WATER GLOBAL PRACTICE Building Urban Water Resilience in Small Island Countries THE CASE OF SOUTH TARAWA, KIRIBATI OCTOBER 2019 About the Water Global Practice Launched in 2014, the World Bank Group’s Water Global Practice brings together financing, knowledge, and implementation in one platform. By combining the Bank’s global knowledge with country investments, this model generates more firepower for transformational solutions to help countries grow sustainably. Please visit us at www.worldbank.org/water or follow us on Twitter at @WorldBankWater. About GWSP This publication received the support of the Global Water Security & Sanitation Partnership (GWSP). GWSP is a multidonor trust fund administered by the World Bank’s Water Global Practice and supported by Australia’s Department of Foreign Affairs and Trade; the Bill & Melinda Gates Foundation; The Netherlands’ Ministry of Foreign Trade and Development Cooperation; the Rockefeller Foundation; the Swedish International Development Cooperation Agency; Switzerland’s State Secretariat for Economic Affairs; the Swiss Agency for Development and Cooperation; and the U.K. Department for International Development. Please visit us at www.worldbank.org/gwsp or follow us on Twitter #gwsp. Building Urban Water Resilience in Small Island Countries The Case of South Tarawa, Kiribati OCTOBER 2019 © 2019 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org This work is a product of the staff of The World Bank, with external contributions. The findings, interpretations, and con- clusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of 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 of The World Bank concerning the legal status of any territory, or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Please cite the work as follows: World Bank. 2019. “Building Urban Water Resilience in Small Island Countries: The Case of South Tarawa, Kiribati.” World Bank, Washington, DC Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; Fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover photo: Stephane Dahan/World Bank. Cover design: Duina Reyes. Contents Foreword vii Acknowledgments ix Executive Summary xi Abbreviations xvii Chapter 1  Introduction 1 Chapter 2  What is Water Supply Resilience in a Context of Water Scarcity? 3 Chapter 3  Why Does Water Supply Resilience Matter for South Tarawa? 5 3.1. Geographic, Physical, and Socioeconomic Fragility 5 3.2. Chronic Water Supply Insecurity 6 3.2.1. Historical Reliance on Three Sources of Water 6 3.2.2. Strategies for Coping with Inadequate Water Supply 11 3.3. Increasing Water Scarcity and the Need for New Resources 13 3.4. Water Sector Vulnerabilities 15 3.4.1. Droughts 16 3.4.2. Seawater Overtopping 18 3.4.3. Anthropic Risks 19 3.4.4. Technological and Operational Risks 21 3.5. How Resilient Will the Future Water Supply Be? 23 Notes 26 Chapter 4  How Can South Tarawa Improve Its Water Supply Resilience? 27 Potential Measures for Reducing Water Stress during Supply Crises 4.1.  27 4.1.1. Encouraging Water Conservation 27 4.1.2. Optimizing the Use of Rainwater Harvesting 28 4.1.3. Improving Urban Groundwater Through Managed Recharge 31 4.1.4. Direct Nonpotable Water Reuse 35 4.1.5. Increasing PUB Emergency Storage Capacity 37 4.1.6. Increasing Desalination Capacity 38 4.1.7. Managing Water Reserves in a Sustainable Manner 41 4.1.8. Other Supply Augmentation Measures 42 4.1.9. Assessing the Proposed Resilience Measures 43 Notes 46 Regulation of Water Supply and Sanitation in Bank Client Countries iii Chapter 5  Recommendations and Lessons Learned 47 5.1. Priority Measures and Complementary Analyses 47 5.2. Lessons for Water-Scarce Cities in Small Islands and Elsewhere 47 Note 50 Bibliography 51 Appendix A Multiple Water Sources and Their Uses in South Tarawa—Sociocultural Analysis 55 Appendix B  Rainwater Harvesting 73 Appendix C  Urban Lenses: Water Balance 87 Appendix D Conservation of Freshwater Lenses on Bonriki and Buota, through Watershed Management 89 Boxes 3.1. Ancestral Thrifty Water Use Traditions in Kiribati Households 12 3.2. Water Sharing–A Social Duty Deeply Anchored in Kiribati Culture 13 3.3. Water Crisis in Malé 22 4.1. When Cultural Norms and Requirements Reinforce Water Hardships 31 4.2. Improvised Graywater Reuse in Kiribati Households 37 Figures ES.1. Contribution of Water Resources to Water Supply, 2019 xi ES.2. Per Capita Freshwater Availability in South Tarawa, 1947–2017 xii ES.3. Contribution of Water Resources to Water Supply (2030, after STWSP) xii ES.4. Freshwater Availability 2030, in Normal and Various Crisis Scenarios xiii ES.5. Overview of Resilience Enhancement Measures xvi 3.1. Prevalence of Rainwater Harvesting across Buildings in Betio, Bairiki, and Bikenibeu 8 3.2. Per Capita Freshwater Availability in South Tarawa, 1947–2018 14 3.3. Five-Year Cumulative Rainfall in South Tarawa, Actual/70-Year Average Ratio 16 3.4. Two- Year Cumulative Rainfall in South Tarawa, Actual/70-Year Average Ratio 16 3.5. Return Period of Rainwater Harvesting System Dry-Ups Exceeding 1 Month 18 3.6. Residential Water Consumption from Water Utilities Compared to Countries’ GDPs 24 3.7. Impact of Water Crisis Scenarios on Production Capacity, by Water Resource (in percent) and Water Availability (in lpcd), 2030 25 4.1. Key Contributing Factors to the Water Balance of Urban Lenses 32 4.2. Balance of Freshwater Inflow and Outflow by Urban Lens Surface 32 4.3. Potential Relative Contribution of Greywater 33 4.4. Schematic Lens Recharge System (Greywater Disposal+Rainwater) 34 4.5. Options for Greywater Reuse Management in Low-Density Areas 36 iv Building Urban Water Resilience in Small Island Countries 4.6. Cost-Effectiveness of an Emergency Storage Solution for Various Storage Sizes, Durations, and Intensities (lpcd gap) of Emergencies 39 4.7. Desalination Backup Capacity Required to Cover Crisis Supply Gap 40 4.8. Preferred Areas for a Backup Desalination System 40 4.9. Improvement in Water Supply in Crisis Scenarios due to Proposed Measures 44 4.10. Overview of Resilience Enhancement Measures 46 A.1. Answers to the Questions “Do You Ever Use PUB Water? Water from a Well? Rainwater?” 56 A.2. Domestic Water Sources Allocated for Various Uses 57 B.1. Individual and Collective Rainwater Harvesting Systems (dark dots): Identification through Aerial Imagery (Betio) 82 B.2. Individual and Collective Rainwater Harvesting Systems (dark dots): Identification through Aerial Imagery (Bairiki) 82 B.3. Individual and Collective Rainwater Harvesting Systems (dark dots): Identification through Aerial Imagery (Bikenibeu) 82 D.1. Percentage of Unauthorized Households, by Years of Occupation, 2014 90 Maps 3.1. Location of Tarawa and Kiribati in the Pacific Ocean 5 3.2. South Tarawa: Population Density and Main Physical Features 6 3.3. Bonriki Water Reserve Infiltration Galleries and Seawater Overtopping 20 Photos 3.1. PUB Water Connection 7 3.2. Shallow Well Next to a Private Toilet and Pig Enclosure 10 4.1. Private Rainwater Tank and Laundry Area 29 4.2. Favorable Setup for Direct Reuse 38 Tables 3.1. Water Use by Source (lpcd): Current Situation (2019) 9 3.2. Water Use by Source (lpcd): Future Situation (2030) 15 4.1. Rainwater Harvesting: Current and Potential Increase Capacity 28 4.2. Temporal Characteristics of Water Crisis Triggers 43 4.3. Overview of Resilience Enhancement Measures 45 5.1. Indicative Action Plan, 2020–30 48 B.1. Summary of Rainwater Harvesting Programs implemented in Kiribati 74 B.2. Storage Capacity and Costs for Rainwater Harvesting Development Programs 78 B.3. Collective Rainwater Harvesting Systems Survey—Selected Indicators 83 C.1. Current and Future Water Balance across South Tarawa Populated Areas 88 D.1. Summary of the LMD Census of Illegal Households in Bonriki 90 Building Urban Water Resilience in Small Island Countries v D.2. Potential Prohibited Uses 99 D.3. Potential Permitted Uses on the Water Reserves 100 D.4. Water Reserves Governance Options 101 D.5. Toolkits for Community Engagement: Content 105 D.6. Matrix of Options with Cost Estimates 105 vi Building Urban Water Resilience in Small Island Countries Foreword Small island developing states are on the front lines Pacific and beyond. The study also draws special atten- of the climate crisis. They face enormous climate tion to the role of social resilience during water-supply adaptation challenges stemming from rising sea lev- crises, a dimension often overlooked by water sector els coupled with more frequent and intense storms policymakers and yet critical in most contexts. and droughts. Many island states are already water Water challenges in a city such as South Tarawa, constrained due to small watersheds and limited Kiribati’s capital, matter not only because the lives opportunities for groundwater and surface water and prosperity of its own inhabitants are at stake, but storage. Their water quality is all too frequently also because issues such as climate variability, degraded because of generally high population den- anthropogenic pressure on water resources, and sys- sities, low levels of sanitation, and salt water intru- temic infrastructure weaknesses foretell the human, sion. Advanced water technologies, such as physical, and economic challenges many other cities desalinization or wastewater reuse, can potentially are likely to face later this century. In this way, places help address these challenges but are financially like Kiribati are the canary in the coal mine when it and institutionally difficult to sustain in many coun- comes to the perils of climate change. Because South tries  and in the event of breakdowns may even—­ Tarawa’s climate risks are more imminent and more perversely—increase water security risks. threatening, and because its capacity to address Kiribati faces all of these challenges and more, and these risks is more constrained than elsewhere, water this report aims to shed light on improved water man- management models that prove resilient and sustain- agement approaches to boost resilience and increase able in such a context today could help unlock solu- water security. We believe the proposed resilience tions to future urban water security challenges ­ measures—many of which are low cost and nature globally. We believe this report makes a valuable con- based—are applicable to other small islands across the tribution to sharing these lessons. Jennifer Sara Benoît Bosquet Global Director Regional Director, Sustainable Development Water Global Practice East Asia and Pacific Region Regulation of Water Supply and Sanitation in Bank Client Countries vii Acknowledgments This report was prepared by a team of water special- Collaboration with many partners and stakeholders ists, led by Stephane Dahan, and including Manuel in the water sector has been essential throughout the Marino and Helene Le Deunff. preparation of this study. The team particularly The peer reviewers were Greg Browder, Global wishes to thank the Ministry of Infrastructure and Lead, Water and Resilience; Lizmara Kirchner, Senior Sustainable Energy of Kiribati, and the Public Utilities Water and Sanitation Specialist; and Dominick de Board, for their precious contribution to data collec- Waal, Senior Economist, all from the World Bank’s tion and analyses, as well as the Asian Development Water Global Practice. Bank and the New Zealand Ministry of Foreign Affairs Contributions from Clementine Stip, Edkarl Galing, and Trade, for their review of the document. and Van Anh Vu Hong from the World Bank, as well as Finally, the team is grateful to the Global Water the Secretariat of the Pacific Community (SPC), Ian Security & Sanitation Partnership (GWSP) for its sup- Hay Consulting, Tony Falkland, and Jon Metcalfe port to this work. have been critically important in the preparation of this work. Regulation of Water Supply and Sanitation in Bank Client Countries ix Executive Summary What Is Water Supply Resilience in a is available in several areas that are served by sewer- Context of Water Scarcity? age networks. Resilience is the capacity of a system to absorb the In this context, the population of South Tarawa has shocks or stresses imposed by climate change and other developed sophisticated strategies for coping with factors. Water supply services around the world are inadequate water supply. Households follow several increasingly being affected by shocks or stresses such advanced water scarcity management principles, as droughts, flooding, or human-induced disasters. such as diversification of water resources, fit-for-pur- Building water supply resilience requires long-term pose water use, thrifty water consumption, and planning that takes into account the uncertainties adaptive strategies of water use, depending on the faced by decision makers. local level of water stress. Collectivism and water sharing also play a critical role in reducing affordabil- Why Does Water Supply Resilience Matter ity challenges, water-related conflicts, and vulnera- for South Tarawa? bility to drought. Water scarcity challenges, and the effects of climate South Tarawa presents striking features of geographic, change are increasing, and they underline the need for physical, and socioeconomic fragility. Kiribati is one of additional water resources. Inherent limitations in the smallest, most remote, and most geographically freshwater resources, climate variability, sea level dispersed countries in the world. Its capital, South rise, and continuous population growth are exposing Tarawa, has uniquely fragile water resources due to South Tarawa to increasingly severe water supply its small size, lack of natural capacity for water stor- age, and competing land use. FIGURE ES.1. Contribution of Water Resources to Water supply insecurity is chronic across South Water Supply, 2019 Tarawa. Despite the presence of a centralized water supply system operated by the public utilities board PUB: Groundwater, 20% (PUB), and fed primarily by groundwater from the water reserves located in Bonriki and Buota, access Rain-water, to safely managed, piped water supply services Sea-water, 10% 14% remains very limited in South Tarawa: Only two thirds of the population has PUB house connections that supply 10 liters per capita per day, on average. As shown in figure ES.1, rainwater harvesting (RWH) has emerged as an important coping strategy, but its quantitative contribution to water use remains mar- ginal. Urban groundwater is generally brackish and contaminated bacteriologically, but is still widely Well water used for nonconsumptive purposes through a net- 56% work of private, shared, and communal shallow Source: World Bank data. wells. A piped supply of seawater for toilet flushing Note: PUB = public utilities board. Regulation of Water Supply and Sanitation in Bank Client Countries xi FIGURE ES.2. Per Capita Freshwater Availability in South Tarawa, 1947–2017 25,600 Per capita freshwater availability (m3/year/person) 12,800 6,400 3,200 No water stress 1,600 Water stress 800 Water scarcity 400 Absolute water scarcity 200 100 57 62 67 72 77 82 87 92 97 02 07 47 52 12 17 20 20 19 19 19 19 19 19 19 19 19 19 19 20 20 Sources: Rainfall data: Kiribati Meteorological Services; Historic population: Hoy et al. 2014. deficits. Per capita freshwater availability in South FIGURE ES.3. Contribution of Water Resources to Water Supply (2030, after STWSP) Tarawa has dramatically declined in recent decades, as shown in figure ES.2, primarily due to a six-fold population increase over the last 50 years. Since the end of the 1990s, freshwater availability has hovered PUB: around the absolute water scarcity threshold. Desalination, 48% In recognition of this deepening challenge, the South PUB: Ground-water, Tarawa Water Supply Project (STWSP) is planning a 22% 4,000 cubic meter (m3) per day seawater desalination system to address water supply deficits in the long term. The centralized water supply system will rely Rain-water, on a combination of desalinated seawater and Sea-water 6% 11% groundwater abstracted from the Bonriki and Buota Well water 13% freshwater lenses, as shown on figure ES.3. In addi- 1 tion to the improved PUB water services, South Source: World Bank data. Note: PUB = public utilities board; STWSP = South Tarawa Water Supply Project. Tarawa’s population is expected to continue to rely to some extent on shallow wells and RWH for a variety Despite the PUB water system upgrade and the addi- of reasons, including affordability, taste preferences, tional production of desalinated water, the water sec- and attachment to cultural practices. tor will continue to present key vulnerabilities, both xii Building Urban Water Resilience in Small Island Countries natural and artificial. Severe droughts, such as those strengthen PUB’s operational capacity with private experienced in the past, could lead to a significant sector support. However, once people become accus- decrease in yields at water reserves, drive well-wa- tomed to an improved reliability of water supply and ter salinity above acceptable levels—even for nonpo- to higher levels of water consumption, traditional table water use—and dry up most of the RWH water scarcity management strategies may fall in dis- systems. Sea level rise will aggravate the likelihood use, undermining resilience to water scarcity crises. and impact of seawater overtopping the water Once the PUB water system is upgraded, three water reserves, threatening salinization of the lenses. In crisis scenarios stand out as having potentially the the long term, the contamination of the water greatest impact on water availability. For this study, reserves whether by settlers, or by airport-related each event (and combination thereof), and the effect activities, could affect water quality to a degree it would have on the production capacity of each that requires very costly treatment solutions, partic- water resource has been estimated, and this is illus- ularly if encroachment by settlers continues trated in figure ES.4. For example, a breakdown of unmitigated. the desalination plant would reduce freshwater avail- The modernization of the PUB water system may also ability by more than 50 percent, or even more if com- introduce new risks for South Tarawa. Water supply bined with a major drought or pollution event. In all will be vulnerable to the systemic risk of a temporary of the crisis scenarios, the availability of potable breakdown of seawater desalination due to technical water would, however, remain above 10 liters per issues. To mitigate such risks, STWSP will help capita per day (lpcd), which can be considered a FIGURE ES.4. Freshwater Availability 2030, in Normal and Various Crisis Scenarios 90 Single-event cases Multiple-event cases 80 Total water available (lpcd) 70 79% 77% 76% 60 70% 66% 50 58% 49% 40 30 26% 29% 29% 20 10 0 19 30 ht ng n n T n n n n n tio ow tio w tio ow w O g 20 pi 20 SW do do ou llu llu llu kd kd op k k al Dr al Po Po Po t+ ea ea ea ea rt rm rm e br Br Br Br gh t+ + ov No No T n t+ + + ou gh er O io T n SW Dr gh at at tio O ou SW lin aw ou Dr llu sa Dr Se Po Da Source: World Bank data. Note: The percentage figures show the ratio of water volumes available during an event vs. during the normal situation. Percentages shown are the ones remaining in case of crisis events vs. the “normal” level. SWOT = seawater overtopping. Building Urban Water Resilience in Small Island Countries xiii minimum acceptable level for human consumption with rainwater overflow from tanks could improve (that is, for drinking and cooking) in an emergency. the availability and quality of well water, making this Efforts to build water supply resilience in South resource a more relevant backup solution for non- Tarawa should therefore seek to improve availability consumptive use when other sources fail. Increased during these types of events, in order to reduce the PUB water supply delivery will boost greywater gen- gap between availability and the water consumption eration and enhance the recharging of lenses, espe- levels South Tarawa will have grown used to. cially during droughts. Pilot programs will be required to test the design of simple treatment and How Can South Tarawa Improve its infiltration systems that can maximize the water Water Resilience? quality of nearby wells. This will also improve South A first measure, critical to the resilience of South Tarawa’s environment by reducing the unmanaged Tarawa’s water sector, is strengthening the utility’s discharge of greywater in backyards. Policy measures overall capacity and sustainability. PUB’s ability to plan could also contribute to the management and protec- for and react to any water crisis will greatly depend tion of urban groundwater quality. These could on its operational performance and its financial sta- include the establishment of quality guidelines for bility. Furthermore, the efficient management of its and monitoring of well water; prohibiting the use of distribution network will be essential in order to min- soaps and detergents that are harmful to the environ- imize leaks and water shortages during a crisis. ment; and an amendment of the Kiribati Building The Kiribati inhabitants’ tradition of thrifty water use Code to prescribe the direction of overflow of RWH is a major asset that the authorities should seek to systems for recharging wells. encourage and sustain. The delivery of safe and plen- There is also significant scope for managing PUB’s tiful PUB water will require careful management of infrastructure in a way that is more resilient to short- changes in people’s water-use habits by building term breakups of the desalination system or the water enhanced awareness of water conservation princi- reserve production system. Protection of the Bonriki ples, and applying appropriate water tariffs. water reserve, stands as a priority measure for help- Opportunities to improve water availability during cri- ing to weather brief breakdowns of the desalination ses, through RWH, are related more to improving water plant, or other parts of the system, but the adoption resource management than to expanding infrastructure. of planned management procedures for bulk PUB The current stock of hard-roofed structures could water storage facilities, and prudent asset manage- allow significant expansion of RWH, but low cost-­ ment strategies that are designed to gradually effectiveness remains a barrier. When yields from the respond to such events, are also important. water reserves decline due to drought or a pollution Direct reuse of nonpotable water, already practiced crisis, advising the public that they should keep informally in South Tarawa, could reduce water demand their rainwater tanks full, if necessary by progressively and thereby minimize the impact of decreased water filling them with PUB water, will provide a critical availability. In low density areas without a seawater safety cushion of a several-week supply if the desali- network for toilet flushing, direct reuse of greywater nation plant happens to fail in that same period. at the household level could help reduce the reliance There is significant scope for better managing, on PUB water and/or minimize the burden of fetching recharging, and using urban groundwater. The recharg- well water. Similarly, direct reuse of nonpotable ing of urban lenses with treated greywater and/or water could be considered in lieu of seawater xiv Building Urban Water Resilience in Small Island Countries flushing in high density areas where sewerage sys- •• If the above measures are implemented to their tem expansion is envisaged. Across South Tarawa, fullest potential, two options could be considered, we estimate that up to 40 percent of households to bridge the remaining water deficit: the imple- could benefit from the implementation of a direct mentation of bulk PUB water storage in Bonriki; nonpotable water reuse scheme. or the construction of a back-up desalination unit. Various measures could help reduce pollution risks in If desalination system breakdowns are expected the water reserves. Diffuse pollution from people liv- to last less than 2 weeks, then bulk water storage ing on the reserves could be mitigated through a vari- would be technically feasible and sufficient to ety of low-cost watershed management approaches. weather the crisis, and would be the more cost-­ Similarly, potential pollution linked to fuel spillage effective of the two options. If desalination system from the airport could be contained, facilitating breakdowns for more than 2 weeks are considered remediation, through continuous groundwater credible, however, then the back-up desalination monitoring. plant would become more cost-effective, and Another option for improving resilience could be the could help address water shortage crises over time. construction of a backup seawater desalination system, •• Finally, the infill of borrow pits3 in Bonriki, along independent from the existing ones. There is no tech- with RWH and groundwater transfer from North nical limit to the production capacity of such a Tarawa appear to be the costliest measures being backup system, and in theory it could address the considered by far. There could still be justification supply gaps in all of the crisis scenarios. However, for RWH, however, if it is determined that house- keeping a seawater desalination plant in standby holds will be willing to pay a large premium for pri- mode to cope with emergencies requires regular and vate ownership of this quality water resource, careful maintenance, at a nonnegligible cost. which is anchored in Kiribati tradition and appreci- A clear prioritization of measures emerges from a ated for its taste. comparison of their cost-effectiveness. 2 What Can Water-Scarce Cities on Small •• As shown in figure ES.5, water conservation mea- Islands and Elsewhere Learn from this sures, tank crisis management, direct greywater Study? reuse, deep-rooted palm tree clearing, and pollution The sophistication of urban water scarcity manage- mitigation can all significantly augment (or prevent ment in South Tarawa, shaped by long exposure to the loss of) water supplies at a very limited cost, and water hardship, illustrates the value that small island can be considered as no-regret measures. countries can bring to global discussions on urban •• The managed recharge of urban water lenses is water resilience. The convergence of droughts, also expected to require very limited investment, anthropic pressure on water resources, and systemic but the benefits in terms of incremental water infrastructure failures represents a clear societal availability at acceptable nonpotable water quality threat, even more immediate than the potential levels are not known at this stage, and will require effects of sea level rise. An important finding of this piloting and monitoring to ascertain. Any mean- study is that, after the planned modernization of ingful increase of well-water availability at an PUB’s water system, water supply security can be acceptable level of quality would make this mea- significantly boosted by local, low-tech, low-cost sure more cost-effective than the following ones. measures that make the most of existing resources. Building Urban Water Resilience in Small Island Countries xv FIGURE ES.5. Overview of Resilience Enhancement Measures Unit cost (US$ million per incremental lpcd across South Tarawa) Lowest 10 priority N. Tarawa measures transfer Small private systems Rain- Borrow pits infill water harvest. Large collective systems 1 Design to cover long (>30 days) events Back-up Complex housing retro tting desalination Water reserves palm clearing PUB storage 0.1 Water Direct reserves mgt grey- Design to cover short water (<5 days) events reuse Tanks mgt. Highest Water 0.01 conservation priority No or minor plumbing measures 0 5 10 15 20 25 > 25 Incremental lpcd delivered across South Tarawa—2030 (maximum potential) Measures: Increasing potable water availability Avoiding a decrease of potable water availability Increasing availability of water close to potable quality Increasing non-potable water availability Source: World Bank data. Note: Greywater and rainwater recharge to urban water lenses is not illustrated, due to uncertainties about the unit costs and benefits. Finally, the preservation of traditional water man- Notes agement practices, which are both frugal and collec- 1. A freshwater lens is a convex-shaped layer of fresh groundwater that tivist, and are shaped by a long history of water floats above the denser saltwater, usually found on small coral or limestone islands and atolls. scarcity, will also be critical in maximizing the resil- 2. Due to insufficient data availability, full cost-benefit analyses were ience of communities to future water-supply shocks. not carried out. Sustaining social resilience should remain a priority 3. Pits resulting from the excavation of material used for the construc- when investing in the water resilience of urban tion of the Bonriki airport airstrip. infrastructures. xvi Building Urban Water Resilience in Small Island Countries Abbreviations ADB Asian Development Bank $A Australian dollar CAPEX capital expenditure CSO community services obligation DBO design build operate DFAT Department of Foreign Affairs and Trade EC electrical conductivity ENSO El Niño Southern Oscillation EU European Union GDP gross domestic product GEF Global Environmental Facility GoK Government of Kiribati HCI Human Capital Index HDI Human Development Index HIES Household Income and Expenditure Survey IPCC Intergovernmental Panel on Climate Change KAP Kiribati Adaptation Program KIRIWATSAN Water and Sanitation in Kiribati outer islands LMD Land Management Department lpcd liters per capita per day M&E monitoring and evaluation MELAD Ministry of Environment, Land and Agriculture Development MFAT Ministry of Foreign Affairs and Trade MHMS Ministry of Health and Medical Services MICTTD Ministry of Information, Communication, Transport and Tourism Development MISE Ministry of Infrastructure and Sustainable Energy NOAA National Oceanic and Atmospheric Administration O&M operation and maintenance OPEX operational expenditure Regulation of Water Supply and Sanitation in Bank Client Countries xvii PUB public utilities board PV photovoltaic RCP Representative Concentration Pathway RWH rainwater harvesting SAPHE Sanitation Public Health and Environment SMP Sustainable Management Plan SODIS solar water disinfection SPC Secretariat of the Pacific Community STWSP South Tarawa Water Supply Project SWOT seawater overtopping TA technical assistance TBD to be determined UNDP United Nations Development Program WASH water supply, sanitation, and hygiene WHO World Health Organization WSS water supply and sanitation xviii Building Urban Water Resilience in Small Island Countries Chapter 1 Introduction The water sector in South Tarawa, the capital city of This report was conducted in parallel with the Kiribati, is entering a time of extensive transition. On preparation of the South Tarawa Water Supply the one hand, climate change and the pressures Project (STWSP), to provide a diagnostic of the placed on fragile water resources by overpopulation vulnerabilities that South Tarawa’s water supply ­ are converging to create unprecedented risks for sector may still face after the completion of the water-supply security. On the other hand, an unpar- project. Its main objective is to inform decisions by ­ alleled effort by the government of Kiribati and the government of Kiribati about how to address donors to support modernization of the city’s reticu- these vulnerabilities. Chapter 2 sets the stage with a lated water supply system will drastically improve brief presentation of  the concept of water supply water supply services, although it may also introduce resilience in a context of water scarcity. Chapter 3 some new systemic water security risks. provides a summary of water supply conditions in In a small island context, risks can materialize more South Tarawa, and analyzes the impacts of various quickly than elsewhere, and can have disproportion- stresses and shocks on freshwater availability. ate consequences. Strengthening water sector resil- Chapter 4 discusses potential measures to reduce ience is therefore critical both to people’s welfare and water supply deficits during major crises. Finally, to the economy. Traditional perceptions of risk, and chapter 5 offers an overview of these measures in the resilience mechanisms adopted by households the form of an action plan, and proposes several les- and communities based on centuries of experience sons that can be applied to water-scarce cities on with water hardship, may no longer apply in these other small islands, and elsewhere. changing times. Building Urban Water Resilience in Small Island Countries 1 Chapter 2 What is Water Supply Resilience in a Context of Water Scarcity? 1.  Water supply and sanitation (WSS) services around give proper consideration to both climate risks and the world are increasingly being affected by external opportunities when planning for WSS services and shocks, including climate-related shocks. Rapid urban- water resources. ization is compounding the effects of a lack of urban 3.  Resilience is the capacity of a system to absorb the planning with substantial impacts on water availabil- shocks or stresses imposed by climate change and other ity and quality both within and beyond city boundar- factors, and in the process to develop greater robust- ies. These include the overexploitation of water ness. Projects planned with resilience as a goal are resources; decreased water security; increased vul- designed, built, and operated to better handle not only nerability to floods and other natural disasters; and the range of potential climate change and climate-in- various water-related health issues. Climate change duced natural disasters, but also to promote contin- is exacerbating these impacts and worsening water gencies that will lead to efficient, rapid adaptation to a scarcity. The consequences of climate change will less vulnerable future state (Bonzanigo et al. 2018). continue to be felt through more frequent, and/or more severe, extreme weather events, changing rain- 4. In light of increasing water scarcity, building fall patterns and temperatures, seasonal shifts, and water-supply resilience will require long-term planning sea level rise. that takes into account the uncertainties faced by deci- sion makers. To establish what resilience really means 2.  In the case of the Pacific islands, population growth for a water system, stakeholders must first agree on in urban centers and increasing vulnerability to storm the performance objectives that the system must surges, seawater overtopping (SWOT), and variability in meet when confronted with strains and shocks. The rainfall patterns could pose serious threats to habit- resilience of each element of the system can then be ability, and to the availability of water resources. The assessed against these objectives in possible future built environment is not always equipped to weather scenarios, which can be constructed by considering these shocks, which leads to vulnerability in the WSS the various risks faced by the system. The system’s infrastructure, and in the urban structures that affect ability to meet the performance objectives under var- service provision. At the same time, tensions between ious types of stress will determine its current resil- municipal and other water users are growing over ience, and help identify measures to strengthen it in access to water resources. Because climate change the future. This requires analyzing various scenarios amplifies existing uncertainties and threats it should and the associated responses to them, and develop- not be evaluated as a stand-alone impact. In order to ing a publicly communicated plan of how to handle promote water security, it is critically important to both the triggers and the responses. Building Urban Water Resilience in Small Island Countries 3 Chapter 3 Why Does Water Supply Resilience Matter for South Tarawa? 3.1.  Geographic, Physical, and 6.  South Tarawa has uniquely fragile water resources Socioeconomic Fragility due to its small size, lack of natural capacity for water 5.  The Republic of Kiribati is one of the smallest, most storage, and competing land use. South Tarawa is the remote, and most geographically dispersed countries in country’s main urban center: it is located on the atoll the world. The country consists of one raised coral of Tarawa, and spans a string of coral islets connected island and 32 low-lying coral islands distributed in by several causeways. Its population is expected to three groups: the Line Islands, the Phoenix Islands, grow from 58,000 in 2016 to 96,000 in 2040. No more and the Gilbert Islands. The capital city, South than two meters above mean sea level, South Tarawa Tarawa, which is in the Gilbert Islands, is located stretches for more than 30 kilometers and is on aver- about 4,000 kilometers from the major trade markets age less than 250 meters wide. Only three of the islets of Australia and New Zealand, as illustrated in are wide enough to hold meaningful groundwater map 3.1. South Tarawa presents features of striking storage volumes: the urban center of Betio, and the geographic, physical, and socioeconomic fragility. islets of Bonriki and Buota. As shown in map 3.2, the population density exceeds 7,500 inhabitants per square kilometer in many of the islets (a remarkably high number considering the absence of multistory MAP 3.1. Location of Tarawa and Kiribati in the residential buildings). This exerts constant pressure Pacific Ocean on the integrity of the water reserves. 7. These geographic features create significant chal- lenges for human development and economic growth. According to the last available Household Income KIRIBATI and Expenditure Survey (HIES), conducted in 2006, TARAWA 24 percent of the population of South Tarawa lived under the poverty line (Kiribati National Statistics Office 2006). Opportunities for cash employment and consumption, and access to higher education and specialized social services that are not available else- where in Kiribati, have made South Tarawa a magnet for internal migration from the Outer Islands. This further increases the challenges of population den- sity and urban development. In 2012 it was estimated that half of South Tarawa’s population was living in IBRD 44312 | MARCH 2019 informal areas. Only around 20 percent of the coun- Source: IBRD 44312 | March 2019 try’s population is formally employed in the cash Building Urban Water Resilience in Small Island Countries 5 MAP 3.2. South Tarawa: Population Density and Main Physical Featuresa Buota Population density (inhabitants per km2) < 5,000 Water reserves Bonriki < 5,000–10,000 Uninhabitable areas > 10,000 (swamp, sh farm, tree farm) Bikenibeu Betio 2 km Bairiki 2 km Source: World Bank data. Note: Heteromorphic map to aid visualization. a. The islet of Buota is administratively located in North Tarawa, but due to its integration into South Tarawa’s economy and water system, it is broadly understood in this report to be included in “South Tarawa.” economy, with 70 percent of the jobs provided by the every 48 hours, and at very low pressure. Connected public sector (World Bank 2018). Despite improve- households consume an average of 10 liters per cap- ments in revenues in recent years, Kiribati’s Human ita per day (lpcd), while nonresidential customers Development Index (HDI) ranking is 137, and its consume about 3 lpcd. Water is chlorinated at vari- Human Capital Index (HCI) ranking is 113; both are ous points in the system (and on demand, through among the lowest in the Pacific region. water tanker delivery), but negative pressure in the distribution pipelines is leading to groundwater infil- 3.2. Chronic Water Supply Insecurity tration and recurrent bacteriological contamination. 3.2.1. Historical Reliance on Three Sources of Water Water at the tap fails to comply with microbial water 8. Despite the presence of a centralized water sup- quality standards two thirds of the time.2 Water is ply system, access to safely managed water supply currently supplied for free to most customers, except ­ services remains very limited in South Tarawa. About in three pilot areas (Nanikaai, Tebikenikora, and 90 percent of the urban population in Kiribati Tanaea), where network upgrades under the Kiribati (which largely overlaps with the South Tarawa pop- Adaptation Program (KAP)-III now allow the delivery ulation) has access to a basic water supply service. 1 of a continuous water supply.3 The purchase of fuel, South Tarawa’s reticulated water supply system pro- such as wood and kerosene, presents an additional vides service to about two thirds of its population expense for the majority of households (58 percent), through house connections (as shown in photo 3.1), which currently boil water before consuming it (ADB and is operated by the Public Utilities Board (PUB). 2018). Many households rely on multiple water Water is abstracted primarily from two freshwa- sources, including rainwater collected from roofs, ter lenses, located in the Bonriki and Buota water and local groundwater from household wells. Both reserves. Services are provided for up to 2 hours reticulated water and groundwater show high levels 6 Building Urban Water Resilience in Small Island Countries PHOTO 3.1. PUB Water Connection Source: © Stephane Dahan/World Bank. of bacterial contamination. A baseline survey carried which set up a revolving fund in 2001 to provide loans out by the World Bank in 2018 found that two thirds for public servants to purchase rainwater collection of households spend more than half an hour a day on and individual storage devices. This scheme led to the water collection tasks. This situation is expected to 4 installation of more than 1,000 water storage tanks improve significantly with the donor-funded project (which typically hold 6 cubic meters each) across that is due to be approved in 2019, as explained in South Tarawa. An analysis of drone imagery and field section 3.3. surveys conducted for this study suggests that more than 2,000 buildings (that is, about 20 percent of resi- 9.  Although rainwater harvesting (RWH) has emerged dential buildings, and 38 percent of nonresidential as an important coping strategy, much of its potential is buildings) are now equipped with RWH systems unrealized. The use of RWH, historically practiced in across South Tarawa, as shown in figure 3.1. The Kiribati, was significantly boosted by the develop- uptake of residential RWH systems remains limited ment of modern roofing that allows rainwater cap- due to installation costs, and to the prevalence of tra- ture. More recently it was expanded by the Sanitation ditional housing (with its thatched roofs), which Public Health and Environment (SAPHE) project, Building Urban Water Resilience in Small Island Countries 7 FIGURE 3.1. Prevalence of Rainwater Harvesting across Buildings in Betio, Bairiki, and Bikenibeu 25 20 Cumulative roof area (ha) 15 10 5 0 0 0 0 0 0 0 0 0 0 0 00 90 40 50 70 30 10 20 60 80 00 ,0 0– 0– 0– 0– 0– 0– 0– 0– 0– 1, >1 80 60 40 0– 30 20 10 50 70 90 Building size (roof area, m2) Without rainwater harvesting With rainwater harvesting Source: World Bank data. represented two thirds of all residential buildings in generally perceived by the people as the highest qual- 2013 (ADB 2013). Nevertheless, around 60 percent of ity water resource, its use for nonconsumptive pur- households report using rainwater on a regular basis, poses is not viewed positively by the people. Overall, with our survey revealing that: (1) one third owning our survey suggests that rainwater consumption rep- RWH systems; (2) one third using rainwater that is ­esidential resents 5 lpcd, about half of which is for r shared with neighbors (on average two households water use, as shown in table 3.1. Household rainwater per system); and (3) the rest purchasing rainwater is largely used for human consumption despite the from nonresidential buildings (for example, from almost complete absence of disinfection systems. community halls or churches). Unfortunately, even when the system is well main- tained and clear of leaves, the presence of birds, rats, 10.  Rainwater is perceived locally as the most precious and other animals/vectors can be a source of contami- water resource, and it is used primarily for drinking nation (and potential corrosion of the roof can add water, despite its substandard bacteriological quality. heavy metals to the mix). Weekly random water qual- The total RWH catchment area is estimated at ity testing by the Ministry of Health and Medical 30 hectares. However, only a limited fraction of the Services (MHMS) across South Tarawa in 2017 showed water collected, generally less than 50 percent in col- the presence of E. coli more than 90 percent of the lective buildings and about 10 percent in individual time.6 houses, is eventually consumed.5 The rest overflows during rain events. Indeed, stored rainwater is per- 11. The current stock of RWH systems illustrates the ceived as a lifeline option in case of significant disrup- high value the South Tarawa population assigns to tions in PUB services, and it is used sparingly, even water  quality and security. The estimated value when tank water levels are high. As rainwater is (at  installation) of the RWH assets exceeded US$10 8 Building Urban Water Resilience in Small Island Countries TABLE 3.1. Water Use by Source (lpcd): Current Situation (2019) Residential Other Non consumptive Consump­tive Laundry/ Toilet Animals, All uses washing flushing gardening Total Total 5 18 6 7 6 42 PUB 4 4 <1 <1 2 10 Rain 2 <1 0 0 3 5 Urban well <1 14 6 7 1 28 Seawater 7 Source: World Bank data. Note: Color scale: the darkness of the blue is proportional to the amount of lpcd. ­ on average, about US$200 per South Tarawa million ( their middle, groundwater is generally brackish. It is inhabitant). In a large majority of cases, these systems also almost systematically contaminated bacteriologi- were financed by the building owners, sometimes cally, and is used primarily for laundry, bathing, gar- with the support of microcredit schemes (such as the dening, and so on. It is widely recognized in South SAPHE revolving fund). As discussed in section 3.1.1., Tarawa that urban groundwater is not suitable for the associated average cost over the 20-year life of the human consumption, although its salinity is a stron- equipment has exceeded US$6 per cubic meter. People ger deterrent to consumption than any information lacking access to rainwater from their own, or a neigh- about microbial or other types of contamination. bor’s, household system can purchase rainwater from There has been a gradual shift away from using well nonresidential buildings that sell water to the public. water for drinking or cooking purposes over the past The most common market price is US$18 per cubic two decades. This has been prompted by the increased meter (or $A 0.50 per bucket). This very high level of salinization of the local lenses, (which is probably due willingness-to-pay demonstrates the high value to increasing water abstraction and urban density); as placed by the population on this resource. In the pres- well as by the availability of alternative sources of ence of other free water resources, it illustrates the water (PUB water and RWH). However, our survey ­ uality— high premium people put on the organoleptic q shows that 8 percent of households still report using that is, the taste—of rainwater (compared to the often well water for drinking or cooking in the absence of brackish taste of well water, or even PUB water), and accessible or affordable alternatives.7 Abstraction is on the security represented by full rainwater tanks usually done manually, although more than 20 per- (compared to the erratic PUB water supply). cent of households were found to use electric pumps in a 2011 household survey (ADB 2011). The last major 12. Urban groundwater is abundantly tapped into for urban groundwater quality assessment campaign was nonconsumptive uses through a network of private, carried out in 2011 in Betio, Bairiki, and Bikenibeu. shared, and communal shallow wells, such as the one Results of the analysis showed significant microbio- illustrated in photo 3.2. Outside of the few islets that logical contamination; a high nitrate content (up to are wide enough to hold a small freshwater lens in 800 milligrams per liter) in the most densely Building Urban Water Resilience in Small Island Countries 9 PHOTO 3.2. Shallow Well Next to a Private Toilet and Pig Enclosure Source: © Stephane Dahan/World Bank. populated areas; and petroleum-related industrial almost as much organic pollution as the human pop- pollution in several areas (ADB 2011). ulation.10 Kiribati’s infant mortality rate is among the 13. Inadequate sanitation is a major issue in South highest in the Pacific, at 44 per 1,000 live births, and Tarawa. Only 49 percent of the population has access is partly attributable to infantile diarrhea (UNICEF to at least basic sanitation services. The other 51 per- 8 2013). In 2012, one in every two persons was treated cent uses shared sanitation facilities; unimproved for a waterborne disease in a hospital or clinic in onsite sanitation systems such as pit latrines without South Tarawa (ADB 2013). a slab or platform; or practices open defecation in the sea (near-shore), with 60 percent of the population 14.  Major efforts remain to be undertaken toward the resorting to this at least occasionally (ADB 2013). conceptualization and further development of san- Fecal contamination of the environment due to the itation services. South Tarawa has three sewerage human waste stream is aggravated by the unsafe dis- systems, coupled with seawater supply networks posal of pig manure. There were as many as 16,075 for flushing, in the original settlement centers. pigs in South Tarawa in 2015, and they generated 9 About 25 percent of the population is connected to 10 Building Urban Water Resilience in Small Island Countries these  systems. PUB offers a vacuum truck service habits, and the emotional cost of negotiating access that comes to households to empty septic tanks. to water with neighbors or relatives. People who have All sewage (from both sewerage systems and septic to walk the “last mile” to get better quality water do tanks) is discharged through submarine outfalls into not always walk it. More than half of households boil the deep ocean, beyond the reef, after screening: water before consuming it. There is also limited use dilution, dispersion, and die-off are relied upon to of the solar water disinfection (SODIS) method.11 avoid pollution of near-shore areas. Except in areas 16. Households in South Tarawa follow several key that have sewerage systems, households generally water scarcity management principles. In the context discharge greywater locally, taking advantage of of unreliable sources of water (whether from PUB, or the high infiltration capacity of coral sand. There is from rainwater) and low-quality sources (groundwa- currently no consensus among sector stakeholders ter), the survey conducted as part of this study on the most appropriate sanitation option for South showed that the vast majority of households in South Tarawa as a whole, as a universal sewerage system Tarawa use a variety of sources to meet their daily cover would be unaffordable. A concept study is cur- water needs. Our survey found that 50 percent of rently being carried out with financial support from respondents use three different sources of water on a New Zealand to identify appropriate sanitation mod- regular basis; 43 percent use two sources; and only els, and to prepare for their implementation. While 6 percent use a single source. This almost systematic seawater flushing helps minimize freshwater use, it reliance on multiple sources of water is consistent is suspected that leaks from seawater networks are with findings from other Pacific island countries, affecting the salinity of urban lenses. such as the Marshall Islands or the Solomon Islands Strategies for Coping with Inadequate 3.2.2.  (Elliott et al. 2017). This allows fit-for-purpose water Water Supply use depending on water quality—as explained in the 15. Water use practices in South Tarawa are largely previous paragraph and illustrated in box 3.1—and aligned with techno-scientific definitions of water qual- adaptive water use, depending on the availability of ity, although choices about water use are rarely con- PUB water and the abundance of rainfall. Local water nected to health considerations. The practice of boiling storage helps buffer variability of supply. More than water is very common, and children’s water con- 75 percent of the population stores water in one of sumption is often different from that of adults. the household storage tanks that were distributed in However, when people acknowledge that there is a the 2000s as part of the SAPHE project, while the rest link between water and illness, they typically refer to collect water in buckets, small drums, or pots. an external source of information (for example, Monitoring of the water level in rainwater tanks plays health professionals) rather than stating it as a per- an early warning role, drawing attention to fluctua- sonal or deeply ingrained insight. Well water is tions in the water supply, and increasing at least the avoided as much as possible for consumptive uses, stated awareness of residents regarding the need to but this is mostly due to its salinity, or its organolep- save water. Finally, many households already prac- tic properties, rather than its bacteriological contam- tice greywater reuse to some extent for gardening or ination. In practice, the people’s satisfaction with for their pigs, closing the water cycle. Thus, South their water is less connected to concerns of water Tarawa provides a remarkable illustration of how quality than to convenience of access, entrenched best practice water management principles in a Building Urban Water Resilience in Small Island Countries 11 BOX 3.1. Ancestral Thrifty Water Use Traditions in Kiribati Households Ms. N. is a 49-year-old working mother living cleaning the dishes, cooking, and hand washing. in the urban village of Bairiki. She lives with her Ms. N tells her children not to shower with PUB husband and their extended family in a household water: “We need to conserve it.” Like many people of 10 persons. Ms. N. is satisfied with her water in South Tarawa, she does not like the taste of PUB situation: she has access to all three sources of water, but she drinks it anyway, and boils it for tea water available in South Tarawa. The family uses when necessary. But her favorite source of drinking water sparingly, and carefully matches different water is rainwater: “It is easier to drink,” she says. sources with intended uses. Their well is of The family owns a 3,000-liter rainwater tank. The particular importance in their daily life, and Ms. N. priority use for that water is drinking, but when it feels sorry for those who don’t own a private well: is abundant, rainwater can also be used for other “A well is a basic need,” she says. Although the purposes. “Messages are broadcast on the radio water is brackish—Ms. N. knows it is, because the to inform the population when it is time to ration soap does not lather when they shower—water is water, but we can see it by ourselves.” Although used for multiple purposes: dishwashing, bathing, Ms. N. does not use kitchen water for gardening, gardening, laundry, toilet flushing, and feeding the the most nutritious part of it containing leftover family’s five pigs. A significant part of the water food is carefully set aside for the pigs and mixed used in the household comes from the public water with their feed. supply network (PUB). PUB water is reserved for Source: Face-to-face interview, July 2018. context of water scarcity can be applied at the house- culturally significant customs in contemporary hold level (World Bank 2017). Tarawa, such as cultural norms of generosity, exchange, and reciprocity. These play a central role in 17.  Collectivism and water sharing play a critical role in maintaining social cohesion, cultural identity, and reducing affordability challenges, water-related con- urban survival. The prevalence of kin relationships in flicts, and vulnerability to droughts. As in many Pacific Kiribati society is reflected in the networks of water cultures, people in Kiribati hold to a collectivist worl- sharing in South Tarawa. Kin-based sharing can, dview, in which people “identify as being part of a however, also exacerbate the vulnerability of indi- large interdependent group comprising the nuclear viduals, and consuming shared water is almost never family, extended family and lineage members from the preferred option. Birth position and gender often their village community” (Kolandai-Matchett et al. determine the degree of control over shared water 2017). The powerful ties that bind society together sources, and can affect the location of houses relative through water exchange offers important informa- to the water infrastructure. tion about the everyday uses of water at the domestic and community level. A significant portion of access 18.  While the sharing of water with relatives, friends, to water in South Tarawa is probably attributable to or neighbors helps alleviate exposure to water risks, it water sharing between households, as illustrated in can also exacerbate individual vulnerability and social box 3.2. This is particularly significant in areas that tensions. Observations conducted for this report are densely populated. Water sharing helps maintain also found that the need to depend on relatives for 12 Building Urban Water Resilience in Small Island Countries BOX 3.2. Water Sharing–A Social Duty Deeply Anchored in Kiribati Culture Rotia, a 30-year-old woman, lives in the Betio access to rainwater for free from a neighbor’s tank, village, in the center of the island, together with but in limited quantities (four 12-liter buckets for her husband and their two children. The family 3 days). Even when these sources of water are shares a connection to the PUB system with not sufficient, Rotia says that they don’t resort to three other households, totaling 21 persons. This drinking water from their personal well, which they means that when the water is distributed to their share with more than 30 persons in the community. rationing zone every other day, the family needs Instead, she occasionally buys bottled water, or to be present, and negotiate with their neighbors asks for water from a relative living nearby who to make sure they can store enough water to last has “a good well.” until the next distribution time. Rotia also has Source: Face-to-face interview, July 2018. drinking water can significantly increase the dis- Increasing Water Scarcity and the 3.3.  tances that have to be covered to fetch water. Need for New Resources Furthermore, within family water-sharing arrange- 19. Inherent limitations in freshwater resources and ments, sisters and younger siblings may have less continuous population growth are exposing South control over the water than their older brothers. Tarawa to a growing water supply deficit. Freshwater Such dependence can become a stigmatizing and availability12 across South Tarawa has dramatically stressful experience. We noticed that social and cul- declined in recent decades, as shown in figure 3.2, tural obligations to share water are increasingly primarily due to a six-fold population increase over being questioned. As a result, several owners and the last 50 years. Freshwater availability now hovers users of rainwater tanks mentioned tensions caused around the absolute water scarcity threshold,13 and it by the quantity of water that is being shared. reached an all-time low of 277 lpcd in 2018. Bonriki Cultural factors can also increase hardship in the and Buota’s rainfall-fed groundwater lenses, which case of collective RWH systems. In some cases, peo- are used for producing PUB drinking water, have a ple choose to spend more time fetching water in combined sustainable yield of approximately 2,000 order to ensure that it is from their own parish. cubic meters per day. The estimated sustainable Population growth, the geographic fragmentation of freshwater yield of the urban water lens in Betio is kin groups, environmental changes, and the marginal, at 25 cubic meters per day (White 2010). rise of individualist values are contributing to dis- Even with a major reduction in water losses mantling the cooperative living system that used to ­ (currently at 60 percent), and with consumption be intrinsic in society, and increasing the perceived limited to basic needs, basic water demand is ­ price of conforming to water-sharing norms. expected to exceed groundwater production capac- The improvement of PUB services will probably fur- ity by 2,500 cubic meters per day in 2020, increasing ther undermine the role of traditional kin-based to a range of 3,300–4,800 cubic meters per day in water sharing. 2040 (ADB 2017a). Building Urban Water Resilience in Small Island Countries 13 FIGURE 3.2. Per Capita Freshwater Availability in South Tarawa, 1947–2018 25,600 Per capita freshwater availability (m3/year/person) 12,800 6,400 3,200 No water stress 1,600 Water stress 800 Water scarcity 400 Absolute water scarcity 200 100 57 62 67 72 77 82 87 92 97 02 07 47 52 12 17 20 20 19 19 19 19 19 19 19 19 19 19 19 20 20 Sources: Kiribati Meteorological Services (precipitation data); South Tarawa Office of Te Beretitenti and T’Makei Services, 2012 (population data). Note: A logarithmic scale is used; population data available at: http://www.climate.gov.ki/wp-content/uploads/2013/01/6_South-Tarawa-revised-2012.pdf. 20. This deepening water deficit along with emerging system to address water supply deficits in the long climate-related challenges require the consideration of term, and build resilience to climate change. The proj- new resources for South Tarawa’s potable water supply. ect, which is under preparation with donor support, The government of Kiribati adopted the National Water will implement a 4,000-cubic-meter-per-day seawa- Resources Policy in 2008. Demonstrating the impor- ter desalination plant, and a solar photovoltaic (PV) tance of this sector, it also developed the Tarawa Water array to offset the energy requirements of the sys- Master Plan 2010–30 and the Tarawa Water and tem. The aim is to guarantee consumption of 59 lpcd Sanitation Roadmap 2011–30. Considering all water (52 lpcd of which is for residential water demand) production options and the potential for reduction in with a water system that is resilient to climate-re- physical losses of water, these plans identified noncon- lated threats such as droughts and seawater overtop- ventional water sources, such as desalination, as the ping14 (SWOT). The seawater desalination system will only option that will be able to meet South Tarawa’s be expandable to 6,000 cubic meters per day should water demand over the long run, along with a diversifi- residential water demand further increase in the cation of sources, including urban groundwater and future. The system will be in Betio, which is the only RWH, for risk mitigation and increased resilience. available land option identified during project prepa- These studies encouraged the use of rainwater, but ration. STWSP will also rehabilitate and upgrade point out that it cannot be relied on during droughts. South Tarawa’s water distribution system to achieve universal coverage, which is expected by 2022, and 21.  The South Tarawa Water Supply Project (STWSP), to will drastically reduce physical losses, which are be initiated in 2019, will finance a seawater desalination 14 Building Urban Water Resilience in Small Island Countries TABLE 3.2. Water Use by Source (lpcd): Future Situation (2030) Residential Other Non consumptive Consump­tive Laundry/ Toilet Animals, All uses washing flushing gardening Total Total 10 39 10 9 12 80 PUB 8 37 8 2 8 63 Rain 2 3 5 Urban well 2 2 7 1 12 Seawater 10 Source: World Bank data estimates. Note: Color scale: the darkness of the blue is proportional to the amount of lpcd. expected to fall below 25 percent after rehabilitation washing, mopping, gardening, and for pigs. Overall, of the system. The centralized water supply system current consumption from RWH and from shallow will rely on a combination of desalinated seawater wells is estimated, respectively, as 2 lpcd and 20 lpcd. and groundwater abstracted from the Bonriki and Table 3.2 shows the expected shift toward PUB water Buota freshwater lenses. It will also incorporate two under the new water supply system, with households desalination plants, which can produce 500 and 100 still relying to some extent on shallow wells for non- cubic meters per day (about 5 lpcd overall), one at the consumptive uses, and on rainwater. The behavior in western end of the Betio causeway and the other at communities that already have a continuous water the Tungaru Central Hospital. These are currently supply under the KAP III pilot project confirms that under completion with financial support from the households continue to rely, albeit to a lesser extent, government of New Zealand. Groundwater resources on these local water resources. Actual reliance on the are expected to contribute 26 percent to PUB’s water various sources depends on factors such as PUB sources by 2030. water tariffs, and the population’s level of confidence in the potability of PUB water. 22.  In addition to using upgraded PUB water services, South Tarawa’s population is expected to continue to rely to some extent on shallow wells and RWH. While 3.4. Water Sector Vulnerabilities no precise quantitative information is available on 23.  Despite the PUB system upgrade, the water sector current consumption from these sources, various still presents key vulnerabilities. Water supply vulnera- surveys15 provide a general appreciation of their con- bilities are both natural and anthropic. South Tarawa’s tribution to people’s mix of water sources to date water balance is vulnerable to the effects of climate (with PUB water services still limited to 2 hours every change (droughts and SWOT); to the impact of human 48 hours). As indicated earlier, rainwater is, together activities on freshwater resources (encroachment and with PUB water, the preferred source of drinking pollution in water reserve areas); to possible systemic water, and to a lesser extent it is used for cooking. Its water infrastructure (desalination plant breakdowns); use for other purposes is and will remain marginal. and to an unforeseen escalation in water demand. Shallow wells are recognized as the resource of low- From a probabilistic perspective, these hazards are est quality, and are used primarily for water for independent of each other. It should be noted that the Building Urban Water Resilience in Small Island Countries 15 critical water infrastructure is being designed under cumulative precipitation has been plotted since the STWSP to withstand SWOT). It should also be noted 1950s, for 5- and 2-year periods, respectively. South that due to its proximity to the equator, South Tarawa Tarawa experiences occasional severe droughts, is never on the path of cyclones. This section provides which are often associated with La Niña events. In an overview of the relevant vulnerabilities for South 1998–2000, during one of these events, the amount Tarawa, and proposes summary scenarios to assess of precipitation was a third of the annual average potential impacts on the current system in each case. (2,020 millimeters per year). Although climate change scenarios are projecting an increase in 3.4.1. Droughts annual and seasonal mean rainfall with high confi- 24.  There is little understanding of how droughts and dence (PACCSAP 2015), groundwater recharge is climate change are likely to affect South Tarawa in the expected to be negatively affected by additional future. The climate of Kiribati shows high year-to- evapotranspiration caused by an increase in the year variability (particularly for rainfall), driven pri- intensity and frequency of days of extreme heat. marily by the El Niño Southern Oscillation (ENSO), Long-term droughts (those lasting more than a year) as can be seen in figures 3.3 and 3.4, where the occur on average every 10 years, and are understood FIGURE 3.3. Five-Year Cumulative Rainfall in South Tarawa, Actual/70-Year Average Ratio 180 160 140 120 Percent 100 80 60 40 20 0 1947 1957 1967 1977 1987 1997 2007 2017 Source: Kiribati Meteorological Services. FIGURE 3.4. Two-Year Cumulative Rainfall in South Tarawa, Actual/70-Year Average Ratio 200 150 Percent 100 50 0 1947 1957 1967 1977 1987 1997 2007 2017 Source: Kiribati Meteorological Services. 16 Building Urban Water Resilience in Small Island Countries to be almost fully driven by ENSO. There is no con- Organization (WHO) guidelines suggest a maximum sensus on how climate change will affect ENSO. salinity (EC) value for nonpotable purposes of 2,500 There is more consensus, however, on the expected Siemens per centimeter.16 At this value, water micro-­ sea level rise due to climate change—22–49 centime- can be considered as mildly brackish. In September ters by 2064—but the thickness of the Bonriki 2011, when conductivity surveys were carried out in groundwater lens is not expected to vary due to sea wells across Betio, Bairiki, and Bikenibeu, 51 percent level rise, at least not until 2050 (Mack 2015). of the samples exceeded this threshold. South Tarawa had, at that time, received cumulative precipitation 25.  Extreme droughts could lead to a significant decrease over the previous 2 years, comparable to the 1947– in yields from water reserves. Severe droughts, if they 2017 average, as can be seen in figure 3.4. increase in frequency and length, could lead to saline intrusion at the perimeter of the lens, and to a signifi- 27.  Major droughts will continue to cause salination of cant deterioration of water quality at many of its infiltra- urban groundwater lenses. The small urban freshwater tion galleries if abstractions are not reduced accordingly. lenses of South Tarawa from which urban wells are fed The Bonriki water lens is estimated to have a hydraulic (excluding the Bonriki and Buota lenses) are estimated residence time of 4–5 years (White, Falkland, and Scott to have hydraulic residence times of between a few 1999). As shown in figure 3.3, every 10 years on average months and two and a half years, with the times gen- the cumulative amount of precipitation over the previ- erally higher the further away they are from the coast ous 5 years drops by more than 20 percent below the (ADB 2011; White et al. 1999). As shown in figure 3.4, 1947–2017 average (40 percent during the recent 1998– every 6 years on average, the cumulative amount of 2000 drought). Consistent with this reduction in cumu- precipitation over the previous 2 years drops by more lative precipitation, hydrogeological modeling has than 50 percent below the 1947–2017 average (and it suggested a decrease of 40 percent in the sustainable dropped by 75 percent during the 1998–2000 drought). yield of the Bonriki lens in the event of such a severe That severe drought event resulted in dramatic ­ cenario does drought event (Jolliffe 2017). While such a s increases in the salinity of domestic wells and in the not currently appear in climate projections, it is consis- death of trees (Jolliffee 2017). In the absence of long- tent with South Tarawa’s history of drought. A  more term monitoring data to quantify the impact of extreme scenario, which has not been experienced so droughts on the urban groundwater lenses of South possibility—that is, one far but cannot be excluded as a ­ Tarawa, it is difficult to say precisely what the impact involving two such droughts in a 20-year period—would of such an event has been on the quality of well water lead to a reduction of up to 50 percent in the sustain- across South Tarawa’s islets. However, it can be esti- able yield. Given the range of uncertainties surrounding mated with a high level of confidence that a 75 percent the influence of climate change on ENSO, and the fre- reduction in cumulative precipitation for a 2-year quency and intensity of future droughts, a 50 percent period would cause salinity in most wells to strongly reduction in the yields from the water reserves will be exceed the WHO guidelines on maximum salinity val- considered as the worst-case scenario. ues for nonpotable purposes. 26.  Urban well water salinity is also affected by major 28.  In the absence of alternatives, temporary deterio- drought events, and can be expected to significantly ration of urban groundwater quality does not signifi- exceed acceptable standards for nonpotable water cantly impact the use of this resource. Although it is during these events. The 1993 World Health common knowledge locally that groundwater is Building Urban Water Resilience in Small Island Countries 17 subject to saline intrusion during droughts (as well as FIGURE 3.5. Return Period of Rainwater Harvesting System Dry-Ups Exceeding 1 Month during high tides), awareness of the recent changes in salinity levels is limited, since most people refrain 10,000 from drinking well water anyway. Variations in salin- 9,000 ity levels are more often noticed, for example, by women when their detergents or soaps do not lather 8,000 Storage volume (liters) as easily as usual. Anecdotal evidence that some 7,000 households have shifted entirely to PUB water or 6,000 rainwater due to excessive groundwater salinity has 5,000 been gathered during this research. However, as ye rs a s ye 4,000 ar shown by the household survey conducted in June 50 s ar rs 10 ye 2018, during a significant drought event (with a a 3,000 ye 5 2 rs 57 percent drop in 2-year precipitation compared to ea 1y 2,000 the normal situation), 90 percent of the population 1,000 continued to use well water. This illustrates the lack 6 8 10 12 14 16 18 20 22 24 Number of users per system of water supply options during episodes of severe water scarcity, and it does not represent a desirable Source: World Bank data. level of service. 29.  Most RWH systems cannot withstand major drought Our detailed assessment of the 40 largest systems sup- events, as was seen during the 2017–18 dry spell. plying 5,600 people suggests that more than 60 per- A simulation of tank-operating regimes over 50 years cent of them (in terms of number of users) are likely to of monthly precipitation records suggests that, if used run dry once every 10 years. by one household at an average of 5 lpcd, the common residential RWH systems found in South Tarawa17 3.4.2. Seawater Overtopping could withstand any of the droughts recorded in the 30. Sea level rise will aggravate both the likelihood past 50 years. This outcome is, however, very sensitive and the impacts of SWOT on the water reserves. SWOT to the number of households that use the system could be generated in South Tarawa by a combination (­ typically 1–3 households), and to the size of the tank of storm surges triggered by low atmospheric pres- (typically 3–7 cubic meters), as shown in figure 3.5. sure; high-level, long-term sea level rise; and severe Across this spectrum of system configurations, about wave conditions created by distant weather forma- 80 percent of residential systems would not withstand tions, in particular in the North Pacific. Currently the a 10-year return drought without running dry for at risk of significant inundation of the Bonriki water least 1 month. This situation should now improve, reserve leading to a significant impact on the lens is since the 2017 Kiribati National Building Code defines negligible, and such an event has never been household storage requirements based on the number recorded. However, over time this risk is expected to of users, and is designed to prevent system failure grow with the projected rising of seawater levels, more often than 1 month in 10 years. Nevertheless which is predicted to be 22–49 centimeters by 2064, the performance of large, nonresidential RWH sys- across a range of climate scenarios studied by tems is highly variable, given the multiplicity of the Intergovernmental Panel on Climate Change ­ building types, water uses, and number of purchasers. (IPCC) and the National Oceanic and Atmospheric 18 Building Urban Water Resilience in Small Island Countries Administration (NOAA).18 Modeling of a 28-centime- 3.4.3. Anthropic Risks ter sea level rise (which could be reached before 33. The most serious anthropic risk in the long term 2050) suggests that SWOT that temporarily (that is, is contamination of the water reserves to a degree for up to 5 years) were to reduce yields from the that would require costly treatment, in addition to cur- Bonriki lens by 54 percent would have a 20-year rent means of treatment being used (aeration and return period (Mack 2015). Map 3.3.a shows the chlorination). There are currently close to 80 unau- location of infiltration galleries in Bonriki, while ­ thorized dwellings on the Bonriki water reserve, and map 3.3.b illustrates the extent of SWOT across the 20 on the Buota reserve. Overall, this represents a water reserve in that scenario. Following such an 300 percent increase compared to the situation in event, it would be necessary to significantly limit 2005. Encroachment by settlers and squatters is con- abstraction for up to 5 years in order to allow the tinuously increasing because of South Tarawa’s rap- lenses to recharge. Should drought occur at the same idly growing population, the scarcity of land, and the time, the limitations would be even more restrictive. lack of enforcement of existing regulations. Human 31.  The risks of SWOT on the Buota water reserve and in settlement and activity on these reserves could affect urban areas cannot be quantified precisely at this stage. water quality through sand and gravel mining; the The conditions—sea level rise, with swelling on the digging of wells; the use of graveyards; the raising of lagoon side generated by storms in the North Pacific— pigs; and the growing of crops and taro swamp fields, that were simulated in the study of the Bonriki water which use animal manure and fertilizers; as well as reserve would also be relevant to the Buota water direct pollution of the freshwater lenses through reserve, and to the freshwater lenses in urban areas poor sanitation practices. As indicated in the Tarawa from which private wells draw water. In the absence of Water Master Plan (2010), “continued settlement [in modeling based on detailed topographic data, it is not the water reserve areas] greatly increases the risk of a possible to infer the risk of significant inundation and severe disease outbreak, such as the 1977 cholera what its impact would be on these local groundwater outbreak.” It is worth noting that South Tarawa’s two lenses. As a first, and probably conservative approxi- other water reserves, in Betio and Teaoraereke, mation, it will be considered that the impact of a major which are much smaller than the one in Bonriki, SWOT event on the freshwater yield of Buota and were abandoned in the past due to similar encroach- urban water lenses would be similar to those expected ment issues. The proposed expansion of Bonriki to affect the Bonriki water lens. International Airport, with airplane hangars, work- shops, and possibly fuel stations within the reserve, 32. Severe droughts represent a more immediate would also entail pollution risks that would require threat than SWOT. While the yield reduction for over- careful management. The key environmental risks at topping appears comparable in size and probability airports relate to fuel storage, transport and refuel- to that caused by severe droughts, the risks associ- ing, electrical substations, and the storage of chemi- ated with SWOT would only materialize following a cal products (Nunes et al. 2011). significant rise in sea level. Analyses undertaken as part of the STWSP preparation suggest that climate 34. Any contamination of the Bonriki water lens change-induced SWOT and drought could together would affect water supply, as any treatment process lead to a drop in the lenses’ yield to nearly zero for a would be very costly and time consuming. While bacte- significant period of time (ADB 2017b). riological contamination could be addressed through Building Urban Water Resilience in Small Island Countries 19 MAP 3.3. Bonriki Water Reserve Infiltration Galleries and Seawater Overtopping a. In iltration galleries and monitoring network Monitoring Bores In iltration Gallery Gallery Pumps Location BN36 Airport Pond 17 Coastline BN11B 19 BN34 BN11 16 15 BN1B BN20B BN1 BN20 14 BN2 BN2B BN236BN23B PUB water treatment plant BN19 18 BN15 13 BN4 BN4C BN27 BN15B 11 10 9 5 BN24 12 8 7 BN7 BN21 BN28 BN35 4 BN13B BN7B 3 BN22 2 22 BN26 BN13 1 BN29 21 BN32 BN25 BN33 20 b. Inundation Max. Water Level (m) <0.1 <0.25 <0.5 <0.75 <1.0 <1.5 <2.0 >2.0 Sources: panel a, Galvis-Rodriguez et al. 2016; b, Fenton and Laplante 2018. Note: Simulation for Representative Concentration Pathway (RCP) 8.5, 2064 sea level rise, offshore scenario, 20-year return period event. 20 Building Urban Water Resilience in Small Island Countries chlorination, pollution by nitrates (fertilizers, pig breakdown and service interruption. The water distri- manure) or toxic substances (oils, fuel, metals, bution network is in an advanced state of disrepair. ­ chemicals) would not be treatable with the methods Physical non-revenue water is estimated to exceed currently used in Kiribati. Once the water lens is pol- 60 percent. Inadequate quality of water services luted, there is no available or affordable technology (regarding water pressure, continuous supply, and to remedy it. This pollution risk would probably so on), together with the effect of customers’ lack materialize gradually if related to human settle- of willingness to pay; their tampering with the net- ments, but in a short timeframe in the context of an works; and water leaks have been self-reinforcing in industrial or airport contamination incident. 19 a vicious cycle over the past two decades, against the backdrop of an increasing freshwater deficit. Facing 3.4.4. Technological and Operational Risks widespread discontent by the population, in 2013 35.  In the long term, any analysis of potential risks to PUB’s line ministry requested that it stop charging water supply availability must consider a temporary residential customers. In the absence of a residential interruption in seawater desalination due to technical water tariff, PUB water and sewerage revenues now issues. Desalination technology with solar offset mostly come from nondomestic customers, and from requires constant maintenance and the availability on-demand services such as delivery by tanker and of spare parts, including expensive membranes that septic tank emptying. Overall, from 2010 to 2017, must be replaced every 5 years. Some of the spare PUB’s income has been generated from electricity (85 parts are unavailable locally, and their procurement percent); water supply and sewerage (6 percent); and delivery from overseas, together with the nec- water supply subsidies in the form of community ser- essary repairs and possibly the need for external vice obligation (CSO) (7 percent); and other income technical support in the case of a major breakdown, (2 percent). The current average operation and main- could leave the capital city without desalinated tenance (O&M) costs of water services amounts to water for some weeks before the plant would be US$1.7 per cubic meter (m3) sold, a high level given back to full capacity. While the plant is being the system’s simple physical features. This is driven designed to take into account the risks of SWOT, notably by (1) high physical losses; (2) high electric- other calamities, such as fires, could also leave PUB ity20 and chlorine supply costs (26 percent of total without the capacity for desalination. This was percent of total costs); and (3) high labor costs (47  ­ recently experienced in the capital city of the costs). Some improvements in operational efficiency Maldives, as described in box 3.3. In such an event, have, however, been achieved since 2015 with which is entirely possible, the centralized water sys- enhanced donor support. tem would need to be supplied entirely from the 37.  The water network efficiency of PUB is critical to water reserves on the far eastern side of South maintaining water availability during a drought event. Tarawa, as these would be the only remaining The new water system implemented under STWSP is sources of PUB water. This risk underlines again the designed to cope with physical losses within the net- need to preserve the water reserves, and also to main- work of up to 25 percent, provided that water con- tain the capacity of Bonriki’s pumping systems to sumption does not exceed its design level. With an operate under this hydraulic regime. almost entirely renewed water supply network it is 36.  PUB’s current operational and financial underper- anticipated that water losses will be much lower. formance could lead to a significant risk of systemic Such a parameter could have an indirect impact on Building Urban Water Resilience in Small Island Countries 21 BOX 3.3. Water Crisis in Malé On December 4, 2014, a fire crippled the shops that had rationed mineral water, while seawater desalination plant in Malé, the capital hotels indicated that their supplies were running of the Maldives. This was the only potable water out. In response to the crisis, countries such as production system on the island. The city’s China, India, and Sri Lanka stepped in, shipping 120,000 inhabitants were instantly left without or airlifting in bottled water, as well as small access to potable water services, prompting the emergency desalination units, to help address declaration of a state of emergency. The only vital needs and calm public discontent. The crisis remaining water sources were the limited water lasted more than a week before the plant was put volumes stored in the city’s reservoirs. Malé’s lens, back into operation. As President Abdulla Yameen which had been depleted and salinized by decades declared to the media: “We did not have any of over-abstraction, could not be used for drinking fallback plan for any disaster of this magnitude.” or washing purposes. After several days without Sources: Freyberg 2014; “Maldives Faces Water Crisis after Fire at its Sole Desalination Plant,” DownToEarth, July 4, 2015, water supply services, the local media started https://www.downtoearth.org.in/news/maldives-faces-water- to report that angry residents were attacking crisis-after-fire-at-its-sole-desalination-plant-47757 the city’s resilience to a water crisis: (1) physical phase of the DBO contract, PUB could consider guar- losses could further reduce the availability of water anteeing repair or supply contracts with regional for supply, creating hardships for the population suppliers, in order to minimize response times in the during normal times, and even more during an emer- event of a crisis. gency; and (2) such losses would undermine PUB’s 39.  Assessing PUB’s future operational cost recovery financial position and its capacity to react, an aspect remains difficult in the absence of a water tariff, but that is not modeled in this analysis. preliminary analyses suggest that the water business 38.  STWSP will help strengthen PUB’s O&M capacity branch of the utility could become financially viable as through private sector support under Design-Build- a result of STWSP. A volumetric water tariff is already Operate (DBO) contracting arrangements. In addition applied in the three KAP-III pilot areas, illustrating to ensuring reliability of the desalination plant and the government’s willingness to discontinue the free the water supply network over the 5-year operation distribution of water. The Ministry of Infrastructure period (2023–27), the DBO contract will play a key role and Sustainable Energy (MISE) is now commission- in helping to improve PUB’s operational and financial ing a study to build on this experience, and identify a performance, and will help modernize its manage- tariff that would be applicable across South Tarawa ment systems and build its capacity to undertake after the water system has been upgraded by STWSP. preventive, predictive, and breakdown maintenance A preliminary cost-recovery analysis shows that, of the plant and network, and ensure sound asset with a tariff that is consistent with local socioeco- management. In such a remote location, and with a nomic parameters and an expressed willingness to water system that relies primarily on one desalina- pay, 155 percent of operating expenditure (OPEX) tion plant, keeping a generous supply of quickly (without depreciation) could be recovered during the available spare parts is advisable. Beyond the O&M O&M phase of the DBO contract, and 117 percent after 22 Building Urban Water Resilience in Small Island Countries the completion of STWSP. This is to be compared to and will largely depend on economic growth, the evo- the current cost coverage ratio of 32 percent. lution of cultural norms around water use, and the extent to which housing can accommodate modern 3.5.  How Resilient Will the Future Water plumbing and appliances. However, some factors sug- Supply Be? gest that there is likely to be a progressive loss of thrifty, resilient water practices over time. Water 40. Water resilience can be defined as the system’s resource diversification at the household level, and ability to meet defined delivery objectives despite the fetching of well water and transporting of rainwa- shocks. In the case of water supply, these objectives ter to the point of use, can be laborious and time con- could be simplified to maintaining a minimal, accept- suming. Also, traditionally, sharing water has not been able level of residential water availability for a city in so much a matter of decision, as part of a broader set of a prolonged emergency—that is, one lasting several social and cultural norms expressed in water use hab- weeks or more. The absolute minimum water con- its. As these norms are increasingly questioned in the sumption guidelines, such as those provided by obligation to share contemporary urban context, the ­ WHO, would not be acceptable in a capital city that water will likely be questioned as well. needs to sustain its social and economic activities, and where the population has become accustomed to 42.  In the long run, overall water consumption could a certain level of service (WHO 2013). During the reach 100 lpcd, that is, more than two and half times worst of its 2017 water crisis, the city of Cape Town, the current level. Considering STWSP design South Africa reduced its water demand by 45 percent, assumptions, and the two new desalination plants putting a significant strain on the population, but that were implemented with financing support from avoiding critical socioeconomic disruptions. During New Zealand, PUB water use is expected to reach its 2008–09 drought, Cyprus lowered household 64 lpcd by 2030. This is consistent with the water water allowances by up to 30 percent below normal consumption levels usually found in systems with consumption levels, and preferred to ship in fresh- one tap at each house. It is also estimated that the water from Turkey at the prohibitive cost of US$8 per population will continue to complement the PUB cubic meter rather than impose further restrictions. supply with rainwater and urban groundwater by 41.  South Tarawa’s population is likely to become more another 10–15 lpcd. It should be noted that the water reliant on an abundant water supply, and this will shape supply network is being designed with the assump- its vulnerability to water scarcity crises. The extent to tion that consumption from PUB could rise to which the population will accept a decrease in water 100 lpcd in the long term, which corresponds to the consumption during a prolonged crisis will be largely level typically supplied through multiple household influenced by its level of dependence upon an abun- taps. International benchmarking with other Pacific dant and reliable water supply once the new system is Island countries in a comparable Gross Domestic in place and has been operating for some time. The Product (GDP) per capita range suggests that resi- current level of PUB water service makes it difficult to dential water consumption from PUB in South infer with a high degree of certainty the level of water Tarawa will approach 100 lpcd in the long term, as use that households will adopt once safe water is sup- shown in figure 3.6. For this analysis, water con- plied continuously to homes in a reliable manner. sumption in 2030 (5 years after STWSP completion) South Tarawa’s water use trajectory is multifactorial, has been considered to reach 80 lpcd (see table 3.2), Building Urban Water Resilience in Small Island Countries 23 FIGURE 3.6. Residential Water Consumption from Water Utilities Compared to Countries’ GDPs 500 400 Residential consumption lpcd R2 = 0.36 300 200 100 0 100 1,000 10,000 100,000 GDP per capita (US$) Countries Paci c Islands utilities Trendline-Paci c Islands utilities Kiribati-PUB STWSP design Source: IBNet (accessed from www.ib-net.org on April 10, 2019). Note: GDP = gross domestic product; PUB = Public Utilities Board; STWSP = South Tarawa Water Supply Project. and the minimum acceptable level of water avail- water quality of the latter resource. Drought or SWOT ability 30 percent lower, at 56 lpcd. To satisfy any events that primarily affect traditional water sources water demand increase beyond 80 lpcd, an addi- (water reserves, urban groundwater, rainwater) would tional augmentation of PUB’s water production lead to a reduction of less than 30 percent of total capacity would be required after STWSP. water availability, an impact deemed acceptable on a temporary basis. Even two-event combinations 43. Once the PUB water system is modernized, the involving a drought, pollution, and/or SWOT, as most severe water crisis scenarios, in terms of decreased shown in figure 3.7, could almost be weathered due to water availability, will be those involving a temporary the contribution of seawater desalination and net- interruption of the desalination plant. As indicated in work improvements resulting from STWSP. The gaps box 3.3, the worst-case single-event scenario would would be limited to about 10 lpcd, which could be be a temporary interruption of the desalination sys- closed by using well water. tem due to technical breakdown, which would leave the South Tarawa population without adequate 44. In all crisis scenarios, potable water availability freshwater availability, and more critically so if would remain above 10 lpcd, which can be considered a ­ combined with pollution of the water reserve, SWOT, minimum acceptable level for human consumption or a drought. In such cases, gaps of 38–41 lpcd below (that is, for drinking and cooking) in an emergency. As acceptable thresholds would occur. It would not be illustrated in figure 3.7, in all except one scenario possible to absorb these gaps relying on RWH or well (pollution of the Bonriki water reserve, combined water, given the significant limitations in terms of with a breakdown of the Betio desalination plant), 24 Building Urban Water Resilience in Small Island Countries this threshold could be met by PUB water services supplemented by rainwater. Public messages recom- relying on the two other desalination plants and the mending the boiling or chlorination of rainwater Buota water reserve. In this worst-case scenario, PUB would then need to be reinforced, in order to ensure would be able to supply 8 lpcd, which could be that it would meet potability standards. FIGURE 3.7. Impact of Water Crisis Scenarios on Production Capacity, by Water Resource (in percent) and Water Availability (in lpcd), 2030 Single-event scenarios Combined-events scenarios Drought SWOT Events Pollution Tech. breakdown Betio desalination –100 –100 –100 –100 PUB-NZ desal. Impacts on production Bonriki lens –50 –54 –100 –100 –100 –50 –100 –54 –100 capacity of water Buota lens –100 –54 –100 –100 –100 –50 –100 –54 –100 resources Urban lenses Quality impact Quality… ...impact Rainwater –70 –70 –70 –70 Total water 90 available Normal water supply (lpcd) in 80 Water supply Total water available (lpcd) 2030 70 availability reduced 60 but still acceptable 50 Water supply availability 40 reduced below 30 acceptable levels 20 10 0 30 ht g n n T n n n n n tio w io ow tio ow ow O n ug pi 20 SW do t llu llu llu kd kd kd op o k Dr al Po Po Po + ea ea ea ea rt rm ht ve Br Br Br Br + + No ug ro t T + + + gh O o t T n e SW Dr gh at tio O ou SW aw ou Dr llu Dr Se Po Betio desal. plant Bonriki lens Rainwater harvesting PUB-NZ desal. plant Buota lens Urban lenses Source: World Bank data. Note: The contribution of urban lenses to water availability is provided for illustration purposes, because (1) water is usually brackish and does not comply with bacteriological standards; (2) the impact of events on water quality cannot be quantified; and (3) people’s future willingness to rely on water of various salinity levels for nonpotable use is not known. NZ = New Zealand; PUB = public utilities board; SWOT = seawater overtopping. Building Urban Water Resilience in Small Island Countries 25 12. Ratio between the internally generated surface water annual runoff Notes and groundwater recharge derived from precipitation falling within  1. WHO/UNICEF JMP 2017, https://washdata.org. A household has South Tarawa’s boundaries. access to basic water supply service when a water point is available and the collection time is no more than 30 minutes for a round trip, 13. Absolute water scarcity relates to areas whose freshwater availability including queuing time. is below 500 lpcd, according to the Water Stress Index (also known as the Falkenmark Indicator).   2. According to random weekly water testing carried out by the Ministry of Health and Medical Services across South Tarawa, 2017. 14. Seawater overtopping occurs when a storm surge combined with high winds, a high tide, large waves and sea level rise project signifi-   3. Current residential water tariffs are US$1.4 per cubic meter (m3) per cant volumes of seawater inland. month from 0 to 2.5 m ; US$4.2 per m per month up to 10 m ; and 3 3 3 US$12.6 per m beyond this. A review of this tariff is planned in the 3 15. Surveys providing information on water consumption from shal- short term, under STWSP. low wells and rainwater harvesting include: Ministry of Finance   4. The survey was carried out in June 2018; baseline information was and Economic Development of the Republic of Kiribati, 2016. collected from 300 households in South Tarawa. Population and Housing Census Report; White and Falkland. 2009. Tarawa Water Master Plan. Kiribati Adaptation Program II;   5. Calculation is based on a review of monthly precipitation from 1947- Jacobs, Scope Pacific Ltd. 2017. South Tarawa Desalination 2008, and a standard runoff coefficient of 0.85 of the reported water Economic and Social Feasibility Study; New Zealand Ministry of consumption data from our survey. Foreign Affairs and Trade; a survey of 292 households conducted   6. Raw water quality data communicated in 2017. as part of this study.   7. However, due to the stigmatization of well water consumption, and 16. Calculated based on chloride guidelines (250 mg/l). Recent editions fluctuations in access to safe water, the exact extent of the practice is of WHO guidelines only focus on potable water standards. likely to be higher. 17. With a 75 square meter (m2) roof, a 5 m3 tank, a 0.85 runoff coefficient,  8. WHO/UNICEF JMP 2017, https://washdata.org. A basic sanitation and 5 liters per capita per day. facility is defined as one that hygienically separates human excreta from human contact and is not shared between households, 18. Including IPCC’s RCP6, RCP8.5 scenarios and NOAA’s Intermediate High scenario.   9. 2015 census. 10. Based on a daily production of 130 g of DBO5 per pig, as proposed by 19. A pollution event in urban groundwater lenses has not been consid- E. P. Taiganides in Pig Waste Management and Recycling: The ered in this analysis: its scale would be limited to one of South Singapore Experience (1992). Tarawa’s islets, and it would not have systemic impacts on the city’s water supply security. 11. Under this method clear PET bottles are filled with the water and set out in the sun for six hours to kill viruses, bacteria, and parasites 20. This is driven by high power supply costs (US$0.5/kWh), com- (giardia and cryptosporidium). Source: www.sodis.ch. pounded by old electromechanical equipment. 26 Building Urban Water Resilience in Small Island Countries Chapter 4 How Can South Tarawa Improve Its Water Supply Resilience? 4.1.  Potential Measures for Reducing commonly show water consumption levels of 100 Water Stress during Supply Crises lpcd, sometimes even up to 150 lpcd. Communication 45.  Several approaches can be considered for improv- efforts should also promote safe water use at times of ing the resilience of the South Tarawa water supply in reduced water availability, considering the greater response to the various potential hazards presented in risk of water-borne disease, with fewer (frequently the previous section. This section provides an over- zero) hand washing sources available; and greater view of these approaches and assesses their potential ­ ower-quality water sources (Elliott et al. 2017) use of l contribution to improved resilience of the system. 47.  Demand management awareness campaigns should A first measure, critical to the resilience of public take into consideration and build on existing practices ­ utilities board (PUB) and South Tarawa’s water sec- that are consistent with the logic of water conservation tor, is the strengthening of the utility’s capacity and advocacy. It is critically important to encourage water sustainability. conservation by crafting outreach messages that approach water use from the perspective of local 4.1.1. Encouraging Water Conservation expertise and priorities: that give credence to direct 46. The I-Kiribati tradition of thrifty water use is a experience rather than the knowledge of experts in major water security asset that the authorities should assessing potability and quality of water; and that seek to encourage and sustain. The delivery of safe place prime importance on self-reliance and respon- and abundant freshwater through South Tarawa sibility in water management at the household level. Water Supply Project (STWSP) will require careful Widespread reuse of kitchen water for gardening and management of changes in people’s water use habits. pig rearing could be extended to reusing good quality The project will entail significant communication laundry and shower greywater to irrigate gardens; and awareness-building efforts concerning water the rationing of high-value water could also be sys- conservation. In addition, PUB and MISE have tematically extended to all three types of water demonstrated their willingness to enforce block tar- sources, including well water; and attempts could be iffs that will discourage wasteful water use, and to made to apply the logic of the monitoring of rainwa- allow for limited residual pressure at water distribu- ter tanks, which already informs water management tion points to help households better control their efforts in many households to assessing the quality of water use. Constant monitoring will be of critical domestic well water and understanding groundwater importance in order to ensure that PUB water use supply in relation to rainfall and demand. does not significantly exceed planned allocations (64 lpcd), with corrective tariff measures and/or 48.  Harnessing the competitive mindset of communi- communication campaigns ready for launching if ­ ties in Kiribati could also help drive progress in achiev- need be. As indicated in Section 3.5, countries ing water conservation goals. One possibly effective with a ­ similar level of socioeconomic development strategy for involving local communities may be Building Urban Water Resilience in Small Island Countries 27 through the awarding of prizes, gifts, and competi- permanent buildings with an impermeable roof sur- tions to stimulate citizens’ interest in water conser- face, and that they should collect water from the vation, and to make engagement in it entire roof area. Production capacity could theoreti- attractive–provided these are approved by local cally increase by almost 300 percent in residential authorities. In this spirit, community outreach mes- buildings, and by 23 percent in nonresidential build- sages should reflect a collective identity, focusing on ings, as shown in table 4.1. family and the household as a functional unit, and on 51. However, due to the high cost, maximizing RWH community rather than on individuals. production capacity is difficult. The capital cost of a 49.  Volumetric water tariffs will remain a highly effec- typical residential RWH system2 is around US$2,000. tive tool to promote water conservation and help sus- Over the estimated 20-year lifetime of the equip- tain thrifty water practices. Increasing the block tariff ment, the long-term average supply cost is US$8.1 per structure can be particularly useful for communicat- cubic meter (75 percent of which corresponds to the ing the value of water to users. In general, pricing sig- purchase of the tank). On large nonresidential build- nals such as tiered-rate structures seem to be more ings opportunities for economies of scale are limited efficient than conservation mandates (World Bank because both the water supply capacity and the opti- 2017). Setting tariff levels drastically higher than peo- mal storage capacity (which represent the bulk of the ple are willing to pay in the higher-tariff blocks can capital costs) increase almost linearly with the roof also help to effectively target the desired consump- surface area. The savings potential associated with tion levels. the purchase of a larger water tank remains limited by the current local market, which only offers indi- 4.1.2. Optimizing the Use of Rainwater Harvesting vidual storage capacity up to 10 cubic meters. For a 50.  The current stock of hard-roofed structures could, typical 500-square-meter communal hall, the long- in principle, allow a significant expansion of rainwater term average cost becomes US$6.1 per cubic meter. harvesting (RWH) as part of South Tarawa’s portfolio of In comparison, the PUB water tariff is on average water resources. This could be achieved by: US$1.4 per cubic meter for typical household con- (1) rehabilitating out-of-order systems (about 10 per- sumption levels. cent of the total stock according to field visits), and 52.  If RWH is to further expand, it will be driven by its ensuring that gutters on all of the sides of large build- perceived benefits in terms of resilience. The existing ings are connected (these systems capture on average only 75 percent of roof areas as found in our survey); TABLE 4.1. Rainwater Harvesting: Current and (2) expanding coverage from the current 20 percent Potential Increase Capacity to all houses with suitable roofs, which represent Current Maximum potential about 70 percent of houses in South Tarawa; and (3) production production capacity expanding RWH coverage for nonresidential build- capacity (lpcd) (lpcd) ings, from the current level of about 50 percent to up Residential buildings 1.9 7.4 to 90 percent of buildings.1 This would be consistent Nonresidential 2.3 2.8 buildings with the provisions of the 2017 Kiribati National Total 4.2 10.3 Building Code, which stipulates that rainwater stor- Source: World Bank data. age and roof drainage should be provided on all Note: Average liters per capita per day across the population of South Tarawa. 28 Building Urban Water Resilience in Small Island Countries PHOTO 4.1. Private Rainwater Tank and Laundry Area Source: © Helene Le Deunff/World Bank. rainwater storage infrastructure could contribute to the yield from the water reserve lenses, the public communities’ resilience to water crises, for example could be advised to maintain storage tanks close to the combination of a temporary breakdown in the maximum water levels by filling them frequently desalination system with a reduction in the yield with PUB water. However, there are three main chal- from water reserves lenses due to drought, seawater lenges in operationalizing such a system: (1) ensur- overtopping (SWOT), or pollution events. In total, ing access for everyone to the tanks in a time of rainwater tanks across South Tarawa currently rep- crisis (currently 37 percent of the population never resent an estimated storage capacity of 15,000 cubic uses RWH, which might in some cases be due to meters, with about 11,000 cubic meters (equivalent affordability or accessibility (that is, distance) to about 200 liters per capita) available for residen- issues); (2) ensuring adequate water quality, since tial water consumption. This could provide an RWH systems rarely comply with bacteriological ­average of 8 lpcd for 1 month. While technical break- quality standards, and are almost never chlorinated; downs cannot be predicted, droughts or pollution and (3) encouraging the filling of water tanks in a incidents usually develop gradually, and can be staged manner that does not overwhelm PUB water monitored. In the case of a significant decrease in production capacity. Building Urban Water Resilience in Small Island Countries 29 53.  Improving the equity of water security requires a chlorination is rare (ADB 2018). To aim for universal denser network of storage facilities, as well as incen- treatment of rainwater, it is recommended that the tives to owners to contribute to the resilience of the public awareness campaigns on safe water supply system. A full mapping of water tanks across South and hygiene planned under STWSP will promote the Tarawa could help identify the areas where inhabi- 3 systematic simple disinfection methods of tanks, and tants do not have access to at least 200 liters of stor- the boiling of rainwater before human consumption; age capacity within a reasonable distance (up to 200 incorporate messages about the maintenance and meters). Expansion of storage capacity should priori- suggest cleaning proper use of RWH systems; and ­ tize those areas. Along with this, continuous aware- methods that both respect and protect the widely ness campaigns will be needed in order to ensure that perceived purity and value of rainwater. everyone has access to water from tanks installed on 55.  Economic benefits related to the hedonic value of private property, particularly during droughts. rainwater, along with the desire to retain cultural iden- Currently water sharing is determined by proximity tity, may contribute to sustained reliance on this of relationship; close relatives have priority. Because resource. The value that people place on the organo- of recent urbanization, people who are not related to leptic qualities of rainwater is likely to evolve with each other or who do not even originate from the the improvement of the PUB water supply, which same island are now living side by side, and may not combines water from the Bonriki and Buota lenses be inclined to participate in systematic water sharing. with desalinated water. The factors that can influ- The marginal cost of connecting tanks to residential ence willingness to change behavior are poorly RWH systems is estimated at US$1.3 per cubic meter understood, but ethnographic studies suggest that (less than US$1.0 per cubic meter for large institu- the I-Kiribati are reluctant to conform to anything tional buildings). This additional cost is financially that is perceived as not local (Brewis 1991). Thus, justified in relation to the current PUB water tariff of integrating the cultural values that local people US$1.4 per cubic meter (in the lower tariff block). To attach to the water resources they control and use incentivize the filling of storage tanks with PUB water into the efforts to adapt to the stresses on these water in situations of enhanced water lens vulnerability, resources is of critical importance (Kuruppu 2009). temporary rebates on the higher PUB water tariff The cultural processes negotiated by people in their blocks could be offered to institutional rainwater daily lives in Kiribati have been shown to reinforce tank owners. and reproduce water hardships, as illustrated in 54.  Efforts to enhance public awareness about the safe box 4.1. The sense of vulnerability associated with management of RWH and storage will be required. Two environmental change and urbanization might also main factors contribute to substandard water quality produce an urge to preserve cultural values (Jones from RWH systems: (1) the insufficient maintenance 1996). As culture is so fundamental to I-Kiribati and cleaning of roofs and storage tanks; and (2) the behaviors and outlook on life, the perception of PUB lack of first-flush and disinfection systems. Even if water as being “modern” might prevent its full adop- tanks are used to store high-quality PUB water, health tion even when it is abundant, accessible, safe, and of hazards will remain due to the possible contamina- good quality: an enduring attachment to rainwater as tion of water tanks. Currently, only 58 percent of a culturally valued resource may be maintained. households boil water prior to consumption, and This must be considered if the I-Kiribati are to be 30 Building Urban Water Resilience in Small Island Countries BOX 4.1. When Cultural Norms and Requirements Reinforce Water Hardships Water sharing has benefits, but it also entails at a high psychological cost. Mr. A., for instance, additional burdens. When members of the same who is 30 years old, moved from North Tarawa so kin group live far apart, the distances to collective he could live with his young small family close to water distribution points can be a strain. Mr. R., a the house of relatives in Bikenibeu. He reports that man in his 50s living in Ambo, has to beg for one depending on others for water creates emotional bucket of rainwater a day from relatives. To collect distress: “I am so tired of asking for rainwater from it he must also walk for 10 minutes to the other the neighbors!” he says. side of the islet. Exchanging water can also come Source: Face-to-face interview, July 2018. encouraged to fill their tanks with PUB water as a pre- scenario for urban groundwater). This schematic ventative tactic for crisis management during periods water balance is presented in the absence of detailed of low rainfall, as they may be reluctant to mix the knowledge, adequate monitoring, or accurate model- two sources. ing of the flows of freshwater in the lenses. Given the current tree cover in urban areas, evapotranspiration Improving Urban Groundwater Through 4.1.3.  is considered to be 65 percent of precipitation.4 Managed Recharge Greywater returned to the lens is, in the absence of a 56.  Even with improved PUB services, urban ground- dedicated sewerage system,5 assumed to represent water, like rainwater, will remain a last- resort water 80 percent of total consumption, and leaks are pro- resource anchored in the I-Kiribati’s deep cultural rela- jected to be a maximum of 25 percent of PUB water tionship to the land. The improved management of production in 2030. urban groundwater, therefore, has a multilayered 58.  Improvements in the availability and efficiency of impact on water security, at both a material and a PUB water supply services may have significant, posi- symbolic level. tive impacts on the recharge of urban lenses. As shown 57. Human activities significantly influence the avail- in figure 4.2, which shows the inflows and outflows ability and quality of urban groundwater, particularly at the surface of a lens in a dense urban area like Betio during droughts. Along with rainwater infiltration, during a severe drought, the net inflow due to natural leaks from PUB networks and greywater discharge factors (+ 11 lpcd) increased threefold in 2019 (up to + contribute to the recharge of urban lenses. On the 33 lpcd) thanks to anthropic factors, and would be other side of the equation, freshwater is drawn out of expected to increase almost six-fold by 2030 (up to lenses through well water abstraction, evapotranspi- 64 lpcd). The increased contribution of greywater ration, outflow, and dispersion at the bottom and discharge is of particular importance to future water sides of the lenses. The respective contributions of balance, and it is made possible by a massive transfer these different factors are illustrated in figure 4.1, of desalinated seawater without which the net assuming conditions of severe drought lead to a freshwater inflow to the lenses would be almost nil. ­ 75 percent decrease of the 2-year cumulative rainfall This analysis highlights the potential benefits of (which is identified in Section 3.4.1 as the worst-case managing greywater discharge in situ, rather than Building Urban Water Resilience in Small Island Countries 31 FIGURE 4.1. Key Contributing Factors to the Water Balance of Urban Lenses Precipitation: + 197 lpcd Evapo- transpiration: Well water –128 lpcd Greywater abstraction + 12 lpcd PUB discharge: leakages + 62 lpcd + 10 lpcd Freshwater zone Transition zone Out ow, dispersion Source: World Bank data. Note: Average flows across South Tarawa, drought conditions, 2030. PUB = public utilities board. FIGURE 4.2. Balance of Freshwater Inflow and Outflow by Urban Lens Surface Natural factors Anthropic factors - 2019 Anthropic factors - 2030 Net in ow: 64 Ipcd 70 +10 60 Net in ow: +54 Liters per captia per day 50 33 Ipcd 40 +34 +15 30 20 +28 10 –22 0 –12 –10 –22 –20 Rain EvapoT Well abs. Greywater PUB Well abs. Greywater PUB in ltr. disc. leakages disc. leakages Source: World Bank data. Note: Drought conditions in Betio, 2019 (left and middle sections) and 2030 (left and right sections). Anthropic factors (related to human activity) include well water abstraction, graywater discharge, and PUB network leaks. discharging it into the sea (treated or untreated), population), greywater flow is comparable to the nat- mixed with toilet flushing and other high-organic ural recharge that occurs when precipitation is close content wastewater. to the average level, as illustrated in figure 4.3. While freshwater yields would remain very limited (up to 59. The recharging of urban lenses with greywater 2 lpcd in Betio, 5 lpcd in Bairiki, 2 lpcd in Bikenibeu, during severe droughts can maintain hydrogeological and probably similar, or even less, elsewhere), and conditions comparable to a normal precipitation would only be available toward the centers of the regime. In communities with a high population den- islets, the overall quality of groundwater across the sity (from about 20,000 people per square meter, top of the lens would be significantly improved which represents about 80 percent of South Tarawa’s (White 2010). For most South Tarawa inhabitants, the 32 Building Urban Water Resilience in Small Island Countries FIGURE 4.3. Potential Relative Contribution of Greywater 120 Nanikai 100 Average Banraeaba precipitation Betio regime Teaoraereke 80 Taborio Eita Bairiki Abarao Bikenibeu Percent 60 Antebuka Ambo 40 Bonriki Tanaea Severe drought: Temaiku cumulative precipitation at 25% of average 20 0 10,000 20,000 30,000 40,000 50,000 Population density (people per square kilometer) Hydrogeological conditions with recharge with greywater during severe drought Hydrogeological conditions without recharge with greywater during severe drought Source: World Bank data. Note: Potential relative contribution to total lenses of freshwater inflows, depending on population density—in case of a severe drought in 2030. recharge of lenses with increased graywater volumes the ground. People sometimes dig holes to bury could therefore translate into reduced salinity in oth- smelly liquid substances, or bring them to the beach erwise brackish groundwater. for discharge into the sea. These practices have the potential of becoming aggravating nuisances once 60. As currently managed, greywater could pose supply volumes significantly increase. The contami- increasing challenges to the quality of the environ- nation of the lenses can also be expected to increase ment, and possibly of the groundwater, in the long due to the pollutants they will carry when improved term. Currently greywater is discharged without water availability triggers new consumption behav- treatment by households directly into their plots, iors, leading to an increase in the bacteria, deter- where it infiltrates quickly due to the high absorption gents, and nutrients that were already found in capacity of South Tarawa’s sandy ground. Water from excessive quantities in some areas of the lenses in the kitchen (used for cooking and dishwashing) is 2010 (White 2010). Their discharge into the ocean generally shared with pigs when it is free from soap through deep submarine outfalls (as is currently or detergents; otherwise it is used for gardening, or to being done, together with the rest of household water plants. Water used for hygiene (shower, toilet- sewage in Betio, Bairiki, and Bikenibeu) would ­ ing), and water from the laundry are left to infiltrate avoid these problems. However, it would miss the Building Urban Water Resilience in Small Island Countries 33 opportunity to use these low-pollution, low-salinity figure 4.4. Such solutions would be better suited to waters for recharging urban lenses, and for increas- areas with moderate population density. Where ing the resilience of communities to droughts; and it houses are traditional and lack plumbing systems, could also potentially negatively affect coastal simple concrete collection basins (such as the one ecosystems. shown in photo 4.1) would be required to receive washing and laundry greywater and transfer it 61.  Simple, inexpensive solutions could help improve through underground piping to the treatment sys- greywater management, and well water salinity and tems. The design of these systems would also take bacteriological quality. Reducing water salinity in into consideration the capacity of local coral sands shallow wells requires the disposal of greywater in to effectively remove pathogens within a very short proximity to them. But water recharged from point distance from the discharge point (Institute of sources usually stays relatively close to the point of Environmental Science and Research Limited 2016). discharge, as has been found from pollution studies at and near point sources in other atolls (Falkland 62. Several measures are of critical importance in 2001). Current greywater disposal practices, which order to better regulate water use from urban lenses, ignore this insight, would be unlikely to notably and to reduce the impacts of greywater discharge. The improve the quality of the water in shallow house- establishment of quality guidelines for well water, hold wells. Given the nature of the pollutants con- and a monitoring system could help clarify the tained in greywater, simple facilities such as grease advisable nonpotable water use conditions; define traps and leaching pits could be considered for targets for the management of urban lenses; and treating centralized household greywater, and for rationalize recommendations for the population. ensuring its dispersion into the ground, as shown in Salinity limits could build on the 1993 edition of the FIGURE 4.4. Schematic Lens Recharge System (Greywater Disposal+Rainwater) Rainwater tank Collection basin Well Gravel filter Grease trap and Leaching pit Source: World Bank team. 34 Building Urban Water Resilience in Small Island Countries WHO guidelines (2,500 μS/cm), while guidelines for nearby latrines or septic tanks. Piping rainwater over- bacteriological quality could be based on the latest flows through a gravel filter adjacent to the well WHO guidelines for safe recreational water environ- appears advisable in order to ensure that leaves and ments. Limiting the use of soap and detergents to other debris on roofs and in gutters are not flushed those that are free of inorganic phosphates, and that directly into the well (Falkland 2001). Pilot testing in contain only surfactants that minimize environ- South Tarawa is recommended before rollout. This mental pollution would contribute significantly to should include long-term monitoring of the impact the protection of urban groundwater quality. on water quality in wells with and without such an A national regulation restricting or disincentivizing installation, to confirm the concepts of both rainwa- the purchase or importation of detergents that do ter and greywater recharge, and to quantify their not comply with such requirements is recom- benefits. mended from a water security perspective. Also, the inclusion in the Kiribati Building Code of require- 4.1.4. Direct Nonpotable Water Reuse ments to direct RWH system overflows to the wells 64.  In low-density areas without a seawater network could be considered, along with the measure pre- for toilet flushing, direct reuse of greywater could help sented in the next paragraph. Finally, investigations save PUB water and/or minimize the burden of fetching to identify leaks from the seawater flushing net- well water. According to the household survey that work, and their impact on urban lenses is critical in was conducted as part of this study, 60 percent of order to avoid groundwater salinity hotspots. households in Bairiki, Betio, and Bikenibeu, and 63. RWH can also help maximize the freshwater probably a smaller percentage in other areas of South recharge of shallow wells, and reduce salinity. As was Tarawa, have access to private flush toilets—a num- successfully shown in the outer islands of the ber that is likely to increase in the coming decade. Maldives, directing rainwater into or near shallow Washing areas for showering and bathing are fre- wells could help improve their salinity (Falkland quently located close to toilets, a configuration that 2001). Together with greywater recharge near the lends itself to the reuse of greywater for toilet flush- well, the injection of rainwater would lead to the ing, as illustrated in figure 4.5. A small ground tank development of “domes” of groundwater of reduced (less than one cubic meter) could be installed to salinity around the well. Such systems would target receive greywater from the washing area and could overflows from RWH systems, which represent close feed into the toilets through a manual pump. With an to 90 percent of rainwater collected in average pre- expected water use of about 20 lpcd (comparable to cipitation years, and more than 70 percent even average household washing needs), such a system during severe drought periods. Where no RWH sys- would save almost 5 cubic meters from the monthly tem is installed, this approach can still be used by PUB water bill (which could amount to US$5–10 per both households and community buildings at a rela- month, considering the post-STWSP PUB water sys- tively low cost by installing gutters and piping to tem operating costs), or the lifting of 5 tons of water nearby wells. In addition to reducing the salinity of from wells, which most households carry out manu- well water, this method can improve the bacteriolog- ally. This conservation of water would be particularly ical quality of the water by increasing the water level critical during severe droughts, where any reduction locally, and slowing the migration of pollutants from in well water abstraction can weigh significantly on Building Urban Water Resilience in Small Island Countries 35 FIGURE 4.5. Options for Greywater Reuse Management in Low-Density Areas Kitchen water Laundry Washing Toilet toileting Over ow Pigs Gardening Well Leaching pit Onsite sanitation PUB water Rainwater Well water 2 lpcd 5 lpcd 10 lpcd 20 lpcd Source: World Bank data. Note: Left panel: without direct reuse; Right panel: with direct reuse in a shower-to-toilet configuration. the lenses’ freshwater balance, as was shown in reuse area. A rapid review of a typical household ­ figure 4.2 above. layout suggests a difference between the modern type of housing—where the places that water is con- 65.  Similarly, direct reuse of nonpotable water could sumed are close to each other in a single building— be considered in lieu of seawater flushing in high-­ and more traditional settings, where water can be density areas where sewerage system expansion is accessed and used in different parts of an open com- envisaged. (Such expansion is considered a ­ priority pound. While some households have been reusing in the areas of Betio, Bairiki, and Bikenibeu that are greywater for gardening, as illustrated in box 4.2, not currently covered by the sewerage network.) such use should not be encouraged without appro- In-house direct reuse could enable major savings, priate prior treatment. considering the cost of seawater flushing systems (up to US$800 per beneficiary) (GHD 2017); operat- 66.  Across South Tarawa, up to 40 percent of house- ing (pumping); and maintenance costs, not to men- holds could benefit from the implementation of a direct tion the risks of groundwater salinization due to reuse scheme for nonpotable water. Such measures possible leaks in the seawater network. Depending could lead to an overall incremental water use of on the house’s configuration, greywater from the 8 lpcd from sources normally used for toilet flushing kitchen or laundry could be used for that purpose as outside of the areas that are already served by a sew- well; and in more modern houses dedicated plumb- erage system. Brochures designed to raise awareness ing could ensure that greywater is transferred to the of and describe possible reuse arrangements for a 36 Building Urban Water Resilience in Small Island Countries BOX 4.2. Improvised Graywater Reuse in Kiribati Households Mr. N. is a 68-year-old microentrepreneur rainwater in buckets from his metal roof. Mr. N. who retired a few years ago from his job in the is very proud of the ingenious plumbing system shipbuilding industry. The house he shares with he has designed to channel greywater from the his wife is conveniently located on the main kitchen sink to the banana pit behind the house. road. This enables him to run a small kiosk The two young banana trees, mulched with where he sells fresh fruits and vegetables from organic material and fed with kitchen water, his garden and a mixed basket of dry products. have already produced fruits that he has sold to With his protected well and a connection to the his customers. The water he uses to boil fish and water supply system, Mr. N. has access to two to clean rice is also carefully set aside every day of the three types of water usually accessed in to feed his pigs. South Tarawa. When it rains, he also collects Source: Face-to-face interview, July 2018. variety of household layouts, and possibly a dedi- 4.1.5. Increasing PUB Emergency Storage Capacity cated revolving fund, could help spur the rollout of 67.  Expanded storage of treated water could help to this solution. Photo 4.2 shows an example of a bath- effectively address short-term emergencies related to room setup that is well suited to direct sink-to-toilet technological breakdowns. Available space is the main reuse. According to our household survey, well water constraint in the western part of South Tarawa, where is being used by 90 percent of households for toilet the main desalination plant will be located. The flushing, and PUB water is used in the remaining 10 implementation of aboveground water tanks could percent of cases. It is likely that the use of PUB water be considered for water reserves that have a minimal for toilet flushing will increase after the PUB system impact on aquifer recharge. Any location near the is upgraded. Moreover, the actual scope of greywater Bonriki headworks would seem particularly appro- reuse will largely depend on the solutions imple- priate for taking advantage of existing chlorination mented to improve sanitation across South Tarawa. facilities. Sizing would be limited by investment costs As indicated in Section 3.2.1, a concept study is cur- (about US$200–300 per cubic meter), and ensuring rently being carried out with financial support from acceptable water quality (subject to simple rechlori- New Zealand for the purpose of identifying appropri- nation) in the tanks could also be a challenge in the ate sanitation models and preparing for their imple- case of extended storage duration. Also, the water mentation. Solutions that require minimal freshwater transmission system in the eastern part of South input (for example, dry sanitation, vacuum sewers, Tarawa has an insufficient conveyance capacity for and seawater flushing) would need to be compared pumping well beyond the normal production levels with opportunities for greywater reuse in toilets. The in Bonriki and Buota. An upgrade of the pumping and rollout of this solution could, for example, be transmission system from Bonriki may be required in financed by a revolving fund, through a scheme sim- order to allow the simultaneous operation of these ilar to the one that was used in the past for the expan- systems during emergencies at the Betio desalination sion of RWH. plant. Building Urban Water Resilience in Small Island Countries 37 PHOTO 4.2. Favorable Setup for Direct Reuse Source: © Helene Le Deunff/World Bank. 68.  The economic relevance of this solution, compared operating the desalination plant will help in better to other previously discussed solutions, strongly appreciating the nature and frequency of potential depends on the type and expected duration of a water breakdowns and reaction times, and will facilitate crisis. As shown in figure 4.6, the cost-effectiveness of the prioritization of solutions. water storage would be more economical for small-storage capacity (5,000 cubic meters or less), 4.1.6. Increasing Desalination Capacity and for emergencies that could be addressed within a 69. Implementing the new seawater desalination few days, than low-cost solutions such as pollution system will significantly improve the resilience and ­ mitigation measures, or clearing vegetation or palm operation of PUB water systems. The solar-powered trees from the groundwater reserves (US$0.10 and desalination system planned under STWSP is US$0.13 million per additional lpcd respectively). For designed to meet a water demand of 59 lpcd until long-term crises of 2 weeks or more, the solution 2030, with a 4,000 cubic meter per day production appears less cost-effective than a backup desalina- capacity. The plant will be located in Betio, at the tion system would be (US$0.3–0.5 million per western end of South Tarawa, and its water transmis- ­ incremental lpcd across South Tarawa). Experience sion and brine disposal systems are designed to allow 38 Building Urban Water Resilience in Small Island Countries FIGURE 4.6. Cost-Effectiveness of an Emergency Storage Solution for Various Storage Sizes, Durations, and Intensities (lpcd gap) of Emergencies 45 40 35 US$0.9/lpcd Cost-e ectiveness Days of emergency supply 30 25 US$0.6/lpcd 20 15 US$0.3/lpcd 10 US$0.1/lpcd 5 0 5 10 15 20 lpcd delivered Storage capacity: 20,000 m3 15,000 m3 10,000 m3 5,000 m3 Source: World Bank data. future expansion of production capacity to 6,000 70. Investment costs for the engineering, procure- cubic meters per day. While the marginal cost of ment, and construction of desalination plants of this increasing desalination capacity on this site would be capacity can be expected to be in the range of much less than the construction of a new unit, the US$2,000–3,000 per cubic meter per day of production introduction of a new desalination system in a differ- ­capacity. These cost ratios are at the higher end of ent location would help reduce the systemic impacts normal prices for reverse osmosis desalination plants of a local failure or disaster affecting one plant. Even (which start at US$500 per cubic meter per day), due though in both cases the source of water remains to the small scale of the units and the remoteness of identical (seawater), the use of independent desali- Kiribati. The architecture of PUB’s distribution sys- nation systems represents a diversification of PUB’s tem would in principle allow for the installation of an portfolio mix, and the hedging of water supply additional desalination unit across South Tarawa’s risks. A new 3,300 cubic-meter-per-day desalination transmission system. However, a review of the backup system could, in principle, fully address the Bonriki pumping system would be needed in order to freshwater supply gaps described in figure 4.7 for all ensure its capacity to operate in this altered hydrau- crisis configurations. The introduction of other, more lic regime, without the Betio desalination plant and effective measures could, however, mitigate the cost-­ with the backup production system. As shown in need for such a large additional desalination capac- figure 4.8, from a hydraulics perspective it would be ­ ity, and should be considered in priority. preferable for such a system to be located in the Building Urban Water Resilience in Small Island Countries 39 FIGURE 4.7. Desalination Backup Capacity Required to Cover Crisis Supply Gap 90 Normal water supply 80 70 Total water available (lpcd) Water availability 60 acceptability limit 1,000 50 1,600 m3/day m3/day 3,300 3,000 3,000 40 m3/day m3/day m3/day 30 20 10 0 t n n 30 L in r O + tio + ow + tio + ow + ow + gh tio ow pp te 20 A SW ht llu ht kd ht llu OT kd T kd on RM g T n n n n n ou to a ea WO llu kd er eaw g Po oug ea ug ea ti Po SW ou Dr Br llu NO Po ea Br ro S Dr Dr S Po Br D Br ov Single-event cases Multiple-event cases Source: World Bank data. Note: Production capacity needed to fill the supply gap in various scenarios (well water is excluded from the portfolio mix). SWOT = seawater overtopping. FIGURE 4.8. Preferred Areas for a Backup Desalination System 1,500 m3/day 1,000 m3/day 1,800 m3/day 600 m3/day 400 m3/day Buota 300 m3/day 200 m3/day Bonriki 4,000 m3/day Preferable area for the location of the back-up plant Betio Approximative boundary of Betio desalination plant’s 500 m3/day zone of in uence 2 km 2 km Water demand Water production Transmission main Source: World Bank data. 40 Building Urban Water Resilience in Small Island Countries western part of South Tarawa (subject to land avail- These plans can include a range of measures, such as: ability), in order to minimize the need for upgrades of (1) clearly defining permitted activities and land use; the transmission system. (2) clearly defining prohibited activities and land use, and the regulatory measures that will be enforced in 71.  Keeping a seawater desalination plant in standby cases of infringement; (3) creating public aware- mode to cope with emergencies requires regular and ness-raising and behavior change campaigns regard- careful maintenance. Even if a large new desalination ing the impact of human activities on groundwater unit were built within the PUB water system to cope quality, along with warning signs that describe the with emergencies, the Bonriki and Buota freshwater penalties for noncompliance; (4) establishing an lenses would probably continue to be relied upon in active surveillance and compliance regime, with the the first instance, given their lower associated oper- equipment and technology needed to regulate and ating costs.6 To ensure that the plant was in good prevent further encroachment on the reserves; working condition and capable of being put into (5) filling in the existing pits through which the lenses operation immediately, if for example there were a are most vulnerable to contamination; (6) imple- failure of the Betio desalination system, it would menting small-scale infrastructure improvement need to be operated in “hot standby” mode: this sug- (sanitation, drainage); (7) creating low-cost perimeter gests that it should be briefly run on a frequent basis. barriers, and erecting educational and signs around Such a strategy was adopted during the Millennium the water reserves; and (8) enhancing groundwater Drought by the government of Queensland, Australia, monitoring so that it can detect early signs of ground- with the use of a standby desalination plant in the water pollution. Even if settlers were relocated away Gold Coast to cope with acute drought events or the from the reserves, some of these actions would failure of other water production systems.7 Together remain critical in order to avoid any further inflow of with investment costs, long-term operation and people. Further details are provided in appendix D. maintenance could represent up to US$0.5 million per additional lpcd of water availability across 73. Industrial pollution from the airport must be pre- South Tarawa. vented, and remediation measures prepared in case of an accident. In addition to observing normal airport 4.1.7.  Managing Water Reserves in a Sustainable infrastructure design and operation standards, it is Manner critical to adopt prevention measures, including the 72.  Diffuse groundwater pollution generated by human installation of observation wells around the most settlers in water reserves can be mitigated. Organic exposed areas (where fuel is stored and handled), and nitrate pollution appears to be the most immedi- and monitoring these wells in order to allow for the ate threat related to the presence of settlers in the timely detection of possible leaks of oil products water reserves. The best way to preserve the yield into the lens. If contamination can be identified from these reserves is to control urbanization over before it spreads out in the lens, affected soil can be them, and to properly manage the activities that take removed and abstraction of polluted groundwater place in their proximity. However, pollution risks to can be carried out from remediation boreholes dug the water lenses can also be minimized through around the contaminated site (Svoma and Houzim ­ sustainable water management plans designed to 1984). In such a case, only the nearby infiltration mitigate risks and prevent further encroachment. ­ gallery(s) would have to be temporarily interrupted, Building Urban Water Resilience in Small Island Countries 41 and the impact on PUB water production would be 76. The transfer to South Tarawa of groundwater limited. abstracted on North Tarawa’s main islets could also be considered in order to diversify water resources. The 74. The costs associated with these various measures Tarawa Water Master Plan (White 2010) and later the are particularly difficult to estimate at this stage. Water and Sanitation Roadmap 2011–30 (ADB 2011) Significant technical assistance will be required for discussed the development of groundwater produc- those elements of the sustainable management tion systems in several islets of North Tarawa, and plan that are associated with mitigating house- cross-lagoon water transfers to South Tarawa. Aside hold-borne pollution, or the preparation and man- from the complexity and cost of securing abstraction agement of resettlement processes. The main rights from North Tarawa landowners, this solution infrastructure costs are related to the fencing of the was found to be very onerous, with a unit cost of reserves, if this is deemed necessary. The cost of a about US$11.5 per cubic meter delivered. sanitation infrastructure would probably be negli- gible. The main costs associated with mitigating 77. While this report promotes the consideration of industrial pollution would be the drilling of bores, urban aquifer recharge with greywater, future densifi- and equipment for monitoring them. An overall cation and modernization of South Tarawa could in the cost estimate of US$2 million over the next 20 years long term require adjustments to this strategy. Changes is assumed for the various measures considered in of lifestyle, and an increase in pollution loads could this analysis. at some point overwhelm local greywater treatment capacity. Alternative greywater management strate- 4.1.8. Other Supply Augmentation Measures gies would then need to be considered, including 75.  Two measures described in previous studies could centralized treatment, provided that flushing sys- increase the yield of the Bonriki water lens. The tems do not rely on seawater, which would under- removal of deep-rooted palm trees from the central mine conventional treatment solutions. A wastewater portion of the water reserve to reduce transpiration reclamation system would eliminate the need to dis- losses is described in the Tarawa Water Master Plan charge the effluent into the ocean, and, together with as allowing an increase of 250 cubic meters per day of a nonpotable water supply network, it could, in a cir- the lens’s yield, at a cost of about US$325,000 (White cular manner, supply households with the recycled 2010). The infill of borrow pits located on the edge of water for flushing and other nonconsumptive uses. the reserve that are filled with brackish water could In the context of South Tarawa, this approach would, also increase the yield by 250 cubic meters at a cost of however, face several constraints: (1) the cost of around US$3.0 million. Both approaches would, building a second pressurized water network and however, require negotiations with landowners, and doubling PUB’s O&M and commercial efforts to man- might also require the payment of compensations to age both services (potable and nonpotable); (2) the those who are using the trees or the pond for eco- city’s elongated shape, which stretches for 30 kilo- nomic activities. From a water security perspective, meters, would complicate the centralization of these measures would be most beneficial during wastewater at a single reclamation plant; and (3) the water crisis events that do not affect the water introduction of another technological solution, reserves (for example, a temporary breakdown of the with the associated risks of systemic failure. On the desalination system). other hand, the energy requirements for wastewater 42 Building Urban Water Resilience in Small Island Countries reclamation are generally lower than for seawater All the measures listed below can be programed inde- desalination. Should technology improvements in pendently and their benefits added up cumulatively. wastewater reclamation make it financially competi- The exception is tank crisis management policy tive with the alternatives described in this report, an (Measure 2), the specific benefits of which would be initial pilot in Betio could be considered in the long increased if the RWH infrastructure were expanded term, since: (1) population density in this area is likely (Measure 3), with a combination of these two mea- to make decentralized greywater management more sures yielding a greater benefit than the sum of each difficult; and (2) a sewerage system is already in place individual measure. Figure 4.9 shows the combined there, centralizing wastewater in a location close to effect of these measures on water supply availability the future Betio desalination plant. in the various crises scenarios, confirming that water availability thresholds can be met even in the most 78. The recharge of the Bonriki water lens has been dire scenarios. analyzed as part of this study, but the technical and economic feasibility of this remains uncertain. Artificial 80.  During a water shortage crisis, the relevance of cer- recharge of the Bonriki water lens could be tain measures is strongly conditioned by the temporal particularly helpful in speeding up recovery of its ­ characteristics of the events at play. Certain events sustainable yield after severe droughts or SWOT (droughts, pollution) can be monitored as they unfold events. However, such a solution would only be sen- and can trigger enhanced preparedness, but their out- sible if water production in Bonriki has been fully comes can be impossible to predict. The breakdown of interrupted: reducing abstraction from the lens the desalination system cannot be anticipated, but we would otherwise be preferable, in order to improve can assume that its resolution would most likely take hydrogeological conditions. If the Bonriki water its ­ place within a few days (or weeks at worse), making production system is stopped, it is unlikely that water water storage solutions relevant. Water conservation production from other plants would cover all of efforts that minimize people’s reliance on abundant South Tarawa’s needs and leave any surplus that water are particularly useful for helping people cope could be used for water recharge. with prolonged crises. During short emergencies, altered water availability would likely be perceived as 4.1.9. Assessing the Proposed Resilience Measures exceptional, time-bound, and more or less acceptable 79. The resilience measures presented so far in regardless of usual water reuse. Table 4.2 summarizes Section 4.1 cover a broad spectrum of cost-­ effectiveness and types of benefits. Table 4.3 provide an overview TABLE 4.2. Temporal Characteristics of Water for each of these measures, of the potential incre- Crisis Triggers mental water availability; its quality; the situation in Time before the which it could be mobilized; and the long-term unit Event Onset recovery of water resources cost per cubic meter. Costs were found to vary Droughts Slow (months) Long (months to years) between US$0.02 million per incremental lpcd Seawater overtopping Sudden Long (months to years) delivered across South Tarawa for improved tank ­ Pollution Slow (years) Long (months to years) management (which relies primarily on policy deci- Desalination breakdown Sudden Short (days to weeks) sions and awareness raising), to US$1–3 million per Note: The onset of pollution events can be sudden with, for example, incremental lpcd for decentralized RWH schemes. accidental fuel spillage, but this would be unlikely to affect the entire lens. Building Urban Water Resilience in Small Island Countries 43 FIGURE 4.9. Improvement in Water Supply in Crisis Scenarios due to Proposed Measures 90 Normal water supply 80 Total water available (lpcd) 70 Water availability 60 acceptability limit 50 40 30 20 10 0 t n n 30 L in r O + tio + ow + tio + ow + ow + gh tio ow pp te 20 A SW ht llu ht kd ht llu OT kd T kd on RM g T n n n n n ou to wa ea WO llu kd g Po oug ea ug ea ti Po SW ou Dr a Br ollu NO Po ea Br o S Se Dr Dr Dr Br P er Br ov Backup desalination Greywater direct reuse Urban groundwater before measures Urban groundwater with managed recharge Tanks crisis management Total freshwater before measures Source: World Bank data. Note: The contribution of urban groundwater to water availability, and the split between urban groundwater and backup desalination are illustrative only and cannot be ascertained at this stage. the temporal characteristics of the four main water acceptable level of quality would make this mea- crisis triggers. sure more cost effective than the following ones. 81.  A clear prioritization of measures emerges from a •• If the above measures are implemented to their comparison of their cost-effectiveness. It follows the fullest potential, two options could be considered following principles: to bridge the remaining water deficit: the imple- mentation of bulk PUB water storage in Bonriki; or •• Water conservation measures, tank crisis manage- the construction of a backup desalination unit. If ment, direct greywater reuse, clearing of deep- desalination system breakdowns are expected to rooted palms in water reserves, and pollution last less than 2 weeks, then bulk water storage mitigation can significantly augment (or prevent would be technically feasible and sufficient to the loss of) water supplies at a very limited cost. weather the crisis, and the more cost-effective of These can all be considered as no-regret, or low-re- the two options. If scenarios of desalination ­system gret, measures. breakdown for more than 2 weeks are considered •• The managed recharge of urban water lenses is also credible, then the backup desalination plant expected to require very limited investment, but becomes more cost-effective and can help address the benefits in terms of incremental water availabil- water shortage crises without capacity limitation. ity at acceptable nonpotable water quality levels are •• Finally, the infill of borrow pits in Bonriki, RWH, not known at this stage, and piloting and monitor- and groundwater transfer from North Tarawa ing will be required in order to ascertain them. Any appear to be the costliest measures by far. There meaningful increase of well water availability at an could still be justification for RWH based on the 44 Building Urban Water Resilience in Small Island Countries willingness of households to pay a large premium plumbing supply stores. They also favor the use of for private ownership of this quality water resource, nature-based solutions to filter and store freshwater which is firmly anchored in Kiribati tradition and locally underground. In the context of remote, devel- appreciated for its taste. oping small island states, these features are essential in building the resilience of communities to water-­ 82.  South Tarawa can, to a large extent, rely on low- related shocks. tech solutions to boost its water resilience. All the measures described in this report, except for seawa- 83.  A full cost-benefit analysis of all measures could ter desalination, consist of optimized management not be conducted at this stage, due to a lack of data. of existing water sources, and improved water use Most measures entail externalities that are not habits. The need for backup seawater desalination c aptured in the calculations presented in table 4.3. ­ is still to be clarified in some of the worst-case sce- Positive externalities include, for example, the narios. The management-related measures can be health and economic impacts of major water supply implemented and maintained with materials and shortages, and a healthier urban environment equipment that are usually available in local (improved greywater management), while negative TABLE 4.3. Overview of Resilience Enhancement Measures Impacts on water resources across South Tarawa Unit cost Potable quality Benefits possible/visible during: Measure Incremental (US$ million per Comment Can Break- Normal lpcd Yes No Drought Pollution incremental lpcd) be down regime 1 Water conservation Up to 50 • • • • • • • 0.01 (awareness Avoids wasteful water use campaigns) 2 Tanks crisis Up to 8 • • 0.02 (awareness Need for enhanced management campaigns) awareness about disinfection needs and 3 Rainwater harvesting Up to 6 • • • • 1.0–3.0 (max for techniques expansion private systems) 4 Managed urban TBD through • • TBD Positive health and lenses recharge pilot environmental externalities 5 Direct greywater Up to 8* • • • • • 0–0.15 (based on *Depends on how much PUB reuse plumbing needs) water is used for flushing 6 PUB water storage Up to 20* • • • • • 0.08–1.0 (< 0.4 for *Depends on crisis duration crises under 2 weeks) (20 lpcd if <1–2 weeks) 7 Backup desalination No limit • • • • 0.3–0.5 (CAPEX + Requires strong and system(s) standby OPEX. Based continued PUB financial on size) position 8 Water reserve Up to 18 • • 0.07 Avoids reduction in pollution mitigation production capacity 9 Water reserve palm Up to 3 • • • 0.13 Rental/compensation to clearing water reserves landowners not included in the costs 10 Infill of borrow pits Up to 3 • • • 1.1 11 North Tarawa water Up to 22 • • • • 5.0 transfer Source: World Bank data. Note: CAPEX = capital expenditure; OPEX = operational expenditure; TBD = to be determined. Building Urban Water Resilience in Small Island Countries 45 FIGURE 4.10. Overview of Resilience Enhancement Measures Unit cost (US$ million per incremental lpcd across South Tarawa) Lowest 10 priority N. Tarawa measures transfer Small private systems Rain- Borrow pits infill water harvest. Large collective systems 1 Design to cover long (>30 days) events Back-up Complex housing retro tting desalination Water reserves palm clearing PUB storage 0.1 Water Direct reserves mgt grey- Design to cover short water (<5 days) events reuse Tanks mgt. Highest Water 0.01 conservation priority No or minor plumbing measures 0 5 10 15 20 25 > 25 Incremental lpcd delivered across South Tarawa—2030 (maximum potential) Measures: Increasing potable water availability Avoiding a decrease of potable water availability Increasing availability of water close to potable quality Increasing non-potable water availability Source: World Bank data. externalities could be impacts on local ecosystems Notes (infill of burrow pits, palm tree clearing, greywater 1. Baseline values are based on the surveys carried out in the present infrastructure solutions); brine release (in the study; future values are proposed as an ambitious but possibly achievable target. desalination system); or the social impacts associ- 2. With a 5 m3 water storage capacity and 75 m2 guttered roof, allowing ated with changes in land use or relocation of inhab- the supply of 292 m3 of rainwater for household consumption over itants (water reserve management, burrow pit 20 years. 3. In this study such analysis focused on Betio, Bairiki, and Bikenibeu. infills, palm clearing, water abstraction in North The prevalence of rainwater tanks is likely to be less in other areas, as Tarawa). The valuation of these externalities would they have been populated more recently. require socioeconomic and environmental data that 4. Estimate is based on Falkland 1992. were not available in the context of this study. 5. The sewerage systems in Betio, Bairiki, and Bikenibeu are for the moment collecting blackwater only. 84.  Figure 4.10 provides an overview for each of these 6. The solar PV power generation will feed into PUB’s grid and will not be specifically assigned to the desalination plant. Non-electricity-related measures, of the potential incremental water availabil- costs are minimal at the Bonriki and Buota production systems. ity; its quality; the situation in which it could be mobi- 7. https://www.seqwater.com.au/water-supply/water-treatment​ lized; and the long-term unit cost per cubic meter. /desalination. 46 Building Urban Water Resilience in Small Island Countries Chapter 5 Recommendations and Lessons Learned 5.1.  Priority Measures and Complementary local freshwater cycle—such as integrating greywater Analyses management, managed aquifer recharge, use of 85.  Table 5.1 summarizes the list of measures identified groundwater for specific purposes, and rainwater in the previous sections and the complementary stud- harvesting (RWH)—could be very relevant for other ies that should be carried out to confirm their feasibil- island states in the Pacific, and at the urban utility ity. It assigns a tentative timeline that takes into scale for other water-scarce areas of the world as account, in particular, the parallel implementation of well, if properly adapted to local conditions and cul- South Tarawa Water Supply Project (STWSP) and its tural norms. key prerequisites. These prerequisites include com- 87. Droughts, anthropic pressure on water resources pletion of the new desalination system before (pollution), and systemic infrastructure failures repre- Measure 2 (tank crisis management) can be carried sent more immediate societal threats than the pro- out, and the completion of managed recharge pilots jected effects of sea level rise. Over the next 20 years before the backup desalination system can be sized, South Tarawa appears vulnerable to the degradation among others. A number of the measures proposed and loss of fragile freshwater lenses due to pollution, for the first years of the action plan could be con- over-abstraction, drought-induced salination, and ducted under STWSP, under which activities such as seawater overtopping (SWOT). These growing threats water, sanitation, and hygiene awareness or sanita- increase reliance on seawater desalination, the tech- tion pilots are being considered. nological complexity of which creates a new vulnera- bility of its own. While the solution being 5.2.  Lessons for Water-Scarce Cities in implemented through STWSP is necessary in order to Small Islands and Elsewhere address PUB’s water supply issues, along with this, 86.  The level of sophistication already reached in the efforts should be made to strengthen South Tarawa’s management of Kiribati’s urban water scarcity illus- resilience in the face of these emerging threats. trates the value that small island countries can bring to 88. System efficiency is key for resilience. PUB cur- global discussions about urban water resilience. rently incurs significant losses, with more than Principles such as diversification of resources, fit-for- 60 percent of the water produced not reaching final purpose water use, water sharing, and adaptive water consumers. This issue will be addressed, through management, which underpin the water scarcity STWSP, with the modernization of the entire distri- management strategies of cities around the world, bution network. However, network mismanagement are already at play in South Tarawa. In the absence of is a gradual process that can develop unnoticed, a functional centralized water system, these princi- ­ leading over time to significant losses that, in this ples, which are usually most relevant to water utili- case, would translate into a need for additional ties, are applied here at the household level. The desalination capacity at a significant cost. If this were measures proposed to help households optimize the combined with increased physical losses, the various ­ Building Urban Water Resilience in Small Island Countries 47 TABLE 5.1. Indicative Action Plan, 2020–30 See Water generated/ Financial cost- 0–3 3–6 6–10 Leading Measure Cost para. saved (lpcd) effectiveness years years years agency A. Water conservation measures Awareness campaigns and tariff regulation 50 + Up to 50 A • • • PUB B. Tank crisis management Awareness campaign (when to fill with PUB water) 56 + Up to 8 • MISE Map rainwater tank gaps, support expansion there 57 ++ A • • MISE Review incentive needs for institutional tank owners 57 − • MISE C. Rainwater harvesting expansion Enhance safe rainwater awareness 58 + Up to 6 • MISE D Expand private and institutional rainwater harvesting 40 +++ • Private D. Managed recharge of urban water lenses Restrict usage of non-groundwater-friendly detergents 66 − • MELAD Study and pilot for greywater recharge around wells 65 + B • MISE To be confirmed Pilot rainwater recharge around shallow wells 67 + To be confirmed • MISE based on results Wells salinity and bacteriological quality monitoring 67 + of pilots based on results • • • MISE of pilots Set-up of guidelines for non-potable water quality 66 − • MHMS Roll-out of rainwater, greywater treatment and recharge systems 67 +++ • MISE E. Direct greywater reuse Awareness campaign on benefits, set-up options 70 + • MISE Up to 8 B Setup of revolving fund 70 + • MISE F. Bulk PUB water storage Assess needs based on performance of other measures 85 − • MISE Up to 20 C Design and implement if needed +++ • PUB G. Backup desalination system(s) Assess needs to cover deficits after D and F measures 85 − • MISE No limit C Design and implement (if needed) +++ • PUB H. Water reserve management Implement relocations (if preferred) 76 ++ • MELAD Implement sustainable management plans (if needed) 76 ++ Up to 18 B • • • MELAD Monitor wells around airport fuel handling areas 77 ++ • MICTTD Water reserve palm clearing 79 + Up to 3 C • MELAD Infill of borrow pits 79 +++ Up to 3 D • MELAD Source: World Bank data. Note: MELAD = Ministry of Environment, Lands and Agriculture Development; MICTTD = Ministry of Information, Communication, Transport and Tourism Development; MISE = Ministry of Infrastructure and Sustainable Energy; MHMS = Ministry of Health and Medical Services; PUB = Public Utilities Board. Lpcd estimate based on 2030 population. The table was elaborated based on preliminary calculations of long-term costs (including capital and operational expenditures), and information gathered from MISE, PUB, and the Asian Development Bank’s Project Preparation Technical Assistance during the preparation of STWSP and of this study. Cost (US$ million) Cost-effectiveness (US$ million per incremental lpcd [or avoided loss thereof]) - Less than 0.1 A less than 0.05 + 0.1–0.5 B 0.05–0.1 ++ 0.5–1 C 0.1–0.5 +++ More than 1 D more than 0.5 48 Building Urban Water Resilience in Small Island Countries water crisis scenarios described in this report could level shows promising potential for increasing local challenge South Tarawa’s resilience sooner and more resilience. The differentiation of sources means that acutely than is predicted here. This vividly illustrates effluents are separated at the household level, which the role that system efficiency can play in resilience: provides a good starting point for differentiated treat- it is a silent but important contribution that is often ment and use based on the quality of the effluent. The overlooked in favor of more visible measures, such as formalization of greywater reuse practices can be a first augmentation of supply. step in normalizing reuse behaviors as part of a com- prehensive sanitation and water resilience strategy. 89.  Water supply security can be significantly boosted by local, low-tech, and low-cost measures that make 91. The still largely informal urban infrastructure in the most of existing resources. Many of the priority Kiribati offers opportunities to establish improved measures identified rely on changes in water use water management solutions at the household and practices, nature-based solutions, and other local community levels. Retrofitting plumbing, and water arrangements that can be undertaken at the house- and wastewater management systems to differenti- hold level with very limited need for technical capac- ate various types of effluents can be very costly in ity or additional investment in technology. This modern communities and homes, and is even more ensures that these solutions can be sustained more expensive in multistory buildings. Introducing such easily by the population than complex infrastructure approaches in a city that is still transitioning to for- approaches can be; this is a key feature of long-term mality and modernity can help establish good stan- resilience in a fragile context. The fact that solutions dards and practices before technological lock-in such as RWH and greywater reuse, which are consid- reduces the opportunities to increase resilience. ered “nonconventional” in some contexts, are Municipal land use planning, building codes, and already commonplace at the household level in associated financing mechanisms can be a powerful Kiribati emphasizes the importance of service pro- tool to stimulate the development of good practices, viders understanding household water management as well as to strengthen the protection of water practices when considering their own broader strate- resources located within municipal boundaries. gies. It is, however, important to note that addressing 92.  The preservation and revitalization of traditionally PUB’s water supply issues through the planned frugal and collectivist water management practices, investments under STWSP is a prerequisite for many shaped by a long history of water scarcity, will be criti- of these measures. cal in maximizing the resilience of communities to 90. Although conversations about sanitation often future water supply shocks. In a context of severely focus on blackwater and solids, greywater manage- 1 constrained water resources, the I-Kiribati have ment should be considered an integral part of sanita- already incorporated into their cultural norms and tion strategies. In the case of Kiribati, those who have behaviors key features of resilient water manage- access to improved sanitation mostly rely on seawa- ment. The delivery of safe and reliable water at the ter flushing and sewerage network discharge into the tap by PUB may weaken this precious legacy, eroding sea. While the idea of blackwater reuse in South people’s perception of water as a scarce resource that Tarawa was discarded early on given the cost, com- requires constant conservation efforts at the individ- plexity, and land requirements associated with such ual level, and solidarity at the community level. This treatment, greywater management at the household would undermine their capacity to cope with the Building Urban Water Resilience in Small Island Countries 49 future temporary but drastic decreases in water of water use are more likely to be owned, honored, and availability that can be expected. Awareness cam- sustained by the population. Ethnographic accounts of paigns that promote the continuation of traditional social and cultural dynamics provide a valuable starting water conservation practices, and an adequate tariff point for reviewing the options for improvement of policy will play a critical role in retaining and sustain- water resilience. Such an informed approach implies a ing positive cultural norms around water use. focus on context and on embedding analysis of local perceptions and uses of multiple water sources in the 93. Effective paths toward improved water resilience, intervention design. - therefore, need to recognize not only the technical­ economic drivers of water supply security, but also people’s broader social and cultural relationship with ­ Note water. Interventions that build on the lived experiences 1. Wastewater from toilets. 50 Building Urban Water Resilience in Small Island Countries Bibliography ADB (Asian Development Bank). 2011. “South Tarawa Water and Sources and Their Use in Remote Communities with Evidence from Sanitation Roadmap 2011–2030.” TA-7359 (KIR). Asian Development Pacific Island Countries.” Water Resources Research 53: 9106–17. https:// Bank, Manila. doi​.org/10.1002​/2017WR021047. ———. 2013. Economic Cost of Inadequate Water and Sanitation. South Falkland, T. 1992. Review of Tarawa Freshwater Lenses, Republic of Tarawa, Kiribati. Manila: Asian Development Bank. Kiribati. Report HWR92/681. Hydrology and Water Resources Branch, ACT Electricity and Water, prepared for the Australian International ———. 2017a. “TA-9200 KIR: South Tarawa Water Supply Project Development Assistance Bureau. (49453-001) – Project Preparatory Technical Assistance (PPTA) Demand Forecast Report.” Asian Development Bank, Manila. ———. 2001. Report on Integrated Water Resources Management & Sustainable Sanitation for 4 Islands – Republic of Maldives. Ecowise ———. 2017b. “TA-9200 KIR: South Tarawa Water Supply Project Environmental for UNICEF & Maldives Water and Sanitation Authority. (49453-001) – Project Preparatory Technical Assistance (PPTA) Draft Final Report.” Asian Development Bank, Manila. Fenton, Dennis, and Benoit Laplante. 2018. “Climate Change (Impact) Assessment of the South Tarawa Water Supply Project.” Report prepared ———. 2018. “Consideration of Funding Proposals – Addendum III – for the Asian Development Bank, Manila Funding Proposal Package for FP084.” GCF/B.20/10/Add.03. Asian Development Bank, Songdo, South Korea. Freyberg, T. 2014. “Desalination Dependence Highlighted by Maldives Water Crisis.” WaterWorld, December 8. https://www.waterworld.com​ Aslin, Heather J., and Valerie A. Brown. 2004. Towards Whole of /articles/2014/12/desalination-dependence-highlighted-by-maldives-wa- Community Engagement: A Practical Toolkit., Canberra: Australian ter-crisis.html. Government, Department of Agriculture, Fisheries and Forestry. Galvis-Rodriguez, Sandra, Amandine Bosserelle, Peter Sinclair, Vincent Bandaragoda, D. J. 2005. “Stakeholder Participation in Developing Post, and Adrian Werner. 2016. “Abstraction and Climate Impacts on the Institutions for Integrated Water Resources Management: Lessons from Bonriki Freshwater Lens, Tarawa, Kiribati: An Optimization Approach to Asia.” Working Paper. International Water Management Institute, Sri Lanka. Maximize the Resource during Droughts.” Presentation to the 31st meet- ing of the Science, Technology and Resources (STAR) Network, June 6–8, Binoka, Danfung. 2002. “Case Study: Community Development & Nadi, Fiji. Participation Initiatives Project.” SAPHE Project, Republic of Kiribati, High Level Pacific Regional Consultation on Water. July 29 – August 3. GHD. 2017. “South Tarawa Wastewater System Expansion Concept Design Sigatoka, Fiji. Report.” New Zealand Ministry of Foreign Affairs and Trade. 51/33920. Boakye, M. K., and O.B. Akpor. 2012. “Stakeholders’ Participation in GoK MELAD (Government of Kiribati Ministry of Environment, Land and Water Management: A Case Study of the Msunduzi Catchment Agriculture Development). 2008. “National Water Resources Policy: Management Forum of KwaZulu-Natal, South Africa.” Journal of Water for Healthy Communities, Environments, and Sustainable Sustainable Development 5 (6). Development.” National Water and Sanitation Committee, through the Ministry of Infrastructure and Sustainable Energy. GoK MELAD, Tarawa. Bonzanigo, L. Rozenberg, J. Felter, G. Calner, R. Lempert, and P. Reed. 2018. “Building the Resilience of WSS Utilities to Climate Change and ———. 2014a. “Governance Road Map: Immediate Actions Plan, Parts I, II Other Threats: A Road Map.” Washington, DC, World Bank. and III. Kiribati Adaptation Program Phase III, Buota and Bonriki Water Reserves Task Force.” GoK MELAD, Tarawa. Brewis, Alexandra Avril. 1991. “Age and Infertility: An Ethnodemographic Study from Butaritari Atoll, Kiribati.” Dissertation. University of Arizona, ———. 2014b. “Final Baseline Situation Assessment Report. Kiribati USA. Adaptation Program Phase III, Buota And Bonriki Water Reserves Task Force.” GoK MELAD, Tarawa. Carpenter, Clive and Paul Jones. 2004. An Overview of Integrated Water Resources Management in Pacific Island Countries: A National and ———. 2014c. Cabinet Memorandum: Removal of Buota/Bonriki Water Regional Assessment. Fiji: SOPAC. Reserve Occupants. Tarawa: GoK MELAD. Chifamba, Ephraim. 2013. “Community Participation in Integrated Water ———. 2015. “Draft Sustainable Management Plan for Buota and Bonriki Resources Management in the Save Catchment, Zimbabwe.” Journal of Water Reserves. Kiribati Adaptation Program, Phase III, Buota and Environmental Science and Water Resources 2 (10): 360–74. Bonriki Water Reserves Task Force.” GoK MELAD, Tarawa. Elliott, M., M. C. MacDonald, T. Chan, A. Kearton, K. F. Shields, Gunkel-Grillon, P., E. Roth, C. Laporte-Magoni, and M. Le Mestre. 2015. J. K. Bartram, and W. L. Hadwen. 2017. “Multiple Household Water “Effects of Long-Term Raw Pig Slurry Inputs on Nutrient and Metal Building Urban Water Resilience in Small Island Countries 51 Contamination of Tropical Volcanogenic Soils. Uvéa Island Overmars, Marc, and Sasha Beth Gottlieb. 2009. “Adapting to Climate (South Pacific).” Science of the Total Environment 533: 339–46. Change in Water Resources and Water Services in Caribbean and Pacific Small Island Countries.” Washington, DC: Global Environment Facility; Hoverman, Suzanne, Ingrid De Lacy, Helen Ross, and Terry Chan. 2009. Suva, Fiji: SOPAC. “Vanuatu Integrated Water Resource Management Case Study Sarakata Catchment, Espiritu Santo Island, Vanuatu.” Australian Water Resource PACCSAP. 2015. “Current and Projected Future Climates of Kiribati.” https:// Facility, International Water Centre. www.pacificclimatechangescience.org/wp-content​ /uploads/2013/06​ /5_PCCSP_Poster_Kiribati.pdf. Hoy, D., T. Kienene, B. Rieher, and K. Vieney. 2014. “Battling Tuberculosis in an Island Context with a High Burden of Communicable and Non- Ross, Helen, Bronwyn Powell, and Suzanne Hoverman. 2008. “Public Communicable Diseases: Epidemiology, Progress, and Lessons Learned Participation and Community Engagement for Water Resource in Kiribati, 2000 to 2012.” International Journal of Infectious Diseases 30: Management in the Pacific.” Australian Water Research Facility Policy 135–41. doi:10.1016/j.ijid.2014.11.025. Brief. IWC Australia, Brisbane. Institute of Environmental Science and Research Limited. 2016. Coral Sewell, B. 1983. Atoll Economy: Social Change in Kiribati and Tuvalu. Sands Research Report. Ministry of Foreign Aid and Trade, New Zealand. No.3. Development Studies Centre. Canberra: Australian University. Jiménez, A., and H. Le Deunff. 2015. “Accountability in WASH: A Reference SOPAC. 2007. “National Integrated Water Resource Management Guide for Programming.” Accountability for Sustainability Partnership: Diagnostic Report. Sustainable Integrated Water Resources and UNDP Water Governance Facility at SIWI and UNICEF. Stockholm, Sweden. Wastewater Management in Pacific Island Countries.” GEF, UNDP, UNEP, SOPAC, Niue. Jolliffe, James. 2017. “Cost Benefit Analysis of Sustainable Management Options of the Bonriki Water Reserve, Tarawa, Kiribati.” Climate and SOPAC/UNDP/UNEP/GEF. 2010. “Developing a Pacific Regional Project Abstraction Impacts in Atoll Environments (CAIA) series. Secretariat of for Community-Led Water Resources Management.” Second Meeting of the Pacific Community, Suva, Fiji. the Regional Project Steering Committee, and Inception Workshop for the Project, Implementing Sustainable Water Resource and Wastewater Jones, P. 1996. “Changing Face of the Islands.” Australian Planner 33 (3): Management in Pacific Island Countries. Palau. 160–3. ———. 2012. “Options for Strengthening Community Involvement in IWRM ———. 1997. “Towards an Understanding of the Role of the Sociocultural via Development and Implementation of a Partnership with the GEF Order on Urban Management in the Pacific Islands.” PhD Dissertation, Small Grants Programme.” Fourth Meeting of the Regional Project University of Queensland. Steering Committee, Fiji. ———. 2016. “The Emergence of Pacific Urban Villages: Urbanization Trends Švoma, Jan, and Vladimír Houzim. 1984. “Protection of Groundwater in the Pacific Islands.” Pacific Studies. Manila: Asian Development Bank. from Oil Pollution in the Vicinity of Airports.” Environmental Geology and Kiribati Adaptation Program Phase III (KAP III). “Component C1.4 Water Sciences 6: 21–30. https://doi.org/10.1007/BF02525566. Improved Water Reserves Governance (Buota and Bonriki Reserves).” Briefing Note (22-Feb-15). Tarawa. Taiganides, E. Paul. 1992. Pig Waste Management and Recycling: The Singapore Experience. Ottowa: International Development Research Centre. Kolandai-Matchett, K., E. Langham, M. Bellringer, and P. A.-H Siitia. 2017. “How Gambling Harms Experienced by Pacific People in New Zealand The South Tarawa Sanitation Improvement Sector Project TA 7359. 2011. Amplify When They are Culture-Related.” Asian Journal of Gambling The Water and Sanitation Roadmap 2011–2030. Fraser Thomas Partners, Issues and Public Health 7: 1–20. New Zealand. Kuruppu, N. 2009. “Adapting Water Resources to Climate Change in UNEP (United Nations Environment Programme). 2012. Integrated Water Kiribati: The Importance of Cultural Values and Meanings.” Environmental Resources Management Planning Approach for Small Island Developing Science & Policy 12: 799–809. doi:10.1016/j.envsci.2009.07.005. States: Guidelines, Methods and Tools. New York: UNEP. Nunes, L., Y-G. Zhu, T. Stigter, J. Monteiro, and M. Teixeira. 2011. UNICEF (United Nations Children’s Fund). 2013. “Kiribati: Tracking “Environmental Impacts on Soil and Groundwater at Airports: Origin, Progress in Maternal and Child Survival. A Case Study Report, 2013.” Contaminants of Concern and Environmental Risks.” Journal of _Kiribati​ UNICEF, Suva, Fiji. http://www.unicef.org/pacificislands/14-02-2014​ _ Environmental Monitoring: JEM 13: 3026–39. doi:10.1039/c1em10458f. Case_Study_For_Delivery_to_UNICEF_8-29-2013_conversion.pdf. Mack, Phoebe. 2015. “Bonriki Inundation Vulnerability Assessment Wade, Robert. 1987. “The Management of Common Property Resources: (BIVA) Summary Report.” Secretariat of the Pacific Community, Suva, Collective Action as an Alternative to Privatization or State Regulation.” Fiji. http://biva.gsd.spc.int/. Cambridge Journal of Economics 11: 95–106. Ostrom, E. 2009. “Beyond Markets and States: Polycentric Governance of Weiss, Kenneth R. 2015. “Kiribati’s Dilemma: Before We Drown We May Complex Economic Systems.” Nobel Lecture. Die of Thirst.” Scientific American, October. 52 Building Urban Water Resilience in Small Island Countries White, I. 2010. “Tarawa Water Master Plan: Te Ran, Groundwater.” Kiribati WHO (World Health Organization). 2013. “How Much Water Is Needed in Adaptation Program II. Government of the Republic of Kiribati, Tarawa. Emergencies.” Technical Notes on Drinking Water, Sanitation and Hygiene in Emergencies. No. 9. WHO, Geneva. White, I., T. Falkland, P. Crennan, P. Jones, T. Metutera, B. Etuati, and E. Metai. 1999. “Groundwater Recharge in Low Coral Islands: Bonriki, World Bank. 2017. Water Scarce Cities: Thriving in a Finite World. South Tarawa, Republic of Kiribati. Issues, Traditions and Conflicts in Washington, DC: World Bank. Groundwater Use and Management.” IHP Humid Tropics Programme, ———. 2018. “Program Document for a Proposed Development Policy Grant IHP-V Theme 6. Paris: UNESCO. in the Amount of SDR 3.6 Million (US$5 Million Equivalent) to the White, I., T. Falkland, P. Perez, A. Dray, T. Metutera, E. Metai, and Republic of Kiribati for the Fifth Economic Reform Development Policy M. Overmars. 2007. “Challenges in Freshwater Management in Low Coral Operation.” World Bank, Washington, DC. Atolls.” Journal of Cleaner Production 15: 1522–28. Wutich, A. 2011. “The Moral Economy of Water Reexamined: Reciprocity, White, I., T. Falkland, and D. Scott. 1999. “Drought in Small Coral Islands: Water Insecurity, and Urban Survival in Cochabamba, Bolivia.” Journal of Case Study, South Tarawa, Kiribati. IHP-V.” Technical Documents in Anthropological Research 67: 5–26. Hydrology No. 26. UNESCO, Paris. Building Urban Water Resilience in Small Island Countries 53 Appendix A Multiple Water Sources and Their Uses in South Tarawa—Sociocultural Analysis A. Introduction 2. Integrate multiple water sources and uses into a water governance framework that goes beyond a This paper reports on multiple household water mere focus on drinking water; source usage in South Tarawa. The study employed a mixed-method approach, 3. Seek solutions that are embedded in and reflective triangulating data from 38 semi-direct interviews; a of social relations and norms, allowing for a robust survey of 239 households; a system mapping of shared commitment to emerge. access to water; and observations of participants in Material Conditions of Access to B.  the study. Drinking Water in South Tarawa The objective of the study was to understand the Domestic water use in South Tarawa is primarily sociocultural context within which high-risk and influenced by material conditions of access, that is, high-consumptive multiple uses of water are taking the quality, quantity, equity, and ease of access to place in South Tarawa in order to assess the relevance local sources of water. of various fit-for-purpose water supply strategies in Quality: Water is accessed through multiple sources the local context, and the need for supporting behav- of varying quality. In South Tarawa, five main sources ior change and awareness campaigns. of water are available for domestic use: raw water The research found that although local practices from domestic wells; treated water from the public are in line with contemporary good practices of water supply system (PUB); rainwater; bottled water; and management, they are independently formed and seawater. Three of these sources are used on a regular maintained. The organization of the way households basis for both consumptive and nonconsumptive use multiple water sources in South Tarawa is not purposes1: well water, reticulated water (PUB water), only a response to water insecurity: cultural norms and rainwater (See figure A.1). Our survey found that also play an important role in the people’s choices a vast majority of households in South Tarawa use concerning water use. Cultural and social factors can multiple sources to meet their daily water needs: also create specific dynamics of vulnerability that 50 percent of the respondents use three sources of affect access to water in South Tarawa. The study water; 43 percent use two sources; and only 6 percent found that following local tradition, households use only a single source.2 The majority of respon- often manage their water resources quite carefully. dents (90 percent) use water from a domestic well; The recommendations include proposals to: 87 percent use reticulated water, and 63 percent use 1. Acknowledge people’s cultural and material rainwater.3 resources for dealing with water stress, including Quantity: Water sources and uses vary over time. One the existence of informal networks working along- recurring word in the data garnered from individual side public utilities board (PUB) in providing access interviews about choices of water sources was to water; “sometimes.” In other words, water availability Building Urban Water Resilience in Small Island Countries 55 FIGURE A .1. Answers to the Questions “Do You Ever Use PUB Water? Water from a Well? Rainwater?” Well water 90 10 PUB water 87 13 Rain water 63 37 0 20 40 60 80 100 Percent Use (%) Don’t use (%) Note: PUB = public utilities board. varies, and this variability is not captured by studies single-family dwellings in the 2000s as part of the that are done at a single point in time. As a result, the Asian Development Bank (ADB) Sanitation for Public boundaries between potable and nonpotable water Health and Environmental Improvement (SAPHE) are under constant negotiation, depending on water Project. Observations have confirmed that many of availability. Two types of factors influence changes in the tanks still play a central role in everyday water access to the sources of water and how it is used: the use. However, the tanks are getting old and need to availability of PUB water, and the abundance of rainfall. be replaced, as noted in the case of Betio. Intermittent supplies of treated water often force peo- The differences in storage volume have implica- ple to turn to alternative sources, such as rainwater or tions for access to both PUB water and rainwater. It is well water; and variations in rainfall create changes in crucial for households to be able to store water that is the type of water that is used for drinking. Access to distributed only intermittently. Furthermore, in the quality drinking water is improved during the wet absence of sufficient storage, households cannot use season. Several of the households that do not have a the services of water trucks for the delivery of addi- private rainwater tank reported that they drink rain- tional PUB supply. The lack of sufficient storage also water when heavy rainfalls enable them to collect it precludes the possibility of water storage over time. in buckets. The purposes for which rainwater is used Therefore, families without a rainwater tank or can also vary, with owners of private harvesting tanks adequate storage are more affected than ­ ­ others by prioritizing rainwater for drinking, and to a lesser variability in rainfall. In periods of drought, sporadic extent for cooking in the dry season. access to piped water can translate into high-risk Equity of Access: Changes in water availability affect water uses. In several interviews, drinking raw households with limited storage facilities. The quanti- resort underground water was described as a last-­ tative survey shows that in more than 33 percent of response to conditions of absolute water scarcity households, water is collected in buckets, small “when running out of rainwater.” drums, or pots, while 77 percent of the sample store One respondent in Bairiki, where five households water in one of the storage tanks distributed to share one PUB connection, “would like to have more 56 Building Urban Water Resilience in Small Island Countries tanks if possible, to keep rainwater,” and added that Three intersecting themes emerge from our data as “right now we use buckets and basins to catch rain- significant aspects of the underlying local context of water.” Similarly, a dweller in Betio (who shares a access to water and its influence on water use: the PUB connection with 35 persons) says she needs a understanding of what “good water” is; beliefs in col- bigger tank so she can store PUB water. lectivism and water sharing; and the importance of Ease of Access. In this urban atoll setting, access self-reliance. Understanding of “good water” points are generally quite close to the residents’ homes. Drinkability is a fundamental criterion in the local This was confirmed by the distances reported in the definition of water quality. A clear distinction is made quantitative survey, where the maximum distance to between primary and secondary water sources and water was said to be 289 meters. However, users may uses in South Tarawa. The survey data, which have to collect water several times a day, and carry recorded four different water sources, and 10 uses for buckets weighing between 26 and 30 kilograms. each source, confirms that water for human ingestion needs to be of the best possible quality (see figure A.2). C. Sociocultural Determinants of Water Use •• Rainwater and PUB water are used for drinking, Domestic water use4 in Tarawa is not only deter- cooking, and producing drinks for selling in the mined by the material conditions of access to water form of ice-blocks; but also by the social and cultural dimensions of •• Nonconsumptive uses are overwhelmingly pro- water users’ lives. vided by lower-quality water (well water). FIGURE A .2. Domestic Water Sources Allocated for Various Uses Drinking 6 73 21 Cooking 8 85 6 Ice blocks 22 47 31 Dishwashing 52 47 1 Bathing/washing 71 28 2 Gardening 83 17 Laundry 84 15 Pig rearing 87 13 Mopping 88 12 Flushing 89 7 4 0 20 40 60 80 100 Percent Well water PUB water Rain water Sea water Source: World Bank data. Note: PUB = public utilities board. Building Urban Water Resilience in Small Island Countries 57 In addition, several practices are compatible with •• An unclear understanding of bacteria does not pre- techno-scientific definitions of water quality: vent appropriate responses to water contamination. Some respondents mentioned “germs” or “bacte- •• Boiling water is very common. In most cases, drink- ria” that come through the air; overpopulation; or ing water is boiled in the morning, and sometimes the proximity of wells to toilets. They responded several times a day. Boiled water is kept in the ket- to pollution by avoiding well water, or by boiling tle or poured into a bucket covered with a lid, or a drinking water, even without being able to explain thermos that is used as a water dispenser, making precisely how contamination occurs. water readily available for tea and cold drinks throughout the day. One respondent in Banraeaba However, locally, the notion of “good water” differs said that his family drinks tea with hot water in the from the scientific or technical definition of potable morning “until it is finished, then cold drinks for water. In South Tarawa, health concerns are not a sig- the rest of the day.” The consumption of cold or hot nificant dimension in the people’s definition of “good drinks influences the choice of water. PUB water is water.” often described as being good enough for tea and •• Health issues are rarely connected to choices about coffee (since it needs to be boiled anyway), while water consumption. Interviews show that gastric rainwater (which is often not boiled) is used for diseases are common in Tarawa, but they are not cold drinks. seen as being linked to water contamination. Our •• Children’s water consumption is often different from opinion poll revealed that 94 percent of the sample that of adults. For instance, bottled water is often can’t name a water-related health issue in their bought for serving to the children. In some cases, community. Only 2 percent blame water for inci- Solar water disinfection (SODIS) is reserved for the dences of diarrhea. The interviews disclose that children; sometimes, on the contrary, it is specifi- diarrhea is most often seen as being connected to cally seen as not for them. Rainwater is sometimes food poisoning or epidemics. The primary water-­ boiled for the children, but not for adults; or PUB related concerns are about scarcity rather than the water is boiled only for the children. In some cases, impact of water on health. adults and children also use different kinds of •• The main problem with well water is its salinity, water for bathing: adults might shower with well rather than its contamination. Salinity is the most water while children bathe in PUB water. common reason people give for not drinking water •• There is a high level of literacy on good water prac- from domestic wells. On the other hand, they tices. Interviewees are generally aware that they acknowledge that they consume underground are not supposed to consume well water; and it can water, and argue that this water is “not salty.” Two take many follow-up questions to uncover the of the persons in that situation—including one in practice when it exists. Similarly, respondents tend Betio—mentioned that they have access to two to over-report desirable behaviors: one inter- types of well water: one is acceptable for drinking viewee in Ambo, for instance, asserted that chil- water, and the other one is for other uses. Salinity is dren in her family consumed exclusively SODIS the immediate constraint to using well water that is water, while our observations contradicted this directly experienced by people. When asked how assertion. they know about the salt content, people say they 58 Building Urban Water Resilience in Small Island Countries have tasted the water, either intentionally or not, •• Satisfaction is not often connected to the chemical when showering. Salinity is also noticed in relation quality of water. When asked to explain why they to the fact that detergents or soaps do not produce are satisfied with their water conditions, people as much lather with underground water as with said that they value for instance: PUB water or rainwater. In one instance, an inter- –– Not having to share water with people from viewee mentioned that the first bucket of water outside their family; used for the laundry was filled with PUB water –– Regularity of service (hardship is often identified “otherwise there is no lather.” Another person said with irregular PUB flows); they use rainwater for cleaning the dishes “because –– The amount of water in relation to household the water mixes well with the soap,” something need; that the salty well water does not permit. –– That water is readily available; •• When a link is made between water and illness, it is –– That water is free. usually in reference to an external source of informa- tion. When respondents mention that they don’t The only health-related aspect that contributes to drink well water because it is contaminated, it is users’ satisfaction with their water is the convenience stated as information provided by the Ministry of of not having to boil or treat water before consuming it. Health. If further questions are asked, respondents •• In practice, convenience of access, entrenched hab- may report the presence of “germs” or “worms,” its, and the emotional cost of gaining access to water without adding any further details. It is rare to find override concerns about water quality. People in evidence of direct experience of water contamina- households that are left to walk the “last mile” for tion other than in terms of its taste. In one case, a water quality tend not to walk it. There is, for woman said that her family does not consume instance, a low level of uptake of the SODIS (SOlar water from their domestic well “because it is dirty. water for DISinfection)5 method. Some people It has a brown color.” She also said it is “not salty.” have tested the technique but abandoned it. Only In contrast, the water from their neighbor’s well is three cases of regular use were reported in the sur- “good.” Another case is a resident of Betio who vey. One respondent in Banraeaba said he was mentioned that well water sometimes smells bad. informed about the technique by “Environmental The special nature of well water is also reported as Health.” Similarly, families who have been recom- being experienced through contact with the skin, mended to clean their rainwater tanks do not which feels uncomfortable after a shower. Some always do so. Shame and embarrassment are people are not sure of the reason why they are not important emotional experiences connected with supposed to drink well water. access to water that can prevent users from solicit- •• In several cases, respondents witnessed well water ing water from unrelated kin groups. consumption and reported that it does not have any serious impact on health: “People get sick and Collectivism and Water Sharing they recover. That’s part of life.” (These firsthand The combination of water scarcity, a growing urban experiences contradict the messages received from population, and inadequate delivery of water might health care personnel, and contribute to the per- have been expected to result in the emergence of an petuation of high-risk practices. informal sector of private water vending through Building Urban Water Resilience in Small Island Countries 59 kiosks or other vendors, as it has in larger developing water services. Studying the reasons for water sharing countries. However, the growth of this sector in reveals the types of hardship being faced by some South Tarawa has so far been limited to the reselling households. Depending on others for drinking water of rainwater for fundraising purposes by churches can be a response to seasonal scarcity; a lack of punctu- and community maneaba (the traditional meeting ality of delivery (when PUB water has not been deliv- houses). ered on time); or it can be structural in nature (for Water stress is still managed in large part without example, in cases of failures of local infrastructure). financial transactions, through the use of domestic Benefits and potential harms of water sharing. The wells and water sharing. But while interventions in practice of water sharing without payment reduces South Tarawa have typically addressed the problems some of the challenges documented in other coun- connected with well water consumption, the central tries with respect to affordability, water-related role played by collective social support structures in conflicts, and resilience to drought. Within a family access to water has been insufficiently acknowledged. enclave, the boundaries of cooperation and obliga- As in many Pacific cultures, the inhabitants of Kiribati tion that include requests for water are extended to hold to a collectivist worldview, in which people include the households of relatives. In a settlement “identify as being part of a large interdependent group of 38 persons in Betio that is comprised of the comprising the nuclear family, extended family and houses of three brothers and a cousin, the older lineage members from their village community” brother took on the responsibility for the allocation (Kolandai-Matchett et al. 2017). The powerful ties that and management of rainwater. He reported making bind society together through the sharing of water sure that water is apportioned with increased cau- offers important information about everyday uses of tion when there is a drought warning. However, the water at the domestic and community level. 6 kin-based sharing that lies at the heart of water A significant part of access to water in South Tarawa is access in South Tarawa, and the moderate inequity likely attributable to water sharing across households. of access, can exacerbate individual vulnerability. Evidence collected through this research indicates Analyses of household-level access in three com- that water sharing is widespread in the urban settle- munities show that ordinal position of birth and ments of South Tarawa. In the course of the 38 semi- gender not only determine the degree of control structured interviews we conducted, there were only over shared water sources, but can also affect the 9 cases of households that did not have any form of spatial disposition of houses in relation to the water sharing relationships with other households. water infrastructure. One respondent explained Water sharing is more common and far reaching in that the cost of water gauged by the meters on the areas that are densely populated such as Betio and household doesn’t distinguish the amount of water Bairiki. In urban settlements, water points are shared use by each of several large family units. Having with many different households. Urban families also only a single meter for several different family generally share a greater number of water sources, units placed pressure on her, as a single individual, while in semiurban areas, inhabitants tend to share for paying the water bill, in the absence of a break- only one source with others. down of charges incurred. But as the youngest sis- Water sharing arrangements are primarily a coping ter, she felt that she was not in a position to mechanism for households that have inadequate negotiate with her siblings. 60 Building Urban Water Resilience in Small Island Countries The hybrid water provisioning system of Tarawa nor- Improvement in the water infrastructure seems to malizes water sharing. When the usual source of drink- affect the likelihood of soliciting or providing rainwa- ing water is exhausted, households swiftly move on to ter from or to a relative or neighbor. Sharing of well the next source. If the second type of water is not water do not seem to be affected by access to an available within the household, people rely on outside improved water supply. sources. One respondent in Bikenibeu for instance, Sharing relations can be complex. In some cases, one begs for water from her neighbors’ rainwater tank household may let other families access their well when all the PUB water they have managed to store while depending on relatives for their own drinking has been used. On a practical level, fetching water water. from a relative’s rainwater tank is very similar to Water sharing helps to maintain a culturally signifi- everyday access from any other water point, using the cant custom in contemporary Tarawa. Water sharing same devices and without the need for a financial practices are part of broader set of socioeconomic transaction. Every household in Tarawa has a profu- activities based on norms of generosity, exchange, sion of buckets at its disposal. Buckets are used both and reciprocity that play a central role in social cohe- for collecting and storing drinking water. The most sion and cultural identity in Kiribati. Gifts of water common type of bucket on the atoll is a white plastic result in the preservation of the highly valued princi- bucket with a lid, initially used as a container for ples of an “egalitarian society.” Water sharing may crackers. When filled with water, it contains more involve households with different social ranks, and than 10 liters. The quantities of water exchanged are with a diversity of assets and livelihoods. For some counted in numbers of buckets per day, or jugs or tea- families, ties to well-off relatives are crucial in man- pots per week. aging water shortages. Different sources of water are shared differently. It is The prevalence of kin relationships in Kiribati society quite common to share well water among family and is reflected in the networks of water sharing in Tarawa. neighbors in Tarawa. Some parts of the country also Households are expected to provide for immediate have a long tradition of sharing other water sources. kin in need through the system of bubuti, a “social On islands in the drier southern part of the Gilberts form of reciprocal aid most common between family archipelago, inhabitants have long had access to two members and friends” (Jones 1997). This form of rec- types of wells, including some shared ones that are iprocity is prevalent in Kiribati as well as in other situated in the widest part of the island, where the parts of the Pacific and the world, and is a way of freshwater lens is deepest and where water stays safeguarding subsistence in a context of chronic inse- fresh even during droughts. Several cases were iden- curity of resources. According to traditional concepts tified in South Tarawa of proud owners of “good of decency, requests for water can only rarely be wells” readily letting their neighbors come and fetch refused. In the interviews, customary norms and water. The contemporary sharing of water from rain- bonds of kinship were commonly mentioned to jus- water tanks may depend on such past experiences tify providing water to others, but the notion of with sharing water. However, unlike with wells, it is Christian charity was never used as a reason to pro- not unusual for quantities of rainwater to be strictly vide water to fellow citizens. In fact, it is considered limited. PUB connections can also be shared among shameful to ask for water from someone outside of many households, especially in urban villages. one’s own family network. One result of this is that Building Urban Water Resilience in Small Island Countries 61 water sharing, while it perpetuates and sustains kin- rainwater tank with a neighbor family, and a house- ship relations, can also extend the distance that must hold in Banraeaba allows their neighbors to take a be traveled to access water quite significantly. few teapots of rainwater from their tank each week. However, traditional kin-based relationships do con- Still, sharing appears to be a solution of last resort. tinue to provide a model of organization for social The perception of dependence on others is stressful, exchanges in urban Tarawa that are helpful in coping and can lead to power entanglements, as was sum- with water stress (Jones 2016). Water sharing rela- marized by a household in Bikenibeu. tionships can also be formed across unrelated house- Anecdotal evidence suggests that people’s water holds within a settlement when families share close sharing practices may be changing. Increasing changes ties, and can sometimes be maintained for several in the abundance and quality of water due to popula- generations. These relationships are often described tion growth and environmental change, as well as the as “family like” (for example, “our grandmothers rise of individual values and a growing number of were like sisters”) or “partnerships.” In the heteroge- urbanized communities that are comprised of strang- neous and densely populated areas of South Tarawa, ers are gradually dismantling the formerly built-in where many new settlers are not related to their cooperative living system and related water sharing neighbors, networks of water sharing are increas- arrangements (Jones 2016), thereby eroding water ingly being extended to ensembles of sustained sharing practices in South Tarawa. However, in an sharing and neighborly cooperation, or among mem- ­ interview conducted for a separate project, one bers in a group or community of interest, such as a woman reported, for instance, that she could not risk parish. compromising her family’s well-being by sharing her Independence vs. cooperation: consuming shared sharing rainwater with neighbors. Traditionally, ­ water is never the preferred option. Although there is water is not an individual decision, but rather one apparent pride over the insular model of communi- manifestation of a broader set of social and cultural ties bound together by social cohesion, and strong norms. As these norms are increasingly being ques- cultural norms of sharing, this sometimes comes into tioned in the contemporary urban context, the obli- conflict with the reluctance to depend on others gation to share water is being questioned as well. (including depending on PUB). The emotional cost of sharing was also evident in interviews. One respon- Self-Sufficiency dent from Bikenibeu, who lives in a household that is The priority given to self-reliance, and belief in one’s totally dependent on others for their water supply, own effectiveness in responding to water stress, are says he is “tired of having to ask for rainwater from major determinants of water use in South Tarawa. the neighbors.” It often took some time in the inter- The importance of independence explains the use of view and several follow-up questions to uncover sit- multiple resources for water. Participants in our study uations of dependence. In other instances, sharing referred to the importance of the diversification of arrangements were disclosed by neighbors, not the sources in securing access to water. The wealthier interviewees themselves. On the other hand, house- households that participated in the study were mid- holds that are sharing water with others would be dle-class urban dwellers with two or more incomes. more open about it: for example, one family in These families typically make sure they can secure Bikenibeu shares their PUB connection and private access to at least two, and ideally three, different 62 Building Urban Water Resilience in Small Island Countries sources of water. One middle-aged civil servant said: times of the day for a multitude of domestic “I worry about water because we rely on only one purposes. source of drinking water. What would happen if the The primacy given to self-reliance drives the inhabi- pipes were damaged?” Respondents in Eita and tants of South Tarawa to meet their water needs. Water Baraeaba expressed the same concern. Another par- users in South Tarawa are actively engaged in gaining ticipant who has access to PUB services 24/7 said: access to water through sharing water; relying on “We would like to have a rainwater tank to drink. We noncentralized water supplies; getting water from could use it if there were problems with PUB.” illegal connections; and using electric pumps to draw The idea of self-reliance is central to the sustained water out of municipal pipes. Kiribati families have use of well water. Domestic wells have played a traditionally played a central role in securing their long-standing role in demonstrating the link between own water. Their practices are grounded in commu- landownership, kin, and residence in Kiribati. nal solutions that are devised to manage scarce Digging a well is one of the first material and moral resources. They believe that water should be fetched claims made on land. Owning a well also helps meet and shared within the household: “Traditionally, obligations to the family, an important factor for individual ownership of goods or services was absent maintaining social recognition and self-esteem in (in the Pacific islands), with high value placed on the traditional cultures. One respondent in Ambo, an kin groups working together and sharing the fruits of elderly man born on Aurorae, a drought-prone island, their collective efforts” (Jones 2016). In contempo- explains: “I have tried to dig at least 10 wells since I rary South Tarawa, this happens within the context moved here but they were all salty (…). With a well, of a high fluidity of households marked by frequent you are not dependent on others, you don’t have to departures and returns, and households of a large wait for others to deliver water. (…) If my well yielded size largely exceeding the average seven or eight potable freshwater (not salty), I would stop using people. PUB right away.” In practice, well water seems to Domestic efforts to be self-reliant are consistent with function as a safety net. “I am not worried because I the absence of well-defined lines of accountability in have a well” says a man in Bikenibeu. And a 72-year- the provision of drinking water. PUB water is perceived old man in Bairiki says: “We need the well in case as belonging to entities that are not clearly linked to there is a problem with PUB water.” Many of those the general public: the government; PUB; or the local who disclosed that they consume well water for landowners of Bonriki and Buota. The perception of drinking or cooking purposes, said that they do so inadequate access to water does not translate into “when [they] run out of rainwater and PUB water.” political demands for better services. To solve a prob- One noted “We need to use the well because PUB is lem of water supply in a local community, contacting unpredictable.” The overall importance of well water the MP or a councillor does not seem to be consid- is palpable in numerous statements: “I feel sorry for ered an option. Complaints to PUB have arguably those who don’t have a well;” “A well is necessary increased over the past years7 but dissatisfied water because you always need water;” “In our situation, users that we talked to in Tarawa had either given up well water is necessary;” and “It would be a problem after several unsuccessful attempts or did not even without a well.” The central role of wells in everyday try to approach the utilities. Participants in the study life is also visible in practice. Water is bailed at all expressed a general lack of trust in services provided Building Urban Water Resilience in Small Island Countries 63 by public agents outside of their own personal net- their access to water. Those who reported water works or family: “People work for their family first. hardships tend to be those who share water with a It’s difficult to work outside families, because there is large number of people, drawing on sources that no reciprocity. It is changing, but this is still very are limited. One respondent shares one PUB con- strong in Kiribati.” The lack of trust in public services nection with 5 other households (or about 32 peo- in this society translates into a lack of action to hold ple); another one, in Ambo, shares two ineffective service providers accountable for standards of water PUB connections with 20 people: both of these delivery. individuals indicated that they would like to have Respondents tend to minimize water-related hard- more PUB water. One respondent said they “need a ships. This is illustrated by the answer to the question lot more water, more rainwater.” Another respon- about problems experienced with water supply: 145 dent—a member of one of 4 households, or 21 per- of the 237 respondents (61 percent) said they had no sons, sharing all three water sources—complained problems at all. In longer interviews, participants about the poor quantity and irregular services would almost always start by stating that they are available from PUB. Some people talked about the “satisfied” or “happy” with their water situation. But hard work, the time, and the number of people that then they described very poor conditions of access. need to be mobilized in order to collect and store For example, in Bairiki, a household of 9 experienced PUB water: “When water flows, the whole family is recurring difficulties in covering the cost of the 20 or called. Everybody needs to interrupt whatever 50 cents charged by the nearby church maneaba they were doing and mobilize all the containers (meeting hall) for a bucket of rainwater. Nonetheless, available.” they often say they are “satisfied” with their water sit- uation. To understand this apparent contradiction, it Local Dynamics of Vulnerability should be understood that the I-Kiribati don’t will- There are multiple signs of vulnerability to water inse- ingly share information, whether positive or negative, curity associated with the socioeconomic and demo- with strangers. Furthermore, as underlined by several graphic characteristics of households in South Tarawa. local experts, “This is their normal life.” In other The capacity to deal with water stress is constrained words, they are used to these conditions and see no for many families by the absence of access to a PUB urgent need to change them. There is no evidence of connection: 13.2 percent of the sample reported not an understanding of what acceptable conditions using any PUB water. In the qualitative study, which would be. More fundamentally, as one participant was limited to questions about water storage capac- said, “People have water and they work hard for it.” ity, the lack of money to pay for rainwater punctually, Securing water for the household is still a responsibil- or to buy kerosene to boil the water were also men- ity for each family to shoulder. Despite the shift to an tioned. Several persons interviewed said they would economy based on wages, failure to meet one’s water like to have a rainwater tank but that they can’t afford needs is still often seen as reason for shame. It is per- the cost. ceived as a sign of laziness and therefore is not readily In addition to differences in wealth, profession, and disclosed, especially to outsiders. education, some factors of unequal access to water are However, in follow-up questions, some people specific to South Tarawa. Local norms, values, atti- reported that they worried about some aspects of tudes, and aspirations constantly shape and 64 Building Urban Water Resilience in Small Island Countries influence the way residents of South Tarawa access in Bikenibeu say their geographical proximity to and use water. the water reserves explains why they receive bet- ter service. At the island level, the water pressure •• Situation among siblings. Analyses of house- in neighborhoods farthest from the distribution hold-level access in three communities (family main and service reservoirs is not kept at the plots in Ambo, Eita, and Bairiki) show that the threshold required to reach the homes. communities located along the water supply net- Communities in this situation are often new set- work are heterogenous and difficult to categorize tlers, but also households with vulnerable mem- in socioeconomic terms. Within family land plots, bers, such as the elderly, or female-headed ordinal position of birth and gender determine the households. The quality of underground water in degree of control over shared water sources and periurban areas is experienced as more acceptable can affect the spatial disposition of houses in rela- than in the most urbanized areas, which fosters tion to the water infrastructure. consumptive use of well water. •• Situation of new settlers. With new settlements •• Social obligations. Obligations to community and increasingly flourishing on state and freehold kin, and especially financial obligations to the lands (Jones 2016), the already low water pressure, church, are often fulfilled ahead of household local cuts in services after the renovation of the expenses that could improve water security. One main road, the cost of new connections, and rules respondent in Eita said that some families may be that make illegal settlers ineligible to apply for contributing up to several hundreds of dollars a water connections preclude many new settlers week to their church, including at church fundrais- from accessing the PUB network. Furthermore, ing events that are organized every payday. These new settlers are by definition not equipped with contributions are likely to reduce the capacity of the 500 liter storage tanks that were distributed by the poorest homes to invest in strategies of diversi- the SAPHE project. fication to improve their access to water. •• The intermittent water supply system creates and reinforces inequalities. The PUB schedule produces •• Urbanization trends. In the villages of Betio and differences in the ability of families to access water. Bairiki, which are the most densely populated areas Residents who own storage tanks can fill and store in South Tarawa, households sometimes lack suffi- enough water to last until the next day. For house- cient space to install water tanks near their homes. holds without storage, or who share a tap with In addition, some people living in government many other families, the uncertain access to water housing are entitled to a rainwater tank but can’t get generates anxiety. They need to negotiate with one because their houses have thatched roofs. others and be physically present in order to collect •• Intersectional differences in gender-differentiated water at the time it is supplied. roles require attention. •• Multiple forms of spatial differences in water access. –– There is no clear gender skewing for the fetching Different households experience different levels of of water. Men and women share the burden of hardship linked to uneven water pressure and the fetching water for their domestic needs, both for variable quantities of water that are supplied large quantities of drinking water from distant through the network across the atoll. Respondents sources and for more frequent and limited Building Urban Water Resilience in Small Island Countries 65 quantities of water from closer sources. This Water Conservation insight gathered from direct observation was •• There is a link between the quantities of water that confirmed in the answers to direct questions are demanded per source and the number of sources about gender sharing of labor. The outcome available for domestic use. When more than one is consistent with the observation that men source is accessed, household water management and women seem to have similar levels of is complex, and the sources are matched with knowledge about water resources (knowledge intended uses according to whether they require of the PUB delivery schedule, and familiarity more or less water. Therefore, the quantities with water storage, treatment, and uses). The and types of uses are one indicator of the value level of stress and worry linked to the vagaries and accessibility of each source. When it is eas- of access to water, therefore, do not seem to be ily available, rainwater is generally the preferred gender specific. source. The preference for rainwater is not always –– However, gender inequalities are found in the verbalized in situations where households have differentiated uses of water, and in practical no private rainwater tank. In everyday water use, arrangements that discriminate against women. rainwater is reserved for high-value uses such Observations confirm that women and girls as drinking and cooking, for which more lim- are primarily responsible for preparing water ited quantities are necessary. Even in the rainy for domestic use, household sanitation, and season, rainwater is rarely mentioned for water-­ hygiene, including caring for sick family intensive,  nonconsumptive activities such as members. gardening, flushing, or washing clothes. bathing, ­ –– Mothers, wives, sisters, and daughters depend on contrast, well water supports the majority In ­ the decisions made about water by male relatives. of nonconsumptive uses at any given time and The convenience of water use is significantly across seasons. The number of different noncon- improved for women in households where the sumptive uses of well water and types of uses male head of the family has invested in even explain the fact that, although it is less visible, rudimentary plumbing systems and pumps. well water is still very present in everyday life. Male family members are often the ones •• The equipment in rainwater tanks fosters attentive- who make the decisions with regard to PUB ness to fluctuations in the water supply and increases connections, or who purchase or build water- residents’ stated awareness of the need to save related technology for household use. water. People who own a rainfall tank can track the abundance or scarcity of rainfall by being aware of D. Water Conservation and Reuse the volume of water in storage. “There are mes- The use of multiple sources for water translates into sages on the radio to warn us about drought. We specific patterns of its domestic management, includ- see it also ourselves.” This knowledge informs how ing practices of water conservation and graywater people manage their water supplies. There are reuse. The various values of the different water some reports of rainwater being apportioned ahead sources have an impact on how much water will be of droughts. People who simply harvest rainwater used from each source, and whether it will be con- 8 opportunistically do not report planning their con- served or not. sumption to the same extent. 66 Building Urban Water Resilience in Small Island Countries •• The use of devices designed for centralized water gives her pigs a shower. (This statement was vali- schemes such as flush toilets, washing machines, dated by the observation that her three pigs were or pressure hoses is still not widespread in South indeed wallowing in a muddy pen.) Additional Tarawa. water is also needed to clean the trough and to •• There is no notion of responsibility to a common-pool clean pens that have concrete flooring. (This type resource. Water has been traditionally managed at of enclosure is, however, not usual in the city, and the household or kinship group level in Kiribati. the cleaning of these sites is not always regular). The changes introduced by the networked water •• Cultural norms of water rationing are multiple and system do not seem to have transformed the peo- changing. Historically, the landowners of South ple’s relationship to water sources. The vulnerabil- Tarawa have had ready access to fresh under- ity of the water reserves of Buota and Bonriki is ground water. Today, the city is home to residents sometimes presented as a potential threat, but no from the outer islands, who have different types of direct link is made to household water use. water resources and precipitation patterns, and are still attached to the traditions of the rural societies •• Pigs are significant water consumers. A vast major- from which they came. Teachers who participated ity of the surveyed sample (70 percent) answered in the study also tell of a new generation that is less the question about their water practices in relation prone to conserving water. As a result, the evi- to pigs. This is loosely consistent with the 2015 dence of behavior patterns remains somewhat census, according to which 80 percent of house- mixed. However, the following general remarks holds in South Tarawa rear pigs. According to the can be made: 2015 census, there were 16,075 pigs in South Tarawa. Given the numbers, the amount of water –– Practices of water conservation differ according to consumed by pigs is significant. In the great major- the source. Rainwater is conserved to a greater ity of households interviewed for this study and extent than other sources, especially by families for a separate research project, respondents who do not have their own rainwater harvesting reported giving their pigs liquid feed, generally (RWH) systems. Well water is not discussed twice daily. Self-reported measures of the quanti- in terms of scarcity or saving. For instance, in ties of water fed to pigs show that adult pigs get Ambo, a male respondent in his fifties with around 10 liters of water per animal per day. Of the 9 seven pigs and very limited PUB water resources 165 households in the sample that reported owner- said that he showers twice a day: “It’s not ship of backyard pigs, 87 percent said that the pigs wasting because it is well water.” are given well water. –– The conservation practices for PUB water are •• Many pigs get extra fluid in addition to their feed. It more complex. Evidence of rationing treated is common in South Tarawa to give pigs extra water water is reported by families whose water from a variety of sources “for drinking at lunch pressure is low; by teachers; by families that time,” “sometimes,” “when it is too hot,” “in the are entirely dependent on the PUB network; ­ morning,” “when the children come back from and in households where the “user pays” cost- school,” “when we have a bucket full” or “every recovery pilot program is being implemented. day.” One respondent mentioned that she even A woman in Bairiki recalled the signs “Kawakina Building Urban Water Resilience in Small Island Countries 67 te ran” (“Conserve water”) that were put on taps soapy water for washing (combined water and dish in Betio in the 1980s, when connections were detergent); and a third one for rinsing.” Several res- metered and PUB water had to be paid for. In idents reported having abandoned gardening other households, parents say they encourage because of a lack of water. their children to save water. Several parents •• Pigs are part of domestic greywater reuse. The lived mentioned that they scold their children for realities of water used for pigs that came to the fore wasting PUB or rainwater. One respondent through interviews and observations point to the forbids her children to bathe using PUB water, important role played by backyard pigs in domes- and insists that they use well water instead. tic water management. Most interviews revealed The children in another respondent’s family that, in addition to their wet feed, pigs are fed a use PUB water for showering only at the time variety of liquid kitchen waste: water used to rinse of distribution, after all containers have been the rice, to wash and prepare vegetables and fruits, filled. to gut and clean fish before its preparation, and –– However, when no one is home to collect the fish stock or soup. The fish is cleaned in a small water at the time of its distribution, the liquid is basin and the water used for this purpose, along left to seep into the ground (as seen in Ambo), with fish guts, blood and scales, is then boiled for unless prior arrangement for collecting the water the pigs. A teacher in Bairiki says this is “an advice has been made with the neighbors. What could from old people, otherwise the mothers will kill be seen as a sign of indifference to waste can be their babies.” In addition, pigs are also served left- understood first and foremost as a sign of strong over drinks. As recycled liquids, these different flu- respect for neighbors’ integrity in Kiribati: one is ids consist in a mix of different types of water that king on one’s land. extend beyond well water. Water Reuse •• Water transformed by backyard pigs plays a role in •• Kitchen water is used as a source of irrigation water. the contamination of underground water. Raw pig The most advanced example we saw in Tarawa of a slurry (which is a mixture of animal feces, urine, low-tech system that uses gravity to increase the and water) is spread onto the soil without any productivity of a backyard fruit-growing business treatment. Pig slurry is well known to be a source was in Ambo. In this instance, sink water is drained of organic matter that has micronutrients, but it from the kitchen to a single large banana pit. In also has high levels of copper, zinc, iron, manga- other cases, good-quality water is infiltrated into nese, cobalt, and cadmium (Gunkel-Grillon et al. the ground by throwing kitchen water together 2015). The high hydraulic conductivity of the coral with other valuable waste products (leaves and sands means that pig waste is rapidly transported other organic material) on fruit trees and ornamen- through the soil to groundwater (White et al. 2007). tal plants. The process of dishwashing is described The long-term application of raw pig slurry onto as follows: “There are three basins used for wash- coral soils can lead to an increase of the total nitro- ing the dishes: one in which food residue is scraped gen, phosphorus, and other metal content, with a from dishes and rinsed a first time with clear water: high level of contamination factors (Gunkel-Grillon this is the pig food. Then there is a basin filled with et al. 2015). 68 Building Urban Water Resilience in Small Island Countries E. Recommendations quality problems and take actions to address them. This section summarizes key recommendations •• Establish more equal conditions of storage, and drawn from the study of water uses in South Tarawa. increase water security by equipping households It aims to provide practical ideas to guide culturally with RWH tanks and promoting the storage of PUB appropriate strategies for increasing and sustaining water, including through the distribution of new the health and livelihood benefits of improved access containers to the most vulnerable households, and to safe water. The recommendations are referenced replacing the old ones. at the point where the findings of evidence and anal- ysis were made. Insights from the Analysis of Sociocultural 2.  1.  Insights from the Review of Material Conditions Determinants of Demand for Water of Access to Water The cultural resources already mobilized by the The review of the material conditions that determine inhabitants of South Tarawa for dealing with water water usage offers lessons in how development part- stress show that cultural norms of equality; the prev- ners can help foster water security. alence of nonmonetary exchanges of water; and The expansion and improvement of drinking water ubiquitous access to domestic well water all provide supply services in South Tarawa will no doubt recon- a safety net that explains the general absence of the figure residents’ access to safe water, notably by commodification of water in Tarawa. As a result, reducing vulnerability to contamination and sea- even the poorest citizens have access to water. sonal variability. However, the study clearly shows However, there are concerns about the correlation that residents will likely continue to supplement between these practices and water-related illnesses, piped water with various noncentralized supplies of as well as about overconsumption. water in the foreseeable future. Key recommendations in this regard are to: Key recommendations in this regard are to: •• Work “with the grain” in community awareness work. •• Integrate the multiple water sources into data col- Many interventions have made great efforts to lection efforts and assist practitioners in collecting build on traditional forms of governance in order and using the data in the monitoring of water qual- to inform inhabitants about projects, seek their ity and water use practices. approval, and collect views of members of the •• Involve residents of Tarawa in the management and communities (White et al. 1999). Most community monitoring of rainwater and the urban lenses under outreach projects are conducted in the maneaba, a Betio, Bairiki, and Bikenibeu, and along South convenient platform from which to train and edu- Tarawa. Several residents indicated that tests had cate people. But these hybrid forms of participa- been performed in their wells, and that they were tion do not always mean that local participants, interested in being informed about the results. bureaucrats, and external support agencies under- One option is to make information about under- stand and acknowledge each other’s motivations. ground water quality regularly available in com- In addition, data collected in the interviews shows munities. A simple total coliform test and a that many people engaged in income-generating participatory approach could also encourage activities do not have time to participate in com- communities to learn more about their water munity outreach meetings; that concepts being Building Urban Water Resilience in Small Island Countries 69 discussed are often misunderstood; and that pro- integration into the work of health and water pro- ­ posed solutions are not adopted. 10 fessionals in water monitoring and communica- •• Use existing forms of community organization to tion to the public. Advice concerning water use reach out to local residents about their water use. should flow from multiple sources, including from The water supply, sanitation, and hygiene (WASH) elder to younger members of the community, and community mobilization program, conducted as in cooperation with the churches. part of the South Tarawa Sanitation Improvement •• The water needs of pigs need to be factored in, within Sector Project (STWSP),11 has successfully strength- a management framework of multiple water sources. ened or encouraged the creation of community Pigs will remain an integral part of residents’ water groups that are modeled on traditional forms of use and reuse in the near future, and this use is decision-making bodies at the community level. linked to the quality of the underground freshwater Building on the vibrant participation of individuals lenses. Measures to improve animal health and wel- in these community activities, and on the compet- fare would improve their survival rate. This, com- itive mindset of the I-Kiribati, communities are bined with restrictive policies about pig housing in brought together in beach cleanup operations; or certain areas, could translate into reduced numbers they compete with other communities for the of animals and thereby decreased the volume of financing of projects.12 water being consumed and transformed by pigs. •• Harness the competitive mindset of communities in •• Enhance the level of trust between users and service Kiribati to help drive progress in reaching develop- providers. This study shows that in South Tarawa, ment goals. One possible effective strategy for the formal institutions of water supply are coex- involving local communities is through the awarding isting with dense informal networks made up of of prizes, gifts, and competitions (subject to approval strong traditional ties. These tight-knit communi- by local authorities) in order to stimulate interest, ties not only demand a great deal from insiders, and make the process of engagement attractive. but they tend to fragment society overall, because •• Create community outreach messages that reflect trust is confined to a close circle of kinship. In the the importance of a collective identity, with a focus context of inadequate delivery of water services, on family and/or the household as a functional this particularized trust fuels distrust for the enti- unit, and on community rather than individuals. ties responsible for the allocation and provision of •• Make choosing high-quality water attractive, in line water to the general public. In addition, there with local motivations. Data collected in South could be a perception that the opportunity for dia- Tarawa has highlighted the fact that communities logue on water supply is being usurped by the are already very well aware of public health mes- international donors’ role in assistance to the sages emphasizing better health. Promoting the water sector. External support agencies need to healthy use of water within the local context of work toward bringing water users to hold PUB Tarawa may be more effective if more emphasis is accountable for the quality of water services. Key placed on the social norms of dignity, self-reliance, recommendations in this regard are to: and quality of life, and less emphasis on health. –– Clarify the accountability framework in water This strategy would also benefit from better supply for domestic stakeholders (Jiménez and 70 Building Urban Water Resilience in Small Island Countries Le Deunff 2015). Provide water users with clear, This approach exposed situations of vulnerability accessible, comprehensible information on water that otherwise might have gone unnoticed. It also services and users’ rights and responsibilities enabled the triangulation of information received through understandable contracts and detailed from interdependent individual households. water bills. •• Appropriate selection of data collection strategies to –– Operators can counteract the lack of trust reduce bias. For example, face-to-face interviews by improving the functioning of complaint were conducted, to reduce respondent discomfort mechanisms. This can be achieved through due to the presence of bystanders or focus group providing users with better information about members; also, several different interviewers, who those mechanisms; assuring that there is represented different demographic characteristics sufficient transparency of the process; and (age, gender, nationality) were used. The latter may providing feedback on outcomes within clear have helped to reduce the overall effect of inter- and appropriate time limits. viewers’ expectations, whether real or inferred. –– Additional channels can be put into place •• The tendency of residents to minimize water-related for handling complaints through existing hardship argues for the triangulation of various institutional mechanisms for oversight (such sources of data—direct observations, and both as ombudsmen) that function as watchdogs to qualitative and quantitative methods—in order to ensure accountability. more carefully examine the tensions between cul- –– Establish opportunities for dialogue on water tural and physical constraints, and the gaps services (public hearings, community water and between verbal accounts and actual practices. sanitation mapping, and the like.) 3.  Insights from the Descriptions of Local Dynamics Insights from the Descriptions of Water 4.  of Vulnerability Conservation and Reuse Practices The study highlights some of the complexities and This research has identified a variety of water conserva- dynamics of the cultural and social factors that can tion and reuse practices taking place in the capital of affect access to water in South Tarawa. This calls for Kiribati, together with the careful use of various quali- closer attention to local enabling and constraining ties of water to satisfy different kinds of water demands. contextual factors in the monitoring and evaluation Advocacy work in promoting water conservation of interventions. should harness and support these local practices. Key recommendations in this regard include: Key recommendations in this regard are: •• System mapping of water access. The system-based •• Water management awareness campaigns should mapping approach was found to provide a good take into consideration and build on existing practices complement to face-to-face interviews in the that are consistent with the principles of water con- context of collective water access in South Tarawa. ­ servation being advocated. For example, widespread The mapping enabled identification and descrip- kitchen water reuse for gardening and pig rearing tion of the many social, cultural, and economic could be extended to include the reuse of laundry interconnections, as well as the multiple sources of and shower graywater to irrigate gardens. The water that are integral to water access in the capital. rationing of better-quality water could be Building Urban Water Resilience in Small Island Countries 71 systematically extended to all three types of water 5. Water disinfected with the SODIS method is made drinkable using the rays of the sun. Clear PET bottles are filled with the water and set sources, including underground freshwater. out in the sun for 6 hours. The UV-A rays in sunlight kill germs such Attempts could be made to apply the logic of water as viruses, bacteria, and parasites (giardia and cryptosporidia). (Source: www.sodis.ch) monitoring of rainwater tanks that is already  6. The significance of reciprocal bonds for urban survival in South informing water management in many households Tarawa will likely increase “as global climate change is expected to to measuring the quality of domestic well water and intensify chronic ecological risks, including flash-flooding, heat understanding groundwater supply in relation to stress, drought, and water insecurity” (Wutich 2011). rainfall and demand.   7. Personal observation of the author. •• Encourage water conservation by crafting outreach   8. It is difficult to assign precise quantities of water consumed in South Tarawa. The measure of volume of water used per source was not messages that consider water use from the perspec- amenable to the interview technique. Respondents were, however, tive of local expertise and priorities. Priority should asked to name and discuss the source of water from which they used the largest quantity. Very few interviewees disclosed quantities be given to direct experience rather than expert expressed in terms of liters. As an illustration, many respondents knowledge in assessing the potability and quality who have access to a well said they thought that the greatest quanti- ties of water used in the household came from domestic wells; but of water; and the primacy of self-reliance and there were almost as many who thought that piped water repre- responsibility in water management at the house- sented the largest volume consumed. hold level should be recognized and appreciated.   9. The research finds that self-reporting on the quantities of water used are very unreliable in South Tarawa, making both the quantitative survey and the interviews poor tools for serious efforts to measure Notes the volume of domestic water demand. In the case of pig feeding, 1. Drinking and cooking are defined as “consumptive” uses; all other however, the quantities of water are reported in terms of buckets. uses are “nonconsumptive.” Given that most buckets are of the same type in Tarawa, the reporting 2. Percentage calculated on a total of 233 responses, i.e. not including regarding this water use provides a higher level of reliability of the those who did not respond, and three persons who declared that data. they don’t use any of the three sources. 10. For example, tippy-tap, SODIS. 3. Percentages calculated on the basis of valid responses, that is, 235 respondents for PUB water, and 237 for well water and rainwater. 11. South Tarawa Sanitation Improvement Sector Project Additional Financing Project Administration Manual (PDF) Accessed October 22, 4. The aspects of domestic water use that are analyzed here are house- 2017. hold choices about water; the allocation of water to various uses; the quantities of water consumed; and water conservation practices. 12. Interview with a representative from the WASH NGO “Te Maeu.” 72 Building Urban Water Resilience in Small Island Countries Appendix B Rainwater Harvesting Data collection and analyses were conducted through further validated by discussions held with the two consultations with stakeholders, household surveys, local suppliers of polytanks in Kiribati. About 70 per- and aerial imagery analysis by a team of water spe- cent of their business is generated from donor and cialists of the World Bank, and by Ian Hay Consulting, NGO-supported projects; 20 percent is from govern- with the support of Ministry of Infrastructure and ment ministries; and roughly 10 percent involve indi- Sustainable Energy (MISE). vidual household-level engagements. Program funding, selection, and implementation. A. Review of Rainwater Harvesting (RWH) The review scoped seven major RWH programs that Programs in Kiribati have been implemented since 2000 in Kiribati. Scope of past and current programs. According to the Specific interventions in RWH in South Tarawa were 2015 Census, rainwater is the main source of drinking supported by the KAP Programs (about 20 systems); water for 3,901 households, roughly 50 percent of the the MFAT Sector Program (about 40); and the Kiribati households in South Tarawa. Dividing South Tarawa Housing Corporation, through which at least 710 into Betio and the rest of South Tarawa, it appears loans for the construction of RWH systems were pro- that outside of Betio, the majority of households vided. The rest of the programs and projects imple- (53 percent) use rainwater as their main source of mented systems in North Tarawa and the Outer drinking water (not PUB [public utilities board] water, Islands of the Gilberts group. Details about these pro- as most do in Betio). Based on a quick review of past grams are provided in table B.1. and existing rainwater harvesting (RWH) initiatives KIRIWATSAN Phase 1 & 2. The European Union (EU)- in South Tarawa, including interviews with various funded Water and Sanitation in the Outer Islands of organizations conducted during this study, there are the Republic of Kiribati, or KIRIWATSAN, series of at least 750 RWH systems installed in South Tarawa. projects have supported the construction of several (This estimate covers the main programs imple- RWH systems, both in North Tarawa (Phase 1) and in mented over the last 15 years). One system usually the Outer Islands (Phases 1 & 2).1 KIRIWATSAN-1 pri- consists of 1–10 tanks, ranging from 500 to 10,500 marily financed the full installation of RWH systems liters each. The most common systems consist of 2–4 in priority villages identified by MISE, which targeted tanks of 5,000 liters each. These systems were 141 RWH systems in total (17 in North Tarawa, and installed mainly on: (1) communal buildings (most 124 in 12 of the Outer Islands), including 117 that were maneabas); (2) govern- commonly village and church ­ completed as of July 2017. About 250 tanks of 5,000 ment buildings, including ministerial offices, hospi- or 6,000 liters each were installed under tals and clinics, and schools (public and private); and KIRIWATSAN-1. Under KIRIWATSAN-2, the project a small number were installed on private houses. adopted a more holistic approach in assessing water Most households, therefore, rely on collective RWH resource alternatives, focusing on eight of the Outer schemes. The inventory of RWH programs was Islands. Implementation has also evolved from an Building Urban Water Resilience in Small Island Countries 73 TABLE B.1. Summary of Rainwater Harvesting Programs implemented in Kiribati Program Name / Implementing Implementation Arrangements (Application, Description of RWH Projects Funded Project Cost Funding Agency Procurement, Management, O&M) Kiriwatsan I / EU MISE-UNICEF Primarily financed new systems; MISE Full installation of RWH systems (material, provided a list of the locations, with shipping, labor, installation) was contracted a target number of 141 RWH systems out to a company; project could afford in N Tarawa and 12 Outer Islands. 117 to buy additional tanks at the end of the were completed as of July 2017. implementation period, with the agreement that the beneficiary councils would take over and finish the installation (but they did not). Kiriwatsan II / EU MISE-SPC Project worked on existing RWH Trained and assisted by 2 construction catchments, and took a participatory supervisors and 1 water technician from MISE, approach, involving the communities the communities took an active role in the in the construction or rehabilitation of implementation and O&M of the systems. the RWH systems. About 30 systems are to be built in 8 Outer Islands KAP II / GEF DFAT, Office of the Initial target of KAP II for RWH The roofs were selected based on an MFAT, President and systems was 20. Only 4 were built, on assessment conducted on more than 100 line ministries churches/schools, due to an appraisal communal building roofs. Criteria for selection process that underestimated the costs unknown. of the water and coastal works under the project. KAP III / GEF-LDCF, Office of the This project finances RWH systems In South Tarawa, the project has adopted an Average cost for small GFDRR, Japan President and in both North and South Tarawa, as application process approach. Communities system in South Tarawa PHRD, DFAT line ministries well as on the Outer Islands. In South can apply for, and need to comply with, a list is AUD 7,000 Tarawa, the grant provides materials of criteria (e.g., existing roof, minimum of 10 Average cost for larger and TA (through technical works households served, available space to install system in villages in supervisors) to the selected applicants, the system, etc.); present an O&M plan; commit North Tarawa is AUD with a maximum of five 5,000-liter to providing free labor; and be responsible for 25,000. tanks, regardless of the potential maintenance. Successful grantees are usually capacity of the RW catchment—in some church groups, or communities composed of instances more than 25,000 liters. In a limited number of neighboring households. all, 14 systems serving 419 households Design is provided by MISE. TA is provided by a (around 3,700 people) are being works supervisor (through the project). While installed in South Tarawa. In North all grantees agreed to charge for water, it was Tarawa, 4 villages were identified observed during the research that community to benefit from RWH investments, or church maneabas charged for water and systems were installed in (50 c/bucket). 6 different sites. In North Tarawa, the selection is at the project level, based on a water resource assessment, and one system, providing 25,000 liters of capacity storage, is built for about 60 households (or 200 people). The installation is contracted out to a company (contract available, collected during the research). On the community side, formal arrangements are introduced through the creation of a legally binding village constitution that creates and specifies the roles and responsibilities of various table continues next page 74 Building Urban Water Resilience in Small Island Countries TABLE B.1. continued Program Name / Implementing Implementation Arrangements (Application, Description of RWH Projects Funded Project Cost Funding Agency Procurement, Management, O&M) committees (welfare, mangrove plantation, rainwater harvesting, groundwater abstraction, etc.) For instance, the rainwater and groundwater committees are entitled to and responsible for setting the price of water, the O&M rules, etc. They charge a fixed rate of $A1/week for RW and 50 c/week for groundwater. Kiribati Housing Since 2002, at least 710 loans Application for a loan. Max. of $A1,500/ Corporation for household RW tanks (and the Nature of applicants unknown: individual loan associated gutters and downpipes) households, possibly others. have been arranged through the RKL offers a 20-year warranty on their tanks Corporation in South Tarawa. The and provided some of the systems; they have loans (a maximum of $A1,500) initially confirmed that they have never had to provide were provided from a revolving fund, maintenance services. and repayment could be made over a period of 1–2 years depending on Some of the houses and RWH systems are household income. Current modalities visible from the main road at Bairiki. unknown. MFAT The New Zealand Aid Program financed The project was implemented in two rounds: the installation or rehabilitation of at the first one was based on an application least 40 RWH systems (estimate of process, while the second one followed a 110 tanks of 6,000 liters on average, checklist approach built on the lessons learned so a total storage capacity of 660,000 from the first round. liters), mostly on community or church maneabas, schools, hospitals, and clinics. KOIFWAP / IFD MISE Total of 277 RWH systems in Church members were charged a water fee; 4 outer islands; RWH systems were members were willing to pay for water. rehabilitated in existing church maneabas. Note: RWH = rainwater harvesting; MISE = Ministry of Infrastructure and Sustainable Energy; SPC = Secretariat of the Pacific Community. external private company contracted to provide KAP II & III Programs. The World Bank has been sup- material supply, shipping, labor, and installation, to a porting the WASH sector with investments through more participatory approach employing a private/ the Kiribati Adaptation Program (KAP) series of oper- community partnership model, with the latter taking ations that began in the early 2000s. KAP II was a more active role in village promotions, and the con- funded by the Global Environment Facility (GEF), the struction, rehabilitation, and operation and mainte- Australia Department of Finance and Trade (DFAT), nance (O&M) of the systems. To date about 30 RWH and the New Zealand Ministry of Finance and Trade systems have been built through KIRIWATSAN-2 in (MFAT), and was implemented by the Office of the eight Outer Islands. President and key line ministries. KAP II had an initial Building Urban Water Resilience in Small Island Countries 75 target of 20 RWH systems to be installed in govern- Water Reserve. It was designed to allow for future ment or communal buildings, but only four systems connection to a piped supply of PUB water, which were built in South Tarawa due to underestimated will eventually allow for more tap stands. costs for the water and coastal works components. In South Tarawa, the project has adopted an appli- KAP-III, through its Resilience Fund facility, has been cation process approach, where participating com- supporting the creation of small community-based munities are rated based on a set of eligibility criteria RWH systems in South Tarawa, and larger village-­ (adequacy of existing roofs, number of beneficiary ­ roundwater-based reticu- based RWH systems and g households served (a minimum of 10), availability of lated systems (where suitable groundwater lenses are space to install the system, etc.); and are required to available) in North Tarawa and a few of the Outer present an O&M plan, provide free labor, and commit Islands. The implementation of RWH systems in both to maintaining the system. The grant provides tech- North and South Tarawa have adopted two different nical assistance (TA) by making technical works approaches, as follows: supervisors and materials available, with a maximum In North Tarawa, four villages were identified to be of five 5,000-liter tanks, regardless of the potential covered by the installation of RWH systems and the capacity of the rainwater catchment (which is in construction of two village-based groundwater some instances more than 25,000 liters). In total, abstraction systems, in Notoue and Tabonibara, pro- 14 systems have been installed, serving 419 house- viding a total of 16,000 liters per day of potable water. holds, or around 3,700 people. The average project Selection was made at the project level following a cost is $A 7,000, following the standard design pro- water resource assessment. The typical RWH system vided by MISE. Successful grantees of the Resilience can benefit as many as 60 households, or about 200 Fund are usually church groups. The rest are small people, through collective distribution, and provide communities composed of a limited number of 25,000 liters of storage capacity. Facility installation neighboring households: RWH systems are in this is contracted to a company with formally agreed- case built on one or more of the host households (still upon community arrangements through the creation with a maximum of five tanks per community). Water of a legally binding village constitution that creates is always distributed through a collective tap, with no and specifies the roles and responsibilities of various reticulation. committees (welfare, mangrove plantation, rainwa- Kiribati Outer Island Food and Water Project (UNDP). ter, groundwater, etc.). For instance, the RWH and A total of 277 RWH systems were installed in four of groundwater committees are responsible for setting the Outer Islands. The works often consisted of reha- the price of water, general operating rules, assigning bilitation of existing systems attached to church O&M committees charged with looking after facility maneabas. upkeep, and so on. MFAT Sector Program. Between 2012 and 2016 In North Buota, KAP III also financed the construc- around 40 RWH systems were installed in South tion of an RWH system with a reticulation to two vil- Tarawa, including at least 110 tanks (most of which lages with 19 tap stands (3–5 households per tap were 6,000 liters). The project was implemented in stand), providing 800 liters per day of potable water. two rounds, the first based on an application process, The system includes roofed structures and RWH and the second employing a checklist approach tanks within the northwestern boundary of the Buota that built on lessons learned from the first round. 76 Building Urban Water Resilience in Small Island Countries MFAT developed a preliminary research concept note be given to the other members. The Resilience Fund of to carry out baseline research that will collate avail- KAP III took a participatory approach in South Tarawa, able information on RWH in the Pacific. Its scope will involving the communities in the construction, reha- cover rainwater collection, storage, and treatment, bilitation, and O&M of the RWH systems, by running but will not include the distribution of stored water an open application process requiring the provision of to consumers. This research will make design recom- free labor for construction and rehabilitation, and mendations with the objective of contributing to a requesting a viable O&M plan from the applicant com- better infrastructure, and improvement in O&M. The munity. MISE provides the necessary TA to beneficiary World Bank is closely coordinating with MFAT to communities, including training and support of design avoid potential duplication of activities. and construction supervisors, and water technicians Kiribati Housing Corporation. Since 2002, at least 710 who are available for technical advice after the system loans for household rainwater tanks (and the associ- is built. ated gutters and downpipes) have been arranged in Implementation Costs. The average cost for the most South Tarawa. These loans (a maximum of $A 1,500 common type of water tank (5,000 liters) ranges from per application) were initially provided from a revolv- $A900–1,000. Full installation, including material ing fund, and repayments could be made over a supply, shipping, and labor is around $A1,500–2,000 period of 1–2 years, depending on household income. 2 per tank. For the communal systems installed under Some of the houses and RWH systems are visible KAP III, the total cost is about $A7,000–10,000 (with a from the main road in Bairiki, and they seem to be in maximum of five tanks installed). The Resilience good condition. RKL, which offers a 20-year warranty Fund has a grant ceiling amount of $A10,000 that on their tanks, and which provided some of the sys- supports communities and church groups comprised tems, confirmed that they have never had to provide of at least 10 families. For the bigger systems, which maintenance services. were financed by KAP III in North Tarawa or the Outer Islands, the total cost for one system varied, from Design, Construction, Governance, B.  $A24,000–27,500 (for a capacity of 25,000 liters). Operation, and Maintenance of Table B.2 provides the costing details per program. Rainwater Harvesting Systems Current roles and responsibilities of the government Governance, management structure, and O&M. Most, if concerning the quality of construction, and water qual- not all, of the RWH systems installed though aid pro- ity. MISE is responsible for overseeing and monitor- grams are collective installations for churches, hospi- ing the technical design and quality of RWH systems tals, clinics, and schools; or they are for distribution to built in Kiribati. Some RWH guidelines (for design, a defined community of households. In the latter case, construction, and O&M) were developed by MISE the system can be installed on a communal building under KAP II for application throughout the country. such as a maneaba, and households affiliated with the MISE also provides TA for O&M of the systems, espe- maneaba (as well as others) can come and collect rain- cially in the Outer Islands, where there is at least 1 water. In some cases, the system is installed on the technician trained and available on each island. The private property of one of the community members, Ministry of Health is responsible for regulating drink- with a clear agreement that the member will provide ing water quality, and its Environmental Health unit the land and look after the system, and that access will is responsible for monitoring the quality of the five Building Urban Water Resilience in Small Island Countries 77 TABLE B.2. Storage Capacity and Costs for Rainwater Harvesting Development Programs Typical storage capacity installed per Cost of installation Cost of water sold by the Program system of one system community Kiriwatsan I 10,000 liters (total of 1,375,000 liters, No data available No data available (No system in South Tarawa) installed through 117 systems) Kiriwatsan II No data available No data available No data available (No system in South Tarawa) KAP II From 10,000 to 24,000 liters (total of No data available No data available System in South Tarawa 68,000 liters, installed through four systems) KAP III in South Tarawa 20,000–25,000 liters (total of 295,000 $A7,000 tbc Free of charge when installed on liters, installed through 14 systems) an individual household $A0.5/bucket (of 10–15 liters) when installed on a community or church maneaba KAP III in North Tarawa 25,000 liters $A25,000 AU $1/week (regulated by Rainwater Committee) KAP III in North Tarawa, Buota, No data available No data available No data available through reticulated system Kiribati Housing Corporation No data available No data available No data available DFAT Estimated 16,000 liters (estimated total of No data available No data available 660,000 liters, installed through 41 systems) types of water that are available in Kiribati (well 2. First flush device. The RWH guidelines for Kiribati water, rainwater, PUB reticulated water, ocean water, recommend a manual system; however, KAP III in and lagoon water). They sample and test these five North Tarawa installed a semiautomatic device, sources every 8 weeks from fixed sites, and perform based on past experiences. It is worth examining physical, chemical, and microbiological tests. which of the two works better. The semiautomatic However, in practice, because of the limited resources device still requires a manual intervention, but it is available to conduct this monitoring, it is not clear a simple maneuver. how often it is actually done, and what actions are 3. Optimizing the potential of the catchment area. taken based on the results. During the visit to the Mwegaraoin Nanikaai com- Technical Issues Identified during the Site Visits: munity, it was noticed that gutters and pipes were 1. Height of maneabas. The design for RWH systems installed on only about half of the surfaces of the associated with maneabas takes into account the roofs, missing the opportunity to capture twice the height of the roof, considering that gutters on quantity of water. It is not clear whether this was maneabas are generally quite low (< or > 1.5 an issue of the limited number of tanks per appli- meters). This will require additional pumping or cation (a maximum of five); or of the difficulty of excavation so that the rainwater tanks can be clearing the leaves on one side of the roofs; or installed below ground level. Accordingly, under whether it was an oversight of the community and/ KIRIWATSAN, two types of “generic” designs were or the project when the application was submitted developed. and reviewed. In any case, from this experience, 78 Building Urban Water Resilience in Small Island Countries KAP III has confirmed that they now do a more wish they had more tanks because some of the roof detailed onsite inspection for each application, in catchment areas could effectively hold more tanks. order to better assess all technical aspects, includ- Some members in the communities are now buying ing the catchment areas. additional tanks to add to the system for their own 4. Concrete tanks and cisterns: KIRIWATSAN has private use (not to be shared with the rest of the introduced some concrete tanks (in addition to the community). This demonstrates the capacity for water tanks) in the RWH systems to provide more leverage that some publicly-funded initiatives can storage; and MISE is in the process of inspecting have. the first ones. Some leaks were noticed, and are Charging for rainwater. In South Tarawa, some of the being further assessed. RWH systems financed by KAP III were built on pri- 5. Need for treatment/filter. The few tests conducted vate lots or households, but they serve a larger com- by the Department of Health, as well as other munity, usually composed of a limited number of assessments completed under different projects, households living around the system (and predefined have shown that RWH facilities have a high risk of during the application process). These beneficiary getting contaminated in South Tarawa. Even if the communities initially planned to charge for each system is well maintained and clear of leaves, the bucket of water, in order to support a maintenance presence of birds, rats, or other animals/vectors fund; however, in actual practice the caretaker who is can be a source of contamination (and potential a member of the community does not charge any- corrosion of the roof can add heavy metals to the thing. Caretakers rely on a common but informal mix). This is the reason it is recommended that understanding with the neighbor community mem- rainwater be boiled, which some residents do, and bers that when a repair or a maintenance issue comes that chlorine also be used when needed: but this up money will be raised, and each household will was never heard by the team during the visits. need to contribute. Ideally rainwater should be periodically tested and On the other hand, in the case of an RWH system the results should be used to take necessary action. installed on a communal building, such as a church 6. Applying the Rain Tank Calculator. Most of the pro- maneaba, where there seems to already be another grams worked on existing catchment areas, which fee collection system established, a fixed rate of means that the area of roof catchment was a param- $A0.50 cents per bucket is collected by an appointed eter that was more or less set. In principle, RWH sys- member of the community, and guarded by the chair- tems should be designed at full capacity, optimizing man of the community who manages all funds col- all of the available roof catchment area. It was lected (for water as well as other things). In Naanikai, observed that: (a) some government offices could church maneabas are charging $A0.20 cents per 10 capture and store more rainwater if they had had liter-bucket. In KAP III beneficiary villages in North more tanks installed, but the system was designed Tarawa, the committees established under the vil- only to serve the people working in the office; lage constitutions supported by the project are the (b) KAP III provided grants at a ceiling of US$10,000 ones setting up prices and collecting the money, at a and a maximum number of five 5,000-liter tanks per fixed rate of $A1 per week for RW and $A0.50 cent per grant; some beneficiary communities expressed the week for groundwater (where there is a reticulated Building Urban Water Resilience in Small Island Countries 79 groundwater system). Finally, on the private side it or (2) piecemeal shipping of materials, combined was learned that the Toyota facility in Eita in South with poor inventory and onsite safekeeping resulted Tarawa sells desalinated water for $A2.50 per 10–15- in missing spare parts by the time construction liter bucket. From all the interviews, it appears that started. Implementation must include a clear plan for nonpayment for water is never an issue in these types purchasing and shipping complete sets of materials; of systems. and full installation of the system must be completed Sustainability of past programs. In general, the pro- before phasing out. grams financed by donors are recent, and the opera- Community involvement is critical. Whether the tion of these systems (for example KAP III) is community is involved in the construction or not, relatively new; and in most cases there seems to be a (depending on the approach taken by the program), it follow-up mechanism the donors themselves use to ­ is of critical importance that they are consulted prior monitor the condition of the systems. For example, to construction; made fully aware of the project; and MFAT is keeping a list of the RWH investments done organized in assigning either a “caretaker” for the by tracking maintenance conditions along with system, or a committee that is charged with looking potential actions to be taken. KIRIWATSAN II like- after O&M of the facility. In cases where communities wise actively monitors the performance of their are involved in the construction work (for example, ­ systems, building on KIRIWATSAN I experience. when there is a free labor requirement in exchange Nevertheless, our visits showed that there were some for the grant, or the provision of an RWH system), the rainwater tanks being left abandoned and with no likelihood of success depends on strong leadership, hardware, and there were some systems that were and the community’s eagerness and demand for rain- becoming rundown. In North Tarawa in particular, water as an alternative source of water. In some where KAP III had financed one RWH system, there cases, despite the expressed commitment on paper were three spare water tanks onsite that had been left during the application process, there were imple- abandoned (two from Australian Aid (AusAID) and mentation issues and delays because the community one from KIRIWATSAN I). failed to provide the labor they had promised. On the issue of charging money for community members Lessons Learned from Past RWH C.  there was an interesting finding: that is, that in col- Programs and Initiatives lective systems, when the system is built on a village, Functionality and sustainability of RWH. Projects community, or church maneaba that already seems should build and turn over complete and operational to be well-functioning, with a chairman or a commit- systems. It was observed in this study that some tee organizing things, money was able to be effec- communities ended up with rainwater tanks but tively collected. On the contrary, when the system is without the supporting hardware (pipes, gutters, built on private property and designed to serve only a taps, etc.) that are necessary to build a complete sys- limited number of neighboring households (pre-­ tem. This was due to: (1) failure to comply with or identified during the project design or grant applica- honor an agreement made between the donor and tion process), even though they may have committed the beneficiary council (for example, the council was to an O&M model where water is to be charged to the supposed to complete the system but never did, users, the caretaker of the system (in this case the while MISE was not aware of or did not feel account- owner of the private property) does not charge for able for or responsible toward the abandoned tanks); water, but relies instead on a common and informal 80 Building Urban Water Resilience in Small Island Countries understanding that the beneficiary households will Working models. Various models have proven suc- contribute if repair or maintenance issues arise. cessful for supplying RWH at the household level: RWH systems need to be visible to the caretakers and (1) when the works are contracted out, the O&M plan communities. Cases of vandalism are most likely to has to be enforceable, for example, through the cre- occur in systems that are not under the direct ation of a legally binding agreement like the one cre- ­ supervision of a caretaker or a host household. ated under KAP III in North Tarawa; (2) when the Furthermore, the communities usually want to be communities participate in the works, strong works able to see the system, either so they can keep the supervision and active participation from the gov- children away from the taps, which is a frequent ernment through the project team is required. In cause of leaks (and for this reason some projects are both cases, consultation with the communities, edu- now considering installing lockable taps); or to be cational and awareness-raising campaigns, and train- able to monitor who is coming to get water. MISE has ing for the caretakers are essential. The involvement a technician in each council (ideally) who is trained of MISE is likewise critically important, as MISE is the and able to provide TA. entity responsible for the quality of the designs, and Promoting water conservation measures. Beneficiary for providing TA to the communities when needed, communities are becoming more and more well edu- through the presence of technicians on the ground. cated on the proper use and conservation of water; in The role of the Ministry of Health in ensuring the particular, they are learning not to use rainwater for quality of drinking water is also an important aspect, washing or anything other than drinking or cooking. which has so far been missing in the implementation However, one issue that repeatedly came up during of past and ongoing programs. Coordination among the interviews in the communities was the improper these key sector agencies could and should be further use of RWH for making kava (a beverage traditionally strengthened. consumed throughout Pacific Ocean countries for For rainwater supply at the institutional level, such medicinal, religious, political, cultural, and social pur- as in government buildings, ministerial offices, hos- poses). Communities do forbid rainwater use for pitals, schools, sport complexes (whether public or drinking kava—this is an explicit rule set by the private), and commercial buildings, the construction Rainwater Harvesting Committees in North Tarawa, of RWH systems that are designed to provide suffi- under the village constitution. Under KIRIWATSAN 2, cient water at least for the building users, are no-re- the village councils have imposed a strict penalty of gret investments that should be supported and $A20–50 for the use of rainwater for kava. encouraged by the government. Strengthening an enabling environment. The experi- ence of KAP III in supporting villages in establishing a Survey Results basic constitution have been helpful in introducing Table B.3 provides a summary of surveys conducted elements of sustainability into their water systems. with large RWH system owners, through both infra- Establishing a legal framework in the form of com- structure diagnostic and interview with owners or munity agreements that help determine the O&M operators. Figures B.1, B.2, and B.3 describes the loca- responsibilities of RWH, and simple reticulated vil- tion of rainwater tanks in Betio, Bairiki and Bikenibeu lage water systems helps strengthen local account- respectively, resulting from the review combining ability and support mechanisms. aerial imagery and site visits. Building Urban Water Resilience in Small Island Countries 81 FIGURE B.1. Individual and Collective Rainwater Harvesting Systems (dark dots): Identification through Aerial Imagery (Betio) FIGURE B.2. Individual and Collective Rainwater Harvesting Systems (dark dots): Identification through Aerial Imagery (Bairiki) FIGURE B.3. Individual and Collective Rainwater Harvesting Systems (dark dots): Identification through Aerial Imagery (Bikenibeu) 82 Building Urban Water Resilience in Small Island Countries TABLE B.3. Collective Rainwater Harvesting Systems Survey—Selected Indicators Building Urban Water Resilience in Small Island Countries % of sides Year 1st flush Volume Type of Industrial Who can use Price of Tank cleaning Has tank run Someone checks Building area (m2) Nb users guttered installed system? (m3) building water use? water? bucket (AUD) frequency dry? on water quality? Saint Joseph Bikenibeu 1226.694 0.5 Cant say No 4,000 100 Community No Can’t say Can’t say Cant say Cant say church Diocese of Kiribati. 1 Cant say No 10,000 100 Community No Members of 1 Can’t say No No the community Tekaibangaki 300.3338 1 2000–04 No 25,000 100 Community Yes Members of Once a year Every 3 months Yes the community or + Otan Bikenibeu 896.3627 1 < 1995 No 5,000 100 Community Yes Everyone Every 2 years 2-3 times per Yes year Tabukintion 259.4451 1 2010–14 No 5,000 11 Community Yes Users of the Can’t say Once a year Yes building, e.g. workers Tebonnano Maneaba. 387.6772 1 2015–18 Yes 4,00,000 73 Community Yes Only members Can’t say Once a year No of the community Tenarikai Maneaba 32 1 2010–14 No 10,000 cant say Community Yes Everyone Cant say No No Santo Rabaere 292 1 2015–18 Yes 10 cant say Community Yes Everyone 0.5 Once a year No Yes Arebwa Omeka Maneaba 76 1 2005–09 No 15,000 cant say Community Yes Everyone 0.5 Cant say Once a year Yes Santo Mikaere 33 0.5 2010–14 No 5,000 100 Community Yes Everyone Once a year 2-3 times per Yes year Santo Betero 134 1 < 1995 No 10,00,000 cant say Community No Everyone 0.5 Once a year Has not run dry Yes Angaieta maneaba 3 52 1 < 1995 No 40 cant say Community Yes Everyone 0.5 Once a year Once a year Yes Tekaotitaeka church 125 1 < 1995 No 25,000 200 Community Yes Everyone Once a year No Yes Tibereti maneaba 370 0.5 2015–18 No 10 300 Community Cant say Everyone 0.2 Cant say No Cant say MYSA building. 2030 0.5 2015–18 No 5,000 80 Community Yes Users of the Once a year No Yes building, e.g. workers Te Tokatarawa 334 0.5 2005–09 No 5,000 13 Community Yes Members of Once a year Every 3 months Yes mwaneaba. the community or + Broadcasting and 66 0.5 1995–99 No 30,000 40 Community No Users of the Every 3+ years No Yes 83 Publications Auth. building, e.g. workers table continues next page TABLE B.3. continued 84 % of sides Year 1st flush Volume Type of Industrial Who can use Price of Tank cleaning Has tank run Someone checks Building area (m2) Nb users guttered installed system? (m3) building water use? water? bucket (AUD) frequency dry? on water quality? Betio Nippon. 38 1 2005–09 No 60,000 10 Community No Users of the 1 Every 3+ years No Yes building, e.g. workers Sport Complex 3340 1 Cant say No 1,00,000 100 Community Yes Users of the 0.5 Once a year No Yes building, e.g. workers Santa Maria. 44 0.5 2005–09 No 15,000 100 Community No Users of the 0.5 Eveiy 3+ years Every 3 months Yes building, e.g. or + workers Marakei Mwaneba. 474 1 Cant say No 5,000 30 Community No Users of the Every 3+ years Every 3 months Yes building, e.g. or + workers Commerce 13 0.5 2010–14 No 5,000 80 Community No Users of the 2+ times per 2-3 times per Yes building, e.g. year year workers Save Up 211 0.5 2015–18 Yes 10,000 20 Community No Members of 0.5 Cant say 2-3 times per No the community year Teiaroo Maneaba. 56 1 2005–09 No 4,000 200 Community No Members of 0.5 Not cleaned Every few years No the community Building Urban Water Resilience in Small Island Countries Beetereem Mwaneaba. 34 0.25 < 1995 No 2,500 30 Community No Members of 1 2+ times per Every 3 months Yes the community year or + Ministry of Education. 407.8516 0.5 2010–14 No 5,000 35 Government No Users of the 2+ times per No No building, e.g. year workers Ministry of Education. 544.7964 0.25 2010–14 No 40,000 50 Government No Users of the Can’t say No No building, e.g. workers KPF building. 1443 1 2005–09 Yes 1,500 24 Government No Users of the Every 3+ years Cant say No building, e.g. workers Betio Hospital. 26 1 Cant say Yes 70,000 cant say Government Yes Everyone Cant say No Yes table continues next page TABLE B.3. continued Building Urban Water Resilience in Small Island Countries % of sides Year 1st flush Volume Type of Industrial Who can use Price of Tank cleaning Has tank run Someone checks Building area (m2) Nb users guttered installed system? (m3) building water use? water? bucket (AUD) frequency dry? on water quality? KIT 2257 1 2005–09 No 40,000 120 Government Yes Users of the Once a year No Yes building, e.g. workers Sacred Heart College. 757.6737 0.5 2010–14 No 1,250 478 Institution. No Members of Cant say 2-3 times per Yes the community year KGV and EBS 682.1387 1 1995–99 No 10,00,000 300 Institution. Yes Users of the Can’t say No Cant say building, e.g. workers St. Maria high school. 901.9878 0.5 2005–09 No 10,000 370 Institution. No Users of the Every 3+ years No Yes building, e.g. workers KGV and EBS 439.1089 1 1995–99 No 10,00,000 300 Institution. Yes Users of the Can’t say No Cant say building, e.g. workers KGV&EBS 396.7749 1 1995–99 No 10,00,000 300 Institution. Yes Users of the Can’t say No Cant say building, e.g. workers KGV & EBS Classroom. 1011.117 1 1995–99 No 10,00,000 300 Institution. Yes Users of the Can’t say No Cant say building, e.g. workers KGV and EBS 400.3573 1 1995–99 No 10,00,000 300 Institution. Yes Users of the Can’t say No Cant say building, e.g. workers St John Bosco Primary 673 0.25 1995–99 No 5,000 100 Institution. No Members of 0.5 Not cleaned No No school. the community Taaken Bairiki Primary 206 1 2015–18 No 20,000 360 Institution. No Members of 2+ times per 2-3 times per Yes School. the community year year Rurubao Primary School. 142 1 2015–18 No 20,000 271 Institution. Yes Users of the 0.5 Not cleaned Once a year Yes building, e.g. workers FSPK 0.25 2010–14 No 3,000 10 NGO No Everyone Every 3+ years Cant say No 85 Father Iotebwa Building 469.9338 0.5 < 1995 No 1,000 5 Residential Cant say Everyone Every 3+ years Eery few years Yes Notes 1. KIRIWATSAN is an EU-supported WASH program implemented by 2. South Tarawa RWH Assessment, GWP Consultants commissioned MISE through EDF-10. Phase 1 was executed with UNICEF, with three under the KAP II Program. subcomponents (RWH, CLTS, and Water Resources Assessment) from 2011–2016; Phase 2 is currently being co-implemented with SPC, exclusively in eight Outer Islands. 86 Building Urban Water Resilience in Small Island Countries Appendix C Urban Lenses: Water Balance Building Urban Water Resilience in Small Island Countries 87 TABLE C .1. Current and Future Water Balance across South Tarawa Populated Areas TOTAL Tanaea Bonriki Temaiku Causeway Bikenibeu Abarao Eita Taborio Ambo Banraeaba Antebuka Teaoraer. Nanikai Bairiki Betio 88 Area (km )2 10.33 0.09 0.70 1.49 0.15 1.81 0.58 0.74 0.26 0.28 1.04 0.20 0.87 0.12 0.46 1.54 Pop 2020 54,901 198 2,829 4,072 1,843 7,575 1,761 3,395 1,443 1,293 2,072 1,615 5,105 1,152 3,218 17,330 Pop 2030 70,321 254 3,624 5,216 2,361 9,703 2,256 4,349 1,848 1,656 2,654 2,069 6,539 1,476 4,122 22,197 Density (inh/km ) 2 6,807 2,818 5,177 3,500 15,738 5,361 3,889 5,876 7,109 5,915 2,552 10,343 7,516 12,296 8,961 14,414 Precipitation (m/yr) 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 2.1 CURRENT - Normal conditions Assumpt (coeff) Runoff lpcd 25% 206 499 272 402 89 262 361 239 198 238 551 136 187 114 157 98 Evapotranspiration lpcd 65% −134 −324 −176 −261 −58 −170 −235 −155 −129 −154 −358 −88 −122 74 −102 −63 Natural recharge lpcd 72 175 95 141 31 92 126 84 69 83 193 48 65 40 55 34 PUB Leakages lpcd 60% 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 15 WW discharge lpcd 0.8 28 28 28 28 28 28 28 28 28 28 28 28 28 28 28 28 Well water abstract. lpcd −22 −22 −22 −22 −22 −22 −22 −22 −22 −22 −22 −22 −22 −22 −22 −22 Total lpcd 93 196 116 162 52 113 147 105 90 104 214 69 86 61 76 55 Natural recharge m3/day 5082 44 344 733 74 890 285 364 128 138 512 98 428 59 226 758 PUB Leakages m3/day 60% 1055 4 54 78 35 146 34 65 28 25 40 31 98 22 62 333 WW discharge m /day 3 0.8 1969 7 101 146 66 272 63 122 52 46 74 58 183 41 115 622 Well water abstract. m /day 3 −1547 −6 −80 −115 −52 −213 −50 −96 −41 −36 −58 −46 −144 −32 −91 −488 Total m3/day 6558 50 420 842 123 1094 333 455 167 173 567 142 565 90 313 1224 2030 - Normal conditions Assumpt (coeff) Runoff lpcd 25% 206 499 272 402 89 262 361 239 198 238 551 136 187 114 157 98 Building Urban Water Resilience in Small Island Countries Evapotranspiration lpcd 65% −134 −324 −176 −261 −58 −170 −235 −155 −129 −154 −358 −88 −122 −74 −102 −63 Natural recharge lpcd 72 175 95 141 31 92 126 84 69 83 193 48 65 40 55 34 PUB Leakages lpcd 25% 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17 WW discharge lpcd 0.8 62 66 66 66 66 54 66 66 66 66 66 66 66 66 54 54 Well water abstract. lpcd −12 −12 −12 −12 −12 −12 −12 −12 −12 −12 −12 −12 v12 −12 −12 −12 Total lpcd 139 245 166 211 102 150 197 155 140 154 264 118 136 111 113 93 Natural recharge m /day 3 5082 44 344 733 74 890 285 364 128 138 512 98 428 59 226 758 PUB Leakages m3/day 60% 1160 4 60 86 39 160 37 72 30 27 44 34 108 24 68 366 WW discharge m3/day 0.8 4388 17 241 346 157 523 150 289 123 110 176 137 434 98 222 1198 Well water abstract. m /day 3 −844 −3 −43 −63 −28 −116 −27 −52 −22 −20 −32 −25 −78 −18 −49 −266 Total 9786 62 601 1103 241 1457 445 672 259 255 700 245 892 164 467 2055 Note: PUB = public utilities board. Appendix D Conservation of Freshwater Lenses on Bonriki and Buota, through Watershed Management A. Introduction and Background arrangements as occurred in the case of acquisi- Introduction tions under Cap 95B. They did receive “one-off” compensation. The Buota and Bonriki freshwater reserves are one of Subsequently, the Buota and Bonriki landowners the main sources of freshwater for the 58,000 resi- pressed for additional compensation, and access to dents of South Tarawa, who comprise more than 50 the water reserves (LMD 2014). This resulted, in the percent of the population of Kiribati (2016 census). late 1990s, in a decision (raised in Parliament, and This report has been carried out as a desk exercise, approved by the Cabinet) that the landowners would and draws on work carried between June and be treated as lessors, and be paid annual “lease” pay- December 2014 under the Kiribati Adaptation ments in the same way as the lessors that were cre- Program (KAP). The 2014 work was contracted by the ated under the State Acquisition of Lands Act. This government of Kiribati as an input to the KAP Phase practice continues. In addition, the right to occupy III, and involved preparation of a roadmap of actions the ocean and lagoon fronts for a depth of up to 50 to address the threats facing the integrity of these meters was established; a road was constructed two water reserves. around the perimeter to mark the 50-meter bound- Background ary; and survey beacons were erected to mark this The key issues affecting sustainability of the Buota boundary as well. and Bonriki water reserves that were identified in the As is the case with “state land” acquisitions in 2014 baseline report included: South Tarawa, the Bonriki and Buota landowners regard themselves as holding the absolute title to the •• Property Rights Relating to the Water Reserves land, while the government has acquired temporary, The complex legal position regarding property rights of or lessor, rights (legally as a water reserve, in practi- the landowners over whose land the water reserves are cal terms, as a lessee). This perception results in established. The Buota and Bonriki water reserves beliefs that the land can be accessed and the produce were not “acquired” as state land in terms of the State thereof (water from wells, coconuts from trees, etc.) Acquisition of Lands Act 95B—as was the case for the gathered, as the “land belongs to them.” In these cir- Betio, Bairiki, and Bikenibeu (township) acquisitions cumstances, it is likely that measures will be required, prior to independence in 1979. Although the existing either as incentives to cooperate (or disincentives to landowners in Buota and Bonriki held ownership not cooperate), in order to secure a sustainable property rights over strips of land (lagoon to ocean) method of managing the reserves. Disincentives have they did not, at the time that the water reserves were proved very difficult to enforce; by the police in established—Bonriki in 1969, and Buota in 1974—have Kiribati generally, and in South Tarawa in particular. all of their land converted to State land, only the designated water reserve area that was created; and ­ •• Differences Between Settlement Patterns on the they did not enter into the same lease payment Buota and Bonriki Reserves Building Urban Water Resilience in Small Island Countries 89 The Land Management Department (LMD) of households on Bonriki had been there 1 year or less, Ministry for the Environment, Lands and Agricultural only 10 percent of Buota residents fell into this cate- Development (MELAD), as the government’s “custo- gory. At the other end of the scale, 40 percent of dian of the land,” is responsible for making and main- Buota residents had been on the reserve for 26 years taining an inventory of unauthorized households and or more; 50 percent for more than 16 years; and dwellings on the two water reserves. In April 2014 35  percent for 6–15 years. It is clear that Buota and there were about 59 such households on the Bonriki Bonriki have very different patterns of settlement, reserve, and 20 on the Buota reserve. The household with Bonriki showing a much higher level of short- size of settlers on the Buota and Bonriki reserves is an stay, transitional households, and Buota a more set- average of around 5 persons, with a small number of tled pattern of long-term occupation. households having 10–20 members. There is a signif- The unauthorized residents in Buota are settled icant difference in the average number of years of wholly along the inside of the 50-meter perimeter, occupation: Buota settlers have been on the water while in Bonriki they are scattered in clumps in the reserve for almost three times as many years—17.6, middle of the reserve, as well as just inside the compared to 6.7 years in Bonriki. perimeter. Analysis of the percentage of households settled LMD has carried out inventories of the unautho- on the reserves, by number of years of occupation rized settlers on the Bonriki water reserve for the (figure D.1) shows that while in 2012 63 percent of years indicated in table D.1. The overall movement of FIGURE D.1. Percentage of Unauthorized Households, by Years of Occupation, 2014 70 60 50 40 Percent 30 20 10 0 <1 yr 2–5 yrs 6–15 yrs 16–25 yrs >26 yrs Buota Water Reserve Bonriki Water Reserve TABLE D.1. Summary of the LMD Census of Illegal Households in Bonriki Year Total Old New Move out Total on site Net movement % increase 2005 24 24 − 2006 34 34 34 +10 29 2007 52 22 18 12 40 +6 15 2008 46 17 29 0 46 +6 13 2009 62 16 37 8 54 +8 15 2011 85 38 22 25 60 +6 10 2012 99 51 13 35 69 +9 13 90 Building Urban Water Resilience in Small Island Countries squatters in and out of the reserve supports the because they have been perceived as receiving observation that Bonriki has a high turnover of set- unwarranted special treatment for the loss of their tlers (roughly 30 percent each year). land. Many landowners in South Tarawa are suffering Although the net increase has dropped from a high the stress of overcrowding, dislocation, and an of 29 percent in 2005–06, it has held at a fairly consis- increasing demand to accommodate and support rel- tent rate of 13–15 percent since then. The number of atives from the outer islands. It has been observed households moving out each year fluctuates, but is that the Bonriki community still has plenty of bread- quite high; on average around 30 percent. fruit trees, and could fish in the lagoon and collect •• Relationship Between Authorities and South Tarawa food off the reef, as their shoreline is less polluted Communities than the more populated areas of South Tarawa. In the past there has been pressure from the community There is a long history of distrust, and at times open on the government not to listen to the Bonriki resi- antagonism between the residents of Buota and dents, and to evict squatters off the reserve. Bonriki and the government authorities responsible for the land, for lease payments, and for the water. To ° South Tarawa: The Bonriki Community Surrounding a significant extent this conflict arises when the the Water Reserve demands of a densely populated and urbanizing In the late 1990s consultations took place with Bonriki South Tarawa come into conflict with more tradition- community landowners.2 It was reported that some ally oriented community values. There is a general residents felt that they had not received sufficient distrust of government officials and their intentions, compensation for the use of their land as a water and a lingering grievance around the feeling that local reserve. They complained that they had not been communities have paid a high price ceding control of made sufficiently aware of the impact that the reserve their land and received too little benefit in return; and, and the airport would have on their lives. Apart from at times, a feeling that their rights to the land and the the loss of the area available for housing and gardens, fruit of the land they own have been unjustly taken their coconut trees and babwai (taro) pits had become from them.1 In the past, this has led to vandalism of nonproductive. It was reported that the village was water supply equipment in the reserves. This fact must previously self-sufficient in food, and was able to sell be taken into account in addressing the future sustain- copra, but since land has been lost to the reserve and able management of the reserves. the airport, they have had to buy coconuts and depend on money from the government and other sources to •• Many Stakeholders, with Different Interests in the live.3 They also complained that the water in their Water Reserves wells had become saline from over-pumping. Some of The key stakeholders with an interest in management the discontent has arisen because the village has of the reserves include: become seriously overcrowded with an influx of rela- ° South Tarawa Communities, including Betio, Bairiki, tives from the Outer Islands (a pressure being experi- and Bikenibeu enced by all of South Tarawa), and the majority of While they are almost certainly not representative of these people are unemployed. Consequently space all members of the community, discussions with the and land are in very high demand, and the cash avail- wider community of South Tarawa in the past have able from the compensation or lease payments has not shown that the people of Bonriki have been resented been sufficient for, or available to, the expanded Building Urban Water Resilience in Small Island Countries 91 population. Although the residents of Bonriki receive –– Landowners believe the compensation a free supply of reticulated water, they do not consider system is unfair because they are not being it an advantage to not pay for this water, since “the compensated for the water that is being water belongs to Bonriki anyway.” removed from beneath their land. South Tarawa: Bonriki Households on the Reserve ° –– Some interviewees were concerned that it is no longer possible to maintain the boundaries In a 2014 interview with the Bonriki councillor it was of the reserve, and suggested that the area of stated that the main drivers of unauthorized settle- the reserve be reduced to allow more living ment were: space for the increasing population. –– Relatives or offspring of landowners living in the area at the time the water reserves were ° North Tarawa: Buota Households on the Reserve declared. Unlike Bonriki, the Buota islet falls under the author- –– Some settlers on the reserve who come ity of North Tarawa. In interviews with members of from other places may be related to the the Eutan Tarawa Council for Buota in June 2014, it landowners. Many are living in the reserve, was stated that: on the northeast ocean side. –– The main driver of unauthorized settlement –– Others come from South Tarawa to take was the “need for more land” from a growing gravel and to cultivate gardens. number of residential families (retired civil –– One group, from the southern Islands has servants with family links to Buota, and so on). been living there for about 5 years; the –– Most unauthorized settlers were landowners people who live behind the control tower are or relatives of landowners. new settlers. –– In regard to uses of the reserve, there was North Tarawa: Buota Community Surrounding the ° some harvesting of plants for traditional Water Reserve medicines, and collection of pandanus Buota residents are concerned with various issues (palm-like tree) for thatching, mats, etc. concerning the water reserve: –– It was well known that settlement on and use of the reserve is prohibited. –– Landowners are worried about future land- leasing rates. ° Government of Kiribati –– There is no more habitable space for the The main government actors with direct interest in growing population in Buota, and many the two water reserves are: families feel they have no choice but to move –– The Public Utilities Board (PUB), which is into the reserve. responsible for the delivery of water and –– The reserve is increasingly dry, and the sanitation services in South Tarawa, and potential for the community’s subsistence is management of the two water reserves and decreasing. their resources. PUB is under the authority of –– There is a need for freshwater in Buota, the Ministry of Infrastructure and Sustainable especially in the northeast, where well water Energy (MISE), which is responsible for is saline, and about 200 people are struggling the maintenance of government buildings, to have enough drinkable water. regulating public utilities services, construction 92 Building Urban Water Resilience in Small Island Countries and maintenance of roads, technical services, allow the compulsory purchase of land for water and energy management. reserves, this has never been done by the govern- –– The Attorney General, who is responsible for ment, because of the fundamental importance of providing legal advice to the government, land ownership in Kiribati. Instead, the government and providing legal representation for the currently pays affected landowners annual lease government; payments.”4 –– The Ministry for Health and Medical Services Numerous reports recommend that a specific manage- (MHMS) which is responsible for health ment arrangement for the two water reserves be estab- inspection services, monitoring of water lished, and that this involve community participation: quality, and environmental health; and ° “Maintaining the water reserves at Bonriki and –– The MELAD which is responsible for land Buota will require commitment by government management (lease payments to landowners, and the involvement of local communities. compensation payments to unauthorized In order to address the encroachment of occupiers, and land use (planning regulation) settlers onto the Bonriki and Buota Water on the water reserves. Reserves and to include local landowners in the ° Urban and Island Councils process, a community-government Committee for the Management of Water Reserves in Bonriki –– The Teinainano Urban Council is responsible and Buota was proposed in 2000 as a lead-in for urban planning and development control to the SAPHE project. This Committee was on the Bonriki islet; and North Tarawa’s Eutan planned to have representatives from the water Tarawa Council is responsible for the Buota reserve villages, from the Unimwane (traditional islet. elders) of Tarawa and from the lead government ° Committees, Task Forces, and Councils agencies and was to be facilitated by the Ministry A number of committees and task forces have been of Home Affairs and Rural Development with created over the past 20 years to address water supply secretariat provided by the Land Management issues in Tarawa in general, and the Buota and Bonriki Division within the Ministry. The Committee water reserves in particular. These have been of vari- met in February 2002 but is now defunct. Some able durability, and have achieved limited results. government agencies are still uncomfortable Committees and task forces have thus far not proven to with the notion of community participation. be a sustainable or effective means of achieving long- The reactivation of this Committee is seen as term management and governance objectives. essential.” Tarawa Water Master Plan 2010–2030 •• Previous Initiatives to Conserve and Protect the (December 2010). Water Reserves •• Another strategy proposed to improve the man- The Tarawa Master Plan (2010) states that “The cur- agement and protection of the water reserves and rent water reserves in South Tarawa were declared to involve local landowners in the process was, over land that was privately owned. This has been instead of paying them lease fees, paying them to highly controversial and has generated long-standing be custodians and managers of the water reserves, disputes between authorities and landholders with the lease payments being linked to perfor- and their communities. Although PUB regulations mance criteria such as the absence of settlers, Building Urban Water Resilience in Small Island Countries 93 houses and animals; gravel and sand mining; crop are productive, and vegetation is thriving. In Bonriki, planting; the infilling of wells, babwai pits and the coconut palms and babwai pits are not produc- mining pits; and the removal of dwellings and tive: in 2012, 63 percent of the 59 Bonriki households other domestic infrastructure, and the absence of had been there for 1 year or less; and the percentage new burials on the water reserves. The proposal, turnover of households was around 30 percent, with which is politically sensitive, has never been con- a net increase each year of about 13 percent. The main sidered (White et al. 1999). driver of this change is a combination of temporary ° The KAP II Buota Consultations: (December and long-term demands for land on which to settle— 16–17, 2010) states, with reference to commu- from resident families as well as new migrants from nity involvement in protection of the water the Outer Islands. reserves, “There used to be a committee making Land, livelihoods, and access to resources. In consul- sure that people do not build houses on the tations held with the Bonriki community in 1996–97 reserve but it has not been active for at least 10 it was reported that the government was obliged to years now” and “The Unimwane strongly compensate for losses associated with the use of the encourage the government to involve all the land, but not for anything found below ground level, Unimwane of Buota on any decision making to including the water; hence the focus by landowners prevent future damage to the infiltration galler- on the effect of water extraction on vegetation, and ies and the water reserve.” consequent loss of livelihood or subsistence. In addi- •• Risks tion to the agitation for redress, the villagers have Contamination. The most serious risk is contamina- taken practical steps to reduce their difficulties. tion of the water reserves that could result in a very Bonriki residents moved onto the reserve, built serious outbreak of a disease such as cholera, result- houses, and grew vegetables. In some cases, the plots ing in many deaths among the population of South were sold or rented to them by the landowners. In Tarawa. There is also a risk of contamination of the other cases, the landowners could not refuse their reserves by toxic substances (oils, metals, chemicals relatives, due to the custom of the bubuti, even reaching the water lens) that would not be eliminated though they had been compensated by the govern- by adding chlorine to the water. A high incidence of ment for nonuse of their land. water-related diseases that result in clinic or hospital Because the water reserves are perceived as “gov- visits is now found in South Tarawa. ernment land,” they are often raided for their Increasing settlement on and around the reserves. resources. Settlers at the Bonriki end of South Tarawa The risks associated with settlement on the reserves have few opportunities for generating income. Sand are clear; however, it is also clear that the risks are and gravel are currently in very short supply in Tarawa. different in Buota and Bonriki. In Buota, all of the Therefore, sand and gravel mining, although illegal, existing structures built on the reserve are immedi- provides South Tarawa residents with an opportunity ately inside the 50-meter boundary, and the majority to earn some income, and to many, the Bonriki water of the 20 households are resident families and retired reserve appears to be an ideal source for materials. But civil servants with family links to Buota: 50 percent the impact of sand and gravel mining can be seen in of them have been on the reserve for more than 16 aerial photos, and the practice is endemic across the years. Coconut palms and babwai pits on the reserve reserve. Mining also has significant impacts on the 94 Building Urban Water Resilience in Small Island Countries water reserve: destruction of vegetation; increased Conclusions vulnerability of groundwater to pollution (as less soil Taking into consideration the points made in the overlies the water table); increased direct evaporation present appendix, the following conclusions are losses from the water table; and damage to pumping proposed: stations and salinity monitoring boreholes. Inspections •• The Bonriki and Buota freshwater reserves and of the reserve show that the impacts of mining are lenses are critical to the economic, social, and becoming progressively worse, and that they consti- environmental existence of South Tarawa, which tute a significant threat to South Tarawa’s reticulated is the capital of Kiribati as well as its economic and water supply. political hub; and home to about half of the Increased settlement on the two water reserves will population. lead to inappropriate land use through, for example: •• The nature and extent of settlement on the Buota •• The digging of open wells, exposing the groundwa- water reserve is very different from that on the ter to direct contamination, and creating algal Bonriki reserve. The Bonriki reserve is more vulnera- blooms in the water. ble, more accessible, and more degraded environ- •• Active graveyards. There are several active grave- mentally. The Buota islet is, administratively and yards on the Bonriki water reserve, which are developmentally, a more traditional rural environ- exposing the shallow groundwater to the risk of ment, and Bonriki a more “urban” environment. pollution. •• Effective protection and conservation of the water •• The raising of pigs. There are about 2.4 pigs per reserves has been inadequate since they were estab- household in Kiribati. The fecal contaminant from lished in 1969 and 1974; it is a high priority to put pigs on the reserve poses a significant threat to into place the institutional and organizational frame- water quality. works that will provide effective protection and con- •• Growing crops. Prevalent on the lagoon side of the servation of them. Bonriki reserve, the use of fertilizers and animal •• Although the most likely additional source of pota- waste as crop nutrients is a significant pollution ble water is desalination of saltwater, and this will threat. need to be added to the water supply resources in •• Growing babwai. Babwai pits are excavated into the the immediate future, it does not detract from the water table, and fertilizer and animal wastes are importance of the Buota and Bonriki water reserves being added directly to the groundwater. Babwai as a long-term, sustainable source of freshwater for cultivation is a major threat to groundwater qual- a substantial number of Kiribati’s people; for its ity, and is also increasing evaporative losses. economy; and for its environmental sustainability. •• Direct pollution. The terminal wells of gallery pump •• A wide range of actions may be taken to protect stations can be opened and used as rubbish dumps and conserve the Buota and Bonriki water reserves. or toilets. These involve taking measures in the following •• Vandalism to the infrastructure. Vandalism is a areas: potential threat in the water reserves: mitigation ° Initial agreement by stakeholders as to future will involve improved management, education, actions, sequencing, priorities, revenues, costs, and awareness. and management, including different strategies Building Urban Water Resilience in Small Island Countries 95 and actions for residents on and surrounding process, and agree on their training requirements, the Buota and Bonriki islets and reserves, and in and how they will be supervised. Tarawa. 5. Identify and secure the required resources. ° Immediate actions to be taken to educate, raise 6. Initiate a facilitated consultative process with key awareness, and inspire the behavior changes stakeholders to agree on a roadmap on directions needed to increase water yield; to physically (strategy) and actions (including what, when, who, identify and protect the water reserves; to and with what resources). eliminate vandalism; and to increase education and general public awareness of these issues. •• Agreement: A Roadmap for Directions and Actions ° Agreement of a sustainable management plan (SMP). Define strategies and actions for residents on and Public participation and community engagement surrounding the Buota and Bonriki islets and are encouraged. reserves. ° Development of policy, laws, and regulations when and where required to implement the 1. Identify immediate actions, education, awareness, above points. and the types of behavior change that is being ° Find ways to increase revenues and fund costs sought. which arise from addressing the previous points. 2. Physically identify and protect the water reserves. 3. Review and agree on the strategies and actions B. Measures to Mitigate Risks described below, and address questions such as: Initial Agreement by Stakeholders What can be agreed on as a basis for discussion? The starting point for putting into place measures to What principles and criteria should be applied in mitigate risks to the Buota and Bonriki water reserves order to achieve the agreement of all stakeholders? is divided into two sequential activities: Are there any “red lines” (nonnegotiable boundar- •• Initial Discussions ies) within which discussion must take place?7 The following matters should be considered: Immediate Actions, Education, C.  1. Hold initial meetings with the government (execu- Awareness, and Behavior Change tive & administrative5) and communities (Buota, •• Remedial Actions to Protect the Water Lens Bonriki, and South and North Tarawa) to identify To reduce the potential for contamination or pollu- and agree on the approach to be taken and identify tion of the lens, and to put into place (as far as possi- key stakeholders6 and representatives. ble) a minimum8 depth of ground cover above the 2. Identify and gather current, relevant, useful data water lens, the following actions should be taken: (from LMD, the Attorney General, the PUB, the Fill in any existing pits more than 1 meter in depth. Ministry of Environment, Lands and Agriculture These exist as a result of excavation to extract, sand, Development (MELAD), local councils, etc.). aggregates and rocks, or to build wells and pit latrines; 3. Determine the main processes, inputs, and out- and 2. Dig up existing tracks and roads to make them puts for negotiating SMPs for both reserves. impassable; keep only the limited access routes lead- 4. Identify the person(s) who will facilitate an inde- ing to the water treatment works and other sites pendent, respected, capable 12-month (or longer) needed for maintenance and/or security. 96 Building Urban Water Resilience in Small Island Countries Note: The existing cemetery in the central (northeast) new infiltration galleries may involve increased land part of the reserve should not be moved. rental payments. Consideration may be given to the establishment of a Note. It is assumed that this activity will not be imple- new cemetery site for the Bonriki and Buota village mented. The Water and Sanitation Roadmap identi- communities outside of the water reserve perimeter. fied the estimated cost (in 2011) as $A2,500,000, and •• Increasing the Water Capacity of the Bonriki Reserve. noted that the New Zealand Aid Program was pro- Tree removal in Bonriki. Removal of about 1,700 deep- posed to fund it. rooted coconut trees from the central portion of •• Actions to Create Barriers to Accessing the Water Bonriki will increase the sustainable yield of the Reserves reserve by 250 m3/day—about 12.5 percent of the cur- Potential physical perimeter barriers and interior rent combined Buota/Bonriki yield. Only minor mod- ground-cover barriers include: ification of the existing infrastructure would be Perimeter barriers. Cost is a factor, as are longevity, required. The advantages and disadvantages of this resilience in the face of human ingenuity, and main- option are explained in the Tarawa Water Master tenance. An additional factor is whether perimeter Plan, and preliminary capital cost estimates are barrier construction and maintenance can be part of outlined. community engagement, and can result in financial Some members of the community may perceive or other benefits to the community. Walls and fences the loss of useful assets (such as timber), although are likely to be too costly, too difficult to maintain, there is evidence from community leaders and and will involve little community engagement or local politicians that the coconuts are not produc- benefit, while providing opportunities for income tive. This measure should be explored as a negoti- that result in the theft of barrier materials. The prin- ating point, testing acceptance or rejection by the ciple usually adopted is to provide multiple barriers. community at an early stage. While the benefits of Suggestions include: removing the trees are reasonably clear, the local resistance could be mitigated by employing local ° Initially marking the boundary using dead or workers, or by allowing community participation in old coconut trees on the reserve. revenues generated. If, however, resistance is sub- ° Planting rows of dense vegetation, preferably stantial, it may be preferable not to pursue this with thorns. measure. ° Erecting RWH panels on the reserve boundary, Infilling ponds, western end of Bonriki. During con- near the villages. struction of the airport runway at Bonriki, borrow ° Erecting informational and warning signs at pits were excavated at the western (lagoon) end of regular intervals along the boundary. the islet. These have introduced salinity into the freshwater lens. By cleaning the ponds of organic Interior Ground-Cover Barriers. Such barriers can pre- matter and infilling with clean, dredged sand, the vent easy access to the reserves (for sand mining, etc.), area and the sustainable yield of the Bonriki reserve either by making it easier to see intruders, or physi- could be increased by an additional 250 m /day. This 3 cally more difficult for them to gain access to the option will require negotiation with landowners and ground surface because of defensive vegetation. Three with the Bonriki community; and the installation of options have been identified (for Bonriki only): Building Urban Water Resilience in Small Island Countries 97 Cut all plant cover down to a low level. Pay the ° does it need to be in 1, 5, and 10 years? How do local community to slash and/or remove the we get there? vegetation once or twice a year. ° Objectives: overall messages and emerging Plant extensively, taking care not to affect other ° themes. permitted uses; or, plant dense, low-growing ° Identification of the various public audiences plants or bushes that will prevent access. and stakeholders in the context of the protec- Plant shrubs or trees that will not draw much ° tion and conservation of the Buota and Bonriki water from the lenses, or attract human access, water reserves, taking into account the sus- but will provide a useful product. tainable management policies, laws, regula- tions, procedures, and applicable operating •• Actions to Prevent Vandalism processes. The following actions for preventing vandalism have ° Goals, strategies, and actions. Messages to key been identified: audiences are drafted, considering the follow- Select members of the Buota and Bonriki com- ° ing questions: What do we want the public, munity to report vandalism and/or unautho- audiences, and stakeholders to know and rized land use, (working with village councils, remember? What are the goals of a SMP? How church groups, Unimwane, etc.). are they measured? How are the government Install notices around the reserve, warning of ° and the community planning to reach their penalties. goals?) Engage youth and communities in promoting ° ° Implementation, budget, performance targets, education, awareness raising, and behavior measurement, and evaluation of the imple- change. mentation phase. It is suggested that an education, awareness-rais- Medium and Long-Term Land Use ing, and behavior change plan be prepared and •• Different Approaches for Buota and Bonriki implemented, and that regulatory warning signs be Agreement as to the definitions of prohibited and displayed. The signs would provide information to permitted land use within the Buota and Bonriki the public, both on-site and off-site, of the laws (reg- water reserves will not be the same at the two ulations, by-laws, and so on), and the penalties that reserves. There are good arguments to suggest that result from contravention of the laws. the Bonriki reserve could be used productively for The development of the plan, as part of the SMP, nontraditional uses that generate revenues and/or would incorporate the following elements: that contribute to the development of South Tarawa. Education, awareness-raising, and behavior ° The existing traditional uses (the raising of coconut change processes, incorporating a “Forum” palms, etc.) are too degraded to be productive. But approach including open discussions in public there is an equivalent argument to suggest that the meetings. Buota reserve should be protected and conserved in °Background and situation analysis that its natural state as much as possible, in order to pro- addresses the following questions: Where is tect the reserve and to provide plant resources for the management of the water reserves now? Where Buota community. 98 Building Urban Water Resilience in Small Island Countries •• Land Uses Not Permitted Within the Water Reserves (1) generate revenue; and (2) provide a barrier to For both the Buota and Bonriki reserves, it is sug- accessing the main inner area above the lens? gested that only very limited human contact should Within the perimeter area, limited, noncontami- be permitted. Under no circumstances should any of nating uses could be considered such as: warehous- the following be permitted (see also table D.2):9 ing, recreational facilities, places of worship, police stations, banks, post offices, magistrate courts, craft 1. Potential contamination of the water resource with markets, and so on. In this scenario, the inner use liquid or solid wastes. could occupy, say 80 percent of the land over the 2. Any kind of toxic or contaminating liquid or solid reserve and its use could be limited mainly to open that may percolate through the soil into the water space, vegetation, and other uses that are limited and lens. that involve very little human or animal contact. 3. Any kind of hole that reduces the ground cover over the lens. Policy, Laws, and Regulations 4. Any areas of extensive hardstanding that place a Policy issues are those that may require cabinet-level barrier between rain and the lens, including build- decisions. They may include issues relating to the ings, or any other use that involves the removal of strategies to be adopted in order to implement policy. soil over the lens. Some of the policy issues may need to be addressed and resolved before the commencement of other •• Permitted Uses activities; others may arise during the process of Potential permitted uses on the water reserves are deciding the mitigation measures and the manage- briefly reviewed in table D.3. ment and governance arrangements (that is, during the preparation stage); still others will be required •• Perimeter and Inner Reserve Use (Bonriki only) before the measures and arrangements agreed upon Is it feasible, in terms of sustaining the water resource, can be initiated (i.e., before implementation). to consider an additional perimeter area, say This report has not identified the policy issues that 50 meters in width (for a total of 100 meters), in may arise during the preparation and implementa- which some limited uses may be permitted, that will: tion phases; these will be identified during the prepa- ration phase, and in concluding agreements on water TABLE D.2. Potential Prohibited Uses reserves SMPs. The policy issues that will need to be addressed Prohibited uses—To be further developed as part of water before commencement fall into three categories: reserve regulatory framework Solid waste—landfill, incineration, recycling, composting, etc. 1. Policies and laws regarding illegal activities on the Liquid waste disposal—Septic tanks, sludge ponds, wastewater water reserves. The households currently occupy- reticulation and treatment ing the water reserves are illegal in terms of Public Industrial uses—Anything involving toxins. All sand, gravel, and rock mining. Utilities Ordinance Chapter 83 of 1977. PUB has the Residential uses that may involve septic tanks, domestic animals, power to “remove any structure or fill in any pit” human waste. upon the reserves. The public participation and Commercial uses that may involve the introduction of toxins, community engagement approach proposes that septic tanks, or hardstanding negotiations take place with the households Building Urban Water Resilience in Small Island Countries 99 TABLE D.3. Potential Permitted Uses on the Water Reserves Uses Advantages, Disadvantages and Comments Buota and Bonriki Communities (including households located on the reserves) Approved access rights Low impact, small numbers of humans. Potential rights to access “fruits” of the land and policing duties. Strategic Infrastructure (for Bonriki only) Solar/Photovoltaic Panels Low impact, small numbers of humans dealing with operation and maintenance (O&M). Relatively small areas occupied. Minimal hardstanding. Industrial uses Frees land in South Tarawa for housing. Uses with minimal people onsite. Dependent on use. No toxic liquids or substances. No oils, etc. Minimal hardstanding. (For example warehouses or storage facilities with low impact.) Telecommunications masts Low impact, small numbers of humans dealing with O&M. Relatively small areas occupied. Minimal hardstanding. Electricity substations Low impact, small numbers of humans dealing with O&M. Relatively small areas occupied. Recreational uses: (for Bonriki only) Football fields Attractive to local communities. Located on perimeter, with restricted access to field only & toilets provided Volleyball, basketball, etc. Agriculture/Horticulture/Fisheries/Industrial: (for Bonriki and Buota) Medicinal plants & herbs Used by the whole of South Tarawa, but could be tended by local community Low impact as long as no fertilizers is used. Roof thatch & matting Access for pandanus leaves collection. Strictly limited to women and local residents. Commercial (Bonriki only) Possibly limited warehousing that involves minimal human presence. No septic tanks, etc. Minimal hardstanding. Existing Uses (for Bonriki and Buota) Water treatment works Existing Cemeteries (Bonriki only) Too difficult to move, and of relatively little negative impact. Dedicated access to be provided to existing facilities. No further cemetery expansion within the reserves. concerned, and with the Bonriki and Buota com- different, and it is recognized now that desalina- munities. PUB and the government may need to tion is the only available option that will help fully either decide not to enforce the law, and/or be pre- meet South Tarawa’s water demand. In these cir- pared to change the law, if the outcome of these cumstances, and bearing in mind the inevitable negotiations is an agreement between the communi- growth of South Tarawa’s population, adoption of ties and the government that requires changes in a policy of “managed decline” of the water reserves policy, law, or regulation. could be considered. 2. Policy and strategies that differentiate between the 3. A SMP involving agreement between the govern- two water reserves. The government’s policy is to ment and communities. There is no specific existing protect and conserve freshwater sources for the policy or legal framework that covers the end public water supply. Existing law reinforces this result proposed in this report, which is a legally ­ policy direction. The economic, social, and envi- binding agreement (contract) involving govern- ronmental contexts in Buota and Bonriki are quite ment authorities and communities, who may both 100 Building Urban Water Resilience in Small Island Countries have managerial and governance responsibility; government may wish to determine in advance of a and whose authorities may share revenues and consultative and facilitated process? Can the same costs. Does the government agree with this open- question be applied to costs? Can costs be shared ended approach? That is, is mutual agreement to between the government and local communities? If be achieved through a consultative and facilitated communities receive financial benefits, can they also process in order to achieve effective management be expected to share costs? and governance of the water reserves? If so, can the existing agreement be recognized in law? Is Long-Term Reserve Management D.  Governance Options there any potential conflict with existing laws? Principles Revenues and Costs Internationally, a wide range of approaches for the The mitigation measures, and the management and management of natural resources have been tried governance arrangements that will be finally agreed and tested.10 Many of these have successfully accom- upon cannot be predicted in advance. Can any of the modated traditional or customary approaches to the revenues arising from use of the Buota and Bonriki ownership and management of common property water resources, or the land and vegetation above resources with those involving state ownership and the water lenses, as well as the ocean and lagoon management. Some of the options and their poten- frontages, be shared? If so, by and with whom? Also, tial application to the Buota and Bonriki water are there any exclusions, or strategies, that the reserves is reviewed in table D.4. TABLE D.4. Water Reserves Governance Options Options Description of option Applicability: Pros and cons State ownership and The government owns the resource–in this case a This approach has been in place for 40 years. It has not management source of water–and manages the resource through succeeded in resolving the risks to the water reserves state organizations and the application of laws and created by adjacent settlement, encroachment, and regulations. illegal occupation. State ownership The government owns the resource and contracts out This is almost certainly an expensive option with limited and contracted out management to another, usually a private firm. The long-term sustainability, and probable difficulty in management state uses laws and regulations as the institutional finding a “manager” with the appropriate combination framework for the arrangement. of technical and locally applicable social skills. State ownership The government owns and manages the resource This is the most viable option given the need to and community through state organizations, or by contracting out. The combine management of different freshwater sources, involvement management arrangements include mechanisms for the importance of the water supply to a substantial community involvement in advising; in decision making; population, the opportunities and risks arising from and in receiving benefits (financial or in-kind). The aim communities living around the reserves; and the is to incentivize communities to protect and conserve complex property and resource rights involved. It will the resource. involve difficult, lengthy, and complex negotiations to achieve agreement for sustainable management. Community Ownership of the resource is transferred to the This is unlikely to be acceptable politically or ownership community, provided the community achieves administratively, as it would almost certainly result set criteria and manages the resource, receiving in unsustainable tensions between the communities the benefits—within a set of laws and regulations involved in “ownership” of the resources, and those established by the state.a benefiting from access to the freshwater supply. a. As adopted, for example, in Zimbawe’s CAMPFIRE program. Building Urban Water Resilience in Small Island Countries 101 In developing long-term reserve management apply them. But the consultative aspects, or means to and governance options, and bearing in mind their negotiating a solution (or agreement) that everyone implementation, a number of principles are thought supports and can accept and enforce are the missing to be important in determining the approach elements. The Kiribati context emphasizes: of all stakeholders for securing water resources in •• The technical (water supply) in terms of the main Tarawa generally, and to the management of these organizations responsible for water supply and use two water reserves in particular. The following prin- (O&M) of the water reserves in South Tarawa, being ciples for guiding the preparation of a SMP are the PUB and the MISE and; proposed: •• The law in terms of protection and conservation of 1. Fairness to both of the communities involved. The the water reserves—laws (and subsidiary regula- residents of South Tarawa need to have access to tions) relating to land occupation and tenure, sufficient clean water to support good health, sur- physical planning, development control, and envi- vival, and livelihoods. The communities of Buota ronmental protection—Ministry of Environment, and Bonriki that live on and next to the water Lands and Agriculture Development (MELAD). resource are giving up (in part) their land and live- lihoods, in order to protect the water resource. What is missing is an institutional and organiza- A solution that considers the different interests of tional framework for reaching agreement with all the two parties can only be found through a fair involved parties, and especially with the communi- process that brings these communities together to ties involved–both those that are on the land illegally, achieve a solution that will be supported by all. and those who have an interest in the land–as to a 2. Consultation and participation. The key to being way forward that will ensure a clean water supply for “fair to all” is full and open participation of all all of South Tarawa’s communities. stakeholders–the communities involved, govern- Reasons for public participation and community ment agencies, groups representing particular engagement. The importance of public participation sections of society, political and traditional repre- ­ in environmental management and planning has sentatives—in discussions held at all stages of the been well accepted internationally for at least 40 process that result in agreement, on decisions years. The reasons are: made. •• The opportunity to make better decisions, by open- 3. Process management. To achieve effective partici- ing up the decision-making process to a wide set of pation, consultation, and decision making, a care- people with expertise that includes local experi- fully prepared, step-by-step process that combines ence, traditional knowledge, and various forms of thorough preparation and implementation with technical knowledge. honest communication among all stakeholders is •• Better public acceptance of and compliance with the necessary. decisions made, because people have the opportu- Aspects of the Approach. The formal institutional nity to be heard, and are included in the deci- and organizational framework is reasonably clear: sion-making process. that is, the policies, laws, and regulations are well •• Social justice—the principle that those who will be defined and clarify how to apply them; and who will affected by a decision deserve to have input into it. 102 Building Urban Water Resilience in Small Island Countries In addition, community engagement enables the •• Identify the benefits. It is important to consider public to go beyond participation in a decision each participant’s perspective: Why should they and become motivated to support a new perspective or participate? How can they gain from having a say? issue, and take appropriate actions themselves. •• Commit sufficient resources to conduct the process When people are motivated to change their own well, and enable people to participate (for example, practices, they can often achieve far more than is by providing travel costs and/or replaced wages). possible through legislation, policies, or programs Practitioners generally agree that there is no single alone. “recipe” for a good participation process: it is best to Public Participation and Community Engagement customize the process according to local circum- There is a large body of international experience on stances. It is particularly useful to seek and accept good processes for public participation and commu- local advice while designing the process. nity engagement, notably Ross, Powell and Hoverman 2008. It is particularly important that the processes Sustainable Management Plans are well designed, and that they include genuine Sustainable Management Plans for the Buota and intent to listen to the public and take their advice. Bonriki water reserves are expected to: Key recommendations include: •• Be formal agreements that are signed by all stake- •• Be inclusive. The processes should be open to all holders that have the powers and responsibilities relevant members of society, including women, defined in the agreements; all age sets, and minorities. They should be com- •• Be prepared through a public participation and com- fortable for the people involved. They should suit munity engagement approach that adopts princi- their cultures, languages, and convenience, and ples and processes that are formally agreed upon be held in locations where they feel at ease to by the stakeholders in advance; speak freely. •• Be customized to take into account the different •• Balance differences in power as much as possible circumstances on the two water reserves; through preparation of the parties, providing the •• Be facilitated by one or more facilitators, either disadvantaged with appropriate resources, and jointly for both water reserves, or with separate facilitating discussions that encourage the less facilitators, one for each of the two SMPs; advantaged to speak out. Structure and content. The structure and content of •• Allow enough time for the process, so that all can be the SMPs is expected to be as a brief, concise, and reached, and heard. People may need extra time, easy to understand as possible, and may contain: and repeated visits, to understand complex issues. •• Provide facilitation. A neutral facilitator, or a team •• Vision and objectives; of facilitators, who can combine process skills with •• Governance arrangements (strategic direction and knowledge of the issue, can be helpful in the executive decision making); process. •• Organizational arrangements: day-to-day adminis- •• Build capacity. Some, or all, of the parties may trative support to the executive, and to the facilita- need assistance in order to participate effectively, tor(s), including for the management of financial through preparatory briefing and discussion. resources; Building Urban Water Resilience in Small Island Countries 103 •• Agreements and arrangements relating to: (1) 1–2 years, and will involve employing one or more households on the water reserves; (2) the Bonriki facilitators who will have primary responsibility; will and Buota communities; and (3) the South Tarawa provide critical continuity and support for indepen- (and possibly North Tarawa) communities; dence; and will be supported administratively by a •• Actions to restrict access to the water reserves; small unit attached to the Office of the President. •• Actions to optimize water resources; The preparation phase will be divided into three parts: •• Sustainable land use (prohibited, permitted, and development control). •• Initial discussions and agreements; •• Agreement resulting in a roadmap on directions Governance arrangements. Governance arrangements and actions; are arrangements that will provide for executive deci- sions regarding preparation and implementation of the •• Preparation of SMPs for the Buota and Bonriki Buota and Bonriki SMPs. Executive decisions include water reserves. those that involve strategic direction, as well as deci- Facilitator training. The key qualities of the facilita- sions involving, for example, SMP annual budgets, and tor(s) employed to develop the SMPs are that they will performance and actions involving government policy, be seen as neutral by the stakeholders, and will have a laws, or regulations. It is expected that the SMPs will reputation for good judgment and communication. include terms of reference that detail the governance It not expected that they will have previously com- arrangements. Representation in the executive body pleted a similar public participation and community may include members of parliament and local govern- engagement process successfully. For this reason, it is ment councillors representing South Tarawa, Buota, expected that: (1) they will require training in preparing and Bonriki; key government representatives community-based action plans; and (2) they will have (from PUB, MELAD, MISE, and the Attorney General’s input into the process of developing the roadmap. office); and traditional community representatives Examples of SMP preparation. A number of published (Unimwane), as well as representatives of women and guidelines provide toolkits for community engage- youth. ment and public participation in the development of Management arrangements. Management arrange- integrated water resource management plans that are ments will provide first for the preparation of the two relevant in the case of Kiribati. The processes sug- SMPs, and within the SMPs, for the subsequent gested in the guidelines are outlined in table D.5. implementation of the agreements reached in them. SMP implementation costs. Estimates for various It is expected that the preparation phase will take SMP options are provided in table D.6. 104 Building Urban Water Resilience in Small Island Countries TABLE D.5. Toolkits for Community Engagement: Content Toward Whole of Community Engagement: Integrated Water Resources Management: Planning Approach for Small Island Developing A Practical Toolkit (Australian Government) Nations (United Nations Environmental Program, UNEP) Starting Up Pillars What does good community engagement mean? The planning cycle Who is involved? Stakeholders, communities & others Stakeholder participation and dialogue Difference between engagement & consultation? Planning cycle Knowledge systems and communication strategies Step 1. Diagnostics to determine entry point Community decision makers and stakeholders Identify the entry point and lead agency; carry out stakeholder analysis; establish means of coordination and facilitation; carry out situation analysis (economic, environmental, and social The community engagement process (including gender, poverty). Values, principles, and criteria Selecting Tools and Techniques Step 2. Visioning General public involvement and participation tools Problem tree analysis, objective tree analysis Negotiation and conflict resolution tools Step 3. Strategizing Information, education, and extension tools Scenario development—selection of sustainable management plan options (framework identification, content identification, strategy preparation) Rapid and participatory rural appraisal tools Stakeholder analysis and social profiling tools Step 4 Planning Survey and interview tools Plan preparation, action plan and budget, responsibility matrix, scheduling, setting and measuring performance targets and indicators. Planning and visioning tools Team building and leadership tools Step 4. Implementation Participatory action research tools Administrative and financial arrangements (annual work plans and budgets, reporting, recordkeeping, procurement, budgeting and expenditure), adoption of policy, laws and Deliberative democracy tools regulations, capacity development. Lobbying and campaigning tools Participatory M&E tools Step 6. Monitoring, Evaluation and Documentation Data collection and monitoring, reporting. Documenting lessons learned. TABLE D.6. Matrix of Options with Cost Estimates Lead Community Community Annual Costs Initial Costs No. Government consultation— participation— (AUD) — Revenues generated (AUD)a Agencies Preparation implementation operational 1 Immediate Actions, Education, Awareness and Behavior Change 1.1 Remedial actions to protect the lens PUB/MISE Yes Preferable 100,000 10,000 Possible but limited in amount 1.2 Increasing water capacity of the Yes Essential 2,500,000b TBD Bonriki reserve 1.3 Barriers to access the water PUB/MELAD/ Yes Preferable 50,000 TBD reserves Attorney General 1.4 Prevent vandalism Yes Essential Include in TBD No 1.5 1.5 Education, awareness raising, Yes Essential 200,000 over 1st 5 years public relations & behavior change 2 Medium and Long-Term Land Use table continues next page Building Urban Water Resilience in Small Island Countries 105 TABLE D.6. continued Lead Community Community Annual Costs Initial Costs No. Government consultation— participation— (AUD) — Revenues generated (AUD)a Agencies Preparation implementation operational 2.1 Land use planning (to develop MELAD/Local Yes Essential 50,000 Not High potential (Bonriki planning policies, strategies, and Government applicable only) regulation) 2.2 Development control Yes Preferable See 2.1 25,000 Limited potential 3 Management and Governance 3.1 Implement Sustainable New institution Yes Essential 100,000 Revenues to High potential Management Plan—Bonriki pay costs 3.2 Implement Sustainable Yes Essential 100,000 Limited potential Management Plan—Buota 4 Policies, Laws, Regulations, Revenues, and Costs 4.1 Policies, laws, and regulations PUB, MISE, Yes Limited 50,000 Not applicable MELAD, Attorney 4.2 Revenues and costs Yes Preferable 50,000 As above General + new institution a. Initial and annual costs are indicative, and assume that an initial five-year budget of approximately $A500,000 is available for initial (preparation) costs. b. Source: ADB 2011. (Note: It is assumed that this activity will not be implemented, since the estimated costs are greater than the potential benefits.) Notes   1. Traditionally, resources under and on the land, such as water and  6. Key stakeholders may include: households on the reserves, the coconut trees, are owned by the landowner as part of the land. This Buota community, the Bonriki community, the South Tarawa com- customary ownership can be in conflict with the approach taken in munity, PUB/MELAD, local and national political representatives, states governed by laws that reserve certain natural resources, both and others (representatives of airport, etc.). below and above ground, to the State.   7. “Red lines” may include specific aspects of government policies as   2. Bonriki – Temaiku Consultations. South Tarawa Sanitation well as the provisions mandated by laws and regulations. The gov- Improvement Sector Project Quarterly Report No. 1. January 2013. ernment will need to identify which are nonnegotiable, and which Snowy Mountains Engineering Corporation. may involve flexibility. Resolution of these concerns should not delay the start of consultations; but the final agreement will need the  3. In the Bonriki water reserve, the potential to harvest coconuts, support of all stakeholders. toddy, and babwai on the reserves has mostly or even completely dis- appeared. Many residents believe this is because of excessive   8. To be defined. extraction of freshwater.   9. Access is required for operation and maintenance of the water galler-   4. At the end of 2014 the total annual payment to Buota and Bonriki ies, water treatment works, and for policing of the site and mainte- landowners was around $A1 million. nance of site protection works.  5. Politicians (cabinet, ministers, members of parliament) and civil 10. A theoretical argument for the management of common resources servants. may be found in Ostrom (2009). 106 Building Urban Water Resilience in Small Island Countries SKU W19051