Wat. Sci. Tech. Vol. 40, No.2, pp. 69--76,1999
Published by ElsevierScienceLIdon behalfof lhe IAWQ Printedin GreatBritain. Allrightsreserved 0273-1223/99 $20.00 + 0.00
ARSENIC IN DRINKING WATERPROBLEMS AND SOLUTIONS T. Viraraghavan*, K. S. Subramanian** and 1. A. Aruldoss*** • Faculty ofEngineering, University ofRegina, Regina, Canada S4S OA2 .. Environmental Health Directorate, Health Canada, Ottawa, Canada ••• Stanley Associates Engineering Limited, Calgary, Canada
ABSTRACT The current United States maximum contaminant level for arsenic in drinking water is set at 50 Ilgll. Because of the cancer risks involved, Canada has already lowered the maximum contaminant level to 25 Ilgll; the United States Environmental Protection Agency is reviewing the current allowable level for arsenic with a view of lowering it significantly. Various treatment methods have been adopted to remove arsenic from drinking water. These methods include 1) adsorption-coprecipitation using iron and aluminum salts, 2) adsorption on activated alumina, activated carbon, and activated bauxite, 3) reverse osmosis. 4) ion exchange and 5) oxidation followed by filtration. Because of the promise of oxidation-filtration systems, column studies were conducted at the University of Regina to examine oxidation with KMnO. followed by filtration using manganese greensand and iron-oxide coated sand to examine the removal of arsenic from drinking water; these results were compared with the data from ion exchange studies. These studies demonstrated that As (111) could be reduced from 200 ug/l to below 25 Ilgll by the manganese greensand system. In the case of manganese greensand filtration, addition of iron in the ratio of 20: I was found necessary to achieve this removal. I!:> 1999 Published by Elsevier Science Ltd on behalf of the lAWQ. All rights reserved.
KEYWORDS Arsenic removal; drinking water; manganese greensand; iron-oxide coated sand.
INTRODUCTION Water is one of the major means of transport of arsenic in the environment. Arsenic in the aquatic environment is predominant in places with high geothermal activities. Soil erosion and leaching have been 8 reported to contribute to 612 and 2380 x 10 g/yr of arsenic in the dissolved and suspended form to the oceans, respectively (Mackenzie et al., 1979). Soil erosion and agricultural runoff are large contributors to the arsenic concentration in sediments. High arsenic levels (8.6-13.2 mg/g) have been reported to be associated with sediments and a potential exists that it may be released in hazardous amounts to the overlying waters (Viraraghavan et al., 1992). Industrial effluents are a major source of arsenic to the environment. Arsenic and arsenical compounds are found in effluents from the metallurgical industry, glassware and ceramic industries, dye and pesticide manufacturing industries, petroleum refining, rare earth industry and other organic and inorganic chemical industries. It finds application in the manufacture of herbicides and pesticides. Other industries using arsenic include wood and hide preservative; lead shot manufacture; phosphate detergent builder and presoaks used in many fertilizers (patterson, 1985).
T. VIRARAGHAVAN et al.
Environmental chemistry of arsenic Arsenic generally exists in the inorganic form in water supplies. Under different redox conditions, arsenic is stable in the +5, +3, -3 and 0 oxidation states. The pentavalent (+5) or arsenate species are AS04,3, HAs04'2, and H2As04-. The trivalent (+3) or the arsenite species include As(OH)3, As(OH)4-, As020H,2 and AS03,3. The pentavalent arsenic species are predominant and stable in the oxygen-rich aerobic environments, whereas the trivalent arsenite species are predominant in moderately reducing anaerobic environments such as groundwater (Ghosh and Yuan, 1987). The pH determines the predominant arsenate or arsenite species. The stability and the predominance of the arsenic species in the aquatic environment at different pH ranges is shown in Table I (Gupta and Chen, 1978). Table 1. Stability and predominance of arsenic species in varying pH ranges in the aquatic environment (Gupta and Chen, 1978) pH As(III) pH As(V)
0-9 H3As03 0-2 H3As04
13 HA S0 32.
7-11 H3As03 2.
Arsenic in its soluble form generally occurs in its +3 and +5 oxidation states. Both organic and inorganic forms of arsenic can be detected in natural water systems. Methylated or organic arsenic occurs at concentrations less than 1 J.lgll and is not of major significance in drinking water treatment (Edwards, 1994). Arsenic in its various chemical forms and oxidation states is released into the aquatic environment by natural erosion processes and industrial discharges. On release to the aquatic environment, the arsenic species enter into a methylationldemethylation cycle, while some are bound to the sediments or taken up by biota where they could undergo metabolic conversion to other organoarsenicals (Fowler, 1983). Several fungi and bacterial species have been demonstrated to methylate inorganic arsenic by an initial reduction to arsenite, and the addition of methyl groups (Fowler, 1983). Arsenic toxicology Arsenic has acquired an unparalleled reputation as a poison, with arsenic trioxide, a tasteless and odorless inorganic arsenic compound, constituting a convenient agent for homicide (Pershagen, 1983). High levels of arsenic (0.9 to 3.4 mg/l) in well waters in the Cordoba region of Argentina were responsible for 165 deaths. A high proportion of deaths were due to cancer of the respiratory system and gastrointestinal tract (pershagen, 1983). High arsenic levels have also been found in the water supplies in Chile and Ghana. The toxicity of arsenic is dependent on its oxidation state, chemical form and solubility in the biological media (Subramanian, 1988a). The toxicity scale of arsenic decreases in the order: arsine > inorganic As(III) > organic As(III) > inorganic As(V) > organic As(V) > arsonium compounds and elemental arsenic. The toxicity of As(III) is about ten times that of As(V) (Pontius et al., 1994). An acute high dose of arsenic by oral intake causes gastrointestinal irritation resulting in difficulty in swallowing, thirst, abnormally low blood pressure and convulsions. The lethal dose for adults has been noted to be 1-4 mg As/kg (Pontius et al., 1994). A recent study on cancer risks from arsenic in drinking water indicates that arsenic could cause liver, lung and kidney/bladder cancer in addition to skin cancer (Smith et aI., 1992). The study showed that the lifetime risk of dying from cancer of the liver, lung, kidney or bladder on consumption of 1Vday of water containing 50 J.lg/l of arsenic (current US Environmental Protection Agency (USEPA) maximum contaminant level in drinking water), could be as high as 13 per 1000 persons. In the United States, over 350,000 people may be drinking water containing more than 50 ug/l of arsenic and over 2.5 million people could be supplied with water having arsenic levels over 25 ug/l, Based on the average arsenic levels and water consumption patterns in the United States, the risk was estimated to be around 1 per 1000. A survey of 114 wells in the Tainan region of the southwest coast of Taiwan showed arsenic concentrations ranging from 0.6 to Z.O mg/l (Shen, 1973). Blackfoot disease, a peripheral disorder characterized by
Arsenic in drinking water
gangrene of the extremities, especially the foot, was the cause of 244 deaths . A chemical factory manufacturing several chemicals including the insecticide Paris-Green (acetocopper arsenite), was responsible for the contamination of wells in the southern part of Calcutta, India (Chatterjee et al., 1993). Over seven thousand people were consuming the arsenic-contaminated water for several years, but this fact remained unnoticed until September 1989. A few died, and some of the victims were hospitalized, while symptoms of arsenic poisoning were evident in many families living in the area. Water samples analyzed for arsenic indicated extremely high levels of contamination, with total arsenic concentrations ranging from as low as 0.002 to as high as 58 mgll. Arsenic in groundwater has been found above the maximum permissible limit (0.05 mgll) in six districts of West Bengal, India; these six districts have an area of 34,000 km2 with a population of 30 million (Chatterjee et al., 1995; Das et al., 1995). It was estimated, based on a survey of small areas of these affected districts, that at least 800,000 people could be drinking water high in arsenic 'with more than 175,000 people showing arsenical skin lesions that are. the late stages of manifestation of arsenic toxicity. The source of arsenic is geological. Most of the water samples contained a mixture of arsenite and arsenate and in none of them was methylarsonic or dimethylarsinic acid detected. A recent study of groundwater samples collected in an area of about 270 km 2 from Madras City, India, showed that the arsenic levels exceeded the maximum permissible limit over the entire city and a positive correlation of arsenic with other toxic metals showed that all these toxic elements are anthropogenic in origin (Ramesh et al., 1995). In 1984, two projects for improving the quality of drinking water for 100,000 people were completed in the Xinjiang Uighur Autonomous Region of China, an area of endemic arsenic toxicosis (Lianfang and Jianzhong, 1994). A follow-up study of the area showed that an improvement in the symptoms and signs of arsenic poisoning in humans had occurred as a result of the new drinking water sources; it is clear that the disease can be controlled by supplying arsenic-free water in areas of endemic arsenic toxicosis (Lianfang and Jianzhong, 1994). A survey conducted by the American Water Works Association (AWWA) for inorganic contaminants in water supplies in the United States revealed 34 violations for arsenic (maximum contaminant level (MCL) 0.05 mg/l), with concentration values ranging from 0.52 to 0.190 mgll and a mean concentration ofO.083g/l (American Water Works Association Committee, 1985). A combined AWWA-USEPA data base revealed 46 MCL violations. Most of the violations were reported in New Mexico , Texas and Oklahoma. Isolated cases of arsenic violations occurred in Alaska, North Carolina, New Hampshire, Virginia and D1inois. Arsenic contamination of well waters in Nova Scotia, Canada was first identified in 1976 when a resident of Waverly (Halifax county) was diagnosed as being intoxicated with arsenic . The well water in the resident's home was found to contain 5 mgll of arsenic, which was ten times the maximum acceptable concentration in Canada of 0.05 mg/l. A follow-up study of 198 wells in the Waverly area indicated that 34 wells were contaminated with arsenic above the maximum permissible limit (Grantham and Jones, 1977). A survey of ground water supplies in Halifax county revealed consistent MCL violations in various communities (Meranger et al., 1984). In at least 10% of the samples analyzed, the arsenic contamination was in excess of 500 J.lg/l, while 23% of the samples analyzed were found to contain arsenic higher than 250 J.lgll (Meranger et al., 1984). Guidelines for arsenic in drinking water Due to the carcinogenicity of some arsenic compounds, the objective should be to reduce its exposure to a level as close to zero as possible, taking into consideration its health effects and toxicology, occurrence and human exposure, availability and cost of the treatment technology, the practical quantitation limit of analytical techniques and the estimated risk for cancer at the low concentrations of arsenic normally found in drinking water. Based on these cons iderations, some regulatory agencies have revised the maximum contaminant level for arsenic in drinking water . An interim maximum acceptable concentration (!MAC) value of 25 J.lg/l has been established in Canada (Toft et al., 1990). The US Environmental Protection Agency had initially considered lowering the arsenic MCL from the existing 50 Jlg/l to between 10 and
T. VIRARAGHAVAN et al. ~gtl
(AWWA Committee, 1993). A further downward revision of the MCL to a range between 2 and is being considered (Reid, 1994).
Treatment technologies A survey of arsenic in surface water and groundwater systems showed higher levels in groundwater systems (pontius et al., 1994). A community being supplied with water containing an inorganic contaminant such as arsenic higher than the MCL has two possible solutions: change to a new water source, or incorporate treatment methods either at the point-of-entry or at the point-of-use to meet the contaminant MCL (Fox and Sorg, 1987; Rozelle, 1987). In many cases, treatment of the source water may be the only option available. The following treatment methods have been adopted to remove arsenic from drinking water under both laboratory and field conditions: adsorption-coprecipitation using iron and aluminum salts; adsorption on activated alumina/activated carbon/activated bauxite; reverse osmosis; ion exchange; and oxidation followed by filtration. The United States Environmental Protection Agency (USEPA) has summarized that coagulation with iron and aluminum salts, and lime softening are the most effective treatment processes for removing arsenic from water to meet the interim primary drinking water regulations standard of 0.05 mg/l (USEPA, 1978). Oxidation of As(III) to As(V) and removal using the above processes has been recommended. Jekel in a recent review also concluded that if As (Ill) is present in a water source, oxidation must be applied, followed by the precipitation of As(V) by ferric salts (Jekel. 1994). Because of the promise of oxidation-filtration systems, it was decided to conduct column studies at the University of Regina to examine arsenic removal by oxidation (using KMn04) followed by manganese greensand and iron oxidecoated sand filtration and compare with an ion exchange system. MATERIALS AND METHODS Materials Tap water supplemented with appropriate amounts of As (III) was used as the raw water source for all the experiments. The major physicochemical characteristics of the water were: pH, 7.4 to 7.7; conductivity. 524 f!S/cm; sulfate, 366 mgll; total dissolved solids, 321 mgll; hardness. 184 mgll as CaC03; total alkalinity, 117 as CaC03; and sodium, 42; potassium, 3.8; magnesium, 21; calcium, 39; chloride, 12; total arsenic, <0.001; iron, 0.064; and manganese, 0.018 mgll. The pH of the water was not adjusted for any of the experiments. The As(III) dosing solution was prepared from ultrapure AS203. The KMn04 solution was prepared from ACS-certified material. The greensand, supplied by Watergroup Canada Ltd., Regina, was produced by Inversand Co. in New Jersey. The physicochemical characteristics of greensand, a zeolite-type glauconite mineral, have been given in a previous publication (Subramanian et al., 1995). Manganese greensand (MGS) was produced by treating the glauconite sand with KMn04 according to the recommendations of the Saskatchewan Environmental Department (Saskatchewan Environment and Public Safety, 1987). Successive treatments produced a granular material which was coated with a layer of active hydrous manganese dioxide and other high oxides of Mn. For the column studies, the conditioning was effected in situ with the required quantity (0.005 m 3 greensand +10 g KMn04 in 972 ml water) ofKMn04 added to the column and allowed to stand overnight. The column was backwashed for over an hour prior to starting the run. MGS has a finite capacity, so that once the oxidative capacity is exhausted, it must be regenerated. Red flint filter sand manufactured by American Materials Corporation, Wisconsin, USA was supplied by Watergroup Canada Ltd, Regina. Iron oxide coating in columns was performed in situ, with a known quantity of FeCh.H20 added and allowed to stand overnight. The column was backwashed for over an hour before starting the run. Amberlite IR-120 Plus cation exchange resins manufactured by Rohm and Haas Company, Pennsylvania, USA and supplied by Watergroup Canada Ltd, Regina were used for this study. The ion exchange resins were activated with Fe III ions by mixing 300 g of the resin in O.IM FeCh.7H20 solution using a paddle mixer for approximately four hours. The activation of the ion exchange resins for the column studies was done in batch mode.
Arsenic in drinking water
Speciation and detennination of arsenic Total arsenic was determined based on stabilized temperature platform furnace atomization atomic absorption spectrometry (STPF-AAS) involving nickel matrix modification. The speciation determination of As(lII) and As(V) was based on the procedure developed by Subramanian (1988b). A Varian spectra AA600 graphite furnace atomic absorption spectrometer equipped with a Zeeman-effect background correction system, a graphite tube atomizer (GTA), a PSD-IOO programmable sample dispenser and an Epson FX8709 dot matrix printer were used for the measurement of arsenic. Column studies Two IO-cm diameter columns, each 180 em high, were used in the studies (Subramanian et al., 1995). Arsenic-dosed tap water was pumped into the column filters from a 205-1 tank using a submersible pump (Model #3E-12NT, Little Giant Pump Co., Oklahoma). A filtration rate of 1.0 to 1.5 Vmin/m 2 (2 to 3 gpm/fr') was achieved for MGS using a flow controIler. The empty bed contact time (EBCT) is the average residence time in minutes and is obtained by dividing the bed volume, including voids (I), by the flow rate in I min", The EBCT for all the experiments was found to be 5.64 min. The percent column efficiency for arsenic removal (also caIled arsenic removal efficiency) was calculated as foIlows: percent column efficiency = 100 - 100x (total effluent arsenic concentration/total influent concentration, where the concentration is expressed in gil). The column capacity for arsenic removal (also referred to as media capacity or arsenic removal capacity) for the oxidation and removal of As(lII) was calculated according to the expression: column capacity (g/l) = amount (g) of As removed/volume (I) of media, including voids. The amount of As removed (g) was calculated as foIlows: influent arsenic concentration (gil) - effluent arsenic concentration (g/I) x volume (I) of water filtered. The procedural details are given below. Tap water supplemented with 200 ug/l As(lII) was used. Break-through curves were developed and on exhaustion, the columns were backwashed with tap water, and the elution of As from the media was studied. Potassium pennanganate was continuously fed using a chemical injection pump (Model # CI530LP, Blue & White Industries, California) to oxidize the Fe(II), Mn(II) and As(III), and to generate the column. The KMn04 requirement was calculated using the foIlowing formula: KMn04 concentration (mgll) = lAx As (Ill) concentration (mg/I) + 2 x Mn(II) concentration (mg/l) + Fe(lII) concentration (mg/I), based on its stoichiometric reaction to oxidize As (III), Mn(II) and Fe(lI) present in the raw water. A 10% excess of KMn04 feed was applied to the dosage obtained from the above equation to ensure that sufficient oxidant was present to continuously regenerate the greensand filter. The above experiment was repeated by adding Fefll) in the form of FeS04.6H20 to the influent to give Fe/As ratios of 10 [200 1.lg/1 As (III) and 2 mg/I Fe (In], 20 [100 1.lg/1 As (III) and 2 mg/I Fe (II)] and 7 [50 ug/l As (III) and 2 mgll Fe (II»). Column studies were also conducted using intermittent regeneration of the column using KMn04. Tap water was spiked to an initial As(IIn concentration of 200 ug/l and continuous column runs using iron oxide-coated sand media were conducted to breakthrough at a filtration rate of 1.0 to 1.5 Vmin/m2 (2 to 3 gpmJft2) and on exhaustion, the column was backwashed with tap water and the elution of arsenic from the media studied. Tap water was spiked to an initial As (III) concentration of 200 I.lgll and continuous column runs using ion exchange resin 2 were conducted at a filtration rate of2.0 to 2.25 Vmin/m (4 to 4.5 gpm/ft') to breakthrough. RESULTS AND DISCUSSION Results of column studies showed that continuous regeneration of the column using KMn04 produced better results than intermittent regeneration; details are available elsewhere (Viraraghavan and Aruldoss, 1996). Table 2 shows a comparison of the results obtained using three treatment media with continuous regeneration of the manganese greensand media. Figures 1, 2 and 3 show arsenic removal by manganese greensand (Fe:As ratio = 20:1), iron oxide coated sand and ion exchange resin respectively. In the presence of iron, manganese greensand proved to the best available treatment option for the removal of As (Ill); overaIl, manganese greensand with an Fe/As ratio of 20:1 was found to be the optimum performer. A detailed discussion of the use of manganese greensand for removal of arsenic in drinking water is available in an earlier publication (Subramanian et al., 1997).
T. VlRARAGHAVAN et al.
Table 2. Comparison of column experiment results for the three treatment methods Manganese greensand
Filtration method --+ Arsenic concentration (/.lgll) Fe:As Total throughput volume (I) Overall removal efficiency (%) 3
Overall column capacity (gift ) Throughput volume to MAC (1)
200 0: I 1003 41.3 9.3
200 10: 1 1440 83.3 26.7
100 20: 1 1440 81.8 14.1
50 7: 1 2168 47.6 6.5
768 49.7 10.5
851 37.6 4.1
.§ ~ c
IS 25 w
g § §
Iron oxide coated sand 200
IS w 'o....As(T)
Figure1. Effiuent arsenic and manganese concentration usingmanganese greensand filtration (Fe:As = 20:I) 1I5 ~--------------~
Time (hours) Figure2. Effiuent arsenic concentration using ironoxidecoated sandfiltration.
II flO •... + I
_ __'__ _-'-_..J
Time (hours) Figure 3. Effiuent arsenic concentration using ionexchange resinsactivated with Fe)+
General discussion The occurrence of As (III) and As(V) of geologic origin in the raw water supply, especially in groundwater, is a major problem in certain regions of Asia such as West Bengal, India (Chatterjee et aI., 1995; Das et al., 1995) and Xinjiang, China (Lianfang and Jianzhong, 1994), and economic removal of arsenic from drinking water using simple and appropriate technologies offers one of the best solutions. The other option, of course, is to search for a water source without arsenic. Conventional water treatment plants (generally used in large
Arsenic in drinking water
communities) employing coagulation, sedimentation and filtration will be usually effective in arsenic removal. The problem becomes serious in small communities in developing countries of Asia and other parts of the world, where treatment systems for groundwater are either non-existent or simple with only disinfect ion. In these circumstances, simple treatment systems using filtration techniques with local media such as natural oxides (hematite, manganese and iron oxides) may be the best option (Hsia and Lo, 1994; Subramanian et al., 1995; Saskatchewan Environment and Public Safety, 1987; Viraraghavan and Aruldoss, 1996; Subramanian et al., 1997). In this context, manganese greensand media may be appropriate in many cases. Simplicity and cost are the two major factors that should influence the selection of a treatment system for arsenic removal in rural communities dependent on groundwater. CONCLUSIONS Based on the results of the study, the following conclusions are made: 1. Manganese greensand was found to be the best available treatment option for removal of arsenic. 2. Arsenic removal capacity of the three media based on column studies decreased in the following order: manganese greensand> iron oxide coated sand> ion exchange resins activated with ferric ions. 3. Iron to arsenic ratio had a significant effect on the removal of arsenic by manganese greensand filtration. ACKNOWLEDGEMENTS The authors acknowledge major support from Health Canada for this study. REFERENCES American Water Works Association Committee (1985). An AWWA survey of inorganic contaminants in water supplies . Journal American Water Works ASSOCIation, 77(5), 67-72. AWWA Committee (1993). Research needs for inorgan ic contaminants. Journal American Water Works Association, 85(5), 106-113. Chatterjee, A., Das, D. and Chakraborti, D. (1993). A study of groundwater contamination by arsenic in the residential area of Behala, Calcutta due to industrial pollution. Environmental Pollution , 80(1), 57-65. Chatterjee, A., Das, D., Mandai, B.K., Choudhury, T.R., Samanta, G. and Chakraborti , D. (1995) . Arsenic in groundwater in six districts of West Bengal, India: the biggest arsenic calamity in the world, Part I - Arsenic species in drinking water and urine of the affected people . Analyst, 120(3), 643-650. Das, D., Chatterjee, A., Mandai, B.K., Samanta, G., Chakraborti, D. and Chanda, B. (1995) . Arsenic in groundwater in six districts of West Bengal, India: the biggest arsenic calamity in the world, Part 2 - Arsen ic concentration in drinking water, hair, nails, urine, skin-scale and liver tissue (biopsy) of the affected people. Analyst, 120(3),917-924. Edwards, M. (1994) . Chemistry of arsenic removal during coagulation and Fe-Mn oxidation . Journal American Water Works Association, 86(9) , 64 - 78. Fowler, B.A. (1983) . Arsenic metabolism and toxicity to freshwater and marine species, In: Biological and Environmental Effects ofArsenic, B.A. Fowler (ed.), Elsevier Science Publishers, Amsterdam, 6,155-170. Fox, K.R. and Sorg, TJ. (1987) . Controlling arsenic, fluoride, and uranium by point-of-use treatment. Journal American Water Works Association, 79(10), 81-84. Ghosh, M.M. and Yuan, J.R. (1987) . Adsorption of arsenic and organoarsenicals on hydrous oxides. Environmental Progress, 6(3), 150-157. Grantham, D.G. and Jones, J.F. (1977) . Arsenic contamination of water wells in Nova Scotia. Journal American Water Works Association, 69(12),653-657. Gupta, S.K. and Chen, K.Y. (1978). Arsenic removal by adsorption. Journal Water Pollution Control Federation, 50(3), 493-506. Hsia, T.H. and Lo, K-L. (1994) . The pollution problems and treatment methods of arsenic in water. Water Supply, 8(3/4), 32-44. Jekel, M.R. (1994) . Removal of arsenic in drinking water treatment. In: Arsenic in the Environment: Pan I Cycling and Characterization, J.O. Nriagu (ed.), John Wiley and Sons, New York, N.Y., 119-132. Lianfang, W. and Jianzhong, H. (1994) . Chronic arsenism from drinking water in some areas of Xinjiang, China . In: Arsenic in the Environment: Pan 1/ Human Health and Ecosystem Effects, J.O. Nriagu (ed.), John Wiley and Sons, Inc., New York, N.Y., 159-172. MacKenzie, F.T., Lantzy, R.J. and Paterson, V. (1979) . Global trace metal cycles and predictions . Journal Int. Assoc. Math. Geol., 6,99-142. Meranger, J.e., Subramanian, K.S. and McCurdy, R. F. (1984). Arsenic in Nova Scotian groundwater. The Science of the Total Environment, 39(1/2), 49-55. Patterson, J.W . (1985) . Arsenic. In: Industrial Wastewater Treatment Technology, Butterworth Publishers. pp. 11-21.
T. VlRARAGHAVAN et al.
Pershagen, G. (1983). The epidemology of human arsenic exposure. In: Biological and Environmental Effects ofArsenic, B.A. Fowler (ed.), Elsevier Science Publishers, Amsterdam, 6, 199-232. Pontius, F.W., Brown, K.G. and Chen, C.J. (1994). Health implications of arsenic in drinking water. Journal American Water Works Association, 86(9), 52-63. Ramesh, R., Shiv Kumar, K., Eswaramoorthi, S. and Purvaja, G.R. (1995). Migration and contamination of major and trace elements in groundwater of Madras City, India. Environmental Geology, 25(2), 126-136. Reid, J. (1994). Arsenic occurrence: USEPA seeks clearer picture. Journal American Water Works Association, 86(9), 44-51. Rozelle, L.T. (1987). Point-of-use and point-of-entry drinking water treatment. Journal American Water Works Association, 79(10),53-59. Saskatchewan Environment and Public Safety (1987). Conditioning a manganese greensand filter. Clear Water Reporter 1(4). Shen, V.S. (1973). Study of arsenic removal from drinking water. Journal American Water Works Association, 65(9), 543-548. Smith, A.H., Hopenhayn-Rich, C., Bates, M.N., Goeden, H.M., Picciotto, I.H., Duggan, H.M., Wood, R., Kosnett, M.J. and Smith, M.T. (1992). Cancer risks from arsenic in drinking water. Environmental Health Perspectives, 97, 259-267. Subramanian, K. S. (1988a). Arsenic. In: Quantitative Trace Analysis ofBiological Materials, H.A. McKenzie and L.E. Smythe (eds), Elsevier Science Publishers, Amsterdam, 573-587. Subramanian, K. S. (I 988b). Determination of arsenic in urine by graphite platform-in-furnace atomic absorption spectrometry. Can. J. Spectrosc. 33,173-181. Subramanian, K. S., Viraraghavan, T., Tanjore, S. and Phommavong, T. (1995). Manganese greensand for removal of arsenic in drinking water treatment. In: Proceedings ofthe Water Quality Technology Conference, pp. 1063-1089. American Water Works Association, New Orleans. Subramanian, K. S., Viraraghavan, T., Tanjore, S. and Phommavong, T. (1997). Manganese greensand for removal of arsenic in drinking water. Water Quality Research Journal ofCanada. 32(3), 551-561. Toft, P., Tobin, R.S., Meek, M.E. and Wood, G.C. (1990). Guidelines for Canadian drinking water quality. Coping with the Guidelines in the 1990s - Proceedings of the Fourth National Conference on Drinking Water, Toronto, Ontario, Canada, R.S. Tobin and W.J. Robertson (eds), pp. 1-9. United States Environmental Protection Agency (USEPA) (1978). Manual of Treatment Techniques for Meeting the Interim Primary Drinking Water Regulations, EPA-600/8-77-005. Viraraghavan, T., Jin, V.C. and Tonita, P.M. (1992). Arsenic in water supplies. International Journal of Environmental Studies, 41(1/2), 159-167. Viraraghavan, T. and Aruldoss, J. A. (1996). Removal of arsenic in drinking water by manganese greensand filtration and ion exchange treatment. Report preparedfor Health Canada, University of Regina, Regina, Canada.