Report No. 31303 Towards a More Effective Operational Response Arsenic Contamination of Groundwater in South and East Asian Countries (In Two Volumes) Volume I: Policy Report March 28, 2005 Environment and Social Unit South Asia Region Water and Sanitation--WSP Document of the World Bank Table of Contents Acknowledgments v AbbreviationsandAcronyms vi Key Pointsof This Study viii Executive Summary ix 1. Background andIntroduction 1 2. Objectives andAudienceof the Study 1 3. Methodology 2 4. Key Findings 2 4.1 Continued Uncertaintyabout Epidemiology: How BigI s the Arsenic Threat? 2 4.1.1 Current Estimatesand Projections ofNumber ofArsenicosis Patients inAsia 4 4.1.2 Standards for Arsenic Concentrations inWater 6 4.2 What i s the Global and Regional Distribution of Arsenic Contamination? 7 4.3 Mechanisms o f Arsenic Mobilization: How DoesIt Get into the Groundwater? 10 4.4 How I s GroundwaterQuantity Relatedto Groundwater Quality? 11 4.5 Technical Options and Social Considerations:What Can and ShouldBe Done? 13 4.5.1 Sequencing 13 4.5.2 Technology Options 14 4.5.3 Social and Cultural Considerations 16 4.5.4 OperationalResponsesUndertakenby Countries So Far 17 4.6 The Economicsof Arsenic: Investment Choices- What and When? 19 4.6.1 Using Cost-Benefit Analysis (CBA) to InformArsenic Decisionmaking 19 4.6.2 Demand-SideManagement 21 4.6.3 The Economicsof Arsenic Mitigation 21 4.7 The PoliticalEconomy ofArsenic: What Are the Prospects for Action? 21 5. What Should Governments, DevelopmentPartners, andthe World BankDo? 24 5.1 Project-Level Action 25 5.2 National-Level Action 25 5.3 Global-Level Action 25 6. Conclusionsand Recommendations 26 Annex 1.PolicyMatrix: OperationalResponsesto Arsenic Contaminationin South andEast Asian Countries 28 References 30 ... 111 Tables Table 1. Scale ofArsenic Contamination: Selected Countries inSouth andEast Asia 3 Table 2. Current Population Identifiedwith Arsenicosis inEast and SouthAsianCountries 4 Table 3. Bangladesh: Estimated HeaIth Impact o fArsenic Contamination o f Tubewells 5 Table 4. EstimatedAnnual Deaths from Diarrheal Diseaseof Children under Five 6 Table 5. Current National Standards of Selected Countries for Arsenic inDrinkingWater 7 Table 6. Water Supply Options for Arsenic Mitigation 15 Table 7. Operational ResponsesUndertakenbyEast and SouthAsianCountries 18 Table 8. Conceptualized Incentive Matrix: Stakeholder Incentives for Action on Arsenic Issues 22 Figures Figure 1.World Distributiono fArsenic inGroundwater andthe Environment 8 Figure2. Locations ofHigh-Arsenic Groundwater Provinces inSouth andEast Asia 9 Figure3. Classification of Groundwater Environments Susceptible to Arsenic Contamination 12 Figure4. Practical Steps for Project-Level Responsesto Arsenic Contamination inGroundwater 13 Figure5. PrivatePublic InvestmentsinTubewells inBangladesh inthe Last 70 Years 17 Box Box 1.Results of the Economic CBA inthe Case of Bangladesh 20 ContentsofVolume11: TechnicalReport Paper 1.Arsenic Occurrence inGroundwater inSouth andEast Asia: Scale, Causes, andMitigation Paper 2. An Overview of Current Operational Responsesto the Arsenic Issue inSouth andEast Asia Paper 3. Arsenic Mitigation Technologies inSouth andEast Asia Paper 4. The Economics o fArsenic Mitigation iv Acknowledgments This study was conducted by the World Bank and the Water and Sanitation Programs o f South and East Asia, with additional financing from the Bank Netherlands Water Partnership Program and the UKDepartment for InternationalDevelopment (DFID).Itwas preparedby ateam composed ofKarin Kemper and Khawaja Minnatullah (co-task leaders); Amal Talbi, Ede Ijjasz-Vasquez, and Carla Vale (World Bank); Stephen Foster and Albert Tuinhof (World Bank Groundwater Management Advisory Team - GWMATE); and Jan-Willem Rosenboom (WSP-East Asia). The background papers were preparedby Pauline Smedley (British Geological Survey), Amal Talbi (World Bank), Feroze Ahmed (Bangladesh University of Engineering and Technology) and Phoebe Koundouri (Reading UniversityAJniversity College London and GWMATE). Valuable comments were provided by Caroline van den Berg, John Briscoe, and Nadim Khouri (peer reviewers); and Junaid Ahmad, Guy Alaerts, Ejaz Ghani, Rachel Kauhann, Smita Misra, Rick Pollard, Jamal Saghir and Luiz Tavares. In addition, the study benefited fiom numerous comments from participants at the session held at the World Bank Water Week in Washington, D.C. in February 2004 and at the Regional Operational Responses to Arsenic Workshop held in Kathmandu, Nepal, in April 2004. The team would like to express their gratitude to Jeffrey Racki (Acting Director, South Asia Social and Environment Unit), Alastair McKechnie (Country and Regional director, South Asia Region) and Shantayanan Devarajan (Chief Economist, South Asia Region) for their guidance during the study. The team would also like to thank Shyam Ranjitkar (World Bank, Nepal Office) and Dibya Ratna Kansakar from the Department o f Irrigation in Nepal for hosting the workshop in Kathmandu, the respondents o f the study survey and the many colleagues in the World Bank, WSP and numerous organizations who providedinformation for this study. -The study documentation was edited by John Dawson. Vandana Mehra (WSP) manageddesign andpublication arrangements. This volume i s a product o f the staff o f the World Bank and o f the Water and Sanitation Program. The findings, interpretations, and conclusions expressed inthis paper do not necessarily reflect the views of the Executive Directors o f The World Bankor the governments they represent. The World Bank does not guarantee the accuracy o f the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part o f The World Bank concerning the legal status o f any territory or the endorsement or acceptance o f such boundaries. V AbbreviationsandAcronyms The following list includes all abbreviations and acronyms used throughout Volumes Iand I1of the report. AAN AsianArsenic Network AAS atomic absorption spectrometer/spectrometry A E S atomic emission spectrometry AIIH&PH All IndiaInstitute ofHygiene andPublic Health APSU Arsenic Policy Support Unit (Bangladesh) As arsenic ASV anodic stripping voltammetry AusAID Australian Agency for International Development AWWA American Water Works Association BAMWSP Bangladesh Arsenic Mitigation Water Supply Project BGS British Geological Survey BUET Bangladesh University o f Engineering and Technology CBA cost-benefit analysis CCA chromated copper arsenate CEPIS Pan American Center for Sanitary Engineering and Environmental Sciences (Peru) CGIAR Consultative Group on International Agricultural Research Danida DanishAgency for International Development DF discount factor DOC dissolved organic carbon DPHE Department o f Public Health Engineering(Bangladesh) DWSS Department o f Water Supply and Sewerage (Nepal) EAWAG Swiss Federal Institute for Environmental Science and Technology EPA Environmental ProtectionAgency (United States) GDP gross domestic product GF-AAS graphite furnace-atomic absorption spectrometry GPL General Pharmaceutical Ltd. GPS global positioning system GWMATE World Bank Groundwater Management Advisory Team HG-AAS hydridegeneration-atomic absorption spectrometry HG-AFS hydridegeneration-atomic fluorescence spectrometry ICP inductively coupled plasma ICP-MS inductively coupled plasma-mass spectrometry IRC International Water and Sanitation Center (formerly International Reference Center for Community Water Supply) JICA Japan International Cooperation Agency Lao PDR Lao People's Democratic Republic MDG MillenniumDevelopment Goal M S mass spectrometry N A M I C National Arsenic Mitigation Information Center (Bangladesh) NASC National Arsenic Steering Committee (Nepal) NGO nongovernmental organization vi NPV net present value NRC NationalResearch Council (UnitedStates) NRCS Nepal Red Cross Society NTU nephelometric turbidityunit PAHO PanAmerican HealthOrganization P V present value SDDC silver diethyldithiocarbamate SORAS solar oxidation andremoval of arsenic TCLP toxic characteristic leachingprocedure UNCHS UnitedNations Centre for HumanSettlements UNDP UnitedNations Development Program UNICEF United Nations Children's Fund WHO World Health Organization WRUD Water Resources UtilizationDepartment (Myanmar) UnitsofMeasurement I-Lg microgram m g milligram k g kilogram L liter I-LgL- I micrograms per liter m g milligrams per liter cm centimeter m meter km kilometer Unitsof Currency (January 2004) 1US$ =58 taka (Tk) (Bangladesh) 1US$ =48 rupees (Rs) (India) vii Key Points of This Study The following are the key points to emerge from this study o f operational responses to arsenic contamination o f groundwater inSouth and East Asia: Many millions o f people throughout South and East Asia inhabit areas where certain hydrogeological processes mean that groundwater may be contaminated naturally with levels of arsenic that constitute a danger to humanhealth. A considerable amount of research has been carried out into the causes and effects of this contamination and possible mitigation measures, but significant uncertainties remain which have to be factored inwhen attempting to define a balancedpolicy response. A number of operational responses have already been implemented.This study reviews the current status o fboth research and operational responses. Unfortunately, the responses to arsenic contamination have so far lacked cohesion, and the problem needs to be addressed in a much more integrated and strategic manner in future, primarily within the water supply sector. For example, arsenic mitigation needs to be a primary consideration in any new water supply or irrigation interventions in the identified areas. The same consideration needs to be applied to institutional approaches to developing arsenic mitigation strategies, which need to take account o f the importance o f building capacity and providing incentives for different actors to respond to the arsenic problem. Arsenic is not the only problem relating to drinking water supply. Not only may other inorganic consituents (such as iron and manganese) be present, but another major problem is poor bacteriological water quality, which in fact claims many more lives annually and over time than arsenic contamination. Thebe problems occur at scales whose resolution i s beyond current available resources; it will therefore be necessary to consider trade-offs that take into account the costs and benefits o f a range o f mitigation measures. This study suggests a methodology by which a cost-benefit analysis can help resolve this difficult issue. The complexity o f the arsenic problem is such that mitigation measures cannot wait for definitive answers to the issues. Mitigation activities will, in many cases, have to proceed against a background o f uncertainty. This report outlines what can be done at project, national, and global levels. At all levels it is important that governments overcome the constraints related to such a politically sensitive issue and drive forward measuresthat can mitigate the effects o f arsenic contamination. ... Vlll Executive Summary BackgroundandIntroduction i. Thedetrimentalhealtheffectsofenvironmentalexposuretoarsenichavebecomeincreasingly clear in the last few years. High concentrations detected in groundwater from a number o f aquifers across the world, including in South and East Asia, have been found responsible for health problems ranging from skin disorders to cardiovascular disease and cancer. ii. Theproblemhasincreasedgreatlyinrecentyearswiththegrowinguseoftubewellstotap groundwater for water supply and irrigation. The water delivered by these tubewells has been found in many cases to be contaminated with higher than recommended levels o f arsenic. In the study region, countries affected include Bangladesh (theworst affected), India, Myanmar, Nepal, and Pakistan (South Asia); and Cambodia, China (including Taiwan), Lao People's Democratic Republic, and Vietnam (East Asia). iii. Thisstudy concentrates onoperationalresponsesto arsenic contaminationthat maybeof practical use to actors who invest in water infrastructure in the affected countries, including governments, donors, development banks, and nongovernmental organizations (NGOs). Objectives andAudience of the Study iv. The objectives o f this study are (a) to take stock o f current knowledge regarding the arsenic issue; and (b) to provide options for specific and balanced operational responses to the occurrence o f arsenic in excess o f permissible drinking water limits in groundwater inAsian countries, while taking into account the work that has already been carried out by many different stakeholders. v. The study provides information on (a) occurrence o f arsenic in groundwater; (b) health impacts o f arsenic; (c) policy responses by governments and the international community; (d) technological options for and costs o f arsenic mitigation; and (e) economic aspects o f the assessment and development o f arsenic mitigation strategies. The focus o f the study i s on rural rather than urban areas, due to the particular difficulties associated with applying mitigation measures inscattered rural communities. vi. The study is structured as follows: Volume I:Policy Report. This report summarizes the main messages o f Volume 11, and highlights the policy implications o f arsenic mitigation. Volume I1comprises four specialist papers: Paper 1. Arsenic Occurrence in Groundwater in South and East Asia: Scale, Causes, and Mitigation Paper 2. An Overview o f Current Operational Responses to the Arsenic Issue in South and East Asia Paper 3. Arsenic Mitigation Technologies in South and East Asia b Paper 4. The Economics o f Arsenic Mitigation The Scale of the Arsenic Threat vii. In South and East Asia an estimated 60 million people are at risk from high levels of naturally-occurring arsenic in groundwater, and current data show that at least 700,000 people in the region have thus far been affected by arsenicosis. However, although the negative health effects o f arsenic ingestion in general, and the specific impact o f ingestion o f arsenic- contaminated groundwater, have both been widely studied, there is still no clear picture o f the epidemiology o f arsenic in South and East Asia, and uncertainty surrounds such issues as the spatial distribution o f contamination; the symptoms and health effects o f arsenic-related ix diseases, and the timeframe over which they develop; and the impact o f arsenic compared to other waterborne diseaseswhose effects may be more immediate. viii. While arsenic is clearly an important public health threat, it needs to be noted that morbidity and mortality due to other waterborne diseases is also a serious health issue. Therefore, mitigation measures to combat arsenic contamination in South and East Asia need to be considered within the wider context o fthe supply o f safe water. ix. Due to the carcinogenic nature o f arsenic, the World Health Organization (WHO) recommends a maximumpermissible concentration for arsenic in drinkingwater o f 10 pg L-' (micrograms per liter), which has been adopted by most industrial countries. Most developing countries still use the former WHO-recommended concentration o f 50 1-18L-'as their national standard, due to economic considerations and the lack o f tools and techniques to measure accurately at lower concentrations. Further studies are needed to assess the relationship between levels o f arsenic and health risks in order to quantify the inevitable trade-offs at different standards between such considerations as health risks, the ability o fpeople to pay for safe water, andthe availability o f water treatment technology. Distributionof Arsenic Contamination x. The concentration o f arsenic in natural waters, including groundwater, is usually below the WHO guideline value of 10p g L-'.However, arsenic mobilization is favored under some specific hydrogeochemical conditions, especially highly reducing (anaerobic) conditions, which can bring about the dissolution o f iron oxides and the associated desorption o f arsenic. In South and East Asia such conditions tend to occur in the shallower parts of Quaternary aquifers underlying the region's large alluvial and deltaic plains (Bengal basin, Irrawaddy delta, Mekong valley, Red River delta, Indus plain, Yellow River plain). (Some localized groundwater arsenic problems relate to ore mineralization and mining activity, which are not the focus o f this study.) Recent hydrogeochemical investigations have improved our knowledge o f the occurrence and distribution o f arsenic in groundwater, although some uncertainty remains regarding the source, mobilization, and transport o f the element in aquifers. xi. One o f the important findings o f recent detailed aquifer surveys has been the large degree o f spatial variability in arsenic concentrations, both with depth and even laterally at the same depth over distances o f a few hundred meters. Temporal variability also occurs, though insufficient monitoring has been carried out to establish a clear picture o f variations inarsenic levels over different timescales. Arsenic MitigationMeasures xii. Arsenic mitigation requires a sequence o f practical steps involving enquiry and associated action. Assessing the scale o f the problem (now and over time) involves field testing, laboratory testing, and monitoring; identifying appropriate mitigation strategies involves technological, economic, and sociocultural analysis o f possible responses; and implementation involves awareness raising and direct action by governments, donors, NGOs, and other stakeholders at local, national, and regional levels. Sustainability in the long run remains a major challenge. xiii. The two main technological options for arsenic mitigation are (a) switch to alternative, arsenic-free water sources; or (b) remove arsenic from the groundwater source. Alternatives in the first category include development of arsenic-free aquifers, use of surface water and rainwater harvesting; alternatives in the second category involve household-level or community-level arsenic removal technologies. For each option there will be a wide range of design specifications and associated costs. xiv. Despite continuing uncertainty regarding arsenic occurrence and epidemiology, the lethal nature and now well-established effects o f arsenic exposure in South and East Asia make i t necessary that informed choices and trade-off decisions are made to address arsenic X contamination o f drinking water sources and the scope and extent o f mitigation measures, within the context ofthe development ofthe water sector andthe wider economy. xv. Accordingly, a simple cost-benefit methodology has been developed that takes into account data limitations and provides decisionmakers with an approach for rapid assessment o f the socioeconomic desirability o f different mitigation policies under various scenarios. In particular, the methodology permits an analysis o f options in order to choose between different approaches indealing with (a) the riskthat arsenic might be found inan area where a project is planned; and (b) the risk mitigation options where a project's goal is arsenic mitigation per se. xvi. Demand-side perspectives are an important consideration for designing arsenic mitigation measures that meet the requirementso f households and communities. For example, are users willing to pay for an alternative such as piped water? Demand preferences can be assessed through contingent valuation or willingness to pay studies and can provide important guidance to decisionmakers. There i s a need to strengthen institutional capacities in the countries to carry out such assessments. The Political Environment of Arsenic Mitigation xvii. Arsenic has become a highly politicized topic in the international development community and within some affected countries due to its carcinogenic characteristics and due to the earlier failure to consider it as a possible natural contaminant in groundwater sources. This factor makes rational analysis o f the issue difficult and highlights the fact that application o f mitigation measuresneeds to consider the political as well as the social and economic climate. The scattered rural communities most affected by arsenic contamination often have limited political presence and are inparticular need o f support. xviii. Governments that want to address the arsenic issue will therefore have to take a stronger lead role in their countries and on the international plane. This goes both for more strategic research and knowledge acquisition regarding arsenic in their countries, as well as for the choice and scope o f arsenic mitigation activities. The Importance of an Effective Operational and Strategic Approach xix. Significant strides have been made since arsenic was first detected in drinking water tubewells in Eastern India and Bangladesh in the early 1980s and early 1990s, respectively. However, a range o f factors - including projected populationgrowth inthe region, continuing private investment in shallow tubewells, and the drive towards achievement o f the MillenniumDevelopment Goal related to safe water supply- add to the urgency o f adopting a more strategic approach for effective action at project, national, and international levels. xx. A t project level, any interventions that consider using groundwater as a source must involve an assessment of whether occurrence o f arsenic would affect the outcome o f the project. Such an assessment would include consideration o f technical factors (such as screening and possible mitigation technologies), social and cultural factors, and economic factors (including a cost-benefit or least-cost analysis). xxi. Some countries have taken arsenic to the national level o f attention, including Bangladesh, Nepal, and Cambodia. Others, such as India, Pakistan, and China, have only started to address the issue, while in others international organizations such as UNICEF and local NGOs and universities are the focal points for arsenic-related activities. Although the characteristics o f arsenic contamination are unique to each affected country, study results suggest that three simple steps would help governments more effectively address the problem now and in the future: (a) encourage further research in potentially arsenic-affected areas in order to better determine the extent o f the problem; (b) ensure that arsenic is included as a potential risk factor in decisionmaking about water-related issues; and (c) develop options for populations inknown arsenic-affected areas. xi I xxii. A t the global level, focused research on the chemistry o f arsenic mobilization and the dose- response relationships for arsenic are o f vital importance in formulating a more effective approach. If governments and the international community are to achieve the MDGs inwater supply and sanitation then the knowledge gaps regarding arsenic need to befilled, notably by (a) further epidemiological research directly benefiting arsenic-affected countries; (b) socioeconomic research on the effects o f arsenicosis, understanding behavior and designing demand-based packages for the various arsenic mitigation techniques; and (c) hydrogeological and hydrochemical research xxiii. It also needs to be made clear that, due to the nature of arsenic itself, in the not-so-distant future there will be diminishing returns on investments in scientific arsenic research to reduce uncertainty. The important challenge will be to identifythose areas where improved research- level data collection is likely to provide a major return. For other areas the main question will be how to manage inthe face o f unavoidable and continuing uncertainty. xxiv. Accordingly, the international dialogue should shift towards targeted research priorities that address these issues. This would also include the pursuit o f the research agenda regarding arsenic in the food chain. Both the World Bank and a number o f development partners are contributors to the Consultative Group on International Agricultural Research (CGIAR) and this organization would lend itself to buildingup a coherent and focused research agenda on this topic in order to provide decisionmakers with guidance regarding arsenic-contaminated groundwater. xii 1. BackgroundandIntroduction The detrimental health effects o f environmental exposure to arsenic have become increasingly clear in the last few years. Drinking water constitutes one o f the principal pathways o f environmental arsenic exposure in humans. High concentrations detected in groundwater from a number o f aquifers across the world, and specifically in South and East Asia, have been found responsible for health problems ranging from skin disorders to cardiovascular disease and cancer. Food represents a further potential exposure pathway to arsenic in instances where crops are irrigated with high-arsenic groundwater, or where food is cooked with arsenic-contaminated water. However, the relative impact on humanhealth is not as yet quantified and is inneed o f further study. With groundwater-based water supply and irrigation projects being implemented across the arsenic-affected regions o f Asia, there is a serious need to address this issue not only for a single country like Bangladesh - the most well-known and dramatic case - but also in a regional context, as more countries in the region have been reported to have higher than the permissible standards o f arsenic in groundwater. In South Asia, other countries affected by arsenic include India, Myanmar, Nepal, and Pakistan. In East Asia Cambodia, China (including Taiwan), Lao People's Democratic Republic, and Vietnam are affected. The increasing recognition o f the wide geographic spread o f the problem has provided the motivation to carry out t h s study at a cross-regional scale. Current literature available on arsenic tends to be conceptual, analytical, or prescriptive in terms of standard setting, with little coverage o f concrete operational responses for those actors who invest in water infrastructure in these countries, such as governments, development banks, nongovernmental organizations (NGOs),and donors. Since the potential health hazards o f arsenic are now known it is necessary to frame and implement responses in operational terms, outlining steps that minimize the health risks presented by water supply projects whose intended benefits may be negated by the harmfulmedium or long-term effects o f arsenic exposure. 2. Objectives andAudience of the Study The objectives o f this study are to (a) take stock o f current knowledge regarding the arsenic issue; and (b) provide options for specific and balanced operational responses to the occurrence o f arsenic inexcess o f permissible limits ingroundwater inAsian countries. It also aims to provide stakeholders with tools and guidance to analyze the extent o f the arsenic contamination in their respective countries and regions and help them to develop appropriate responses while taking into account the work that has already been carried out by many differentstakeholders. The study thus provides information on (a) state-of-the-art knowledge about natural occurrence o f arsenic in groundwater, including spatial distribution and hydrogeochemical aspects; (b) current state o f knowledge regarding known and potential health impacts o f arsenic; (c) previous policy responses by governments and the international community (development partners, civil society, and academia); (d) technological options for and costs o f arsenic mitigation; and (e) economic aspects o f the assessment and development o f arsenic mitigation strategies. The study also indicates steps to be taken by decisionmakers regarding investment projects that use groundwater, both in terms of specific considerations during project design and implementation and interms o f relevant upstream sector analysis. Thus, the principal target audiences of this study are governments and their development partners, including international development banks,bilateral donors, and development NGOs who are active in water-related issues in the region. Within these groups, it is expected that decisionmakers and managers will primarily focus on the Policy Report (Volume Io f this study), which provides a synthesis o fthe comprehensive review and an analysis o fthe subject matter, and distils the policy implications. - 1 - Arsenic Contamination of Groundwater in South and East Asian Countries -Volume I-Policy Report Technical staff and water sector professionals will also have an interest in Volume 11, the Technical Report, which comprises the detailed study background papers providing a wealth o f state-of-the-art information and references to specialized literature. Volume I1 includes four papers, namely: 0 Paper 1. Arsenic Occurrence in Groundwater in South and East Asia: Scale, Causes, and Mitigation Paper 2. An Overview o f Current Operational Responses to the Arsenic Issue in South and East Asia Paper 3. Arsenic Mitigation Technologies in South andEast Asia Paper 4. The Economics o f Arsenic Mitigation While the papers are complementary, they have been prepared as stand-alone products in order to serve as reference literature for readers who require more detail about each o f these topics. It is expected that academics and a wider civil society audience who are involved in water resources development issues, and specifically water supply and sanitation, will also benefit from the study. 3. Methodology Comprehensive literature reviews were undertaken for all background papers. Papers 1 and 3 largely draw on the body o f internationally available existing analysis, and provide state-of- the-art overviews. Paper 4 also draws on available information, but in addition develops a simple and pragmatic methodology for decisionmakers to deal with the economics o f interventions regarding arsenic and to make rational choices between potential (different) strategies. Paper 2 is based on an extensive literature review and on a survey administered to government officials, international organizations, NGOs, and researchers. All papers draw on the feedback received at the session organized during the World Bank Water Week in Washington, D.C. inFebruary 2004, and on the results o f the Regional Operational Responses to Arsenic Workshop subsequently held inKathmandu, Nepal, during 26 and27 April 2004. The study thus links these strings o f information more closely with one another, and draws out the broader policy and institutional implications for issues currently under discussion. The study also evaluates in-depth experience from one country, Bangladesh, the most affected and active int h s regard, and compares its findings with that o f other countries where appropriate, to understand the impact o f past and current interventions on arsenic mitigation and knowledge generation, and to provide policy recommendations for future interventions inthis area. 4. Key Findings 4.1 ContinuedUncertaintyabout Epidemiology:How BigI s the Arsenic Threat? Over the years, a number o f studies have been conducted to assess and quantify the impact o f ingesting arsenic-contaminated groundwater.' However, a surprising finding o f the present study has been that - in spite o f more than a decade o f research, studies, and other interventions regarding arsenic in South and East Asia - no clear picture has yet emerged o f the epidemiology of arsenic inthe region. Estimates o fthe (future) health impacts of arsenic ingestion are mostly based on and extrapolated from data for the United States o f America and Taiwan (China), and their validity for interpretation at a wider scale is therefore frequently questioned. See Papers 1and2 inVolume I1for more detailedinformation. - 2 - Arsenic Contamination of Groundwater in South andEast Asian Countries -Volume I-Policy Report Globally, the only large-scale screening, carried out in Bangladesh through the Bangladesh Arsenic Mitigation Water Supply Project (BAMWSP) and other Government o f Bangladesh funding sources and donors, which included patient identification, indicated that far fewer people show signs o f arsenicosis* than could be expected from extrapolation o f the United States and Taiwan epidemiological data.3 The negative health effects o f arsenic ingestion have been documented for the last 200 years. Inspite ofthe uncertainty regarding exact numbersit is clear that there are major effects, but it is not yet clear how widespread or serious these are or what the relationship of disease to exposure is in different settings. It is, however, obvious that millions o f people are at risk from arsenic-induced diseases. Table 1 summarizes, for the currently affected countries for which data are available, the estimated area andpopulation at risk, and the levels o f arsenic in groundwater. _____ Table 1. Scale o fArsenic Contamination: Selected Countries inSouth andEast Asia Location Areal extent (h') Population at risk" Arsenic range (pg L-') Alluvial/deltaic/lacustrine plains Bangladesh 150,000 35,000,000 <1-2,300 China (Inner Mongolia, Xinjiang, Shanxi) 68,000 5,600,000 404,400 India (West Bengal) 23,000 5,000,000 <10-3,200 Nepal 30,000 550,000 <10-200 Taiwan (China) 6,000 (?)10,OOOb lo-1,800 Vietnam 1,000 10,000,000` 1-3,100 Myanmar (?)3,000 3,400,000 Cambodia (?)<1,000 320,000d -Not available. a. Estimatedto be drinkingwater with arsenic >50 pgL-'.From Smedley 2003 and data sources therein. b. Beforemitigation. c. UnitedNations Children's Fund(UNICEF) estimate. d. Maximum. Source: Regional Operational Responsesto Arsenic Workshop inNepal, 26-27 April 2004. Table 1 shows that the estimated population at risk from natural arsenic contamination in groundwater in Asian countries is at least 60 million. What i s not clear is (a) how many people in these risk areas will be affected by arsenic-related disease and withm which timeframe (especially compared with other waterborne diseases where effects may be more immediate, such as diarrhea in under-five-year-olds, which is often fatal); and (b) what exactly the health effects are going to be; there is still uncertainty whether skin lesions, Arsenic has various health effects ranging from arsenicosisto skin cancers and intemal cancers. So far there is still no widely accepted definition of what constitutes arsenicosis. Generally, the term i s usedfor the pattem of skin changesthat occur after chronic ingestionof arsenic. The BAMWSP database is currently being analyzed with regardto its consistency. At a local scale, BAMWSP data were corroborated by a controlled local study inAzaihar District carried out by the University of Columbia in2003 (Van Geenandothers2003). - 3 - Arsenic Contamination of Groundwater in South and East Asian Countries -Volume I-Policy Report typically the most visible expression o f arsenicosis, are the first symptom or if internal cancers and other ailments can also bepresent inthe absence o f skinlesions. Generally, it can be said that far more rural than urbanpopulations are at risk.This is due to the fact that it i s easier and more affordable to implement arsenic removal technologies in urban areas. For rural domestic water supply the situation is completely different; the distinctive feature o f arsenic contamination o f groundwater in South and East Asia is the very large number o f scattered small communities affected, constituting a major financial and management challenge. Accordingly, this report primarily addresses the rural dimension o f the arsenic issue. 4.1.1 CurrentEstimatesand Projectionsof Numberof Arsenicosis PatientsinAsia The estimates o f the current number o f patients with arsenicosis for countries o f East and South Asia are summarized intable 2. Table 2. Current PopulationIdentifiedwith Arsenicosis inEast andSouth AsianCountries Regionlcountry Numberof arsenicosis Year of first patients identifiedso far discovery East Asia Cambodia - 2000 China provinces: 522,566 InnerMongolia 1990s Xinjiang 1983 Jilin, Shanxi, Ningxia, Qinghai, Anhui, Beijing 2001-2002 Taiwan 1960s Lao PDR Myanmar 1999 Vietnam 1998 South Asia Bangladesh 10,000 (partial results) 1993 India(West Bengal) 200,000 1978 Nepal 8,600 1999 Pakistan 242 cases per 100,000 2000 people basedon the results of 10districts -Not available. Table 2 shows that there are approximately 700,000 people who have been affected by arsenicosis. For Bangladesh inparticular, with regard to projected future cases, an estimate o f the arsenic-related health burden, provided in Ahmed (2003) and adjusting data from the UnitedStates Environmental Protection Agency (EPA) to Bangladesh conditions, concluded that total skin cancer would affect 375,000 people. Using data from a more detailed survey o f the data currently available inthe literature, Maddison, Luque, and Pearce (2004) estimated the impact on health o f arsenic inBangladeshas indicated intable 3. - 4 - Arsenic Contamination o f Groundwater in South and East Asian Countries -Volume I-Policy Report Table 3. Bangladesh: EstimatedHealth Impact of Arsenic Contaminationo fTubewells Impact on healthhype of illness Males Females Combined Cancer cases: Fatal cancerdyear 3,809 2,718 6,528 Nonfatal cancerdyear 1,071 1,024 2,095 Total cancer fatalities accumulated over 50 190,450 135,900 326,400 years Arsenicosis casesa: Keratoses 277,759 74,473 352,233 Hyperpigmentation 654,718 316,511 971,230 cough 21,823 68,887 90,712 Chest sounds 144,831 67,025 211,858 Breathlessness 93,247 176,874 270,122 Weakness 132,927 240,176 373,104 Glucosuria 67,887 63,55 1 131,439 Highbloodpressure 94,396 88,366 182,762 Total arsenicosis cases in each year 1,487,588 1,095,863 2,583,460 a. Figures indicate average number of cases occurring ineach year (not number of new cases). Source: Maddison, Luque, and Pearce 2004, p. 32. The estimates suggest that in Bangladesh 6,500 people will die from cancer every year, a total o f 326,000 people in a period o f 50 years, while 2.5 millionpeople will develop some kindo f arsenicosis over that period. So far, these two figures are the only quantification o f the potential arsenic-related health burden. They depend heavily on epidemiological assumptions and demonstrate how the lack o f reliable epidemiology information adds uncertainties to the projectednumber o fpeople at risk. When comparing morbidity and mortality due to arsenic with that o f other waterborne diseases, bacteriological contamination is a much more serious issue. A study conducted by the World Health Organization (WHO) and UNICEF in2000 indicated that approximately 4 billion cases o f diarrhea are reported globally every year, causing 2.2 million deaths, mostly among children under five, and intestinal worms infect about 10% o f the population in the developing world (WHO-UNICEF 2000). Diarrhea and worm infestation are two major waterborne public health threats in South Asia. A 2000 survey in Bangladesh by the Bureau o f Statistics and UNICEF indicated that about 110,000 children under five die due to diarrhea every year (Bureau o f Statistics-UNICEF 2002). The situation is similar or even worse in Nepal, India, and Pakistan. Table 4 shows the estimated annual deaths o f children under five due to diarrheal disease in the countries under study. Estimated deaths vary from 650,000 to 1.3 million per year, depending on whether the assumption is made that 15% or 30% o f total deaths are due to diarrheal disease. The figures cannot be directly compared to arsenicosis or arsenic-induced - 5 - Arsenic Contaminationof Groundwater in South and East Asian Countries -Volume I-Policy Report fatalities because they are estimates for the entire areal extent o f the countries, not just for those areasthat are arsenic affected. Nevertheless, they show the magnitude o f the burden due to diarrheal disease, as an indicator o fthe impact o f inadequate water supply. Two conclusions can be drawn here. First, the public health effects o f arsenic are areality and they need to be taken seriously. As the effects o f arsenic are long term it is likely that arsenic- related disease, with and without fatal outcomes, i s going to increase over the coming decades, affecting hundreds o f thousands o f people. Second, with waterborne disease claiming so many lives annually, it is important to integrate arsenic considerations into a rational approach within the overall context o f waterborne public health threats. Further investment in safe water supply i s a necessity and arsenic is but one o f the considerations in this regard. Table4. Estimated Annual Deaths from Diarrheal Disease o fChildren under Five Country Region Annual total Low estimate High estimate mortality of (15%ofchild (30% of child children under the mortalityunder 5 mortalityunder 5 age of fivea years dueto years due to diarrhea) diarrhea) Bangladesh South Asia 323,000 48,450 96,900 Cambodia East Asia 65,000 9,750 19,500 China EastAsia 735,000 110,250 220,500 India South Asia 2,346,000 351,900 703,800 Lao PDR East Asia 20,000 3,000 6,000 Myanmar SouthAsia 129,000 19,350 38,700 Nepal South Asia 74,000 11,100 22,200 Pakistan South Asia 579,000 86,850 173,700 Vietnam EastAsia 64,000 9,600 19,200 South Asia total 3,451,000 517,650 1,035,300 EastAsia total 884,000 132,600 265,200 Total 4,335,000 650,250 1,300,500 a. Data from UNICEF website. 4.1.2 Standardsfor Arsenic ConcentrationsinWater Due to the carcinogenic nature o f arsenic, the WHO has issued a provisional guideline for maximum permissible concentration o f arsenic in drinkingwater o f 10 pg L-' (microgramper liter). WHO guidelines are intended as a basis for setting national standards to ensure the safety o f public water supplies and the guideline values recommended are not mandatory limits. Such limits are meant to be set by national authorities, considering local environmental, social, economic, andcultural conditions. The WHO-recommended maximum permissible value is usually related to acceptable health risk, defined as that occurring when the excess lifetime risk for cancer equals (that is, 1 person in 100,000). However, inthe case of arsenic, the United States EPA estimates that this risk would mean a standard as low as 0.17 pg L-',which is considered far too expensive to - 6 - Arsenic Contaminationof Groundwaterin Southand EastAsian Countries-Volume I- Policy Report achieve, even for industrial countries such as the United States. The EPA thus conducted an economic study with concentrations o f 3, 5, 10, and 20 pg L-' and concluded that for the United States a standard o f 10 yg L-' represents the best trade-off among health risks, the ability o f people to pay for safe water, and the availability of water treatment technology. Thus, even this stricter standard, which has been adopted by most industrial countries, is a compromise. Most developing countries still use the former WHO-recommended concentration o f 50 pg L-' as their national standard for arsenic in drinking water, partially due to economic considerations and the lack o f tools and techniques to measure accurately at such low concentrations (table 5). Here, it is important to note that even though the exact health effects o f an arsenic concentration o f 50 pg L' have not been quantified, many correlations between internal cancer and lower concentration o f arsenic have also beenfound. Therefore, while the respective current national standards are valid and followed by international agencies such as the World Bank, epidemiological studies at these lower concentrations are o f utmost importance in providing a better basis for decisionmakers in developing countries to understand the risks they are taking by adhering to their hgher national standards and the trade-offs involved ininvesting inarsenic mitigation compared to other development needs. Table 5. Current National Standards o f Selected Countries for Arsenic inDrinkingWater ~ Countryiregion Standard: pg L-' Country Standard: pgL-' Australia (1997) 7 Bangladesh(1997) 50 EuropeanUnion (1998) 10 Cambodia 50 Japan (1993) 10 China 50 USA (2002) 10 India 50 Vietnam 10 LaoPDR (1999) 50 Canada 25 Myanmar 50 Nepal 50 Pakistan 50 Regarding arsenic concentration in irrigation water, neither international agencies nor individual countries propose any recommended maximum permissible values and hrther research is needed to come to conclusive recommendations inthis regard (see section 5). 4.2 What is the GlobalandRegionalDistributionof Arsenic Contamination? The concentration o f arsenic innatural waters, including groundwater, is typically below the WHO provisional guideline value for arsenic in drinking water o f 10 pg L-'. However, arsenic mobilization in water is favored under some specific geochemical and hydrogeological conditions and concentrations can reachtwo orders o f magnitude higher than this in the worst cases. Most o f the extensive occurrences o f high-arsenic groundwater are undoubtedly of natural origin, that is to say they involve the mobilization o f arsenic naturally present in the ground and not the discharge o f pollutants at the land surface, although the extent to which mobilization can be accelerated by groundwater pumping i s still open to some question. Figure 1shows the distribution o f documented cases o f arsenic contamination in groundwater and the environment worldwide. Many of these cases are related to areas o f mineralization andmining activity and a few are associated with geothermal sources. While these cases can be severe, with high concentrations o f arsenic in waters, sediments, and soils, their lateral scale is usually limited. Other areas with recognized high-arsenic groundwater are not - 7 - Arsenic Contaminationof Groundwater inSouth andEastAsian Countries-Volume I-PolicyReport associated with obvious mineralization and mining or geothermal activity. Some o f these occur in major aquifers and may be potentially much more serious because they occupy large areas and can provide drinkingwater to large populations. This study deals with these areas rather than those where arsenic release is due to miningor geothermal activities. Figure1.World Distribution o fArsenic inGroundwater and the Environment IBRD33757 Arsenic-aflected aquifers 0 Arsenicrelatedtominingand mineralization D Geothermalwaters Source: Modifiedafter SmedleyandKinniburgh2002. Note:InChina, arsenic has furtherbeenidentifiedinthe provincesof Jilin, Qinghai, Anhui, Beijing,and Ningxia(reportedat RegionalOperationalResponses to Arsenic Workshop inNepal,26-27 April 2004). InIndia, further affectedstates are Assam, ArunachalPradesh,Bihar, Manipur, Meghalaya,Nagaland, UttarPradeshandTripura. Major alluvial plains, deltas and some inland basins composed o f young sediments are particularly prone to developing groundwater arsenic problems. Several o f these aquifers around the world have now been identified as having unacceptably high concentrations o f arsenic. These include not only the alluvial and deltaic aquifers in parts o f Asia, but also inland basins in Argentina, Chile, Mexico, the southwestern United States, Hungary, and Romania. Important differences exist between these regions, but some similarities are also apparent. The majority o f the high-arsenic groundwater provinces are in young unconsolidated sediments, usually o f Quaternary age, and often o f Holocene deposition o f less than 12,000 years in age. These aquifers do not appear to contain abnormally high concentrations o f arsenic-bearing minerals but do have geochemical and hydrogeological conditions favoring mobilization of arsenic and its retention in solution. Manyo fthe world's aquifers with higharsenic levels are located inthose areasof Asia where large alluvial and deltaic plains occur, particularly around the perimeter of the Himalayan mountain range.In South Asia, naturally occurring arsenic in groundwater was initially identified West Bengal, India and in Bangladesh in the early 1980s and 1990s respectively. - 8 - Arsenic Contamination of Groundwater in South and East Asian Countries -Volume I-Policy Report Since then governments, donors, international organizations, NGOs, and research institutions have increased testing o f groundwater sources. As a result, naturally occurring arsenic has now been identified in the groundwater o f all the countries in South and East Asia that are the subject o f this study. Figure 2 shows the locations o f high-arsenic groundwater provinces inthe countries of South and East Asia. There may be other Quaternary aquifers with high groundwater arsenic concentrations that have not yet been identified, but since awareness o f the arsenic problem has grown substantially over the last few years, these are likely to be on a smaller scale than those already identified Figure2. Locations o f High-Arsenic Groundwater Provinces inSouth and East Asia Many of the health consequences resulting from contaminated groundwater have emerged in relatively recent years as a result o f the increased use o f groundwater from tubewells for drinkingand irrigation. Interms o f numbers o f groundwater sources affected and populations at risk problems are greatest in Bangladesh, but have also been identified in India (West Bengal, and more recently Assam, Arunachal Pradesh, Bihar, Manipur, Meghalaya, Nagaland, Tripura and Uttar Pradesh), China, including Taiwan, Vietnam, Thailand, Cambodia, Myanmar, and Nepal. Occasional high-arsenic groundwaters have also been found inPakistan, althoughthe occurrences there appear to be less widespread. - 9 - Arsenic Contaminationof GroundwaterinSouthandEastAsianCountries-Volume I-PolicyReport Hence, much o f the distribution i s linkedto the occurrence o f young (Quaternary) sediments in the region's large alluvial and deltaic plains (Bengal basin, Irrawaddy delta, Mekong valley, Red River delta, Indus plain, Yellow River plain). Although groundwater arsenic problems have been detected insome middle sections o f the Indusand Mekong valleys, such problems have apparently not emerged in the lower reaches (deltaic areas). Whether this represents lack o f testing or whether arsenic problems do not occur there i s as yet uncertain. However, the young Quaternary aquifers most susceptible to developing groundwater arsenic problems appear to be less used in these areas as a result o f poor well yields or high groundwater salinity. Other Quaternary sedimentary aquifers in Asia have not been investigated and so their arsenic status is unknown. Some localized groundwater arsenic problems in South and East Asia relate to ore mineralization and mining activity (for example inpeninsular Thailand andMadhyaF'radesh, India). 4.3 Mechanisms of Arsenic Mobilization:How Does It Get into the Groundwater? One of the key hydrogeochemical advances o f the last few years has been in the better understanding o f the diverse mechanisms o f arsenic mobilization in groundwater, as well as its derivation from different mineral sources. The most important mineral sources in aquifers are metal oxides (especially iron oxides) and sulfide minerals (especially pyrite). Release o f arsenic from sediments to groundwater can be initiated as a result o f the development o f highlyreducing (anaerobic) conditions, leading to the desorption of arsenic from iron oxides with the breakdown o f the oxides themselves. Such reducing conditions are usually found in recently-deposited fine-grained deltaic and alluvial (andsome lacustrine) sediments. Release o f arsenic can also occur in acidic groundwaters under oxidizing (aerobic) conditions. This tends to occur inarid and semiarid settings resultingfrom extensive mineral reaction and evaporation. High-arsenic groundwaters with this type o f association have not been reported inQuaternary aquifers inSouth andEast Asiabut are found insome aridinlandbasins inthe Americas (western United States, Mexico, Argentina). Analogous conditions could occur in some aridparts of the region, such as northern China or western Pakistan, but there is as yet no evidence for this. Despite the improved understanding o f the occurrences and distribution o f arsenic in groundwater, there remains some uncertainty as to the precise nature o f the source, mobilization, and transport o f the element in aquifers. It is only in the last few years that detailed hydrogeochemical investigations have been carried out in some o f the affected regions. Earlier responses to water-related arsenic problems typically involved engineering solutions or finding alternative water sources, with little emphasis on research. It is worthy o f note that, despite the major epidemiological investigations that have been carried out in Taiwan since the discovery o f arsenic-related problems there in the 1960s, there has been little hydrogeochemical research carried out in the region. Even today, the aquifers o f Taiwan are poorly documented and the arsenic occurrence little understood. One o f the important findings o f recent detailed aquifer surveys has been the large degree o f spatial variability in arsenic concentrations in the affected parts o f aquifers, even over lateral distances o f a few hundredmeters. This means that predictability o f arsenic concentrations on a local scale is poor (and probably will always be so). Hence, blanket testing o f individual wells in affected areas is necessary. This can be a major task in countries like Bangladesh where the contamination i s extensive andthe numbero f wells i s very large. There is also uncertainty regarding the temporal variability o f arsenic concentrations in groundwater as very little groundwater monitoring has been carried out. Some studies have noted unexpectedly large temporal variations over various timescales but the supporting data are often sparse and inaccessible and so these reports cannot be relied upon. More controlled monitoringo f affected groundwaters is required to determine their variability inthe short term , (daily), inthe medium term (seasonally), and inthe longterm (years, decades). - 10- Arsenic Contamination of Groundwater inSouthand East Asian Countries -Volume I-Policy Report 4.4 How I s Groundwater Quantity Related to Groundwater Quality? The emerging arsenic problem has revealed the dangers o f groundwater development without consideration o f water quality in tandemwith water quantity. Improvedunderstanding o f the risk factors involved indevelopment of groundwaters has enabled targeting of those aquifers perceived to be most susceptible to developing arsenic problems inrecent years. However, the toxicity o f arsenic i s such that it should also be afforded greater attention in other aquifers used for drinlungwater supply. There is an argumentfor routine testing for arsenic in all new wells provided in major groundwater development projects, regardless o f aquifer type. Randomized reconnaissance-scale sampling for arsenic i s also recommended for existing public supply wells inall aquifer types where no arsenic data currently exist inorder to obtain basic statistics on the distribution o f arsenic concentrations. Groundwater development in previously unexploited but potentially susceptible sedimentary aquifers needs to be preceded by detailed hydrogeological andhydrochemical investigations to ensurethat groundwater will be o f sufficiently high and sustainable quality. The scale of investigations should be commensurate with the scale o fproposeddevelopment. Figure 3 illustrates a tool for an initial risk assessment o f the susceptibility o f an aquifer to arsenic contamination. The shaded boxes indicate the most susceptible pathway. The figure also indicates that significant knowledge about the geography o f arsenic has been created in past years, which permits a strategic response to arsenic contamination. - 11- Arsenic Contamination of Groundwater in South and East Asian Countries -Volume I Policy Report - Figure 3. Classification of Groundwater Environments Susceptible to Arsenic Contaminationa province E I N Geothermallyinfluenced Low-temperaturegroundwater V groundwater I R I 0 Nonminingareas N mineralizedareas M E N T (deltas, closedbasin) I Mixing/ Reducing: Oxidizing: Mineral 1 dilution Reductive Desorption(Fe dissolutione.g. desorptionand oxides) pyrite oxidation dissolution Evaporation Oxidizingor (Fe oxides) mildly reducing Confinedaquifers I I I Increased Low Eh(<50 mv) HighpH(>8) HighFe, SO4 temperature No dissolvedoxygen Highalkalinity(>500 mgL-') Possiblylow pH I Increasedsalinity HighFe, Mn,NH4 PossiblyhighF, U,B, Se, Presenceof N (Na, C1) LOW so4(500 mg~'l) Increasedsalinity metals (Cu, Ni, I HighpH>7 PossiblyhighDcc(>lomgL-') HighEh,DOC C A T E.g. Bangladesh; China 0 ( h e r Mongolia); Taiwan; Possibly: Somepartsof R I WestBengal inIndia; northemChina S Possibly: Cambodia; some parts ofnorthern China; LaoPDR, Vietnam a. For further details see Paper I,Volume 11. b. Not all indicators of low flushing ratesnecessarily apply to all environments. Source: Smedleyand Kinniburgh 2002. -12- 1 Arsenic Contaminationof Groundwater in SouthandEast Asian Countries-Volume I-Policy Report regional scale, with young sediments in alluvial and deltaic plains and inland basins, and areas ofmining activity andmineralization, as obvious target areas for firther evaluation. The guidelines for improving understanding o f the arsenic problem and how to go about dealing with it are broadly the same in any region at increased risk from arsenic contamination. Firstly, the scale o f the problem needs to be assessed. Secondly, where problems exist, it is necessaryto find out whether or not the situation is becoming worse with time. Thirdly, where problems exist, it is necessary to identify the potential strategies or alternatives that are most appropriate for supplyingsafe (low-arsenic) water. Central to these issues i s arsenic testing. In any testing program, it is important to distinguish between reconnaissance testing, which is necessary for establishing the scale o f a groundwater arsenic problem, and blanket testing, which is required for compliance and healthprotection. Blankettesting involves the analysis o f a sample of water from every well used for dnnlung water. For reconnaissance testing, the numbers o f samples need not be large; they should however be collected on a systematic basis. Some monitoring (repeat sampling of a given water source in order to assess temporal changes over a given timescale as distinct from repeat testingto cross-check analytical results) may also be required. Regardless o f the scale o f arsenic contamination in water, there are two ways to measure it. The first method i s to use a field test kit, and the second is to conduct laboratory chemical analysis. The field test measures are more qualitative than quantitative, thus the choice o f one method versus the other depends on several parameters, including the precision o f measurement required. The quality o f analytical results i s also paramount; analysis o f arsenic inwater is by no means a trivial task, yet reliable analytical data are key to understanding the nature and scale o f groundwater arsenic problems as well as dealing with them. Instigation o f any new arsenic testing or monitoring program requires consideration o f the analytical capability o f the local laboratories. In some cases, development o f laboratory capability (for example quality assurance procedures, training, equipment upgrades, increased throughput) may be required and should bebuiltinto the testingprogram. Appropriate mitigation responses for arsenic-affected regions will necessarily vary according to local hydrogeological conditions, climate, population affected, and infrastructural factors. Surface water may or may not be available as an alternative. Other groundwater aquifers at different depths or in different locations may be available for use and need additional assessment. Decisions about what action to take in respect o f the arsenic-affected aquifer depend on factors such as percentage o f wells o f unacceptable quality and range in concentrations (degree by which standards, for example 50 p g L-' or 10 pg L-', are exceeded). 4.5.2 Technology Options The two main technological options are (a) to switch to alternative, arsenic-free water sources; or (b) to remove arsenic from the groundwater source. Table 6 illustrates that there is a range o f technological options that can be used to mitigate arsenic exposure. They vary in terms o f cost (total and per capita), need for operation and maintenance, and expected sustainability. The cost figures provided in the table have been collected from those countries where these options are implemented (mainly from Bangladesh) in order to provide an approximate idea o f costs, but they will vary between countries. The financial and sociocultural sustainability o f any options chosen will depend on the same factors as other typical water supply interventions, again highlightingthat arsenic mitigation needs to be integrated inthe sector. - 14 - Arsenic Contaminationof Groundwater in SouthandEastAsian Countries -Volume I-Policy Report Table 6. Water SupplyOptions for Arsenic Mitigation Technology Tech life Annualized Operation& Water Unit cost capital maintenance production (us$im3) recovery cosb'year (m3) (US$> (US$) Water Supply Technologies: Rainwaterharvesting 15 30 5 16.4 2.134 Deephandtubewell 20 120 4 820 0.151 4,500 0.028" Pondsand filter 15 117 15 820 0.161 2,000 0.066a Dugiring well 25 102 3 410 0.256 1,456 0.072" Conventionaltreatment 20 2,008 3,000 16,400 0.305 Pipeddistribution 20 5,872 800 16,400 0.375 73,000 0.084" Arsenic treatment (households) basedon: Coagulation-filtration 3 3 25 16.4 1.70 Ironcoatedsand/brick 6 0.9 11 16.4 0.73 Dust 5 3 1 16.4 0.24 Ironfillings 5 1.2 29 16.4 184 Synthetic medid 4 3.2 36 16.4 2.39 activatedalumina Arsenictreatment (community) based on: Coagulation-filtration 10 44 250 246 1.21 Granulatedferric 10-15 500-600 450-500 820-900 1.20 hydroxide/oxide Activated alumina 10-15 30-125 500-520 164-200 3.20 Ionexchange 10 50 35 25 3.40 Reverse osmosis 10 440 780 328 3.72 As-Fe removal(air 20 32,000 7,500 730,000 0.054 oxidation-filtration) a. Developmentof full potential ofthe system. Source:Paper3, Volume 11. - 1 5 - Arsenic Contaminationof Groundwaterin South andEastAsian Countries-Volume I-Policy Report Regarding arsenic removal specifically, a number o f treatment technologies have been successfully deployed in many industrial and also some developing countries. These technologies are also very expensive and therefore lend themselves to economies o f scale, makingthemmore suitable to high-population urbancenters thanto lower population density rural areas. Small-scale arsenic removal technologies, especially handpump-mounted ones, are being developed, field-tested, and validated invarious countries. Bangladesh has been the front runner for such an extensive technology validation, and demonstrates the complex nature o f the process and duration. After about three years o f field testing a few technologies were provisionally validated by the government, with recommendations for further testing for a similar duration before final certification. Paper 3 provides a detailed presentation o f available technologies for arsenic screening and arsenic removal and their approximate costs andmanagement requirements at household and community levels. As in the water supply sector ingeneral, the main challenge is sustainability. While it would conceivably be possible to install community arsenic removal plants insmall urban areas and invillages withpipedcommunitywater supply(keeping inmindthe economy of scale), these units would have to be maintained inorder to be effective inthe long run.This is also true for small community and household-level units. Thus, the equation does not only include financing, but also social and cultural factors. The extreme long-term toxic nature o f arsenic, combined with the fact that it has essentially no physical parameters for detection (colorless, tasteless, and odorless), and is very difficult to analyze inthe field at concentrations twice the WHO guideline value (10 pg L-I),mean that arsenic treatment units require very sensitive monitoring and maintenance arrangements. This nature o f arsenic thus calls for carefully sequenced and highly effective mitigation measures, such as screening o f sources for arsenic (local, countrywide, and regional levels), awareness raising about the nature o f arsenic poisoning, and implementation o f arsenic mitigation measures, from the immediate (switching to safe sources for drinking and cooking water supply) to the ultimate provision o f a long-term viable arsenic-safe water supply. 4.5.3 Social andCulturalConsiderations A number o f social, cultural, economic, and political factors come into play in deciding the sequencing and implementation strategy for effective mitigation. As in any other context these factors vary by country, and even w i t h n countries. Issues include suggestions for sharing o f arsenic-safe wells (opinions vary as to whether households who have their own handpumps are really amenable to long-term sharing o f their water source with neighbors whose source is contaminated). Inthe case o f Bangladesh, it seems that households interpret the shift to shared communal systems (installation o f pond sand filters and maintenance- intensive rainwater harvesting) as a step backwards, compared to the convenience o f the shallow handpumps they have grown accustomed to (and invested in) over the past 30 years. Data collected by the BAMWSP show that increased investment in shallow tubewells has taken place over the past 30 years, including in the five years preceding the BAMWSP screening program, which ended in 2003 (figure 5). This happened in spite o f the widely known hazards o f arsenic contamination and stands incontrast to Cambodia, where rainwater harvesting has beenpart o f rural culture, even inrecent decades. Contingent valuation studies can be useful in identifying people's preferences and in designing an appropriate menu o f arsenic mitigation options for individuals and communities, but this also involves an institutional change inattitude towards listeningto communities. Stigmatization o f arsenicosis victims i s prevalent in a number o f countries. Anecdotally, it is considered a serious social side effect o f arsenic contamination. It affects entire families and has an adverse impact on, for example, marriage prospects for young people and on income- earning opportunities. Interestingly, little research has been carried out on the social aspects o f arsenic mitigation, and only a few scientific papers provide sufficient rigor and depth to prepare any guidance on the matter. - 1 6 - Arsenic Contamination of Groundwater in South and East Asian Countries -Volume I- Policy Report Figure5. PrivatePublic Investments inTubewells inBangladesh inthe Last 70Years 1,400,000 1,200,000 g 800,000 3v1 l,ooo,ooo k 600,000 z 400,000 200,000 0 Source: BAMWSP-NAMIC Database, 2004. 4.5.4 OperationalResponsesUndertakenby Countries So Far The results of this study have shown that most countries in the region have carried out some of the concrete operational steps described above. Table 7 summarizes the operational responses that the countries have undertaken so far. Bangladesh and West Bengal, India, have been the most dynamic, primarily because they were the first ones to detect arsenic in their groundwater. Only Bangladesh and West Bengal have implemented these measures at a larger scale, while other countries have started to become active in more limited areas and regions. In addition, especially in the smaller East Asian countries, NGOs and international organizations seem to have been the main drivers, rather than government entities. An interestingpoint is that virtually no country has taken major steps towards active and strategic monitoring of arsenic in groundwater, and much o f the action has focused on provision of technological options to address the arsenic issue. Paper 2 provides a detailed account of these activities ineach country. - 17- X X X X X X X x x x x x X X X X X x x X "x X X x x x x x x x k x x x X X X x x x a X X X X I e X x x x x I X <.-9 c X W % -0 9 .-E X X X 3 e, Y 2e, 5 M .e E "8 5 4.6 The Economicsof Arsenic: InvestmentChoices-What andWhen? 4.6.1 UsingCost-Benefit Analysis (CBA)to InformArsenic Decisionmaking Inspite of the uncertainty regarding arsenic epidemiology, the lethal nature and now well- established effects o f arsenic exposure in South and East Asia compel governments, intemational financing institutions, donors, and NGOs in the water field to make informed choices and trade-off decisions to address arsenic contamination o f drinking water sources andthe scope andextent o fmitigationmeasures. At the same time, investments inarsenic screening andmitigation needto be assessedfrom a wider development perspective. Given the huge investment needs that countries are facing in areas such as basic health care, education, transport, and agriculture, are arsenic-related investments justified? And how can this question be rationally answered, taking into consideration the host o f uncertainties mentioned earlier? Accordingly, a simple cost-benefit methodology has been developed that explicitly takes into account data limitations and provides decisionmakers with an efficient and readily applicable methodology for rapid assessment o f the socioeconomic desirability o f different mitigation policies under various scenarios. Paper 4 provides a general introduction to the way o f thinking about costs and benefits of mitigating (natural) pollutants, including considerations of trade-offs in decisionmaking with respect to the allocation o f financial resources in a budget-constrainedenvironment. Inparticular, the methodology permitsan analysis of options, enabling a choice to be made between different approaches in dealing with (a) the risk that arsenic might be found in an area where a project i s planned; and (b) the risk mitigation options when a project's goal is arsenic mitigation per se. The suggested approach estimates benefits o f mitigation activities as the sum o f saved output productivity and foregone medical costs achieved through the reduction o f arsenic exposure. The present value o fthese benefits is then compared with the present value o f costs o f various mitigation measures in order to determine when and which mitigation policies pass a C B A (that is, produce a positive change insocial welfare). As an illustration, the model was then applied to the case of Bangladesh and clearly confirmed the value o f arsenic mitigation measures that are being undertaken in the country (box 1). The model results also show that not all mitigation technologies pass the CBA unless they are assumed to be 100% effective. Moreover, rainwater harvesting (combined with dug wells or deep handtubewells during the dry season) is not welfare increasing, even at 100% level o f effectiveness. This result points to the need for careful evaluation o f the mitigation measures to be implemented, and indicates that it is not true that any mitigation technology can be applied. In addition, the results indicate that at the project level a least-cost analysis needs to be carried out. The results o f this simple model were also compared with those o f a paper by Maddison, Luque, and Pearce (2004), which usedmore sophisticated data, drawing on the growing body of Bangladesh-specific data and on the best available epidemiological estimates. The results o f the comparison between the simple model and the Maddison, Luque, and Pearce applications are very similar, clearly indicating the applicability o f the model in counties where available data are even more limitedthan inBangladesh. 19 Arsenic Contaminationof Groundwater in Southand EastAsian Countries-Volume I-Policy Report Box 1.Results of the Economic CBA inthe Case of Bangladesh Ten different arsenic mitigation technologies were analyzed, ranging from dug wells to pond sand filters and piped village water supply systems. The surprising result o f the analysis was that the net present value ranged from US$8.2-l.l billion, to US$22.3-11.1 billion, to US$71.8-87.9 billion as the discount rate ranged from 15%, to lo%, and to 5% respectively. The variation under the same discount rate reflects the varying costs o f different technology options. The net present value arising from these calculations canbe as large as 11%o f current Bangladesh gross domestic product (GDP). A sensitivity analysis was undertaken, assuming more realistic scenarios under which only 70% and 50% o f mitigation activities would be effective. Under scenario 1, the relevant net present value (discounted at 10%) amounted to approximately US$9.5 billion, which constitutes around 4% o f current Bangladesh GDP. Under scenario 2 the relevant net present value (discounted at 10%) amounted to approximately US$5 billion, still constituting around 2% o f Bangladesh GDP. It is important to note that inthe calculation (a) the environmental benefits of mitigation strategies were not taken into account (mainly due to lack o f precise data); and (b) the calculated health expenditures represent lower bounds o fthe relevant magnitudes. Thus, while the latter are the current actual expenditures made, they may not be really sufficient for the treatment o f arsenic-related illnesses in Bangladesh. The calculated net benefits from arsenic mitigation are therefore underestimates o f the true benefits and should be used as a very conservative measure o f the welfare increases to be derived from implementing the various mitigationpolicies. With the exception of the option of rainwater harvesting (supplemented by a dug well for the dry season) when discounted at a 10% rate, all other considered mitigation technologies are welfare increasing (that is, they pass a CBA) under all three levels o f effectiveness at both 5% and 10% discount rates. However, when discounted at a 15% rate many o f the mitigation technologies do not pass a C B A at lower than 100% level o f effectiveness. Moreover, rainwater harvesting (+ dug well) and rainwater harvesting (+deep tubewell) are not welfare increasing even at 100% level of effectiveness. Thus, proposed mitigation measures need to be carefully evaluated and it does make a difference which mitigation technology can be applied in a specific context. Moreover, these results indicate that at the project level one may want to carry out a least-cost analysis. The use o f pond sand filters (30 households/pond sand filter), taking account o f the level o f service, turns out to be superior to other technologies. However, even though the analysis concludes that pond sand filters are the most economically efficient option, two real-life caveats make this option less attractive. First, pond sand filters are often very polluted. To take this into account in a C B A a risk-weighting factor should be included in the methodology to indicate the increase in child morbidity and mortality due to water source contamination. The second caveat is the lack o f space in Bangladesh for accommodating so many ponds. In earlier years space was not an issue, but increasing population density has reduced available land in any given village, or people use the ponds for fish farming, a significant source o f income inrural Bangladesh. This situation makes the shadow price involved inusing the pond very high, as it should include the price o f the land where the pond will be situated. It can even be the case that the corresponding land has to be purchased through an actual money transaction, which makes the relevant price an explicit one. Overall, no significant discrepancies among technologies are documented. The more dramatic effects on the desirability o f different mitigation technologies emerge by the changes in the choice of discount rate o f the future flow o f cost and benefits. This exercise highlights the significance o f the choice o f the discount rate, as well as the importance o f the ability to predict the degree of effectiveness o f a proposed policy, which in turn is related to the need to listen to communities and find out their true demandfor the respective arsenic mitigation options. 20 Arsenic Contaminationo fGroundwaterin South andEast Asian Countries-Volume I-Policy Report 4.6.2 Demand-Side Management Different mitigation technologies may be effective, but a key question is whether people find these desirable and are willing to adopt and sustain them (for example, moving from 50% to 90% o f successful implementation). Therefore demand-side perspectives are an important consideration for designing appropriate arsenic mitigation measures. No matter what the solution is in terms o f technology, if it does not meet the preferences o f households and communities the adoption, usage, or scaling up o f the technology will not occur. Indeed, results o f studies carried out by the Water and Sanitation Program in Bangladesh suggest that communities are not only seeking arsenic-free water sources but are also prepared to pay for alternatives that are as convenient as the traditional tubewell, for instance piped water (Ahmad and others 2003). Demand preferences can be assessedthrough contingent valuation or willingness to pay studies and can provide important guidance to decisionmakers. These studies, for instance, have provided the background for preparation o f the Bangladesh Water Supply Program Project, which started implementation in 2005, with financing from the World Bank. 4.6.3 The Economics of Arsenic Mitigation The results o fthe economic analysis powerfilly illustrate four points. First, from an economic point o f view, arsenic mitigation interventions inBangladesh are very well justified. Second, even in a situation o f limited data, economic analysis can and should be carried out. As mentioned above, the results o f the purposefully simple model developed here are very similar to those based on a far more detailed analysis, providing confidence that in countries with much less available information than Bangladesh relatively simple calculations can provide decisionmaking support. This is a clearly needed contribution to the politicized arsenic debate addressed insection 4.7. Third, the findings make an economic case for up-front investment inscientific data (aquifer investigation and screening programs). The logic behind this necessity is the following. The long-run nature o f project-specific developments means that initial screening costs will be discounted over a long-run horizon; hence, these costs will be relatively small innet present value terms irrespective o f their absolute initial value. On the contrary, the effects o f arsenic contamination could be detrimental to both the economy and health o f the inhabitants o f an area over a much shorter horizon. Moreover, one should keep in mind that the decision to develop a particular area is irreversible inpractical terms. This characteristic o f irreversibility necessitates great caution about the decision to develop or not, hence such decisions should be taken underminimumrisk conditions. The combined result o f these three effects increases the net potential benefit to society that can be achieved through gathering information regarding the extent and existence o f arsenic contamination prior to any other project-related appraisal. Fourth, not only should the demand-side perspective be incorporated, but well-established methodologies exist for such assessments. To increase effectiveness o f arsenic mitigation measures, it i s important to strategically employ them up-front. Thus these methodologies, when applied on a case-by-case basis for the different countries, may provide guidance as to the trade-offs between (a) a variety o f arsenic-related investments (for example, screening versus implementation o f different mitigation options); and (b) arsenic-related investments compared to other investments in water supply and sanitation, which would also save lives from other waterborne diseases. There is a clear need to strengthen institutional capacities in the countries to carry out such assessments. Of course, economic analysis can only contribute one building block to the development o f an operational response to arsenic contamination; ethical, social, and political considerations will also play a role insuch deliberations. 4.7 The PoliticalEconomy ofArsenic: What Are the Prospectsfor Action? Arsenic has become a highly politicized topic in the international development community and within some affected countries due to its carcinogenic characteristics and, more 21 importantly, due to the earlier complete failure to consider it as a possible natural contaminant ingroundwater sources. The slow nature of arsenic poisoning, accompanied by eventual very visible marks o f arsenicosis, including gangrene and skin keratoses, i s very striking for the media, and such cases are highlypublicized incountries o f higharsenic occurrence. It was not possible inthe course of the present study to get deeper into this area and there are very few hard data about this issue -which, however, is at the core o f effectively dealing with arsenic. A case in point is the fact that a number of knowledgeable people in the different countries who were approached by the study team to fill out the study survey (Paper 2, annex 1) were not willing to respond because they felt that the topic was too sensitive. On the one hand, this has made the survey instrument less useful than anticipated, but on the other it underlined the fact that arsenic i s a. very sensitive issue with deep political significance and, therefore, technocratic solutions alone are not likely to be successful. Politicization o f the arsenic issue induced a strong bias in reporting o f data and studies, with activists making claims that are often weakly substantiated and sensational, while skeptics are being intimidated into not reporting their data and findings. Table 8 shows an incentive matrix, developed in an attempt to analyze the incentives that different stakeholders - notably governments, donors, intemational agencies, and NGOs - face indealing with arsenic. Itis obvious that onthe government side, urgency regarding arsenic comes about only when it i s shown to be a really crucial issue, when compared with the many other development issues affecting a country. This may explain why some governments have not been as active as some actors might have expected. Table 8. Conceptualized Incentive Matrix: Stakeholder Incentives for Action on Arsenic Issues Donors/ Incentive factors Govemment intemational NGOs agencies Number of people at risk Number o f arsenicosis patients I I I I 22 Arsenic ContaminationofGroundwater inSouth and EastAsian Countries-Volume I-Policy Report However, a further issue is related to the fact that in most o f the countries - including BangladeshandNepal - groundwater was rightlypromoted by governments and development partners as a safe water source compared to surface water, due to the very high public health risk of waterborne disease caused by pathogens. While this policy helped to reduce the disease burden due to bacteriological contamination, it was, however, detrimental to the health o f a part o f the population due to ignorance regarding arsenic contamination, and official acknowledgment o f ths constitutes not only a loss o f face, but also highlights an issue that i s difficult to resolve as there are no clear alternatives. As mentioned in section 4.5, technologies for arsenic removal exist but - especially in rural contexts - they are often too expensive or too difficult to maintain to be considered as effective alternatives. The other options - using other sources, such as deep groundwater in some countries or a return to surface water or rainwater harvesting - are fraught with other health-related problems. Thus, governments may prefer to avoid dealing with the arsenic issue. Clearly, this presents a difficulty because awareness raising is an important way to give affected people the tools to protect themselves, even though it i s not a complete solution. At the same time, politicians are ina dilemma as they fear promoting another solution that, inthe longrun,mightbe detected to be inappropriate or detrimental. This is highlighted by the ongoing debate about the use o f the deep (old) aquifers in Bangladesh and Nepal. These aquifers - for a number o f hydrogeochemical reasons not yet entirely understood - are not susceptible to arsenic contamination. They clearly constitute an alternative as a safe water supply source in rural areas where the overlying shallow aquifers are contaminated, and where surface waters are suboptimal due to the associated microbial pollution risks. Yet there is a risk (the assessmento f the extent o f this riskvaries significantly depending on the interlocutor) that these deep aquifers might also become locally polluted if the wells tapping them were inadequately constructed or if there was a sudden surge in irrigation abstraction from these aquifers. Not surprisingly, in Bangladesh, politicians have been reluctant to promote this alternative and insteadprefer to promote other noncontroversial options in spite o f their short-term health risks, lack o f effectiveness, and low social acceptability among the arsenic-affected populations. This stalemate situation is now showing signs o f resolution as a more structured approach to the investigation is available and controlled use o f the deep aquifers isbeing developed. On the other hand another stakeholder group, donors and international finance institutions, have quite a strong incentive to deal with arsenic, as they have been under close and serious scrutiny for the quality and effectiveness o f their water supply investments inthe region. This hasbecome all the moreclear since the lawsuit against the BritishGeological Survey by some affected patients from Bangladesh. Especially during the Water Decade (1981-1990) international aid agencies strongly promoted groundwater as a safe source, particularly in ruralareas, and financed andpromoted water supply projects wholly reliant on groundwater. The detection o f naturally occurring arsenic in large parts o f Bangladesh and in West Bengal came as a very unwelcome surprise, and there were no clear-cut strategies for quickly and effectively addressing the situation. It is therefore not surprising that the development partners play a very active role in financing arsenic-related interventions, including research, support to policy formulation, and mitigation. However, most partners are at present focusing on detection o f arsenic, awareness building, mitigation measures, and action research. Arsenic considerations are still not fully integrated into water supply sector decisionmaking. According to the study team's knowledge, only the Australian Agency for International Development (AusAID) has taken the initiative to develop specific guidelines to address the issue o f arsenic inits fundedprojects. Clearly, though, while these international institutions, public sector agencies, and NGOs have an incentive to act, they also need not worry about reelection intheir constituencies, and thus they are less risk averse thanthe elected governments. 23 Arsenic Contaminationof Groundwater in South andEastAsian Countries-Volume I-Policy Report A further group of stakeholders consists of a variety of research institutions. As a sensitive and relatively new topic, arsenic is attractingresearchers from all over the world. Inline with the above, however, most research has been financed from outside the region, though Asian researchers have been involved on the teams. The only major study financed withn a country, conceptualized and carried out bynational researchers, i s probably the one inChina. It is also notable, as mentioned earlier, that most researchhas focused onhydrogeology rather thanonepidemiology and social aspects, although avariety of international conferences have pointed out those glaring gaps (Ahmed 2003). This again highlights the lack o f government leadership and direction indealing with the issue. It may also reflect the desire o f donors and international finance institutions to cover a serious lapse through ostensive action, rather than takingamore comprehensive operational view ofthe issue. Finally, it must be said that the arsenic crisis has opened a new market, especially for NGOs, but also for investors in the water sector. The crisis mode and labels such as "the greatest mass poisoning inhistory", which has often beenrepeated inthe literature, permits a growing number o f actors to lobby for certain types o f investments, notably those involving a return to various types o f surface water resources with treatment (thus taking investments out o f households' private hands) in both water and irrigation supply, or promotion o f arsenic removal technologies o fvarious kinds. In summary, the political economy is such that many actors continue pursuing their own interests, not necessarily in a cost-effective manner conducive to solving the issue or to the benefit o f those affectedby arsenic. This latter stakeholder group suffers from the well-known problem faced by large groups with many free riders, in that a large amount o f mainly rural people are potentially affected, but due to a lack o f knowledge, social standing, and resources they are not developing the political clout to demand or implement effective solutions. Poverty certainly plays a role, given that wealthier households - even ruralones - do have the means to look for alternative sources. Governments that want to addressthe arsenic issue will therefore need to overcome their own hesitancy and take a stronger lead role in their countries and on the international plane in order to address the issue. This includes more strategic research and knowledge acquisition regarding arsenic in their countries, appropriate choice and scope o f arsenic mitigation activities, and internal capacity building o f the relevant water supply and water resource agencies. Such action can be supported by the knowledge accumulated in the past decade regarding arsenic, arsenic mitigation, and tools for options analysis that can be used by decisionmakers to analyze local and national options. 5. What ShouldGovernments,DevelopmentPartners,andthe World BankDo? Significant strides have been made since arsenic was first detected in dnnking water tubewells in Eastern India and Bangladesh in the late 1980s and early 1990s. However, more needs to be done and it needs to be done in a more strategic manner, at project, national, and international levels. This section summarizes the remaining action and knowledge gaps and what could be done by different stakeholders inorder to enhance the operational responsesto the arsenic issue inAsian countries. I t i s clear that arsenic consideration needs to be embedded in overall water supply investments and cannot be seen as an isolated issue. In fact, similar considerations apply to other toxic trace elements (such as fluoride, manganese, and boron) that are found in groundwater. The recommendations put forward here can also be applied to those elements. The differences lie primarilyingeographic occurrence, scale, andpoliticization o f the topic. Arsenic contamination i s a long-term issue and, with extended screening, more affected areas are likely to be found in the future, if not at the same scale as those so far located. Interventions and action by governments and their development partners should therefore take 24 Arsenic Contaminationof Groundwater in South andEast Asian Countries-Volume I-Policy Report place at three different levels simultaneously: (a) project and local level; (b) national level; and (c) global level. 5.1 Project-Level Action The findings o f this study make it clear that the occurrence o f arsenic ingroundwater sources must henceforth be taken as a strong possibility inthe countries of the region. Therefore, in any project interventions that consider using groundwater as a source, decisionmakers needto make a judgment if occurrence o f arsenic would affect the outcome o f the project and make provisions accordingly. In general, this would be the case for all water supply projects, but would also include education and health projects that use groundwater as sources for schools and hospitals, and irrigation projects (where wells are often also used for domestic water supply). As pointed out earlier, there are currently no guidelines for arsenic in irrigation water. International study results are not conclusive as to the impacts o f irrigating crops with arsenic-contaminated water. For this reason, arsenic in irrigation wells should be tested for and documentedinorder to have information available for possible future use (see section 5.3 on global-level action). Sequencing and integration are important. Following the simple sequence of steps outlined in figure 4 would ensure that investments adequately intemalize arsenic as another factor that has to be taken into account in interventions in water supply and irrigation. Obviously, possibilities for t h s will be conditioned by a number o f factors, including the political economy involved. This has to be considered inmaking investment commitments. 5.2 National-Level Action Some countries have taken arsenic to the national level o f attention, including Bangladesh, Nepal, and Cambodia. Others, such as India, Pakistan, and China, have only started to address the issue, and in others, international organizations such as UNICEF and local NGOs and universities are the focal points for arsenic-related activities (see table 7 and Paper 2). Since each country has a unique situation interms o f knowledge, scale, and scope o f the problem, generalized and sweeping recommendations onwhat to do will not be useful. Study results suggest, however, that the countries would benefit from (a) encouraging further research in potentially arsenic-affected areas in order to better determine the extent o f the problem; (b) ensuring that arsenic is included as a potential risk factor in decisionmaking about water-related issues (see section 5.1 on project-level action); and (c) developing viable options and coping mechanisms for populations in known arsenic-affected areas. With these three steps, governments -whether at national or at provincial and state level-would be able to address currently affected populations and prevent future investments having negative impacts on their citizens. The arsenic issue has shown that there has been qn underinvestment in groundwater monitoring. While arsenic is now identifiedand being tested for, there are other elements that also need attention. Governments and development partners should actively work to link water supply and water resources management investments in order to address the issue o f groundwater quality monitoring up-front and buildthe requisite capacity. At the national level, governments thus need to take more assertive action in defining their countries' needs and developing strategic actions to deal with arsenic ingroundwater. 5.3 Global-Level Action Focused research on the chemistry o f arsenic mobilization and the dose-response relationships for arsenic are o f vital importance informulating a more sensible approach. Ifthe Millennium Development Goals (MDGs) in water supply and sanitation are to be achieved, then the glaring knowledge gaps regarding arsenic need to be filled, notably by (a) further epidemiological research in directly arsenic-affected countries; (b) socioeconomic research on 25 Arsenic Contaminationof Groundwater inSouth andEastAsian Countries-Volume I-Policy Report the effects o f arsenicosis, understanding behavior and designing demand-based packages for the various arsenic mitigation techniques; and (c) hydrogeological and hydrochemical research. Inaddition, it is likely that inthe near future there willbe diminishingreturns on investments inscientific arsenic researchto reduce uncertainty. The important challenge willbeto identify those areas where improved research-level data collection i s likely to provide a major return and for other areas the main question will be how to manage in the face o f unavoidable and continuinguncertainty. Accordingly, the international dialogue should shift towards targeted research priorities addressing these issues. This would also include the pursuit o f the research agenda regarding arsenic in the food chain. Both the World Bank and a number o f the other development partners are contributors to the Consultative Group on International Agricultural Research (CGIAR), and this organization would lend itself to building up a coherent and focused research agenda on this topic in order to provide decisionmakers with guidance on arsenic- contaminated groundwater inirrigation. These suggested operational responses and their expected outcomes are summarized in annex 1. 6. Conclusions and Recommendations The present study has shown that naturally occurring arsenic in groundwater is more widespread in South and East Asian countries than is generally recognized and that, with continuous testing, more contaminated groundwater aquifers are bound to be identified, ifnot at the same scale as previously. At least 60 million people are currently estimated to live in arsenic risk-prone areas. This, along with projected population growth in the region, continuing private investments in shallow tubewells, and consideration o f the MDGs related to safe water supply, will have considerable impact on government and development community engagement inwater supply andpossibly also inirrigation. Although our ability to predict arsenic concentrations in groundwater from a given area or aquifer is still rather limited, knowledge o f its occurrence and distribution has improved greatly over the last few years. Enough is therefore probably known about where high concentrations tend to occur to make reasonable estimates o f likely at-risk aquifers on a regional scale. Young sediments in alluvial and deltaic plains and inlandbasins and areas o f miningactivity andmineralization are obvious target areas for further evaluation. There are still considerable knowledge gaps regarding arsenic, notably on the epidemiological side. While microbial contamination is undoubtedly o f a larger scale and has more immediate impacts, notably on children, the scope o f the public health threat that arsenic poses in the medium and long terms i s not yet clear. This is especially true when compared with other development challenges faced by the countries inthe region. The issue itself, and the political economy that has developed around it, i s such that clear and easy answers are not likely to be available inthe near future. Nevertheless public health effects o f arsenic are a reality and they need to be taken seriously. As the effects o f arsenic are long term it is likely that arsenic-related disease, with and without fatal outcomes, is going to increase over the coming decades, affecting hundreds o f thousands o f people. At the same time, with other waterborne diseases, notably diarrhea, still claiming so many lives annually, it is important to integrate arsenic considerations into a rational approach to reflect the overall context o f waterbome public health threats. Further investment in safe water supply i s a necessity and arsenic is but one o f the considerations in this regard. It is thus recommended that governments and development partners make use of the information and experience generated over the past two decades and actively include arsenic into assessments when investing in projects that use groundwater as a source (such as water 26 Arsenic Contaminationof GroundwaterinSouth andEastAsianCountries-Volume I-Policy Report supply, irrigation, and education infrastructure) and support institutional strengthening at various levels inorder to deal with arsenic and other groundwater pollutants. At national as well as provincial and state levels, governments would benefit from (a) supporting and originating further research on arsenic occurrence in their territories; (b) making sure that arsenic is taken into account when water-related investments are made and that trade-offs are adequately analyzed; and (c) making their voices heard in developing a cross-regional and international research agenda that would strategically address the remaining knowledge gaps. To make these actions effective, the institutional arrangements within countries, provinces, and states will need to be reviewed and, ifnecessary, improved andstrengthened. 27 Arsenic Contamination o f Groundwater in South and East Asian Countries -Volume I- Policy Report Annex 1. PolicyMatrix: OperationalResponsesto Arsenic Contamination inSouthandEastAsian Countries Level ictivity Expectedoutcome tesponsibility Project hseminate information on rechnical staff are well informed jovemments; level irsenic to national and ind can incorporate arsenic issues levelopment nternational project staff ippropriatelyin investment rganizations, including xojects and studies Ievelopment banks, Ionors, andNGOs :ncorporate arsenic considerations 4rsenic is effectively incorporated jovemments; nall projects usinggroundwater into upstream decisionmalung and ievelopment 3s a (potential) drinking water design regarding investment xganizations, including ;ource (including water supply, projects and studies ensuring levelopment banks, :ducation, irrigation projects, benefits from interventions are ionors, and NGOs iealth projects) in South and East achieved 4sia [farsenic is afactor inthe Arsenic issues are effectively rechnical staff in xoposed project, develop incorporated into investment :ovemments, appropriate activities to be projects and studies, ensuring ievelopment incorporated into the project, such benefits from interventions are aganizations, and NGOs as reconnaissance testing, blanket achieved testing, monitoring, arsenic mitigation investments, social and economic assessments, willingness-to-pay studies (see Volume 11) National Encourage further research in Knowledge base increases and Governments level potentially arsenic-affected areas provides inputto decisionmaking inorder tobetter determine the at different levels extent o f the problem and ensure that data are publicly available as soon as possible Ensure that arsenic is included as Investments inwater supply will Government authorities a potential risk factor in take into account arsenic as a risk decisionmaking about water- factor so that any occurrence can related issues, for example by be mitigated up-front issuing guidelines Develop and implement options Affected people will receive (and Government authorities for and with populations in known participate in) effective mitigation with, ifrequested, arsenic-affected areas (including measures and reduce their support from awareness raising and training exposure, leading to positive health development partners; programs, alternative water benefits NGOs supplyoptions) Develop and implement Increased knowledge o f Government authorities integrated groundwater groundwater resources and aquifer: with, ifrequested, management programs, including to permit more effective support from aquifer mapping, testing, decisionmaking regardingwater development partners; monitoring, and publicly supply investments and necessary NGOs accessible databases inorder to arsenic mitigation measures support the water-using sectors Capacity to effectively address and institutionbuilding arsenic and other pollutants is strengthenedwithin the country at local, regional, and national levels 28 Arsenic Contaminationo f Groundwater in South and East Asian Countries -Volume I- Policy Report Level Activity Expected outcome Responsibility Integrate arsenic as one factor in Arsenic issues become integrated 3ovemments national policies regarding water into sector policies and will be supply-related activities, more effectivelyaddressed by including research and drawing on existing institutions investments and knowledge Global Develop and implement a more Knowledge gaps are diminished Governments, level strategic global research agenda and arsenic-inherent uncertainties development partners, to the benefit o f arsenic-affected are more strategically addressed , NGOs countries, including: Feedback loop into projects and Targeted epidemiological national-level activities improves research project and policy outcomes Social research on the effects o f arsenicosis, understanding behavior and designing demand-based packages for the various arsenic mitigation techniques Geohydrological and hydrochemical research in countries and inthe region Research on arsenic inthe food chain, for example through CGIAR network Ensure that data and analyses Research results are disseminated Governments; research carried out by external research effectively and ina timely manner institutions and organizations and universities are and can be putto use as soon as universities made available to the respective possible countries as soon as possible 29 Arsenic Contamination of GroundwaterinSouth and East Asian Countries-Volume I- Policy Report References Ahmad, J., B N. Goldar, S. Misra, and M. Jakariya. 2003. Fighting Arsenic: Listening to Rural Communities - Willingness to Payfor Arsenic-Free, Safe Drinking Water in Bangladesh. WSP-South Asia. Ahmed, M. F. 2003. Arsenic Contamination: Bangladesh Perspective. ISBN984-32-0350-X. Dhaka, Bangladesh: ITNBangladesh. Bureau of Statistics-UNICEF (United Nations Children's Fund). 2002. Child Nutrition Survey of Bangladesh 2000. Bangladesh Bureau of Statistics, Statistics Division, Ministry o f Planning, Government of Bangladesh, andUNICEF. Maddison, D., R. C. Luque, and D. Pearce. 2004. The Economic Cost of Arsenic Contamination of Groundwater in Bangladesh. Water and SanitationProgram. Smedley, P. L. and D. G. Kinniburgh. 2002. "A Review of the Source, Behaviour and Distribution of Arsenic inNatural Waters." Applied Geochemistry 17:517-568. Smedley, P. L. 2003. "Arsenic in Groundwater - South and East Asia." In: A. H. Welch and K. G. Stollenwerk, eds., Arsenic in Ground Water: Geochemistry and Occurrence 179-209. Boston, Massachusetts:Kluwer Academic Publishers. Van Geen, A. and others. 2003, "Spatial Variability of Arsenic in6000 Tube Wells ina 25 km2Area of Bangladesh." Water Resour. Res. 39(5): 1140. DOI: 10.1029/2002WR001617. WHO-UNICEF (World Health Organization and United Nations Children's Fund). 2000. Global Water Supply and Sanitation Assessment 2000 Report. 30