UNDP Rural Energy 2008 Eng

Residential energy demand
in rural Cambodia:
An empirical study for Kampong Speu and Svay Rieng
Cambodia
March 2008
UNDP Cambodia Energy and Environment Programme
Rural energy preparatory assistance project
Background:
Cambodia is an agricultural country. More than 85 percent of Cambodians live in rural ar-
eas with agriculture as their primary livelihood, including farming, fishing, raising livestock,
and harvesting forest and non-forest products. Most poor rural families rely on firewood for
cooking. Increasing time is needed to collect scarce firewood and indoor smoke pollution is
a burden for rural women. The pressure on forest resource depletion continues. Less than 20
percent of rural households have access to grid-quality electricity services. Almost all people
in rural areas use firewood for cooking, kerosene for lighting, and lead-acid rechargeable bat-
teries for household appliances. Energy use is inefficient, inconvenient, costly and environ-
mentally unsustainable. Inefficient use of batteries and kerosene are the other main energy
sources for rural families.
Purpose:
The preparatory activities aim to establish a systematic data monitoring system for rural en-
ergy and to formulate a full size project to develop rural livelihoods through improved access
to sustainable energy services. Access to sustainable energy will be critical for Cambodia to
reach the Millennium Development Goals as well as the targets set in the National Strategic
Development Plan for the reduction of fuel wood dependency and poverty. To plan an ef-
fective intervention for rural energy development, systematic baseline data for monitoring
needs to be in place. A rural energy balance sheet will capture the demand and supply of rural
energy services and can be used as an effective tool to formulate a rural energy development
strategy at both national and provincial levels.
© 2008, UNDP Cambodia

Residential energy demand in rural Cambodia: An empirical study for Kampong Speu and Svay Rieng
DISCLAIMER
The responsibility for opinions in this publication rests solely with the authors. Publication does not constitute an endorsement by the
United Nations Development Programme or the institutions of the United Nations system.
Residential energy demand
in rural Cambodia
An empirical study
for Kampong Speu and Svay Rieng
Cambodia
TITLE
Residential energy demand in rural Cambodia
TABLE OF CONTENTS
Acknowledgments .............................................................................................................................................. iii
Acronyms and Abbreviations .......................................................................................................................... iv
Foreword ................................................................................................................................................................... 1
1 Introduction ............................................................................................................................................ 3
1.1 Background ............................................................................................................................. 3
1.2 Energy policy framework of the RGC ............................................................................. 3
1.3 UNDP’s country strategy and rationale for involvement in energy .................... 4
1.4 Objectives ................................................................................................................................ 5
1.5 Approach and methodologies ......................................................................................... 5
2 Survey results . ........................................................................................................................................ 9

2.1 The provinces of Kampong Speu and Svay Rieng ..................................................... 9
2.2 Social classes ........................................................................................................................ 10
2.3 Energy used for cooking .................................................................................................. 12
2.4 Energy for lighting and appliances .............................................................................. 14
2.5 Energy use by social class ................................................................................................ 16
2.6 Supply cost and affordability ......................................................................................... 21
2.7 Energy Transition ................................................................................................................ 22
3 Future demand in surveyed provinces . ..................................................................................... 25

3.1 Forecasting methodology . ............................................................................................. 25
3.2 Demand and supply modelling using historical data ........................................... 25
3.3 Refined modelling . ............................................................................................................ 29
3.4 Demand forecasts .............................................................................................................. 31
3.5 Extrapolation to national level ...................................................................................... 35
4 Lessons learned .................................................................................................................................. 38

4.1 Rural energy supply in Cambodia ................................................................................ 38
4.2 Next steps . ............................................................................................................................ 39
UNDP Cambodia i
TITLE
Acknowledgments
The study was jointly undertaken by the Ministry of Industry, Mines and Energy (MIME) and UNDP
Cambodia as part of the project development exercise “Preparatory Activities for a Rural Energy
Development Program”. Technical assistance was provided by an international non-governmen-
tal organisation (NGO) Groupe Energies Renouvelables, Environnement et Solidarités (Group for Re-
newable Energy, Environment and Solidarity − GERES)
The MIME team was directed by H.E. Tun Lean, General Director of Energy and included Mr. Heng
Kunleang and Mr. Cheap Sour. GERES’ Cambodia team comprised Mr. Rogier van Mansvelt (Tech-
nical Advisor) Mr. Jan Simon Brouwer (International Communication Advisor), Ms. Anna Ritzen
(Energy Expert) and a group of electrical engineering students led by Mr. Kong Pagnarith.
Photos used in the report were taken by GERES and UNDP Cambodia.
The final report was reviewed and edited by Gerhard Zieroth, International Energy Consultant.
UNDP Programme Analyst Kati Susanna Veijonen managed the project and oversaw the prepa-
ration of this report.
UNDP Cambodia iii

TITLE
ACRONYMS AND ABBREVIATIONS

ADB Asian Development Bank
BAU Business as Usual
CPAP UNDP Country Programme Action Plan
DIME Department of Industry, Mines and Energy
EC European Commission
EdC Electricité du Cambodge (Electricity of Cambodia)
EDFM Energy Demand Forecasting Model
GDP Gross Domestic Product
GERES Groupe Energies Renouvelables, Environnement et Solidarités (Group for Re-
newable Energy, Environment and Solidarity)
GJ Giga Joule = 1000 Mega Joules
JBIC Japanese Bank for International Cooperation
JICA Japan International Cooperation Agency
kWh Kilowatt Hours
LEAP Long Range Energy Alternative Planning
LPG Liquified Propane Gas
MDG Millennium Development Goals
MIME Ministry of Industry, Mines and Energy
NGO Non-Governmental Organization
NIS National Institute for Statistics
NSDP National Strategic Development Plan, 2006 - 2010
REE Rural Electricity Enterprise
REP PoR Regional Energy Programme for Poverty Reduction
RGC Royal Government of Cambodia
SME Small and Medium Scale Enterprises
TJ Tera Joule = 1000 Giga Joules
UNDAF United Nations Development Assistance Framework
UNDP United Nations Development Programme
WB World Bank
iv UNDP Cambodia
FOREWORD
For the past decade, there has been mounting pressure on Cambodia’s rural households, which
struggle to cope with the depletion of traditional biomass energy sources and increasing costs
for fuels such as kerosene and diesel. Although considerable progress has been made to ex-
pand electrification into rural areas, a large segment of our rural households remain without
electricity and rely on energy forms that are inefficient and pollute the domestic environment.
Lack of modern energy is closely linked to rural poverty and lack of opportunities for our popu-
lation in rural and remote areas.
With financial support from the United Nations Development Programme (UNDP), the Minis-
try of Industry, Mines and Energy (MIME), with technical assistance from the international non-
governmental organisation (NGO) Groupe Energies Renouvelables, Environnement et Solidarités
(Group for Renewable Energy, Environment and Solidarity − GERES), Cambodia launched a
project in 2006 entitled “Preparatory Activities for a Rural Energy Development Program”. The
project included a study on rural household energy demands. This report confirms with de-
tailed data that meeting their basic energy needs is a daily struggle for many families in rural
areas. Most families have neither electricity nor gas for lighting and cooking. Rural inhabitants
often spend hours each day gathering wood and other biomass. These hours might otherwise
be spent on productive rural business, agriculture or learning. Indoor pollution through open
fires is a serious health hazard. At the same time, forest resources are under ever increasing
pressure from firewood extraction and charcoal production.
Based on the results of this study, we now have to take up the challenge and use the insights
gained. Together with UNDP and other development partners, we have to move ahead and
simultaneously improve the policy framework for rural energy while developing and imple-
menting projects.
I would like to thank UNDP for its valuable support to the study and ensure that the Ministry
will fully support the development of the programme that UNDP is developing in cooperation
with other donors to improve rural households’ access to more modern, sustainable and ef-
ficient rural energy services at an affordable price to all, including the poor.
H.E. Tun Lean

Director General
Ministry of Industry, Mines and Energy
UNDP Cambodia 1
TITLE
1. INTRODUCTION Cambodia
has a low
electrification
1.1 Background rate and high
More than 85 percent of Cambodians live in rural areas with agriculture as their primary means electricity costs.
of livelihood. Rural families in Cambodia make a living from farming, fishing, livestock, and
harvesting forest and non-forest products. Since the end of internal conflicts in 1991, Cam-
bodia has made important progress in ensuring peace and security, rebuilding institutions
and infrastructure, establishing a stable macroeconomic environment, and putting in place a
liberal investment regime. In spite of these achievements, rural development remains a ma-
jor challenge. With a per capita Gross Domestic Product (GDP) of approximately US$571 in
2007, Cambodia remains one of the poorest countries in the world. Poverty is still widespread,
particularly in rural areas. In 2004 it was estimated that 34.7 percent of the population lived
below the poverty line, although that number is down from 47 percent ten years ago as a re-
sult of the RGC’s poverty reduction strategy. Inequality appears to be increasing and recent
economic expansion has been narrowly based. Economic growth has not led to a significant
reduction in rural poverty. Against this background, it has been widely acknowledged by the
Royal Government of Cambodia (RGC) and development partners that development progress
in Cambodia will depend on achievements in rural areas.
Rural development requires opportunities for income generation and the provision of infra-
structure such as roads, safe water supplies and affordable energy. While progress has been
made in recent years in the area of rural infrastructure, Cambodia has one of the lowest elec-
trification rates in Asia (less than 20 percent of households have access to grid electricity) and
has among the highest electricity costs in the region (US$0.35-0.90/kWh).
1.2 Energy policy framework of the RGC
Access to sustainable energy services is a critical element for Cambodia to reach the Millen-
nium Development Goals (MDGs) as well as the targets set in the National Strategic Develop-
ment Plan, 2006-2010 (NSDP) for the reduction of fuel wood dependency and poverty. Rural
energy development supports all four pillars of growth in the RGC’s Rectangular Strategy: ag-
ricultural productivity, promotion of small and medium scale enterprises (SMEs), energy in-
frastructure development and greater gender equity. In its Rectangular Strategy the RGC has
acknowledged the link between poverty alleviation and energy supply.
In 1999, the RGC approved the Cambodia Power Sector Strategy 1999-2016. The objectives of
this policy are:
1. to provide an adequate supply of energy throughout Cambodia at reasonable and afford-

able price;
2. to ensure a reliable, secure electricity supply at prices, which facilitate investment in Cam-
bodia and development of the national economy;
3. to encourage exploration and environmentally and socially acceptable development of
energy resources needed for supply to all sectors of the Cambodian economy;
4. and to encourage efficient use of energy and to minimize detrimental environmental ef-
fects resulting from energy supply and use.
Based on this Power Sector Strategy, the RGC approved in 2006 the Rural Electrification by Re-
newable Energy Policy as an integral part of the Government’s overall agenda for the Energy
UNDP Cambodia 3
TITLE
Sector. The main objective of this policy is to create a comprehensive enabling framework
for renewable energy technologies to increase access to electricity in rural areas. The policy
acknowledges the Master Plan Study on Rural Electrification by Renewable Energy in the King-
dom of Cambodia as the guiding document for the implementation of projects and programs.
The Master Plan envisions: 1) To achieve the 100 percent level of village electrification includ-
ing battery lighting by 2020; 2) To achieve the 70 percent level of household electrification
with grid quality electricity by 2030. While the policy framework for rural electrification is quite
comprehensive, there is a policy gap with respect to other types of energy that are essential for
the rural population. In other words, cooking energy that is supplied by biomass, or fuels such
as Liquified Petroleum Gas (LPG) is not yet covered by a specific policy.
Donors such as the World Bank (WB), Asia Development Bank (ADB), Japanese Bank for In-
ternational Cooperation (JBIC), Japan International Cooperation Agency (JICA) and, more re-
cently, UNDP have pledged to support the strategic energy sector goals of the RGC.
1.3 UNDP’s country strategy and rationale for involvement in energy

In recognition of the daunting problems of rural energy, UNDP Cambodia has included the
energy sector in the current Country Programme Action Plan (CPAP) that outlines the coopera-
tion between the two partners from 2006 until 2010. UNDP sees rural energy as a logical ex-
pansion of the organization’s traditional support for a variety of rural development initiatives
over the last 15 years. Rural energy also has multiple linkages to the approved United Nations
Development Assistance Framework 2006-2010 (UNDAF) which, in turn, is consistent with the
Government’s Rectangular Strategy and with its NSDP covering the same period. The UNDAF
identifies four major priorities where the UN can make a significant contribution:
1. good governance and human rights,
2. agriculture and rural poverty,
3. capacity building and human resource development, and
4. support for the design and implementation of the NSDP.
In recent years UNDP Cambodia has established itself as one of the key partners supporting the
RGC in its efforts to establish conditions for sustainable rural development, economic growth
and poverty alleviation. Given the close link between rural development and sustainable en-
ergy supply UNDP, in consultation with the RGC, decided to launch a rural energy initiative.
This new initiative will contribute to achieving the NSDP macro indicators, especially Indicator
25: reduction of fuel wood dependency of households.
UNDP, in partnership with other donors, is recognized and accepted in Cambodia as a key cata-
lyst for rural development and poverty alleviation. The key value-added stems from UNDP’s
ability to promote a policy framework that can facilitate integrated rural development solu-
tions with energy as a theme that cuts across sectors such as agriculture, infrastructure, health,
education and rural business development. Given the importance of securing rural energy
services, UNDP Cambodia took a first step in developing a rural energy program with partici-
pation in the Regional Energy Program for Poverty Reduction (REP-PoR), which seeks broad-
based interventions to improve access to energy services, promote efficient use of energy
and increase access to financing for sustainable energy. The regional process was followed by
country-level stakeholder consultations with both government and civil society to determine
specific energy-poverty linkages and related gaps, and to suggest mechanisms and modali-
ties to close them. The REP-PoR analysis revealed that the lack of a pro-poor vision framework
hampers the effectiveness of the ongoing energy programs in meeting overall socioeconomic
development goals. This is exemplified by inadequate program focus on energy sources other
than grid electricity, while it is clear that a large portion of Cambodia’s rural population will not
enjoy grid access in the foreseeable future.
4 UNDP Cambodia
Charcoal production, and the ensuing deforestation in Kampong Speu, threatens the Aural Wildlife Sanctuary.
Following participation in the regional REP-PoR, UNDP Cambodia launched the national ini-
tiative ‘Preparatory Activities for a Rural Energy Development Program’. While at the national
level considerable work has been done in the area of energy policy development, strategic
planning and capacity development, comprehensive planning efforts have not been made
to reach out to the communes and villages where pro-poor rural energy projects need to be
planned and implemented. As part of the preliminary rural energy activities, this study aims
at providing empirical data and baseline information that allow effective planning of rural
energy projects and efficient monitoring of their results.
1.4 Objectives
The objective of the study presented here is to provide a framework exploring options with the
government, donors and other development partners for Cambodia’s future directions in rural
energy development. Collection and analysis of energy supply and demand data for two pilot
provinces provide a profile of existing conditions in the rural energy sector, and highlight areas
for priority attention to achieve sustainable energy supply. These priorities include enabling
the energy sector to expand its contribution to rural incomes and employment. The baseline
data compiled will also allow better design, planning, and implementation and monitoring of
rural energy projects. The study is not a stand-alone exercise in social research. It provides the
basis for the development of a major rural energy project that UNDP intends to implement
together with the RGC and other development partners.
1.5 Approach and methodologies
The study presented here was undertaken by the Ministry of Industry, Mines and Energy (MIME)
with funding support from UNDP and technical assistance provided by the international NGO
GERES (Groupe Energies Renouvelables, Environnement et Solidarités – Group for Renewable En-
ergy, Environment and Solidarity). A multidisciplinary team performed a variety of tasks over
UNDP Cambodia 5
TITLE
a period of approximately 12 months. The study explores empirically and conceptually the
connections between energy supply and demand and poverty in Cambodia. It aims to show
that securing a sustainable energy supply should be considered a core component of any rural
poverty alleviation strategy. It emphasizes how and why rural energy interventions must be
carefully designed to consider poverty concerns. Thus, the methodology aims to capture the
relationship between access to modern energy, affordability and the potential of income gen-
eration through energy-related rural activities. The study’s ultimate goal is to lay the ground-
work for the design of a poverty-oriented rural energy project that would be supported by
the RGC, UNDP and a variety of other partners. Under the activity ‘Preparatory Activities for a
Rural Energy Development Programme’, all necessary primary and secondary data have been
collected to assess residential rural energy demand with the view to formulate a ‘Rural Energy
Development Project’. Survey work for this study was accompanied by training of local staff
in data management. Capacity development included practical training of MIME staff in the
General Department of Energy and staff in the provincial Departments of Industry, Mines and
Energy (DIME) in Kampong Speu and Svay Rieng.
This publication is the result of a combination of field research, desk studies, policy analysis,
capacity development efforts and stakeholder consultations. Empirical work was based on
surveying a sample of households using questionnaires. The study also reflects the contri-
butions of government staff, policymakers, academics and community representatives. The
National Institute of Statistics (NIS) analyzed existing secondary data. Various national sur-
veys that have been executed by NIS since 1993 were used to identify trends and cross-check
empirical data. Based on these data, energy trends have been derived showing changes in
energy use patterns for Phnom Penh, and provincial urban and rural areas.
Survey teams visited households in Kampong Speu and Svay Reign to collect data on rural energy use.
Field research has been conducted in two provinces: Kampong Speu and Svay Rieng. A total
of 2,159 households were randomly selected and interviewed about energy use for cooking,
lighting and other appliances (e.g., rice cooker, radio/TV). Prior to field research, surveyors
were trained. The survey teams spent three weeks in each province. The level of confidence of
6 UNDP Cambodia
the sample size is more than 95 percent, and 1,050 households in Kampong Speu and 1,109 in
Svay Rieng were surveyed with respect to their energy use. Table 1 displays the sample sizes
used for the household survey. It should be noted, however, that accurate measurement of
the units in which energy is used can be critical to data quality. The difficulty of accurately cap-
turing this information varies substantially across energy forms. Some are measured in formal
scientific units such as kg LPG or kWh of electricity. However, respondents often think more in
terms of wood bundles or in terms of cylinders and bottles in which a fuel is sold. Kerosene,
for example, is regularly sold in a variety of bottles having different contents. The wood con-
sumption figures in particular involve a considerable uncertainty. This problem was overcome
by providing the enumerator with scales so that they could measure fuel wood samples and
ensure that the correct weight was entered.
Securing a sustainable energy supply should be considered a core component of any rural poverty alleviation
strategy.
A distinction between urban and rural households was also made. The percentages of urban
and rural households in the sample were defined according to the statistics on urban and rural
households in the NIS database. The data were analyzed with the statistical software SPSS and
further energy specific calculations and simulations were performed using spreadsheet soft-
ware and the software tool Long Range Energy Alternative Planning (LEAP). The LEAP system
is a Windows-based software system for energy and environmental policy analysis. It is widely
used for integrated energy planning and climate change mitigation analysis and has been
applied in hundreds of different organizations in over 140 countries. LEAP is developed and
supported by the U.S. Center of the Stockholm Environment Institute, a non-profit research
institute based at Tufts University in Somerville, Massachusetts.
Table 1 Household and sample size in selected provinces

Urban households % Urban Rural Households % Rural Total
Total Cambodia 382,114 15% 2,217,054 85% 2,599,168
Kampong Speu 8,875 6% 128,253 94% 137,128
Sample Size 62 6% 988 94% 1,050
Svay Rieng 3,517 3% 9,522 97% 113,039
Sample Size 138 12% 971 88% 1,109
UNDP Cambodia 7
TITLE
The rationale for selecting the two provinces was based on a variety of considerations:
• First, the two provinces cover all three topographical zones found in Cambodia and
their associated agro-ecological characteristics: alluvial plains, undulating terrain and
mountainous areas.
• Second, the provinces represent different endowment levels with respect to natural
resources. While Svay Rieng represents a typical lowland paddy rice-producing area
with a high population density and very little in natural forest resources, Kampong Speu
shows significantly more diversification in economic activities including forest products,
palm sugar and wine. Serving the urban charcoal markets, Kampong Speu is an energy
exporter, while resources in Svay Rieng are insufficient to satisfy the province’s wood
fuel needs. Both provinces require support to ensure a sustainable energy supply.
While the study’s main goal is to establish baseline data for project preparation and monitor-
ing, the study also extrapolates findings to the national scale. Based on the data in Kampong
Speu and Svay Rieng the total energy demand in Cambodia is projected and forecast. The
forecast is based on historic trends established by the NIS and empirical data from the survey.
Forecasting rests on the assumption that rural areas will gradually follow the energy develop-
ment path of provincial urban centres, and provincial urban areas that of the Phnom Penh
area. In addition to the statistical analysis, a detailed study on wood and charcoal flow to
Phnom Penh and wood and charcoal production was conducted in Kampong Speu. Data was
gathered by monitoring trucks and interviewing 130 wood and charcoal producers.
Data collected in interviews established a baseline for project preparation and monitoring, as well as allowing the
findings to be extrapolated to a national scale.
8 UNDP Cambodia
2. Survey results
Given the role that energy services play in poverty reduction, the RGC desires to implement
policies and allocate resources that will accelerate the transition from the use of traditional
biomass fuels to clean modern fuels and electricity. As a consequence, policymakers increas-
ingly seek empirical evidence of the relationship between investments made in energy supply
infrastructure, the energy policies they implement, and tangible improvements at the house-
hold level. The survey results discussed below provide such evidence.
2.1 The provinces of Kampong Speu and Svay Rieng
Administratively, Cambodia is divided into 24 Provinces (including four municipalities). At the

next level there are 183 districts, then approximately 1,600 communes, and 13,400 villages. In
rural areas, there are typically eight villages in each commune, with approximately 150 house-
holds in each village. In the rural area, the average household size is 5.2 occupants. Both
Kampong Speu and Svay Rieng are classified as rural provinces. With respect to their resource
endowment, their agricultural production and their energy economy, the two provinces show
significant differences. Table 2 summarizes key data on the two provinces.
Table 2 Key features of the surveyed provinces

Feature Kampong Speu Svay Rieng
Total Population 714,000 532,000
Number of Households 138,000 110,000
Surface Area km2 7,017 2,966
Population Density persons/km2 102 180
Average Household Size 5.2 4.8
Number of Districts 8 7
Kampong Speu
Kampong Speu is located to the west of Phnom Penh. It borders Kampong Chhnang and Pur-
sat to the north, Phnom Penh to the east, Kampot and Takeo to the south and Koh Kong to the
west. The area of the province is 7,017 square kilometres. Kampong Speu’s topography is vari-
able, from a large area of lowland paddy fields in the east to lowland and upland forested areas
in the west. These forest areas provide the feedstock for the province’s important charcoal pro-
duction. Cambodia’s highest mountain, Phnom Aural, is located here. In 2004 the estimated
population was 714,000. The population of children aged under-five-years was 95,000, which
was 13 percent of the total provincial population (NIS 2004). The total number of households
in 2004 was 138,000, giving an average household size of 5.2 persons.
The people of Kampong Speu live in eight districts composed of 87 communes and 1,358 vil-
lages. The population density of the province is 102 persons per square kilometre compared
to an average population density for Cambodia of 75 per square kilometre. In Kampong Speu
the rice balance falls far below the minimum rice food needs of the population. Among the
communes of the province, only 9 percent of communes produce enough rice to meet mini-
mum food needs, while 91 percent produce less than minimum food needs.
UNDP Cambodia 9
TITLE
Svay Rieng
Svay Rieng is located in southeastern Cambodia. The entire province is in the low-lying alluvial
plain, which comprises about 25 percent of Cambodia’s total area. It borders Kampong Cham
to the north, has a long border with Vietnam from east to south, and borders Prey Veng to the
west. The area of the province is 2,966 square kilometres. The topography is mostly lowland
paddy fields. Upland and lowland forests are only found in the northern district of Romeas
Hek. In 2004 the estimated population was 532,000. The population of children aged under-
five-years was 57,000 thousand, which was 11 percent of the total provincial population. The
total number of households in 2004 was 110,000, giving an average household size of 4.8
persons.
The people of Svay Rieng live in seven districts composed of 79 communes and 690 villages.
With 180 persons per square kilometre, the province is comparatively densely populated. The
high population density and the lack of forest resources in Svay Rieng are responsible for a
critical supply situation with respect to biomass supply for energy purposes. With a rice pro-
duction of 177 percent of its needs, Svay Rieng province produces a significant rice surplus.
Figure 1 Map of Cambodia showing the two pilot provinces
2.2 Social classes
Rural families in Cambodia face numerous challenges and risks that limit rural earnings and so-
cial progression. The predominance of subsistence agriculture and dependence on common
property natural resources limit incomes, especially for the poorest households. Poor rural
10 UNDP Cambodia
households operate in a risky environment in which the incidence of shocks and crises such as
floods, crop and animal losses through disease, health crises, and land grabbing by powerful
groups or individuals is high. Low and seasonal agricultural revenues force rural households
to continually search for work or other income-generating activities, the majority of which
would be temporary and poorly rewarded. Many activities are directly or indirectly related to
the rural energy sector, in particular the wood fuel sub-sector. Significant improvements in
how natural wood and forest resources are managed are essential to ensure sustainable avail-
ability of natural resources.
In order to better understand the rural poverty-energy link and to allow an analysis of rural
energy problems from a poverty perspective, the study categorized the surveyed population
according to social classes. In line with common practice in rural social surveys, the quality
of housing has been used to define four social classes. Figure 2 displays the variety of house
types found in rural Cambodia.
Figure 2 House types

Brick Half brick, half wood Wood with tile roof Wood with iron Thatch with iron
or asbestos roof or thatch roof

In order to rationalize social classes to a manageable number, only four housing categories
were used as a proxy for social class. Table 3 displays the social classes together with the num-
ber of households that fall into the respective housing categories in each province.
The social class categorization does not provide accurate information on household income.
It is, however, safe to assume that there is a strong correlation between housing standard and
family income. Interestingly, households had difficulties stating their monthly incomes. The
reason may be a combination of a number of factors including unwillingness to disclose in-
come, significant fluctuations in income (as a function of season, harvests, and casual jobs),
and a changing variety of income sources.
However, the households were in a position to provide aggregated monthly expenditure fig-
ures. With monthly expenditures averaging US$47, Svay Rieng appears to be a poorer prov-
ince than Kampong Speu where monthly expenditures average US$76. The relative poverty
of Svay Rieng is also supported by the lower housing standard found in this province. In Kam-
pong Speu, on the other hand, brick and wood houses were more common and no houses in
the lowest housing category were found there (see Table 3 and Figure 3).
Table 3 Sample size social classes

No. Class House type Households Households
Kampong Speu Svay Rieng
1 Upper social class Brick, brick/wood 161 78
2 Middle social class Wood + tile/iron/asbestos 652 555
3 Lower social class Thatch + iron 148 306
4 Lowest social class Clay + iron/thatch 0 31
UNDP Cambodia 11
TITLE
Figure 3 Social classes in Kampong Speu and Svay Rieng
2.3 Energy used for cooking
Energy for cooking is the most basic energy need in Cambodia where both urban and rural
households still rely on biomass fuels. Exposure to smoke from the burning of traditional fuel
wood for cooking increases the risk of disease, most notably acute lower respiratory infection,
in both children and adults. Women and children are disproportionately affected because of
women’s primary role in food preparation and child care, which exposes them to indoor air
pollution.
Cooking in urban areas
Urban households typically use several energy sources for cooking. The most common are
firewood, LPG, charcoal and electricity, the latter mainly used for rice cooking. The survey
results reveal significant differences between the two provinces: in Svay Rieng even the urban
households collect most of their firewood themselves; in Kampong Speu most of the firewood
demand of urban households is met by purchasing. The user percentages for purchased fuel
wood, LPG and electricity for rice cookers are significantly higher in Kampong Speu than in
Svay Rieng. The widespread use of electric rice cookers – 60 percent of the urban households
in Kampong Speu use these appliances – is another distinguishing feature that underlines
the more sophisticated energy economy of Kampong Speu. In Svay Rieng only 13 percent of
urban households use electric rice cookers.
Table 4 depicts the urban energy mix used for cooking. The figures illustrate significant differ-
ences in the two provinces. In Kampong Speu, 42 percent of the households use commercial
firewood. Consumption averages 90 kg per month. LPG use is widespread, with 66 percent of
the households using an average quantity of 10 kg per month. Households that use LPG may
also use purchased wood and many families also have electric rice cookers. Although 13 per-
cent of urban Svay Rieng households own rice cookers, this appliance is used less frequently
there where electricity is relatively cheap. With an average consumption of only four kWh per
month, rice cookers would only be used two or three times a week, whereby a consumption of
ten kWh in Kampong Speu suggests a much more frequent use of the cookers there.
The results of the survey show clearly that households are using more than one source of en-
ergy for cooking. Although Svay Rieng has very little forest resources, even urban cooking is
12 UNDP Cambodia
still dominated by firewood. Thirty percent of the households use commercial firewood and
67 percent of the families collect firewood themselves. Even in the absence of closed forest
cover, it seems that the biomass resources of Svay Rieng and its neighbouring provinces are
still sufficient to support an energy economy dominated by fuel wood. There are, however,
strong indications that the current biomass usage rate is not sustainable in Svay Rieng. While
the fuel wood cycle has not been studied in detail, it appears that urban fuel wood demand in
Svay Rieng is satisfied by a combination of harvesting of woody biomass (mostly bushes and
dead wood from isolated trees) and of wood imports from forested areas in the north of Svay
Rieng and from neighbouring provinces. A baseline study for the National Biodigester Pro-
gram showed that 67 percent of the people collecting firewood had to go further and further
every year.
Figure 4 Share of households using particular energy sources
Table 4 Energy sources used for cooking in urban Kampong Speu and Svay Rieng
Urban cooking Kampong Speu Svay Rieng
Energy source HH User % Units per HH per month HH User % Units per HH per month
Firewood purchased 42% 90 kg 30% 147 kg
Firewood collected 5% 40 kg 67% 143 kg
Palm fronds 0% 0 8% 70 kg
Charcoal 61% 28 kg 35% 16 kg
LPG 66% 10 kg 38% 8 kg
Electricity (rice cooker) 60% 10 kWh 13% 4 kWh
Cooking in rural areas
Household energy consumption in rural areas follows a pattern that is similar to cooking prac-
tices in urban areas. Table 5 indicates that low cost energy sources are more prominent in both
provinces, both in absolute and in percentage terms. Eighty-seven percent of Svay Rieng’s rural
households use collected firewood. However, with 145 kg, the quantities used per household
per month are lower than in Kampong Speu where families that collect firewood consume an
average of 175 kg per month. While this finding supports the hypothesis that Svay Rieng faces
resource constraints with respect to firewood, the consumption figures for both collected and
commercial firewood are still high.
UNDP Cambodia 13
TITLE
Energy for cooking is the most basic energy need in Cambodia where both urban and rural households still rely on
biomass fuel.
User percentages for modern fuels are significantly higher in rural Kampong Speu where 12
percent of the households use LPG and 2 percent use electricity. Charcoal consumption is also
significantly more prominent. This higher penetration of modern fuels in Kampong Speu is
most probably explained by higher cash incomes of rural households. It could also indicate
local shortages of freely accessible firewood. The consumption pattern in Svay Rieng sug-
gests that rural families resort to the collection of any type of biomass to satisfy their cooking
needs. The increased use of low-grade fuels, such as palm fronds and animal dung, indicates
constraints in the availability of proper fuel wood. Although a small proportion of the families
use only animal dung, households that resort to this low-grade fuel use significant quantities
(169 kg per household per year). The LPG plus electricity combination, undoubtedly the top
combination on the so-called energy ladder, moved from being a prevalent energy mix in ur-
ban Kampong Speu to a much lower position in rural areas.
Table 5 Energy sources used for cooking in rural Kampong Speu and Svay Rieng
Energy source HH User % Units HH User % Units
per HH per month per HH per month
Firewood purchased 38% 149 kg 13% 148 kg
Firewood collected 53% 175 kg 87% 145 kg
Palm fronds 22% 79 kg 15% 80 kg
Charcoal 24% 37 kg 6% 10 kg
LPG 12% 9 kg 8% 4 kg
Electricity (rice cooker) 2% 9 kWh 0.5% 7 kWh
Animal dung 1% 83 kg 1% 169 kg
2.4 Energy for lighting and appliances
Urban lighting
Urban areas in the two provinces show energy consumption patterns similar to those found
in other Asian countries. Besides lighting, a considerable variety of appliances can be found
in households ranging from rice cookers to refrigerators, fans, TVs and other electronic equip-
ment including the occasional computer. In both provinces urban electrification rates are high
with 98 and 97 percent in Kampong Speu and Svay Rieng respectively. Surprisingly, there is
14 UNDP Cambodia
still widespread use of candles and kerosene, particularly in Svay Rieng where electricity is pro- Batteries are still
vided by cross-border connections from Vietnam. Cross-border supply from Vietnam is sold at very popular as
standard Electricité du Cambodge (EDC – Electricity of Cambodia) rates and thus significantly an electricity
cheaper than Rural Electricity Enterprise (REE) supplied diesel electricity or self-generation. source
Although kerosene consumption is rather low in absolute terms, the finding that 40 percent of
households in Svay Rieng still use kerosene for lighting is not easy to explain. Lack of reliability
and frequent power outages may be one reason for the households to maintain a diversified
energy mix. The same reason may be behind the common use of candles in the province.
Households in Kampong Speu that are served by REE may have to use candles, batteries and
kerosene because electricity is only supplied for a couple of hours each day at a scheduled
time. Through the use of kerosene they may also want to save on costs, as REE electricity rates
are quite high.
Table 6 Energy sources used for lighting and other appliances in urban areas
Energy source HH User % Unit per HH per month HH User % Unit per HH per month
EDC (grid electricity) 32% 48 kWh 97% 31 kWh
REE 66% 67.5 kWh – –
Own Generator – – – –
Battery 2% 2.4 kWh 9% 1.8 kWh
Kerosene 8% 1 kg 40% 1.2 kg
Candles 24% 0.6 kg 38% 0.47 kg
Rural lighting in Kampong Speu and Svay Rieng
In the absence of grid supply, rural households in

both provinces satisfy their lighting needs using
lead acid batteries and kerosene. With only 23 per-
cent and 7 percent of the rural households served
by either EDC or REEs, the most prevalent energy
mix for rural lighting and appliances is clearly a
combination of batteries, kerosene and candles.
Table 7 illustrates that higher electrification rates

in Kampong Speu – 27 percent of households
have EDC, REE or their own generators – reduces
the prevalence of batteries and kerosene. While
62 percent of the Kampong Speu households use
batteries, the figure climbs to 84 percent in Svay
Rieng where only 7 percent are supplied by grid
electricity.
The results also demonstrate the importance of

electricity to rural households: in both provinces
rural households make sure that they have some
In the absence of grid supply, rural households in
access to electricity even if it is only based on 12- both provinces satisfy their lighting needs using
volt batteries. Electricity consumption from lead lead acid batteries and kerosene.
acid batteries averages at 3.8 kWh per months in
Svay Rieng. This is a surprisingly high value.
UNDP Cambodia 15
TITLE
Assuming that the standard battery holds a usable charge of not more 0.5 kWh, this figure im-
plies that a family in this province buys two battery charges a week. Despite the widespread
use of batteries, kerosene is still a very important source for lighting showing user percentages
of 62 percent and 75 percent respectively for Kampong Speu and Svay Rieng. This again dem-
onstrates that rural households maintain a portfolio of energy options, probably in an attempt
to satisfy energy service needs at a minimum cost.
Table 7 Energy sources used for lighting and other appliances in rural areas
HH User % Quantity per HH per month HH User % Quantity per HH per month
EDC 2% 27.71 kWh 6% 27.62 kWh
REE 21% 20.57 kWh 1% 11.09 kWh
Own Generator 4% 5.61 kWh 2% 3.07 kWh
Battery 62% 1.38 kWh 84% 3.80 kWh
Kerosene 62% 1.88 kg 75% 1.34 kg
Candles 21% 0.74 kg 9% 0.63 kg
2.5 Energy use by social class
This section explores the relation between energy consumption and the households’ social
class. Although the survey performed did not include questions with respect to household
income, the division of the households into three classes in Kampong Speu and four classes
in Svay Rieng can be considered as a proxy for household income. This is underlined by the
results of the energy expenditure survey that shows significant differences between the four
classes identified. Figure 5 displays the results of this survey.
Figure 5 Monthly energy expenditure by social class
Table 8 confirms the close correlation between house type and energy expenditure: the bet-
ter the housing standard, the higher the energy expenditure. Even within the social classes
defined in the survey, this correlation holds.
The expenditures depicted in Table 8 correspond with current average prices for energy ser-
vices in the two provinces as shown in Table 9. The actual cost for the energy services listed in
Table 9 varies quite considerably. Lead acid batteries show the highest cost fluctuations per
unit of energy. Not only do charging costs vary significantly, the batteries are often in poor
16 UNDP Cambodia
shape and hold significantly less usable electricity than their nominal rating would suggest.
The cost of privately generated electricity can also vary quite considerably per unit (kWh) as
generator efficiency and fuel supply costs fluctuate from application to application.
Table 8 Monthly expenditure per house type

Social Type of house Monthly expenditure Type of house Monthly expenditure
Class $ $
1 Brick 146 Brick 74
Brick/Wood 113 Brick/Wood 73
2 Wood/Tile 79 Wood/Tile 63
Wood/Corrugated Asbestos 67 Wood/Corrugated iron 47
Wood/Corrugated Iron 66 Wood and thatch 44
3 Thatch 35 Thatch/Corrugated iron 33
4 Clay/Corrugated iron or thatch 23
Total average 76 Total average 47
Table 9 Average cost per unit of energy

Energy source Average cost per unit
EDC grid power 720 Riel per kWh (650 Riel per kWh for Svay Rieng)
Privately generated electricity 2000- 3800 Riel per kWh
Battery 1000 Riel per 50 Ah battery (4000 Riel per kWh)
Gas (LPG) 3500- 3750 Riel per kg
Kerosene 3000-3300 Riel per litre
Charcoal 300-500 Riel per kg
Firewood 200 per kg (or bundle)
Cooking energy and social class
A similar correlation can be seen in Table 10, this time in an inverted manner: in both provinces,
the poorer the household, the more likely it is that firewood is being collected. What is a real
surprise is that survey results suggest that households that do collect firewood in Kampong
Speu spend more than double the time of their counterparts in Svay Rieng. This phenomenon
is not easy to explain, especially in the light of the figures presented in Table 5, which suggest
that the amount of fuel wood used per household is practically identical for both provinces.
Anecdotal evidence suggests that many villages in Svay Rieng have a considerable number of
trees scattered in the vicinity of the dwellings, allowing families to collect tree branches as fuel
wood without travelling very far. The uncertainty with respect to biomass resources clearly
indicates that the data on biomass requires verification and further analysis.
Even households in the upper social classes (25 percent of the households in Kampong Speu
and 60 percent in Svay Rieng) collect their firewood themselves. Only the highest social class
shows a relatively low user percentage (around 30 percent in both provinces). The other social
classes, especially in Svay Rieng, show extremely high user percentages. In that province, 88
percent of the middle social class, 89 percent of the lower social class and 90 percent of the
lowest social class collect fuel wood. In Kampong Speu these percentages are lower, but still
considerable: 55 percent for the middle social class and 80 percent for the lower social class.
UNDP Cambodia 17
TITLE
Table 10 Average hours spent collecting firewood per social class

Social Type of house % per Hours per day Type of house % per Hours per day
class type of type of
house house
1 Brick 7% 2.33 Brick 66% 0.83
Brick/Wood 31% 2.37 Brick/Wood 57% 1.17
2 Wood/Tile 59% 2.88 Wood/Tile 85% 1.11
Wood/Corrugated 61% 2.49 Wood/Corrugated iron 84% 0.97
Asbestos
Wood/Corrugated 59% 2.44 Wood and thatch 91% 0.98
Iron
3 Thatch 80% 3.33 Thatch/Corrugated iron 86% 0.83
4 Clay/Corrugated iron 100% 0.87
or thatch
Total (average) 57% 2.81 Total (average) 84% 0.95
Further analysis of the data obtained provides some interesting insights into the modalities of
firewood collection. Contrary to conventional wisdom, men seem to be the main providers of
fuel wood. Figure 6 illustrates that this holds for both provinces and all social classes. In Svay
Rieng this percentage (averaging 50 percent) is a little higher than in Kampong Speu (averag-
ing 40 percent). Women seem to take on a third of the workload, with percentages increasing
in the lower social classes. This is more prominent in Svay Rieng.
The demand on children to collect

firewood differs between the two Cooking practices observed
provinces. In Kampong Speu 27 per-
cent of the children in the upper so- In order to verify the figures obtained in the survey
on energy consumption for cooking purposes, ten
cial class are responsible, while this is
households in Vor Sor village (Samrong Tong district,
21 percent for the lower social class. Kampong Speu province) were selected for comple-
In Svay Rieng the same decrease in mentary small scale research. Household energy
participation is noticeable, but the consumption was monitored for five days. The fami-
difference between the upper so- lies consisted of five to seven household members.
cial class and lowest social class is The observations coincided with the times when the
more significant. In 29 percent of families had their meals (breakfast, lunch and din-
the households a child is the main ner). The households were using different energy
responsible for collecting firewood sources (wood, charcoal and LPG) and tools (stone
in the upper social class, while this is stoves, LPG with one or two burners, traditional Lao
stoves) for cooking.
only 10 percent for the lowest social
class. The results show more similarities than differences
in cooking practices between the types of houses
Charcoal, a refined and cleaner form (brick/wood/thatch). In all the households, women
of fuel wood, is mostly used by high- make decisions about cooking and also carry out the
er income households. Charcoal actual cooking. Firewood is the main energy source
production involves the removal used for cooking, and the supply is stored in or close
of volatile matter from woody bio- to the kitchen area. Only two families used LPG and
mass, providing a fuel that is high in only one used charcoal, both LPG and charcoal were
pure carbon. Charcoal burns almost used in limited numbers. All families had a prefer-
ence for firewood since it is cheaper than LPG and
smokeless and is less polluting than
charcoal. They all indicated they would rather use
wood. Forty-four percent of high-in- LPG or charcoal but that it is too expensive for them.
come households in Kampong Speu One kg of firewood costs 200 Riel, one kg of charcoal
and 24 percent in Svay Rieng use 700 Riel and 200 gram of LPG costs 800 Riel.
charcoal. The percentage decreases
18 UNDP Cambodia
rapidly with class. In Svay Rieng the middle, lower and lowest social classes use hardly any
charcoal at all. LPG is mainly used by higher social classes. On average the user percentage
is 13 percent for Kampong Speu and 8 percent for Svay Rieng. When LPG is used, the higher
social classes use the highest quantity per month.
Figure 6 Percentage of responsible household members collecting firewood per social

class
Rural middle class housing
Lighting and appliance energy and social class
Poor families in rural areas in Cambodia are highly dependent on kerosene as an energy source
for lighting. The relatively high cost of kerosene is a burden for most families, but better al-
ternatives are not always available or are too costly. Especially for households without access
to electricity and batteries, kerosene is the best available energy source at the moment. This
is reflected by the higher user percentage and quantities used per month by the lower social
classes.
UNDP Cambodia 19
TITLE
Rising kersosene Table 11 Kerosene use per social class per month
prices are a Kampong Speu Svay Rieng
burden for poor Social Type of house % per type Litres per Type of house % per type Litres per
rural households class of house month of house month
1 Brick 24% 1.75 Brick 53% 1.13
Brick/Wood 57% 1.83 Brick/Wood 72% 1.49
2 Wood/Tile 63% 1.79 Wood/Tile 75% 1.22
Wood/Corrugated asbestos 63% 1.83 Wood/Corrugated iron 72% 1.50
Wood/Corrugated iron 64% 1.82 Wood and thatch 82% 1.29
3 Thatch 80% 2.16 Thatch/Corrugated iron 82% 1.29
4 Clay/Corrugated iron or thatch 84% 1.63
Total (average) 64% 1.88 Total (average) 76% 1.35
Table 11 illustrates that the kerosene user percentages are higher in Svay Rieng than in Kam-
pong Speu, indicating a less sophisticated energy economy in the former province. Again a
very strong correlation between user percentages and social class becomes obvious: the lower
the social class and income, the higher the user percentage for kerosene.
Electricity is used in two forms: grid power from either EDC or REE and battery power from lead
acid batteries. User percentages in the upper social class in Kampong Speu show a significant
percentage of high-income households using both forms, which indicates a high diversifica-
tion of energy options. Scheduled electricity supply in Kampong Speu, where electricity is
mostly provided by REE for a couple of hours a day, could be a reason for high-income house-
holds to maintain redundancy.
Table 12 Cumulative average consumption figures and user percentages from

Kampong Speu
Electricity Batteries Kerosene Firewood Firewood Charcoal LPG
purchased collected
Social User KWh/ User Capacity User Litres/ User Kg/ User Kg/ User Kg/ User Kg/
class % month % Ah % month % day % day % day % month
1 54% 26.7 70% 95 48% 1.82 52% 4.54 29% 5.81 44% 1.34 30% 10.87
2 22% 18 80% 75 64% 1.80 41% 5.00 55% 5.88 23% 1.15 11% 8.45
3 2% 10 60% 61 80% 2.16 15% 4.87 80% 5.41 11% 1.19 2% 6.20
In Svay Rieng the figures indicate that grid power and batteries are the only sources of electric-
ity. High-income households also use significantly more kWh per month than lower classes.
The same holds for the average size of battery. Battery size should not, however, be confused
with electricity consumption per time unit. It is the frequency of charging that determines the
amount of electricity consumed. A maximum value extractable from a 100 Ah 12 volt battery
would be approx 3 kWh per week when well managed and charged properly.
The absolute consumption figures for electricity in Svay Rieng are higher for all social classes
that use power. This can be explained by the fact that power would be EDC supplied from
cross-border connections with Vietnam, making unit cost significantly cheaper than in Kam-
pong Speu. There seems to be a certain degree of price elasticity here. This is supported by
consumption figures of LPG, another modern urban fuel. Though retail prices for LPG are iden-
tical in both provinces, Svay Rieng LPG users use less LPG than their Kampong Speu counter-
parts. This seem to indicates that electricity may – for some applications, such as rice cooking
– be a more convenient and perhaps cheaper choice in Svay Rieng where power is supplied at
comparatively low rates by EDC.
20 UNDP Cambodia
Table 13 Cumulative average consumption figures and user percentages Svay Rieng
Electricity Batteries Kerosene Firewood Firewood Charcoal LPG
purchased collected
Social User KWh/ User Capacity User Litres/ User Kg/ User Kg/ User Kg/ User Kg/
class % month % Ah % month % day % day % day % month
1 35% 33.4 67% 77 60% 1.29 29% 4.54 63% 4.81 24% 0.36 17% 6.46
2 8% 21 90% 63 75% 1.38 13% 4.97 88% 4.93 6% 0.33 10% 3.24
3 4% 12 84% 44 82% 1.29 8% 4.81 89% 4.42 2% 0.39 3% 3.81
4 0% 0 74% 42 84% 1.63 0% 0 90% 3.84 3% 0.03 0% 0
Table 14 shows energy expenditure for actual users. This provides some very interesting in-
sights: although all forms of energy are purchased and used, fuel wood is still the predominant
form of energy throughout all social classes. For the upper and middle social classes, LPG
also plays an important role – more prominent in Kampong Speu than in Svay Rieng. In both
provinces, both of the middle social classes spend most on firewood, LPG and charcoal and the
lowest social class spends most on kerosene and batteries. The lowest variation is found with
kerosene: it is used across the classes and provinces and it appears that quantities consumed
by individual households are in the same order of magnitude. Another interesting result is the
importance of charcoal as a fuel in Kampong Speu. Charcoal users in all three classes spend
nearly US$4 a month, an amount similar to that spent by electricity users for grid power.
Table 14 Average expenditure per month per energy type per social class (US$)
Electricity Batteries Kerosene Firewood (purchased) Charcoal LPG
Kampong Speu
Kampong Speu
Kampong Speu
Kampong Speu
Kampong Speu
Social class Kampong Speu

Svay Rieng
Svay Rieng
Svay Rieng
Svay Rieng
Svay Rieng
Svay Rieng
1 4.81 5.43 2.55 1.70 1.46 1.07 6.81 6.81 4.02 1.08 10.05 5.98
2 3.24 3.41 1.70 1.70 1.44 1.22 7.50 7.46 3.45 0.99 7.82 3
3 1.80 1.95 1.70 1.19 1.73 1.07 7.31 7.22 3.57 1.17 5.74 3.52
4 - - - 1.19 - 1.71 - - - 0.09 - -
2.6 Supply cost and affordability
In order to gauge the significance of energy expenditure and its relevance for poverty allevia-
tion, the average expenditure on energy sources as a percentage of total household expendi-
ture has been analyzed. Table 15 depicts results that are somewhat surprising as all surveyed
households consistently reported a figure of 11 to 12 percent in Kampong Speu and 8 to 10
percent in Svay Rieng. One could speculate that the 10 percent figure might be a household
target. The option to collect free firewood in case of a cash shortage allows a household to
keep energy expenditure in this range. In other words, cash management is helped by the
availability of the wood collection option.
UNDP Cambodia 21
TITLE
Woodlots Table 15 Share of energy costs of total monthly expenditure per social class
generate income Monthly total expenditure Monthly energy expenditure Energy expenditures as
and fuel at the $ per household $ per household % of total expenditure
same time Social class Kampong Speu Svay Rieng Kampong Speu Svay Rieng Kampong Speu Svay Rieng
1 122 74 13.41 6.93 11% 9%
2 73 52 7.72 4.05 11% 8%
3 35 33 4.04 2.66 12% 8%
4 - 23 - 2.32 - 10%
In absolute terms, the picture looks dismal for the poorest households in Svay Rieng which,
after energy expenditure, have just US$20 to cover all other expenses such as food, cloth-
ing, education, transport, etc. This
picture may deteriorate rapidly in the
future if depletion of natural resources
is not accompanied by rural growth
based on both extending areas under
agricultural production and ensur-
ing sustainable availability of natural
resources. While the current natural
resources management framework
has provided very limited benefits in
terms of rural growth and poverty re-
duction, there are several indications
of rapidly decreasing availability of
biomass resources. Both intensifica-
tion and expansion of agricultural ac-
tivities will put additional pressures on
the resource base. This indicates the
need to establish viable land-use plan-
ning mechanisms, and develop social-
ly acceptable mechanisms for allocat-
ing biomass use rights to agricultural
households. Secure access to natural
resource assets will be an essential fix
to broad-based rural growth and re- Charcoal production is a thermal process that involves the
duction of vulnerability. removal of volatile matter from woody biomass, providing a fuel
that is high in pure carbon.
2.7 Energy Transition
A significant body of literature on household energy transitions that has focused on how
households ascend what has been called an energy ladder supports the findings of this study.
At the bottom of the ladder, many people use low-grade biomass fuels (e.g., straw, dung, and
wood) in inefficient and unhealthy ways. At transitional points along the ladder, consum-
ers purchase commercial fuels such as charcoal and kerosene. Finally, consumers change to
cleaner, efficient and more convenient forms of energy such as LPG and electricity. The ladder
model assumes that households move in incremental steps on the ladder, a theory that is sup-
ported by the findings of the household surveys performed in Kampong Speu and Svay Rieng.
However, at the same time there are many indications that moving up is, in reality, a complex
process. Various factors determine the energy transition of households in Cambodia. Access,
income, price, and degree of urbanization are clearly critical parameters that determine speed
and degree of improvements in living conditions.
22 UNDP Cambodia
While increases in household incomes are an obvious Figure 7 Energy transition High up-front
driver, very often access or availability forces households cost of modern
to move on the ladder. As rural settlements urbanize, fuel energy systems
wood becomes scarcer or unavailable while commercial obstacle for rural
fuels become more readily accessible as a result of im-
poor
proved economies of scale in transport and fuel distribu-
tion. However, the surveys have also discovered that the
pathways to a better energy supply are often multifaceted
and do not necessarily always involve a smooth transition
from one rung to the next. Households in Cambodia seem
to maintain a range of fuel options leading to the variety
of possible transitional conduits shown in Figure 7.
This is an important insight and should be seriously con-

sidered when designing rural energy projects. When de-
ciding on their energy options, households aim to mini-
mize risk by considering reliability of supply together with
the cost for the various options. Other variables include
substitutability of energy sources, degree of convenience,
level of safety, and pollution emitted into the home as-
sociated with the various energy sources. These insights should be taken into consideration
when designing policies or programs. For poor households, initial investment or start-up cost
is certainly one of the most critical obstacles when moving up the energy ladder. Policy and
project designs must focus on addressing these barriers if pro-poor energy interventions are
to be successful. While a multi-pathway energy ladder model is certainly a valid concept for
long-range planning, the model also indicates that energy utility or service might be more
important to consumers than previously thought.
Rural fuel traders
As income levels are very difficult to examine, the parameter expenditure was used to compare
the different social classes and evidently shows that Kampong Speu is a richer province (in
monetary terms) than Svay Rieng where households on average spend only US$47 per month
per family of five to six members. This is significantly lower than the poverty level, which is as-
sumed to be US$1 per day per person. A strategic response to this situation is promoting the
establishment of woodlots. This has the potential to generate income and alleviate serious
UNDP Cambodia 23
TITLE
Traditional biomass shortages at the same time. Fast growing species could substantially increase income
cooking methods levels of poor rural families.
are inefficient
and unhealthy The second priority should be the improvement of the hygienic situation in kitchens. The
popular ‘three stone stoves’ are very inefficient and generate irritating and hazardous smoke
that mainly affects woman and children. Improved cook stoves can save up to 50 percent of
fuel wood compared to cooking with the ‘three stone stove’.
Water filters also improve the hygienic situation in kitchens and save fuel wood at the same
time. Water filters costs US$7.50, which is the equivalent of about three months fuel wood
supply used for boiling drinking water. Other energy efficiency technologies are best pro-
moted through increases in household incomes. The study results suggest that once families
are able to spend more money they will instantly start an energy transition towards LGP gas
for cooking and batteries for lighting.
24 UNDP Cambodia
3. Future demand in surveyed provinces With a total of

41,000 Giga
Joules (GJ) in
3.1 Forecasting methodology Kampong Speu,
these rural
In this section, the energy demand is forecasted for the provinces of Kampong Speu and Svay electricity
Rieng. This forecast forms the basis of a nationwide energy demand/supply model described enterprises
in Section 3.5. In the demand forecast the distinction between cooking and lighting has been
supply about 70
maintained. The forecast is based on probable trends derived from historical data from NIS.
times more
The forecast rests on the assumption that provincial urban areas will follow the trend of Phnom
Penh and rural areas the trend of the provincial urban areas. The final energy demand is calcu- energy to their
lated with LEAP software that included GDP figures, population and income growth. customers than
REEs in Svay
The forecast presented here can be considered as the baseline or business as usual (BAU) sce- Rieng
nario. It does not consider increasing scarcity of resources or price increases for energy ser-
vices. The BAU also does not consider the reasons why a particular household lacks access to
a modern fuel or electricity. It takes into account neither supply-side failures nor demand-side
failures, the latter being the case when a household chooses not to use the available service
because it is too expensive, culturally unfamiliar, or otherwise inappropriate. Because the
policy implications of these supply- and demand-side failures vary significantly, more data to
differentiate between the two situations would be necessary.
Table 16 depicts the basis from which energy forecasting has been performed in the frame-
work of this study. In terms of energy units, it is very clear that cooking energy is much more
important than lighting and appliance use. The base table also shows the relative importance
of REEs in Kampong Speu. With a total of 41,000 Giga Joules (GJ) in Kampong Speu, these rural
electricity enterprises supply more than eight times the energy to their customers than REEs
in Svay Rieng.
Table 16 Total energy use in Kampong Speu and Svay Rieng (GJ/year)
Cooking Lighting/Appliances
Source Kampong Speu Svay Rieng Source Kampong Speu Svay Rieng
Firewood purchased 1,412,631 406,395 EDC 9,671 13,059
Firewood collected 2,231,390 2,617,870 REE 41,025 595
Palm tree branch 415,945 254,909 Own Generator 1,291 255
Charcoal 449,509 30,167 Battery lighting 4,610 15,286
LPG 115,813 24,680 Battery other 16,157
Electric rice cooker 1,321 140 Kerosene 80,486 60,951
Animal Dung 17,137 19,262 Candles 9,945 3,347
Total 4,643,745 3,353,423 163,184 93,495
3.2 Demand and supply modelling using historical data
In the framework of the research undertaken, MIME has developed an ‘Energy Demand Fore-
casting Model’ (EDFM). Such a customized model seemed necessary in order to overcome lim-
itations of existing tools such as LEAP. LEAP uses variables such as GDP, population growth and
income growth for demand forecasting, but does not allow the inclusion of parameters such
as behavioural change in order to balance supply and demand. Thus, EDFM was developed
to simulate a dynamic change in the energy mix by including fuel switches as a response to
UNDP Cambodia 25
TITLE
income growth. EDFM merges existing historical demand data with the correlation between
income and the preference for a type of energy source.
The EDFM uses forecasting results of LEAP where GDP, income and population growth param-
eters can be included. EDFM then simulates the shift in energy supply sources in the future.
The results are highly relevant for policy responses and the design of interventions aimed to
ensure sustainability of energy supplies. It should be noted, however, that the model is still
under trial and has its own limitations. It is, as is any other forecasting tool, based on certain as-
sumptions and forecasting and does not include all drivers relevant for the balance of supply
and demand. It does not, for instance, include increasing scarcity and price elasticity changes
over time. The model is, however, flexible enough to consider some of these parameters in a
scenario analysis. The distinguishing characteristic of EDFM as a specialized modelling tool is a
greater depth of energy planning from the perspective of energy policy analysts. Analysts can
use such information to quantify the benefits of electric lighting for rural households, using
methods involving available cash and willingness to pay.
EDFM simulation rests on the following assumptions:

• The income level of people living in rural areas will gradually increase over the years;
• Income level directly influences a household’s choices regarding the use of an energy
source;
• The future shift between energy sources in the provincial urban areas will follow the
historical shift between energy sources in Phnom Penh;
• The shift between energy supply in the rural areas will follow the historical shift be-
tween energy sources of provincial urban areas;
• The energy transition from now to 2030 will follow the same progression that can be
observed in the NIS income quintiles from poor to rich.
NIS regularly conducts a Socio-Economic Survey including questions about energy use. Data
is available from the years 1993, 1996, 1998, 2000 and 2003/4. NIS analyzed existing data for
the purpose of this study in order to establish historical trends for Phnom Penh and all provinc-
es. Cooking energy and energy sources for lighting and other appliances were distinguished
in the trend analysis. The NIS data also contains information on the use of different energy
sources per income category for Phnom Penh, provincial urban areas and rural areas. The
population is divided into five income categories or income quintiles as shown in Table 17.
Table 17 NIS income quintiles

Quintile Average income per household per month (riel) Average income per month (dollar)
1 264,677 64.56
2 570,484 139.14
3 978,959 238.77
4 1,810,044 441.47
5 8,128,360 1,982.53
EDFM probably provides the most accurate energy model for rural Cambodia at the moment.
However, the trends derived from historical data are based on data that is not sophisticated
enough and may cause distortions. These distortions are assumed to cause deviations on the
order of 15 percent in 2015 and of 25 percent in the year 2030. Further refinement of data will
allow increasing accuracy of the model.
26 UNDP Cambodia
Kampong Speu province
Information derived from the historical trend in the urban cooking energy mix in Kampong
Speu shows that fuel wood still is the main energy source. Its importance started to decrease
after 1999 together with an increase in the use of charcoal and LPG. Historical shifts have oc-
curred in the energy mix for lighting and other appliances. The role of kerosene has declined
dramatically since 1997 and privately provided electricity has increased along with EDC power
and the use of batteries.
The modelling results for the urban energy mix in Kampong Speu are shown in Figure 8. Sig-
nificant changes are expected for both cooking and lighting. The decrease of wood and later
charcoal as cooking fuels is mirrored by an increase in LPG use. At the same time, kerosene
is rapidly decreasing and batteries are pushed out and substituted by either EDC or privately
generated electricity. It is interesting that urban Kampong Speu shows an energy mix that
deviates significantly from Phnom Penh with fuel wood retaining an important role for poor
families, but also for families in higher income groups. The use of charcoal will rapidly increase
in the near future (40 percent in 2010), but will steadily decline after 2013, reaching 30 percent
in 2030. LPG is hardly used at the moment, but will show a constant increase until 2030, mak-
ing up 50 percent of the energy mix in 2030.
Figure 8 Urban energy mix projection Kampong Speu
Changes in the lighting and appliances mix follow a more complex pattern. The use of kero-
sene seems to be fading. In 1997 the share was 100 percent, but in 1993 it was only 30 percent.
The total use of kerosene will drop in the near future stabilizing at 5 percent for ten years and
then drop to nil by 2030. For all income groups, batteries remain an important energy source.
For the two lowest income categories, the share is between 30 to 40 percent, for income cat-
egories three and four the share is between 40 and 70 percent. Only for the highest income
category is the use of batteries decreasing. The model predicts a significant fluctuation in the
UNDP Cambodia 27
TITLE
share of privately provided electricity that can be explained by the transitional nature of self-
generation. It becomes obsolete the moment grid power is provided. In 2030 it is predicted
that all income groups will use publicly provided electricity, although its share in the energy
mix is smaller in lower income categories than it is in higher income categories.
The predicted development of the rural energy mix in Kampong Speu is depicted in Figure 9.
The dependence on fuel wood is more prominent than in urban areas showing that 100 per-
cent of the households use this fuel in 2007 irrespective of income categories. However, wood
shows a rapid decline in the distant future, to be used by only 35 percent of the households in
2030. Similar to urban areas, LPG becomes more important and 15 percent of the households
are expected to use it by 2030. Lighting energy, which is currently still dominated by kerosene,
will gradually be replaced by battery power and later by an increasing share of grid electricity.
A moderate 10 percent is expected by 2020 and beyond. Electricity use will increase from just
2 percent now to about 20 percent in 2025. Thereafter, penetration will increase more rapidly,
reaching 50 percent in 2030. This penetration of electricity is responsible for the phasing out
of battery use, dropping from 50 percent in 2025 to less than 10 percent in 2030.
Figure 9 Rural energy mix projection Kampong Speu
Svay Rieng province
For Svay Rieng a picture emerges when supply and demand is forecast using historical trends.
As an example, Figure 10 shows the development of the rural energy mix for lighting and ap-
pliances. Again, a rapid transition from a kerosene-dominated household sector to publicly
and privately provided electricity characterizes the results based on historical trend.
28 UNDP Cambodia
Figure 10 Rural energy mix projection Svay Rieng
3.3 Refined modelling
In order to eliminate possible distortions that are inevitable when using historical trends, this
section presents filtered data that not only consider the historical trends derived from NIS
data, but also the results of the questionnaire-based household survey conducted in 2007.
The forecast models are divided into ‘rural’ and ‘urban’, and within those categories a further
distinction has been made between energy sources used for ‘cooking’ and ‘lighting and other
appliances’.
Charcoal in Kampong Speu ready to be transported to Phnom Penh markets.
Kampong Speu
Figure 11 shows the modelling of cooking energy for urban and rural Kampong Speu. In the
urban sector LPG clearly is the predominant fuel with 68 percent of the households using
it now, increasing to about 85 percent in 2030. LPG gradually replaces wood and charcoal,
which will be used by only 15 percent and 35 percent of the households respectively. Elec-
UNDP Cambodia 29
TITLE
tricity is used by 60 percent of the households and will remain at this level until 2030. Rural
cooking in Kampong Speu will follow a similar path with a sharper drop in the use of firewood
during the next eight years. LPG becomes a cooking fuel for 50 percent of the households in
2030. Surprisingly, the modelling of charcoal consumption shows that rural households in
Kampong Speu do not exactly follow the trend of urban areas (i.e., charcoal will never become
a predominant fuel as households transition directly to LPG).
Figure 11 Energy use in Kampong Speu
Svay Rieng
Compared to the results for Kampong Speu, the energy transition in Svay Rieng follows a sim-
ilar, while more biomass-dominated, path. Fuel wood remains the main energy source for
cooking for both urban and rural households. LPG substitutes wood and charcoal at a much
slower rate than in Kampong Speu, reaching 65 percent penetration in urban and 25 percent
in rural Svay Rieng. According to the modelling, 60 percent of urban households will still use
wood in 2030, as opposed to 45 percent in rural areas.
The model also produces interesting results with respect to the energy use for lighting and
appliances in rural Svay Rieng. Accordingly, batteries retain an unchanged penetration rate of
80 percent over the entire projection timeframe, while kerosene will only used by 50 percent
of the households in 2030. What is more striking is the electricity use model. While one would
normally expect a steady rise in the use of publicly-provided electricity as a result of govern-
ment driven electrification programs, electricity use fluctuates and shows a significant drop
after 2017.
30 UNDP Cambodia
Figure 12 Energy use Svay Rieng
3.4 Demand forecasts
Cooking
Figure 13 shows the forecasted urban energy use for cooking in both provinces. Accordingly
the mix of energy sources used will change over time. Charcoal and LPG will become the
main energy sources, while firewood first initially increases, but declines towards the end of
the forecasting period. This decline starts earlier in Kampong Speu than in Svay Rieng. Even
though the use of firewood as a direct cooking energy source is decreasing, a decrease in the
total wood demand for charcoal and firewood is not expected. Without interventions, wood
demand will actually increase dramatically as 6 kg of wood are needed to produce 1 kg of char-
coal. Even when considering the higher heating value of charcoal, traditional charcoal-making
UNDP Cambodia 31
TITLE
technologies involve very low energy efficiencies. Up to 80 percent of the energy content of
the wood used for charcoal burning in earth mound kilns is lost in the process. This trend is a
serious concern as forest resources are already under serious pressure and modern fuels such
as LPG will not penetrate the urban markets fast enough to avoid deterioration of biomass
resources.
Figure 13 Total urban energy use cooking (GJ)
Figure 14 shows that total energy use, expressed in energy content or Giga Joules (GJ), in rural
Kampong Speu for cooking will decline by about half from 2007 to 2030. The use of fuel wood
is expected to decline dramatically. Similar to the forecasted development in urban areas,
charcoal use will increase, particularly in Kampong Speu (i.e., even in rural areas the pressure
on woody biomass will increase). LPG demand also shows a five-fold increase from 2007 to
2030. In Svay Rieng a decline in total energy use is noticeable, but it is less dramatic than in
Kampong Speu. Firewood remains the main energy source.
Phnom Penh baseline
The EDFM simulation assumes that provincial urban areas will follow the energy demand/sup-
ply trend of Phnom Penh where some dramatic changes in energy use have occurred over the
last ten years. The use of fuel wood for cooking has decreased rapidly in ten years time with
households shifting to LPG. Forty percent of lower income households use LPG, rising steadily
with income to approximately 70 percent in the highest income category. However, fuel wood
is still a substantial energy source for cooking, especially for low-income households. Charcoal
is also still an important cooking fuel in Phnom Penh with LPG apparently also substituting for
charcoal. Contrary to the use of firewood, the share of charcoal seems to be comparable for all
families regardless of their income. Only for the highest income households does charcoal not
seem to play a role as a fuel.
Grid electricity and privately generated electricity (85 percent and 10 percent respectively) are
almost the sole energy sources for lighting appliances. Households in lower income categories
are more likely to use privately generated electricity, indicating that the EDC grid does not yet
serve some low cost urban areas in Phnom Penh. Kerosene, an energy source that was important
in 1993 (20 percent share), dropped to nil ten years later, indicating a high consumer preference
for safe and clean electricity.
32 UNDP Cambodia
Figure 14 Total rural energy use for cooking (GJ)
Lighting and appliances
With increasing standards of living, total urban energy use for lighting and other appliances
will increase sharply in both provinces (Figure 15). Kampong Speu is expected to experience
a significant expansion of EDC services. Despite this expansion, REE will remain the predomi-
nant supplier of electricity. In urban Svay Rieng EDC will remain the most important supplier
of energy. This is due to the fact that households have access to relatively cheap electricity
imported from Vietnam. It is impossible for diesel-based REEs to compete with cross border
supplies. However, a change in policy could give REE a more prominent role in Svay Rieng.
Instead of generating power using expensive diesel and low efficiency equipment, REEs could
become bulk buyers of cross-border electricity supply and retail the power in local areas.
Figure 15 Urban energy use for lighting and other appliances (GJ)
UNDP Cambodia 33
TITLE
Figure 16 Rural energy use for lighting and other appliances
Figure 16 displays the trends in rural areas. In Kampong Speu, a sharp increase in demand for
EDC and REE electricity is expected. With declining battery and kerosene use, electricity will
become the main energy source by 2030. Villages along the main roads are expected to be-
come urbanized centres with dynamic transitions to a more urban energy use pattern. While
other energy sources (batteries, kerosene and candles) will still be used, their contribution to
the average energy mix will constantly fall over the forecasting period. Nevertheless, the num-
ber of households connected to grid electricity will remain low. In quantitative terms, batter-
ies and kerosene are low intensity energy supplies (i.e., the total energy use figures might be a
weak indicator for actual improvements in the energy mix of the majority of the population).
Thus, total household energy use data is insufficient for extensive energy policy analysis.
Household access to modern energy
A better measure for the actual impact of energy transitions on the population is an analy-
sis of accessibility (i.e., a forecasting of household numbers that use certain types of energy).
The results of this analysis are depicted in Figure 17. Accordingly, cooking habits will change
rapidly during the projection timeframe. In Kampong Speu, for instance, urban use of LPG
will increase rapidly from 5,000 households in 2007 to 41,000 households in 2030. The use of
charcoal and of electric rice cookers will also increase, albeit at a slower rate. Wood use will
increase slowly until 2020 and decline slowly thereafter. Svay Rieng, the more rural province,
will experience a similar transition, but with a time lag as the number of urban households us-
ing LPG will initially increase gradually, then rapidly after 2026. In contrast to Kampong Speu,
wood will remain the most important energy source in Svay Rieng until 2030.
The rural cooking energy mix shows a different transition path. In both provinces the number
of rural households using firewood will decrease gradually over the entire projection time-
frame. While LPG will penetrate rural areas rapidly, the use of electricity for cooking will re-
main insignificant, indicating limited access to grid electricity in rural areas. In 2030 most rural
households in Kampong Speu will exhibit an almost urban energy mix, while Svay Rieng will
still have a fuel wood dominated cooking energy mix. Approximately 54,400 households will
still be using wood in 2030, with charcoal and LPG following with 47,500 and 38,700 respec-
tively.
34 UNDP Cambodia
Figure 17 Household sources of energy for cooking Cross border

supplies from
Vietnam to
improve
electricity supply
in Svay Rieng
The forecasted energy mix for lighting and appliances show similar trends towards moderniza-
tion. Access to EDC or REE grid power in urban Kampong Speu will gradually increase from
2,900 EDC supplied households in 2007 to 32,100 households in 2030, and for REE supply from
5,800 households in 2007 to 32,100 in 2030. In urban Svay Rieng a rapid increase of the num-
ber of households using EDC is expected due to the expansion of cross-border supply from
Vietnam.
Figure 18 Rural transition in Kampong Speu
For rural Kampong Speu the forecasting yields some interesting results, as depicted in Figure
18: from a high number of households using batteries in 2007, battery use will increase to a
peak of 146,100 households in 2022. Thereafter battery use will drop due to a substitution
for battery power by EDC and REE. However, towards the end of the projection timeframe
the importance of REE starts to decline due to the availability of cheaper and EDC power. The
UNDP Cambodia 35
TITLE
Despite the shift number of households using candles remains more or less constant. Kerosene use shows a
to modern steep drop until 2016 and then gradually decreases.
energy, demand
for wood fuels,
kerosene and 3.5 Extrapolation to national level
batteries
Based on the energy forecasting for Kampong Spey and Svay Rieng, an indicative extrapola-
remains strong
tion can be simulated for Cambodia’s total energy demand. In order to show a band in which
demand will develop, simulations were calculated based on Kampong Speu and on the Svay
Rieng data. Table 18 shows the results for the simulation based on Kampong Speu, a province
with a more urban characteristic and a high use of electricity, charcoal and LPG. The Svay
Rieng based simulation is displayed in Table 19. It shows a more rural energy use pattern with
a slower penetration of LPG and electricity and a significantly higher demand for fuel wood
and kerosene.
Table 18 Total residential energy use in Cambodia based on Kampong Speu

Year 2007 2010 2012 2015 2017 2020 2022 2025 2027 2030
Total electricity demand (GWh) 750 932 1,098 1,217 1,296 1,472 1,613 1,887 2,026 2,222
Total charcoal demand (thousand tonne) 321 424 535 801 807 770 756 729 719 724
Total firewood demand (thousand tonne) 4,696 4,013 3,400 2,236 2,099 1,993 1,880 1,508 1,286 973
Total LPG demand (thousand tonne) 61 77 83 96 110 147 180 232 255 283
Total kerosene demand (thousand tonne) 26 24 16 11 10 9 8 10 10 9
Total candle demand (tonne) 5 5 4 5 5 5 5 5 6 6
Total animal dung (thousand tonne) 25 28 28 27 28 29 29 30 31 32
Table 19 Total residential energy use in Cambodia based on Svay Rieng

Year 2007 2010 2012 2015 2017 2020 2022 2025 2027 2030
Total electricity demand (GWh) 279 303 327 374 478 503 532 569 602 663
Total charcoal demand (thousand tonne) 42 58 70 71 73 90 103 119 138 198
Total firewood demand (thousand tonne) 4,889 4,825 4,767 4,872 4,916 4,476 4,092 4,004 3,940 3,639
Total LPG demand (thousand tonne) 22 24 26 40 45 55 60 70 80 103
Total kerosene demand (thousand tonne) 24 29 27 26 22 21 22 23 24 23
Total candle demand (tonne) 2 3 3 3 3 3 4 4 4 4
Total animal dung (thousand tonne) 87 643 940 468 261 270 449 631 414 145
The reality of Cambodia’s energy balance will probably be between the two scenarios devel-
oped with data from Kampong Speu and Svay Rieng. Although there is a massive shift to-
wards modern energy sources in both scenarios, the demand for firewood, charcoal, kerosene
and batteries will remain high over the next decades.
Simulated energy transitions are shown in Figure 19, which displays the number of house-
holds using particular types of energy.
36 UNDP Cambodia
Figure 19 Projection of households using different sources of energy
Again, a more urban style transition is represented by the Kampong Speu based model, while
the model using Svay Rieng data as a basis exhibits the characteristics of a more rural transi-
tion. In any of the scenarios, households use a variety of energy sources for cooking, lighting
and other purposes. The Kampong Speu based forecast shows a significant decrease in fuel
wood use together with an increased use of charcoal. Although EDC and REEs are expected to
connect about four million households by 2030, the use of batteries will remain high. Charcoal
demand will also increase. This will put additional pressure on biomass resources as current
conversion technologies require six kg of wood to produce one kg of charcoal. Under this sce-
nario, interventions to ensure a sustainable supply of fuel wood become critically important. It
should be noted that not only the residential fuel wood market needs to be supplied, but that
small industrial consumers such as palm sugar manufacturers and distilleries rely on biomass
fuel as well. Additional pressure on forests is expected from demand for export timber. With-
out massive efforts to boost wood production, the search for energy will remain a daily grind
for rural Cambodians. In many remote locations these people have neither electricity nor LPG
to cook their food. Women and children from these families will continue to spend hours each
day gathering low grade fuels – time they might otherwise be able to spend on productive
work or education.
Table 20 displays the projected rural energy demand for Cambodia in TJ. The replacement of
inefficient energy sources results in a continuous drop in the aggregated demand from 84,345
TJ in 2007 to 64,679 in 2030. However, this transition will not occur automatically, at least not
in a smooth manner. A smooth shift from environmentally unsustainable and unhealthy en-
ergy sources to modern fuels will have to be managed and guided by an appropriate policy
framework. Such a policy framework needs to acknowledge the needs of the poor and care-
fully balance social objectives with the economic need to price fuels at their supply cost.
Table 20 Estimated rural energy demand in Cambodia (TJ)

2007 2010 2012 2015 2017 2020 2022 2025 2027 2030
Total electricity 1,787 2,155 2,493 2,790 3,118 3,476 3,777 4,328 4,636 5,095
Total charcoal 5,107 6,796 8,532 12,323 12,441 12,156 12,153 11,998 12,137 13,112
Total firewood 73,637 67,920 62,785 54,742 54,038 49,749 45,866 42,355 40,178 35,522
Total LPG 1,885 2,313 2,512 3,135 3,591 4,671 5,570 7,020 7,763 8,964
Total kerosene 1,108 1,166 978 843 735 673 674 731 767 717
Total candle 139 144 140 148 154 164 171 186 194 203
Total animal dung 683 4,117 5,945 3,030 1,763 1,821 2,915 4,026 2,703 1,068
Total 84345.5 84610.5 83383 77009 75838.5 72707 71125 70642 68376.5 64679.5
UNDP Cambodia 37
TITLE
Energy sector 4. Lessons learned

development has
favored urban
areas 4.1 Rural energy supply in Cambodia
The empirical data gathered in the framework of the project and its subsequent use for en-
ergy modelling provides the groundwork for both further research and project planning. The
results show that households are using a mix of different sources for cooking like firewood,
charcoal and LPG, and for lighting are using kerosene and car batteries and sometimes also
grid electricity, depending on availability and income. The analysis has also shown that the
development of the energy sector has strongly favoured urban areas, leaving rural areas be-
hind. Phnom Penh in particular has a modern energy sector with most people connected to
affordable grid electricity and a high percentage of LPG use for cooking. The provincial urban
areas are following the trend of Phnom Penh, with provincial urban areas growing fast. Rural
areas along major roads also seem to urbanize, particularly in Kampong Speu. Svay Rieng is
a remoter, poorer and more rural province, and traditional fuels dominate its energy sector
characteristics.
In both provinces lower income

households use mainly firewood
and sometimes charcoal for cook-
ing, while higher income families use
mainly LPG. However, even high-
income households still use charcoal
to prepare certain dishes. A similar,
special purpose use for charcoal ex-
ists in rich Western countries where
charcoal is used for barbeques. Poor
households use mostly kerosene for
lighting, while higher income house-
holds use grid electricity if available.
Rural electricity enterprises, like this one in Kampong Speu, are
In the absence of grid power high-in- bringing electricity to areas where grid electricity is unavailable.
come households use small genera-
tors, car batteries or a combination of
both. In Kampong Speu, 60 percent use kerosene and batteries for lighting, 20 percent have
grid electricity. In Svay Rieng, 75 percent use kerosene, 84 percent use batteries and only 6
percent of the households have access to grid electricity.
As local REE grids often only provide electricity during peak hours in the evening, households
often use batteries to ensure availability of electricity beyond grid service hours. The data
analysis revealed that households in the two surveyed provinces use more or less a constant
share of their available cash to procure energy services. This share is approximately 10 percent
of total monthly expenditures for all social classes (for details, refer to Table 15). Clearly, this
shows that increasing household income would be a way to promote a transition to a more
modern rural energy sector in Cambodia.
For many households, however, access is as much of a problem as affordability. Cambodia

has one of the lowest electrification rates in Asia and among the highest electricity costs in
the world. The problem is not necessarily that people are unwilling to pay for clean modern
energy. Results of this survey suggest that households will spend a significant share of their
incomes on clean energy, which in turn improves the quality of their life or enables them to
become more productive. Privately operated REE have been supplying electricity to rural cus-
38 UNDP Cambodia
tomers for years, but many enterprises are challenged by low customer demand and rising Cooking needs
operating costs. REE-type expansion of electricity supply is thus limited to areas where at least are quantitatively
a minimum of demand density can be found. Remote areas that show low population densi- more important
ties are unlikely to be covered by an expansion of private supply. Without access to electricity, than lighting
poor households are denied a host of modern services such as high quality electric lighting
and appliances
and refrigeration. The use of lead acid batteries has been widely adopted in Cambodia to pro-
vide a minimum of power, however, at unit costs that are several times the cost of grid electric-
ity provided by the national utility.
The supply of sustainable cooking energy may, however, be more critical for rural households
than the availability of electricity services for the foreseeable future. In most rural areas fuel
wood extracted from common resources is still the dominant energy source for cooking, al-
though in Kampong Speu, 38 percent of the households already purchase firewood and 24
percent are using charcoal. In Svay Rieng, nearly every household uses wood, while 13 percent
purchase wood and only 6 percent use charcoal. Wood and charcoal will remain the main en-
ergy sources for cooking till 2030. It is expected that without massive interventions, forests will
disappear at alarming rates, especially in Kampong Speu where fuel wood extraction for direct
use and charcoal production has already degraded vast forest areas. While difficult to quantify,
the economic costs of fuel wood extraction and use are considerable. At the same time, tra-
ditional fuels have become an
important rural industry pro-
viding a livelihood for many
families. It is estimated that
in Kampong Speu alone, ap-
proximately 8,000 families are
producing more than 100,000
tons of charcoal per year and
2,700 families are cutting and
selling more than 200,000 tons
of fuel wood. It will be neces-
sary to find innovative ways to
engage those already working
in the rural energy industry to
implement strategies to boost
Improved cook stoves help to reduce the burden of wood collection or the the production of fuel wood
expenditure households have to make for energy purposes. and increase the efficiency of
both conversion and end use
technologies. To promote energy saving technologies will also result in a cleaner and healthier
cooking environment. The high exposure risk to indoor air pollution particularly affects wom-
en and children. The proximity of the pollution source and recipients exacerbates the risks.
The concentration of harmful substances in the air of a traditional kitchen is typically orders
of magnitudes greater for cooking-fire generated pollution than for pollutants produced from
outside sources such as power plants. Energy efficient stoves will also reduce the burden of
wood collection or the expenditure households have to make for energy purchases.
4.2 Next steps
The problems of rural energy in Cambodia should certainly not be dealt with in isolation.
Poverty and dependence on low-grade fuels are closely related, and increases in household
incomes normally trigger a switch to modern fuels if these are accessible. A purely energy
sector scheme for improving rural energy may, therefore, not succeed if economic growth is
UNDP Cambodia 39
TITLE
prevented. In an appropriate framework of policies that promote economic development in

rural areas, one of the most powerful ways to ensure sustainable energy supplies is to guaran-
tee that the energy market is determined by consumers’ choices. Energy prices should reflect
total supply cost including externalities such as environmental degradation and productivity
loss. Obviously, a transition to prices that reflect supply cost cannot be achieved immediately
without hurting the poor. Policy must recognize the need for price adjustments to go hand in
hand with increases in rural households’ incomes. Government policies should also encour-
age competition and consumer choice on a level playing field for different stakeholders in the
energy sector including utilities, private companies, and enterprises set up by the local com-
munities.
Unlike in other developing countries, the rural energy sector in Cambodia has enjoyed quite
a liberal environment in the past with little or no rules and regulations that would have stifled
the emergence of REE, fuel traders or battery charging businesses. Consequently, most ru-
ral energy is either provided by private enterprise or by household members collecting low-
grade fuels. On the other hand, surprisingly little activity has developed in the area of wood
production. Given the fact that wood is still an extremely popular fuel, which is increasingly
commercialized, this can only be explained by the fact that in many parts of the country, for-
ests are considered a free common resource. Even when traded, the price for wood normally
only reflects the cost of harvesting and transportation and does not include resource costs (i.e.
the cost of replanting).
Therefore, a high priority should

be given to three-pronged inter-
ventions that address the lack of
sustainable management and
protection of existing forest re-
sources, the promotion of wood
production schemes, and the
dissemination of more efficient
technology simultaneously.
Integrated, large-scale wood
production programs would
not only reduce the pressure
on natural forest resources, but
they would also create employ-
ment and income. Sustainable Leucaena Cephalous trees, a fast growing species, are being used in ru-
wood production would allow ral electricity generation.
charcoal production to continue
and the introduction of gasification technology to generate electricity. A sustainable wood
energy supply that also addresses land management and ownership issues would produce
significant benefits for rural communities. Community forestry laws are already in place and
communities need assistance to develop management plans for community forestry projects.
Such a program has the potential to be cost efficient and could be replicated throughout the
country.
The problems of rural energy in Cambodia have long been recognized and the results of the
research performed for this study reconfirm the need for quick and coordinated action. While
UNDP can play an important catalytic role, real progress in tackling the threat of a rural energy
crisis is critically linked to the implementation of policy and institutional changes, mobilization
of local private sector, donor and NGO resources, and the development of innovative respons-
es to the energy needs of Cambodia’s rural population.
40 UNDP Cambodia
Tables
Table 1 Household and sample size in selected provinces.................................................................7
Table 2 Key features of the surveyed provinces......................................................................................9
Table 3 Sample size social classes.............................................................................................................. 11
Table 4 Energy sources used for cooking in urban Kampong Speu and Svay Rieng.............. 13
Table 5 Energy sources used for cooking in rural Kampong Speu and Svay Rieng................. 14
Table 6 Energy sources used for lighting and other appliances in urban areas....................... 15
Table 7 Energy sources used for lighting and other appliances in rural areas.......................... 16
Table 8 Monthly expenditure per house type....................................................................................... 17
Table 9 Average cost per unit of energy.................................................................................................. 17
Table 10 Average hours spent collecting firewood per social class................................................ 18
Table 11 Kerosene use per social class per month................................................................................. 20
Table 12 Cumulative average consumption figures and user percentages
Kampong Speu................................................................................................................................. 20
Table 13 Cumulative average consumption figures and user percentages
Svay Rieng........................................................................................................................................... 21
Table 14 Average expenditure per month per energy type per social class (US$)..................... 21
Table 15 Share of energy costs of total monthly expenditure per social class............................ 22
Table 16 Total energy use in Kampong Speu and Svay Rieng (GJ/year)........................................ 25
Table 17 NIS income quintiles....................................................................................................................... 26
Table 18 Total residential energy use in Cambodia based on Kampong Speu........................... 36
Table 19 Total residential energy use in Cambodia based on Svay Rieng..................................... 36
Table 20 Estimated rural energy demand in Cambodia (TJ).............................................................. 37
Figures
Figure 1 Map of Cambodia showing the two pilot provinces........................................................... 10

Figure 2 House types........................................................................................................................................ 11
Figure 3 Social classes in Kampong Speu and Svay Rieng................................................................. 12
Figure 4 Share of households using particular energy sources........................................................ 13
Figure 5 Monthly energy expenditure by social class.......................................................................... 16
Figure 6 Percentage of responsible household members collecting firewood
per social class................................................................................................................................... 19
Figure 7 Energy transition.............................................................................................................................. 23
Figure 8 Urban energy mix projection Kampong Speu....................................................................... 27
Figure 9 Rural energy mix projection Kampong Speu......................................................................... 28
Figure 10 Rural energy mix projection Svay Rieng.................................................................................. 29
Figure 11 Energy use in Kampong Speu..................................................................................................... 30
Figure 12 Energy use Svay Rieng.................................................................................................................... 31
Figure 13 Total urban energy use cooking (GJ) ....................................................................................... 32
Figure 14 Total rural energy use for cooking (GJ)..................................................................................... 33
Figure 15 Urban energy use for lighting and other appliances (GJ)................................................. 33
Figure 16 Rural energy use for lighting and other appliances............................................................ 34
Figure 17 Household sources of energy for cooking.............................................................................. 35
Figure 18 Rural transition in Kampong Speu............................................................................................. 35
Figure 19 Projection of households using different sources of energy............................................ 36
UNDP Cambodia 41
TITLE
Cambodia
United Nations Development Programme
No. 53, Pastuer Street
P.O. Box 877
Phnom Penh
Cambodia
March 2008
www.un.org.kh/undp

UNDP Rural Energy 2008 Eng

Uploaded by

Copyright:

Available Formats

UNDP Rural Energy 2008 Eng

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

UNDP Rural Energy 2008 Eng

Uploaded by

Copyright:

Available Formats

Residential energy demand

© 2008, UNDP Cambodia

2 Survey results . ........................................................................................................................................ 9

3 Future demand in surveyed provinces . ..................................................................................... 25

4 Lessons learned .................................................................................................................................. 38

UNDP Cambodia iii

ACRONYMS AND ABBREVIATIONS

H.E. Tun Lean

1.2 Energy policy framework of the RGC

1. to provide an adequate supply of energy throughout Cambodia at reasonable and afford-

1.3 UNDP’s country strategy and rationale for involvement in energy

1.5 Approach and methodologies

Table 1 Household and sample size in selected provinces

2.1 The provinces of Kampong Speu and Svay Rieng

Administratively, Cambodia is divided into 24 Provinces (including four municipalities). At the

Table 2 Key features of the surveyed provinces

Figure 1 Map of Cambodia showing the two pilot provinces

2.2 Social classes

Figure 2 House types

Table 3 Sample size social classes

Figure 3 Social classes in Kampong Speu and Svay Rieng

2.3 Energy used for cooking

Cooking in urban areas

Figure 4 Share of households using particular energy sources

Cooking in rural areas

2.4 Energy for lighting and appliances

Rural lighting in Kampong Speu and Svay Rieng

In the absence of grid supply, rural households in

Table 7 illustrates that higher electrification rates

The results also demonstrate the importance of

2.5 Energy use by social class

Figure 5 Monthly energy expenditure by social class

Table 8 Monthly expenditure per house type

Table 9 Average cost per unit of energy

Cooking energy and social class

Table 10 Average hours spent collecting firewood per social class

The demand on children to collect

Figure 6 Percentage of responsible household members collecting firewood per social

Rural middle class housing

Lighting and appliance energy and social class

Table 12 Cumulative average consumption figures and user percentages from

Social class Kampong Speu

2.6 Supply cost and affordability

2.7 Energy Transition

This is an important insight and should be seriously con-

Rural fuel traders

3. Future demand in surveyed provinces With a total of

3.2 Demand and supply modelling using historical data

EDFM simulation rests on the following assumptions:

Table 17 NIS income quintiles

Kampong Speu province

Figure 8 Urban energy mix projection Kampong Speu

Figure 9 Rural energy mix projection Kampong Speu

Svay Rieng province

Figure 10 Rural energy mix projection Svay Rieng

3.3 Refined modelling

Charcoal in Kampong Speu ready to be transported to Phnom Penh markets.

Figure 11 Energy use in Kampong Speu

Figure 12 Energy use Svay Rieng