Document Sample
MDP_EnergyPaper_Sept1_NoFigures Powered By Docstoc
					                                      Energy services for the poor
                   (Commissioned paper for the Millennium Project Task Force 1)

                                          Draft August 31, 2004

                                  Prepared by Vijay Modi, Professor
                     Earth Institute and Department of Mechanical Engineering
                                         Columbia University
                                          Ph: 212 854 2956
                                          Fax 212 854 7306

    (thanks to many for fruitful discussions: Guido Schmidt-Traub (Millennium Development Project
     Secretariat), Edwin Adkins (Columbia), Pepukaye Bardouille (McKinsey), Antony Bugg-Levine
   (McKinsey), Klaus Lackner (Columbia, EEE), Sebastian Morris (IIMA), David Nissen (Columbia,
SIPA/CEMTPP), Pedro Sanchez (Columbia, EI), Susan McDade (UNDP, BDP), Minoru Takada (UNDP,
  BDP), Macartan Humphreys (Columbia, Ctr Glob. & Dev., involved in MDG task force 1), Dato Lee
Yee-Cheong (Task Force 10), Albert Wright (TF Water), Marco Quinones (Sasakawa Foundation), Robin
Sears (CERC), Chandrika Bahadur (MP Secretariat), Marc Levy (CIESIN), Deborah Balk (CIESIN), and
 students who helped prepare figures: Sean White (Columbia, ME), Alex Whitworth (Columbia, EEE.),
                     Eric Slosberg (Columbia, ME), Vinita Venkatesh (Columbia, ME)

At the outset the author wishes to clarify that this report specifically addresses the role of energy in the
poorest of the poor countries. In many of these countries, particularly in rural areas, a substantial fraction
of the population relies on biomass or dung for cooking fuel and heat; on kerosene wick lamps, batteries
or candles for lighting; and on human or animal energy for motive power (e.g. for tilling/weeding of land,
agro-processing, or transport). In considering the question of whether 2 billion people can transition from
solid fuels to cleaner burning fuels, it is worth noting that Brazil raised the availability of modern cooking
fuels such as LPG from 16% in 1960 to 78% in 1985. Similarly, while about 1.6 billion are without
electricity access in the world, Tunisia’s electrification program managed to expand services from 6% of
the population in 1976 to 88% in 2001, including bringing electricity to 35% of the people living in rural
areas by 2001. So while the idea of reducing the number of people who are ―energy poor‖ by 50% by
2015 appears to be a daunting proposition, there are examples that show that it can be done.

There is no specific target for energy services—such as a specified quantity of modern cooking fuel, or
lux or lumen levels for lighting—in the framework of the Millennium Development Goals (simply called
―MDGs‖ hereafter). And although there are three indicators that are energy related, they do not address
the issue of access to energy services. Yet, modern energy services are an important enabler for many of
the goals. They can:
      Reduce health risks by replacing solid fuels for cooking with modern cooking fuels;
      Save time and effort (e.g. LPG saves time and effort otherwise spent searching for biomass and
         other fuel);
      Lead to more efficient use of fossil fuel (electric lighting is nearly one to two orders of magnitude
         more efficient than light from a kerosene wick-lamp, even when the electricity is produced from
         an inefficient diesel generator operating at part-load);
      Provide significantly improved quality of service (e.g. modern fuels and electric lighting are both
         more efficient and effective than biomass and wick lamps, respectively);
       Provide opportunities for productive and income generating activities. This can be from SME
        (e.g. an engine or an electric motor can enable one to start a business around a maize grinding
        mill) to large enterprises.
       Result in more productive use of land for agriculture (e.g. use of manure and crop residues for
        cooking can deprive land of otherwise valuable soil nutrient enhancement)
       Possibly reduce local land degradation (e.g. use of motive power for tilling can lower the
        dependence on draft power based on free grazing livestock),
       Save money by simply providing a lower unit cost option.

This report illustrates some of the linkages between energy and the MDGs, goals that include targets on
reduction of poverty, income, health, education and gender equality. The role of urbanization, geography
and technology in meeting the energy needs of a significant proportion of those that lack such services is
addressed. Some approximate cost estimates are provided. The role of place-specific analysis and
solutions is highlighted. Finally some policy recommendations are provided.

                                                                             Table of Contents
INTRODUCTION .......................................................................................................................................................4
LINKAGES WITH THE MILLENNIUM DEVELOPMENT GOALS ............................................................... 15
    ROLE OF WOMEN (TIME POVERTY, WATER, FIREWOOD, TRANSPORT, HEALTH) ........................................................ 16
    HUNGER .................................................................................................................................................................. 19
    EDUCATION ............................................................................................................................................................. 19
    HEALTH ................................................................................................................................................................... 20
    HIV/AIDS IMPACTING LABOR SHORTAGE ............................................................................................................... 21
    ENERGY AND ENSURING ENVIRONMENTAL SUSTAINABILITY ................................................................................... 22
    URBAN AREAS: NUCLEI OF GROWTH, PERI-URBAN AND SLUM DWELLERS ............................................................... 24
SCALE-UP ................................................................................................................................................................. 26
MDG-BASED POVERTY REDUCTION STRATEGIES ......................................................................................... 26
    MODERN COOKING FUELS ....................................................................................................................................... 28
             Box 1: Liquefied petroleum gas subsidies in Brazil: 1973-2003 .......................................................................... 29
             Box 2: Deepam Scheme for encouraging LPG use in Andhra Pradesh, India ................................................... 30
    ELECTRICITY ........................................................................................................................................................... 32
    ELECTRIFICATION FOR RURAL AREAS: IMPORTANCE OF AGGREGATION OF DEMAND ............................................... 32
             High density dispersed settlements ................................................................................................................................. 33
             Box 3: Electricity Services in Urambo Village, Tanzania ....................................................................................... 37
             High density nucleated settlements ................................................................................................................................. 38
             Low Density (both nucleated and dispersed) Settlements ............................................................................................... 38
        Urban Areas ....................................................................................................................................................... 40
        Some Examples of Scale-up and Lowered unit costs.......................................................................................... 41
             South Africa................................................................................................................................................................... 41
             Tunisia ........................................................................................................................................................................... 41
        Lowering Unit Costs .......................................................................................................................................... 42
             Box 4: Diesel Powered Multifunctional Platform in Mali......................................................................................... 43
    TRANSPORT ............................................................................................................................................................. 44
      Other cross-sectoral benefits of improved transport.......................................................................................... 47
    EDUCATION, TRAINING AND RESEARCH CAPACITY ................................................................................................. 48
    TRANSITIONAL TECHNOLOGIES (LADDER-CLIMBING VS. LEAPFROGGING) .............................................................. 49
    IMPORTANCE OF REGIONAL INTEGRATION OF INFRASTRUCTURE ............................................................................. 50
    ENERGY SECURITY .................................................................................................................................................. 51
APPENDICES............................................................................................................................................................ 53
    APPENDIX I: A PRELIMINARY ENERGY NEEDS ASSESSMENT FOR THE MDGS. .................................................... 53
    SETTING TARGETS FOR ENERGY SERVICES.............................................................................................................. 53
    ESTIMATING RESOURCE NEEDS ............................................................................................................................... 54
    ONGOING REFINEMENTS OF THE ENERGY NEEDS ASSESSMENT .............................................................................. 55
    PRELIMINARY COUNTRY RESULTS ........................................................................................................................... 56


A measure of the some of the essential attributes of a ―decent‖ life, and how a substantial proportion of
those who do not enjoy such a life at present can achieve it by 2015, is embodied in the Millennium
Declaration and the Millennium Development Goals (MDGs). Energy is an enabling service that is
required to live a decent life and concurrently can allow the capacity of our intellect and physical
strengths to amplify our productive abilities.

Without a certain absolutely essential amount of energy, human existence would be nearly impossible—
for example, the energy needed to cook food and the energy needed to stay warm in cold climates. But
one needs more than this life-sustaining amount of energy to live a decent life. How much more energy,
and in what form, is sufficient to meet that goal? One reason this question is difficult is that, firstly, one
needs affordable and reliable access to energy in order to carry out many of the specific activities that
generate and sustain economic growth, which in turn increases income, initiating a bootstrapping process
toward this decent life. Secondly, one also needs energy to alleviate many of the broader conditions that
can prevent people from contributing to and benefiting from economic growth.

It is critical to recognize that energy alone is not sufficient for creating the conditions necessary for
economic growth to occur. A central message of the Millennium Project is that a whole set of integrated
strategies—for reducing poverty, hunger, disease, illiteracy, environmental degradation, and
discrimination against women—is required to meet the MDGs and to ensure that no one single factor
becomes an impediment to development. Hence the following exercise considering energy needs for the
poor is carried out as an effort to ensure that lack of adequate energy services (both in amount and kind)
does not become an impediment to achieving the MDGs.

While an attempt to define what a decent life is certainly fraught with difficulty, a relationship between
primary energy consumption and the Human Development Index (HDI), an arbitrary but nonetheless
useful measure of well-being, shows a sharp increase in HDI with small increases in energy consumption
at low consumption levels (Fig. 1).

Fig. 1 Human Development Index (UN HDR, 2004) vs. Annual Per Capita Commercial Energy Consumption
(UN Common Database, 2000, in kilograms of oil equivalent).

Even though attempts to establish a causal relationship between the two have not been conclusive, higher
GDP per capita values are also associated with higher energy consumption, as can be seen in Figure 2.

Figure 2: A log-log plot of per capita commercial energy consumption in kgoe and GDP for year 2000 (in
$PPP Intl.) (Sources: UN/WB).

Together, these figures illustrate that increased energy consumption correlates closely with both welfare
generally and economic growth. These relationships appear to become decoupled at higher levels of
wealth and development, when further increases in consumption do not appear to impact the HDI and
economic growth becomes less energy-intensive. This illustrates the consumptive nature of some uses of
energy. It is entirely possible that due to geographic isolation and economic disparities within a country
one could have large consumptive uses of commercial energy without benefits actually reaching the poor.
However, since the focus in this analysis is energy’s relationship to poverty and the MDGs, we are
primarily concerned with the far less ambiguous regime where, for example, an increase from 30 to 300

kgoe1 in primary commercial energy consumption does have a strong association with dramatically
improved living standards.

In Table 1, on the following page, the annual per capita commercial energy consumption and electricity
consumption (in kWh) are shown for several countries. Also shown are measures of poverty such as the
per capita income in Purchasing Power Parity Dollars and the percent population below $1/day and
$2/day where income is in $PPP. The data show that there is considerable variation within the poor
countries; however these differences are small relative to the energy consumption of wealthy countries.
Note that biomass energy sources are not included here; that a significant portion of the energy
consumption is not household consumption; and that commercial energy consumption across households
also varies.

    kilograms oil equivalent:1 kgoe is about 42 MJ

 Table 1: GDP per capita, energy consumption and poverty in selected countries, with emphasis on Sub-Saharan Africa.
Sources: UN Human Development Report, 2004:;
United Nations Common Database:
                                                                 Electricity      Energy consumption,
 HDI                                                            consumption         commercial, per                          Population below
rank         Country               GDP Per Capita                per capita              capita                           income poverty line (%)
                                                                                                                          c                d
                                   US$           PPP US$                                                        $1 a day         $2 a day        Nat’l poverty line
                                                                                                                         b                b                    b
                                   2002             2002          (kWh) 2001        kg oil equivalent, 2000    1990-2002        1990-2002          1990-2001
8      United States           36006 (a)         35750 (a)      13,241            7725
9      Japan                   31,407            26,940         8,203             3730
28     Korea, Rep. Of          10,006            16,950         6,632             3284                         <2              <2                ..
72     Brazil                  2,593             7,770          2,122             717                          8.2             22.4              17.4
94     China                   989               4,580          1,139             561 (b)                      16.6            46.7              4.6
119    South Africa            2,299             10,070         4313 (c)          2649 (c)                     7.1             23.8              ..
127    India                   487               2,670           561              318                          34.7            79.9              28.6
130    Cambodia                321               2,060             18             14       (a)                 34.1            77.7              36.1
131    Ghana                   304               2,130           404              120                          44.8            78.5              39.5
146    Uganda                  236               1,390            66              26                           82.2 (a)        96.4 (a)          44
148    Kenya                   393               1,020          140               96                           23              58.6              52
157    Senegal                 503               1,580          151               128.                         26.3            67.8              33.4
162    Tanzania, U. Rep. of    267               580              85              41                           19.9            59.7              35.7
163    Côte d'Ivoire           707               1,520          233               174                          15.5            50.4              36.8
165    Malawi                  177               580              76              27     (a)                   41.7            76.1              65.3
167    Chad                    240               1,020            12              5      (a)                   ..              ..                64
170    Ethiopia                90                780              30              29                           26.3            80.7              44.2
174    Mali                    296               930              34              18     (a)                   72.8            90.6              63.8
176    Niger                   190               800              41              33     (a)                   61.4            85.3              63
                               Source: GDP and population       Source:           Source: UN Common            Source: World Bank. 2004. World Development
                               data from World Bank. 2004.      correspondence,   Database, code: 17510        Indicators 2004. CD-ROM. Washington, DC.
                               World Development                UN DESA,          a: Estimate, UN Statistics   a. HDR 2003/WB Dev. Indicators, 2003.
                               Indicators 2004. CD-ROM.         Statistics        Division                     b. Data refer to the most recent year available during
                               Washington, DC.                  Division, NY.     b: does not include Hong     the period specified.
                               a. practical issues arising in   03/2004.          Kong, Macao and Taiwan.      c. Poverty line is equivalent to $1.08 (1993 PPP
                               the calculation of the PPP                         c: South African Customs     US$).
                               US dollar GDP prevent make                         Union.                       d. Poverty line is equivalent to $2.15 (1993 PPP
                               these values different.                                                         US$).

Although the selection of countries in Table 1 was rather arbitrary, it shows important general patterns
among both the nations of sub-Saharan Africa and higher-income developing countries. The three
columns of poverty data show that while Brazil, China and India, enjoy, on average, considerably higher
GDP per capita when compared to the poorest countries, they nonetheless have significant poor
populations. Considered broadly, these data suggest that per capita commercial energy consumptions
level of about 500 kgoe—a value based on a nation’s total, not just residential, consumption—are not
unrealistic to associate with an expected 50% reduction in the number of poor for the poorest countries.
Simplistic as this analysis is, a review of studies that attempt rigor in obtaining the precise energy-
economic growth relationships have failed to provide estimates that are much better. This is due to
variations in other key factors, such as the structure of the economy, geography, which domestic energy
resources and technologies are available and a consumer’s costs of energy.

Starting with this national-level data, and proceeding to more specific discussions of the relationship
between energy services and many human needs, this report argues that it is important to address energy
poverty as a part of a comprehensive plan to meet the MDGs. Moreover, the choice of the energy
services system—including the choice of specific infrastructure, end-use technology and institutions for
delivery of services—as well as the business plan—including who pays for upfront costs, and how, and
how recurrent costs are financed—have an impact on the ability of the poor to benefit. Proper design of a
system that allows the poor to access these services can have a high proportional return with respect to
achieving the MDGs.

Figure 3, on the next page, shows IEA2 data regarding the contribution of various energy sources to the
total average per capita final energy consumption of one hundred developing and transition countries.
The countries are grouped according to their share of population living under with less than $2/day.

Figure 1: The contribution of different energy sources to the total average per capita energy consumption of
100 developing and transition countries, grouped according to the share of population living under with less
than $2/day, IEA.

Once again, Figure 3 suggests an increase in the total per capita commercial energy consumption (i.e.
excluding biomass) from about 50 kgoe (for >75% living below $2/day) to 400 kgoe (for 40-75% living
on less than $2/day). This is a similar level of primary energy consumption that was suggested earlier
based on the experience of India, China and Brazil as indicated by Table 1. This discussion provides us
with a rough quantitative estimate of the overall level of primary energy consumption that includes all
modes of consumption and not just household consumption.

It is also instructive to observe which modes of final energy consumption grow, and by how much, with
decreasing poverty. What is evident here is the dramatic difference in the topmost red segment of the bar
graph between a representative country with >75% living below $2/day and one with 40-75% living on
less than $2/day. This segment represents liquid fuels which are used primarily for transportation and, to a
lesser extent, in industry and households. While Figure 3 masks the differences in energy use patterns in
individual countries, it illustrates, in an aggregate sense, the importance of transportation and fuel needs
within a country’s combined energy sources.

Along with liquid fuels, electricity deserves special consideration as an energy carrier. Electricity is
primarily produced from motive power. Motive power can be derived from high-grade heat—from a

 International Energy Agency’s World Energy Outlook 2002, Chapter 13, ―Energy and Poverty‖, Fig. 13.2, p. 11.

multitude of possible sources including chemical energy carriers like coal, oil, natural gas and biomass;
geothermal heat; or heat from nuclear fission—or from the potential energy of water at an elevation in
hydroelectric power plants. It is also becoming increasingly easy to generate electricity directly from
wind (where strong sustained winds are available) and from sunlight. While generation of electricity
directly from sunlight using photovoltaic (PV) cells is clean and scalable down to a few watts, making it
ideal for household consumption, PV has a high initial and recurrent costs at present due in part to the
need for batteries for storage.

As Reddy3 (2000) has pointed out, what the end user or consumer cares about is the service that energy
enables. For this reason the focus of this paper is on energy services and not merely energy infrastructure.
Thus, it is important to move from raw data at the national level toward some conception of how much a
given amount of energy may translate into services consistent with meeting the MDGs and how the
population at large can benefit from these services.

We begin with a discussion of the energy needs at the household level. One measure of ―energy poverty‖
at the household level could be the inability to cook with modern cooking fuels and the lack of a bare
minimum of electric4 lighting to read for a few hours each evening. Note that this roughly corresponds to
about 50 kgoe of annual commercial energy per capita, an estimate based on the need for approximately
40 kgoe per capita for cooking and the remainder as fuel for electricity. This represents energy needs just
for household cooking and lighting needs. Hence this estimate is considerably smaller than the earlier
annual per capita consumption of 400 to 500 kgoe. The availability of electricity and is also crucial for
industry, commercial enterprises, small and medium businesses, educational institutions, government
offices, public transportation systems, health clinics, water supply systems, communication infrastructure
and street lighting.

To understand where the energy poor are, we first look at the map of world by country of current
electricity consumption (Fig. 4). This figure illustrates total national electricity consumption at present,
further highlighting differences between equatorial and non-equatorial regions generally, and Sub-
Saharan Africa and the rest of the world in particular. Figure 5, provides data and future projections over
several decades on the change in the number of people without electricity, by region. The figure provides
a clear illustration not only of the absolute disparity in access to electricity for the world’s poorest in Sub-
Saharan Africa and India and others worldwide, but also the challenges ahead in reversing longstanding

Figure 4: World Map of Electricity Use Per Capita

Figure 5: Number of people without electricity, 1970-2030 (IEA World Energy Outlook, 2002)

For cooking food, process heat and transportation, other energy carriers are generally more suitable than
electricity. These include natural gas or liquefied petroleum gas (or LPG, a gas that liquefies when
pressurized in the specialized canisters used for its storage and transportation), liquids such as kerosene (a
petroleum-derived fuel that liquid at atmospheric pressure and does not need a specialized container, but
not as clean burning as LPG) or DME (dimethyl ether, an energy carrier that may emerge in the near
future) and diesel or petrol for transportation. Frequently chemical energy carriers are also a feedstock for
other products such as fertilizers and plastics. High-grade heat from energy carriers is also used for

  in World Energy Assessment, UNDP, 2000.
  Electricity is a much better energy carrier for these needs even though one can for example and many do obtain
lighting by burning kerosene in lanterns.

materials processing in many industrial applications. In colder climates, liquid fuels provide much of the
heat necessary for households. Motive power is also of specific interest, and its direct use in some
contexts allows for the possibility of bypassing the intermediate conversion to electricity. Motive power
produced in engines has a variety of valuable uses: transportation, irrigation, intensification of agriculture
(tractors and farm equipment), processing of agricultural products (moving, crushing, grinding), and a
multitude of other uses. These systems rely on the availability and affordability of common
transportation fuels, such as diesel.

Specifically for cooking, gaseous and liquid fuels have many benefits in addition to convenience and are
increasingly preferred (but not exclusively so) with higher incomes to traditional fuels such as solid
biomass. However, the direct use of solid biomass—a variety of solid fuels such as charcoal, fuelwood or
dung—is widespread in the poorest parts of the world. So another way to observe where the energy poor
are, one can look at geographical and quantitative data on the number and distribution of people who cook
with traditional biomass fuels worldwide as shown in Figures 6 and 7.

Figure 6: World Map of Traditional Fuel as a percentage of total energy consumed.

Figure 7: Number of people relying on Traditional Biomass for Cooking and Heating in Developing
Countries, 2000 (from IEA World Energy Outlook, 2002)

The minimum basic energy service that we anticipate at the household level is i) electricity-based lighting
and communication (perhaps for a radio and a cell phone) and ii) access to modern cooking fuel. Where
access to modern cooking fuel is prohibitively expensive, an intermediate goal may be a significantly
improved household cooking environment. This may consist of higher-quality fuelwood or charcoal, of
consistent availability, from an environmentally sustainable source (as for example from woodlots, agro-
forestry or managed charcoal production). It may also include the use of improved cookstoves in
environments with better ventilation. In the spirit of the MDGs, the goal would be to target at least 50%
of those who do not currently have these services by 2015.

The annual amount of energy required for cooking varies with the type of food, the fuel and stove used
and the specific cooking practices of a household. More than 80% of the heat generated while cooking
with wood on a traditional three stone fire does not end up in the pot, whereas with kerosene and LPG
nearly half can, doubling the cooking efficiency per unit of energy consumed. Diet is also a factor in
energy needs: Cecelski5 reports that coastal communities in Peru and Ghana where fish consumption is
high use much less cooking fuel than inland villages relying on hard staples such as maize, cereals,
potatoes and cassava. Generally, with few exceptions, when most of the food eaten is prepared at home,
about 1 GJ of energy per capita into the pot is needed per year (within a factor of two). If one were to
assume that the needs for cooking fuel were to be met exclusively (not always the case) by either LPG or
kerosene, and accounting for energy efficiency of these fuels, this works out to about 40 kg of LPG (or
about 45 kg of kerosene) per person per year.

With respect to household lighting, a minimum level of about 20 lux over a 10 foot diameter spot at
ground level is proposed as a standard. This can correspond to a total light output of as little as 150
lumens if the light is well directed, or more typically about 300 to 400 lumens. The only clean and
efficient source of lighting currently available is electricity-based. This minimum light level can be
produced currently from a 7W source using a compact fluorescent lamp (CFL), directed by a metal

 Cecelski Elizabeth, Energy and rural women's work: crisis, response and policy alternatives. International Labour
Review, 1987, vol. 126, pp. 41-64.

reflector. With additional needs for communication devices amounting to 3W, one can anticipate a 10W
need for 3 to 4 hours a day, corresponding to about 12 to 15 kWh per year per household. The actual
consumption level would vary by household vastly depending upon the cost of service, individual ability
to pay and form of payment.

This analysis anticipates several issues regarding lighting. First is the initiation of a discussion of lighting
as a basic need and of the essential minimum level to meet the MDGs. Second is a consideration of how
to ensure this level of service for those who currently have no lighting or those who use inefficient and
smoky kerosene wick-lamps—which, due to very low efficiency, tend to provide 10 to 20 lumens and
consume 20 to 30 liters of kerosene per year when used 3 to 4 hours per day. Third is a discussion of the
means to achieve this minimum need in circumstances where a large fraction of a population does not
currently have access to such a service. Fourth, and finally, it is asked how can one ensure that the cost to
the state of providing this service is low and not wasteful so that subsidies can be effectively target those
who cannot afford this basic level of service. , and to plan for energy interventions as a part of the
MDGs, it is helpful to consider change over time in electricity consumption in various regions of the
world. Figure 4, below, provides data and future projections over several decades on the change in the
number of people without electricity, by region. The figure provides a clear illustration not only of the
absolute disparity in access to electricity for the world’s poorest in Sub-Saharan Africa and India and
others worldwide, but also the challenges ahead in reversing longstanding trends.

These proposed initial minimum levels—for cooking, lighting, and other energy needs—are not intended
as a level of consumption for the long term. Instead, one would hope that with the synergies of rising
welfare and economic growth anticipated to accompany meeting the MDGs, those consuming at these
minimum levels would soon enjoy the benefits of larger energy consumption, which they could then
afford and pay for.

What is different about this paper on the energy needs of the poor and its relevance to the MDGs? One
key underlying assumption here is that the poor countries will commit to good governance. Second, in
this analysis, energy services are seen as enabling success in meeting an entire set of objectives related to
the MDGs and can exploit synergies in the process. Third, the goals are to be achieved in a relatively
short time frame of now to 2015, and the targets will require a substantial scale-up—not incremental
increase. This, along with the commitment of the rich countries to provide much larger financial
assistance, can permit economies of scale, reducing unit costs. Lower unit costs will help to bring a larger
fraction of the population within the fold of ability to pay, while reducing the total outlay for subsidies (in
some form) targeted to the poorest.

The broad guiding principles that are used in developing the recommendations are:

     1. The Millennium Development Goals (MDGs) derived from the Millennium Declaration provide
        the operational framework for international development cooperation. At their core, the MDGs,
        which are to be achieved at the national level, call for a rapid scaling up of investments in social
        services, basic infrastructure, gender equality, and environmental sustainability.
     2. Fortunately, specific interventions are known to meet each of the MDGs. Since the goals are
        interdependent, these interventions need to be integrated into broad-based strategies addressing
        the full range of targets.
     3. To meet the MDGs, countries need to develop medium-term action plans through to 2015 based
        on rigorous assessments of their needs in terms of infrastructure, human resources, and financing
        to meet the Goals. These national action plans need to take account of each country’s specificity
        and propose interventions for meeting the Goals at scale.

While the provision of modern energy services generally improves the local environment of the
immediate user, these services frequently rely on use of fossil fuels which can have a variety of impacts.
The extraction and transport of fossil fuels can lead to severe land and water degradation far from the end
use. The safe and effective use of fossil fuels may require more stringent accident prevention measures
than other energy forms. Fossil fuel consumption inevitably leads to some greenhouse gas emissions that
contribute to global warming. Energy security may become an issue for those nations that have ample
fossil fuel reserves today but are exporting the fuel at a rapid pace. The demands for increased energy
consumption and better transport can also lead to a stalemate due to the inter-related and competing needs
of poverty alleviation, environmental sustainability and equity.

Furthermore, many past projects have either neglected or paid only lip-service to the ―software‖ of energy
services and transport infrastructure. Some of the reasons for failure have been: insufficient resources
overall, and particularly for operation and maintenance; political expediency that favors one-time project
allocations and fails to provide for recurrent costs; lack of long-term planning, transparency and public
participation during the planning, contract and implementation process which may, among other
problems, fail to ensure that the needs of the poor who rely on their local natural resources for a living are
addressed or may inadequately compensate those who are adversely affected; and inadequate institutions
and technical capacity to carry out planning studies, establish performance standards, create
environmental impact norms and monitor established rules.

The paper advocates that the stalemate can be broken if one addresses these ―software‖ issues and allows
for adaptation to increasingly cleaner energy technologies with growth as opposed to an expensive
solution that ―leapfrogs‖ to the cleanest possible technology immediately. An example can clarify this: A
transition from an old and polluting coal plant to solar photovoltaic implies a cost increase of nearly two
orders of magnitude in electricity costs. A modern coal or natural gas plant can achieve significant
reductions in pollution with costs that are may be two or three times that of old coal. For electricity
generation, transition from an older diesel generator to a newer technology gas, fuel oil or even diesel
generator with modern emission control technologies dramatically reduces impacts on the local
environment. To quote from a recent UK Department for International Development (DFID) report6,
―Consideration of the positive environmental aspects of renewable energy sources must be balanced
against meeting practically, quickly and efficiently the immediate energy needs of the poor with whatever
energy services are accessible.‖ The dangers of not allowing the stalemate to be broken could be
devastating for a population that is already weakened by poor health. Moreover the technology choices
made in meeting the immediate energy needs of the poor need not be permanent.‖

Hence the advantages and disadvantages of fossil fuels, particularly the environmental impacts, need to be
weighed in a broader context of needs. An accompanying paper by Robert Watson addresses this issue in
further detail. It is noted here that the per capita use of fossil fuels in the poorest countries is vastly
smaller (and likely to remain so in near future) when compared to the use by the developed nations.

 Energy for the poor, underpinning the Millennium Development Goals, August 2002, p. 23.

Executive Summary and Recommendations

The key recommendations made in this report fall into several categories as summarized below:

Energy needs for meeting the MDGs
    1. Energy services are a ―missing MDG‖ that needs to be addressed if countries are to halve
       poverty and meet the other Goals. Specifically, countries need to improve household access to
       improved cooking fuels and stoves, as well as electricity. In addition, energy services need to be
       part of a multi-sectoral approach that can meet the needs of social services (such as clinics,
       schools and drinking water-related infrastructure); telecommunication, irrigation and agro-
       processing services; and important government facilities and small and medium-sized enterprises
    2. In preparing the energy needs assessment, countries need to identify minimum ―MDG-
       compatible‖ energy needs of households and specify coverage targets to be achieved by 2015.
       Moreover, plans for energy services need to recognize that the demand for these services will
       grow as access is increased.
    3. In the poorest countries, energy services should be provided using known technologies that are
       cost-effective and have been proven to work at scale. Where low-carbon technologies offer the
       same services at a lower cost these should be given preference.
    4. The poorest countries should have available at their disposable the entire spectrum of primary
       energy source options so that they are not excessively constrained during their very early stages
       of industrial and agricultural growth. There has been strong emphasis in the last two decades on
       the environmental impact of fossil fuels with the suggestion that donor assistance encourage the
       poor countries to ―leapfrog‖ to newer renewable energy technologies (RETs). While these are
       appropriate in many circumstances, their high cost (e.g. for solar home systems) and limited
       geographical coverage (e.g. for micro-hydro and wind) can become an impediment to rapid
       scale-up. An approach that allows the poor to climb-up the energy ladder, with all options
       available, may allow wider segments of the population to benefit. As consumption grows and
       newer technologies become available, the poor can transition to technologies with lower
       environmental impact. Newer technologies take time to develop and test, a phase during which
       it is easier for a customer with either the financial means or a niche need to afford and use them.
    5. Local productive enterprises, local energy resources, technical improvements in production,
       efficiency improvements, emissions control and sustainable land-use practices can all add up and
       lead to productivity enhancements, improved use of an otherwise potentially harmful energy
       source and simultaneous reduction in unit costs. One example of this is the use of fuel wood in
       conjunction with improved cook stoves and improved cooking environment. Another example is
       the efficient and controlled production of charcoal from wood obtained from sustainable
       managed woodlots or agroforestry, combined with use of efficient charcoal stoves.
    6. It is crucial to recognize the importance of motive power as an energy service. Two examples of
       innovative ideas that demonstrate the importance of motive power are the multi-functional
       platform (for agro-processing and other tasks) and the treadle pump (for small scale irrigation or
       use of water harvesting combined with drip irrigation).

Means of implementation
   7. Recognize the central role that women have in the energy-transport-poverty nexus. Money in the
       hands of women goes further; saving of time and effort of women and girls has greater rewards
       to household income and education. There may be benefits to maternal health and child mortality
       that also feedback to lowered fertility.
   8. Recognize the importance of policies that facilitate delivery of energy services, for example,
       those that lead to the creation and strengthening of businesses and institutions that will provide
       energy services and those that promote the kind of training that people who own, manage and

          work in these businesses will need. Promote actions by national government to reduce import
          duties on energy generation technologies.
    9.    Policies, institutional priorities and donor assistance should recognize and support the energy
          delivery systems that are evolving where organized effort is not reaching the poor. The poor in
          rural areas frequently use batteries, transporting them to charging points, to address their small
          and distributed needs for electricity. However, due to a lack of formal recognition and
          institutional support to build the necessary small-scale infrastructure for this approach, the poor
          end up spending considerably more for low quality, improperly charged batteries and hence
          require frequent charging and replacement. Similarly the lack of recognition of mini-grids or
          stand alone diesel power can also impede access. For example, the cumulative nameplate
          capacity of all imported diesel gensets in Ethiopia over the last ten years is over 1 GW and
          exceeds the current installed hydropower capacity of the country. These standalone systems and
          mini-grids are addressing the need for access in areas without transmission and distribution
          capacity. Utilities, NGOs, donors and planners must recognize the role of these services and the
          private sector in development and ensure that these solutions are supported through financial
          assistance, technical support structures and appropriate standards.
    10.   A combination of large scale and small-scale approaches will be needed to provide energy
          services for meeting the MDGs. Large scale approaches can promote efficient supply and
          market mechanisms for the purchase and movement of energy carriers and materials for
          infrastructure—at the district and national level—that is commensurate with the need for modern
          energy services in households, businesses and public facilities—at the village, town or city level.
          Small-scale approaches will be needed that recognizes that the poor are likely to be excluded
          from modern energy services if such services can only be provided on the basis of full cost
          recovery. Where cost recovery approaches are being used, the penetration rates for the service is
          generally low.
    11.   Build highways that connect population centers to transport hubs and ports; improve port
          facilities and operation; reduce border transit time; where volumes justify, strengthen railroad
          infrastructure; and reduce the cost/time of doing business by regulatory reform and improved
          information and communications technology (ICT).
    12.   Work on regional agreements to form regional power pools, transmission line networks across
          countries to better utilize capacity, reduce variability of supply and use the optimal mix of
          primary energy sources at any time. Create joint agreements for sharing the costs of study,
          research and development of new regional sources of energy such as geothermal and hydro
          power that cut across geographic regions.
    13.   Enhancing local research will lead to capacity building, technological developments that emerge
          out of the local need and practice, emergence of standards, and provide a forum for south-south
    14.   Vocational training can engage youth in the learning process and in the role of infrastructure
          extension while creating a favorable environment for creation of small businesses.

    15. A ―bottom-up‖ approach will be needed that recognizes that the poor are likely to be excluded
        from modern energy services if such services can only be provided on the basis of full cost
        recovery. Where cost recovery approaches are being used, the penetration rates for the service is
        generally low.
    16. Even the poor can generally afford to pay the cost of fuel and/or amortized generation costs of
        electricity in grid based systems. However they are unable to pay for the upfront costs of
        extending the wire (and associated transformers and meters) to their home. Suitable technologies
        (e.g. low cost concrete poles in India and use of Single Wire Earth Return or SWER technology),
        appropriate standards (e.g. lower transformer capacity per rural consumer) and aggregation, can
        lower the upfront costs substantially. In addition, at their low level of consumption, per kWh

         tariffs may not be enough to cover the cost of billing and collection. There are technological
         solutions (e.g. prepaid cards in South Africa) or community-based approaches (e.g. a fixed
         monthly fee per customer for a fixed wattage) that can address the later problem. Still, in
         dispersed areas, the upfront costs are unlikely to be affordable by the poor even if everyone in
         the community was provided with a connection. Yet the social and economic benefits of
         building the necessary grid and delivery infrastructure with public/donor assistance are likely to
         make such an investment desirable. There are innovative mechanisms by which such assistance
         could be channeled through the private sector as well.
     17. Perhaps even more than lighting, the upfront and recurring costs of modern cooking fuels are
         difficult to afford if the distribution and sale of these fuels is entirely on a cost basis. The
         multitude of short and long-term benefits to health, productivity, agriculture, environment and
         women’s welfare generally that modern cooking fuels provide suggests that a rapid scale-up of
         modern cooking fuel distribution systems should be important part of the strategy to meet the
         MDGs. A combination of top down interventions such as lowered transport and distribution
         costs (through improved road and port infrastructure), better handling and storage facilities at
         ports, bulk purchases of fuels, and impetus from the government through suitable regulatory
         reform can assist in lowering the costs of LPG and kerosene. In addition bottom-up approaches
         can encourage market development through smaller LPG cylinders, upfront payments that are
         spread out over a longer period and consumer education. In areas where the bottleneck is
         upfront capital costs (for example for a LPG stove-cylinder or for a new electrical connection)
         and not the ongoing costs, immediate opportunities exist for scale-up.
         For the case of modern cooking fuels, it is difficult for households making less than $2/day7 to
         afford the recurrent cost of fuel immediately since household fuel costs alone could be as much
         as $0.30/day8. Suitably designed targeting to identify those populations, an exit strategy from
         subsidies as people earn and consume more, and bill collection schemes that minimize cost of
         collection can ensure that subsidies are indeed reaching the poor with least leakage or waste.
     18. Improvements in infrastructure, financing mechanisms and institutional framework can reduce
         the ―overhead‖ of providing a service in areas where an ability to pay for the service already
         exists—if these overhead costs are reduced to levels comparable to those in developed countries
         with high penetration of services.
     19. Ensure that the framework for oil/gas revenue management is transparent as well as sensitive to
         the energy security of the country.

Energy services and the Environment
    20. Efficient and judicious use of fossil fuels for cooking can actually allow the immediate
        environment within a home to improve. At the low levels of consumption needed for cooking,
        the benefits of this improvement may outweigh the global impact of use.
    21. Fossil fuels remain the convenient energy carriers for motorized transport. Promoting fuel
        efficiency for this end-use can improve the local environment as well as reduce GHG emissions
        that impact global environment.

 This is in real USD, corresponding to a per capita income in PPP terms of about $1 to $3 per day.
 This assumes daily consumption of 0.50 kg, LPG costs of $300/ton and transport/filling/distribution costs of

Linkages with the Millennium Development Goals

It is well understood that energy and transport are enablers of both the social and economic dimensions of
growth; however, quantitative understanding of the relationship is just beginning to emerge. Some of
these linkages are discussed below. The issues of gender inequality are so pervasive that they are present
in nearly every linkage discussed here.

Poverty reduction and savings in time

At the household scale, some of the ways that modern energy services may directly contribute to
economic growth and poverty reduction are as a direct amplifier of human capacity, by creating the
possibility of small enterprises and by reducing unit costs. Indirect contributions to economic growth may
come in the form of free time for other productive activities, improved health and education, improved
access/supply of clean water, and reduced local environmental degradation.

The amplification provided by energy services is visible in the developed world. Electricity can allow
one to operate computers and servers, and use telephones and the internet, which can in turn lead to a
large increase in the productivity of human labor. A single trained professional can provide services to a
much larger number of people with improved transport and telephony.

Many productive uses of electricity rely on motors for producing motive power. However one can also
directly produce motive power from an engine. A UNDP/UNIDO project in Mali has shown that the use
of a multi-functional platform (an diesel engine-driven source of motive power that can power a variety of
end-use devices for food and agriculture product processing, small scale water pumping, carpentry and
possibly for electricity using a generator) resulted in multiple benefits. The platform is owned by a
women’s co-operative, which mobilizes funds for its acquisition and elects a management committee with
the skills to ensure the technical and economic viability of the platform. 9 A single platform increased
annual incomes of several hundred women from $40 to $100 and freed up between two to six hours of
time per service-use, depending upon the specific services provided by the platform. In a study of the
impact of these platforms, the food processing functions of the platform alone led to time savings, better
quality products, lower product loss in processing, and increased processing capability10. There has also
been improved attendance of girls in schools with more time available for studying. The project shows
that with a participatory demand-driven approach, ownership, management and operation by women, and
micro-finance, energy technologies can lead to poverty reduction and associated benefits. Abeeku Brew-
Hammond of the Kumasi, Ghana based KNUST/KITE11 is also examining the possibility of using other
liquid biofuels to substitute for the diesel.

A UNDP concept paper about the platform notes that the use of diesel may be an obstacle for acceptance
within environmental management and sectoral programs that favor renewable energy technologies
(RETs). One should not ignore the possibility that much research has gone into mass-produced diesel
engines that are much cleaner today. Diesel engines now power more than 40% of the automobiles in
continental Europe. Increased incomes today can also make it possible to transition to a cleaner
technology tomorrow.

  Energy For Poverty Reduction: The Concept Of The Multi-Functional Platform (MFP) As A Regional Project,
UNDP Concept Paper, May 2001.
   Impact Study Of The Multi Functional Platform On The Living Conditions Of Women, Mali, June 2001
   Kwame Nkrumah University of Science and Technology/Kumasi Institute of Technology and Environment

A recent comprehensive effort to measure the social benefits of rural electrification in the Philippines12
has attempted to quantify the benefits of electricity access. In spite of the fact that the task is fraught with
analytical difficulties, the authors find significant benefits in terms of opportunity costs from time saved,
lower cost lighting and improved productivity of home business. Other studies also point to the positive
relationship between educational opportunities made possible by electric lighting, and higher lifetime
earnings, as well as public safety13.

Role of women (time poverty, water, firewood, transport, health)

It would be hard to imagine in the developed world today a family spending one or more hours every day
gathering biomass such as wood, agricultural residues, and dung when one could instead buy cooking fuel
for the same purpose at a price that reflects a mere two or five minutes of income from work. Yet this is
the burden of women in the developing world. The disproportionate amount of daily time and effort spent
by women and young girls gathering solid fuels and water for household chores could be used for other
income-producing activities. The time spent gathering biomass varies with geographic location, land
ownership, the time of the year, climatic events and loss of control over local resources. Figure 8, below,
shows data from Tanzania.
Figure 8: Time spent (in hundreds of hours) and the transport burden (measured in tons-km), in Tanzania,
both per person and per year, disaggregated by gender and tasks.
From UNDP “Energy After Rio: Prospects and Challenges,” Chapter 2, 1997, available at [the original source may be Barwell, et al, “Household Time Use
and Agriculture Productivities in Sub-Saharan Africa. A Synthesis of I. T. Transport Research (London: I.
T. Transport Ltd, November 1987)]

A recent ESMAP study in rural India14 found collection time for wood to be 37 hours/month. The
benefits to health may be even larger (evidence for this is discussed separately under ―Health‖ later in this
document.) It is also worthwhile quoting verbatim the conclusions of a recent study15 on the impact of
energy on women's lives in rural India.

     1.   The long, unrecognized hours spent on arduous, unhealthy, and unpleasant tasks have sometimes been cited
          as a development outcome for rural energy projects. But the main motivation for rural energy programs
          justifiably has been generally improving rural quality of life, conserving fuel, and alleviating deforestation.
          This study shows that the impact of adopting modern energy services by rural women may be even more
          significant than was previously assumed by development researchers. As a consequence, in designing rural
          energy programs it is not unreasonable to pay more attention to the problems women have in obtaining a
          reliable and efficient energy supply.

     2.   Women in India typically spend much of their time on the hard work involved in caring for their families.
          They often work 12 to 14 hours a day, most of which is unpaid and recognized mainly within the family. In

    UNDP/World Bank Energy Sector Management Assistance Programme (ESMAP), ―Rural Electrification and
Development in the Philippines: Measuring the Social and Economic Benefits,‖ Report ESM255/02, May 2002.
    Fitzgerald, K.B., D. Barnes, and G. McGranahan, 1990: Interfuel Substitution and Changes in the Way
Households Use Energy: The Case of Cooking and Lighting Behavior in Urban Java. Industry and ,Energy
Department, World Bank, Washington, DC, 13 June.
   UNDP/World Bank Energy Sector Management Assistance Programme (ESMAP), ―Energy strategies for Rural
India: Evidence from Six States,‖ August 2002.
   Impact of Energy on Women's Lives in Rural India‖, carried out recently by the Joint UNDP/World Bank ESMAP
program, The International Bank for Reconstruction and Development and THE WORLD BANK.

          addition, some of those hours are spent in an extremely unhealthy environment. Respiratory illness and eye
          problems are common among women who cook on traditional chulhas (Smith 1998). Infant mortality may
          also be higher among children raised in such homes (Hughes and Dunleavy 2000; Claeson, Bos, and
          Padmanathan 1999; Mishra, Retherford and Smith 1997). This study supports the findings that women who
          use biofuels lead the most burdensome lives. Of the three biofuels, firewood involves the most drudgery in
          terms of time and effort needed to collect it on a regular basis, but in terms of time spent cooking it is a better
          alternative than agricultural residues or dung. In any event, most women use a combination of these fuels to
          meet their energy needs.

     3.   The use of LPG or kerosene stoves relieves women of much of the most arduous tasks involved in cooking
          for their families and permits them to lead a relatively comfortable and healthy life. Household electrification
          also has positive consequences for women in terms of their general quality of life, including an increased
          likelihood that they will read, watch television, and earn income. Having lights at night increases their ability
          to read in the evening after dark. However, the advantages of electrification could be exploited even more, as
          there still is a puzzling underinvestment in appliances such as mixers, grinders, blenders, and others that could
          help women with their daily household work.

     4.   While the Government of India in association with many other nongovernmental private organizations has
          instituted various programs to address rural energy problems, the execution, pace, and rigor of
          implementation has been uneven. The results of this study indicate that such programs are essential for
          bringing about greater independence of women, principally through reducing the time spent on such tasks as
          fuelwood collection, food preparation, and cooking. In addition, the subordinate position of women in rural
          society needs to be recognized in the development and implementation of rural energy programs. Many recent
          studies indicate that the consultation with and participation of those benefiting from development programs
          lead to a greater likelihood of their success.

     5.   Kerosene, LPG, electricity, and improved stoves do appear to have a significant impact in terms of reducing
          arduous tasks by women in rural households. Besides improving the quality of life, in some cases electricity
          can be used for productive and income-producing activities inside of the home. Although not all households
          or all women in rural households will take advantage of the benefits made possible by the modern use of
          energy, the benefits for the majority of households are cumulative and worthwhile. In this context, energy
          policymakers need to pay more attention to the impact of modern rural energy services and how they affect
          the lives of women.

Rosen and Vincent report16 that households (primarily women) spend an average of 134 minutes per day
collecting water, and that the time saved by bringing water supplies closer to households is likely to
dominate estimates of the benefits of improving rural water supplies. Modern energy services through use
of electric or fuel-operated pumps can make it easier to bring water supply closer to home. In addition to
time costs there may be energy costs as well. The time spent collecting fuel wood also reduce the
proportion of daylight hours otherwise available, which may be critical to other income generating
activities. This could be further exacerbated due to absence of electricity access.

In addition to the time and effort spent gathering fuel, there are also related needs that arise from the need
to fetch water and carry supplies and products to and from markets. Frequently water is fetched by girls
and women in plastic containers that are either head loaded or carried strapped on the back, from a water
source (river, spring or a stream) likely to be at a lower elevation or lifted from a well. Mechanical
energy—perhaps from a windmill, a diesel generator or an electric motor—can provide the means to lift
the water to a storage tank.

  Rosen, S., J. R. Vincent. "Household Water Resources and Rural Productivity in Sub-Saharan Africa: A Review
of the Evidence." Harvard Institute for International Development Paper DDP No. 673, February 1999.

Equally important are transport services for women. In a study for the ILO, Geoff Edmonds17 has
examined the effects of poor access. Studies by Malmberg-Calvo (1994)18, Urasa (1990)19 and Howe et
al20 (1987) show the critical importance of gender-responsive solutions. In the latest of these studies, the
high and disproportionate transport burden of women in rural Tanzania (see Figure 3), one begins to
comprehend the scale of the problem. In a World Bank technical paper Calvo21 summarizes the results of
household surveys (from Burkina Faso, Uganda, Zambia, Ghana and Tanzania) that show that 87 percent
of trips in rural Africa take place on foot and that women bear more than 65 percent of the household time
and effort spent on transport. The daily transport burden of a typical adult woman was equivalent to
carrying a load of 20 kg for 1.4−5.3 kilometers. The paper also reports that since women are frequently
the principal farmers and key providers for their families22 reducing unnecessary time spent on transport
would free up the time for other beneficial economic and social activities.

Time for domestic effort also impacts a women’s ability to access health services. Fetching water and
fieldwork23, daily chores essential to the welfare of the family, affect a woman’s ability to access prenatal
care. Similarly, one study has shown that women who report that their ability to fetch water was
negatively impacted by a need to visit the clinic were two times more likely to utilize prenatal care
services at a low level than other women. Mwaniki24 reports that utilization of health facilities for
maternity services was significantly influenced by the number of children and distance to health facility in
that, as number of children increased, utilization of maternity services fell and mothers living less than
five kilometers to a health facility utilized the services better than those living beyond this range. One of
the major constraints experienced by the mothers as they sought for the services included lack of transport
and lack of money for transport. Porter25 points out that while women often cannot afford transport, even
when they live in settlements with good road access, the burden faced by women who reside off-road is
additionally heavy because of the roles they are commonly expected to play in sub-Saharan Africa as
porters and (particularly in West Africa) as produce traders. In addition to their many household duties,
women are widely required to transport their husbands’ produce, as well as their own goods, to the
nearest good road, or directly to bulking markets (which are now mostly situated on paved roads). In
coastal Ghana and elsewhere they may also have to travel to richer roadside settlements for other
household tasks such as taking maize to the grinders for food preparation, since poorer off-road
settlements often do not have their own grinding machines. Kaur26 provides a rare on-road/off-road

   Wasted Time: The Price of Poor Access, By Geoff Edmonds (for the International Labour Organization, 1998,
Development Policies Department International Labour Office Geneva).
   C. Malmberg-Calvo: Case Study on the Role of Women in Rural Transport - Access to Domestic Facilities. World
Bank. 1994.
   I. Urasa: Women and Rural Transport - An Assessment of Their Role in Sub-Saharan Africa. Rural Travel and
Transport Project, SSATP, International Labour Organisation, Geneva., 1990.
   Barwell. I. J., Howe, and P. Zille, Hourehold Time Use and Agriculture Productivities in Sub-Saharan Africa. A
Synthesis of I. T. Transport Research (London: I.T. Transport Ltd., November 1987).
   World Bank Technical Paper No. 411. ―Options for Managing and Financing Rural Transport Infrastructure,‖
Christina Malmberg Calvo.
   Cleaver, Kevin, and Götz Schreiber. 1994. Reversing the Spiral: The Population, Agriculture and Environment
Nexus in Sub-Saharan Africa. Directions in Development. Washington, D.C.: World Bank.
   McCray, T. M. "An issue of culture: the effects of daily activities on prenatal care utilization patterns in rural
South Africa." Social Science & Medicine In Press, Corrected Proof.
   Mwaniki, P. K., E. W. Kabiru, et al. (2002). "Utilisation of antenatal and maternity services by mothers seeking
child welfare services in Mbeere District, Eastern Province, Kenya." East African Medical Journal 79(4): 184-187.
   Porter, G. (2002). "Living in a Walking World: Rural Mobility and Social Equity Issues in Sub-Saharan Africa."
World Development 30(2): 285-300.
   NEED TO COMPLETE THIS REF: 2000, cited in Booth et al., 2000, p. 70

comparison of women’s income which illustrates this point. Women in a village on a main road in
Cameroon were found to be able to make an average income over double that earned by women in an
isolated village 90 minutes travel from the road, because of the greater time they had available to produce
food to sell. In common with women resident in roadside locations, off-road women have less funds than
men to pay fares when transport is available. However, in addition (as discussed above) the off-road fares
tend to be substantially higher than those for journeys on paved roads.

Rural transport charges are higher in sub-Saharan Africa than in any other region in the world. For
instance, Ellis and Hine27 (1998) compare villages in Zimbabwe, which have an average of one motorized
vehicle per 300 people, with Sri Lanka, where the level is five times as great. For journeys of as much as
30 km they report transport charges as up to two and a half times more expensive in Africa than Asia.


As reported by Task Force 2, ―replenishing soil fertility at the lowest possible cost‖ is a key entry point in
Africa. Use of modern fuels or improved stoves can allow a greater proportion of biomass (in form of
crop residues and manure) to be returned back to the soil. This allows nutrient replenishment directly as
well as indirectly by providing additional soil carbon that in turn can reduce leaching of soils. Modern
cooking fuels could also indirectly increase farm productivity by freeing up women’s time and ensuring
that the physical effort of biomass collection by child-bearing women is not detrimental to their health.

In addition to fuel, electrical power could help raise agricultural productivity and income of smallholder
farms by enabling them the use of information and communication infrastructure for weather and market
information and pumps to lift water collected through rainwater harvesting.

High transport costs due to inadequate road network density and poor road conditions can lead to high
costs of agricultural inputs, affecting farmers’ incomes. Lowering of transport costs at all levels (to ports,
within the country on national roads, regional roads and rural feeder roads) can also significantly lower
the cost of transporting fuel.


For school-age girls, modern energy services could mean more time to go to school and time for after-
school study. King and Alderman in their brief28 summarize studies that show that investments in
infrastructure that save time spent collecting water and fuelwood benefit all household members, and, in
particular, mean fewer interruptions to women’s paid work and to girls’ schooling. Access to transport,
may also play a role in education. A study of Morocco shows that the presence of a paved road in the
community improves the schooling outcome of rural children. The study notes that in the absence of
paved road, 21% of rural girls, as compared to 58% of rural boys, ever attend school. If a paved road
exists, the schools participation rate increases to 48% for girls and 76% for boys29. Schultz30 also

   Ellis, S. D., & Hine, J. L. (1998). The provision of rural transport services: Approach paper. World Bank SSATP
Working Paper 37, Washington, DC.
   King EM, Alderman H., ―Education,‖ Brief No. 6 in 2020 Focus 6 Empowering Women to Achieve Food
Security, Quisumbing AR and Meinzen-Dick RS, Eds., International Food Policy Research Institute (IFPRI), 2001.
   Khandker, Shahidur R., Victor Lavy and Deon Filmer (1994) Schooling and Cognitive Achievements of Children

suggests that girls are constrained in their schooling in part by the demands placed on their time and
suggest that the use of electricity and refrigeration could reduce households' dependence on the labor of
girls. Reflecting the complexity of the problem, Glick and Sahn31 argue that an increase in incomes is also
an important factor since even when electricity access is available in urban areas domestic work
obligations continue to limit female schooling for the very poor.


There is increasing evidence that the use of solid biomass fuels for cooking in indoor environments,
especially using poorly designed stoves in inadequately ventilated spaces, can lead to an increased disease
burden. Once again, where women—including mothers with young children—carry out a
disproportionate amount of cooking activity they are also likely to share a disproportionate disease
burden. There has been substantial recent progress made in measuring, examining, documenting and
attempting to identify quantitative links between use of solid cooking fuels and the associated disease
burden. Several leaders in this field—Kirk Smith, Dan Kammen, Majid Ezzati and Nigel Bruce—have
carried out much of the work and reviewed recent progress32,33,34. Moreover organizations such as the
WHO, the World Bank and ITDG35 have also actively tried to document these linkages.

Smoke produced during the combustion of solid fuels contains a number of pollutants such as
particulates, carbon monoxide, benzo[]pyrene, formaldehyde, and nitrogen dioxide. In households with
limited ventilation (as is common in many developing countries), exposures experienced by household
members—particularly women and young children who spend a large proportion of their time indoors—
have been measured at levels many times higher than health-based WHO guidelines and national
standards (Bruce et al. 200036; Smith et al 200337). Exposure to small particulates (less than 10 microns in

in Morocco, World Bank Discussion Paper 264, Washington DC: World Bank. http://www-
   Schultz, T. Paul. "Returns to Women's Education," Yale University Economic Growth Center Discussion Paper
No. 603, June 1990.
   Glick P; Sahn D E. ―Schooling of girls and boys in a West African country: the effects of parental education,
income, and household structure,‖ Economics of Education Review, Volume 19, Issue 1, February 1999.
   Ezzati M and Kammen DM (2001) ―Indoor air pollution from biomass combustion as a risk factor for acute
respiratory infections in Kenya: an exposure-response study‖ Lancet, 358(9282): 619-624 (erratum 358(9287)).
   Y. von Schirnding, N. Bruce, K. Smith, G. Ballard-Tremeer, M. Ezzati, K. Lvovsky, Addressing the Impact of
Household Energy and Indoor Air Pollution on the Health of the Poor – Implications for Policy Action and
Intervention Measures. CMH Working Paper Series Paper No. WG5 : 12.
   Addressing the Impact of Household Energy and Indoor Air Pollution on the Health of the Poor: Implications for
Policy Action and Intervention Measures. Paper Prepared for the Commission on Macroeconomics and Health.
(Working group 5 : Improving Health Outcomes of the Poor). Based in parts on the proceedings of a WHO-USAID
Global Consultation on the Health Impact of Indoor Air Pollution and Household Energy in Developing Countries,
Washington DC, 3-4 May 2000, by Y. von Schirnding, WHO, Geneva, N. Bruce, University of Liverpool, United
Kingdom K. Smith, University of California, Berkeley, USA, G. Ballard-Tremeer, Eco Ltd, United Kingdom, M.
Ezzati, WHO, K. Lvovsky, World Bank.
   Smoke – the Killer in the Kitchen- Indoor Air Pollution in Developing Countries by Hugh Warwick and Alison
Doig, ITDG Publishing 2004.
   The health effects of indoor air pollution exposure in developing countries by Nigel Bruce, Rogelio Perez-Padilla
and Rachel Albalak World Health Organization 2002.
   Smith KR, Mehta S, Feuz M, Indoor smoke from household solid fuels, in Ezzati M, Rodgers AD, Lopez AD,
Murray CJL (eds) Comparative Quantification of Health Risks: Global and Regional Burden of Disease due to
Selected Major Risk Factors, Geneva: World Health Organization, 3 volumes, in press 2003.

diameter) is believed to be a risk factor for acute respiratory infections (ARI) and acute lower respiratory
infections (ALRI). There also appears to be association of exposure with chronic bronchitis [assessed by
symptoms] and chronic obstructive pulmonary disease [COPD - progressive and incompletely reversible
airways obstruction] particularly among women. Smith et al (2003) also report evidence from China on
increased evidence that exposure to coal smoke in the home markedly increases the risk of lung cancer,
particularly in women. Evidence, though tentative, is also reported of other important child and adult
health problems, such as low birth weight babies and blindness in adults.38

A recent World Bank (2004) report39 on policies and actions for achieving the MDGs, states that in 1999
some 10 million children under the age of five died in low-income countries—2.1 million in India alone.
Using Indian health survey data, World Bank researchers40 have concluded that investments targeted at
improving environmental conditions can substantially reduce child mortality. In rural India, per 1,000
live births, universal access to private piped water would save 10.4 children, electricity would spare 5.5,
and separate kitchens with clean cooking fuels would save the lives of 33.6. The analysis also shows that
girls face a higher mortality risk than boys after the first month of birth (suggesting a significant gender
bias in household resource allocation) and that child mortality is higher among those born to mothers with
no education. Universal female primary education would reduce the under-five mortality rate from 99.9
to 77.0 deaths per 1,000 live births.

HIV/AIDS impacting labor shortage

It is sometimes assumed that labor is plentiful throughout poorer communities and is not a constraint on
development programs. In research related to understanding the dynamics of soil degradation in sub-
Saharan Africa, Muchena et al41 emphasize that ―the myth of abundant cheap labor and/or family labor …
is increasingly being questioned even in densely populated areas such as Kisii in Kenya (800 persons km-
  ); a number of farms are already experiencing shortage of labor as off-farm incomes become more
attractive.‖ The impact of AIDS on available labor for work in the livestock42, non-timber forestry43 and
other agricultural sectors has been devastating throughout sub-Saharan Africa. The Joint United Nations
Program on HIV/AIDS [UNAIDS] 199944 reports that, in Zimbabwe, the output of communal agriculture
fell by half in the preceding five years, largely due to AIDS. FAO estimates that, by 2020, nearly 20% of
Kenya’s agricultural labor force could be lost to AIDS.45 The affect of AIDS on African women will play
a crucial role in the continent’s future, as women fill vital roles both as agricultural workers and as
caregivers. Women account for 70 percent of the agricultural workforce in sub-Saharan Africa, and a
recent study in Tanzania found that a woman whose husband is ill spends 60 percent less time on

  Biomass Cooking Fuels and Prevalence of Blindness in India by Vinod K. Mishra, Robert D. Retherford, and Kirk
R. Smith, Journal Of Environmental Medicine, 1: 189–199 (1999)
     Global Monitoring Report 2004, World Bank.
   Van der Klaauw, Bas, and Limin Wang. 2003. ―Child Mortality in Rural India: Determinants and Policy
Implications.‖ World Bank, Washington, D.C.
   Muchena, F. N., D. D. Onduru, et al. "Turning the tides of soil degradation in Africa: capturing the reality and
exploring opportunities." Land Use Policy, (March, In Press?) July 2004.
   Engh, I.-E., et al. HIV/AIDS in Namibia: The impact on the livestock sector. Sustainable Development
Department (SD), Food and Agriculture Organization of the United Nations (FAO), February 2000.
   Barany, M., A.L. Hammett, A. Sene, B. Amichev. ‖Nontimber Forest Benefits and HIV/AIDS in Sub-Sharan
Africa.‖ Journal of Forestry, December 2001.
   Joint United Nations Program on HIV/AIDS (UNAIDS). 1999. A review of household and community responses
to the HIV/AIDS epidemic in the rural areas of sub-Saharan Africa. Geneva.
   Food and Agriculture Organization of the United Nations (FAO). 2001. AIDS: a threat to rural Africa. Rome.

agricultural work.46 In contexts such as these where labor is scarce, energy services can act as a substitute
for hours of taxing manual labor, raising productivity dramatically.

Energy and ensuring environmental sustainability

A set of concerns regarding the increased use of fossil fuels for energy services relates to the associated
impacts on the environment. This includes effects on natural resources (especially the local resources on
which the poor depend for their livelihood), the local and global environment, and on health (which is
effected by changes in the local environment). Fossil fuel use, exploration, transportation, transformation
and distribution will have some unavoidable detrimental effects on the environment. These activities, and
related technologies and infrastructure, will also inevitably present some risk to people—risks that can be
minimized but perhaps not completely eliminated.

Sometimes these risks and effects present competing and contradictory benefits and problems. One
example is cooking fuels: While it is desirable to replace biomass as a cooking fuel with LPG, the former
is a renewable resource whereas the later is not. Similarly, while rural roads are important for growth,
there are potential impacts on the environment (see the report of the taskforce on environmental
sustainability) and health (accidents and possible higher HIV rates) as well. As suggested at the outset,
responding to these challenges will require considerable community participation and empowerment with
real say in decision making, careful study of impacts on the environment, institutions that carry out these
tasks, fair compensation to those who are displaced and engineering ingenuity to ensure that the problem
of local environmental degradation is addressed.

There are opportunities to learn from the mistakes of the past and benefit from technological expertise
developed in higher energy consuming countries. For example, modern diesel engines in Europe are
much cleaner today than the higher emission engines used in the developing world. Natural gas is a much
cleaner fuel than oil and can replace liquid fuels for power generation and even transportation. Pipelines
may be a better way to transport aggregated fuel supplies than by road. Transport management measures
in urban areas such as dedicated lanes for public transport and strict enforcement of road safety rules can
help to balance the needs of urban transportation and the local environment.

In considering environmental impacts and fossil fuel use, greenhouse gas (GHG) emissions are a key
concern. At the global level, there is growing evidence that continued use of fossil fuels may lead to
unpredictable and possibly devastating effects on climate. This evidence also suggests that the ―winners‖
and ―losers‖ in this process are uncorrelated to those who emit the most and those who do not. In this
discussion, it is essential to draw a distinction between fossil fuel use in the poorest, least developed
countries—where energy consumption and GHG emissions are low both per capita and in aggregate, and
the primary concern will be the local environment—as compared with countries such as China, India,
Brazil and South Africa—where energy consumption is generally one or two orders of magnitude higher
per capita, and aggregate use may affect the environment beyond the local scale. Much of the growth in
fossil fuel consumption in developing countries will come from economy-driven growth to meet
industrial and transportation needs. As acknowledged by IPCC (1996) and UNFCCC (1992), this growth
will permit economic transitions, which will decrease the vulnerability that arises from their present high
dependence on land-based production activities—such as agriculture and fisheries. With the exception of
South Africa, the existing fossil fuel consumption levels in sub-Saharan Africa (SSA) are so low that even
if they increased at an annual rate of 10% (the annual rate at which China’s consumption grew during the
1971 –1997 period), by 2015 the associated per capita GHG emissions will remain at levels that are at
less than 5% of those in the high-income countries today.
  Baier, E., 1997. The impact of HIV/AIDS on rural households/communities and the need for multisectoral
prevention and mitigation strategies to combat the epidemic in rural areas. Rome: FAO.

The distinction becomes even greater when the least developed countries are compared to industrialized
nations. By way of example, United States CO2 emissions per capita are nearly 200 times those of
Ethiopia. This is due largely to the fact that Ethiopia’s cooking fuel consists predominantly of biomass,
and much of its electricity is produced from hydropower (both nominally renewable resources). So even
if Ethiopia’s fossil fuel consumption were to increase six-fold by year 2015 (this would imply a rate of
growth far higher than that achieved anytime in even the fastest growing countries) its per capita CO2
emissions would still be thirty times less than that of the United States. Countries that experienced
dramatic economic growth in the last two centuries can now afford to make large-scale investments in
modern energy services that do not rely on fossil fuels, biomass or hydroelectric power. It is these
countries that bear the primary responsibility for reducing GHG emissions and hence must primarily share
the burden of addressing the problem.

Overall, the increased GHG emissions from SSA are not likely to have any significant impact on the
world climate, nor have any mechanisms been identified that suggest that the region’s own GHG
emissions will have feedback effects on climate locally. And considering emissions more broadly, natural
gas and LPG produce far less potentially detrimental emissions, at least to the immediate user, than wood
or dung. Furthermore, a switch to modern cooking fuels would not be limited by world resources or by
GHG emissions, as recognized by Kirk Smith’s observation47 that ―even if all 2 billion people shifted to
LPG for household fuel, it would add less than 2% to global greenhouse gas (GHG) emissions from fossil
fuels. In terms of human health, a shift to LPG would actually result in a net reduction of human
exposures to air pollution that would be substantially larger than today’s total exposure from all fossil fuel

There may, however, be significant impacts of global GHG emission triggered climate change on SSA’s
energy sector due to the region’s high reliance on biomass and hydropower. For example, changes in
precipitation can be felt through loss or variability in hydroelectric potential, variations in runoff which
can lead to siltation (with subsequent impact on hydroelectric power generation), and impacts on
fuelwood (and hence charcoal) production, which represents the dominant sources of primary energy in
the household sector.

In some circumstances it may be prudent for the poorest to consider leapfrogging to newer and/or
renewable technologies even if they may not be strictly cost-competitive at present. Such circumstances
may arise when other factors such as cost of reliance on imports, domestic job growth, involvement of
local supply chains and effects on the environment and tourism are taken into account. Goldemberg48
(2003) argues that Brazil’s commitment to ethanol from biomass—a move that involved high initial
costs—did eventually pay off as costs fell with increasing production and associated learning rates.
Similar long term benefits from investment in distributed generation and/or biomass-based technologies
may accrue in the case of other countries as well. An analysis of these benefits does need to be carried
out however before one can recommend such an approach. But in the immediate future—at least for the
next few decades, particularly in the case of transportation and cooking fuels—fossil fuel based options
appear to be the most cost-effective way for the least developed countries to meet their energy needs.

Another important question from the standpoint of environmental sustainability is whether fuelwood
collection—for use in rural areas, for sale in urban areas or for charcoal production—causes or

   Smith K., ―In Praise of Petroleum?‖ Science, by Smith 298 (5600): 1847, 2002.
   Goldemberg, J., S. T. Coelho, et al. (2004). "How adequate policies can push renewables." Energy Policy 32(9):

contributes to deforestation. Most past studies49 report that more forest destruction has been caused by
conversion to agricultural use and through the nationalization of forests—denying unorganized,
underprivileged people, particularly women, access to natural resources. Depletion of the forests may
also result from timber sales or commercial charcoal production. But by and large it appears that, rather
than fuelwood being the cause of deforestation, fuelwood scarcity is a result of deforestation caused by
other factors. It may happen that when a forest resource is severely depleted—due to reasons other than
fuel wood collection—continued fuelwood collection may indeed make the problem worse. This
circumstance, along with already depleted soils, may create a situation where biomass waste or dung that
could help restore the soil is used as cooking fuel. Under these circumstances, interventions with modern
fuels may offer a way out of this trap.

Mahiri and Howorth,50 in reviewing two decades of fuelwood crisis in Kenya, confirm the local
specificity of forest depletion. In Kenya, they suggest that the problem of indiscriminately felling trees
and resulting environmental degradation may exist near urban centers. However, in rural areas the
situation is intricately tied in with land use and control, where those who can consolidate land by building
a fence are able to increase their wood supply, but those who cannot have declining access to fuelwood.
Geist and Lambin51 in a recent meta-study analyze the frequency of proximate causes and underlying
driving forces of deforestation, including their interactions, as reported in 152 sub-national case studies
from Asia (55 studies), Africa (19 studies) and Latin America (78 studies). Only in Africa does wood
extraction for fuelwood occur with any significance as a possible cause of tropical deforestation.

Urban areas: Nuclei of growth, peri-urban and slum dwellers

City size and density have been important factors in determining social complexity and technological
evolution.52 The size and density of a city enable specialization, bring together people and lead to new
businesses and products. To ensure that these engines of growth flourish it is important that their unique
technical problems of communication, transport, and energy are addressed. Many of the businesses that
rely on newer communication and computer technologies demand very reliable and ―clean‖ electric
power supply. For these businesses to remain competitive in the global marketplace it is critical that
quality of electricity infrastructure is not sacrificed for the sake of cost. In a sample survey of private
businesses in 69 countries,53 the effects of poor infrastructure respondents listed most often were loss of
productivity and impediments to new investment and business creation. High population densities in

    e.g. Sarin, Improved stoves, women, and domestic energy, Environment and urbanization, 1991, 3(2), 51-56;
Leach, G. and R. Mearns (1988). Beyond the Woodfuel Crises: people, land, and trees in Africa. London, Earthscan;
UNDP (2000). World Energy Assessment: Energy and the Challenge of Sustainability (New York, United Nations
Development Program [] ); the recent CIFOR report
emphasizes the importance of population growth and includes a regional analysis of the fuel-wood issues.
   in Land Degradation and Development, vol 12, pp. 205-215, 2001.
    Helmut J. Geist And Eric F. Lambin Proximate Causes and Underlying Driving Forces of Tropical Deforestation,
BioScience ,Vol. 52, pp 143, February 2002.
   AUSUBEL, J.H., The Rockefeller University, New York, New York 10021-6399 (pp. 1-21 in Cities and Their Vital Systems: Infrastructure, Past, Present,
and Future, J.H. Ausubel and R. Herman, eds., National Academy, Washington DC, 1988.) If this isn’t right, the
author’s full list of pubs is available here:
   Brunetti, Aymo and Kisunko, Gregory and Weder, Beatrice (1997) Institutional Obstacles do Doing Business –
Region-by-Region Results from a Worldwide Survey of the Private Sector, Policy Research Working Paper No.
1759, World Bank.

urban areas lead to higher densities of energy consumption and will allow economies of scale54.
According to the UNDP urbanization is increasing in the developing world, and, of the world’s 1.3 billion
poor people, about 40 percent lived in peri-urban agglomerations in 1999. The low cost of connecting the
peri-urban to the grid can be exploited as an advantage leading to a win-win situation for the consumers
and the utilities55.

It should be easier to distribute cooking fuels such as LPG—a fuel with higher effective energy density in
terms of both volume and weight per unit energy into the pot—than to transport fuelwood and charcoal to
the city. In fact, in many urban areas of sub-Saharan Africa (e.g. Accra in Ghana) charcoal and fuel-wood
are almost the same in cost per unit energy into the pot56. Thus urban areas permit a lower barrier to entry
for modern cooking fuels if upfront costs can be subsidized and fuel payments, which are often lumpy
relative to income, can be spread out. Increased use of LPG in urban areas can also allow larger bulk
imports of LPG thus reducing unit costs, helping to pay for import and storage infrastructure and reducing
the high ―overhead‖ that would otherwise be associated with lower total LPG demand.

Irregular tenure to dwellings, shared spaces, ill-defined responsibilities for payment and low electricity
consumption levels have sometimes made it difficult for utilities to serve slum dwellers. Slum dwellers
are also frequently ignored and bypassed in favor of rural populations in spite of their active participation
in the economic growth of the city. The urban poor who do not have electricity tend to pay high prices
both for relatively poor kerosene based light and for low quality biomass cooking fuels.

   Marchetti, C. 1975. Transport and storage of energy: RR-75-038. Laxenburg, Austria: International Institute for
Applied Systems Analysis.
   ESMAP Report: Peri-Urban Electricity Consumers—A Forgotten but Important Group: What Can We Do to
Electrify Them? October 2001.
   Wisdom Ahiataku-Togobo, Ministry of Energy, ―Household Energy For Cooking In Ghana‖ Oct., 2002

Providing energy services for the MDGs: Assessing needs and planning for scale-up

MDG-based poverty reduction strategies

The core recommendation of the UN Millennium Project is to put the MDGs at the center of national and
international poverty reduction strategies through a series of specific and practical efforts by developing
countries and their development partners. The Project’s recommended strategy begins at the country
level. Every country in the world that suffers from extreme poverty – including middle-income countries
with pockets of extreme poverty – is recommended to adopt an MDG-based Poverty Reduction Strategy
(PRS). The MDG-based PRS should set a serious 10-year timetable to 2015, with policies, governance
strategies, and public investment plans. In order to achieve the MDGs, countries need to take the 2015
targets and time horizon seriously. This implies a major a shift in thinking and language: from planning
around a marginal expansion of social services and infrastructure to planning around a long-term
investment strategy to achieve the MDGs. The world thus needs to change its discussions from a focus on
―accelerating progress towards the Goals‖ to ―achieving the Goals.‖

The MDGs require the scaling up of key public investments in infrastructure, human capital,
empowerment of women, and environmental sustainability. Each specific MDG (poverty, hunger,
education, health, etc.) requires multiple public investments, which we group into ―investment clusters.‖
The Project has identified seven priority clusters, with details provided in the 10 Task Force Reports that
accompany the Global Plan:
    1. Health
    2. Education
    3. Rural Development
    4. Urban Development
    5. Gender Equality
    6. Science and Technology
    7. Cross-border Cooperation

As argued above, investments in improved energy services are critical for supporting each of the seven
priority clusters identified above. The Project is therefore recommending to explicitly include investments
in energy services as part of the rural and urban development strategies. The remainder of this document
focuses on identifying the need for improved energy services, and outlining strategies for delivering the
necessary investments in low-income countries.

As a first step towards developing MDG-based poverty reduction strategies, countries need to work
backwards from the MDG outcome targets to concretize the operational set and scale of public
interventions – including human resources, infrastructure, and financial resources – needed as inputs in
order to achieve the Goals by 2015. We call this quantification of required investments the ―MDG needs
assessment.‖ In almost every area of the MDGs, and notably in the goals involving hunger, mortality,
disease control, and access to basic infrastructure, a well-known and widely tested set of interventions is
available for a low-income country to meet the respective challenges. The terms ―costing‖ and ―needs
assessment‖ are often used interchangeably in the policy discussion even though they describe different
methodologies. We use the term ―needs assessments‖ to underscore the importance of not focusing
exclusively on the financial cost of achieving the MDGs when assessing a country’s needs. A national
MDG needs assessment should include three basic sections:

    1. An identification of the key interventions (including human resources, infrastructure, and
       physical inputs) a country needs to achieve the MDGs;

       2. A detailed assessment of the ―coverage targets‖ that the interventions need to reach as they are
          scaled up to meet the MDGs;
       3. An assessment of the resources – in terms of infrastructure, human resources, and financing –
          required to provide the necessary interventions at MDG scale;

A more detailed description of MDG needs assessments is provided in a background paper provided by
the Millennium Project.57

Energy needs assessment for the MDGs

Historically, the needs for infrastructure in the poorest countries have generally far exceeded what was
possible with available funds. In this climate, planning was done based on what investments are available
and not what are needed to achieve a specific goal. While it is difficult to ascertain with accuracy what
energy services are necessary to achieve the MDGs with precision, we make an attempt to provide a
framework for decision-making based on a needs assessment of energy services for meeting the MDGs.
Full details are provided in Appendix 1 and 2. Here we concentrate on outlining the key elements of an
energy needs assessment.

The needs assessment starts by defining targets for energy services that are to be achieved by 2015. They
include minimum household consumption of improved cooking fuels and electricity that is deemed
―MDG compatible‖. In addition, countries need to identify, which types of cooking fuels and stoves are
deemed compatible with meeting the health and environment objectives. Using these minimum
consumption targets countries can specify how many people will need to have access to ―MDG-
compatible‖ energy services by 2015 in order for poverty to be halved and other MDGs to be met. A
similar assessment needs to be carried out for social service institutions, including schools and health
facilities, as well as other key government facilities. Finally, the energy needs of the private sector,
including agriculture and food processing, need to be addressed as part of the needs assessment.

Such a ―planning approach‖ to providing access to energy services based on a detailed needs assessment
can allow for significant reductions in unit costs. The exercise allows one to:
        i) explore specific means of delivering services at low initial consumption levels (for example
             initial use of LPG may not fully replace other fuels and may benefit from smaller cylinder
        ii) develop specific mechanisms to target the poor, who would otherwise not avail of the service
             e.g. with LPG in Brazil or electricity in South Africa;
        iii) exploit demand nodes from the productive use sector that can pay for the service but could
             benefit from the convenient, clean, reliable and cost-effective service that offers unique
             possibilities for business expansion;
        iii) ensure unit cost reductions due to aggregation, lower transaction costs and cost reductions
             along entire supply chain (e.g. with lowering transport costs);
        iv) create the ability to invite competitive bidding from the private sector to lower the costs of
             infrastructure borne by the government (e.g. in Argentina with electric supply to the rural
        v) develop technical innovations or place specific technical standards (e.g. use of SWER and
             lower capacity standards for transformer sizing in Tunisia);
        vi) create partnerships that can allow a country or region-wide effort to eliminate barriers to entry
             through education and training, market organization, consumer finance and regulatory
             framework (e.g. the LPG Challenge); and

     Available online at

        vii) allow multi-sectoral synergies to be exploited.

Modern Cooking Fuels

The multitude of short and long-term benefits to health, productivity, agriculture, environment and
women’s welfare that modern cooking fuels provide suggests that a rapid scale-up of modern cooking
fuel distribution systems should be important part of the strategy to meet the MDGs. A combination of
top down interventions such as lowered transport and distribution costs (through road and port
infrastructure), improved handling and storage facilities at ports, bulk purchases of fuels, and impetus
from the government through suitable regulatory reform can assist in lowering the costs of LPG and
kerosene. In addition bottom-up approaches can encourage market development through provision of
smaller LPG cylinders, upfront payments that are spread out over a longer period and consumer
education. In areas where the bottleneck is upfront capital costs (for example for a LPG stove-cylinder or
for a new electrical connection) and not the ongoing costs, immediate opportunities exist for scale-up.

Perhaps even more than lighting, the upfront costs and the recurring costs of modern cooking fuels are
difficult to afford if the distribution and sale of these fuels is entirely on a cost basis. It is difficult for
households making less than $2/day (in real USD, corresponding to a per capita income in PPP terms of
about $1/day to $3/day) to afford the recurrent cost of fuel immediately, since household fuel costs alone
could be as much as $0.30/day (assuming daily consumption of 0.50 kg and LPG costs of $300/ton and
transport/filling/distribution costs of $300/ton). Suitably designed targeting to identify those, an exit
strategy from subsidies for those who consume and earn more, and bill collection schemes that minimize
cost of collection can ensure that subsidies are indeed reaching the poor with least leakage or waste.

The ―low hanging fruit‖ in reaching these populations is the substitution of solid biomass fuels with LPG
(or kerosene, which is not as clean burning but easier to transport). If on the basis of energy into the pot,
solid biomass fuel is nearly the same cost, then a market opportunity exists that may require an
intervention of the kind advocated by the LPG Challenge. This may not however be sufficient to reach
the requisite scale-up by the year 2015.

There is wide agreement on either a direct subsidy or a form of lease/finance mechanism for the initial
costs of the stove and the cylinder as well the desirability of a range of canisters sizes with particular
emphasis on easy to carry smaller canisters that can reduce lumpiness of payments. There is also
agreement on the desirability of the fuel itself from health, time-saving and environmental perspectives,
even though it a fossil fuel. Where is there is still considerable disagreement is regarding the role of
subsidies for recurrent costs of fuel. This question needs further scrutiny in light of two rather large-scale
programs for fuel subsidy, one in Brazil which is viewed as a success and the one in India which is
thought of as a failure.

Brazil’s energy strategy serves as a useful example of how a government subsidy program can positively
and dramatically affect the rate and extent of penetration of modern energy services. Jannuzzi58,59
presents data on the penetration of modern energy services, and concomitant decline of traditional fuel
use, in the Brazilian residential sector between 1960 and 1985. In this period, as electricity connection
charges were eased in urban areas for low-income households, connection rates rose, from 38% of
Brazilian households in 1960, to 80% by 1985. Over the same time, penetration of fuelwood and
kerosene fell from 61 and 20 percent, to 28 and 7 percent, respectively, indicating a shift away from these

   Jannuzzi, Gilberto De Martino. Residential energy demand in Brazil by income classes Issues for the energy
sector. Energy Policy, June 1989, p. 255.
   Jannuzzi, Gilberto De Martino. ―LPG subsidies in Brazil: an estimate‖ unpublished, provided by
correspondence, 2004.

fuels for cooking and lighting. Initial patterns of penetration tended to favor adoption of modern energy
services in urban areas, in specific geographical regions and by those of somewhat higher incomes.
Nonetheless, the overall data trends clearly indicate comprehensive improvement in use of modern energy

Similar trends for adoption of LPG in Brazil are described in the Box below. It is noteworthy that during
the 30 year period beginning in 1973, the inflation adjusted per capita subsidy (based on the entire
participating population) was less than $1/year. Brazil also transitioned from a collective subsidy to a
targeted subsidy for the poor—in effect an exit strategy once the broad goals of poverty reduction were
indeed achieved. Under the targeted subsidy plan, even though the subsidies are about $16/beneficiary (or
about $60/family) the program’s total outlay is smaller than under a broader subsidy. This is because
only those families with a monthly per capita income no more than one-half the minimum-wage qualify,
or about 20% of the population. At this low income level it would be difficult to imagine how one could
afford the unsubsidized cost of LPG.

Box 1: Liquefied petroleum gas subsidies in Brazil: 1973-2003

In Brazil, penetration of LPG services was aided substantially by government programs and subsidies for the past
three decades. Throughout this period, LPG subsidies maintained the cost of energy at essentially stable prices, in
real terms. Particularly for LPG, the results of the program were dramatic, allowing penetration of LPG (and city
gas) to rise from 18% nationwide in 1960, to 98% of households in 2004. The penetration in rural areas, at 93%, is
particularly impressive given the difficulty of reaching what are often low-density populations.

Jannuzzi’s description focuses on a three decade period between 1973 and 2003 during which an effective cross-
subsidy scheme played a key role in this rapid and comprehensive adoption of LPG. During the program’s initial
period, from 1973-2001, prices for LPG and other petroleum derivatives were administered by the central
government and kept uniform across the country for all customers. The LPG subsidy varied over time, averaging
around 18% of the retail price. Jannuzzi estimates the cumulative value of this portion of the program, corrected for
inflation and domestic LPG prices, at US$ 2.9 billion and (at an average per capita LPG consumption of 2 GJ, or
roughly 40 kg/yr), a per capita, inflation-adjusted annual subsidy of US$ 0.73.

In 2001, as part of a broader deregulation of markets for petroleum products, LPG prices were liberalized, and
collective subsidies were eliminated. This was accompanied establishment by a voucher program subsidizing only
those families with a monthly per capita income no more than one-half the minimum-wage income. (Currently one
minimum wage is roughly R$ 240, or US$ 76.50.) As of 2002, the number of families participating in the targeted
subsidy plan was 6.7 million (at a cost of US $349 million), and this increased to 7.9 million families—or roughly
20% of Brazilian population—in 2003 (at a cost of US$ 462 million). Considering an average family size of 3-4
people (IBGE, 2004), this averages to US$ 16 per capita annual subsidy cost.

Overall, the average per capita subsidy rose throughout the program—from approximately US$ 1 to $ 16, after price
liberalization. However, with liberalization in 2001, the net expenditure of the government decreased almost by half
due to the smaller number of participating, low income families. Overall, the author found these costs to be low
relative to the benefits of providing greater access to a better cooking fuel. Furthermore, after subsidies helped to
establish an LPG market in Brazil, it has been most sensible to redirect the subsidy toward only the low income
consumers who are most affected by changing prices.

Source: G. Jannuzzi, “LPG subsidies in Brazil: an estimate” unpublished, provided by correspondence.

LPG subsidies in India provide another instructive, though less successful, example. During the last
decade, the annual LPG subsidy in India has varied between $0.50/capita (in mid nineties) to $1.50/capita
(in 2002). However, the bulk of the LPG is consumed by about 30% of the population—by income,
nearly the top half and the top 20% of the urban and rural populations, respectively. This indicates that
the annual LPG subsidy per beneficiary in 2002 was about $5/beneficiary (or about $25/family) per year,
with the total annual subsidy being $1.3 billion. The leakage rates for LPG are quite low in India.

A comparison of government programs in India promoting use of ―clean‖ fuels, such as kerosene and
LPG, illustrates some differences in results from ongoing versus ―first cost‖ subsidies60. In the state of
Himachal Pradesh, subsidies for LPG, combined with subsidies for use of pressure cookers to increase
energy efficiency, resulted in greater penetration of clean fuels into rural areas. However, a universal
price subsidy may present problems in the long-term, such as continued high costs on the state
government as well as the fact that as much as 80-90% of the benefit of the subsidy may accrue to urban
households, rather than the rural poor.

The Deepam scheme undertaken recently in Andhra Pradesh, was an attempt to both relieve the burden of
ongoing, universal fuel subsidies on the state, while encouraging increased LPG use in targeted poor
populations, particularly women in rural areas. It is described in greater detail in Box 2.

Box 2: Deepam Scheme for encouraging LPG use in Andhra Pradesh, India

One focus of energy interventions in poor areas around the world has been to encourage women to switch their
primary cooking fuel from biomass—which has dramatic impacts on indoor air quality and requires substantial time for
fuel-gathering—to other cleaner and less labor-intensive forms of energy. The Deepam scheme, undertaken in the
Indian state of Andhra Pradesh, hopes to assist millions of poor families in making this switch by providing LPG
“connections” in the form of a refillable fuel canister, free of charge.

The primary objectives of the program are to improve the health and reduce the workload of women and children, who
are most affected by biomass fuel collection and use. Another key objective is to test the effectiveness of a subsidy
targeted only at the one-time capital costs of the initial transition to LPG use, and thus avoid the ongoing costs of fuel
subsidies to the state. The program’s goal is to provide 3 million canisters in total to low-income women,
predominantly in rural areas, who are members of community self-help groups. The canisters, which ordinarily cost
about 1,000 Rupees each (or $ 20-22 USD), are provided free, though recipients must themselves supply a roughly
equal investment for the stove and connection accessories necessary for a full LPG cooking system. Slightly more
than half this number had been distributed between the scheme’s initiation in summer of 1999 and spring of 2002.

Thus far, this approach has been effective in the distribution of millions of LPG canisters, and LPG use has increased,
but survey data indicates that the primary goal has not been achieved: the recurrent costs of LPG refills are too high
for the majority of families to truly switch from biomass to LPG as their primary cooking fuel. The program has,
however, been more successful in motivating fuel switching in urban areas where biomass fuel is less plentiful and
thus more expensive. Due to this lack of fuel switching, the beneficial outcomes of the Deepam scheme with respect
to the Millennium Development Goals—particularly health—are not seen at the desired scale. For example, women
who substitute LPG for biomass in cooking do report decreased eye and respiratory irritation, but these benefits are
necessarily limited by the extent of LPG uptake and use.

A final significant factor in the Deepam scheme is that it is occurring within an overall context of petroleum sector
deregulation. While the scheme is designed around a targeted, one-time subsidy, it nonetheless maintains an
ongoing, universal subsidy related to fuel. A World Bank study has noted that “At the state level, the Government of
AP surrenders 5 liters per month of PDS [Public Distribution System, i.e. subsidized] kerosene for every new Deepam
beneficiary enrolled in exchange for increased allocation of subsidized LPG by the Central Government to cover the
scheme” [emphasis added]. Thus, the program is, in essence, trading off one ongoing fuel subsidy for another. Also
the subsidy distribution system favors state-owned oil companies, rather than private sector LPG distributors, despite
the fact that the latter may have better market penetration in rural areas.

The Deepam scheme provides important lessons for planning energy interventions to advance the MDGs in that it
shows that in spite of the initial cost subsidy, the intervention may be unsuccessful if recurring costs remain too high.
The program still holds important possibilities, particularly if energy sector deregulation and the involvement of private-
sector vendors allows for dissemination of smaller canister sizes, through additional distribution outlets, and otherwise
takes advantages of opportunities and efficiencies that are not a current focus of public-sector LPG distributors.

Source: World Bank, ESMAP newsletter, Issue No. 6, March, 2002.

   Viswanathan, Brinda, K.S. Kavi Kumar. Cooking fuel use patterns in India: 1983-2000. Energy Policy. Article
in Press, 2003, p. 14.

The use of subsidies to encourage penetration of clean and modern cooking fuel has shown a range of
effectiveness. Examples from Brazil and India show both successful and unsuccessful elements, with
results varying according to income level, geographical region, and whether the setting is urban or rural.
The uneven effectiveness of particular programs, however, does not appear to argue against the use of
subsidies per se. Instead, it suggests that a variety of factors, including the manner in which subsidies
target specific populations, the percentage of a subsidy relative to both the fixed and recurring costs of the
energy service, and the institutional framework through which a subsidy is implemented, as well as other
factors, all play key roles in outcomes.

The example of subsidies used in Brazil show that both across-the-board and targeted subsidies can
advance the penetration of modern energy services to the poor, including those in rural areas. Depending
upon the depth of poverty for a given country, however, the percentage of the population in need of
subsidy support, as well as the size of that subsidy, will certainly vary. However, the Brazilian example,
particularly for LPG, shows that effective scale-up can occur, and be revised to become more efficient as
the program progresses.

Other interventions, aside from switching to LPG or kerosene, can help reduce household exposure to
indoor air pollution. The most widely implemented of these is the promotion of improved stoves. Not all
stoves with lower fuel-wood consumption however lead to lower harmful emissions. Development of
stoves combined with modifications such as chimneys, windows and eaves spaces and improved cooking
practices will need longer term sustained research effort. Efficient charcoal production practices—
including sustainable production of wood through wood lots, agroforestry, managed firewood
plantations—combined with improved charcoal stoves can use less fuelwood than a simple three-stone
fire, while also reducing the level of harmful emissions in the person cooking. Charcoal is a more
convenient cooking fuel than fuelwood, is already an accepted means of cooking and permits local
production and employment. The approach advocated here is that, in addition to the options of LPG,
kerosene and improved cookstoves, one also addresses improvements in current practice that can offer an
alternate possible cooking fuel, such as charcoal, even though this may be a transitional fuel towards
gaseous and liquid fuels.

This issue of cooking with biomass has future technological challenges as well. Many of the
technological advances in the field of energy during the last two centuries have had to do with taking poor
quality sources of energy (e.g. coal or biomass) and converting them into a form of energy that is of much
higher quality at least to the immediate user. Developments in technology may make it possible to do this
with cooking fuels and electricity at a household or community scale even though this is often not cost
effective at a household scale.

A public-private partnership between the UNDP and the World LP Gas Association, called the LP Gas
Rural Energy Challenge, is addressing the problem of high first costs in rural areas (these can be as high
as $45 to $70). The description here is abstracted from the UNDP document ―LP Gas Rural Energy
Challenge‖. The program points to the importance of developing partnerships that draw on the strengths
of the public sector (village or district government, national, international); the private sector (SMEs,
companies and WLPGA61); and local institutions (community organizations, NGOs) to create viable new
markets. Approaches to address the first cost problem include bank financing, manufacture and
distribution of smaller LP gas cylinders and subsidies from host governments for assisting narrowly
targeted new rural users. The program illustrates the considerations that form the basis of a ―business
plan,‖ such as financing, community involvement in repayment of loans, use of Energy Service
Companies (ESCOs) that provide energy to customers on a fee-for-service basis, local partners that can

     World Liquid Petroleum Gas Association

assist with local labor resources and development of production and distribution chains, and working
closely with LP Gas companies that can set up sales organizations, conduct market research and identify
the services that customers want and can afford.


There is wide agreement on the importance of electrification to the development of productive activities
as well for health, educational, commercial and ICT development. Electricity is an ideal medium for end-
uses such as lighting, motive-power (e.g. for agro-processing or irrigation) and for communication
devices. However, it is difficult technologically to produce electricity at scales smaller than a few kW
except with newer technologies such as solar photovoltaics (PV). There is also agreement on the need to
replace the widespread use of kerosene wick lamps, disposable alkaline cells and poorly charged lead-
acid batteries for lighting since the effective unit cost of lighting (e.g. in units of $/kilolumen-hr) using
these means is nearly one to two orders of magnitude higher than the most expensive thermal generation
of electricity.

The ―low hanging fruit‖ as regards to electricity is addressing the problem of low levels of penetration in
areas where grid electricity is already present. Here the barrier to entry is the upfront cost associated with
bringing the wire to a home. In urban, peri-urban areas and slums these costs are low and with
aggregation could be further lowered. This issue is discussed in greater detail in Appendix 2.

The challenge for electricity service delivery (or lighting and ICT service delivery) is in rural areas where
population densities as well as consumption levels are lower. In these areas the initial costs of grid
electrification can be as high as $1000/household. One strategy for reaching these populations is to
exclusively promote decentralized solutions that are based on RETs, a leapfrogging approach in the sense
that it bypasses fossil fuel based technologies. However, these technologies tend to have high unit costs
or limited geographic availability. Another is to strengthen transitional low-cost solutions that are already
being used by the poor. An example is diesel-powered mini-grids which can provide points for charging
of batteries that are then carried to households. These transitional solutions will provide rural areas with
an increased range of options to choose from for immediate scale-up until either grid technologies
become viable or RETs become more affordable. Another example is the multi-functional platform
powered by a diesel engine for low-cost rural motive power (discussed in greater length below). The low
upfront costs of such solutions allow them to be easily replaced. The higher recurrent costs of such
technologies are less of an issue since consumption levels are likely to be low during this transition. For
example, generation and fuel costs (even though they vary with technology) of diesel based electricity do
not exceed $0.40/kWh if diesel costs are below $1/litre. Even at this high generation cost the recurrent
electricity costs would be lower than the recurrent cost of kerosene for lighting and would be affordable
by the poor.

Electrification for rural areas: importance of aggregation of demand

Aggregation of demand is a crucial factor in efforts to extend electricity services to poor households. A
combination of geographic and economic factors determine both the capital expenditures and recurrent
costs of extending power from a pre-existing electrical grid, establishing a local ―mini-grid‖ with its own
generation capacity, or creating a more distributed design—for example, household systems based on
solar PVs. Moreover, a needs-assessment based approach to planning can anticipate the provision of
services to not only those who can already afford to pay but also to the very poor, possibly with subsidies.
This can allow at the outset a higher penetration rate thus reducing unit costs for all.

The primary factors for predicting initial capital costs are geographic in nature, specifically the patterns of
overall population density, household spacing within a community, and distance between a community
and any pre-existing grid power. These factors determine the cost of transmission and distribution
infrastructure which is often the primary capital expense of any new grid-based system. Generation costs
are a small fraction of the total costs.

Population density per square kilometer—averaged over space—is a useful general guide in estimating
distances between households within a community and thus the cost of low voltage distribution lines.
However, equally important is whether households are concentrated (―nucleated‖) into villages or towns
or spread more evenly over the landscape, as one might find if each household were situated on a family
farm. Furthermore, the distance from the community to a pre-existing national or regional electrical grid
is a decisive factor in whether it is preferable for a community to purchase power from an external utility,
or to establish a local mini-grid and maintain its own power generation equipment. A combination of
economics and geography in aggregation of demand has, thus far, created an urban bias in the extension
of electricity services to the poor. Factors of higher population density, greater nucleation of households,
higher cash incomes, and lower availability of traditional (biomass) substitutes have generally favored
urban and peri-urban areas as the lowest-cost regions for early electrical grid expansion.

However, options do exist for the rural poor. Here we discuss various electricity technologies within
simplified settlement pattern modalities. The following sections describe a general approach to analyzing
and costing these options based on observed geographic and economic factors in specific rural

Rural areas worldwide show a range not only in average population density but also in the settlement
patterns (nucleated versus dispersed) seen at a given density. Even at high average population densities
(here we use densities of 300 persons/km2 as representative value), the settlement pattern can range from
highly dispersed, with each household located on a farm with little local aggregation (e.g. much of rural
East Africa), to a nucleated village pattern where most households are located in close proximity to each
other (e.g. much of rural India). At low population densities one could also have a range of settlement
patterns from dispersed households (e.g. in the African Sahel) to nucleated villages with their centers
spaced far apart (less arid parts of West Africa).

Of these settlement patterns, some are less easily reached by current systems than others, a crucial
concern in efforts to achieve the MDGs. The ―high density nucleated‖ rural populations are, in many
ways, similar to peri-urban areas, and so many have already benefited from grid electrification. Other
settlement modalities tend to present greater challenges. The people who live in ―high density dispersed‖
rural settlements represent a large fraction of the population in need of electricity services. Those who
live in ―low density dispersed‖ patterns represent perhaps the greatest challenge for electrification in that
neither population nor dwellings are concentrated together. In these circumstances of very disaggregated
demand, costs of providing services at the household level may be so high that a viable option to reach the
poor may be through aggregated public spaces, such as a schools or health clinics, which can also house
ICT services.

High density dispersed settlements

To anchor the discussion here we will talk about a part of Western Kenya that the author is familiar with
where there are about 1000 households in a 7.5/km2 area, without electricity and primarily using kerosene
wick lamps. This area is a hunger hotspot with 60 to 70% of the population living below the poverty line.
Most of the people living in the vicinity of Lake Victoria live at these high densities, in a dispersed

In this setting, it is useful to consider the basic minimum household electricity needs for lighting. Table 2,
below, illustrates the results of simple cost estimates for a variety of household lighting technologies with
different yearly usage levels. The options explored here include: portable batteries which would be
carried from public charging stations to households for use; a small electrical grid (―minigrid‖) with
diesel generation, transmission, and distribution to local households provided locally; solar PV home
systems installed separately for each household; and, finally, a full connection for each household to the
national or regional electricity grid. Note that all of these options include only use in households,
although one option (the portable batteries) would rely on charging at public use points (i.e. in schools,
health clinics or other community settings). This table illustrates the results of one approach to cost
estimation, system-level planning and selection of the lowest cost electricity technology for a specific
community. The upper portion of the table consists of estimated total costs for specific technologies,
summed over five- and ten-year periods. The lowest cost option for each level of consumption is bolded
and shaded, and, in the lower portion of the table, this cost is listed per household and per person. These
calculations exclude any costs of service for public use.

Table 2: Total costs for household electricity consumption summed over five- and ten-year intervals,
excluding public consumption.
                                                Costs per household ($US) for Given Levels of
                                                 Annual Household Electricity Consumption
                                                   10 kWh          100 kWh         500 kWh
   Grid Power (central generation at 5
 cents/kWh, MV at $15/m, LV at $5/m)
                              5 years                $   551             $    573         $    673
                             10 years                $   553             $    598         $    798

         Solar Photovoltaic (systems at
       $10/Wp and batteries 0.50/kWh)
                                5 years              $   399             $ 3,540          $17,499
                               10 years              $   474             $ 4,290          $21,249

     Minigrid (diesel $0.70/liter, genset
                  $1000/kW, wire $2/m)
                                  5 years            $   194             $    380         $ 1,209
                                10 years             $   206             $    505         $ 1,834

    Portable Battery. Grid or generator
         charging (higher cost shown)
                                5 years              $    94                 NA               NA
                               10 years              $   162                 NA               NA

                                                Annual cost per household, lowest cost option
                                 5 years            $    19           $ 76         $ 135
                                10 years            $    16           $ 51         $     80
                                                   Annual cost/person, lowest cost option
                                                       assuming five person/household
                                 5 years             $    4           $ 15          $ 27
                                10 years             $    3           $ 10          $ 16
These results are estimates, based on known equipment, generation and other costs, as well as specific geographic
circumstances, such household spacing and distance from the main electricity grid. As estimates, they are intended not as
specific technology recommendations, but rather as an illustration of a method for selecting appropriate technologies
based upon conditions which can be adjusted for specific geographies and cost environments. Some assumptions used
here include: low voltage line at $5/m; medium voltage lines at $15/m; a distance of 1,000 meters from a grid connection
point in the community to the national/regional electricity grid; a total of 1,000 households in a 7.5 square kilometer area,
giving a density of 133 households/km2; a connection cost of $100 per house. “NA” indicates a technology is not likely to
be practical at a given consumption level.

Under the minimum lighting specifications, as described in the introductory section of this paper, one
could replace kerosene lamps with CFL based lighting and a radio for three to four hours every evening
with as little as 10 kWh of electricity per year. These are consumption levels more typical of systems
relying on charged batteries than grid-based electricity. Moreover, grid electrification would cost nearly
$100/capita even following best practices. While grid electricity is desirable in the long run, this is
frequently, at present, not a viable option for the poor. Hence portable batteries, a transitional solution,
would be feasible for a few dollars per capita annual costs. Some or all of the costs of batteries could be
borne by the households that already spend nearly $12/year on kerosene for wick lamps. Providing
support for such transitional energy systems would require the provision, possibly by the government, of
well designed and safe charging stations. These charging stations would be geographically distributed and
located at public facilities such as community schools and health clinics, whose baseline demand for

electricity they would also meet. Alternately this could be done through a business model operating as a
SME that would serve both the public facilities as well as the household needs with the appropriate
financing for the public consumption provided through the government. While there are numerous small
businesses that already provide charging services as well mini-grid operators running on diesel gensets,
the lack of institutional support (technical and financial) has not allowed this market to develop its full

Technical support would ensure that batteries are lightweight (a rechargeable nickel metal hydride
(NiMH) battery pack for the purpose described above could be as light as 0.5 kg), efficient charge control
does not reduce battery life (500 cycles is easily feasible), quality design and choice of lamps and fixtures
permit the maximum lighting benefit from a given charge (250 lumens of light is a good guideline for
illuminating a 10 foot diameter spot on the floor for reading), and that reliable low-maintenance systems
for either generating electricity locally or for grid connections to the charging stations are available. It is
critical to recognize that because of its low lighting efficiency, a kerosene wick lamp produces nearly one
order of magnitude lower light output per unit of power consumed (lumen/Wth) compared with a CFL
lamp operated by a portable battery that is charged by an inefficient generator.

There are many advantages of a stand-alone solar home system (SHS) or a grid connection at home. Both
are far more convenient than a portable battery. Emerging technologies could reduce the cost of solar
panels and related components. Some would be able to afford the higher costs of SHS or grid regardless.
Supportive policies such as tax incentives, technical standardization and possible CDM funding would
encourage this market development, which in turn would eventually reduce the costs of these options for
the poor as well. Where grid electrification has occurred, lifeline rates have been used to assist the poor
with at low consumption levels. The level of consumption below which lifeline rates apply have however
been frequently set too high, effectively subsidizing higher income consumers. This in turn makes it
difficult for the utility to meet its own generation costs and thus jeopardizes service to the truly needy.

The costing exercise presented above also demonstrates that as soon as consumption levels grow to 100
kWh/year and beyond (still a very low level of consumption), the cost of RETs such as solar PV could
become prohibitive. That is, the cost of solar PV would not be much lower than the cost of a grid
connection, and these sunk costs would be lost if indeed the consumption were to grow to 100 kWh/year,
outstripping the capacity of a typical PV home system. This suggests that it may not be worthwhile to
make large investments in a system that would be sufficient for, at most, the next few years.

The electricity system in Urambo Village, Tanzania (see Box 3 below) represents a case where an
independent rural co-operative is successfully operating a diesel mini-grid at a consumption level that
averages about 400 kWh per connection per year. The minigrid serves a village with a dispersed
settlement pattern, with 250 co-op members requiring 11 km of transmission line. At these consumption
and cost levels only about 10% of the community has electricity access. In this setting, a possibility for
extending service beyond this 10% and on to the poorest in the community include a state supported co-
op operated program which could either provide a life-line rate with limited consumption (they mention 2
to 3 kWh/month) or low cost battery charging services.

Box 3: Electricity Services in Urambo Village, Tanzania

While rural electrification efforts have been ongoing in Tanzania for over 40 years, it is estimated that, as of 2001,
only 2% of the nation’s rural households had access to electricity services. As a strategy for accelerating rural
electrification, a study conducted from 1989-1991 by Tanzania’s national electric utility (TANESCO) and the
Stockholm Environmental Institute (SEI) recommended the creation of rural electricity cooperatives that would be
managed independently of the national utility, setting their own tariffs, and either operating their own generators or
purchasing power in bulk from the state utility to supply remote or otherwise underserved consumers.

The first such cooperative in Tanzania, the Urambo Electric Consumers Co-operative Society (UECCO)—based in
Urambo Village in the Tabora region—was established in 1993 by the Urambo District Council with financial,
technical and other assistance from the SEI, TANESCO and the Swedish International Development Agency (Sida).
The cooperative’s earliest tasks were to rehabilitate the infrastructure of the pre-existing, but insolvent, local electricity
system created by TANESCO in 1985 and to establish the financial and institutional foundation necessary for the
cooperative’s sustained operation. This included a variety of tasks. Some were technological, such as repair and/or
replacement of three diesel generator sets (two 85 kW and one 108 kW) and the distribution system’s ~11 km of
power lines. Others were procedural and institutional, such as the writing of by-laws and creation of a development
committee to expand the consumer base. Still others were financial and administrative, including the hiring and
training of local employees, and setting share prices, connection fees and tariffs.

UECCO began operation in 1994, and by 2002, the coop employed three workers (two technicians and one
accountant) and provided electricity service to an estimated 2000 persons, or 10% of the village. The cooperative
consisted of approximately 250 cooperative members, most of which are households. Customers maintained an
average consumption (electrical load) per connection of ~250W, for around 4 hours per day, totaling around 35 kWh
per month. Throughout, the cooperative succeeded in most key aspects of the system’s operation and management.
Technological components, such as maintenance of the power plant, additional connections, and metering of the
majority of connections, were handled by the system’s two plant workers. The cooperative also enhanced its
institutional capacity by slightly expanding its development committee and holding periodic meetings to adjust tariffs.
Financial and administrative tasks, such as record keeping, billing, and managing income from tariffs and
expenditures (for fuel, maintenance, salaries, etc.) were performed primarily by the cooperative’s sole accountant.

The system has maintained operation for nearly 10 years, serving an expanding customer base through a growing
distribution system. The coop’s service is reliable and efficient, with electricity provided for 97% scheduled hours in
2002, and a diesel-to-electricity conversion efficiency equal to or better than that achieved by TANESCO’s
comparable plants. The service has relatively fair and effectively managed finances, in which most tariffs for most
customers are billed according to consumption, metered at the household level. In fact, given the choice between a
flat-rate system, with circuit-breakers to ensure individual households do not consume above a certain limit, and a
system with individual meters for each household, the consumers chose the latter. Through this system, the coop’s
tariff income generally meets its operating expenses, which consist primarily (80%) fuel costs, with the remainder
going to maintenance and salaries.

Still, important challenges remain: Committee meetings, at which tariffs are established or changed, tend to be far
too infrequent to effectively respond to wide swings in fuel costs, leading to budget deficits for the cooperative. Also,
though tariffs effectively cover most of the day-to-day expenses of the cooperative, they do not recover sufficient
capital for large scale, occasional repairs. Generator breakdown and other similar events have required additional
donor assistance. Finally, there remain some users who, due to lack of available meters at the time of their
connection, are charged a flat-rate, and are probably substantially over-consuming relative to their tariff payments.

Still, in a day-to-day sense, the system is largely technologically and financially viable, even at a penetration rate of
only 10% within the village. It is reasonable to predict that unit costs would drop with a higher penetration rate, and
that more households would participate if financial support were available. The possibility of establishing a “lifeline”
service, as low as 2-3 kWh per month may be an option, though users have, in the past, preferred metered
consumption to flat-rates with an upper limit on consumption.

As is, at least 10% of the village inhabitants appear able to afford both the purchase of a share in the cooperative, as
well as the recurring costs, as they are currently distributed over the base of consumers. It is noteworthy that
UECCO’s tariff as of October, 2002, was 0.47 USD/kWh, more than ten times the rate for electricity supplied through
TANESCO’s grid. Authors note that the viability of UECCO’s system at this price indicates the likelihood that
TANESCO’s service is perhaps too heavily subsidized.

Source: Ilskog, E., et al, Electrification co-operatives bring new light to rural Tanzania, Energy Policy, in

High density nucleated settlements

Nucleated villages like those in rural India would require considerably lower upfront costs for grid
electrification since the households are spaced closely. In fact, a government funded rural electrification
effort now reaches more than 95% of the villages of India. The low quality of service to the poor in India
(in spite of the high number of rural communities with grid access, a significant portion of the rural
population continues to use kerosene for lighting) has much to do subsidy structures that did not allow
state-run utilities to recover the costs of generation undermining service to the poor, as discussed earlier.
Another reason for the low quality service to the poor in spite of government programs to bring them
electricity access is the high (relative to the consumption) cost of bill collection. Innovations such as those
used in South Africa using prepaid cards can allow the collection costs to be low and subsidies to be well

Low Density (both nucleated and dispersed) Settlements

Many parts of Africa do not have the high average population densities of the Lake Victoria region
discussed above. We use an arbitrary value of 100 persons/km2 and below as a guideline for low density
areas. This corresponds to about 10 to 20 households/km2. However, it is crucial to note that this
conventional measure of density, which averages over space, hides any patterns of clustering that may be
present. That is, for a given area at a given average population or household density, there may be an
extreme clustering pattern in which hundreds of households live in a nucleated village (e.g. several
hundred households living in a cluster that is 5 to 10 km from another cluster). Or, at the other extreme of
the same average density, a dispersed settlement pattern prevails, with little village-style aggregation.

In the former scenario, some of the electrification approaches discussed earlier may be appropriate. In the
latter, more dispersed case, it is unlikely that grid based solutions, targeting higher consumption levels,
will be immediately available for the poor. It may be difficult to find sufficient degree of aggregation to
meet the household electricity needs with any technologies but solar PV. Considering a ten-year time
horizon for basic-lighting level consumptions, a SHS for basic lighting would cost about $10/capita per

If these costs are deemed too high, or other priorities do not justify this expense for household lighting
alone, a possible means to meet other fundamental energy needs of the poor is to ensure that basic social
services such as health, education, agriculture/health extension and, where needed, public facilities for
improved water supply/purification have access to electricity and/or motive power. Such electric access
could also be provided in conjunction with productive use functions such as agro-processing as has been
done with the multi-functional platforms (MFP) in West Africa. Education and health facilities could be
powered by stand-alone systems that could be based on either solar PV or, where consumption levels are
higher and access to fuel is feasible, on generators. Generators may also permit use for other productive
functions on a fee for service basis.

Table 3, on the following page, is similar to Table 4 in that it illustrates an approach to estimating the
lowest cost electricity technology for different levels of total annual consumption. However, it considers
use only in public contexts, including schools and health clinics. The upper portion of Table 3 shows cost
estimates for this public use, on a per household basis. The lower portion of the table provides the annual
estimated cost per household and per capita for the lowest cost technology. Note that, although these
calculations exclude costs for household electricity provision.

Table 3: Total costs for public electricity consumption summed over five- and ten-year intervals, excluding
household consumption (assumptions described in text).
                                                  Costs ($US) of Public Annual Electricity Consumption
                                                         Divided Across All Households Served
Technologies:                                    1,000 kWh             5,000 kWh              10,000 kWh
  Grid Power: (central generation at 5
   cents/kWh, MV at $15/, LV at $5/m)
                              5 years               $ 45                      $ 49                    $    54
                             10 years               $ 46                      $ 54                    $    64

Solar Photovoltaic (systems at $10/Wp
              and batteries 0.50/kWh)
                               5 years              $ 48                     $ 238                    $ 475
                              10 years              $ 55                     $ 275                    $ 550

     Minigrid (diesel $0.70/liter, genset
                 $1000/kW, wire $2/m):
                                  5 years           $ 19                      $ 59                    $ 108
                                10 years            $ 24                      $ 84                    $ 158

                                                      Annual cost per household, lowest cost option
                                5 years             $ 4                  $ 10                  $ 11
                               10 years             $ 2                   $ 5                  $     6
                                                  Annual cost/person, lowest cost option, assuming five
                                5 years             <$1                   $ 2                   $ 2
                               10 years            < $ 0.50               $ 1                   $ 1
As in table 2, these results are estimates based on known costs and realistic geographic and infrastructure conditions.
They are intended to illustrate of a method for technology selection based upon specific conditions that will vary on a
case-by-case basis. Some of the assumptions used for these calculations include: low voltage line at $5/m; medium voltage
lines at $15/m; a distance of one km from a connection point in a community to the national/regional electricity grid; and
1,000 households for the community. The household density and connection cost per house are not relevant, since only
service to community centers (schools and clinics) is considered. The relevant assumptions for public use are: 4
community centers per 1000 households, at a connection cost of $500 each.

Renewable energy technologies, particularly solar PV (now available at scales as small as 35 Watt panels)
with battery storage may be the least-cost option where power demands are small and fuel is hard to
obtain reliably. A wealth of knowledge is now accumulating regarding best practices in structuring the
role of government and donors, growth and nurture of service provider institutions and NGOs, lowering
or eliminating tax burdens, standardization and certification of systems, supporting programs for the
training in design, maintenance and safe use of these systems and credit and delivery mechanisms.
Development of business models where customers pay for electricity use rather than equipment and the
supplier retains the flexibility of using the best suited technology and cost structure for the local
conditions may provide opportunities for scale-up. This way, as demand increases, the service providers
could move the hardware to new locations. Such scale-up will require people who understand the local
conditions and culture, can manage businesses, understand the complexities of developing cost estimates
and negotiate with local and government officials.

Decentralized energy systems may also permit use of local biomass to produce electricity locally (e.g.
through biomass gasification). If such systems can be scaled down to one or few households and can be
low-maintenance, they hold the potential for becoming the source of electricity for rural areas. For

example a prototype of a low-maintenance, external combustion, 500W Stirling engine that can also
purify water was demonstrated recently by Dean Kamen. If the costs of such systems can be driven down
with mass production the use of biomass for electricity could provide ―dispatchable‖ power, thus
eliminating the need for batteries and the unit can easily be placed at a distance from the home to prevent
smoke and particulates from entering the house. It is electricity that is harder to transmit in small
quantities over long distances and thus using biomass for electricity (rather than for cooking) and using
modern fuels that are more easily transported in small quantities and are ―divisible‖ for cooking (rather
than for electricity) may make more sense. Generation of electricity can be easily placed at a distance
from the house whereas cooking is an activity that is generally carried out within a home.

Urban Areas

Reaching the urban poor remains a challenge and will likely grow with increasing urbanization.
However, because of higher levels of aggregation and other reasons mentioned earlier, reaching the urban
poor with electricity, cooking fuels and transport services often represents an easier task than providing
those same services to the rural poor. Efforts to reduce the barriers to entry for services, such as
regularization of slum dwellers for housing security, use of smart cards for prepaid electricity services for
a fixed number of kWh, provision of smaller LPG cylinders and improved public transport can address
the problem of reaching the urban poor. A detailed discussion of the low costs of urban electrification is
provided in Appendix 2.

The growth of secondary cities also presents sufficient density to allow for inexpensive grid electrification
for substantial populations, offering another means to provide electricity services to a greater proportion
of a nation at lower cost. The percent population that lives in cities in excess of 50,000 in some selected
countries is shown below.
Table 4: Settlements with populations greater than 50,000 persons, by country, with estimated MV line
length (or length of paved road) needed to connect the settlements. (Source: CIESIN)

                 Number                                             Total National    % of total
            secondary cities Total Population in   Area   MV line    Population     population in
Country     or larger ( > 50K)  these cities      (km ) length (km)      1999      secondary cities
Ethiopia            19            4,911,529      1104300   4,581     61,672,000          8%
 Ghana              21            4,824,098      238533    2,238     19,678,000         25%
Tanzania            44            8,759,890      883749    6,236     32,793,000         27%
 Kenya               9            3,850,313      580367    2,285     29,549,000         13%
 Niger               8            1,748,978      1267000   3,184     10,400,000         17%
Senegal             13            3,976,522      196722    1,599      9,278,617         43%
 Chad                4             941,973       1284000   2,266      7,458,000         13%

Some Examples of Scale-up and Lowered unit costs

South Africa and Tunisia (among other countries) carried out national electrification programs at
remarkable rates, in a cost-effective manner. The details of these programs offer valuable lessons
regarding innovative approaches to serving low-consumption consumers in rural environments.

South Africa

Electrification in South Africa grew from about 36% of households in 1990 to 67% in 2000, with more
than 3 million new customers. Gaunt reports on meeting electrification's social objectives in South
Africa, and implications for developing countries62. A key observation here is the dramatic reductions in
the capital investment costs per customer in rural electrification, suggesting that appropriately planned
rural systems need not be much more expensive than urban systems. Between 1996 and 2001, the
national average cost per rural electric connection decreased by 40% in current terms and 70% after
taking into account inflation, eventually becoming the same as an urban connection cost. The savings
were achieved by adopting designs that match the network technology and capacity more closely to the
requirements of the customers, (greater application of single-phase instead of the traditional three-phase
distribution at medium and low voltage), broad application of prepayment metering, and revised industry
standards and implementation procedures. Using low capacity, low cost grid connections, South Africa's
rural electrification program can supply substantially more energy than PV systems for a similar or lower

The South African experience with prepayment metering, is a development of significant note since this
can allow the consumers to purchase a service in small quantities and at the same time ensure low cost of
bill collection, dramatically reducing the fraction of costs that are purely administrative for servicing a
household with low consumption.


A recent program undertaken by Tunisia Electricity and Gas Company (STEG) achieved similar
success.63 As part of a multi-sectoral approach to extension of a range of infrastructure and services to
rural communities, Tunisia’s electrification program expanded services from 6% of the population in
1976 to 88% in 2001, including bringing electricity to 35% of the people in living in rural areas by 2001.
The technical strategy of the program was to use a combination of three-phase and single-phase power
lines, preferentially extending less-expensive single-phase wire to rural communities. Between 1981 and
2001, the percentage of LV lines relative to the total system lines (including LV and MV) rose from 16%
to 51%. In roughly the same period (from 1980 – 2002), the percentage of single-phase stations relative
to the system’s total rose from 20% to 72%. This saved an estimated 30-40% over the cost of MV lines,
15-20% on MV/LV substations, and between 18-24% on the system overall relative to what the same
expansion would have cost with the previous ratio of MV and LV lines. In addition, the program carried
out other aggressive cost-cutting measures, such as use of single wire earth return (SWER) design, shorter
poles (saving 20% on cost), equipment standardization, and bulk buying. Administrative innovations,
such as decentralized planning and improved corporate management practices, contributed to efficiency at
the institutional level. Key economic factors also played a role, including adjusting tariffs frequently to
keep the utility’s budget in balance; basing of tariffs on long-run marginal costs (lower for HV, industrial
users; higher for LV, household users); using urban tariffs to cross-subsidize rural rates, and establishing

   Gaunt, C. T. "Meeting electrification's social objectives in South Africa, and implications for developing
countries." Energy Policy In Press, Corrected Proof.
   Cecelski, Elizabeth, et al. ―Low Cost Electricity and Multisector Development in Rural Tunisia,‖ World Bank
Energy Lecture Series, Washington DC, February 11, 2004.

peak, off-peak, and ―lifeline‖ tariff levels (capping the latter at < 50 kWh/month). As a indication of the
program’s success, Tunisia has set rural electrification for all as a minimum standard for public service,
and has set a goal of 100% electrification, through a variety of grid-based and off-grid technologies, by

Lowering Unit Costs

As alluded to in programs in South Africa and Tunisia, lower transmission line costs can significantly
impact the costs of rural electrification. Figure 9 from an ESMAP study64 shows a comparison of Medium
Voltage line costs per km varying from $2000 for India to $18,000 for Mali for the same configuration.
What is clear is that there are considerable variations in materials and labor costs across countries. Some
of the variation is because of availability of low cost materials (such as inexpensive of cement in India for
concrete poles), some are due to varying standards, and some due to topography and transport. However
bulk purchase and lowered transport costs alone could lead to significant savings. A combination of high
penetration rates ensured by government policies and lowered line costs can have dramatic effects on the
cost of a new connection.

Figure 9: Variation in Total Three-phase Line Cost (Labor and Materials) by country.
Figure 10 illustrates how the estimated cost per electricity connection scales with household (connection)
density and different cost of low voltage line, which is generally a dominant cost factor, particularly at
low household densities. This estimate illustrates how household connection density of 25/km2 and a
wire cost of $10/m can lead to a connection cost of $3000, whereas a density of 100/sqkm and $2/m for
wire can reduce this cost to about $500.

Figure 10: Connection Cost per Household with Varying Connection Densities and Low Voltage Line Costs.
The calculation for these estimates also assumes the following: at $15 per meter costs for medium voltage line; a distance
of 1000 m from the national/regional electricity grid to the community being connected, and a base cost of $100 per
connection (for labor, equipment at the household, and other factors that are required for every connection and do not
scale with household density).

Motive Power

Productive uses of electricity, especially those that benefit women can end up providing both social and
economic benefits simultaneously. One example of this is the successful experience with the multi-
functional platform in Mali. This intervention, described in the box below, points to the importance of
addressing the specific time- and effort-consuming tasks that women are engaged in and how these can be
carried out with greater efficiency and better product quality with an energy service that provides motive

  From p. 10 of Reducing the Cost of Grid Extension for Rural Electrication ESM227, Energy Sector Management
Assistance Programme, Report 227/00, February 2000, UNDP/World Bank, c/o Energy, Mining and
Telecommunications Department, The World Bank, 1818 H Street, NW, Washington, DC 20433, U.S.A.)

Box 4: Diesel Powered Multifunctional Platform in Mali

By many measures, Mali is one of the poorest and least developed countries in the world. As of 2001, more than 70%
of the population survived on less than $1 (PPP) per day, and the nation falls very near the bottom (172 of 175) of the
UNDP’s Human Development Index (HDI). Geography and energy are key factors in Mali’s poverty: Nearly three-
quarters of Mali’s population of roughly 12 million lives in semi-arid rural areas, where poverty is most severe and yearly
variations in rainfall can have drastic economic impacts. Rural villages tend to be small (1000 - 2000 people) and
dispersed, and electrification virtually nonexistent. Biomass constitutes the vast majority (90%) of the country’s energy
supply, particularly in rural areas, where women and girls are responsible for the time-consuming and labor-intensive
work of fuel collection.

Beginning in 1993, the UN Industrial Development Organization (UNIDO) and the International Fund for Agricultural
Development initiated a program to decrease the burden of fuel collection, supply labor-saving energy services, and
promote the empowerment of women through the provision of the “multi-functional platform” (MFP) to rural villages. The
MFP is a 10 horsepower diesel engine with as many as twelve modular components in an integrated system that can
supply a variety of services. These include motive power for time- and labor-intensive work such as agricultural
processing (milling, de-husking), and electricity for lighting (approximately 200-250 small bulbs), welding or pumping

Although the benefits of these services are shared by many in the villages, those who acquire, own, manage and control
the platform itself are women’s organizations. This is not only to ensure that women and children benefit directly, but
also to create a group with the necessary skills to ensure the MFP’s long-term viability, while building capacity and
empowering women generally. Women’s groups are responsible not only for owning and managing the system, but also
for covering between 40 to 60% of the MFP’s initial cost, as much as 2600 $US of the 4,300 $US total cost for
construction and installation (~3,000,000 Fcfa). The remaining $1700 or so is provided by the program, which is funded
by UNDP, the Norwegian government, and various local partners (NGOs, social clubs and other donors), who also help
with financing platforms and training costs. In total, transaction costs for the program are high, totaling 8.3 million for the

In 1998, the UNDP and the Government of Mali had begun supporting the program’s implementation phase, which has,
from 1999 to 2004, installed nearly 400 platforms, reaching an expected 8,000 rural women in villages. Each MFP
intervention—from a feasibility study, to installation, to operation—lasts about two years. Capacity building and
institutional support is strongest in the earlier phases, then tapers off, leaving the rural women’s village groups in charge
of the platform’s operation, relying on a network of private suppliers, technicians and partners. A similar project is being
planned in Senegal, where a national MFP program hopes to install 1,000 platforms between 2005 and 2010.

Recent studies have concluded that the presence and use of platforms appears to have beneficial impacts, both direct
and indirect, with respect to the MDGs. Benefits observed in 12 villages studied include:
         Poverty Reduction and Alleviating hunger (MDG-1): The motive power provided by the platforms, used for
          mechanization of tasks that were previously carried out by women, by hand, reduced the time required for
          labor-intensive tasks from many hours often to a matter of minutes. The time and labor saved could be shifted
          directly to other, income-generating activities, leading to an increase in women’s income by an average of 0.47
          $US per day. In addition, rice production and consumption increased after the introduction of MFPs, an indirect
          benefit arising from time saved.
         Education (MDG-2) and Gender Equity (MDG-3): Girl-to-boy ratios in schools and the proportion of children
          reaching grade five both improved with the installation of the platforms, as youg girls were needed less for
          time-consuming and laborious chores.
         Reduction in Child and Maternal Mortality (MDGs 4 & 5): Increases in time and the mother’s socio-economic
          status accompanying the introduction of MFPs correlate with improvements in women’s health and increases in
          the frequency of women’s visits to local clinics for prenatal care.
Overall, the MFP program in Mali offers compelling evidence that time saved in the lives of women and children,
combined with the added socio-economic and capacity benefits to women’s groups of controlling and managing the
MFP as a resource, can confer substantial benefits to health and welfare.

Mali Case Study by Silvia Fracchia, Remi Lang, Bridget Anderson Columbia University, SIPA, 2004.
Multifunctional Platform For Village Power, Nalini Burn and Laurent Coche, 2000:
Impact Study of the Multifunctional Platform on the Living Conditions of Women, by Moussa Diagan for the MLI/99/002
         project: Multifunctional Platform for Poverty Alleviation.” Mali, June 2001.
MFP project website:


Transport services play a critical role in economic growth as well as in efficient delivery of social
services. The rural poor are particularly affected by the cost of transport. They receive lower prices for
their products because of high transport costs, must buy inputs at higher cost, need to travel further over
difficult roads for social services and cannot access some services at all if transport services do not exist
or are too expensive.

Motorized transport services depend on a variety of factors: quality of roads, cost and availability of fuel,
costs of vehicles and risks associated with operating the vehicle. The cost of delivery of energy services is
also higher where transport services are more expensive. High transport costs impact energy costs at
every level, from the cost and maintenance of decentralized energy systems, to the cost of bill collection
for centralized systems.

A dominant factor is the existence and quality of roads, typically measured as road density (or length of
paved roads in km per 1000 persons). These data are shown below, in Figure 11 for Africa.

Figure 11: Map of Paved Road (km) per 1000 people

Densities using this measure for many of the landlocked countries in sub-Saharan Africa are an order of
magnitude lower than those of countries such as Brazil, China and India; and two-orders of magnitude
lower than those of the high-income countries.

In India transportation by road using 10-ton truckloads costs about 3 cents/ton-km with full truckloads for
distances over 500 km to locations where a truck can find payloads in the return direction. Under similar
assumptions in much of sub-Saharan Africa the transportation costs can be as much as 10 cents (estimates
for road tankers from UNDP/World Bank ESMAP study65). When transporting fuel, trucks frequently
cannot carry other products in the return direction thus possibly as much as doubling these costs.

The impact of high transportation costs and taxation can be very significant on the cost of fuel. A very
rough estimate shows that to move fuel a distance of 1000 km by road (as may be the case in land-locked
countries such as Niger, Chad, CAR, or Mali) the cost of transport alone could add $0.20/liter to fuel cost
(assuming roundtrip trucking costs based on 20 cents/ton-km). In Figure 8 where cost of gasoline and
diesel in SSA are shown in 2002 (data from GTZ66) what we observe is a 30 or 40 cent per liter premium
for these countries. While data for the movement of bulk LPG over 1000 km are not available, it would
not be unreasonable to expect them to be any lower. One measure of how low transportation costs can be
is by examining these costs in the United States where the average cost of moving 1 ton over 1 km is 1.45
cents. In 1890, this cost in the United States was 11 cents in 2001 dollars67. In addition, high
transportation cost for fuel further exaggerates the problem of distributing fuel over large distances in
quantities where pipelines are not cost effective.

   Sub-Saharan Petroleum Products Transportation Corridor Analysis and Case Studies, March, 2003.
   Deutsche Gesellschafft fur Technische Zusammenarbeit
   Harvard Institute of Economic Research Discussion Paper Number 2014, Cities, Regions and the Decline of
Transport Costs, by Edward L. Glaeser and Janet E. Kohlhase, July 2003.

While the importance of major road and rail lines (especially those connecting land-locked countries to
the sea) for economic growth is well known, equally important are local roads and motorized vehicle
access. A World Bank Report68 describes how inadequate pricing and marketing policies can impact rural
areas severely and suggests that ―more attention should be given to policies affecting the availability and
cost of transport services at the local level as well as deal with intermediate means of transport (i.e.,
intermediate between headloading and motorized transport) which are generally underdeveloped.‖

As pointed out by Hine and Ellis69, while road investment has an important role, improving short lengths
of feeder roads may have little impact if no change in transport mode occurs. According to them,
upgrading a 5 km feeder road from earth to gravel standard might only increase farmgate prices by about
one tenth of one per cent. In comparison bringing new motor vehicle access 5km closer to a village (or
farm) when the alternative was head loading by hired labor could increase farmgate prices by over a
hundred times as much. Motor vehicles and roads that allow motor vehicle access are both necessary to
permit a quantum reduction in costs of transport. The ―software‖ of new road construction needs to be
considered as well. This includes the need for integrated multi-sectoral planning and associated
considerations of land use, enforcement of road rules, footpaths and budgetary provisions for

Rural Roads in Sub-Saharan Africa, Lessons from World Bank Experience, by John Riverson, Juan Gaviria, and
Sydney Thriscutt.
   World Bank report ―Agricultural Marketing and Access to Transport Services‖ by J. L. Hine and S. D. Ellis, TRL
Limited. http://www.transport-

Table 5: Data for selected countries: Traditional fuel, CO2 emissions, Paved Roads, Urban/Rural Electrification
   Sources for data: UN Human Development Report, 2004: ;
                         Traditional fuel
       Country            consumption          CO2 emissions      Paved Roads                        Urban vs. Rural Electrification
                                                                                                                        Number of Households
                        % of total energy        Per capita       km per 1000          Electrification levels of       connected to the National
                         requirements           (metric tons)       people                  Households                       Grid (estimates)
                                                                                      Urban %:          Rural %:
                              2001                  2000             '99-'01         1999-2002            2001            Urban             Rural
   United States                4                   19.8              13.13
   Japan                       1.2                   9.3               7.04
   Korea, Rep. of              1.9                   9.1               1.38
   Brazil                     35.7                   1.8               0.56
   China                       7.8                   2.2               1.01
   South Africa              12.9 (g)                7.4               1.67          77 - 84 (*)         49.0               NA           2,095,229
   India                      24.3                   1.1               1.49
   Cambodia                   95.1                   (.)               0.17
   Ghana                      84.5                   0.3               0.43            61 AT            4.3 AT
   Uganda                      95                    0.1               0.08             20                1.0            126,000          35,428
   Kenya                      70.6                   0.3               0.26             20               1.70            394,985          71,718
   Senegal                    72.5                   0.4               0.45             13                73
   Tanzania, U.
   Rep. of                      92.8                 0.1             0.11               39                1.0            239,000          46,159
   Côte d'Ivoire                75.1                 0.7             0.31              13 AT            73 AT
   Malawi                       86.6                 0.1             0.51               47               0.05             14,300           1,090
   Chad                         97.8                 (.)             0.03
   Ethiopia                     95.2                 0.1             0.06                13              0.70               NA            78,573
   Mali                         88.3                 0.1             0.17
   Niger                        77.3                 0.1             0.07
                      g. refers to the South   Source:                              Sources:
                      African Customs          WB Development Indicators, CD        --Unless noted: AFREPREN, African Energy Data and
                      Union, (BWA, LSO,        ROM, 2004.                           Terminology Handbook Year 2003 - 2004;
                      SWZ & NAM)                                                    --(*) Karekezi, S. and L. Majoro (2002). "Improving modern
                      Source: calculated                                            energy services for Africa's urban poor." Energy Policy 30(11-
                      using data from UN                                            12): 1015-1028.
                      correspondence, UN                                            --AFREPREN gives ave. for Sub-Saharan Africa (SSAf) of:
                      DESA, Statistics                                              39.96;
                      Division                                                      IEA Energy Outlook 2002: SSAf total: 22.6% (Urban: 51.3%;
                                                                                    Rural: 7.5%), AT Alan Townsend

Other cross-sectoral benefits of improved transport

Improved transport can both directly and indirectly assist in a rural health and education. It can reduce
the isolation of rural areas, helping them attract and retain health, education and services professionals.
There may also be direct benefits in form of access to emergency health care facilities, (e.g. need for
emergency vehicular transport during child birth). Direct benefits to education were reported in
Morocco70 where the presence of a paved road improved the schooling outcomes of rural children.

A DFID study71 reports that women located in a village on a main road in Cameroon were able to spend
more time producing food to buy and sell, and made an average income of $570—more than twice the
$225 earned by women in an isolated village located 1.5 hours from the road. Improved roads with
adequate provision for the use of footpaths may also make it possible to use IMT (Intermediate Modes of
Transport) and benefit from technological progress in the form of motorized transport. This would reduce
the time and effort spent in transport. The possibility of women-owned and operated motorized vehicles
that are perhaps shared among several tasks should be explored. Kessides et al72 report that women are
responsible for producing 80% of Uganda’s food, provide about 70% of total agricultural labor and carry
out agricultural tasks without the benefit of technological innovation, inputs or finance while men are not
constrained by competing claims on their labor time.

Fan73 reports that in India state expenditure on agricultural research and extension, improved roads,
irrigation and education had all contributed significantly to productivity growth in agriculture and in
agricultural wages. Many other reports (FAO74 ERD75, DFID76) cite the benefits of rural roads to poverty

In each of the three countries analyzed in the article, ―Fertilizer market development: a comparative
analysis of Ethiopia, Kenya, and Zambia,‖77 transport and handling costs accounted for 50% or more of

   Khandker, Shahidur R., Victor Lavy and Deon Filmer (1994) Schooling and Cognitive Achievements of Children
in Morocco, World Bank Discussion Paper 264, Washington DC: World Bank. http://www-
   Reference to come—may match 77.
   Kessides, C., Malmberg-Calvo, C., Farvacque-Vitkovic, C. Decentralization of Infrastructure, December 18, 2000
   Fan, Shenggen and Hazell, Peter (1999) ―Are Returns to Public Investment Lower in Less-favored Rural Areas ?
An Empirical Analysis of India,‖ EPTD Discussion Paper No. 43, International Food Policy Research Institute,
Washington, D.C.
Fan, Shenggen, and Hazell, Peter and Thorat, Sukhadeo (1999) ―Linkages between Government Spending, Growth
and Poverty in Rural India,‖ Research Report 110, International Food Policy Research Institute, Washington, D.C.
Fan, Shenggen, and Zhang, Linxiu, and Zhang, Xiaobo (2002) ―Growth, Inequality and Poverty in China: The Role
of Public Investments,‖ Research Report 125, International Food Policy Research Institute, Washington, D.C. 37.
    FAO, IFAD, and WFP (2002): ―Reducing Poverty and Hunger: The critical role of financing for food,
agriculture and rural development‖, FAO, Rome, Italy.
   ERD POLICY BRIEF NO. 13, Infrastructure and Poverty Reduction, What is the Connection?
by Ifzal Ali and Ernesto M. Pernia.
   Making Connections: Infrastructure and Pro-Poor Growth: Implications of Recent Research, December 2002.
Paper prepared by Christopher Willoughby under contract to Oxford Policy Management for the UK Department for
International Development.
   Jayne, T. S., J. Govereh, et al. (2003). "Fertilizer market development: a comparative analysis of Ethiopia, Kenya,
and Zambia." Food Policy 28(4): 293-316.

total domestic marketing margins. The sum of importer, wholesaler and retailer profit margins generally
account for less than 10%.

Education, Training and Research Capacity

To enable both the top-down large national and regional infrastructure as well as the bottom-up service
delivery systems for distribution to the consumer, there will be a need for commensurate education and
training programs. The place specific nature of infrastructure needs and appropriate solutions will also
require research that is carried out in the region. For example, in order to exploit the geothermal
resources in the Rift valley of East Africa one would need people with training in geology as well as
engineering to carry out everything from exploration and drilling of test wells to design, construction and
maintenance of the power plants. To carry out the detailed analysis and management of environmental,
social and economic impacts of hydroelectric power plants would need experts in hydrology, earth
science, engineering, economics and social sciences. If one were to take the example of agriculture,
private industry or rich-world research enterprise has failed in the past to carry out the necessary research
to develop new seeds and technology for developing world needs. These were largely developed by the
CGIAR system of research institutions. More importantly, these institutions also played a critical role in
training the people within the region who became part of the country specific agriculture research efforts.
A similar research enterprise may also be needed for the development of rural infrastructure and for the
support of rural industries and technology. Such institutions also provide a forum for the exchange and
cross-fertilization of ideas and scientific knowledge, exploiting the international networks that already
exist through professional societies within each discipline. In particular there may be much to gain from
south-south interactions in adapting cost effective technologies that may be more suitable for small initial
consumption levels.

One successful model for this institution and capacity building may be the land grant universities that
were created in each of the states of the US. Such institutions would play a critical role in identifying the
basic and applied research needs that are specific to a certain region. For example, the Tigray region in
Ethiopia desperately needs the understanding of low-cost labor-based technologies for stabilization of
eroded land, water harvesting and purification, construction and repair of year round vehicle-accessible
roads and small scale distributed power generation.

Utilities also need people for distribution, transmission and bill collection. For example a typical electric
utility with a low penetration rate as is the case in sub-Saharan Africa may have as many as one employee
per 100 customers. (South Africa has one employee per 375 customers and for the top ten utilities in
Japan, the number is one employee per 575 customers). If say grid electrification were to reach 50 million
additional households (half of the roughly 100 million households without electricity in SSA, or about 5
million new connections per year in SSA in the next 10 years) one would need 10,000 trained electrical
engineers just for the design of the distribution network. This estimate is made on the basis of data
provided by KPLC in Kenya, where one electrical engineer is able to design and follow the needed
construction for 500 new connections per year. More detailed exercises of this nature will be needed to
make judicious estimates of the needs reasonable

In contrast to these estimated needs, technical capacity is one of the least developed aspects of most
enterprises throughout sub-Saharan Africa. A recent UNIDO report78 has emphasized the reluctance of
African manufacturers and other firms to employ trained scientists, engineers and technicians. Even in
Zimbabwe, which has historically had the second-most advanced industrial sector outside of South
Africa, scientific and technical staff tend to make up less than two percent of the workforce, and these

  Industrial Development Report 2004: Industrialization, Environment and the Millennium Development Goals in
Sub-Saharan Africa - The New Frontier in the Fight Against Poverty. UNIDO, 2004.

were overwhelmingly concentrated in food-processing firms, due to that industry’s quality control and
testing needs. Small African firms tended to have even lower formal skill levels within their workforces.
In comparison, Asia’s lowest level of technical training (three percent) is seen in Sri Lanka, and its
highest in South Korea (twelve percent). One reason given for Africa’s lower technical staffing levels is
the fact that few African governments offer supports such as substantial tax benefits and other policy
incentives to encourage employee training that are common throughout Asia. Instead of formal education
and training, African enterprises tend to rely upon traditional apprenticeship systems that transmit skills
that may be unsuited to modern, technological production. Problems also exist within the public sector
training system, which tends to offer narrow and obsolete courses that are ill-suited for private sector

To reach the large and dispersed rural populations of East Africa with rural infrastructure, an extension
approach—not unlike agriculture extension—may be needed to carry out the dissemination, monitoring,
maintenance, organization and mobilization (for labor based techniques). These workers could, for
example, be responsible (with support from larger R&D organizations) for maintaining the transport,
water, energy and infrastructure associated with health and education facilities in a rural area. This could
include environmental restoration, water harvesting, school and clinic building, electricity and water
needs, monitoring and maintenance of rural roads. These extension workers could work as a team in
association with health and agriculture extension workers or could be based in say secondary schools. In
the later scenario, they could also play an educational role in developing training programs in vocational
skills for the schools. Intermediate technical skills for youth to participate and contribute in the growth of
infrastructure as well as the associated job creation benefits can play a critical economic as well as social
transformation function. Such infrastructure extension workers, of necessity, will have to be ―generalists‖,
perhaps with basic technical skills. University educational programs will need to adapt their curriculum
to meet the demands for such training. Moreover, for rural areas, if alternative ownership and market
structures such as private companies, cooperatives, local consumer associations, public-private joint
ventures, or local government initiatives are to develop, trained people who can create and manage these
institutions will be needed. Decentralization of services will also create the need for training in the
financial, management and technical aspects of developing proposals and operating a ―service provider‖
organization. Policies and financial assistance will need to be implemented in parallel to ensure that
educational programs provide such training within existing or new institutions.

Transitional Technologies (Ladder-climbing vs. Leapfrogging)

With focus on technology rather than local needs and conditions, and on global rather than local
environmental concerns, there is now considerable emphasis on ―leapfrogging.‖ This relates to the idea
that developing countries can and should adopt new technologies rapidly, bypassing the intermediate
technologies that are now considered obsolete by wealthier nations. However newer technologies are not
always low-cost and when this is the case, even wealthier nations are slow at adopting them. It is also
important to recognize that promotion of high-cost technologies can slow adoption of solutions that then
may be unaffordable at a later date. When a newer technology does in fact meet the needs based on local
conditions, e.g. mobile telephones, they do lead to higher adoption rates and can replace or add to existing
infrastructure such as land lines.

In contrast, transitional, lower-cost, shorter-term technologies help to preclude technological lock-in,
allowing households or communities to change technologies along the energy ladder according to their
needs. For example, while a low-cost LPG appliance can provide advantages for a household that
currently relies on biomass as cooking fuel, other households may be ready to make a change from to
DME or natural gas. Also, portions of a given energy technology may remain useful as other components
of a system change. For instance, a pre-existing grid infrastructure that formerly relied upon generation

using coal or oil may allow for easier adaptation to gas or hydropower or wind. In these and other cases,
the flexibility of a low-cost, intermediate solution may be limited if the initial emphasis is on
leapfrogging. Thus, the ideas of leapfrogging and technological lock-in should be evaluated carefully
before being pursued for the poorest countries.

Importance of regional integration of infrastructure

The growth and integration of regional infrastructure and markets can be a driver of economic and social
development in Africa. The formation of regional power pools, transmission line networks across
countries, and pipelines for fuels can lead to better utilization capacity (and hence lower costs), reduced
variability of supply and use the optimal mix of primary energy sources at any point in time. As is clear
from Fig. 12 of river basins in Africa, below, hydropower and geothermal resources do not follow
political boundaries. The creation of joint agreements for sharing the costs of study, research and
development of new regional sources of energy such as these that cut across physical and political
geographies can help reduce costs and aid development across nations. The development of shared oil &
gas and electricity infrastructure and markets—such as the Southern Africa Power Pool (SAPP), the West
Africa Power Pool (WAPP) and the West Africa Gas Pipeline (WAGP)—are expected to bring multiple
benefits to participating nations.

Figure 12: Map of Africa's hydrological basins (
For many reasons, national governments and international lending and development institutions have
targeted energy systems, particularly for electricity and petroleum products, for interconnection at the
regional level. Several of these relate to economic efficiency and expansion of energy services. The
sharing of electricity generation capacity allows some countries to meet energy needs without relying
entirely upon their local primary energy resources or having to expand their own generation. Also, there
are great price differentials in Africa: household electricity tariffs range from roughly 16 cents/kWh US
in Mali to roughly 5 cents/kWh in Ghana.79 This suggests the potential for trade, expansion in generation
and supply for those with comparative advantage, and lower cost for buyers. Other benefits relate to the
nature and size of the energy markets themselves. A larger market encourages private investment and
allows for projects that are larger-scale, lowering supply costs and reducing the need for redundant
facilities. By providing light, heat and motive power to enhance and enable economic activity beyond the
subsistence level, energy interconnection at the local level can help develop markets where they are
limited. 80

Integration of energy infrastructure can also have beneficial impacts on regional politics, peace and
security. It tends to decrease the macroeconomic risks of particular countries by helping them to diversify
their energy sources. Integration also fosters economic cooperation and increases the costs of conflict.81
In contrast, a lack of crucial infrastructure, or limited energy systems with high technical and ―non-
technical‖ losses (often due to theft), are widely recognized as an impediment not only to the provision of
services offered by the infrastructure itself, but also other crucial elements of growth, such as trade.82

A frequently cited example of potential gains from integration of energy infrastructure—for Africa
generally, but Central Africa in particular—is the potential for hydroelectricity generation and
distribution. Some estimates for the hydropower in Africa include 300 GW of continuous energy for the
continent as a whole and 90 GW of continuous energy for the Democratic Republic of Congo. The latter

   A Brighter Future? :
   Stryker, J.K. et al, pp. 194 in Lavergne.
   OECD/AfDB, 2002, W.T. Oshikoya and M. Nureldin Hussain, p. 77-78.

estimate equates to nearly 800 TWh/year, a potential amount that is third behind estimates for China
(1,320 TWh/yr) and Russia (1,096 TWh/yr), and ahead of both the United States (700 TWh/yr) and
Canada (530 TWh/yr). The power generation guaranteed during low water output years is estimated to be
80% of the installed capacity for the Democratic Republic of Congo, and this is the highest guaranteed
amount of any nation.83,84 Power experts from around the world have argued for this resource as a basis
for regional and additional continental interconnections, based on a five-region plan. These would
enlarge the extent of Africa’s interconnections—currently limited to three, in the West, South, and East—
while expanding Africa’s energy supply, creating opportunities for continental energy trade and export to

Those experienced with Western lending for infrastructure emphasize the need to learn from past lessons
and reform the process of project planning, implementation and revenue management in key ways to
bring increased emphasis on poverty reduction. These include efforts to target the proper sort of
infrastructure for given conditions, since water, transportation and/or energy can have differing impacts
on GDP growth at differing economic levels; to be attentive to the urban bias in infrastructure and
provision of related services; to be aware of the need to strengthen, integrate and expand local markets,
since they serve as a basis for strong and sustained exports; and to establish sound legal, regulatory and
institutional frameworks to ensure that public priorities—poverty reduction in particular—remain the
focus throughout. These and other provisions can help to avoid the pitfalls of the past: ―white elephant‖
projects that do not meet local needs, corruption and lack of transparency, adverse social and
environmental impacts, and others85 86. Some experts note the potential for integration of infrastructure in
Africa, and related lending from OECD countries, to have similar effects as the European Investment
Bank in the history of Western Europe’s integration and enduring peace87 .

Energy Security

Fossil fuels used in the developed world frequently originate within the least developed countries. The
Gulf of Guinea in Africa is poised to become a major source of oil and natural gas. Similar conditions for
export exist in Bolivia with gas and Chad with oil, both landlocked countries. Regional cooperation,
refineries and intra-continent cross-border fuel pipelines and fuel transportation corridors based on
anticipated demand and external assistance may be the best way to ensure that Africa and Latin America
can benefit from lower energy prices due to resource availability within their regions. Once the
economies of the poor countries grow and their energy needs continue to rise, they will have an increasing
ability to adopt and develop non-fossil fuel based energy sources and address the GHG emission
problems associated with use of fossil fuels. However, continued rapid use by the developed-world of
fossil fuels that frequently originate in poorer countries can rapidly close this window of opportunity for
the poor countries. They could be left with the prospect of trying to kick-start their economies with
increasingly higher cost fossil fuels, or instead depend upon newer emerging technologies that are yet to
mature and hence costly.

Not all countries have the same set of conditions and hence the same priorities. For a sparsely populated
country such as Chad—with settlements far apart, an extremely low annual average commercial energy
consumption (in the single digits in kgoe), and a nearly nonexistent road network—the priority may be a
road network, a mini-refinery and stand-alone diesel generators for the short term. These could power
decentralized mini-grids or women-owned and operated multi-functional platforms of the kind first

   Sarfoh, J.A. 1993. (Cited in Lavergne, p. 13).
   Hammons, T.J., et al. African Electricity.
   OECD/AfDB, 2002, Oshikoya and Hussain, ―Infrastructure and Economic ..,‖ p. 85-101.
   OECD/AfDB, 2002, Johnson, H.F. ―Poverty, Infrastructure and …,‖ p. 171-175.
   OECD, 2002, Tutty, M.G. ―Enhancing Infrastructure in Africa,‖ pp. 181-187.

introduced in Mali (through a UNDP-UNIDO program). The mini-refinery would allow Chad to use,
without paying large transportation costs, some of the billion barrels of oil that it has begun exporting.
This may help to ameliorate the irony that Chad has an electric power generation capacity just for
extraction/pumping of oil (destined for export) that is several times that of the power generation capacity
of the entire country.


Appendix I: A preliminary energy needs assessment for the MDGs.

In collaboration with McKinsey & Co. the Millennium Project has developed a preliminary energy needs
assessment for a number of countries. The following section draws on a more detailed discussion
provided in a background paper by the Project.88 The targets and some of the assumptions presented
below have been revised in the meantime. We are currently updating the results of the needs assessment.

Setting Targets for Energy Services
Targets for energy services need to address two dimensions of ―access‖. The first dimension covers
minimum or ―MDG-compatible‖ energy consumptions per household, while the second, sets out coverage
targets for access to energy services to be achieved by 2015.

1. Consumption targets
Clean and efficient energy for cooking and reading light are important for meeting the MDGs related to
health, gender equality and education. Households with access to electricity should also be able to power
a radio. Similarly, schools and health facilities require sufficient energy for lighting, cooking,
communication, as well as refrigeration and electrical equipment as necessary. Based on these minimum
needs we propose the MDG-compatible consumption targets summarized in Table 6. The targets for
household energy consumption outline strictly minimum requirements exclude space heating, a second
light bulb, as well as energy needs for productive use and income generation.

Annual consumption                  Lighting/electrification*           Cooking*
Household                           75kWh or equivalent                 1 gigajoule useful energy
School                              2,000kWh                            7,000kg LPG
Hospital                            50,000kWh                           10,000kg LPG
Clinic                              8,000kWh                            1,000kg LPG
Health post                         2,000kWh                            400kg LPG
Table 6: Minimum MDG-compatible targets for energy consumption 89
For comparison, Reddy90 suggests in a recent UNDP report that an electricity consumption on a
continuous basis of 18W/capita as being adequate for home electrification for lighting, space comfort,

     Available online at
     We are currently reviewing these targets and propose the following objectives:
Annual consumption per              Electrification                      Cooking*
Household                           75kWh or 15kWh w/battery+            1 gigajoule useful energy
School                              2,000kWh or 500kWh w/solar           7,000kg LPG
Hospital                            50,000kWh                            10,000kg LPG
Clinic                              8,000kWh                             1,000kg LPG**
Health post                         2,000kWh                             400kg LPG**
+ 75 kWh is sufficient to provide reading light in one room for 4 hours each day. Households that receive
electricity through batteries charged at local schools will only be able to access enough electricity to cover
minimal household lighting needs (15 kWh).
   Schools accessing electricity through solar panels will have limited generation capacity.
* For households, a range of fuels may be able to provide adequate cooking energy. Schools and health
centers will require enough LPG thermal energy services to cook meals for patients or one meal for each
student each day.

  Chapter 4, Energy Technologies and Policies for Rural Development of Energy for Sustainable Development: A
Policy Agenda, Editors: Thomas B. Johansson and José Goldemberg, 2002

food preservation, and entertainment in tropical countries. Assuming a five-person household, the annual
electricity consumption would be about 900 kWh/household using this measure. In the same work,
Robert Williams of Princeton University also provides an estimate of 3.75 watts/capita for five CFLs for
lighting, 3.13 watts/capita for a color television and 13.65 watts/capita for a refrigerator adding up to a
similar estimate for annual electricity consumption.

2. Coverage targets
We propose the following coverage targets:
   (i) Reduce by half, between 2005 and 2015, the proportion of urban and rural households
        without access to MDG-compatible lighting;
   (ii) Reduce by half, between 2005 and 2015, the proportion of urban and rural households
        reliant on cooking methods that are not MDG-compatible; and
   (iii) By 2015, provide adequate.91 clean and efficient energy services to all educational and
        health facilities.

These energy targets are consistent with the poverty reduction MDG. In our assessment they describe the
minimum required to achieve other MDGs, including gender equality, environmental sustainability,
halving hunger, and improving human health.

Estimating Resource Needs
We have applied a simple gap analysis in each of the three countries to develop an estimate of the
resources required to deliver the energy services required to achieve the MDGs. To this end we used the
Long-range Energy Planning model (LEAP), an energy systems scenario planning and modeling tool
designed by the Stockholm Environment Institute, to develop energy scenarios.92

In LEAP we defined the household energy services in each country differentiating between urban and
rural populations. This projections cover electrification rates, the proportion of households using one of
six lighting methods and one of six cooking methods. 93 For each country, we estimated four groups of
costs: end-use devices, fuel consumption, electrical connections, and power plants (see Table 7). Overall
per capita energy program costs averaged $19 (ranging from $17-23). These per capita values reflect total
MDG-energy program costs divided by total population of the country, i.e. not only the portion of
population actually reached by the program.

   With minimum consumption levels as defined above.
   LEAP is available free of charge to developing country governments and non-governmental organizations. More
information on the model is available at
   The lighting methods are: candles and open fires, kerosene wick lamps, kerosene hurricane lamps, gas lamps, and
electric bulbs. The six cooking methods include: unprocessed biomass in stone fires, improved wood stoves,
charcoal, kerosene, gas, and electricity.

Type of cost        Share of total       Components
                    program cost
Fuels               40-53%               All fuels used for lighting and cooking (except electricity):
                                         wood, coal, charcoal, kerosene, LPG
Electrical          20-33%               Central grid connections
Connections                              Mini-grid, Solar, home systems
End-Use             12-19%               Light bulbs
Devices                                  Kerosene lamps (hurricane and wick)
                                         Kerosene stoves
                                         Gas stoves
Power plants        3-22%                Additional generating capacity required for energy program
Table 7: Breakdown of Total Energy Program Costs
As can be seen from the table, cooking fuels make up the bulk of total needs. Electrical connections,
which are the only ―one-off‖ or fixed costs in our model, account for roughly one quarter of program
costs. Where central grid expansion is used, population dispersion and distance from the central grid are
the most important drivers of unit costs. Unit costs for new household connections accordingly range
from $500 (in urban Ghana) to $1,500 (the average cost of a solar home system connection). The
expected growth of urban populations acts as a key driver of investment needs. For example, Ghana will
need to spend more than 40 percent of its total electrification budget to maintain its high urban
electrification coverage rate of 78 percent in the fact of rapid population growth.

End-use device costs, and costs associated with construction and operation of power generation facilities
both typically make up approximately 20 percent of total program costs. Meanwhile, expected
investments in generation capacity are primarily driven by industrial energy consumption since residential
use will only account for 10 percent of total consumption.

     Box 1: Interventions missing from preliminary resource estimates for Energy
          Electricity services for agriculture and other productive activities,
          Thermal energy for space heating and electricity for household ventilation,
          Policy interventions to support private sector participation in electricity and fuels
          Repairs and maintenance to existing electricity grid system,
          Demand-side efficiency interventions (to reduce electricity wastage) at household,
         commercial, and industrial levels,
          Training and capacity building to support implementation of technology/end-use
         device interventions, and
          Training and capacity building to support electrification.

A stand-alone rollout of diesel generators to all health and education facilities would increase the overall
energy program cost by approximately 3 percent. The savings potential in connecting nearby households
to a mini-grid could make the rollout of diesel generators to all health and education facilities cost neutral.

Ongoing Refinements of the Energy Needs Assessment
The Millennium Project is continuing to refine its needs assessment for energy services. In particular, we
are preparing a simple excel-based investment model to complement detailed energy planning in LEAP.94
 Together with all sectoral needs assessment tools, this model will be made available as soon as possible on

It will allow countries to adapt the needs assessment methodology to their needs and priorities, and to
review assumptions and results in a fully transparent manner. In addition, we are reviewing the list of
interventions and targets and will produce updated needs assessments over the coming months. In
particular we will include energy services powered by batteries that households can recharge at generator

Preliminary country results

Electrification rates in Bangladesh are extremely low relative to the rest of Asia. While only 25 percent of
the urban population and 10 percent of the rural population in Bangladesh is electrified, approximately 41
percent of the South East Asia population and 87 percent of East Asia is electrified. The vast majority of
Bangladesh’s population, therefore, is dependent on inefficient and relatively low luminosity fuels, such
as kerosene, for lighting. On the cooking side, only about one half of the urban population has access to
modern fuels. In rural areas, charcoal and unprocessed biomass dominate. In regard to supply, power
plants are presently underutilized, with plants operating at about 2/3 of maximum capacity.

On the energy supply side, current installed power generation capacity totals 3.6GW, composed of
hydropower and thermal facilities, and annual electricity generation is estimated to be 14,000 GWh.
Households consume only 10 percent of electricity generated. Power plants are presently underutilized,
with plants operating at about 2/3 of maximum capacity.

Applying the coverage targets laid out above to Bangladesh results in an increase in electrification rates to
39 percent in urban areas, and 21 percent in rural areas—reaching an additional 4.9 million households in
total between 2005 and 2015. On the cooking side, the proportion of the population with access to cleaner
cooking methods will grow by 12 millions households. Increased demand for electricity from households,
educational- and healthcare facilities, and industry will require a 65 percent increase in annual power
supply. Table 8 summarizes the projected investments needs in the energy sector for Bangladesh.

Table 8: Cost of key Energy interventions in Bangladesh.
Total per capita costs are particularly high at the beginning of the period since Bangladesh will need to
build additional generation capacity to expand access to electricity. The costs for additional capacity have
been projected linearly across the 11-year period. The results show that investments in rural energy would
require approximately three quarters of total resources.

Not included in these summary results is the cost of increasing generation capacity and providing
electricity for industrial needs. We estimate the corresponding resource requirements to amount to
roughly $7.2 billion dollars for the period from 2005 to 2015. Due to capacity constraints, these
investments will need to be initiated as early as 2005.

Cambodia has no oil, natural gas, or other energy resources, although a few waterfalls in remote mountain
areas could produce hydroelectric power. Because the vast majority of Cambodians cannot afford
imported fuels, wood remains the principal source of energy throughout the country. 82 percent of the
energy source comes from fuelwood, while 92 percent of the population depends on this energy (CSD
2003). The present energy situation is unsustainable, however, as forest coverage shrinks and demand for
energy increase. Alternative energy sources such as solar power and wind power are being investigated

and are in the early stages of development on a small scale. The electrification rate in Cambodia is
approximately 15 percent, which is significantly lower than the South East Asia average of 41 percent.

Table 9 summarizes preliminary resource estimates for the energy sector. Unfortunately, the available
energy data for Cambodia was extremely limited. For this reason, the thermal energy patterns used to
generate cost numbers were derived from our work on Bangladesh. Specifically, we assume that in urban
areas, energy for cooking is provided by unprocessed biomass (70 percent), charcoal (15 percent) and
kerosene (15 percent). In rural areas, 85 percent of the population is assumed to cook on biomass, and 15
percent on charcoal. On the lighting side, 25 percent of urban- and 10 percent of rural areas are assumed
to be electrified, which is consistent with official statistics reporting that 15 percent of the total population
has access to electricity. Unelectrified urban- and rural households are assumed to rely on kerosene
hurricane and kerosene wick lamps in equal proportion.

Table 9: Cost of Energy interventions in Cambodia.
On the supply-side, given that insufficient data was available concerning existing power generation
capacity, all numbers were calculated assuming that all incremental demand has to be served by adding
new capacity. In other words we assume that there is no over capacity to meet future increases in energy
demand. We also do not include fuel costs or O&M expenditures for existing power plants, which will
lead to an understatement of true costs. Additionally, it should be noted that the model only makes
provision for power generation to serve electricity consumed by households, schools and health facilities.
We have not included industrial electricity consumption.

While urban rates of electrification in Ghana, at 78 percent, are relatively high compared to the rest of
sub-Saharan Africa (23 percent), rural access (18 percent) remains comparatively limited. The vast
majority of Ghana’s rural population depends on inefficient and relatively low luminosity fuels, such as
kerosene, for lighting. In addition, only a very small proportion of the population has access to modern
fuels for thermal energy: in urban areas, although LPG has made inroads, charcoal is the most widespread
cooking fuel. In rural areas, unprocessed biomass dominates. Recent droughts in Ghana have caused
significant power shortages that underscored the need for additional power sources. Currently, electricity
is mainly produced from hydro sources. Households consume only 10 percent of all electricity generated
in the country.

Applying our coverage targets set out above to Ghana requires the country to increase electrification rates
to 82 percent in urban areas, and 28 percent in rural areas—reaching an additional 0.6 million households
in total between 2005 and 2015. On the cooking side, the proportion of the population with access to
cleaner cooking methods will increase by 2 million households. Increased demand for electricity from
households, educational and healthcare facilities, and industry will require a 45 percent increase in annual
power supply. Table 10 below summarizes our resource estimates for the energy sector.

Table 10: Cost of key Energy interventions in Ghana.
Due to the high rates of electrification – particularly in urban areas – Ghana’s investment needs in the
sector are lower than in other countries we have studied. In absolute terms, though, they vary between $12
and $15 over the period. Not included in this summary is the cost of increasing generation capacity and
providing electricity for industrial needs. We estimate the corresponding resource requirements to amount
to roughly $495 million dollars for the period from 2005 to 2015.


Tanzania’s electrification rates of 10 percent in urban areas and 1 percent in rural areas are significantly
lower than the sub-Saharan average of 23 percent, but comparable to those in Uganda. The vast majority
of Tanzania’s population is therefore dependent on inefficient and relatively low luminosity fuels, such as
kerosene, for lighting. Similarly, in thermal energy, only a very small proportion of the population has
access to modern fuels: in urban areas, charcoal is the most widespread cooking fuel, while in rural areas,
unprocessed biomass dominates. Biomass accounts for over 90 percent of the primary energy used (URT

Achieving the energy targets for Tanzania would imply an increase in electrification rates to 27 percent in
urban areas and 13 percent in rural areas—reaching an additional 1.3 million households between 2005
and 2015. For thermal energy targets, the proportion of the population with access to cleaner cooking
methods will have to grow by 3.6 million households over the same period. Increased demand for
electricity from households, educational- and healthcare facilities, and industry will also require a 30
percent increase in annual power supply. Our resource estimates are summarized in Table 11. The results
show that, due to very low levels of access to improved energy in rural areas, significant investments will
be needed in order to provide the non-urban population with improved energy services. Not included in
these summary results is the cost of increasing generation capacity and of providing electricity for
industrial needs, which we estimate to be roughly $218 million dollars for the period from 2005 to 2015.
Since current capacity is sufficient to meet current industrial demand, most investments to satisfy the
industrial sector’s energy needs will need to be made later in the period, as demand increases.

Table 11: Cost of key Energy interventions in Tanzania.

The country’s electrification rates of 10 percent in urban areas and 1 percent in rural areas are
significantly lower than the sub-Saharan average of 23 percent. These rates are comparable to those for
Tanzania. The vast majority of Uganda’s population is therefore dependent on inefficient and relatively
low luminosity fuels, such as kerosene, for lighting. Similarly, for thermal energy, only a very small
proportion of the population has access to modern fuels. In urban areas, charcoal is the most widespread
cooking fuel, while in rural areas unprocessed biomass dominates.

The MDG-consistent energy targets require Uganda to increase electrification rates to 27 percent in urban
areas and 13 percent in rural areas—reaching an additional 1 million households between 2005 and 2015.
For thermal energy, an additional 3.1 million families will need to gain access to cleaner cooking methods
over the same period. Increased demand for electricity from households, educational- and healthcare
facilities, and industry will require a 70 percent increase in annual power supply.

Table 12: Cost of Energy interventions in Uganda.
The table above presents the summary analysis of costs associated with these interventions. The majority
of investments will need to be made in rural areas, focusing on improved fuel inputs. Our projections
show that with increasing coverage rates, operating costs will rise sharply, contributing to a steep increase
in per capita costs from $5.9 in 2005 to $16.6 in 2015. Not included in these summary results is the cost
of increasing generation capacity and providing electricity for industrial needs, which we estimate at
roughly $350 million dollars for the period from 2005 to 2015. Since current capacity is sufficient for
meeting industrial demand, most investments to satisfy the industrial sector’s energy needs will need to
be made after 2007.


Shared By: