Reviews of Environmental Contamination Volume 197: Arsenic Pollution and Remediation: An International Perspective (Reviews of Environmental Contamination and Toxicology) (Vol. 197)

Reviews of Environmental Contamination and Toxicology VOLUME 197

Reviews of Environmental Contamination and Toxicology Arsenic Pollution and Remediation: An International Perspective Editor

David M. Whitacre Volume Editors

Hemda Garelick and Huw Jones

Editorial Board Lilia A. Albert, Xalapa, Veracruz, Mexico • Charles P. Gerba, Tucson, Arizona, USA John Giesy, Saskatoon, Saskatchewan, Canada • O. Hutzinger, Bayreuth, Germany James B. Knaak, Getzville, New York, USA James T. Stevens, Winston-Salem, North Carolina, USA Ronald S. Tjeerdema, Davis, California, USA • Pim de Voogt, Amsterdam, The Netherlands George W. Ware, Tucson, Arizona, USA

Founding Editor Francis A. Gunther

VOLUME 197

123

Coordinating Board of Editors DR. DAVID M. WHITACRE, Editor Reviews of Environmental Contamination and Toxicology 5115 Bunch Road Summerfield North, Carolina 27358, USA (336) 634-2131 (PHONE and FAX) E-mail: [email protected]

Springer New York: 233 Spring Street, New York, NY 10013, USA Heidelberg: Postfach 10 52 80, 69042 Heidelberg, Germany Library of Congress Catalog Card Number 62-18595 ISSN 0179-5953 Printed on acid-free paper. # 2008 Springer Science + Business Media, LLC All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science + Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.

ISBN: 978-0-387-79283-5 DOI 10.1007/978-0-387-79284-2 Springer.com

e-ISBN: 978-0-387-79284-2

Special Foreword

Exposure to arsenic-contaminated drinking water is a major threat to human health. Millions of people across the world are exposed to arsenic-contaminated drinking water with concentrations far in excess of the 10 mg/L maximum permissible level established by the World Health Organization (WHO). The major arsenic exposure pathway is believed to be via natural (geological) sources of contaminated groundwater. In addition, arsenic is introduced into the environment from anthropogenic sources, primarily metal mining and smelting activities, which pollute soils, sediments, and surface waters and groundwater worldwide. The implications for human health of arsenic exposure are serious, but neither are these implications fully understood nor are solutions for mitigation adequately evaluated or communicated. The purpose of the six papers comprising this volume is to address this knowledge gap. These papers result from a project supported by the Chemistry and the Environment Division (VI) of the International Union of Pure and Applied Chemistry (IUPAC). They are consonant with and underpin the key IUPAC objectives of advancing the chemical sciences and the application of chemistry in service to mankind. IUPAC, in its role as an objective scientific, international, and nongovernmental body, in collaboration with international governmental bodies [e.g., United Nations Educational, Scientific and Cultural Organization (UNESCO) and the WHO] addresses many global issues involving the chemical sciences as well as issues that transcend pure science and have important sociopolitical implications. Arsenic contamination clearly has such implications. The papers presented in this volume aim to review and analyze the status of arsenic pollution and consequential human exposure and to provide a practical guide to available arsenic remediation technologies. Moreover, we endeavor to advise on tools that support informed decision making when choosing avenues for arsenic mitigation. Such decision making cannot be solely concerned with arsenic treatment technologies, and the papers therefore seek to highlight and provide guidance on arsenic treatment technologies in the context of varying scenarios that can inform effective mitigation policies. The authors of these papers have a diversity of knowledge, research experience, and interests, all of which contributed to assembly of this volume. The team’s expertise in epidemiology (Harry Caussey); risk assessment and toxicology (Nick v

vi

Special Foreword

Priest); environmental chemistry (Hemda Garelick, Huw Jones, and Zolta´n Galba´cs); environmental geochemistry (Eugena Valsami-Jones and Agnieska Dybowska); analytical chemistry (Joerg Feldmann); bioremediation (Pornsawan Visoottiviseth); environmental engineering (Feroze Ahmed, Rita Fo¨lde´nyi, Nora Kova´ts, and Ga´bor Borbe´ly); and environmental management (Bryan Ellis, Hemda Garelick, and Md. Khoda Bux) was critical in analyzing effects of and solutions to arsenic water pollution on exposed populations. Key points addressed by each successive paper are these: l

l

l

l

l

l

Health risks of arsenic contamination, with reference to the technical challenges associated with optimizing arsenic remediation approaches that are acceptable to arsenic-polluted communities. Overview of the global status of arsenic pollution sources, both natural and anthropogenic, and behavior of arsenic in groundwater and surface waters. Information is provided on modes of formation and release of arsenic and the corresponding implications to environmental mobility and toxicity of different arsenic chemical species. Effects of high spatial and temporal variation of arsenic contamination and the consequential need for cheap, quick, onsite (field kits) analytical techniques that accurately portray the degree and nature of contamination so critical to remediation efforts are discussed. A variety of potential remediation technologies for arsenic removal are described. To be effective, particularly in developing countries with the greatest arsenic contamination, such methods must be reliable, cost-effective, and sustainable. The range of mitigation options available for arsenic reflects the complexity of its chemistry. Appraising suitable arsenic remediation technologies is itself a sizable challenge. In this paper, we address, through multi-criteria approaches, the factors relevant to evaluating mitigation options. The final paper of the series shares the challenges faced by three countries with arsenic-contaminated regions in addressing and remediating sources of arsenic contamination. ‘‘Access to safe water is a fundamental human need and, therefore, a basic human right. Contaminated water jeopardizes both the physical and social health of all people. It is an affront to human dignity. Yet even today, clean water is a luxury that remains out of the reach of many.’’

These words, spoken by Kofi Annan, then Secretary General of the United Nations, on World Water Day, March 22, 2001, sadly remain equally relevant in 2007. We, the authors, believe that this situation cannot be allowed to persist and hope that this series of papers will redress it, in some small way. November 2007

Hemda Garelick and Huw Jones

Acknowledgements

The authors thank the IUPAC division of Chemistry and the Environment and its president, Dr Ken Racke, for its financial support and for providing a framework cogent to developement and implementation of the project which led to this volume. Special thanks are extended to Dr Yehuda Shevah whose drive and commitment made this project possible. We thank John Koushappas, Paula Newland, and Yvette Brown from the School of Health and Social Sciences Learning and Technical Unit at Middlesex University for their support in managing the web-based activities for our team and for their technical support in production of figures. We, herewith, express our full appreciation to the independent referees of the papers—Prosun Bhattacharya, The Royal Institute of Technology, Stockholm, Sweden; Eton Codling, USDA, ARS, Beltsville, MD, USA; Karen HudsonEdwards, Birkbeck College, London, UK; Jake Peters, USGS, Atlanta, GA, USA; and David Polya, The University of Manchester, UK—who provided comprehensive reviews and constructive feedback. Very special thanks are owed to R.E.C.T. editor Dr. David Whitacre, who tirelessly and meticulously reviewed the manuscripts and who provided the team with outstanding editorial support.

vii

Contents

Introduction to Arsenic Contamination and Health Risk Assessment with Special Reference to Bangladesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Deoraj Caussy and Nicholas D. Priest Arsenic Pollution Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Hemda Garelick, Huw Jones, Agnieszka Dybowska, and Eugenia Valsami‐Jones Onsite Testing for Arsenic: Field Test Kits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Joerg Feldmann Technology for Remediation and Disposal of Arsenic . . . . . . . . . . . . . . . . . . . . . . . 77 Pornsawan Visoottiviseth and Feroze Ahmed A Multi-Criteria Approach for Assessing Options to Remediate Arsenic in Drinking Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Bryan Ellis and Hemda Garelick Case Reports: Arsenic Pollution in Thailand, Bangladesh, and Hungary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Huw Jones, Pornsawan Visoottiviseth, Md. Khoda Bux, Rita Fo¨lde´nyi, Nora Kova´ts, Ga´bor Borbe´ly, and Zolta´n Galba´cs Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

ix

Introduction to Arsenic Contamination and Health Risk Assessment with Special Reference to Bangladesh Deoraj Caussy(*) and Nicholas D. Priest

I II III IV V

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Exposure Pathways for Arsenic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Health Effects of Exposure to Inorganic Arsenic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Bioavailability of Ingested Inorganic Arsenic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Application of the Health Risk Paradigm for Arsenic Contamination in the Bengal River Basin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 A Hazard Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 B Dose–Response Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 C Exposure Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 D Risk Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 VI Risk Mitigation for Arsenic Contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 A Control Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 B Socioeconomic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 C Legal and Ethical Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

I Introduction Arsenic is a metalloid element that occurs in nature in both organic and inorganic compounds. The valence of arsenic is three in arsenite (As (III)) and five in arsenate (As (V)) compounds. Arsenic compounds are poisonous and can be categorized into D. Caussy World Health Organization, Regional Office for South-east Asia, Ring Road, New Delhi, 110 002 India N.D. Priest School of Health and Social Sciences, Middlesex University, The Burroughs, London NW4 4BT UK

D.M. Whitacre (ed.), Reviews of Environmental Contamination Volume 197. doi: 10.1007/978-0-387-79284-2_1, # Springer Science þ Business Media, LLC 2008

1

2

D. Caussy, N.D. Priest

one of three broad categories, all with different toxicities: organic arsenic compounds, inorganic arsenic compounds and their solutions, and arsine gas. Of these, inorganic compounds are of relatively greater toxicological significance for human health than are organic compounds. Toxicological profiles for arsenic compounds have been published both in standard texts (Ishinishi et al. 1986) and in reports produced for international and governmental agencies (IARC 1982; ATSDR 2000; IPCS 2001). The reader is referred to these for a comprehensive description of the toxicokinetics and toxicity of arsenic. The purpose of this chapter is to review the salient health risks posed by arsenic contamination, with special reference to the technical challenges for optimizing arsenic remediation in ways that are acceptable to communities in Bangladesh.

II

Exposure Pathways for Arsenic

Humans may be exposed to inorganic arsenic via air, water, food, and soil. Arsenic is widely distributed in nature as sulfides in minerals, as dissolved salts in groundwater, surface waters, and seawater, and in soils, sometimes consequent to its anthropogenic extraction and usage. Arsenic compounds are extracted from minerals and have been widely used as therapeutic agents, now mostly in the form of antiparasitic drugs, as pesticides, in the manufacture of glass, as an alloying agent, in the manufacture of some dyes, and as gallium arsenide for production of crystals in the semiconductor industry (for more information on arsenic sources, see ‘‘Arsenic Pollution Sources,’’ later in this volume). Most food products usually contain less than 250 mg/kg arsenic. However, seafood such as demersal fish, crustaceans, and marine algae may contain up to 100 mg/kg arsenic. The low levels in plants contrast with the much higher levels (40 mg/kg) in soil and, under normal soil conditions, may reflect the insolubility of many arsenic-containing minerals such as pyrite. The average U.S. daily dietary intake for humans is estimated to be 10–20 mg arsenic. In Japan, however, where the seafood content of the diet is high, intakes are much larger (70–370 mg/d). In addition to ingestion from food and drinking water, air is also a potential source for human exposure to arsenic. Significant intakes by inhalation occur in residents living near industrial sources where exposures to arsenic trioxide are possible; however, this is unlikely to be a significant source of exposure to natural forms of arsenic in Table 1 Reported levels of arsenic in groundwater in selected countries Location Recorded arsenic concentration (mg/L) Nova Scotia, Canada 50 California, Romania, New Zealand 40–1,300 Japan 1,700 Cordoba, Argentina 3,400 Taiwan, China 1,800 Bangladesh 300 USA domestic water sources <10 (99%) Source: IPCS 2001.

Introduction to Arsenic Contamination and Health Risk Assessment

3

Fig. 1 Extent of arsenic contamination in Asia. Courtesy British Geological Survey (BGS) (Smedley 2003)

Bangladesh. Consumption of arsenic-contaminated water, however, poses a significant threat to human health in some parts of the world. The global nature of arsenic contamination in groundwater has been summarized in the report of the British Geological Survey (BGS 2001). Table 1 summarizes the arsenic levels found in groundwater in selected parts of the world. Arsenic contamination is particularly notable in deltaic plains of river basins in Asia, including the Brahmaputra Gangetic River and the Red River-Mekong deltas (Fig. 1). Groundwater contamination is further discussed later in this volume (see ‘‘Arsenic Pollution Sources’’). Bangladesh is among the countries most affected by arsenic contamination in the WHO (World Health Organization) South-East Asia Region. The origin of arsenic contamination in the Bengal River Basin can be traced to the deposition of arsenicladen sediment in the Bengal deltas. This contaminated sediment has been carried by the Ganga and Brahmaputra Rivers from the Himalayan mountains to the delta over millions of years (McArthur 2002). These arsenic-containing sediments adhered to rocks and eventually formed the impervious layer of the aquifers in the delta. Increasing population growth led to the search for a source of drinking water free of microbial contamination, and, hence, the exploitation of aquifers that were subsequently found to be contaminated. Aquifer exploitation reached its

4

D. Caussy, N.D. Priest

Fig. 2 Extent of arsenic contamination in Bangladesh

peak during the green revolution (McLellan 2002), when demand was compounded by the need for irrigation water. To meet growing needs, tube wells were sunk into the aquifers; in Bangladesh alone some 10 million tube wells were drilled into arsenic-contaminated aquifers. The consequence and tragedy was that some 40 million people were subsequently exposed to toxic levels of arsenic, sometimes exceeding the WHO Guideline value by a factor of 20 or more. Of the 64 districts within Bangladesh, 59 have reported unsafe levels of arsenic in groundwater (see the last chapter in this collection for more information on this topic). However, arsenic contamination is not homogeneous, and the correlation between arsenic concentrations in water, and disease prevalence, is less than perfect. Considerable geographic variability is apparent (Fig. 2). Most of the heavily contaminated areas are in the lower deltaic plain of the Brahmaputra River basin.

III

Health Effects of Exposure to Inorganic Arsenic

The health outcomes resulting from arsenic exposure depend on the dose, modality, and duration of exposure, as well as the source and chemical type of arsenic (Caussy et al. 2003a). The major health effects of arsenic have been reviewed elsewhere (IPCS 2001). Arsenic is toxic following both acute and chronic intakes. However, drinking contaminated water is only likely to produce effects under conditions of chronic intake. Chronic effects produced by the ingestion of inorganic arsenic include skin lesions, disturbances of the peripheral nervous system, anemia and leukopenia, liver damage, circulatory disease, and cancer. Many of these effects have been observed in populations that consumed contaminated water, including populations in Taiwan, Argentina, and Bangladesh (IPCS 2001).

Introduction to Arsenic Contamination and Health Risk Assessment

5

Under conditions of chronic intake, skin lesions are characterized by keratosis and melanosis of varying severity. These lesions are usually manifested on the palms of the hands and soles of the feet. Keratotic lesions may, at some stage, become malignant (by a process that remains unclear and may involve induction by factors other than arsenic) (IARC 1982) and result in squamous cell skin cancer (Caussy 2003a). Such tumors are commonly multifocal and develop throughout the body. Basal cell carcinoma has also been described in cases that display chronic arsenical dermatitis. Chronic exposure to arsenic may also damage the peripheral nervous system. Such damage is characterized by a peripheral neuritis affecting mainly the upper and lower limbs. This neuritis results in a reduced sense of touch, in numbness and in paraesthesia, characterized by a ‘‘pins-and-needles’’ sensation. Nevertheless, such effects have only been seen in arsenic-exposed workers and may not occur in populations exposed only to lower arsenic concentrations in drinking water. Not surprisingly, such neuropathy has not been widely reported from Bangladesh. Similarly, mucous membrane lesions and liver damage have only been described in arsenic-exposed workers. Disturbances in the hematopoietic system have been noted in subjects with arsenic-induced skin lesions, and in these subjects the skin lesions may be accompanied by anemia and leukopenia. Arsenic-induced anaemia is not associated with iron deficiency and is aplastic. In addition to blood damage, arsenic exposure also damages other components of the circulatory system. Toxic myocardial effects and peripheral blood vessel damage leading to atrophic acrodermatitis and gangrene of the extremities have been described. Gangrene of the extremities, known as blackfoot disease, has been seen, but only in Taiwanese populations that have ingested arsenic. Therefore, the linkage between blackfoot disease and arsenic, as a sole causal agent, is uncertain.

IV

Bioavailability of Ingested Inorganic Arsenic

The bioavailability of arsenic is a key determinant in health outcomes of exposure, because it is related to the ability of arsenic to be liberated from ingested matrices (e. g., soil, water, and food) and thus enter into the bloodstream where it exerts its toxic effects (Caussy et al. 2003a). Most information on the bioavailability of arsenic is derived from human and animal observations. Studies from human volunteers and animal experiments show about 90% of ingested, dissolved, inorganic arsenic salts are absorbed from the gut and enter the bloodstream. This percentage is much higher than for most other nonessential elements and, in part, is a feature of the similar chemical properties of the arsenate (AsO42) and phosphate (PO42) ions. Phosphate is an essential body component and is avidly absorbed from the gut. Arsenate appears to follow many metabolic pathways that exist for phosphate, including deposition in the body at sites of phosphate incorporation. Arsenic also avidly binds to sulfhydryl groups on proteins and other biomolecules. Nevertheless, the exact

6

D. Caussy, N.D. Priest

fraction of arsenic uptake tends to depend upon gut contents and the presence, or otherwise, of phosphate absorption inhibitors such as aluminum salts, calcium carbonate, and lanthanoid salts, all of which inhibit arsenic uptake. In contrast, the fractional uptake of arsenic incorporated within organic compounds (organic arsenic) is reported to be lower than for dissolved inorganic species. Arsenic, absorbed within the gastrointestinal tract, first passes through the liver and then enters the bloodstream and is distributed to the body (Caussy 2003a). Some arsenic accumulates in tissues and the remainder is excreted, mostly in urine. Although the affinity of arsenic for tissues, its kinetics of deposition, redistribution, and excretion, are arsenic species dependent, all species are rapidly cleared from blood and become evenly distributed among body tissues. Only a few tissues selectively concentrate or retain arsenic: liver, lungs, skin, nails, hair, and the skeleton (Table 2). Because skeletal muscle mass comprises a large proportion of body weight, a considerable amount of body arsenic is present in this tissue, despite its rather low concentrations. The data presented in Table 2 reflect exposure of the general Japanese population to organic arsenic intake from seafood, rather than tissue distributions acquired from ingestion of inorganic arsenite (As (III)) and/or arsenate (As (V)) in contaminated drinking water. Following the intake of inorganic arsenic, more is likely to be deposited in liver and the skeleton, but less in kidneys. However, the deposition and retention of arsenic species is complicated by metabolic and redox processes within the body. These processes result in the oxidation of arsenite to arsenate, the metabolic reduction of arsenate to arsenite, and the methylation of arsenite in the Table 2 Measured tissue concentrations of arsenic and implied organ content for human organs collected in Japan Tissue Arsenic concentration Tissue content (mg/kg) (mg) Bone 96 450 Brain 34 41 Hair 174 3 Heart 41 11 Large intestine 25 8 Small intestine 22 12 Kidney 41 11 Liver 42 65 Lung 47 40 Muscle 29 700 Nail 892 3 Pancreas 20 2 Skin 64 180 Spleen 21 3 Stomach 22 3 Teeth 78 3 Uterus 36 2 Source: Ishinishi et al. (1986).

Introduction to Arsenic Contamination and Health Risk Assessment

7

liver to form monomethyl arsenic acid (MMA) and dimethyl arsenic acid (DMA). These transformations affect both tissue deposition and retention. Results of human volunteer studies conducted with oxidized inorganic arsenic salts, such as those present in most contaminated drinking water, suggest that arsenic administered as As(V) is excreted according to the following pattern: 0.66 with a biological retention half-time (Tb½) of 2.1 d; 0.3 with Tb½ of 9.5 d; and 0.037 with Tb½ of 38 d. In addition, arsenate substitutes for phosphate and is deposited in the skeleton, where it may be retained for a much longer period, perhaps as long as 5,000 d. Of the excreted fraction, about half comprises inorganic ions and about half as DMA/MMA, with a preponderance of DMA; however, these proportions vary with age, gender, and intake composition. Unchanged arsenite is usually excreted faster than arsenate; the Tb½ of organic arsenic in the body is about 20 hr. However, this apparent difference between arsenate and arsenite may be an artefact of the acidic conditions in the stomach, which adjusts the oxidation state of ingested ions. More exhaustive studies will probably reveal little or no difference between the bioavailability and biokinetic behavior of ingested soluble arsenate and arsenite ions.

V

Application of the Health Risk Paradigm for Arsenic Contamination in the Bengal River Basin

The health risk paradigm can be viewed as comprising two components (Fig. 3): health risk assessment and risk management. The risk assessment paradigm classically involves a four-step process consisting of hazard identification, dose–response assessment, exposure assessment, and risk characterization (NRC 1983). The risk assessment process uses data from many sources, including animal experiments, in vitro studies and human epidemiological observations. The epidemiological gaps for applying the health risk assessment paradigm to arsenic contamination has been reviewed elsewhere (Caussy 2003b, c; Caussy and Than Sein 2006). Two uncertainties, linked with exposure and health effects, temper the risk assessment process, and these are discussed next.

A

Hazard Identification

The purpose of hazard identification is to qualitatively characterize the relationship between arsenic exposure and probable adverse health outcomes. This analysis primarily relies on data derived from epidemiological and toxicological databases. Global epidemiological studies (of cross-sectional, case-control, and cohort study designs) have shown that exposure to unsafe levels of arsenic represents a definite health hazard, including arsenicosis and cancer (ATSDR 2000). In the absence of animal data, epidemiological data have been used to calculate the lowest observable adverse effect level (LOAEL) for arsenicosis. This LOAEL value is defined as the lowest dose needed to induce melanosis or keratosis, or both, and various sources estimate the value as

8

D. Caussy, N.D. Priest

Risk Management of arsenic poisoning is a two step process Risk assessment

3: EXPOSURE ASSESSMENT

Risk management

CONTROL OPTIONS

4: RISK CHARACTERIZATION

2: DOSERESPONSE ASSESSMENT 1: HAZARD IDENTIFICATION

LEGAL CONSIDERATIONS

RISK MANAGEMENT DECISIONS

OTHER ECONOMIC AND SOCIAL FACTORS

Fig. 3 Health risk paradigm. (Adapted from NRC—National Research Council 1983)

10–20 mg/kg/d (ATSDR 2000). Regional variations in the derived LOAEL value exist. The lowest LOAEL value (0.04 mg/kg/d) was observed in Mexico (Cebrian et al. 1983), whereas the highest value (18 mg/kg/d) was observed in exposed West Bengal populations (Chakraborty and Saha 1987). For an Asiatic 60-kg adult, this dose corresponds to a daily intake of 900 mg, an amount equivalent to a water concentration of about 600 mg/L, assuming a daily consumption of 1.5 L. Existing LOAEL data suggest that arsenic intakes are much higher than those derived using the WHO water quality standard; this may reflect continuing uncertainty in the LOAEL and a protective conservatism built into the WHO standard.

B

Dose–Response Assessment

In dose–response assessment, the dosage of arsenic needed to induce predefined adverse health effects is quantified. Both experimental animal data and epidemiological data are routinely used to build the dose–response curve. Statistical modelling has been used with cancer endpoints to assess dose response following arsenic exposure and concomitant intake. For example, Chen et al. (1985) and Chen and Wang (1990) used an ecological study to investigate the association between arsenic and cancer mortality. The relationship they studied was between arsenic concentration in water from more than 83,000 wells and reported cancer incidence from a cancer registry. Using regression analysis, the investigators

Introduction to Arsenic Contamination and Health Risk Assessment

9

demonstrated a dose-dependent association between cancer mortality and arsenic exposure. These data, collected in Bangladesh and West Bengal, were inadequate to estimate dose–response functions for skin or internal cancers, largely because the duration of exposure to tube well water is less than the 20-yr induction period typically needed. However, this limitation has been addressed by using surrogate data on cancer incidence from Taiwan to exposure dose from Bangladesh (Yu and Ahsan 2004). Results show that the risk of skin and non-skin cancer steadily increased with increasing dose of arsenic exposure.

C

Exposure Assessment

Because most arsenic is eventually excreted in urine, urinary excretion levels may, with reservation, be used to estimate total arsenic intake from inhaled, ingested, and skin absorption sources. Such intake measurements provide a meaningful estimate for toxicity assessment, unless the diet contains significant levels of much less toxic organic arsenic. Under conditions of normal arsenic exposure, levels of arsenic in urine range from about 5 to 50 mg/L, implying total intakes of about 10–100 mg/d, primarily from the consumption of either drinking water or seafood. The speciation of arsenic, in natural sources of drinking water, is variable and depends upon dissolved oxygen levels, pH, and other factors. In general, arsenic is extracted from its aquifer under reducing conditions and reaches the surface as As (III), but on exposure to air is oxidized to As(V). Organic arsenic concentrations are low in most groundwaters but may be higher in biota-containing surface waters. Data for arsenic exposure assessment in Bangladesh have been largely derived from environmental monitoring of water in tube wells. Monitoring by various agencies in Bangladesh indicates that 59 of the 64 districts have arsenic concentrations in groundwater in excess of the prevailing national standard of 50 mg/L (BGS 2001). The predicted intake is high, and in some places, water levels exceed the WHO guideline values by a factor of 30. Such high levels pose serious implications for remediation technologies. Although tube wells have been in use for many years, neither the concentration nor duration of exposure to individuals can accurately be determined, partly because arsenic concentration data are often unavailable for the entire exposure period, as well as because the exposed population obtained drinking water from multiple sources. Furthermore, the concentration of arsenic fluctuates in the same tube well with time (NRC 1999). Moreover, when arsenic concentration was measured, testing methods used were either qualitative, semiquantitative, or constituted methods that did not conform to approved procedures. Because many of the methods utilized were not validated, most exposure data must be regarded as anecdotal. Biological monitoring of human beings, through measurement of arsenic levels in hair and nails, is limited, partly as a result of the cumbersome nature of the test and partly because of the inability of the markers to detect arsenic from exposure events more than 9 mon before analysis (NRC 1999; IPCS 2001).

10

D. Caussy, N.D. Priest

Such samples are also subject to environmental contamination that conflates the assessment of body burden. The existing exposure data are also deficient in being limited to a single matrix: water. A proper exposure assessment requires food chain monitoring capable of delivering insights on dose, species, and bioavailability of arsenic in commonly consumed food, particularly rice, which may contain significant levels of inorganic arsenic.

D

Risk Characterization

The main risk of arsenic to human health is noncancerous and cancerous skin lesions (NRC 1999; IPCS 2001). The first and most common manifestation of arsenic poisoning is dermal lesions in the form of keratosis and melanosis. However, correctly diagnosing arsenic-induced dermal lesions from similar lesions mimicking arsenicosis requires special skills. Until recently, no universal method existed for diagnosing and classifying arsenicosis. Misdiagnoses of arsenic-induced skin lesions should be minimized with the introduction of uniformity of case definition (Caussy 2005). It has been concluded from epidemiological assessment of arsenic poisoning in the Bengal Basin that some 12 of 40 million exposed subjects are projected to develop some form of skin lesion within 10–20 yr of exposure (Caussy and Than Sein 2006). An alternative method, based on dose–response analysis, projects the count of arsenicosis cases in Bangladesh to amount to 1,864,000 (Yu et al. 2003). To date, the reported clinical cases fall short of these projected numbers, either because cases are undiagnosed, are unreported, or exposed persons have not yet manifested the disease. In part, the estimates vary because standard diagnostic criteria are lacking. The picture is further confused by data that show poor correlation between arsenic exposure levels in drinking water and manifestation of arsenicosis, i.e., low incidence of disease in areas with apparently high water levels of arsenic and vice versa. If such variability is confirmed, then additional factors are involved, either in the induction of disease or in the assessment of arsenic intake (e.g., inadequate consideration of arsenic levels in food). A strategy for effective remediation requires sufficient future work to establish reliable causation and dose response for arsenicosis in Bangladesh. Arsenic-induced skin cancers require a latency of about 20 yr before manifestation. Hence, skin cancers have not yet been widely reported in Bangladesh populations exposed to arsenic. To reliably link arsenic as the causal agent for skin cancer, a cancer incidence rate statistically greater than the prevailing background level is required. It is commonly assumed (by regulators) that carcinogens induce their effects in a linear function with dose. Accordingly, exposure to arsenic may be assumed to present a finite risk of skin cancer irrespective of exposure. Experience with patients treated with Fowler’s solution (which contains arsenic) shows the absolute risk of skin cancer to be about 4%/g of arsenic intake. Although fraught

Introduction to Arsenic Contamination and Health Risk Assessment

11

with uncertainty, one can extrapolate this risk to low-exposure environmental scenarios. The absolute risk of developing skin cancer after 30 yr of arsenic ingestion at the 10 mg/L WHO drinking water limit is 6  10–4, which is a low but significant level of risk. Epidemiologically based estimates of cancer incidence from arsenic exposure have been completed for Taiwan (Yu et al. 2003). Results produced by these authors indicate that the principal chronic debilitating conditions resulting from arsenic poisoning are cancers of the skin, lungs, and bladder (Chen et al. 1992). Estimates of the projected number of cancers from Bangladesh have been reported (Yu et al. 2003). To obtain these projections, the authors used the exposure rate from Bangladesh, the cancer incidence rate from Taiwan, and the dose–response data from West Bengal. It was estimated that 125,590 skin cancers and 3,250 internal organ cancers may occur in Bangladesh (Molla et al. 2004). Life-table analysis by this author, also using the exposure data from Bangladesh and the cancer incidence rate from Taiwan, has shown that there will be at least a doubling of lifetime mortality risk of internal cancer as a result of drinking arsenic-contaminated water. By virtue of its chronic nature, the morbidity and mortality associated with arsenic disease have serious debilitating consequences on those affected. The consequences of morbidity and mortality to arsenicosis can be captured in a single measurement called the disability-adjusted life years lost (DALY). One estimate found a total of 1908 DALY to be associated with arsenicosis, thus highlighting the importance of the problem (Molla et al. 2004).

VI

Risk Mitigation for Arsenic Contamination

Results of the risk assessment process form the basis for risk management. With the exception of cancer, all arsenic-induced toxic effects are likely to have exposure thresholds that can be managed by the imposition of appropriate exposure limits (e.g., WHO guideline limit for drinking water). Risk management relies on legislative mandates, cost-benefit considerations, sociocultural acceptance of control measures, and availability of suitable control options (NRC 1983).

A

Control Options

Water, suitable for drinking and cooking, is a basic necessity in Bangladesh where contamination is widespread in rural areas and remote villages. Although control options are available and control technology is improving, each option has advantages and limitations, and no single control option is practical under all circumstances. Among the most widely used options for securing safe water is exploitation of deep aquifers. Aquifers generally yield microbiologically safe water, but they are prone to contamination by arsenic and other heavy metals. Therefore, in addition to

12

D. Caussy, N.D. Priest

the cost of sinking a tube well in an aquifer, one must consider the risk of finding impure water in the deep aquifer and the continuing cost to regularly monitor the aquifer for water purity during its natural life. The latter measure is important because arsenic-free aquifers can become contaminated at any time. Although supplies of piped arsenic-free water are rapidly becoming available to all parts of West Bengal, the high cost limits its use in many areas. Rainwater harvesting is also viable in areas with abundant rainfall but is of limited use in arid areas. Classical methods for purifying water of arsenical contamination rely on filtration, adsorption, and precipitation. Simple domestic filters, such as the ‘‘three-jug filter,’’ are commonly used in the region and are reasonably effective provided that the initial concentrations of arsenic in the water are not too high. Similarly, ‘‘teabag’’ adsorption and medium-level purification plants have been effectively employed. Although these methods can be applied widely, certain epidemiological and risk assessment issues must be resolved before their use. First, the methods must undergo environmental technology verification (ETV) by competent internationally recognized bodies and national control authorities. These entities validate that the particular method or technology employed accomplishes its aim. Furthermore, performance of each technology or method must be validated with cogent coprecipitant molecules such as iron, phosphates, or nitrates that may compete with arsenic during water purification (Chen et al. 1999). Second, the lower and upper concentration limits of arsenic that can be removed by each remediation method must be realistically defined. The current WHO drinking water guideline is 10 mg/L, but the prevalent national standard in Bangladesh is set at the higher and more achievable level of 50 mg/L. The success of risk mitigation depends on both technological feasibility and cost. Although, in countries of the Bengal River Basin, technologies may reach the standard of 50 mg/L for drinking water, it is important that, wherever possible, such technologies should be capable of yielding arsenic-safe water to a lower limit of at least 10 mg/L, in anticipation of future requirements. Furthermore, the cost of newer remediation technologies must be affordable to consumers. Nevertheless, a technology that removes 50% of arsenic from water should not be rejected only because it is incapable of reducing arsenic levels to the drinking water guideline value, because a reduction of 50% comprises a significantly lower risk of adverse effects including cancer.

B

Socioeconomic Factors

Even if mitigation technology is feasible and affordable, ultimate success is dependent on acceptance by the target population. Epidemiological observations from Bangladesh demonstrate that consumption of arsenic-free water could be managed by simply directing villagers to switch from an arsenic-positive (red-painted) to an arsenic-negative (green-painted) tube well in the same village. However, cultural

Introduction to Arsenic Contamination and Health Risk Assessment

13

beliefs prevent widespread acceptance of this approach (Hanchett et al. 2002; van Geen et al. 2002). A study of water supplies in Bangladesh shows that tube wells are regarded by local populations to be inherently safe because their historic advent coincided with the disappearance of waterborne bacterial pathogens (Caldwell et al. 2003). Studies sponsored by the World Bank have shown that arsenic-affected communities are willing to pay for arsenic-safe water. The challenge, then, is to establish the acceptability to communities of remediation technologies, taking into account their beliefs, attitudes, and practices.

C

Legal and Ethical Considerations

Risk management aimed at reducing arsenic concentration to safe levels carries both legal and ethical implications. If risk mitigation measures are predicated on national governmental policy, then national authorities must have the infrastructure and manpower to monitor and enforce the imposed standards. Similarly, the affected community has a right to demand uniform application of such standards. In the absence of national policies or guidelines for managing arsenic risk, the international vendors that market arsenic removal technologies must be ethical to ensure that their products conform to international standards.

Summary The problem of arsenic contamination in the Bengal River Basin illustrates a classic conundrum in environmental health, namely, that development projects can have double effects: on one hand development of tube wells eliminated bacterial pathogens and on the other it exposed the population to poisoning from arsenic. Thus, in future development projects the full health risk of a project must be considered during the planning, implementation, and decommissioning phases (Caussy 2003b; Caussy et al. 2003b). If such a holistic approach would have been followed, the mass contamination in the Bengal River Basin, in which millions of people were and are exposed to unsafe levels of arsenic, could have been averted. Although definite knowledge gaps in applying risk assessment steps for arsenic contamination exist, arsenic clearly poses a serious health problem and economic consequences to the affected population of the Bengal River Basin. It is binding on the international community to alleviate the problem through remediation measures to reduce arsenic exposure. One Environmental Sustainability Millennium development goal is to increase the proportion of population with sustainable access to an improved water source (Bartram et al. 2005). Providing water with safe levels of arsenic to affected communities of the Bengal River Basin will directly contribute to improved community health.

14

D. Caussy, N.D. Priest

References ATSDR (2000) Toxicological Profile for Arsenic (update). U.S. Department of Health and Human Services, Atlanta. Bartram J, Lewis K, Lenton R, Wright A (2005) Focussing on improved water and sanitation for health. Lancet 365:810–812. BGS (2001) Arsenic contamination of groundwater in Bangladesh. In: Kinniburgh DG, Smedley P (eds) British Geological Survey, Keyworth. Caldwell BK, Caldwell JC, Mitra SN, Smith W (2003) Searching for an optimum solution to the Bangladesh arsenic crisis. Soc Sci Med 56:2089–2096. Caussy D (2003a) Case Studies of the impact of understanding bioavailability: arsenic. Ecotoxicol Environ Saf 56:164–173. Caussy D (2003b) Health Risk Assessment: Perspectives from the WHO. RISK Newsletter 9–10. Caussy D (2003c) Normative role of WHO in mitigating health impacts of chronic arsenic exposure in South-east Asia Region. In: Chappel W, Abernathy CO (eds) vol V. Elsevier, Amsterdam, pp 437–445. Caussy D (ed) (2005) A Field Guide for Detection, Management and Surveillance of Arsenicosis Cases. World Health Organization, New Delhi. Caussy D, Than Sein U (2006) Health Risk Assessment Arsenic Contamination in the South East Asia Region: Evidence-based practices in norm setting. In: Naidu R, Smith E, Owens G, Bhattacharya P, Nadebaum P (eds) CSIRO, Melbourne, pp 483–493. Caussy D, Gochfeld M, Gurzau E, Neagu C, Ruedel H (2003a) Lessons from case studies of metals: investigating exposure, bioavailability, and risk. Ecotoxicol Environ Saf 56:45–51. Caussy D, Kumar P, Than Sein U (2003b) Health impact assessment needs in South-East Asian countries. Int J Public Health 81:439–443. Cebrian ME, Albores A, Aguilar M, Blakely E (1983) Chronic arsenic poisoning in the north of Mexico. Hum Toxicol 2:121–133. Chakraborty AK, Saha KC (1987) Arsenical dermatosis from tubewell water in West Bengal. Indian J Med Res 85:326–334. Chen C-J, Wang C-J (1990) Ecological correlation between arsenic level in well water and age-adjusted mortality from malignant neoplasms. Cancer Res 50:5470–5474. Chen C-J, Chuang Y-C, Lin T-M, Wu H-Y (1985) Malignant neoplasms among residents of a blackfoot disease-endemic area in Taiwan: high arsenic Artesian well water and cancers. Cancer Res 45:5895–5899. Chen C-J, Chen CW, Wu MM, Kuo TL (1992) Cancer potential in liver, lung, bladder and kidney due to ingested inorganic arsenic in drinking water. Br J Cancer 66:888–892. Chen HS, Frey MM, Clifford D, McNeil S, Edwards M (1999) Arsenic treatment considerations. J Am Water Works 91:74–85. Hanchett S, Nahar Q, Agthoven AV, Geers C, Rezvi FJ (2002) Increasing awareness of arsenic in Bangladesh: lessons from a public education programme. Health Policy Plan 17:393–401. IARC (1982) Evaluation of the Carcinogenic Risk of Chemicals to Humans, Industrial Processes and Industries Associated with Cancer in Human. International Agency for Research on Cancer, Lyon. IPCS (2001) Environmental Health Criteria 224. Arsenic and Arsenic Compounds, 2nd Ed. World Health Organization, Geneva. Ishinishi N, Tsuchiya K, Vahter M, Fowler BA (1986) Chapter 3 Arsenic. In: Handbook on the Toxicology of Metals. Eds: Friberg L, Nordberg GT, Vouk VB. Elsevier, Amersterdam, pp 43–83. McArthur JM (2002) A layman’s guide to arsenic pollution of groundwater in Bangladesh and West Bengal (personal communication). McLellan F (2002) Arsenic contamination affects millions in Bangladesh. Lancet 359:1127.

Introduction to Arsenic Contamination and Health Risk Assessment

15

Molla A, Anwar KS, Hamid SA, Hoque ME, Haque AKZM (2004) Analysis of disability adjusted years (DALYs) among arsenic victims: a cross-sectional study on health economics perspective. Bull Bangladesh Med Res Council 30:43–50. NRC (1983) Risk Assessment in the Federal Government: Managing the Process. National Academic Press, Washington, DC. NRC (1999) Arsenic in Drinking Water. National Academic Press, Washington. Smedley P (2003) Arsenic in groundwater: South and East Asia. In: Welch A, Stollenwerk K (eds) Kluwer, Boston, pp 179–209. van Geen A, Ahsan H, Horneman AH, Ahmed KM, Gellan A, Stute M, Wimpson HJ, Wallace S, Small C, Parvez F, Lolacono N, Becker M, Cheng Z, Momota H, Shanewaz M, Seddique AA, Graziaon JH (2002) Promotion of well-switching to mitigate the current arsenic crisis in Bangladesh. Bull WHO 80:732–737. Yu Chen Ahsan H (2004) Cancer burden from arsenic in drinking water in Bangladesh. Am J Pub Health 94:741–744. Yu Winston H, Harvey CM, Harvey CF (2003) Arsenic in groundwater in Bangladesh: a geostatistical and epidemiological framework for evaluating health effects and potential remedies. Water Resour Res 39:1146–1162.

Arsenic Pollution Sources Hemda Garelick(*), Huw Jones, Agnieszka Dybowska, and Eugenia Valsami-Jones

I II III

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Arsenic Speciation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Natural Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 A Arsenic Minerals in the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 B Arsenic Occurrence and Concentrations in Rock-Forming Minerals . . . . . . . . . . 21 C Arsenic Occurrence and Concentrations in Rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 D Groundwater Contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 IV Anthropogenic Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 A Modern and Historical Uses of Arsenic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 B Mining as a Source of Arsenic in the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 C Arsenic Emissions from Coal Burning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 D Agricultural Use of Arsenic Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 E Wood Preserving Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

H. Garelick Department of Natural Sciences, School of Health and Social Sciences, Middlesex University, The Burroughs, London NW4 4BT, UK, e-mail: [email protected] H. Jones Institue of Social and Health Research, School of Health and Social Sciences, Middlesex University, The Burroughs, London NW4 4BT, UK, e‐mail: [email protected] A. Dybowska Department of Mineralogy, The Natural History Museum, Cromwell Road, London SW7 5BD, UK E. Valsami-Jones Department of Mineralogy, The Natural History Museum, Cromwell Road, London SW7 5BD, UK

D.M. Whitacre (ed.), Reviews of Environmental Contamination Volume 197. doi: 10.1007/978-0-387-79284-2_2, # Springer Science þ Business Media, LLC 2008

17

18

H. Garelick et al.

I Introduction Arsenic is a group V element, together with nitrogen, phosphorus, antimony and bismuth. Its electronic configuration is [Ar]3d104s24p3; it has an atomic weight of 75 and commonly occurs naturally in two oxidation states, þ5, þ3, and, more rarely, in the 0 or –3 state. Arsenic is classified as a metalloid in that it has a chemical nature intermediate between that of metals and nonmetals. The environmental presence of arsenic derives from both natural and anthropogenic sources. Many natural processes contribute to environmental background concentrations of arsenic, including pedogenesis, dust storms, volcanic eruptions, geothermal/hydrothermal activity, and forest fires. Arsenic is widely distributed and is present in the Earth’s crust at an average abundance of about 5 mg/kg. It occurs naturally in more than 200 mineral forms, of which approximately 60% are arsenates, 20% sulfides and sulfosalts, with the remaining 20% comprising arsenides, arsenites, oxides, silicates, and elemental arsenic (Onishi 1969). The main mineral hosts of arsenic are arsenopyrite, orpiment, and realgar; these are As(III) compounds that were formed under reducing, subsurface conditions. When exposed to surface conditions of the Earth, such compounds are oxidized to form pentavalent forms such as iron arsenate and, in calcareous soils, as calcareous arsenolite. Arsenic, in its most recoverable form, occurs as a trace constituent in iron pyrite, galena, and chalcopyrite (Goldschmidt 1954). Arsenic contamination in the environment is causing a significant global human health problem. It has been estimated that 60–100 million people, in India and Bangladesh, are currently at risk of arsenic-related disease as a result of drinking arsenic-contaminated waters (Ahmad 2001; Chakraborti et al. 2002). Arsenicosis is also prevalent in certain areas of China including Shanxi, Xinjiang and Inner Mongolia (Wang et al. 2000; Guo et al. 2001) and in Taiwan (Chen et al. 1999), Vietnam (Berg et al. 2001) and Nepal (Neku and Tandukar 2003). In these areas of endemic contamination, the major arsenic exposure pathway is believed to be from drinking groundwater contaminated with natural (geological) sources of arsenic. Similarly, geologically based arsenic contamination occurs in areas where geothermal fluids rich in arsenic enter surface waters, e.g., geothermally active regions of Yellowstone National Park in the United States (US) and the Taupo Volcanic Zone in New Zealand (Webster and Nordstrom 2003). Significant amounts of arsenic are also introduced into the environment from anthropogenic sources, among which metal mining and smelting are the most important. Arsenic pollution of soils, sediments, surface waters, and groundwater is reported from mining districts worldwide. Mining-related arsenic pollution has been reported in the US (Moore and Woessner 2003), Australia (Smith et al. 2003), Brazil (Borba et al. 2003), Northern Chile (Oyarzun et al. 2004), Czech Republic (Filippi et al. 2004), Northern Peru (Bech et al. 1997), Taiwan (Shih and Lin 2003), Hetao area of China (Zhang et al. 2002), Portugal (Patinha et al. 2004), Iran (Modabberi and Moore 2004) and Mexico (Carillo-Cha´vez et al. 2000). In some parts of the world, coal mining and burning of arsenic-rich coals have been reported as an important source of arsenic contamination. Elevated concentrations of

Arsenic Pollution Sources

19

arsenic from coal burning were reported in Guizhou Province (Dai et al. 2005) and other areas of China (Sun 2004). The major exposure pathway for the population in these areas of China has been linked to ingesting food contaminated by arsenic deposited during burning of coal for cooking, crop drying and heating (Liu et al. 2002). Arsenic has also been introduced into the environment through extensive use of arsenical compounds in agriculture. Until the 1970s, approximately 80% of arsenic production was for the manufacture of pesticides in the form of simple inorganic salts. Today, approximately 50% of arsenic production is dedicated to making pesticides, with organic arsenic compounds comprising the more important forms (Matschullat 2000). Wood preservatives account for another 30% of the world arsenic market. Arsenic has long been an important constituent of wood preservative formulations, among which the most commonly used was chromated copper arsenate (CCA), which contains 22% pure arsenic. Many sites used by the timber industry to treat lumber, dispose of treated lumber, or use treated timber in construction (including children’s playgrounds) have been contaminated with arsenic.

II

Arsenic Speciation

Similar to other metalloids, arsenic exists in a number of forms or species. Both bioavailability and toxicity are influenced by the type of arsenic species present. The predominant form of arsenic, in solution, is controlled by both redox potential (Eh) and pH. The stability of different arsenic aqueous species is shown as a function of pH for As(III) (Fig. 1a) and As(V) (Fig. 1b). The stability of arsenic is shown in a combined Eh–pH diagram (Fig. 2), which allows predictions of the most stable form of arsenic. The system modeled also includes oxygen, water, and sulfur. Soluble species and solids with solubilities sufficiently low to occur in the system are illustrated. Regions of solubility less than 10–5.3 mol/L are shaded. From Fig. 2 it can be seen that arsenates are stable (H3AsO4, H2AsO4–, HAsO42–, and AsO43–) at the high Eh values of oxygenated (surface) waters. The following arsenite species occur at mildly reducing conditions (i.e., low Eh values): H3AsO3, H2AsO3–, and HAsO32–. Arsenic oxides are too soluble to appear on the diagram. Under conditions in which S2– is stable in its reduced form, the minerals realgar (AsS) and orpiment (As2S3) may be formed at pH values below 5.5 and when Eh values are near 0. HAsS2(aq) is stable at low pH in the presence of sulfide and arsenic concentrations of up to 10–6.5 mol/L. At low Eh values As0, i.e., pure arsenic in its metallic form, is stable and, at very low Eh values, arsine (AsH3) may be formed. Arsenic concentrations lower than 10–5.3 mol/L (Fig. 2) will diminish the predominance of solid species. Sulfide concentrations influence the boundary between arsenic sulfides and arsenic metal. Organic species are not shown in Fig. 2 because they are stable only at extremely low Eh values (Gupta and Chen 1978).

20

H. Garelick et al.

Fig. 1 Arsenic (III and V) predominance diagrams. (Modified from Gupta and Chen 1978.)

Eh–pH diagrams provide no information on kinetics. The oxidation of arsenite to arsenate in the presence of oxygen is reported to be very slow at neutral pH but proceeds faster in strongly alkaline or acid solutions.

III A

Natural Sources Arsenic Minerals in the Environment

Arsenic occurs as a major constituent in more than 200 minerals including elemental arsenic, arsenides, sulfides, oxides, arsenates, and arsenites. The most commonly found arsenic minerals are listed in Table 1.

Arsenic Pollution Sources

21

Fig. 2 Redox (Eh)-pH diagram for arsenic at 25 C and 1 atm with total As =10–5.3 mol/L and total S = 10–3 mol/L. Solid species depicted in parenthes-es in grey (shaded) area. (Adapted from Ferguson and Gavis 1972.)

The greatest concentrations of these minerals occur in mineralized areas associated with the presence of cadmium, lead, silver, gold, antimony, phosphorus, tungsten and molybdenum. The most abundant arsenic ore mineral is arsenopyrite, FeAsS, which contains 46% arsenic by mass.

B

Arsenic Occurrence and Concentrations in Rock-Forming Minerals

Sulfide Minerals Arsenic is present in varying concentrations in many common rock-forming minerals. The typical ranges of recorded concentrations are presented in Table 2. Because the chemistry of arsenic has some similarities to that of sulfur, the greatest concentrations of the element tend to occur in sulfide minerals, of which pyrite is the most abundant. Concentrations in pyrite, chalcopyrite, galena and marcasite can be quite variable, even within a given grain, but in some cases exceed 10% by mass (Smedley and Kinniburgh 2002). Pyrite can be formed in low-temperature sedimentary environments under reducing conditions. It is present in the sediments of many rivers, lakes, and oceans and also occurs in many aquifers. Pyrite commonly forms preferentially in zones of intense reduction such as around buried plant roots or other nuclei of decomposing organic matter. Pyrite is not stable in aerobic systems and oxidizes to Fe oxides, with the release of large amounts of sulfate ions, acidity and associated trace constituents, including arsenic. The presence of pyrite as a minor constituent in

22

H. Garelick et al.

Table 1 Arsenic minerals occurring in nature Mineral Composition Occurrence Native arsenic As Hydrothermal veins and deposits that contain other arsenic minerals Niccolite NiAs A minor component of Ni-Cu ores in hightemperature hydrothermal veins Hydrothermal veins Safflorite (Co,Fe)As2 Realgar AsS Vein deposits often associated with orpiment, clays, and limestones, deposits from hot springs Hydrothermal veins, hot springs, volcanic Orpiment As2S3 sublimation products Cobaltite CoAsS Medium-temperature hydrothermal deposits, metamorphic rocks Arsenopyrite FeAsS The most abundant arsenic mineral, dominantly in veins of hydrothermal origin, found in pegmatites, high-temperature gold, quartz and tin veins, in contact metamorphic sulfide deposits, also in gneisses, schists and other metamorphic rocks Hydrothermal veins, accessory mineral in igneous Arsenian pyrite Fe(As,S)2 rocks, pegmatites and contact metamorphic deposits Mesothermal deposits associated with other Lollingite FeAs2 sulfides and calcite gaunge Hydrothermal veins and contact metamorphic Tennantite (Cu, Fe)12As4S13 deposits Hydrothermal vein deposits formed at medium Enargite Cu3AsS4 temperatures Secondary mineral formed by oxidation of FeAsS, Arsenolite As2O3 native arsenic and other arsenic minerals Secondary mineral formed by oxidation of Scorodite FeAsO42H2O arsenic-bearing sulfides Secondary mineral formed by the alteration of CoAnnabergite (Ni,Co)3(AsO4)28H2O Ni-bearing arsenides and sulfides, in the oxidized zone of hydrothermal mineral deposits Secondary mineral, in limestone blocks and Hoernesite Mg3(AsO4)2H2O volcanic tuff Secondary mineral in the oxidized zone of some Symplesite Fe2+3(AsO4)28H2O arsenic-rich hydrothermal mineral deposits Secondary mineral in the oxidized zone of Cu Conichalcite CaCu(AsO4)OH deposits, an alteration product of enargite Pharmacosiderite Fe3(AsO4)2(OH)35H2O Secondary mineral formed by oxidation of FeAsS and other arsenic-bearing sulfides Source: Modified from Goldschmidt (1954); Anthony et al. (1990); Azcue and Nriagu (1994); Anthony et al. (2000); Smedley and Kinniburgh (2002).

Arsenic Pollution Sources

23

Table 2 Typical arsenic concentrations in rock-forming minerals Mineral Arsenic concentration (mg/kg) Sulfide minerals Arsenopyrite 46% by mass Pyrite 100–77,000 Pyrrhotite 5–100 Marcasite 20–126,000 Galena 5–10,000 Sphalerite 5–17,000 Chalcopyrite 10–5,000 Oxide minerals Haematite Up to 160 Fe oxide (undifferentiated) Up to 2,000 Fe III oxyhydroxide Up to 76,000 Magnetite 2.7–41 Ilmenite <1 Silicate minerals Quartz 0.4–1.3 Feldspar <0.1–2.1 Pyroxene 0.08–1.17 Carbonates Calcite 1–8 Dolomite <3 Siderite <3 Other minerals Apatite <1–1,000 Fluorite <2 Source: From Smedley and Kinniburgh (2002).

sulfide-rich coals is ultimately responsible for arsenic pollution in the vicinity of coal mines and in areas of intensive coal burning.

Oxide Minerals High arsenic concentrations are also found in many oxide minerals and hydrous metal oxides, either as part of the mineral structure or as sorbed species. Concentrations in iron oxides can attain weight percent values, particularly where they form as oxidation products of primary iron sulfide minerals, which have abundant arsenic. Adsorption of arsenate to hydrous iron oxides is particularly strong (Goldberg 1986; Manning and Goldberg 1996; Hiemstra and van Riemsdijk 1996). Adsorption to hydrous aluminum and manganese oxides may also be important, if these oxides are present in quantity (Peterson and Carpenter 1983; Brannon and Patrick 1987). Arsenic may also sorb to the edges of clays and on the surface of calcite (Goldberg and Glaubig 1988), a common mineral in many sediments. However, all these loadings are much smaller on a weight basis compared to arsenic loadings on iron oxides.

24

H. Garelick et al.

Phosphate and Other Minerals Arsenic concentrations in phosphate minerals are variable but can reach high values, up to 1000 mg/kg in apatite (Smedley and Kinniburgh 2002). However, phosphate minerals are much less abundant than oxide minerals and thus make a correspondingly smaller contribution to the arsenic loading of most sediments. Arsenic can also substitute for Si4þ, Al3þ, Fe3þ, and Ti4þ ions in many mineral structures and is therefore present in many other rock-forming minerals, although at much lower concentrations. Most common silicate minerals contain approximately 1 mg/kg or less arsenic, and carbonate minerals usually contain less than 10 mg/kg of arsenic. Comprehensive data on arsenic concentrations in common rock forming minerals are presented by Smedley and Kinniburgh (2002).

C

Arsenic Occurrence and Concentrations in Rocks

Igneous Rocks Arsenic concentrations in igneous rocks are usually low, and there is relatively little difference between concentrations in different igneous rock types. Among igneous rocks, volcanic rocks (especially ashes) are often implicated in the generation of high aqueous arsenic concentrations. Arsenic concentrations in most metamorphic rocks are similar to the average crust concentrations of 5 mg/kg or less. Higher arsenic concentrations are typically associated with pelitic rocks (slates, phyllites) with average values of arsenic of 18 mg/kg (Goldschmidt 1954).

Sedimentary Rocks The concentration of arsenic in sedimentary rocks is typically in the range 5–10 mg/kg (Webster 1999), and average sediment concentrations are enriched in arsenic relative to igneous rocks. Sands and sandstones tend to have the lowest concentrations, reflecting the low arsenic concentrations of their dominant mineral components, i.e., quartz and feldspars. The average sandstone arsenic concentration is approximately 4 mg/kg. Argillaceous deposits have a broader range and higher average arsenic concentration than do sandstones, with a typical value of approximately 13 mg/kg (Ure and Berrow 1982). The higher values reflect the larger proportion of sulfide minerals, oxides, organic matter and clays. Black shales have relatively high arsenic concentrations, mainly because of their enhanced pyrite content. Arsenic concentrations in coals and bituminous deposits are variable but often high. Some coal samples have been found with extremely high arsenic concentrations, up to 35,000 mg/kg (Belkin et al. 2000), although

Arsenic Pollution Sources

25

generally low concentrations (2.5–17 mg/kg) were reported by Palmer and Klizas (1997). Carbonate rocks typically have low concentrations, reflecting arsenic concentrations of their constituent minerals (3 mg/kg). Some of the highest observed arsenic concentrations are found in ironstones and Fe-rich rocks; concentrations up to 2900 mg/kg were reported by Boyle and Jonasson (1973). Phosphorites are also relatively enriched in arsenic, with reported values of up to 400 mg/kg (Smedley and Kinniburgh 2002).

D

Groundwater Contamination

Occurrence and Sources The occurrence, origin and mobility of arsenic in groundwater are primarily influenced by local geology, hydrogeology, and geochemistry of the aquifer sediments. Elevated arsenic concentrations may arise from natural weathering and leaching of arsenic from geological materials containing arsenic (e.g., arsenopyrite, base metal sulfides, realgar, orpiment), and drainage from thermal springs and geysers, as well as atmospheric deposition. Arsenic sulfides, arsenic-rich pyrite, and arsenic-rich iron oxyhydroxides are the most commonly found natural sources of arsenic contamination in groundwater worldwide. Various studies have suggested that the primary source of arsenic in groundwater derives from dissolution of arsenic-bearing sulfide minerals, or desorption from, or reductive dissolution of, arsenic-rich Fe oxides. Arsenic contamination of groundwater from natural sources worldwide has been attributed to several geochemical processes. These pathways include oxidation of arsenic-bearing sulfides, desorption of arsenic from oxides and hydroxides, reductive dissolution of arsenic from oxides/hydroxides, release of arsenic from geothermal waters and evaporative concentration (Welch et al. 2000), and leaching of arsenic from sulfides by carbonate (Kim et al. 2000). Because these processes are complex, idengification of the mechanism responsible for contamination in a particular case is often difficult. At times, despite the natural origins of contamination (e.g., the presence of arsenic-rich minerals) natural processes of arsenic leaching from the host minerals have been significantly enhanced by anthropogenic activities such as mining. In Bangladesh and West Bengal, the alluvial Ganges aquifers used for public water supplies are polluted with naturally occurring arsenic. Several researchers attribute arsenic contamination in this region directly to oxidation of arsenic-rich pyrite in the aquifer sediments as atmospheric oxygen invades the aquifer in response to lowering of the water level by abstraction (Das et al.1996; Saha and Chakrabarti 1995; Chowdhury et al. 1999; Mandal et al. 1996). Other researchers suggest that arsenic contamination comes from the reductive dissolution of arsenic-rich iron oxyhydroxides, which are derived from weathering

26

H. Garelick et al.

of base metal sulfides (Nickson et al. 1998, 2000; Chowdhury et al. 1999; Bhattacharya et al. 1997; McArthur et al. 2001). Further pathways for release of arsenic into drinking water have been described by Islam et al. (2004). These workers have shown microbial mediation to be a possible mechanism for reductive dissolution of arsenic-bearing hydrated ferric oxides. Other important factors that potentially influence the occurrence of arsenic in water include the presence of organic carbon species that correlate with the arsenic levels found in sand and gravel glacial aquifers in central Illinois (Kelly et al. 2005). Further mechanisms of arsenic release include requires indenting the oxidation of the arsenic-bearing sulfides pyrite and marcasite. This mechanism was the source of arsenic contamination in groundwater from a sandstone aquifer in the Fox River Valley in eastern Wisconsin, USA (Schreiber et al. 2003). Pyrite, with arsenic-rich domains [up to 8.5 arsenic weight percent (wt.%)], was identified in the Mississippian Marshall Sandstone, the principal bedrock aquifer in the region. Arsenic-rich Fe oxyhydroxides (up to 0.7 wt.% arsenic) were also identified in samples of tills derived from the Marshall Sandstone. Depending on redox conditions, Fe oxyhydroxides provide a potential source of arsenic in groundwater. Kolker et al. (2003) concluded that three different mechanisms were responsible for widespread arsenic mobilization in eastern and southeastern Michigan groundwaters: weathering of pyrite in tills derived from the Marshall Sandstone, reductive dissolution of Fe oxyhydroxides in tills, and potential pyrite oxidation in bedrock aquifers. Other data reported in the literature on arsenic contamination of groundwater derived from natural sources are summarized in Table 3. Therefore, the major hypothesized arsenic release mechanisms include sulfide oxidation and reductive dissolution of As-bearing phases (Fe-Mn oxides), as well as microbially enhanced reductive dissolution. In addition, on a local scale, factors such as anion competition (PO43– leaching from fertilizers) enhance mobilization. Many high-concentration arsenic-bearing groundwaters exist in areas where only moderate levels of arsenic occur in the host formation (and often within normal range). Stuben et al. (2003) suggest that the mechanism of solubilization is more important than source type in accounting for such behavior. The mechanism for solubilizing arsenic from relatively low content host formations to create high arsenic levels in groundwater is often related to redox conditions and pH of the water, as previously described. The presence of peat deposits in sediments may contribute to arsenic release, because peat degradation creates acidic and reducing conditions, and, in addition, soluble organics may compete with arsenic for sorption sites.

Arsenic Enrichment in Areas of Geothermal Activity Arsenic is a ubiquitous component of active and fossil geothermal systems. A summary of arsenic concentrations in selected geothermal systems is provided in

41

1,572

1–380

10–210

88

From 5 to >50 Highest detected in metamorphosed marine sediments. 29% wells >10, 14% >20 and 4% >50 g/L

New England, USA

Central Arizona USA (Verde Valley)

992

<0.0003–180

New Hampshire, USA

Supai and Verde formations, organic As complexes common: arsenobetaine, dimethylarsinic acid, As-glutathione complex

Meta-sedimentary bedrock Highest arsenic in groundwater (GW) from metamorphosed marine sediments variable calcareous, with As present in sulfide minerals (pyrite, pyrrhotite) NA

As-enriched pegmatite dikes in the bedrock

Arsenopyrite (FeAsS), westerveldite (FeAs)

NA

0.2–554

Meta-sedimentary bedrock: 3–40

Pegmatites: up to 60

ND

ND

Dissolution of nanocrystalline Asbearing phases Rock–water interaction: dissolution from pegmatites ND

Table 3 Geographic locations, levels, and sources of naturally enriched arsenic contamination in groundwater Location Arsenic concentrations N Identified source Arsenic Mobilization mechanism range/mean (mg/L) concentration (source material) (mg/kg)

(continued)

Robinson and Ayotte 2006 Foust et al. 2004

Ayotte et al. 2003

Peters and Blum 1999; Utsunomiya et al. 2003

Reference

Arsenic Pollution Sources 27

N

73

262

Table 3 (continued) Location Arsenic concentrations range/mean (mg/L)

0.5–278

1–1310 Most common levels: 5–50

Michigan, USA

Arsenopyrite and As-rich pyrite present in Marshall sandstone

Marshall Formation and Coldwater shale, both of Mississipian age: arsenic-rich pyrite, arsenopyrite and glacial deposits (arsenic adsorbed on Fe oxyhydroxides)

Identified source

Carbonation of arsenic sulfides, reduction of As sorbed in the aquifer Weathering of arsenic pyrite Carbonation of sulfide minerals Arsenopyrite oxidation primary mechanism for As release in the vadose zone but not the primary factor controlling As in groundwater

As-rich pyrite: up to 8.5 wt%

FeOOH: up to 0.7 wt% As NA

Reductive dissolution of arsenic-rich iron hydroxide/ oxyhydroxide Dissolution of arsenic sulfide minerals

Mobilization mechanism

NA

Arsenic concentration (source material) (mg/kg)

Slotnick et al. 2003

Kolker et al. 2003

Kim et al. 2002

Reference

28 H. Garelick et al.

India, West Bengal

10–1800 (mean = 500) Mean = 135 <0.15–3590 (median As, from 0.15 to 140) 50–100 in 20% samples, 100–500 in 60% samples, 500–1000 in 15% samples, and >1000 in 5% samples

22

20–1200; mean, 395

North-East Taiwan

NA

NA NA NA

As associated with Fe pyrites

<1,000

NA

Likely to be pyritic material or black shale NA

Sulfide-bearing horizon in sandstone aquifer; As occurs in pyrite, marcasite, but not arsenopyrite, also As-rich FeOOH

126 377 3,900

54

470–897; mean, 671  149 0–1000

South-West Taiwan

5

0.3–166 (Schreiber et al. 2000)

Wisconsin, USA

NA

NA NA NA

NA

NA

NA

NA

ND

Sulfide oxidation (atmospheric oxygen introduced through well boreholes, extensive groundwater withdrawal exposes sulfide-bearing horizon to oxygen) Oxidation of pyrigiferous material Reductive dissolution of As-rich Fe oxyhydroxides Pyrite oxidation and siderite dissolution ND ND ND

(continued)

Das et al. 1994

Kuo 1968 Hsu et al. 1997 Yang et al. 2003

Lin et al. 2006

Wang et al. 2007

Chen et al. 1994

Schreiber et al. 2000

Arsenic Pollution Sources 29

1,420

21

0.74–470

N

50–1250

Table 3 (continued) Location Arsenic concentrations range/mean (mg/L)

As-rich base metal deposits, As-bearing pyrite

Iron pyrites in arsenic-rich layers in the alluvium alongside River Ganga

Identified source

Arsenic concentration (source material) (mg/kg) Aquifer sediments: 1–31 mg/kg As, max. 125 and 196 mg/kg As sorbed to FeOOH: 300–500 Oxidation of As-rich pyrite, arsenopyrite (lowering of water ) Microbially mediated reductive dissolution of FeMnoxyhydroxides (changing redox conditions, presence of organic material) Not clear what drives creation of reducing conditions (flooding, biodegradation of buried peat deposits, human-related activities, pumping, land use, sewage pollution, or intensive GW exploitation that promotes movement of reducing GW in the aquifer

Mobilization mechanism

Stuben et al. 2003

Das et al. 1996

Reference

30 H. Garelick et al.

NA

37.5–450

As in the silty clay and sandy layers of the aquifer sediments – occurs as coatings on mineral grains, occasionally also as arsenopyrite As sorbed on amorphous Fe oxides, carbonates and micas

As in pyrite

20,000

NA

NA

4,800

10–590

50–3700 (mean, 290) Mean, 200; maximum, 3700

Total As in sediment sample: 3.0 (34%) associated with amorphous FeIII oxides and 19% with carbonates/acid volatile sulfides)

11–78 (As in borehole sediments)

NA

Anion exchange of sorbed arsenate and aqueous HCO3– (and to a smaller extent PO43–) Arsenic release from the dissolution of Asenriched carbonates

Oxidation of Asenriched pyrite caused by high GW withdrawal

ND

(continued)

Charlet et al. 2007

Bhattacharya et al. 1997

Chowdhury et al. 1997 Mandal et al. 1996

Arsenic Pollution Sources 31

Bangladesh

3,534

112 45

0.7–640 <10–332

N

<0.5–2500

Table 3 (continued) Location Arsenic concentrations range/mean (mg/L)

Holocene/alluvial deltaic sediments - micaceous quartzo-feldspathic sands Arsenic rich FeOOH

Identified source

As in sediments: 9–28

Arsenic concentration (source material) (mg/kg) Sediments: 1–30 max. 196 mg/kg As (where pyrite locally occurred and scavenged arsenic)

Reductive dissolution of FeOOH driven by microbial degradation of sedimentary organic matter

Microbial dissolution of FeOOH (reducing conditions created mainly by microbial degradation of buried peat deposits and organic waste) Distribution of As pollution controlled by buried peat deposits rather than distribution of As in aquifer sediments Contamination related to well depths (hand-dug wells mostly unpolluted, 30–45 m the highest pollution level)

Mobilization mechanism

Friesbie et al. 2002 Nickson et al. 2000

BGS/DPHE 2001 McArthur et al. 2001 Nickson et al. 1998 Chowdhury et al. 1999

Reference

32 H. Garelick et al.

66

NA

10–593 Mean, 108

Up to 5300

Argentina, Pampean Plain Central Argentina, La Pampa province

NA

470–770

North-west Argentine, Salta and Jujay provinces

Quaternary loess aquifer As associated with Fe oxides and Mn oxides

Tertiary-Quaternary volcanic deposits, post-volcanic geysers and thermal springs Levels in thermal springs: 50-9900 mg/L Loess and loess like sediments

Volcanic sediments, minerals and soil Loess and loess-like sediments (mostly pyroclastic material), and rhyolitic volcanic glass

23 NA

Lacustrine sediments (chalk and gypsum)

53

Maximum, 53.5 Majority of samples, 5–10 <100 to >800

100–316; maximum, 3810

Geological; not specified

Mean, 404

South-east Argentine, Pampa Province of Cordoba

Mexico, Lagunera region Mexico, Rioverde basin Northern Chile

Loess sediments: 3–18 (mean = 8)

NA

Volcanic glass: 6.8–10.4 NA

Loess sediments: 5.5–37.3

NA

NA

NA

Desorption from Fe oxides caused by high pH and competition with P and HCO3–

Not specified

ND

ND

ND

ND

ND

(continued)

Smedley et al. 2005

Farias et al. 2003

De Sastre et al. 1992

Nicolli et al. 1989

Astolfi et al. 1981

Caceres et al. 1992

Planer-Friedrich et al. 2001

Del Razo et al. 1994

Arsenic Pollution Sources 33

11,673

46

Chicugo Plain, 11–370

Kumamoto Plain, 5–66 Fukui and Takatsuki Plains, 11–60

Japan, Kumamoto, Fukui, Takatsuki

67

1–293

Japan, southern region of Fukuoka Prefecture

196

1–3050 Mean, 159

N

Vietnam Red River Basin

Table 3 (continued) Location Arsenic concentrations range/mean (mg/L)

Quaternary sediments (mud, sand, gravel and clay-silt layers, volcanic ash) Arsenic sorbed on Fe-oxides and clay minerals

Quaternary formation with sediments beds of Pleistocene and Holocene age (with peat deposits) As present in iron Oxides phases, not abundant in sulphide phases No detailed source identified

Identified source

Sediments: Chicugo plain, 0.8–33 Kumamoto Plain, <0.5–20.6 Fukui Plain, <0.5– 14 Takatsuki Plain, 2–386

NA

Arsenic concentration (source material) (mg/kg) Sediments: 0.6–33

Elution of arsenite/ arsenate by anion exchange with OH– and reductive dissolution of Asbearing phases Dissolved from the host formation, desorption from Fe oxides by reductive dissolution and competing ions effect (PO4) Stagnant groundwater with reducing and alkaline conditions accelerated As release

Reductive dissolution

Mobilization mechanism

Shimada 1996

Kondo et al. 1999

Berg et al. 2001

Reference

34 H. Garelick et al.

100 >1,000

200

10–500

Mean of 210  310 and a maximum of 1700

1–1610 Mean, 217

Cambodia (Mekong Valley)

73

<1 to 1480

Inner Mongolia, northern China (Huhhot basin)

Holocene sediments

Quaternary alluvial sediments (most likely As sorbed to FeOOH) Holocene and Quaternary sediments (As in Fe bearing phases, carbonates, sulfides, and in organic matter)

Quaternary lacustrine and fluvial sediments Arsenic associated with Fe-oxides

NA

NA

As sorbed on FeOOH: 3–29

Reductive dissolution of oxides, competing ions effect (mainly PO4), diagenetic evolution of Fe oxides (transformation from amorphous to crystalline forms Microbially mediated dissolution of Asbearing FeOOH Microbially mediated reductive dissolution of As-rich hydrous Fe oxides (source of organic carbon driving these microbial processes disputed, e.g., peat layers within the sediments, surfacederived organic matter) Reductive dissolution of As-bearing Fe phases in aquifer sediments

(continued)

Berg et al. 2007

Polya et al. 2004, 2005; Charlet and Polya 2006

Smedley 2003

Smedley et al. 2003

Arsenic Pollution Sources 35

0–300 (mean, 51) (but samples found with up to 2250)

0.42–614 (mean, 41)

Hungary Pannonian Basin

Spain Duero Cenozoic Basin

514

NA

68

N

Pleistocene sediments with arsenic sorbed on amorphous Fe oxides

Pyritic sediments

Identified source

Pleistocene sediments: 0.09–10.8 (extracted with hydroxylamine hydrochloride)

Arsenic concentration (source material) (mg/kg) As-rich pyrite, 0.5%–0.7% Pyritic sediments, 300 Oxidation of pyritic sediments exposed as a result of reduced rainfall, increased groundwater abstraction, and dewatering carried out during urban construction Microbially mediated reductive dissolution of Fe oxyhydroxides, mobilization with organic ligands from the sediment organic matter Desorption from Fe, Mn oxides and organic matter under oxidizing and alkaline conditions; oxidation and dissolution of pyrite

Mobilization mechanism

Versanyi and Kovacs 2006

Appleyard et al. 2006

Reference

Organic-rich sediments of Below detection Gomez et al. 2006 the Middle Miocene limit: 337 clayey Zaratan facies (mean, 23) (As present in Fe and Mn oxides, authigenic pyrite,Ti-Fe oxides, phyllosilicates and organomineral compounds) N, number of measurements; where several concentration ranges were reported for an area, the appropriate reference is provided for each reported range. NA, not available; ND, not discussed.

Up to 7000

Western Australia

Table 3 (continued) Location Arsenic concentrations range/mean (mg/L)

36 H. Garelick et al.

Arsenic Pollution Sources

37

Table 4. Such systems occur throughout the world in one of three general tectonic settings (Webster and Nordstrom 2003): l

l l

Near tectonic plate boundaries [e.g., edge of the Pacific Plate: geothermal fields in New Zealand, Papua (New Guinea), Philippines, Indonesia, Japan, Kamchatka, Alaska, Western USA, Mexico, Central America, and Chile]. Intraplate hot spots: Hawaii, Yellowstone (USA), and the Azores. Rift zones, where the tectonic plates diverge: Gregory Rift Valley in Ethiopia, Kenya, and Tanzania, the Rio Grande Rift Valley in Colorado and New Mexico (USA), and the rift system in Iceland.

Many geothermal systems have been, or will be, developed to generate energy from steam and hot water reservoirs beneath the earth’s surface. Therefore, the potential exists for further mobilization and exposure to arsenic from such development. Arsenic is an important trace constituent in geothermal fluids, where concentrations range from less than 0.1 to nearly 50 mg/kg (Ballantyne and Moore 1988). These wide variations can be attributed to local water physico-chemistry (in particular, low- or high-sulfide fluids), salinity, temperature, and host rock composition. Laboratory experiments have demonstrated that arsenic in geothermal fluids may originate from rock leaching (Criaud and Fouillac 1989). It has also been postulated that some of the arsenic may be of magmatic origin (Ballantyne and Moore 1988). Arsenic occurs predominantly in pyrite, or is associated with Fe oxides in geothermally altered rocks; up to 3.8 wt.% arsenic was reported in pyrite originating from hot springs (Ballantyne and Moore 1988). Arsenic minerals are uncommon in geothermal reservoir rocks. However, a variety of these occur in hot spring deposits including orpiment, realgar, As-rich marcasite, and As-rich stibnite (Ballantyne and Moore 1988). Arsenic (III) predominates in geothermal water directly discharged from an underground reservoir, although it is gradually oxidized to As (V) on the surface. In acid sulfate- and bicarbonate-type springs, As (V) is a predominant species (Ballantyne and Moore 1988). Arsenic (III) can be oxidized to arsenic (V) when a rising geothermal fluid is exposed to atmospheric oxygen or is mixed with another oxidizing fluid such as shallow groundwater. In geothermal waters containing sulfide or tiosulfate, arsenic (III) is preserved until the reduced sulfur is oxidized or volatilized (Webster and Nordstrom 2003). The source of arsenic in geothermal fluids, arsenic speciation, and the fate of arsenic from geothermal fluids are discussed further by Ballantyne and Moore (1988) and Webster and Nordstrom (2003). Geothermal fluids commonly contain elevated arsenic concentrations. A list of these, grouped by geographic location, is presented in Table 4. Part of the arsenic in this matrix is volatilized and released into the atmosphere. Although arsenic speciation in geothermal fluids has been extensively studied, there is a paucity of information on volatile arsenic in geothermal fields. However, Planer-Friedrich et al. (2006) have attempted the quangification of total volatile arsenic and have described its speciation in one of the world’s best known geothermal fields, namely Yellowstone National Park (USA). They measured concentrations of 0.5–200 mg/L at the surface of the hot springs and identified chloro- and thioarsine species, which have not been

38

H. Garelick et al.

Table 4 Arsenic concentration ranges in hot springs of various regions. Country Geothermal field Arsenic concentration (mg/L) United States Yellowstone National 160–3600 (hot springs) (USA) Park Lassen Peak, California 2200–24300 (thermal waters)

Reference Stauffer and Thomson 1984

Geyser Bight, Umnak Island, Alaska

up to 3800

White et al. 1963

Valles Caldera, New Mexico

21–2400 (spring water)

Criaud and Fouillac 1989

Honey Lake Basin, California Coso Hot Springs, California Imperial Valley, California Long Valley, California Steamboat Springs, Nevada

up to 2600

Welch et al. 1988

Taupo Volcanic Zone, Waiotapu and Rotokawa geothermal fields

0.71–6.5 mg/kg 0.64–1.34 mg/kg (hot spring waters)

Webster 1990

Kawerau geothermal field

20–30 (river Tarawera receiving input of geothermal fluids)

Mroczek 2005

Wairakei, Broadlands, Orakei Korako and Atiamuri geothermal fields

3800 (groundwater from Waikarei) up to 121 (river and lake waters with inputs from geothermal fields)

Robinson et al. 1995

Bulgaria

Southwestern Bulgaria

3–9.8

France

Chassolle group, Cezallier

240–509

Criaud and Fouillac 1989

Dominica

Wooten Waven Valley of Desolation

0.3–775 <0.3–650

Mexico

Los Azufres, Mexico

5–30,000 (drainage water from different evaporation basins) 8-8200 (springs and surface runoffs in and outside geothermal field)

New Zealand

up to 7500 up to 15,000 up to 2500 up to 2700

Birkle and Merkel 2000

Arsenic Pollution Sources

39

19,000 (geothermal brine) 160 mg/kg (max in pond sediments) 24,000 (pond water)

Birkle and Merkel 2002

Philippines

Mt Apo Philippines

31–6200 (hot spring water)

Webster 1999

Japan

Kyushu geothermal fields

700–4600 (geothermal water)

Yokoyama et al. 1993

Russia

Kamchatka

2.0–3.6 mg/kg

Karpov and Naboko 1990

100–5900 (thermal waters)

White et al. 1963

Ischia island

0.5–1558 (shallow thermal wells and thermal spring waters)

Lima et al. 2003

Phlegrean Fields, Volcano island, Mt Etna, Mt Vesuvius

0.1–6940 (groundwaters)

Aiuppa et al. 2003

Nafmafjall, Krafa hightemperature fields

0.001–0.048 mg/kg

Arnorrson and Linvall 2001

50–120 (thermal waters)

White et al. 1963

Italy

Iceland

Chile

El Tatio system (Antofagasta region)

45,000–50,000

Ellis and Mahon 1977

Germany

Wiesbaden spa (thermal springs)

37–110 (public thermal water taps)

Schwenzer et al. 2001

Turkey

Geothermal fields in Turkey: Alibeykoy (Canakkale), Kızılkoy (Balıkesir), Kukurtlu (Bursa), Cumali-Seferihisar (Izmir), DoganbeySeferihisar (Izmir), and Balcova (Izmir)

130–280 (geothermal water)

Baba and Armannsson 2006

previously documented in the natural environment. The toxicity, mobility, and degradation products of these arsenic species are not known, nor are the processes that lead to their formation, although microbial transformation is implicated.

40

H. Garelick et al.

Volcanic degassing also represents an important natural source of arsenic release to shallow aqueous systems. Groundwaters circulating in active volcanic areas may contain appreciable amounts of arsenic as a result of their interaction with deeprising fluids or leaching of ore deposits (Welch et al. 1988; Yokoyama et al. 1993; Aiuppa et al. 2003). Volcanic aquifers often constitute an important water resource for people living near them. For example, the Etnean aquifer represents the main water supply for drinking and agricultural purposes for about 1 million people living in eastern Sicily, Italy (Aiuppa et al. 2003). The potential therefore exists for exposure of these communities to elevated arsenic concentrations under conditions favorable to its release to potable water supplies. In general, surface waters are more susceptible to arsenic contamination than corresponding groundwaters in areas of geothermal systems. Mixing of geothermal waters with groundwater and surface waters is known to increase the arsenic concentrations in the latter (Webster 1999; Robinson et al. 1995). Some soil contamination also occurs in such areas, but tends to be local; in contrast, water flowing through the surface and subsurface catchments of the geothermally active areas may transport arsenic beyond the boundary of the geothermal field. Geothermal waters also affect the arsenic content of plants in the vicinity of hot springs (Loppi and Bonini 2000). Arsenic levels and geothermal activity are so closely related that arsenic anomalies in soils and waters have been used as geochemical indicators for geothermal areas (Bingqiu et al. 1986; Shiikawa 1983). Elevated arsenic concentrations in surface waters, which may be locally used for drinking and irrigation, and accumulation of arsenic in edible aquatic plants in river systems receiving geothermal fluids, constitute the main exposure pathways for local populations in geothermally active areas. The solubility of arsenic entering surface waters from geothermal fluids is affected by the extent of arsenic adsorption onto particle surfaces. Under the higher rock: water ratios present in aquifers, arsenic concentrations may be reduced through adsorption onto clays and other silicate surfaces, particularly in the presence of Fe-Mn oxides. Appropriate pH conditions, as determined by the pH of the zero point of charge (ZPC) of the available sorbing surfaces (Manning and Goldberg 1997), facilitate this process. In addition, the solubility of geothermally derived arsenic may be limited by the precipitation of orpiment if high sulfide concentrations accompany arsenic in geothermal fluids (Ballantyne and Moore 1988). Little can be done to control arsenic inputs from geothermal springs; indeed, future geothermal power developments may increase amounts of arsenic entering the aquatic biosphere. Hazards may increase as geothermal fields are more often tapped for power generation because of the corresponding increases in rates and volumes of geothermal fluids reaching the surface. Disposal of wastewaters from operations of such plants will exacerbate the problem because of arsenic enrichment of the wastewater. In developed geothermal fields, wastewater is reinjected into the field to avoid environmental pollution. Reinjection is expensive and reduces productivity of the field. Therefore, where not required by law, wastewaters are discharged into the local environment, resulting in contamination of soils and waterways. Examples ´ rmannsson (2006). of such problems are provided in Mroczek (2005) and Baba and A

Arsenic Pollution Sources

IV

41

Anthropogenic Sources

Anthropogenic arsenic sources include the following: l

l l

l l

l l

A

Metal acquisition and processing: Ore production and processing, ore melting and roasting in nonferrous smelters, metal melting in iron works; metal treatment admixture in bronze, lead, and copper alloy production; galvanizing processes; ammunition factories; battery plate production. Energy: High-temperature oil and coal burning; and operation of power plants. Agriculture and livestock: Use of arsenical insecticides, algaecides, and defoliants; debarking trees, soil sterilants, feed additives, cattle and ship dips, intense husbandry disinfectants; compost and dung surplus (arsenic from animal feeding). Wood preservatives. Medicine: Pharmaceutical substances: antisyphilitic drugs; treatment of trypanosomiasis, amebiasis, and sleeping sickness. Waste material: Waste incineration; household waste disposal. Other industrial uses/sources: Glassware and ceramics production; use in decoloring agents; electronics industries; solar cells, optoelectronic devices, semiconductor applications, light-emitting diodes (digital watches); dyes and colors; pyrotechnics; drying agents for cotton; oil and solvent recycling; cement works; angifouling agents; catalysts.

Modern and Historical Uses of Arsenic

Although there is no agreement as to the first intentional use of arsenic in antiquity, arsenic poisoning was among the first occupational diseases to be recognized by mankind (Azcue and Nriagu 1994). The first documented historical uses of arsenic were pharmaceutical and medicinal. Arsenic was also commonly used in pigments, poisons, and in the manufacturing of glass. A major modern use for arsenic was as pesticides in agriculture: monosodium methylarsenate (MSMA), disodium methylarsenate (DSMA), dimethylarsinic acid (cacodylic acid), and arsenic acid (Azcue and Nriagu 1994). Arsenic acid is also used in the manufacture of wood preservative salts. Arsenillic acid is used as an additive in poultry, swine, and veterinary feeds. Until the 1970s, approximately 80% of the arsenic produced was used to manufacture simple inorganic salts for use as pesticides. Today, pesticides account for approximately 50% of the arsenic consumed, with organic arsenic compounds now dominating pesticide production. Another 30% of the world arsenic market is accounted for by wood preservatives. The remaining arsenic is used to manufacture glass, alloys, electronics, catalysts, fodder additives, and veterinary chemicals (Matschullat 2000). Currently, the main arsenic producers are located in China, Chile, Peru, Mexico, Russia, Kazakhstan, and France. These countries account for more than 90% of world arsenic production (USGS 2006).

42

B

H. Garelick et al.

Mining as a Source of Arsenic in the Environment

Mining, smelting, and ore beneficiation processes are major sources of arsenic contamination worldwide. Arsenic is a natural component of lead-, zinc-, copper-, and gold-bearing ores and consequently may contaminate soils, sediments, and groundwater during mining and smelting operations. The most common source of arsenic in such environments is from weathering of arsenic sulfides such as arsenopyrite and other arsenic-bearing metal sulfides exposed to atmospheric conditions (for example, by placement on a heap or waste rock dump). Arsenic can be leached from sulfide ores by mine water and may be further concentrated during the metal sulfide beneficiation process, in which an arsenichosting ore is chemically treated before smelting. The concentration of arsenic liberated by sulfide oxidation and smelting is dependent on several factors: l

l l

The arsenic concentration available in the ore or deposit (related to geology of the original hydrothermal deposit) The Eh–pH regime of the environment from which the arsenic was liberated The chemistry of natural waters that promote release or attenuation of arsenic

Arsenic is a common constituent of many sulfide minerals, and thus no specific arsenic-enriched mineral such as arsenopyrite need be present. However, all sulfide mining and smelting sites are potential locations of future arsenic contamination (Madhavan and Subramanian 2000). The contamination of soils, waters, and sediments by arsenic originating from mining and smelting activities has been widely reported in the literature, and several examples are discussed next. Because arsenic is associated with gold deposits, gold mining can contribute to arsenic contamination. In fact, gold mining activities have been recognized as the main source of arsenic pollution in many locales. Gold is an important commodity that has been mined extensively within Australia, particularly in the state of Victoria, where elevated arsenic concentrations are associated with gold mining residues (Hinwood et al.1998; Lamb et al. 1996). Concentrations of arsenic reported from gold mining areas varied among regions. In addition, the concentrations at specific sites were found to be dependent on the type of ore body being mined, the type of ore processing used, and the means of waste disposal (Ellice et al. 2001). High concentrations of arsenic have been reported in mullock heaps and mining waste disposal areas in Victoria, where concentrations ranged from 280 to 15,000 mg/kg in areas of former mine waste disposal (Hinwood et al. 1998; Ellice et al. 2001). Arsenic contamination related to gold production has also been reported in Brazil (Borba et al. 2003). These authors estimated that 390,000 t arsenic was discharged to the environment during three centuries of gold mining in the Iron Quadrangle mining district. Arsenic concentrations up to 4,800 mg/kg were found in gold deposits in this area. The highest arsenic concentrations in water and stream sediments were reported in the vicinity of mining areas. Up to 300 mg/L was found in surface water, whereas the arsenic content in stream sediments was in the range of 20–4,000 mg/kg. Although no arsenic poisoning cases were reported from the area, there are arsenic hazards from

Arsenic Pollution Sources

43

dispersion of old tailings by flooding, construction of settlements on arsenic-contaminated soils, and occasional consumption of contaminated groundwater. Contamination arising from gold mining activities in the Takab area of Iran resulted in severe pollution of a local river system (Modabberi and Moore 2004). Arsenic-gold is the main ore deposit in the area, and its mining history dates back hundreds of years. Sediments from a stream (Zashuran) contained arsenic residues ranging from 125 to 125,000 mg/kg; levels in sediments from a river (Sarouq) ranged from 27 to 148 mg/kg. The range of arsenic concentrations detected in waters associated with these sediments was 28–40,000 mg/L in the Zashuran stream and 25–280 mg/L in the Sarouq River. Arsenic enrichment in sediments from the Elqui watershed river system in Northern Chile was reported to result from gold mining activities (in the El-Timbro District) as well as geological processes (Oyarzun et al. 2004). Arsenic levels were very high in both stream (55–485 mg/kg) and the Early Holocene sequence lacustrine sediments (119–2344 mg/kg), suggesting that contamination, in this region of Chile, derives both from industrial (mining operations) and major geological processes (long-term erosion from arsenic-rich epithermal ores and alteration zones). Historical mining activity has led to significant anthropogenic contamination of soils and estuaries in Cornwall and Devon, SW England. Arsenic-rich polymetallic ores were processed in these areas by heat treatment to remove arsenic, leaving residues of metalloids and metals. At some sites, arsenic ore (arsenopyrite) was processed (smelted and refined) and As2O3 was produced (Barton 1971). Very high arsenic concentrations were reported from several mining sites in the area: Poldice, 432–37,600 mg/kg (Thornton and Farago 1997), Roseworthy, 815–67,300 mg/kg with extreme values of 10% and 16% by mass (Frizzeli 1993), and Devon Great Consols Mine, 173–52,600 mg/kg (Kavanagh 1998). Poisoning from arsenic release to the environment from mining activities is well documented in Thailand. Approximately 1,000 people in Southern Thailand have been diagnosed with arsenic-related skin disorders. People living in and near Ron Phibun town were particularly affected (Williams 1997; Choprapawon and Rodcline 1997). The affected area lies within the South-East Asian Tin Belt. Arsenic was found at concentrations up to 5,000 mg/L in shallow groundwaters from Quaternary alluvial sediment that had been extensively dredged during tin mining operations. Deeper groundwaters from older limestone aquifers were less contaminated (Williams et al. 1996), although high arsenic concentrations were found in a few cases, presumably from mining activities. The mobilization of arsenic is believed to be caused by oxidation of arsenopyrite and is exacerbated by former tin mining activities. Ghana, an important gold-mining country, has been active in mining since the late 19th century. The most important mining area in Ghana is the Ashanti Region. The gold found there is associated with sulfide minerals, particularly arsenopyrite. Arsenic is mobilized from arsenopyrite by oxidation, which is either induced or exacerbated by mining activity. High regional arsenic concentrations have been found in soils near the mines and the associated treatment works (Bowell 1992, 1993). In some cases, high concentrations of arsenic were found in river waters from nearby mining

44

H. Garelick et al.

activity (Smedley et al. 1996). However, for mechanistic and physico-chemical reasons peculiar to a region, high arsenic levels in soils and bedrocks near mines do not necessarily contaminate local groundwater. For example, Smedley et al. (1996) found that many groundwaters of the Ghanaian mining areas had low arsenic concentrations, with a median concentration in tube well waters of just 2 mg/L. Arsenic contamination from mining activities has been identified in numerous areas of the US, many of which have been summarized by Welch et al. (1988, 1999). Groundwater from some areas has been reported to have very high arsenic concentrations locally (up to 48,000 mg/L). Well-documented cases of arsenic contamination include the Fairbanks gold mining district of Alaska (Wilson and Hawkins 1978; Welch et al. 1988), the Coeur d’Alene Pb-Zn-Ag mining area of Idaho (Mok and Wai 1990), Leviathan Mine in California (Webster et al. 1994), Kelly Creek Valley, Nevada (Grimes et al. 1995), Clark Fork River, Montana (Welch et al. 2000) and Lake Oahe in South Dakota (Ficklin and Callender 1989). Many other areas of the world have high concentrations of arsenic in soils, sediments, and waters as a result of mining activity. Documented cases include the Lavrion region of Greece, associated with lead and silver mining activity (Komnitsas et al.1995) and the Zimapa´n Valley of Mexico, South Africa, Zimbabwe and Bowen Island, British Columbia (Boyle et al. 1998). Madhavan and Subramanian (2000) investigated arsenic enrichment in the mining and smelting areas of the Khetri (copper) and Zawar (lead-zinc) deposits in Rajasthan, India. The concentration of arsenic in Zawar zinc and Khetri copper tailings was found to be 1,519 and 1,179 mg/kg, respectively. Tailings at a local metal sulfide smelter contained 10,727 mg/kg (zinc), 61 mg/kg (lead), and 12 mg/kg (copper). The mean value of arsenic in the mine water inlet of the Zawar region was 3 mg/L; in the mine water outlet, 130 mg/L was reported. The mean value found for arsenic in groundwater samples from this region (5.8 mg/L) was below the current MAC (maximum allowable concentration) for arsenic. Generally, arsenic concentrations found in inlet and outlet mine waters and in groundwaters from the Khetri region were low (all values below 10 mg/L). Although severe contamination of the environment has been documented in many mining districts, it is surprising that, with the exception of Thailand, little or no impact on the health of the local populations was reported.

C

Arsenic Emissions from Coal Burning

Elements such as arsenic, selenium, fluorine, and mercury are commonly enriched in coal deposits (Finkelman 2004). Coal, when combusted, volatilizes these elements and makes them available for inhalation, contamination of food and crops, uptake by livestock, and bioaccumulation in birds and fish. Volatile elements such as arsenic can also condense after coming in contact with ambient air to form minerals (arsenolite, orpiment, realgar), which then fall to the ground. These mineral contaminants may then be leached by rainwater and transported into

Arsenic Pollution Sources

45

waterways. During coal combustion at power plants, most arsenic in coal volatilizes, with only a minor portion remaining in bottom ash. Most of the escaping arsenic is captured by fly ash. Because 97%–99% of the fly ash is collected by electrostatic precipitators, the atmospheric emission of arsenic (solid phase and gaseous) is usually assumed to be minor (10%–30% of initial arsenic levels in coal). However, arsenic can be leached from waste fly ash to create environmental problems. In the natural environment, arsenic is readily leached from acid (SiO2-rich) bituminous coal ashes, but little is leached from alkali (CaO-rich) lignite ashes (Yudovic and Ketris 2005). Estimates put the proportion of arsenic derived from fuel combustion to total anthropogenic arsenic emissions at 2–6% (Nriagu and Pacyna 1988), or up to 17% (Kizilstein 1997, as cited in Yudovic and Ketris 2005). The amount of arsenic released into the atmosphere depends on the particular combustion regime and pollution control systems employed. Yudovic and Ketris (2005) estimate that 2%–47% of the arsenic content in coal is released into the atmosphere from power plants. The extent of arsenic volatility from coal is affected by modes of arsenic occurrence in the coal (i.e., different forms of As such as those associated with sulfide minerals or present as sorbed arsenate ions), with the dominant arsenic forms being pyritic, that fraction of arsenic associated with the organic fraction of coal and inorganic arsenate (Yudovic and Ketris 2005). The presence of chlorine in coal can favor formation of gaseous arsenic compounds; the presence of illite or CaCO3 favors immobilization of arsenic by the formation of refractory arsenates such as KAsO4 and Ca3(AsO4)2, with high melting temperatures, above 1,000 C (Yudovic and Ketris 2005). The average arsenic content for bituminous coals and lignites in the world is 9.0 (0.8) mg/kg and 7.4 (1.4) mg/kg, respectively; on an ash basis this equates to 50 (5) and 49 (8) mg/kg, respectively (Yudovic and Ketris 2005). Arsenic may be found in individual coal beds at levels as high as 3,845 mg/kg (Miocene bituminous feed coal from Western Turkey) (Karayigit et al. 2000) and 35,000 mg/kg (Guizhou Province in China) (Ding et al. 2001). Yudovic and Ketris (2005) suggest that a relationship exists between the arsenic content in coal and its mode of occurrence in coals. Typically, in coals with high arsenic content, arsenic-enriched sulfides dominate (pyrite and other rare sulfides), whereas, in those with low arsenic content, arsenic is associated with organic matter. There are four generic types of arsenic accumulation in coal (Yudovic and Ketris 2005): l l

l

l

Chinese type: hydrothermal arsenic enrichment Dakota type: hypergene enrichment from groundwaters draining arsenic-bearing tufa host rocks Bulgarian type: arsenic enrichment resulting from arsenic-bearing waters entering coal-forming peat bogs from sulfide deposit aureoles Turkish type: volcanic input of arsenic in coal-forming peat bog as exhalations, brines, and volcanic ash.

An increased exposure to arsenic from combustion of arsenic-rich coal has been reported in Slovakia (Keegan et al. 2002; Pesch et al. 2002) and the Guizhou

46

H. Garelick et al.

Province of China (An et al. 1997; Liu et al. 2002; Ng et al. 2003; Finkelman et al. 1999; Dai et al. 2005). In the Guizhou Province of China, an estimated 3,000 people were found to be suffering from severe arsenic poisoning and approximately 200,000 people were believed to be at risk of arsenic overexposure (Liu et al. 2002). The primary source of arsenic exposure in the Guizhou Province was consumption of chili peppers and other food dried over fires fueled with higharsenic coal. Coal, when burned inside the home in open pits for daily cooking and crop drying, produces a high concentration of arsenic in indoor air. High arsenic residues from the air then permeate food during its preparation. Liu et al. (2002) reported arsenic concentrations from the affected region of China to be 70 mg/kg in smoked chili peppers (compared with 0.04 mg/kg in uncontaminated peppers) and 3.40 mg/kg in smoked corn.

D

Agricultural Use of Arsenic Compounds

Arsenic has been widely used in insecticides and other pesticides. Arsenical pesticides were used extensively in the orchards of many countries from the late 1800s until the introduction of DDT in the early 1960s. The extensive past use of the horticultural pesticides lead arsenate (PbAsO4), calcium arsenate (CaAsO4), magnesium arsenate (MgAsO4), zinc arsenate (ZnAsO4), zinc arsenite [Zn (AsO2)2], and Paris green [Cu(CH3COO)23Cu(AsO2)2] in orchards has contributed to soil arsenic contamination (Folkes et al. 2001; Embrick et al. 2005; Brouwere et al. 2004; Wang and Mulligan 2006). Such pesticide use led to extensive contamination of surface and subsurface soils in most Australian states (Smith et al. 2003). Although a majority of this contamination is of a diffuse nature, there are incidences of point-source contamination, largely at existing or former pesticide storage sites. Arsenic concentrations up to 30 times background levels and ranging from <0.5 to 115 mg/kg were reported at contaminated sites in Australia. Similarly, there are reports that arsenic leached through soil profiles at some Australian orchard sites (Smith et al. 2003). Other inorganic arsenicals, primarily sodium arsenite, were widely used in agriculture, since about 1890, as weed killers or nonselective soil sterilants (Reigart and Roberts 1999). Arsenic acid was also used extensively as a cotton desiccant. Moreover, arsenic compounds such as H3AsO4, 3-nitro-4-hydroxy phenylarsonic acid, and 4-nitrophenylarsonic acid were used as animal feed additives (US Department of Agriculture 1970; Thompson 1973). The use of sodium arsenite (NaAsO2) to control aquatic weeds has contributed to the arsenic burden in New Zealand lakes (Tanner and Clayton 1990). Instances of soil arsenic contamination have also been attributed to the extensive use of arsenical insecticides in sheep and cattle dips to control ticks, fleas, and lice (McBride et al. 1998; McLaren et al. 1998). Arsenic residues in surface soil (0–10 cm) of historical sheep dip sites range between 37 and 3542 mg/kg. Arsenic at such sites can penetrate deeply into soil; values up to 2282 mg/kg have been reported at the 20–40 cm depth (McLaren et al. 1998). Smith et al. (2003) report the presence

Arsenic Pollution Sources

47

in Australia of more than 1,000 arsenic-contaminated sites previously used as cattle dips. Arsenical pesticides, principally lead, calcium, and sodium arsenate, were widely applied on apple, blueberry, and potato crops in New England during the first half of the 20th century. Extensive use of these compounds in an area correlate positively with levels of arsenic contamination in adjoining stream sediments; arsenic residue values ranging from 0.3 to 93 mg/kg were reported in 1,600 screened stream sediment samples (Robinson and Ayuso 2004). Murphy and Aucott (1998) estimated amounts of arsenical pesticides historically applied to cropland, turf, and golf courses in New Jersey (USA) to identify areas with high arsenical pesticide residues that would require soil monitoring and possible remediation. Arsenical pesticides were extensively used in the US beginning with the second half of the 19th century. Paris green, a copper and arsenic compound, was first used in the US in 1865 for control of the Colorado potato beetle; by 1868 its use had been well established. Another arsenical insecticide, calcium arsenate, was extensively used for boll weevil control in US cotton production between 1920 and 1950. Calcium arsenate was also used as a broad-spectrum insecticide on a variety of vegetable crops and on white potatoes throughout the US (Murphy and Aucott 1998). Lead arsenates were also used, primarily in sprays and dusts, for protection against insects attacking fruits and vegetables. Lead arsenate was recommended for the routine spraying of apples throughout the growing season and its use continued until 1964, when it was superseded by synthetic organic insecticides (primarily DDT). Furthermore, Murphy and Aucott (1998) estimated that between 1900 and 1980, 25 million kg lead arsenate and 9 million kg calcium arsenate were applied to soils in the state of New Jersey. The consequence is that residential developments built in the area on former apple orchards are contaminated with arsenic and lead. Beginning in the late 1980s, the US EPA (US Environmental Protection Agency) banned the use of many inorganic arsenic-based pesticides (US EPA 1992a). However, organoarsenical compounds such as monomethylarsenic acid (MMA) and dimethylarsenic acid (DMA) are allowed to be used, because they are generally considered to be noncarcinogenic. However, the lack of sufficient data in the literature to define the stability of these compounds in the soil environment is of concern (Datta et al. 2006); due to the potential for future release of arsenic from these compounds following their breakdown or transformations in soil.

E

Wood Preserving Industry

Arsenic has been commonly used in the wood preserving industry as a component of preservatives (US EPA 2006). Timbers not naturally durable are treated with preservatives to prevent decay by wood-boring Crustacea, mollusks, and fungi. The most widely used wood preservative for timbers exposed to aquatic or moist environments was CCA. The metal elements in CCA are present in the form of

48

H. Garelick et al.

oxides including CuO, Cr2O3 and As2O5. Three CCA formulations were developed (Hingston et al.2001): l l l

Type A: CuO (18%), Cr2O3 (65%), and As2O5 (16%) Type B: CuO (20%), Cr2O3 (35%), and As2O5 (46%) Type C: CuO (19%), Cr2O3 (48%), and As2O5 (34%)

CCA-treated wood has been extensively used for more than 60 yr in Europe and North America (Hingston et al.2001). Since the late 1980s, US production of CCAtreated lumber has averaged approximately 5  108 ft3 yr–1 (Solo Gabriele and Townsend 1999). Because of the toxicity of arsenic and chromium used in CCA treatment, regulatory and public attention has recently focused on the potential risks from this exposure source, in particular, risks to children exposed to CCA-treated wood decks and play sets (Kwon et al. 2004; Ursitti et al. 2004; Shalat et al. 2006; Zartarian et al. 2006). Many sites used by the timber industry to treat wood, or to dispose of waste, or sites where treated timber was used are contaminated with arsenic (Rice et al. 2002; Hopp et al. 2006). In addition, timber treatment effluent has been identified as an important source of arsenic contamination in aquatic and terrestrial environments (Breslin and Adler-Ivenbrook 1998; Bolan and Thiyagarajan 2001; Hingston et al. 2001; Rice et al. 2002). The well-documented human and environmental toxicity of arsenic-based wood preservatives led the European Commission to restrict use of these products (European Commission 2003). Since 2004, CCA-treated wood cannot be used for residential purposes or in marine water installations, and professional and industrial uses are only permitted under strict conditions (Hopp et al. 2006). The US EPA also banned the use of CCA-treated wood for residential purposes after 2003 (USEPA 2006). Despite the reduction in CCA-treated wood production, the disposal of waste CCA-treated wood in landfill raises concerns because of the potential for leaching of arsenic. Townsend et al. (2005) investigated arsenic leaching from weathered CCA-treated wood collected from demolition sites and disposal facilities. The authors found the average leachable arsenic concentration (using the standard Toxicity Characteristic Leaching Procedure, or TCLP) (US EPA 1992b) to be 3.2–13 mg/L. In 60 of 100 samples tested, the TCLP arsenic concentrations exceeded the regulatory value of 5 mg/L, thus confirming the need to dispose of weathered CCA-treated wood as a hazardous waste in the US Mandal and Suzuki (2002) and Mukherjee et al. (2006) reviewed the literature and summarized data (Table 5) on incidents of arsenic contamination from various industrial and mining sources around the world.

Summary Arsenic is a widely dispersed element in the Earth’s crust and exists at an average concentration of approximately 5 mg/kg. There are many possible routes of human exposure to arsenic from both natural and anthropogenic sources.

Arsenic Pollution Sources Table 5 Arsenic pollution from anthropogenic sources worldwide Country Arsenic source Arsenic concentrations Mindanao Island, Geothermal power plant Matingao and Marebl Philippines suspected source rivers contaminated, 0.1 mg/L Nakajo, Japan Waste water from a 0.025–4.00 mg/L in factory producing local well water arsenic sulfide Toroku and Matsuo Arsenious acid and Not available villages, Japan white arsenic produced from FeAsS Behalla area, South Industrial effluent 0.05–58 mg/L in well Calcutta, India discharge from a water factory manufacturing copper acetoarsenite (Paris-Green) Czech Republic Burning of arsenic Not available contaminated coal Toronto, Ontario Secondary lead smelters Vegetation and soils Canada contaminated Guizhou Province, Burning of arsenic rich Not available China coal Srednogorie, Copper smelter Air pollution and soil Bulgaria processing high contamination, arsenic sulfide ores local river: 0.75–1.5 mg/L Anaconda, Montana Copper smelters As2O3 emitted to the air, contamination of local vegetation and soils, arsenic in soils: 212–236 mg/ Kg Ghana Gold mining, arsenic 0.90–8.25 mg/L arsenic rich mine tailings in water samples; 942–10,200 mg/kg arsenic in sediments Lavrion, Greece Lead mining and Soils contamination. smelting arsenic in garden soils and house dusts: up to 14,800 and 3800 mg/kg respectively Mae-Moh Valley, Lignite mining and Arsenic-rich sediments Northern power generation in streams, 1.2–325 Thailand g/ L arsenic in surface waters, groundwater from deep tube wells in Mae Moh mine area – average 364 g/ L As.

49

Reference Webster 1999

Terado et al. 1960a Hotta 1989a, Tsuda et al. 1990a Guha Mujumder et al. 1988

Bencko and Symon 1977a Temple et al. 1977a Dai et al. 2005 Nilsson et al. 1993

Hwang et al. 1997

Bennerman et al. 2003, Serfor-Armah et al. 2006 Stavrakis et al. 1994a

Bashkin and Wongyai 2002

(continued)

50 Table 5 (continued) Country Nitra Valley in central Slovakia

Australia

Japan

SW England

Patancheru, AndhaPradesh state, India

a

H. Garelick et al.

Arsenic source Coal burning power station

Manufacture of: fertilizers, paint, electronics; gas works; mining (mainly gold deposits); power stations; transport; and wood preservatives Wastewater from a factory producing arsenic sulfide, production of arsenious acid, manufacture of insecticides Mining and smelting activities

Production of veterinary chemicals and pharmaceuticals, pesticide industries

Arsenic concentrations 8.8–139.0 mg/kg arsenic in garden soils and 2.1–170 mg/kg in house dusts Soil contamination

Reference Keegan et al. 2002

Not available

Mukherjee et al. 2006

Over 700 km2 of agricultural and urban land estimated as contaminated with arsenic. Soil contamination with concentrations up to 10% by mass reported at some mining sites Elevated arsenic concentrations in stream waters around former mining sites (5–30 g/L total As) Arsenic in groundwater: 0.14–7.35 mg/L; arsenic in surface water: 0.30–8.95 mg/L

Abrahams and Thornton 1987

Smith et al. 2003

Farago et al. 1997, Kavanagh 1998 Rawlins et al. 2003

Chandra Sekhar et al. 2003

Cited in Mandal Suzuki (2002).

Arsenic occurs as a constituent in more than 200 minerals, although it primarily exists as arsenopyrite and as a constituent in several other sulfide minerals. The introduction of arsenic into drinking water can occur as a result of its natural

Arsenic Pollution Sources

51

geological presence in local bedrock. Arsenic-containing bedrock formations of this sort are known in Bangladesh, West Bengal (India), and regions of China, and many cases of endemic contamination by arsenic with serious consequences to human health are known from these areas. Significant natural contamination of surface waters and soil can arise when arsenic-rich geothermal fluids come into contact with surface waters. When humans are implicated in causing or exacerbating arsenic pollution, the cause can almost always be traced to mining or miningrelated activities. Arsenic exists in many oxidation states, with arsenic (III) and (V) being the most common forms. Similar to many metalloids, the prevalence of particular species of arsenic depends greatly on the pH and redox conditions of the matrix in which it exists. Speciation is also important in determining the toxicity of arsenic. Arsenic minerals exist in the environment principally as sulfides, oxides, and phosphates. In igneous rocks, only those of volcanic origin are implicated in high aqueous arsenic concentrations. Sedimentary rocks tend not to bear high arsenic loads, and common matrices such as sands and sandstones contain lower concentrations owing to the dominance of quartz and feldspars. Groundwater contamination by arsenic arises from sources of arsenopyrite, base metal sulfides, realgar and orpiment, arsenic-rich pyrite, and iron oxyhydroxide. Mechanisms by which arsenic is released from minerals are varied and are accounted for by many (bio)geochemical processes: oxidation of arsenic-bearing sulfides, desorption from oxides and hydroxides, reductive dissolution, evaporative concentration, leaching from sulfides by carbonate, and microbial mobilization. Arsenic enrichment also takes place in geothermally active areas; surface waters are more susceptible than groundwater to contamination in the vicinity of such geothermal systems, and evidence suggests that increased use of geothermal power may elevate risks of arsenic exposure in affected areas. Past and current mining activities continue to provide sources of environmental contamination by arsenic. Because gold- and arsenic-bearing minerals coexist, there is a hazard of mobilizing arsenic during gold mining activities. The Ashanti region of central Ghana currently faces this as a real risk. Historical arsenic contamination exists in Cornwall, UK; an example of a recent arsenic pollution event is that of Ron Phibun town in southern Thailand, where arsenic-related human health effects have been reported. Other important sources of arsenic exposure include coal burning in Slovakia, Turkey, and the Guizhou Province of China; use of arsenic as pesticides in Australia, New Zealand, and the US; and consumption of contaminated foodstuffs (China) and exposure to wood preserving arsenicals (Europe and North America).

References Abrahams PW, Thornton (1987) Distribution and extent of land contaminated by arsenic and associated metals in mining regions of Southwest England. Trans Inst Miner Metall Section B Appl Earth Sci 96:B1–B8.

52

H. Garelick et al.

Ahmad K (2001) Report highlights widespread arsenic contamination in Bangladesh. Lancet 358:133. Aiuppa A, D’Alessandro W, Federico C, Palumbo B, Valenza M (2003) The aquatic geochemistry of arsenic in volcanic groundwaters from southern Italy. Appl Geochem 18:1283–1296. An D, He YG, Hu QX (1997) Poisoning by coal smoke containing arsenic and fluoride. Fluoride 30:29–2. Anthony JW, Bideaux RA, Bladh KW, Nichols MC (1990) Handbook of Mineralogy, vol I. Elements, Sulfides, Sulfosalts. Mineral Data Publishing, Tucson, AZ. Anthony JW, Bideaux RA, Bladh KW, Nichols MC (2000). Handbook of Mineralogy, vol IV. Arsenates, Phosphates, Vanadates. Mineral Data Publishing, Tuscon, AZ. Appleyard SJ, Angeloni J, Watkins R (2006) Arsenic-rich groundwater in an urban area experiencing drought and increasing population density, Perth, Australia. Appl Geochem 21:83–97. Arnorrson S, Linvall R (2001) The distribution of arsenic, molybdenum and tungsten in natural waters in basaltic terrain North Iceland. In: Cidu R (ed) Proceedings, 10th Water Rock Interaction Symposium, Abingdon, Balkema, pp 961–964. Astolfi EAN, Maccagno A, Garcia-Fernandez JC (1981) Relation between arsenic in drinking water and skin cancer. Biol Trace Elem Res 3:133–143. Ayotte JD, Montgomery DL, Flanagan SM, Robinson KW (2003) Arsenic in groundwater in Eastern New England: occurrence controls and human health implications. Environ Sci Technol 37:2075–2083. Azcue JM, Nriagu JO (1994) Arsenic: historical perspectives. In: Nriagu JO (ed) Arsenic in the Environment. Part I. Cycling and Characterization. Wiley, New York, pp 1–17. ´ rmannssonn H (2006) Environmental impact of the utilization of geothermal areas. Baba A, A Energy Sources Part B 1:267–278. Ballantyne JM, Moore JN (1988) Arsenic geochemistry in geothermal systems. Geochim Cosmochim Acta 52(2):475–483. Bannerman W, Potin-Gautier M, Amoureux D, Tellier S, Rambaud A, Babut M, Adimado A, Beinhoff C (2003) Mercury and arsenic in the gold mining regions of the Ankobra River basin in Ghana. J Phys IV Fr 107:107–110. Barton DB (1971) Essays in Cornish Mining History, vol 2, D. Bradford Barton, Truro, pp 101–125. Bashkin VN, Wongyai K (2002) Environmental fluxes of arsenic from lignite mining and power generation in Northern Thailand. Environ Geol 41(8):883. Bech J, Poschenrieder C, Llugany M, Barcelo J, Tumea P, Tobias FJ, Barranzuela J, Vasquez R (1997) Arsenic and heavy metal contamination of soil and vegetation around a copper mine in Northern Peru. Sci Total Environ 203:83–91. Belkin HE, Zheng B, Finkelman RB (2000) Human health effects of domestic combustion of coal in rural China: a causal factor for arsenic and fluorine poisoning. In: 2nd World Chinese Conference on Geological Sciences Extended Abstracts, Stanford University, Stanford, CA, pp 522–524. Berg M, Tran HC, Nguyen TC, Pham HV, Schertenleib R, Geiger W (2001) Arsenic contamination of groundwater and drinking water in Vietnam: a human health threat. Environ Sci Technol 35(13):2621–2626. Berg M, Stengel C, Trang PTK, Viet PH, Sampson ML, Leng M, Samreth S, Fredericks D (2007) Magnitude of arsenic pollution in the Mekong and Red River Deltas: Cambodia and Vietnam. Sci Tot Environ 372:413–425. BGS/DPHE (2001) Arsenic Contamination of Groundwater in Bangladesh, vol 2. Final Report. In: Kinniburgh DG, Smedley PL (eds) BGS Technical Report WC/00/19. British Geological Survey, Keyworth UK. http://www.bgs.ac.uk/arsenic/bangladesh/reports.htm. Bhattacharya P, Chatterjee D, Jacks G (1997) Occurrence of arsenic-contaminated groundwater in alluvial aquifers from the Delta Plain Eastern India: options for a safe drinking water supply. Water Resour Dev 13:79–92.

Arsenic Pollution Sources

53

Bingqiu Z Jinmao Z Lixin Z Yaxin Z (1986) Mercury, arsenic, antimony, bismuth and boron as geochemical indicators for geothermal areas. J Geochem Explor 25:379–388. Birkle P, Merkel B (2000) Environmental impact by spill of geothermal fluids at the geothermal field of Los Azufres Michoacan Mexico. Water Air Soil Pollut 124(3-4):371–410. Birkle P, Merkel B (2002) Mineralogical-chemical composition and environmental risk potential of pond sediments at the geothermal field of Los Azufres Mexico. Environ Geol 41 (5):583–592. Bolan NS, Thiyagarajan S (2001) Retention and plant availability of chromium in soils as affected by lime and organic amendments. Aust J Soil Res 39:1091–1103. Borba RP, Figueiredo BR, Matschullat J (2003) Geochemical distribution of arsenic in waters sediments and weathered gold mineralized rocks from Iron Quadrangle Brazil. Environ Geol 44:39–52. Bowell RJ (1992) Supergene gold mineralogy at Ashanti Ghana: implications for the supergene behaviour of gold. Miner Mag 56:545–560. Bowell RJ (1993) Mineralogy and geochemistry of tropical rain forest soils: Ashanti Ghana. Chem Geol 106:345–348. Boyle RW, Jonasson IR (1973) The geochemistry of As and its use as an indicator element in geochemical prospecting. J Geochem Explor 2:251–296. Boyle DR, Turner RJW, Hall GEM (1998) Anomalous arsenic concentrations in groundwaters of an island community, Bowen Island, British Columbia. Environ Geochem Health 20:199–212. Brannon JM, Patrick WH (1987) Fixation transformation and mobilization of arsenic in sediments Environ Sci Technol 21:450–459. Breslin VT, Adler-Ivanbrook L (1998) Release of copper, chromium and arsenic from CCA-C treated lumber in estuaries. Estuar Coast Shelf Sci 46(1):111–125. Brouwere KD, Smolders E, Merck R (2004) Soil properties affecting solid–liquid distribution of As(V) in soils. Eur J Soil Sci 55:165–73. Caceres L, Gruttner E, Contreras R (1992) Water recycling in arid regions: Chilean case. Ambio 21:138–144. Carrillo-Cha´vez A, Drever JI, Martı´nez M (2000) Arsenic content and groundwater geochemistry of the San Antonio-El Triunfo, Carrizal, and Los Planes aquifers in southernmost Baja California Mexico. Environ Geol 39(11):1295–1303. Chakraborti D, Rahman MM, Paul K, Chowdhury UK, Sengupta MK, Lodh D, Chanda CR, Saha KC, Mukherjee SC (2002) Arsenic calamity in the Indian subcontinent. What lessons have been learned? Talanta 58:3–22. Chandra Sekhar K, Chary NS, Kamala CT, Venkateswara Rao J, Balaram V, Anjaneyulu Y (2003) Risk assessment and pathway study of arsenic in industrially contaminated sites of Hyderabad: a case study. Environ Int 29(5):601–611. Charlet L, Chakraborty S, Appelo CAJ, Roman-Ross G, Nath B, Ansari AA, Lanson M, Chatterjee D, Basu Mallik S (2007) Chemodynamics of an arsenic hotspot in a West Bengal aquifer: a field and reactive transport modeling study. Appl Geochem 22:1273–1292. Charlet L, Polya DA (2006) Arsenic in shallow, reducing groundwaters in Southern Asia: an environmental health disaster. Elements 2(2):91–96. Chen S-L, Dzeng SR, Yang M-H, Chiu K-H, Shieh G-M, Wai CM (1994) Arsenic species in groundwaters of the Blackfoot disease area Taiwan. Environ Sci Tech 28:877–881. Chen CJ, Hsu LI, Tseng CH, Hsueh YM, Chiou HY (1999) Emerging epidemics of arseniasis in Asia. In: Chappell WR, Abernathy CO, Calderon RL (eds) Arsenic: Exposure and Health Effects. Elsevier Science, Amsterdam, pp 113–121. Choprapawon C, Rodcline A (1997) Chronic arsenic poisoning in Ronpibool Nakhon Sri Thammarat, the Southern Province of Thailand In: Abernathy CO, Calderon RL, Chappell WR (eds) Arsenic: Exposure and Health Effects. Chapman & Hall, London, pp 69–77. Chowdhury TR, Mandal BK, Samanta G, Basu GK, Chowdhury PP, Chanda CR, Karan NK, Lodh D, Dhar RK, Das D, Saha KC, Chakraborti D (1997) Arsenic in groundwater in six districts of West Bengal India: the biggest arsenic calamity in the world: the status report up to August

54

H. Garelick et al.

1995 In: Abernathy CO, Calderon RL, Chappell WR (eds) Arsenic: Exposure and Health Effects. Chapman & Hall, London, pp 93–111. Chowdhury TR, Basu GK, Mandal BK, Biswas BK, Samanta G, Chowdhury UK, Chanda ChR, Lodh D, Roy SL, Saha KCh, Roy S, Kabir S, Quamruzzaman Q, Chakraborti D (1999) Arsenic poisoning in the Ganges Delta. Nature (Lond) 401:545–546. Criaud A, Fouillac C (1989) The distribution of arsenic(III) and arsenic(V) in geothermal waters: examples from the Massif Central of France, the Island of Dominica in the Leeward Islands of the Caribbean, the Valles Caldera of New Mexico, United States, and Southwest Bulgaria. Chem Geol 76(3-4):259–269. Dai S, Ren D, Tang Y, Yue M, Hao L (2005) Concentration and distribution of elements in Late Permian coals from western Guizhou Province China International. J Coal Geol 61(12):119–137. Das D, Chatterjee A, Samanta G, Mandal B, Chowdhury RT, Samanta G, Chowdhury PP, Chanda C, Basu G, Lodh D, Nandi S, Chakraborty T, Mandal S, Bhattacharya SM, Chakraborti D (1994) Arsenic contamination in groundwater in six districts of West Bengal India: the biggest arsenic calamity in the world. Analyst 119:168–170. Das D, Samanta G, Mandal BK, Chowdhury TR, Chanda CR, Chowdhury PP, Basu GK, Chakraborti D (1996) Arsenic in groundwater in six districts of West Bengal India. Environ Geochem Health 18(1):5–15. Datta R, Sarkar D, Sharma S, Sand K (2006) Arsenic biogeochemistry and human health risk assessment in organo-arsenical pesticide-applied acidic and alkaline soils: an incubation study. Sci Total Environ 372:39–48. De Sastre MSR, Varillas A, Kirschbaum P (1992) Arsenic content in water in the Northwest area of Argentina. International Seminar Proceedings: Arsenic in the Environment and its Incidence on Health. Universidad de Chile, Santiago, Chile, pp 91–99. Del Razo LMJ, Hernandez JLG, Garcia-Vargas GG (1994) Urinary excretion of arsenic species in human populations chronically exposed to arsenic via drinking water. A pilot study. In: Chappell WR, Abernathy CO, Cothern CR (eds) Arsenic: Exposure Health Effects. Science and Technology Letters, Northwood, pp 91–100. Ding Z, Zheng B, Long J, Belkin HE, Finkelman RB, Chen CH, Zhou D, Zhou Y (2001) Geological and geochemical characteristics of high arsenic coals from endemic arsenosis areas in southwestern Guizhou Province, China. Appl Geochem 16(11-12):1353–1360. Ellis AJ, Mahon WAJ (1977) Chemistry and Geothermal Systems. Academic Press, New York. Ellice MC, Dowling K, Smith K, Smith E, Naidu R (2001) Abandoned mine tailings with high arsenic concentrations: a case study with implications for regional Victoria Arsenic in the AsiaPacific Region. Abstracts CSIRO Land and Water, Adelaide, South Australia, p 124. Embrick LL, Porter KM, Pendergrass A, Butcher DJ (2005) Characterization of lead and arsenic contamination at Barber Orchard, Haywood County, NC. Microchem J 81(1):117–121. European Commission (2003) Commission Directive 2003/2/EC of 6 January 2003 relating to restrictions on the marketing and use of arsenic (tenth adaptation to technical progress to Council Directive 76/769/EEC). Official Journal of the European Communities, Brussels. Farago ME, Thornton I, Kavanagh PJ, Elliott P, Leonardi G (1997) Health aspects of human exposure to high arsenic concentrations in soil in South-West England In: Abernathy CO, Calderon RL, Chappell WR (eds) Arsenic: Exposure and Health Effects. Chapman & Hall, London, pp 211–226. Farias SS, Casa VA, Vazquez C, Ferpozzi L, Pucci GN, Cohen IM (2003) Natural contamination with arsenic and other trace elements in ground waters of Argentine Pampean Plain. Sci Total Environ 309(1-3):187–199. Ferguson JF, Gavis J (1972) A review of the arsenic cycle in natural waters. Water Res 6(11) 1259–1274. Ficklin WH, Callender E (1989) Arsenic geochemistry of rapidly accumulating sediments, Lake Oahe, South Dakota. In: US Geological Survey Toxic Substances Hydrology Program: Proceedings of the technical meeting, Phoenix, Arizona, September 26–30, 1988. Mallard GE,

Arsenic Pollution Sources

55

Ragone SE (eds) US Geological Survey Water–Resources Investigations Report 88-4220:217– 222. Filippi M, Golias V, Pertold Z (2004) Arsenic in contaminated soils and anthropogenic deposits at the Mokrsko, Roudny, and Kasperske, Hory gold deposits, Bohemian Massif (CZ). Environ Geol 45:716–730. Finkelman RB (2004) Potential health impacts of burning coal beds and waste banks. Int J Coal Geol 59(1-2):19–24. Finkelman RB, Belkin HE, Zheng B (1999) Health impacts of domestic coal use in China. Proc Natl Acad Sci USA 96:3427–3431. Folkes DJ, Kuehster TE, Little RA (2001) Contributions of pesticide use to urban background concentrations of arsenic in Denver, Colorado, USA. Environ Forensics 2(2):127–139. Foust RD, Mohapatra P, Compton-O’Brien AM, Reifel J (2004) Groundwater arsenic in the Verde Valley in central Arizona USA. Appl Geochem 19(2):251–255. Friesbie SH, Ortega R, Maynard DM, Sarkar B (2002) The concentrations of arsenic and other toxic elements in Bangladesh’s drinking water. Environ Health Perspect 110(11):1147–1153. Frizzeli PM (1993) Verification of a Preliminary Site Inspection Guidance Manual for the Idengification of a Potentially Contaminated Land. M.Sc. Thesis, University of London. Goldberg S (1986) Chemical modelling of arsenate adsorption on aluminum and iron oxide minerals. Soil Sci Soc Am J 50:1154–1157. Goldberg S, Glaubig RA (1988) Anion sorption on a calcareous montmorillonitic soil-arsenic. Soil Sci Soc Am J 52:1297–1300. Goldschmidt VM (1954) Arsenic. In Muir A (ed) Geochemistry. Clarendon Press, Oxford, pp 468–475. Gomez JJ, Lillo J, Sahun B (2006) Naturally occurring arsenic in groundwater and idengification of the geochemical sources in the Duero Cenozoic Basin, Spain. Environ Geol 50:1151–170. Grimes DJ, Ficklin WH, Meier AL, McHugh JB (1995) Anomalous gold antimony arsenic and tungsten in groundwater and alluvium around disseminated gold deposits along the Getchell Trend, Humboldt County, Nevada. J Geochem Explor 52:351–371. Guha Mujumder DN, Chakraborty AK, Ghosh A, Gupta JD, Chakraborty DP, Dey SB, Chattopadhyay N (1988) Chronic arsenic toxicity from drinking tubewell water in rural West Bengal. Bull WHO 66:499–506. Guo XJ, Fujino Y, Kaneko S, Wu KG, Xia YJ, Yoshimura T (2001) Arsenic contamination of groundwater and prevalence of arsenical dermatosis in the Hetao plain area, Inner Mongolia, China. Mol Cell Biochem 222:137–140. Gupta SK, Chen KY (1978) Arsenic removal by adsorption. J Water Pollut Control Fed 50:493–506. Hiemstra T, van Riemsdijk WH (1996) A surface structural approach to ion adsorption: the charge distribution CD model. J Colloid Interface Sci 179:488–508. Hingston JA, Collins CD, Murphy RJ, Lester JN (2001) Leaching of chromated copper arsenate wood preservatives: a review. Environ Pollut 111(1):53–66. Hinwood A, Bannister R, Shugg A, Sim M (1998) Environmental arsenic in rural Victoria: an update. Water 25:34–36. Hopp L, Peiffer S, Durner W (2006) Spatial variability of arsenic and chromium in the soil water at a former wood preserving site. J Contam Hydrol 85:159–178. Hsu KH, Froines JR, Chen CH (1997) Studies of arsenic ingestion from drinking water in Northestern Taiwan: chemical speciation and urinary metabolites. In: Abernathy CO, Calderon RL, Chappell WR (eds) Arsenic: Exposure and Health Effects. Chapman & Hall, London, pp 190–209. Hwang Y-H, Bornschein RL, Grote J (1997) Environmental arsenic exposure of children around a former copper smelter site. Environ Res 72(3):72–81. Islam FS, Gault AG, Boothman C, Polya DA, Charnock JM, Chatterjee D, Lloyd JR (2004) Role of metal-reducing bacteria in arsenic release from Bengal delta sediments. Nature (Lond) 430:68–71.

56

H. Garelick et al.

Karayigit AI, Gayer RA, Querol X, Onacak T (2000) Contents of major and trace elements in feed coals from Turkish coal-fired power plants. Int J Coal Geol 44:169–184. Karpov GA, Naboko SI (1990) Metal contents of recent thermal waters mineral precipitates and hydrothermal alteration in active geothermal fluids, Kamchatka. J Geochem Explor 36:57–71. Kavanagh PJ (1998) Impacts of High Arsenic Concentrations in South West England on Human Health and Agriculture. Ph.D. Thesis, University of London. Keegan T, Hong B, Thornton I, Farago M, Jakubis P, Jakubis M, Pesch B, Ranft U, Nieuwenhuijsen MJ (2002) Assessment of environmental arsenic levels in Prievidza district. J Expos Anal Environ Epidemiol 12(3):179–185. Kelly WR, Holm TR, Wilson SD, Roadcap GS (2005) Arsenic in glacial aquifers: sources and geochemical controls. Ground Water 43(4):500–510. Kim M, Nriagu J, Haack S (2000) Carbonate ions and arsenic dissolution by groundwater. Environ Sci Technol 34:3094–3100. Kim M, Nriagu J, Haack S (2002) Arsenic species and chemistry in groundwater of southeast Michigan. Environ Pollut 120:379–390. Kizilstein LY (1997) Geochemistry of berillium in coal: an ecological aspect, Geokhimiya 3:336–342. (cited in Yudovic and Ketris 2005). Kolker A, Haack SK, Cannon WF, Westjohn B, Kim M-J, Nriagu J, Woodruff LG (2003) Arsenic in southeastern Michigan. In: Welch AH, Stollenwerk KG (eds) Arsenic in Ground Water. Kluwer, Boston, pp 281–294. Komnitsas K, Xenidis A, Adam K (1995) Oxidation of pyrite and arsenopyrite in sulfidic spoils in Lavrion minerals. Engineering 12:1443–1454. Kondo HY, Ishiguro Y, Ohno K, Nagase M, Toba M, Takagi M (1999) Naturally occurring arsenic in the groundwaters in the southern region of Fukuoka Prefecture Japan. Water Res 33(8):1967–1972. Kuo TL (1968) Arsenic content of artesian well water in endemic area of chronic arsenic poisoning. Rep Inst Pathol Natl Taiwan Univ 20:7–13. Kwon E, Zhang H, Wang Z, Jhangri GS, Lu X, Fok N, Gabos S, Li X-F, Le XC (2004) Arsenic on the hands of children after playing in playgrounds. Environ Health Perspect 112 (14):1375–1380. Lamb IA, Hughes MJ, Hughes CE (1996) Dispersion of arsenic in soil and water of the Ballarat Goldfield, Victoria. International Conference on Contaminants in the Soil Environment in the Australasian-Pacific Region, February 18–23, Adelaide, South Australia, pp 275–276. Lima A, Cicchella D, Di Francia S (2003) Natural contribution of harmful elements in thermal groundwaters of Ischia Island (southern Italy). Environ Geol 43:930–940. Lin YB, Lin YP, Liu CW, Tan YC (2006) Mapping of spatial multi-scale sources of arsenic variation in groundwater on Chia Nan flood plain of Taiwan. Sci Tot Environ 370:168–181. Liu J, Zheng B, Vasken Aposhian H, Zhou Y, Chen M-L, Zhang A, Waalkes MP (2002) Chronic arsenic poisoning from burning high-arsenic-containing coal in Guizhou China. Environ Health Perspect 110(2):119–122. Loppi S, Bonini I (2000) Lichens and mosses as biomonitors of trace elements in areas with thermal springs and fumarole activity (Mt Amiata, central Italy). Chemosphere 41 (9):1333–1336. Madhavan N, Subramanian V (2000) Sulfide mining as a source of arsenic in the environment. Curr Sci 78(6):702–708. Mandal BK, Suzuki KT (2002) Arsenic round the world: a review. Talanta 58:201–235. Mandal BK, Chowdhury TR, Samanta G, Basu GK, Chowdhury PP, Chanda CR, Lodh D, Karan NK, Dhar RK, Tamili DK, Das D, Saha KC, Chakraborti D (1996) Arsenic in groundwater in seven districts of West Bengal India: the biggest arsenic calamity in the world. Curr Sci 70 (11):976–986. Manning BA, Goldberg S (1996) Modelling competitive adsorption of arsenate with phosphate and molybdate on oxide minerals. Soil Sci Soc Am J 60:121–131.

Arsenic Pollution Sources

57

Manning BA, Goldberg S (1997) Adsorption and stability of arsenic (III) at the clay mineral-water interface. Environ Sci Technol 31:2005–2011. Matschullat J (2000) Arsenic in the geosphere: a review. Sci Tot Environ 249:297–312. McArthur J M, Ravenscroft P, Safiullah S, Thirlwall MF (2001) Arsenic in groundwater: testing pollution mechanisms for sedimentary aquifers in Bangladesh. Water Resour Res 37 (1):109–117. McBride GL, Holland P, Wilson K (1998) Investigation of contaminated sheep dipping sites in the Waikato. In: Proceedings of Waste Management Institute, New Zealand, Conference. Waste Management Institute, New Zealand, Inc., Rotorua, pp 129–137. McLaren RG, Naidu R, Smith J, Tiller KG (1998) Fractionation and distribution of arsenic in soils contaminated by cattle dip. J Environ Qual 27:348–354. Modabberi S, Moore F (2004) Environmental geochemistry of Zarshuran Au-As deposit, NW Iran. Environ Geol 46:796–807. Mok W, Wai CM (1990) Distribution and mobilization of arsenic and antimony species in the Coeur d’Alene River, Idaho. Environ Sci Technol 24:102–108. Moore JN, Woessner WW (2003) Arsenic contamination in the water supply of Milltown Montana. In: Welch AH, Stollenwerk KG (eds) Arsenic in Ground Water: Geochemistry and Occurrence. Kluwer, Norwell, MA, pp 329–350. Mroczek EK (2005) Contributions of arsenic and chloride from the Kawerau geothermal field to the Tarawera River, New Zealand. Geothermics 34:223–238. Mukherjee A, Kumar Sengupta M, Hossain A, Ahamed S, Das B, Nayak B, Lodh D, Rahman MM, Chakraborti D (2006) Arsenic contamination in groundwater: a global perspective with emphasis on the Asian scenario. J Health Popul Nutr 24(2):142–163. Murphy EA, Aucott UM (1998) An assessment of the amounts of arsenical pesticides used historically in a geographical area. Sci Total Environ 218:89–101. Neku A, Tandukar N (2003) An overview of arsenic contamination in Nepal and its removal at household level. J Phys IV 107:941–944. Ng JC, Wang J, Shraim A (2003) A global health problem caused by arsenic from natural sources. Chemosphere 52:1353–1359. Nickson RT, McArthur JM, Burgess WG, Ravenscroft P, Ahmed KZ, Rahman M (1998) Arsenic poisoning of Bangladesh groundwater. Nature (Lond) 395:338. Nickson RT, McArthur JM, Ravenscroft P, Burgess WB, Ahmed KZ (2000) Mechanism of arsenic poisoning of groundwater in Bangladesh and West Bengal. Appl Geochem 15:403–413. Nicolli HB, Suriano JM, Gomez Peral MA, Ferpozzi LH, Baleani OM (1989) Groundwater contamination with arsenic and other trace elements in an area of the Pampa Province of Co´rdoba, Argentina. Environ Geol Water Sci 14:3–16. Nilsson R, Jha AN, Zaprianov Z, Natarajan AT (1993) Chromosomal aberrations in humans exposed to arsenic in the Srednogorie area Bulgaria Fresenius. Environ Bull 2:59–64. Nriagu JO, Pacyna JM (1988) Quantitative assessment of worldwide contamination of air, water and soils by trace metals. Nature (Lond) 333:134–139. Onishi H (1969) Arsenic. In: Wedepohl KH (ed) Handbook of Geochemistry, vol II. SpringerVerlag, New York, p 33. Oyarzun R, Lillo J, Higueras P, Oyarzun J, Maturana H (2004) Strong arsenic enrichment in sediments from the Elqui watershed, Northern Chile: industrial (gold mining at El IndioTambo district) vs. geologic processes. J Geochem Explor 84:53–64. Palmer CA, Klizas SA (1997) The chemical analysis of argonne premium coal samples. US Geological Survey Bulletin 2144. US Geological Survey, Reston, VA. Patinha C, Silva Da EF, Fonseca EC (2004) Mobilisation of arsenic at the Talhadas old mining area, Central Portugal. J Geochem Explor 84(3):167–180. Pesch B, Ranft U, Jakubis P, Nieuwenhuisen M, Hergemoller A, Unfried K, Jakubis M, Miskovic P, Keegan T, EXPASCAN study group (2002) Environmental arsenic exposure from a coal burning power plant as a potential risk factor for nonmelanoma skin carcinoma: results from a case control study in the district of Prievidza, Slovakia. Am J Epidemiol 155(9):798–809.

58

H. Garelick et al.

Peters SC, Blum JD (1999) The source and transport of arsenic in a bedrock aquifer New Hampshire, USA. Appl Geochem 18(11):1773–1787. Peterson ML, Carpenter R (1983) Biogeochemical processes affecting total arsenic and arsenic species distributions in an intermittently anoxic Fjord. Mar Chem 12:295–321. Planer-Friedrich B, Armienta MA, Merkel BJ (2001) Origin of arsenic in the groundwater of the Rio Verde basin Mexico. Environ Geol 40:1290–1298. Planer-Friedrich B, Lehr C, Matschullat J, Merkel BJ, Nordstrom DK, Sandstrom MW (2006) Speciation of volatile arsenic at geothermal features in Yellowstone National Park. Geochim Cosmochim Acta 70:2480–2491. Polya DA, Rowland HAL, Gault AG, Diebe NH, Jones JC, Cooke DA (2004) Geochemistry of arsenic-rich shallow groundwaters in Cambodia. Geochim Cosmochim Acta 68(11): A520–A520. Polya DA, Gault AG, Diebe N, Feldman P, Rosenboom JW, Gilligan E, Fredericks D, Milton AH, Sampson M, Rowland HAL, Lythgoe PR, Jones JC, Middleton C, Cooke DA (2005) Arsenic hazard in shallow Cambodian groundwater. Min Mag 69(5):807–823. Rawlins BG, O’Donnell K, Ingham M (2003) Geochemical Survey of the Tamar Catchment (south-west England). British Geological Survey Report, CR/03/027. Reigart JR, Roberts JR (1999) Recognition and Management of Pesticides Poisoning. http://www. epa.gov/pesticides/safety/healthcare. Rice K, Conko KM, Hornberger GM (2002) Anthropogenic sources of arsenic and copper to sediments in a suburban lake, northern Virginia. Environ Sci Technol 36(23):4962–4967. Robinson BH, Brooks RR, Outred HA. Kirkman JH (1995) The distribution and fate of arsenic in the Waikato River System North Island New Zealand. Chem Speciation Bioavail 7(3):89–96. Robinson GR, Ayotte JD (2006) The influence of geology and land use on arsenic in stream sediments and ground waters in New England, USA. Appl Geochem 21(9):1482–1497. Robinson GR, Ayuso RA (2004) Use of spatial statistics and isotopic tracers to measure the influence of arsenical pesticide use on stream sediment chemistry in New England, USA. Appl Geochem 19:1097–1110. Saha AK, Chakrabarti C (1995) Geological and geochemical background of the As bearing groundwater occurrences of West Bengal. In: Proceedings of the International Conference on Arsenic in Groundwater: Cause, Effect and Remedy, Calcutta, p 42–43. Schreiber ME, Simo JA, Freiberg PG (2000) Stratigraphic and geochemical controls on naturally occurring arsenic in groundwater eastern Wisconsin, USA. Hydrogeol J 8:161–176. Schreiber ME, Gotkowitz MB, Simo JA, Freiberg PG (2003) Mechanism of arsenic release to ground water from naturally occurring sources Eastern Wisconsin. In: Welch AH, Stollenwerk KG (eds) Arsenic in Groundwater: Geochemistry and Occurrence. Kluwer, Boston, pp 259–280. Schwenzer SP, Tommaseo CE, Kersten M, Kirnbauer T (2001) Speciation and oxidation kinetics of arsenic in the thermal springs of Wiesbaden spa Germany Fresenius. J Anal Chem 371:927–933. Serfor-Armah Y, Nyarko BJB, Adotey DK, Dampare SB (2006) Levels of arsenic and antimony in water and sediment from Prestea, a gold mining town in Ghana and its environs. Water Air Soil Pollut 175:181–192. Shalat SL, Solo-Gabriele HM, Fleming LE, Buckley BT, Black K, Jimenez M, Shibata T, Durbin M, Graygo J, Stephan W, Van De Bogart G (2006) A pilot study of children’s exposure to CCA-treated wood from playground equipment. Sci Total Environ 367:80–88. Shih C-J, Lin C-F (2003) Arsenic contaminated site at an abandoned copper smelter plant: waste characterization and solidification/stabilization treatment. Chemosphere 53:691–703. Shiikawa M (1983) The role of mercury arsenic and boron as pathfinder elements in geochemical exploration for geothermal energy. J Geochem Explor 19:337–338. Shimada N (1996) Geochemical conditions enhancing the solubilization of arsenic into groundwater in Japan. Appl Organomet Chem 10:667–674.

Arsenic Pollution Sources

59

Slotnick MJ, Meliker J, Nriagu J (2003) Sources of arsenic in Southeastern Michigan groundwater. J Phys IV 107:1247–1250. Smedley P (2003) Arsenic in groundwater: south and east Asia. In: Welch AH, Stollenwerk KG (eds) Arsenic in Groundwater: Geochemistry and Occurrence. Kluwer, Boston, pp 179–209. Smedley PL, Kinniburgh DG (2002) A review of the source behaviour and distribution of arsenic in natural waters. Appl Geochem 17:517–568. Smedley PL, Edmunds WM, Pelig-Ba KB (1996) Mobility of arsenic in groundwater in the Obuasi area of Ghana. In: Appleton JD, Fuge R, McCall GJH (eds) Environmental and Geochemical Health. Geological Society Special Publication 113. Geological Society, London, pp 163–181. Smedley PL, Zhang M, Zhang G, Luo Z (2003) Mobilisation of arsenic and other trace elements in fluviolacustrine aquifers of the Huhhot Basin, Inner Mongolia. Appl Geochem 18(9):1453–1477. Smedley PL, Kinniburgh DG, Macdonald DMJ, Nicolli B, Barros AJ, Tullio JO, Pearce JM, Alonso MS (2005) Arsenic associations in sediments from the loess aquifer of La Pampa, Argentina. Appl Geochem 20:989–1016. Smith E, Smith J, Smith L, Biswas T, Correll R, Naidu R (2003) Arsenic in Australian environment: an overview. J Environ Sci Health Part A Toxic/Hazardous Substances Environ Eng A38 (1):223–239. Solo-Gabriele HM, Townsend T (1999) Disposal practices and management alternatives for CCA treated wood waste. Waste Manag Res 17(5):378–389. Stauffer RE, Thompson JM (1984) Arsenic and antimony in geothermal waters of Yellowstone National Park, Wyoming, USA. Geochim Cosmochim Acta 48(12):2547–2561. Stuben D, Berner Z, Chandrasekharam D, Karmakar J (2003) Arsenic enrichment in groundwater of West Bengal India: geochemical evidence for mobilization of As under reducing conditions. Appl Geochem 18(9):1417–1434. Sun G (2004) Arsenic contamination and arsenicosis in China. Toxicol Appl Pharmacol 198(3):268–271. Tanner CC, Clayton JS (1990) Persistence of arsenic 24 years after sodium arsenite herbicide application to Lake Rotoroa, Hamilton, New Zealand. N Z J Mar Freshw Res 24:173–179. Townsend T, Dubey B, Tolaymat T, Solo-Gabriele H (2005) Preservative leaching from weathered CCA-treated wood. J Environ Manag 75(2):105–113. Thompson WT (1973) Agricultural Chemicals Book. 1: Insecticides. Thompson Publication, Washington (Cited in Mandal and Suzuki 2002). Thornton I, Farago ME (1997): The geochemistry of arsenic. In: Abernathy CO, Calderon RL, Chappell WR (eds) Arsenic: Exposure and Health Effects. Chapman & Hall, London, pp 1–16. Ure A, Berrow M (1982) The elemental constituents of soils. In: Bowen HJM (ed) Environmental Chemistry of the Elements. The Royal Society of Chemistry, London, pp 94–203. Ursitti F, Vanderlinden L, Watson R, Campbell M (2004) Assessing and managing exposure from arsenic in CCA-treated wood play structures. Can J Public Health 95(6):429–433. US Department of Agriculture (1970) The Pesticides Review, 1970. USDA, Washington, DC (Cited in Mandal and Suzuki 2002). US Environmental Protection Agency (1992a) Chemical-specific parameters for toxicity characteristic contaminants. Document 600/S, 3-91/004. US Environmental Protection Agency (1992b) Method 1311: Toxicity Characteristic Leaching Procedure. http://www.epa.gov/epaoswer/hazwaste/test/pdfs/1311.pdf. US Environmental Protection Agency (2006) Chromated Copper Arsenate (CCA). http://www. epa.gov/oppad001/reregistration/cca/ (Accessed February 2006). US Geological Survey (USGS) Mineral Commodity Summaries for Arsenic (2006). http://minerals.usgs.gov/minerals/pubs/commodity/arsenic/arsenmcs06.pdf (Accessed January 2006). Utsonomiya S, Peters SC, Blum JD, Ewing RC (2003) Nanoscale mineralogy of arsenic in a region of New Hampshire with elevated As-concentrations in the groundwater. Am Mineral 88:1844–1852.

60

H. Garelick et al.

Versanyi I, Kovacs LO (2006) Arsenic, iron and organic matter in sediments and groundwater in the Pannonian Basin, Hungary. Appl Geochem 21:949–963. Wang GQ, Huang YZ, Gang JM, Wang SZ, Xiao BY, Yao H, Hu Y, Gu YL, Zhang C, Liu KT (2000) Endemic arsenism fluorosis and arsenic-fluoride poisoning caused by drinking water in Kuitun, Xinjiang. Chin Med J 113:524–534. Wang S, Mulligan CN (2006) Occurrence of arsenic contamination in Canada: sources, behavior and distribution. Sci Total Environ 366:701–721. Wang SW, Liu CW, Jang CS (2007) Factors responsible for high arsenic concentrations in two groundwater catchments in Taiwan. Appl Geochem 22:460–476. Webster JG (1990) The solubility of As2S3 and speciation of As in dilute and sulfide-bearing fluids at 25 and 90 C. Geochim Cosmochim Acta 54:1009–1017. Webster JG (1999) The source of arsenic and other elements in the Marbel-Matingao river catchment, Mindanao, Philippines. Geothermics 28(1):95–111. Webster JG, Nordstrom DK (2003) Geothermal arsenic. In: Welch AH, Stollenwerk KG (eds) Arsenic in Ground Water: Geochemistry and Occurrence. Kluwer, Boston, pp 101–126. Webster JG, Nordstrom DK, Smith KS (1994) Transport and natural attenuation of Cu Zn As and Fe in the acid mine drainage of Leviathan and Bryant Creeks. In: Alpers CN, Blowes DW (eds) Environmental Geochemistry of Sulfide Oxidation. American Chemical Society Symposium Series 550. ACS, Washington, DC, pp 244–260. Welch AH, Lico MS, Hughes JL (1988) Arsenic in groundwater of the western United States. Ground Water 26(3):333–347. Welch AH, Helsel DR, Focazio MJ, Watkins SA (1999) Arsenic in ground water supplies of the United States. In: Chappell WR, Abernathy CO, Calderon RL (eds) Arsenic: Exposure and Health Effects. Elsevier, Amsterdam pp 9–17. Welch AH, Westjohn DB, Helsel DR, Wanty RB (2000) Arsenic in ground water of the United States: occurrence and geochemistry. Ground Water 38(4):589–604. White DE, Hem JD, Waring GA (1963) In: Fleischer M (ed) Data of geochemistry chemical composition of sub surface waters. US Geological Survey Professional Paper 440-F. Williams M (1997) Mining-related arsenic hazards: Thailand. Case-study Summary Report. British Geological Survey Technical Report WC/97/49. Williams M, Fordyce F, Paijitprapapon A, Charoenchaisri P (1996) Arsenic contamination in surface drainage and groundwater in part of the southeast Asian tin belt, Nakhon Si Thammarat Province, southern Thailand. Environ Geol 27:16–33. Wilson FH, Hawkins DB (1978) Arsenic in streams stream sediments and ground water, Fairbanks area, Alaska. Environ Geol 2:195–202. Yang CH, Chang C, Tsai SS, Chuang HY, Ho CK, Wu TN (2003) Arsenic in drinking water and adverse pregnancy outcome in an arseniasis-endemic area in northeastern Taiwan. Environ Res 91:29–34. Yokoyama T, Takahashi Y, Tarutani T (1993) Simultaneous determination of arsenic and arsenious acids in geothermal water. Chem Geol 103:103–111. Yudovich YE, Ketris MP (2005) Arsenic in coal: a review. Int J Coal Geol 61(3-4):141–196. Zartarian VG, Xue J, Ozkaynak H, Dang W, Glen G, Smith L, Stallings C (2006) A probabilistic arsenic exposure assessment for children who contact CCA-treated playsets and decks. Part 1: Model methodology, variability results and model evaluation. Risk Anal 26(2):515–531. Zhang H, Ma D, Hu X (2002) Arsenic pollution in groundwater from Hetao Area, China. Environ Geol 41:638–643.

Onsite Testing for Arsenic: Field Test Kits Jo¨rg Feldmann

I II III

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Criteria for Field Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Analytical Methods for Water Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 A Laboratory-Based Reference Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 B Field Testing Kits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 C Analytical Methods for Sludge and Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 D Analytical Methods for Volatile Arsenic in Gas Samples . . . . . . . . . . . . . . . . . . . . . 73 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

I Introduction Arsenic, a widely distributed element, is toxic and carcinogenic. It occurs naturally at readily detectable concentrations in soils and sediments and can contaminate freshwater and biota. Additionally, anthropogenic influences from mining, leather and glass production, and the use of pesticides, including wood preservatives, can contribute to environmental contamination by arsenic. Arsenic derives from various sources, has a highly diverse distribution pattern, and has a rather complex environmental behavior resulting from its tendency to transform into different species, each of which has different mobility and toxicity. The variable distribution of arsenic creates challenges when arranging sampling programs adequate to establish an accurate picture of the true status of arsenic as an environmental contaminant. Even when reliable sampling is achieved, success still depends on accomplishing rapid onsite sample analysis at an acceptable cost. In recent years, the main focus of arsenic monitoring has been on drinking water, although soil and staple food monitoring is also important, because these contribute significantly to the arsenic burden of people living in affected areas. Furthermore, solids produced from remediating arsenic-contaminated water also require onsite J. Feldmann Department of Chemistry, University of Aberdeen, Aberdeen, UK, e-mail: [email protected]

D.M. Whitacre (ed.), Reviews of Environmental Contamination Volume 197. doi: 10.1007/978-0-387-79284-2_3, # Springer Science þ Business Media, LLC 2008

61

62

J. Feldmann

monitoring (MTI Diagnostic 2005). In fact, proper assessment of arsenic contamination requires onsite measurements of water, soil, sludge, and foodstuffs, although such monitoring must fulfill certain criteria. It is the intent of this chapter to review these criteria.

II

Criteria for Field Monitoring

In this chapter, we emphasize water analysis but also make relevant comments that concern analysis of solids and gases. In many countries, the maximum contaminant level (MCL, 50 mg/L) for total arsenic in groundwater or potable water is a critical index to judge safety (WHO 1993). Recently, the World Health Organization (WHO) guideline value (10 mg/L) recommended for drinking water has been approved by many countries. In the United States, this 10 mg/L value has been enforced since the beginning of 2006. Unfortunately, neither the MCL nor WHO guidelines differentiate among arsenic species, despite the fact that the toxicity of arsenic varies enormously with its speciation. The situation is not as straightforward for soil, sludge, and foodstuffs as it is for water. None of these solid media is strictly regulated, although a few guidelines exist. In the UK, environmental quality criteria exist for soil and vary according to use: domestic gardens, allotments, and play areas have a critical value of 10 mg/kg; landscapes and building hard coverings have a threshold of 40 mg/kg (Visser 1993). In Canada, criteria for soil assessment in contaminated areas start at 12 mg/kg (CCME 2003); remediation criteria are, again, use dependent; e.g., the upper limit for agricultural use (soils) is 20 mg/kg, whereas limits are 30 and 50 mg/kg for residential and industrial land, respectively. Guidelines recommended for arsenic content in foodstuffs are contentious. At present, only Australia has established a guideline (1 mg/kg) for total arsenic (no discrimination among species) in foodstuffs (National Food Authority 1993). Although this level is considered as to be too generous by many, only China has thus far introduced maximum levels for a variety of foodstuffs (USDA 2006). Interestingly, to date, China is also the only country to incorporate arsenic speciation into their guidelines (e.g., 0.15 mg inorganic arsenic per kg rice). It has long been recognized that human intake of arsenic from drinking contaminated water increases the risk of contracting cancer. The link between intake and cancer has fomented an enormous undertaking to develop field kits adequate to analyze arsenic in water samples. Today, a range of these kits are commercially produced and available for sale. In contrast, there are few methods for direct arsenic measurement in solids. The only exception is the handheld XRF (X-ray fluorescence) probe, which enjoys reasonable success for screening arsenic content in brownfield soils. Alternative instruments such as SERS (surface-enhanced Raman spectroscopy) or LIBS (laser-induced breakdown spectroscopy) are currently not commercially available (Melamed 2005).

Onsite Testing for Arsenic: Field Test Kits

63

Field kits have been developed to detect arsine gas and are mainly used in occupational settings (Dra¨ger Safety 2006). These are not covered at length in this review. In summary, field data on arsenic contamination can only be effectively collected if commercially available instruments and methods meet the following criteria: these must l l l l l

be sensitive, accurate, and sufficiently precise for decision making be sufficiently reliable (do not routinely require backup or retesting) have a reasonably short duration of analysis be cost-effective be easy to use

III A

Analytical Methods for Water Samples Laboratory-Based Reference Methods

Standard laboratory methods for determining total arsenic utilize one type or another of atomic spectrometric detection: atomic absorption spectrometry (AAS), atomic fluorescence spectrometry (AFS), inductively coupled plasma atomic emission spectrometry (ICP-AES), and/or mass spectrometry (ICP-MS). All these methods are expensive and require skilled operators and modern laboratory services, such as cooling water, an electricity supply, and purified gases. Thus, these methods are generally unsuitable for use in the field. All the foregoing methods detect only arsenic, the element, but provide no information on arsenic speciation, i.e., the molecular forms of arsenic. Hence, these techniques must be coupled with other species-selective processes to be complete. Species-selective methods include certain pretreatments, such as use of hydride to separate arsenic species that form volatile arsines, or utilization of high performance liquid chromatography (HPLC) or gas chromatography (GC). Moreover, anodic stripping voltammetry (ASV) can be directly used to detect arsenic species, although this method is in decline because it suffers from severe matrix interferences that only skilled operators can overcome. Another consideration is that, when hydride generation methods are used, arsenic should be in a form that readily and quantitatively produces arsines. When used under strict QA/QC (Quality Assurance/Quality Control) regimes, these aforementioned methods can be effective in determining arsenic at sub-mg/L levels in water with an accuracy of 10%. However, this level of sensitivity and accuracy comes at the expense of considerable time and cost. The effectiveness of field test kits are adjudged by comparing the results they generate with those generated using the more sophisticated (and expensive) laboratory-based methods.

64

B

J. Feldmann

Field Testing Kits

In general, colorimetric techniques or electrochemical sensing are the methods used in field kits for arsenic detection. Colorimetric Methods There are a variety of commercial colorimetric field kits for determining arsenic. All of these kits rely on the 100-yr-old Gutzeit reaction (Gutzeit 1891) for their success. This reaction generates arsine (AsH3) gas by reducing arsenic under acidic conditions with addition of zinc powder, thus: AsðOHÞ3 þ 3Zn þ 6HCl ! AsH3 þ 3ZnCl2 þ 3H2 O H3 AsO4 þ 4Zn þ 8HCl ! AsH3 þ 4ZnCl2 þ 4H2 O Only a few metals or metalloids can form volatile compounds that can readily be stripped from solution; hence, this reaction introduces a degree of selectivity. The generated arsine gas is trapped by reagents to form a colored complex. The intensity of the color is used for quantification. The color intensity obtained is evaluated by comparison with a standard color chart to achieve semiquantitative measurements, or a simple spectrophotometer is used to assess the nature of complexes when they are formed. Commercial kits commonly employ impregnated mercury (II) bromide paper and/or a solution of diethyldithiocarbamate (DEDTC) (Rowe et al. 1973) as reagents. The Gutzeit-type field kit uses test paper, which produces a yellow stain after reacting with arsine. More arsine engenders a darker color, and it is suggested that a different complex may be formed, although the structure of the formed complex is unknown (Raham et al. 2004). The shade of color on the impregnated paper correlates to the amount of arsine produced and, therefore, to the concentration of arsenic in the original sample. The color produced during analysis is compared with colors on a reference test sheet, from which the arsenic concentration in each sample can be read. The method is easy to use and is capable of detecting arsenic concentrations as low as 1 mg/L. However, the analysis is often only semiquantitative. The major reason for this is that the method lacks specificity because other elements such as antimony and sulfur can also produce color on the impregnated paper. Antimony, as stibine, and sulfur, as sulfides, are volatile compounds and can react with the paper as well. The interference by antimony is negligible, because its mobility and concentration are usually one-tenth of that of arsenic. Sulfides can be removed before reaching the impregnated test paper by being absorbed onto filters. Alternatively, gaseous arsine can react with silver DEDTC, which forms an intense red complex of unknown chemical structure. Photometric measurement of the purplish-red color produced by the colloidally dispersed silver at an absorption maximum (535–540 nm) is correlated to the concentration of arsenic in solution. Another detection method, based on chemo-trapping arsine by a different approach, is sometimes used. In this method, arsine forms a hetropolymolybdate blue complex that can be detected spectrophotmetrically at 840 nm, although not

Onsite Testing for Arsenic: Field Test Kits

65

directly. Arsine is oxidized by hypobromite to arsenate during the trapping process. The arsenate then reacts with ammonium molybdate in the presence of hydrazine sulfate to form a complex that is spectrophotometrically analyzed. Unfortunately, none of these field kits is able to distinguish between arsenic species, with the exception that arsenite and arsenate can be differentiated. Although dimethylarsinic acid DMA(V) and methylarsonic acid MA(V) can be volatilized as their respective methylarsines, neither they, nor other organoarsenicals, are addressed by the field kits just discussed. This limitation is unfortunate because methylated arsenic species are common in water samples and in wastewaters that have high microbial activity. Besides occurring naturally, methylated arsenics in groundwater can originate from anthropogenic sources, such as insecticides or fungicides (see Garelick et al. 2008, this volume). Therefore, because of insensitivity to organic arsenic species, test kits may under report arsenic concentrations and be inherently negatively biased. Table 1 provides an overview of most colorimetric-based commercially available field kits. Performance of Commercial Colorimetric Test Kits The Environmental Technology Verification (ETV) Program was formed to facilitate the implementation of innovative new technologies for environmental monitoring, and it is a concept that the US EPA supports (US EPA 2006). The ETV manager works in partnership with testing organizations, vendors, and developers as well as potential buyers. In July 2002, the ETV initiated tests of several commercially available arsenic field kits. Testing was conducted under field conditions with both trained technical and untrained (nontechnical) operators. Table 2 presents the results of this testing and includes assessment of the ‘‘new generation’’ of field testing kits. Table 1 Summary of selected commercially available colorimetric field kits for analysis of arsenic in water Name Method MDL (mg/L) Time Capital cost per 100 tests Merck Gutzeit HgBr2 1–20 or 30 min– $50 5–500 120 hr Test-chart Hach EZ arsenic test 10–500 20–40 min $80–50 kit GPL test kit Bromide strips 10–2500 20 min $43 Ditital Arsenator Gutzeit method with 2–100 30 min (Wagtech Int.) photometric readout Visual As detection Gutzeit 10–500 30 min kit 200 Wagtech Test chart or photometry NIPSOM field kit Gutzeit 10–700 10 min $18 Test chart MDL, method detection limit.

Accuracy

0% to 84%

7% to 91%

16%–24% (TO) NTO: Slope 0.59 Offset 0.095 R ¼ 0%–38% 0.98 TO: Slope (NTO) 0.66 Offset–0.30 R = 0.96

93% to 104% (TO) 67% to 81% (NTO)

61 % to þ10% (TO) 77 to 96% (NTO)

QuickTM Low Range II Compuscan

QuickTM II color chart 3.6 to 7

0.5 to 3.9

0.7 to 2.1

No significant effects

No significant effects

No significant effects

No significant effects

1.2 to 1.5

98% to –27% (TO) 76%–9% (NTO)

No performance differences for ob

No significant effects

33 (NTO) 28 (TO)

Slope: 1.27 Offset: 2.25 R = 0.990 TO: Slope 0.79 Offset0.33 R = 0.99 NTO: Slope 0.71 Offset 7.29 R = 0.90 NTO: Slope 0.42 Offset 0.29 R = 0.96 TO: Slope 0.71 Offset 2.45 R = 0.98 NTO: Slope 0.52 Offset 3.37 R = 0.98 TO: Slope 0.88 Offset –1.82 R = 0.98

11%–38% (NTO) 12%– 71% (TO) 0% to 55%

Better performance of NTO Unit differences are not significant Better performance of NTO Unit differences are smaller but significant Better performance of NTO

Better performance of NTO

No performance differences for ob

Interunit Reproducibility (ub), operator bias (ob)

15–50 (NTO) 20– No significant 40 (TO) effects

Method detection Matrix effects or limit (MDL) interferences4 (mg/L)

Slope: 0.99 Offset: 2.41 R = 0.977 (NTO)

Linearity3

0%–40% (NTO) 0%– 26% (TO)

Precision2

QuickTM Low Range II Arsenic scan

2%–17% (at 10 PeCo (Peters Engineering) With mg/L1) 1%– visual testing 113% (at 23–93 mg/L) 10 mg/L1  As 75 (Peters Engineering) With 1%–157% >10 electronic testing mg/L  6%–310% QuickTM Low 92% to –8% (TO) 74%–74% Range II color (NTO) chart

Products

Fp: 0% Fn: 19% TO Fn: 24% NTO

Fp: 0–3% Fn: 52–67% TO Fn: 9% NTO

Fp: 0% Fn: 62% TO Fn: 38% NTO

Fp: 2% NTO 13% TO fn: all 0% Fp: 0% Fn: 62% TO Fn: 33% NTO

Fp: 0% NTO 3% TO Fn: all 0%

15‐min analysis 50 samples for $220

15-min analysis 50 samples for $350 Capital cost: $1,600

15-min analysis 50 samples for $350 Additional to $1,600

Cost of tester $330 Additional 100 cost $60 15-min analysis 50 samples for $350

100 samples for $200

Costs and time Rate of false positive (fp) and false negative (fn) at 10 mg/L

Table 2 Environmental Technology Verification (ETV) joint statement for arsenic field test methods (June 2002, August 2003)1

66 J. Feldmann

38% to þ239% (TO) 81% to 579% (NTO)

93% to 99% (TO) 86% to 66% (NTO)

74% to 31%

13%–91% Accuracy was sample dependent: (NTO) 3%– 37% (TO) 3% to 83%

QuickTM Low range color chart

QuickTM Low Range Arsenic scan

PDV 6000 with VAS 2.1

NanoBandTM Explorer Slope 1.28 Offset –10.73 R¼ 0.956

12.1–14.2

5.8–8.6

Effects from drinking water matrix and low and high levels of iron and sulfide No apparent matrix effect

Op: requires technical skilled personnel

Ub exist on 5% significance level

Positive bias with No significant difference (ob) No higher levels of sodium chloride, ub sulfide, and iron

4.0 to 7.2 mg/L

No significant differences (ob) Significant unit biased (ub)

Better performance of NTO No ub

Positive bias with Better performance of higher levels of sodium chloride, NTO sulfide, and iron

No significant effects

No significant effects

3.1 to 6.7

3.7 to 18.2

4.5 to 6.1

2

ETV Joint Verification Statement (US EPA, Battelle), http://www.epa.gov/etv/verifications/vcenter1-21.html. NTO, nontechnical operator; TO, technical operator. 3 Linearity: results ¼ slope  reference value þ offset. 4 Tested for high levels of sodium sulfate, iron, or acidity.

1

71% to þ96% (TO) 82% to 108% (NTO)

QuickTM II Compu-scan

11%–44% (TO) NTO: Slope 1.39– 0.85 Offset R ¼ 13%–38% 0.93 TO: Slope (NTO) 0.73 Offset–0.55 R ¼ 0.92–0.93 10%–58% (TO) NTO: Slope 0.40 Offset 3.05 R ¼ 16%–108% 0.91 TNO: Slope (NTO) 0.67 Offset 1.39 R ¼ 0.92 0%–10% (TO) NTO: Slope 0.90 0%–23% Offset 2.78 R ¼ (NTO) 0.98 TO: Slope 0.83 Offset þ2.61 R ¼ 1.00 5%–23% (TO) NTO: Slope 0.68 Offset 0.99 R ¼ 0%–42% 0.966 TO: Slope (NTO) 0.85 Offset 0.83 R ¼ 0.997 6%–16% Slope 1.17 Offset –1.56 R ¼ 0.995

78% to –4% (TO) 85% to – 22% (NTO)

QuickTM II Arsenic scan

15‐min analysis 50 samples for $220 Capital cost: $1,600

15‐min analysis 50 samples for $220 Additional to $1,600

Fp: 0% NTO Fp: 13% TO Fn: 22% NTO Fn: 7% TO

Fp: 0% Fn: 38%–42%

List price of instrument $8,000

List price of instrument: $7,900

Fp: 0%–3% Fn: 15‐min analysis 14%–19% TO 50 samples for Fn: 10% NTO $220 Capital cost: $1,600

Fp: 3% TO Fp: 15‐min analysis 12.5% NTO Fn: 50 samples for 80 0% TO Fn: 14% NTO

Fp: 3%–9% TO and 0% TO Fn: 38%–10% TO Fn: 14% NTO

Fp: 0% Fn: 19%–33% TO Fn: 29% NTO

Onsite Testing for Arsenic: Field Test Kits 67

68

J. Feldmann

The table presents the new commercial systems, in particular those test kits developed to improve sensitivity (Merck, Hach tests). These new colorimetric methods have detection limits below 10 mg/L, which is sufficient to make decisions on whether water contains more or less than 50 mg/L (the guideline value for drinking water in countries such as Bangladesh and India). Results from older test kits, such as Merckoquant, are inadequate to allow such decision making. Although the new colorimetric test kits can detect less than 10 mg/L, and do have good reproducibility, they often give false results, primarily false positives (i.e., concentrations <10 mg /L are reported to be >10 mg/L). Oddly, and for unknown reasons, test results indicated that nontechnical operators consistently generated better results than did technical operators (US EPA 2006). Moreover, this occurred not with one but with a series of kits. Concerns have been raised for the safety of operators because of the high levels of toxic arsine gas generated during use of some test kits (Hussam et al. 1999). Because arsine is the most toxic molecular form of arsenic, it is advisable to use the field kits only in well-vented rooms or outdoors. Surveys indicated that the unit sample costs of field kits were an issue when more than 1,000 water samples were analyzed. The cost of the test kits, themselves, may also be a consideration. The overall costs for consumables vary between $2 and $7 per sample. With tested kits, the addition of photometric readouts and a computer ($1,600) seems unjustified, because they have shown no significant improvement in colorimetric kit performance. Electrochemical Sensors for Water Sample Analysis In addition to colorimetric test field kits, those based on electrochemical principles are used for arsenic analysis. Most commercially available electrochemical sensors are based on anodic stripping voltammetry (ASV) (Table 3). Table 3 Commercially available electrochemical sensors for onsite measurements of arsenic in water Name Method MDL Interference Positives Negatives (mg/L) ASV 13 Sensitivity strongly High sensitivity Difficult method Nanodepends on No Requires BandTM Explorer matrix (Hg, generation of trained Cu, Zn) toxic waste operator Expensive <1 Sensitivity strongly High sensitivity Commercial MetalyzerTM ASV 5000 depends on No product was matrix (Hg, generation of discontinued Cu, Zn) toxic waste PDV 6000 ASV 8 Sensitivity strongly High sensitivity Difficult method depends on No Requires matrix (Hg, generation of trained Cu, Zn) toxic waste operator Expensive ASV, anodic stripping voltammetry.

Onsite Testing for Arsenic: Field Test Kits

69

These methods use anodic stripping to quantify all dissolved arsenic (as arsenite and arsenate), at a potential of þ145 mV, with respect to the standard calomel electrodes from a conditioned gold-plated electrode. The analysis by ASV involves three steps: 1. A glassy carbon electrode is conditioned by plating a thin film of gold onto the electrodes 2. The sample is acidified 3. The dissolved arsenic is reduced on the electrode surface Antimony and bismuth are potentially interfere with arsenic, when using the ASV method, even in small concentrations; copper interferes only when its concentration is greater than 100 times that of arsenic (Cavicchioli et al. 2004). However, interference levels as high as 5 ppm of antimony or 10 ppm sulfide did not affect measurement using NanoBandTM Explorer (Abbgy et al. 2002).

Performance of Commercial Electrochemical Test Kits The EPA found that many arsenic test kits suffer from a large, mainly positive, systematic error; errors exceeding 10,000% higher than reference concentrations are possible. Only one kit [Industrial Test Systems (ITS) QuickTM Ultra Low II Kit] produced results with a negative bias (i.e., the actual concentration is higher than the one determined by the kit). The sensors also showed an analytical bias. Although the Nano-BandTM Explorer showed only a positive bias, up to 500%, the PDV 6000 produced biases of both a negative and positive nature (–74% to 31%). From a human health exposure perspective, false-negative determinations are more serious because they may result in unsafe levels of arsenic being unknowingly consumed. Conversely, false-positive determinations may result in mitigation/remediation actions when they are unneeded. Repeatability measurements for colorimetric tests range from 0% to 139% (relative standard deviation, RSD) of the original measurement, whereas the electrochemical sensors show a slightly better repeatability (3%–91%) for the NanoBand Explorer and 3%–16% for the PDV 6000 units. Colorimetric field tests show very limited matrix effects when sulfide filters are used (Spear et al. 2006). Sodium chloride, iron, sulfate, and acidity do not influence analysis results. High iron and/or hydrogen sulfide may influence performance of the electrochemical sensors, in particular that of PDV 6000 (US EPA 2006). The cost of the analysis can vary, depending on the number of analyses, the sophistication of the instrument, and the cost of consumables. For some kits, the cost of one analysis is approximately $1, although in others costs may be as high as $8–$10. The initial cost of electrochemical sensors is high (~$8,000). Because the cost for consumables is modest, one can amortize the cost of such sensors over several thousand samples in long campaigns, which renders them cost-effective.

70

J. Feldmann

Independent Field Assessments In recent years, a few papers have reported arsenic concentrations in polluted areas that compare field- versus laboratory-derived values. Such studies have chiefly focused on Bangladesh (Hussam et al. 1999; Pande et al. 2001; Rahman et al. 2002; Brickson 2003; Jalil and Feroze 2003; Dhar et al. 2004; Khandaker 2004; Deshpande and Pande 2005; van Geen et al. 2005). In addition, two excellent reviews on field kits and onsite monitoring have been published (Kinniburgh and Kosmus 2002; Melamed 2005). Since 1997, the World Bank, UNICEF, WHO, and other organizations have pressed for the testing of tube wells in Bangladesh and West Bengal. The field kits described herein are currently used in the affected areas. In Bangladesh alone, more than 1.3 million tube wells have been tested and decisions made from results garnered using these field kits; one result of testing was that tube wells containing water with arsenic concentrations higher than 50 mg/L were painted red, while those with levels lower than this value were painted green (ESCAP–UNICEF–WHO 2001). Most decisions made before 2001 were based on results from field kits, which had a specified minimum detection limit of 100 mg/L (Merckoquant1 arsenic kit, 1.10026.0001, Merck). The Bangladesh Arsenic Mitigation Water Supply Project (BAMWSP) (Chowdhury 2002), and UNICEF (DPHE–UNICEF 2001) tested more than 1 million wells; other organizations tested a smaller cohort of wells (BRAC 2000; Fact Sheet 12 2002; NGO Forum 2002). However, the study of Hussam et al. (1999), which showed a very poor correlation between results from a field test kit (Merck) and those from laboratory-based methods (R2 = 0.2–0.5), triggered a further series of investigations. Rahman et al. (2002) retested the samples with laboratory-based reference methods using HG-AAS and found that 45% (representing 2,866 wells) had been misclassified by being placed in the lower range (<50 mg/L). Since 2001, only the Merck (Merckoquant, sensitive, 1.17926.0001) or Hach kits, both with a method detection limit (MDL) of 10 mg/L, have had sufficient sensitivity and reproducibility to be used in ongoing surveys. Recently, Merck introduced a highly sensitive kit with an MDL of 5 mg/L (Merckoquant, highly sensitive, 1.17927.0001 As). In a comprehensive study using three field kits (Merckoquant, NIPSOM, and GPL), more than 290 wells were tested against reference methods (Rahman et al. 2002). At arsenic concentrations below 50 mg/L, the rates of false positives were acceptable: 9.2% (NIPSOM) and 6.5% (GPL). In the range between 50 and 100 mg/L, the rate of false positives increased dramatically: 35% and 18%, respectively for NIPSOM and GPL. As a result, using NIPSOM, one-third of the unsafe tube wells were colored green, i.e., safe. The rate of false negatives reported at sensitivities between 50 and 100 mg/L was 57% and 68% for NIPSOM and GPL, respectively. If results of both test methods were combined, the false-negative rate was even higher, meaning that up to two-thirds of the wells painted red were actually safe. Above a method sensitivity of 100 mg/L, the percentage of false-negative

Onsite Testing for Arsenic: Field Test Kits

71

results remained at 26% for NIPSOM and 17% for GPL. This assessment shows that 45% of mislabeled wells are not explained by variability of arsenic concentration or lack of QA/QC; rather, the errors are intrinsically linked to the methodologies used. This study clearly demonstrated the need for a stringent testing procedure to generate reliable data; such a procedure should make costly retesting unnecessary. It has been shown that these test kits generate a hazardous amount of arsine during testing; the generated levels exceed the threshold limit value in occupational air by a factor of 30 (Hussam et al. 1999). The poor performance of the tests created a strong incentive for field test kit developers to improve performance by several low-cost manipulations. The mixing chamber was optimized, and errors of visual inspection were eliminated by using spectrophotometers (Kinniburgh and Kosmus 2002). Using the new generation of field kits (Hach EZ arsenic kit, #2822800) appeared to improve the accuracy of the field methods. van Geen et al. (2005) reported that 88% of 800 tested wells were assigned correctly using the 50 mg/L drinking water standard in one district of Bangladesh. A previous BAMWSP study classified more than 66% of wells incorrectly; wells were then retested using an altered protocol (increasing the reaction time from 20 to 40 min), which reduced incorrect classifications to 34%. These results show improved data are generated from field kits when wells are classified using the local standard (50 mg/L). However, if the WHO guideline of 10 mg/L is considered, new challenges arise; some new-generation arsenic field kits are able to detect 1/10 of the WHO guideline, but the time required to perform testing is unacceptable. Furthermore, no arsenic field kit has been shown to be capable of determining arsenic species, including the water-soluble organoarsenicals [DMA(V) and MA(V)] in groundwater and wastewater. Commercial electrochemical sensors have been developed with adequate sensitivity for field use (US EPA 2006). However, these sensors have matrix-effect problems and are not user friendly, although testing of these sensors under rugged conditions has, nevertheless, been encouraged. Research efforts are needed that explore new avenues for detection of arsenic. Such work should include arsenic species recognition and address matrix interferences.

Prospects for Future Technological Developments Only a few reports exist on the development of new techniques. One explores microlever technology for arsenic, which has been shown to work for phosphate. An alternative approach is to use molecular recognition coupled to an enzymelinked immunosorbent assay (ELISA) assay. However, these technologies are in the development phase and are not anticipated to provide commercial field kits within the next 5 yr.

72

C

J. Feldmann

Analytical Methods for Sludge and Plants

Laboratory-Based Methods for Arsenic in Sludge and Biota X-ray fluorescence techniques have been developed for routine use in analyzing arsenic in soil, sediments, and other solids. Plants are usually easy to digest, which makes solution-based methods more suitable. Such digests can easily be determined by the laboratory-based methods previously discussed.

Arsenic Field Kits for Solids and Biota The degree of success achieved in remediating drinking water can be measured by analyzing the arsenic content of plants used for phytoremediation, or by measuring arsenic content in sludge/sediment, if arsenic was remediated by precipitation from the water phase. Few field kits are designed to do such testing, although kits used for water may be used to assess the arsenic content in solid samples. Solids must undergo acidic dissolution before their arsenic content can be tested. One caution is that arsenic may be strongly bound to solid matrices (as occurs with organoarsenicals) and may not necessarily form volatile species. The only reliable technology on the market for arsenic analyses in solids is XRF. Samples are irradiated with x-rays or gamma rays, which are generated at characteristic wavelengths by a photoelectrical effect. For arsenic, a sealed radioactive source (109Cd) is often used as the generator. Arsenic is identified using an energy-dispersive detection system. Portable XRF spectrometers are commercially available.

ETV Program Assessment While the previously discussed ETV program (US EPA 2006) was in progress, a number of XRF units were tested in 1997. Each instrument was portable, weighed less than 10 kg, and was able to measure 100 mg/kg in soil with a drift of 15%. There are limited reports available about the reliability of these instruments. Results of an onsite analysis of abandoned buildings in England was compared with laboratory results; the arsenic concentration in one solid sample (60 mg/kg) was positively identified (Potts et al. 2002). However, there is a paucity of systematic studies comparing results of laboratory-based methods with field methods. Sensitivity could be increased by onsite digestion of solid samples and preconcentration of the arsenic species on a resin, which could then be determined directly. Another study showed that determining speciation may be possible when species fractionation is combined with XRF detection (Le et al. 2000).

Onsite Testing for Arsenic: Field Test Kits

D

73

Analytical Methods for Volatile Arsenic in Gas Samples

Analysis of arsenic in gas is largely an unstudied area. However, it may be an important one, because volatilization of arsenic occurs from water, sludge, and contaminated soil. There is no established laboratory method available for analysis of arsenic in gas, although the Occupational Safety and Health Administration in the U.S. recommends a method for arsine in air, in which arsine is trapped on coconut shell charcoal and subsequently determined by using atomic absorption spectrometry (NIOSH 1994). Arsenic can also be trapped/absorbed in wash solutions, which are then analyzed by spectrometric methods mentioned previously. For example, recently it has been demonstrated that organoarsenicals such as trimethylarsine can be absorbed onto silica tubes impregnated with silver nitrate (Krupp et al. 2007). Two commercial field test methods exist for volatile arsine analysis in air samples and are normally used in occupational safety settings (Dra¨ger Safety 2006; Gastec Corporation 2007). When using Dra¨ger tubes (Nr. 0.05/a), gas is pumped through a column filled with a goldimpregnated support material. The AsH3 reduces the Au3+ to colloidal Au, which is visually detected within 6 min as it changes from white to grey-violet. An atmospheric concentration of 0.05 to 3 ppm (v/v) of AsH3 can be measured. Unfortunately, phosphines and stibines interfere with this method. Using Gastec tubes (Nr. 19LA), arsine reacts with mercuric dichloride to give an arsenic mercury chloride complex (As(HgCl)3) and hydrogen chloride. The latter is than neutralized by a basic compound, which gives a color change from yellow to red. The MDL is around 0.02 ppm (v/v) in ambient air.

Summary The performance of existing field test kits for arsenic has generally been unsatisfactory. Reports of false-negative and false-positive results exceeding 30% are not unusual, although more recent techniques appear to be more reliable. However, studies using these recent techniques had only to meet the local water standard of 50 mg/L. If the new WHO guideline (10 mg/L) is adopted as a decision-making criterion, the sensitivity of most arsenic testing kits is not sufficient, particularly in the hands of nontrained operators. New developments with sophisticated electrochemical sensors may deliver the needed sensitivity but suffer from matrix effects, even with trained operators. A failing of all available commercial methods is that they do not determine organoarsenicals, despite the fact that, in some cases, organic species may be the predominant ones present.

References Abbgy A, Kelly T, Lawrie C, Riggs K (2002) NanoBand Explorer Portable water Analyzer. ETV Report. http://www.epa.gov/etv/pdfs/vrvs/01_vr_nano-band.pdf. BRAC: Combating a Deadly Menace (2000) Early Experiments with a Community-based Arsenic Mitigation Project, Bangladesh. Research Monograph Ser. No. 16, Dhaka, Bangladesh.

74

J. Feldmann

Brickson BE (2003) Field kits fail to provide accurate measure of arsenic in groundwater. Environ Sci Technol 37:35A–39A. Cavicchioli A, Scalea La MA, Gutz IGR (2004) Analysis and speciation of traces of arsenic in environmental food and industrial samples by voltammetry: a review. Electrophoresis 16:697–711. CCME (Canadian Council of Ministers of the Environment) (2003) Canadian Environmental Quality Guidelines. http://www.ccme.ca/publications/pubs_updates.html. Chowdhury AQ (2002) Local Government Division, Ministry of Local Government, Dhaka, Bangladesh, Jan. 14–16, 2002. Deshpande LS, Pande SP (2005) Development of arsenic testing field kit: a tool for rapid on-site screening of arsenic contaminated water sources. Environ Monit Assess 101:93–101. Dhar RK Zheng Y, Rubenstone J, Geen van A (2004) A rapid colorimetric method for measuring arsenic concentrations in groundwater. Anal Chim Acta 526:203–209. DPHE-UNICEF (2001) Monthly Progress Report. Dhaka, Bangladesh, Dec. 2001. Dra¨ger Safety (2006) In: AFC International, Inc. Your Gas Detection Instrumentation http://www. afcintl.com/pdf/draeger/CH25001.pdf. ESCAP–UNICEF–WHO (2001) Economic and Social Commission for Asia and the Pacific, Geology and Health: Solving the Arsenic Crisis in the Asia Pacific Region: ESCAP– UNICEF–WHO Expert Group Meeting, Bangkok, May 2–4, 2001. Fact Sheet 12 on Arsenic (2002) A Disaster. Forum Publishing, Dhaka, Bangladesh. Gastec Corporations (2007) http://www.zefon.com/analytical/ download/19la.pdf. Gutzeit H (1891) Pharm Zeitung 36:748–756. Hussam A, Alauddin M, Khan AH, Rasul SB, Munir KM (1999) Evaluation of arsine generation in arsenic field kit. Environ Sci Technol 33(20):3686–3688. Khandaker NR (2004) Limited accuracy of arsenic field test kit. Environ Sci Technol 38:479A. Kinniburgh DG Kosmus W (2002) Arsenic contamination in groundwater: some analytical considerations. Talanta 58:165–180. Krupp EM, Johnson C, Rechsteiner C, Moir M, Leong D, Feldmann J (2007) Investigation into the determination of trimethylarsine in natural gas and its partitioning into gas and condensate phases using (cryotrapping)/gas chromatography coupled to inductively-coupled plasma mass spectrometry and liquid/solid sorption techniques. Spectrochim Acta B 62:970–977. Jalil MA, Feroze AM (2003) Arsenic detection and measurements by field test kits. In: Feroz AM (ed) ITN-Bangladesh, Center for Water Supply and Waste Management, Dhaka, Bangladesh, p. 442. Le XC, Yalcin S, Ma MS (2000) Speciation of submicrogram per liter levels of arsenic in water: on‐site species separation integrated with sample collection. Environ Sci Technol 34:2342–2347. Melamed D (2005) Monitoring arsenic in the environment: a review of science and technologies with the potential for field measurements Anal Chim Acta 532:1–13. MTI Diagnostic (2005) Evaluation of the MTI OVA 5000 for continuous arsenic monitoring at the Vineland Chemical Company Superfund site. http://www.mtidiagnstics.com/case_studies. html. National Food Authority (1993) Australian Food Standard Code. Australian Government Publication Service, Canberra. NGO Forum for Drinking Water Supply and Sanitation (2002) Dhaka, Bangladesh, Jan. 2002. NIOSH (National Institute for Occupational Safety and Health (1994) Manual of Analytical Methods (NMAM1), Schlecht PC, O’Connor PF (eds) 4th Ed. DHHS (NIOSH) Publication 94-113, August, 1994. Pande SP, Deshpande LS, Kaul SN (2001) Laboratory and field assessment of arsenic testing field kits in Bangladesh and West Bengal, India. Environ Monit Assess 68:1–18. Potts PJ, Ramsey MH, Carlisle J (2002) Portable X-ray fluorescence in the characterisation of arsenic contamination associated with industrial buildings at a heritage arsenic works site near Redruth, Cornwall, UK. J Environ Monit 4:1017–1024.

Onsite Testing for Arsenic: Field Test Kits

75

Raham MM, Fujinaga K, Seike Y, Okumura M (2004) A simple in situ visual and tristimulus colorimetric method for the determination of trace arsenic in environmental water after its collection on a mercury(II)-impregnated paper. Anal Sci 20(1):165–170. Rahman MM, Mukherjee D, Sengupta MK, Chowdhury UK, Lodh D, Chanda CR, Roy S, Selim M, Quamruzzaman Q, Milton AH, Shahidullah SM, Rahman MT, Chakraborti D (2002) Effectiveness and reliability of arsenic field testing kits: are the million dollar screening projects effective or not? Environ Sci Technol 6(24):5385–5394. Rowe JJ, Fournier RO, Morey GW (1973) Chemical analysis of thermal waters in Yellowstone National Park, Wyoming, 1960–1965. Geol Surv Bull (US) 1973:1303–1334. Spear JM, Zhou Y, Cole CA, Xie YFF (2006) Evaluation of arsenic test kits for drinking water analysis. J Am Water Works Assoc 98(12):97–105. USDA (2006) Foreign Agricultural Service Global Agriculture Information Network Report CH6064. China Peoples Republic: of FAIRS Product Specific Maximum Levels of Contaminants in Foods. US EPA (2006) ETV Joint Verification Statement (US EPA, Battelle). Arsenic Test Kits. http://www.epa.gov/etv/verifications/vcenter1-21.html. van Geen A, Cheng Z, Seddique AA, Hoque MA, Gelman A, Graziano JH, Ahsan H, Parvez F, Ahmed KM (2005) Reliability of a commercial kit to test groundwater for arsenic in Bangladesh. Environ Sci Technol 39:299–303. Visser WJF (1993) Contaminated land policies in some industrialized countries.TCB R02, Technische Commissie Bodembescherrning, Den Haag. WHO (1993) Arsenic in Drinking Water. Committee on Toxicology, Board on Environmental Studies and Toxicology, Commission on Life Sciences. National Academy Press, Washington, DC.

Technology for Remediation and Disposal of Arsenic Pornsawan Visoottiviseth(*) and Feroze Ahmed

I II

III

IV

V

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 A Passive Sedimentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 B In Situ Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 C Chemical Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 D Solar Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Coagulation and Filtration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 A Bucket Treatment Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 B The Star Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 C Fill and Draw Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 D Arsenic Removal Unit Attached to Tube Wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 E Iron-Arsenic (Fe-As) Removal Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 F Lime Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Sorptive Filtration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 A Activated Alumina . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 B Granular Ferric Hydroxide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 C Read-F Arsenic Removal Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 D Iron-Coated Sand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 E Shapla Arsenic Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 F Sono Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 G SAFI Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 H Activated Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 I Indigenous Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 J Cartridge Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Ion Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

P. Visoottiviseth Department of Biology, Faculty of Science, Mahidol University, Rama VI Road, Bangkok 10400, Thailand, e-mail: [email protected]. F. Ahmed Department of Civil Engineering, Bangladesh University of Engineering and Technology, Dhaka 1000, Bangladesh, e‐mail: [email protected]

D.M. Whitacre (ed.), Reviews of Environmental Contamination Volume 197. doi: 10.1007/978-0-387-79284-2_4, # Springer Science þ Business Media, LLC 2008

77

78

P. Visoottiviseth, F. Ahmed

VI

Membrane Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 A Techno-Food Water Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 B MRT-1000 and Reid System, Ltd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 C Low-Pressure Nanofiltration and Reverse Osmosis . . . . . . . . . . . . . . . . . . . . . . . . . 107 D Mobile Reverse Osmosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 E Combined Sand and Nanofiltration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 VII Bioremediation by Algae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 VIII Phytoremediation of Arsenic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 IX Comparison of Arsenic Removal Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 X Conventional Filtration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 XI Alternative Water Supply Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 A Deep Tube Well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 B Dug/Ring Well . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 C Surface Water Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 D Piped Water Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 XII Operational Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 A Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 B Technology Verification and Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 XIII Disposal of Generated Arsenic Waste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

I Introduction Arsenic contamination of water is a major public health problem in many countries worldwide. Symptoms of arsenic exposure have no known effective treatment, but drinking arsenic-free water can reduce risk to affected populations and alleviate symptoms of arsenic toxicity. In areas where the drinking water supply contains unsafe levels of arsenic, two main options have been identified: either find a safe source (mitigation) and/or remove arsenic from the contaminated source (remediation). Substantial effort has been invested in developing techniques for removing arsenic from water. Some of these techniques have been field implemented, while others are only performed in the laboratory. For any effective technology to be appropriate for use in affected areas of developing countries, it should ideally be simple, low cost, versatile, transferable, and should use local resources. Most importantly, such technologies must be accessible to local communities and especially to women. All over the world, women collect and carry water for their families, use water for cooking and cleaning and for growing food. Therefore, women should be at the forefront as users of arsenic treatment technologies. Das et al. (2004) concluded that, in villages of India and Bangladesh, even a highly successful technology may not succeed in rural areas unless it fits well with the rural circumstances and is well accepted by the local population. Several methods are available for removal of arsenic from water in large conventional treatment plants. The most commonly used processes include oxidation and sedimentation, coagulation and filtration, lime treatment, adsorption onto sorptive media, ion exchange, and membrane filtration (Cheng et al. 1994; Hering et al. 1996,

Technology for Remediation and Disposal of Arsenic

79

1997; Kartinen and Martin 1995; Shen 1973; Joshi and Chaudhuri 1996). A review of these well-established arsenic removal technologies is presented by Sorg and Logsdon (1978). Jakel (1994) has documented several advances in arsenic removal technologies. In view of the lowering of the drinking water standards by the United States Environmental Protection Agency (US EPA), a review of arsenic removal technologies was made to consider the economic factors involved in implementing the lower standards for arsenic (Chen et al. 1999). Many arsenic removal technologies have been discussed, in detail, in the American Water Works Association (AWWA) reference book (Pontius 1990). A comprehensive review of low-cost, well water treatment technologies for arsenic removal, with the list of companies and organizations fostering them, has been compiled by Murcott (2000). Arsenic removal technologies have also been reviewed by Ahmed et al. (2000, 2001), the World Bank (2005), and Ahmed (2003). Other low-cost technologies that have been investigated are presented in Table 1. Some of these technologies can be reduced in scale and can be conveniently applied at the household and community level for removal of arsenic from contaminated water. During the last 2–3 yr, many small-scale arsenic removal technologies have been developed, field tested, and used in action research programs in some Asian countries. The following review of these technologies briefly updates what is known of the technological development in arsenic removal, and discusses the current status of the problem, as well as prospects and limitations of different treatment processes; this review also delineates areas needing further improvement for successful implementation and adaptation of technologies to rural conditions.

II

Oxidation

Arsenic is present in groundwater as As(III) and As(V), in different proportions. Most treatment methods are effective in removing arsenic in pentavalent form, and hence include an oxidation step as a pretreatment to convert arsenite to arsenate. Arsenite can be oxidized by oxygen, ozone, free chlorine, hypochlorite, permanganate, hydrogen peroxide, and Fulton’s reagent; in developing countries oxygen, hypochloride, and permanganate are commonly used for oxidation. The following reactions describe oxidation by oxygen, hypochloride, and permanganate: H3 AsO3 þ 1=2O2 ! H2 AsO4  þ 2Hþ

ð1Þ

H3 AsO3 þ HClO ! HAsO4  þ Cl þ 3Hþ

ð2Þ

½H3 AsO3 þ 2KMnO4 ! 3HAsO4  þ 2MnO2 þ þ 2Kþ þ 4Hþ þ H2 O

ð3Þ

The oxidation processes convert predominantly noncharged arsenite to charged arsenate, which can be easily removed from water. Aeration is the simplest means of oxidation, and many treatment processes depend on oxidation by air. However, air oxidation of arsenic is a very slow process, often requiring weeks (Pierce and Moore 1982). Air oxidation of arsenite

80

P. Visoottiviseth, F. Ahmed

Table 1 Performance and cost of some specific technologies for removing arsenic from contaminated water Technology Performance Cost (% removed) PRECIPITATION (Taking advantage of naturally70%–80% Final As conc. None occurring Fe and/or Mn 20–36 ppb precipitation) (University of Kalyani, West Bengal) SEDIMENTATION Used with precipitation or coagulation COAGULATION Iron salts (Murcott 1999) >90% $0.06/yr (40 L/d at 15 mg/L dose) 90% Final As conc. 30 ppb Low Alum (Ahmed et al. 2000) BUETa ‘‘Bucket’’ or ‘‘Tea-Bag’’ 80%–99% Final As conc. $0.05/packet treats 10 L WHO-SEA method 50–70 ppb (cost is less if mass produced) Iron filings (zero valance) >94%–99% $0.22/yr (Ramaswami et al. 2001) University of ColoradoDenver Other coagulants: tablets, lime, >78% Low natural or synthetic polymers CONVENTIONAL FILTRATION Cloth, sand, charcoal, other 20%–75% Low native material media – coconut husks, peanut shells, water hyacinth, rubber leaves, etc. ADSORPTION Activated alumina metal oxide 90%–96% $0.02–$0.03/20 L (Project Earth Inc.) Final As conc. 10–25 ppb Capital cost <$100 per unit (cheaper if produced locally) Iron filings and sand 90% $0.06/L (AsRT – Univ. of Connecticut) Final As conc. <27 ppb for Capital cost = $900 for >1000 pore volume of pilot unit; full-scale eluent unit from about $10. (two columns treating 3.8 L/min) Laterite 50%–90% None or low Low Other adsorbents: Bijoypur clay, hematite (Fe2O3), fly ash

Technology for Remediation and Disposal of Arsenic Table 1 (continued) Technology

Performance (% removed) Final As conc. <2 ppb

Ion (anion) exchange (Clifford et al. 1998) Univ. of Houston Activated alumina Final As conc. <50 ppb (West Bengal Engineering College / Water for People) Final As conc. <50 ppb Ferric hydroxide [Fe(OH)3] or ferric hydroxide-coated newspaper pulp (Khair et al. 1999) Univ. of Dhaka OXIDATION Aerationa 25% Photochemical oxidationa SOLAR DISTILLATION Final As conc. 0 ppb Solar Still (Young Associates)a MEMBRANE Fe(III) coagulation þ Final As conc. <2 ppb microfiltration depended on pH, Fe (Clifford et al. 1998) Univ. dose of Houston Reverse osmosis (RO) 86% Membrane filtration and RO 90%–93% (100–150 ppb) (Wanichapichart 2005) Prince of Songkla Univ., Thailand Electrodialysis 80% Nanofiltration BIOREMEDIATION Absorption by immobilized 85%–90% green alga, Chlorella vulgaris (Visoottiviseth and Lauengsuchonkul 2004) Mahidol Univ., Thailand. PHYTOREMEDIATION Wetland treatment 80%–90% a Bangladesh University of Engineering and Technology. Source: Modified from Murcott (2000).

81

Cost Expensive

$1,400/unit Each unit serves 200–300 households Low

Potentially very low

$0.02 per person per day Expensive

Expensive Moderate

Expensive Expensive Very low

Very low

can be catalyzed by bacteria, strong acidic or alkaline solutions, copper, powdered activated carbon, and high temperature (Edwards 1994). Chemicals such as chlorine and permanganate can rapidly oxidize arsenite to arsenate under a wide range of conditions. Hypochloride is readily available in rural areas, but the potency (available chlorine) of the hypochloride declines under poor storage conditions; chlorine escapes from hypochloride when it comes in contact

82

P. Visoottiviseth, F. Ahmed

with air. A residual concentration of 0.2 mg/L free available chlorine in water is required for oxidation of arsenite. Potassium permanganate is also readily available in developing countries, is more stable than bleaching powder, and has a long shelf life. Potassium permanganate effectively oxidizes arsenite and ferrous compounds. Ozone and hydrogen peroxide are very effective oxidants but their use in developing countries is limited. Filtration of water through a bed containing solid manganese oxides can rapidly oxidize arsenic, without releasing excessive manganese in the filtered water.

A

Passive Sedimentation

Passive sedimentation has received considerable attention because of the habit of rural peoples of drinking stored water from pitchers. Routine activities associated with collecting and storing water in homes may reduce arsenic concentrations. Experiments conducted in Bangladesh showed passive sedimentation to variably reduce arsenic concentrations in drinking water. Arsenic reduction by sedimentation appears to be dependent on water quality, particularly on the presence of precipitating iron. Ahmed et al. (2000) reported >50% reduction of arsenic concentrations by sedimentation in tubewell water containing 380–480 mg/L alkalinity as CaCO3. However, this process was unreliable in reducing arsenic to levels less than the World Health Organization (WHO) maximum contaminant level (MCL) of 10 mg/L. High alkalinity and the presence of iron in tubewell water increase arsenic removal during storage. Most studies showed reductions up to 25% of the initial arsenic concentrations found in groundwater from these wells. However, passive sedimentation failed to reduce arsenic to the desired level (50 mg/L) in Bangladesh when arsenic content of tubewell water was high (BAMWSP, DFID, and Water Aid Bangladesh 2001).

B

In Situ Oxidation

In situ oxidation of arsenic and iron in aquifers has been attempted in a Department of Public Health Engineering–Danish International Development Agency (DPHEDanida)-funded Arsenic Mitigation Pilot Project in Bangladesh. The aerated tubewell water was stored in a 500-L capacity feed water tank (Fig. 1) and released back into the aquifer through the tube well. The dissolved oxygen in the water oxidizes arsenite to the less mobile arsenate and ferrous to ferric iron; this results in a reduction of the arsenic content in the tubewell water. The oxidation of arsenite to arsenate reaction is shown in Equation 1 above. Subsequent reactions of arsenate and ferric hydroxide are shown in Equations 7 to 9, below. Experimental results confirm that such in situ oxidation reduces the concentration of arsenic in tubewell water to about half of its original value. This effect results from both underground precipitation and adsorption on ferric iron. The study revealed that the in situ oxidation method is effective in reducing arsenic content to meet the Bangladesh standard of 0.05 mg/L when the

Technology for Remediation and Disposal of Arsenic

83

1400 mm 1400 mm 1100 mm

X

X

Tray (75 mm slope) 600 mm

GI sheet Tray

1100 mm

Inlet to TW

35 mm GI pipe

150mm Wash out 75 mm

Top View

Fig. 1 Feed water tank for in situ arsenic removal by oxidation in the aquifer

concentration of arsenic in tubewell water is less than 0.10 mg/L. At higher concentrations, such water treatment may not achieve a final concentration sufficient to meet the required standard. Treatment efficiency improves if the dissolved oxygen content of recharge water, or quantity of water recharged in the aquifer, is increased. Usually, the excess water pumped from the tube well during the day is stored in the feed water tank. It is then recharged at night back into the tube well, where it resides undisturbed overnight. The method is simple, chemical free, and may be well accepted by people using it; unfortunately, it is not highly effective in removing arsenic.

C

Chemical Oxidation

In situ chemical oxidation of arsenite to arsenate and ferrous iron into ferric iron by chemical oxidation in the aquifer, and subsequent coprecipitation and adsorption, can immobilize subsurface arsenic. Matthess (1981) injected 29 t potassium permanganate directly into 17 contaminated wells to reduce contamination in an aquifer containing high concentrations of arsenite and ferrous iron. The water of this aquifer also had a low pH value. Arsenate was precipitated with ferric oxides, and arsenic content in water was reduced from 13,600 to 60 mg/L. However, there may be problems and uncertainties in effectiveness of in situ remediation of arsenic by chemical oxidation. The introduction of reactive chemicals and microbes in the aquifer may have unforeseen effects on subsurface ecology and groundwater chemistry. Moreover, in a dynamic system of continuous recharge, water movement and withdrawal in a treated aquifer may induce their own effects. Long-term effectiveness of such remediation is, therefore, not assured.

D

Solar Oxidation

Solar oxidation and removal of arsenic (SORAS) is a simple method for oxidizing arsenic (III) by irradiating drinking water in transparent bottles with sunlight to

84

P. Visoottiviseth, F. Ahmed

reduce dissolved arsenic (Wegelin et al. 2000). Ultraviolet radiation catalyzes the oxidation of arsenite in the presence of oxygen (Young 1996). Experiments in Bangladesh show that the process can reduce the arsenic content of water by about one-third.

III

Coagulation and Filtration

Coagulation and flocculation removes arsenic from solution by three mechanisms (Edwards 1994): (1) precipitation—the formation of insoluble compounds, (2) coprecipitation—the incorporation of soluble arsenic species into a growing metal hydroxide phase, and (3) adsorption—the electrostatic binding of soluble arsenic to external surfaces of the insoluble metal hydroxide. Precipitation, coprecipitation, and adsorption by coagulation with metal salts and lime, followed by filtration, is the method most frequently employed for removing arsenic from water. This method can effectively remove arsenic and other suspended and dissolved solids (e.g., iron, manganese, phosphate, fluoride, and microorganisms) from water. Moreover, it achieves additional health and aesthetic benefits because it may improve turbidity, color, and odor, resulting in significant water quality improvement. Chemical coagulation is effective in removing arsenic from drinking water, but the dose requirement is several times higher than that required for conventional water treatment, especially for destabilization and removal of colloidal particles. Alum (Al2(SO4)318H2O), ferric chloride (FeCl3), and ferric sulfate (Fe2(SO4)37H2O) are common coagulants used for removing arsenic from water. Ferric salts are more effective in removing arsenic than alum on a weight basis and are effective over a wider range of pH. In both cases, pentavalent arsenic is more effectively removed than is the trivalent form (Ahmed and Rahaman 2000). In the coagulation-flocculation process, aluminum sulfate, ferric chloride, or ferric sulfate is added to, and dissolved in, water, with stirring for 1 to a few minutes. Aluminum or ferric hydroxide micro-flocs are rapidly formed. The water solution is then gently stirred for several minutes to allow agglomeration of micro-flocs into larger ones that settle more readily. During this process, microparticles and negatively charged ions become attached to the flocs by electrostatic forces. Arsenic is also adsorbed onto coagulated flocs. Because trivalent arsenic is nonionic, it is not significantly removed during this treatment process. Therefore, As(III) must be oxidized to As (V) for removal by this process, which can be achieved by the addition of bleaching powder (chlorine) or potassium permanganate (Equations 2 and 3, above). The chemical equations describing alum coagulation are as follows: Alum dissolution: Al2 ðSO4 Þ3  18H2 O ! 2Al3þ þ 3SO4 2þ þ 18H2 O

ð4Þ

Aluminium precipitation (acidic): 2Al3þ þ 6H2 O ! 2AlðOHÞ3 þ 6Hþ

ð5Þ

Technology for Remediation and Disposal of Arsenic

85

Coprecipitation (nonstoichiometric, nondefined product): H2 AsO 4 þ AlðOHÞ3 ! Al  AsðcomplexÞ þ other products

ð6Þ

Arsenic precipitated as Al(AsO4) or adsorbed on aluminum hydroxide flocs as the Al–As complex is partially removed by sedimentation; filtration may be required for complete removal of flocs. Similar reactions take place for ferric chloride and ferric sulfate, resulting in the formation of a Fe–As complex as an end product; this complex is then removed by sedimentation and filtration. The reactions for arsenate with hydrous iron oxide are shown below, where [FeOHo] represents an oxide surface site (Mok and Wai 1994; Hering et al.1996). FeðOHÞ3 ðsÞ þ H3 AsO4 ! FeAsO4  2H2 O þ H2 O

ð7Þ

FeOHo þ AsO43 þ 3Hþ ! FeH2 AsO4 þ H2 O

ð8Þ

FeOHo þ AsO43 þ 2Hþ ! FeHAsO4 þ H2 O

ð9Þ

Effective immobilization of arsenic by hydrous iron oxide (Equations 7 to 9) requires oxidation of arsenic species to As(V). Freshly formed hydrous ferric oxide (HFO) and hydrous aluminum oxide (HAO) have maximum arsenic adsorption capacities of approximately 0.1 M As(V)/M Fe or Al (i.e., 46 mg As/g of ferric chloride or 23 mg As/g alum). When the sorbents are formed in situ, adsorption capacities are much higher [(~0.5–0.6 M As(V)/M Fe or Al]. The difference reflects effects of coprecipitation (preformed hydroxides only remove arsenic through adsorption), while in situ formation also leads to coprecipitation (Edwards 1994). Arsenic removal by coagulation is mainly controlled by pH and coagulant dose. Adsorption is theoretically favored at a pH below the sorbent point of zero charge, because positively charged surfaces of the sorbents attract arsenate anions. Laboratory tests have shown arsenate adsorption to be optimal on HFO below pH 8 and on HAO below pH 7 (Sorg and Logsdon 1978; Edwards 1994; Hering et al. 1996). In alum coagulation, removal is most effective at a pH 7.2–7.5. Efficient iron removal by coagulation is achieved at a pH range of 6.0 to 8.5 (Ahmed and Rahaman 2000). Cations and anions are very important in arsenic removal by coagulation. Anions compete with arsenic for sorptive sites and lower the rates of removal. Manning and Goldberg (1996) expressed the comparative theoretical affinity at neutral pH for anion sorption on metal oxides as: PO4 > SeO3 > AsO4 > AsO3 >> SiO4 > SO2 > F > BðOHÞ3 PO4 has the highest affinity for metal oxides and is most likely to compete with arsenic for adsorption sites. Dissolved silicates can interfere with removal of both arsenite and arsenate. The presence of more than one anion may have a synergistic

86

P. Visoottiviseth, F. Ahmed

effect on arsenic removal. Addition of either silicate or phosphate has some effect on arsenic removal, but presence of both can reduce arsenate removal by 39% and arsenite removal by 69% (Meng et al. 2000). Results of arsenic removal studies in Bangladesh (Meng and Korfiatis 2001) concluded that elevated levels of phosphate and silicate, in well water, dramatically decreased adsorption of arsenic by ferric hydroxides. A Fe/As mass ratio greater than 40 was required to reduce arsenic concentration to less than 50 mg/L. Elevated sulfate and carbonate levels slightly reduce arsenite removal but have little effect on arsenate removal. Natural organic matter can also reduce arsenic removal efficiency.

A

Bucket Treatment Unit

The Bucket Treatment Unit (BTU), developed from the DPHE-Danida Project, uses the principles of coagulation, coprecipitation, and adsorption. This unit consists of two 20-L buckets placed one above the other. Chemicals are manually mixed with arsenic contaminated water in the upper (red-colored) bucket by vigorous stirring for 30–60 sec, and are then flocculated by gentle stirring for about 90 sec. After mixing, the water is allowed to settle for 1–2 hr. Thereafter, water from the top bucket is drained into the lower (green-colored) bucket via a plastic pipe, then through a sand filter installed in the lower bucket. Flow is initiated by opening a valve fitted above settled sludge in the bottom of the red bucket; thus, inflow of sludge in the sand filter is avoided. The DPHE-Danida bucket treatment unit is shown in Fig. 2a. The BTU units utilize chemical doses of 200 mg/L aluminum sulfate and 2 mg/L potassium permanganate, supplied in crushed powder form for water treatment. Their performance in removing arsenic under field and laboratory conditions is reported to be good (Sarkar et al. 2000; Kohnhorst and Paul 2000). The performance of the BTU was studied (BAMWSP, DFID, and Water Aid Bangladesh 2001) with mixed results. Under rural operating conditions in Bangladesh, units often failed to remove arsenic to the target level (0.05 mg/L). Poor mixing and variable water quality (particularly pH) appeared to cause the poor performance. Bangladesh University of Engineering and Technology (BUET) modified and improved performance of the BTU by using 100 mg/L ferric chloride and 1.4 mg/L potassium permanganate. Water so treated gave arsenic values below 20 mg/L, and values never exceeded 37 mg/L, compared with arsenic concentrations in the tubewell water of 375–640 mg/L. The BUET-modified BTU is depicted in Fig. 2b. Further field testing of the modified units is underway in a rural area of the Comilla district of Bangladesh. The modified BTUs are very effective in removing iron, manganese, phosphate, and silica. Initially, fecal coliform bacteria were found in treated waste, probably derived from contact with contaminated human hands. This biocontamination was eliminated by adding bleaching powder to the chemical packet used in the BTU. The BTU is a promising technology for economic arsenic removal at the household level. It can be locally built using available materials and is effective if operated properly.

Technology for Remediation and Disposal of Arsenic

87

Top Bucket

Flexible Plastic Pipe

Top Bucket Flexible Plastic Pipe Filter

Bottom Bucket

Cloth Screen

Bottom Bucket

Sand Filter

PVC Slotted Screen

a

b

Fig. 2 Double-bucket household arsenic treatment unit. (a) BPHE-Danida* Unit; (b) BUET** Modified Unit. (* Department of Public Health Engineering-Danish International Development Agency; ** Bangladesh University of Engineering and Technology.)

The Mennonite Central Committee (MCC) in Bangladesh also experimented with a range of arsenic removal technologies (from tube water) including the DPHE-Danida BTU. MCC replaced permanganate with bleaching powder to achieve oxidation in the units they tested; alum was used for coagulation-sedimentation of arsenic. The DPHE-Danida bucket treatment units they tested were found to remove more than 90% of arsenic present in tubewell water.

B

The Star Filter

The Star Filter developed by the Stevens Institute, USA, also uses two buckets, one to mix chemicals (iron coagulant and hypochloride) supplied in packets and the other to separate flocs by sedimentation and filtration. The second bucket has an inner bucket with side slits Fig. 3) to help sedimentation and retain the sand bed. Visible large flocs are formed after chemical packets are added and stirred into the water. After treatment, clean water is collected after filtration through cloth to prevent entry of sand. The sand bed is quickly clogged by flocs and requires washing at least twice a week. An assessment showed that the technology was

88

P. Visoottiviseth, F. Ahmed Chemicals mixing stick

Transfer of chemical mixed water Main bucket Interior bucket Slits Outlet with cloth filter Filter sand

Plastic pipe to deliver treated water

Fig. 3 Star Filter developed by Stevens Institute Technology

effective in reducing arsenic levels to less than 0.05 mg/L for 80%–95% of the samples tested (BAMWSP, DFID and Water Aid Bangladesh 2001).

C

Fill and Draw Units

The fill and draw unit is a community-type treatment unit designed and installed during the DPHE-Danida Arsenic Mitigation Pilot Project. It comprises a 600-L (effective) capacity tank with a slightly tapered bottom for collection and withdrawal of settled sludge. The tank is fitted with a manually operated mixer that has flat-blade impellers. The tank is filled with arsenic-contaminated water, and the required quantity of oxidant and coagulant is then added. The water is then mixed for 30 sec by rotating the mixing device at the rate of 60 rpm and is then left overnight to allow sedimentation. The water takes some time to become completely still, which helps flocculation. The floc formation is caused by the hydraulic gradient of the rotating water in the tank. The settled water is then drawn through a pipe fitted a few inches above the bottom of the tank, is passed through a sand bed, and is finally collected through a tap (Fig. 4). The mixing and flocculation processes in this unit are controlled to better effect higher removal of arsenic.

D

Arsenic Removal Unit Attached to Tube Wells

The principles of arsenic removal by alum coagulation, sedimentation, and filtration have been employed in a compact unit for water treatment at the village level in West Bengal, India. The arsenic removal plant is attached to a tube well fitted with a hand pump (Fig. 5). This unit is effective in removing 90% of the original concentrations (300 mg/L) of arsenic from tubewell water. The treatment process involves addition of sodium hypochloride (Cl2) and dilute aluminum alum, followed by mixing, flocculation, sedimentation, and up-flow filtration.

Technology for Remediation and Disposal of Arsenic

89

Gear system Cover Impeller

Handle Filtration unit

Tank

Treated water

Sludge withdrawal pipe

Fig. 4 DPHE-Danida fill and draw arsenic removal unit

A - Mixing; B - Flocculation; C - Sedimentation; D - Filtration (Up-flow) A

B

D

B C

Fig. 5 Arsenic removal plant for use with tube wells (designed and constructed in India)

E

Iron-Arsenic (Fe-As) Removal Plants

The use of naturally occurring iron precipitates in Bangladesh groundwater is a promising method for removing arsenic. Tubewell water, used in 65% of the area of Bangladesh, contains iron in excess of 2 mg/L. In some areas, the concentration of dissolved iron is higher than 15 mg/L, which is unacceptable for domestic water supplies. Although there is no clear link between natural concentrations of iron and arsenic, they commonly coexist as contaminants of groundwater. Most tubewell water samples that satisfy the Bangladesh Drinking Water Standard for Iron (1 mg/L) also satisfy the standard for arsenic (50 mg/L). Only about half the samples with 1–5 mg/L iron satisfy the arsenic standard; 75% of the samples with iron content above 5 mg/L are unsafe because of high concentrations of arsenic. Iron precipitates [Fe(OH)3] formed by oxidation of dissolved iron [Fe(OH)2] present in groundwater have an affinity for adsorbing arsenic. Aeration and sedimentation of tubewell water rich in dissolved iron does, indeed, remove arsenic. In Bangladesh, iron removal plants (IRPs) constructed on the principles of aeration, sedimentation, and filtration successfully remove arsenic without using other che-

90

P. Visoottiviseth, F. Ahmed

micals. The conventional community-type IRPs work as arsenic removal plants (ARPs) as well. A typical plant of this sort is shown in Fig. 6. This plant was designed by BUET and first used in rural areas about 20 yr ago for iron removal from groundwater. The DPHE-Danida project also installed experimental up-flow Fe-As removal plants after the design provided by BUET. The treatment unit shown in Fig. 7 is

100 mm PVC pipe (slotted) 25 mm thick Slab Plug

A

200

Tubewell 125 D

Outlet

C

625 Pitcher (Kalshi)

E

75 A B C D E

SECTION X -X

Aeration Initial sedimentation Adsorption Filtration Final sedimentation 125

A

175

B

75

C X

550

D E

Platform

X

75

25

300

75

275

125

PLAN

Fig. 6 A typical iron-arsenic removal plant, designed by BUET, that also removes arsenic

Technology for Remediation and Disposal of Arsenic

91

Aeration Tray

Treated Water

Tubewell

Filter Media

Washout Valve

Fig. 7 Fe-As removal unit fitted with an upflow filter

attached to a tube well and is rather simple to operate. The As-Fe removal unit relies on oxidation, precipitation, adsorption, and filtration to remove arsenic. Such units were installed to remove arsenic from tubewell water having concentrations between 0.1 and 1.0 mg/L. The efficiency of the units depends on the arsenic and iron content of water. Results from unit operation show them to operate reliably at 50% efficiency (arsenic removal). Dahi and Liang (1998) suggest that As(III) is oxidized to As(V) in the IRPs, which augments arsenic removal in IRPs constructed in Noakhali. The relationship between removal of iron and arsenic when using IRPs is shown in Fig. 8. Results show that most IRPs can lower arsenic content of tubewell water by 50%–80% of original concentrations. The efficiency of these community-type plants can be improved by increasing contact time between arsenic species and iron flocs. Medium-scale Fe-As removal units (capacities of 2000–3000 m3/d) have been constructed in district towns using the aforementioned principle. Some production wells that supply water to urban areas are also contaminated with arsenic. The working principles of a medium-sized Fe-As unit are shown in Fig. 9. The main processes effecting treatment are aeration, sedimentation, and rapid sand filtration, with provision for addition of chemical if required. These units utilize the natural iron content of water and have rather low (40%–80%) efficiency. A downside of these plants is that the water requirement for washing the filter beds is very high. The experience of operating small- and medium-sized Fe-As removal plants in

92

P. Visoottiviseth, F. Ahmed 100 90

Arsenic Removal ,%

R2 = 0.69 80 70 60 50 40 30 20 20

30

40

50

60

70

80

90

100

Iron Removal , % Fig. 8 Correlation between Fe and As removal in treatment plants

Aeration

Overhead water tank Chlorination Filter bed

Inlet pipe

Pump

Pump Back washing

Water supply

Fig. 9 Fe-As removal plant suitable for use by small towns

Bangladesh suggests that arsenic removal by coprecipitation and adsorption on natural iron flocs has good potential, if arsenic content of water does not exceed 0.10 mg/L.

F

Lime Treatment

Water treatment by addition of quicklime, CaO, or hydrated lime, Ca(OH)2, also removes arsenic. Lime treatment is a process similar to that of coagulation with metal salts. The precipitated calcium hydroxide [Ca(OH)2] acts as a sorbing

Technology for Remediation and Disposal of Arsenic

93

flocculent for arsenic. An excess of lime will not dissolve but remains as a thickener and aid to coagulation. The excess lime, along with precipitates, must be removed by sedimentation and filtration. Arsenic removal by lime is usually between 40% and 70% effective. The highest removal is achieved at pHs between 10.6 and 11.4. McNeill and Edward (1997) studied arsenic removal by water softening and found that the main mechanism of arsenic removal was sorption onto magnesium hydroxide solids that form in situ. Trace levels of phosphate were found to slightly reduce arsenic removal below pH 12, whereas arsenic removal efficiency at lower pH is increased by addition of iron. The disadvantage of arsenic removal by lime is that it requires large quantities (800–1200 mg/L), which produce a large volume of sludge. Obviously, water to which lime has been added requires secondary treatment to properly adjust the pH level. Lime softening may primarily serve as a pretreatment before the use of alum or iron coagulation.

IV

Sorptive Filtration

Arsenic can be removed from water to very low levels by filtration through sorptive media. Activated alumina, activated carbon, iron- or manganese-coated sand, kaolinite clay, hydrated ferric oxide, activated bauxite, cerium oxide, titanium oxide, silicon oxide, and many other natural and synthetic substances have been used as sorptive media to remove arsenic from water. The efficiency of sorptive media depends on the use of oxidizing agents to aid arsenic sorption. Media eventually saturate with contaminants removed from water; the specific sorption affinity of the medium for components present determines sorptive life. Saturation is reached when active sites of the media are exhausted and the media cannot remove further impurities.

A

Activated Alumina

Activated alumina (Al2O3) has a good sorptive surface area, in the range of 200–300 m2/g, and is an effective medium for arsenic removal. When water passes through a packed column of activated alumina, all impurities, including arsenic, are adsorbed on the surfaces of the media grains. The column will eventually become saturated, first at the top and later toward its bottom. Arsenic removal by activated alumina is influenced by both the pH and the arsenic content of the water. Arsenic removal is optimum at pHs between 5.5 and 6.0, when the surface of the medium is positively charged; removal efficiency drops as the point of zero charge is approached. When the surface of the medium becomes negatively charged at pH 8.2, the removal capacities are only 2%–5% of the capacity at optimal pH (Clifford 1990). Regeneration of saturated alumina is performed by exposing the medium to 4% caustic soda (NaOH), either in a batch process or by column flowthrough. Either process produces arsenic-contaminated caustic wastewater. The residual caustic

94

P. Visoottiviseth, F. Ahmed

soda is then washed and the medium is neutralized with a 2% solution of sulfuric acid. During the process, about 5%–10% alumina is lost and the capacity of the regenerated medium is reduced by 30%–40%; activated alumina must be replaced after three or four regenerations. As with coagulation processes, prechlorination dramatically improves the column capacity. The activated alumina-based sorptive media used in Bangladesh include BUET Activated Alumina, Alcan Enhanced Activated Alumina, and media employed in the Apyron Arsenic Treatment Unit. Each of these three units/processes is reviewed below.

BUET Activated Alumina The BUET activated alumina arsenic removal unit (ARU) consists of subunits for oxidation-sedimentation, filtration, and activated alumina adsorption. Oxidation and sedimentation is performed in a 25-L plastic bowl. Approximately 1 mg/L potassium permanganate is added to water in the bowl to oxidize As(III) to As(V); the mixture is stirred vigorously with a wooden stick and then allowed to settle for about 1 hr. The settled water is filtered through a sand bed and is then passed through the activated alumina column. The unit is very effective in removing arsenic and iron from tubewell water. One practical problem with the ARU is that women have difficulty in raising water to the level required for gravity flow through the subunits. The problem has been addressed by design modification. The modified BUET activated alumina ARU is shown in Fig. 10.

Oxidation Sedimentation Unit

Sand Filtration Unit

Activated Alumina

Fig. 10 The BUET activatedalumina arsenic removal unit

Technology for Remediation and Disposal of Arsenic

95

Inlet (Water from Tubewell)

Tubewell

Outlet (Treated Water)

Fig. 11 An Alcan enhanced activated alumina unit

Alcan Enhanced Activated Alumina Unit In this process, water from a tube well is allowed to pass through an enhanced activated alumina bed and the treated water is collected as shown in Fig. 11. The unit has a simple and robust design. No chemicals are added during treatment, and the process relies entirely on the active surface of the media for adsorption of arsenic. Other ions present in water, such as iron and phosphate, may compete for active sites on alumina and thereby reduce the arsenic removal capacity of the unit. Iron present at elevated levels in shallow tubewell water will eventually accumulate in the activated alumina bed and interfere with water flow. The unit can produce more than 3,600 L arsenic-safe drinking water per day, enough for 100 families. Alcan’s enhanced activated alumina unit is designed for single use and therefore saturated media must be replaced after use. Environmentally safe disposal of spent activated alumina (~40 kg per treatment cycle) is required.

Apyron Arsenic Treatment Unit Apyron Technologies Inc. (ATI), USA, has developed an arsenic treatment unit (ATU) in which Aqua-Bind media is used to reduce arsenic in groundwater. The ATU consists of a cylindrical adsorber vessel containing Aqua-Bind media. This media consists of nonhazardous aluminum oxide (Al2O3) and manganese oxide (Mn2O3) that can selectively remove As(III) and As(V) from water. The column receives water under slight positive pressure from a manually operated lift pump (Fig. 12). Water flows downward through a two-chamber housing capable of capturing particulate iron and adsorbing arsenic. Discharge water exits into the designated container at a rate of approximately 15 L/min. Experimental units installed in India and Bangladesh are reported to consistently reduce arsenic in water to less than 10 mg/L. The proponents of ATI’s Aqua-Bind media (Senapati and Alam 2001) claim that it (1) removes both arsenic (III) and arsenic (V); (2) successfully treats arsenic levels from 25 to more than 4,000 ppb in the presence of

96

P. Visoottiviseth, F. Ahmed

Aqua-Bind Filter Media

Raising Hose Pipe

Deliver Hose Pipe

Lift Pump

Aqua-Bind Arsenic

Bucket Containing Treated Water

Fig. 12 An Apyron arsenic treatment unit

up to 15 ppm of iron; (3) reduces contact times (ideal for point-of-use systems), (4) operates over a wide pH range (6–8) and temperatures (0 –100 C), (5) is nonleachable, allowing safe disposal of spent media (as per the Toxic Characteristic Leaching Procedure test); (6) is NSF 61 certified for use in drinking water applications; (7) resists microbe growth; and (8) is highly selective for As, even with competing ions (sulfates, silica, Ca, etc.).

B

Granular Ferric Hydroxide

Granular ferric hydroxide (AdsorpAs) is a highly effective adsorbent used for removal of arsenate, arsenite, and phosphate from natural water and wastewater. AdsorpAs treatment capacity ranges from 40,000 to 60,000 BV (bed volumes), until adsorbed arsenic exceeds the permissible level of 0.01 mg/L. AdsorpAs has 0.2–2.0 mm grain size, 72%–77% porosity, 250–300 m2/dm3 specific surface, 1.22– 1.29 kg/dm3 bulk density, and 52%–57% active substances (Fe(OH)3 and b-FeOOH). It has an adsorption capacity of 45 g/kg for arsenic and 16 g/kg for phosphorus on a dry weight basis (Pal 2001). The granular ferric hydroxide reactors are fixed-bed adsorbers that operate as conventional filters. The units require iron removal by pretreatment to avoid clogging the adsorption bed. A typical granular ferric hydroxide-based arsenic removal unit is shown in Fig. 13. Water containing high dissolved iron and suspended matter is pretreated by aeration and filtration through a gravel/sand filter bed, and is then passed through AdsorpAs in the adsorption tower for removal of arsenic. M/S Pal Trockner (P) Ltd of India and Sidko Limited of Bangladesh installed several granular ferric hydroxide-based arsenic removal units in India and Bangladesh. Proponents claim AdsorpAs has very high arsenic removal capacity, 5- to 10 fold higher than activated alumina. The unit, therefore, produces less residual spent solids; typically the residual mass of spent AdsorpAs is 5–25 g/m3 of water treated. The spent granular ferric hydroxide is a nontoxic solid waste. Under normal conditions, arsenic does not leach from spent AdsorpAs.

Technology for Remediation and Disposal of Arsenic

Gravel Filter Bed

97

Adsorption Tower

Contaminated Water Inflow

Treated Water Outflow

Fig. 13 A granular ferric hydroxide-based arsenic removal unit

C

Read-F Arsenic Removal Unit

READ-F is an adsorbent produced and promoted by Nihon Kaisui Co. Ltd, Japan and Brota Services International, Bangladesh for arsenic removal in Bangladesh. Read-F is selective for arsenic ions under a range of conditions, effectively adsorbing both arsenite and arsenate. Oxidation of arsenite to arsenate is not needed for arsenic removal in this method, nor must pH be adjusted before or after treatment. The READ-F is composed of an ethylene-vinyl alcohol copolymer (EVOH) and hydrous cerium oxide (CeO2nH2O), with the latter acting as the adsorbent. The material contains 60% water, with a 0.7-mm average particle size and a 1.6 g/ml specific weight. The material contains no organic solvent or other volatile substance and is not classified as a hazardous material. Laboratory tests at BUET and field testing of the materials at several sites under the supervision of the Bangladesh Arsenic Mitigation Water Supply Project (BAMWSP) showed that the adsorbent is highly efficient in removing arsenic from groundwater. Two units utilizing READ-F technology are available commercially in Bangladesh: one is a household treatment unit and the second is a community treatment unit (Fig. 14). The units remove iron by sand filtration to avoid clogging the resin bed. The household unit has sand and resin beds arranged in one container, whereas these two beds exist separately in the community unit. According to proponents, READ-F is approved by the Japan Ministry of Health and Welfare for treatment of potable water and has provided efficient and dependable arsenic removal for water treatment facilities at Ibaraki, Japan for 3 yr. READ-F is regenerated by adding sodium hydroxide, then sodium hypochloride, and is finally rinsed with water. The regenerated READ-F is neutralized with hydrochloric acid and flushed before reuse. After neutralization, wastewater is treated with a small amount of adsorbent for safe disposal.

98

P. Visoottiviseth, F. Ahmed

Fig. 14 An arsenic removal unit (ARU) based on READ-F materials

Fig. 15 A household arsenic removal unit based on iron-coated sand

Plain sand

Iron Coated Sand

D

Iron-Coated Sand

BUET has tested a unit that utilizes iron-coated sand for removal of arsenic from groundwater (Fig. 15). Pretreatment for removal of excess iron, to avoid clogging of the active filter bed, is required. Pretreatment consists of precipitating iron by air oxidation. The water is then filtered through sand to trap excess iron. This sand filter, about 10 cm in depth, is placed in a 15-cm-diameter PVC chamber having perforations at its base. Water flows from the top of the bucket into the sand filter

Technology for Remediation and Disposal of Arsenic

99

250 As (III) vs Bed Volume As (V) vs Bed Volume

Arsenic Conc. µg / L

200

150

100

50

0 0

50

100

150

200

250

300

350

400

450

Bed Volumes

Fig. 16 Arsenic removal capacity of iron-coated sand

via a replaceable plastic pipe. A 1- to 2-cm-thick gravel bed is placed at the bottom to retain sand. The water then passes through a second 40-cm-deep iron-coated sand filter that is responsible for removing arsenic. Water enters into a strainer placed in iron-coated sand and eventually flows to the tap. Figure 16 shows the arsenate and arsenite removal capacities of iron-coated sand (Ali et al. 2003). Raw water containing 300 mg/L arsenic is effectively cleaned of arsenic when filtered through iron-coated sand (Fig. 16). It was found that ironcoated sand will process about 350 BV, each satisfying the Bangladesh drinking water standard (50 ppb) before becoming exhausted. Saturated media is regenerated by passing 0.2 N sodium hydroxide through the column or soaking the sand from the column with 0.2 N sodium hydroxide, followed by flushing with distilled water. Bed volumes, which represent the ratios of volume of water treated to volume of media, continued to remove arsenic successfully after five regeneration cycles. Iron-coated sand is effective in removing both As(III) and As(V).

E

Shapla Arsenic Filter

The Shapla Arsenic Filter, a household arsenic removal unit, has been developed and promoted by International Development Enterprises (IDE), Bangladesh. The unit media constitutes iron-coated brick chips manufactured by treating such chips with a ferrous sulfate solution; the media works on the same principles as ironcoated sand. Water from contaminated tube wells is allowed to pass through earthen containers filled with the filter media; the containers are fitted underneath with a drainage system. A drawing of the Shapla Filter is shown in Fig. 17.

100

P. Visoottiviseth, F. Ahmed

Lid

Flexible Water Delivery Pipe

Iron Coated Crushed Brick Particles Cloth Filter on Perforated Plate

Suppor

Treated Water in a Bucket

Fig. 17 Shapla filter for arsenic removal in households

It has been claimed that 20 kg Shapla filter media can clean (to nondetectable levels) up to 3,000 L tubewell water having arsenic concentrations of 0.3–0.4 mg/L. Currently, the cost of the filter, including the reusable earthen container, is 350 Tk (the Bangladeshi Taka equals about $US 5). IDE estimates that 20 kg filter material can be produced at a retail price of Tk 100, and claims that the exhausted filter media is nontoxic and can be disposed of safely. Experimental units have been installed for field testing in rural areas of Kachua, Sonargoan, Noakhali, and Rajshahi in Bangladesh in collaboration with UNICEF (United Nations International Children’s Emergency Fund), Danida, and SDC (Sustainable Development Commission of UNICEF). Reports say the unit is effective in arsenic removal and affordable for the majority of the rural population.

F

Sono Filter

The Sono filter uses zero valent iron filings (cast iron turnings), sand, brick chips, and wood coke for removing arsenic and other trace metals from groundwater in Bangladesh (Munir et al. 2001; Khan et al.2000). The filtration system originally consisted of a 3-kalshi unit (burned clay pitchers), widely used in Bangladesh for water storage and for drinking and cooking. Three kalshis were arranged vertically one above the other on a steel or wooden frame (Fig. 18a). The top kalshi contained 3 kg cast iron turnings, and this layer is covered with 2 kg sand. The middle kalshi contained 2 kg sand, 1 kg charcoal, and 2 kg brick chips. Brick chips are also placed around the holes to retain finer materials. Tubewell water is poured in the top kalshi, filtered through the top and middle kalshis, and the treated drinking water is collected from the bottom kalshi. Nikolaidis and Lackovic (1998) showed that

Technology for Remediation and Disposal of Arsenic

Raw Water

101

Top Bucket (Red)

Filter Media 1 Coarse Sand, Composite Iron Matrix(CIM) Coarse Sand Brick Chips

Filter Media 1 Sand, Iron Filings & Brick Chips

Filter Media 2 Sand, Charcoal & Brick Chips

Bottom Bucket (Green)

Filter Media 2 Coarse Sand, Charcoal Fine Sand Brick Chips

Filtered Water Filtered Water

a

b

Fig. 18 The Sono Filters for arsenic removal from groundwater. (a) Sono three-kalshi arsenic filter. (b) Modified Sono arsenic filter. (From Hussan 2003.)

97 % arsenic can be removed by adsorption on a mixture of zero valent iron filings and sand, and recommended that arsenic species can be removed through formation of coprecipitates and mixed precipitates and by adsorption onto the ferric hydroxide solids. The Sono 3-Kalshi unit was very effective in removing arsenic, but the onetime-use unit is rapidly clogged if groundwater contains excessive iron. Field observations indicated that, over time, the iron filings bond with the solid mass, rendering cleaning and replacement of materials difficult. To overcome these problems, the filter media are housed in two plastic buckets (Fig. 18b). The cast iron turnings have been processed into a complex iron matrix (CIM), which is capable of maintaining its active CIM integrity for years. Manganese in CIM catalyzes oxidation of As(III), and all As(V) is removed by a surface-complexation reaction between the surface of hydrated iron (FeOH) and arsenic species (Hussan 2003). In 2007, the Sono Filter competed for and won the $1 million ‘‘Grainger Challenge Prize’’ for sustainability; the prize was awarded for innovative solutions in removing arsenic from drinking water.

G

SAFI Filter

The SAFI filter is a type of household candle filter. The candle is made of composite porous materials such as kaolinite and iron oxide on which hydrated ferric oxide is deposited by sequential chemical and heat treatment. The candle filter works on the principles of adsorption, filtration, and on chemically treated active porous composite materials. The oxyhydroxides of Fe, Al, and Mn assist in the removal of

102

P. Visoottiviseth, F. Ahmed

Table 2 Some important features of SAFI filters claimed by the proponent Characteristics of filter Claims by the proponent Candle lifetime 2 yr (treatment capacity, 4000 L with As content 1.5 ppm) Flow rate 40 L/d (small type), 60–80 L/d (standard type) Leaching No leaching of arsenic from candle up to pH 11 Regeneration Can be regenerated three times (cycles) at a cost of Tk.70 per regeneration Clogging If clogged, flow can be restored by treatment with 5% H2SO4 Candle cost Tk 600 (standard type), Tk 250 (small type)

arsenic, iron, and bacteria. Some of the relevant features claimed for the SAFI filter are shown in Table 2. The SAFI filter was reported to have good arsenic removal capacity initially, but efficiency declined with time. Moreover, the filter media became clogged, and the unit suffered rapid erosion from mechanical cleaning and was generally regarded to show poor workmanship. The filter candle, in many cases, was found to leak at joints and to disintegrate because of inadequate strength.

H

Activated Carbon

Granular activated carbon (GAC) removes arsenic by adsorption to some extent, depending on the pH of the raw water. Studies conducted by All India Institute of Hygiene and Public Health (AIIH & PH), using granular activated carbon, revealed that GAC could be used to remove arsenic from groundwater, but the process was not economically viable.

I Indigenous Filters There are several filters available in Bangladesh that use indigenous material as arsenic adsorbents. Red soil, rich in oxidized iron, brick chips, clay minerals, iron ore, iron scrap or filings, and processed cellulose materials are known to adsorb arsenic. Some filters manufactured from these materials include Granet Homemade Filter, Chari Filter, Adarsha Filter, Bijoypur Clay, and Processed Cellulose filter. The Garnet home-made filter contains relatively inert materials such as brick chips and sand. No chemical is added to the system. Air oxidation and arsenic adsorption on iron-rich brick chips, plus natural iron flocs (in groundwater) may be the main mechanisms by which arsenic is removed from groundwater using this process. The Garnet home-made filter was previously evaluated (BAMWSP, DFID and Water Aid Bangladesh 2000). The unit produced an inadequate quantity of water and did not show reliable results under various operating conditions in different areas of Bangladesh.

Technology for Remediation and Disposal of Arsenic

103

The Chari filter is a modified version of the three-kalshi filter, in which open pitchers or Charis were used to allow easier access to the absorbent materials for washing and replacement. The top Chari filter uses brick chips and iron filings, while sand is placed in the middle Chari. The unit was developed by Dhaka Community Hospital (DCH) and has been used extensively in their arsenic mitigation programs. It appears that the ‘‘open concept’’ design of the Chari filter renders long life to the filter, when compared with regular three-kalshi units, which are replaced every 3–6 mon. Open units, however, are more vulnerable to contamination. During field visits, most Chari filters have been found to exist in unhygienic conditions. The effectiveness of this process to remove arsenic, bacteria, and other contamination is not known. The Adarsha filters employ clayey material to form filter candles. The Adarsha filter participated in an assessment conducted in Bangladesh but it failed to meet the arsenic reduction criterion (BAMWSP, DFID and Water Aid Bangladesh 2001). Aluminum-rich Bijoypur clay and treated cellulose were also found to adsorb arsenic from water (Khair 2000). However, no commercial unit has been constructed to utilize these materials for arsenic removal in Bangladesh.

J Cartridge Filters Cartridges filled with sorptive media or ion-exchange resins are available in the market. These units remove arsenic and other dissolved ions present in water. Cartridge filters are unsuitable for water having high impurity levels and/or iron because such ions have high affinity for media and can quickly saturate it, requiring frequent regeneration or replacement. The Chiyoda Arsenic Removal Unit (Japan), available in Bangladesh, was tested at BUET Laboratory. This Chiyoda Removal Unit treated 800 BV and still met the WHO guideline value (10 mg/L), or 1,300 BV, and still met the Bangladesh Standard (50 mg/L), when the feed water arsenic concentration was 300 mg/L (Ahmed et al. 2000).

V

Ion Exchange

The ion-exchange process utilizes similar principles to that of activated alumina but employs synthetic resins of enhanced ion-exchange capacity. The synthetic resin is based on a cross-linked polymer skeleton called the matrix. The charged functional groups are attached to the matrix through covalent bonding and fall into strongly acidic, weakly acidic, strongly basic, and weakly basic groups (Clifford 1990). The resins are used to remove specific undesirable cations or anions from water. The strongly basic resins can be pretreated with anions such as Cl– and are used for removal of negatively charged species, including arsenate. The capacity of ion exchange to remove arsenic is dependent on the sulfate and nitrate content of raw water because these ions have a higher affinity for ion exchange than does arsenic. Compared to other types of media, the ion-exchange

104

P. Visoottiviseth, F. Ahmed

process is less dependent on the pH of water. Arsenite is uncharged and is therefore not removed by ion exchange; hence, preoxidation of As(III) to As(V) is required for removal of arsenite. The oxidant must be removed from ion-exchange media to avoid damage to the resins. Arsenic removal would be enhanced if ion-exchange resins were developed that could selectively remove arsenic species. Ion-exchange capacity is similar to adsorption capacity in that both are measured by the number of active sites per unit of media, usually expressed by milliequivalents (mEq) per mL. Typical theoretical exchange capacities for strong base anionexchange resins range from 1 to 1.4 mEq/mL (Clifford 1990) or 3.0 to 4.2 mEq/g dry wt. The maximum sorption capacity for arsenic (molecular weight of 75) is 315 mg As/g. Actual sorption capacities under field conditions are much lower. After they are exhausted, ion-exchange resins can be easily regenerated by treatment with NaCl solutions. The equations that describe this regeneration are as follows: Arsenic exchange: 2R-Cl þ HAsO4 ! R2 HAsO4 þ 2Cl

ð10Þ

R2 HAsO4 þ 2Nþ þ 2Cl ! 2R-Cl þ HAsO4 þ 2Naþ

ð11Þ

Regeneration:

where R stands for ion-exchange resin. Tetrahedron Technology Tetrahedron (USA) promoted an ion exchange-based arsenic removal technology in Bangladesh. The technology proved its arsenic removal efficiency, even at high flow rates. Figure 19 shows the schematic diagram behind this technology. This Chlorine Source Sieve Stabilizer Column Head

Tap

Stone Chips

Stand

Resin Column (Ion Exchanger)

Fig. 19 Depiction of the Tetrahedron arsenic removal technology

Technology for Remediation and Disposal of Arsenic

105

process utilizes a stabilizer and an ion exchange (resin column) along with facilities for chlorination using chlorine tablets. Tubewell water is pumped or poured into the stabilizer through a sieve containing a chlorine tablet. The water, mixed with chlorine, is stored in the stabilizer and subsequently flows through the resin column when the tap is opened. The purpose of the chlorine is to kill bacteria and oxidize arsenic and iron. The stabilizer smoothes flow pulses from the pump and traps iron and other hydroxide precipitates formed in water. Finally, the ion-exchange media adsorbs and cleans arsenic, sulfate, and phosphate from the water. This Tetrahedron filter was tested in Bangladesh (BAMWSP, DFID and Water Aid Bangladesh 2001) and demonstrated promising results. The residual chlorine minimized bacterial growth in the media. The saturated resin can be regenerated by NaCl solution. Liquid wastes from the process, including salt and arsenic produced during regeneration, require safe disposal.

VI

Membrane Techniques

Synthetic membranes can remove many contaminants from water, including bacteria, viruses, salts. and various metal ions. Usually, two types of membrane filtration are used: (1) low-pressure membranes such as microfiltration (MF), and (2) highpressure systems, such as ultrafiltration (UF), nanofiltration (NF), and reverse osmosis (RO). The pore size of membranes and size of materials to be separated are shown in Fig. 20. It is apparent (Fig. 20) that RO and NF have the appropriate

Size, Micron Relative Size of Various Materials Present In water

0.001

Aqueous salts

0.01

0.1

1.0

10

100

1000

Bacteria

Viruses

Algae Humic acids

Metal ions

Separation Reverse Osmosis Ultrafiltration Processes Nonofiltration

Cysts

Clay

Silt

Sand

Conventional Filtration Processes Microfiltration

Fig. 20 Sizes of impurities present in water and pore sizes of various membranes. (From Najm and Trussell 1999.)

106

P. Visoottiviseth, F. Ahmed

pore sizes for removal of arsenic. In recent years, a new generation of RO and NF membranes have been developed that are less expensive and operate at lower pressures. Arsenic removal by membrane filtration is independent of pH and presence of other solutes but can be adversely affected by presence of certain colloids. Also, iron and manganese, if present, can lead to scaling and membrane fouling. Once fouled by impurities, the membrane cannot be successfully backwashed but must be replaced. Therefore, water with high suspended solids content requires pretreatment for arsenic removal. Most membranes, however, cannot withstand oxidizing agents. The US EPA (2002) reported that NF was capable of more than 90% removal of arsenic, while at ideal pressures RO provided removal efficiencies greater than 95%. Water rejection (about 20%–25% of the influent) may be an issue in water-scarce regions (US EPA 2002).

A

Techno-Food Water Technology

Techno-food (Bangladesh) Co. is a manufacturer and supplier of water demineralization units that use membrane technology to clean industrial water supplies. This organization markets several domestic water purification systems with water treatment capacities varying from 60 to 1,200 L/d. Techno-food Bangladesh have also introduced new-generation NF and RO membranes for arsenic removal in Bangladesh. The Techno-Food water treatment units operate at 50–150 psi, and remove 95%–98% of total dissolved solids, including arsenic. In this method, arsenic removal is independent of pH. In addition to arsenic, membrane filtration removes many other impurities, including bacteria. The membrane does not utilize chemicals and does not accumulate arsenic as do other adsorbing materials; hence, disposal of used membranes is not a threat to the environment. Operation and maintenance of these units is simple, requiring only periodic wiping to clean membranes. The seller claims that the RO and NF membrane-separation technology is among the safest of arsenic removal systems. Installation of small community or large production units for public water supplies is possible at affordable costs with this technology.

B

MRT-1000 and Reid System, Ltd

Jago Corporation, Ltd promoted a household RO water dispenser (MRT-1000) manufactured by B & T Science Co., Ltd, Taiwan. This system was tested at BUET and demonstrated an arsenic (III) removal efficiency greater than 80%. A wider-spectrum RO system, named Reid System, Ltd was also promoted in Bangladesh. Experimental results showed that this system could effectively reduce arsenic and other impurities in water. The capital and operational costs of the RO system are relatively high.

Technology for Remediation and Disposal of Arsenic

C

107

Low-Pressure Nanofiltration and Reverse Osmosis

Oh et al. (2000) applied RO and NF membrane processes to treatment of arsenic contaminated water in which needed pressure was applied with a bicycle pump. A NF membrane process coupled with a bicycle pump can operate under conditions of low recovery and low pressure (from 0.2 to 0.7 MPa). The rejection rate for arsenite is lower than for arsenate in ionized forms; hence, water containing higher arsenite levels requires preoxidation to achieve acceptable arsenic removal. Tubewell water in Bangladesh has an average ratio of arsenite to total arsenic of 0.25. However, RO coupled with a bicycle pump operating at 4 Mpa can remove arsenic because of its high rate of rejection. The study concluded that low-pressure NF with preoxidation or RO (coupled to a bicycle pump) could successfully treat arsenic-contaminated groundwater in rural areas (Oh et al. 2000).

D

Mobile Reverse Osmosis

The Prince of Songkla University, Thailand (Wanichapichart 2005) has tested a mobile RO machine capable of removing arsenic from water. This machine (Fig. 21) produced drinking water at a rate of 208 L/hr and could fill a 750-L container daily. It operates by passing pipeline water through a 0.5-mm filter to an RO spiral-wound membrane. The quality of water before and after filtration, and user attitudes toward the project, are shown in Table 3. Houses nearest the RO machine benefited most

Fig. 21 A mobile reverse osmosis (RO) unit

108

P. Visoottiviseth, F. Ahmed

Table 3 Operating parameters of, and user attitudes toward, use of the reverse osmosis (RO) technology Elements Feed Permeate Standard Summarized (%) from 238 families pH 4.5–6.1 4.5–6.1 6.5–8.5 51% faced with shortage of drinking water in Hardness (as CaCO3) 3.1–11 <1.0 <100 summer Iron 0–0.13 <0.01 <0.3 Sulfate Chloride Nitrate Arsenic (ppb)

<25 <5 1.0 100-150

<25 <5 0.6 <10

<250 <250 <4.0 10

42% paid for bottled water 34.6% obtained free water from the RO machine 97.5% in the district requested an RO‐ machine Treated-water output of 15– 500 gal/mon/family for cooking and drinking

from the project; others more distantly located did not use water from the system. Users in the community agreed that membrane technology was useful in providing drinking water of acceptable quality during the summer months. Communities in the district may consider a mobile RO machine as an alternative to provide arsenicfree water, both for their community and for nearby districts. However, the membrane in such units must be changed every 3 mon, which entails further expense.

E

Combined Sand and Nanofiltration

The efficiency of combining oxidation, sand filtration, and nanofiltration (NF) to remove arsenic was studied jointly by Rajshahi University, Bangladesh and Hokkaido University, Japan. A pilot plant with equipment for (1) oxidation by air and/or NaClO, followed by (2) two sand filter columns, each filled with 1-m-high manganese sand, and (3) a NF unit for further removal of arsenic to trace level, was constructed. This plant was field tested for 8 mon using tubewell water contaminated with arsenic (98–170 mg/L), iron (2,470–9,900 mg/L), and manganese (455– 700 mg/L). Oxidation and one-stage filtration through manganese sand reduced arsenic content of water to 44 mg/L, which is within the limits acceptable in Bangladesh. The NF unit was installed after sand filtration to attain a high level of arsenic removal. Arsenic content, in treated water following NF, was 0.4 mg/L or less (Rahman and Magara 2002). The NF system gave a high water recovery rate of 60%, although the process is costly.

VII

Bioremediation by Algae

Remediation of arsenic by an alga, Chlorella vulgaris, has been explored by many investigators. Metal accumulation by algae is influenced by a number of biotic and abiotic factors (Genter 1996). Abiotic factors include chemical speciation of As,

Technology for Remediation and Disposal of Arsenic

109

metal concentration, duration of exposure, concentration of other ions (e.g., Ca, Mg, P), pH, presence of complexing and chelating agents, redox conditions, temperature, light, and flow rate of water. Biotic factors influencing performance include species-specific characteristics, algal biomass, extracellular products, stage of organism development, and cellular activity. Results indicated that Chlorella vulgaris grew better in a medium with arsenate at concentrations up to 2,000 mg/L and accumulated arsenate at levels up to 50,000 mg As/kg dry cell wt (Maeda and Sakaguchi 1990; Maeda and Ohki 1998). Visoottiviseth and Lauengsuchonkul (2004) reported that the green alga Chlorella vulgaris exhibited good growth when exposed to very high concentrations of arsenic. The authors immobilized this alga on alginate beads and tested its efficiency. Results indicated that efficiency depended on the numbers of beads and duration of exposure (Fig. 22). They then constructed an arsenic filter using the immobilized Chlorella vulgaris in combination with other adsorbent materials (Fig. 23). Tests of this algal filter in arseniccontaminated areas of Thailand showed that it performed well, removing at least 95% of arsenic present. This filter is economic, although the alginate beads must be changed every 3 mon. The used beads can be further used as fish food.

VIII

Phytoremediation of Arsenic

Because soils are a major source of arsenic contamination, removal of arsenic from soil should help reduce arsenic pollution of surface waters. In 2001, the arsenic hyperaccumulating fern, Pteris vittata, was discovered by Ma et al. (2001), and later Visoottiviseth et al. (2002) discovered Pitylogramma calomelanos, another species of arsenic hyperaccumulating fern. Both ferns have been tested for their efficiency in removing arsenic from water and soil (Tu et al. 2004; Alkorta et al.

70.00

675 beads 60.00

% As removal

2025 beads 50.00

3375 beads

40.00 30.00 20.00 10.00 0.00 0 min

5 min

10 min

15 min

30 min

1 hr

3 hr

9 hr

24 hr

7 day

Time

Fig. 22 Removal of arsenic from water using different numbers of algal beads at various exposure times

110

P. Visoottiviseth, F. Ahmed

Fig. 23 Arsenic filter using immobilized Chlorella vulgaris and an adsorbent material

2004; Huang et al. 2004). In Thailand, a field trial on phytoremediation using the fern Pitylogramma calomelanos was planned in an area highly contaminated with arsenic. However, use of this fern was rejected by the local population because they could not discern any economic benefit to them from its use. Thus, plants selected for phytoremediation must meet rather diverse criteria. They must accumulate high levels of As, have high biomass, have a short life cycle, tolerate high concentrations of arsenic, and be economically important to users. Marigolds are plants that possess such characteristics and have been used in a field trial (Fig. 24). Each marigold plant has at least 10 flowers and can be sold for $2.00 per 100 flowers. The flowers can be planted, harvested, and sold in a 45-d cycle. In addition to achieving arsenic removal, it was estimated that growers of marigold plants could each earn at least $57,500 per ha per yr (Chintakovid et al. 2007).

IX

Comparison of Arsenic Removal Technologies

Advances in using modern technology to remove arsenic from rural water supplies have been remarkable during the last 2–3 yr. A comparison of extant processes is shown in Table 4. All technologies described have merits and demerits and are continually being refined. The major objectives of these technological refinements are to (1) improve efficiency in arsenic removal, (2) reduce capital and operating costs, (3) improve user friendliness, (4) achieve easier maintenance, and (5) resolve arsenic sludge and concentrate management problems. Arsenic removal technologies must be economically sound to be accepted by users.

Technology for Remediation and Disposal of Arsenic

111

Fig. 24 Marigold plants grown in the high arsenic-contaminated areas of Thailand

X

Conventional Filtration

The performance of conventional filtration methods is variable, and the efficiency of arsenic removal achieved can range widely (from 4.5% to 96%), depending upon conditions and the particular adsorbent used (Jiang 2001). Criteria for selecting a suitable adsorbent include cost of the medium; ease of operation and handling; adsorption capacity (breakthrough point) and potential for re-use and regeneration. In a recent review of arsenic treatment technologies published by Vu et al. (2003), iron filings, ferric salts, granular ferric hydroxide, alumina-manganese oxide, Aquabind, and kimberlite tailings are all listed as potentially low-cost adsorbents. Properly handled, all can remove arsenic in a relatively short time, and the adsorbents themselves can then be removed from water by filtration. Moreover, metalloaded polymers are new prospective sorbents for removing As(III) and As(V) from water and may have promise (Dambies 2004). Other sorbent materials for arsenic removal have also been reviewed: granular activated carbon (Bissen and Frimmel 2003; Daus et al. 2004; Jiang 2001); iron-coated sands (Jiang 2001; US EPA 1999); manganese greensand (Viraghavan et al. 1999); manganese-coated sand (Bajpai and Chaudhuri 1999); iron hydroxide granulates (Bissen and Frimmel 2003; Daus et al. 2004; Driehaus et al. 1998; Pal 2001); natural materials (natural zeolites, volcanic stone, catcaceous powder) (Elizalde-Goznales et al. 2001); zero valent iron (iron filings) (Nikolaidis et al. 2003; Su and Puls 2001, 2003), and kimberlite tailings (Dikshit et al. 2001). Elizalde-Gonzales et al. (2001) reviewed naturally occurring solids as agents for arsenic adsorption from water. These authors conclude that, although such solids are cheap—indeed, they can be obtained free of charge—their rates of removing arsenic from water are usually low. Naturally

112

P. Visoottiviseth, F. Ahmed

Table 4 A comparison of advantages and disadvantages of major arsenic removal technologies Technology Advantages Disadvantages Oxidation and sedimentation l

Air oxidation

l

Relatively simple, lowcost but slow

l

Partial removal of arsenic

l

Chemical oxidation

l

Relatively simple and rapid

l

Used as pretreatment for other processes

l

Oxidizes other impurities and kills microbes

Coagulation and filtration l

Alum coagulation

l

Relatively low capital cost

l

Produces toxic sludges

l

Iron coagulation

l

Relatively simple

l

Low removal of As(III)

Utilize commonly available chemicals

l

Preoxidation is required

l

Removal efficiencies may be inadequate to meet strict standards

l

Sorption techniques l

Actvated alumina

l

Well known and commercially available

l

Produces arsenic-rich liquid and solid wastes

l

Iron coated sand

l

Well defined

l

Replacement/regeneration is required

l

Ion exchange resin

l

Many possibilities and large development effort

l

High-tech operation and maintenance

l

Relatively high cost

l

Other sorbents Membrane techniques

l

Nanofiltration

l

Well defined; high removal efficiency

l

High capital and running costs

l

Reverse osmosis

l

No toxic solid wastes produced

l

High-tech operation and maintenance

l

Electrodialysis

l

Capable of removing other contaminants

l

Arsenic-rich water effluent is produced

Algal filter

l

Produce no toxic wastes

l

Light is required for photosynthesis

Phyto-remediation

l

High efficiency (>95%)

l

Environment friendly

l

Low capital cost

l

Provides income to users

Bioremediation l l

l

Used algal beads can be fed to fish Source: Based on Ahmed 2003; The World Bank 2005.

Technology for Remediation and Disposal of Arsenic

113

occurring solids are often employed for household-level water treatment in underdeveloped countries, such as Bangladesh. Saha et al. (2001) studied a wide range of adsorbents (mainly natural materials) for their ability to remove As(III) and As(V) from water. They compared the following materials: Kimberlite tailings, water hyacinth, wood charcoal, banana pith, coal fly ash, spent tea leaves, mushroom, saw dust, rice husk ash, sand, activated carbon, bauxite, hematite, laterite, iron oxide-coated sand, and hydrous granular ferric hydroxide. Activated alumina was used as the reference in this study. Removal efficiencies for As(III) varied from 5% to 92% and for As(V) from 12% to 99%. Iron oxide-coated sand and hydrous granular ferric oxide performed best and was studied in detail. In the field portion of the study, columns containing these two adsorbents were run for 20 d with influent water containing 0.32 mg/L arsenic. The column with granular ferric hydroxides supplied 45 L water/d with less than 0.01 mg/L arsenic.

XI

Alternative Water Supply Options

Although shallow tube wells recently dug in South Asia to tap alluvial aquifers do provide low-cost drinking water, this water is often contaminated with arsenic. Such contaminated water poses a hazard to millions of people. The problem is exacerbated because tube wells with high levels of arsenic are in the same areas where contaminated tube wells predominate. In the absence of an alternative water source, people drink arsenic-contaminated water without considering the consequences. Alternatively, people who avoid tubewell water by drinking surface water are at high risk of contacting water-borne diseases. Unless arsenic contaminated water is cleaned before use, the only alternatives for access to arsenic-safe water include deep tube well, dug/ring well, rainwater harvesting, treatment of surface water, and piped water supply.

A

Deep Tube Well

The aquifers deposited at different geological times are usually stratified. Deeper aquifers, separated by relatively impermeable strata, are relatively free from arsenic contamination. Study results show that only about 1% of tube wells with depths exceeding 150 m are contaminated with arsenic levels above 50 mg/L. Similarly, only 5% of deep tubewell water exceeds arsenic levels of 10 mg/L (BGS and DPHE 2001). Therefore, deep aquifers separated from shallow contaminated aquifers by the prerequisite impermeable layer can be dependable sources of arsenic-safe water. To avoid percolation of arsenic-contaminated water, annular spaces of deep tubewell boreholes must be sealed, at least to the level of the impermeable strata (Fig. 25). It is very difficult to seal a small-bore tube well, although technological refinements using clay as a sealant are underway. A protocol for installation of deep tube wells adequate to achieve arsenic mitigation has been developed in Bangladesh. Because of stratification problems, such deep tube wells may initially

114

P. Visoottiviseth, F. Ahmed

Fig. 25 Deep tube well diagram Manually Operated Deep Tubewell

Silty clay Arsenic Contaminated Shallow Aquifer (Sandy silt )

Water Table

Clay Seal Clayey Layer Relatively Impermeable Strainer

Arsenic-safe Deep Aquifer (Fine to medium sand)

Sand Trap

yield arsenic-safe water but later see an increase of arsenic content from mixing of contaminated and uncontaminated waters. Such mixing can be minimized by recharging the deep aquifer with water filtered through coarse media or by replenishment from horizontal movement of water. Experience in the design and installation of tube wells showed that reddish sand produces water with optimum amounts of dissolved iron and arsenic. The reddish sand color results from oxidation of iron on sand grains to ferric form. Such sand will not release arsenic or iron into groundwater; rather, ferric iron-coated sand will adsorb arsenic from groundwater. The Dhaka water supply, although contaminated by arsenic, is probably protected by its local red-colored soil. Hence, installation of tube wells in reddish sand, if possible, should be undertaken to protect against arsenic contamination.

B

Dug/Ring Well

Dug wells are the oldest method of procuring groundwater. Water from dug wells is relatively free from dissolved arsenic and iron, even in locations with contaminated tube wells. Why such dug wells avoid high arsenic levels is not fully known. The following are among explanations given for low arsenic content of dug well water: (1) arsenic and iron precipitates because of oxidation from open air exposure and

Technology for Remediation and Disposal of Arsenic

115

agitation during water withdrawal, and (2) dilution with freshwater from seepage into the well from replenishing rainwater, surface water, or groundwater (from the top layer of a water table). Such water would be cleaned by percolation through aerated soil. Dug well collects water from surface of an aquifer, which is close to ground surface where microbial pollution and presence of organic matters are higher. Water from such aquifers smells bad, has high turbidity, and contains ammonia. The water from shallow aquifers is also susceptible to bacterial contamination. Satisfactory protection against bacteriological contamination is possible by sealing the well top with a watertight concrete slab. Water may be withdrawn by installing a manually operated hand pump. Completely closed dug wells have good sanitary protection, but a dearth of oxygen can adversely affect the water quality. Construction and operational difficulties are often encountered when dug wells are sunk into silty and sandy soils. ‘‘Sand boiling’’ interferes both with digging and operation of dug wells and can lead to well wall collapse. Water in the well must be chlorinated for disinfection after construction, and lime may also be added to improve water quality.

C

Surface Water Treatment

Sand Filter (SF) One option for a community-type surface water-based system is a slow sand filter (SSF), commonly known as a pond sand filter (PSF) in Bangladesh, where it was originally designed for filtration of pond water. It is a package-type slow sand filter unit developed to treat surface waters, usually low-saline pond water, for domestic water supply in coastal areas. Water from the pond or river is manually pumped from a tube well to the filter bed, which rests above ground. Treated water is collected for distribution from taps. Treated water from a PSF is normally bacteria free or within safety limits. The average period between maintenance for the PSF unit is usually 2 mon, after which the sand in the bed must be cleaned. The construction of a typical PSF is shown in Fig. 26. The operating conditions necessary for slow sand filters include low turbidity (<30 nephelometric turbidity units, NTU), low bacterial count, no algal bloom, absence of Cyanobacteria, and free from bad smell and color. A protected surface water source is ideal for slow sand filtration. If the foregoing conditions are broached, the following problems are likely to be encountered: low discharge, need for frequent cleaning, and poor effluent quality. Community involvement in proper operation and maintenance of these small units is absolutely essential if the system is to remain operational. The package-type SSF provides low cost and very high efficiency for removal of turbidity and bacteria. It is preferred for use by medium-size settlements in their water supply systems. Although PSF has very high bacterial removal efficiency, it fails to reduce bacterial count to acceptable levels when surface water is heavily contaminated. In such cases, treated water may require chlorination to meet

116

P. Visoottiviseth, F. Ahmed

Handpump for water supply to filter

SURFACE WATER CLEAR WATER

Filter sand Coarse aggregate Under drainage system

Raw water from pond

Fig. 26 Pond sand filter for treatment of surface water

drinking water standards. Roughing (pretreatment) filters are combined with a SSF when water turbidity exceeds 30 NTU. Roughing filters are responsible for removing turbidity and color to a level adequate for efficient operation of the SSF. Although the cost is higher, communities may construct small-scale conventional surface water treatment plants that utilize coagulation-sedimentation-filtration and disinfection to cope with variable raw water quality. Wetland Treatment Wetland treatment can remove arsenic from surface water using aquatic or wetland plants commonly found in a contaminated area. In Asian countries, where sunlight is available all year, wetland treatment may be more suitable than other technologies, because it is inexpensive to establish and maintain and is environmentally friendly. Among plants commonly used in this method are alum, phragmites, vetiver grass, and cattail. Among these plants, alum (Colocasia esculenta) was best at removing arsenic. Alum retains the highest concentration of As it absorbs it in its underground stem (Aksorn and Visoottiviseth 2004). As plants grow and absorb arsenic, they are harvested, and new plants are seeded to repeat the process (Fig. 27). Wetland treatment in Thailand removed more than 90% of arsenic (accumulating As at concentrations from 300 to >1000 mg As/L) from surface water.

Rainwater Harvesting (RWH) Rainwater harvesting (RWH) can be used as a source of clean water in South Asian countries. The advantages and disadvantages of RWH systems are presented in

Technology for Remediation and Disposal of Arsenic

117

Fig. 27 Elephant ear or alum, Colocasia esculenta, growing in a wetland

Table 5 Advantages and disadvantages of rainwater collection systems Advantages Disadvantages l

The quality of rainwater is comparatively good

l

The initial cost may prevent a family from installing a rainwater harvesting system

l

The system is independent and therefore suitable for scattered settlements

l

Water availability is limited by rainfall intensity and available roof area

l

Local materials and craftsmanship can be used in construction of the rainwater system

l

Mineral-free rainwater has a flat taste, which may not appeal to all persons

l

No energy costs are incurred in running the system

l

Mineral-free water may cause nutrition deficiencies in people on mineral-deficient diets

l

Easy to maintain by owner/user

l

Poorer segments of the population may not have suitable roofs for rainwater harvesting

l

The system can be located very near consumption points Source: GOB (2002).

Table 5. If one relies on a rainwater supply system, both the volume of water required in the community and probable availability of rain (intensity and distribution) must be determined in advance. In particular, a storage tank and catchment area of the proper size must be constructed. The unequal distribution of annual

118

P. Visoottiviseth, F. Ahmed

rainfall in Asian countries is offset by constructing large storage tanks; such tanks constitute the main cost of the system. In Bangladesh, the most commonly used household water container is a round concrete tank about 1.6 m high and 2 m in diameter. It contains approximately 3,200 L water and is connected via a gutter to collect rainwater from the roof. The container can supply water to two families, comprising 10 people, for about 64 d (assuming 5 L/person/d). Because this amount (5 L/person/d) is less than is normally consumed, rainwater harvesting is inadequate by itself to meet the needs of an average household. In Thailand, water containers hold only 2,500–3,000 L, but each family may use multiple containers. Such containers were given to residents by the Thai Government to mitigate the arsenic pollution problem in the Ron Phibun district of southern Thailand. Unfortunately, people living in the affected area prefer to drink well water rather than rainwater because they like its taste better. The quality of rainwater is relatively good, but the water is not free from impurities. Analysis of stored rainwater has shown some bacteriological contamination. The cleanliness of roofs and storage tanks is critical in maintaining good rainwater quality. Initial runoff from the roof should be discarded to prevent entry of impurities into the stored water. If the storage tank itself is clean, bacteria or parasites from rainwater will tend to die off. Some devices have been offered, or good practices suggested, that either contain or divert the first ‘‘foul’’ flush from roofs away from the storage tank. When such flows cannot be successfully diverted, regular roof and gutter cleaning, before the rainy season, is needed, and regular maintenance thereafter; such cleaning enhances the quality of stored rainwater. The storage tank must be cleaned and disinfected annually or when the tank is emptied. The rather large tanks are difficult to clean, and to do so effectively requires someone to actually climb inside the tank. In Thailand, women do not like to do this because it is thought to bring bad luck to the head of the family. Another downside to RWH is that minerals, generally regarded as essential for good tasting drinking water, are essentially lacking in rainwater. The roof constitutes the normal catchment area for rainwater collection. Rainwater can be collected from any type of roof, although concrete, clay tiles, and metal give the cleanest water. The poorer sections of Bangladesh are not well positioned to utilize rainwater. These people lack roofs or have only small thatched ones. A thatched roof covered with polyethylene can be used as a catchment for rainwater but requires application of certain skills to properly guide water to the storage tank. In coastal areas of Bangladesh, people are known to use their clothes, fixed at four corners with a pitcher underneath, to collect rainwater. A plastic sheet (Fig. 28) has also been tried as a catchment for rainwater harvesting by people who have no good alternatives. Land surfaces have been used for rainwater catchment as well, and the resultant water has been stored underground in gravel/sand-packed reservoirs. In such cases, water is channeled toward the reservoir and allowed to pass through a sand bed before entering the reservoir. This process is analogous to recharge of an underground aquifer during the rainy season for utilization in the dry season.

Technology for Remediation and Disposal of Arsenic

119

Fig. 28 Plastic sheet catchment for rainwater harvesting (RWH)

D

Piped Water Supply

Having a supply of piped water is preferred by consumers because (1) water can be delivered in close proximity to consumers, (2) the water is protected from external contamination, (3) it is usually monitored for quality, (4) it is likely to benefit from institutional operation and maintenance, and (5) the required quantity of water is delivered. Only piped water can compete in convenience, storage, and use with water from tube wells. Moreover, piped water is feasible for clustered rural settlements and settlements in many urban fringes. The piped water can be connected to houses or to spigots near the house, or at other standposts depending on the financial condition of consumers. Piped water can come from sinking a deep tube well in a arsenic-safe aquifer, or treatment of contaminated surface or tubewell water in community treatment plants. In Bangladesh, supplying piped water to rural areas has been examined, and a large number of pilot schemes are being implemented there by various organizations. Piped water is a more problematic and costly option when populations live in the scattered rural areas of Bangladesh and West Bengal.

XII

Operational Issues

There are many challenges in reducing arsenic contamination to the desired levels. Analysis of arsenic at low concentration is difficult, as is performance monitoring of water treatment systems. Validating the claims made by sellers concerning the

120

P. Visoottiviseth, F. Ahmed

performance of treatment technologies or arsenic measurement devices is also important, but challenging. Safe disposal of toxic sludge and spent media is a continuing environmental concern. Technologies that rely on patented media or processes and imported components may not be easily procured when such materials and components are needed. Many of the same issues are also important in the operation of small water treatment facilities used in households, or at the community level. It is not possible to make broad arrangements for operation, repair, and maintenance of small water supply systems. The know-how and participation of people at the local level are vital for keeping small treatment systems operational. Many small-scale systems have failed for lack of local initiative, commitment, and ownership. The unit cost of the small systems may be higher because of the necessity of scaling down larger conventional water treatment system designs.

A

Costs

Cost is a key factor when considering the use and sustainability of arsenic removal technology in rural areas. Cost varies with materials used, quantity of media/ chemicals used, quality of groundwater, etc. Most technologies have only been installed and operated during either field or pilot-scale testing. Hence, the costs of installation, operation, and maintenance are not well known or standardized to meet various local conditions (The World Bank 2005). Costs of alternative water supply systems are presented in Table 6. The unit cost of water produced by different systems, expressed as annualized capital recovery, is presented in Table 7. The cost of arsenic mitigation depends on which technology is used. An evaluation of comparative technology costs demonstrates that deep tube wells can provide water at nominal cost in many cases but cannot provide Table 6 Costs for installing, operating, and maintaining alternative technologies to secure water with acceptable levels of arsenic Alternative No. Unit cost, Operation and Comments technology household $US maintenance cost/ (s)/ unit yr, $US Rainwater 1 200 5 Low reliability harvesting Dug/ring wells 25 800 3 Depth about 8 m Deep tube well 50 900 4 Depth about 300 m Pond sand filters 50 800 10–20 Slow sand filter process Surface water 1,000 15,000 3,780 Conventional process treatment Piped water supply 100 8,000 500 Systems utilize sources of arsenic-safe groundwater 1,000 40,000 800 Source: Based on GOB (2002).

Technology for Remediation and Disposal of Arsenic

121

Table 7 Comparison of operating and maintenance costs, water production/unit, and technical life of each technology used to secure water safe to drink and use Cost of Technology Tech Annual capital Operation and Water life recovery, $US maintenance cost/ production, water/m3, $US annum, $US m3 Alternative water supply Rainwater 15 30 5 16.4 2.134 harvesting Deep tube well 20 120 4 820 0.151 4,500 0.028 Pond sand filter 15 117 15 820 0.161 2,000 0.066 Dug/ring well 25 102 3 410 0.256 1,456 0.072a Conventional 20 2,008 3,780 16,400 0.353 treatment Piped water 15 5,872 800 16,400 0.407 73,000 0.091 Arsenic treatment (no. households) Shapla 6 0.9 11 16.4 0.73 SONO 45-25 5 3 1 16.4 0.24 Read-F 5 1.2 29 16.4 1.84 Magic Alkan 4 3.2 36 16.4 2.39 Bucket treatment 3 3 25 16.4 1.70 a At full development potential of the system. Source: Based on Ahmed (2004); The World Bank (2005).

arsenic-free water at all locations. Dug/ring wells can provide water at moderate cost, but whether desired water quality can be maintained is unclear. Piped water can be provided, but at a higher economic cost, though convenience and health benefits are quite high. An important consideration for piped water is that an increase in the local household density reduces unit cost to any single household. A study by the World Bank suggests that the cost of having a piped water supply will be lower than other options for a medium village of 500 households (The World Bank 2005). The cost of installing RWHs, at the household level, is very high. Installation of similar community-scale RWHs may be cheaper, although effective management of such systems may be difficult. The capital recovery/amortization factor for the computation of annual capital recovery in Table 7 has been calculated using the following formula: Capital Recovery=Amortization Factor ¼ where i ¼ interest rate and N ¼ number of years.

ð1 þ iÞN ðð1 þ iÞN  1Þ=i

122

P. Visoottiviseth, F. Ahmed

Verification of technologies in Bangladesh shows that their performance is dependent on pH and the presence of phosphate and silica in natural groundwater. Most of these technologies fall short of meeting the treatment capacity promised by sellers. A reduction in rated capacity increases the unit cost of water treatment.

B

Technology Verification and Validation

Much development work has transpired during the past 5 yr to meet the demand for arsenic mitigation technology. Appropriate validation of the performance of these technologies is needed to help buyers select the one that meets their requirements. The US EPA has developed protocols for validation of arsenic treatment technologies and arsenic field test kits as part of the Environmental Technology Verification (ETV) Program. The protocols are in use for validation of technologies in collaboration with the ETV program in Canada and at Battelle Laboratories in the U.S. (USEPA 2002, 2003). The WHO has developed generic validation protocols for adoption in South East Asia regional countries (WHO 2003). The Bangladesh Council for Scientific and Industrial Research (BCSIR) has been given the responsibility for verification and validation of arsenic removal technologies under the Bangladesh Environmental Technology Verification–Support to Arsenic Mitigation (BETV-SAM) Program. Sono, Read-F, Sidco, and Alcan filters have been verified and provisionally certified for deployment in Bangladesh, and some are in the process of verification. Other arsenic removal technologies are undergoing verification in Bangladesh, presently.

XIII

Disposal of Generated Arsenic Waste

Arsenic removal units produce a variety of arsenic-rich solids and semisolids, such as arsenic-saturated hydrous ferric or aluminum oxides and other filter media. Regeneration of activated alumina and ion-exchange resins results in various liquid wastes that may be acidic, caustic, saline, and/or too arsenic rich for easy disposal. Hence, disposal of sludge, saturated media, and liquid wastes rich in arsenic is of environmental concern. Large treatment plants must discharge contaminated brine streams, resulting from the RO/NF technologies, into large bodies of water. Inland treatment plants would either pretreat waste prior to discharge, or would discharge to a sanitary sewer. Discharge to sanitary sewers may also require pretreatment to remove high arsenic levels. The waste stream produced by ion exchange (IE)/activated alumina (AA) technologies contains highly concentrated brine with high total dissolved solids (TDS). These brine streams may require pretreatment before discharge to either a receiving body of water or sanitary sewer. Hazardous wastes are often blended into stable waste or engineering materials such as glass, brick, concrete, or cement block. There is a possibility of air or water pollution downwind or downstream from kilns burning arsenic-contaminated

Technology for Remediation and Disposal of Arsenic

123

sludge. In Hungary, experiments showed that some 30% of arsenic in coagulated sludge was lost to the atmosphere in this way (Johnston et al. 2001). Sludge or spent filter media with low arsenic content can be disposed of in landfills without significant increase in the background concentration of arsenic. Wastes with high concentrations of arsenic may need solidification or confinement before final disposal. Regeneration of AA columns results in a toxic waste containing very high concentrations of soluble arsenic. The effluent from regenerating AA columns contains acid and caustic rinses, which can be mixed and disposed of on a prepared bed of cow dung in shallow holes dug into the earth. The microorganisms in cow dung transform the arsenic to gaseous arsine, which is released into the surrounding air. The Toxic Characteristic Leaching Procedure (TCLP) test was developed by the USEPA to identify wastes likely to be hazardous for disposal in a landfill. The permissible level for TCLP leachate is generally 100 times higher than the maximum contaminant level (MCL) in drinking water. The TCLP test was conducted on different wastes collected from arsenic treatment units and materials in Bangladesh (Eriksen-Hamel and Zinia 2001; Ali et al. 2003). It has been observed that, in almost all cases, amounts of arsenic leached from such wastes were very small. Arsenic leaching testing was conducted at Bangladesh University of Engineering using different extraction fluids. The results show that for all extractants, arsenic concentration in the column effluents were initially very high, but afterward dropped sharply (Ali et al. 2003). Several researchers also conducted TCLP tests on sludge resulting from arsenic removal by coagulation using aluminium and ferric salts. Results indicated that arsenic content in leachate ranged between 0.009 and 1.5 mg/L (Brewster 1992; Chen et al. 1999). Such levels are well below those requiring classification as hazardous wastes. It appears that most sludges would not be considered as hazardous, even if the WHO guideline value of 0.01 mg/L for arsenic in drinking water were used. Scientists who work with arsenic mitigation have long recognized the high capacity of the sea as a sink for arsenic containments. The arsenic in the sea is absorbed mostly by marine algae and marine animals. Inorganic arsenic in marine waters is transformed into organic arsenic compounds by marine fauna and flora. Seaweeds, as well as seaweed-eating animals, contain high concentrations of arsenosugars whereas other marine filter-feeding animals, such as shrimps and lobsters, contain high concentrations of arsenocholine. These organic arsenicals are considered nontoxic and are quite safe for marine animals. Phytoremediation studies have demonstrated that organic arsenic residues accumulated in plants, even at high concentrations, are sufficiently low in toxicity to allow them to be disposed in the sea.

Summary Groundwater contaminated with arsenic must be treated to meet stringent drinking water standards or guideline values. In recent years, several reliable, cost-effective, and sustainable treatment technologies have been developed, although improve-

124

P. Visoottiviseth, F. Ahmed

ments will continue to emerge as work continues. All treatment technologies work by concentrating arsenic at some stage of treatment. Large-scale use of arsenic removal systems generates arsenic-rich treatment wastes, and indiscriminate disposal of these sizable wastes may lead to environmental pollution. Safe disposal of arsenic-rich media is a growing environmental concern that needs to be addressed. For the developing world, arsenic-contaminated water requires some form of treatment to be sufficiently safe for consumption by local populations. Such treatment is particularly important where arsenic [particularly as As(III)] levels in raw water exceed 200 mg/L. At this level and above, >95% removal efficiency is required to produce water that meets international standards, an unlikely result in many locations. Alternative sources for securing safe water may also include rainwater harvesting, use of uncontaminated (filtered) surface waters, and water extraction from new deep tube wells and dug wells. There are disadvantages attendant to using these alternative water sources. For example, rainwater has few mineral salts and is subject to contamination from air pollution or by microbes, including pathogens. Similarly, surface waters, e.g., pond waters, or water from dug wells may require purification before use. Excessive pumping from deep tube wells may lower the water table sufficiently to allow entry of arsenic-contaminated waters from shallower horizons. Bioremediation and phytoremediation are more suitable to developing countries where sunlight is plentiful. In such countries, plant biodiversity is also great and may allow identification of plants suitable for bioremediation. In addition to removing arsenic from water, phytoremediation can also provide economic benefit to the people who apply the methods.

References Ahmed MF (2003) Treatment of arsenic-contaminated water. In: Ahmed MF (ed) Arsenic Contamination: Bangladesh Perspective, 2nd Ed. ITN-Bangladesh Centre for Water Supply and Waste Management, BUET, pp 354–403. Ahmed MF (2004) Cost of water supply options for arsenic mitigation. In: People-Centred Approaches to Water and Environmental Sanitation. Preprint, 30th WEDC International Conference, Vientiane, Lao, PDR, pp 600–603. Ahmed MF, Rahaman MM (2000) Water Supply and Sanitation: Low Income Urban Communities. International Training Network (ITN) Centre, BUET. Ahmed MF, Jalil MA, Ali MA, Hossain MD, Badruzzaman ABM (2000) An overview of arsenic removal technologies in BUET. In: Ahmed MF (ed) Bangladesh Environment-2000. Bangladesh Poribesh Andolon, pp 177–188. Ahmed MF, Ashraf AM, Adeel Z (2001) Technologies for Arsenic Removal from Drinking Water. A compilation of papers presented at the International Workshop on Technologies for Arsenic Removal from Drinking Water, Bangladesh University of Engineering and Technology, Dhaka, Bangladesh and the United Nations University, Tokyo. http://www.unu.edu/env/arsenic/BUETWorkshop.htm. Aksorn E, Visoottiviseth P (2004) Selection of suitable emergent plants for removal of arsenic from arsenic contaminated water. Sci Asia 30:105–113. Ali MA, Badruzzaman ABM, Jalil MA, Ahmed MF, Al-Masud A, Kamruzzaman M, Rahman AR (2003) Fate of arsenic in wastes generated from arsenic removal units. In: Ahmed MF, Ashraf,

Technology for Remediation and Disposal of Arsenic

125

Zafar A (eds) Fate of Arsenic in the Environment. Bangladesh University of Engineering and Technology and United Nations University, pp 147–160. Alkorta I, Hernandez-Allica J, Garbisu C (2004) Plants against the global epidemic of arsenic poisoning. Environ Int 30:949–951. Bajpai S, Chaudhuri M (1999) Removal of arsenic from groundwater by manganese dioxidecoated sand. J Environ Eng ASCE 125(8):782–784. BAMWSP, DFID, and Water Aid Bangladesh (2001) Rapid Assessment of Household Level Arsenic Removal Technologies: Phase-I and Phase-II. Final Report, W.S. Atkins International Limited. BGS and DPHE (2001) Arsenic contamination of groundwater in Bangladesh. In: Kinniburgh DG, Smedley PL (eds) British Geological Survey. Technical Report, WC/00/19, 4 vols. British Geological Survey, Keyworth. http://www.bgs.ac.uk/arsenic/Bangladesh/Reports/ChapWorldReview. Bissen M, Frimmel FH (2003) Arsenic—a review. Part II: Oxidation of arsenic and its removal in water treatment. Acta Hydroch Hydrob 31:97–107. Brewster MD (1992) Removing arsenic from contaminated wastewater. Water Environ Technol 4:54–57. Chen HW, Frey MM, Clifford D, McNeill LS, Edwards M (1999) Arsenic treatment considerations. J Am Water Works Assoc 91:74–85. Cheng RC, Wang CC, Beuhler MD (1994) Enhanced coagulation for arsenic removal. J Am Water Works Assoc 86:79–90. Chintakovid W, Visoottiviseth P, Khokiattiwong S, Lauengsuchonkul S (2007) Potential of the hybrid marigolds for arsenic phytoremediation and income generation of remediators in Ron Phibun District, Thailand. Chemosphere 70:1532–1537. Clifford DA (1990) Ion exchange and inorganic adsorption In: Pontius F (ed) Water Quality and Treatment. American Water Works Association, McGraw Hill, New York. Clifford DA, Ghurye AG, Tripp AR (1998) Development of a novel anion exchange process for arsenic removal from water. In: Proceedings of the 3rd International Conference on Arsenic Exposure and Health Effects. Society of Environmental Geochemistry and Health/EPA/ERPI, San Diego, CA, 12–15 July. Dahi E, Liang Q (1998) Arsenic removal in hand pump connected iron removal plants in Noakhali, Bangladesh. Presented at International Conference on Arsenic Pollution of Ground Water in Bangladesh: Causes, Effect and Remedies, Dhaka, 8–12 February. Dambies L (2004) Existing and prospective sorption technologies for the removal of arsenic in water. Sep Sci Technol 39:603–627. Das D, Chatterjee A, Samanta G, Chowdhury TR, Mendal BK, Dhar K, Chanda CR, Lodh D, Chowdhury PP, Basu GK, Biswas BK, Chowdhury UK, Rahman MM, Paul K, Chakraborti D (2004) A simple household device to remove arsenic from groundwater and two years performance report of arsenic removal plant for treating groundwater with community participation. School of Environmental Studies, Jadavpur, Culcutta, India. Daus B, Wennrich R, Weiss H (2004) Sorption materials for arsenic removal from water: a comparative study. Water Res 38:2948–2954. Dikshit K, Saha CJ, Bandyopadhyay M (2001) Comparative studies for selection of technologies for arsenic removal from drinking water. In: Ahmed FM, Ashraf Ali M, Adeel Z (eds) Technologies for Arsenic Removal from Drinking Water. A compilation of papers presented at the International Workshop on Technologies for Arsenic Removal from Drinking Water. Bangladesh University of Engineering and Technology, Dhaka, Bangladesh and the United Nations University, Tokyo. Driehaus W, Jekel M, Hildebrandt U (1998) Granular ferric hydroxide– a new adsorbent for removal of arsenic from natural water. J Water Supply Res Technol–Aqua 47:30–35. Edwards M (1994) Chemistry of arsenic removal during coagulation and Fe-Mn oxidation. J Am Water Works Assoc 86:64–78.

126

P. Visoottiviseth, F. Ahmed

Elizalde-Gonzales MP, Mattusch J, Einicke WD, Wennrich R (2001) Sorption on natural solids for arsenic removal. Chem Eng J 81:187–195. Eriksen-Hamel N, Zinia KN (2001) A study of arsenic treatment technologies and leaching characteristics of arsenic contaminated sludge. In: Ahmed MF, Ashraf M, Zafar A (eds) Technologies for Arsenic Removal from Drinking Water. Bangladesh University of Engineering & Technology (BUET) and United Nation University (UNU), pp 207–213 (http://www. unu.edu/env/arsenic/Proceedings.htm). Genter RB (1996) Ecotoxicology of inorganic chemical stress to algae. In: Stevenson RJ, ML Bothwell ML, Lowe RL (eds) Algal Ecology: Freshwater Benthic Ecosystems. Academic Press, San Diego, CA, pp 403–468. GOB (2002) Ahmed MF, Ahmed CM (eds) Arsenic Mitigation in Bangladesh. Local Government Division, Ministry of Local Government Rural Development and Cooperatives, Government of Bangladesh. Hering JG, Chen PY, Wilkie JA, Elimelech M, Liang S (1996) Arsenic removal by ferric chloride. J Am Water Works Assoc 88:155–167. Hering JG, Chen P, Wilkie JA, Elimelech M (1997) Arsenic removal from drinking water during coagulation. J Environ Eng ASCE 123:800–807. Huang JW, Poynton CY, Kochian LV, Elless MP (2004) Phytofiltration of arsenic from drinking water using arsenic-hyperaccumulating ferns. Environ Sci Technol 38:3412–3417. Hussan A (2003) Development and Deployment of Arsenic Filter for Groundwater. A pre-proposal of a three page narrative submitted for consideration by the Grainger Challenge. http://www. nae.edu/nae/grainger.nsf/weblinks/MKEZ-6XYR4U?OpenDocument. Jekel MR (1994) Removal of arsenic in drinking water treatment. In: Nriagu JO (ed) Arsenic in the Environment. Part 1: Cycling and Characterization. John Wiley & Sons Inc., New York, pp 119–132. Jiang JQ (2001) Removing arsenic from groundwater for the developing world: a review. Water Sci Technol 44:89–98. Johnston RH, Heijnen H, Wurzel P (2001) Chapter 6: Safe Water Technology. Final Draft. Arsenic in Drinking Water. WHO Drinking Water Series. http://www.who.int/watersanitationhealth/ dwq/en/arsenicun6.pdf. Joshi A, Chaudhury M (1996) Removal of arsenic from groundwater by iron-oxide-coated sand. ASCE J Environ Eng 122(8):769–771. Kartinen EO, Martin CJ (1995) An overview of arsenic removal processes. J Desalination 103:79–88. Khair A (2000) Factors responsible for the presence of arsenic in groundwater: Bangladesh context. In: Ahmed MF (ed) Bangladesh Environment–2000. Bangladesh Poribesh Andolon, pp 198–209. Khair A, Azad AK, Paul GN (1999) Arsenic removal from drinking water by low-cost materials. Presentation at the 1999 Dhaka Conference, Dhaka, Bangladesh. Khan AH, Rasul SB, Munir AKM, Alauddin M, Habibuddowlah M, Hussam A (2000) On two simple arsenic removal methods for groundwater of Bangladesh. In: Ahmed MF (ed) Bangladesh Environment–2000. Poribesh Andolon, Bangladesh, pp 151–173. Kohnhorst A, Paul P (2000) Testing simple arsenic removal methods. In: Water, Sanitation and Hygiene: Challenges of the Millennium. Pre-prints of the 26th WEDC Conference, Dhaka, Bangladesh, pp 177–181. Ma QL, Komar KM, Tu C, Zhang W, Cai Y, Kennelley E (2001) A fern that hyperaccumulates arsenic. Nature (Lond) 409:579. Maeda S, Ohki A (1998) Bioaccumulation and biotransformation of arsenic, antimony, and bismuth compounds by freshwater algae. In: Wong Y-S, Tam NFY (eds) Wastewater Treatment with Algae. Springer-Verlag and Landes Bioscience, New York, pp 73–92. Maeda S, Sakaguchi T (1990) Accumulation and detoxification of toxic metal elements by algae. In: Akatsuka I (ed) Introduction to Applied Phycology. SPB Academic Publishing, The Hague.

Technology for Remediation and Disposal of Arsenic

127

Manning B, Goldberg S (1996) Modelling competitive adsorption of arsenate with phosphate and molybdate on oxide minerals. Soil Sci Soc Am J 60:121–133. Matthess G (1981) In situ treatment of arsenic contaminated groundwater. Sci Total Environ 21:99–104. McNeill LS, Edward M (1997) Arsenic removal during precipitative softening. J Environ Eng 123:453–460. Meng XG, Korfiatis GP (2001) Removal of arsenic from Bangladesh well water using household filtration system. In: Ahmed MF, Ashraf M, Zafar A (eds) Technologies for Arsenic Removal from Drinking Water. Bangaldesh University of Engineering & Technology (BUET) and United Nation University (UNU), pp 121–130 (http://www.unu.edu/env/arsenic/Proceedings. htm). Meng XG, Bang SB, Korfiatis GP (2000) Effects of silicate, sulfate, and carbonate on arsenic removal by ferric chloride. Water Res 34:1255–1261. Mok WM, Wai CM (1994) Mobilization of arsenic in contaminated river water. In: Nriagu JO (ed) Arsenic in the Environment. John Wiley & Sons Inc., New York, pp 99–117. Munir AKM, Rasul SB, Habibuddowla M, Alauddin M, Hussam A, Khan AH (2001) Evaluation of performance of Sono 3-kolsi filter for arsenic removal from groundwater using zero valent iron through laboratory and field studies. In: Ahmed MF, Ashraf M, Zafar A (eds) Technologies for Arsenic Removal from Drinking Water. Bangladesh University of Engineering & Technology (BUET) and United Nation University (UNU), pp 171–189 (http://www.unu.edu/ env/arsenic/Proceedings.htm)). Murcott S (1999) Appropriate Remediation Technologies for Arsenic-Contaminated Wells in Bangladesh. Arsenic in Bangladesh Ground Water. February 27–28, Wagner College, Staten Island, New York. Available on line: http://phys4.harvard.edu/wilson/arsenic/remediation/ arsenic_removal/murcott.html. Murcott S (2000) A comprehensive review of low-cost, well-water treatment technologies for arsenic removal. http://phys4.harvard.edu/wilson/murcott2.html. Najm I, Trussell RR (1999) New and emerging drinking water treatment technologies. In: Identifying Future Drinking Water Contaminants. National Academy Press, Washington DC. pp 33–42. Nikolaidis NP, Lackovic J (1998) Arsenic remediation technology-AsRT. Presented at International Conference on Arsenic Pollution of Ground Water in Bangladesh: Causes, Effect and Remedies, Dhaka, 8–12 February. Nikolaidis NP, Dobbs GM, Lackovic JA (2003) Arsenic removal by zero-valent iron: field, laboratory and modelling studies. Water Res 37:1417–1425. Oh JI, Yamamoto KK, Kitawaki H, Nakao S, Sugawara T, Rahaman MM, Rahaman MH (2000) Application of low-pressure nanofiltration coupled with a bicycle pump for the treatment of arsenic-contaminated groundwater. J Desalination 132:307–314. Pal BN (2001) Granular ferric hydroxide for elimination of arsenic from drinking water. In: Ahmed MF, Ashraf M, Zafar A (eds) Technologies for Arsenic Removal from Drinking Water. Bangaldesh University of Engineering & Technology (BUET) and United Nation University (UNU), pp 59–68 (http://www.unu.edu/env/arsenic/Proceedings.htm). Pierce ML, Moore CB (1982) Adsorption of arsenite and arsenate on amorphous iron hydroxide. Water Resources 16:1247–1253. Pontius FW (ed) (1990) Water Quality Treatment: A Handbook of Community Water Supplies. American Water Works Association. McGraw-Hill, New York. Rahman MH, Magara Y (2002) The results of the pilot plant study for arsenic removal. Preprint of the symposium on the Study of Arsenic Mitigation in Chapai Nawabgonj, Bangladesh, Dhaka, March 12–17. Ramaswami A, Tawachsupa S, Isleyen M (2001) Batch mixed iron treatment of high arsenic waters. Water Res 35:4474–4479. Saha JC, Dikshit K, Bandyopadhyay M (2001) Comparative studies for selection of technologies for arsenic removal from drinking water. In: Ahmed MF, Ashraf Ali M., Adeel Z (eds)

128

P. Visoottiviseth, F. Ahmed

Technologies for Arsenic Removal from Drinking Water. A compilation of papers presented at the International Workshop on Technologies for Arsenic Removal from Drinking Water. Bangladesh University of Engineering and Technology, Dhaka, Bangladesh and the United Nations University, Tokyo. Sarkar A, Chowdhury UK, Rahman MH, Chowdhury TR (2000) Bucket treatment unit for arsenic removal. In: Water, Sanitation and Hygiene: Challenges of the Millennium. Preprints of the 26 WEDC Conference, Dhaka, Bangladesh, pp 308–310. Senapati K, Alam I (2001) Apyron arsenic treatment unit: reliable technology for arsenic safe water. In: Ahmed MF, Ashraf M, Zafar A (eds) Technologies for Arsenic Removal from Drinking Water. Bangaldesh University of Engineering & Technology (BUET) and United Nation University (UNU), pp 146–157 (http://www.unu.edu/env/arsenic/Proceedings.htm). Shen YS (1973) Study of arsenic removal from drinking water. J Am Water Works Assoc 65:543–548. Sorg TJ, Logsdon GS (1978) Treatment technology to meet the interim primary drinking water regulations for inorganics: Part 2. J Am Water Works Assoc 70:379–393. Su C, Puls RW (2001) Arsenate and arsenite removal by zerovalent iron: kinetics, redox transformation and implications for in situ groundwater remediation. Environ Sci Technol 35:1487–1492. Su C, Puls RW (2003) In situ remediation of arsenic in simulated groundwater using zerovalent iron: laboratory column tests on combined effects of phosphate and silicate. Environ Sci Technol 37:2582–2587. The World Bank (2005) Towards A More Effective Operational Response: Arsenic Contamination of Groundwater in South and East Asian Countries, vol II. Technical Report No. 31303, Paper 3. The World Bank, Washington, DC, pp 165–207. Tu S, Ma LQ, Fayiga AO, Zillioux EJ (2004) Phytoremediation of arsenic-contaminated groundwater by the arsenic hyperaccumulating fern Pteris vittata. Int J Phytoremed 6:35–47. US EPA (1999) Technologies and costs for removal of arsenic from drinking water. EPA 815-P01-001. U.S. Environmental Protection Agency, Washington, DC. US EPA (2002) Arsenic in drinking water-treatment technologies: removal. http://www.epa.gov/ ogwdw000/ars/treat.html. US EPA (2003) Environmental Technology Verification Program (www.epa.govt/etv/verifications/vcenter1-21.html). Viraghavan T, Subramanian KS, Aruldoss JA (1999) Arsenic in dinking water: problems and solutions. Water Sci Technol 40:69–76. Visoottiviseth P, Lauengsuchonkul S (2004) Removal of arsenic from water by the isolated Chlorella vulgaris. LUCED International workshop on management of resources in urban and industries (MANURE) and chemical assessment of the environment (CHASE), 15th–17th June, Suratthani, Thailand. Visoottiviseth P, Francesconi K, Sridokchan W (2002) The potential of Thai indigenous plant species for the phytoremediation of arsenic contaminated land. J Environ Pollut 118:453–461. Vu KB, Kaminski MD, Nunez L (2003) Review of Arsenic Removal Technologies for Contaminated Groundwaters. Argonne National Laboratory, Argonne, IL. The University of Chicago. http://www.ipd.anl.gov/anlpubs/2003/05/46522.pdf. Wanichapichart P (2005) Drinking Water Production System for Arsenic Removal: Case Study in Moo 2, Ron Piboon Sub-district, Ron Piboon District, Nakhon Sri Thamarat. Thai Environ Eng J 21(1):121–129. Wegelin M, Gechter D, Hug S, Mahmud A, Motaleb A (2000) SORAS–a simple arsenic removal process. (http://phys4.harvard.edu/wilson/arsenic_project_remediation_technology.html). World Health Organization (2003) United Nations Synthesis Report on Arsenic in Drinking Water. http://www.who.int/water_sanitation_health/dwq/arsenic3/en/. Young E (1996) Cleaning up arsenic and old waste. New Scientist magazine. 14 December: p 22.

A Multi-Criteria Approach for Assessing Options to Remediate Arsenic in Drinking Water Bryan Ellis and Hemda Garelick

I II III IV

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 Multi-Criteria Approaches (MCA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 Defining Criteria and Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Defining Arsenic Treatment Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 A Treatment Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 B Source Mitigation Technology Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 V MCA Performance Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 A Benchmarking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 B Utility Scoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 VI Source Exposure Vector Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 A Raw Surface Water KPI (Key Performance Indicators) . . . . . . . . . . . . . . . . . . . . 138 B Groundwater KPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 VII Health Risk Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 A Toxicological KPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 B Infectious Disease Risk KPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 C Other Chemical Pollution KPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 VIII Cost Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 A Capital Cost KPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 B Operation and Maintenance Cost KPI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 IX Community Acceptance Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 X Technical and Organizational Skill-Base Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 XI Community Location Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 XII Aggregating KPI Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 A Factor Weighting and Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 B Exclusion Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

B. Ellis and H. Garelick Urban Pollution Research Centre, Middlesex University, The Burroughs, London, NW4 4BT, UK.

D.M. Whitacre (ed.), Reviews of Environmental Contamination Volume 197. doi: 10.1007/978-0-387-79284-2_5, # Springer Science þ Business Media, LLC 2008

129

130

B. Ellis, H. Garelick

I Introduction Arsenic exposure threatens millions of people in India, Bangladesh, Thailand, China, and Mexico who drink groundwater contaminated at levels (0.86–1.86 mg/L) far in excess of the 10 mg/L World Health Organization (WHO) maximum permissible level. To address this serious problem, the Division of Chemistry and the Environment of the International Union of Pure and Applied Chemists (IUPAC) established a project designed to provide a simple and practical guide for decision making on arsenic remediation technologies (IUPAC 2003). When selecting technologies appropriate to remediate arsenic, one must consider aspects other than water quality; microbiological contamination, cost, availability of technical expertise to achieve local remediation, and social acceptance factors must also be factored into the remediation equation. Consideration of such multiple criteria involves many stakeholders at a plethora of institutions and organizational levels. Successful arsenic remediation benefits from, and often utilizes, best practices technology. Identifying and selecting best practices requires careful evaluation of environmental, technological, and economic factors, as well as regional social values; frequently, there is also an interdisciplinary and interinstitutional context to be considered (Wenzel 2001). Any procedure designed to resolve arsenic remediation must address many complex problems in a manner that is comprehensible to, and verifiable by, stakeholders. Multi-criteria analysis (MCA) provides such an approach in that it facilitates optimal decision making when divergent-minded stakeholders are involved. The methodology outlined in this review was developed to (1) address complexity arising from contention among multiple stakeholders, and (2) apply multi-criteria to achieve successful arsenic remediation.

II

Multi-Criteria Approaches (MCA)

A structured multi-criteria, multi-objective methodology was proposed for evaluation of sustainable water resource systems in the 1998 UNESCO (United Nations Educational, Scientific and Cultural Organization) International Hydrological Program Report (ASCE and UNESCO 1998). When constructing water resource systems, this report recommended a holistic approach to decision making, a main feature of which is sustainability (economic, social, and environmental). More recently, others have presented approaches for management of urban runoff that also utilize sustainability criteria to define constraints on treatment technology best practices (Ashley et al. 2001; Ellis et al. 2004, 2005). An approach similar to these can be used to identify strategies for arsenic treatment that allows end-users to rank or select ‘‘best solutions,’’ after considering limitations and stakeholder perspectives. MCA is a decision-making tool originally developed to help evaluate options that rely on values that are not easily assignable. An example is cost–benefit analyses, wherein a range of influencing factors (or defining criteria) important to

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

131

stakeholders are considered. MCA also allows users to consistently process large amounts of complex information and enables stakeholders to see the ‘‘bigger picture,’’ thereby facilitating discussions and increasing transparency of the decision-making process. For each criterion, MCA procedures utilize a performance matrix (consequence table) in which options for arsenic remediation are presented (and described) in columns, and performance values for options are presented in rows. Various forms of data may be entered into the matrix. Data may be numerical, semiquantitative, and/or descriptive, with performance of each option being aggregated against all the criteria; this aggregation is then used as the basis for eventual comparison. Examination of matrix inputs may enable decision makers to more clearly see that one option consistently performs better or worse than others for any particular criterion. Moreover, in more detailed analyses, performance data in the matrix are converted into numerical values through use of preference scale scoring and criterion weighting. To preserve validity, it is essential that the preferences expressed are mutually exclusive. If such criteria are mutually independent, and if statistical uncertainty is not formally built into the MCA, then a simple linear additive model can describe the results (ODPM 2003). This linear, nonparametric model combines the product of the value score for each criterion, with the weight of the criterion to achieve a weighted score. The approach highlights the ‘‘extremes’’ and achieves higher discriminatory power. However, the linear additive method is the preferred MCA approach for selecting arsenic treatment methodology, and it has a record of providing robust and effective support to decision makers. Figure 1 is a flow chart that illustrates the key stages and progression of the MCA procedure. In this procedure, objectives and available technologies are first defined, after which development of criteria and specification of performance can proceed. Specification enables options to be ranked, and risk analysis of alternative remediation options to be undertaken, before final selection.

III

Defining Criteria and Indicators

The primary components of any decision-making process include defining performance criteria and their relationship to the multi-criteria decision-making parameters; criteria are here defined as the major factors on which a judgment or decision is based. Table 1 provides a survey of primary criteria essential in defining key performance indicators (KPIs) for arsenic treatment/mitigation. KPIs provide means by which the performance of individual criteria (e.g., exposure to microbial contamination as a surrogate for infectious disease/health risk) are evaluated. More limited criteria were used by Jakariya et al. (2003) to create their matrix of arsenic treatment options for two ‘‘upazila’’ (or district) communities in Bangladesh. Similar restricted criteria were used by Boerschke and Stewart (2001) for developing arsenic technology verification. However, a broader set of KPIs were considered by the Ontario Centre for Environmental Technology

132

B. Ellis, H. Garelick STAGE 1 DEFINE OBJECTIVES FOR DECISION-MAKING

Water quantity Amenity / Ecology

STAGE 2 GENERATE MITIGATION / REMEDIATION OPTIONS

Alternative water sources; treatment options

STAGE 3 DEVELOP SUSTAINABILITY CRITERIA, INDICATORS AND BENCHMARKS

Generic listing of criteria and indicators for each option

STAGE 4 COLLECT DATA / INFORMATION FOR MULTI-CRITERIA ANALYSIS

Spread sheet compilation Laboratory

Field Literature

Modelling

STAGE 5 MATCH DATA WITH CRITERIA AND INDICATORS FOR EACH OPTION More Indicators and / or benchmarks needed?

STAGE 6 RANK OPTIONS

Spread sheet analysis for options

STAGE 7 SENSITIVITY AND / OR RISK ANALYSIS OF OPTIONS

STAGE 8 APPLY SELECTED / PREFERRED OPTION

STAGE 9 POST-PROJECT MONITORING AND / OR ACCREDITATION EVALUATION

Fig. 1 Multi-criteria analysis flow chart

Table 1 Criteria and key performance indicators (KPIs) used in Multi-Criteria Analysis (MCA) for the evaluation of arsenic mitigation options Criteria (CRI) KPIs Source/exposure vector Groundwater supply Surface waters Health risk Toxicology (behavior/form/NOELa, etc.) Infectious disease risk Other chemical pollution Cost Capital cost Operation and maintenance (O & M) costs Community acceptance Technology acceptance Technical and organizational-skill base Local competencies, training, etc. Community location Urban Rural a NOEL = no observable effect level.

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

133

Advancement (OCETA 2001) in their protocols for verifying arsenic treatment options. The KPI scoring, similar to those of Jakariya et al. (2003), used grades 1–5, without weighting; OCETA stressed that grade aggregation is inappropriate to their method. Loetscher and Keller (2002), in their multi-criteria sanitation model, used a much more limited ‘‘triple-bottom-line’’ approach, based on technical, institutional, and sociocultural criteria. Another important determinant in selecting criteria and deciding what is technologically appropriate depends on whether treatment is to occur in urban or rural locations. The KPIs shown in Table 1 constitute a preliminary, but sufficiently flexible, list that can be refined to meet needs of differing organizations, regulations, and end-users. These KPIs can help discriminate among relevant and appropriate properties for each listed criterion. The final decision may be suboptimal, because it is driven, or constrained, by specific local, social, or financial considerations.

IV

Defining Arsenic Treatment Options

There are two mitigation options against which KPIs are evaluated. The first option consists of conventional and advanced physicochemical treatment (or remediation) technologies, and the second comprises a smaller group of source ‘‘technologies’’ that utilize natural processes to mitigate arsenic contamination. A full description of various arsenic treatment technologies can be obtained from the Massachusetts Institute of Technology (2001) website.

A

Treatment Options

Chemical Treatments Precipitation, Coprecipitation, Adsorption, and Coagulation These treatment techniques normally employ metal salts and/or lime and are frequently followed by filtration. Included are household treatment systems, such as Bucket Treatment Units (BTU), ‘‘three- and four-jug’’ and ‘‘tea-bag filters, ‘‘fill and draw’’ units, and free-standing iron-arsenic (Fe-As) removal plants.

Ion-Exchange/Reverse Osmosis This approach is based on use of a synthetic regenerative resin (or matrix) of a cross-linked polymer to remove specific undesirable cations or anions from water.

134

B. Ellis, H. Garelick

The method includes household cartridge filter units filled with ion-exchange resin or other sorptive media. Physical Treatments l

l

l

l

Filtration includes various sorptive filtration techniques that utilize activated alumina, carbon, iron- and manganese-coated sand, kaolinite or bauxite clay, and ferric, titanium, and silicon oxides as well as other natural and synthetic media. Other filtration techniques included in this category are small householdscale units that utilize iron-coated brick chips and/or wood coke (Shapla and Sono 3-Kalshi filters), ‘‘candle’’ filters made of porous materials (SAFI filters), and other filters utilizing locally sourced iron-coated sand, brick chips, iron filings, cellulose, etc. (such as the Garnet, Chari, Adarsh, Bijoypur filters as described by Visoottiviseth and Ahmed 2008, in this volume). Membrane technology (MT) covers a range of synthetic membrane filtration techniques that include low-pressure microfiltration (MF), ultrafiltration (UF), and high-pressure nanofiltration (NF). It also includes water demineralization using MT. Slow sand filters comprise conventional aeration, sedimentation, and filtration techniques that often use upflow filters with capacities up to 3000 m3/d. A range of alternative media have also been used in slow sand filters, including volcanic rock, kimberlite tailings, coconut/rice husks, and peanut shells. Also in this category are pond sand filters, an option for small community treatment; household units can utilize iron-coated sand with slow sand pretreatment to remove excess iron. Solar stills are simple methods for solar oxidation of arsenic in transparent containers/bottles. Their effectiveness is based on ultraviolet radiation catalysis.

Bioremediation These treatment techniques rely on algal (Chlorella vulgaris) remediation and phytoremediation with ferns or, for larger-scale treatments, remediation by artificially constructed wetlands.

Hybrid Technologies This category utilizes a combination of techniques such as conventional sand filtration followed by MT treatment.

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

B

135

Source Mitigation Technology Options

Dug Wells Manually dug or ring wells comprise one of the oldest methods of obtaining water; nearly 1 million exist in Bangladesh alone. Such shallow wells (<10 m deep) are known to have long-term resistance to arsenic contamination because of natural oxidation/precipitation and frequent fresh surface water replenishment. However, such sources are contaminated with bacteria and ammonia and have high turbidity and color. Deep Tube Wells (DTW) Tube wells with depths greater than 120–150 m generally have less than 50 mg/L arsenic, with the majority having levels less than 10 mg/L. The existence of an impermeable layer separating the deep from shallow (contaminated) aquifer layer is a prerequisite for obtaining arsenic-safe water from deep boreholes. Even without an impermeable layer, infiltration recharge is likely to keep the inflowing groundwater arsenic free, even after prolonged abstraction. Rainwater Harvesting Rainwater can be an independent source of good-quality water. Construction of the harvesting apparatus may be accomplished by local craftsman. Once constructed, such harvesting equipment can be operated without any energy cost. However, water quantities are limited by rainfall intensity and availability of sufficient impermeable roof area. Because collected water is largely mineral free, it may cause nutritional deficiencies. Household water collection jars of 2000- to 3000-L capacity have been used in both Bangladesh and Thailand, but bacterial contamination problems have been encountered.

V

MCA Performance Matrix

One can construct a performance matrix from the previously described treatment methods, based on selected criteria (CRI) and the KPIs listed in Table 1. The resultant matrix is presented in Table 2; cells define the performance (i.e., the measure of how well a particular technology option meets its objective) of treatment options. The performance prescribed in any cell is obtained by benchmarking individual KPIs to derive utility scores, which then are aggregated to rank the attractiveness of treatment options. The KPIs and CRI can be weighted, if required, to incorporate specific stakeholder preferences.

Rural

Urban

O&M costs including waste arisings disposal Technology acceptance Local competence

Other chemical pollution Capital cost

Toxicology (Behavior/form) Infectious risk

Raw surface waters

Groundwater supply

Dug Wells 7 4 5.5 3 8.5 5 7 3 2 1 6 2

3.5 1 6 4 6 3 1 1

Rainwater Harvesting 2.5 2 6 4 3.5 3 7.5 4 5 3 9 5

7 4 7 4 2 2

Deep Tube-Well Source 6.5 4 8 4 9 5 9.5 5 8 4 8 3

9 5 3 1 10 5

3 3 5 3 7.5 4 8.5 4 7 4 4 1

8 4 4 3 4 2 3 3

Ion Exchange/ Reverse Osmosis 1 1 1 1 1 1 2 1 9 5 9 5

7 3 9 5 3 1 0 1

7.5 4 9.5 5 4 3 6 3 1 1 8 3

4.5 2 6 4 6 3 1 2

Bioremediation

Physical Treatment

2 2 3 2 2 2 2 1 6 3 9 5

8 4 9 5 9 5 10 5

2 2 4 3 1 1 4 2 4 2 6 2

9 5 5 2 4 2 9 5

Cell values in bold represent KPI grade values (1–5). The other values represent KPI utility scores

TOTAL (%) (sum of score x weight)

Community acceptance Technical skill base Community location

Cost

Health risk

Source exposure vector

a

Chemical Treatment

Precipitation/ Coagulation/ Adsorption

KPI

Membrane Technology

CRI

Solar Still

Table 2 The MCA performance matrix

Filtration 5 3 4.5 3 5.5 3 8.5 4 4 2 8.5 4

8 4 3.5 2 4 2 5 3

Slow Sand Filters 6 4 7 4 4.5 3 8 4 5 3 8 3

7 3 5 3 5 3 7 4

Hybrid Technology 4 3 9 5 10 5 3 1 10 5 3.5 1

9 5 10 5 10 5 10 5

100%

KPI

Weighting

100%

CRI

136 B. Ellis, H. Garelick

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

A

137

Benchmarking

The matrix in Table 2 was created by evaluating and then entering various available mitigating technologies along with the KPIs, which were benchmarked through preference scoring. The essence of this benchmarking is to compare technology mitigation options (in terms of function and/or process) with existing best practices, to ensure that optimum solutions for mitigation are chosen. Benchmarking is best done in collaboration with stakeholders to support the circumstances and processes that underpin superior performance. Benchmarking helps to raise awareness about performance and provides greater insight on relative strengths and weaknesses of options. Benchmarking also helps increase willingness of stakeholders to reach for, and share solutions to, problems common to arsenic mitigation in different regions or with different technologies. It also helps stakeholders gain a broader perspective of the interplay among factors (or enablers) that facilitate implementation of good practices. There can be difficulties in getting agreement on which KPIs are appropriate for deriving a performance score. In the end, key drivers that have the greatest impact on performance are critical to identifying and implementing optimal mitigation technologies.

B

Utility Scoring

The first step in setting benchmark scores is to establish an appropriate scale with justified reference points; normally, the worst performance is allocated a 0 score, and the best is arbitrarily given a score of 100 (or, a 1 to 10 scoring system may be used). Some workers recommend using intervals spanning ranges of 0 to 33, 33 to 66, and 66 to 100 to represent low, mean, and high descriptors, respectively. If such a semiquantitative scale is adopted, the midvalue of the interval range is selected for evaluation, if no other evidence or end-user preference is available. Use of such scoring results in standardized ratings, which have this general relationship: x ¼ f ðwhere f is user input or a default grade valueÞ; such that 0  x  1ðor 10Þ where 0 equals the worst and 1 (or 10) equals the best outcome (or option). Alternatively, the relationship can be expressed as 0 = no and 1 (or 10) = yes; or, 0 = very low and 1 (or 10) = very high, if a qualitative dimension is sought. The various technology options are then scored for performance against each KPI, based on results of published data or the experiences of those who have applied the technology. If data are unavailable or speculative in nature, an acceptable performance score may be derived through collaborative stakeholder discussions or inputs by questionnaire. The derived utility scores for mitigating technologies are then directly recorded into the performance matrix (see Table 2). However, because of the probable

138

B. Ellis, H. Garelick

uncertainty that surrounds assigned utility scores, another alternative is to use a broader rating scale (1–5). If used, such default values are entered into the matrix rather than using the derived utility score. Jakariya et al. (2003), in the preference matrix they developed, used such a coarse grading scale (1–5) to define the performance of arsenic treatment options, although no adequate justification was provided for the grade values they gave. When using such broader scales, rating outcomes tend to be compressed and provided a narrower range when aggregated; the exception is dug wells, which emerge as the preferred option, by far. This situation undoubtedly results from community familiarity with indigenous dug well technology, which is scored highly on every criterion. Although utility scores and grades may appear to be explicit, the method is actually analogous to that of ‘‘fuzzy logic’’ because allocations are essentially estimates. Estimates can be given more credence through a process called ‘‘repeated-measures design,’’ which has been applied to some KPIs, such as costing. Each KPI is inevitably influenced by a range of interactions and competing factors. Among the most important of these are costs, social acceptance, and the chemical and microbiological composition of the arsenic-polluted water source. The dynamic interplay of these factors affects arsenic removal efficiencies and contributes to the uncertainty of outcomes when various control technologies are applied. No attempt has been made here to identify bounds of uncertainty for utility scores, although variability in source data may increase confidence in the methods selected. Both utility scores and grade values are presented in Table 2; grade values are shown in bold font beneath utility scores. Higher utility scores (1–10) and higher grades (1–5) reflect better performance. Thus, 10 and 5, respectively, mark maximum achievable performance. Although scores should be regarded as consistent within each criterion and KPI, the scale is not necessarily linear; i.e., a utility score of 8 is not twice as good as a score of 4. However, it is possible to use a direct rating procedure, based on expert judgment, and thereby provide relative weightings between 0 and 100 for each option of a particular KPI. Scoring consistency is very important when using such a procedure; i.e., a value of 50% should be twice as good as 25%.

VI A

Source Exposure Vector Criteria Raw Surface Water KPI (Key Performance Indicators)

The presence of arsenic in raw water is not readily apparent, because arsenic does not alter the taste, smell, or color of the water. Nonetheless, arsenic poisoning induces neurological damage when consumed at contamination levels greater than 10 mg/L. Raw surface waters comprise an important drinking water source in several countries where arsenic contamination exists, including Mexico, Argentina, the western United States, and South Thailand. Such sources are vulnerable to

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

139

leaching of arsenic from mine tailings and industrial (ore dressing) discharges (e.g., in South Thailand, levels of 165–985 mg/L have been recorded in headwaters of the Ron Phibun catchment). However, such levels are frequently highly localized because of their proximity to a pollution source. Baseline concentrations of arsenic in river waters are in the range of 0.1–0.8 mg/L, rising to 2.0 mg/L when local lithology and groundwater conditions foster higher concentrations (Smedleigh and Kinniburgh 2001; Garelick et al. 2008). River waters in areas of geothermal activity, e.g., the western U.S. and New Zealand, may carry concentrations of 10–70 mg/L, rising to highs of 350 mg/L during low flow periods. Rivers with flow dominated by arsenic-contaminated regional groundwater also carry high concentrations. Significant levels have also been recorded in urban rivers as a result of industrial and/or sewage discharges (Smedleigh and Kinniburgh 2001). Organic forms of arsenic usually exist at low levels in raw surface waters, although they can increase in areas of high bacterial activity. The benchmark selected for the raw surface water KPI reflects the potential for arsenic removal for each treatment option for this source exposure vector. Generally, the more basic and smaller systems are less likely to effectively remove arsenic. Normally, 99% removal is required to achieve the WHO Guideline for Drinking Water Quality (GDWQ) of 10 mg/L, and 90% removal achieves the less strict guideline of 50 mg/L, established in countries such as China, Bangladesh, Vietnam, and, until recently, the United States. Hybrid technologies, including piped water supply from conventional treatment plants, meet the goal of obtaining safe water for consumers and represent the optimum desired endpoint vis-a`-vis utility score. Treatment using hybrid technology justifies allocation of the maximum score (and default grade), as illustrated in Fig. 2. Such technology constitutes a feasible option for clustered rural settlements

10

+++ HT

WHO Guidelines

9

+ Limited removal ++ 50%+ removal +++ 80%+ removal (<0.005)

8

Utility Score

5

7

++ BIO DW

4 ++ SF

6 3

5

+/++ CG/P/A

4 3

2

+ F SS

2 + DTW

1

1

+++ IE MT

++ RH

HT Hybrid Technologies CG/P/A Coagulation/Precipitation/Adsorption IE Ion Exchange MT Membrane Technology BIO Bioremediation SF Sand Filtration F Filtration SS Solar Distillation DTW Deep Tube Well DW Dug Well RH Rainwater Harvesting

0 0

10

20

30

40

50

60

70

80

90

100

Percentage Arsenic Removal

Fig. 2 Plot of arsenic removal utility score and grade values for raw surface waters. The scale for utility scores (horizontal axis) is 1–10, and for grade values (numbers within circles) is 1–5; see text for explanation of scores and grades

140

B. Ellis, H. Garelick

and urban fringes, and received priority for arsenic mitigation in Bangladesh, under a project of the 2004–2009 World Bank Water Supply Program, established to provide piped water to at least 300 villages. However, this hybrid option is both difficult and costly to implement for dispersed populations in rural areas of Bangladesh and West Bengal. Tubewell abstraction from deep aquifers effectively competes with hybrid piped water technologies in mitigating arsenic but is excluded from this KPI because it does not constitute a source of raw surface water. However, shallow dug wells are relatively free from dissolved arsenic, even at locations where deep tube wells may be contaminated. This natural mitigation that occurs in dug wells is at least partly attributable to continued replenishment by rainwater and surface waters, which are ‘‘treated’’ during percolation through the zone of aeration in the subsoil. In addition, freshwater recharges have a diluting effect on background contamination of groundwater in the well. Thus, dug wells are included in the raw surface water KPI and have been assigned an above-average utility score of 6 (or default grade of 4). The coagulation process is inefficient in removing arsenic, and without a final filtration step only has an efficiency of about 30%–35%; hence, coagulation is assigned an average utility score of 4 and a default grade of 3 (see Fig. 2). By comparison, membrane, ion-exchange, and reverse osmosis technologies can remove 90%–99% of both arsenite and arsenate, although synthetic ion-exchange resins sometimes require an initial oxidation step. Therefore, these technologies are assigned high utility scores (>8) and a grade value of 5. Rainwater harvesting is considered to be a raw surface water supply. Although rainwater harvesting possesses natural arsenic mitigation potential, it can be contaminated from other pollutant sources. Wetland phytoremediation (using Typha, Thalia, and other emergent species) has been attempted in South Thailand; results show up to 60% removal of arsenic from surface waters containing initial arsenic concentrations of 300–1000 mg/L. Thus, bioremediation technology options have been assigned a utility score of 6 (and default grade of 4) in the graphical plot (see Fig. 2). Small-scale infiltration and solar distillation treatment units have been assigned low utility scores and grades because they have a low flow rate and serious volume restrictions. Sand filtration is a commonly used option for treating raw surface waters; some 400 units, serving 2000 households per unit, operate in the coastal Bangladesh region. However, they only have average (50%+) efficiency in arsenic removal, and concentrations up to 150 g/L can remain in the ‘‘treated’’ water. The allocation of utility scores to treatment technology options is somewhat subjective and is dependent on individual user experience, preferences, and prejudices; however, the methodology provides for stakeholder discussion to arrive at a scoring consensus. The plot shown in Fig. 2 depicts a linear relationship between percentage arsenic removal and assigned utility scores. Some may argue that a nonlinear plot is more appropriate, with much lower utility scores being assigned to removals below a 50% value. Such scoring would emphasize the nonacceptable performance of mitigation options having low removal potential.

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

B

141

Groundwater KPI

The utility and grade scores assigned to the groundwater KPI are shown in Table 2. Because groundwater arsenic levels may be inherently low, most remediation technologies are likely to yield water substantially cleansed of arsenic. By virtue of restricted flow volumes and retention times, as well as reflushing potential, bioremediation treatment has a higher risk of ‘‘nonacceptable’’ performance than do shallow dug wells (which have a higher potential for contamination). However, bioremediation can convert arsenic to an organic form, thus rendering it less toxic; this benefit is reflected in the final scores allocated to bioremediation compared to dug wells.

VII A

Health Risk Criteria Toxicological KPI

Significant positive correlations exist between the concentration of total watersoluble arsenic and its bioavailability. The potential toxicity of arsenic is dependent on both its total concentration and bioavailability, defined as the potential of a contaminant to bind to a cell. Arsenate [as As(III), As(IV), and As(V)], arsenite methylarsonic acid (MMAA), and dimethyl arsenic (DMAA) acid are all normally found in water bodies; these arsenic entities have differential uptake (ability of a cell to accept a bioavailable species), as follows: As(III) > MMAA > AS(V) > DMAA. As(III) is, by far, the most mobile and toxic species, reducing organism growth at concentrations less than 3 mg/L and causing neurological damage in humans at concentrations as low as 100 mg/L. As much as 90% of a single dose of As(III) can be absorbed from the gastrointestinal tract, in comparison to other forms, which have less efficient uptake. Because oxygen content and prevailing redox conditions control oxidation state, they serve as primary controls on mobility and toxicity of arsenic. Thus, knowing what arsenic species are present is essential to understanding its probable bioavailability. Bioavailability varies substantially between sampling sites and is not equal from site to site when comparing the concentration of total water-soluble arsenic (Price and Pilcher 2005). Experiments show that ageing and sequestration of arsenic occurs in contaminated aquifers and soils, which renders it progressively less bioavailable with time. Site-specific evaluations of bioavailability are needed to produce realistic estimates of exposure and risk if accurate measures of required ‘‘clean-up’’ are to be derived. Options can be semiquantitatively evaluated for their effectiveness in removing pentavalent and trivalent arsenic, as illustrated in Table 3.

142

B. Ellis, H. Garelick

Table 3 Relative potential for removing arsenic species with different treatments Treatment technology Arsenic removal potential As(III) Arsenite As(V) Arsenate l

Coagulation/precipitation Alum Iron

– ++ –

+++ +++ +++

+++

+++

l

Ion exchange

l

Membrane technologies

l

Fe-Mn oxidation

?

++/+++

l

Solar still

+

+

l

Porous media sorbents

++

++

KEY +++ >90% ++ 60%–90% + 30%–60% – <30% ? Insufficient information

Tubewell water, abstracted from deep aquifers, can be highly contaminated with both inorganic and organic arsenic species; arsenopyrites are thought to be the principal source of arsenic species found in rock strata at 15–90 m below the surface. At this depth, arsenopyrites are subject to oxidation and reduction in the presence of anaerobic metal-reducing bacteria as water is withdrawn from wells for domestic and irrigation purposes. Such oxidation-reduction is alleged to be primarily responsible for the presence of soluble As(III) species, concentrations of which exist up to 30 mg/L versus the permissible 0.01 mg/L limit. Thus, tubewell sources must carry a low utility score, although it is recognized that deep sources below 100 m can be comparatively free of arsenic species and thereby comprise a safe source of drinking water. Hybrid technologies may also justify a high utility score, because they can provide safe and reliable water free from dissolved species. Shallow dug wells may contain relatively low levels of As(III), even at locations where deep tube wells are contaminated. This natural mitigation in shallow dug wells is attributed to continual replenishment by rainwater and surface waters, which are, to some extent, ‘‘treated’’ by percolation through soil. The fresh inflow to groundwater will also have a diluting effect on background contamination in well water; thus, dug wells have been assigned an above-average utility score of 6. Bioremediation may convert arsenic to an organic form and render it considerably less toxic. Such action offers an above-average removal performance and is, therefore, assigned an above-average utility score of 6. Conventional sand filtration has limited potential for removing soluble arsenic species, although the presence of Fe/Al hydroxides in the sand filter can enhance adsorption potential; thus, sand filtration has been allocated an average utility score. Rainwater harvesting will not

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

143

remove any As species, although it may have a natural arsenic mitigating potential and an above-average ‘‘removal’’ performance, which justifies a medium to high utility score of 7. Figure 3 shows a plot of the utility scores and grades assigned to each of the treatment options for the toxicological KPI.

B

Infectious Disease Risk KPI

The Bonn Charter for safe drinking water (IWAHQ 2005) identifies sources, treatment, and distribution as key criteria that require appropriate monitoring and management strategies. The Bangladesh national policy for arsenic mitigation (NAMIC 2006) also indicates that preference should be given to piped surface water, rather than groundwater, as a primary water supply source. Therefore, any treatment option for arsenic remediation must consider the potential risks of microbial infection as a KPI; indeed, this factor is included in the MCA performance matrix (see Table 2). Table 4 shows the potential health significance of waterborne bacteria in the context of existing WHO guidelines (WHO 2006). All pathogen groups present a major risk of infection that can persist once a water supply becomes contaminated. Microbial removal, through disinfection processes, can achieve typical 3 log inactivation with CT (concentration  time needed to kill a defined portion of a

10

5

9

Utility Score

8

4

7 6

3

5 4

2

3 2

1

1

lls

e

We

ng ep

Tu

be

cha De

Ex Ion

ion

Dis tilla tion ar So

Filt rat

/A /P GC

ion Filt rat nd Sa

iati

on

ells gW Du

rem ed Bio

g arv est in rH

Ra

inw

ate

eT ran mb Me

Hy bri

dT

ech n

ech no

olo

log

gy

y

0

Treatment Technology

Fig. 3 Arsenic toxicology utility plot for various treatment technologies. See text for explanation of scores and grades

144

B. Ellis, H. Garelick

Table 4 Health significance of waterborne pathogens Pathogen group Health Resistance in Resistance significance water supply to chlorine Enteric bacteria

High

Enteric viruses Protozoa Helminths

Relative infectivity

Mostly low

Low to moderate

High

Short to moderate (some may multiply) Long

Moderate

High

High High

Long to moderate Moderate

High Moderate

High High

Important animal sources In many

Generally not In some No

microbial population) values of 2–30 mg/min/L (WHO 2004) for enteric bacteria and viruses. Thus, by analyzing the infectious organism survival values, one can derive a relevant utility score plot for the infectious disease risk KPI, based on a benchmark that defines the microbial removal efficiency of each potential treatment technology. Figure 4 presents the scores that result from this microbial risk analysis approach: ion-exchange (IE) and solar still (SS) technologies are given zero efficiencies (utility score ¼ 0), whereas membrane technology (MT) and hybrid technologies (HT), which incorporate disinfection, are given near 100% efficiencies (utility score ¼ 10). Filtration and coagulation/flocculation clarification (CG/P/A) options had variable efficiencies and were given the lowest utility scores. Slow sand filtration (SSF) ponds are also used to wash clothing and dishes and to water livestock. These activities may produce high bacterial loads that require considerable community effort and discipline to prevent. If the removal performances illustrated in Table 4 and Fig. 4 can be achieved, perhaps in combination with follow-up chlorination, they can provide acceptable water quality; thus, a final utility score of 7 was assigned to this treatment option. On the other hand, up to 90% of shallow dug wells are known to be contaminated by fecal and other bacteria; such contamination results in very low utility (and grade) scores. Deep tube wells are likely to possess very low bacterial/viral risks and justify high utility (and grade) scores, as indicated in Fig. 4. The KPI plot from these scores describes a nonlinear relationship, which is also reflected in the grade value scores. These grade scores are entered in bold beneath the direct utility score for each treatment option in the Table 2 performance matrix.

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

145

10 9

5

MT HT DTW

SS

8 Utility Score

7

SF

4

6 F

5 3

4

CG/P/A

3 2 1

RH

2 BIO 1

0 0

DW IE 10

20

30

40

50

60

70

80

90

100

Percentage Microbial Removal

Fig. 4 Plot of utility score and grade values for infectious disease risk, based on microbial removal performance. See Fig. 2 for key to abbreviations and text for explanation of scores and grades

C

Other Chemical Pollution KPI

Sources of arsenic in the environment have both anthropogenic and natural origins. The main anthropogenic sources are: industrial wastes; phosphate ores; fertilizers; burning fossil fuels (mainly coal) and oil; mine tailings; smelting; ore processing; metal extraction and purification; production of chemicals, cement, glass, leather, textiles, alkali, alloys, pigments, desiccants; petroleum refineries; catalysts; and pesticides (wood preservatives, insecticides, herbicides). Anthropogenic arsenic pollution is rarely caused by a single pollutant; rather, it is accompanied by other chemical pollutants (e.g., heavy metals and hydrocarbons, etc.). In the natural environment, arsenic occurs in more than 200 different minerals. As a result, it is found in conjunction with other polluting elements that are source dependent (e.g., arsenic-rich base metal sulfides such as As-rich pyrites). Arsenic, in its most recoverable form, is found in various types of metalliferous deposits, and it is common in iron pyrite; galena and chalcopyrite are very important sources of arsenic in the environment. These minerals constitute the major sources of arsenic contamination of groundwater in the most polluted areas of Bangladesh and West Bengal. The presence of other polluting elements may influence the choice of water treatment or mitigation technology. There is evidence that an elevated concentration of iron (Fe/As molar ratio of 10–15) enhances the removal of arsenic from groundwater. On the other hand, the presence of phosphate and silicate may reduce

146

B. Ellis, H. Garelick

removal efficiencies (Sharma et al. 2003). Both data deficiencies and the sitespecific nature of nonarsenic metal pollution constrain the ability to designate KPIs for this ‘‘Other Chemical Pollution’’ group; these KPIs are not presented in Table 2 and would need to be completed by stakeholders who have local knowledge of pollution sources and levels.

VIII A

Cost Criteria

Capital Cost KPI

The costs of arsenic removal are central to adoption and sustained use of any treatment technology, particularly in rural areas. Given the infancy of many treatment options, total capital costs are often unknown and cannot be fully standardized. Table 5 summarizes capital costs (primarily dependent on plant/unit size, pH, and phosphate/silica content of raw water or groundwater) obtained from the literature. Other cost variables include flow rates, system design, and construction type as well as specific site conditions. Based on the range and distribution of capital costs shown in Table 5, it is possible to construct a utility curve using a three-stage interval scale: 0–33 (low cost; utility value of 9), 33–66 (medium cost; utility value of 5), and 66–100 (high cost; utility value of 3). On this scale, the maximum utility value is assigned to the lowest cost option and the minimum utility score to the most expensive option. The utility scores are shown in parentheses in the next to last column of Table 5. The allocation of utility scores for specific treatment options has taken into consideration the estimated life for each technology, as well as unit cost range (Table 5). Deep tubewell technology, although carrying a high initial cost for design and installation, becomes relatively ‘‘affordable’’ because its cost is amortized over a 20-yr period; it is given a utility score of 6.5 in the performance matrix. In comparison, MT, which has low capacities and a 2- to 3-yr life expectancy, appears as a high cost option with a utility score of 2. The costs of the filtration options vary widely (Table 5), with activated alumina filters having relatively high initial costs. However, such costs are partly offset because they are used as ‘‘compact’’ household filtration units. Thus, a mid-utility score of 5 has been assigned to the filtration option. As shown in Table 5, there is no ‘‘cheap’’ treatment option, although capital costs are clearly related to the operational scale of specific treatment technologies. Therefore, the resulting utility score plot is nonlinear. The utility scores and grade values shown in Fig. 5 have been entered into the Table 2 performance matrix, with the latter entries shown in bold font.

Filtration

Deep tube well

Slow sand filtration

Membrane technology Solar distillation Rainwater harvesting

20–30

5–10

20

15–20

2–3 2–4 15

0.24–3.11/m3 0.05/10L 0.091–0.407/m3

0.028–0.151/m3

0.066–0.353/m3

0.2/person/d 2.134/m3

Table 5 Capital costs associated with arsenic treatment Life (yr) Unit cost t5.26 Treatment option (US$/m3 or L) Coagulation3–10 1.21–1.7/m3 precipitation 0.02–0.03/20 L 0.06/40 L Adsorption media 3–4 Bioremediation 5–10

153,000–305,000

25–55/household 200/family 33,913a 169,566b 15–20/household 800/50 households 25,719a 128,593b 257,166c 20–65/household 900/50 households 21,561a 107,803b 215,607c 5–10/household 18,000–51,500

700

8,000–40,000

39,600–58,500

154,700–228,309

249,081

Installation costs (US$) 71,767

Total costs (US$)

Medium cost (5)

Low-medium cost (6.5)

Medium cost (6)

Medium cost Medium-low cost (7.5) High cost (2) High cost (2) High cost (2.5)

Comment (utility score) High cost (3)

3 (continued)

3

4

4

2 2 2

4

3

Grade value

t5:1

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water 147

Treatment option

Table 5 (continued)

10–15 5–10

Life (yr)

4.07/m3

Unit cost (US$/m3 or L)

290,521 15/household 800/25 households

17,653a 44,534b 85,024c

Total costs (US$)

t5.31

Source: Koundouri (2005); Ahmed (2003); Murcott (1999); OCETA (2001). a Based on village of 100 households. b t5.32 Based on village of 500 households. c t5.33 Based on village of 1,000 households.

t5.30

Hybrid technology (piped supply with chlorine disinfection) t5.28 Ion exchange t5.29 Dug well

t5.27

t5.26

77,574

Installation costs (US$)

Very high cost (1) Low-medium cost (7)

Comment (utility score) Medium-high cost (4)

1 4

Grade value

148 B. Ellis, H. Garelick

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

149

10

5

9 8

Utility Score

7

4

6 5

3

4 3

2

2 1

1

0 BIO

DW

DTW

SF

F

HT

CG / P / A

RH

MT

SS

IE

Treatment Technology

Fig. 5 Plot of utility score and grade values for capital costs associated with arsenic treatment options. See Fig. 2 for key to abbreviations and text for explanation of scores and grades

B

Operation and Maintenance Cost KPI

Deep tubewell sources can provide water at fairly nominal operation and maintenance (O and M) costs, but these are not feasible everywhere. Hybrid piped water supplies are clean, convenient, and healthy for users and have a relatively low O and M cost. The unit O and M costs for hybrid technologies, tube wells, and sand filtration units decrease as the number of households served increases (Table 6). Activated alumina filtration can also have relatively high running costs, which reduces its overall utility score. O and M costs for rainwater harvesting and dug well sources are relatively high, because management of such systems is difficult and requires regular health inspections for bacteria and other pollutants. Neither monitoring equipment nor laboratory analysis costs are included in the O and M costs shown in Table 6, but these are known to be relatively high. Analysis costs can be reduced by using semiquantitative arsenic field test kits for preliminary screening. All treatment technologies ultimately concentrate arsenic in sorption media, sludges, or liquid media, which then require safe disposal. The US EPA has recommended a maximum leachate level of no more than 100 times the Guideline for Drinking-Water Quality (GDWQ) level (50 mg/L). Thus, sludges that leach more than 5 mg/L arsenic are considered to be hazardous and would require disposal in a special landfill. However, tests conducted in Bangladesh (Ahmed

150

B. Ellis, H. Garelick

Table 6 Operational and maintenance costs for arsenic treatment options Treatment option Life O&M costs Comment (yr) (US$/yr) (utility score) Coagulation-precipitation 3–10 500–23 3.5/ Medium (5) family/yr Adsorption media 3–4 Bioremediation 5–10 Very low (9.5) Membrane technology 2–3 High (3) Solar distillation 2–4 Medium-high (4) Rainwater harvesting 15 5 Medium-low (6) 3,523/100 households 17,616/500 households 35,232/1,000 households Slow sand filtration 15–20 10–20 Low (7) 0.4–0.9/family/yr 2,242/100 households 11,212/500 households 9,322/1,000 households Deep tube well 20 4 Low (8) 1,083/100 households 5,415/500 households 10,831/1,000 households Filtration 5–10 10–36 Medium-high (4.5) 2–6/family/yr Hybrid technology 20–30 500/100 households Low (9) (Piped water supply 800/1,000 with chlorine households disinfection) Ion exchange 10–15 1,000 Very high (1) Dug well 5–10 3 Medium (5.5) Source: Koundouri (2005); Ahmed (2003); Murcott (1999); OCETA (2001).

Grade value 3

5 2 3 4

4

4

3 5

1 3

et al. 2001) indicated that most treatment option sludges and elutriates would not be considered hazardous, even at the lower WHO GDWQ value of 10 mg/L. Nevertheless, concern has been expressed about contaminated sludges from major coagulation, precipitation, and filtration treatment plants that would require a large receiving water dilution factor. Ion exchange also yields highly concentrated brine streams that require pretreatment before disposal.

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

151

The utility scores shown in the next to last column of Table 6 require adjustment to reflect increased costs, if sludge disposal is included as an O and M cost. Such costs are considered to be site specific and were not included in the current analysis.

IX

Community Acceptance Criteria

Not all communities have the same priorities, or technology awareness, or are affected equally by treatment option costs. Hence, the feasibility, effectiveness, and acceptability of arsenic treatment options will vary from place to place. For many reasons, all options, apart from dug wells, are either technically inefficient or are disliked by communities. The reasons for this dislike include convenience, access, monitoring requirements, slow flow rates, and cultural attitudes (Jakariya et al. 2003). Many projects have failed because social and cultural relationships were not adequately considered (Safiuddin and Karim 2003); failures resulted from factors such as lack of ease of system use by women, access and waiting time, and local feasibility of equipment, materials, and upkeep/maintenance. Filtration (F), IE, MT, and coagulation/precipitation/adsorption (CG/P/A) techniques are said to be among the most commonly used arsenic treatment options (Visoottiviseth and Ahmed 2008, this volume), and have high community acceptance, particularly where they can be scaled down for household and small community use. However, the technology associated with their application and maintenance is not simple. The introduction of reactive chemicals (and possibly microbes) in CG/ P/A treatment may have unforeseen effects on the oxidation process, which can reduce arsenic treatment efficiency. In addition, filtration techniques that are based on patented media and imported technology can pose operational difficulties for rural communities. Inappropriate management of technology can lead to a lack of community commitment and failure in system ownership. Poor mixing and variable water quality affects the performance of small bucket, fill-and-draw, and three- or fourjug treatment units, although they represent promising low-cost treatment options at the household level. Iron-arsenic (Fe-As) removal plants offer potentially efficient treatment options, although community participation in plant O and M is essential for their effectiveness (see Visoottiviseth and Ahmed 2008, this volume). Community acceptance of most small (household)-scale filter units is uncertain because of clogging and leakage problems as well as hygiene issues. The reputation and acceptance of such household units suffers as a consequence. For these reasons, and despite their alleged popularity, simple household technology options have been assigned relatively low utility scores (Table 7). SSF may appear to be socially acceptable given the potential output volumes (serving >60 households per unit) and lack of required chemical treatment. However, SSF from filter ponds may also have a conjunctive use, such as fish culture, which can foment contamination. In addition, many ponds are plagued by algal

152

B. Ellis, H. Garelick

Table 7 Community acceptance of various technology treatment options Treatment option Comment Utility score Coagulationprecipitationadsorption Filtration

Slow sand filtration

Deep tube well Dug well

Ion exchange Membrane technologies Bioremediation

Rainwater harvesting

Hybrid technologies

Solar distillation

Relatively simple; low cost; common chemical usage; requires systematic and sustained community input Well-known; readily available/simple techniques; user friendly; sludge and mixing problems; low flow rates; small batch rates Simple construction; no chemicals used; large volumes; easily contaminated; high turbidity; queuing and access problems; requires land Assumed to be ‘‘safe’’ water; restricted to land ownership Indigenous technology; low-arsenic or arsenic-free source; field test kit use by households; regular health inspections; restricted to those who own land Unfamiliar technology; lack of knowledge and awareness Well-defined techniques; lack of familiarity at local level; low flow rates; O&M requirements Low cost; simple technique; environmentally friendly; requires land Simple technology; low O&M; uses local materials and craftsmanship; novelty not always acceptable; uncertain source and supply Piped supply regarded as totally safe source; assured supply and control; ease of access and use; little local O&M required; highly centralized Unfamiliar technology; safety fears; lack of awareness and knowledge

High (7.5)

Grade value 4

Mediumhigh (5.5)

3

Medium (4.5)

3

Very high (9)

5

High (8.5)

5

Low (1)

1

Low (2)

2

Low-medium (4)

3

Low-medium (3.5)

3

Very high (10)

5

Low (1)

1

growths and high turbidity, which considerably reduces their effectiveness as a reliable water source. Such units also frequently require community awareness and participation to achieve basic maintenance such as sand washing and drying. Therefore, the social acceptability for obtaining clean water using SSF is not high, as reflected by the assigned average score of 4.5. Both dug wells and deep tube wells represent well-established technologies in rural communities and are readily accepted as reliable sources of water for drinking and cooking. Sanitary dug wells (with filters and hand pumps) are capable of delivering as much as 1000 L/d.

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

153

Tube wells and hybrid piped water are regarded to be ‘‘safe’’ water supplies, and most communities are generally willing to pay for such sources (Caldwell et al. 2003). Therefore, both options have been assigned high utility scores (Table 7). Rainwater harvesting, on the other hand, represents a more novel, innovative option for many areas, and may not be readily available to households in communities having thatched or small roofs. In addition, such sources are not available during dry periods unless sufficient storage capacity has been constructed; rainwater harvesting has, therefore, been assigned a low to medium score (3.5). However, in those regions such as coastal Bangladesh where rainwater harvesting is a familiar water source, the technology is generally socially acceptable and would justify a higher utility score.

X

Technical and Organizational Skill-Base Criteria

The issue of household versus community-level technological competencies and organizational skills is an important consideration for the selection of arsenic treatment options. In addition to the fit and community acceptance of a technology, local organizational skills, cost, and technical feasibility of the project must also be considered. The most acceptable technologies are frequently more expensive, because these are thought to produce ‘‘superior’’ and ‘‘safe’’ sources of drinking water. However, further work is often needed to make these options affordable and socially acceptable in communities where they are to be used. The transfer of new technologies has proven to be extremely slow in developing countries. It is not only technical equipment that provides a barrier for local rural communities, but also the lack of ‘‘know-how’’ and a dearth of organizational capabilities and familiarity with managerial procedures. Such local skills are essential for successfully managing community treatment systems over the long term. Table 8 presents the advantages and disadvantages of various arsenic treatment technologies in the context of local know-how and skills needed to implement them.

XI

Community Location Criteria

Sustainable remediation is directly linked with prevailing local social and economic conditions. Therefore, solutions in rural settings may be different from those in urban/peri-urban areas. Nearly three-quarters of the world’s poor live in rural areas, and a large part of this population lacks access to clean water. These figures regularly change with advancing urbanization in both developed and developing countries. It is predicted that, by 2017, the proportion of the world’s population living in urban areas will exceed those living in rural areas (United Nations 2004). This rapid urbanization,

Table 8 Local expertise and competency ratings for implementation and management of arsenic mitigation technologies Treatment option Advantages Disadvantages Utility score CoagulationRelatively simple; common chemical Maintenance? High-very high precipitationusage; no complex or sophisticated (8.5) adsorption skill base necessary Filtration Well-known; readily available technology Maintenance High-very high (8.5) Slow sand filtration Continuous operation; construction from Local O&M capabilities; tension with High (8) local materials; no chemical treatment community fish farming use required Deep tube well Subsidized training often available; Pump maintenance Very high (9.5) trained O&M personnel often supplied Dug well Traditional competencies, community Maintenance? Medium-high (7) organization and knowledge Ion exchange High (2) Membrane Well-defined technology High technical expertise; dependence on Very low (2) technologies skilled personnel Bioremediation Simple technology; low O&M Long-term operation issues; low volumes Medium (6) requirements Rainwater Low O&M requirements; local materials Problem of dry weather periods; potential Medium-high harvesting infectious risk (7.5) Hybrid technologies Large volume capability; imported Technical expertise and organizational Low (3) technologies capacity will be often required for installation and O&M Solar distillation Good control of infectious disease Very low volumes; skill base necessary Low-medium (4)

2

1

4

3

2 1

3

5

4

4

Grade values 4

154 B. Ellis, H. Garelick

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

155

Table 9 The influence of urban community location and setting on sustainability of arsenic mitigation technologies Treatment option Advantages Disadvantages Utility Grade score value CoagulationSuitable for urban Unsuitable for Medium (7) 4 precipitationcommunities? households? adsorption Filtration Not appropriate for Maintenance? 4 2 single households Slow sand filtration Suitable for small Land uptake 5 3 communities Deep tube well More appropriate for Volumes? Medium-high 4 extended family (8) groups Dug well Household Costs; queuing; time Low (2)? 1 applications as well as community use Ion exchange Good control Costs; large-scale 9 5 supply? Membrane Suitable for large Restricted volumes 6 3 technologies communities? Bioremediation Simple technology Land requirements Very low (1) 1 Rainwater Low O&M Infectious disease Medium (5) 3 harvesting requirements; risks; individual ease of household construction application Piped water supply Traditional urban Cannot cover large 10 5 technology distances safely Hybrid technologies Networked system Costs; operation; 9 5 O&M; skill base Solar distillation Not fully appropriate Low volumes; costs 4 2 at community level

particularly in the developing world, is associated with widespread poverty, inadequate health, poor water supply and sanitation, and environmental degradation. In Tables 9 and 10, an attempt has been made to disaggregate the relative advantages and limitations of arsenic remediation in rural versus urban locations. Dug wells have traditionally comprised the technology of choice for households in rural areas and, therefore, justify a higher than average score. However, the scores given to dug wells are limited because of their potential for bacterial contamination. Dug wells are much less feasible (queuing times, storage conditions, and costs) in crowded urban locations. Given such limitations, their utility value is much less than average, as is reflected in Tables 9 and 10.

156

B. Ellis, H. Garelick

Table 10 The influence of rural community location and setting on sustainability of arsenic mitigation technologies Treatment option Advantages Disadvantages Utility Grade score value CoagulationSuitable for Unsuitable for 4 1 precipitationcommunities? households? adsorption Filtration Suitable for single Maintenance? 8.5 4 households Slow sand filtration Suitable for small 8 3 rural communities? Deep tube well Traditional use Expensive; O&M 8 3 Dug well Traditional Limitations of water 6 2 household use quality Ion exchange Good individual Cost; O&M issues 9 5 control and costs Membrane Suitable for Technology 9 5 technologies communities and competency and small family use lack of skill base Bioremediation Simple technology Fish farming use High (8) 3 Rainwater Low O&M; Dry weather supply? High-very 5 harvesting individual high (9) application Piped water supply External supply Cannot cover large Low-medium 1 distances safely (3–4) Hybrid technologies External supply and Costs; operational 3–4 1 imported skills required technologies Solar distillation Individual Costs; skill base Medium (6) 2 application

XII

Aggregating KPI Scores

The following formula is used to aggregate KPIs into a single criterion: KPIs on the maximal markðMaxm Þcj ðai Þ ¼ ð

X

where ai = the individual technology alternative cj = criterion fmi = ratio between the sum of partial marks (fmi) Maxm = KPIs on the maximal mark

mi Þ=ðMaxm Þ  100

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

157

Table 11 Ranking of technology alternatives for removing arsenic from water, using the ratings presented in Table 2 Hybrid technologies Deep tube wells Membrane technologies Slow sand filtration Rainwater harvesting Precipitation/coagulation/adsorption options Bioremediation Filtration Dug wells Solar distillation Ion exchange

At this stage, the end-user is not invited to attribute any weighting factor to the KPI, because this is done later for all criteria. To fill the performance matrix, aggregation is completed for each technology alternative and for each criterion. Assuming that an equal weighting is given to each of the KPIs shown in Table 2, the aggregate scores (cj, aj) for each treatment technology can be readily calculated. Based on the grading values, this would derive the ranking of alternative arsenic treatment technologies as shown in Table 11. The MCA procedure also allows a weighting score to be applied to the six generic criteria that are distinct from the 10 KPIs. Based on the KPI scorings shown in Table 2, it is clear that hybrid solutions such as piped water supplies and deep tube wells are the preferred remediation options. However, if higher weightings were given to the cost criterion and associated KPIs, these two options would give way to others such as slow sand filtration.

A

Factor Weighting and Sensitivity

The assigning of weights to criteria (and/or KPIs) is a contentious issue within MCA and, undoubtedly, outcomes can be substantially influenced by the weighting procedure (Simos 1990; Figueira and Roy 2002). Given the ‘‘fuzzy logic’’ surrounding KPI data, an outranking MCA procedure could be employed that incorporates uncertainty into the evaluation of the capability and efficiency of alternative treatment options. Such procedures use pseudo-criteria as employed by the ELECTRE III software model, in which differences in performance between alternatives are compared by assigning preference, veto, and indifference (or neutral) thresholds to criteria (Martin et al. 2007). The utility scales used for evaluating each alternative KPI do not necessarily equate on the various scales used. End-user preferences and opinions may often by fuzzy, conflicting, open to change during the decision process, or influenced by the

158

B. Ellis, H. Garelick

modeling outcomes. Therefore, weighting is required to equate the scales, and this is equivalent to judging the relative importance of the scales used. Weights can be assigned using a ‘‘swing-weighting’’ method that determines how the swing from 0 to 10 (or 0 to 100) on one scale compares to 0 to 10 (or 0 to 100) on another scale, and also how much that difference affects outcomes. For example, cost might be an important factor in choosing a technology option, but if the difference in cost between all options is low, then the cost criterion could receive a low weight (because the difference between highest and lowest costs is small and does not affect the final decision). Hence, the weight of a criterion reflects both the range of difference between options and how much that difference matters. To apply the swing method, it is first necessary to identify criteria with the greatest relative 0–100 (or 0–10) swing in preference. A pairwise comparison may be necessary to achieve this, always retaining the criterion with the biggest ‘‘swing.’’ The criterion that emerges as having the largest swing is assigned the maximum value (e.g., 10 or 100), and this becomes the standard to which all others are compared. It is given a weight of 50, if the swing of the next criterion is judged to be half that of the first. This process is open to debate, and the weights selected should be those which best represent the views of the stakeholders as a whole, although the final decision may well be political. If consensus cannot be reached, it may be best to accept two or even more sets of weights, because agreement on choice of options can sometimes be reached without agreement on weights. To calculate the final weighted scores, the option score on a criterion (or KPI) is multiplied by its weight; this is done for each criterion, and the products are added together to give the overall preference score for that option. An important additional step in the MCA could be the application of a simple sensitivity analysis, which involves reanalysis with different utility scores and weights to see how much the differences influence the outcome; this may be done using both the derived utility score and grade value. This process may help address concerns of stakeholders who disagreed over what constituted appropriate scores and weights by allowing their preferences to be evaluated. Sensitivity analysis may show that two or three options are always preferred and that there is little difference between/among them. In this way, the MCA matrix can be used as a negotiating tool in the decision-making process. Sensitivity analysis can also be seen as a way of highlighting areas of agreement rather than disagreement, which tends to be the focus of most stakeholders discussions. It may also crystallize areas where further investigation or creation of new options are needed; such inputs may achieve costeffective ‘‘trade-offs’’ between objectives, i.e., a solution that is almost as beneficial but is much less costly or more socially acceptable. Clearly, some data programmed into certain KPI utility scoring may be arbitrary. Thus, it is feasible that a range of parameters may lead to the same acceptable treatment. Even if an acceptable solution is gained, a sensitivity analysis can help verify the robustness of the final rankings. In some cases, the technology performance is not sufficiently distinctive to constrain the final ranking, with criteria weighting failing to influence this ranking. This outcome would suggest that, before commencing any decision-making process and seeking a stakeholder consensus,

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

159

there is a need to undertake a more complete evaluation to assess the performance efficiency of each technological option under varying degrees of factor influence.

B

Exclusion Limits

The procedure for selecting arsenic remediation technologies will possibly deem certain treatments as unsuitable for pursuit at a specific location. In the majority of such cases, the reason for exclusion is likely to be environmental or scientific and relate to potential impairments in system performance based on physical treatment limits, availability of technical capacity, or extremely high or low source concentrations, etc. However, technological exclusion may occasionally result from overriding economic or social parameters. Such exclusions can be identified in the performance matrix by entering zero values for the utility (and/or grade) scores that would eliminate those particular technologies from the final evaluation.

Summary The identification of best practice technologies to remediate arsenic-enriched drinking water involves the resolution of several technical, environmental, economic, and social factors. Multi-criteria analysis (MCA) provides a procedure to sort through diverse influencing factors as a means of facilitating the stakeholder decision-making process. The primary key MCA criteria used to define arsenic treatment options are expressed as source–exposure vector, health risk, cost, social acceptance, and technical competency. MCA not only can handle a complex mix of quantitative and qualitative information but also fosters means to resolve conflicting stakeholder opinion (or strategies). The MCA procedure involves construction of a performance matrix from utility scores for each key performance indicator (KPIs) that may influence outcomes. Data in the performance matrix are converted into numerical values through application of a specific utility scale scoring and weighting technique for each criterion. Inspection of the performance matrix scores facilitates decision making because they summarize arsenic treatment options numerically for all important criteria and KPIs. The weighting procedure enables stakeholder preferences (or strategies) to be incorporated into the selection process. Given the ‘‘fuzzy logic’’ nature of the KPI information, uncertainty may influence data outcome; this can be addressed by using an outranking procedure such as ELECTRE III or a simpler ‘‘swing’’ pairwise preference method. Sensitivity analysis can also be performed by reiterating the analysis using different utility scores and/or weights to assess influence on performance matrix outcomes. This approach enables the MCA methodology to be used as a negotiating tool in the decision-making process and allows areas of stakeholder agreement and disagreement to be highlighted.

160

B. Ellis, H. Garelick

References Ahmed MF, Ashraf AM, Adeel Z (2001) Technologies for Arsenic Removal from Drinking Water. International Workshop, Bangladesh University for Engineering & Technology, Dhaka and United Nations University, Tokyo. Ahmed MF (2003) Treatment of arsenic contaminated water. In: Ahmed MF (ed) Arsenic Contamination: Bangladesh Perspective. ITN-Bangladesh Centre for Water Supply and Waste Management, BUET, pp 354–403. ASCE and UNESCO (1998) Sustainability Criteria for Water Resource Systems. Report of Joint ASCE/UNESCO Task Committee on Sustainability Criteria. Project M-4.3, IHP-IV. American Society of Civil Engineers, Reston, VA. Ashley RM, Smith H, Jowitt PW, Butler D, Blackwood DJ, Davies JW, Gilmour DJ, Foxon T (2001) A multi-criteria analysis/risk management tool to assess the relative sustainability of water/wastewater systems: SWARD (Sustainable Water Industry Asset Resource Decisions). In: Pratt CJ, Davies JW, Perry JL (eds) Proceedings, 1st National Conference on Sustainable Drainage. Coventry University, Coventry, UK, pp 221–231. Boerschke RK, Stewart DK (2001) Evaluation of arsenic mitigation technologies for use in Bangladesh. In: Ahmed MF, Ashraf AM, Adeel Z (eds) Technologies for Arsenic Removal from Drinking Water. International Workshop, Bangladesh University for Engineering & Technology, Dhaka and United Nations University, Tokyo, pp 214–230. Caldwell BK, Caldwell JC, Mitra SN, Smith W (2003) Tubewells and arsenic in Bangladesh: challenging a public health success story. Int J Popul Geogr 9:23–38. Ellis JB, Deutsch J-C, Mouchel J-M, Scholes L, Revitt DM (2004) Multi-criteria decision approaches to support sustainable drainage options for the treatment of highway and urban runoff Sci Total Environ 334/335:251–260. Ellis JB, Deutsch J-C, Legret M, Martin C, Revitt DM, Scholes L, Seiker H, Zimmerman U (2005) The Day Water decision support approach to the selection of sustainable drainage systems: a multi-criteria methodology for BMP decision makers. Proceedings, 10th International Conference on Urban Drainage, Copenhagen. August 2005, IWA, London. Figueira J, Roy B (2002) Determining the weights of criteria in the ELECTRE type methods with a revised Simos’ procedure. Eur J Operational Res 139:317–326. Garelick H, Jones H, Dybowska A, Valsami-Jones E (2008) Arsenic pollution sources. Rev Environ Contam Toxicol (this volume). IUPAC (2003) Remediation technologies for the removal of arsenic from water and wastewater. www.iupac.org/projects/2003/2003-017-2-600.html. IWAHQ (2005) Bonn Charter for Safe Drinking Water. http://www.iwahq.org/uploads/bonn% 20charter/IWA_DWS_2005.pdf. Koundouri P (2005) The economics of arsenic mitigation. I: Arsenic Contamination of Groundwater in South and East Asian Countries, vol II. Technical Report, Water & Sanitation Program, World Bank, Washington DC, pp 210–262. Jakariya M, Chowdhury AMR, Hossain Z, Rahman M, Sarker Q, Khan RI, Rahman M (2003) Sustainable community-based safe water options to mitigate the Bangladesh arsenic catastrophe: a experience from two upazilas. Curr Sci 85(2):141–146. Loetscher T, Keller J (2002) A decision support system for selecting sanitation systems in developing countries. Socio-Econ Plan Sci 36:267–290. Martin C, Ruperd Y, Legret M (2007) Urban stormwater drainage management. The development of a multi-criteria decision aid approach for Best Management Practices. Eur J Operational Res 181(1):338–349. Massachusetts Institute of Technology (2001) Arsenic Remediation Technologies. http://web.mit. edu/murcott/www/arsenic. Murcott S (1999) Appropriate remediation technologies for arsenic-contaminated wells in Bangladesh. Arsenic in Bangladesh Ground Water, February 27–28, Wagner College, Staten

A Multi-Criteria Approach to Remediate Arsenic in Drinking Water

161

Island, New York. Available on line: http://phys4harvardedu/wilson/arsenic/remediation/ arsenic_removal/murcotthtml. NAMIC (National Arsenic Mitigation Information Centre) (2006) Bangladesh Arsenic Mitigation Water Supply Project. http://www.bamwsp.org/. OCETA (2001) ETV-AM Screening, Testing and Evaluation Protocols, vols 1–6. Ontario Centre for Environmental Technology Advancement, Toronto, Canada. ODPM (2003) Multi-criteria Analysis Manual. Office of the Deputy Prime Minister, London. UK. www.dter.gov.uk/about/multicriteria. Price RE, Pilcher T (2005) Distribution, speciation and bioavailability of arsenic in a shallowwater submarine hydrothermal system, Tutum Bay, Ambitle Island, PNG. Chem Geol 224 (3):122–135. Safiuddin M, Karim MM (2003) Water resources management in the remediation of groundwater arsenic contamination in Bangladesh. In: Murphy T, Guo J (ed) Aquifer Arsenic Toxicity and Treatment. Backhuys, Leiden, The Netherlands, pp 1–17. Sharma AK, Tjell JC, Mosbaek H (2003) Removal of arsenic using naturally occurring iron. J Phys IV Fr 107(2):1223–1226. Simos J (1990) Evaluer l’Impact sur l’Environnement: Une Approche Originale par l’Analyse Multicritere et la Negociation. Presses Polytechniques & Universitaire Romandes, Lausanne, Switzerland. Smedleigh PL, Kinniburgh DG (2001) Source and behaviour of arsenic in natural waters. In: Synthesis Report: Arsenic in Drinking Water. World Health Organisation, Switzerland, pp 2–61. United Nations (2004) World urbanisation prospects: the 2003 revision. Data tables and highlights. United Nations, New York. Visoottiviseth P, Ahmed F (2008) Technology for remediation and disposal of arsenic. Rev Environ Contam Toxicol (this volume). Wenzel V (2001) Integrated assessment and multi-criteria analysis. Phys Chem Earth (B) 26 (7/8):541–545. WHO (2004) Inactivation (disinfection) processes. In: LeChevallier MW, Au KK (eds) Water Treatment and Pathogen Control: Process Efficiency in Achieving Safe Drinking Water. http://www.who.int/water_sanitation_health/dwq/en/watreatpath3.pdf. WHO (2006) Guidelines for drinking-water quality. www.who.int/water_sanitation_health/dwq.

Index

Activated carbon, arsenic removal, 102 Adarsha filter, arsenic removal, 103 Agricultural use, arsenic, 46 Alcan activated-alumina, arsenic removal (illus.), 95 Algae, arsenic removal in Thailand, 172 Algal bioremediation, arsenic removal, 108 Alumina filtration, arsenic removal, 93 Analysis of arsine, gas samples, 73 Analytical methods, arsenic in water, 63 Anodic stripping voltammetry, arsenic analysis, 63, 69 Anthropogenic arsenic contamination, sources, 41 Anthropogenic pollution sources, arsenic (table), 49 Apyron treatment unit, arsenic removal (illus.), 96 Arsenic accumulation in coal, types, 45 Arsenic adsorption method, layered double-hydroxides, 183 Arsenic bioavailability, inorganic form, 5 Arsenic concentrations, minerals (table), 23 Arsenic concentrations, volcanic effects, 40 Arsenic contamination from mining, US, 44 Arsenic contamination source, mining, 42 Arsenic contamination, anthropogenic sources, 41 Arsenic contamination, Asia (illus.), 3 Arsenic contamination, Bangladesh (illus.), 4 Arsenic contamination, Bengal River Basin, 7 Arsenic contamination, environment, 18

Arsenic contamination, geothermal effects, 37 Arsenic contamination, mobilization mechanism (table), 27 Arsenic contamination, risk mitigation, 11 Arsenic contamination, surface water, 40 Arsenic detection, colorimetric test kits (table), 65 Arsenic detection, field test kits, 64 Arsenic dose-response, cancer, 8 Arsenic emissions, coal burning, 44 Arsenic enrichment, geothermal effect, 26 Arsenic exposure, risk characterization, 10 Arsenic field test kits, solids and biota, 72 Arsenic field test kits, Environmental Technology Verification (table), 66 Arsenic health effects, inorganic form, 4 Arsenic health risk, key criteria, 141 Arsenic in Bangladesh, water purification, 11 Arsenic in drinking water, mitigation approach, 130 Arsenic in drinking water, remediation options, 129 ff. Arsenic in food, Thailand, 166 Arsenic in groundwater, sources and levels (table), 27 Arsenic in Hungarian drinking water, health effects, 182 Arsenic in Hungarian drinking water, mitigation, 182 Arsenic in water, analytical methods, 63 189

190

Arsenic in water, colorimetric testing (table), 65 Arsenic in water, Thailand, 166 Arsenic levels, hot springs (table), 38 Arsenic mitigation in Bangladesh, pond sand filters, 177 Arsenic mitigation methods, Bangladesh, 176 Arsenic mitigation methods, community location effect (table), 155 Arsenic mitigation methods, criteria flow chart (table), 132 Arsenic mitigation options, dug or deep tube wells, 135 Arsenic mitigation options, rainwater harvesting, 135 Arsenic mitigation, dug or deep tube wells, 114 Arsenic mitigation, Thailand, 168 Arsenic monitoring, criteria, 62 Arsenic occurrence, minerals, 21 Arsenic pollution in Hungary, drinking water, 179 Arsenic pollution in Thailand, history, 164 Arsenic pollution of drinking water, Hungary (illus.), 181 Arsenic pollution, anthropogenic sources (table), 49 Arsenic pollution, Bangladesh, 175 Arsenic pollution, gold mining, 42 Arsenic pollution, sources in Thailand, 165 Arsenic pollution, Thailand, Bangladesh and Hungary, 163 ff. Arsenic remediation methods, community acceptance, 151 Arsenic remediation, rainwater harvesting, 116 Arsenic remediation, surface water treatment, 115 Arsenic removal from drinking water, membrane filtration, 185 Arsenic removal method, bucket treatment unit, 86 Arsenic removal method, oxidation, 79 Arsenic removal method, Star Filter (illus.), 88 Arsenic removal methods, comparative utility (illus.), 139

Index

Arsenic removal methods, comparison (table), 112 Arsenic removal methods, performance and cost (table), 80 Arsenic removal technologies, ranking alternatives (table), 157 Arsenic removal unit, bucket treatment (illus.), 87 Arsenic removal unit, Read-F, 97 Arsenic removal unit, tube well (illus.), 89 Arsenic removal, activated carbon, 102 Arsenic removal, Adarsha filter, 103 Arsenic removal, algal bioremediation, 108 Arsenic removal, cartridge filters, 103 Arsenic removal, Chari filter, 103 Arsenic removal, coagulation and filtration, 84 Arsenic removal, ferric hydroxide, 96 Arsenic removal, fill and draw unit (illus.), 89 Arsenic removal, in situ oxidation (illus.) 83 Arsenic removal, ion exchange, 103 Arsenic removal, iron-arsenic plant (illus.), 90, 92 Arsenic removal, iron-coated sand (illus.), 99 Arsenic removal, legal implications, 13 Arsenic removal, lime treatment, 92 Arsenic removal, membrane techniques, 105 Arsenic removal, nanofiltration and reverse osmosis, 107 Arsenic removal, phytoremediation, 109 Arsenic removal, SAFI filter, 101 Arsenic removal, sedimentation, 82 Arsenic removal, Shapla filter, 99 Arsenic removal, social implications, 12 Arsenic removal, Sono filter, 100 Arsenic removal, sorptive filtration, 93 Arsenic removal, technologies, 78 Arsenic removal, Tetrahedron technology, 104 Arsenic removal, treatment options, 133 Arsenic removal, water contamination (table), 80 Arsenic residues, humans (table), 6 Arsenic salts, human studies, 7 Arsenic speciation, characteristics, 19

Index

Arsenic speciation, predominance diagrams (illus.), 20 Arsenic species removal, methods comparison (table), 142 Arsenic treatment methods, capital costs (table), 147 Arsenic treatment methods, toxicology utility score (illus.), 143 Arsenic treatment options, comparative costs (table), 150 Arsenic treatment, costs (table), 120 Arsenic treatment, defining options, 133 Arsenic use, wood preserving industry, 47 Arsenic waste in plants, Thailand, 174 Arsenic waste, disposal, 122 Arsenic, agricultural use, 46 Arsenic, Bangladesh, 1 ff. Arsenic, characteristics, 18 Arsenic, concentration in rocks, 24 Arsenic, contamination, 1 ff. Arsenic, electrochemical detection (table), 68 Arsenic, environmental media analysis, 72 Arsenic, environmental mineral sources, 20 Arsenic, exposure pathways, 2 Arsenic, field monitoring criteria, 62 Arsenic, field test kits, 61 ff. Arsenic, future detection methods, 71 Arsenic, groundwater contamination (table), 2 Arsenic, groundwater contamination, 25 Arsenic, health effects in Thailand, 167 Arsenic, health risk assessment, 1 ff. Arsenic, health risk paradigm (illus.), 8 Arsenic, historical uses, 41 Arsenic, human uptake, 6 Arsenic, mineral content (table), 22 Arsenic, pollution sources, 17 ff. Arsenic, remediation and disposal, 77 ff. Arsenic, spectrometric detection, 63 Arsenic-contaminated mining waste, Thailand, 168 Arsenic-induced skin cancer, Bangladesh, 10 Arsenic-rich coal, China, 46 Arsine gas, analysis, 64 Arsine gas, analytical methods, 73 Asia, arsenic contamination (illus.), 3

191

Atomic absorption spectrometry, arsenic detection, 63 Atomic emission spectrometry, arsenic detection, 63 Atomic fluorescence spectrometry, arsenic detection, 63 Bangladesh arsenic contamination, source in groundwater, 25 Bangladesh drinking water, arsenic-free options, 118 Bangladesh risk mitigation, arsenic, 11 Bangladesh, arsenic contamination (illus.), 4 Bangladesh, arsenic contamination, 1 ff. Bangladesh, arsenic contamination, 18 Bangladesh, arsenic exposure, 9 Bangladesh, arsenic mitigation measures, 176 Bangladesh, arsenic pollution, 163 ff. Bangladesh, arsenic pollution, 175 Bangladesh, arsenic-induced skin cancer, 10 Bengal Basin, arsenic risk, 10 Bioavailable, inorganic arsenic, 5 Bioremediation for arsenic removal, algae, 108 Bioremediation with algae, arsenic removal, 172 Bioremediation, arsenic removal option, 134 Bucket treatment unit, arsenic removal (illus.), 87 Bucket treatment unit, arsenic removal method, 86 BUET activated-alumina, arsenic removal (illus.), 94 Cancer, arsenic dose-response, 8 Carcinogenicity, arsenic in Bangladesh, 10 Cartridge filters, arsenic removal, 103 CCA use, arsenic pollution, 48 Chari filter, arsenic removal, 103 Chemical treatment, arsenic removal option, 133 China, arsenic in coal, 46 Chlorella vulgaris, arsenic removal in Thailand, 172 Chlorella vulgaris, arsenic removal, 110

192

Coagulation, arsenic removal method, 84, 85 Coal burning, arsenic emissions, 44 Colorimetric field test kits, arsenic detection (table), 65 Colorimetric methods, arsenic detection, 64 Community acceptance, arsenic remediation methods, 151 Comparative utility, arsenic removal methods (illus.), 139 Concentration in groundwater, arsenic sources, 25 Concentration in rocks, arsenic, 24 Concentrations of arsenic, mineral sources (table), 23 Contamination, arsenic in Bangladesh, 1 ff. Criteria, arsenic mitigation (table), 132 Deep tube well, arsenic mitigation option, 135 Deep tube well, arsenic mitigation, 114 Deep tube wells, arsenic mitigation in Bangladesh, 178 Detections methods for future, arsenic, 71 Dimethyl arsenic acid, transformation, 7 Disposal technology, arsenic, 77 ff. Disposal, arsenic waste, 122 Drinking water in Hungary, arsenic pollution, 179 Drinking water pollution in Hungary, health effects, 182 Drinking water remediation, options for arsenic, 129 ff. Drinking water, arsenic removal costs (table), 120 Dug wells, arsenic mitigation in Bangladesh (table), 176 Dug wells, arsenic mitigation option, 135 Electrochemical detection of arsenic, water analysis (table), 68 Electrochemical test kits, performance with arsenic, 69 Environmental contamination, arsenic, 18

Index

Environmental media analysis, arsenic, 72 Environmental Technology Verification, method verification (table), 66 Environmental toxicity, arsenicals, 48 Epidemiology assessment, arsenic, 10, 11 Exposure assessment, arsenic and Bangladesh, 9 Exposure pathways, arsenic, 2 Ferric hydroxide, arsenic removal, 96 Field kits, arsenic detection in solids, 72 Field monitoring criteria, arsenic, 62 Field test kits, arsenic detection, 65 Field test kits, arsenic, 61 ff. Field test kits, costs, 66 Field testing kits, arsenic detection, 64 Field vs. laboratory methods, arsenic detection, 70 Fill and draw units, arsenic removal (illus.), 89 Filtration, arsenic removal in Thailand, 170 Filtration, arsenic removal method, 84 Filtration, arsenic removal, 93 Food contamination in Thailand, arsenic, 166 Geothermal activity, arsenic effects, 26 Geothermal effects, arsenic contamination, 37 Gold mining, arsenic pollution, 42 Groundwater contamination, arsenic (table), 2 Groundwater contamination, arsenic, 25 Groundwater sources and levels, arsenic (table), 27 Health effects in Thailand, arsenic, 167 Health effects, arsenic in Hungarian drinking water, 182 Health effects, inorganic arsenic, 4 Health hazard, arsenic 7 Health risk assessment, arsenic, 1 ff. Health risk criteria, arsenic, 141 Health risk paradigm, arsenic (illus.), 8 Historical uses, arsenic, 41 Hot springs, arsenic levels (table), 38 Human health risk, arsenic, 7

Index

193

Human poisoning, metal mining, 43 Human residues, arsenic (table), 6 Human studies, arsenic salts, 7 Human uptake, inorganic arsenic, 6 Hungarian drinking water pollution, arsenic, 179 Hungarian drinking water, arsenic removal by membrane filtration, 185 Hungarian drinking water, mitigating arsenic, 182 Hungary, arsenic pollution location (illus.), 181 Hungary, arsenic pollution, 163 ff.

Mitigation measures, arsenic in Bangladesh, 176 Mitigation of arsenic in drinking water, Hungary, 182 Mitigation of arsenic, Thailand, 168 Mobilization mechanism, arsenic contamination (table), 27 Monitoring criteria, arsenic, 62 Monomethyl arsenic acid, transformation, 7 Multi-criteria approaches, arsenic in drinking water, 130 Multi-criteria assessment, arsenic remediation, 129 ff.

India, arsenic contamination, 18 Infectious disease risk, waterborne organisms (table), 144 Inorganic arsenic, bioavailability, 5 Inorganic arsenic, health effects, 4 Ion exchange, arsenic removal, 103 Ion-exchange treatment, arsenic removal, 133 Iron-Arsenic removal, plants (illus.), 90 Iron-coated sand, arsenic removal (illus.), 99

Nanofiltration, arsenic removal, 107

Laboratory vs. field methods, arsenic detection, 70 Layered double-hydroxides, arsenic adsorption method, 183 Lime treatment, arsenic removal, 92 Mass spectrometry, arsenic detection, 63 Membrane filtration, arsenic removal from drinking water, 185 Membrane techniques, arsenic removal, 105 Membrane technology, arsenic in Thailand, 170 Methods for arsenic removal, performance and cost (table), 80 Mineral sources, arsenic (table), 22 Mineral sources, arsenic, 20 Mining pollution in the US, arsenic contamination, 44 Mining waste management, arsenic in Thailand, 168 Mining, arsenic contamination source, 42

Onsite testing, arsenic field kits, 61 ff. Options for treatment, arsenic contamination, 134 Oxidation, arsenic removal methods, 79, 82, 83 Physical treatment, arsenic removal, 134 Physical-chemical properties, arsenic, 18 Phytoremediation, arsenic removal in Thailand, 172 Phytoremediation, arsenic removal, 109 Plant analysis, arsenic, 72 Plant waste in Thailand, arsenic disposal, 174 Pollution by arsenic, anthropogenic sources, 18 Pollution sources, arsenic, 17 ff. Pond sand filters, arsenic mitigation in Bangladesh, 177 Rainwater harvesting, arsenic mitigation in Bangladesh, 178 Rainwater harvesting, arsenic mitigation option, 135 Rainwater harvesting, arsenic remediation, 116 Rainwater jars, arsenic in Thailand (illus.), 170 Read-F, arsenic removal unit, 97 Remediation technology, 77 ff.

194

Remediation, arsenic contaminated drinking water, 129 ff. Reverse osmosis, arsenic removal in Thailand, 171 Reverse osmosis, arsenic removal, 107 Reverse osmosis, operating parameters (table), 108 Risk mitigation, arsenic exposure, 11 SAFI filter, arsenic removal, 101 SAFI filter, features (table), 102 Sedimentation, arsenic removal in Thailand, 170 Sedimentation, arsenic removal technology, 82 Shapla filter, arsenic removal, 99 Sludge analysis, arsenic, 72 Solar oxidation, arsenic removal method, 83 Sono filter, arsenic removal, 100 Sorptive filtration, arsenic removal, 93 Sources of arsenic pollution, Thailand, 165 Star Filter, arsenic removal method (illus.), 88 Surface water contamination, arsenic, 40 Surface water treatment, arsenic remediation, 115 Technology alternatives to remove arsenic, ranking (table), 157 Technology for disposal, arsenic, 77 ff. Tetrahedron technology, arsenic removal, 104 Thailand arsenic pollution, managing mining waste, 168 Thailand arsenic pollution, rainwater jars (illus.), 170 Thailand health effects, arsenic, 167

Index

Thailand metal mining, arsenic poisoning, 43 Thailand, arsenic mitigation, 168 Thailand, arsenic plant waste management, 174 Thailand, arsenic pollution history, 164 Thailand, arsenic pollution sources, 165 Thailand, arsenic pollution, 163 ff. Thailand, arsenic removal by reverse osmosis, 171 Thailand, biotic removal of arsenic, 172 Tube well, arsenic removal unit (illus.), 89

Volcanic effects, arsenic contamination, 40 Water analysis of arsenic, electrochemical detection (table), 68 Water contamination, arsenic in Thailand, 166 Water contamination, arsenic removal (table), 80 Water monitoring criteria, arsenic, 62 Water purification, arsenic in Bangladesh, 11 Water remediation, arsenic in Thailand, 170 West Bengal arsenic contamination, source in groundwater, 25 Wood preserving industry, arsenic use, 47 Wood treatment, arsenic pollution, 48

X-Ray fluorescence, arsenic monitoring, 62