Arsenic removal during conventional aluminium-based drinking-water treatment_2.pdf

PII: S0043-1354(00)00424-3 Wat. Res. Vol. 35, No. 7, pp. 1659–1664, 2001 # 2001 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0043-1354/01/$ - see front matter ARSENIC REMOVAL DURING CONVENTIONAL ALUMINIUM-BASED DRINKING-WATER TREATMENT JAN GREGOR* Institute of Environmental Science and Research Limited, PO Box 29-181, Christchurch, New Zealand (First received 21 March 2000; accepted in revised form 25 August 2000) Abstract}The changing forms and concentrations of arsenic
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Wat. Res.  Vol. 35, No. 7, pp. 1659–1664, 2001 #  2001 Elsevier Science Ltd. All rights reservedPrinted in Great Britain0043-1354/01/$-see front matter PII: S0043-1354(00)00424-3 ARSENIC REMOVAL DURING CONVENTIONALALUMINIUM-BASED DRINKING-WATER TREATMENT JAN GREGOR* Institute of Environmental Science and Research Limited, PO Box 29-181, Christchurch, New Zealand (First received 21 March 2000; accepted in revised form 25 August 2000) Abstract } The changing forms and concentrations of arsenic through aluminium-based coagulationtreatment processes were tracked for three drinking-water treatment plants. This has provided directevidence of where and how arsenic is removed. In general, soluble As(V) is converted to particulate As(V)by adsorption during rapid mixing, and is removed along with naturally occurring particulate arsenicpredominantly by clarification. Soluble As(III) tracks through the treatment processes and is converted tosoluble As(V) during final chlorination. The ability of a water treatment process to achieve the maximumacceptable concentration for arsenic in drinking water is dependent on the concentration of As(III) in thesource water. #  2001 Elsevier Science Ltd. All rights reserved Key words } arsenic, forms, removal, aluminium, coagulation INTRODUCTION For decades, arsenic has been fortuitously removedfrom contaminated waters during the coagulationprocess, generally to a level acceptable for drinking-water purposes. Previously in New Zealand, theacceptable value had been 0.05mg/L, which very fewsources of drinking-water exceeded and thus the needfor specific treatment was unwarranted. The currentUS-EPA maximum contaminant level (MCL) is0.05mg/L (USEPA, 1998), but a value of 0.005mg/L is proposed (USEPA, 2000). In 1995, New Zealandadopted a maximum acceptable value (MAV) forarsenic in drinking-water of 0.01mg/L. This isconsidered to be the concentration below which thepresence of arsenic does not result in any significantrisk to a consumer over a lifetime of consumption(Ministry of Health, 1995). With the lowering of theacceptable concentration, specific treatment forarsenic removal has become necessary.Arsenic exists in natural waters in a variety of forms including the trivalent (As(III)) and pentava-lent (As(V)) oxidation states, and soluble, particulateand organic-bound. Effective treatment must targetall of these forms, either directly or after conversionto a more readily removed form. The proportions of these forms depend on factors such as pH, particleand natural organic matter concentrations andcompositions, biological activity, and degree of oxygen saturation, and thus will be spatially andseasonally dependent. Chen  et al  . (1999) havecharacterised the forms of arsenic in source watersamples, and Hering  et al  . (1997) have carried outbench-scale studies with As(III) and As(V) to identifythe characteristics of source waters that havepronounced effects on arsenic removal.By far the most commonly used treatment in NewZealand is aluminium-based coagulation with disin-fection by chlorination. For river waters withsignificant inherent arsenic concentrations, the degreeto which this treatment removes arsenic is howeversomewhat variable, as gauged by field measurementof total arsenic concentrations before and aftertreatment (Davies  et al  ., 1994; McLaren and Kim,1995). At best, the 0.01mg/L MAV was achieved formost of the year, but seasonal arsenic concentrationfluctuations in the river water were reflected in thetreated water. At worst, less than 15% of the arsenicwas removed. Similarly in the USA, McNeill andEdwards (1995) reported a wide range in decreases insoluble As(V) concentrations for five alum coagula-tion treatment plants (6–74%).In this study, the changing forms and concentra-tions of arsenic through aluminium-based coagula-tion treatment processes were tracked for threedrinking-water treatment plants that abstract waterfrom the same river over a distance of no more than40km. The findings provide insight into the criticalsteps of coagulation, and identify when treatmentadditional to aluminium-based coagulation is re-quired in order to meet the MAV for arsenic. *Author to whom all correspondence should be addressed.Tel.: +64-3-3516-019; fax: +64-3-3510-010; e-mail: jan.gregor@esr.cri.nz1659  MATERIALS AND METHODOLOGY Source water Arsenic concentrations in the Waikato River, one of New Zealand’s major sources of drinking-water, range fromapproximately 0.01mg/L at its source (Lake Taupo) to0.03–0.06mg/L after natural and anthropogenic geothermaldischarges. The three drinking-water supplies selected forthis study (Hamilton City, Pukerimu Rural, Lake Karapiro)draw essentially the same quality water (Table 1, duplicatesamples were collected on different days in the same week)and use aluminium-based coagulation. The one notableexception to water quality consistency is the higher pH atLake Karapiro.Both Hamilton City and Pukerimu Rural water suppliesabstract water from the run of the river. The Lake Karapirowater supply is different, being from an artificial lakeformed on the river. Lake Karapiro is the last of eighthydroelectric lakes formed on the Waikato River with anarea of 7.7km 2 and a retention period of 2.8 days(Magadza, 1979). Treatment plants and sampling sites The Hamilton City water supply serves approximately100,000 people, whereas Pukerimu Rural and Lake Kar-apiro serve small communities of 3000 and 10,000 peoplerespectively. Alum is used as the coagulant at bothHamilton City and Lake Karapiro, and polyaluminiumchloride is used at Pukerimu Rural. All three treatmentplants use a synthetic flocculant aid. Lake Karapiro pre-chlorinates prior to rapid mixing. Both Hamilton City andLake Karapiro pH-adjust upwards immediately post-filter.Hamilton City adjusts from approximately pH 6.8 to 8.0,and Lake Karapiro adjusts from approximately pH 7 to8.0–8.5 in the winter and to pH 7.5 in the summer.Five sampling sites were identified for each watertreatment plant: source water, during the initial stages of coagulation (rapid mix micro-floc formation, 2 to 3minmixing), after flocculation/clarification (macro-floc develop-ment, 60–120min slow mixing), after filtration, and afterfinal chlorination. Sample collection and analysis At each treatment plant, and for each sampling site,water samples were collected twice during the summer andtwice during the winter to coincide with extremes of riverflow and composition, and the demand for treated water(treatment plant flow rate). The duplicate samples werecollected on different days in the same week.At each treatment plant, and for each sampling site, theconcentrations of total arsenic, soluble arsenic (andparticulate-bound arsenic, by difference), soluble As(III)(and soluble As(V), by difference), and natural organicmatter (NOM)-bound arsenic were determined. Totalarsenic was determined from an acid preserved sample byICP-MS, and soluble arsenic was determined from a field-filtered (0.45 m m) acid-preserved sample. Soluble As(III) wasdetermined from a field-filtered sample that was subse-quently eluted through an anion exchange column (Amber-lite CG-400 resin). The As(V) was retained by the resin, andthe non-ionic As(III) passed through. NOM-bound arsenicwas determined from dialysis against Milli-Q water througha Spectra/Por MWCO3,500 membrane for the source wateronly, with the NOM-bound arsenic being retained outsidethe dialysis bag and inorganic arsenic reaching equilibriuminside and outside the dialysis bag. Ultraviolet absorbance(UV) measurements indicated that only approximately 1%of the UV-active NOM was small enough to pass throughthe dialysis membrane. Details of this method, used foralginate bound-aluminium, have been reported previously(Gregor  et al  ., 1996). RESULTS The changing forms and concentrations of arsenicthrough the three aluminium-based coagulationtreatment processes are shown in Fig. 1 and Table2. The duplicate results for each sampling site(Source, Rapid Mix, Clarified, Filtered, and Chlori-nated) have been averaged. The agreement in resultsfor duplicate samples is considered reasonable, giventhe experiment scale of an operational water treat-ment plant. In the figure, the height of each barrepresents the total arsenic in the water. This is thensubdivided into particulate arsenic, soluble As(III),and soluble As(V).Arsenic has been removed to a concentration  5 m g/L during summer and winter at HamiltonCity and Pukerimu Rural, and during winter at LakeKarapiro. For all three treatment plants, betterpercent removals and lower concentrations in thetreated water were achieved in winter (range 5 1–3 m g/L) than in summer (range 3–10 m g/L).The total arsenic concentration in the WaikatoRiver in summer was 1.5 times higher than in winter.Regardless of season and except for Pukerimu Ruralwinter, the arsenic was predominantly soluble(85–100%), and except for Lake Karapiro summer,90% or more of the soluble arsenic was solubleAs(V). The predominance of soluble arsenic wasmimicked in the treated water. Between source and Table 1. Waikato river water qualitypH Turbidity (NTU) Absorbance at 270nm a Total arsenic ( m g/L) Hamilton city Summer 8.2, 7.6 3.0, 1.8 0.03, 0.03 27, 28Winter 7.2, 7.5 1.5, 1.4 0.05, 0.05 18, 18 Pukerimu rural  Summer 7.5, 7.3 1.0, 1.4 0.03, 0.02 29, 29Winter 7.5, 7.5 0.7, 0.5 0.03, 0.04 16 Lake Karapiro Summer 8.4, 8.5 2.5, 2.1 0.03, 0.03 25, 28Winter 8.7, 8.5 1.7, 1.0 0.04, 0.04 18, 18 a An estimate of natural organic matter content. Jan Gregor1660  treated water, the soluble As(V) component wasconverted to a particulate form and removedpredominantly by settling (clarification). The solubleAs(III) travelled through the treatment process untilchlorination, where it was mostly oxidised to As(V).Unlike Hamilton City and Pukerimu Rural, asignificant portion (22%) of the Lake Karapirosummer source sample was soluble As(III). Thissoluble component travelled through the remainingtreatment, resulting in treated water that struggled toachieve the MAV of 0.01mg/L. The oxidation stateof this soluble arsenic changed several times duringtreatment. It was converted to soluble As(V) duringrapid mixing due to pre-chlorination. Between rapidmixing and the exit of the clarifier, it was reduced tosoluble As(III), then finally converted back to solubleAs(V) during filtration and chlorination.Approximately 4 m g/L of NOM-bound arsenic wasfound in the three source waters (the only samplingsite for which sufficient NOM was present to makethis measurement), regardless of season. Given thatthe total arsenic concentration in summer was 1.5times higher than in winter, the fraction of totalarsenic bound to NOM was lower in the summer, co-incident with the lower NOM content as estimated byabsorbance at 270nm (Table 1). DISCUSSION Achieving the MAV  In addition to the intended purpose of particulateand natural organic matter (NOM) removal, alumi-nium-based coagulation has a role to play in reducingthe concentration of arsenic. The goal is to consis-tently reduce the concentration of arsenic to belowthe acceptable value for drinking water. Two of thethree water treatment plants in this study achieved Fig. 1. Changing forms and concentrations ( m g/L) of arsenic during treatment. Duplicate results havebeen averaged.Arsenic removal by aluminium coagulation 1661  this goal based on the New Zealand MAV of 0.01mg/L, and the third achieved it for part of theyear. However, with the possibility of the acceptableconcentration being lowered even further in thefuture, as it is proposed in the USA, it is importantto understand whether existing treatment processesare capable of doing better. Comparison of arsenic fractionation in New Zealand and other source waters The arsenic in the Waikato River srcinates fromnatural geothermal sources and from geothermalpower station discharges. The seasonal variation inconcentrations of arsenic in the Waikato River foundin this study was similar to that reported at theHamilton City intake in 1993/1994 (McLaren andKim, 1995). The samples of the present study weretaken in March and September 1999, and theconcentrations of 28–29 m g/L and 13–18 m g/Lmatched the 30 m g/L and 18–20 m g/L respectivelyreported by McLaren and Kim (1995) for the sametime of the year.It was found in this study that the dominant formof arsenic in the Waikato River is soluble. This is incontrast to the results of a recent USA nationalsurvey of drinking water sources which found forsurface waters typically 23–54% of the total arsenicto be particulate (Chen  et al  ., 1999). The highproportion of soluble arsenic in the Waikato Rivermay be related to its geothermal origin and theconsequential effect of water composition. Forexample, mono-silicic acid has been identified as thecomponent of geothermal bore water that inhibitsarsenic adsorption to onto hydrous ferric oxidesurfaces (Swedlund and Webster, 1998). An elevatedconcentration of mono-silicic acid in the WaikatoRiver could thus inhibit adsorption of arsenic ontoiron-bearing colloids and particles.The results of this study are in agreement withprevious studies that As(V) predominates in theWaikato River (Aggett and Aspell, 1980). That theproportion of As(III) increases in spring/summer in alake setting (Lake Karapiro) is also in agreementwith previous studies. This has been attributed tobiologically mediated reduction in the sediments(Aggett and Aspell, 1980; Kuhn and Sigg, 1993)and anoxic conditions and thermal stratification of the water trapping the reduced As(III) (Aggett andKriegman, 1988). An oxygen deficit at depth has beenreported for Lake Karapiro during summer months(Magadza, 1979), which will sustain the presence of As(III).  Anabaena oscillaroides , present in WaikatoRiver water, is also reported to reduce As(V) toAs(III) (Freeman, 1985). Taste and odour problemswere reported in the Lake Karapiro water supply atthe time of summer sampling. The chemicalsresponsible for these problems derive from bacteriaand algae present in the water or sediment, support-ing the presence of biological activity.In summary, the Waikato River water containsseasonally-variable concentrations of arsenic, con-tains very little particulate arsenic, and solubleAs(III) can contribute significantly to the totalarsenic concentration during spring/summer. Theconsequences for treatment effectiveness of the Table 2. Changing forms and concentrations ( m g/L) of arsenic during treatment. Duplicate results have been averagedSource Rapid mix Clarified Filtered Chlorinated Hamilton city summer Particulate As 0.6 24.8  5 0.5  5 0.5  5 0.5Soluble As(V) 25.5  5 0.5  5 0.5  5 0.5 4.4Soluble As(III) 1.6 3 4.3 4.8 0.5 Hamilton city winter Particulate As 2 16.4 0.8  5 0.5  5 0.5Soluble As(V) 14.8  5 0.5  5 0.5  5 0.5 1Soluble As(III) 0.8 0.6 0.8 0.6 0.6 Pukerimu rural summer Particulate As 4 25.4 2  5 0.5  5 0.5Soluble As(V) 24.2  5 0.5  5 0.5  5 0.5 2.7Soluble As(III) 0.6 1.8 2.6 3.1  5 0.5 Pukerimu rural winter Particulate As 4 16.4 1.5  5 0.5  5 0.5Soluble As(V) 8.2  5 0.5  5 0.5  5 0.5  5 0.5Soluble As(III) 0.5 0.8 1 1.4 0.5 Lake Karapiro summer Particulate As  5 0.5 21.6 2.2 0.8  5 0.5Soluble As(V) 23 4.8 1.8 6.4 8.4Soluble As(III) 5 0.6 6.4 2 0.6 Lake Karapiro winter Particulate As  5 0.5 15.4 1.8  5 0.5  5 0.5Soluble As(V) 17.4  5 0.5 1.4 1.7 2Soluble As(III) 0.6 0.5 0.5 0.5 0.5 Jan Gregor1662
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks