr, but athigher HLR, pH has a small but weak effect with pH 6 giving slightly higher floc settling(removal) rates than pH 5. This could be due to a small difference in coagulating speciespresent, in view of the suggestion made earlier that at pH 5, Al coagulates NOM purely bycomplexation – charge neutralization while at pH 6, although complexation – chargeneutralization is predominant. At higher HLR shearing would tend to reduce floc aggregatesizes and hence cause slower settling rates.The results presented in Figure 5, pertaining to Nyong River water (pH 5.9), confirmthe metal cation effect on settling rates. At all HLR values greater than 2m/hr, the percentremoval of Al-NOM flocs is higher than that of Fe(III)-NOM flocs, indicating that Al-NOMflocs have a higher settling rate compared to Fe(III)-NOM flocs for HLR above 2m/hr. Butfor HLR below 2m/hr, the difference between the two coagulants reduces rapidly so thatpercent removal and hence settling rates are comparable at HLR below 1m/hr. As in Figure 4,the percent removal of NOM, for both coagulants and for HLR greater than 5m/hr, is quitelow (< 30%). Interestingly, for greater than 50% NOM removal, it is necessary to operate atHLR values less than 2m/hr, that is, at quite low HLR as observed in practice at the NyongRiver water treatment plant. NOM removal values of 80% or more can be achieved at HLRvalues of 1m/hr or less. The need for very low HLR values to achieve high percent removalof NOM points to low density of metal – NOM flocs in the absence of significant turbidity inthese waters. It has been observed from the results presented above that, Al – NOM flocsgenerally settle faster than Fe (III) – NOM flocs, the difference being more evident for NyongRiver water.
  11. 11. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME128Figure 4 Effect of hydraulic loading on the removal of NOM from 25 mg/L NOM watercoagulated at pH 5 (a) and pH (6.5) (b) with Al sulphate (●) and Fe (III) sulphate (○)Figure 5 Effect of hydraulic loading on NOM removal from Nyong river raw water (20mg/L NOM) coagulated at pH 5.9 with Al sulphate (●) and Fe (III) sulphate (○)(b)Hydraulic loading rate (m/hr)0 5 10 15 20 25 30%NOMremoval020406080100(b)Hydraulic Loading rate (m/hr)0 5 10 15 20 25 30%NOMremoval020406080100(a)Hydraulic loading rate (m/hr)0 2 4 6 8 10 12 14%NOMRemoval020406080100
  12. 12. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME1293.3. Coagulation MechanismsReaction of NOM with Al and Fe (III) depends on the metal hydrolysis speciespresent and the degree of dissociation of carboxyl and phenolic groups of the NOM. Thehydrolysis of Al and Fe (III) in the presence of humic acids and NOM is limited by thepresence of the latter but still dependent on solution pH. Masion et al. [46] studied thehydrolysis of Al in the presence of small organic acid ligands using small angle X –rayscattering and found that speciation of Al in the pH range of 3 – 8 was limited to monomers,dimers and small oligomers (trimers). Further, Vilgé-Ritter et al. [45] found that Fe (III)hydrolysis in the presence of lake and river NOM at pH 5.5 and 7.5 was limited to smalloligomers (trimers) due to hindrance of the hydrolysis of Fe (III) by NOM. Rose et al. [47]who studied Fe speciation in NSIMI river water from the same tropical forest zone as NyongRiver in Cameroon, stated that NOM hindered Fe hydrolysis even at pH 6, Fe was poorlypolymerized due to complexation with NOM which blocks the Fe growth sites. Hence thepresence of NOM prevents formation of polymers and oxyhydroxides of Fe at pH below 7.5.In view of the foregoing, it can be inferred that the mechanisms of coagulation of Al and Fe(III) sulphate with NOM as suggested by the trends in Fig. 3 are as follows.In the case of Al sulphate, there are three segments. For pHi 5.5 – 4.5 correspondingto final pH (pHf) 4.3 – 4.5, reaction between positively charge Al ionic species (monomers,dimers and trimers) and ionized carboxyl of NOM occurs by electrostatic attraction followedby complexation –charge neutralization. The effective pH (4.3 – 4.5) is too low for theformation of a significant amount of Al hydroxide and it is also known that at least 80 % ofthe carboxyl groups are ionized in this pHf range [19,48]. The high level of residuals in thispH range is probably due to soluble metal NOM complexes and non-complexed NOM. In therange of pHi 5.5 – 6.5 (pHf 4.5 – 5.0), the amount of NOM residuals roughly stabilizes asdoes the ccc. This could be due to the fact that most of the NOM carboxyl groups are ionized.However, in this pHf range, dimer and trimer, aluminum species with different numbers ofsulfate ions also forms [48-50] and so there is formation of Al – NOM complexes by reactionof NOM with various Al ionic species. Coagulation probably occurs by complexation chargeneutralization. At pHi 6.5 – 7.7 (pHf 5.0 – 5.4) all the carboxyl groups of NOM and some ofthe phenolic groups have ionized and dimeric and trimeric aluminium species has becomepredominant. In this region, coagulation also occurs principally by complexation-chargeneutralization. The fall in NOM residuals as pHi increases above 6.5 in Figure 3 is consistentwith the total ionization of carboxyl and partial ionization of phenolic groups of NOM. Asregards removal of NOM by Fe (III) sulphate, Figure 3b shows only two segments. NOMresiduals increase as pH falls (pHi 5.5 – 4.5, pHf 4.5 – 3.9). In this pH zone, coagulationoccurs by complexation – charge neutralization, similar to the case of Al sulphate. Increase inNOM residuals as pH falls in this zone is probably due to the formation of soluble Fe (III) –NOM species. Indeed, from studies of coagulation of humic acid by Fe (III) salts usingfluorescence quenching, Cheng [44] stated that under low pH conditions, dissolvedcomplexes of Fe (III) – organic matter were found in solution. For pHi 5.5 – 7.5, that is at pHf4.5 – 6.3, the amount of NOM residuals is constant and independent of pH, thus suggestingstrongly that in this pH range, the mechanism of NOM removal by Fe is independent of pH.Since, as stated above, there is no evidence of Fe (III) hydroxide in the presence of NOM[45, 47], coagulation occurs by complexation – charge neutralization. It is clear thatthroughout the pH range (pHi 4.5 – 7.5, pHf 3.9 – 6.3) explored in our studies, removal ofNOM by Fe (III) takes place by complexation – charge neutralization. The size of humicsubstances is also known to depend on pH: a stretched configuration occurs at alkaline pH
  13. 13. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME130due to intramolecular electrostatic repulsions, whereas small humic aggregates can be formedbelow pH 5 [51, 52].Thus the aggregation of humic acid with hydrolyzed-Fe species can beascribed as the restructuration of flexible humic network during association with coagulantspecies. As shown in recent electron energy loss spectroscopy and pyrene fluorescenceexperiments [29, 22], conformational rearrangements can also be evidenced during thecoagulation of negatively charged humic colloids with cationic hydrolyzed metal species.4. CONCLUSIONSStudies of NOM coagulation by Al and Fe (III) sulphate in soft, low alkalinity, lowturbidity high NOM containing tropical water using Jar test show that coagulant demand ismuch higher (twice or more) for Al compared to Fe (III) salt. Whereas, column settlingexperiments show that Al gives better floc settling rates than Fe (III) except at very lowhydraulic loading rates. The difference between Al and Fe (III) sulphate in terms of coagulantdemand (ccc) and removal of NOM from soft, low alkalinity and low turbidity humic waterobserved is due to differences in metal hydrolysis species present in the pHi range concerned.In the pHi zone where NOM residuals are minimized (pHi > 5.5) Al and Fe (III) coagulatesNOM by complexation – charge neutralization.REFERENCE[1] Thurman. E. M., (1985), ‘‘Organic geochemistry of natural waters’’, MartinusNijhoff and W. Junk Publishers, Dordrecht, p. 497.[2] Zumstein. J., Buffle. J., (1989), ‘‘Circulation of pedogenic and aquagenic organicmatterin an eutrophic lake’’, Water Res., Vol. 23, pp. 229–239.[3] Peuravuori. J., (2005), ‘‘NMR spectroscopy study of freshwater humic material inlight of supramolecular assembly’’, Environ. Sci.Technol. Vol. 39, pp. 5541–5549.[4] Malcolm. R. L. (1985), Geochemistry of stream fulvic and humic substances, in: G.R. Aiken, D. McKnight. M., Wershaw. R. MacCarthy. L., P. (Eds.), Humic substancesin soil, sediment and water. Geochemistry, isolation and characterisation, Wiley-Interscience, New York, pp. 181-210.[5] Steelink. (1983), ‘‘Humic substances in soil, sediment, and water’’, John Wiley andSons Inc., New York, pp. 457-476,[6] Jekel. M. R., and Heinzmann. B., (1989), ‘‘Residual aluminium in drinking watertreatment’’, J. Water SRT-Aqua., Vol. 38, pp. 281 – 288.[7] Bernhardt. S. H., (1993), ‘‘Control of flocculants by use of streaming current detector,J. Water SRT-Aqua, Vol. 42, p. 239.[8] Owen. D.M., et. al. (1995), ‘‘NOM characterization and treatability’’, J. AWWA, Vol.87, pp. 46–63.[9] Lin. C.F., Lin. T.Y., Hao. O.J., (2000), ‘‘Effects of humic substance characteristics onUF performance’’, Water Res. Vol.34, pp.1097–1106.[10] Shon. H.K., et. al.(2009), ‘‘A study on the influence of ionic strength on the elutionbehaviour of membrane organic foulant using advanced separation tools’’,Desalination and Water Treatment, Vol.11, pp. 38–45.[11] Suffet. I. H., MacCarthy P., (1989), ‘‘Aquatic humic substances influence on fate andtreatment of pollutants’’, American Chemical Society, Washington DC,.
  14. 14. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME131[12] Countway. R. E., Dickhut. R. M., Canuel. E.A., (2003), ‘‘Polycyclic aromatichydrocarbon (PAH) distributions and associations with organic matter in surfacewaters of the York River’’, VA Estuary. Org. Geochem. Vol. 34 , pp. 209-224.[13] Edwards. G. A., Amirtharajah. A., (1985), ‘‘Removing color caused by humic acids’’,J. Am. Water Works Assoc. Vol.77, pp.50-57.[14] Hundt. T.R. and O’Melia. C.R. (1988), ‘‘Aluminium fulvic acid interactions:mechanisms and applications’’, J. Am. Water Works Assoc., Vol. 80, p.176.[15] van Benschoten. J.E., Edzwald. J.K., (1990), ‘‘Chemical aspects of coagulation withaluminium chloride’’, Water Res. 24, pp.1527 – 1537.[16] Matilainen. A., Lindqvist. N., Tuhkanen. T., (2005), ‘‘Comparison of the efficiencyof aluminium and ferric sulphate in the removal of natural organic matter duringdrinking water treatment process’’, Environmental Technology, Vol. 26, pp. 867-875.[18] Mazet. M., Angbo. L., Serpaud. B., (1990), Adsorption of humic substances onpreformed aluminium hydroxide flocs, Water Res., Vol. 24, pp.1509-1518.[19] Edzwald. J.L., (1993), ‘‘Coagulation in drinking water treatment: particles, organicsand coagulants’’, Water Sci. Technol., Vol. 27, pp.21-35.[20] Masion. A., Vilgé-Ritter. A., Rose, J., Stone, W. E. E., Teppen Brian, J., D.Rybacki,J.-Y. Bottero, (2000), ‘‘Coagulation-flocculation of natural organic matter with Alsalts: speciation and structure of the aggregates’’, Environ. Sci. Technol., Vol. 34,pp.3242-3246.[21] Narkis. N., Rebhun. M., (1983), Inhibition of flocculation processes in systemscontaining organic matter, J. WPCF Vol.55, pp. 947-954.[22] Siéliéchi. J.M., et al. (2008), ‘‘Changes in humic acid conformation duringcoagulation with ferric chloride: Implications for drinking water treatment’’, WaterRes., Vol. 42, pp.2111-2123.[23] Zhang. P., et al. , (2008), ‘‘Coagulation characteristics of polyaluminum chloridesPAC-Al30 on humic acid removal from water’’, Separation and PurificationTechnology, Vol. 63, pp. 642–647.[24] Bose. P. and Reckhow. D.A., (1998), ‘‘Adsorption of natural organic matter onpreformed aluminum hydroxide flocs’’, J. Environ. Eng., Vol. 124, pp. 803–811.[25] Eikebrokk. B., (1999), ‘‘Coagulation-direct filtration of soft, low alkalinity humicwaters’’, Water Sci. Technol., Vol. 40, pp. 55-62.[26] Eikebrokk. B., (1996), ‘‘Removal of humic substances by coagulation’’, in H.H.Hahn, E. Hoffmann, and H. Odegaard (Eds.), Chemical Water and WastewaterTreatment V, Springer Verlag, Berlin, pp.173 – 187.[27] Crozes. G., White P., Marshall M., (1995), ‘‘Enhanced coagulation. Its effect onNOM removal and chemical costs’’, J. Am. Water Works Assoc. Vol. 87, pp.78 – 89.[28] Jung. A.V., et al., (2005), Coagulation of humic substances and dissolved organicmatter with a ferric salt: An electron energy loss spectroscopy investigation, WaterRes., Vol. 39, pp.3849–3862.[29] Kazpard. V., et al. (2006), ‘‘Fate of coagulant species and conformational effectsduring the aggregation of a model of a humic substance with Al13 polycations’’, WaterRes., Vol. 40, pp.1965 – 1974.[30] Yang. Z.L., Gao. B.Y., Yue. Q.Y., Wang. Y., (2010), ‘‘Effect of pH on thecoagulation performance of Al-based coagulants and residual aluminum speciationduring the treatment of humic acid–kaolin synthetic water’’, Journal of HazardousMaterials, Vol. 178, pp.596 – 603.
  15. 15. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME132[31] Duan. J. and Gregory J., (2003), ‘‘Coagulation by hydrolysing metal salts’’, Adv.Colloid Inter. Sci. Vol.100-102, pp. 475-502.[32] Yan. M., et al. (2006), ‘‘Enhanced coagulation in a typical North-China watertreatment plant’’, Wat. Res., pp.3621 – 3627.[33] Krishnan. P., ‘‘Column settling tests for flocculant suspension Discussion’’. J.Environ. Eng. Div. ASCE 102 (1976), pp. 227 – 229.[34] Yao. K.M., (1979), ‘‘Column settling test and tube settling’’, J. Am. Water WorksAssoc., Vol. 71, pp. 109 – 112.[35] Zanoni. A.E., Blomquist. M.W, (1975), ‘‘Column settling tests for flocculantsuspensions’’, J. Environ. Eng. Div. ASCE. 6, pp.309 – 318.[36] Jackson. (1968), ‘‘Soil Chemical Analysis-Advanced Course’’, 4thedition.Department of Soil Science, University of Wisconsin, Madison.[37] Degremont. (1989), Mémento technique de l’eau, Technique et Documentation,Lavoisier, Paris, pp. 23-38.[38] Korshin. G. V., Li. C.W., Benjamin. M.M., (1997), ‘‘Monitoring the properties ofnatural organic matter through UV spectroscopy: A consistent theory’’, Water Res.Vol. 31, pp.1787-1795.[39] Zanoni. A.E., Blomquist M.W., (1975), ‘‘Column settling tests for flocculantsuspensions’’, J. Environ. Eng. Div. ASCE. 6, pp.309 – 318.[40] Edwards G. A., Amirtharajah A., (1985), Removing color caused by humic acids. J.Am. Water Works Assoc., Vol. 77, pp.50-57.[41] Cheng. W.P., Chi. F.H., Li. C.C., Yu. R.F., (2008), ‘‘A study on the removal oforganic substances from low-turbidity and low-alkalinity water with metal-polysilicate coagulants’’, Colloids Surf. A, Vol.312, pp.238–244.[42] Black. A.P., Singley. J.E., Whittle, G.P., Maulding, J.S. (1963), ‘‘Stoichiometry of thecoagulation of color-causing organic compounds with ferric sulphate’’, J. Am. WaterWorks Assoc., Vol. 10, pp.1347 – 1366.[43] Jekel. M. R., (1986), ‘‘Interactions of humic acids and aluminum salts in theflocculation process’’, Water Res., Vol. 20, pp.1535-1542.[44] Cheng .W. P., (2002), ‘‘Comparison of hydrolysis/coagulation behavior of polymericand monomeric iron coagulants in humic acid solution’’, Chemosphere, Vol. 47, pp.963-969.[45] Vilgé-Ritter. A., Rose., J., Masion. A., Bottero. J.-Y., Lainé. J.-M., (1999),‘‘Chemistry and structure of aggregates formed with Fe-salts and natural organicmatter’’, Colloids Surf. A: Physicochem. Eng. Aspects, Vol. 147, pp.297-308.[46] Masion. A., Bottero. J.Y., Thomas. F., Tchoubar. D., (1994), ‘‘Chemistry andstructure of Al(OH)/organic precipitates. A small-Angle X-ray scattering study. 2.Speciation and structure of the aggregates’’, Langmuir, Vol. 10, pp. 4349 – 4352.[47] Rose. J., Vilgé. A., Olivier-Lauquet. G., Masion. A., Frechou. C., Bottero. J. Y.,(1998), ‘‘Iron speciation in natural organic matter colloids’’, Colloids Surf. A:Physicochem. Eng. Aspects, Vol. 136, pp.11-19.[48] Lu. X., Chen. Z., Yang. X., (1999), ‘‘Spectroscopic study of aluminium speciation inremoving humic substances by Al coagulation’’, Water Res., Vol. 33, pp.3271 – 3280.[49] Urabe. T., Tsugoshi. T., Tanaka. M., (2008), ‘‘Electrospray ionization massspectrometry investigation of the blocking effect of sulfate on the formation ofaluminum tridecamer’’, Journal of Molecular Liquids 143, pp.70–74.
  16. 16. International Journal of Advanced Research in Engineering and Technology (IJARET), ISSN0976 – 6480(Print), ISSN 0976 – 6499(Online) Volume 4, Issue 4, May – June (2013), © IAEME133[50] Wang. Y., et al. (2009), ‘‘Characterization of floc size, strength and structure invarious aluminum coagulants treatment’’, Journal of Colloid and Interface Science,Vol. 332, pp. 354–359.[51] Lead. J.R., Wilkinson. K.J., Starchev. K., Canonica. S., Buffle. J., (2000),‘‘Determination of diffusion coefficients of humic substances by fluorescencecorrelation spectroscopy: role of solution conditions’’, Environ. Sci. Technol., Vol.34, pp.1365–1369.[52] Plaschke. M., Romer. J., Klenze. R., Kim. J.I.,. (1999), ‘‘In situ AFM study of sorbedhumic acid colloids at different pH’’, Colloids Surf. A Physicochem. Eng. Aspects,Vol. 160, pp. 269–279.[53] P.S. Thué, J. M. Siéliéchi, P.P. Ndibewu and R. Kamga, “Physico-Chemical Studieson The Adsorption of Atrazin on Locally Mined Montmorillonites”, InternationalJournal of Advanced Research in Engineering & Technology (IJARET), Volume 4,Issue 1, 2013, pp. 79 - 95, ISSN Print: 0976-6480, ISSN Online: 0976-6499.[54] R Radhakrishanan and A Praveen, “Sustainability Perceptions on WastewaterTreatment Operations in Urban Areas of Developing World”, International Journal ofCivil Engineering & Technology (IJCIET), Volume 3, Issue 1, 2012, pp. 45 - 61,ISSN Print: 0976 – 6308, ISSN Online: 0976 – 6316.