87052 THE WORLD BANK GROUP MONGOLIA: HEATING STOVE MARKET TRENDS IN POOR, PERI-URBAN GER AREAS OF ULAANBAATAR AND SELECTED MARKETS OUTSIDE ULAANBAATAR Stocktaking Report of the Mongolia Clean Stoves Initiative With generous support from the Australian Government TABLE OF CONTENTS Abbreviations and Glossary of Mongolian Terms vii Acknowledgements ix Executive Summary xi Key results xiii Road Map xvii Chapter 1. Introduction 1 1.1 Study Context 1 1.2 Air Pollution in Ulaanbaatar 2 1.3 Low-emission Stove Initiatives in Mongolia 3 1.4 Study Background and Objectives 8 1.5 Study Method and Data Sources 8 1.6 Structure of this Report 9 Chapter 2. Housing and Household Socio-Economic Information 11 2.1 General Characteristics of Ger Districts 11 2.2 Type and Tenure of Dwelling 13 2.2.1 Characteristics of Gers 13 2.2.2 Characteristics of Detached Homes 14 2.3 Characteristics of Households in the Ger Areas 15 2.4 Household Experience with Banking and Credit 16 2.4.1 Experience with Credit/Loans 17 2.4.2 Knowledge of Lending Products and Banking Services 17 2.4.3 Knowledge and Interest in Mobile Banking 18 2.5 Conclusion 18 Chapter 3. Heating Stove Ownership and Use 21 3.1 Total Number of Heating Stoves Owned and Used 21 3.2 Age of Stoves Used During the Previous Winter Months 23 3.3 Heating Stove Use by Dwelling Type 23 3.3.1 Heating Stoves Used in Gers 24 3.3.2 Heating Systems and Stoves Used in Detached Houses 24 3.4 Low-emission Stove Ownership and Use 26 3.4.1 Low-Emission Stove Ownership in Different City Districts 27 3.4.2 Characteristics of Households that Bought Low-Emission Stoves 29 3.5 Conclusion 30 Chapter 4. Household Preferences and Perceptions of Heating Stoves and Air Pollution 33 4.1 Heating Habits 33 iii Stocktaking Report of the Mongolia Clean Stoves Initiative 4.2 Cooking Habits 35 4.3 Household Preferences and Perceptions of Heating Stove Performance 36 4.3.1 Preferred Characteristics and Design Features of Stoves 36 4.3.2 Households’ Perceptions of the Performance of Their Existing Stoves 37 4.4 Willingness to Change Stoves 39 4.4.1 Reasons for Resistance to Changing Stoves 40 4.4.2 Perceived Difficulties with Using Low-Emission Stoves 40 4.5 Source of Information and Knowledge about Low-Emission Stoves 41 4.6 Household Attitudes toward Air Pollution 42 4.7 Conclusion 43 Chapter 5. Heating Fuel Consumption and Expenditures 45 5.1 Types of Fuels Used by Households 45 5.2 Total Household Heating Fuel Expenditures 47 5.3 Estimated Coal Usage and Expenditures 50 5.3.1 Comparison of Coal Usage between Households Living in Different Types of Dwellings with Different Heating Systems 51 5.3.2 Comparison of Coal Usage between Low-emission and Traditional Stoves 52 5.4 Estimated Firewood Consumption and Expenditure 52 5.5 Conclusion 54 Chapter 6. Heating Stove Demand and Supply in Four Selected Cities 57 6.1 Study Background and Objectives 57 6.2 Study Methodology and Data 57 6.3 Overview of the Four Aimag Centers Studied 57 6.4 Characteristics of the Demand for Stoves in the Aimag Centers 59 6.4.1 Demand for Low-Emission Stoves and Potential Willingness to Pay 59 6.4.2 Demand for Low Pressure Boilers (LPBs) 60 6.5 Characteristics and Limitations of Stove Supply Chains 60 6.6 Linkages of Stove Markets in Ulaanbaatar and the Aimag Centers 63 6.7 Perceptions of Air Pollution in the Aimag Centers 65 6.8 Conclusion 65 Chapter 7. Conclusions and Implications for Stove Replacement Efforts 67 References 73 Annex A: Sampling Plan 74 TABLE OF FIGURES Figure 2.1. Ger-Area Households in Surveyed Districts, by Primary Winter Residence 13 Figure 2.2. Size of Ger (As Indicated by Number of Walls), 2012 13 Figure 2.3. Insulating Felt Coverage of Gers 14 Figure 2.4. Proportions of Detached Houses with Different Types of Heating Systems 15 Figure 2.5. Sizes of Detached Houses with Different Types of Heating Systems 15 Figure 2.6. Share of Households with At Least One Member with a Bank Account 17 Figure 2.7. Type of Bank Account or Other Financial Account Held by Households 17 iv Table of Contents Figure 2.8. Source of Credit or Financing to Ger-Area Households 17 Figure 2.9. Purpose of Loan 17 Figure 2.10. Share of Households that Have Heard of Lending Products Specifically for Heating and other Household Energy Products 18 Figure 3.1. Number and Age of Stoves 23 Figure 3.2. Heating Systems by Dwelling Type, Ger-Area Households, 2012 24 Figure 3.3. Percentage of Traditional and Low-Emission Stoves Used for Heating in Gers 24 Figure 3.4. Heating Systems and Stoves Used in Detached Houses, 2012 24 Figure 3.5. Survey-Estimated Market Share of Low-Emission Stoves, by Dwelling Type 26 Figure 3.6. Income of Customers Purchasing Low-Emission Stoves from XacBank 30 Figure 4.1. Percentage of Households Using Different Stoves for Cooking During Winter 35 Figure 4.2. Appliances Used for Cooking by Traditional and Low-Emission Stove Owners During the Winter* 36 Figure 4.3. Stoves Used for Cooking by Female Respondents During the Winter* 36 Figure 4.4. Importance of Stove Design Features to Ger-Area Consumers 37 Figure 4.5. Rating of Stove Heating Abilities 38 Figure 4.6. Views of Stove Performance among Traditional and Low-Emission Stove Users 38 Figure 4.7. Household Reasons to Not Change Their Stove 40 Figure 4.8. Perceived Difficulties with Using a Low-emission Stove (Comparing those who have and have not used low-emission stoves already) 40 Figure 4.9. Where Did You Hear about Low-Emission Stoves? 41 Figure 4.10. Knowledge about Low-Emission Stoves 42 Figure 4.11. Opinion of Air Pollution Problem 43 Figure 5.1. Sources of Coal Used by Households 47 Figure 5.2. Heating Fuel Expenditure as Percentage of Monthly Income 48 Figure 5.3. Comparison of Average Coal Consumption per Household by Income Categories 51 Figure 5.4. Comparison of Average Coal Consumption per Household by Dwelling Type and Heating System 51 Figure 6.1. Location of the Four Cities included in the Rapid Assessment of Stove Markets 58 Figure 6.2. Stoves for Sale in Central Markets of Darkhan and Bayankhongor 61 Figure 6.3. Artisan Welder in Bayankhongor 61 Figure 6.4. Low Pressure Boilers in the Aimag Centers 63 Figure 6.5. Market Structure and Supply Chains of Stoves outside Ulaanbaatar 64 Figure 7.1. Reductions in Population-Weighted Exposure to Air Pollution in Ulaanbaatar Resulting from Lower Ger Stove Emissions 68 TABLE OF TABLES Table 1.1. Average annual concentrations of particulate matter in Ulaanbaatar 2 Table 1.2. Selected List and Subsidized Low Emission Stove Prices, Winter 2011-2012 4 Table 2.1. Socio-Economic Information of Households in the Ger Areas 16 Table 3.1. Total Number of Stoves Owned in Ger Areas, by Stove Type, 2012 22 Table 3.2. Total Number of Stoves Used in Ger Areas, by Stove Type, 2012 22 Table 3.3. Average Age of Stoves in Ger Areas, by Stove Type, 2012 23 Table 3.4. Low-emission Stove Sales by District and Heating Season 28 v Stocktaking Report of the Mongolia Clean Stoves Initiative Table 4.1. Supplemental Heating Devices 34 Table 4.2. Number of Times per day Households Fire their Stoves and Add Fuel 34 Table 4.3. Percentage of Households that Keep their Stove Burning Continuously in the Morning, Afternoon, Evening, and Night 35 Table 4.4. Preferred Replacement Models for Traditional Stove Users 39 Table 5.1. Heating Fuels Used by Households 46 Table 5.2. Raw Coal and Processed Fuel Use by Households in Six Districts 46 Table 5.3. How Often Households in Bayangol District Use Processed Fuels 46 Table 5.4. Average Household Monthly Spending on all Heating Fuels by Income Quintile 48 Table 5.5. Average Household Monthly Spending on all Heating Fuels by Types of Dwelling and Heating System 49 Table 5.6. Average Household Monthly Spending on Heating Fuels by Types of Stove 49 Table 5.7. Household Coal Usage and Expenditure by Income Quintile (September 2011 through April/May 2012) 50 Table 5.8. Household Coal Usage and Expenditure by Type of Dwelling and Heating System (September 2011 through April/May 2012) 52 Table 5.9. Household Coal Usage and Expenditure by Type of Heating Stove (September 2011 through April/May 2012) 53 Table 5.10. Household Firewood Usage and Expenditure by Income Quintile (September 2011 through April/May 2012) 53 Table 5.11. Household Firewood Usage and Expenditure by Type of Dwelling and Heating System (September 2011 through April/May 2012) 54 Table 6.1. Estimated Technical Potential of Demand for Improved Heating Stoves in Four Select Cities 59 Table A.1. Total Number of Households (HH) and Population of Selected Districts and Living in Ger Area at the End of 2011 74 vi ABBREVIATIONS AND GLOSSARY OF MONGOLIAN TERMS ADB Asian Development Bank aimag Administrative region of Mongolia, equivalent to a province ASTAE Asia Sustainable and Alternative Energy Program AQS Air Quality Standard CAF Clean Air Fund CDM Clean Development Mechanism CHP Centralized Heating Plant CSI Clean Stove Initiative duureg Administrative district of Ulaanbaatar City EAP East Asia and Pacific ger Wood-framed, felt-covered dwelling traditionally used by nomadic households hashaa Small fenced-in plot typically occupied by a single household or family IDA International Development Association (The World Bank Group) khoroo Administrative subdivision of Ulaanbaatar City below the level of duureg kW Kilowatt LPB Low Pressure Boiler LPG Liquefied Petroleum Gas MCA Millennium Challenge Account MCC Millennium Challenge Corporation MJ Megajoule MVR monitoring, verification and reporting PM Particulate Matter SEET Stove Emissions and Efficiency Testing soum Administrative division below the level of aimag, equivalent to a county or township Tg Mongolian Togrog (currency) UBCAP Ulaanbaatar Clean Air Project WHO World Health Organization μg Microgram The findings, interpretations and conclusions expressed in this report are entirely those of the authors and should not be attributed in any manner to the Australian Government, the World Bank, or its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publications and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, other information shown on any map or figure in this volume do not imply on the part of the World bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. vii ACKNOWLEDGEMENTS This report is a summary of activities conducted in ger areas of Ulaanbaatar to access heating for the initial phase of the Clean Stoves Initiative appliances producing less particulate matter (CSI) in Mongolia. Funded with generous emissions. It will also provide technical assistance support from the Australian Government, the for the development medium-term particulate CSI is forum for promoting knowledge-sharing matter abatement measures in coordination with and collaboration to improve household access development partners, including city greening, to modern cooking and heating solutions in East district heating, affordable housing, and power Asia and the Pacific. The initiative is comprised plant emissions controls. of four country-specific programs, in China, The World Bank would like to thank its Indonesia, Lao PDR, and Mongolia. counterparts for the CSI in Mongolia, the The initial phase of the CSI builds on the unique Ministry of Energy and Municipality of experience that Mongolia has already gained in Ulaanbaatar, for their excellent cooperation and providing access to clean heating technologies oversight of the CSI. The stocktaking benefited and will share lessons learned with the other greatly from the inputs of a wide range of countries. Activities for the CSI stocktaking stakeholders, for which the authors are most consisted of a survey of 1,000 households in grateful, including the National Committee poor, peri-urban ger areas of Ulaanbaatar and on Reducing Air Pollution (NCRAP), Clean a rapid appraisal of market trends in four other Air Fund, Ministry of Environment and Green cities, Darkhan, Ondorkhaan, Bayankhongor, Development, Asian Development Bank, Japan and Khovd. The 1,000-household survey International Cooperation Agency, Millennium provides an updating of the World Bank’s Challenge Corporation, Millennium Challenge baseline survey of heating stove market trends Account-Mongolia, XacBank, Khentii Aimag in Ulaanbaatar, presented in Mongolia: Outputs Governor’s Office, Khovd Aimag Governor’s of the stocktaking, including this report, are Office, local universities, heating stove retailers, intended to inform the development of a national heating utility executives, ger-area leaders, and strategy in Mongolia for providing access to many others who contributed their time and clean household energy for heating and cooking energy. during the next phase of CSI. The preliminary The co-authors of the report are Mr. Gailius J. findings of the stocktaking were shared with Draugelis, Mr. Voravate Tuntivate, and Mr. Chris stakeholders in December 2012. The draft report Sall. The report was prepared under the guidance was disseminated for comment at a workshop in of Mr. Draugelis, Task Team Leader of the CSI Ulaanbaatar held in January 2013. in Mongolia and UBCAP. Mongolia Marketing In addition to managing the CSI, the World Bank Consulting Group LLC (MMCG) carried out is also supporting the Ulaanbaatar Clean Air the surveying work in Ulaanbaatar from July to Project (UBCAP) with a $15-million soft credit August 2012. The rapid appraisal in Darkhan, financed by the International Development Ondorkhaan, Bayankhongor, and Khovd was Association, the World Bank’s concessional completed in August 2012. Mr. Robert van der assistance arm. UBCAP will enable consumers Plas and Ms. Wang Xiaoping provided expert ix Stocktaking Report of the Mongolia Clean Stoves Initiative peer review and comments. Mr. Dejan Ostojic Bank; Ms. Badamkhorloo, former UBCAP provided overall guidance as coordinator of the Project Management Unit (PMU) Director; Mr. East Asia Clean Stoves Initiative, World Bank. Enkhbold, current Director UBCAP PMU, and Mr. Otgonbayar Yadmaa and Mr. Nomuuntugs Ms. Tsenduren, UBCAP PMU Manager, for Tuvaan provided invaluable logistics support. Ms. their stalwart support and advice throughout the Zolzaya Tuguldur provided much appreciated process. Finally, the Mongolia CSI team thanks logistical and publications support. The team Mr. Dejan Ostojic, Coordinator of the overall gives special thanks to Ms. Coralie Gevers, regional CSI program and guidance provided by World Bank Mongolia Country Manager, Mr. Mr. Mark Lundell, Sector Manager, China and Mark Lundell, Sector Manager, China and Mongolia Sustainable Development. Mongolia Sustainable Development Unit, World x EXECUTIVE SUMMARY Ulaanbaatar is the coldest capital of the world and campaigns about health problems associated remains one of its most polluted. Coal and wood with air pollution. burning for heating of individual residences in The report relies primarily on results of a survey ger areas are essential for survival but contribute among 1,000 households in ger areas of the about 60 percent of the fine particulate (PM2.5) six central districts (duureg) of Ulaanbaatar concentrations in the city. These levels of conducted in July 2012 and a rapid assessment exposure are very harmful to health and exceed of the market for heating stoves in four cities World Health Organization (WHO) standards outside Ulaanbaatar performed in August 2012, many-fold. as well as information received from XacBank The heating appliances causing the pollution are and the Millennium Challenge Account-Mongolia both traditional stoves that have been used for (MCA-Mongolia) on stove sales in a recent stove generations and, increasingly, coal fired stove- replacement program. Results from a similar furnaces (called Low Pressure Boilers or LPBs) market assessment by the World Bank in 2007- used by wealthier households when constructing 20091 are used as a baseline and backdrop for larger homes. The overwhelming majority of insights into the changes that have happened to households in the ger areas of Ulaanbaatar the market since then. (informal settlements surrounding the city), With funding support from the Australian however, are poor, and the population continues Government, this study is part of the World to grow as job prospects in Ulaanbaatar attract Bank’s East Asia and Pacific (EAP) Clean more migrants. The World Bank estimates Stove Initiative (CSI) comprising four country- that a reduction of 80% of emissions from ger specific programs (China, Indonesia, Lao PDR, area heating could achieve a 48% reduction and Mongolia) and a regional forum promoting in population weighted exposure to PM2.5. collaboration, learning, and knowledge-sharing To achieve this, poor households need to be on access to modern cooking and heating convinced to permanently switch to less polluting solutions at the household level. The EAP CSI heating solutions, an effort that will require follows a 2011 report, One Goal, Two Paths: a multi-year, coordinated set of policies and Achieving Universal Access to Modern Energy in programs. East Asia and the Pacific, which presents the This study takes stock of recent developments broader goals and work to be done to provide and provides market information on affordability, universal access to modern energy in the attitudes, fuel consumption, and other market region. Achieving this goal in the EAP Region information for stoves and fuels inside and will be challenging given that more than 1 outside Ulaanbaatar. It provides insights for billion people-or every second household in the solutions to the important challenges that remain Region-lack modern cooking/heating solutions. to achieve a sustainable market transformation to low-emission stoves. Other programs on related 1 World Bank, Mongolia - Heating in Poor, Peri-urban Ger issues could benefit, such as those for new Areas of Ulaanbaatar. ASTAE (Asia Sustainable and Al- fuels and insulation, as well as public awareness ternative Energy Program) Report (Washington, DC: World Bank, 2009). xi Stocktaking Report of the Mongolia Clean Stoves Initiative Achieving the universal access to modern about 55 percent of all households living in the energy services by 2030 is also the goal set by central ger areas of Ulaanbaatar had purchased the UN, which declared the Year 2012 as the low-emission stoves, including 69 percent of Year of Sustainable Energy for All. If this goal is those households in the targeted khoroo where to be achieved by 2030, access to modern energy subsidized stoves were sold. The MCA-Mongolia must be scaled up massively. funded stove replacement project was results- based, providing consumer subsidies that The findings presented in this study capture disbursed only upon sale and verified installation some of the initial outcomes of a recent stove of eligible stoves (see chapter 1). The program’s replacement program in the Ulaanbaatar rapid sale of new, imported stoves differentiated ger areas. As part of a major effort to reduce it from programs in other countries that outdoor air pollution in Ulaanbaatar, the donor have experienced a more gradual market community organized small pilots in 2009-2011. development and have included some support of In 2011-2012, MCA-Mongolia financed a scale- local stove supply. As there are many pathways up of its pilot program in selected sub-district that point to a solution-such as focusing on (khoroo) of the five districts that comprised either local production or imports from regional Ulaanbaatar City’s Air Pollution Reduction Zone. or even global suppliers-the experiences and The Clean Air Fund (CAF) provided additional developments of the market for low-emission subsidies for stoves after the launch of the stoves in Ulaanbaatar form an important case scale-up in 2011. An unprecedented 97,877 low- study for the design and support of cleaner emission stoves were sold from June 2011 to cooking and heating solutions internationally. November 2012. By the end of November 2012, xii KEY RESULTS The results of the survey among Ulaanbaatar storeys; (ii) stoves attached to heating walls in ger area households and the rapid assessment detached homes; and (iii) LPBs in detached of stove markets outside Ulaanbaatar lead to the homes with radiators in most rooms and/or conclusion that the significant results of stove piped hot water systems. Most households use replacement efforts to date could be jeopardized their heating stoves to do most of their cooking, if it is discontinued. There has been some especially during the winter months, although discussion of curtailing the program after 2013 some households with LPBs also have another due to its large sales figures, but results indicate cooking appliance. that if this happens, households are likely to Since 2007, the number of households in the six revert to purchasing and using traditional stoves. central ger area districts has grown by 36 percent The demographic forces driving growth in or 43,990 households. According to the household the demand for stoves in the ger areas are still survey, about 44 percent of households have a at play. Supply chains for low-emission stoves ger in their hashaa (a small fenced-in plot), about continue to be heavily dependent on the subsidy 42 percent have a detached home of one or two program, and the traditional stove market, storeys, and about 14 percent have both a ger and though clearly reduced in Ulaanbaatar, continues a detached home. The average size of dwellings to thrive outside Ulaanbaatar without subsidies. has increased in all market segments since 2007, If a longer-term intervention is not put in place, correlating with increases in income (see chapter the number of traditional stoves used by ger- 5). The average floor area of a five-walled ger is area households will start pushing upward again 28m2. The averaged detached house without a along with emissions. The following five findings heating wall that relies on a space-heating ger support this conclusion and summarize the key stove has a living area of 39 m2. The average insights presented in this report. home with a heating wall is slightly larger, with an area of 41 m2, and homes with LPBs are the Finding 1: Demographics and household dwelling largest (73 m2). Fuel use correlates with dwelling situations are changing size; ger homes reported using 3.9 tons of coal One of the main contributions of this report is per year, compared to 4.84 tons for homes with the detailed household, dwelling, and stove use heating walls and 6.3 tons for homes using LPBs information on households in the six Ulaanbaatar (see section 3.2.) ger areas, presented in chapters 2 and 3. In As people become wealthier and move to larger general, compared with 2007, the markets have dwellings, they also upgrade heating amenities. grown in absolute numbers and have changed Heating walls continue to be the most common in line with evolving household preferences for form of heating system, although the proportion larger and more comfortable housing, driven by of detached homes with heating walls dropped rising incomes. Markets for heating appliances, from 70 percent in 2007 to 62 percent in 2012. This both in and outside Ulaanbaatar, consist of three market share was almost entirely taken by LPBs, distinct market segments: (i) stoves used in gers considered the high-end of the coal-fired heating and detached homes of one or occasionally two xiii Stocktaking Report of the Mongolia Clean Stoves Initiative appliance market (and a different technology from than traditional stove users (13 percent) (section traditional stoves). The portion of detached homes 4.2). Among the top difficulties perceived by with LPBs rose from 16 percent to 23 percent. users of low-emission stoves are the need to Homes with stoves without heating walls, a step make a fire twice per day (70 percent compared above living in gers, maintained a stable share at to 45 percent of traditional stove users), 15 percent of total detached-home households. difficulties with refueling (60 percent compared The survey furthermore shows that around 67% of to 42 percent of traditional stove users) and time traditional stove users are interested in switching required to heat dwelling (44 percent compared to a low-emission stove (section 4.3) and that to 25 percent of traditional stove users). This fewer households are buying traditional stoves- is particularly important because the survey the average age of traditional stoves in use by results also indicate that the ability of the stove households has risen significantly to 9 years, from to heat a room and stay warm (meaning also less 6 years in 2007. Because stoves have an average refueling, not a popular task) is seen as its most lifecycle of 10 years, many of the remaining important aspect by consumers. Significantly, households with traditional stoves will likely respondents who said they had learned about buy replacement stoves in the near term. There low-emission stoves from friends, neighbors, or continue to be large, un-served segments of the relatives were about 35 percent less likely to be ger-area market that use relatively more coal. low-emission stove owners. If the performance of the low-emission stoves is not well-appreciated, it Finding 2: Consumer preferences must guide may be very difficult to get more households to stove development, marketing, and stove use them. New entrants to the market will need replacement initiatives to be careful not to overpromise or under-deliver on stoves to avoid a negative feedback loop. The ger household heating and cooking preferences and habits, presented in chapter 4, As market segments more amenable to current can inform the development, market entrance, low-emission stove offerings saturate, the stove or introduction of new stove types. New switching program can face stronger head technologies are more readily accepted by ger- winds selling to households less inclined to buy area consumers-and more properly used-when them, such as households in detached homes they align more closely with the traditional using heating walls. Heating walls, which use patterns of cooking and heating behavior. the flue gas from the stove to heat a larger area, are self-constructed or made by local artisans. Although four different types of stoves are Installation of low-emission stoves with heating offered in the current program, room for more walls, requiring the purchase of connectors, competition exists. Apart from producing has been challenging (see section 3.2). Another less pollution, low-emission stoves are also area that needs more attention is the increased seen by their users as saving fuel, one of the use of low pressure boilers (LPBs), which are characteristics that consumers ranked as being like coal fired furnaces distributing hot water most important in their preferences for stoves through pipes. Some LPBs are pre-fabricated in (section 4.2.1). However, while about two-thirds China while others are made by local artisans. of low-emission stove users in the survey were LPBs are completely different technologies satisfied with the ability of their new stoves than stoves and need to be tested and evaluated to heat their homes, the same was not true for separately to set minimum requirements for their cooking. Only 52 percent of low-emission stove emissions performance in the next phase of stove users agreed their stoves were easy to cook with, replacement activities. compared to nearly 90 percent of traditional stove users. The survey also reveals that during the Finally, related to fuel use (discussed in chapter winter, households with low-emission stoves tend 5), if policies to introduce alternative fuels such to cook more with electric hot plates (33 percent) as semi-coke coal are pursued, it is imperative xiv Key Results that the alternative fuels are tested, and that the Results of the rapid assessment of markets results are made public, so that estimated impacts outside Ulaanbaatar indicate that clear linkages on emissions can be evaluated against program exist between the stove market in Ulaanbaatar costs. Mandating new fuels may also require and markets in other cities, especially the introducing new types of stoves that are designed aimag centers. Even if traditional stove makers specifically for use with those fuels. If the right in Ulaanbaatar go out of business, producers stove is not paired with the right fuel, emission outside Ulaanbaatar could potentially fill the reduction benefits may be negated. So far, the market with cheaper, traditional stoves when government has provided significant support to subsidies on low-emission stoves are lifted. It is alternative fuel producers (especially for semi- also likely that some of the low-emission stoves coke coal). Yet, the survey reveals that only 5 purchased by households in Ulaanbaatar will percent of all households in the surveyed ger flow out of the city through resale markets, and areas reported using semi-coke coal on a regular that those households will buy traditional stoves basis, including only 49 percent of households in as replacements. The structure of the stove the Bayangol District, where raw coal is formally markets outside Ulaanbaatar is similar to those banned. Comparatively, 98 percent of households in Ulaanbaatar, but they are of a smaller scale in the six districts report using coal, while 6 and more dispersed in nature. These market percent burn dried animal dung. characteristics challenge the viability of local production of low-emission stoves, unless costs In sum, while the current low-emission stove are significantly reduced. Stoves are supplied by offerings have been accepted by the market, a artisanal producers, who would have difficulties wider range of product offerings will be needed ensuring consistent quality when producing to maintain the program’s penetration rates. the high quantities that are needed to address the also growing concerns over air pollution Finding 3: Heating fuel expenses remain a high exposure in the provincial capitals (aimag proportion of monthly income, especially for very centers) heard during the rapid assessment. low-income families A nationwide stove replacement program could Although incomes have risen and the spending help address the risk of traditional heating on fuels as a share of monthly income has stoves out-competing low-emission alternatives. declined by 5 percent overall, heating fuel The existing distribution channels for stoves expenses continue to be very high, especially and materials between Ulaanbaatar and outside when compared with subsidized district markets could be exploited to distribute low- heating enjoyed by residents in the city center. emission stoves from larger producers or Households in the bottom income quintile spent importers in Ulaanbaatar. 31 percent of their monthly income on heating fuels during the winter of 2011/12 (compared Finding 5: Technical capacity and materials for to 42 percent in the winter of 2007/08). By maintenance of low-emission stoves are needed comparison, households in the top quintile spent only 6 percent of their monthly income Private entrepreneurs must be attracted to on heating (chapter 5). All households-whether provide low-emission stoves at lower costs. better-off or poor-continue to be exposed In addition, necessary skills and spare parts to the extremely high levels of particulate inventories for the repair and maintenance concentrations. of the new, low-emission stoves need to be developed so that their supply and use can be Finding 4: Markets in and outside Ulaanbaatar sustainable. As discussed below as part of the are intricately linked, suggesting the possible need road map, skills of the artisanal producers could for a scaled-up program to reach areas outside the be upgraded and programs addressing private capital sector needs could be adjusted. xv ROAD MAP The road map first presented in the 2009 World than fuel consumption alone, and set interim Bank report, describing a market-based approach performance targets that can eventually be for introducing new technologies that can burn accepted, after practical experience is gained, fuel (new or old) more cleanly in Ulaanbaatar, as new standards, following the Mongolian remains largely relevant today and offers a regulatory process. Mongolia has three main framework for understanding good practices standards (MNS5216:2002, MNS 5041:2001, and assessing options to address sustainability MNS 5679:2006) with some updates for challenges of the current program. The seven stoves, boilers and furnaces, and solid fuels- steps of this road map-with new information all of which could be improved. Areas for based on recent findings-are: improvement are summarized in the World Bank’s Mongolia: Heating in Poor, Peri-Urban • Estimate the impact of the proposed mitigation Ger Areas of Ulaanbaatar (October 2009). measures on the overall air pollution to ensure the government’s investment will meet Standards and certification mechanisms need to expectations. The World Bank’s Mongolia: Air be developed after sufficient practical experience Quality Analysis of Ulaanbaatar: Improving is obtained with testing and product performance Air Quality to Reduce Health Impacts in the field. A rush to legal standards may (December 2011) estimates that a 30% to 80% inhibit innovation, or worse, set ineffective or emissions reduction from ger area heating can unrealistic parameters. Households can deviate produce a 18% to 48% reduction in population from stove manufacturer’s instructions, which weighted exposure to PM2.5. Under the affect stove emissions performance. These Ulaanbaatar Clean Air Project (UBCAP), the deviations can happen especially if the stove city will start to implement a program that, differs significantly from traditional products. among other requirements, that proposes to A more pragmatic approach to standards set a maximum of 70 micrograms of PM2.5 development would divide tasks into building per net megajoule (MJ) of heat emitted for blocks: (i) using stove replacement activities ger stoves, for access to subsidies. This is to introduce interim standards, or eligibility about a 90% reduction in emissions compared criteria, as a starting point for developing to traditional stoves (1,000 μg PM2.5/net MJ). standards; (ii) establishing sufficient, and Technical assistance for air quality analysis in preferably local, laboratory capacity to set Ulaanbaatar currently provided by the Japan initial emissions requirements and measure International Cooperation Agency (JICA) compliance during program implementation and an evaluation of the MCC-funded stove- (perhaps, too, establishing links with labs switching project being conducted by MCC outside Mongolia may help build capacity); (iii) and MCA-Mongolia may provide an updated building awareness and understanding of the forecast of impacts of mitigation measures. impacts of various parameters that could be included in the eventual standards and emission • Develop a testing program for equipment, requirements by publishing testing reports and with a focus on safety and emissions rather explaining results to stove producers, taking xvii Stocktaking Report of the Mongolia Clean Stoves Initiative into account variations in performance due to • Set an enforcement mechanism to address possible “misuse” of stoves; and (iv) adjusting non-compliant products and manufacturers. protocols and testing interpretations based on Stove switching eligibility criteria for feedback from household use (using surveys, new stoves and vendors should include observations, repeated lab tests, and field tests warranty requirements and minimum after- of emissions). Monitoring data will provide sales service standards. (Warranties of two confidence that the eventual standard is realistic, years were provided by low-emission stove achieves its objective, and is enforceable in a cost vendors that received subsidies from MCA- effective manner. However, while field testing Mongolia and CAF.) During the warranty is important, it is very difficult to obtain reliable periods, participating producers should agree results; field testing results should be interpreted to random spot checks of stoves sold (i.e. by experienced analysts. Also, all work towards random lab testing) and monitoring of their the development of standards, including stove service performance. In addition, they must testing and monitoring, should involve the agree to a penalty system for non-compliance, Mongolian Agency for Standardization and including permanent exclusion from the Metrology, who will be ultimately responsible for program in cases of persistent violations. developing and enforcing the eventual standard. Further, as local sellers of imported stoves Finally, inputs from private stove developers, might not be available, a likely need will exist households, and other stakeholders should also for skilled workmen and the availability of be collected and considered during the standards good quality spare parts to service the low- development process. emission stoves. Under UBCAP, technical assistance is available through a new Stove In 2011, an Asian Development Bank-financed Development Center, which could be a technical assistance project implemented by source for training. Partnerships with the the Ministry of Mineral Resources and Energy imported stove producers are encouraged established a Stove Emissions and Efficiency so that appropriate knowledge is with local Testing (SEET) Laboratory, which conducted service providers after warranties expire. The over eighty tests of various stove models, center can also provide short term technical including some of the Turkish and Chinese-made support for producers encountering design or models that were subsidized by MCA-Mongolia manufacturing problems. and CAF. Based on this work, Ulaanbaatar City developed eligibility criteria for stoves and • Establish a targeted and well-justified subsidy producers. These criteria already address some program to make new systems affordable. of the deficiencies in the current set of standards, Subsidy arrangements should consider: and require that stoves and fuels be tested (i) the end point policy objective of the together for emissions. Although the SEET lab government, which in this case is reducing did not receive support after the ADB program air pollution (especially particulate matter) was completed and is temporarily dismantled as exposure; and (ii) the consumers’ willingness of the writing of this study, the lab will continue to switch to new products. Although a well- to receive support from the City of Ulaanbaatar designed stove replacement program usually through UBCAP, supported by the International brings large net benefits in reduced health Development Association, the concessional risks and fuel savings, the right amount and assistance arm of the World Bank. MCA- types of subsidy will depend on the ability of Mongolia conducted field tests but the results are the market to sustain the transformation to not publicly available at the time of writing this the use of cleaner stoves. Based in part on report. Once available, these testing results will the available data presented in this report, be valuable inputs to understanding variations in key principles of an appropriate subsidy emissions between lab and home use. program include: (i) the ability of consumers xviii Road Map to choose among several stove models, as centers in designated khoroo (see chapter not only the stove price (which was in fact 1) to control quality and prevent the sale of ranked lower than other positive features “copies” - similar but lower quality versions. of low-emission stoves; see chapter 4), but XacBank, the main distributor of MCA- also by fuel savings and other features; (ii) Mongolia and CAF-subsidized stoves, is the payment of subsidies only upon verified also implementing a Clean Development installation of the new equipment; (iii) the Mechanism (CDM) project for stove possibility of rewards for the use of stoves switching. The CDM requires strict that achieve even larger emission reductions monitoring of stove installation and use, (i.e. considering that the government is following an agreed protocol. It also requires willing to subsidize products to achieve independent validation to verify emissions public objectives, consumers should be reductions. rewarded for using stoves that achieve Although sales are an important milestone, proportionally better public outcomes); the final development outcome is achieved (iv) the use of subsidy levels that reflect only by the sustained use of the low-emission the affordability of target households and stoves among area households. Although their willingness to start using another households sign an agreement promising to type of stove before the end of their current use the subsidized stove, low-emission stoves stove’s useful life; and (v) the use of a final are priced below market prices of traditional consumer’s price that is not too low (or free) stoves, creating an incentive for reselling. to ensure stoves are sold to people who will Evidence of large-scale reselling does not use and maintain it. exist, however, perhaps because people who During the winter of 2011-2012, the stove might buy re-sold stoves expect the subsidies replacement program included subsidies of to continue. It is extremely difficult to measure around 249,800 Tg to 408,450 Tg (equivalent the occurrence of leakage. The study team did, to US$195-319) for low-emission stove models. however, identify a few low-emission stoves Subsidy amounts included two separate in the aimag and soum centers, outside the elements, financed by MCC/MCA-Mongolia program boundaries of Ulaanbaatar. To further and later by CAF. The subsidies reduced the mitigate this risk of low-emission stoves not sales price of low-emission stoves to roughly being used by the purchasing households in 20-25 percent of the price for traditional ger Ulaanbaatar, possible mechanism include: (i) stoves, not including accessories. The prices of verifying installation and training consumers in larger low-emission stoves ranged from 30-40 using the stoves before releasing the subsidy percent of traditional stoves used with fittings for (already applied in the current program); (ii) heating walls. The pricing strategy employed by using site visits to customers during the winter the recent stove replacement program achieved (either by producers or by program staff) to its goal of high and rapid sales rates; however, verify stoves are in use, provide supplementary sustaining such levels of demand without the training, and check on customer satisfaction; high subsidy elements is unlikely, leaving and, (iii) establishing a transparent reporting sustainability challenges for the next phase of protocol linked to regular monitoring, providing the program. information to stakeholders about program developments and possible adjustments needed. • Establish credible supply chains, in partnership with the private sector, and use a “third party” • Remove old equipment from use. To achieve verification system for installation and use sufficiently large and sustained emission of new stoves. The recently implemented reductions, the support for cleaner stove replacement program used a clear and alternatives must happen in parallel with credible supply chain, involving product the removal of polluting heating systems. xix Stocktaking Report of the Mongolia Clean Stoves Initiative The July 2012 survey reveals an estimated subsidies. Concerns of the private sector 212,708 stoves are owned by 164,127 should be effectively addressed in the next households, of which only 172,055 stoves phase of the program. are in use. The recent stove replacement Although the current program already includes program required submission of old stoves and has tested features of market based as part of the sales transaction, but program approaches, gaps remain. The program can be participants noted that about 7 percent of supported by the following policy options: households received exemptions, and that the stoves that households turned in tended • Continue the program, including to be in worse condition as the sales season maintaining credible supply chains, progressed.2 Stove replacements need to verification procedures, and consumer be carefully monitored because households subsidies. A publicly announced may be tempted to provide unused stoves (or timetable to scale down the subsidy even poor-condition stoves purchased from could encourage households on the fence other households) to meet the requirement. to participate; however, curtailing the While no fail-safe method exists to ensure program too quickly could jeopardize honest trade-ins, mitigation measures include results. the independent verification mechanism discussed above, as well as continued public • Publish eligibility criteria for stoves and awareness campaigns on the health impacts producers. The current program did not of the pollution caused by the older stoves. publish eligibility criteria. Consequently, the private sector in Mongolia was only • Market-based approaches to disseminate able to participate by selling imported cleaner heating systems are preferred over stoves (the only products that were nonmarket-based approaches such as stove identified as being eligible for subsidies). giveaways. Broad consensus seems to exist Publishing criteria and undertaking a that artisanal stove manufacturing is not transparent and objective evaluation and going to have sufficient capacity to design enforcement processes will be critical and supply the market with low-emission to encouraging greater private sector stoves quickly enough and at a large enough participation. scale to bring down unit costs to parity with traditional stoves. The recent stove • Encourage development of technical replacement program collaborated with standards for a wider range of products, foreign producers and Mongolian private including heating walls, low pressure sector distributors to achieve very rapid boilers, and combinations with new fuels. penetration rates. The partnership with the This can be done through support of the private sector is critical to the success of SEET Laboratory. any market-based program as it provides the • Develop skills and spare parts supply chains professional know-how, customer orientation, for servicing current stoves in partnership and innovation needed to provide good with the current import suppliers. customer services and respond to market demand. There have been reports that major • Continue to monitor and evaluate private sector participants are no longer program impacts through household interested in participating, which presents surveys, air quality monitoring, and a major risk to the program, especially as stakeholder consultations. On the latter, it faces challenges to find ways to support coordination among all stakeholders, a transformed, clean stove market without donors and government agencies, is a critical factor to ensuring focused and 2 Author’s phone interview with XacBank representative, sustainable outcomes. November 1, 2012. xx CHAPTER 1. INTRODUCTION 1.1 Study Context With funding support from the Australian practical applications of intervention strategies Government, this study is part of the World to develop sustainable, cleaner stove markets- Bank’s East Asia and Pacific (EAP) Clean through experience with stove replacement Stove Initiative (CSI) comprising four country- efforts in Ulaanbaatar-and to develop a scalable specific programs (China, Indonesia, Lao PDR, strategy, if agreed with counterparts, for and Mongolia) and a regional forum promoting extending stove replacement to other key stove collaboration, learning, and knowledge-sharing markets in Mongolia. Mongolia-CSI counterparts on access to modern cooking and heating are the Mongolian Ministry of Mineral Resources solutions at the household level. and Energy and, given the dominance of the city’s stove market, Ulaanbaatar Municipality. The CSI follows a 2011 World Bank report, One Goal, Two Paths: Achieving Universal The first phase of the Mongolia-CSI takes stock Access to Modern Energy in East Asia and the of the current market situation in Ulaanbaatar Pacific, which identified the necessary work for and other select cities. Phase 1 in particular governments in the EAP region that must be focuses on stove replacement activities in carried out on two paths simultaneously, toward Ulaanbaatar City, including the MCC-funded meeting the enormous challenge of providing project, which is the most significant market universal access to modern energy. On the intervention in the city’s recent history. It also first path, achieving universal electricity access included a rapid appraisal of the market for requires accelerating both grid and off-grid heating stoves in four provincial capitals (aimag programs while employing appropriate policies centers). Phase 1 will be completed after the and innovative technical solutions to reduce dissemination of this study and participation in a costs, improve reliability, and provide timely regional cross-country workshop in March 2013. service to all households. On the second path, a major push is needed to increase access to clean Upon agreement with the two counter parts, fuels and advanced cooking and heating stoves the second phase of the Mongolian-CSI will be for poor, primarily rural, households that are launched in 2013. Phase 2 will likely consist of likely to continue relying on solid fuels for most a series of activities to monitor and generate of their cooking and heating needs beyond 2030. knowledge on stove replacement efforts in Achieving universal access to modern energy Ulaanbaatar, as well as support engagement services by 2030 is the challenging goal set by on developing a strategy to scale up activities in the UN in 2012 as part of its Sustainable Energy Ulaanbaatar to other stove markets in Mongolia. for All initiative. These activities could include: (i) periodic assessments of the stove market in Ulaanbaatar; Closely aligned with the regional program (ii) developing South-South and North-South activities, the Clean Stove Initiative for Mongolia networking and knowledge sharing among (the Mongolia-CSI) was launched in March stove program practitioners (such as dialogues 2012. Its objective is to strengthen knowledge of among technical and scientific communities, 1 Stocktaking Report of the Mongolia Clean Stoves Initiative twinning arrangements and other collaborative population densities in the world, with only models, and study tours and distance learning 2.8 million people and a land area equivalent to activities) to build stronger linkages among Britain, France, Germany and Italy combined. relevant scientific and technical communities These low economic densities create high in Mongolia, international counterparts, and, infrastructure costs and barriers to modern stove replacement program participants; and services; in 2007, only about 54 percent of rural (iii) developing a national clean stove scale up households report to have access to electricity.3 strategy, including stakeholder consultations and About 68 percent of the country’s people live in the study of related issues such as institutional urban areas, a share which is a high compared arrangements and long term alternatives to to other East Asian countries. Almost all of current heating and cooking solutions that might this is attributed to the growth of Ulaanbaatar, be attainable as Mongolia’s economy continues its Mongolia’s capital. The city’s 1.2 million fast-paced growth. residents account for more than 40 percent of Mongolia’s total population, and more than 60 Mongolia has made great strides in promoting percent of its urban population. energy access through rural electrification, for example by providing off-grid access to solar Ulaanbaatar remains nearly unlivable at certain home systems for herders. This study aims to times during the winter due to severe air provide empirical evidence and lessons learned pollution. Pollution of very small particulate to support the replacement of traditional heating matter (PM), which is inhaled deep into the appliances with affordable, cleaner alternatives. lungs and causes major health damage, is severe. Improving household access to modern energy In a recent World Bank air pollution study, the forms part of a broader engagement by the worst recorded annual average concentration was World Bank with the Mongolian government, more than 10 times higher than the Mongolian private sector, and donor partners on priorities Air Quality Standard (AQS) for PM10 and 25 for Mongolia’s energy sector. times higher than the Mongolian AQS for PM2.5.4 These numbers are extremely high, especially 1.2 Air Pollution in Ulaanbaatar considering that the Mongolian AQS is already 2-3 times higher than WHO standards.5 As shown Mongolia is a large country with severe winters in table 1.1, concentrations in ger areas (where and a small population. Temperatures across the vast majority of the urban poor live) are more the country during winter time routinely fall severe than in the city center. below -20oC and can reach below -40oC on the coldest days. The country has one of the lowest Table 1.1. Average annual concentrations of particulate matter in Ulaanbaatar345 PM10 PM2.5 Exceedance Ratio to Air Quality Standards Area μg/m3 μg/m3 Mongolian WHO Central city areas 150–250 75–150 3–6 7–15 Ger areas 350–700 200–350 7–14 17–35 3 World Bank and Public-Private Infrastructure Advisory Facility, Mongolia: Foundation for Sustainable Development: Re- thinking Infrastructure Services in Mongolia (Washington, DC, June 2007). 4 World Bank, Mongolia: Air Quality Analysis of Ulaanbaatar: Improving Air Quality to Reduce Health Impacts (World Bank Sustainable Development Series Discussion Paper, December 2011). 5 The Mongolian annual ambient air quality standards are 50 μg/m3 and 25 μg/m3 for PM10 and PM2.5, respectively, while the WHO interim targets for developing countries are 70 μg/m3 for PM10 and 35 μg/m3 for PM2.5. See World Bank, Mon- golia: Air Quality Analysis of Ulaanbaatar. 2 CHAPTER 1. Introduction The extreme seasonal variability in air quality- City of Ulaanbaatar in five districts: Bayanzurkh, the summer months are relatively clean while Chingeltei, Khan-Uul, Songinokhairkhan, and wintertime air quality is one of the worst Sukhbaatar. Additional subsidies were provided recorded anywhere-reduce annual average by the Mongolian Clean Air Fund (CAF). A concentrations and divert attention from the total of around 68,850 low-emission stoves were fact that Ulaanbaatar is one of the world’s sold and delivered to households in the five most polluted capital cities. The alarming PM districts from June 2011 to June 2012.8 By the concentrations in Ulaanbaatar lead to significant end of November 2012, 97,877 stoves had been health impacts, valued at between US$177 and sold,9 reaching 55 percent of all the 177,933 US$727 million each year (with an average of households living in the central ger areas of US$463 million).6 Emissions from heating in Ulaanbaatar as of the end of 2012,10 including 69 ger areas comprise about 60 percent of total percent of those households in the khoroo that estimated contributions to population-weighted were targeted by the MCA-Mongolia project. exposures to fine particulates on an annual The low-emission stoves were imported from basis, but can be much higher, even extreme, Turkey and China and sold exclusively (with during winter months in certain parts of the some exceptions) through dedicated distribution city. For example, the monthly average PM10 centers called “Product Centers.” The product concentration measured at a railway station centers were established in selected khoroo. in a ger area (the Zuun station) was 1,850 μg/ Each household was eligible to buy one m3 in January 2009, while the four highest daily subsidized stove, at the authorized product average measurements at the same station were center in its khoroo. Registries of households in the range of 3,612-4,360 μg/m3. The highest living in the khoroo, maintained by khoroo PM2.5 concentrations were measured at the administrators, served as the master list for Bayanhoshuu station, also in a ger area, where eligible households that could purchase stoves. the monthly average figure was about 1,500 μg/ m3 and the five highest daily concentrations were The majority of the stoves (about 88 percent) in the range of 2,310-4,060 μg/m3.7 were distributed through product centers operated by XacBank, a commercial bank with 1.3 Low-emission Stove Initiatives in roots in micro-finance in Mongolia that was Mongolia contracted by MCA-Mongolia for its services. The bank supplied the sales and support As part of a major effort to reduce the outdoor air staff and offered a micro-loan to finance the pollution in Ulaanbaatar, the donor community purchase of the stove if a household could not organized a number of small pilots in 2009-2010. pay for the stove with its own funds. XacBank Pilots were financed by the Asian Development staff initiated the sale in the product center, Bank (ADB), Deutsche Gesellschaft for Internationale Zusammernarbeit (GIZ, formerly 8 As noted below, no stoves were sold in Bayangol District due to a ban on raw coal there. GTZ), United States Millennium Challenge 9 MCA-Mongolia sales data provided to authors via email Corporation (MCC), the World Bank, and World with MCC representative, December 27, 2012. Vision, and implemented in partnership with 10 Overall market share is presented as the percentage of the Ministry of Mineral Resources and Energy ger-area households in the six districts covered by the survey that purchased low-emission heating stoves as and City of Ulaanbaatar. In 2011, with financing of December 2012. The total number of households in- from the MCC, the United States Millennium cludes those households with low-pressure boiler sys- Challenge Account–Mongolia (MCA-Mongolia) tems as well households in khoroo not targeted by the MCC-funded project, with the understanding that these launched a scaled-up stove-switching project in households should nonetheless be counted among those “air pollution reduction zones” designated by the that have yet to acquire low-emission heating appliances. The total number of households reflects official statis- 6 World Bank, Mongolia: Air Quality Analysis of Ulaanbaa- tics for the ger-area population as of the end of 2011, the tar. latest data available as of the time of writing. See Ulaan- 7 Ibid. baatar Statistics Department, http://statis.ub.gov.mn/. 3 Stocktaking Report of the Mongolia Clean Stoves Initiative checking household eligibility and signing a user producers provided training to household on agreement with each household, which inter how to use its new stove. Delivery was confirmed alia included a commitment by the household to in the field by staff (via an innovative mobile use the stove according to owners’ instructions. SMS verification system developed by XacBank). Importantly, the household as required to hand Costs of XacBank services were covered in their old stove with the purchase of a new substantially by MCA-Mongolia. XacBank has one, to remove polluting stoves from use. The initiated a Clean Development Mechanism household made payment at a XacBank branch, (CDM) project that is expected to finance the which transferred funds to the stove producer product centers in the future. upon verified delivery. A subsidy was then Households were offered a subsidized price transferred by MCA-Mongolia to the producer for the stoves. Stove models were on display at also upon verification of delivery and submission the product centers with price sheets and other of documentation. XacBank also managed the informational materials. Price sheets displayed receipt and destruction of the old stoves and, the actual price, the discount, and the final price. upon confirmation of payment, the distribution of Table 1.2 gives the listed and final prices of stove the new stove, which was delivered to household models sold during the winter 2011/12 sales soon after purchase. Upon delivery, the stove season. Table 1.2. Selected List and Subsidized Low Emission Stove Prices, Winter 2011-2012 MCA-Mongolia Mongolia Clean Air Stove Model List Price Customer Price Subsidy Fund Subsidy (Tg) (Tg) (Tg) (Tg) Silver Mini 325,000 250,700 50,000 24,300 Silver Turbo 459,250 208,450 200,000 50,800 Royal Single 275,000 209,800 40,000 25,200 Royal Double 330,000 245,000 40,000 45,000 Note: Clean Air Fund subsidy was introduced after the start of the sales season; Source: authors’ interviews and correspondence with XacBank representatives, Ulaanbaatar, Mongolia, March 2012. Not including accessories, after-subsidy prices for The four stove models sold under the MCA- smaller capacity low-emission stoves (Silver Mini Mongolia project constituted the bulk of stove and Royal Single) were roughly 70-75 percent switch-outs in the years and months leading below the market price for a traditional ger stove up to the July 2012 survey. Most of the stoves in Ulaanbaatar; the after-subsidy prices for the (the Silver models) were supplied by a Turkish large models (Silver Turbo and Royal Double) manufacturer and imported and distributed were about 60-70 percent lower than the going by a private Mongolian company with strong price for a traditional ger stove with a heat wall distribution capabilities. Other stoves (the fitting.11 Households also needed to purchase Royal models) were manufactured in China. chimneys and heating wall connecters for certain No domestically-manufactured stoves were models, which added costs of up to around 25,000 eligible under the program. MCA-Mongolia did Tg to costumers during the initial months of the not disclose its selection criteria for stoves and program but were eventually given for free. producers.12 12 11 Traditional ger stove prices are estimated on the basis of This does not imply that selection criteria were absent, asking prices quoted by stove traders at Narantuul Mar- or that MCA-Mongolia did not test the emissions per- ket in Ulaanbaatar, and in the four other cities visited by formance of stoves. MCA-Mongolia tested 14 models of the study team. domestic and imported stoves that were commercially vi- 4 CHAPTER 1. Introduction Small pilots were conducted during the 2010- additional insights and develop stove and producer 2011 winter season under an ADB technical eligibility criteria, which were incorporated into assistance project13 and the MCA-Mongolia the Ulaanbaatar Clean Air Project (UBCAP), a project. The separate ADB and MCA-Mongolia US$15 million IDA soft credit project, implemented pilots tested the use of a subsidy mechanism to by the city of Ulaanbaatar to sustainably increase sell more-efficient, emission-reducing heating low emission stove penetration after the MCA- stoves to ger-area consumers via neighborhood Mongolia program expires in September 2013 or product centers. The ADB pilot, designed and earlier (see box 1.1). launched jointly with XacBank and World Vision The UBCAP stove emissions criteria are publicly International, sold a model of “improved stove” available.15 Although emission factors under model that was developed by GTZ and produced laboratory settings can diverge from those in in Mongolia; MCA-Mongolia piloted the sale the field,16 for the purposes of this study, the of imported stoves (see box 1.2 for definitions). stoves sold under the Mongolia-MCA program The product centers, located near high-traffic are referred to in the report as low-emission areas such as water kiosks and bus stops, were stoves. During the stove switch-out scale-up, also used to sell insulating covers to improve the MCA-Mongolia conducted field testing of the thermal efficiency of gers.14 stoves with technical input of Lawrence Berkeley The ADB project, implemented by the Ministry of National Labs, but the testing results are not yet Mineral Resources and Energy, also established publicly available. the Stove Emissions and Efficiency Testing To test alternative solutions to stoves, the (SEET) Laboratory. It was the first-of-its-kind stove government declared “no raw coal zones” in emissions testing facility in Mongolia, capable Bayangol District. MCA-Mongolia and CAF of reliably measuring particulate emissions. subsidized stoves, which are designed to About eighty tests were performed on various burn raw coal, were not allowed to be sold in models and a testing protocol was developed and Bayangol.17 The alternative fuel to raw coal independently evaluated by a European testing offered to Bayangol households is mainly semi- company. The SEET lab tested the Turkish coke coal briquettes (semi-coke mixed with stoves, showing that the stoves’ emissions were a binder; if lay is used, ash content can be as significantly less than those of traditional stoves, high as 50 percent18). However, the Bank team provided households used the stove according is unaware of any reliable, publicly available to the users’ manual. All heating stoves have two emissions tests performed using semi-coke coal fueling stages-lighting and re-fueling. Traditionally, briquettes in traditional stoves. Stoves should be households would add more coal onto an ongoing matched for a specific fuel to ensure the fuel is fire. The new Turkish stoves required a cold re- burned as completely as possible.19 In the case fuel for best results, a deviation from traditional refueling practices. The ADB project helped gain 15 Emission testing in the field is much more difficult than in the laboratory; however, communications from the able. Testing was performed by the Mongolian Universi- testing team seemed to confirm that field and lab tests ty of Science and Technology. MCC and MCA-Mongolia, yield approximately the same results. 16 written comments provided to World Bank authors, Feb- This was identified as a risk in the safeguards report for ruary 16, 2013. UBCAP. Ulaanbaatar Services Improvement Project II 13 The ADB project partnered with the Ministry of Mineral Project Implementation Unit, “Mongolia: Ulaanbaatar Resources and Energy, Ulaanbaatar City, World Bank, Clean Air Project Safeguards Report” (internal manage- GTZ (now GIZ), and WorldVision International (an ment report, September 29, 2011). 17 NGO). For details and project documents, see “43177- Product centers were still operated in Bayangol to sell 012: Ulaanbaatar Clean Air, Project Data Sheet,” http:// other energy efficiency products, such as ger blankets www.adb.org/projects/43177-012/details. and vestibules. 14 18 XacBank was an early pioneer of the product center So far, no test results are available. Author’s interview model, having first operated centers in ger neighbor- with representatives from MAK Mongolyn Alt Corp, hoods during the winter of 2009-2010. XacBank, “Market World Bank office Ulaanbaatar, Mongolia, April 8, 2008. 19 Mechanisms for MCA Energy and Environment Project” World Bank, Mongolia: Heating in Poor-Peri Urban Ger (presentation, Ulaanbaatar, Mongolia, March 15, 2010). Areas. 5 Stocktaking Report of the Mongolia Clean Stoves Initiative that fuel switching does not produce significant The MCA-Mongolia program funded new emissions reductions, the stoves will need to be purchases until the end of November 2012. The switched to match the new fuel. CAF and UBCAP are expected to continue stove replacement activities to maximize low emission stove penetration and address sustainability issues. Box 1.1: Ulaanbaatar Clean Air Project (UBCAP) The objective of the Ulaanbaatar Clean Air Project (UBCAP), launched in 2012, is to develop selected medium-term particulate matter abatement measures in Ulaanbaatar in coordination with development partners. The project supports the implementation of short-term measures and the development of medium- term measures through three components: (a) the implementation of a stove and low pressure boiler replacement program to reduce particulate matter emissions in the ger areas; (b) the preparation of action plans and policy recommendations to reduce particulate matter in central Ulaanbaatar through district heating, city greening, and other medium-term abatement measures; and (c) public awareness raising and coordination. The project is financed with a US$15 million soft credit from the International Development Association, the World Bank’s soft credit arm. As part of the first component, particulate matter mitigation in ger areas, the UBCAP will cover as many households as possible in the stoves and low pressure boiler market segments that were not covered by previous subsidy programs, and will provide additional market development support needed to help sustain the clean stove technology market. The UBCAP stove replacement program will use the sales and distribution mechanism used by MCA-Mongolia and CAF to maintain continuity and avoid duplication. The UBCAP subsidy program will have clear working arrangements between potential financing sources prior to implementation. The proposed main implementation and institutional arrangements for stove replacement include: • Stove and vendor eligibility. Selection criteria for stoves and vendors are summarized below. Emissions performance will be tested at the Stove Emissions and Efficiency Testing (SEET) laboratory. The Project Management Unit (PMU) for UBCAP will determine the eligibility of stove models and producers, based on the selection criteria in the operations manual, approved by the Project’s Steering Committee. An annual review of the selection criteria is planned, including involvement of stove producers, consumers and other stakeholders. The PMU is encouraged to regularly consult producers and stakeholders on the criteria. A Participation Agreement shall be signed between the PMU and each eligible stove producer, defining the rules for participation in the UBCAP stove replacement program. Key Stove Eligibility Criteria o Fuel: Fuel choice is open; however, the stove and fuel combination needs to be tested. o Capacity of stoves: More than 3 kW average over test conditions with up to 90% of fuel burned; for houses, with or without heating wall, 5-9 kW depending on size; low pressure boilers, 5-15 kW. o Particulate emissions: 70 mg/net MJ for PM2.5 (net MJ of heat emitted from the stove). o Carbon monoxide emission: 7 grams/net MJ (complying most of the time with the Mongolian National Standard 5679:2006, i.e., 2% CO/CO2 emissions). o Cooking time: a cooking vessel of 400 mm diameter with 8 liter water capacity should reach boiling condition between 30 and 60 minutes after the fire is lit (from cold start). o Workmanship and durability: safe handling and adequate strong components for at least 5 years. Key Stove Producers’ Criteria o Legal entity: Producers must have a valid business license. o Participation Agreement: Producers must sign a Participation Agreement which confirms agreement with terms and conditions of the project, including penalties for non-compliance. o Production capacity: Producers must be able to supply a minimum of 100 eligible stoves within a four week period after signing a supply contract. 6 CHAPTER 1. Introduction o Quality control: Producers must sign a participation agreement which gives permission to UBCAP PMU to periodically or at random intervals verify quality and performance of stoves produced and give permission to publish production and price statistics. o Warranty: Producers must provide a warranty to the user, guaranteeing that the stove will be repaired or replaced during a 2 year period following the purchase, if there are defects, breakage or component malfunctions due to manufacturing defects. o Safeguards: Producers must demonstrate compliance with project safeguards requirements. o Training Customers: Producers must develop a basic user manual for their stove and deliver a brief hands-on training to the purchaser when the stove is installed. • Regulating supply chain through Product Centers. Eligible vendors sell eligible stoves through Product Centers based generally on a consignment business model. This will reduce the risk of copy-cat models sold outside the accepted distribution channel. Eligible models will be shown in the Product Centers, similar to those operated under the MCA-Mongolia project. • Subsidy Mechanism. Eligible stoves will be sold at a subsidized price. The subsidy levels are calculated based on the following principles: (a) equalize the price of the new stove with the price of traditional model (basically, only one model is used in all households); (b) a temporary, additional subsidy to promote more rapid adoption; and (c) a third subsidy, not piloted but proposed for UBCAP, to buy back stoves currently in use. The subsidy is expected to be based on experience gained from the MCA-Mongolia stove replacement project. Criteria which will guide the final value of the UBCAP subsidy may include: (a) consumers should have the freedom to purchase the eligible stove of their choice; (b) consumers should be rewarded proportionally, to the extent possible, for using the stoves that have better emission properties; and (c) consumers’ price for the new stoves should not be too low or zero (i.e., no giveaways). The elements considered in the subsidy are: (a) the subsidy should promote a rapid penetration of stoves in a large part of the market (e.g., 90%); (b) the subsidy should include payment for the old stoves consumers will be asked to provide as part of this switch-out program; (c) the pricing provided should not be smaller for eligible models than those already offered, signaling to the market that low prices will not be sustained. • Financing Mechanism. Households will purchase stoves at Product Centers. Households may pay the balance out of pocket or from a XacBank micro-credit. XacBank is utilizing an ongoing Dutch grant-financed project intended to bring down costs of micro-loans to offer financing for household contributions. Payment to the stove vendors is made in two stages: household contributions are first transferred directly to vendor accounts by XacBank. Subsidies are then transferred upon households meeting the following conditions: (a) signed purchase agreement between HHs and the Project (including XacBank) agreeing to purchase and use the stove for purposes intended, hand over the old stove, and monitoring and verification; (b) household counterpart contribution paid in full (XacBank receipt); and (c) verification of installation of the stove in the household. Funds flow would be managed through the commercial bank’s internal funds management systems, which can be audited easily. • Training, monitoring, verification and reporting (MVR). A two stage MVR system is envisioned. XacBank plans to verify installation of the stoves in households because it intends to use the sales toward carbon credits. A consulting firm, most likely an NGO, will be selected to verify stove installation, check on proper operation, and perform quality control checks on the supplied products. The quality control mechanism includes a ‘three strikes’ system including: (a) a warning to first-time violators; (b) temporary exclusion from the program and public disclosure for second-time violators; and (c) removal from the program for three-time violators of the participation agreement. 7 Stocktaking Report of the Mongolia Clean Stoves Initiative Figure B1.1 Overview of UBCAP financing and stove distribution Client chooses product and is issued an invoice that they take to the Branch to pay Product Client deposits or takes a loan Center to pay contribution at local Branch and takes receipt to Product Center ÕàñBank Home l r al e s ts fo eek u w req ea u b sidy ts onc s s roduc mit p Product Sub vered Directly deposits Center e l i d client contribution Client turns in receipt and old stove (if PIU pays bulk subsidy applicable, fills out a customer agreement to each producer Producer form and receives product PIU based on XacBank Account NGO collects old stoves on weekly basis submission and sells them to scrap metal producer 1.4 Study Background and Objectives market in Ulaanbaatar and how clean stoves might be able to be sold outside Ulaanbaatar In 2007-2009, the World Bank conducted a series of stocktaking activities to establish policy • To share lessons learned with other countries guidance and program design recommendations on stove switch-out programs, in particular for transforming Ulaanbaatar’s stove market. other countries involved in the World Bank’s Included in these activities was a 1,000 household multi-phase technical assistance project, survey conducted between December 2007 “East Asia and Pacific Clean Stoves Initiative”, and February 2008. A follow-up survey of 1,000 namely China, Indonesia, and Lao PDR. households was conducted in July and August The report makes use of analytical reports listed 2012. This report takes stock of the household below in addition to the Bank team’s experience heating situation in the central ger districts of gained from the preparation of the World Bank- Ulaanbaatar and draws comparisons with the supported Ulaanbaatar Clean Air Project (IDA 2007/2008 baseline. The report also summarizes credit of $15 million) and overall engagement findings of a rapid assessment of stove markets on air pollution reduction efforts since mid- outside of Ulaanbaatar. 2007. Given the expected evaluations that shall The report’s objectives are: be conducted by the Millennium Challenge Corporation and other donors, the Ulaanbaatar • To describe the characteristics of the stove clean stove work is likely to generate a unique markets in and outside Ulaanbaatar in set of analytical work supporting the low- Mongolia in July 2012, comparing these to emission stove agenda in many other countries. the characteristics of the stove market in 2007/2008 before major efforts were made to promote low-emission stoves. 1.5 Study Method and Data Sources • To provide insights for the design of stove The study relied on both primary and secondary switching programs, especially focusing on sources of data and information. The main characteristics of a sustainable clean stove primary sources of data are: 8 CHAPTER 1. Introduction • A household survey conducted in six ger as data from the Ulaanbaatar Statistical Department districts including Khan-Uul, Bayanzurkh, and Mongolia Population and Housing Census. Bayangol, Sukhbaatar, Chingeltei and In addition, the study makes use of ADB reports Songinokhairkhan; from its own “Ulaanbaatar Clean Air Project”, as well as several recent reports including Heating in • Company sales data for individual customers Poor, Peri-Urban Ger Areas of Ulaanbaatar (World and households that have purchased low Bank Asia Sustainable and Alternative Energy emission stoves from XacBank, and official Program, December 2009), Mongolia: Air Quality sales statistics from MCA Mongolia; Analysis of Ulaanbaatar: Improving Air Quality to • A rapid appraisal conducted in four selected Reduce Health Impacts (World Bank Sustainable cities, Darkhan, Ondorkhaan, Bayankhongor, Development Series Discussion Paper, December and Khovd; and 2011); and the Project Appraisal Document: Ulaanbaatar Clean Air Project (February 2012, • Interviews with government officials, experts, Report No. 66081-MN). The Bank team worked government officials, and other stakeholders in close collaboration with the Ulaanbaatar Clean in the field. Air Project Management Unit, which represented Secondary sources of data comprise of a range of the city of Ulaanbaatar as the counterpart for the published information, including reports for the Ulaanbaatar part of the Mongolia CSI, and with the World Bank’s Ulaanbaatar Clean Air Project and Ministry of Energy, which is the counterpart for Ulaanbaatar Services Improvement Project, as well the overall Mongolia CSI. Box 1.2: Terminology In this report, unless otherwise noted: • Stoves refer broadly to traditional space-heating stoves used in gers and detached homes, as well as low-pressure boilers. • A low-emission stove refers to a clean-burning, energy-efficient stove that meets criteria for indoor and outdoor air pollution, safety, and combustion and fuel efficiency. • An improved stove refers to an improved version of a traditional stove and is not considered a low-emission stove because it does not meet all of the criteria described above. • Coal refers to raw coal. 1.6 Structure of this Report also explores households’ willingness to replace their traditional stoves. Chapter 5 discusses The structure of this report reflects the household consumption of heating fuels and directional organization of the study. Chapter expenditures. The chapter offers detailed 2 offers a detailed demographic profile of analysis of differences in fuel consumption by households in the ger-districts, including socio- households with and without low-emission economic characteristics, dwelling types, stoves, as well as usage rates for semi-coke and past experience with banking and credit. coal in areas where raw coal has been banned. Chapter 3 discusses stove ownership and market Chapter 6 assesses heating stove demand and penetration rates for low-emission stoves in stove supply chains, as well as linkages between the ger areas. Chapter 4 provides insight into the stove markets in Ulaanbaatar and the aimag household perceptions of heating stoves and air centers, drawing from the findings of the rapid pollution, along with household heating habits appraisal conducted in four select cities. An and preferred stove design features. The chapter overall conclusion is presented in chapter 7. 9 CHAPTER 2. HOUSING AND HOUSEHOLD SOCIO-ECONOMIC INFORMATION This chapter provides detailed demographic of about 42,000 since the end of 2007.20 The and socio-economic information about the actual number of households is likely higher, households living in the six ger districts/areas at around 185,000, because of new arrivals included in the survey. Because type of dwelling and households not counted in the registries is closely associated with the type of home maintained by khoroo administrators.21,22 Given heating system and fuel type used, the analyses the size of the ger areas and current economic of the type and size of the homes and type of status of most residents, it is foreseeable that heating stoves and heating system are described these informal settlements will persist for many in this chapter. Details on the kinds of insulation years to come. used for different types of dwelling unit are also The ger areas consist mainly of small fenced-in provided, since heating requirements during plots (hashaa) with felt-covered tents (ger) and/ the winter months are closely associated with or simple detached homes. Although the vast the level of insulation of the dwelling unit. majority of households (more than 95 percent) Additional information on the socio-economic in the ger areas have access to grid electricity, profiles of the owners or occupants and none have running water. Ger-area households information of household experiences with must purchase water at the 550 or so kiosks banking and credit are presented in the last that are scattered throughout ger areas. Most section. This is because any promotion of the households are located within 100-500 meters of heating stove improvement program would the nearest kiosk. Apart from indoor plumbing, likely involve investments in new heating stoves households also lack sewer connections, so and/or other heating equipment by home most people rely on individual outhouses owners or occupants. in their hashaa and public bathhouses. The characteristics of the six ger districts surveyed 2.1 General Characteristics of Ger Districts are similar to other ger areas in Ulaanbaatar The ger areas of Ulaanbaatar, occasionally (e.g., Nalaikh, Bagakhangai, and Baganuur). referred to as informal settlements, have been a magnet for rural to urban migration since 20 This does not include ger-area households living in apart- even before market transition. Ulaanbaatar’s ments, dormitories, or other improved housing. 21 According to stakeholders from the Mongolian Clean ger areas have grown considerably over the Air Fund (CAF) and National Committee for Reducing past two decades, both in terms of population Air Pollution (NCRAP), presentations at “Conference for and area, and continue to grow. According to Sustainability of Ulaanbaatar Air Pollution Reduction Ac- tivities,” workshop hosted by Millennium Challenge Cor- official numbers from the Ulaanbaatar Statistics poration and Millennium Challenge Account-Mongolia, Department, at the end of 2011 about 164,127 Ulaanbaatar, Mongolia, January 29, 2013. 22 households were living in the ger areas of the The total number of households in the surveyed areas is reported as 164,127 throughout the report to be consis- six districts surveyed for this study, an increase tent with existing official estimates. 11 Stocktaking Report of the Mongolia Clean Stoves Initiative 12 CHAPTER 2. Housing and Household Socio-Economic Information 2.2 Type and Tenure of Dwelling Note: Figure includes only those villages and districts represented in the sampling frame for the 2012 survey. Where do ger area households live during the See Annex A for description of sampling. Estimates for winter months? Do they own or rent? What is the total number of households by dwelling type for 2007 based on shares of dwelling types observed in 2007 survey the condition of insulation? These are important and extrapolated to the sampling frame of the 2012 survey. questions for heating and cooking stove market Population is as of December 2007 and December 2011. development. The size of dwellings and the level of insulation largely determine the type and capacity 2.2.1 Characteristics of Gers of stove needed (if stoves are too powerful, people will use less fuel in them which can reduce Nearly all ger households are home owners. combustion efficiency and increases emissions). About 92 percent of the households living in gers Together with income (which was also surveyed during the winter heating season own their ger; and will be presented later in the report), the type only a small minority (3 percent) is renting. The and tenure of the dwelling are indicators of the remaining 5 percent live rent-free by some other quality of life households have come to expect, arrangement. including cooking and heating preferences (also surveyed and presented later in the report). The average size of ger homes has increased slightly since the December 2007 survey. The The survey reveals that about 44 percent of size of ger homes is reflected by the number of households have a ger on their hashaa, and about walls they have: the more walls, the larger the 42 percent have a detached home with one or two ger.23 In 2007, about 29 percent of gers had four stories. The remaining 14 percent have both a ger walls, and 64 percent had five or more walls. As and a detached home. Households with both a shown in figure 2.2, in 2012, about 25 percent of ger and detached home tend to live in their ger gers had four walls, while 75 percent had five or during the summers and move into their houses more walls. during the winter heating season. As shown in figure 2.1, about 47 percent of households live in Figure 2.2. Size of Ger (As Indicated by a ger during the winter months, while the rest live Number of Walls), 2012 in detached homes. Compared to 2007/08, the proportion of households that live in gers during Morethanfive walls(6,401), the winter has increased slightly, while the share 7.9% of households in detached homes has declined. Fourwalls Figure 2.1. Ger-Area Households in Surveyed (19,859),24.6% Districts, by Primary Winter Residence 200,000 Fivewalls 160,000 (54,490),67.5% GerͲareahouseholds Detachedhome 120,000 87,645(53%) Detachedhome 80,000 69,742(57%) 40,000 Ger Ger Source: World Bank Clean Stove Initiative Survey, July 76,482(47%) 52,403(43%) 2012. 0 Winter2007/8 Winter2011/12 Source: ASTAE/World Bank: Baseline Fuel Consumption, Heating Stove, and Household Perception Survey, December 2007; World Bank Clean Stove Initiative Survey, 23 July 2012; Ulaanbaatar Statistics Department A typical five-walled ger has an area of 28m2, with a diam- eter of 5.6 to 6 meters depending on the erected height of the ger. 13 Stocktaking Report of the Mongolia Clean Stoves Initiative The 2012 survey results suggest that thermal 2.2.2 Characteristics of Detached Homes insulation in gers has worsened, as the portion of gers with only a single felt covering for insulation About 53 percent of ger-area households in the has increased by more than 9 percent (up from 21 six districts are living in a detached single-family percent in 2007). The insulation of a ger may be home as their main dwelling during the winter improved by covering of the wall and roof with felt months.24 Detached homes in the ger areas are blankets, laying rugs or carpets on the floor, adding relatively small. Most have two rooms and a a skylight around the chimney to seal the opening kitchen. The average total living space, excluding at the top of the ger, and attaching a vestibule. As the kitchen, is only 48 square meters. About 88 shown in figure 2.3, 30 percent of gers have only a percent of detached homes have only one story, single felt cover for the walls and roof, while most while two-story homes account for 12 percent of (61 percent) have two or more covers for the walls detached homes. Also, as with households living and roof. Virtually all (over 98 percent) have a in gers, most households in detached homes skylight, and about one-third have a vestibule. are homeowners (96 percent). Only 1 percent of households in detached homes are renting. The Similar trends are seen in flooring that suggest a remaining 3 percent live rent-free by some other widening gap between bigger, more permanent arrangement. gers with thicker insulation, and smaller, less permanent gers with poorer insulation. Most Detached homes can be classified into three gers continue to have wood flooring (66 percent types based on the type of heating system used: in 2012); however, the share of gers with just earthen floors has increased by about 10 percent • detached houses that rely on just a stove since 2007, to 24 percent. At the opposite end of for space heating; the spectrum, about 9 percent of gers now have • detached houses with a heat wall cement floors, up from only 2 percent in 2007. As connected to a space-heating stove;25 and expected, the share of gers with single felt covers • detached houses with a low-pressure that also have earthen floors is much higher than boiler (LPB) and radiator system.26 the overall average at around 40 percent. The survey results (see figure 2.4) show that heat The findings imply that low-emission stoves, which walls continue to be the most common type of provide more heat with less fuel, can provide significant benefits in terms of more comfort at heating system for detached homes, accounting lower or equivalent fuel costs especially to smaller for 62 percent (a decline of about 8 percent ger households whose dwellings are less well- compared to five years ago). About 23 percent insulated and less permanently constructed. of detached homes use LPBs (an increase of about 7 percent compared to five years ago). Meanwhile, the proportion of detached houses Figure 2.3. Insulating Felt Coverage of Gers 24 A detached single family home is commonly referred to as a “separated home,” using the direct Mongolian to Other*(14%) English translation. 25 A heat wall is a hollow, interior, heat-retaining wall Singlecoverfor wallsandroof through which flue gases from the stove escape into the (30%) atmosphere; this is used to heat the house. The heating wall is a typical extension of stoves in Mongolia. The wall Doublecover is made of brick or a hollow metal sheet, with the hot forwallsand fumes from the attached stove led through ducts in the roof(56%) masonry (in the case of brick) or through metal ducts, to heat the whole wall. Since bricks retain heat and metal ab- sorbs it, the heat slowly passes through it to the surround- ing environment. The wall makes the stove more effective Source: World Bank Clean Stove Initiative Survey, July 2012. by providing a large emitter surface for energy absorbed Note: *Single cover for walls and double cover for roof, or from the smoke, which would otherwise be lost. 26 double cover for walls and single cover for roof. LPBs are a heating system designed for larger homes, where radiators can be installed in rooms. 14 CHAPTER 2. Housing and Household Socio-Economic Information with space-heating stoves and no heat walls Figure 2.5. Sizes of Detached Houses with has increased by only 1 percent-well within the Different Types of Heating Systems margin of error from a statistical point of view. 80 74 Sizesofhousesinsquaremeters 2007 2012 65 Figure 2.4. Proportions of Detached Houses 60 with Different Types of Heating Systems 43 42 38 39 80% 40 70% Percentofdetachedhomes 62% 2007 2012 60% 20 40% 0 23% Detachedhome, Detachedhome, Detachedhome, 20% 14% 15% 16% noheatwall withheatwall withLPB 0% Source: ASTAE/World Bank: Baseline Fuel Consumption, Detachedhome, Detachedhome, Detachedhome, Heating Stove, and Household Perception Survey, noheatwall withheatwall withLPB December 2007; and World Bank Clean Stove Initiative Survey, July 2012. Source: ASTAE/World Bank: Baseline Fuel Consumption, Heating Stove, and Household Perception Survey, December 2007; and World Bank Clean Stove Initiative 2.3 Characteristics of Households in the Ger Survey, July 2012. Areas A comparison of the data from the 2007 and The number of people in a household, gender of 2012 surveys suggests that more households in the head of household, level of schooling, and the ger areas are moving into bigger homes-or household income provide important inputs for expanding their existing homes-and upgrading marketing programs, financiers, and even stove their heating systems (figure 2.5). Not only were designers. These household characteristics can more homes using LPBs in 2012 versus 2007, the guide the way that messages are communicated average size of a detached house with an LPB or impact the choice for certain stove design has also increased, from 65 square meters in features, such as the size of pots that can be used 2007 to 73.5 square meters in 2012. By contrast, on the stove (it is more convenient to have a large homes that rely on space-heating stoves and heat pot to cook for a larger family, for example). walls are much more size-constrained-it does not The average size of households in the surveyed make sense to build a heating wall for a home areas was just over 4 persons in 2012 (see beyond a certain size. So, it is not surprising table 2.1), slightly lower than it was in 2007. that the average size of these homes is about the Reported monthly income for households over same in 2012 as it was in 2007. The average size the previous 12 months rose to 750,082 Tg on of detached homes that use space-heating stoves average, which is about 80 percent higher in real without heat walls was between 38 and 39 square terms than what was reported in 2007.27 Similar meters in 2007 and 2012. Detached houses with to the situation in 2007, about one-fifth of ger- heat walls are slightly larger than those without area households in 2012 are headed by females. heat walls because heat walls provide a larger heat-emitting surface and can warm a larger area than just a space-heating stove. The average size 27 Average monthly income reported by ger-area house- holds in the 2007 survey was around 408,000 Tg (in year of those homes was between 42 and 43 square 2011/12 prices). Prices are adjusted using the average meters in 2007 and 2012. monthly consumer price index data from the Interna- tional Monetary Fund for January to December 2007 and July 2011 to April 2012 (more recent data not yet avail- able at the time of writing). See IMF, “Data and Statis- tics,” http://www.imf.org/external/data.htm#data. 15 Stocktaking Report of the Mongolia Clean Stoves Initiative Average income and level of formal schooling percent higher, and twice as many household varies among ger households and households heads living in detached homes with LPBs have living in detached homes. Ger households, some post-secondary education. The positive which continue to represent the largest number relationships between income, size of dwelling, of households living in the surveyed areas, are and educational attainment suggest a common generally the poorest and have the lowest levels characteristic in most housing markets whereby of reported educational attainment for household higher income households start to seek a higher heads. By comparison, income for households quality of life, including more sophisticated and living in detached homes with LPBs is about 50 convenient cooking and heating solutions. Table 2.1. Socio-Economic Information of Households in the Ger Areas Detach house Detach Detach house Ger without heat house with with low Total wall heat wall pressure boiler Total Number of Households 76,483 13,623 54,326 19,695 164,127 Household Monthly Income 634,770 695,613 850,738 957,906 750,082 (Tg) Family Size (persons) 4.1 4.0 4.5 4.3 4.2 Female Head of Household (%) 22.5 28.9 21.4 12.5 21.4 Head of Household with post- 18.1 25.3 29.9 35.9 24.7 secondary schooling (%) Source: World Bank Clean Stove Initiative Survey, July 2012. 2.4 Household Experience with Banking and also open doors to other banking services. Credit XacBank not only offers micro-credit toward the purchase of stoves, it also uses its financial The survey also investigated the linkages management services to process stove payments between informal and formal banking sources, to stove suppliers. The bank also offers loans for what households borrow for, and their interest other energy efficient products. and use of mobile banking (banking by mobile phone, which might be a powerful way of The survey finds that most households in the transacting stove purchases in the future). six surveyed ger districts have experience with Access to finance is an important feature to formal banking, and about half of the households consider in designing stove switching programs have experience with formal credit. With respect especially in markets with poor households. to banking, about 86 percent of the households Well-designed micro-credit schemes can (around 141,500) have at least one member help households spread out payments for with a bank account (figure 2.6). As shown in stoves, which can make up a significant figure 2.7, debit and savings accounts are most share of monthly income for households with common. Only one-tenth of households have little savings capacity. Informal channels of a credit card. Furthermore, about 23 percent finance are important sources of funds in of households have an account at a pawn shop, these communities and should be explored to indicating that semi-formal financial institutions understand how households might collect funds continue to have a presence in the ger areas, but to pay for a new stove. Yet, formal banking can are less popular now than formal banking. 16 CHAPTER 2. Housing and Household Socio-Economic Information Figure 2.6. Share of Households with At Least Figure 2.8. Source of Credit or Financing to One Member with a Bank Account Ger-Area Households Relativesorotherfamilymembers 49% No (22,650) Commercialbank 48% 14% Friend 39% Pawnshop 33% Privatecompanythatsellsgoods 9% Yes (141,477) PrivateMoneyLender 9% 86% Micofinanceorsavinggroup 7% 0% 20% 40% 60% Percentofhouseholds Source: World Bank Clean Stove Initiative Survey, July Source: World Bank Clean Stove Initiative Survey, July 2012. 2012. land and/or home property, business Figure 2.7. Type of Bank Account or Other investments, and previous debts. As shown in Financial Account Held by Households figure 2.9, about half of the households borrow Debitcardaccount 58% money for food and living expenses, and about a third have borrowed money for consumer goods. Savingaccount 50% Pawnshop 23% Figure 2.9. Purpose of Loan Creditcardaccount 9% Foodandbasiclivingexpenses 50% Microfinance/savinggroup 6% Tobuyconsumergoods 35% Checkingaccount 1% Paymedicalbillsand/orhealthcare 18% 0% 10% 20% 30% 40% 50% 60% Buylandand/orhome 15% Percentofhouseholds Investmentand/ortoexpandbusiness 12% Source: World Bank Clean Stove Initiative Survey, July Payoffotherdebts 8% 2012. 0% 20% 40% 60% Percentofhouseholds 2.4.1 Experience with Credit/Loans Source: World Bank Clean Stove Initiative Survey, July 2012. The survey reveals that informal credit continues to dominate, and relatives and friends are the most important source of loans; however, the 2.4.2 Knowledge of Lending Products and use of formal credit is expanding. About half Banking Services of the households surveyed have borrowed money at some point from a commercial bank. From the point of view of a commercial bank, On the semi-formal market, about 33 percent reaching ger-area households that have past of households have borrowed money from a experience with formal banking and/or credit pawnshop before. Only 7 percent have ever is crucial. Already, some commercial banks borrowed from a micro-lending or savings group. in Mongolia have created lending products Findings are shown in figure 2.8.\ specifically designed to help households overcome the upfront costs of modern heating The main purposes for which households solutions and other energy products. These borrow money are to pay for: food and basic include micro-loans for low-emission stoves, living expenses, consumer goods, health care, super-insulating ger blankets, vestibules, 17 Stocktaking Report of the Mongolia Clean Stoves Initiative heating floors, energy efficient flues, and solar 2.5 Conclusion photovoltaic panels. The survey reveals that overall more than half of the households have Analysis of socio-economic characteristics of heard about these loan products offered to households in the ger areas confirms that a consumers by banks. large number of households are still relatively poor, though income levels have risen. The majority of poorer households live in gers, Figure 2.10. Share of Households that Have while households that are financially better off Heard of Lending Products Specifically for tend to live in detached houses. The portion Heating and other Household Energy Products of households living in gers versus detached LowͲemissionheatstove 68% homes has also remained the same. Thus, the Heatingfloor 58% overall profile of households in the ger areas is Energyefficientflue 56% largely unchanged since the last survey carried Ecoconsumerloanforgerblanket 53% out in 2007. What has changed is the number Vestibules 45% of households in these areas. Since 2007, the Solarpanel 41% population in the ger areas of the six districts 0% 10% 20% 30% 40% 50% 60% 70% 80% surveyed increased by about 42,000 households, Percentofhouseholds which is significant from the perspective of air Source: World Bank Clean Stove Initiative Survey, July pollution in Ulaanbaatar because virtually all 2012. of these households are reliant on coal-burning stoves for heating. 2.4.3 Knowledge and Interest in Mobile Within the overall picture of ger-area households, Banking there are also some disparities that are emerging since the 2007 survey. On one end of the Aside from knowledge of specific lending spectrum, a portion of better-off households in products, the survey also explored households’ detached homes are building larger homes-or knowledge and experience regarding the use expanding their existing homes-and upgrading of mobile phones for banking transactions. their heating systems from traditional stoves Survey results show that every household has to LPBs, which burn much more fuel and emit at least one mobile phone, with an average of more pollution (the report later discusses fuel about 3.3 mobile phones per household. This use). On the other end of the spectrum, more of indicates that most household members have the households living in small, four-walled gers their own personal mobile phone. However, have homes that are less well-insulated and less only a small portion of the households use text permanently constructed. These findings suggest messaging on a regular basis, and only slightly a widening gap in housing conditions. more (17 percent) have ever used a mobile phone for banking. That said, many are aware With respect to household banking and credit, of the possibility for bank customers to conduct the 2012 survey finds that the vast majority of transactions with their mobile phones, and more households in the six surveyed ger districts have than half (57 percent) of respondents indicated experience with formal banking, and about half they are interested in using their mobile have experience with formal credit. Apart from phones for banking. This finding is important, relatives and friends, commercial banks are the since small lending products designed to help most important source of credit or financing for households gain access to heating devices ger-area households. The top two reasons for typically involve high transaction costs. Use of borrowing money identified by households are mobile phones may reduce transaction costs to to pay for food and basic living expenses and consumers and administrative costs to banks, buy consumer goods. In terms of the potential potentially lowering the price of loans. for offering financial products to encourage the 18 CHAPTER 2. Housing and Household Socio-Economic Information purchase of low-emission stoves, perhaps the of financial products specifically designed most important information from the survey for household heating and energy products. is that about 44 percent of households have This suggests that more households may be already experienced borrowing money from a comfortable buying low-emission stoves on commercial bank or a retailer to buy consumer credit, perhaps for some even through mobile goods, and most are aware of the existence banking. 19 CHAPTER 3. HEATING STOVE OWNERSHIP AND USE Stoves and low pressure boilers are the primary price (see also section 1.3). This project ran heating devices for virtually all households in at full scale from August 2011 to the end of the six central ger area districts. However, one November 2012. As of July 1, 2012-just before type of stove does not match the needs of all surveying was done for this study-around households. Stoves serve a dual purpose and are 69,000 low-emission stoves had been sold to used for both heating and cooking during the the households in the ger areas. An additional cold season from early or late September to late 29,000 stoves were sold in the fall and winter, April or early May. For space heating, stoves can bringing the total number of stoves distributed be used by themselves or, in the case of detached under the project to around 98,000. homes, with attached heat walls to increase Based on the household survey conducted the distribution of heat to warm a larger area. in July 2012, the following sections provide a Some detached homes have furnaces (or stoves detailed descriptive analysis of heating stove attached to a water tank) with piping to circulate ownership and use in ger areas. The estimated hot water to radiators (or to showers) in different total number of stoves owned and used is also rooms throughout the house. These “stoves” are provided to show the level of market penetration commonly referred to as a low pressure boiler by low-emission stoves and the size of the (LPB). remaining market. The chapter also describes Since around 2008, low-emission and energy- key stove user characteristics. efficient stoves have been discussed as a tool to reduce air pollution in Ulaanbaatar. 3.1 Total Number of Heating Stoves Owned As discussed in chapter 1, pollution source and Used inventories showed that emissions from traditional heating stoves in the ger areas The 2012 survey asked households about the accounted for around an estimated 60 percent number of stoves they own, and how many they of the concentrations of very small particulates actually used for heating during the previous (PM2.5) in the city’s air.28 By providing better winter months. From the survey, it is estimated combustion efficiency, low-emission stoves that in 2012, households in the ger areas can lower the level of pollutants emitted and surveyed owned about 212,700 stoves in total offer an effective means to reduce pollution in (see table 3.1), which is more than the total the city. As a result, in 2011 the Government of number of households (164,127 households) in Mongolia and MCA-Mongolia began to provide the survey area. About 9 percent of households financing to import and sell low-emission stoves reported owning two or more stoves. to ger residents in Ulaanbaatar at a subsidized The average number of stoves owned per 28 household in 2012 was 1.30. By comparison, the Based on population-weighted exposure to PM for Ulaan- baatar City in 2009, calculated using pollution dispersion average for the baseline survey in 2007 was 1.03. modeling. World Bank, Mongolia: Air Quality Analysis of The difference may be the result of several factors. Ulaanbaatar: Improving Air Quality to Reduce Health Im- pacts (Sustainable Development Series Discussion Paper, December 2011). 21 Stocktaking Report of the Mongolia Clean Stoves Initiative Table 3.1. Total Number of Stoves Owned in during the previous winter for heating. Although Ger Areas, by Stove Type, 2012 in practice only one stove is usually used for heating in ger-area homes, about 5 percent of Number Share respondents said they used more than one stove. Traditional stoves 128,182 60% Some households may have used additional Low-emission stoves 46,449 22% stoves to heat workshops, garages, or other Locally-made LPBs 17,069 8% structures on their hashaa; and some may have Imported LPBs 3,939 2% switched to using another stove during the Other stoves 17,069 8% middle of the winter heating season (e.g., by Total 212,708 100% purchasing a low-emission stove). Furthermore, Source: World Bank Clean Stove Initiative Survey, July many stoves counted in the “other” category in 2012. table 3.1 above are not for heating, including smaller stoves used for cooking by ger First, the questionnaire used for the 2012 survey households during the warm season. contained more detailed questions on stove ownership and usage than the questionnaire used The number and types of different stoves that in the 2007 survey. While the questionnaire for were used by ger-area households during the 2012 survey differentiated between number the previous winter are shown in table 3.2. of stoves owned and used, the 2007 survey only Traditional stoves accounted for 65 percent of asked respondents to identify the stoves that all stoves used for heating during the winter of their households were using for heating at the 2011/12, while low-emission stoves accounted time of survey. Second, the total number of for 22 percent. Compared with four years ago, stoves owned by ger-area households may have reliance on traditional stoves has dropped by increased somewhat as a result of the low-emission 11 percent. Meanwhile, the survey reveals stove exchange program. Under the program, that LPBs account for about 11.6 percent of all households had to exchange their old stoves in heating stoves that households reported to be order to receive new low-emission stoves. Old using during the previous winter, which is about stoves were removed from the households and 3 percent higher than in the winter of 2007/08. taken to be destroyed. However, the study team was given the impression through interviews of Table 3.2. Total Number of Stoves Used in Ger participants that a few households that bought new Areas, by Stove Type, 2012 low-emission stoves received an exemption (about 7 percent), including newly-wed couples that had Number Share been given traditional stoves as wedding gifts and Traditional stoves 112,427 65% households that used brick stoves that could not Low-emission stoves 38,077 22% easily be removed. Furthermore, as the program Locally-made LPBs 16,249 9% progressed, the stoves that were exchanged Imported LPBs 3,775 2% tended to be older and in worse condition than those that were turned in at the start of the Other stoves 1,477 1% program.29 It is therefore conceivable that some Total 172,055 100% old heating stoves were exchanged that were not Source: World Bank Clean Stove Initiative Survey, July being used by households, and that the stove 2012. exchange program indirectly created a market for old stoves. Of the 212,700 stoves owned by households in 2012, about 172,000 were reportedly used 29 Author phone interview with XacBank representative, November 1, 2012. 22 CHAPTER 3. Heating Stove Ownership and Use 3.2 Age of Stoves Used During the Previous Figure 3.1. Number and Age of Stoves Winter Months Traditionalstoves ImportedLPBs LocallyͲmadeLPBs 16,000 Table 3.3 presents an overview of average age 14,000 of different stove types. An analysis of stove age 12,000 11,815 13,459 numberofstoves 11,325 12,963 reveals that the traditional stoves have been used 10,000 for an average of just over 9 years.30 The age of 8,000 4,759 traditional stoves has increased remarkably since 6,000 4,000 the previous survey in 2007, when traditional 821 821 2,000 821 328 stoves had been used for less than 6 years on 1,312 Ͳ 3,118 2,462 1,805 1,805 492 1,149 0 average. This finding has a couple of important Lessthan1 1year 2years 3years 4years 5years year implications. First, it is evident that very few numberofyearsstoveshavebeeninuse households bought new traditional stoves within Source: World Bank Clean Stove Initiative Survey, July the past two years. As shown in figure 3.1, the 2012. number of traditional stoves acquired each year Note: Survey was conducted in August 2012; one year old has dropped significantly from about 13,500 five stoves were defined as stoves acquired before August or September 2011. Stoves less than one year old were defined years ago to around 4,800 in 2011. Of the 112,427 as stoves acquired after January 2012. traditional stoves estimated to be in use during the winter of 2011/12, only 6,072 were acquired within the past two years. Low-emission stoves 3.3 Heating Stove Use by Dwelling Type have almost completely replaced traditional stoves Heating is an important part of everyday life on the market, which is not surprising because for ger-area households. In the areas of the six the subsidized price of the low-emission stoves districts surveyed, there around 168,000 stoves was much lower than the price for most traditional were in use during the winter of 2011/12, a stoves. Second, because about 90 percent of sizeable increase since 2007/08. The types of households replace their stoves within 10 years stoves used by households are closely associated (as was seen in the 2007/08 baseline survey), with the size and type of the dwelling and can be most of the traditional stoves that remain in use grouped into the following categories: will likely be replaced within the next couple of years. If stove replacement efforts end in 2013, • Ger using space heating stove and the unsubsidized price of low-emission stoves • Detached house low pressure boiler remains the same, then households will have no (LPB) alternative but to buy another traditional stove. • Detached house with heat wall. • Detached house without heat wall (using Table 3.3. Average Age of Stoves in Ger Areas, a space-heating stove). by Stove Type, 2012 Figure 3.2 illustrates the relative share of these Number of different heating systems among the total Type of Stove Years in of Use number of ger-area households. The following Traditional stoves 9.2 sub-sections provide a detailed description of Low-emission stoves 0.8 stoves used for these different types of heating Locally-made LPBs 5.5 systems and dwelling types. Imported LPBs 3.1 Source: World Bank Clean Stove Initiative Survey, July 2012. 30 This is about the same as the average age of traditional stoves owned but not used. 23 Stocktaking Report of the Mongolia Clean Stoves Initiative Figure 3.2. Heating Systems by Dwelling Type, 3.3.2 Heating Systems and Stoves Used in Ger-Area Households, 2012 Detached Houses DetachedhouseswithLPBs As noted in chapter 2, of the 87,643 households 12%(19,695) that in 2012 live in the detached houses within DetachedhouseswithspaceͲheating the surveyed area, about 62 percent use space- 47%(76,483) gerstovesandheatwalls heating stoves with heat walls, 15 percent use 33%(54,326) stoves without a heat wall, and 23 percent DetachedhouseswithspaceͲheating gerstoves,butnoheatwalls of households use LPBs (figure 3.4). The proportion of detached houses with a heat wall 8%(13,623) GerswithspaceͲheatingstoves declined by about 6 percent compared to the previous survey, while the proportion of detached Source: World Bank Clean Stove Initiative Survey, July houses with LPBs increased by about 6 percent 2012. over the same period. The proportion of detached houses without heat wall remained the same. 3.3.1 Heating Stoves Used in Gers The results indicate that more households are upgrading from relatively simple heating stoves All of the ger households in the areas surveyed to LPBs, which, as will be discussed in chapter 5, rely on heating stoves (including LPBs). has significant implications for fuel consumption. As shown in figure 3.3, about 69 percent of households that lived in a ger during the Figure 3.4. Heating Systems and Stoves Used in previous winter months used some form of Detached Houses, 2012 traditional heating stoves, including metal sheet, cast iron, and sawdust or brick stoves. While the Traditional stoveused proportion of ger households using traditional directly, stoves has declined significantly from 96 percent withoutheat wall,7% in 2007, there are still about 52,000 households living in the ger areas of the six districts LowͲpressure boiler(LPB), surveyed that continue to rely on traditional 23% stoves; only about 24,000 households living in LowͲemission stoveusedwith gers have switched to low-emission stoves. heatwall,8% Traditional stovewithheat wall,54% Figure 3.3. Percentage of Traditional and LowͲemission Low-Emission Stoves Used for Heating in Gers stoveused directly, 96% withoutheat 100% wall,8% 90% (ofthoselivingingers) percentofhouseholds 80% 69% 2007 2012 70% Source: World Bank Clean Stove Initiative Survey, July 60% 2012. 50% 40% 31% 30% 20% With regard to the types of heating stove these 10% 4% 0% 0% households used, the survey reveals that low- 0% emission stoves have not penetrated into this Traditionalstove Improvedversion LowͲemission oftraditional stove market segment as much as it has for households stove living in gers. Only about 19 percent (16,700) of the households that live in the detached houses Source: World Bank Clean Stove Initiative Survey, July 2012. Note: Low-emission stoves had only just been introduced use low-emission stoves, compared to 59 percent in 2011; prior to 2011 only improved version of traditional (51,400) that use traditional stoves, and 21 stoves were available in the market. percent (18,100) that use LPBs. 24 CHAPTER 3. Heating Stove Ownership and Use Heating walls have different technical have done some repair or maintenance of their characteristics that can influence the heat walls. performance of stoves, when compared to stoves LPBs appear to have emerged as a preferred in gers or detached homes without heating walls. type of stove for high-end consumers, but there About 54,300 detached homes that have heat is still not much known about how households walls, equal to 62 percent of detached houses have been using them or about their emissions or 33 percent of all households in the surveyed factors. As mentioned above, from the survey ger areas. Of these houses with heat walls, the results, it is estimated that about 23 percent of survey reveals about 45 percent have a “vertical all detached houses in the surveyed ger areas heat wall,” which means the heat walls are used LPBs for heating during the previous constructed to allow flue gases to move vertically winter. Since detached houses with LPB tend to exchange heat before exiting the chimney. to be bigger than detached houses with heat About 27 percent of heat walls are “horizontal,” walls and those that use stoves directly for space which means that flue gases move horizontally to heating and hot tap water, some households exchange heat before exiting. Another 18 percent with LPBs also use other coal heating stoves for have a combination of vertical and horizontal supplementary heating and cooking. The survey channels. It is important to note that about 10 finds that about 2,000 households living in percent of the households that live in detached detached homes with LPBs also own traditional houses with a heat wall do not know the type of stoves, and that a few hundred households used heat wall they have. both low-emission stoves as well as LPBs during The vast majority of heat walls are made of brick. the previous winter. Heat walls made of exposed metal sheeting are The survey finds that the LPBs that are being use not widely used because the metal can heat to a are relatively new: about 60 percent of LPBs were very high temperature and poses a burn hazard. installed within the past four years. With respect Furthermore, because heat walls are relatively to the type of LPB used, the survey reveals that simple to make, the survey finds that about half more than three-quarters of LPB stoves are made were constructed by the home owner, while locally, the remaining are imported, which cost another 41 were constructed by local masons. significantly more than the locally made LPB. The remaining nine percent of households do not know who built their heat walls. On the basis If Mongolia’s current rates of economic growth of observation, quality of construction is reported continue, it is expected that demand for LPBs to be highly variable. will continue to rise as more households are able to pay for LPBs and the internal piping Heat walls require regular maintenance and needed, to increase their comfort and heat cleaning because soot build-up can block flue their (increasingly larger) new homes. About gases from escaping and cause the wall to one-third of respondents with LPBs said that crack or even explode. About 49 percent of the their LPB is the first heating system used in households indicate that their households clean their homes. (In other words, these are new the heat wall once a year, while 20 percent homes.) Another 31 percent of LPB users used reported cleaning it more than once a year, 11 to have a stove with a heat wall, indicating that percent clean it once every two or three years, these households have recently upgraded their and 2 percent clean it only once every four or heating systems. About 11 percent indicated that five years. About 18 percent of households had their existing LPBs were replacements for older never cleaned their heat wall at all. There is no LPBs. Although the market segment for LPBs significant difference in the reported frequency is small, it is growing rapidly compared to four of cleaning between households with and without years ago. Because LPB stoves require more low-emission stoves. In addition, the survey coal (see chapter 5) and emit more pollution shows that about 58 percent of the households 25 Stocktaking Report of the Mongolia Clean Stoves Initiative than traditional space-heating stoves, attention The contrast in adoption rates for households must be paid to this market segment to control living in detached homes with and without air pollution in Ulaanbaatar. Of course, any heat walls is remarkable. While the survey intervention to phase out high-emission LPBs did not include questions to explore why will need to be targeted differently than previous households living in detached houses with stove replacement programs because LPBs heat walls were more resistant to adopting low- and ger stoves represent two distinct market emission stoves, it did reveal other differences segments with two distinct supply chains. between households that may shed light on the Virtually all of the households that use LPBs discrepancy in adoption rates. A comparison of said they bought their stove from the Ail Market, household monthly income between households a construction materials market. The LPB living in detached houses with and without heat market could be considered as a subset of the walls suggests that affordability is probably not construction and plumbing market. the reason for higher or lower adoption rates. Reported monthly income for households in 3.4 Low-emission Stove Ownership and Use detached houses with heat walls was 850,737 Tg compared to 695,613 Tg for those living in Low-emission stoves, which were introduced to detached houses without heat walls and 634,770 the Ulaanbaatar market at subsidized prices via a Tg for those living in gers, yet households living large-scale stove replacement program launched in detached houses without heat walls or gers in June 2011, account for 22 percent of the heating were much more likely to acquire low-emission stoves used by households in the ger areas stoves. The same is true with education. surveyed. Although the estimated number of low- emission stoves households reported owning Figure 3.5. Survey-Estimated Market Share of at the time of survey in July 2012 is 46,449, the Low-Emission Stoves, by Dwelling Type number of low-emission stoves reportedly used NolowͲemissionstoves WithlowͲemissionstoves during the previous winter months was estimated Allhouseholds at 38,077. The discrepancy of 8,372 stoves 73% 27% between the number of stoves owned and used Gers 66% 34% is due primarily to the fact that a large number Alldetachedhouses 79% 21% of households acquired their low-emission stoves DetachedhouseswithspaceͲheating 48% 52% gerstovesandnoheatwalls toward the end of the winter or after the heating DetachedhousewithspaceͲheating season was already over. The survey reveals that gerstovesandheatwalls 83% 18% about 6,400 low-emission stoves were acquired DetachedhouseswithLPBs 92% 8% between mid-April and the end of June 2012. 0% 20% 40% 60% 80% 100% As can be seen in figure 3.5, the market share Source: World Bank Clean Stove Initiative Survey, July for low-emission stoves varies considerably by 2012. dwelling type. As of July 2012, 31 percent of ger Note: Low-emission stoves in detached houses with households owned low-emission stoves. The LPB refers to the typical low-emission space heating market share for households living in detached stoves, not the not the low-emission LPB stoves. homes without heat walls was even higher, at 52 Market shares are estimated on the basis of survey percent. However, only 18 percent of households data and not official sales statistics. See footnote 27. living in detached homes with heat walls and 8 percent of households in detached homes with Reported rates of post-secondary schooling LPBs had acquired low-emission stoves.31 were higher for households living in detached homes with heat walls, yet their adoption rates 31 were much lower. These findings suggest that A total of 9,519 households living in detached houses with heating walls that adopt low emission stoves ac- socio-economic characteristics were probably count for only 5.7 percent of all households in the six not decisive factors in explaining the low rates surveyed ger districts. 26 CHAPTER 3. Heating Stove Ownership and Use of low-emission stove ownership for households in Chingeltei and Sukhbaatar Districts, sales with heat walls. Rather, as will be discussed in centers were operated in all of the khoroo with chapter 4, perceptions about the functionality of sizeable ger-area populations. By the end of the low-emission stoves and their use with heat walls winter, 79 percent of the ger-area households may be more influential in shaping the decision in Chingeltei and 64 percent of the ger-area by households with heat walls to not acquire the households in Sukhbaatar had acquired low- new stoves. Further market studies are needed emission stoves. On the other side, subsidizes to closely examine the reasons for this decision stoves were not sold in Bayangol District due to in order to increase adoption rates for this group. that district’s ban on raw coal (though product centers were still operated in Bayangol to sell 3.4.1 Low-Emission Stove Ownership in other energy efficiency products such as ger blankets and vestibules).33 Different City Districts An additional 29,027 low-emission stoves were The Mongolia-MCA project sold 68,850 low- distributed during the second phase of Mongolia- emission stoves during the 2011/12 heating MCA and CAF subsidies in the fall and winter of season, reaching 42 percent of the ger-area 2012. The second phase targeted khoroo farther households in the six central districts of from the city center, particularly in Bayanzurkh Ulaanbaatar.32 An additional subsidy element and Songinokhairkhan, where 12,250 and 13,034 was financed by CAF. Stoves were sold only to stoves were sold, respectively. No stoves were households registered in designated khoroo of sold in Chingeltei, and only one sales center was five districts where sales centers were operated. operated in Sukhbaatar. As with the previous year, These khoroo were identified as “air pollution stoves were not sold in Bayangol due to the raw reduction zones” by the City of Ulaanbaatar. coal ban. In all, the program reached 55 percent Due to the project’s targeting of particular areas, of all households living in the ger areas of the six as can be seen in Table 3.4, the market share central districts, or 69 percent of the households of low-emission stoves (expressed as a portion living in those khoroo where stoves were sold. of all households living in each district) varied considerably across districts. On the one side, 32 Sales statistics provided to authors via email by MCC representative, December 27, 2012. Note that the World Bank survey estimated that 46,449 households in the surveyed areas owned low-emission stoves as of July 2012. The survey likely underestimated the total number of low-emission stoves due to two factors. First, the sur- vey was intended to provide an overall view of the ger-ar- ea market for heating stoves. Sampling was not stratified or weighted according to the khoroo in which the MCA- Mongolia and CAF program was operated. Sales were concentrated in designated khoroo. By chance, the ran- dom sampling across all the khoroo under-represented those khoroo where the stoves were sold. Whereas the khoroo where low-emission stoves were sold accounted for 22 percent of the observations in the survey, they rep- resent 27 of the total number of households in the ger ar- eas of the six districts covered in the sampling frame (as of the end of 2011). Second, estimates for the total num- ber of stoves in the survey were extrapolated according to official population statistics for the ger areas. As was discussed in chapter 1 of this report, the actual number 33 of households in the ger areas is probably larger. The Raw coal is the appropriate fuel for which the models of survey does not purport to make any quantitative infer- low-emission stoves distributed by the program are de- ences about the possibility of leakage (i.e. households signed to burn. Other fuels, such as semi-coke coal bri- selling or giving away their low-emission stoves to others quettes, are not intended for use with the stoves and may outside the Ulaanbaatar). affect emissions performance. 27 Stocktaking Report of the Mongolia Clean Stoves Initiative Table 3.4. Low-emission Stove Sales by District and Heating Season Total share of Low-emission stoves Ger-area households in households with low- sold district* District emission stoves Winter Fall/Winter End of End of Winter Fall/Winter 2011/12 2012 2011 2012 2011/12 2012 Bayangol 0 0 11,855 11,638 0% 0% Bayanzurkh 12,585 12,250 45,083 55,890 28% 44% Chingeltei 22,680 0 28,639 29,242 79% 78% Khan-Uul 2,642 2,960 15,802 16,479 17% 34% Sukhbaatar 12,640 783 19,852 19,260 64% 70% Songinokhairkhan 18,303 13,034 42,896 45,424 43% 69% Total 68,850 29,027 164,127 177,933 42% 55% Source: MCC and MCA-Mongolia stove sales data; Ulaanbaatar Statistics Department population data. Notes: * ger-area households as of the end of 2011 and 2012, excluding khoroo outside the central area of Ulaanbaatar City; ** cumulative market share is estimated as a share of all households in the district as of the end of 2011 and 2012, including those not in designated khoroo where stoves were sold, to provide a fuller view of the remaining market size. 28 CHAPTER 3. Heating Stove Ownership and Use 3.4.2 Characteristics of Households that high upfront cost of the stoves was offset by the Bought Low-Emission Stoves subsidies provided by MCA-Mongolia and CAF. The survey finds that 58 percent of households The survey results show that there is no that bought low-emission stoves live in gers. significant difference in reported income or size Reported monthly income for households that of ger between households that use low-emission bought low-emission stoves was about 725,550 stoves and households that do not use low- Tg, versus 758,785 Tg for households that did emission stoves. The average total household not own low-emission stoves. The difference is monthly income for ger households that use low- not statistically significant. emission stoves is 646,221 Tg, compared with 629,334 Tg for ger households that do not use By contrast, there is a much larger difference low-emission stoves. in reported income between households that said they used their low-emission stoves and However, it appears that the ger households that households that acquired low-emission stoves use low-emission stoves live in better-insulated gers but do not use them. The average monthly than ger households that do not use low-emission income of households that use low-emission stoves. The survey finds that about 70 percent of stoves was 687,816 Tg, compared to 1,025,180 ger households that use low-emission stoves have Tg for those that own but do not use their low- two layers of insulating roof covers, while only emission stoves. This implies that many of the 52 percent of ger households that do not use low- households that acquired their stoves late in the emission stoves have gers with two layers of roof season but did not use them were higher-income covers. Similarly, 40 percent of ger households households. From the data, it is impossible to say that do use low-emission stoves have two layers whether these households purchased new stoves of insulating wall covers, while only 22 percent of with the intention to use the stoves, or whether ger household that do not use low-emission stoves they simply had extra cash on hand and were have a double layer of wall covers. This condition motivated to seize on the opportunity to buy a suggests that many ger households that purchased highly-discounted stove while they could before low-emission stoves also may have purchased the sales season ended. At any rate, the difference subsidized insulation and other ger retrofit materials reveals two distinct groups of consumers: poorer that were sold concurrently with the stoves. households that bought their stoves earlier in the An analysis of customer data maintained by winter sales season and used them to heat their XacBank provides additional insight into the homes, and better-off households that waited until characteristics of the individual consumers the end of the season to make their purchase but who purchased the new stoves. XacBank’s have not yet used their stoves. monitoring and verification database contains There are several reasons for why poorer demographic information for around 51,000 of households in gers tended to be early adopters the people who bought stoves from XacBank and users of low-emission stoves. First, heating product centers in the six districts where is a basic necessity for Mongolian households, surveying was conducted for this study between not a luxury good. Second, although gers are June 2011 and July 2012.34 Customer information more compact, they are generally not as good as detached houses for retaining heat. Third, 34 According to the XacBank sales data, a total of 63,043 households that live in the gers may be more stoves were purchased from XacBank-operated product centers and delivered to households in the six districts conscious about the overall costs of heating between June 1, 2011 and June 30, 2012 (excluding their home because they tend to spend a higher 1,827 customers for which district of residence was not portion of their income on heating each winter. clearly listed). Of these 63,043 customers, demographic information was not provided or incomplete for 11,619 Using a more efficient low-emission stove could customers. Sales data provided to authors via email with reduce spending on heating fuels, while the XacBank representative, November 23, 2012. 29 Stocktaking Report of the Mongolia Clean Stoves Initiative includes district and khoroo of residence, gender Figure 3.6. Income of Customers Purchasing (of both the customer and head of household), Low-Emission Stoves from XacBank income, and years living in current residence. 50% 47% 45% The customer data show that 40 percent of the Percentageoftotalcustomers 40% customers who bought stoves from the XacBank 35% 30% product centers were female. Most of these 25% 22% 20% female customers (67 percent) were heads of 20% 15% household. What this implies is that female- 10% 10% headed households were somewhat more 5% likely to purchase low-emission stoves than 0% 150,000orless 150,001Ͳ350,000 350,001Ͳ550,000 Morethan550,000 male-headed households. While female-headed Reportedmonthlyincomeofindividualcustomer(Togrog) households accounting for only 21 percent of all households in the ger areas of the six districts Source: XacBank low-emission stove sales data, June 1, surveyed, the share of female-headed households 2011 to June 30, 2012. among stove-buying customers was around 27 percent. Yet, female customers also tended to 3.5 Conclusion delay their buying decision until farther into the winter sales season, purchasing stoves about two The recent low-emission stove replacement weeks later than male customers on average (a program has had a significant impact on stove statistically significant difference).35 ownership among households living in the ger areas. However, it is important to emphasize that Reported income of individual customers in the by the end of the winter just prior to the survey, XacBank database shows a similar trend to that more than half of all households living in the ger observed in the survey data. As illustrated in areas were still using traditional stoves. Given figure 3.6, most customers (about 70 percent) the changing demographics and revelations that reported a monthly income of 350,000 Tg or less. some households have more than one stove, As for the timing of purchase, customers in the it is difficult to estimate the remaining market. lowest income bracket bought their stoves an It is evident that around 20,000 households in average of 8-12 days earlier in the sales season detached homes that use LPBs will require than customers in the middle income brackets. attention in the next phase, with additional Customers in the highest income bracket also investigations on this segment of the market. bought their stoves earlier than those in the The trends in other market segments also lead to middle income brackets. Furthermore, the the conclusion that more efforts will be required XacBank customer data show that households in to ensure that the majority of households will the highest-income bracket were not any more completely switch to and use low-emission stoves. likely to purchase their stoves after the end of the heating season than households in the lower A comparison of adoption rates for low-emission income brackets. Households in the highest- stoves among different segments of the ger- income bracket account for only 11 percent of area market reveals that households living in those households that received their stoves the detached houses were less likely to have after April 30, 2012. Unlike the survey data, purchased low-emission stoves than households the XacBank customer data does not provide living in gers. Only 18 percent of households information on whether households in the living in detached houses with heat walls have highest income bracket were more or less likely acquired low-emission stoves versus 31 percent of to use their stoves after purchasing them. households living in gers. This finding suggests that households living in detached houses with 35 The exact date of stove purchase is not provided in the heat walls are much more resistant to changing XacBank database. Approximate purchase dates are based on information about the delivery or installation their heating systems. Adoption rates were even date. 30 CHAPTER 3. Heating Stove Ownership and Use lower for the LPB market segment, which has not that the condition of the old stoves exchanged for yet been targeted by stove replacement programs. new low-emission stoves grew worse and worse However, heating walls and LPBs use more coal as the program progressed. It is conceivable that and the LPB market is growing rapidly as more the stove exchange program created an indirect households are moving to bigger houses (or market for old stoves that were not used by expanding their existing houses) and upgrading households. This unintended consequence of their heating systems. To ensure that low- household behavior is not necessarily unusual, emission stoves can achieve higher penetration especially if the traditional stoves that household rates in under-represented market segments, a must turn in for destruction are still usable and in clear understanding of household preferences and good condition. While the exchange program put reasons why households are likely or not likely to in place rules to prevent households from offering adopt low-emission stoves is needed. up unusable stoves-and these rules were tightened over time-measures to improve actual replacement To ensure that for each low-emission stove added to rates are needed. This could be done by training the market one traditional stove is taken out of use, and educating households of the harmful effects of it is important that traditional stoves are removed traditional stoves in addition to random checking from the homes of households buying low-emission during winter time to identify secondary stove stoves. Removal rates for old stoves under the stove use and perhaps assigning an additional value for exchange program were around 93 percent for the turning them in. 2011/12 sales season. However, it was observed 31 CHAPTER 4. HOUSEHOLD PREFERENCES AND PERCEPTIONS OF HEATING STOVES AND AIR POLLUTION The successful and sustainable promotion low-emission stoves. The third section of the of low-emission stoves depends largely on chapter discusses household knowledge of the acceptance of consumers. Stoves must low-emission stoves, sources of information be designed to fit the needs and preferences about low-emission stoves, attitudes toward air of households, the ultimate users of low- pollution, and perceptions of the causes of air emission stoves. Households have different and pollution. This information is used to assess evolving preferences as incomes rise and other the effectiveness of information dissemination characteristics change. Encouraging the growth methods, experience that can be used for a of innovative private enterprises that offer a future program. variety of good-quality products will increase the likelihood for the rapid, widespread, and 4.1 Heating Habits sustainable uptake of low-emission stoves by consumers. Otherwise the risk exists that cheap Heating habits are important for stove designers, low-emission stoves are purchased for re-sale emissions testing protocol developers, and stove rather than for household use. switching program managers because adoption of new technologies is easier when these This chapter begins by describing how technologies more closely match with traditional households utilize their stoves and follows patterns of cooking and heating behavior. In by exploring which stove design features are addition, understanding what affects and does considered to be important from the point not affect heating habits can also offer insights of view of consumers. This discussion on into constraints and opportunities for adaptation important design features is supplemented to new technologies. Winters in Mongolia are by an examination of consumers’ perceptions cold and long. The heating season typically of the performances of their existing stoves. begins around September and last through Comparisons are drawn between households April or mid-May. Sub-zero temperatures begin that are using traditional stoves and households in mid - to late November and last until March. that have recently purchased and are using The average daily temperature for Ulaanbaatar low-emission stoves. The main objective of this in winter is about -13°C, with temperatures analysis is to gain a better understanding of the dropping to as low as -40°C at night, making preferred characteristics and design features Ulaanbaatar the coldest capital city in the consumers are interested in. The second part world. With such cold weather, it is very costly of the chapter gauges households’ willingness for households living in the ger areas to keep to switch from traditional to low-emission their homes warm and comfortable, especially stoves. The reasons for why households are compared with those who live in apartment willing or unwilling to change their stoves are buildings and have access to district heating. examined as well as the foreseeable obstacles Most ger-area households rely on raw coal that may prevent households from adopting which they burn in their main heating stove to 33 Stocktaking Report of the Mongolia Clean Stoves Initiative keep warm. Only a small number of financially 84 and 70 percent of the households keep their better-off households (about 9 percent) use stoves burning continuously throughout the supplemental heating devices, mainly electric morning and evening, respectively (table 4.3). space heaters (table 4.1). Slightly less than one Only about 27 percent of the households keep percent use heat pumps or gas space heaters. their stoves burning continuously through The portion of households that use supplemental the afternoon, and only 9 percent keep their heating is only 3 percent higher than in 2007/08, stoves burning through the night. As expected, which is within the statistical margin of error. during the peak winter months, just about every household keeps its stove burning continuously Table 4.1. Supplemental Heating Devices through the morning and evening (96 and 94 percent, respectively); about three-quarters keep Percent their stoves burning through the afternoon (77 Supplemental Number of of all percent); and just over half keep their stoves heating devices households households burning through the night (58 percent). Electric space heater 13,622 8.3% Heat pump 492 0.3% These findings confirm that level of heating Gas space heater 821 0.5% stove utilization coincides with variation in Total households 164,127 the severity of air pollution in the city. Hourly Source: World Bank Clean Stove Initiative Survey, July measurements of PM2.5 from June 2008 to May 2012. 2009 at the Western Cross monitoring station located in the city center, fairly close to the ger areas in the north and west, showed that Each day, ger-area households must repeatedly winter concentrations increased sharply in the fire their stoves and add fuel to maintain or morning and evening hours when most ger- increase room temperature. As shown in table area households were using their stoves. Source 4.2, during the shoulder months from September apportionment methods and pollution dispersion to November 2011 and from March 2012 to May modeling have confirmed that ger stove 2012, households reportedly fired (cold-started) emissions are the primary source of very small their stoves an average of between 1 and 2 times particulate matter in the city air, representing 60 and added fuel another 1 to 2 times over a 24- percent of population-weighted exposure to PM2.5 hour period. During the peak winter months on average.36 The World Bank study also showed from December 2011 to February 2012, they that peak pollution coincides with the lighting and fired their stoves more than two times a day and refueling phases. added fuel three times. The survey finds no relationship between Table 4.2. Number of Times per day household income and the number of times Households Fire their Stoves and Add Fuel households fire their stoves or add fuel. This means that financially better off households do Number of Number of times Month cold starts per day fuel is not fire their stoves and/or add fuel more often per day added than lower- income households. This finding Sept - Nov 2011 1.6 1.3 confirms that households only fire their stoves Dec 2011 - or add fuel when needed. 2.3 2.9 Feb 2012 Mar - May 2012 1.7 1.3 Mornings and evenings are the peak times for heating. During the shoulder months, about Source: World Bank Clean Stove Initiative Survey, July 2012. 36 World Bank, Mongolia: Air Quality Analysis of Ulaan- baatar: Improving Air Quality to Reduce Health Impacts (Sustainable Development Series Discussion Paper, De- cember 2011). 34 CHAPTER 4. Household Preferences and Perceptions of Heating Stoves and Air Pollution Table 4.3. Percentage of Households that During the summer, the vast majority of Keep their Stove Burning Continuously in the households either use electric stoves, LPG Morning, Afternoon, Evening, and Night stoves, or firewood for cooking. Some households cook outside their gers or houses, Month Morning Afternoon Evening Night either by moving their main heating stove out Sept - Nov 84 27 70 9 of their homes or by using another stove, which 2011 they may keep in a small shed on their hashaa Dec 2011 - that serves as a summer kitchen. The survey 96 77 94 58 Feb 2012 confirms that 92 percent of households use Mar - 84 28 70 9 electricity as a source of fuel for cooking during May 2012 the summer months, 19 percent use LPG, and Source: World Bank Clean Stove Initiative Survey, July 17 percent use firewood. Other less common 2012. cooking fuels in the summer are coal, sawdust and briquettes, each accounting for less than 1 4.2 Cooking Habits percent of households. During the winter months, the vast majority, Figure 4.2 shows that low-emission stove users 76 percent, of households use their coal- are much less likely than traditional stove users burning stoves for both heating and cooking. to cook with their heating stoves. About half This is normal practice in Mongolia and is the of low-emission stove users say they cook on major reason why households prefer stoves their heating stoves versus about 80 percent of that are good for heating as well as cooking. traditional stove users. Households with low- Some households also use electric stoves for emission stoves also tend to cook more with cooking, and, recently, some have even begun electric hot plates during the winters than those to use liquefied petroleum gas (LPG) stoves. with traditional stoves. As will be discussed in a Households either use these electric and LPG subsequent section, the difference suggests that stoves for supplementary cooking, with their the low-emission stoves currently are less easy to heating stove being their main stove for cooking, cook with than traditional stoves. Low-emission or vice versa. In other words, most households stove users have already begun to adjust their do not rely on a single stove for all their cooking. cooking habits in the winters to their new stoves. About the same proportion of traditional and low- emission stove users use the electric hot plate Figure 4.1. Percentage of Households Using for cooking during the summers. Different Stoves for Cooking During Winter 90% Recognizing that women traditionally do most of 80% 7% the cooking in ger-area households, the survey 70% Sometimes Always also took a more granular look at the cooking Percentofhouseholds 60% appliances that female respondents said they 50% use during the winter (Figure 4.3). About 90 40% 76% percent of women who use traditional stoves 30% said that they always use their heating stove to 13% 20% 11% do their cooking, and another 2 percent said 10% 18% 13% 6% they sometimes use their stove for cooking. By 6% 0% contrast, 61 percent of women with low-emission Heatingstove Electrichotplate Electricstove LPGstove stoves said they always use their heating stoves Source: World Bank Clean Stove Initiative Survey, July to do their cooking, while 7 percent said they 2012. sometimes use their stoves for cooking. Female respondents whose households have purchased 35 Stocktaking Report of the Mongolia Clean Stoves Initiative low-emission stoves tend to use electric hot 4.3 Household Preferences and Perceptions plates to do more of their cooking than those of Heating Stove Performance with traditional stove users. Thirty-six percent of female respondents whose households own While price is an important element in a low-emission stoves said they always use an purchasing decision, it alone does not guarantee electric hot plate to do their cooking, versus only rapid penetration, nor does it guarantee use after 16 percent of women whose households rely on purchase because all consumers, including the traditional stoves for heating. poor, have different preferences. Low-emission stoves are not expected to be perfect substitutes for traditional stoves. As written in a report for the Figure 4.2. Appliances Used for Cooking by Traditional and Low-Emission Stove Owners World Bank’s Ulaanbaatar Clean Air Project, “No During the Winter* one local model may meet all user preferences... All clean-stove technologies per se are imported Traditionalstoveusers LowͲemissionstoveusers because the traditional technology used ubiquitously in Ulaanbaatar is highly polluting. 83% Useheatingstovetocookduringwinter Therefore, the impacts of the importation of 55% stoves have less to do with the actual production and more to do with the way in which Mongolian 13% Useelectrichotplatetocookduringwinter 33% consumers and/or producers absorb the technology and accept it in their daily life. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Changing the technology requires new ideas from Percentofhouseholds outside to be adapted in the Mongolian context.”37 Note: * includes households who “always” or “sometimes” Achieving rapid adoption rates for low-emission use the above appliances to cook stoves will depend not only on price but also Source: World Bank Clean Stove Initiative Survey, July on the features of the stoves and how they are 2012. perceived by consumers in the local market. With this in mind, the survey asked households about Figure 4.3. Stoves Used for Cooking by Female stove performance based on a list of key features Respondents During the Winter* aside from price. Survey results for the perception 100% of low-emission stove features are compared with 92% 90% Femaletraditionalstov results for the perception of traditional stoves Percentoffemalerespondents 80% FemalelowͲemissionst because it is assumed that new products would 70% 67% need to at least match or outperform existing 60% stoves across a wide range of features. 49% 50% 40% 4.3.1 Preferred Characteristics and 30% 22% Design Features of Stoves 20% 10% Designing a stove to fit the needs of consumers 0% is essential to ensuring acceptance by the local Heatingstove Electrichotplate market. Stoves that do not meet the needs of consumers or lack key features are likely to Note: * includes respondents who “always” or “sometimes” be rejected. To assess preferences, the survey use the above appliances Source: World Bank Clean Stove Initiative Survey, July 2012 asked respondents about 13 different stove design features they might consider important 37 Ulaanbaatar Services Improvement Project II Project Implementation Unit, “Mongolia: Ulaanbaatar Clean Air Project Safeguards Report” (internal management re- port, September 29, 2011). 36 CHAPTER 4. Household Preferences and Perceptions of Heating Stoves and Air Pollution when purchasing a stove. Respondents ranked Figure 4.4. Importance of Stove Design the features as: (i) very important, (ii) important, Features to Ger-Area Consumers or (iii) not so important (see figure 4.4). Veryimportant Important Retainsheatforlongtime 71% 28% The results indicate that the ability to heat Fuelefficiency 64% 34% a room and stay warm is seen as the most Heatsdwellingquickly Stovemadeofthickironmaterial 64% 56% 39% 35% important aspect of stoves by consumers. Abilitytouseforbothheatingandcooking 55% 40% Goodqualityanddesign 48% 49% As shown in figure 4.4, about 70 percent of Provisionofsubsidy 47% 43% respondents ranked heat retention as a very Stoveproduceslesssmokeandsoot 44% 53% Easeofuse 34% 62% important design feature; about 65 percent said Price 33% 46% Easeofstarting/firing 29% 65% that the ability of the stove to heat up quickly Shapeandappearance 24% 65% was very important. This finding underscores the 0% 20% 40% 60% 80% 100% percentofrespondentswhoagreed importance of stoves for heating in a very cold climate. The third most-important characteristic Source: World Bank Clean Stove Initiative Survey, July is “fuel consumption.” Thickness of the iron 2012. material, and ability to use the stove for both heating and cooking are ranked as the fourth and 4.3.2 Households’ Perceptions of the fifth most-important characteristics. The least Performance of Their Existing Stoves important features were shape and appearance, and ease of lighting. In summary, households Replacing traditional stoves with low-emission look for something that works, is convenient stoves requires that the new stoves perform (retention of heat means less need for refueling, as well or better than households’ existing not a popular task in early hours), and saves on stoves on important features, so it is important fuel expenditures. to understand what households think of their current stoves. To gauge how satisfied In addition, it is important to note that only households are with their existing stoves, a third of the households think that price is the survey asked respondents to rank the very important when buying a new stove. This performance of their stoves on four key aspects may be due partially to a few factors. First, related to heating: (i) the ability to retain heat for the high level of the subsidy, including the a long time; (ii) the ability to heat a large room; elements financed by MCA-Mongolia and CAF, (ii) the ability to control heat to a comfortable significantly reduced the price of low-emission level; and (iv) the ability to heat up quickly. The stoves. Second, heating stoves are an absolute results are presented in figure 4.5. The survey necessity for ger-area households, and good finds that about two-thirds of the households are stove performance is crucial to staying warm satisfied with the heating performance of their and comfortable through the long winters. current stoves, with the exception of the ability A corollary of this argument is that even if to control heat. This finding reflects that most low-emission stoves are distributed for free, heating stoves used in Mongolia are relatively households may still be unwilling to give up simple and are not equipped with heating control their old stoves if the new stoves lack important devices. There are no significant differences design features. The government did try stove in perceptions of their system’s heating ability giveaways in previous programs several years between households using traditional stoves ago that failed for this reason. and those using low-emission stoves. There was also no significant difference in opinion between households in different dwelling types or across different income quintiles. 37 Stocktaking Report of the Mongolia Clean Stoves Initiative Figure 4.5. Rating of Stove Heating Abilities users. As can be seen in figure 4.6, views among Good Fair Poor traditional and low-emission stove users on the safety, ease of lighting/firing, and durability of 67% Abilitytoretainheatforalongtime 8% 24% their stoves were mostly similar. 67% Abilitytoheatdwellingquickly 9% 23% By comparison, however, there was a much 66% larger divergence in opinion among traditional Abilitytoheatalargespace 24% 9% and low-emission stove users on the amount of 36% Abilitytocontrolheattocomfortablelevel 28% 30% smoke and soot created by their stoves. Nearly 0% 20% 40% 60% 80% 90 percent of traditional stove users agreed that percentofrespondentswhoagreed their stoves create a lot of outdoor air pollution, versus only 45 percent of the low-emission stove Source: World Bank Clean Stove Initiative Survey, July 2012. users. Around 55 percent of traditional stove- users agree that their stoves create a lot of smoke indoors, compared with only 10 percent of In addition to heating ability, the survey also low-emission stove users. These findings confirm collected information of household perceptions that most low-emission stove users are confident of stove performance on five other aspects: (i) that their stoves do not contribute to outdoor safety; (ii) ease of lighting/firing; (iii) durability; or indoor air pollution. They also indicate that (iv) smoke and soot production; and (v) use cleaning and maintenance is easier with the low- for cooking. Views of stove performance were emission stoves because they produce less soot. compared for traditional and low-emission stove Figure 4.6. Views of Stove Performance among Traditional and Low-Emission Stove Users Traditionalstoveusers LowͲemissionstoveusers HEATINGABILITY Stovecanheatroominashorttime 67% 57% SAFETYFEATURES Stovehasgoodsafetyfeatures 73% 69% EASEOFFIRING Stoveisnotdifficulttolight 76% 75% DURABILITY Lidontopofstovewon'tbreakeasily 86% 78% Handleonashboxwon'tbreakeasily 82% 78% SMOKEANDSOOT Stovecreatesalotofindoorairpollution 54% 10% Stovecreatesalotofoutdoorairpollution 88% 45% Stovecreatesalotofsmokeandsoot 61% 17% Sootmustbefrequentlycleanedfrom… 56% 34% COOKING Stovecanbeusedforcookingwithlargepots 60% 41% Itiseasytocookwiththestove 85% 52% 0% 20% 40% 60% 80% 100% Percentwhoagreed Source: World Bank Clean Stove Initiative Survey, July 2012. 38 CHAPTER 4. Household Preferences and Perceptions of Heating Stoves and Air Pollution Differences were also seen among traditional Among the subset of households that have and low-emission stove users on ease of cooking. yet to acquire low-emission stoves, 72 percent As was discussed above, the ability to cook with of respondents living in gers, 85 percent of heating stoves is important because the vast respondents living in detached houses without majority of households in the ger areas use their heat walls, 64 percent of respondents in homes stoves for both heating and cooking during the with heat walls, and 60 percent of those with winter months. Nearly 90 percent of traditional LPBs said they would be interested in replacing stove users agreed their stoves are easy to cook their stoves. with, compared to only half of low-emission Among the traditional stove users who indicated stove users. Furthermore, about 60 percent of interest in switching out their stoves, the traditional stove users agreed that their stoves “Silver Mini” appears to be the most preferred can be used for cooking with a large pot or wok, replacement model (named by 29 percent of versus only 41 percent of low-emission stove respondents), followed by the “Silver Turbo,” users. Views of stove cooking performance were and the “Royal Single” (both named by 14 similar among female and male respondents. percent of respondents). Aside from low- Respectively, 90 and 85 percent of female and emission stoves, about 20 percent of households male traditional stove users agreed that it is still using traditional stoves are interested in easy to cook with their stoves, versus only 52 replacing their stoves with LPBs. Another 2 percent of both female and male low-emission percent would like to change to low-emission stove users. Similarly, 63 percent and 59 percent LPBs. Only a very small minority (4 percent) of of female and male traditional stove users traditional stove users said they are interested respectively agreed that their stoves can be in replacing their existing stoves with another used to cook with large pots, versus 59 and 42 traditional stove. percent of female and male low-emission stove users. (The differences are within a statistical margin of error.) Perceived difficulties with Table 4.4. Preferred Replacement Models for cooking may reflect the fact that new models of Traditional Stove Users low-emission stoves recently introduced to the Number of Ulaanbaatar market have smaller combustion Type of stove Percent Households chambers. Because large pots are common in Silver mini 24,455 29 traditional Mongolian cooking and preferred Low pressure boiler 17,069 20 by ger households, introducing a stove with a Silver turbo 11,981 14 large burner or combustion chamber that can Royal single 11,653 14 accommodate the large pots would be more Other 6,073 7 suited to household cooking habits and customs. Locally made low- 5,580 7 emission stove 4.4 Willingness to Change Stoves Traditional stove 3,447 4 Royal double 2,626 3 The survey asked whether households were Low-emission low interested in replacing their current stoves. pressure boiler 1,641 2 About 60 percent (97,656) of all households stove expressed interest, either in the short or long Total 84,525 100 term. Among those households still using Source: World Bank Clean Stove Initiative Survey, July traditional stoves, the share of respondents 2012. interested in changing their stoves was higher, at 67 percent (equal to 84,525 households)- of which 83 percent said they are interest in replacing their stoves in the near term. 39 Stocktaking Report of the Mongolia Clean Stoves Initiative 4.4.1 Reasons for Resistance to Changing 4.4.2 Perceived Difficulties with Using Low- Stoves Emission Stoves For households that are not interested in Perceptions of difficulties using low-emission changing their stoves, several reasons were stoves differ between traditional stove users noted for their reluctance. As shown in figure and households already using them (figure 4.6, the two most cited reasons were that they 4.7). In most cases, fewer than half of traditional are used to their existing stoves (92 percent) stove users believed they would encounter and that their existing stoves are still in good difficulties in using a low-emission stove. On working condition (87 percent). Only a third of the one hand, traditional stove users tended to respondents agreed that financial reasons would over-estimate, compared with responses from prevent them from replacing their stoves, and users, the difficulties with installing the low- about 40 percent cited potential difficulties with emission stoves and operating them. They were installing a new stove in their homes. also uncertain about where they could buy low- emission stoves. On the other hand, they tended Because adoption rates for low-emission to under-estimate difficulties with lighting, stoves among households in detached homes refueling the stoves, and cooking with them. with heating walls were so much lower than Also, while only 25 percent of traditional stove households in detached homes without heating users expected that low-emission stoves would walls, reasons for why traditional stove users in take a long time to heat a ger, about 44 percent these two groups would not want to replace their of households already using a low-emission stoves were also examined. The results show stove agreed that this is a problem. that the perceived difficulty of fitting a heat wall to a new stove is a significant barrier to these households replacing their stoves. About 24 Figure 4.8. Perceived Difficulties with Using a Low-emission Stove (Comparing those who have percent of respondents said that their existing and have not used low-emission stoves already) heat wall would not fit a new stove. HaveusedlowͲemissionstove Haven'teverusedlowͲemissionstove StovemustbecoldͲstartedtwiceaday 70% Figure 4.7. Household Reasons to Not Change 45% Can'tuselargepottoboilwaterorcook 64% Their Stove 48% Stovecan'tberefueledeasily 60% 42% Can'tusestovetoboilwaterorcookatall 51% 41% Weareusedtousingourcurrentstove 92% Havetoreadinstructionstousethestove 48% 62% Currentstoveisstillgoodenough 87% 44% Stoveisslowtowarmupdwelling 25% Newstovewouldbedifficulttoinstall 41% Stovehasasmallfiringchamber 43% 36% Fuelsusedbyhouseholddon'tmatchstove 43% Financialreasons 33% 36% Stovecancauseheatwalltoblowup 32% LowͲemissionstovemodelscurrently 44% 26% beingsolddon'tfitourneeds Highpriceofstove 31% 41% Stovewasinherited 16% Don'tknowwheretobuystove 25% 58% Stoveisdifficulttoinstall 25% Currentstovewasaspecialgift 6% 40% Stoveisdifficulttooperate 21% 26% 0% 50% 100% Increasedfuelexpenditures 18% percentofrespondentswhoagreed 11% (ofthosenotinterestedinchangingstove) 0% 10% 20% 30% 40% 50% 60% 70% 80% percentofrespondentswhoagreed Source: World Bank Clean Stove Initiative Survey, July 2012. Source: World Bank Clean Stove Initiative Survey, July 2012. 40 CHAPTER 4. Household Preferences and Perceptions of Heating Stoves and Air Pollution 4.5 Source of Information and Knowledge portion of households reportedly learned about about Low-Emission Stoves low-emission stoves from billboards (18 percent) or XacBank (9 percent). In terms of The survey’s findings demonstrate that mass effectiveness, direct communication from the media and word of mouth have been the most implementing organization was highly effective prevalent means of reaching consumers (figure in promoting the new stoves: about 70 percent 4.8). In general, households in the ger areas of low-emission stove users said they had heard are already quite familiar with low-emission about the stoves from the project implementer. stoves. Only 1 percent of respondents had Word of mouth was less effective. Respondents never heard of the stoves. About 81 percent who said they had learned about low-emission of households reportedly learned about the stoves from friends, neighbors, or relatives were stoves from radio or TV, and 72 percent said about 35 percent less likely to be low-emission they had heard about the stoves from friends, stove owners. If stove performance is not well- neighbors, or relatives. Other important sources appreciated, it may be very difficult to change of information about low-emission stoves cited minds. New entrants to the market will need to were organizations involved in implementing be careful not to overpromise or under-deliver on the stove replacement program (e.g., MCA- stoves to avoid this negative feedback loop. Mongolia), and the newspaper. Only a small Figure 4.9. Where Did You Hear about Low-Emission Stoves? 81% Radioand/orTV 69% 85% 72% Friends,neighbors,orrelatives 51% 79% 34% Implementingorganization 69% 23% 33% Newspaper 36% 32% 18% Billboard 17% 18% 9% XacBank 10% 9% Total 3% HaveusedlowͲemissionstovealready Stovemaker 2% 3% Haven'tusedlowͲemissionstove 1% NeverheardoflowͲemissionstoves 1% 1% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% percentofrespondents Source: World Bank Clean Stove Initiative Survey, July 2012. Households also have fairly good knowledge agree that low-emission stoves produce less about the features of low-emission stoves. As smoke and soot, use less fuel, retain heat longer, shown in figure 4.9, most households surveyed are easier to start, and are priced low because agree that the stoves have beneficial aspects. For of the subsidy program. As might be expected, example, more than 60 percent of the households households that had already acquired low- 41 Stocktaking Report of the Mongolia Clean Stoves Initiative emission stoves were much more aware of the emission stove users were less likely to agree benefits than households without the stoves, that the stoves were not suitable for use with as well as some of the inconveniences. The some types of heat walls, and were less likely difference in knowledge is quite significant for to say that the stoves were too expensive. On most features, except fuel consumption. About the question of cooking, about 74 percent of the 70 percent of both users and non-users of low- households using low-emission stoves agreed emission stoves agreed that the stoves use less that low-emission stoves are good for heating but fuel than traditional stoves. Moreover, low- not for cooking, versus about half of non-users. Figure 4.10. Knowledge about Low-Emission Stoves Total HaveusedlowͲemissionstove Haven'tusedlowͲemissionstove 75% Emitslesssmokeandsootthanatraditionalstove 90% 70% 74% Stovepriceislowbecauseit'ssubsidized 91% 69% 68% Useslessfuelthanatraditionalstove 68% 69% 65% Retainsheatbetterthanatraditionalstove 82% 60% 63% Createslesssmokewhenfiringthestove 84% 56% 55% Goodforheating,butnotforcooking 74% 49% 54% EasiertocoldͲstartthantraditionalstove 72% 48% 37% Tooexpensive 28% 40% 34% Can'tbeusedwithsometypesofheatwalls 24% 37% 28% Moredifficulttousethantraditionalstoves 30% 27% 16% Requirescleaningchimneymoreoften 26% 13% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% percentofrespondentswhoagreed Source: World Bank Clean Stove Initiative Survey, July 2012. 4.6 Household Attitudes toward Air Pollution percent think that it is high (see figure 4.10). Overall, attitudes toward air pollution in the city Virtually all households in the six surveyed ger among ger-area residents are the same as they districts think that air pollution in Ulaanbaatar were in 2007/08, though about 18 percent more is very bad. In fact, about 90 percent of respondents in 2012 said that air pollution in the households think that air pollution in Ulaanbaatar is extremely high versus in 2007/08. Ulaanbaatar is extremely high and another 9 42 CHAPTER 4. Household Preferences and Perceptions of Heating Stoves and Air Pollution Figure 4.11. Opinion of Air Pollution Problem by a new technology may be difficult to absorb. Without sufficient attention to training users, Low Donotknow Acceptable 0.1% 0.8% households may “misuse” the stove, which can 0.4% then significantly under-perform, use more fuel High than expected and, worse, emit more particulates 8.9% than shown in laboratory emissions testing results. The study confirms that heating ability, including retention of heat, is the most important feature of stoves to consumers, followed by fuel efficiency. The thickness of the metal material Extremelyhigh (durability) and the ability to cook with the stove 89.8% are also important. Preferred design features appear to coincide with the fact that households start fire and add fuel to their stoves only as needed. Price was a much less important factor Source: World Bank Clean Stove Initiative Survey, July cited by households in shaping their purchasing 2012. decision. In general, most households are satisfied with 4.7 Conclusion their current stoves, especially when it comes to Household heating habits are characteristic of durability, ease of lighting, and safety. However, living in cold winters and budget constraints. a comparison of perceived stove performance Households only fire (cold start) their stoves among traditional and low-emission stove users or add fuel as needed, suggesting a good reveals that there are also some important understanding of traditional stove operation. The differences. About 90 percent of traditional main determinants for how often households stove users believe that their stoves are a major fire and add fuel to their stoves are outside contributor to outdoor air pollution, versus only temperature and desired level of comfort within 45 percent of low-emission stove users. About budget constraints. Other determinants are type two-thirds of low-emission stove users are highly of heating system used, and the size and heating satisfied with the ability of their low-emission characteristics of the home. A relevant habit is stoves to heat their homes-more so than the use of heating stoves for both heating and traditional stove users. At the same time, more cooking during the winter months. The survey than half of low-emission stove users say that reveals that about 83 percent of households their stoves are not easy to cook with, while only use the same stove for heating and cooking in about 41 percent of traditional stove users say the winter. This is a common practice among the same. The perceived difficulties of cooking ger-area households and the main reason with the low-emission stoves currently on the households prefer stoves that can be used for market have already begun to change household both heating and cooking. cooking habits, as low-emission stove users are less likely to use their heating stove for cooking Designing stoves to meet the needs of consumers in the winter and more likely to use electricity. is essential to ensuring acceptance by the This may be a risk to sustainability if electricity local market. The ingrained heating habits are prices rise and push households back to more consistent across stove market segments and convenient coal fired cooking stoves. Despite can present barriers to stoves that do not fit well this, as will be discussed in the next chapter, within these habits. Even if new stoves are used, there are no discernible differences in reported heating habits are hard to break so new methods electricity use among low-emission stove users of lighting and refueling, for example, required and non-users. 43 Stocktaking Report of the Mongolia Clean Stoves Initiative The survey confirms that a significant numbers stoves are still in good working condition. of households are still interested in acquiring Only a third cited financial reasons. Perceived new low-emission stoves. About 70 percent of difficulties in switching to low-emission stoves traditional stove users expressed interest in among current non-users included uncertainty replacing their stoves, nearly all of whom wanted about where to buy the stoves, concerns about to acquire a new low-emission stove instead of not being able to use the stoves for cooking, another traditional stove. Of those respondents price, difficulty of installing the stoves, and the who said they are not interested in replacing fear that the stoves would not work properly with their stoves, the most-cited reasons were that heat walls. they are used to their existing stoves or their 44 CHAPTER 5. HEATING FUEL CONSUMPTION AND EXPENDITURES Emissions reductions are optimal when stove to gain an understanding of the actual, on- and fuel match. Previous pilot programs the-ground situation as of August 2012 and focused on fuel saving, rather than combustion reflecting on what happened during the winter efficiency, which is actually more directly or 2011/2012. Results are then compared to the related to emissions reductions. Nevertheless, time of the previous survey in 2007/08. the fuel-saving characteristics of stoves are Raw coal and firewood remain the dominant very important in decisions by households fuels used for heating by ger households in 2012, considering switching to new products, similar to the situation in 2007. Nearly every especially for lower-income households. household in the ger district relies on these two Ulaanbaatar experiences long, harsh winters, fuels for heating and cooking during the winter and households living in the ger districts must months. In practice, households use raw coal as spend a significant percentage of their income their primary heating fuel, while firewood is used (between 16 to 31 percent) on heating fuel to light the fire in their stoves. As shown in Table each winter. While raw coal continues to be the 5.1, a small number of households also use other main heating fuel for 98 percent of households fuels such as processed coal (semi-coke coal and in the ger areas surveyed, a small number coal briquettes), sawdust, sawdust briquette, of households use other fuels. The following dried animal dung, and “anything that burns.” chapter provides a detailed analysis of types The use of “anything that burns” (such as waste, of heating fuels, estimated coal and firewood clothes, bones, and old tires) to supplement consumption, and household expenditures on all coal appears to be a coping mechanism by very heating fuels. poor households. The reported monthly income of these households is less than 380,000 Tg, 5.1 Types of Fuels Used by Households placing them well within the bottom quintile. Emissions reductions are achieved when a stove The use of semi-coke coal and other processed is used with an appropriate fuel and according to fuels such as sawdust briquettes is mainly instructions. The previous chapters have focused determined by the households’ district of on stove use and preferences, but this is only residence. As illustrated in Table 5.2, most half of the equation. Trends in fuel use need to of the households that do use semi-coke be known so that new stove technologies can coal, compressed coal briquettes, or sawdust be designed to match them. There have been briquettes live in Bayangol District, where a ban initiatives by the central government and the on raw coal has been imposed. While only about Municipality of Ulaanbaatar to declare “no raw 8,400 households use semi-coke coal in the ger coal” zones, including one for Bayangol District, areas of the six districts surveyed, 68 percent which was part of the survey. The survey (5,750) of these households live in Bayangol. compared penetration rates for alternatives Sixty-six percent of the households that use to raw coal across the six central ger districts compressed-coal briquettes and 72 percent of the 45 Stocktaking Report of the Mongolia Clean Stoves Initiative households that use sawdust briquettes live in Bayangol. The relatively high rates of processed fuel use in Bayangol belie the fact that only a minority of households in Bayangol use these fuels on a regular basis. Most of the households that do use semi-coke coal or briquettes use them only rarely or occasionally, as evidenced by Table 5.3. Table 5.1. Heating Fuels Used by Households Semi- Compressed Sawdust Animal Anything Coal Firewood coke coal Sawdust briquette dung that burns coal briquette Households 161,009 154,608 8,370 10,012 2,626 9,355 9,027 3,283 Percent 98% 94% 5% 6% 2% 6% 6% 2% Source: World Bank Clean Stove Initiative Survey, July 2012. Note: Some households use more than one type of heating fuel. Table 5.2. Raw Coal and Processed Fuel Use by Households in Six Districts (Percent of households in district that use the following fuels) Songino- Fuel type Bayanzurkh Bayangol Sukhbaatar Khan-Uul Chingeltei khairkhan Raw coal 100% 92% 97% 98% 100% 100% Semi-coked coal 2% 39% 2% 1% 1% 1% Compressed-coal 1% 43% 2% 2% 1% 10% briquettes Sawdust briquettes 2% 44% 0% 2% 0% 8% Source: World Bank Clean Stove Initiative Survey, July 2012. Note: Some households use more than one type of heating fuel. Table 5.3. How Often Households in Bayangol District Use Processed Fuels (Percent of households surveyed in Bayangol) Compressed coal Sawdust Carbonized sawdust Semi-coke briquettes briquettes briquettes coal All the time 9% 14% 10% 18% Sometimes 14% 19% 13% 13% Rarely 14% 9% 12% 6% Haven't used 64% 59% 66% 63% Total 100% 100% 100% 100% Source: World Bank Clean Stove Initiative Survey, July 2012. Note: Some households use more than one type of processed fuel. Fuel characteristics are also important when Ash may be needed to reduce the calorific value designing low-emission stoves and the entire or as a binding material for briquetting, but this burn cycle of the fuel needs to be taken into has the counterproductive effect of reintroducing account. Semi-coke coal may have less ash the ash content that was initially removed during content in bulk, but it may burn too hot and the coking process. In addition, coking reduces damage the stove due to its high calorific value. VOC (volatiles) content. This makes the fuel 46 CHAPTER 5. Heating Fuel Consumption and Expenditures difficult to light, requiring more wood (which Coal from Alagtolgoi has an average heating produces particulates) or an alternative lighting value in the range of 6,200 kilocalories per technique. The World Bank’s urban air pollution kilogram, while both Nalaikh and Baganuur coal study in Ulaanbaatar noted that the cyclical have a typical heating value of 3,500 kcal per kg. nature of peak particulate pollution coincided with In practice, many households-especially those the lighting cycles of stoves.38 So, while coking that buy coal by the bagful-will use coal from products may be comprised of less VOCs and ash, multiple sources. the need for more fuel (e.g. firewood) to ignite Some households may try to buy coal from the the semi-coke coal may negate these benefits. same trader to make sure that they get the same In any case, emission factors cannot be known type of coal, but there is no guarantee of source unless they are tested in a stove. To date, testing or quality. Households that buy coal in bulk results of semi-coke products in traditional or low- could, more or less, use the same type of coal for emission stoves are not in the public domain. most or the part of the winter season, depending With respect to the types and sources of raw on whether they need to replenish their coal and coal, Nalaikh coal remains the most popular use the same coal trader. coal in 2012, as it was in 2008. About 79 percent The size of coal can have important impacts on of the households reported using Nalaikh coal combustion. Reducing the size of raw coal pieces during the 2011/12 winter (compared to 76 and making them more uniform can help coal percent in 2007/08). Alagtolgoi and Baganuur burn more completely-the key objective of a low- coal are a distant second and third. Coal is sold emission coal stove initiative. However, on the in the market either in bulk (typically by the basis of observation, it is found that the current truckload) or in small bags without labeling, so market is far from standardized and the pieces sold households tend to rely on information from coal in bags and in bulk are of many different sizes. traders and/or their own experience to judge the heating value of different types of coal. 5.2 Total Household Heating Fuel Expenditures Figure 5.1. Sources of Coal Used by Households Heating expenditures are seasonal and need 79% to be compared with monthly winter income 80% to understand the real impacts on ger area 70% households, especially the poorer ones with weak Percentofhouseholds 60% 50% savings potentials. Total spending by households 40% on all heating fuels, including raw coal, firewood, 30% 23% semi-coke coal, and briquettes, averaged 637,393 20% 20% Tg during the cold season from September 2011 10% 2% 2% to May 2012 (table 5.4). An analysis of spending 0% by income bracket reveals that households in the Nalaikh Alagtolgoi Baganuur Tavantolgoi Sharyngol (129,660) (38,077) (33,318) (3,775) (3,447) bottom income quintile spent a third (31 percent) Typeofcoalandnumberofhouseholds of their monthly income on heating fuels in the winter. As shown in table 5.4, although spending Source: World Bank Clean Stove Initiative Survey, July on heating fuels increases in absolute terms as 2012. income rises, the proportion of income allocated for heating fuel declines. The top two income quintiles only allocated 10 percent or less of their income on 38 heating fuels during the previous winter months. World Bank, Mongolia: Air Quality Analysis of Ulaan- baatar: Improving Air Quality to Reduce Health Impacts It is clear that heating fuels place a heavy burden (Sustainable Development Series Discussion Paper, De- on very poor households, though the situation has cember 2011). 47 Stocktaking Report of the Mongolia Clean Stoves Initiative improved somewhat compared to 2007/08 when apparently large increase in total spending on households in the bottom income quintile reported heating fuels was due mostly to higher fuel prices spending 42 percent of their monthly income over the past four years, but also to a small increase during the winter months on heating fuel (figure in tons of coal consumed. Regardless, it is clear that 5.2). This improvement has been largely a result the financial burden placed on ger-area households of an increase in cash income, given that reported by heating fuel expenditures is still high when monthly spending on heating fuel increased by compared to households in non-ger areas with 57 percent in real terms since 2007/08.39 This access to district heating. Table 5.4. Average Household Monthly Spending on all Heating Fuels by Income Quintile Total expenditure Expenditure as % Expenditure per Income quintile on heating fuels of monthly income heating month (Tg) (Tg) during winter 380,000 Tg/mo or less 575,812 74,378 31 Valid N 32,661 33,154 33,154 380,001 to 550,000 Tg/mo 610,184 81,540 18 Valid N 33,482 33,482 33,482 550,001 to 758,000 Tg/mo 652,453 84,170 13 Valid N 31,841 31,841 31,841 758,001 to 1,000,000 Tg/mo 669,670 88,630 10 Valid N 35,780 35,944 35,944 > 1,000,000 Tg/mo 680,750 90,599 6 Valid N 29,707 29,707 29,707 Total 637,393 83,798 16 Valid N 163,470 164,127 145,088 Source: World Bank Clean Stove Initiative Survey, July 2012. Note: All heating fuels include raw coal, firewood, briquette, and semi-coke coal. Valid N is valid number of households for the variable shown. Figure 5.2. Heating Fuel Expenditure as Percentage of Monthly Income39 45% 42% Percentofmonthlyhouseholdincome 40% 2007 2012 34% 35% 30% 26% 25% 21% 20% 18% 18% 16% 14% 13% 15% 11% 9% 10% 7% 5% 0% First Second Third Fourth Fifth Allquintiles (bottom) (top) Incomequintile Source: World Bank Clean Stove Initiative Survey, July 2012. Note: See Figure 5.3 for delineation of income quintiles in 2007 and 2012. 39 Adjusting for inflation, average monthly spending on heating fuels for households in the ger areas was 405,858 Tg per month in 2007/08 (expressed in 2011/12 prices). Prices are adjusted based on the IMF’s monthly consumer price index data for Mongolia for the periods of September 2007 to April 2008 and September 2011 to April 2012. IMF, “Data and Statistics,” http://www.imf.org/external/data.htm#data. 48 CHAPTER 5. Heating Fuel Consumption and Expenditures A comparison of total expenditures on heating spend more of their monthly income on heating fuels by dwelling and heating system type reveals fuels because they tend to be the poorest. As that households living in detached houses expected, households living in larger homes with without heat walls and households living in LPBs spent the most on heating fuels in absolute gers spend the least on heating fuels in absolute terms. These patterns of spending on heating terms. Proportionally, however, ger households fuels are highly similar to those seen in 2007/08. Table 5.5. Average Household Monthly Spending on all Heating Fuels by Types of Dwelling and Heating System Monthly Total spending on Spending per Spending on fuels Type of dwelling and heating household heating fuels for heating month as % of monthly system income winter (Tg) (Tg) income Ger 634,770 620,346 81,544 18 Valid N 76,483 76,319 76,483 74,514 Detached house with heat wall 850,738 638,551 83,731 14 Valid N 54,326 53,998 54,326 53,326 Detached house without heat 695,613 592,160 74,957 14 wall Valid N 13,622 13,458 13,623 13,623 Detached house with LPB 957,906 731,187 98,848 16 Valid N 19,695 19,695 19,595 19,695 Total 164,127 637,393 83,798 16 Valid N 163,470 164,127 161,009 Source: World Bank Clean Stove Initiative Survey, July 2012. Households that used low-emission stoves finding confirms that the new heating stoves during the previous winter months spent slightly have allowed households to save money on less on heating fuels than those using traditional fuels-an important design feature for households stoves. As will be discussed further below, this considering replacing their stoves. Table 5.6. Average Household Monthly Spending on Heating Fuels by Types of Stove Traditional versus low- Average monthly Total expenditure per Expenditure as % of emission stoves income heating month (Tg) monthly income Low-emission stoves 770,992 84,753 15 Valid N 126,378 126,378 126,378 Traditional stoves 680,080 80,598 17 Valid N 37,749 37,749 37,749 Total 750,082 81,501 16 Valid N 164,127 164,127 164,127 Source: World Bank Clean Stove Initiative Survey, July 2012. Note: Not all households with low-emission stoves used their stoves throughout the entire winter. A significant number of households acquired low-emission stoves mid-way through the winter. 49 Stocktaking Report of the Mongolia Clean Stoves Initiative 5.3 Estimated Coal Usage and Expenditures As was seen in 2007/08, reported coal consumption by households in 2011/12 was As noted in the previous sections, the vast positively correlated with household income. majority of households in the ger areas use raw Households in the bottom income quintile coal as their main heating fuel. The survey shows consumed the least amount of coal, spending that individual households used about 4.5 tons 376,132 Tg and consuming a total of 3.7 tons for of raw coal and spent a total of 424,373 Tg from the entire winter. By comparison, households September 2011 to May 2012 (see table 5.7). In in the top income quintile spent an average all, it is estimated that the 169,124 households of 482,572 Tg on coal and consumed 5.1 tons in the surveyed area used 735,750 tons of raw for the entire winter. The difference is largely coal during the heating season. Average coal attributable to the fact that higher-income consumption per household is slightly higher than households tend to live in larger homes and the 4.2 tons reported by households in 2007/08. have higher heating requirements, especially Average expenditures on coal were 275,372 Tg in the households in detached homes with LPBs. 2007/08 (expressed in 2011/12 prices), meaning Moreover, the relatively small difference in coal that spending on coal by households increased consumption between bottom and top income by 54 percent in real terms. The increase was due quintiles suggests that demand for coal is mainly to higher fuel prices. relatively inelastic with respect to income. Table 5.7. Household Coal Usage and Expenditure by Income Quintile (September 2011 through April/May 2012) Total spending per Coal used per Total coal used Income quintile household on coal household for entire by all households for winter (Tg) winter (tons) (tons) 380,000 Tg/mo or less 376,132 3.67 119,944 Valid N 32,661 32,661 32,661 380,001 to 550,000 Tg/mo 396,365 4.26 142,502 Valid N 33,482 33,482 33,482 550,001 to 758,000 Tg/mo 412,247 4.54 144,473 Valid N 31,841 31,841 31,841 758,001 to 1,000,000 Tg/mo 457,090 4.92 176,167 Valid N 35,780 35,780 35,780 > 1,000,000 Tg/mo 482,572 5.14 152,664 Valid N 29,707 29,707 29,707 Total 424,373 4.50 735,750 Valid N 163,470 163,470 163,470 Source: World Bank Clean Stove Initiative Survey, July 2012. Figure 5.3 compares the average monthly coal of comfort during the winter months, it could consumption during the winter of 2011/12 be that higher-income households tend to live and 2007/08, showing a slight increase in in larger homes that require more coal to heat the average monthly coal consumption per (especially among LPB users). The findings also household across all income categories. do not account for differences in weather during Although this finding appears to suggest that as the 2007/08 winter and the 2011/12 winter. income rises households demand a greater level 50 CHAPTER 5. Heating Fuel Consumption and Expenditures Figure 5.3. Comparison of Average Coal ger is 28m2, versus 39m2 for a detached house Consumption per Household by Income without a heat wall, 41m2 for a detached house Categories with a heat wall, and 73m2 for a detached house with an LPB. Furthermore, household income is 5.1 5.5 2007 2012 4.9 4.9 also positively correlated with dwelling type and 5.0 4.5 4.8 4.5 Averagecoalconsumptionper 4.5 4.3 4.2 heating system. While average monthly income 4.1 4.0 3.7 3.8 for households living in gers was reported household(tons) 3.3 3.5 as 634,770 Tg, the average for LPB users in 3.0 detached homes was 957,906 Tg per household. 2.5 2.0 Coal consumption varies considerably across the 1.5 1.0 four types of dwellings and heating systems. As 0.5 shown in table 5.8, coal usage of households that First Second Third Fourth Fifth Allquintiles live in ger is the lowest, at only 3.9 tons for the entire (bottom) (top) winter. Households living in detached houses used more coal, especially those living in larger homes Source: World Bank Clean Stove Initiative Survey, July 2012, and ASTAE/World Bank: baseline Fuel Consumption, with better heating systems. Coal consumption Heating Stove, and Household Perception Survey, ranged from 4.1 tons for those in detached homes December 2007. using space-heating ger stoves without heat walls Note: Due to an increase of household income (in nominal to 6.3 tons for those in detached homes with LPBs. terms), the income quintile for 2007 and 2012 are based on Coal consumption by dwelling type and heating the following income categories system showed a similar linear trend for 2007/08, as illustrated in figure 5.4. 2007 survey 2012 survey < 111,330 Tg/ < 380,000 Tg/ Figure 5.4. Comparison of Average Coal First quintile mo mo Consumption per Household by Dwelling 111,331 to 380,001 to Type and Heating System Second quintile 172,660 Tg/mo 550,000 Tg/mo 7 Averagecoalconsumptionperhousehold 172,661 to 550,001 to 6.2 6.3 2007 2012 Third quintile 6 233,990 Tg/mo 758,000 Tg/mo 4.8 5 758,001 to 4.5 4.5 233,991 to 3.9 3.9 4.1 4.2 Fourth quintile 1,000,000 Tg/ 4 3.5 325,869 Tg/mo (tons) mo 3 > 325,860 Tg/ > 1,000,000 Tg/ Fifth quintile 2 mo mo 1 0 Ger Detachedhouse, Detachedhouse, Detachedhouse, Alldwellings 5.3.1 Comparison of Coal Usage between noheatwall withheatwall withLPB Households Living in Different Types of Source: World Bank Clean Stove Initiative Survey, July Dwellings with Different Heating Systems 2012, and ASTAE/World Bank: baseline Fuel Consumption, Heating Stove, and Household Perception Survey, Households’ demand for coal depends on December 2007. several factors, including the price of coal, the price of comparable alternative fuels, household income, type and size of home, type of heating equipment, and the desired level of comfort. This section only compares the level of coal consumption between households that live in different types of dwellings with different heating systems. As discussed in chapter 2, however, the type of dwelling is closely related to dwelling size. The average floor area of a five-walled 51 Stocktaking Report of the Mongolia Clean Stoves Initiative Table 5.8. Household Coal Usage and Expenditure by Type of Dwelling and Heating System (September 2011 through April/May 2012) Average coal Coal used per Average household Type of dwelling and heating expenditure per household for the monthly income system winter (Tg) entire winter (tons) (Tg) Ger 390,914 3.88 634,770 Valid N 76,319 76,319 76,483 Detached house without heat wall 392,786 4.09 695,613 Valid N 13,485 13,485 13,623 Detached house with heat wall 434,185 4.84 850,738 Valid N 53,998 53,998 54,326 Detached house with LPB 548,478 6.28 955,775 Valid N 19,859 19,859 19,859 Total 424,373 4.50 750,082 Valid N 163,470 163,470 164,127 Source: World Bank Clean Stove Initiative Survey, July 2012. 5.3.2 Comparison of Coal Usage between Low- important design features they consider when emission and Traditional Stoves buying a stove (see chapter 4). Currently, the most prominently used low- 5.4 Estimated Firewood Consumption and emissions stoves in the Ulaanbaatar ger areas Expenditure are the imported stoves that were subsidized by MCA-Mongolia and CAF. These low- Generally speaking, firewood is used by ger-area emission stoves use less fuel and burn cleaner households in the winter to light their stoves and than traditional stoves. As shown in table 5.9, ignite coal. Most households buy firewood by the average coal use among households that used bag from the market or roadside traders. A few low-emission stoves during the previous winter buy large logs or other large pieces. On average, months was 4.1 tons, compared to 4.6 tons for households in the six districts consumed 4.9 households that use traditional stoves, a savings cubic meters (m3) of firewood from September of 11 percent. This means that low-emission 2011 to April 2012, which is comparable to the stoves reduced household expenditures on raw 4.7 m3 reported for 2007/08. The difference falls coal by about 16,553 Tg per heating season per within the statistical margin of error. Average household. It is very important to note, however, household expenditures on firewood during this is only an estimated savings because not the winter months is estimated at 189,970 Tg. all low-emission stove users used their stoves Overall, households in the surveyed areas throughout the winter. A significant number consumed a total of 800,758 cubic meters of of households acquired their low-emission firewood during the 2011/12 winter. stoves mid-way through the winter. When controlling for type of dwelling and heating Although firewood is typically used to system, coal consumption by households that complement or supplement coal, total firewood use low-emission stoves is lower for all types of use and associated expenditures are quite dwellings and heating systems. These findings significant. Unlike coal, however, firewood confirm that low-emission stoves outperform consumption appears to be negatively related traditional stoves on fuel efficiency, which is to household income. As shown in table 5.10, important because fuel efficiency was identified firewood consumption declines for households by consumers as being one of the most in the higher income quintiles. 52 CHAPTER 5. Heating Fuel Consumption and Expenditures Table 5.9. Household Coal Usage and Expenditure by Type of Heating Stove (September 2011 through April/May 2012) Average coal used Average expenditure Average household Type of stove per household* per household (Tg) monthly income (tons) Do not use low-emission stoves 428,456 4.6 770,992 Valid N 125,721 119,813 126,378 Use low-emission stoves 410,775 4.1 680,080 Valid N 37,749 39,062 37,749 Total 424,373 4.5 750,082 Valid N 163,470 158,875 164,127 Source: World Bank Clean Stove Initiative Survey, July 2012. Note: The estimated raw coal consumption and expenditure of households that use low-emission stoves are based on raw coal consumption and expenditure from Sept 2011 through April/May 2012. This means that some households that used a low-emission stove may also have used a traditional stove at the beginning of the season and only switched to the low- emission stove at the middle or later in the season. This is due to the fact that not all low-emission stove users acquired their low-emission stove at the beginning of the season; a significant number of households acquired their low-emission stove mid-way through the winter. * Only includes household that use coal (users only). Furthermore, households living in gers consume Households that use more raw coal tend to use more firewood than households living in less firewood and vice versa. This is simple due detached houses (see table 5.11). Consumption to the fact that if households add more coal and patterns for firewood and raw coal were similar keep their fires burning longer, they will use less to those seen in 2007/08, and indicate that these firewood to start their stoves. two complementary fuels are negatively related. Table 5.10. Household Firewood Usage and Expenditure by Income Quintile (September 2011 through April/May 2012) Average firewood Total firewood used Average expenditure Income quintile used per household by all households per household (Tg) (m3) (m3) 380,000 Tg/mo or less 183,344 4.79 158,735 Valid N 33,154 33,154 33,154 380,001 to 550,000 Tg/mo 191,228 5.05 169,154 Valid N 33,482 33,482 33,482 550,001 to 758,000 Tg/mo 217,634 5.33 169,837 Valid N 31,841 31,841 31,841 758,001 to 1,000,000 Tg/mo 184,712 4.93 177,206 Valid N 35,944 35,944 35,944 > 1,000,000 Tg/mo 172,656 4.24 125,825 Valid N 29,707 29,707 29,707 Total 189,970 4.88 800,758 Valid N 164,127 164,127 164,127 Source: World Bank Clean Stove Initiative Survey, July 2012. 53 Stocktaking Report of the Mongolia Clean Stoves Initiative Table 5.11. Household Firewood Usage and Expenditure by Type of Dwelling and Heating System (September 2011 through April/May 2012) Average Average Average coal Monthly Type of dwelling and heating household expenditure used per household system monthly per household household (m3) income income (Tg) Ger 634,770 209,256 5.46 417548 Valid N 76,483 76,483 76,483 76,483 Detached house with heat wall 850,738 178,299 4.56 103,378 Valid N 54,326 54,326 54,326 54,326 Detached house without heat wall 695,613 189,212 4.76 208,552 Valid N 13,622 13,623 13,623 13,623 Detached house with LPB 957,906 147,788 3.59 71,280 Valid N 19,695 19,695 19,695 19,695 Total 164,127 189,970 4.88 800,758 Valid N 164,127 164,127 164,127 Source: World Bank Clean Stove Initiative Survey, July 2012. Households using low-emission stoves remains the most popular type of coal. However, use slightly less firewood than households other fuels are also used by households in the six with traditional stoves. Estimated firewood surveyed ger districts, including semi-coke coal, consumption for low-emission stove users is compressed coal, saw dust, sawdust briquette, 4.8m3, compared to 5.2m3 for traditional stove animal dung, and “anything that burns.” Semi- users. The difference is equal to about a half coke coal is used mainly by households living cubic meter of firewood for the entire winter, or in Bayangol District, where raw coal has been 17,680 Tg in savings on firewood expenditures banned, though usage rates for processed coal for the entire heating season. As with coal are still relatively low with only about half of savings, the estimated firewood savings are households in Bayangol reportedly using semi- a conservative estimate because they do not coke coal on a regular basis. account for the fact that a significant number of In real terms, total household spending on low-emission stove users acquired their stoves heating fuels increased about 57 percent from mid-way through the winter.40 This finding 406,000 Tg in the 2007/08 heating season to suggests that low-emission stoves save both coal about 637,000 Tg in 2011/12. Fuel consumption and firewood. was only slightly higher overall, so most of the increase was due to higher fuel prices. Also, 5.5 Conclusion because average monthly income reported The survey carried out in 2012 confirmed that coal by households increased by 80 percent, is still used by just about every ger-area household spending on heating fuels as a share of total as its main heating fuel and that firewood is monthly income actually declined by 5 percent. used to start stoves and ignite coal. Nalaikh coal While the financial burden placed on ger-area households by heating fuel expenditures has 40 Unlike traditional heating stoves, low-emission stoves re- eased somewhat, it continues to be very high quire the fire to be started as few times as possible; fire when compared to households in the city center should continue to burn at all times at low heating levels with access to district heating, especially for the to keep the room warm. Since firewood is usually used to start fire, firewood usage for low-emission stoves should poorest households. be lower than for traditional stoves. 54 CHAPTER 5. Heating Fuel Consumption and Expenditures Households in the bottom income quintile slightly higher in 2011/12 than it was in 2007/08, spent 31 percent of their total monthly income when households burned an average of 4.2 tons on heating fuels during the winter months of coal for the winter. Average monthly coal of 2011/12 (compared with 42 percent for consumption per household appears to have 2007/08). By comparison, households in the increased in most income categories as well as top income quintile only spent 6 percent of their for all types of dwellings and heating systems. income each month on heating fuels during the With respect to semi-coke coal, the survey winter. reveals that about 8,400 households use semi- As the main heating fuel, raw coal accounts for coke coal, with 5,750 of those households living the largest share of household heating fuel in Bayangol District. The findings also indicated expenditures. The survey shows that ger-area that in Bayangol District, where households are households used about 4.5 tons of raw coal not allowed to use raw coal and are supposed and spent about 424,000 Tg during the heating to use semi-coke coal, compliance with this season from September 2011 to May 2012. It regulation is low. Only 2,790 of the about 12,000 is estimated that households in the surveyed households (or 49 percent) in the ger areas in areas used a total of 735,750 tons of raw coal this district report that their households use during the 2011/12 heating season. In addition, semi-coke coal on a regular basis. The remaining average coal consumption per household was households only use it occasionally or rarely. 55 CHAPTER 6. HEATING STOVE DEMAND AND SUPPLY IN FOUR SELECTED CITIES 6.1 Study Background and Objectives introduced in the first section, followed by a brief overview of the four cities. The chapter then As part of the stocktaking activities in Mongolia examines the demand for stoves in the ger areas under the Clean Stoves Initiative, an assessment was of these cities, as well as the local supply chain. made of the ger-area market for household heating It describes the linkages of the aimag centers to stoves outside Ulaanbaatar. Given Mongolia’s large Ulaanbaatar and identifies the main factors that land size and low population density, a survey of determine the size of the local market and share the same depth and breadth as the one done in of locally-produced stoves. Finally, the chapter Ulaanbaatar was not feasible. Instead, it was decided also presents household perceptions of pollution that a rapid assessment of four selected cities would and knowledge of low-emission stoves, as well be conducted to gather qualitative information about as an estimate of the total technical potential the characteristics and structure of the market for demand for low-emission stoves. stoves. Upon the recommendation of the Ministry of Energy, which, along with the City of Ulaanbaatar, is the counterpart for this study, four cities were 6.2 Study Methodology and Data identified: Darkhan, Ondorkhaan, Bayankhongor, Rapid appraisal methods were used to quickly and Khovd. The four cities were selected based on assess the characteristics and structure of hypotheses about the determinants of market size the market for stoves outside Ulaanbaatar. and structure, including proximity to Ulaanbaatar Information on the market for stoves in Darkhan, and population size. Ondorkhaan, Bayankhongor, and Khovd was The objectives of the assessment were to: (a) gain collected by: (i) visiting the local market and better understanding on the characteristics of shops where stoves are sold and interviewing the supply chain and demand for stoves outside shop owners; (ii) visiting local stove production Ulaanbaatar; (b) assess similarities and differences facilities and interviewing stove producers; (iii) of the local markets with the Ulaanbaatar market; visiting households living in the ger areas of and (c) assess linkages with the Ulaanbaatar each city and interviewing households members; market. In addition, the assessment gauged and (iv) and interviewing public officials in the household perceptions of local air pollution and aimag centers responsible for infrastructure knowledge and interest in acquiring low-emission and planning as well as representatives from the stoves. An estimate of total technical potential aimag heating utility companies. demand for low-emission stoves in the four cities was also made. The assessment was conducted 6.3 Overview of the Four Aimag Centers with a view toward informing deliberations to Studied expand stove replacement activities underway in Ulaanbaatar to other cities in Mongolia. The four cities included in the rapid assessment of stove markets are Darkhan, Ondorkhaan, This chapter summarizes the main findings of Bayankhongor, and Khovd (see figure 6.1). the assessment. Data collection methods are 57 Stocktaking Report of the Mongolia Clean Stoves Initiative Figure 6.1. Location of the Four Cities included in the Rapid Assessment of Stove Markets Darkhan. Darkhan City is located 220 km north km to the southwest of Ulaanbaatar. The city’s of Ulaanbaatar. A total of 20,047 households population has seen rapid growth in past years, reside in Darkhan,41 about 10,000 of which are rising from around 20,000 in 2005 to more than connected to district heating in the central parts 33,000 today. Data from the national population of the city. The remaining households live in the and housing census show that there were 8,389 ger areas and rely on stoves. Of the households households living in the city in 2010. About 7,300 in the ger areas, about 3,300 live in gers and households live in the ger areas. District heating about 6,600 live in detached houses. only provides heating services to about 1,000 households, all of which are living in apartment Ondorkhaan. Ondorkhaan City is the capital of buildings in the city center. Khentii Aimag, and is located about 330 km to the east of Ulaanbaatar. Smaller than Darkhan, Khovd. Khovd city is the capital city of Khovd only 5,502 households were recorded living in Aimag, located far away from Ulaanbaatar at a the city as of 2010. About 600 households live distance of 1,425 km to the west. According to in the city center and are serviced by district the 2010 Census, 7,005 households are living heating, running water, and sewerage. The in the city. A total of 1,420 households in the remaining households live in the ger areas, city center are supplied with district heating requiring stoves for heating. These households from the central heating plant.42 The other 5,585 include about 3,200 households in gers and households live in the ger areas, including about about 1,700 in detached homes. Preparations 2,850 households in gers and about 2,735 in are underway to construct a large centralized detached homes. heating plant on the northeast edge of the city. Based on the number of ger-area households The aimag is also building apartments for 540 in each city, the technical potential demand for households in the city. household heating stoves in the four cities is Bayankhongor. Bayankhongor City is the estimated in table 6.1 below. capital of Bayankhongor Aimag, located 630 42 Because the first plant is already operating at capacity, construction of a second centralized heating plant, locat- 41 National Statistical Office of Mongolia, Population and ed on the southeast edge of the city, was started in early Housing Census (2010). 2012. 58 CHAPTER 6. Heating Stove Demand and Supply in Four Selected Cities Table 6.1. Estimated Technical Potential of Demand for Improved Heating Stoves in Four Select Cities Dwelling Types Number of Technical Total HHs with Potential Detached Number Access to Demand Other of HHs District for Heating Gers Houses & dwellings Heating Stoves Apartments Darkhan City 3,328 16,399 320 20,047 10,000 10,047 Ondorkhaan City 3,174 2,298 30 5,502 600 4,100* Bayankhongor City 6,275 2,923 23 8,389 1,000 7,389 Khovd City 2,849 4,064 92 7,005 1,420 5,585 Total Demand 15,626 27,121 Source: National Statistical Office of Mongolia, Population and Housing Census (2010). * Note: This estimated technical potential demand of 4,100 is taking into account an additional 540 household that will be moving into new apartments being developed by the city government and private investors. 6.4 Characteristics of the Demand for Stoves market. In Khovd, the number of households in the Aimag Centers living in gers is about the same as the number of households that live in detached houses. Although on a vastly smaller scale, the basic structure of demand for heating stoves in the In each of the four cities, the overwhelming four cities is similar to that of Ulaanbaatar. majority of households interviewed purchased Consumers in the ger areas can be grouped into their stoves in the central market or from four market segments based on dwelling type artisanal stove-makers in the aimag center. and heating system: ger households; households Few traveled outside the aimag to buy directly living in detached homes that use space-heating from Ulaanbaatar. The asking retail prices for stoves attached to heat walls; households in traditional ger stoves sold by retailers in the smaller detached homes without heat walls that local markets of the four cities ranged from just use space-heating stoves; and households just under 100,000 Tg to about 150,000 Tg living in larger detached homes with LPBs and depending on size, the thickness of the metal radiator systems. material, customization (i.e., whether the stove was fit with a heating wall attachment), The numbers of consumers in each of the four and workmanship. These asking prices are a market segments varies from city to city. The rough indicator of the affordable price range for size of the ger area and the composition of consumers in the aimag centers. dwelling types and associated heating systems in each city are good determinants for the size 6.4.1 Demand for Low-Emission Stoves and of each market segment. In Ondorkhaan and Bayankhongor, a higher portion of households Potential Willingness to Pay live in gers and smaller detached homes. From Almost all of the households interviewed in the census data and observation, it is estimated four cities were well-informed about the low- that typical space-heating ger stoves account emission stove exchange program that has been for more than half of all stoves bought by implemented in Ulaanbaatar. Because they are consumers in these two cities. By contrast, in still in the same media market as Ulaanbaatar, Darkhan, a much larger portion of households they have been exposed to the same information live in detached homes and heating stoves on television and radio as consumers in retrofitted for use with heat walls dominate the 59 Stocktaking Report of the Mongolia Clean Stoves Initiative Ulaanbaatar. Of the households that had heard market for LPBs will grow. about the stove exchange program, many As shown in chapter 5, households living expressed interest in acquiring the new stoves. in detached houses with LPBs consume The chairwoman of one ger-area bag (district) significantly more coal and firewood than in Darkhan, for example, submitted a request to households living in other dwellings or with the aimag governor on behalf of households in other types of heating systems. Higher coal her district for 1,000 low-emission stoves. (There consumption also means that LPBs emit more are roughly 1,600 households living in her bag.) pollutants into the air, unless these LPB stoves Similar requests were made in Ondorkhaan and are replaced with low-emission models. Any low- Bayankhongor. emission stove program that is expanded into The study team did not have the opportunity to the aimag centers cannot ignore the growing systematically collect detailed household income market for LPBs. data needed to determine the affordability of low- A successfully designed intervention to replace emission stoves in the aimag centers. Despite this, high-emission LPBs in the aimag centers will it is expected that households would be willing also have to address problems with low levels of to pay at least the same price as in Ulaanbaatar, knowledge among consumers. Interviews with given the strong interest demonstrated by LPB suppliers and households that expressed households and the fact that the asking prices an interest in upgrading their heating systems for traditional stoves sold on the local market revealed that potential consumers often do were higher than the subsidized prices for low- not know that LPBs use more coal than space- emission stoves sold under the MCA-Mongolia heating ger stoves. They do not realize that they program. In the four cities, the survey team also must keep their LPB stoves burning throughout spotted a few second-hand low-emission stoves the winter to prevent water in the pipes and that were purchased in Ulaanbaatar under the radiators from freezing, and that this may stove exchange program and either re-sold or increase the number of times they have to add gifted, including a stove being sold in one local fuel to their stoves each day. Furthermore, many shop in Ondorkhaan for 350,000 Tg. households invest in LPBs without improving the thermal insulation of their homes. 6.4.2 Demand for Low Pressure Boilers (LPBs) 6.5 Characteristics and Limitations of Stove Following the same trend seen in ger areas of Supply Chains Ulaanbaatar, more households in the aimag Local stove producers in the aimag centers centers are building new homes (or expanding consist mainly of two types: small workshops their existing homes) and replacing their and individual welders. In addition to stoves, existing stoves with LPBs. At present, the the workshops visited in Darkhan, Ondorkhaan, growing demand for LPBs remains limited and Bayankhongor, all produce other metal to a small segment of high-end consumers. items, such as chimneys, pails, and ash boxes. From interviews with LPB owners, retailers, A few even make wooden ger furniture. Stove and installers, it was learned that the price of production is often seasonal. During the coldest a complete LPB system (including the stove, months of the winter, production might stop radiators, piping, and labor) starts around 1.5 entirely, or the shops might retain one or million Tg. This puts the cost of an LPB system two employees to make accessories such as beyond reach for the majority of households chimneys. Only one of the workshops visited, in living in the ger areas of these cities; however, Bayankhongor, produces stoves throughout the given the recent economic prosperity, it is year. Each of the workshops operates a retail expected that more and more households will storefront (usually no more than a shipping be able to afford LPBs and that the size of the 60 CHAPTER 6. Heating Stove Demand and Supply in Four Selected Cities container) in the main market of the city year. Peak sales season usually runs from the where the workshop sells its stoves alongside late summer to early winter. The customer base stove accessories and other ger furnishings. for these workshops consists of local residents The largest of the stove producers visited, in in the aimag center, households from the Darkhan, claim to make and sell 500-650 stoves surrounding soum centers, herders from nearby per year. Others, in Bayankhongor and Khovd, areas, and even people from neighboring aimag. report production and sales of 200-300 stoves per Figure 6.2. Stoves for Sale in Central Markets of Darkhan and Bayankhongor (Left) Local producer’s stall in Birj Market, Darkhan selling ger stoves and stoves with heat wall attachments; (Middle and right) Stoves on display at local producer’s storefront in Bayankhongor central market. In contrast with the established workshops, business. The customer base for welders appears individual welders in the aimag centers usually to be comprised mainly of herder households only produce stoves to fill orders placed directly from the surrounding areas, at least according by end users. The welders who were interviewed to the welders interviewed. Because the artisans in the four cities explained that once they receive typically do not operate a storefront, customers orders for stoves, the next step for them is then must go directly to the welder’s production yard to locate materials, including scrap materials to place orders. In the same way, the welders also such as old water towers or oil tanks from the produce some made-to-order specialty stoves, Soviet era, or metal sheeting obtained from other such as LPBs, for end users or middlemen (i.e., producers. The supply chain is much less regular, plumbing contractors) that are not made by the and only those welders with connections who local workshops. Individual welders interviewed are able to secure a somewhat reliable supply in the four cities claimed to produce anywhere of materials are able to stay in the stove-making from 50 to more than 100 stoves per year. Figure 6.3. Artisan Welder in Bayankhongor A Bayankhongor welder in his production yard, making stoves. Scrap materials used to make stoves are shown in the center photo. At right is an LPB made by the welder, which he is selling for 350,000 Tg. 61 Stocktaking Report of the Mongolia Clean Stoves Initiative Based on interviews and observation, it can Furthermore, most of the stove retailers in be concluded that stove producers in the four the main markets of the four cities also sell selected cities can all be classified as small a variety of other products, including ger artisanal producers serving only the local frames, felt covers for gers, furniture, electric market. It is possible that the small workshops appliances, and other household goods. Stoves and individual welders in these cities could make might account for a smaller fraction of the low-emission stove models, if provided with retailer’s total business compared to that of local trainings and prototypes to copy. Indeed, at least producers selling in the main market. Evidence one artisan reported having attended trainings from the four cities suggests that despite the for making improved stoves in the past.43 Still, relatively small size of local markets in the aimag artisan producers remain limited by the supply centers, distribution channels are well-organized of materials and the size of the aimag center and established. markets. Local production of low-emission stoves The supply chain for LPBs in the aimag centers might not achieve the economies of scale needed also appears to be well-established, although to be profitable. Ensuring consistent adherence it is structured differently. Unlike traditional with emissions standards would also be difficult. stoves, LPB systems are typically installed by Apart from producers, wholesalers and plumbers or other contractors. The stoves may retailers form the remaining links in the stove be welded by the contractor, or made-to-order by supply chain for stoves in the four cities. As a local artisan. In two of the cities visited, Khovd noted above, local stove producers in larger and Bayankhongor, homeowners imported cities, such as Darkhan, and cities that are prefabricated LPBs directly from China. The located further away from Ulaanbaatar, such mass-produced, Chinese-made LPBs observed as Bayankhongor and Khovd, also own and in households were comparable in price to operate retail shops in the cities’ main markets. locally-welded stoves (around 400,000 to 500,000 In the main markets, local producers compete Tg), but of vastly better quality. The remaining with other well-established retailers that sell components, such as radiators and pipes, are stoves from Ulaanbaatar. In general, there are primarily bought and sold in construction supply two types of retail shops observed in the central stores situated around the central markets in markets that sell stoves from Ulaanbaatar. the aimag centers. According to LPB owners The first type of retailer buys stoves in bulk and retailers interviewed, homeowners often from the Narantuul market in Ulaanbaatar and accompany the plumber or contractor to re-sells them. The retailer is responsible for the stores to buy the different components. arranging transportation to ship the stoves from The implication of this finding is that, as in Ulaanbaatar to the aimag center. The second Ulaanbaatar, any intervention to replace high- type of retailer buys stoves directly from a emission LPBs in the aimag centers will need to producer in Ulaanbaatar. Often, the retailer has use different supply chains than existing stove a pre-existing relationship with the stove maker. replacement efforts, and may need to include The retailer might pay the producer upfront for plumbers and contractors. the entire shipment, or pay in installments as the stoves are sold. 43 An improved stove refers to an improved version of a tra- ditional stove and is not considered a low-emission stove because it does not meet all of the criteria for indoor and outdoor air pollution, safety, and combustion and fuel ef- ficiency of a low-emission stove (see box 1.2). 62 CHAPTER 6. Heating Stove Demand and Supply in Four Selected Cities Figure 6.4. Low Pressure Boilers in the Aimag Centers (Left) Chinese-made low-pressure boiler in Khovd household; (Middle) Locally-made LPB in Darkhan household; (Right) construction supply store in Khovd that sells radiators and piping for LPB systems. 6.6 Linkages of Stove Markets in with factors (ii) and (iii) being interrelated. Ulaanbaatar and the Aimag Centers Households in Ondorkhaan, for example, are dependent primarily on stoves supplied from On the basis of the rapid appraisal results, Ulaanbaatar because Ondorkhaan is within a along with findings from previous studies for few hundred kilometers of the capital, has a the Ulaanbaatar Clean Air Project (UBCAP),44 relatively small population, and no established it is evident that clear linkages exist between manufacturing capacity other than individual the stove market in Ulaanbaatar and markets artisans. Because Darkhan, which is even outside this city. The safeguard report for closer to Ulaanbaatar than Ondorkhaan, has UBCAP estimated that about half of the stoves a much larger population, it is thus able to made by stove producers in Ulaanbaatar were support several larger-scale stove producers as sold to vendors in the Ulaanbaatar market well as individual stove-makers. Consequently, (mainly Narantuul Market), while the other half local producers in Darkhan appear to be were sold to households outside Ulaanbaatar. competing fairly evenly with Ulaanbaatar Interviews with local stove retailers, wholesalers, producers in the aimag center market. producers, and ger households in the four cities Meanwhile, in Bayankhongor and Khovd, which by the rapid appraisal team confirmed that are located further away from Ulaanbaatar, stoves sold in the local aimag center markets retailers sell both locally-made stoves and consist of stoves made in Ulaanbaatar as well as stoves from Ulaanbaatar. Locally-made stoves locally-made stoves. appear to dominate the market, especially in While it was beyond the scope of the rapid Bayankhongor, which has a couple of well- appraisal to estimate the exact share of locally- established manufacturers. made stoves and stoves made in Ulaanbaatar In addition, some local stove producers rely in the local aimag center markets, it was on materials and stove parts from Ulaanbaatar. hypothesized that market share depends largely For example, one workshop in Darkhan that on a combination of three factors: (i) distance produces about 500 to 650 stoves annually from Ulaanbaatar; (ii) size of the local market; sources its stove tops and covers from and (iii) scale of local manufacturing capacity, Ulaanbaatar. Another producer in Bayankhongor, which claims to produce 200 to 300 stoves per 44 Ulaanbaatar Services Improvement Project II Project year, purchases metal sheeting from Ulaanbaatar Implementation Unit, “Mongolia: Ulaanbaatar Clean Air Project Safeguards Report” (internal management re- to bend and weld into stove bodies. By contrast, port, September 29, 2011). 63 Stocktaking Report of the Mongolia Clean Stoves Initiative individual stove makers are much less dependent (perhaps due to repressed demand for traditional on materials from Ulaanbaatar, instead relying stoves from subsidized low-emission stoves), the mainly on scrap materials but also whatever price of traditional stoves outside Ulaanbaatar other materials they can source locally (e.g., by will likely increase, at least in the short run. In buying extra metal sheets that local workshops the longer run, local producers may pick up the ship from Ulaanbaatar). slack and push prices down-or they might decide to keep their prices higher because the stove Given the linkages between the aimag center market has become less competitive than before markets and stove suppliers in Ulaanbaatar, any and they face less pressure from manufacturers disruption of stove production in Ulaanbaatar in Ulaanbaatar. It is also possible in the longer will likely have a direct impact on the supply run that large producers and wholesalers outside of stoves in the aimag centers as well. The Ulaanbaatar may supply traditional stoves to severity of the impact will depend largely on consumers in the Ulaanbaatar market and the dependency of local households on stoves compete with imported low-emission stoves. supplied from Ulaanbaatar and the capacity For smaller cities outside Ulaanbaatar that are of local stove production to replace stoves highly dependent on traditional stoves from from Ulaanbaatar. Due to market linkages, it Ulaanbaatar, adjustment may be slow because is conceivable that if several stove producers large-scale local stove producers do not exist. in Ulaanbaatar decide to go out of business Figure 6.5. Market Structure and Supply Chains of Stoves outside Ulaanbaatar Sour rce of stoves: Market sha s from these are of stoves e two pends on: sources dep Locally mad 1. L nd de stoves, an S 2. Stoves m Ulaanbaatar from ce from Ula 1. Distanc aanbaatar f local marke 2. Size of et o local man 3. Scale of c nufacturing capacity m Size of local market ends on: depe 1. Size of po n the city or Aimag cen opulation in nter opulation in 2. Size of po n the Aimag and Soum nearby, and d 3. Use of the es/no) (in th e market (ye he city or Aimag) A t as a trading ter for cent o region the area or Source: Authors. The linkages between the aimag centers and declines. Given the linkages between Ulaanbaatar also have important implications Ulaanbaatar and the aimag centers, however, it for the long-term success of stove exchange can be expected that a steady supply of new and programs in reducing air pollution in used stoves-including both traditional stoves Ulaanbaatar. The market for subsidized low- and low-emission stoves-will continue to flow emission stoves is expected to be temporary in and out of Ulaanbaatar. The population in and could dry up if subsidies are withdrawn, Ulaanbaatar is far from static. With continued unless the retail price of the stoves significantly migration into the ger areas and the natural 64 CHAPTER 6. Heating Stove Demand and Supply in Four Selected Cities growth of population, the demand for stoves in aimag has also invested in new apartments for the Ulaanbaatar will continue strong for a long 270 households, with plans to build housing time to come. Some households in Ulaanbaatar for another 270 households. Bayankhongor is may sell their low-emission stoves and replace pursuing a similar strategy, with plans to build them with traditional stoves, especially as the a CHP that is expected to cost around 12 billion low-emission stoves reach retirement age. Tg. It is also building two new apartments These low-emission stoves may be re-sold to buildings to house more than 160 households. consumers outside the capital. The remaining Khovd has plans to build a second CHP and has traditional stove producers in Ulaanbaatar set aside land for new housing as well. may try to fill this demand for new stoves in In the immediate term, authorities in three Ulaanbaatar, or producers in other cities may of the four cities expressed interest in try to increase their production and enter the distributing low-emission stoves to ger-area market. Thus, the current market share of residents.46 The aimag governor and city low-emission stoves in Ulaanbaatar cannot be council in Bayankhongor City, for example, assumed to be permanent, and this creation have reportedly approved the purchase of 400 of a temporary market for low-emission stoves low-emission stoves for 100 million Tg. The can only provide a temporary solution to the aimag plans on selling the stoves to ger-area air pollution problem. Only a permanent and households at subsidized prices. Requests for sustained market for low-emission stoves will low-emission stoves were submitted to the ensure that problem of air pollution due to use of governor’s offices of Darkhan and Khentii coal-burning stoves for heating can be solved. Aimag. While the Khovd Aimag governor’s office does not currently have any plans to distribute 6.7 Perceptions of Air Pollution in the Aimag stoves, representatives expressed support for the Centers idea. Interviewees in all four cities reported that air pollution in their cities is getting worse, 6.8 Conclusion to the extent that some interviewees claimed The findings of the rapid appraisal in the four their city’s air pollution was “as bad as in cities show that demand for heating stoves in Ulaanbaatar.”45 Two main contributors to air cities outside Ulaanbaatar is structured similarly pollution identified by respondents are: (1) the to demand for stoves in Ulaanbaatar. Consumers proliferating number of heat-only boilers, and (2) in the aimag center markets can be grouped household stoves in the ger areas. into four market segments linked to dwelling Officials in the aimag government departments type: households living in gers, households responsible for infrastructure planning who were in detached homes with space-heating stoves interviewed for this study all viewed centralized attached to heat walls, households in detached heating and the provision of new housing to ger- homes with space-heating stoves but no heat area residents as the long-term solutions to their walls, and households in detached homes with cities’ air pollution problems. Khentii Aimag has LPBs. The relative size of each market segment issued a moratorium on the construction of new varies from city to city. Also, as in Ulaanbaatar, heat-only boilers in Ondorkhaan and hopes to the current market for LPBs is small but start construction of a large centralized heating growing quickly. plant (CHP) on the edge of the city next year. The plant is expected to cost 8 billion Tg. The 45 46 To the team’s knowledge, only Khovd is actively moni- The team was unable to interview representatives from toring air quality. It was unknown whether Bayankhon- the Darkhan Aimag governor’s office, though the team gor and Darkhan are also conducting monitoring. No did speak with the CEO of the local district heating com- monitoring is being done in Ondorkhaan. pany. 65 Stocktaking Report of the Mongolia Clean Stoves Initiative Ger-area households in the four cities are market for stoves in Ulaanbaatar. In each of highly knowledgeable of the stove replacement the four cities visited, stoves made by local program that has been carried out in producers must compete with stoves from Ulaanbaatar and expressed interest in acquiring Ulaanbaatar. Retailers in the aimag centers low-emission stoves. While it was beyond typically purchase stoves wholesale from the scope of this study to gauge households’ Narantuul Market and then re-sell them, or willingness to pay for the new stoves, from source directly from producers in Ulaanbaatar. interviews and observation of prices in the local While it was beyond the scope of the rapid markets, it is hypothesized that households appraisal to estimate the exact share of locally- in the aimag centers would be willing to pay at made stoves in the aimag center markets least the same price for low-emission stoves versus stoves brought in from Ulaanbaatar, it is as traditional stoves currently being sold in hypothesized that the market share for locally- the aimag centers. Because air pollution is made stoves depends mainly on three factors, reportedly growing worse in the aimag centers namely distance from Ulaanbaatar, the size visited, local authorities have also expressed of the local market, and-related to the second interest in subsidizing the sale of low-emission factor-the existence of local manufacturing stoves to households as a near-term measure capacity. The co-dependency of the Ulaanbaatar for mitigating pollution. However, total technical and aimag markets is further reinforced by the potential demand for low-emission stoves in the fact that larger-scale stove producers in the four cities is estimated to only be about 27,000 aimag centers rely on raw materials sourced stoves. from Ulaanbaatar. On the supply side of the equation, it was Given the connection of the aimag centers to concluded from the rapid appraisal that stove Ulaanbaatar, efforts to replace traditional stoves producers in the four cities are comprised in Ulaanbaatar may have a direct effect on entirely of artisan producers, ranging from households outside Ulaanbaatar. The elimination individual welders to small workshops that might of local production in Ulaanbaatar may drive up employ a handful of people. While individual prices for stoves in the aimag markets in the welders reportedly only make 50 to 100 stoves short run. Also, because the supply chains for per year, the larger workshops might make 500 stoves in the aimag centers and Ulaanbaatar to 650 stoves per year. Production is seasonal are interconnected, it is conceivable that if for most stove-makers. Most also produce other traditional stove producers in Ulaanbaatar items, such as metal pails, ash boxes, chimneys, go out of business and existing subsidies for or even ger furniture. While it is conceivable low-emission products are then lifted, stove that artisan producers could make low-emission producers in the aimag centers could potentially stoves if provided training and prototypes to fill the demand for cheaper traditional stoves in copy, ensuring consistent quality and adherence Ulaanbaatar. This implies that there will always to emission standards would be difficult. Given be traditional stoves available to compete with the small scale of local markets, production of low-emission stoves in Ulaanbaatar and the other low-emission stoves might not be profitable, cities unless a sustainable, nationwide market for unless the cost of making the new stoves can be low-emission heating technologies is developed. reduced and policies are introduced by the cities If a nationwide stove replacement effort is to phase out traditional stoves. pursued, existing distribution channels may be exploited to distribute low-emission stoves from The rapid appraisal also finds that the aimag larger producers or importers in Ulaanbaatar. center markets are intricately linked to the 66 CHAPTER 7. CONCLUSIONS AND IMPLICATIONS FOR STOVE REPLACEMENT EFFORTS The results from the survey among ger area will have been reduced by a large enough households, the rapid assessment of stove margin already for air quality to be improved markets outside Ulaanbaatar, and outcomes by this much, given the observed growth in the from a recent stove replacement program can number of households and stoves in the ger inform future efforts to replace traditional areas since 2008 and the projected market share heating stoves in Mongolia. for low-emission stoves. An on-going program to replace traditional, polluting stoves will be As shown in this report, the recent stove crucial to ensuring that emissions reductions replacement program reached a very large share are deepened. Maintaining these reductions into of the market in a little more than two years. future years will further require the development Nearly 69,000 low-emission stoves were sold at of a more robust and sustainable market for low- subsidized prices in Bayanzurkh, Chingeltei, emission stoves. Khan-Uul, Sukhbaatar, Songinokhairkhan and Bayangol Districts from June 2011 to June 2012. Future stove replacement efforts will need to By the end of November 2012, a total of 97,877 continue to carry out follow-on market research stoves had been sold, bringing adoption rates to monitor changes, and find reasons and for low-emission stoves by households in the ger solutions for lower adoption rates among some areas to around 55 percent.47 market segments, so that adoption rates can be improved across-the-board. The survey shows Despite this initial success, an even higher that penetration rates for low-emission stoves share of the market is required if the goal is to vary considerably among different segments of reduce ger-area stove emissions to the level the Ulaanbaatar market. While sales data show needed for meaningful improvements in air a 55 percent market share as of December 2012, quality. According to previous analysis done by survey data show adoption rates of specific the World Bank of the severity and sources of market segments, such as households living air pollution in Ulaanbaatar, reducing population- in detached homes with heat walls and low weighted exposure to very fine particulate matter pressure boilers, remain relatively lower than for (PM2.5) by about half from the emergency levels ger households. The data indicate there is still measured from 2008 to 2009 would require an large potential for increasing the market share of 80 percent reduction in ger stove emissions (see low-emission stoves. Indeed, the survey confirms figure 7.1).48 It is unlikely that stove emissions that about 70 percent of households with traditional stoves are still interested in replacing 47 As a share of the total number of households in the cen- tral ger-areas of the six districts as of the end of 2012 their existing stoves. 48 At this level, the annual average for population-weighted concen- trations of PM2.5 would be reduced to around 157μg/m3, assum- The emerging market for low pressure ing no change in emissions from other sources such as coal-fired boilers (LPBs) has not been covered by stove power plants and suspended soil and dust compared to 2008/09. These levels are still much higher than Mongolian standards replacement efforts in Ulaanbaatar and also (25μg/m3) and the most lenient targets set by the World Health offers an opportunity for intervention. Organization for developing countries (35 μg/m3). 67 Stocktaking Report of the Mongolia Clean Stoves Initiative Figure 7.1. Reductions in Population-Weighted Future programs will also need to account for Exposure to Air Pollution in Ulaanbaatar the connections between Ulaanbaatar and the Resulting from Lower Ger Stove Emissions aimag center markets. The rapid appraisal 300 of on-the-ground market conditions in four AnnualPM2.5 concentration(ʅg/m3) 260 250 cities, Darkhan, Ondorkhaan, Bayankhongor, 215 and Khovd, revealed that the supply chain for 200 180 stoves in the aimag centers is closely linked 150 135 to Ulaanbaatar. Well-established distribution 100 channels exist to move stoves and stove-making 50 materials between Ulaanbaatar and the aimag 0 centers. Intervening in just the Ulaanbaatar Baseline 30%reductionin 50%reductionin 80%reductionin (6/2008Ͳ5/2009) stoveemissions stoveemissions stoveemissions market may not be sufficient to totally eliminate supply of traditional stoves, especially once the Source: World Bank, Air Quality Analysis of Ulaanbaatar: existing subsidies for low-emission products Improving Air Quality to Reduce Health Impacts, December are lifted. It will be difficult to stop the flow of 2011 some low-emission stoves out of the capital into the aimag centers, where households have Survey data reveal that more households are demonstrated strong interest in acquiring the building bigger homes (or expanding their new stoves. Even if local producers stop making existing homes) and upgrading their heating traditional stoves in Ulaanbaatar, it is conceivable systems to LPBs. Since 2007, the proportion of that traditional stoves could be supplied to detached homes with LPBs has increased from Ulaanbaatar from other markets as households 16 percent to about 23 percent. The average look for replacement stoves. Either way, demand size of homes with LPBs, meanwhile, has grown for stoves in Ulaanbaatar and the aimag centers from 65 m2 to around 74 m2. Households with will likely persist. LPBs burn significantly more coal for heating each winter than other households. LPB users Recognizing the links between Ulaanbaatar and reported using an average of 6.3 tons of raw coal the aimag centers provides an impetus for stove during the winter of 2011/12, compared to 4.7 replacement efforts to be expanded outside tons for households living in smaller detached the capital city. Two different strategies for a homes without LPBs. Higher coal use by LPB nationwide would be: (i) ensuring that stoves users means more pollution, unless low-emission made in the aimag centers comply with new models are introduced to replace existing LPBs. emissions standards by providing assistance At the same time, interventions targeted at to local producers to make low-emission replacing LPBs will need to follow a different stoves and increasing the size and budget of approach than existing stove replacement the inspectorate; or (ii) improving existing activities in Ulaanbaatar, because the supply distribution channels for low-emission stoves chain for LPB systems is structured much to be supplied to the aimag centers from larger differently than that for traditional ger stoves and producers or importers in Ulaanbaatar. Because the technology is more complicated. Detailed the markets outside Ulaanbaatar are relatively technical requirements for installation, use thin (the total technical potential demand for with existing piping and radiator systems, and the four cities visited in this study is only around proper maintenance will need to be introduced 27,000 stoves), expanding distribution channels to ensure the proper functioning and safety of may be a more viable strategy than building local low-emission LPBs. Emissions performance of production capacity. LPBs should be verified with published results The road map first presented in the 2009 of laboratory testing following an appropriate World Bank report, describing a market-based testing protocol. approach for introducing new technologies 68 CHAPTER 7. Conclusions and Implications for Stove Replacement Efforts that can burn fuel (new or old) more cleanly in Standards and certification mechanisms Ulaanbaatar, remains largely relevant today need to be developed after sufficient practical and offers a framework for understanding good experience is obtained with testing and product practices and assessing options to address performance in the field. A rush to legal sustainability challenges of the current program. standards may inhibit innovation, or worse, set The seven steps of this road map-with new ineffective or unrealistic parameters. Households information based on recent findings-are: can deviate from stove manufacturer’s instructions, which affect stove emissions • Estimate the impact of the proposed mitigation performance. These deviations can happen measures on the overall air pollution to especially if the stove differs significantly from ensure the government’s investment will meet traditional products. A more pragmatic approach expectations. The World Bank’s Mongolia: Air to standards development would divide tasks Quality Analysis of Ulaanbaatar: Improving into building blocks: (i) using stove replacement Air Quality to Reduce Health Impacts activities to introduce interim standards, (December 2011) estimates that a 30% to or eligibility criteria, as a starting point for 80% emissions reduction from ger area developing standards; (ii) establishing sufficient, heating can produce a 18% to 48% reduction and preferably local, laboratory capacity to set in population weighted exposure to PM2.5. initial emissions requirements and measure Under the Ulaanbaatar Clean Air Project compliance during program implementation (UBCAP), the city will start to implement a (perhaps, too, establishing links with labs program that, among other requirements, outside Mongolia may help build capacity); (iii) that proposes to set a maximum of 70 building awareness and understanding of the micrograms of PM2.5 per net megajoule (MJ) impacts of various parameters that could be of heat emitted for ger stoves, for access to included in the eventual standards and emission subsidies. This is about a 90% reduction requirements by publishing testing reports and in emissions compared to traditional explaining results to stove producers, taking stoves (1,000 μg PM2.5/net MJ). Technical into account variations in performance due to assistance for air quality analysis in possible “misuse” of stoves; and (iv) adjusting Ulaanbaatar currently provided by the Japan protocols and testing interpretations based on International Cooperation Agency (JICA) feedback from household use (using surveys, and an evaluation of the MCC-funded stove- observations, repeated lab tests, and field tests switching project being conducted by MCC of emissions). Monitoring data will provide and MCA-Mongolia may provide an updated confidence that the eventual standard is realistic, forecast of impacts of mitigation measures. achieves its objective, and is enforceable in a • Develop a testing program for equipment, cost effective manner. However, while field with a focus on safety and emissions rather testing is important, it is very difficult to obtain than fuel consumption alone, and set interim reliable results; field testing results should be performance targets that can eventually be interpreted by experienced analysts. Also, all accepted, after practical experience is gained, work towards the development of standards, as new standards, following the Mongolian including stove testing and monitoring, regulatory process. Mongolia has three main should involve the Mongolian Agency for standards (MNS 5216:2002, MNS 5041:2001, Standardization and Metrology, who will be MNS 5679:2006) with some updates for ultimately responsible for developing and stoves, boilers and furnaces, and solid fuels- enforcing the eventual standard. Finally, inputs all of which could be improved. Areas for from private stove developers, households, and improvement are summarized in the World other stakeholders should also be collected and Bank’s Mongolia: Heating in Poor, Peri-urban considered during the standards development Ger Areas of Ulaanbaatar (October 2009). process. 69 Stocktaking Report of the Mongolia Clean Stoves Initiative In 2011, an Asian Development Bank-financed emission stoves. Under UBCAP, technical technical assistance project implemented assistance is available through a new Stove by the Ministry of Mineral Resources and Development Center, which could be a Energy established a Stove Emissions and source for training. Partnerships with the Efficiency Testing (SEET) Laboratory, which imported stove producers are encouraged conducted over eighty tests of various stove so that appropriate knowledge is with local models, including some of the Turkish and service providers after warranties expire. The Chinese-made models that were subsidized by center can also provide short term technical MCA-Mongolia and CAF. Based on this work, support for producers encountering design Ulaanbaatar City developed eligibility criteria or manufacturing problems. for stoves and producers. These criteria already • Establish a targeted and well-justified subsidy address some of the deficiencies in the current program to make new systems affordable. set of standards, and require that stoves and Subsidy arrangements should consider: fuels be tested together for emissions. Although (i) the end point policy objective of the the SEET lab did not receive support after the government, which in this case is reducing ADB program was completed and is temporarily air pollution (especially particulate matter) dismantled as of the writing of this study, the lab exposure; and (ii) the consumers’ willingness will continue to receive support from the City to switch to new products. Although a well- of Ulaanbaatar through UBCAP, supported by designed stove replacement program usually the International Development Association, the brings large net benefits in reduced health concessional assistance arm of the World Bank. risks and fuel savings, the right amount and MCA-Mongolia conducted field tests but the types of subsidy will depend on the ability of results are not publicly available. Once available, the market to sustain the transformation to these testing results will be valuable inputs to the use of cleaner stoves. Based in part on understanding variations in emissions between the available data presented in this report, lab and home use. key principles of an appropriate subsidy • Set an enforcement mechanism to address program include: (i) the ability of consumers non-compliant products and manufacturers. to choose among several stove models, as Stove switching eligibility criteria for not only the stove price (which was in fact new stoves and vendors should include ranked lower than other positive features warranty requirements and minimum after- of low-emission stoves; see chapter 4), but sales service standards. (Warranties of also by fuel savings and other features; (ii) two years were provided by low-emission the payment of subsidies only upon verified stove vendors that received subsidies from installation of the new equipment; (iii) the MCA-Mongolia and CAF.) During the possibility of rewards for the use of stoves warranty periods, participating producers that achieve even larger emission reductions should agree to random spot checks of (i.e. considering that the government is stoves sold (i.e. random lab testing) and willing to subsidize products to achieve monitoring of their service performance. public objectives, consumers should be In addition, they must agree to a penalty rewarded for using stoves that achieve system for non-compliance, including proportionally better public outcomes); permanent exclusion from the program (iv) the use of subsidy levels that reflect in cases of persistent violations. Further, the affordability of target households and as local sellers of imported stoves might their willingness to start using another not be available, a likely need will exist type of stove before the end of their current for skilled workmen and the availability of stove’s useful life; and (v) the use of a final good quality spare parts to service the low- consumer’s price that is not too low (or free) 70 CHAPTER 7. Conclusions and Implications for Stove Replacement Efforts to ensure stoves are sold to people who will stoves, creating an incentive for reselling. use and maintain it. Evidence of large-scale reselling does not exist, however, perhaps because people who During the winter of 2011-2012, the stove might buy re-sold stoves expect the subsidies replacement program included subsidies of to continue. It is extremely difficult to measure around 249,800 Tg to 408,450 Tg (equivalent the occurrence of leakage. The study team did, to US$195-319) for low-emission stove models. however, identify a few low-emission stoves Subsidy amounts included two separate in the aimag and soum centers, outside the elements, financed by MCC/MCA-Mongolia program boundaries of Ulaanbaatar. To further and later by CAF. The subsidies reduced the mitigate this risk of low-emission stoves not sales price of low-emission stoves to roughly being used by the purchasing households in 20-25 percent of the price for traditional ger Ulaanbaatar, possible mechanism include: (i) stoves, not including accessories. The prices of verifying installation and training consumers in larger low-emission stoves ranged from 30-40 using the stoves before releasing the subsidy percent of traditional stoves used with fittings for (already applied in the current program); (ii) heating walls. The pricing strategy employed by using site visits to customers during the winter the recent stove replacement program achieved (either by producers or by program staff) to its goal of high and rapid sales rates; however, verify stoves are in use, provide supplementary sustaining such levels of demand without the training, and check on customer satisfaction; high subsidy elements is unlikely, leaving and, (iii) establishing a transparent reporting sustainability challenges for the next phase of protocol linked to regular monitoring, providing the program. information to stakeholders about program • Establish credible supply chains, in partnership developments and possible adjustments needed. with the private sector, and use a “third party” • Remove old equipment from use. To achieve verification system for installation and use sufficiently large and sustained emission of new stoves. The recently implemented reductions, the support for cleaner stove replacement program used a clear alternatives must happen in parallel with and credible supply chain, involving product the removal of polluting heating systems. centers in designated khoroo (see chapter The July 2012 survey reveals an estimated 1) to control quality and prevent the sale of 212,708 stoves are owned by 164,127 “copies” - similar but lower quality versions. households, of which only 172,055 stoves XacBank, the main distributor of MCA- are in use. The recent stove replacement Mongolia and CAF-subsidized stoves, is program required submission of old stoves also implementing a Clean Development as part of the sales transaction, but program Mechanism (CDM) project for stove participants noted that about 7 percent of switching. The CDM requires strict households received exemptions, and that monitoring of stove installation and use, the stoves that households turned in tended following an agreed protocol. It also requires to be in worse condition as the sales season independent validation to verify emissions progressed. Stove replacements need to be reductions. carefully monitored because households Although sales are an important milestone, may be tempted to provide unused stoves (or the final development outcome is achieved even poor-condition stoves purchased from only by the sustained use of the low-emission other households) to meet the requirement. stoves among area households. Although While no fail-safe method exists to ensure households sign an agreement promising to honest trade-ins, mitigation measures include use the subsidized stove, low-emission stoves the independent verification mechanism are priced below market prices of traditional discussed above, as well as continued public 71 Stocktaking Report of the Mongolia Clean Stoves Initiative awareness campaigns on the health impacts • Continue the program, including maintaining of the pollution caused by the older stoves. credible supply chains, verification procedures, and consumer subsidies. A publicly announced • Market-based approaches to disseminate timetable to scale down the subsidy could cleaner heating systems are preferred over encourage households on the fence to nonmarket-based approaches such as stove participate; however, curtailing the program giveaways. Broad consensus seems to exist too quickly could jeopardize results. that artisanal stove manufacturing is not going to have sufficient capacity to design • Publish eligibility criteria for stoves and and supply the market with low-emission producers. The current program did not stoves quickly enough and at a large enough publish eligibility criteria. Consequently, the scale to bring down unit costs to parity private sector in Mongolia was only able to with traditional stoves. The recent stove participate by selling imported stoves (the replacement program collaborated with only products that were identified as being foreign producers and Mongolian private eligible for subsidies). Publishing criteria sector distributors to achieve very rapid and undertaking a transparent and objective penetration rates. The partnership with the evaluation and enforcement processes will be private sector is critical to the success of critical to encouraging greater private sector any market-based program as it provides the participation. professional know-how, customer orientation, • Encourage development of technical and innovation needed to provide good standards for a wider range of products, customer services and respond to market including heating walls, low pressure boilers, demand. There have been reports that major and combinations with new fuels. This private sector participants are no longer can be done through support of the SEET interested in participating, which presents Laboratory. a major risk to the program, especially as it faces challenges to find ways to support • Develop skills and spare parts supply chains a transformed, clean stove market without for servicing current stoves in partnership subsidies. Concerns of the private sector with the current import suppliers. should be effectively addressed in the next phase of the program. Continue to monitor and evaluate program impacts through household surveys, air quality monitoring, Although the current program already includes and stakeholder consultations. On the latter, and has tested features of market based coordination among all stakeholders, donors approaches, gaps remain. The program can be and government agencies, is a critical factor to supported by the following policy options: ensuring focused and sustainable outcomes. 72 REFERENCES ADB. “43177-012: Ulaanbaatar Clean Air, Project Data Sheet.” http://www.adb.org/ projects/43177-012/details. Khurelsukh, N. “Clean Air Fund’s Plan and Budget Allocation for Air Pollution Reduction Activities.” Presentation delivered at Conference for Sustainability of Ulaanbaatar Air Pollution Reduction Activities, Ulaanbaatar, Mongolia, January 29, 2013. IMF. “Data and Statistics.” http://www.imf.org/external/data.htm#data. Myagmar, R. “Mongolia’s Air Pollution Priorities and Work Plan.” Presentation delivered at Conference for Sustainability of Ulaanbaatar Air Pollution Reduction Activities, Ulaanbaatar, Mongolia, January 29, 2013. National Statistical Office of Mongolia. 2010. Population and Housing Census. Ulaanbaatar: National Statistical Office of Mongolia. Ulaanbaatar Statistics Department, http://statis.ub.gov.mn/. Ulaanbaatar Services Improvement Project II Project Implementation Unit. “Mongolia: Ulaanbaatar Clean Air Project Safeguards Report” (management unit internal report, September 29, 2011). World Bank and Public-Private Infrastructure Advisory Facility. Mongolia: Foundation for Sustainable Development: Rethinking Infrastructure Services in Mongolia. June 2007. World Bank, Mongolia – Heating in Poor, Peri-urban Ger Areas of Ulaanbaatar. ASTAE (Asia Sustainable and Alternative Energy Program) Report, Washington, DC, 2009. World Bank and Cities Alliances. Mongolia- Managing Urban Expansion in Mongolia; Best Practices in Scenario-Based Urban Planning. Washington, DC: World Bank, 2010. World Bank. “Clean Stove Initiative in the East Asia and Pacific Region.” World Bank Concept Note, Washington, DC, 2011. World Bank. Mongolia - Air Quality Analysis of Ulaanbaatar; Improving Air Quality to Reduce Health Impacts Vol 1 and 2. World Bank Sustainable Development Series Discussion Paper, December 2011. World Bank. Rapid Assessment of Stove Markets outside Ulaanbaatar – Trip Report. Internal report for the project. Washington DC: World Bank, 2012. XacBank. “Market Mechanisms for MCA Energy and Environment Project.” Presentation delivered in Ulaanbaatar, Mongolia, March 15, 2010. 73 ANNEX A: SAMPLING PLAN Ulaanbaatar is divided into nine districts, or six districts of Ulaanbaatar. These districts are duureg. Six of the districts form the central urban Khan-Uul, Bayanzurkh, Bayangol, Sukhbaatar, area of the capital city, fanning out to the ger Chingeltei and Songinokhairkhan. Khoroo areas. Each district is divided into sub-districts, located outside Ulaanbaatar city were excluded or khoroo, of which there are currently 132 in the from the sample frame, including Khoroo 12 city. Each khoroo is the further divided in micro- and 13 in Khan-Uul, Khoroo 20 in Bayanzurkh, districts, or kheseg. Ger areas are located in all of and Khoroo 21 in Songinokhairkhan. Based on the nine districts, usually corresponding to lower the Ulaanbaatar Statistics Department, 164,127 levels of administrative boundary, the khoroo. households (in the highlighted columns) were living in the simple separate houses and gers The target population of the survey included all in the surveyed areas of Ulaanbaatar City as of households that live in the ger areas of the central December 2011 (see table A.1). Table A.1. Total Number of Households (HH) and Population of Selected Districts and Living in Ger Area at the End of 2011 Ger area/ Ger area/simple improved Ger area/ger Total separate houses Districts Total HHs separate houses Population Number Popula- Number Popula- Number Popula- of HHs tion of HHs tion of HHs tion Khan-Uul* 28 602 110 087 110 490 10 296 40 554 5 506 21 890 Bayanzurkh* 71 029 276 933 517 1 933 22 310 87 510 22 773 88 469 Bayangol 48 596 192 111 157 732 6 333 25 969 5 522 21 421 Sukhbaatar 36 343 137 834 187 795 11 810 43 900 8 042 32 905 Chingeltei 36 856 153 117 393 1 628 18 497 82 072 10 142 39 503 Songino- 60 989 251 097 413 1 590 21 852 91 137 21 044 89 774 khairkhan* Total 282 415 1 121 179 1 777 7 168 91 098 371 142 73 029 293 962 Source: Ulaanbaatar Statistics Department, http://statis.ub.gov.mn/ Note: * Excluding khoroo outside of Ulaanbaatar city. An area-based sampling frame was constructed Individual grid cells were then randomly using satellite imagery of the central ger selected from a list. The area-based method was areas of Ulaanbaatar. Khoroo in the frame chosen due to a lack of reliable street addresses were partitioned into smaller grids of 100m2, and household registries in parts of the ger each of which typically contained one to three areas. Surveys were completed for a total of households. Grid cells with more than one 1,000 households. household were assigned unique numbers. 74