9118

Energy in Developing Countries
January 1991
OTA-E-486
NTIS order #PB91-133694
Recommended Ci tati on:
U.S. Congress, Offi ce of Technol ogy Assessment, Energy in Developing Countries,
OTA-E-486 (Washi ngton, DC: U.S. Government Pri nti ng Offi ce, January 1991).
For sal e by the Superi ntendent of Documents
U.S. Government Pri nti ng Offi ce, Washi ngton, DC 20402-9325
(order form can be found i n the back of thi s report).
Foreword
Thi s report was prepared i n the course of the ongoi ng OTA Assessment, Fueling
Development: Energy and Technology in Developing Countries, which is being carried out i n
response to requests from the Senate Commi ttee on Governmental Affai rs; the House
Commi ttee on Energy and Commerce; the Subcommi ttee on Energy and Power of the House
Commi ttee on Energy and Commerce; the Subcommi ttee on Human Ri ghts and I nternati onal
Organi zati ons and the Subcommi ttee on Afri ca of the House Commi ttee on Forei gn Affai rs;
the Subcommi ttee on I nternati onal Devel opment, Fi nance, Trade and Monetary Pol i cy of the
House Banki ng Commi ttee; and i ndi vi dual members of the Senate Envi ronment and Publ i c
Works Commi ttee; the House Sel ect Commi ttee on Hunger; and the Congressi onal
Competi ti veness Caucus.
These commi ttees and subcommi ttees asked OTA to exami ne the rol e of technol ogy i n
provi di ng cost effecti ve energy servi ces that poor countri es need for thei r economi c and soci al
devel opment, whi l e mi ni mi zi ng adverse envi ronmental i mpacts. The commi ttees were
parti cul arl y i nterested i n an anal ysi s of the U.S. rol e i n faci l i tati ng the adopti on of such
technol ogi es. The overal l assessment i s schedul ed for compl eti on i n 1991. Thi s report, the fi rst
of two, was prepared i n response to the requesti ng commi ttees i nterest i n recei vi ng an i nteri m
product. I t exami nes how energy i s suppl i ed and used i n devel opi ng countri es, and how energy
use i s l i nked wi th economi c and soci al devel opment and envi ronmental qual i ty.
OTA recei ved substanti al hel p from many organi zati ons and i ndi vi dual s i n the course of
prepari ng thi s report. We are very grateful for the efforts of the projects contractors, who
prepared parts of the background anal ysi s; members of the advi sory panel ; and workshop
parti ci pants, who provi ded gui dance and extensi ve cri ti cal revi ews; and the many addi ti onal
revi ewers who gave thei r ti me to ensure the accuracy of thi s anal ysi s.
U JOHN H. GI BBONS
Director
. . .
Ill
Advisory PanelEnergy and Technology in Developing Countries
Harry G. Barnes, Jr., Chairman
Critical Languages and Area Studi es Consorti um
I rma Adel man
Department of Economi cs
Uni versi ty of Cal i forni a, Berkel ey
Jeffrey Anderson
I nsti tute of I nternati onal Fi nance
Edward S. Ayensu
Pan-Afri can Uni on for Sci ence and Technol ogy
Ger al d Bar nes
General Motors Corp.
Mohamed T. El -Ashry
Worl d Resources I nsti tute
Eri c Hel l and-Hansen
Uni ted Nati ons Devel opment
Pr ogr amme
Carl N. Hodges
Envi r onmental Resear ch Labor ator y
Uni versi ty of Ari zona
Donal d Jest
Sun Refi ni ng and Marketi ng Co.
Amory Lovi ns
Rocky Mountai n I nsti tute
Mohan Munasi nghe
Worl d Bank
Henr y Nor man
Vol unteers I n Techni cal Assi stance
Waafas Ofosu-Amaah
Wor l dWI DE
R.K. Pachauri
Tata Energy Research I nsti tute
D. Lou Peopl es
Bechtel Power Corp.
Gabr i el Sanchez-Si er r a
Organi saci on Lati no-Ameri cana de Energi a
Ki rk R. Smi th
Envi ronment and Pol i cy I nsti tute
East-West Center
I rvi ng Snyder
Dow Chemi cal U.S.A.
Thomas B. Steel , Jr.
Natural Resources Defense Counci l
Robert H. Wi l l i ams
Center for Energy and Envi ronmental Studi es
Pri nceton Uni versi ty
Lu Yi ngzhong
Professi onal Anal ysi s I nc. (PAI )
Montague Yudel man
Worl d Wi l dl i fe Fund and The Conservati on
Foundati on
Eugene W. Zel tmann
General El ectri c Co.
Observer
Davi d Jhi rad
U.S. Agency for I nternati onal Devel opment
NOTE: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the advisory panel members.
The panel does not, however, necessarily approve, disapprove, or endorse this report. OTA assumes full responsibility for the
report and the accuracy of its contents.
iv
OTA Project Staff-Energy in Developing Countries
Li onel S. Johns, Assistant Director, OTA
Energy, Materials, and International Security Division
Peter D. Bl ai r, Energy and Materials Program Manager
Project Staff
Joy Dunkerley, Project Director
Samuel F. Baldwin, Senior Analyst
Karen Larsen, Senior Analyst
Robi n Roy, Senior Analyst Paul S. Komor, Analyst
Ni na Gol dman, Research Assistant Sharon Burke, Research Assistant
Roger Chi n, Summer Intern
Administrative Staff
Ti na Brumfi el d Li l l i an Chapman Li nda Long
Contributors
Rosi na Bi erbaum, OTA Oceans and Environment Program
Gretchen Kol srud, OTA Industry, Technology, and Employment Program
Howard Levenson, OTA Oceans and Environment Program
Edward MacDonal d, OTA Food and Renewable Resources Program
Contractors
Russel l deLuci a Vi r endr a Kothar i
deLuci a and Associ ates, I nc. Energy and Envi ronmental Anal ysi s, I nc.
Cambri dge, MA Arl i ngton, VA
Ashok Desai Kari n Li ssakers
Nati onal Counci l of Appl i ed Economi c Research Col umbi a Uni versi ty
New Del hi , I ndi a New York, NY
K.G. Dul eep and Sergi o Ostri a Arjun Makhi jani
Energy and Envi ronmental Anal ysi s, I nc. I nsti tute for Energy and Envi ronmental Research
Arl i ngton, VA Takoma Park, MD
Ahmad Faruqui and Greg Wi kl er Peter Mei er and John Lee
Barakat and Chamberl ai n, I nc. I nternati onal Devel opment and Energy
Oakl and, CA Associ ates, I nc.
Howard Gel l er
Washi ngton, DC
Ameri can Counci l for an Energy Effi ci ent Henry Peski n
Economy Edgeval e Associ ates, I nc.
Washi ngton, DC Si l ver Spri ng, MD
Theodore J. Gorton Vacl av Smi l
Petrol eum Consul tant Uni versi ty of Mani toba
Bethesda, MD Wi nni peg, Canada
Donal d Hertzmark
I ndependent Consul tant
Washi ngton, DC
Environmental Problems and Priorities in Developing Countries
Workshop Participants, April 19, 1990
Mi chael Adl er
U.S. Envi ronmental Protecti on Agency
Lutz Baehr
Uni ted Nati ons
Center for Sci ence and Technol ogy i n Devel opment
Leonar d Ber r y
Fl ori da Atl anti c Uni versi ty
Al Bi nger
Conservati on Foundati on/Bi omass Users Network
Janet Wel sh Brown
Worl d Resources I nsti tute
Lal anath de Si l va
Envi ronmental Foundati on, Ltd.
Sri Lanka
Cl ar ence Di as
I nternati onal Center for Law i n Devel opment
Paul Dul i n
Associ ates i n Rural Devel opment
John J. Gaudet
Robert Goodl and
Worl d Bank
Lupe Gui nand
BI OMA
Venezuel a
Robert I chord
Kari Kei pi
I nter-Ameri can Devel opment Bank
Ananda Kri shnan
Uni ted Nati ons
Center for Sci ence and Technol ogy i n Devel opment
Russel l Mi ttermei er
Conservati on I nternati onal
Hi nd Sadek
Wor l dWI DE
Paul Schwengel s
U.S. Envi ronmental Protecti on Agency
NOTE: OTA is grateful for the valuable assistance and thoughtful critiques provided by the workshop participants. The views
expressed in this OTA report, however, are the sole responsibility of the Office of Technology Assessment.
vi
Outside Reviewers
I rma Adel man
Department of Economi cs
Uni versi ty of Cal i forni a, Berkel ey
Di l i p Ahuja
Envi ronmental Protecti on Agency
Jeffrey Anderson
I nsti tute of i nternati onal Fi nance
Edward S. Ayensu
Pan-Afri can Uni on for Sci ence and Technol ogy
Harry G. Barnes, Jr.
Cri ti cal Languages and Area Studi es Consorti um
Marti n J. Bernard I I I
Argonne Nati onal Laboratory
Vi c Fazi o
Member of Congress
Robert Goodl and
Worl d Bank
Eri c Hel l and-Hansen
Uni ted Nati ons Devel opment Programme
Carl N. Hodges
Envi r onmental Resear ch Labor ator y
Uni versi ty of Ari zona
Davi d Jhi rad
Donal d Jones
Oak Ri dge Nati onal Laboratory
Donal d Jest
Sun Refi ni ng and Marketi ng Co.
Dol ores Kern
Nati onal Coal Associ ati on
Bar bar a Lausche
Worl d Bank
Henr y Nor man
Vol unteers I n Techni cal Assi stance
Waafas Ofosu-Amaah
Wor l dWI DE
Phi l i p OKeefe
ETC (U.K.)
R.K. Pachauri
Tata Energy Research I nsti tute
Gabr i el Sanchez-Si er r a
Organi saci on Lati no-Ameri cana de Energi a
Anjal i Sastry
Rocky Mountai n I nsti tute
Jayant Sathaye
Lawr ence Ber kel ey Labor ator y
Lee Schi pper
Lawr ence Ber kel ey Labor ator y
Ki rk R. Smi th
Envi ronment and Pol i cy I nsti tute
East-West Center
I rvi ng Snyder
Dow Chemi cal U.S.A.
Car l os Suar ez
I nsti tuto de Economi a Energeti ca
Fundaci on Bari l oche
Pamel a Wentworth
Bechtel Power Corp.
Thomas J. Wi l banks
Oak Ri dge Nati onal Laboratory
Robert H. Wi l l i ams
Center for Energy and Envi ronmental Studi es
Pri nceton Uni versi ty
Lu Yi ngzhong
Professi onal Anal ysi s I nc. (PAI )
Montague Yudel man
Worl d Wi l dl i fe Fund and The Conservati on
Foundati on
Eugene W. Zel tzmann
General El ectri c Co.
NOTE: OTA is grateful for the valuable assistance and thoughtful critiques provided by the reviewers. The vi ews expressd i n this
OTA report, however, are the sole responsibility of the Office of Technology Assessment.
wi
Contents
Chapter Page
Chapter 1: I ntroducti on and Overvi ew . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter 2: Energy and Economi c Devel opment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Chapter 3: Energy Servi ces i n Devel opi ng Countri es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Chapter 4: Energy Suppl i es i n the Devel opi ng Worl d. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Chapter 5: Energy and the Envi ronment i n Devel opi ng Countri es ...........:. . . . . . . . . . . . . 111
Appendi x A: Gl ossary of Energy Uni ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
.,.
WI
Chapter 1
Introduction and Overview
Contents
Page
I ntroducti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
The Purpose of Thi s Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
The Devel opi ng Worl d and the I ndustri al i zed Worl d . . . . . . . . . . . . . . , . . . . .. 4
Si mi l ari ti es and Di fferences Among Devel opi ng Countri es . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
The Devel opi ng Worl d i n Gl obal Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Anal yti c Focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Overvi ew of the Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Popul ati on Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Economi c Devel opment . . * . . . . . . * . . . . . . . . . . 11
Energy Suppl y Constrai nts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Fi nanci al Constrai nts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Bi omass Suppl y Constrai nts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
I nsti tuti onal Constrai nts on Rapi d Expansi on i n Energy Suppl i es . . . . . . . . . . . . . . . . . . 14
Envi ronmental Degradati on i n Devel opi ng Countri es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Greenhouse Gases and Devel opi ng Countri es q . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Prospects for Effi ci ency I mprovements i n Energy Producti on and Use . . . . . . . . . . . . . . . . 18
Appendi x 1A: Economi c, Soci al , and Energy I ndi cators for Devel opi ng Countri es . . . . 20
Figures
Figure Page
1-1. Di fferences Between Devel opi ng and I ndustri al Nati ons . . . . . . . . . . . . . . . . . . . . . . . . . 4
1-2.1985 Energy Consumpti on, I ndustri al and Devel opi ng Regi on Fuel Mi x . . . . . . . . . . . 8
1-3. Commerci al Energy Consumpti on, 1973, 1985, and 2020 . . . . . . . . . . . . . . . . . . . . . . . 10
1-4. Per-Capi ta Commerci al Energy Consumpti on, 1973, 1985, and 2020 . . . . . . . . . . . . . 10
1-5. Suspended Parti cul ate Matter Level s i n Sel ected Ci ti es, 1980-84 q . . . . . . . . . . . . . . . 15
1-6. Sul fur Di oxi de Level s i n Sel ected Ci ti es, 1980-84 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Tabl es
Table Page
1-1. Heterogenei ty of the Devel opi ng Worl d: Soci al , Economi c,
and Energy I ndi cators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1-2. Commerci al Energy I mport Dependence i n Devel opi ng Countri es . . . . . . . . . . . . . . . . 7
1-3.1985 Pri mary Energy Suppl i es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1-4. Largest Energy Consumers, 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1-5. Passenger Fl eet Annual Growth i n Sel ected Countri es . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1-6. El ectri c Appl i ance Ownershi p i n Urban Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1-7. Energy I mports, Debt Servi ce, and Export Earni ngs for
Sel ected Devel opi ng Countri es, 1987 q . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1-8. Esti mated Annual Energy I nvestment as a Percentage of Annual Total
Publ i c I nvestment Duri ng the Earl y 1980s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Chapter 1
Introduction and Overview
Introduction
Energy use i n devel opi ng countri es has ri sen more
than fourfol d over the past three decades and i s
expected to conti nue i ncreasi ng rapi dl y i n the future.
The i ncrease i n the servi ces that energy provi des i s
necessary and desi rabl e, si nce energy servi ces are
essenti al for economi c growth, i mproved l i vi ng
standards, and to provi de for i ncreased human
popul ati ons. But fi ndi ng the energy suppl i es to
provi de these servi ces coul d cause major economi c
and soci al probl ems. For many of the devel opi ng
countri es, much of the addi ti onal energy needed wi l l
be suppl i ed by i mported oi l , and ri si ng oi l i mports
wi l l further burden those countri es al ready saddl ed
wi th hi gh oi l i mport bi l l s. Si mi l arl y, bui l di ng dams
or powerpl ants to meet hi gher demands for el ectri c-
i ty coul d push these nati ons even deeper i nto debt.
Energy devel opment and use al so contri bute to l ocal
envi ronmental damage i n devel opi ng countri es,
i ncl udi ng record l evel s of ai r pol l uti on i n some
ur ban ar eas.
The rapi d growth of energy use i n devel opi ng
countri es has wi de i mpacts. The economi c devel op-
ment process has tradi ti onal l y been accompani ed by
rapi d i ncreases i n oi l demand, whi ch, together wi th
ri si ng demand i n the i ndustri al countri es, contri bute
to upward pressures on worl d oi l pri ces. Hi gh l evel s
of i ndebtedness i n the devel opi ng countri es, partl y
energy-rel ated, have al ready contri buted to i nstabi l -
i ty i n the i nternati onal money and banki ng system.
Rapi d i ncreases i n fossi l fuel use i n devel opi ng
countri es al so represent a growi ng contri buti on to
the i ncrease i n l ocal and regi onal ai r pol l uti on as
wel l as atmospheri c concentrati ons of greenhouse
gases such as carbon di oxi de (CO
2
). I nternati onal
efforts to control greenhouse gas emi ssi ons requi re
acti ve parti ci pati on by devel opi ng countri es. Many
devel opi ng countri es coul d be adversel y affected by
cl i mate change, some much more than most i ndus-
tri al nati ons.
An economi cal l y and envi ronmental l y sound
approach to energy devel opment offers potenti al l y
l arge benefi ts both for the devel opi ng countri es and
for the rest of the worl d. I t can contri bute to
economi c growth i n the devel opi ng countri es, l ead-
i ng to hi gher l i vi ng standards, reducti on of hunger
and poverty, and better envi ronmental qual i ty. Thi s
strategy al so hol ds benefi ts for the ri cher countri es.
The devel opi ng countri es are i mportant tradi ng
partners for the Uni ted States. More rapi d economi c
growth i n these countri es coul d sti mul ate U.S.
exports, i ncl udi ng exports of energy technol ogy
products, and, therefore, coul d benefi t the U.S. trade
bal ance. I mproved energy technol ogi es can sl ow the
rate of i ncrease i n greenhouse gas emi ssi ons-a
gl obal benefi t.
The Purpose of This Assessment
Thi s report i s part of an assessment enti tl ed
Fuel i ng Devel opment: Energy and Technol ogy i n
the Devel opi ng Countri es, requested by the Senate
Commi ttee on Governmental Affai rs; the House
Commi ttee on Energy and Commerce; the Subcom-
mi ttee on Energy and Power of the House Commi t-
tee on Energy and Commerce; the Subcommi ttee on
Human Ri ghts and I nternati onal Organi zati ons and
the Subcommi ttee on Afri ca of the House Commi t-
tee on Forei gn Affai rs; the Subcommi ttee on I nter-
nati onal Devel opment, Fi nance, Trade and Mone-
tary Pol i cy of the House Banki ng Commi ttee; and
i ndi vi dual members of the Senate Envi ronment and
Publ i c Works Commi ttee, the House Sel ect Com-
mi ttee on Hunger, and the Congressi onal Competi -
ti veness Caucus.
The Offi ce of Technol ogy Assessment (OTA)
was asked to exami ne the extent to whi ch technol -
ogy can provi de the energy servi ces that devel opi ng
countri es need for economi c and soci al devel opment
i n a cost-effecti ve and soci al l y vi abl e manner, whi l e
mi ni mi zi ng the adverse envi ronmental i mpacts; and
to eval uate the rol e of the Uni ted States i n accel erat-
i ng the adopti on of such technol ogi es by devel opi ng
countri es.
Thi s report, the frost product of the assessment,
exami nes how energy i s currentl y suppl i ed and used
i n the devel opi ng countri es and how energy i s l i nked
wi th economi c and soci al devel opment and the
qual i ty of the envi ronment. Our emphasi s i s pri mar-
i l y on the present status of devel opi ng countri es, and
concerns about current energy trends. Thi s report i s
i ntended to provi de an i ntroducti on to the probl ems,
chal l enges, and opportuni ti es associ ated wi th pro-
-3
4 q Energy in Developing Countries
vi di ng energy servi ces for economi c and soci al
devel opment i n the devel opi ng countri es. These
i ssues are exami ned under four broad topi c areas:
energy and economi c devel opment (ch. 2); energy
servi ces (ch. 3); energy suppl i es (ch. 4); and energy
use and the envi ronment (ch. 5). A subsequent report
wi l l present the resul ts of OTAs assessment of
technol ogi es that can potenti al l y i mprove the effi -
ci ency of both energy producti on and use i n devel op-
i ng countri es; an exami nati on of the technol ogy
transfer process; and ways i n whi ch Congress can
hel p promote the rapi d adopti on of such pol i ci es.
The Developing World and the
Industrialized World
We l argel y fol l ow the defi ni ti on of devel opi ng
countri es-l ow- and mi ddl e-i ncome countri es (fur-
ther di vi ded i nto l ower mi ddl e and upper mi ddl e
countri es)---used by the Worl d Bank
l
(see app. 1A
for a l i st of these countri es), i ncl udi ng al l of the
countri es of Afri ca, Lati n Ameri ca, and Asi a,
excl udi ng Japan.
2
There are wi de di fferences i n average i ndi cators
of soci al and economi c condi ti ons between devel op-
i ng and i ndustri al countri es (fi gure l -l ).
1. Social: The ci ti zens of (OECD) countri es have
a l onger l i fe expectancy (76 years, compared
wi th an average of 62 i n devel opi ng countri es),
l argel y due to l ower i nfant mortal i ty rates (9 per
1,000 l i ve bi rths compared wi th 71 i n devel op-
i ng countri es). A much l arger share of the
Figure l-lDifferences Between Developing and
Industrial Nations (developing nation average
as a share of lndustrlal* nation average)
Real GDP
per capita
Commercial energy
consumption
per capita
Adult literacy
Life expectancy
t I 1 !
0% 25% 50% 75% 100%.
How to Interpret thls flgure. The average values for the industrial
countries are assigned 100 in all cases. The values for the
developing countries are expressed as a share of 100. For
example, average life expectancy in the industrial nations is 76
years and in the developing nations 62 years, or 82 percent of the
industrial country level.
Industrial excludes the U. S. S. R.; based on weighted average of
high-income market economies.
SOURCE: United Nations, Human DevelopmenfReport 199a World Bank
2.
WorldDevelopment Report.
popul ati on has access to secondary and hi gher
educati on and heal th care.
Economic: Average per-capi ta i ncomes
3
(ex-
pressed i n purchasi ng power pari ti es
4
) ar e mor e
than ei ght ti mes hi gher i n OECD countri es than
i n the devel opi ng countri es. Thi s di fference i n
i ncome l evel s refl ects major di fferences i n
economi c structure, parti cul arl y the hi gher
share of agri cul ture i n total producti on i n the
devel opi ng countri es. A much l ower share of
IS=, for ~mple, World B* WOr/dD#elOp~ntRePO~ 1989 (New York+ NY: Oxford university pr~s, 1989), PP. 1*165
2The deftition of developing countries is based prirnarily on per-eapits income levels. This ranking system is rather arbitrary, howeveq if all the
emmtriea of the world are ranked by ascending level of per-eapitaincome, there is no obvious gap in the series to demarcate the two groups of countries.
The World Bank distinguishes six categories of counties: low-income, lower-middle-income, upper-middle-income, high-income oil-exporters,
industrial market economies, including OECD as a subset and non-reporting non-members (the U.S.S.R., North Kor~ East Germany, Angol&
Bulgaria, Albani& Mongo@ and Namibia). The group of developing countries (low- and middlAncome and upper-rniddle-income countries) does not
include Saudi Arabb Kuwait and the United Arab Emirates by virtue of their high per-capita income. The World Bank does, however, include as
developing countries some East and West European countries, such as Poland, Hungary, Yugoslavia, Greeee, and Ibrkey, that qual@ as developing
countries by virtue of their inmme levels, bu~ due to their integration with industrial economies of East and West Europe, do not share other
characteristics of underdevelopmen~ and are therefore not included in this qort. Some other eouniries are excluded due to lack of reported data. Where
group averages of general economic and social indicators are reported direetly from the WorZdlkveZopment Report, these countries are included in the
total. In more detailed analysis, they are excluded. While every effort is made to adhere to these deftitions, it is not always possible, especially when
other sources of data with slightly different deftitions are used.
s~come is USWy m~~ by GrOSS Domestic or Gross National Product. The difference between the two-typiedy S@for most COuntrieS-_iS
that GDPmeasures the total output of goods and serviees witbin the national border of a country, whereas GNPmeasures the output of goods and aerviees
attributable to the nationals of a country wherever that activity oeeurs.
4H~et exc~emt~ me- to convert the GDP of different countries to dollars, average OECD per-pita income app=m to k ovsr20 tim~
higher than average developing country per-capita income. If, however, the comparison of income levels is adjusted to take into aeeount differences in
purchasing power of currencies (i.e., what a unit of currency such as the dollar will buy in different countries) the gap between average per-capita income
levels in developing and OECD countries narrows, and OECD per-capita income levels are 8 times rather than 20 times bigher tban the developing
country average. In either case, the gap in income levels is substantial.
Chapter 1-Introduction and Overview q 5
the popul ati on l i ves i n urban areas-37 percent
i n the devel opi ng countri es compared wi th 77
percent i n the OECD countri es. Popul ati on
growth i s more rapi d i n devel opi ng countri es.
I t i s esti mated to doubl e by 2040, whi l e the
popul ati on of the i ndustri al worl d wi l l i ncrease
by onl y 15 percent over the same peri od.
3. Energy: The economi c and soci al contrasts are
al so refl ected i n energy consumpti on. Per-
capi ta consumpti on of commerci al energy (coal ,
oi l , gas, and el ectri ci ty)
5
i n the OECD countri es
i s on average 10 ti mes hi gher than i n the
devel opi ng countri es. On the other hand, com-
mer ci al ener gy consumpti on i s i ncr easi ng much
faster i n the devel opi ng countri es. Bi omass
energy consumpti on i n the devel opi ng coun-
tri es i s hi gher than i n the OECD countri es and
provi des a much hi gher share of total energy
consumpti on.
Similarities and Differences Among
Developing Countries
The devel opi ng country averages shown i n fi gure
1-1, though adequate to i l l ustrate the broad contrasts
between devel opi ng and i ndustri al countri es, ob-
scure the wi de economi c and soci al di fferences
among devel opi ng countri es. I ndeed, the range of
di fferences between LDCs i s greater than that
between many of them and the i ndustri al countri es.
A generati on of excepti onal l y fast economi c growth
i n the Newl y I ndustri al i zed Countri es (the NI Cs),
combi ned wi th the sl ow growth, or i n some cases,
economi c stagnati on and decl i ne, i n many Afri can
countri es, has wi dened the gap among devel opi ng
countri es. Thus the probl ems, energy or otherwi se,
faced by a rel ati vel y ri ch and devel oped country
such as Brazi l are di fferent from those faced by a
poor country l i ke Ethi opi a, as are the resources
avai l abl e for thei r sol uti on. An appreci ati on of these
di fferences i s necessary for the real i sti c assessment
of energy technol ogi es.
Per-capi ta i ncomes i n the upper mi ddl e-i ncome
devel opi ng countri es (e.g., Brazi l , Argenti na, Al ge-
ri a, Venezuel a, and Korea) are al most seven ti mes
hi gher than i n the l ow-i ncome countri es (tabl e 1-1).
6
The i ncome di fferenti al refl ects major di fferences i n
economi c structure. I n the upper mi ddl e-i ncome
countri es, i ndustry has a much l arger share i n total
output and agri cul ture a much l ower share. I ndi a and
Chi na are excepti ons, wi th atypi cal l y l arge shares of
i ndustry, gi ven thei r l evel s of i ncome. The share of
the total popul ati on l i vi ng i n urban areas i s much
l ower i n the l ow-i ncome countri es. For exampl e, i n
several Afri can countri es onl y about 10 percent of
the total popul ati on i s urban
7
dwel l ers, i n contrast to
countri es l i ke Brazi l , Argenti na, and Venezuel a,
whose l evel s of urbani zati on (about 80 percent of the
popul ati on l i vi ng i n towns) are si mi l ar to those i n the
i ndustri al countri es.
Devel opi ng countri es al so show wi de vari ati ons
i n soci al i ndi cators. Li fe expectancy at bi rth ri ses
from an average of 54 years i n the l ow-i ncome
devel opi ng countri es to an average of 67 i n the upper
mi ddl e-i ncome countri es. I nfant mortal i ty i s twi ce
as hi gh i n the l ow-i ncome countri es (over 100 per
1,000 bi rths compared wi th an average of 50 i n the
upper mi ddl e-i ncome devel opi ng countri es). I ndi a
and Chi na are agai n excepti ons: i n both countri es,
despi te l ower average i ncome, i ndi cators of soci al
devel opment are si mi l ar to those found i n countri es
wi th much hi gher i ncomes. The experi ence of these
two countri es testi fi es to the i mportance of soci al
pol i ci es i n achi evi ng rel ati vel y hi gh l evel s of soci al
devel opment despi te l ow i ncomes.
8
Popul ati on
growth rates al so di ffer wi del y among devel opi ng
countri es. I n recent years these have ranged from
about 1 percent annual l y i n some countri es (e.g.,
Chi na, Uruguay, Korea) to over 3 percent i n several
Afri can countri es.
The wi de vari ati ons i n soci al and economi c
condi ti ons i n devel opi ng countri es are al so refl ected
i n thei r energy use. I n the upper mi ddl e i ncome de-
%eterm commercial energy conventionally applies to cod oi~ gas, and electricity on the basis that they are widely traded in organized markets.
These fuels are distinguished from other fuels such as fuew~ charcoal, and animal and crop wastes, which are described as biomass or
noncommercial fuels. The distinction between them can be misleading, particu.kly in the context of developing countries, as some of the swzdled
noncommercial fuels, such as firewood and charcoal are also widely traded in highly organized markets. ib~this ambiguity we use the term
biomass energy here.
6See also app. 1A.
%timates of urban populations are based on country-specfIc criteria related to size of settlement and presence of urban characteristics.
SW ~eme is develo@ further in a recent publication by the United Nations Development RO~, Human Development Report 1990 (New
York, NY: Oxford University Press, 1990).
9&3j footnote 5.
Table 1-1Heterogeneity of the Developing World: Social, Economic, and Energy Indicators
Low-income India and Lower middle- Upper middle-
Indicators Year countries China income countries income countries
Economic:
GNP per capita ($1987) (ppp)
a
. . . . . . . . . . . . . . . . . . . . . 1987 840.0 900.0 3,000.0 5,420.0
Share agriculture in GDP (%)
b
. . . . . . . . . . . . . . . . . . . . . . 1987* 33.0 30.0 21.0 10.O*
Urban population as share of total (%)
C
. . . . . . . . . . . . . . 1987 24.0 33.0 51.0 66.0
Cars and trucks per 1,000 members of Population
d
. . . . . 1980* 3.2 3.2 19.3 93.3
Social:
Life expectancy at birth (years)
e
. . . . . . . . . . . . . . . . . . . . 1987 54.0 65.0 64.0 67.0
Infant mortality per 100 births
f
. . . . . . . . . . . . . . . . . . . . . . 1987 103.0 62.0 61.0 50.0
Share of age group with secondary education (%)9 . . . . . 1986 25.0 39.0 51.0 59.0
Energy:
Commercial energy consumption per capita
(gigajoules)
h
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1986 4.9 16.3 36.2 58.3
Total energy consumption per capita (gigajoules)
l
.. . . . . 1986 12.3 18.7 41.7 67.0**
Share of traditional energy in total (%}
j.
. . . . . . . . . . . . . . 1986 60.0 13.0 13.0 13.0**
* Estimated.
If Brazil is excluded, the total per capita energy consumption would be 62 gigajoules and the share of traditional energy in total 2 percent.
SOURCES: WOIM Bank, 146ddDeve/opment Report 1989 (New York, NY: Oxford University Press, 1969). Purchasing power parity (ppp) estimate based on data on pp. 164and 222.
blbid,, based on data in table 3.
C
lbid., table 31.
d JoY DunkerleY and Irving t-loch, TfmSpOnEnefgy: Determinants and Po/icy(Washington, DO: Resources for the Future, September 1965), table 51 afld appendiX ta~e 13. 6ae~ on
estimates. Note that totals are unweighed averages and the countries included differ slightly from Work/ Development Report 1989.
e World Bank, op. cit., table 1.
1
lbid., table 32.
g lbid., ta~e ~.
hlbid, ta~e &
lunit~ Nations, Energy ~a~~ Yearbook 1986( New York, NY: United Nations, 1988), table 4.
JThese values for the share of traditional energy are muoh lower than those found in field surveys. These values are presented herein order to have a consistent data set. Estimates
based on field surveys suggest that biomass provides one-third of the energy used by developing countries overall (chs. 3 and 4).
Chapter 1--Introduction and Overview q 7
Table 1-2-Commercial Energy Import Dependence in Developing Countries
Country Number of Number of Number of High Medium Low
income countries energy energy importers importers importers
group in group
a
exporters importers (70-100%)b (30-70%)
b
(0-30%)
b
Low-income . . . . . . . . . . . . . . 38 4 34 29 3 2
China and India . . . . . . . . . . . 2 1 1 0 0 1
Lower middle-income . . . . . . 30 10 20 15 3 2
Upper middle-income . . . . . . 10 6 4 2 1 1
Total . . . . . . . . . . . . . . . . . . 80 21 59 46 7 6
a lnd~es all countries for which import dependence data are Wddh.
b shares of imports in total commercial energy Consumption.
SOURCE: Based on data in the United Nations, 1966 Energy Statistks Yearbook(New York, NY: 1988).
vel opi ng countri es, per-capi ta annual commercial
energy consumpti on (at 60 gi gajoul es
l O
) i s 12 ti mes
hi gher than i n the l ow-i ncome countri es (5 GJ).
l l
Agai n Chi na and I ndi a di ffer from the other l ow-
i ncome countri es, wi th per-capi ta consumpti on of
commerci al energy more than 3 ti mes hi gher than
other l ow-i ncome countri es. Per-capi ta consumpti on
of traditional bi omass fuel s, on the other hand, i s
general l y hi gher i n the poorest countri es, dependi ng
on the bi omass resources avai l abl e.
12
There are si mi l arl y l arge vari ati ons i n energy
resource endowment. Whi l e many countri es have
some energy resources, three-quarters of the devel -
opi ng countri es depend on i mports for part or al l of
thei r commerci al energy suppl i es (tabl e 1-2). Level s
of i mport dependence vary, but i n many countri es,
i mports (al most enti rel y oi l ) provi de nearl y al l
commerci al energy suppl i es. Oi l i mports can be a
consi derabl e strai n on al ready ti ght forei gn ex-
change budgets. I n several countri es, parti cul arl y i n
Afri ca and Central Ameri ca, oi l i mports represent
over 30 percent of forei gn exchange earni ngs from
exports (see app. 1A).
Despi te these di fferences i n aggregate i ndi cators,
there are strong si mi l ari ti es among devel opi ng
countri es wi thi n speci fi c sectors. Energy use i n
tradi ti onal vi l l ages throughout the devel opi ng worl d
i s fai rl y si mi l ar i n terms of quanti ty used, source
(bi omass, muscl e power), and servi ces provi ded
(cooki ng, subsi stence agri cul ture). At the other end
of the scal e, energy use by the economi cal l y wel l off
i s al so reasonabl y si mi l ar between devel opi ng and
i ndustri al countri es, i n terms of quanti ty used (to
wi thi n a factor of 2 or 3), source (oi l , gas, coal ,
el ectri ci ty), and servi ces provi ded (el ectri c l i ghti ng
and appl i ances, i ndustri al goods, pri vate automo-
bi l es, etc.). The l arge di fferences between countri es
are then i n l arge part due to the rel ati ve share of the
total tradi ti onal vi l l agers and the economi cal l y wel l
off i n the popul ati on, and i n the forms and quanti ti es
of energy used by those who are maki ng the
transi ti on between these two extremes. The broad
si mi l ari ti es within speci fi c popul ati on sectors i mpl y
that i t i s possi bl e to make general i zati ons about
technol ogy that are appl i cabl e to a wi de range of
otherwi se di sparate countri es.
The Developing World in
Global Energy
13
The devel opi ng countri es now account for about
30 percent of gl obal energy use, i ncl udi ng both
commerci al and tradi ti onal energy (see tabl e 1-3),
and thei r share i s growi ng rapi dl y. Thei r use of the
di fferent fuel s vary wi del y: they account for 85
percent of bi omass fuel consumpti on but onl y 23
percent of commerci al fuel s (oi l , gas, coal , and
el ectri ci ty). The mai n sources of energy for the
devel opi ng countri es as a group are coal , oi l , and
9
See footnote 5.
10A ~gajo~e (GJ), or 1 b~on jo~es, is about& enqy content of 8 gallons of gasoline. For refeNnCe, ~~per-cwi~ en%Y com~Ption h
the United States is 327 million Britiah thermal units (Mu) or 343 gigajoules. An exajoule (W), or 10
18
joules, ia about the same as a Quad (1.05 EJ =
1 Quad).
llWorld B@ Wor/dDeVezop~ntReport J989 @Je~ York NY: Ofiord University press, 1989). Da@ tirnp. 172, in tOIUES Of Od eqllhkllt (tot?)
converted to gigajoules at 1 toe = 41.9 gigajoules.
12Br=il, d=pite its ~~tivelY hi@ ficome, ~s subs~~ ~ntities of bio~s fiels in modem ~pl i cati o~, wch ss charcoal for stdmdK@ and
ethanol for cars. This contrasts with the use of biomass in the poorer countries, as a cooking fuel using traditional technologies.
ls~e ~~ ~ this ~tion we ~m from the World ~m= Cotiemnce, G/o~/EnergyPer~ectives 2000-2020, 14th Congress, Montreal 1989 (his
1989); and the United Nations, 1986 Energy Statistics Yearbook (New Yorlq NY: 1988), updated to 1987 by data provided by the U.N. Secretariat.
Table 1-3-1985 Primary Energy Supplies (exajoules)
Primary Total Total
Coal Oil Gas electricity commercial Biomass energy
World . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88.7 104.6 58.2 33.0 284.5 36.9 321.3
Industrial countries . . . . . . . . . . . . . . . . . 63.5 77.0 51.7 26.6 218.7 5.5 224.2
Developing countries . . . . . . . . . . . . . . . 25.2 27.7 6.5 6.4 65.7 31.3 97.1
Share of industrial countries . . . . . . . . . . 72% 74% 89% 81% 7770 15% 70%
Share of developing countries . . . . . . . . 28% 26% 11% 19% 23% 85% 30?40
NOTE: As in table 1-1, the values reported for developing country biomass are too low. Field surveys indicate that biomass accounts for roughly one-third of
the energy used in developing countries.
SOURCE: World Energy Conference, G/oba/ Energy Perspecths2M0-2020, 14th Congress, Montreal 1989 (Paris: 1989).
Figure 1-21985 Energy Consumption, Industrial and Developing Region Fuel Mix
Oil
34%
Biomass
electricity y fuels
7% 32%
Gas
23%
Developing nations
Total = 97 exajoules
Industrial nations
Total s 224 exajoules
SOURCE: World Energy Conference, G/oba/Enewy Perspecfives20@2020, 14th Congress, Montreal 1989 (Paris:
1989).
-
-.
bi omass (see fi gure 1-2). However, much of the coal
i s used i n I ndi a and Chi na onl y. The other devel op-
i ng countri es rel y heavi l y on oi l and bi omass for
thei r energy suppl i es. Several devel opi ng countri es
Chi na, I ndi a, Mexi co, Brazi l , and South Afri ca-are
among the worl ds 20 l argest commerci al energy
consumers (see tabl e 1-4). Chi na al one accounts for
al most 10 percent of the worl ds total commerci al
energy use.
Three countri es-Chi n . a, I ndi a, and Brazi l
together account for about 45 percent of total
devel opi ng country consumpti on of both commer-
ci al and bi omass fuel s. And these countri es pl us four
more-I ndonesi a, Mexi co, Korea, and Venezuel a--
account for 57 percent of the total . At the other end
of the scal e are a l arge number of smal l countri es
that, combi ned, account for onl y a smal l part of
gl obal consumpti on. The 50 countri es of Afri ca, for
exampl e, use under 3 percent of total worl d commer-
ci al energy consumpti on. Concerns about gl obal
energy use and i ts i mpl i cati ons focus attenti on on the
l arge consumers, but the energy needs of the smal l
devel opi ng nati ons, though of l esser i mportance to
gl obal total s, are cri ti cal to thei r devel opment
pr ospects.
The devel opi ng countri es are becomi ng i ncreas-
i ngl y i mportant actors i n gl obal commerci al energy.
Thei r share of the total has ri sen sharpl y i n recent
years (see fi gure 1-3), from 17 percent of gl obal
commerci al energy i n 1973 to over 23 percent now.
Despi te thei r much l ower l evel s of per-capi ta
commerci al energy consumpti on, devel opi ng coun-
Chapter l-introduction and Overview q 9
Table 1-4-Largest Energy Consumers, 1987
Per-capita
Total commercial commercial energy
energy consumption consumption
Country (exajoules) (gigajoules)
20 largest commercial energy consumers:
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
10
United States . . . . . . . . 68.1 280
U.S.S.R. . . . . . . . . . . . . 54.7 194
China . . . . . . . . . . . . . . . 23.5 22
Japan . . . . . . . . . . . . . . . 13.4 110
West Germany . . . . . . . 10.0 165
United Kingdom . . . . . . 8.5 150
Canada . . . . . . . . . . . . . 7.5 291
India . . . . . . . . . . . . . . . . 6.5 8
France . . . . . . . . . . . . . . 6.1 109
Italy . . . . . . . . . . . . . . . . . 6.0 105
Poland . . . . . . . . . . . . . . 5.3 141
Mexico . . . . . . . . . . . . . . 4.1 50
East Germany . . . . . . . . 3.8 231
Australia . . . . . . . . . . . . 3.2 201
Brazil . . . . . . . . . . . . . . . 3.2 22
South Africa . . . . . . . . . . 3.2 83
Romania . . . . . . . . . . . . 3.1 136
Netherlands . . . . . . . . . 3.1 213
Czechoslovakia . . . . . . . 2.9 185
Spain . . . . . . . . . . . . . . . 2.4 147
largest developing country energy consumers:
Rank
1 China . . . . . . . . . . . . . . . 23.5
2 India . . . . . . . . . . . . . . . . 6.5
3 Mexico . . . . . . . . . . . . . . 4.1
4 Brazil . . . . . . . . . . . . . . . 3.2
5 South Africa . . . . . . . . . . 3.2
6 South Korea . . . . . . . . . 2.2
7 Argentina . . . . . . . . . . . . 1.7
8 Venezuela . . . . . . . . . . . 1.6
9 Indonesia . . . . . . . . . . . 1.4
10 Egypt . . . . . . . . . . . . . . . 1.0
22
8
50
22
83
52
56
88
8
20
NOTE: Data for the top 10 developing country energy consumers include
only countries listed in app. 1A.
SOURCE: United Nations secretariat.
tri es accounted for one-hal f of the total increase in
gl obal commerci al energy consumpti on si nce 1973.
The i ncreasi ng share of the devel opi ng countri es
i n gl obal commerci al energy consumpti on i s wi del y
predi cted to conti nue. The Worl d Energy Confer-
ence pr ojects an i ncr ease i n thei r shar e to 40 per cent
by 2020 (see figure 1-3), and thi s trend i s confi rmed
i n a l arge number of other studi es.
14
The devel opi ng
countri es are projected to account for al most 60
percent of the gl obal increase (over current l evel s)
i n commerci al energy consumpti on by 2020. Chi na
al one accounts for over one-thi rd of thi s i ncrease.
These i ncreasi ng shares are suffi ci entl y l arge to have
a major i mpact on worl d energy markets. Despi te the
more rapi d rate of growth i n energy consumpti on i n
devel opi ng countri es, thei r per-capi ta consumpti on
of commerci al energy wi l l sti l l conti nue to be far
bel ow the l evel s i n i ndustri al countri es (see fi gure
1-4).
Analytic Focus
The anal ysi s presented i n thi s OTA study has
three i mportant features. Fi rst, rather than concen-
trati ng on energy suppl i es, the anal ysi s focuses on
the servi ces energy provi des. The reason for thi s
approach i s si mpl e. Energy i s not used for i ts own
sake, but rather for the servi ces i t makes possi bl e--
cooki ng, heati ng water, cool i ng a house, heati ng an
i ndustri al boi l er, transporti ng frei ght and peopl e.
Further, there may be many di fferent means of
provi di ng a desi red servi ce, each wi th i ts own costs
and benefi ts. For exampl e, transport i s provi ded i n a
number of ways-bi cycl e, motorcycl e, car, bus,
l i ght rai l , or ai rcraft. The consumer chooses among
these accordi ng to such cri teri a as cost, comfort,
conveni ence, speed, and even aestheti cs. Wi thi n
these consumer constrai nts, a more effi ci ent car may
be pr efer abl e to i ncr easi ng r efi ner y capaci ty i n or der
to reduce capi tal and/or operati ng costs, or because
of envi ronmental benefi ts. More than just engi neer-
i ng and economi cs must be consi dered, i ncl udi ng
soci al , cul tural , and i nsti tuti onal factors. Such fac-
tors are more readi l y i ncl uded i n a servi ces frame-
work than i n a conventi onal energy suppl y anal ysi s.
Second, wi thi n thi s servi ces framework the changes
i n how energy i s used are traced from tradi ti onal
rural areas to thei r modem urban counterparts. Thi s
14~a@Y~i~O@rOj&tiom ~fglo~comai~emr= com~ptionova~next zoy~~~~s. -eandhSckattetiolm,lntermtiomZ
Energy Workshop: Overview of Poll Responses (Stanford University International Energy ProjecL July 1989), reports the results and assumptions of
over 100 projections of global energy consumption and production and provides the means of the different studies. Not all studies report results for all
I@ons. The coverage is nonetheless a comprehensive indicator of how energy forecasters view the future. l%ey suggest that the developing countries
share could rise to over one-third by 2010. Longer term projections in general arrive at similar conclusions. For example, the Emissions Scenurios
document prepared by the Response Strategies Working Group of the Intergovemmental Panel on Climate Change, Appendix Report of the Expert
Group on Emissions Scenarios (RSWG Steering Committee, lhskA), April 1990, concludes that, overa wide range of scenarios, the share of developing
countries (Centrally Planned Ask Afiic& Middle East, and South and East Asia) will increaseffom a 1985 reference level of 23 percent to between
40 and 60percent of global energy in 2100, and that this group of developing countries would account for between60 and 80 percent of the total increase
inenergyconsumption over tbisperiod. Further, developments in the developing countries define much of the difference between the low and high growth
scenarios.
Figure 1-3-Commercial Energy Consumption, 1973,
1985, and 2020 (developing nation energy demand as
100%
76%
60%
26%
o%
a percentage of world total)
EZ Developing regions
Estlmated
1973 1985 2020
Year
SOURCE: VMd Enercw Conference. G/oba/ Enerav Perspectives 2000-
2020, 14th &mgress, Montreal 19S9 (P&is: 1989).
Figure 1-4-Per-Capita Commercial Energy
Consumption, 1973, 1985, and 2020
Gigajoules per capita
250
200
150
100
50
0
- Developing regions
m Developed regions
1973 1985
Year
2020
SOURCE: Vhxfd Energy Conference, Global Enemy Perspwtives 2000-
2020, 14th Congress, Montreal 1989 (P&is: 1989).
progressi on from the tradi ti onal rural to the modern
urban hel ps i l l umi nate the dynami cs of energy use,
and how i t can be expected to change i n the future.
Thi rd, the enti re system needed to provi de energy
servi ces-from the energy resource to the fi nal
energy servi ce, i ncl udi ng producti on, conversi on,
and use-i s exami ned as a whol e. Thi s i s done i n
order to show the total costs and consequences to
soci ety, as wel l as to the i ndi vi dual , of provi di ng
parti cul ar servi ces, and how they mi ght be provi ded
more effecti vel y i n terms of fi nanci al , envi ron-
mental , and other costs. For exampl e, i ncreased
l i ghti ng servi ces can be met by usi ng more conven-
ti onal l i ghti ng and i ncreasi ng the amount of el ectri c-
i ty generated, by i ncreasi ng the use of more effi ci ent
l i ght bul bs, or by a combi nati on of the two. A
systems approach permi ts the compari son of effi -
ci ency and suppl y opti ons i n achi evi ng the desi red
end.
I n our anal ysi s of energy servi ces and systems i t
i s recogni zed that technol ogy adopti on and use i s
embedded i n an i nsti tuti onal framework that pro-
vi des both i ncenti ves and di si ncenti ves to users, and
l argel y determi nes whi ch and how technol ogi es wi l l
be used. Thi s approach has a number of i mpl i cati ons
both for the way technol ogy i s used now and for the
adopti on of new technol ogi es i n the future. Thus, the
energy sector i n many devel opi ng countri es i s
frequentl y characteri zed as i neffi ci ent i n the
sense that more energy i s used to provi de a gi ven
servi ce or output than i s usual i n i ndustri al countri es.
I n a wi der context, however, taki ng i nto account the
many other rel evant factors (fi nanci al , i nfrastructur-
al , manageri al , and i nsti tuti onal ), the technol ogy
may wel l be used to the best of human abi l i ty and
often wi th consi derabl e i ngenui ty and resourceful -
ness. I n many cases, al though energy appears to be
used i neffi ci entl y, energy users may be acti ng
l ogi cal l y gi ven the framework of i ncenti ves and
di si ncenti ves wi thi n whi ch they make thei r deci -
si ons. I t fol l ows therefore that the adopti on of a new
technol ogy wi l l depend not onl y on the i ntri nsi c
superi ori ty of the technol ogy i tsel f but al so on
whether i nsti tuti onal factors favor i ts adopti on. The
pol i cy envi ronment i s of cruci al i mportance to the
adopti on of new technol ogi es.
Overview of the Report
Energy consumpti on i n the devel opi ng worl d has
ri sen rapi dl y i n the past and i s wi del y expected to
conti nue i ncreasi ng rapi dl y i n the future. The Worl d
Energy Conference, for exampl e, projects (i n i ts
moderate economi c growth case) a tri pl i ng i n
consumpti on of commerci al energy i n devel opi ng
countri es between now and 2020. A survey of a l arge
number of projecti ons of commerci al energy use
broadl y confi rms thi s trend. The projected rate of
i ncr ease i n commer ci al ener gy consumpti on i mpl i ci t
i n these forecasts i s l ower than that experi enced
between 1973 and 1985.
15
I ncreased suppl i es of
bi omass fuel s (fuel wood and ani mal and crop
wastes) wi l l al so be requi red. The Worl d Energy
Confer ence pr ojects a 25 per cent i ncr ease i n bi omass
use. Popul ati on growth and economi c devel opment
are the pri nci pal forces dri vi ng the rapi d i ncrease i n
energy use.
Population Growth
In many devel opi ng countri es, ferti l i ty rates (the
number of chi l dren expected to be born to a woman
duri ng the course of her l i fe) have dropped dramati -
cal l y over the past 20 years. Neverthel ess, the
popul ati on of the devel opi ng worl d conti nues to
grow rapi dl y. Over 90 percent of worl d popul ati on
growth i s now occurri ng i n the LDCs. At present, the
popul ati on of the devel opi ng countri es i s about 4
bi l l i on, 77 percent of the worl ds popul ati on. Even
assumi ng conti nued decreases i n ferti l i ty rates, the
popul ati on of these countri es i s projected to ri se to
7 bi l l i on i n 2025,
16
and coul d reach 10 bi l l i on i n
2100, due to the l arge number of women of
chi l dbeari ng age. Devel opi ng countri es woul d then
account for 88 percent of the gl obal popul ati on i n
2100, and for vi rtual l y al l of the i ncrease i n gl obal
popul ati on. The i ncrease i n popul ati on al one i n
devel opi ng countri es woul d account for a 75 percent
i ncr ease i n thei r commer ci al ener gy consumpti on by
2025 even i f per-capi ta consumpti on remai ned at
current l evel s.
Economic Development
Securing hi gher l i vi ng standards for the i ncreas-
i ng popul ati on of the devel opi ng worl d i mpl i es hi gh
rates of economi c growth. The Worl d Energy
Conference, for exampl e, assumes i n i ts moder-
ate growth rate scenari os, an average annual gross
rate of economi c expansi on of 4.4 percent to 2020,
sl i ghtl y l ower than i n the past. Thi s woul d represent
more than a fourfol d i ncrease i n economi c acti vi ty
between now and 2020. Achi evi ng such rates of
growth wi l l certai nl y not be easy, especi al l y i n l i ght
of the hi gh l evel s of debt that have constrai ned
economi c growth i n the 1980s, and i ncreasi ng
competi ti on for forei gn assi stance from the countri es
of Eastern Europe; but fai l ure to achi eve hi gh rates
of growth coul d spel l great hardshi p for the devel op-
i ng countri es, as thei r popul ati ons are growi ng so
rapi dl y.
The process of economi c devel opment that under-
l i es i mprovi ng l i vi ng standards i n devel opi ng coun-
tri es i nvol ves a number of changes, i ncl udi ng hi gher
agri cul tural producti vi ty, growth of manufacturi ng,
constructi on of a modern publ i c works i nfrastruc-
ture, urbani zati on, and i ncreased transportati on (tabl e
1-5 shows the rapi d i ncrease i n the road transport
fl eet). Hi gher standards of l i vi ng al so l ead to
expansi on i n the ownershi p of consumer appl i ances
(tabl e 1-6 i l l ustrates saturati on l evel s for some of the
most wi del y used appl i ances). Al l of these changes
have profound i mpacts on the amounts and types of
energy used.
Commerci al energy consumpti on typi cal l y ri ses
faster than economi c growth as the devel opment
process gets underway, and the share of commerci al
energy i n total energy consumpti on grows as i t takes
the pl ace of tradi ti onal bi omass fuel s. Even though
the rel ati ve share of tradi ti onal fuel s has decl i ned,
the absol ute amounts consumed have conti nued to
ri se, by an esti mated 2.5 percent per year.
17
Despi te the strong connecti on between commer-
ci al energy consumpti on and economi c growth,
there i s evi dence of consi derabl e di fferences among
devel opi ng nati ons i n thei r energy i ntensi ty-the
amount of commerci al energy consumed rel ati ve to
Gross Nati onal Product (GNP). There are exampl es
of countri es wi th si mi l ar per-capi ta i ncomes that
consume qui te di fferent quanti ti es of commerci al
energy. Some of these di fferences resul t from
country-speci fi c physi cal characteri sti cs, but others
are associ ated wi th di fferences i n soci al and eco-
nomi c pol i ci es. Pol i ci es promoti ng heavy i ndustry
and hi gh rates of urbani zati on contri bute to hi gh
energy i ntensi ti es. Si mi l arl y, the energy i ntensi ti es
of countri es change over ti me. I n several i ndustri al
countri es energy i ntensi ti es decl i ned even before
1973 at a ti me when oi l pri ces were fal l i ng, l argel y
Isworld Energy Coderence, Global Energy Perspectives 2000.2020, Mon@e~, 1989 (P~: 1989) projects a threefold kll%ise hl CCI
nsumptionof
commercial fuels in developing countries between 1985 and 2020. The Marine study projects a rise of 250 percent between 1985 and 2010 (see Alan
S. Marine and IAN Schrattenholzer, Intermtional Energy Workshop: Overview of Poll Responses, Stanford University International Energy Projec6
July 1989).
16Rudo~o A. B~@, ~Wd Bos, Patience W. Stephens, and My T. vu, Europe, Middle East, andAfiiCa (EMN) Region popldatiO?l prO@fiO?fS,
1989-90 Edition (Washingto@ DC: World B@ 1990), table 9.
ITWorld Energy Conferenm, Global Energy Perspectives 2000-2020, 14th Congress, Montreal 1989 (pa: 1989).
Table 1-5-Passenger Fleet Annual Growth in Selected Countries (percent)
Passenger Commercial Two and three
Country group cars vehicles wheelers Total
Developing countries:
Cameroon . . . . . . . . . . . . . . . . . . 11.8
Kenya . . . . . . . . . . . . . . . . . . . . . . 3.2
Bolivia . . . . . . . . . . . . . . . . . . . . . . 8.6
Brazil . . . . . . . . . . . . . . . . . . . . . . . 8.9
Thailand . . . . . . . . . . . . . . . . . . . . 8.8
India . . . . . . . . . . . . . . . . . . . . . . . 8.2
China . . . . . . . . . . . . . . . . . . . . . . . 41.6
Taiwan . . . . . . . . . . . . . . . . . . . . . 16.2
Weighted average . . . . . . . . . . 10.0
29.5
3.7
24.5
7.3
4.4
11.2
14.8
5.4
11.4
9.1
4.0
6.9
25.6
9.5
25.4
44.9
10.3
19.1
13.1
3.3
11.6
9.8
8.8
18.4
29.8
11.0
13.9
industrial countries:
Japan . . . . . . . . . . . . . . . . . . . . . . 3.0 4.1 7.0 4.4
United States . . . . . . . . . . . . . . . . 2.4 3.5 -5.6 2.3
West Germany . . . . . . . . . . . . . . . 3.3 0.4 -2.2 2.6
Weighted average . . . . . . . . . . 2.6 3.6 2.4 2.8
SOURCE: Fleet size and growth from Energy and Environmental Analysis, Pohcy Options for Improving Transportation Energy Eff.kiency in Developing
Countries, contractor report prepared for the Office of Technology Assessment, March 1990.
Table 1-6-Electric Appliance Ownership in Urban Areas (percent of households)
Country Radio Tv Fan Washer iron Cooker
China.. . . . . . . . . . . . . . . . . . . . . .
Liberia ... , . . . . . . . . . . . . . . . . . .
Guatemala . . . . . . . . . . . . . . . . . .
Manila ... , . . . . . . . . . . . . . . . . . .
Malaysia ., . . . . . . . . . . . . . . . . . .
Hong Kong . . . . . . . . . . . . . . . . . .
Bangkok ., . . . . . . . . . . . . . . . . . .
Taipei . . . . . . . . . . . . . . . . . . . . . .
39
76
78
80
70
90
66
4
25
78
79
91
96
92
45
56
82
75
96
94
2
1
16
34
5
53
1
74
94
77
87
16
44
91
84
89
KEY: information not available.
SOURCE: Jayant Sathaye and Stephen Meyers, Energy Use in Cities of the Developing Countries, Armua/ F?eviewof Energy, vol. 10, 1985, pp. 109-133,
table 6.
due to i mproved technol ogi es. After the 1973 and
1979 oi l pri ce shocks, the decl i ne i n energy i ntensi -
ti es was experi enced i n al l i ndustri al countri es.
Whi l e the energy i ntensi ti es of the devel opi ng
countri es conti nued to ri se after 1973, the ri se was
l ess sharp than before. These experi ences testi fy to
some fl exi bi l i ty i n commerci al energy use. The
current projecti ons ci ted above, however, al ready
i ncorporate assumpti ons about decl i ni ng energy
i ntensi ty i n devel opi ng countri es.
Energy Supply Constraints
The devel opi ng countri es wi l l face major di ffi cul -
ti es i n tri pl i ng energy suppl i es over the next 30
years. Si gni fi cant obstacl es i ncl ude fi nanci al con-
strai nts, di ffi cul ti es i n i ncreasi ng bi omass fuel
suppl i es, i nsti tuti onal and pol i cy factors, and envi -
ronmental i mpacts.
Financial Constraints
Commerci al energy consi sts of both domesti cal l y
produced and i mported suppl i es. Many devel opi ng
countri es rel y on i mported oi l for vi rtual l y al l of thei r
commerci al energy needs. Further i ncreases i n
energy i mports wi l l i mpose a heavy burden on
l i mi ted forei gn exchange resources, whi ch may
al ready be under pressure because of debt servi ce
payments (see tabl e 1-7).
Fundi ng the devel opment of domesti c energy
suppl i es and i nfrastructure al so poses probl ems.
Energy suppl y faci l i ti es such as el ectri ci ty generat-
i ng stati ons and petrol eum refi neri es are hi ghl y
capi tal i ntensi ve, pl aci ng major demands on the
scarce suppl i es of both domesti c and forei gn re-
sources avai l abl e for capi tal i nvestment. Al ready,
i nvestments i n the commerci al energy suppl y sector
(i ncl udi ng el ectri ci ty, oi l and gas, and coal ) repre-
sented i n the 1980s over 30 percent of publ i c
Table 1-7Energy Imports, Debt Service, and Export Earnings for Selected Developing Countries,
a
1987
Energy imports as share of Debt service as share of Energy imports and debt service
merchandise exports exports of goods and services as share of total exports 1987
Country (percent) (percent) (percent)
Low-income:
Ethiopia . . . . . . . . . . . . . . . . . . . . . . . . . . 55 28.4 83.4
Tanzania . . . . . . . . . . . . . . . . . . . . . . . . . 56 18.5 74.5
Madagascar . . . . . . . . . . . . . . . . . . . . . . . 36 35.3 71.3
Rwanda . . . . . . . . . . . . . . . . . . . . . . . . . . 53 11.3 64.3
Benin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 15.9 112.9
Kenya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 28.8 67.8
Pakistan . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 25.9 51.9
Burma . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 59.3 64.3
Lower middle-income:
Morocco . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 29.9 56.9
Jamaica . . . . . . . . . . . . . . . . . . . . . . . . . . 31 26.6 57.6
Turkey . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 31.7 62.7
Jordan . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 21.8 74.8
Syrian Arab Republic . . . . . . . . . . . . . . . . 40 16.5 56.5
Upper middle-income:
Argentina . . . . . . . . . . . . . . . . . . . . . . . . . . 10 45.3 55.3
Algeria . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 49.0 51.0
a lncludes all nations in which debt service and oil imports combined is greater than 50 percent Of tOtal eXpOrtS.
SOURCE:WorldBank, Wbr/dfJeve/oPmentReport 1989(New Ye*, NY: Oxford University Press, 1989), pp. 172-173 and210-21 1.
i nvestment budgets i n a wi de range of devel opi ng
countri es (see tabl e 1-8), wi th the el ectri c uti l i ty
sector accounti ng for the l i ons share. Despi te these
al r eady l ar ge cl ai ms on capi tal r esour ces, the cur r ent
l evel of i nvestment i n the el ectri ci ty sector maybe
i nadequate. The Worl d Bank esti mates that i nvest-
ments of $125 bi l l i on annual l y (twi ce the current
l evel ) wi l l be needed i n devel opi ng countri es to
provi de adequate suppl i es of el ectri ci ty .18 This
figure represents virtually the entire annual in-
crease in the combined GNP of the developing
countries.
Fi ndi ng the domesti c and forei gn resources needed
to fi nance energy faci l i ti es has al ways posed major
di ffi cul ti es for devel opi ng countri es. Current l evel s
of domesti c resource mobi l i zati on, often rel ated to
l ow l evel s of energy pri ces, are reported i n many
countri es to fal l short of the amounts needed for
system expansi on.
19 In the past, about one-hal f of all
i nvestments i n energy suppl y have been provi ded by
forei gn sources,
20
but hi gh l evel s of current debt i n
many devel opi ng countri es and i ncreasi ng fi scal
di ffi cul ti es i n the i ndustri al countri es make i t
di ffi cul t for many devel opi ng countri es to i ncrease
thei r bor r owi ng fr om abr oad.
Biomass Supply Constraints
I t maybe equal l y di ffi cul t to i ncrease suppl i es of
tradi ti onal bi omass fuel s. Despi te rapi d rates of
urbani zati on i n the devel opi ng worl d, al most two-
thi rds of the total popul ati ons i n poor nati ons l i ve i n
rural areas. These popul ati ons l argel y depend on
bi omass fuel s to produce thei r energy, wi th some
rural el ectri fi cati on where avai l abl e, and smal l but
vi tal quanti ti es of petrol eum products (for i rri gati on,
l i ghti ng, and transport).
Demand for bi omass fuel s (l argel y fuel wood) wi l l
conti nue ri si ng to meet the domesti c needs of the
urban and rural poor, rural i ndustry, and i n some
cases, such as Brazi l , modern i ndustry. Overuse of
bi omass resources al ready contri butes to envi ron-
mental degradati on (see bel ow). Moreover, gather-
i ng tradi ti onal suppl i es of fuel wood i s ti me-
consumi ng, exhausti ng work frequentl y undertaken
by women and chi l dren, who are thus di verted from
other acti vi ti es (educati on and farmi ng) that coul d
18u.s. A8~Y for~er~tio~ DeVelo~e@ P
ower
Shortages inD#eloping Counm.es: ~agni~e,lmpacts, solutions, andthe Role of the private
Sector (Washington.L DC: March 1988), p. 10.
l~wenw J. ~~ )7~rgy p~ce R@~ in De~e/oping Counm.es: ]s~es and options (OSIC Ridge, TN: ()& Ridge National bbomto~, Au~t
1987).
World B@ The Energy Transition in Developing Counm.es (Washington DC: 1983).
Table 1-8-Estimated Annual Energy Investment as a Percentage of Annual Total
Public Investment During the Early 1980s
Over 40 percent 30-40 percent 20-30 percent 10-20 percent 0-10 percent
Argentina Ecuador Botswana Benin Ethiopia
Brazil India China Egypt
Colombia Pakistan Costa Rica Ghana
Korea Philippines Liberia Jamaica
Mexico Turkey Nepal Morocco
Nigeria
Sudan
SOURCE: Mohan Munasinghe, E/ectrk Power Ecvnomics (Imdon: Butterworths, 1990), p. 5.
eventual l y i mprove thei r producti vi ty and l i vi ng
condi ti ons. An esti mated one-thi rd of the popul ati on
of devel opi ng countri es now faces fuel wood defi -
ci ts, and wi l l i ncreasi ngl y rel y on crop wastes and
ani mal dung to meet thei r energy needs.
Institutional Constraints on Rapid Expansion
in Energy Supplies
Over and above the capi tal constrai nts di scussed
above, a wi de range of other factors, customari l y
defi ned as i nsti tuti onal , currentl y i mpede com-
merci al energy sector devel opment. Whi l e defi ni -
ti ons of i nsti tuti onal factors di ffer between observ-
er s, they ar e gener al l y taken to be nontechnol ogi cal ,
encompassi ng a vari ety of economi c, organi zati onal ,
and pol i cy factors that affect the way technol ogi es
perform i n operati onal setti ngs.
21
Some of these
factors (the worl dwi de i ncrease i n i nterest rates, for
exampl e) are outsi de the control of i ndi vi dual
countri es, but others are rel ated to pol i ci es and
procedures i n the i ndi vi dual country.
The el ectri ci ty suppl y system offers an exampl e of
the i mportance of i nsti tuti onal factors. I n most
countri es of the devel opi ng worl d the el ectri ci ty
sector i s government owned, refl ecti ng the i mpor-
tance attached to el ectri c power for meeti ng eco-
nomi c and soci al objecti ves, and i n some cases,
especi al l y i n smal l systems, the advantage of cen-
tral i zati on for securi ng economi es of scal e and
coordi nati on i n pl anni ng and operati ons. However,
government ownershi p can l ead to i nterference and
l oss of autonomy i n day-to-day management of
uti l i ty operati ons and therefore reduced effi ci ency.
A recent Worl d Bank report on the power sector i n
devel opi ng countri es poi nts out that:
Such i nterference has adversel y affected l east cost
procurement and i nvestment deci si ons, hampered
attempts to rai se pri ces to effi ci ent l evel s, mandated
l ow sal ari es ti ed to ci vi l servi ce l evel s, and promoted
excessi ve staffi ng. Thi s i n turn has resul ted i n
i nadequate management, the l oss of experi enced
staff due to uncompeti ti ve empl oyment condi ti ons
and poor job sati sfacti on, weak pl anni ng and de-
mand forecasti ng, i neffi ci ent operati on and mai nte-
nance, hi gh l osses, and poor fi nanci al moni tori ng,
control s and revenue col l ecti on.
22
Manpower probl ems are exacerbated by the l ack
of standardi zati on of equi pment,
23
whi ch makes the
l earni ng process more di ffi cul t. Another di sadvan-
tage of the mul ti pl i ci ty of equi pment i s the di ffi cul ty
of mai ntai ni ng adequate suppl i es of spare parts.
Pri ci ng pol i ci es are frequentl y i denti fi ed as a
major i nsti tuti onal probl em. Energy pri ci ng pol i ci es
vary i n devel opi ng countri es, refl ecti ng di fferences
i n energy resource endowments and soci al and
devel opmental pol i ci es. However, pri ce control s on
energy products, such as kerosene and some el ectri c-
i ty pri ces, are a common feature i n many countri es.
Whi l e l ow pri ces hel p to make energy more afforda-
bl e, they can al so resul t i n a l evel of revenues
i nadequate to cover costs and fi nance future suppl y
expansi on. Many anal ysts have characteri zed thi s as
a common probl em i n the el ectri ci ty sectors of a
wi de range of devel opi ng countri es.
24
Zlworld B-A Review of World Bank Lending for Elecmc Power (Washington, ~: tih 1988), p- 74.
%fohan M~in@q current Power Sector issues in Developing Counrries (WaShirIgtOrL DC: World B*, November 1986), P. 14.
23~ ~i, for ~xwp]e, ~ae we ~ diesel generator sets fro
m
17 different ~~ac~ers (s= U.S. Ag~cy for hte~tional Development, Ekztric
Power Utility Efficiency Improvement Study, draft core report, May 15, 1990).
Mfiu, op. cit., footnote 19; ad Do~d He
rtzmar~ Energy Efficiency and Energy Pricing in Developing Countries, OTA contractor repq June
1990.
Chapter l--Introduction and Overview 15
Figure 1-5-Suspended Particulate Matter Levels in Selected Cities, 1980-84
Shown is the range of annual values at individual sites and the composite 5-year average for the city.
WHO Guideline 60-90 ug/m
3
10 ~ 1 0 0
1,000
I I I I I 1 I I I I I I I I I I I I
Key
1. Kuwait
2. Shenyang
3. Xian
4. New Delhi
5. Beijing
6. Calcutta
7. Tehran
8. Jakarta
9. Shanghai
10. Guangzhou
11. Illigan City
12. Bangkok
13. Bombay
14. Kuala Lumpur
15. Zagreb
16. Rio de Janeiro
17. Bucharest
18. Accra
19. Lisbon
20. Manila
Range of individual site
I I
{ 1.
annual averages
1
2.
t 1
I
3.
- F
/ I
4.
I
5.
Combined site 1 1 ]
6.
average 1980-84
I I
7.
I
8 .
1 I 19.
r
I
I
10.
I 11.
[
I
12.
I I
13.
I
I
14.
I
15.
1 I
16.
I
17.
/
18.
19.
20.
SOURCE: World Health Organization and United Nations Environment Fund, G/oba/ Po//@.on and Health (1.mdon: Yale University Press, 1987), figure 3.
Environmental Degradation in Developing
Countries
Devel opi ng countri es are experi enci ng accel erat-
i ng rates of envi ronmental degradati on and pol l u-
ti on. Whi l e many factors contri bute to envi ron-
mental degradati on, energy producti on and use pl ay
key rol es, especi al l y i n urban envi ronmental qual i ty.
Even at present l evel s of energy generati on and use,
the i mpacts on envi ronmental qual i ty are severe i n
many areas. Addi ti onal l arge i ncreases i n energy use
wi l l exacerbate the si tuati on unl ess steps are taken to
mi ti gate adverse envi ronmental i mpacts. At the
same ti me, energy i s an essenti al i nput to such
envi ronmental control systems as sewage treatment.
The combusti on of fossi l fuel s has l ed to l evel s of
ai r pol l uti on i n ci ti es of devel opi ng countri es that are
among the hi ghest i n the worl d (see fi gures 1-5 and
1-6). The transportati on sector i s the l argest contri b-
utor to ai r pol l uti on i n many ci ti es.
25
The combus-
ti on of oi l or gas i n stati onary sources, such as
el ectri c generati ng uni ts, factori es, and househol ds,
al so contri butes through emi ssi ons of ni trogen
oxi des, parti cul ate, sul fur di oxi de, carbon di oxi de,
carbon monoxi de, and hydrocarbons. The fossi l fuel
mi x has an i mportant i mpact on emi ssi on l evel s.
Coal i s the most del eteri ous of fossi l fuel s i n terms
of emi ssi ons per uni t of useful energy provi ded,
parti cul arl y when i t i s not burned i n modern,
wel l -operated pl ants.
I n addi ti on to the envi ronmental damage caused
by the combusti on of fossi l fuel s, thei r producti on
and transportati on al so i mpose envi ronmental costs,
such as di sturbance of l ands and aqui fers from coal
mi ni ng, and acci dental l eaks and spi l l s duri ng oi l
and gas producti on and refi ni ng. Ai r qual i ty i mpacts
i ncl ude, for exampl e, l ocal ai r pol l uti on from
parti cul ate and other emi ssi ons duri ng coal mi ni ng
preparati on, and transport and the rel ease of meth-
ane, duri ng coal mi ni ng and natural gas producti on
and transportati on.
Non-fossi l energy sources such as hydroel ectri c
devel opment al so causes envi ronmental damage.
Dam constructi on often requi res the cl eari ng of
l ands for access routes and removal of constructi on
materi al , wi th resul ti ng soi l degradati on and erosi on.
Fi l l i ng the reservoi r fl oods l arge tracts of l and,
whi ch usual l y means l oss of agri cul tural l and,
human settl ements, fi sh producti on, forests, wi l dl i fe
~hh~mcities, for example, gasoline-fuekdv ehickx-mostly two and thre wheeler+ar
eresponsible for 85 percent of carbon monoxide and 35
to 65 percent of hydrocarbons, while diesel vehicles-+uses and trucks-are responsible for over 90 percent no NOX emissions. Tats Energy Research
Instit&e, TERIEnergy Data Directory and Yearbook 1988 (New Delhi, India: 1989), p. 250.
.-
Figure 1-6-Sulfur Dioxide Levels in Selected Cities, 1980-84
3
1
Concentration (ug/m
3
)
10
/ \
100
I I I I I I I I I I I I I I I I I I I
I I 1
Key
1. Milan
2.Shenyang
4. Seoul
5. Rio de Janeiro
6. Sao Paulo
7. Xian
8. Paris
9. Beijing
10. Madrid
11. Manila
12. Guangzhou
13. Glasgow
k 14. Frankfurt
15. Zagreb
16. Santiago
17. Brussels
18. Calcutta
19. London
20. New York City
annual averages
Combined site
average 1980-84
I
I
j 7.
I
I 8.
1 I 19.
I
, J
I I
J 10.
1 I
I 11.
i
~ 12.
/
I 13.
14.
I
1 5 .
I
] 18.
] 17.
I I
[ 16.
1 I I 19.
1
[ 20.
SOURCES: Worfd Health Organization and United Nations Environment Fund, G/oba/ Pohtion and Hea/th (Ixmion: Yale University Press, 1987), figure 2.
habi tat, and speci es di versi ty.
26
The dam and reser- where there i s a hi gh popul ati on densi ty or a
voi r i nterrupt the free fl ow of surface water, affect concentrated urban demand for fuel wood, such as
water tabl es and groundwater fl ow, and di srupt the Afri can Sahel .
28
downstream fl ows of water and nutri ent-l aden sedi -
ments. These envi ronmental costs are al l the more
When fuel wood i s i n short suppl y, rural popul a-
onerous i f the l i fespan of hydroel ectri c projects i s
ti ons turn to crop resi dues and dung for thei r fuel
reduced through heavi er than expected si l tati on of
needs. To the extent that these forms of bi omass
reservoi rs from deforested and/or degraded l ands
woul d have been used as ferti l i zers, thei r di versi on
upstream.
27
to fuel contri butes to l owered soi l producti vi ty.
29
Burni ng bi omass, the source of energy used by Fi nal l y, bi omass fuel combusti on has a si gni fi cant
most of the devel opi ng worl ds popul ati on, al so i mpact on ai r qual i ty. Food i s typi cal l y cooked over
causes envi ronmental degradati on. Al though the use open fri es or poorl y vented stoves, exposi ng house-
of bi omass for fuel i s onl y one, and not the pri nci pal , hol d members--parti cul arl y women and chi l dren
cause of deforestati on, i t does add addi ti onal pres- to hi gh l evel s of toxi c smoke. Si mi l arl y, i n col der
sure on forest resources especi al l y i n ari d or cl i mates, homes i n rural areas are often heated by
semi -ari d regi ons where forest growth i s sl ow and open fi res, wi th i ncreased exposures to toxi c smoke.
30
~FOre~pl e, thereservo~fi osofi o Onthe Volta inGha~ withalandrequirementof 8,730km
2
, approaches the size of countries Suchris hbmon
or Cyprus. See R.S. Pan&y (cd.), Man-made Lakes and Human Health (Parsmari lxx University of Suriname, 1979).
~The ~~d ~semoh ~ ~% for emple, wss exFted to be productive for about 110 years, but now has an es-ted Pr~uctive lifetime of
35 years. U.S. Congress, Of3ke of Technology Assessment, Technologies To Sustain Tropical Forests, OIILF-214 (WssMngtoq DC: U.S. Government
Printing Oftlce, March 1984), p. 43.
28~ug~ F. Bmnes, World Bs& Poptition Grow@ Wood Fuels, ~d Resource Problems ~ SU~S_ Mc% ~d~try ~ ~~8Y
Department Working Paper No. 26, March 199Q R.P. Moss and W.B. Morgarq Fuelwood and Rural Energy Production and Supply in the Humid
Tropics, United Nations University, Tycooly International Publishing, Ltd., Dublin, 1981; Daniel Finq Land Use and Abuse in the East Mica
RegiorLAMBIO, vol. 12, No. 6, 1983, pp. 296-301, Dennis Anderson and Robert Fiahwic~ World Bar& Fuelwood Comption and Defomatation
in African Countries, staff working paper No. 704, 19S4.
29~gafic ~tter~soflspovidesm ost of the fitrogen~d m ad ss much= ~the pl l ospho~ n~~~p~n~. It helps the ~ilbhld impo~t
minerals suchasmagnesiurq calchq andpotaasiumthat would otherwise be leached away. It buffera the acidity of soils, and it improves water retention
and other physical characteristics. See Geofffey Barnard and Lars Kristofersoq Agricultural Residues as Fuel in the Third Worli.1, Earthscaq
hlternationalInstitute for Environment and Developmen~ Energy InfonnationProgram, technical report No. 4 (Washington DC andl-mdom EarthscaQ
1985).
= R. Smi@ Biojkels, Air Pollution, and Health:A Global Review (New York NY: Plenum Presa, 1987).
Greenhouse Gases and Developing Countries
Energy use i n devel opi ng countri es al so contri b-
utes to i ncreased emi ssi ons of greenhouse gases and
associ ated gl obal cl i mate change. An i nternati onal
panel of sci enti fi c experts of the I ntergovernmental
Panel on Cl i mate Change (I PCC)
31
recentl y con-
cl uded that:
. . . emi ssi ons resul ti ng from human acti vi ti es are
substanti al l y i ncr easi ng the atmospher i c concentr a-
ti ons of the greenhouse gases: carbon di oxi de,
methane, chl orofl uorocarbons (CFCs) and ni trous
oxi de. These i ncr eases wi l l enhance the natur al
greenhouse effect, resul ti ng on average i n an addi -
ti onal warmi ng of the Earths surface.
32
Based on current model s, the panel predi cts that,
under a busi ness as usual scenari o, the gl obal
mean temperature wi l l i ncrease at a rate of about 0.3
C per decade duri ng the next century, a rate at l east
10 ti mes hi gher than any seen over the past 10,000
years.
33
Atmospheri c concentrati ons of CO
2
have i n-
creased by about 25 percent si nce prei ndustri al
ti mes, l argel y due to emi ssi ons from the burni ng of
fossi l fuel s and from deforestati on. I n 1985, accord-
i ng to another I PCC worki ng group, devel opi ng
countri es contri buted about one-quarter (26 percent)
of annual gl obal energy sector CO
2
emi ssi ons.
%
Under the busi ness as usual scenari o, wi th
expandi ng popul ati ons, rapi dl y i ncreasi ng energy
use, and assumi n g the absence of control measures,
the devel opi ng country share woul d i ncrease to 44
percent of annual energy sector emi ssi ons by 2025.
35
The magni tude of C0
2
emi ssi ons from fossi l fuel
sources i s fai rl y wel l known, but the contri buti on
from deforestati on, whi ch i s vi rtual l y al l from
devel opi ng countri es, cannot be esti mated accu-
ratel y. Thi s makes i t di ffi cul t to cal cul ate wi th
confi dence the devel opi ng country share of gl obal
annual and cumul ati ve emi ssi ons for CO
2
and other
gases. Esti mates of CO
2
emi ssi ons from tropi cal
deforestati on di ffer by a factor of 4.
36
By vari ous
esti mates, deforestati on coul d be the source of
between roughl y 7 and 35 percent of total annual
CO
2
emi ssi ons.
Despi te uncertai nti es, i t i s safe to concl ude that
the devel opi ng countri es al ready contri bute a sub-
stanti al part of current annual gl obal CO
2
emi ssi ons,
and that thei r share wi l l i ncrease i n the future. But,
because of thei r l arge and rapi dl y growi ng popul a-
ti ons, per-capita CO
2
emi ssi ons i n the devel opi ng
countri es wi l l sti l l remai n much l ower than i n the
countri es of the i ndustri al worl d. Devel opi ng coun-
tri es al so account for at l east hal f of the gl obal
anthropogeni c generati on of two other i mportant
greenhouse gases, methane and ni trous oxi de.
Devel opi ng and i ndustri al countri es woul d both
be damaged by the anti ci pated i mpacts of cl i mate
change. I n addi ti on to i ncreases i n mean gl obal
temperature, other major effects of gl obal cl i mate
change woul d i ncl ude i ncreases i n sea l eve1
37
and
shi fts i n regi onal temperature, wi nd, rai nfal l , and
sl~e ~tergovmma~ panel n Climate -e (WCC) is an international task force created by the World Meteorological Ggatition md tie
United Nations Environment Program. Initiated in 1987, this body focuses on analyzing current information available on climate change issues and
devising strategies to address climate change.
qzrntergov-en~ Panel on Climate Change, POliqmakers s~
of the Scientiilc Assessment of Climate Change: Report to IPCC from
Working Group I, June 1990, p. i.
qqId., atp. ii. fhis would m- an incr~ over the preindustrial global average temperature of 2 degrees centigrade by 2025 ~d 4 C by 2@0. TJlis
best estimate prediction has an uncertainty range of 0.2 to 0.5C per decade. This was based on projections derived from another working group that
emissions of C02 could grow from approximately 7 billion tonnes of carbon (BtC) in 1985 to between 11 and 15 BtC in 2025. Methane emissions were
projected to increase from 300 teragrams in 1985 to 500 teragrams by 2025.
34rntergov~nmen~ panel on c~te -q wor~g &OUp m @sponse s@&@S worki ng @up), Policyrnakers S~ Of the
Formulation of Response Strategies: Report Prepared for IPCC, Executive Summary, p. 10, table 2, June 1990.
qsOTAc~c~ation~~n2030 ~@ ~sio~~wer&ofi Scefio, @bles A-21 ~dA-193 h rntergovernmentipmel On Climate -e,
Emissions Scenarios Prepared by the Response Strategies Working Group of the Intergovernmental Panel on Climate Change, Appendix Report of
the Expert Group on Emissions Scenarios (RSWG Steering Committee, Task A), April 1990.
WPCC, Policymakexs Summary of the Formulation of Response Strategies: Report Prepared for JPCC by Working Group III, June 1990, p. 5.
Estimates of C02 emissions from deforestation and land use changes (including wood fhel) were 0.6 to 2.5 PtC in 1980 according to a report prepared
for the WCC Working Group 1. Scientific Assessment of Climate Change: Peer Reviewed Assessment for WG1 Plenary Meeting, May 1990, Apr.
30, 1990, p. 1-9. Mom recent evidence iiom selected countries would indicate deforestation rates now are higher than they were in 1980. Estimates of
C02 emissions in 1987 were 5.7 +0.5 PtC from fossil fuel burning and total emissions were of 7 +1.1 PtC.
qT~e IPCC wor~g ~OUp predi cted an ave~ge mte of global mean sea level rise of about 6 cm per decade over the next century-20 cm by 2030
and 65 cm by the end of the century (with signillcant regional variations). This predicted increase is due primarily to thermal expansion of the oceans
and melting of some land ice.
storm patterns. These effects i n turn woul d sub-
merge some l ow-l yi ng coastal areas and wetl ands,
threaten bui l di ngs and other structures i n these areas,
and i ncrease the sal i ni ty of coastal aqui fers and
estuari es. Such changes woul d di srupt human com-
muni ti es and aquati c and terrestri al ecosystems, and
affect food producti on and water avai l abi l i ty. A
number of devel opi ng countri es woul d be especi al l y
vul nerabl e to ri si ng sea l evel s through threats to
coastal communi ti es and ecosystems, i ncreased
vul nerabi l i ty to storm damage, and i nundati on of
l ow-l yi ng arabl e l ands.
38
The adverse effects of
cl i mate change woul d exacerbate the i mpacts of
i ncreased popul ati ons i n Asi a, Afri ca, and smal l
i sl and nati ons of the Cari bbean, I ndi an, and Paci fi c
Oceans.
39
The ti mi ng, severi ty, and extent of these
potenti al i mpacts remai n uncertai n.
Prospects for Efficiency
Improvements in Energy
Production and Use
I ncreasi ng energy suppl i es i n the busi ness as
usual mode that underl i es the conventi onal projec-
ti ons thus poses formi dabl e chal l enges. Thi s unfor-
tunate fact suggests that al ternati ve approaches for
provi di ng the vi tal energy servi ces needed for rapi d
economi c and soci al devel opment shoul d be i nvesti -
gated, focusi ng on effi ci ency i mprovements.
OTA exami ned the fol l owi ng energy-rel ated serv-
i ces: cooki ng, l i ghti ng, and appl i ances i n the resi -
denti al and commerci al sectors; process heat and
el ectri cal and mechani cal dri ve i n i ndustry and
agri cul ture; and transportati on (see ch. 3). For the
devel opi ng countri es as a whol e, the l argest energy
end use markets are resi denti al /commerci al and
i ndustry, whi ch together account for roughl y 85
percent of the energy used by fi nal consumers,
i ncl udi ng tradi ti onal energy. Consi derabl e di ffer-
ences exi st, however, among devel opi ng nati ons. I n
Afri ca, the resi denti al /commerci al sector consti tutes
a parti cul arl y hi gh share (mostl y i n the form of
bi omass fuel s for cooki ng), whi l e i ndustrys share i s
qui te l ow. I n Lati n Ameri ca, transportati on accounts
for an excepti onal l y hi gh share of the total , whereas
i ts share i n I ndi a and Chi na i s qui te l ow.
I ndustri al process heat and cooki ng are the l argest
ener gy ser vi ces, each accounti ng for about one thi r d
of al l energy consumed (commerci al pl us bi omass)
i n devel opi ng countri es. Thi s pattern of energy use
contrasts wi th the Uni ted States, where transporta-
ti on and space condi ti oni ng are the hi ghest (al though
i n the Uni ted States process heat i s a major user as
wel l ). Much of the energy used for resi denti al
cooki ng and process heat i n the devel opi ng countri es
i s consumed i n Chi na and I ndi a. Together thei r
consumpti on of energy for cooki ng and i ndustri al
process heat accounts for over 40 percent of al l
cooki ng and process heat energy used by devel opi ng
countri es, and for wel l over one-quarter of al l the
energy consumed i n devel opi ng countri es.
A wi de range of technol ogi es are currentl y used to
provi de energy servi ces i n devel opi ng countri es. For
exampl e, cooki ng technol ogi es i ncl ude stoves usi ng
fuel wood, charcoal , kerosene, l i qui d petrol eum gas,
natural gas, and el ectri ci ty, al l wi th di fferent charac-
teri sti cs. These technol ogi es vary wi del y i n thei r
energy efficiency. In an open fi re, for exampl e, onl y
about 15 percent of the energy contai ned i n fuel -
wood goes i nto cooki ng. I n contrast, i n a modern
gas stove about 60 percent of the energy contai ned
i n the gas i s used i n cooki ng. The wi de range of
effi ci enci es i n the current stock of stoves suggests
opportuni ti es for i ncreasi ng effi ci enci es of the stock
and therefore provi di ng more cooki ng servi ces wi th
l ess energy.
There are al so di fferences i n effi ci enci es i n
provi di ng energy servi ces i n the i ndustri al sector
i ndustri al process heat and el ectri c and mechani cal
dri ve. The two l argest devel opi ng country energy
consumers, I ndi a and Chi na, currentl y rel y on
several technol ogi es that are a generati on or more
behi nd the state of the art, and are much l ess
energy-effi ci ent than technol ogi es now bei ng used
i n the Uni ted States and other countri es. I ntegrated
i ron and steel pl ants i n Chi na and I ndi a, for exampl e,
use twi ce as much energy per ton of crude steel
produced as i ntegrated pl ants i n the Uni ted States
and Japan. Lower effi ci enci es are al so frequentl y
observed i n the transportati on sector.
An anal ysi s of the energy suppl y i ndustry i n
devel opi ng countri es si mi l arl y i ndi cates much l ower
ssktergovwnm~~ Panel on Climate Change, Policymalcers SUmmary of the Potential Impacts of Climate Change: Report from Worlcing Group
II to the IPCC, May 1990, p. 8.
39J$D. W* et al,
Environmental and Economic Implications of Rising Sea Level and Subsiding Deltas: The Nile and Bengal Examples,
AMBIO, vol. 18, pp. 340-345, 1989.
operati ng effi ci enci es than i n the i ndustri al coun-
tri es. I n el ectri ci ty generati on, for exampl e, thermal
power pl ants frequentl y operate far bel ow desi gn
capaci ty and effi ci ency. Transmi ssi on and di stri bu-
ti on l osses (i ncl udi ng unaccounted for l osses, un-
metered use, etc.) are frequentl y over 15 percent,
substanti al l y hi gher than l osses i n i ndustri al country
systems.
40 Refi neri es also operate at much l ower
effi ci enci es.
Energy suppl i es i n many devel opi ng countri es are
unrel i abl e, i mposi ng a heavy economi c burden. I n
I ndi a, for exampl e, l osses sustai ned by i ndustry due
to unr el i abl e el ectr i c power suppl i es i n r ecent year s
are esti mated to represent 2 percent of annual GNP,
not i ncl udi ng l osses i n agri cul ture or l osses and
i nconveni ence experi enced by resi denti al and com-
merci al users. Si mi l ar l osses have been esti mated for
Paki stan. Furthermore, el ectri ci ty suppl i es i n many
countri es are of poor qual i ty, di scouragi ng the use of
effi ci ent technol ogi es that are cri ti cal l y dependent
on hi gh-qual i ty energy suppl i es.
I n characteri zi ng i mportant parts of the energy
system as i neffi ci ent, however, i t shoul d be
real i zed that i n many cases users and producers are
acti ng l ogi cal l y gi ven the framework of resources,
i ncenti ves, and di si ncenti ves wi thi n whi ch they
make thei r deci si ons. One of the reasons that poor
househol ds use fuel wood i neffi ci entl y i s that they
l ack the fi nanci al means to buy more effi ci ent
cooki ng systems. I ndustri al users must cope wi th
anti quated machi nery and errati c fuel suppl i es of
uncertai n qual i ty. On the suppl y si de, the record of
poor performance refl ects many factors: poor
repai r and mai ntenance, unavai l abi l i ty of spare
parts, l ow fuel qual i ty, ol der equi pment, unsati sfac-
tory management, l ack of ski l l ed workers, probl ems
of reachi ng di spersed popul ati ons served by i nade-
quate transport systems, and i nappropri ate pri ci ng
and al l ocati on systems.
The exi stence of wi de di fferences between opera-
ti onal effi ci enci es i n reasonabl y standardi zed opera-
ti ons (e.g., cooki ng, steel maki ng, el ectri ci ty genera-
ti on, and petrol eum refi ni ng), both among devel op-
i ng countri es and between the devel opi ng and
i ndustri al countri es, suggests that dramati c i mprove-
ments i n effi ci enci es are possi bl e. However, the
i mportance of factors other than technol ogy must be
recogni zed for the rol e they pl ay i n i mprovi ng
effi ci enci es. The pol i cy envi ronment i n parti cul ar i s
cruci al to the adopti on of new technol ogi es.
More effi ci ent ways of provi di ng energy servi ces
for devel opment, i ncl udi ng both technol ogi es and
the i nsti tuti onal and pol i cy mechani sms determi ni ng
thei r rate of adopti on, wi l l be presented i n a l ater
report of thi s OTA assessment. Attenti on wi l l al so be
pai d to the energy i mpl i cati ons of di fferent devel op-
ment strategi es. Some devel opment strategi es are
associ ated wi th hi gh energy use. But devel opi ng
countri es at the begi nni ng of the devel opment
process may be abl e to capi tal i ze on technol ogy to
devel op toward modern economi es wi thout the hi gh
energy growth that earl i er characteri zed the path to
i ndustr i al i zati on.
ah tie utit~ Stites, for ex~ple, transmission and distribution losses in dense urban service areas are between 6 and 7 percent and irI rural semice
areas nearer 9 to 10 percent.
2
0

q

E
n
e
r
g
y

i
n

D
e
v
e
l
o
p
i
n
g

C
o
u
n
t
r
i
e
s
u
)
(
9
4
)

z
8
6
1

s
=
w
e
s
p
u
e

s
p
o
o
6

J
O

s
~
o
d
x
e

J
O
e
J
e
q
s

e
s
8

e
a
!
w
e
s

I
q
e
a
(
%
)

Z
8
6
1
q
m
d
x
e

e
s
!
p
u
e
q
w
e
w

J
O
a
)
c
o
t
o
I
I I
I

R
%
R
I
-
,
C
D
.

.

.
.

.

.
.

.

.
.

.

.
.

.

.
.

.

.
.

.

.
Appendix 1 A-Economic, Social and Energy Indicators for Developing Countries
2 3 4 5 7 9 10 11 12 13 14
iii

f!!
;

c

::i:
c
0
.!!l
aI CI)

0::
.5 ;:-
..r::
..r:: aI .5 .2-
E- a
CI)
lIIi::i:
.....
III 1::
..... (iI
'0
o.c

E-
.2 aI iii !.-
:&.
l!aI
0.
:0
co
0.
Cl)CO gj
co..r::

;:,d
iii

: -g::i:

1ii
0)1::
..r:: 111-
6'0
!Q
.a .....
if
;':0
0. o>g
oal o,!!!
C<o
OCl)O)
III CI) !Q O!.-
aI aI

g 8m

:a. :a. l1i;::-

CD
g,"i
. 'g.
6 'Q.

1::.!!!
0> .....
+=
aI
0.

&."2

B B
a. co
+=B
._ 00)

1::0 c '0 -0 -0) E iii .....

CD

8.
0
_al*
c. ... c. ...
0:: Cl)o E
'>'
Cl)1::111

E CI) E CI)
111
0.
c1ii

;:,0. ;:,0.
.!B
CI) aI c CI);:'

a..
Cl)o.
CI) .....
1:":
"S.q

15
a..

ale
Cl)CO
:!
0. co
co
,!!! ... Cl)Eco
z z -e-
CI) 0) o CI) o CI)

..r::-
..r::-

00)
(/).5
('I ,It ,n 0
""'''''
a.. ..... a.. III"'" 0.2 0.2
low-Income countries
Average .............. 280w 3w 24w 54w 103w 13,550w 9w 16w 21.9w
1 Ethiopia ........... 130 304 42 12 47 154 77,360 12 8.19 0.67 8.86 92 100 55 28.4
2 Bhutan .... , ....... 150 51 5 48 128 23,310 4 22.31 0.00 22.31 100 100
3 Chad ............. 150 254 43 30 46 132 38,360 6 5.85 0.57 6.42 91 100 3.9
4 Zaire ............. 150 200 32 38 52 98 8.68 3.01 11.69 74 x 2 12.8
5 Bangladesh ........ 160 646 47 13 51 119 6,730 18 2.46 0.03 2.49 99 40 21 24.2
6 Malawi .... , ....... 160 37 13 46 150 11,560 4 7.97 1.77 9.75 82 100 10 23.3
7 Nepal ............. 160 526 57 9 51 128 32,710 25 8.64 0.85 9.49 91 92 31 9.7
8 Lao PDR .......... 170 17 48 110 1,360 19 10.00 3.42 13.42 75 75
9 Mozambique ....... 170 528 50 23 48 141 37950 7 9.59 0.89 10.48 92 100
10 Tanzania .......... 180 291 61 29 53 106 3 9.29 1.26 10.54 88 100 56 18.5
11 Burkina Faso ....... 190 359 38 8 47 138 57,180 6 7.83 0.72 8.55 92 100 7
12 Madagascar ....... 210 370 43 23 54 120 10,000 36 5.87 1.19 7.06 83 100 36 35.3
13 Mali .............. 210 337 54 19 47 169 25,390 7 6.03 0.64 6.67 90 100 32 9.9
14 Burundi ........... 250 338 59 7 49 112 21,120 4 7.20 0.60 7.80 92 100 8 38.5
15 Zambia ........... 250 579 12 53 53 80 7,100 19 13.19 19.58 32.78 40 33 11 13.5
16 Niger ............. 260 320 34 18 45 135 38,770 6 5.44 1.47 6.91 79 100 9 33.5
17 Uganda ........... 260 319 76 10 48 103 21,900 7.01 1.15 8.15 86 100 17 19.5
18 Somalia ........... 290 348 65 36 47 132 16,090 12 7.72 2.63 10.35 75 100 9 8.3
19 Togo ............. 290 489 29 24 53 94 8,720 21 1.88 1.56 3.44 55 100 8 14.2
20 Rwanda ........... 300 388 37 7 49 122 34,680 3 8.59 1.25 9.84 87 100 53 11.3
21 Sierra Leone ....... 300 441 45 26 41 151 13,630 20.00 1.84 21.84 92 100 10
22 Benin ............. 310 531 46 39 50 116 15,940 16 9.77 1.40 11.16 88 x 97 15.9
23 Central African Rep. . 330 436 41 45 50 132 23,070 13 10.74 0.74 11.48 94 100 1 12.1
24 Kenya ............ 330 603 31 22 58 72 10,100 20 14.43 2.94 17.38 83 100 39 28.8
25 Sudan ............ 330 562 37 21 50 108 10,110 20 8.31 2.12 10.43 80 100 38 6.8
26 Pakistan ..... ..... 350 1,165 23 31 55 109 2,900 18 2.19 8.49 10.67 20 39 26 25.9
27 Haiti .............. 360 591 29 55 117 7,180 18 9.51 1.80 11.31 84 93 16 7.0
28 Lesotho ........... 370 795 21 19 56 100 18,610 22 0.00 0.00 0.00 4.4
29 Nigeria ............ 370 565 30 33 51 105 7,980 8.32 4.82 13.14 63 x 310.0
30 Ghana ............ 390 361 51 32 54 90 14,890 35 6.10 5.07 11.18 55 100 14 19.2
31 Sri Lanka .......... 400 1,491 27 21 70 33 5,520 66 4.76 4.82 9.57 50 87 25 19.2
32 Yemen, PDR ....... 420 1,038 16 42 51 120 4,340 0.00 15.22 15.22 0 70 38.2
33 Mauritania ......... 440 539 37 38 46 127 12,110 15 0.00 4.21 4.21 0 100 8 18.2
34 Indonesia .......... 450 1,269 26 27 60 71 9,460 41 7.70 8.47 16.17 48 x 13 27.8
35 Liberia ............ 450 539 37 42 54 87 9,240 17.39 4.78 22.17 78 100 11 2.5
37 Burma ............ 562 24 60 70 3,740 4.07 2.65 6.72 61 1 5 59.3
38 Guinea ........... 24 42 147 57,390 9 5.69 2.46 8.15 70 100
39 Kampuchea, Oem .... 42
87 100
40 Viet Nam .......... 66 46 1,000 43 3.40 3.52 6.92 49 20
China and India
Average ....... 300w 30w 33w 5w 62w 1,640w 25w
5w 10.7
41 China ...... 290 1,489 31 38 69 32 1,000 42 1.70 21.15 22.00 7 x 2 7.1
42 India ...... 300 775 30 27 58 99 2,520 35 3.01 8.34 11.36 27 14 17 18.9
1 1 1 1
.
?HWMHJH i?
mm-cJm-N-mmm-w K
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . .
. . .
:9 : : : : : :&: : : : : : : : : : : : : : : : : :
.
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
. .
, .
.
.
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .(
. . . . . . . . . . . . .
. . . . . . . . . . . . .
:&
. . . . . . . . ......... .m
.
I
x
Chapter 2
Energy and Economic Development
Contents
Page
I ntroducti on and Summary
... .*. *.. ... .,. ... ..**
e
*. **. ..**, *,** ***. *
+
.
e
*
8
****,
v
* *e.
25
Economi c Devel opment and I ts I mpact on Energy
q ....**,.*........****,***.**.****
25
Ri si ng Popul ati ons
*.** *.. ... .+. *.. ... *.. .**. ... ..,
+
**
q
. *,*
+
**
6
* *,
+
* *, ******e,*
26
Hi gher Li vi ng Standards
.*. ..*. ... ... ... ... ... ... **. ...
+o
***. **** .*
c
*. ..** *.
o c
*.
27
Changes i n Energy Consumpti on Wi th Economi c Devel opment . . . . . . . . . . . . . . . . . . . 27
projected Energy Consumpti on i n Devel opi ng Countri es
., *...****.*....,,+,,.**.**,*
33
The Energy Sector and the Macroeconomy
... .*. ..** *,. ... ,.. *.. ... *+ *****6+,****.*
34
Fi nanci ng Energy Suppl i es
*.. ... *.. ... ... ,***. .,** ***
c c
*
Q
**, **, **, **** **. *Q**,*
34
Energy Suppl y Rel i abi l i ty
... ..* ........,.*..**.**+.*****,**,.**,**,*****+**,*,.*
36
Energy Pri ci ng and Demand Management
q *.. ... .*. ... ... ... ..*. ,*+**e* **, **, Q*.,*+
36
Energy and The Tradi ti onal Sector
... ..*. ... .*. ... ... .,. ... ..,
e
,#*
<
. **, **. .
a
*,***.*
39
Seasonal i ty .*. .., *.*. ... ... .*, **. ..** **** **
o
* **** *
ee
. **
+
*co*******,***,++***.*
40
I nequi ti es i n Resource Di stri buti on and Access
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
42
The Rol e of Women
... ... ... ... ... .,. **. ..*. **** ***
o
.**, *
e
. **** **,
Q
*
a
*, ***,
o
**
42
The Rol e of Commerci al Bi omass i n the Rural Economy
q .**.....,.*.....*.,,****
43
Concl usi on ... ..*. *.. ... ,.. ... ... *.. **, ,
e
*
o
#** **** **** *,*
4
**. ***
+
*,
4
.
oo
***
&
*
+
**c@.
44
Boxes
Box
Page
2-A. Factors Affecti ng Popul ati on I ncrease
... ... .*. ... ... ... ... .., .,. ,,*, ,**, ,
ea
#*+
28
2-B. Energy Conservati on I ni ti ati ves i n ASEAN Countri es
q . . . . . . . . . . . . . . . . . . . . . . . . .
39
Figures
Figure
Page
2-1. Worl d Popul ati on Growth, 1750-2100 i n I ndustri al and Devel opi ng Regi ons . . . . . . 26
2-2. Average Annual I ncrease i n Popul ati on Per Decade i n I ndustri al
and Devel opi ng Regi ons, 1750-2100
... ... ... **. ... ..*. .,. *.. ,*, *
26
2-3. Hi stori cal and Projected Gl obal Popul ati on, 1950-2025
q q q q . q q ~ q . . . q . . . . . . . * q q ..27
2-4. Projected Shares of Gl obal Popul ati on, 2000 and 2025
q ..,.....*......,+*,*,.**.
27
2-5. Commerci al Energy Consumpti on and Economi c Devel opment i n
Sel ected Countri es ... ... ..*. ... ... .,. *.. ... ... ,*, ***. .**. **** *
Q
. *
e
** .**
+ o
**,**
31
2-6. Commerci al Energy Consumpti on and Soci al Devel opment
i n Sel ected Countri es
**** **. .*. .
$
, .*. ... ... .*, *.. ,,, ***. **, ,,*
+
*
+
*
vc <a
,,***,**
32
2-7. Energy I ntensi ty and Economi c Devel opment, 1960-2000. . . . . . . . . . . . . . . . . . . . . . . 33
2-8. Hi stori cal and Projected Energy Consumpti on i n Devel opi ng Countri es:
Worl d Energy Conference Moderate Projecti ons
. ....*..,......,.***..*,*****
34
2-9. Seasonal Pattern of Energy Expendi ture on Agri cul tural and Other Essenti al Tasks:
Adul t Farmers, Geni eri Vi l l age
... ... ... ... ... ... .., .*. .*, ,., *o**.,**.+****,*+
41
Tables
Table
Page
2-1. Energy I ntensi ti es i n Sel ected Countri es
. ..........**......+,**.,....@******.*.*
32
2-2. Commerci al Energy Consumpti on, 1985 and Projecti ons for 2010 . . . . . . . . . . . . . . . 35
2-3. Esti mated Annual Energy I nvestment as a Percentage of Annual Total Publ i c
I nvestment Duri ng the Earl y 1980s
. .......,..........*..........,,.*+..*,++***
35
Chapter 2
Energy and Economic Development
Introduction and Summary
Thi s chapter exami nes the two-way l i nkage be-
tween energy and economi c devel opment i n devel -
opi ng countri es--how the process of economi c
devel opment i mpacts energy, and how, i n turn,
devel opments i n the energy sector can affect eco-
nomi c growth.
I n the course of economi c devel opment, commer-
ci al energy consumpti on i s observed to i ncrease
faster than economi c acti vi ty. There are a number of
reasons for thi s: the growth of mechani zed agri cul -
ture and manufacturi ng, the constructi on of a mod-
, i ncreased transpor- em i nfrastructure, urbani zati on
tati on of goods and servi ces, rapi d expansi on i n
ownershi p of consumer appl i ances, and the substi tu-
ti on of commerci al for tradi ti onal fuel s. The absol ute
amount of tradi ti onal energy consumed al so conti n-
ues to ri se, al though i ts share of total energy
consumpti on fal l s.
I n the years to come, hi gh rates of economi c
growth wi l l be needed i n devel opi ng countri es to
provi de thei r rapi dl y growi ng popul ati ons wi th
i mproved l i vi ng standards. I f current trends i n
energy and economi c growth conti nue, commerci al
energy consumpti on i n the devel opi ng countri es
coul d more than doubl e over the next 40 years
accordi ng to most projecti ons. Suppl i es of bi omass
fuel s woul d al so need to i ncrease substanti al l y to
meet the needs of growi ng rural popul ati ons and the
ur ban poor .
Thi s prospect rai ses a di l emma. On the one hand,
i ncreases i n energy suppl i es on thi s scal e coul d
severel y strai n fi nanci al resources i n the devel opi ng
countr i es. The ener gy sector absor bs a l ar ge shar e of
avai l abl e forei gn exchange and capi tal i nvestment.
Consequentl y, energy suppl y pol i ci es have far-
-reachi ng i mpacts on other devel opment pri ori ti es. I n
many devel opi ng countri es, fi nanci al resources may
not be adequate to i ncrease commerci al energy
suppl i es on the scal e projected above.
On the other hand, i nabi l i ty to suppl y needed
energy can frustrate economi c and soci al devel op-
ment. Al ready i n many countri es, the unrel i abi l i ty
and poor qual i ty of energy suppl i es l ead to major
costs to the economy through wasted materi al s,
stoppage of operati ons, and i nvestment i n standby
equi pment.
Energy pri ces are a key factor i n the devel opment
of a countrys energy suppl y i nfrastructure, through
thei r i mpacts on the amount of energy used i n the
economy, the technol ogi es adopted, and, i n some
cases, the di recti on of i ndustri al devel opment. En-
ergy pri ces i n devel opi ng countri es are typi cal l y
subject to pri ce regul ati on throughout the di stri bu-
ti on chai n. The average l evel of energy pri ces,
parti cul arl y i n the el ectri ci ty sector, are reported to
be too l ow i n many countri es to ensure the sectors
fi nanci al vi abi l i ty.
Al though commerci al fuel s attract the most pol i cy
attenti on, two-thi rds of the devel opi ng worl ds
popul ati on l i ve i n rural areas wi th l ow standards of
l i vi ng based on l ow-resource farmi ng. Thi s popul a-
ti on has l i ttl e access to commerci al fuel s and rel i es
l argel y on tradi ti onal sources of energy, gathered
and consumed l ocal l y, and ani mal and human
energy, often used at very l ow effi ci enci es. The mai n
form of tradi ti onal energy used i s wood, an i ncreas-
i ngl y scarce and unsustai nabl e resource. Thi s i m-
poses a speci al hardshi p on those-mai nl y women
and chi l dren-responsi bl e for gatheri ng i t. Dung
and crop wastes, the other forms of energy wi del y
used for cooki ng when wood i s not avai l abl e, have
al ternati ve uses as soi l nutri ents.
Economic Development and Its
Impact on Energy
The pace of economi c growth and l evel of
economi c acti vi ty have major i mpacts on the energy
sector. From 1965 to 1987, for exampl e, the econo-
mi es of the devel opi ng countri es grew at an annual
average of 5.3 percent, and thei r consumpti on of
commerci al energy grew by just over 6 percent.
l
These energy growth rates were hi gher than those i n
the i ndustri al countri es over the same peri od. As a
resul t, the devel opi ng countri es share of gl obal
commerci al energy consumpti on al so rose-from 17
lw~rld Ba~ Wo~ZdDe~elop~ntRepo~t 1989 @ew Yo&, ~: Ofiord universi~ Press, 1989), pp. 167 and 173.
25
33-718 0 - 90 - 2
per cent i n 1973 to 23 per cent i n 1987.
2
Hi gh rates of
economi c growth wi l l conti nue to be needed i n the
devel opi ng worl d to provi de the rapi dl y growi ng
popul ati on wi th i mproved l i vi ng standards.
Rising Populations
The past two centuri es have wi tnessed sharpl y
accel erated growth i n the gl obal popul ati on (see
fi gure 2-l ). The l argest addi ti ons to gl obal popul a-
ti on have been i n the devel opi ng countri es and have
occurred pri mari l y i n the past 50 years (see fi gure
2-2). Box 2-A di scusses factors affecti ng popul ati on
gr owth.
Al though current projecti ons of gl obal popul ati on
growth over the next 35 years di ffer (see fi gure 2-3),
there i s consensus on several major poi nts:
The worl ds popul ati on i s projected to i ncrease
despi te assumpti ons of conti nued decl i nes i n
ferti l i ty rates. The rate of i ncrease i n gl obal
popul ati on, whi l e l ower than i n the past, sti l l
represents a l arge i ncrease i n numbers of
peopl e. Worl d Bank esti mates,
3
whi ch are
si mi l ar to both the Uni ted Nati ons medi um
projecti on and the U.S. Department of Agri cul -
ture projecti on, project an i ncrease i n gl obal
popul ati on from 5.3 bi l l i on i n 1990 to 8.4
bi l l i on i n 2025, an i ncrease of 3.1 bi l l i on.
Vi rtual l y al l of the i ncrease wi l l come from the
devel opi ng countri es. Accordi ng to the Worl d
Bank projecti on, for exampl e, popul ati on
growth i n the i ndustri al countri es--i .e., nati ons
i n the Organi zati on for Economi c Cooperati on
and Devel opment (OECD), the U. S. S. R., and
Eastern Europe-i s expected to add onl y about
125 mi l l i on, or about 4 percent of the gl obal
i ncrease (see fi gure 2-4). The popul ati on of the
devel opi ng countri es i s esti mated to ri se from
i ts present l evel of 4.1 bi l l i on to 7.1 bi l l i on i n
2025, i ncreasi ng thei r share of worl d popul a-
ti on from 77 to 88 percent (see fi gure 2-4).
Popul ati on growth i n Chi na i s projected to be
qui te moderate, as current l ow rates of growth
are assumed to be mai ntai ned. Projecti ons of
Chi nas popul ati on growth are cri ti cal because
of i ts l arge share of the gl obal total .
Accordi ng to Worl d Bank projecti ons, the
bi ggest i ncreases i n popul ati on are predi cted to
Figure 2-1WorId Population Growth, 1750-2100
in Industrial and Developing Regions
Population in billions
12
-
T 7
10
8
6
4
2
0
1750 1800 1850 1900 1950 2000 2050 2100
SOURCE: Thomas MerncK Population Referenee Bureau, World Popula-
tion in Transition, /Jopu/afion Bu//etin, vol. 41, No. 2, April 1986,
update based on United Nations 1989 projections.
Figure 2-2Average Annual Increase in Population
Per Decade in Industrial and Developing Regions,
1750-2100
Population in millions
~ 1 0 0 ~

-
~
40 -
NNN Developing regions
Industrial regions
I
2 0 1 T r l l _ ~ T
-
T - - r I I [ 1 I 1 1 1 I 1 1 1 1 I ! 1
1750 1800 1850 1900 1950 2000 2050 2100
SOURCE: Thomas Merrick, Population Reference Bureau, World Popula-
tion in Transition, Po@ation Bu//etin, vol. 41, No. 2, April 1986,
update based on United Nations 1989 projections.
come from Afri ca and Asi a, whi ch wi l l account
for 30 and 58 percent respecti vel y of the total
gl obal i ncrease. I n Asi a the l arge addi ti on to
popul ati on deri ves from the exi sti ng l arge
popul ati on base; rates of popul ati on growth are
rel ati vel y l ow. I n Afri ca, on the other hand, the
pri mary cause i s the rapi d i ncrease i n popul a-
ti on that, despi te the rel ati vel y l ow popul ati on,
i ncreases i ts share of the total popul ati on from
a cur r ent 12 per cent to 19 per cent i n 2025.
world Energy Conference, GlobaZEnergy Perspective 2000-2020 14th Congress, Montreal 1989 (Paris: 1989), Table 2.
SRMOKOA. B~a~, EdWdBos, patien~ Stephens, and My T VW Europe, MiaWeEast, andA@ca (EMN)Region Population projection: I$WMICI
E&tion, Population and Human Resourees Department, working paper 328 (Washingto~ DC: World Bank), November 1989, table 5.
Chapter 2--Energy and Economic Development q 27
10
8
6
4
2
Figure 2-3-Historical and Projected Global
Population, 1950-2025
Population in billions
U.N. (low, reed, high) ,
--
Worl d Bank
- 1
/
Variations among these projection arise
from dlfferences in projected future fertility rates,
in turn are based on many assumptions, such as the
effects of government policies, GNP per capita,
literacy, and sociocultural factors.
1
1950 1975 2000 2025
Year
SOURCE: Office of Technology Assessment, 1990, based on data in
United Nations, World Population Prospects 1988 (New York,
NY: United Nations, 1989); Rodolfo Bulatao et al., Europe,
Middle East, and Africa (EMN) Region Population Projections,
1989-90 Edition, World Bank, Population and Human Re-
sources Department, Washington, DC, working paper series
328, November 1989.
Though there wi l l no doubt be some unforeseen
di vergence from these popul ati on paths (changes i n
ferti l i ty rates are di ffi cul t to predi ct; the i mpact of
the AI DS epi demi c i n Afri ca on ferti l i ty rates and
popul ati on growth i s unknown), i t i s cl ear that there
wi l l be a l arge i ncrease i n the worl ds popul ati on i n
the decades ahead, accompani ed by a powerful
upward pressure on energy consumpti on. Even wi th
no i ncrease i n per-capi ta energy consumpti on, the
predi cted ri se i n the gl obal popul ati on by 2025
i mpl i es a 75 percent i ncrease i n total commerci al
energy consumpti on.
Higher Living Standards
The major devel opment chal l enge i s to provi de
hi gher standards of l i vi ng for the rapi dl y ri si ng
popul ati ons of the devel opi ng worl d. Thi s task i s al l
the more urgent because of the decl i ni ng l evel s of
per-capi ta i ncome i n many of the countri es of Lati n
Ameri ca and Afri ca i n recent years. Gi ven the
projected ri se i n devel opi ng country popul ati ons
an annual average of about 1.6 percent over the next
Figure 2-4-Projected Shares of Global Population,
2000 and 2025
Billions
10
6
6
4
2
0
I
m Developing regions
I
I 1= i ndust r i al r egi ons I
World Bank projections
1
23%
I I
20%
I
I ,
\ 1
1990 2000 2025
Year
SOURCE: Rodolfo Bulatao et al., Europe, MicWe East, and AMca (EMN)
Region Population Projections, 1989-90 Edition, World Bank,
Population and Human Resourees Department, Washington,
DC, working paper series 328, November 1989.
35 yearn-a ri se i n average per-capi ta i ncomes of,
say, 3 percent per year i mpl i es economi c growth
rates of around 4.6 percent annual l y.
4
I t may not be easy to achi eve such rates of growth.
The current i ndebtedness of many devel opi ng na-
ti ons has added to the al ready di ffi cul t tasks of
economi c management, and threatens to jeopardi ze
prospects of attai ni ng even modest i mprovements i n
standards of l i vi ng. The forei gn debt of devel opi ng
countri es i ncreased rapi dl y i n the 1970s and 1980s
and i n earl y 1989 was esti mated at about $1.3
tri l l i on.
5
As a resul t of thi s i ncrease and the ri se i n
i nterest rates, debt servi ce as a share of total exports
of goods and servi ces i s now doubl e what i t was i n
the earl y 1970s.
6
Changes in Energy Consumption With
Economic Development
The economi c expansi on necessary to achi eve
hi gher standards of l i vi ng for the i ncreasi ng popul a-
ti on of the devel opi ng worl d woul d be expected to
d~e World Energy Coderence modemte growth rate projection is based on annual economic growth rates of 4.4 perC@. Average ~onomic
growthrates of 5.3 percent annually are assumed in the series of projections in Alan S. Marine andb Schrattenholzer, InternationalE nergyWorkshop:
Overview of Poll Responses, Stanford University International Energy Projec6 California, July 1989. The Intergovernmental Panel on Climate Change,
Appendix Report of the Expert Group on Emissions Scenarios (Response Strategies Worl@ Group Steering Committee, Task A), April 1990,
assumes high economic growth rates for the different developing regions of 4 to 5 percent annually, and 2.2 to 3.0 percent annually for the low-growth
case.
SIJfited Nations Development Programm
e, Humun Development Report 1990 (New Yorlq NY: Oxford University Press, 1990), p. 79.
%ese principal and interest repayments are now much higher than new disbursements of long-term debt to developing countries. The net transfer
or outflow of resources from the developing countries amounted to $38 billion in 1987, compared with a net inflow of $35 billion in 1981. See World
Banlq WorZdDeveZopment Report 1989 (Washington DC: Oxford University Press, 1989), p. 18.
Box 2-AFactors Affecting Population Increase
Rates of popul ati on growth are determi ned by the bal ance between bi rth and death rates. Hi stori cal l y, death
rates were the fi rst to decl i ne, due to i mprovements i n nutri ti on and sani tati on, and medi cal advances such as
vacci nes. I f death rates decl i ne, but bi rth rates remai n the same, popul ati on i ncreases. Thi s i s what happened i n the
presentl y devel oped worl d from 1750 to about 1900. Around that ti me, however, bi rth rates started to fal l , resul ti ng
i n an overal l reducti on i n the rate of popul ati on i ncrease. Thi s process, the l agged adjustment of bi rth rates to the
pri or decl i ne i n death rates, i s known as the demographi c transi ti on. I n the devel opi ng worl d, the demographi c
transi ti on i s far from compl ete: death rates have fal l en dramati cal l y--though they are sti l l hi gher than i n the
i ndustri al countri es-but bi rth rates remai n hi gh, l eadi ng to a conti nued rapi d rate of i ncrease i n total popul ati on.
l
Future trends i n popul ati on wi l l si mi l arl y depend on the bal ance between death and bi rth rates. For the
devel opi ng countri es, opportuni ti es sti l l exi st to reduce death rates through i mprovements i n medi ci ne and publ i c
heal th, and further decl i nes are l i kel y and desi rabl e. On the si de of bi rth rates, there i s much greater uncertai nty over
future trends. Bi rth rates are fal l i ng i n the devel opi ng countri es, from 41 crude bi rths per thousand popul ati on i n
the mi d-1960s to 30 per thousand at present. Bi rth rates i n the devel opi ng countri es, however, are sti l l wel l above
death rates, and more than twi ce the bi rth rates i n the i ndustri al countri es (currentl y 13 per thousand).
The number of bi rths depends on three factors: ferti l i ty rates, the age structure of the popul ati on, and the si ze
of the popul ati on base.
The ferti l i ty rate i s defi ned as the number of chi l dren that woul d be born to a woman i f she were to l i ve to
the end of her chi l dbeari ng years and bear chi l dren at each age i n accordance wi th prevai l i ng age-speci fi c ferti l i ty
rates. Ferti l i ty rates i n devel opi ng countri es have fal l en steadi l y, and i n some cases sharpl y, i n recent years. For
the devel opi ng countri es as a whol e, they fel l from 6.1 i n 1%5 to 4 i n 1987, wi th parti cul arl y sharp decl i nes i n Chi na
and I ndi a, Sri Lanka, Korea, and several Lati n Ameri can countri es. Despi te thi s drop, they are sti l l much hi gher than
i n the i ndustri al countri es. There are, however, excepti ons to thi s decl i ni ng trend; ferti l i ty rates have not changed
i n sub-Saharan Afri can countri es, and i n some of these countri es the rates appear to have ri sen.
Ferti l i ty rates are projected to conti nue decl i ni ng unti l the end of the century, when they woul d be 3.3 compared
wi th the current 4. Whi l e thi s assumpti on seems reasonabl e i n the l i ght of hi stori cal trends, the determi nants of
fami l y si ze are not cl earl y known, and there i s i nevi tabl y some degree of uncertai nty over such assumpti ons. Broadl y
speaki ng, ferti l i ty rates decl i ne as l evel s of economi c and soci al devel opment and urbani zati on ri se, womens
educati on i mproves, and knowl edge about fami l y pl anni ng spreads. The connecti ons between these factors are not
wel l -establ i shed, however, as they are hi ghl y correl ated, and i t i s therefore di ffi cul t to di sentangl e the i mpact of any
s i n gl e determi nant. A hi gher share of the popul ati on l i vi ng i n urban areas, other thi ngs equal , maybe of parti cul ar
i mportance i n l oweri ng ferti l i ty rates. I n rural farmi ng communi ti es, many benefi ts accrue to a l arge fami l y. Chi l dren
provi de farm I aborfrom an earl y age chi l dren are abl e to perform si mpl e farm chores. Chi l dren can al so provi de,
i n the absence of soci al i nsurance, some guarantee of ol d age securi ty for parents. These benefi ts of a l arge fami l y
are not so evi dent i n the urban context, where they may al so be outwei ghed by the fi nanci al costs of supporti ng a
l arge fami l y.
Other factors al so i nfl uence ferti l i ty rates. Cul tural and rel i gi ous factors can l ead to hi gher fami l y si ze than
woul d otherwi se be predi cted by i ndi cators of soci al and economi c devel opment and urbani zati on. Al geri a, Li bya,
I ran, and I raq, for exampl e, have ferti l i ty rates near or over 6 despi te thei r rel ati vel y hi gh per-capi ta i ncomes. On
the other hand, aggressi ve government pol i ci es to restri ct fami l i es can l ead to l ower fami l y si zes than predi cted by
other soci al and economi c i ndi cators. For exampl e, ferti l i ty rates i n Chi na, a l ow-i ncome country, fel l dramati cal l y
from 6 i n the mi d-1960s to 2.4 i n 1987l ower than the rate for the i ndustri al countri es 20 years ago-due l argel y
to strong government pol i cy.
Wi th a gi ven ferti l i ty rate, the number of bi rths wi l l be hi gher i f a l arger share of the popul ati on i s i n the
reproducti ve age group. I n the devel opi ng countri es, young peopl e compri se a hi gher share of the popul ati on. Thi s
popul ati on structure gi ves a much greater popul ati on momentum (the tendency for popul ati on growth to
conti nue even after ferti l i ty rates have fal l en to the repl acement l evel ). I n the devel opi ng countri es, the ri si ng share
of women of chi l dbeari ng age i n the popul ati on wi l l conti nue to exert strong upward pressure on the popul ati on,
despi te the expected drop i n ferti l i ty rates.
The si ze of the popul ati on base i s the thi rd determi nant of popul ati on growth. The hi ghest ferti l i ty rates exi st
i n the devel opi ng countri es, whi ch al ready have by far the l argest share (77 percent) of the gl obal popul ati on. Thi s
means that l arger numbers are bei ng added to the worl d popul ati on than woul d be the case i f the hi gh ferti l i ty rates
appl i ed to the i ndustri al countri es, whi ch compri se onl y 15 percent of the gl obal popul ati on.
Iworld B- Wor~DmelOp~ntRepOrt 1989 (New York NY: oxforduniv~i~Press, 1989), pp. 164-165.
l ead to comparabl e i ncreases i n energy consump-
ti on. I n addi ti on, several factors i nherent i n the
devel opment process tend to cause commerci al
energy use i n devel opi ng countri es to ri se more
rapi dl y than the gross nati onal product.
Fi rst, most of the peopl e i n devel opi ng countri es
now rel y pri mari l y on tradi ti onal bi omass fuel s
wood, crop wastes, ani mal dung-for thei r energy
needs. These fuel s are often di ffi cul t and ti me-
consumi ng to gather, are i neffi ci ent and awkward to
use, and can cause si gni fi cant envi ronmental dam-
age. Si mi l arl y, most peopl e i n devel opi ng countri es
rel y pri mari l y on human and ani mal muscl e power
for doi ng thei r work, despi te i ts l ow effi ci ency and
l i mi ted output. Peopl e who are dependent on these
tradi ti onal forms of energy wi l l turn to commerci al
fuel s and technol ogi es i f and when they are avai l abl e
and affordabl e. Thus, the share of tradi ti onal fuel s i n
total energy consumpti on fal l s sharpl y as devel op-
ment proceeds. I n the l ow-i ncome Afri can countri es,
tradi ti onal fuel s account for as much as 90 percent of
total energy use. I n the mi ddl e-i ncome devel opi ng
countri es thei r share fal l s typi cal l y to under 20
percent.
Second, most devel opi ng countri es are now bui l d-
i ng thei r commerci al , i ndustri al , and transportati on
i nfrastructures. Thi s requi res l arge quanti ti es of
energy-i ntensi ve materi al s such as steel and cement.
As a resul t, energy use i n the near- to mi d-term
i ncreases faster than i ncome.
Thi rd, devel opi ng countri es are experi enci ng
rapi d urbani zati on. Urbani zati on has profound ef-
fects on the amount and type of energy consumed.
7
As i ndustry and the l abor force become more con-
centrated i n urban areas, transportati on needs grow.
Food and raw materi al s are haul ed l onger di stances,
and fi ni shed products are marketed over a wi der
area. Urban househol ds purchase a l arger share of
thei r total needs from outsi de the fami l y, compared
wi th rural househol ds, and commerci al provi ders of
goods and servi ces are more l i kel y to use modern
fuel s. Scarci ty of space i n ci ti es
-
encourages the
substi tuti on of modern, compact energy forms for
the bul ki er bi omass fuel s. Fi nal l y, the growi ng food
needs of the ci ti es encour age changes i n agr i cul tur al
technol ogy, whi ch usual l y i nvol ve i ncreased use of
modern fuel s and energy-i ntensi ve ferti l i zers.
Fourth, modern manufacturi ng technol ogi es and
materi al s have si gni fi cantl y l owered the real cost of
goods--from radi os to refri gerators- many consumer
compared wi th costs a generati on ago, and gl obal
di stri buti on systems have i ncreased thei r accessi bi l -
i ty. Peopl e i n devel opi ng countri es can thus pur-
chase many consumer goods at a far earl i er poi nt i n
the devel opment cycl e than di d peopl e i n todays
i ndustri al countri es. Thi s coul d i ncrease energy use
i n the near- to mi d-term both to produce the
materi al s for consumer goods andparti cul arl y for
those that are i ntensi ve energy users, such as
motorcycl es, cars, ai r condi ti oners, and refri gerators
to operate them.
On the other hand, there are factors that may
counterbal ance these trends and si gni fi cantl y mod-
erate the rapi d i ncrease i n energy demand.
Fi rst, the hi gh cost of devel opi ng nati onal energy
i nfrastructures and of purchasi ng energy to support
growi ng energy demands coul d potenti al l y sharpl y
l i mi t economi c growth. Thi s possi bi l i ty i s hi ghl y
undesi rabl e gi ven current l ow, and i n many cases
decl i ni ng, l i vi ng standards i n devel opi ng countri es.
Second, the expected growth i n energy use i n
devel opi ng countri es coul d be reduced through
effi ci ency i mprovements. Energy i s now used much
l ess effi ci entl y i n devel opi ng countri es than i n
i ndustri al countri es. Tradi ti onal fuel s and technol o-
gi es are often much l ess effi ci ent than modern ones:
for exampl e, the effi ci ency of a typi cal wood-fuel ed
cooki ng stove i s just one-fourth that of a modern gas
range. Moreover, the effi ci ency of energy use i n the
?Donald W. Jones, Urbanization and Energy Use in Energy Development Energy JournuZ, vol. 10, No. 4, October 1989.
modern sector i n devel opi ng countri es i s often far
l ower than that commonl y achi eved i n the i ndustri al
countri es. I f devel opi ng countri es adopt the most
effi ci ent technol ogi es now avai l abl e, they mi ght
achi eve average energy effi ci enci es that are hi gher
than those i n i ndustri al countri es that have a l arge
i nstal l ed base of ol der and l ess effi ci ent i nfrastruc-
ture and equi pment.
Thi rd, conti nued economi c devel opment i s at
some poi nt accompani ed by structural changes that
shi ft i nvestment from energy-i ntensi ve i nfrastruc-
ture (roads, bui l di ngs, etc.) to consumer goods
(refri gerators, cars, etc.) and fi nal l y to l ess materi al -
i ntensi ve but hi gher val ue-added goods such as
personal servi ces and el ectroni cs.
Energy use i n devel opi ng countri es wi l l depend
on the net i mpact of these opposi ng factors. At l ow
l evel s of devel opment the fi rst set of factors predom-
i nates, and commerci al energy consumpti on typi -
cal l y ri ses much faster than gross domesti c product
(GDP). Fi gure 2-5 compares per-capi ta commerci al
energy consumpti on wi th per-capi ta GDP for se-
l ected countri es, rangi ng from l owest to hi ghest
i ncome.
8
As thi s fi gure suggests, wi thi n the poorest
countri es commerci al energy consumpti on ri ses
faster than per-capi ta GNP; i n the mi ddl e-i ncome
countri es they ri se at about the same rate; and at the
hi ghest l evel s of i ncome, the i ncrease i n total
commerci al energy consumpti on i s l ess than the
i ncrease i n per-capi ta GNP.
I t i s often argued
9
that GNP per capi ta, a measure
of the val ue of economi c output i n rel ati on to
popul ati on si ze, i s an i nadequate and mi sl eadi ng
i ndi cator of standards of l i vi ng and wel l -bei ng.
10
An
i ndex recentl y devel oped by the Uni ted Nati ons
Devel opment Programme, the Human Devel opment
I ndex (HDI ),
l l
i ncorporates both economi c and
soci al factors. Thi s i ndex i s based on three i ndi ca-
tors: l i fe expectancy at bi rth, adul t l i teracy, and
per-capi ta purchasi ng power.
12
The fi rst two are
sensi ti ve to soci al condi ti ons i n a country and i n
addi ti on refl ect underl yi ng condi ti ons of i ncome
di stri buti on. Average purchasi ng power i n a country
gi ves some i ndi cati on of materi al standards of
l i vi ng. At l ow val ues, the HDI al so shows a cl ose
posi ti ve associ ati on wi th commerci al energy con-
sumpti on (see fi gure 2-6). Hi gher l evel s of HDI ,
however, can be achi eved wi th a wi de range of
commerci al energy consumpti on.
Commerci al energy i s onl y part (and for the
poorest countri es, a very smal l part) of total energy
consumpti on. I f esti mates of tradi ti onal fuel con-
sumpti on are i ncl uded wi th commerci al fuel to
represent total energy consumpti on, the associ ati on
between per-capi ta energy consumpti on and GNP
remai ns cl ose, but at l ower i ncome l evel s the sl ope
i s l ess steep than i n fi gure 2-5.
13
Addi ng commerci al
and tradi ti onal energy together to make total energy
consumpti on does not, however, take i nto account
the l ower conversi on effi ci enci es of tradi ti onal
energy compared wi th commerci al energy. I f tradi -
ti onal energy consumpti on were expressed i n com-
s~e re~tiom~p betw~n GDP pm Capi ti and comrnerci~ energy consumption shown in figure 2-5 is consistent with other std.ieS. A Per-cilpitti
income elasticity for fossil fuels of about 1.5 is given for 13 industrial and developing countries in Gerald Leach et al., Energy and Growth (London:
Butterworths, 1986),p. 25. That is, agivenincreasein per-capita GNP between countries (purchasingpowe rparity) is associated witha 50 percent higher
increase in consumption of fossil fuels. Another study, based on 100 countries, reports a per-capita income elasticity of 1.26 (i.e., a given increase in
per-capita GNP between countries is associated with a 26 percent higher increase in commercial energy consumption). See B.W. Ang, A
Cross-Sectional Analysis of Energy Output Correlation Energy Economics (lmndon: Butterworths, 1987), table 3, p. 280. This elasticity is based on
market exchange rates; comparison of purchasing power parity GNP data increases the elasticity to 1.8. These elasticities are based on cross-sectional
studies (i.e., intercountry comparison at a given point in time, rather than developments in an individual country over time), which are considered to
give a more accurate picture of the long-term relationship between energy consumption and economic growth.
*or example, see Carlos Andrea Perez, Twardk a New Way to hfeasure Development (Caracas, Venezuela: Oftlce of the South Commission in
Venezuel& 1989).
IOG~ my be ~ ~d~te rn~me of so~i~ Wefl-being, but it nonetheless corre~tes strongly with -y SOCid indicators. See P- Dxgup@
Wellbeing and the Extent of Its ReaUzation in Poor Countries, The Economc Journal, supplement (Cambridge, MA: Royal Economic Society Basil
BlackweU, 1990), pp. 1-32. The argument is made here that GNP per capita also correlates closely with political and civil rights.
llunited Nations Development Programme, Human Development Report 1990, op. Cit., fOO~Ote 6, pp. 11-16.
12~ev~um of achimement forthew indimtorswemJ~~s ~eexpec~ncyatbirtho f78, ~ad~tliteraqrateofloopercen$ andtheaverageofflcial
poverty Une income in nine industrial countries, adjusted to take into account purchasing power, of $4,861 per capita
~s~ch op. cit., foo~ote 9, p. 25, gives ~r-=pi~ ~mme elmticities for to~ enqy at about 1.2, lower ~ the 1.5 for fossil fuels alone. B.W.
Ang, A Cross Sectional Analysis of Energy Output Correlation op. cit., footnote 9, reports income elasticities of traditional energy consumption of
-0.95 (i.e., a 10 percent increase in GNP per capita is associated with a 9.5 percent decline in traditional energy consumption) and for total energy
consumption (commercial and traditional) of 1.05.
Chapter 2--Energy and Economic Development .31
Figure 2-5-Commercial Energy Consumption and Economic Development
in Selected Countries
Gi g a j o u l e s p e r c a p i t a
1000 ~
s 24
s 22
s 23
1 0 0 ~
m19 s 21
17
=
s 18
s 20
s 13
s 16
5 6 11
. l ; l ~
5
1 Ethiopia 13 Algeria
2 Nigeria 14 Peru
10 ~ .
4
7
%: 1 0 3 Kenya 15 Costa Rica
4 India
s 2
16 Brazil
5 Zimbabwe 17 Korea
s 3 6 Egypt 18 Argentina
7 Cameroon 19 Venezuela
8 Indonesia 20 Mexico
1 :
9 Morocco 21 Japan
m 1 10 Phi l i ppi nes 22 Fr ance
11 China 23 W. Ger many
12 Thai l and 24 Uni t ed St at es
0 . 1 -
I I I I I I
o 2 4 6 8 10 12 14
Gross domesti c product per capi ta (thousand dol l ars) (ppp)
SOURCES: United Nations, 1986 Energy Statistics Yearbook(New YOIIG NY: United Nations, 1988), table 4; Robert
Summers et al., A New Set of International Comparisons of Real Product and Prioe lsvels, Estimates for
130 Countries, 1950-1985, F/eviewofhworne and Wea/th, Series 34, No. 1, March 1988.
merci al fuel equi val ent,
14
the i ncrease i n energy
consumpti on i n rel ati on to GNP mi ght be somewhat
greater than the i ncrease i n total energy (an energy
el asti ci ty of just over 1.0), but l ess than the i ncrease
i n commerci al energy al one (1.5 or more).
The i ncl usi on of tradi ti onal energy, though neces-
sary to provi de a more compl ete pi cture of the
rel ati onshi p between economi c growth, soci al de-
vel opment, and energy use, rai ses probl ems of i ts
own. Fi rst, the measurement of tradi ti onal fuel s i s
di ffi cul t and prone to underesti mati on. For exampl e,
for I ndonesi a, Mal aysi a, the Phi l i ppi nes, and Thai -
l and, esti mates by the Uni ted Nati ons Food and
Agri cul tural Organi zati on, the Asi an Devel opment
Bank, and the Worl d Energy Conference-three
frequentl y quoted sources of data on tradi ti onal
fuel s-are found i n al most al l cases to be consi der-
abl y l ower than i n other cou.ntry-speci fi c studi es.
15
Second, when consi deri ng tradi ti onal sources of
energy, i t i s di ffi cul t to know whereto draw the l i ne.
Ani mate forms of energy are i mportant i n most
devel opi ng countri es, parti cul arl y the poorest. I f the
bi omass fed to bul l ocks to provi de pl owi ng and
i rri gati on servi ces were i ncl uded, the amount of
tradi ti onal energy consumed woul d i ncrease sub-
stanti al l y. Further, i f the l arge amounts of bi omass
burned i n prepari ng soi l for cul ti vati on i n sl ash and
burn agri cul ture were i ncl uded, per-capi ta energy
use where shi fti ng agri cul ture i s practi ced coul d
concei vabl y be as hi gh or hi gher than i n the
i ndustri al countri es.
l 6
Al though GNP growth i s an i mportant determi -
nant of energy consumpti on, i t i s not the onl y one.
Tabl e 2-1 shows commerci al and total energy
consumpti on rel ati ve to GNP for di fferent l evel s of
i ncome. I f the amount of energy consumed were ti ed
to the l evel of output or devel opment, the energy
i ntensi ti es of al l countri es woul d be the same. As
these fi gures show, however, there are consi derabl e
vari ati ons i n energy i ntensi ty (the rati o of energy
consumpti on to GNP) at al l l evel s of devel opment.
Among the i ndustri al countri es, energy i ntensi ti es
vary wi del y. And among the devel opi ng countri es,
energy i ntensi ti es of countri es wi th the same l evel of
14BOW. ~g,
{
A Method forEs@@ Non<o~erc~Ener~ con~pti on i n ~er-ro~i l ol d s~tor Of Developing COWtriSS, Energy (l.ondoa
Pergamon Press, 1986), p. 423, table 8.
15&, op. cit., footnote 15, P. 423, ~ble 8.
16s= ~R. sfi~me Biofiel T~itio~pa~fic andA~an JOWWZ1 OfEnergy, VO1. 1, No. 1, JSII~ 1987, p. 18, fi~e 5.
Figure 2-6-Commercial Energy Consumption and Social Development
Gi ga j oul e s pe r c a pi t a
o
r
300
250
1
200
-
150
-
100
-
50
-
1 Ethiopia 13 Algeria
. 24
2 Nigeria 14 Peru
3 Kenya 15 Costa Rica
4 India 16 BraziI
5 Zimbabwe 17 Korea
6 Egypt 18 Argentina
7 Cameroon 19 Venezuela
8 Indonesia 20 Mexi co
9 Morocco 21 Japan
10 Phi l i ppi nes 22 Fr ance
s 2 3
11 China 23 W. Germany m 22
12 Thai l and 24 United States
s 21
s 1 9
18
20 s
r
,7
s 13
, 16
7 s r3 s 5 11 s . 1 4 s 1 5
I
1
4. ~m~ g .8
-
s 12
0
, 2 10 m
I
o 0.2 0.4 0.6 0.8 1 1. 2
Human development index
SOURCES: United Nations, 1986 Energy Statistics Yearbook(New York, NY: United Nations, 1988), table 4; United
Nations Development Programme, Human Development Report 1990 (New Yo~ NY: Oxford University
Press, 1990), p. 79.
Table 2-lEnergy Intensities in Selected Countries
Gross domestic Total energy Commercial energy
product, 1985 consumption per GDP consumption per GDP
Country $PPP
a
GJ/$1,000 Index, U.S.= 100 GJ/$1,000 Index, U.S.=1OO
Ethiopia. . . . . . . . . . . . . . . 304 29.6 123 2.3 10
Nigeria . . . . . . . . . . . . . . . . 565 24.8 103 9.0 38
Kenya . . . . . . . . . . . . . . . . 603 29.9 124 5.0 21
India . . . . . . . . . . . . . . . . . 775 15.5 64 11.4 48
Zimbabwe . . . . . . . . . . . . . 954 30.4 126 22.3 94
Egypt . . . . . . . . . . . . . . . . . 1,080 21.3 88 20.6 87
Cameroon . . . . . . . . . . . . . 1,180 16.9 70 9.1 38
Indonesia . . . . . . . . . . . . . 1,269 12.6 52 6.6 28
Morocco . . . . . . . . . . . . . . 1,284 7.8 32 7.3 31
Philippines . . . . . . . . . . . . . 1,352 11.1 46 6.7 28
China . . . . . . . . . . . . . . . . . 1,489 15.4 64 14.3 60
Thailand . . . . . . . . . . . . . . 1,896 12.7 53 7.0 29
Algeria . . . . . . . . . . . . . . . . 2,133 17.8 74 17.4 73
Peru . . . . . . . . . . . . . . . . . . 2,333 9.9 41 8.1 34
Costa Rica . . . . . . . . . . . . 2,712 16.2 67 8.3 35
Brazil . . . . . . . . . . . . . . . . . 3,164 16.4 68 10.7 45
Korea . . . . . . . . . . . . . . . . 3,381 16.6 69 16.1 68
Argentina . . . . . . . . . . . . . . 3,640 17.0 71 15.9 67
Venezuela . . . . . . . . . . . . . 3,723 29.8 124 29.9 126
Mexico . . . . . . . . . . . . . . . . 3,987 13.3 55 12.5 53
Japan . . . . . . . . . . . . . . . . 9,739 12.7 53 12.7 54
France . . . . . . . . . . . . . . . . 10,032 16.1 67 16.0 67
Germany . . . . . . . . . . . . . . 10,959 17.0 70 16.9 71
United States . . . . . . . . . . 12,787 24.1 100 23.7 100
a ppp refers to purchasing power Parity.
SOURCES: United Nations, 1986 Energy StafMcs Yearbook (New York, NY: United Nations, 1988), table 4forenergy consumption data. Robert Summers
et al., A New Set of International Comparisons of Real Product and Price Levels, Estimates for 130 Countries, 1950-1985, Review of home
and Wea/fh, Series 34, No. 1, March 1988 for gross domestic product, purchasing power parity.
devel opment (measuredl y GNP per capi ta) can vary
more than fourfol d.
17
Vari ati ons i n energy i ntensi ty are al so evi dent i n
i ndi vi dual countri es or groups of countri es over
ti me, as shown i n fi gure 2-7. I n the years i mmedi -
atel y before 1973, gl obal energy i ntensi ti es rose.
After 1973, there was a sharp drop i n the i ndustri al
countri es, i n contrast to a conti nued ri se i n the
devel opi ng nati ons, i f at a l ower pace. Experi ence
wi thi n the devel opi ng countri es vari ed consi derabl y,
agai n testi fyi ng to the i mportance of factors other
than economi c growth i n determi ni ng energy con-
sumpti on.
Di fferences i n rates of urbani zati on and i ndustri al
structure account for part of the vari ati on i n energy
i ntensi ti es. Countri es wi th a l arge share of energy-
i ntensi ve i ndustri es, such as steel , paper, chemi cal s,
and al umi num, wi l l tend to have hi gher energy
i ntensi ti es than countri es wi th few energy-i ntensi ve
i ndustri es. Ni geri a, Egypt, Al geri a, and Venezuel a
(wi th l arge oi l dri l l i ng and refi nery operati ons) and
Argenti na, Korea, Zi mbabwe, and Chi na (wi th l arge
metal s i ndustri es, mai nl y i ron and steel ) al l have
rel ati vel y hi gh energy i ntensi ti es.
The i mpact of changes i n i ndustri al mi x i s al so
evi dent over ti me. About hal f of the post-1979
decl i ne i n energy i ntensi ty i n Chi na (whi ch fel l by 40
percent between the l ate 1970s and the l ate 1980s)
can be ascri bed to the l i mi ts on the expansi on of
heavy i ndustri es and to the promoti on of l i ght, and
often export-ori ented, manufactures (e.g., texti l es,
consumer el ectroni cs, processed foodstuffs, and
pl asti cs) .18 The equal l y dramati c decl i ne i n Koreas
energy i ntensi ty i s al so due i n l arge part to the
changi ng i ndustri al mi x. Al though the output of
heavy i ndustri es i n Korea rose sharpl y i n these years,
producti on of l ess energy-i ntensi ve i ndustri es such
as machi nery and transport equi pment grew even
more rapi dl y, resul ti ng i n a decl i ni ng share of heavy
i ndustry i n the total .
Energy i ntensi ti es are al so i nfl uenced by the
technol ogi es used throughout the economy. For
exampl e, the ol der generati on of coal -burni ng tech-
nol ogi es sti l l used i n devel opi ng countri es are much
Figure 2-7Energy Intensity and Economic
Development, 1960-2000
Barrels of oil equivalent per $1,000 of goods and services
8
I I
7. 5-
7 -
8. 5-
>6
8 -
5.5 -
5 ~
1
4.5
I
1 1 1 I
1960 1973 1982 1990 2000
Industrial nations + World average
+
Developing nations
SOURCE: World Bank, Industry and Energy Department, Energy Issues
in the Developing World, Energy Series Paper No. 1, February
19s8.
l ess effi ci ent than modern technol ogi es. Conse-
quentl y, the coal -burni ng countri es (I ndi a and Chi na)
are more energy-i ntensi ve than countri es l argel y
rel i ant on oi l and gas.
Some of the factors that determi ne energy con-
sumpti on (e.g., the si ze of the country and the
l ocati on of natural resources i n rel ati on to i ndustry
and major markets) are country-speci fi c, but others
can be affected by pol i cy deci si ons.
Projected Energy Consumption in
I n the past, the three factors di scussed i n the
precedi ng secti onrapi dl y ri si ng popul ati on, hi gh
economi c growth rates to provi de i mproved stand-
ards of l i vi ng, and structural change as devel opment
gets underway-have been associ ated wi th rapi d
r ates of i ncr ease i n commer ci al ener gy consumpti on
i n devel opi ng countri es. I f these trends conti nue,
i ncreases i n commerci al energy consumpti on i n
devel opi ng countri es coul d be very l arge. Tabl e 2-2
i l l ustrates some of the current projecti ons. The
synthesi s of a wi de range of projecti ons shown i n
thi s tabl e suggests that commerci al energy con-
sumpti on i n the devel opi ng worl d (i ncl udi ng here
OPEC) i n 2010 coul d be 2.5 ti mes hi gher than i t was
i n the base year , 1985, an annual r ate of i ncr ease of
17FrOrn~e~ted sampl e shown i ntabl e 2.1, there does not appear to be a systematic tendency for energy/GNP ratios to rise as levek of development
rise; the energy/GNP ratios of the advanced developing countries are very similar to those of Europe and Japan. However, Ang, op. cit., footnote 9, using
a wider sample, shows a commercial energy/GNP ratio with respect to GNP per capita at 0.80. That is, a 10 percent increase in GNP is associated with
sn 8 percent increase in the energy/GNP ratio.
lsvac~v Srnil, Ctis Energy: A Case Study,
contractor report prepared for the Office of Technology Assessmen4 April 1990.
3.8 percent. Chi na accounts for more than one-thi rd
of the projected i ncrease.
The Worl d Energy Conference forecasts a some-
what sl ower rate of growth i n commerci al energy
consumpti on i n the devel opi ng worl d, an annual
average i ncrease of 3.3 percent. By 2020, however,
consumpti on of commerci al energy i n the devel op-
i ng worl d woul d be three ti mes hi gher than i n 1985,
and consumpti on of tradi ti onal fuel s about 25
percent hi gher (see fi gure 2-8). Popul ati on growth
and ri si ng standards of l i vi ng each account for about
hal f of the total i ncrease.
19
The Energy Sector and the
Macroeconomy
Energy i s wi del y recogni zed as a key economi c
sector i n devel opi ng countri es. Rel i abl e and afford-
abl e suppl i es of energy make major contri buti ons to
economi c and soci al devel opment; conversel y, i n-
adequate or unrel i abl e energy suppl i es frustrate the
devel opment process.
Financing Energy Supplies
Most devel opi ng countri es (59 out of 80) are net
energy i mporters, rel yi ng on i mports for vi rtual l y al l
domesti c commerci al energy consumpti on; for many
countri es, oi l i mports represent 30 percent or more
of total export earni ngs.
20
The share of energy
i mports i n the total export earni ngs of devel opi ng
countri es has fl uctuated wi del y i n recent years as oi l
pri ces have changed. Such wi de fl uctuati ons are
hi ghl y di srupti ve to energy pl anni ng and economi c
devel opment.
Many devel opi ng countri es emphasi ze domesti c
producti on of energy. I nvestments i n hi ghl y capi tal -
i ntensi ve energy suppl y systems often represent a
major share of the total i nvestment budget (see tabl e
2-3), accounti ng i n some countri es for over 40
Figure 2-8-Historical and Projected Energy
Consumption in Developing Countries: World
Energy Conference Moderate Projections
Exajoules
3 0 0 ~
~
100
\
50
1
o +
1970 1980 1990 2000 2010 2020
SOURCE: World Energy Conference, Conservation and Studies Commit-
tee, Global Energy Perspective 2000-2020, 14th congress,
Montreal 1989 (Paris: 1989).
percent of al l publ i c i nvestment.
21
As the predomi -
nant cl ai mant on scarce capi tal resources, devel op-
ments i n the energy sector therefore have a major
i mpact on the amounts of i nvestment avai l abl e for
other economi c and soci al devel opment.
The scal e of future i nvestment demands for the
energy sector i n devel opi ng countri es i s projected to
be very l arge. The Worl d Bank, for exampl e,
esti mates that i nvestments of $125 bi l l i on annual l y
(twi ce the current l evel ) woul d be needed i n
devel opi ng countri es to provi de adequate suppl i es of
el ectri ci ty.
22
Accordi ng to a Worl d Bank esti mate
f
annual aver age expendi tur es on commer ci al ener gy
suppl y faci l i ti es for devel opi ng countri es, el ectri ci ty
accounts probabl y for one-hal f of the total ; oi l ,
i ncl udi ng refi neri es, accounts for about 40 percent;
and natural gas and coal 5 percent each. These
expendi tures do not i ncl ude i nvestment i n smal l -
scal e renewabl e or energy conservati on.
l~e ~t=govemenM panel on Climate change has undertaken projections of energy consumption in developing counties in Append.ixReport
of the Expert Group on Emissions Scenarios (Response Strategies Working Group Steering Committee, Task A), April 1990. Several scenarios are
provided, with different rates of economic grow@ emission coeftlcients, and policies. The high economic growth (growth rates similar to the other
studies quoted here), low emissions scenario forecasts a threefold increase in developing world energy consumption between 1985 and 2025, which is
reasombly similar to the World Energy Conference forecast. The high growt.lL high emissions scenario is similar to the combined forecast results in Alan
S. Marine and h Schrattenholzer, International Energy Workshop: Overview of Poll Responses (Palo Alto, CA: Stanford University, Intern@onal
Energy Project, July 1989).
%Energy imports and debt service together account for over one-third of total export earnings in middle-income countries, and almost 40 percent in
lower-income countries (excluding china and Lndia).
zl~ese &@ cover investmen~ in tie energy sector by public entities. They may include some small amounts of investment in energy COmematiOQ
but virtually all is in energy supplies. They do not include private investment in backup equipment.
~World B* CapiM fiWndi~e for Elecrric Power in the Developing Countries in the 1990 s, World Bank Industry andlMergy Depment
Working Paper, Energy Series Paper No. 21, Washington DC, Februmy 1990.
Table 2-2--Commerciai Energy Consumption, 1985 and Projections for 2010 (EJ=Exajoules)
Increase in
Share of Share of AARG,
a
consumption, Share of
1985 total 2010 total 1985-2010 1985-2010 increase
(EJ) percent (EJ) percent percent (EJ) percent
Developing countries . . . . . . . 69.26 23.3 175.!56 34.5 3.8 106.30 50.2
China . . . . . . . . . . . . . . . . . 21.91 7.4 59.54 11.7 4.1 37.63 17.8
OPEC . . . . . . . . . . . . . . . . . 11.65 3.9 28.91 5.7 2.5 17.26 8.1
Non-OPEC developing
countries . . . . . . . . . . . . 35.70 12.0 87.11 17.1 3.6 51.41 24.3
OECD . . . . . . . . . . . . . . . . . . . 155.83 52.4 215.03 42.2 1.3 59.20 27.9
United States . . . . . . . . . . . . . 73.87 24.8 98.47 19.3 1.2 24.60 11.6
U.S.S.R. and
Eastern Europe . . . . . . . . . 72.32 24.3 118.66 23.3 2.0 46.34 21.9
Total . . . . . . . . . . . . . . . . . . 297.41 100.0 509.25 100.0 2.3 211.85 100.0
a~RG = annual average rate of growth.
SOURCE: Alan S. Marine and Leo Sehrattenholzer, /nternationa/ Energy Workshop: Overview of Po// Responses (Palo Alto, CA: Stanford University,
International Energy Pro@t, July 1989).
Table 2-3-Estimated Annual Energy investment as a Percentage of
Annual Total Public investment During the Early 1980s
Over 40 percent 30-40 percent 20-30 percent 10-20 percent 0-10 percent
Argentina Ecuador Botswana Benin Ethiopia
Brazil India China Egypt
Colombia Pakistan Costa Rica Ghana
Korea Philippines Liberia Jamaica
Mexico Turkey Nepal Morocco
Nigeria
Sudan
SOURCE: Mohan Munasinghe, E/eciric Power Eeortomics (London: Butterwotihs, 1990), p. 5.
About one-hal f of total esti mated energy suppl y
i nvestments are projected to be i n forei gn ex-
change.
23
The forei gn exchange component for oi l
and gas i s typi cal l y hi gh (about two-thi rds of the
total ), as much of the equi pment must be i mported.
On the other hand, forei gn exchange costs for coal
devel opment are l ow (about one-quarter of the total ),
mai nl y because the major coal -usi ng countri es, I ndi a
and Chi na, manufacture coal i ndustry equi pment
domesti cal l y. The projected share of forei gn ex-
change i n el ectri c power vari es wi del y accordi ng to
country. I n countri es wi th devel oped i ndustri al
sectors, the share may be between 5 and 10 percent,
but i n countri es that i mport al l thei r generati ng
equi pment, the share ri ses as hi gh as 70 Percent.
24
Achi evi ng these hi gh l evel s of forei gn resources
for the energy sector i nvestment poses i mmense
chal l enges. Though most countri es are l i kel y to
experi ence di ffi cul ti es, the i ssues wi l l vary from
country to country. The poorest countri es are hi ghl y
dependent on confessi onal ai d (whi ch accounted for
80 percent of thei r total external borrowi ng for the
energy sector i n 1975-80). Thei r success i n acqui r-
i ng funds wi l l depend on the extent of the i ncrease
i n confessi onal fl ows. On the other hand, the
mi ddl e-i ncome countri es depend mai nl y (80 per-
cent) on export-rel ated and pri vate fi nanci al fl ows
for thei r external fi nanci ng of energy i nvestments.
The si tuati on i s parti cul arl y acute i n hi ghl y i ndebted
devel opi ng countri es.
The other hal f of the projected i ncrease i n
i nvestment i n the energy sector comes from domes-
ti c resources, wi th a parti cul arl y hi gh share i n
el ectri ci ty and coal . I n many countri es, however, the
fi nanci al si tuati on of the power sector has deteri o-
rated, as i ncreases i n costs have not been matched by
~~te~titig for tie enq ~tor was projected to come largely (almost threequarters) fium private SOUKXS (supplk dt and privab
commercial loans). Imam horn multilateral agencies and bilateral aid accounted for about onequarter.
~World B@ The Enei-# Transition in Developing Countries @?@@gtOQ w: 1983).
i ncreased revenues.
25
The fi nanci al vi abi l i ty of oi l
refi nery operati ons i n many countri es i s al so com-
promi sed by the structure of petrol eum product
pri ces.
26
The i ssue of domesti c resource mobiliza-
ti on i s of parti cul ar i mportance for coal , where much
of the fi nanci al resources needed are l ocal rather
than forei gn.
There are i ndi cati ons that the devel opi ng coun-
tri es are payi ng i ncreased attenti on to resource
mobi l i zati on for the energy sector. Several countri es
(e.g., Peru, Ecuador, and Col ombi a) are currentl y
openi ng more of thei r terri tori es to oi l expl orati on by
forei gn fi rms. Requi rements for government parti ci -
pati on i n oi l devel opment ventures are bei ng re-
l axed. I mproved fi scal arrangements provi di ng for
the speci al characteri sti cs of gas have been adopted
i n Egypt, Tuni si a, Paki stan, and el sewhere wi th a
dramati c i ncrease i n expl orati on speci fi cal l y di -
rected at gas.
27
Energy Supply Reliability
Just as the presence of rel i abl e suppl i es of
hi gh-qual i ty energy can be a strong i ncenti ve to
economi c devel opment, so unrel i abl e suppl i es can
di scourage devel opment and add substanti al l y to the
cost of usabl e power. El ectri ci ty suppl i es i n many
devel opi ng countri es are characteri zed by frequent
servi ce curtai l ments to customers, i ncl udi ng bl ack-
outs, brownouts, and sharp power surges. Thi s can
have two types of i mpacts:
q I ndustri es and offi ces are unabl e to oper ate,
producti on i s l owered, and raw materi al s are
wasted. I n Chi na, for exampl e, i t i s cl ai med that
el ectri ci ty shortages and di srupti ons duri ng the
1980s were responsi bl e for i dl i ng at l east 20
percent of the countrys i ndustri al capaci ty.
28
For the fi ve publ i c-sector steel pl ants i n I ndi a,
i t has been esti mated that at 1986-87 operati ng
l evel s, i rregul ar and restri cted el ectri ci ty sup-
pl y resul ted i n i ncreased el ectri ci ty consump-
ti on of over 216 gi gawatthours at a cost of $10
q
mi l l i on, and the poor qual i ty of the el ectri ci ty
resul ted i n addi ti onal consumpti on of 412
gi gawatthours at a cost of $18 mi l l i on.
29
Mor e
general l y, l ost i ndustri al output caused by
shortages of el ectri ci ty i n India and Paki stani s
esti mated to have reduced GDP by about 1.5 to
2 percent.
30
Resi denti al consumers are al so
affected.
Many consumers, both resi denti al and i ndus-
trial,
-
are obl i ged to i nvest i n a vari ety of
equi pment-vol tage boosters, standby genera-
tors, storage batteri es, kerosene l amps-i n
order to mi ni mi ze the i mpact of di srupted
suppl i es. Though no data are avai l abl e, expen-
di tures on these devi ces are certai nl y substan-
ti al , addi ng to the cost of provi di ng usabl e
suppl i es.
Such suppl y constrai nts are usual l y associ ated
wi th el ectri ci ty, but there are al so shortages of other
sources of energy. Suppl i es of househol d fuel s i n
many countri es (e.g., I ndi a) are notori ousl y intermit-
tent. Thi s accounts for the exi stence of a wi de range
of cooki ng systems i n many househol ds i n order to
ensure agai nst the shortage of any one fuel . Trans-
portati on servi ces are al so subject to di srupti on
because of unrel i abl e fuel suppl i es.
Energy Pricing and Demand
Management
Energy pri ces pl ay a key rol e i n energy sector
devel opment, through thei r i mpacts on the amount
of energy used i n an economy, the technol ogi es
adopted, and i n some cases, the di recti on of i ndus-
tri al devel opment. The effects on the energy i nfra-
structure are l ong term i n nature, and often di ffi cul t
to reverse.
Energy pri ci ng pol i cy may have several objec-
ti ves: effi ci ent al l ocati on of resources, provi si on of
affordabl e suppl i es to consumers, reasonabl e returns
to energy producers, substi tuti on between fuel s for
fiworld B* Review of World Bank Lending for Electric power,
World Bank Industry and Energy Department Working Paper Energy Series,
Paper No. 2, March 1988.
~Don~d Her&mm~ Energy Efficiency and Energy pricing in Developing ComtrieS,
contracto~ report prepared for the Office of Technology
Assessment, May 1990.
zTfhe~ore J. Gor@q Pe~ole~ in the Developing World, contractor report prepared for the Gi31ce of Twtiology Assessment J~Y 1990.
28Vac~v sd, CC~S ~er=: A Case Study, op. cit., footnote 18.
%hergyandEnvironmentalAnalysis, Conserving Process Heat in Primary Industries of India and C- contractorreportprepared for the Gfi3ce
of Technology Assessmen4 April 1990.
Arm P. San@vi, Impacts of Power Supply Inadequacy in Developing Countries,Journal of Energy Policy (forthcoming).
nati onal securi ty or envi ronmental reasons, promo-
ti on of regi onal devel opment and i ndustri al compet-
i ti veness.
31
The wei ghts of these di fferent objecti ves
i n the formul ati on of energy pol i cy vary among
countri es. The i mportance of the regi onal devel op-
ment objecti ve, for exampl e, vari es from country to
country dependi ng on geographi cal confi gurati on,
pol i ti cs, and hi story. As i n other aspects of economi c
and soci al pol i cy, however , ther e ar e sever al char ac-
teri sti cs of energy pri ci ng that are shared by many of
the devel opi ng countri es:
q
q
q
Governments pl ay a strong rol e i n the commer-
ci al energy sector. I n vi rtual l y al l devel opi ng
countri es the el ectri ci ty sector i s government
owned, and i n many countri es the government
al so owns the coal , and oi l and gas sectors.
Regardl ess of the form of ownershi p, the
government typi cal l y regul ates pri ces of energy
products, frequentl y at several l evel s of the
producti on and di stri buti on chai n.
Soci al objecti ves are an i mportant factor i n
formul ati ng energy pri ci ng pol i ci es. As a basi c
necessi ty of l i fe, energy accounts for a substan-
ti al part of total househol d expendi tures. Gov-
ernments frequentl y ai m to keep the typi cal
cost of househol d fuel s-e. g., kerosene for
l i ghti ng and cooki ng, and i n some cases el ec-
tri ci ty-l ow. The l arge number of poor i n the
popul ati on al so makes pri ce stabi l i ty an i mpor-
tant pol i cy objecti ve. Though soci al equi ty
i ssues are a major preoccupati on i n the pri ci ng
of commerci al fuel s, pri ces of the most com-
mon form of energy used by the poorwood
and charcoal -are usual l y not regul ated.
Economi c objecti ves, notabl y, the desi re to
encourage key strategi c devel opment sectors
i ncl udi ng transportati on and agri cul ture, are
al so refl ected i n pol i ci es desi gned to promote
rural el ectri fi cati on or to keep di esel pri ces l ow.
Pol i ci es that keep key energy product pri ces l ow
can al so produce adverse resul ts. Revenues from
energy sal es may be i nadequate to cover the costs of
suppl yi ng the energy. Thi s probl em i s especi al l y
acute i n the el ectri ci ty generati ng sector i n devel op-
i ng countri es. One study
32
showed that i n 30 out of
37 devel opi ng countri es for whi ch data were avai l -
abl e, el ectri ci ty tari ffs were too l ow to generate the
revenues needed to cover total operati ng costs pl us
al l owances for equi pment repl acement or expansi on
of the system. A survey of el ectri c power projects
fi nanced by the Worl d Bank over a 20-year peri od
33
i ndi cates a consi stent decl i ne i n key fi nanci al
i ndi cators as revenues from sal es of el ectri ci ty
l agged behi nd ri si ng costs.
Petrol eum pri ces are rather di fferent. At present,
subsi di es (defi ned here as pri ces si gni fi cantl y l ower
than those charged i n i nternati onal markets) are
l argel y l i mi ted to oi l -exporti ng (or at l east oi l -
produci ng) countri es. Countri es that i mport al l thei r
suppl i es of petrol eum products, and are therefore
obl i ged to pay current i nternati onal pri ces for thei r
suppl i es, are general l y unwi l l i ng to subsi di ze pri ces
on the domesti c market. I n the oi l -exporti ng coun-
tri es, however, despi te sharp pri ce i ncreases i n
recent years, several petrol eum productsnotabl y,
kerosene and heavy fuel oi l -conti nue to be sol d at
hal f the i nternati onal pri ce or l ess. Low domesti c gas
pri ces, i n combi nati on wi th other factors, di scourage
the devel opment of gas resources and contri bute to
the spectacul arl y hi gh share of fl ared gas i n devel op-
i ng countri es--47 percent of total producti on, com-
pared wi th 4 percent i n the OECD countri es.
%
I n both I ndi a and Chi na, whi ch together account
for 70 percent of al l coal consumpti on i n the
devel opi ng worl d, coal pri ces are kept bel ow pro-
ducti on costs. I n Chi na, two-thi rds of al l coal
enterpri ses l ost money i n 1984.
35
I n the earl y 1980s
the Worl d Bank esti mated Coal I ndi as l osses at
$300 mi l l i on on sal es of $700 mi l l i on.
36
31F0r -m di~m~~ion of the ~co~ and obj~tives of enerw pfic~g, see ~O~n M~~e, Ener~AnalY~s and Policy @ndon: Buttenvortbs,
1990); Lawrence J. Hill, Energy Price Rejorm in Developing Countnes: Issues and Options (Oak Ridge, TN: Oak Ridge National Laboratory, August
1987), and Corazon Siddayao, Criteria for Energy Pricing Policy (London: Oraham and Trotmaq 1985).
sZ~Wence J. Mll, op. cit., footnote 31, pp. 2-10 and 2-20, table 2-3.
Ssworld Ba~ Review of World Bank Lending for Ek%@ic Power,
Industry and Energy Department Working Paper, Energy Series paper No. 2,
March 1988.
~~kKomo, Money tOB~n? Theffigh co~tofEnergy~~~sidies @Twhin@o~ DC: world Resources ~tit@e, 1987), p. 14. Based onInternational
Energy Agency data.
ss~wence J. Him op. cit., footnote 31.
~N&k Kosmo, op. cit., footnote 34, p. 16.
I n some cases, the costs of energy suppl i es are
al so hi gher than necessary. Factors such as excessi ve
staffi ng and poor management i n the el ectri ci ty
sector i ncrease costs, and there are si mi l ar i neffi -
ci enci es i n the oi l suppl y sector. I nsofar as the
popul ati on i s aware of these probl ems, they maybe
rel uctant to agree to pri ce i ncreases that woul d i n
effect subsi di ze the i neffi ci enci es of the suppl y
system. I mproved effi ci enci es on the suppl y si de
mi ght make i ncreases i n pri ces and tari ffs more
pal atabl e and al so hel p to mi ni mi ze the total cost to
consumer s.
I n addi ti on to the general l evel of energy pri ces,
the structure of energy pri ces i s of concern i n both
the el ectri ci ty and petrol eum product markets. Major
di fferences i n the pri ces charged for si mi l ar servi ces--
as i n the case of el ectri ci ty--or for petrol eum
products that can be substi tuted for each other
have gi ven ri se to di storti ons i n product demand.
The subsi di zati on of some fuel s (kerosene and di esel
fuel s) for general economi c and soci al reasons,
combi ned wi th hi gh taxes on others (gasol i ne), l eads
to shortages of the subsi di zed fuel s, surpl uses of the
hi ghl y taxed fuel s, and capi tal i nvestment deci si ons
made on the basi s of energy costs that do not refl ect
the cost of provi di ng that energy.
I n Thai l and i n the earl y 1980s, for exampl e, pri ce
di fferences between gasol i ne, di esel , kerosene, and
l i qui d petrol eum gas (LPG) l ed to shortages and
bl ack markets i n kerosene and LPG; the di versi on of
hal f of the total kerosene suppl y to the transport
sector to adul terate di esel fuel ; di esel i zati on of many
ol der vehi cl es by retrofi tti ng a spark i gni ti on engi ne
to use di esel fuel ; wi despread theft of di esel fuel ; and
surpl uses of gasol i ne as al l vehi cl es used commer-
ci al l y changed over to di esel .
37
Si mi l ar devel op-
ments i n other countri es have contri buted to seri ous
refi nery i mbal ances. I n recent years, Thai l and has
moved to reform i ts petrol eum product pri ci ng
system, but wi de pri ce di fferenti al s persi st i n other
countri es, i ncl udi ng I ndonesi a and I ndi a.
Energy pri ci ng deci si ons are often moti vated by
the need to keep energy affordabl e for l arge popul a-
ti ons of poor househol ds. However, the practi cal
i mpl ementati on of such pol i cy i s di ffi cul t. I t i s often
di ffi cul t to target di sadvantaged groups. Energy
consumpti on surveys i ndi cate that the use of com-
merci al fuel s i s concentrated among mi ddl e and
upper i ncome househol ds, rather than the poor who
rel y mai nl y on wood and charcoal . Moreover, i f
subsi dy programs expand i n scal e, they can l ead to
outcomes that penal i ze the very peopl e they are
desi gned to hel p. Thus, the deteri orati ng revenue
si tuati on of el ectri ci ty systems, attri butabl e i n some
measure to subsi di zed tari ffs, l eads to decl i ni ng
qual i ty and avai l abi l i ty of power suppl i es, whi ch can
cause factori es and workshops to stop operati ons,
thus i ncreasi ng unempl oyment.
Though i mportant, pri ci ng i s just one mechani sm
for i nfl uenci ng energy demand. Others i ncl ude
measures to i nform consumers of cost-effecti ve
opportuni ti es to save energy, the i mposi ti on of
techni cal effi ci ency standards, and sponsorshi p of
energy-effi ci ent technol ogi es.
Devel opi ng countri es, frequentl y ai ded by donor
countri es and organi zati ons, have made some prog-
ress i n demand management and conservati on. For
exampl e, the Associ ati on of South-East Asi an Na-
ti ons (ASEAN) countri es (see box 2-B) have been
parti cul arl y acti ve i n conservati on i n both i ndustri es
and bui l di ngs. I n addi ti on, Chi na has establ i shed
energy conservati on techni cal centers, whi ch have
contri buted to the sharp decl i ne i n Chi nas energy
i ntensi ty. I n Brazi l , energy-savi ng protocol s have
been establ i shed wi th major i ndustri es. Korea has
conducted major audi ts of l arge compani es. Traffi c
management schemes, desi gned mai nl y to al l evi ate
congesti on, but wi th an energy-savi ng bonus, have
been i ntroduced i n Brazi l , Si ngapore, Thai l and, and
Venezuel a. New, more energy-effi ci ent automobi l e
technol ogi es have been i ntroduced i n I ndi a. And
several i mproved wood-burni ng stoves have been
i ntroduced, at l east one of whi ch appears to have
enjoyed consi derabl e success.
38
On the i nsti tuti onal
si de, movements toward deregul ati on of economi c
acti vi ty, as i n Chi na i n the 1980s, have i mproved the
competi ti ve envi ronment under whi ch energy deci -
si ons are taken and thus have contri buted to i m-
proved energy effi ci ency.
37Donald Hert.zmarIG Energy Efficiency and Energy Pricing in Developing Countries, op. cit., footnote 26.
sss~uel Baldwin, Howard Geller, Gautan Dut4 and N.H. Ravindrarnat.4 Improved Wtiburnin
g Cookstoves: Signs of Success, AA4BZ0, vol.
14, No. 4-5, 1985, pp. 280-287.
L
Box 2-BEnergy Conservation Initiatives in ASEAN Countries
In Southeast Asi a, many governments are adopti ng and i mpl ementi ng l aws to encourage energy conservati on
i n bui l di ngs and i ndustry. Desi gn standards have been enact@ or are bei ng consi dered, i n most ASEAN countri es.
I n Mal aysi a, the Mi ni stry of Energy, Tel ecommuni cati ons, and Posts has embarked on devel opment of energy
standards for new bui l di ngs, wi th the goal of reduci ng overal l usage by 10 percent by 1991 (5 to 15 percent for
l i ghti ng, 5 to 10 percent for ai r-condi ti oni ng, and 15 to 20 percent rel ati ng to heat gai n through bui l di ng envel opes).
These standards were wi del y ci rcul ated for revi ew, and were expected to be i mpl emented duri ng 1989. Some energy
audi ts have been commi ssi oned.
l
I n the Phi l i ppi nes, major energy consumers are requi red by l aw to have energy management programs, and
l arge customers must report thei r consumpti on to the Offi ce of Energy Affai rs (OEA) quarterl y. The OEA offers
a wi de range of conservati on servi ces, i ncl udi ng an energy management trai ni ng program, energy conservati on
bri efi ngs, i ndustry-speci fi c publ i cati ons, assi stance to the Energy Management Associ ati on of the Phi l i ppi nes (a
pri vate-sector group), consul ti ng and audi ts, effi ci ency testi ng, and i ndustri al effi ci ency moni tori ng. The Omni bus
Energy Conservati on Law mandates the devel opment of standards for energy use i n commerci al bui l di ngs, for
bui l di ng constructi on materi al s, and for desi gns of commerci al and i ndustri al bui l di ngs pri or to the i ssuance of
permi ts for bui l di ng or for addi ng equi pment such as ai r-condi ti oni ng uni ts.
Si ngapore encourages conservati on through educati onal programs. I ndi rect control s i mposed by the
governments Bui l di ng Control Department standardi ze vari ous desi gn features, such as overhangs and reduced
wi ndow area to decrease demands for ai r condi ti oni ng.
I n Thai l and, the governments Si xth Nati onal Economi c and Soci al Devel opment Pl an (1987-91) speci fi ed
targets for i ncreased effi ci ency i n transportati on, i ndustry, and househol ds. Tax reducti ons and l ow-i nterest l oans
for energy conservati on equi pment are avai l abl e.
l m 1986 s~dards i nsti tuted i u Malaysia areas follows: 1) Buildings whose connected electric service is over 250 kV are reqti
to have separate meters for lighting and outlets and for air conditioning systems. 2) Lighting loads are specified for interior spaces, several
building interiors, and roads and grounds in the vicinity of the building. Lighting controls are specified. 3) For air conditioning, dry bulb
temperatures are set at 25 degrees Celsius and relative humidity at 65 percent plus or minus 5 percent. Automatic setback and shutoff systems
are required.
Energy and The Traditional Sector
and rel ated tradi ti onal technol ogi es general l y have
l ow effi ci enci es and l i mi ted output and producti vi ty
Two-thi rds of the devel opi ng worl ds popul ati on
some 2.5 bi l l i on peopl e--l i ve i n rural areas
39
wi th
l ow standards of l i vi ng based l argel y on l ow-
resource farmi ng. Thi s type of farmi ng i s character-
i zed by hi gh l abor requi rements, l ow producti vi ty
per hectare and, because of the margi nal subsi stence,
strong ri sk aversi on. Rural popul ati ons have l i ttl e
access to commerci al fuel s and technol ogi es and
onl y l i mi ted connecti on wi th the modern economy.
Bi omass fuel s sati sfy the heati ng and cooki ng needs
of these popul ati ons, and muscl e power l argel y
provi des for thei r agri cul tural , i ndustri al , and trans-
portati on energy needs. Al though these energy
sour ces pr ovi de cr uci al ener gy ser vi ces at l i ttl e or no
di rect fi nanci al cost, bi omass fuel s, muscl e power,
l evel s (see ch. 3).
I n many areas, bi omass suppl i es are di mi ni shi ng
due to a host of factors, i ncl udi ng popul ati on growth
and the expansi on of agri cul tural l ands, commerci al
l oggi ng, and fuel wood use (see ch. 5). The poorest
rural peopl e often have l i mi ted access to even these
resources and, therefore, must spend l onger peri ods
of ti me foragi ng for fuel sources--exacerbati ng thei r
al ready di ffi cul t economi c posi ti on.
Tradi ti onal vi l l ages are compl ex, hi ghl y i ntercon-
nected systems that are careful l y tuned to thei r
envi ronment and the harsh real i ti es of survi vi ng on
meager resources.
40
Because the vi l l ages are l argel y
cl osed systems, changes i n any one part affect other
s~orld B@ World Development Report 1989, op. cit., footnote 1.
@See, for e=ple, M.B. Coughenour et al>
Energy Extraction and Use in a Nomadic Pastoral lkosysteq Science, vol. 230, No. 4726, Nov. 8;
1985, pp. 619-625; J.S. Singh, Uma Pandey, and A.K. Tiwari, Man and Forests: A Central - Ehrnalayan Case Study, AMBIO, vol. 13, No. 2, 1984,
pp. 80-87; Anmlya Kumar and N. Reddy, An Indian Village Agricultural Ecosystem-Case Study of Ungra Village, Part II: Discussion Biomuss,
vol. 1, 1981, pp. 77-88.
el ements of vi l l age l i fe. Changes i n agri cul tural
practi ces, for exampl e, change the amount and type
of energy suppl i es avai l abl e. I n turn, energy sector
devel opments, such as rural el ectri fi cati on, can have
major i mpacts on agri cul tural practi ce and i ncome
di stri buti on. Maki ng changes i n rural systems fre-
quentl y proves di ffi cul t due to the l arge ri sks that
changes can pose to popul ati ons l i vi ng on the margi n
of subsi stence.
The fol l owi ng secti ons exami ne four of the major
factors that affect the l i nkages between energy and
the economi c and soci al devel opment of rural
economi es: seasonal i ty; i nequi ti es i n the di stri bu-
ti on of and access to resources; the rol e of commer-
ci al bi omass i n the rural economy; and gender i ssues
i n l abor. Mechani zi ng the mundane tasks of rural
l i fe, a process faci l i tated by the i ntroducti on of
modern fuel s, coul d greatl y i ncrease the producti vi ty
of rural peopl es. To bri ng about i mprovements,
however, wi l l requi re payi ng cl ose attenti on to the
numerous rel ated compl i cati ons, such as seasonal -
i ty, the type of task, cul ture-speci fi c l abor rol es,
chi l drens l abor, and many others.
Seasonality ,
The seasons affect every aspect of rural l i fe: the
avai l abi l i ty of food, fuel , and empl oyment; the
i nci dence of di sease; and even the rates of ferti l i ty
and mortal i ty .41 Labor requi rements for pl anti ng are
seasonal l y peaked to take advantage of l i mi ted
rai nfal l and other favorabl e growi ng condi ti ons.
When rai ns begi n, soi l bacteri a mul ti pl y rapi dl y and
break down the dead pl ant matter i n the soi l l eft by
the dry season; thi s process rel eases a l arge amount
of organi c ni trogen i n the soi l . Crops pl anted qui ckl y
after the rai nfal l can take advantage of thi s ni trogen,
but a short del ay l eaves weeds as the mai n benefi ci -
ari es.
42
Labor requi rements to harvest crops are
si mi l arl y peaked (see fi gure 2-9). Thus, whi l e there
may be a l arge l abor surpl us duri ng most of the year,
l abor shortages occur duri ng the cri ti cal pl anti ng and
harvesti ng seasons. Studi es of Afri can agri cul ture
i ndi cate that l abor i s the major scarce resource i n
food producti on.
43
Modern equi pment coul d reduce the hi gh l abor
demands duri ng pl anti ng and harvesti ng. Even when
the necessary commerci al fuel s are avai l abl e, how-
ever, modern agri cul tural equi pment i s often prohi b-
i ti vel y costl y to purchase or rent due to the very short
peri od i n whi ch i t can be profi tabl y used.
44
Rel a-
ti vel y l ow-cost tradi ti onal technol ogi es face si mi l ar
cost barri ers. For exampl e, the average ani mal -
drawn cart i n Ungra, I ndi a, i s used at just 6 percent
of i ts annual capaci ty .45
Draft ani mal technol ogi es can ease cri ti cal sea-
sonal l abor shortages to some extent. Draft ani mal s,
however, can onl y be used producti vel y for l i ttl e
more than the short growi ng season, yet these
ani mal s requi re food year round. Li mi ted uses for
draft ani mal s, coupl ed wi th thei r hi gh food requi re-
ments, reduce the average draft ani mal effi ci enci es
to just a few percent. Because of the l i mi ted suppl i es
of fodder avai l abl e, farmers often semi -starve draft
ani mal s i n order to save fodder for when the ani mal s
need thei r strength to pl ow the dry-baked ground or
for other purposes.
46
A shortage of draft ani mal s
may l i mi t crops to just one per year-even i n areas
wi th potenti al for doubl e croppi ng.
47
Al though agri cul ture demands very hi gh l evel s of
l abor duri ng the peak seasons, duri ng the remai nder
of the year, rural areas experi ence seri ous under-
41R0~fl
Chambers, Richard bnghurs~ and Arnold Pacey (eds.), Seasonal Dimensions to Rural Poverty (lcmdon: Frances Pinter Publishers, Ltd.,
and Ibttawa, NJ: Allanheld, Osmund2 Co., 1981); Robert Chambers, RuralPoverty Unperceived: Problems and Remedies, WorZdDeve/opment, vol.
9, 1981, pp. 1-19.
42
Robert Chambers, Richard Longhursl and Arnold Pacey (eds.), Seasonal Dimensions to RuraJ Poverty, op. cit., footnote 41, pp. 10-11.
asJeame K~p_ He- F~ing the Cities and Fe-g the p~nts:~t Role for Africas Woma F~ers? wOr~DeveZOpment, VO1. 11,
No. 12, 1983, pp. 1043-1055.
~prabhu ~~, Yves Bigo~ ad H~ p. B-anger, Agricul~ral Mechanization and the Evo~ution of ~ar?ning systems in hb-sdulra?l &ICcl
(Baltimore, MD: Johns Hopkins University Press for the World B* 1987).
asAs a COm~uenceof lowu~tion rates, villagers preferred lower cost wooden wheels OverhigherpdO~ ce pneumatic tires until the depletion
of timber resources caused the price of wooden wheels to rise signit3cantly. At the same time, these price rises led to an active market in second-hand
traditional carts. H.I. Somashekar, N.H. Ravindranath, and Amulya Kumar N. Reddy, Studies on the Ungra Village Agricultural Ecosystem, Part III:
Animal Drawn Carts and Transport (_&mgalore, India: AS-no &te).
*H.I. Somashekar, N.H. Ravindranati and Amulya Kumaz N. Reddy, Studies on the Ungra Village Agricultural Ecosystem Part III: AnimalDrawn
Carts and Transport, op. cit., footnote 45; Jane Bartlett and David Gibbon, Animal Draught Technology: An Annotated Bibliography @mdon: ITDG,
1984); Peter Munzinger, Animal Traction in Afica (Eschborq West ~y: GIZ, 1982).
4TN.H. RZWh&a~ and H.IN. chnaky~ Biomass Based Energy System fOr a South hdian Village, Biomass, vol. 9, No. 3,1986, pp. 215-233.
Chapter 2--Energy and Economic Development . 41
Figure 2-9-Seasonal Pattern of Energy Expenditure on Agricultural and Other
Essential Tasks: Adult Farmers, Genieri Village
Kcal/day
2000
r
F e m a l e s -
1600
1200
I
.
r-
.-
1
1
I
I
w
I
- -
1 :
/
I
. L . :-_ ---- ------
- .
I
?
1
1
-
I
I
1 -
I
t I
J
-
H
1
I

800
2
_ j - - H _ _ - , s = + _ -
I
I
1 !
Female annual
.- - -
average
Male annual
- - --
Y
1 J , , , I , I I I I I I L
I
- -
Males
I l l - -
; - - 7 . - -
I
I
- - . I
I
I
400
I
I
- 1
I
I
I
Rainy season
I
I
4 ~
I
o
Ma y J une J ul y Aug Se pt Oc t Nov De c J a n Fe b Ma r Apr
- Mal es Females
SOURCE: Margaret Haswell, EneWy for Subsistence (London: MacMillan Press, Ltd., 1981).
empl oyment. I n turn, thi s seasonal unempl oyment i n
rural areas propel s a l arge amount of both seasonal
and permanent mi grati on to urban areas.
48
I n Afri ca
and Asi a, where the mi grants are mostl y men,
49
more of the burden for subsi stence crop producti on
i s shi fted to the women who stay behi nd. Mi grati on
to ci ti es i ncreases pressure on forests as wel l , be-
cause urban dwel l ers general l y purchase their wood
suppl i es, whi ch are l i kel y to be deri ved from cutti ng
whol e trees, rather than the gatheri ng of twi gs and
branches as i s more typi cal of rural foragers.
rai ny season, wood i s l ess easi l y obtai ned and more
di ffi cul t to burn than duri ng the dri er months. I n
areas heavi l y dependent on crop resi dues for fuel ,
shortages at the end of the dry season can force the
use of noxi ous weeds as substi tutes, parti cul arl y by
the very poor.
50
The use of bi ogas i s l i mi ted by the
avai l abi l i ty of dung, whi ch i n Ungra, I ndi a, vari es
for adul t cattl e from 3.5 kg/day duri ng the 7-month
dry season to 7.4 kg/day duri ng the wet season.
51
Cor r espondi ngl y, i n mountai nous ar eas or el sewher e
wi th l arge seasonal temperature vari ati ons, fuel
The seasons al so affect the avai l abi l i ty and demands can i ncrease si gni fi cantl y duri ng the
usabi l i ty of r enewabl e ener gy r esour ces. Dur i ng the wi nter. I n a Kashmi r vi l l age, for exampl e, wood
4sM.ic~el P. Todaro, Econonuc Development in the Third World (New York NY: ~x44uuW Inc., 1977); Oerald M. Meier, Leading Issues in
Econonuc Development, 4th ed. (New York NY: Oxford University Press, 1984); Scott M. Swinto~ Peasant Farnung Practices and Off-Farm
Employment in Puebla, Mm-co (Ithac& NY: Cornell University, 1983).
49Mic~el p. T~~, Econo~c Develop~nt in the Third world, op. Ci t., f~tnote 48, pp. 192-193. Note tit in IAtk America more wOmeIl tbl
men now migrate.
%&un Vidyarthi, Energy and the Poor in an I.ndian Village, WorZdDeve20pment, vol. 12, No. 8, 1984, pp. 821-836.
SIN*H. ~v~&~~and HJ q. a%Bio~s Based ~erg System for a SOUth Indian Vikge, op. Cit., fOOtiOte 47.
demands duri ng the wi nter are four ti mes hi gher than
duri ng the summer.
52
Inequities in Resource Distribution
and Access
I n regi ons where bi omass fuel suppl i es are
l i mi ted-parti cul arl y those wi th dry cl i mates and/or
hi gh popul ati on densi ti es-rural peopl e may travel
l ong di stances to col l ect fuel for domesti c use, as
much as 20 mi l es round tri p i n some areas under
speci al condi ti ons. More general l y, when wood i s
scarce they rel y on crop wastes, ani mal dung, or
other materi al s as substi tutes. Esti mates of ti me
spent i n foragi ng range as hi gh as 200 to 300
person-days per year per househol d i n Nepal .
53
Foragi ng i s al so heavy work. I n Burki na Faso,
typi cal headl oads wei gh 27 kg (60 pounds) .54 I n
many regi ons, women and chi l dren shoul der most of
the burden.
Despi te these heavy burdens, vi l l agers often
prefer to i nvest thei r capi tal and l abor i n technol o-
gi es for i ncome-produci ng acti vi ti es, such as yarn
spi nners, rather than i n fuel -conservi ng stoves or
tree-growi ng efforts.
55
Reasons for thi s i nvestment
preference i ncl ude l ack of cash i ncome; the abi l i ty to
mi ni mi ze wood use or to swi tch to al ternati ve fuel s
when wood becomes scarce;
56
confl i cts over owner-
shi p of l and or trees; and easy access to common
l ands. I n addi ti on, vi l l agers often carry out fuel wood
col l ecti on i n conjuncti on wi th other tasks, such as
wal ki ng to and from the fi el ds or herdi ng ani mal s. I n
thi s case, col l ecti ng bi omass resources may prove
l ess burdensome than i t appears.
57
I f wood i s scarce, vi l l agers use crop wastes, dung,
or other l ess desi rabl e fuel s. To the vi l l age user, the
i mmedi ate val ue of these fuel s outwei ghs thei r
potenti al l ong-term envi ronmental costs.
58
I n I ndi a,
for exampl e, a ton of cow dung appl i ed to the fi el ds
produces an esti mated i ncrease i n grai n producti on
worth $8 (U.S. dol l ars), but i f burned el i mi nates the
need for fi rewood worth $27 i n the market.
59
The
di versi on of crop resi dues, previ ousl y used as soi l
enhancers, to fuel use may l ead to a l ong-term l oss
i n soi l ferti l i ty unl ess offset by i ncreased use of
chemi cal ferti l i zers.
Local fuel shortages often have thei r most seri ous
i mpacts on rural l andl ess and/or margi nal farmers
wi th l i ttl e access to fuel suppl i es. The poor may al so
someti mes be deni ed access to thei r tradi ti onal fuel
sources when the market val ue of bi omass ri ses.
60
For exampl e, farm l aborers i n Haryana, I ndi a, are
now someti mes pai d i n crop resi dues for fuel rather
than i n cash, al though previ ousl y they had free
access to these agri cul tural wastes.
6l
The Role of Women
Women shoul der the burden of most domesti c
tasks, i ncl udi ng foragi ng for fuel wood and cooki ng.
I n many areas they al so perform much of the
subsi stence agr i cul tur al l abor . A 1928 sur vey of 140
Sub-Saharan ethni c groups found that women car-
ri ed a major responsi bi l i ty for food farmi ng" i n 85
percent of the cases, and di d al l but the i ni ti al l and
52~J id Huss~ Fuel CO
nsumption Patterns in High Altitude Zones of Kashmir and LadaklL Energy Environment Monitor @iia), vol. 3, No.
2, September 1987, pp. 57-62.
ManandForests: A Central
53JCS. sw Umpmdey, and AK. Tiw~,
Hunalayan Case Study, AMBZO, vol. 12, No. 2,1984, pp. 80-87; Kedar
I-al Shresttq Energy Strategies in Nepal and Technological Options (Nepal: Research Center for Applied Science and Technology, Tnbhuvan
University, for the End-Use Oriented Global Energy Workshop, Sao Paulo, Brazil, June 1984). The World Bank Energy Sector Assessment for Nepal
estimated tbat 16 percent of all labor went for fuelwood and animal fodder collection.
fiE. tis~ Fuel Co
nsumptionAmong Rural Families in Upper VOIW West Afric~ paper presented at Eighth World Forestry Conference, Jakar@
Indonesi% 1978.
SSV~ Vidyd, ~ergy and the Poor in an Indian Village, op. cit., fOotnOte 50.
Mphil OKe~e and Barry MunsIow,
Resolving the Irresolvable: The Fuelwood Problem in Eastern and Southern Africz paper presented at the
ESMAP Eastern and Southern Africa Household Energy Planning Seminar, Harare, Zimbabwe, Feb. 1-5, 1988.
sTfiene T~er, me R~ Rural Energy crisis: Womens Time,
Energy Journal, vol. 8, 1987, pp. 125-146.
5%eoffrey Barnard and Lars Kristoferso~ Agricultural Residues as Fuel in the Third World (Washington, DC, and London: Earthscan and
International Institute for Environment and Developmen
t
% Energy Information progrmq Technical Report No. 4, 1985).
59G.C0 Awm~ ~d NOT s% ~~fimW ad fionomic Rem From ~tfle Dug u ~nure md R@ Energy, vol. 9, No. 1, 1984, pp. 87-9Q
see also G.C. Aggarw~ Judicious Use of Dung in the Third World, Energy, vol. 14, No. 6, 1989, pp. 349-352; Eric Eckhohn et al., Fue2wood: The
Energy Crisis That Wont Go Away (1..ondon: EarthscarL 1984), p. 105; Ken Newcombe, World Bank, Energy Departmen~ AnEconomic Justifkation
for Rural Afforestation: The Case of Ethiopia, 1984.
@V~ Vidyfi, Ener~ and the Poor in an Indian Village, Op. Cit., footnote 50.
61 Cen~e for Science ad ~v~omen~ The State of IndiasEnviro~nt 1984-85: The Second Citizens Report (New De~: 1985).
Chapter 2--Energy and Economic Development . 43
cl eari ng i n 40 percent of the cases.
62
I n contrast,
the Musl i m custom of Purdah, for exampl e, tends to
keep women near thei r homes and away from the
fi el ds i n Bangl adesh.
63
As womens work often does
not produce any cash revenue, opportuni ti es for
i ntroduci ng energy- and l abor-savi ng technol ogi es
for womens work are l i mi ted. I mprovi ng l abor
producti vi ty and energy effi ci ency i n rural areas wi l l
thus requi re speci al attenti on to the rol e of women.
The carel ess i ntroducti on of l abor-savi ng technol -
ogi es coul d i ncrease the burden on women. For
exampl e, the i ntroducti on of ani mal or mechani cal
tracti on for l and preparati on and pl anti ng i ncreases
the area that men can cul ti vate, but does nothi ng to
assi st women i n weedi ng, harvesti ng, post-harvest
food preparati on, storage, and other tasks.
64
The mi grati on of men to l ook for urban work
l eaves women to ful fi l l tradi ti onal mal e rol es as wel l
as thei r own. I n Uttar Pradesh, I ndi a, the mal e:femal e
rati o i n vi l l ages i s 1:1.4 for the worki ng age group of
15 to 50 years.
65
I n Kenya, a quarter of rural
househol ds are headed by womeni n Botswana, 40
per cent.
66
Yet the remi ttances of the mi grants can
make an i mportant contri buti on to rural househol d
fi nances.
Chi l dren, too, pl ay an i mportant rol e i n rural
l abor, freei ng adul ts to perform more di ffi cul t
tasks.
67
I n Bangl adesh, for exampl e, chi l dren begi n
performi ng certai n tasks as earl y as age 4. By age 12,
boys become net producers-produci ng more than
they consume-and are nearl y as effi ci ent i n wage
work as men. By age 15, boys have produced more
than thei r cumul ati ve consumpti on from bi rth, and
by 22 they have compensated for thei r own and one
si bl i ngs cumul ati ve consumpti on.
68
The major rol e
of chi l dren i n farmi ng hel ps expl ai n hi gh ferti l i ty
rates i n rural areas.
The Role of Commercial Biomass in
the Rural Economy
Whi l e much bi omass i s used l ocal l y, rural areas
are al so the source of substanti al amounts of
fuel wood (both fi rewood and charcoal ) used i n
towns.
69
Thi s trade pumps rel ati vel y l arge amounts
of cash i nto the rural economy and provi des much-
needed empl oyment to rural dwel l ers duri ng non-
agri cul tural seasons. To suppl y Ouagadougou, Bur-
ki na Faso, wi th wood duri ng 1975, for exampl e,
requi red some 325,000 person-days of l abor and
generated over $500,000 i n i ncome di rectl y and an
addi ti onal $2.5 mi l l i on i n i ncome through transport
and di stri buti on.
70
Such marketi ng networks can be
qui te extensi ve and compl ex.
71
I n many countri es, peopl e i n the poorest areas,
where condi ti ons do not permi t expansi on of crop or
ani mal producti on and natural woody vegetati on i s
the onl y resource, depend heavi l y on sal es of
GzJ~nne Koopman Hem Feeding the Cities and Feeding the P~
ts: What Role for Africas Women Farmers? op. cit., footnote 43.
6qMmd x cm me Economic Activities of children in a Village in Banglade~ Population and Deve20pment Review, VO1. q, No. s, September
1977, pp. 201-227; GloriaL. Scott and Marilyn Carr, The Impact of Technology Choice on Rurrd Women in Bangladesh World Bt@ Staff Worldng
Paper No.731, Washington DC, 1985.
apeter M~inger, &imaJ Traction Zn Afica, Op. cit., foo~ote 46.
GSJOSOs- Um p~dey, and A.K. Tiw~ Wan and Forests: A Centi ~
yan Case Study, op. cit., footnote 40.
=World B@ population Growth and Policies in Sub-Saharan Afica ~Mhk@OIL DC: l!J W, P. 39.
67rngrid pwer ~ otd.
Childrens labor, especially daughters, is usually more signiilcant tban husbands in easing a work bottleneck for
women. Ingrid Palmer, Seasonal Dimensions of Womens Roles, in Robert Chambers, Richard Longhurst, and Arnold Pacey (eds.), Seasonal
Dimensions to Rural Poverty, op. cit., footnote 41.
68M~d T. Cm ~~rhe Econofic ~tivities of ~dren ~ a Vfiage in Banglade~ op. Cit., foottlote 63.
@The value of commercialized fuelwood and charcoal exceeds 10 percent of the gross domestic product in countries such as Burkina Faso, Ethiop@
and Rwanda and exceeds 5 percent in Libe@ Indonesia, Zaire, Mali, and Haiti. Philip Wade and Massimo Pahnie@ What Does Fuelwood Really
Cost? UNASYLVA, vol. 33, No. 131, 1981, pp. 2023. George F. Taylor, II, and Moustafa Soumare, Strategies for Forestry Development in the West
African Sahel: An Overview, RuraZAficana, Nos. 23 and 24, Fall 1985 and Winter 1986.
nJ.E.M. hold, wood Ehergy and Rural Communities,
Natural Resources Forum, vol. 3, 1979, pp. 229-252.
Tl~fiBefi~~ 4<~ketingNe~orks forForestfiels to supplyufi~c~te~ in the S~el,R~aZAficana, Nos. 23 md24, Fdl 1985 ~d Winter
1986.
44 . Energy in Developing Countries
fi rewood for thei r i ncome.
72
Si mi l arl y, when crops
fai l , charcoal producti on
73
--or, as i n Bangl a-
desh, the cutti ng of wood from farm hedgerows
grown i n part as an economi c buffer to be sol d before
more val uabl e l i vestock and l and
74
provi des al -
ternati ves for earni ng cash. I n I ndi a, for exampl e,
headl oadi ng (i ndi vi dual s carryi ng wood to urban
markets on thei r heads) has become an i mportant
source of i ncome for perhaps 2 to 3 mi l l i on peopl e .75
The response of rural peopl es to fuel shortages
vari es wi del y. Some sel l wood to urban markets and
use the l ower qual i ty resi dues themsel ves. Others
use dung for fuel rather than for ferti l i zer. I n Mal awi ,
to grow suffi ci ent fuel for househol d use on the
typi cal fami l y farm woul d di spl ace mai ze worth
perhaps 30 ti mes more; col l ecti ng free wood
proves much easi er.
76
I n contrast, aeri al surveys of
Kenya have shown that hedgerow pl anti ng i ncreases
wi th popul ati on densi ty-demonstrati ng that vi l l ag-
ers respond to the reduced opportuni ty of col l ecti ng
free wood from communal l ands by growi ng thei r
own.
77
Conclusion
Hi gh rates of economi c growth wi l l be needed i n
devel opi ng countri es to provi de the rapi dl y growi ng
popul ati on wi th i mproved l i vi ng standards. I f pres-
ent trends i n energy and economi c growth conti nue
i nto the future, a sharp i ncrease i n commerci al
energy consumpti on i n devel opi ng countri es wi l l be
requi red. Substanti al i ncreases i n suppl i es of bi o-
mass fuel s wi l l al so be needed. Thi s prospect rai ses
a di l emma. On the one hand, i ncreases i n energy
suppl i es on thi s scal e woul d severel y strai n fi nan-
ci al , manpower, and envi ronmental resources. But
on the other, i nabi l i ty to suppl y needed energy can
frustrate economi c and soci al devel opment.
One way to resol ve thi s di l emma l i es i n di sti n-
gui shi ng between the energy that i s consumed and
the servi ces deri ved from i t. Technol ogi cal i mprove-
ments and other means offer the potenti al of greatl y
i mprovi ng the effi ci ency of energy use-provi di ng
more of the energy servi ces needed for devel opment
whi l e consumi ng l ess energy. The next chapter
provi des an i ntroducti on to the servi ces provi ded by
energy, and how they are currentl y del i vered, wi th a
vi ew to i denti fyi ng potenti al s for i mprovi ng effi -
ci enci es.
TZJ.E.M. Arnold, wood Energy and Rural Communities,
op. cit., footnote 70, Centre for Science and lhvironmen~ The State of Indias
Environment 1984-85: The Second Citizens Report, op. ci~, footnote 61.
T3D.C). Hall and P.J. de Groot, Biomass For Fuel and Food-A Parallel Necessity,
draft for A#vances in Solar Energy, Karl W. Boer (cd.), vol.
3, Jan. 10, 1986; RafiqulHudaChaudhury, The %asonality of Prices and Wages in Banglade~ in Robert Chambers, Richard Imnghurst, and Arnold
Pacey (eds.), Seasonal Dimensions to Rural Poverty, op. cit., footnote 41.
TdRick J. V~ Dm Beld~
S
Upplying Firewood for Household Energy,
in M. Nurul Islaq Richard Morse, and M. Hadi Soesastro (eds.), Rural
Energy To Meet Deve/opnzent Aleeds (Boulder, CO: WestView Press, 1984).
TSCen~e for Science and EINuo~a~ The Smte of IdiasEnvironrnent 1984-85: The Second Citizen sReport, op. cit., footnote 61, p. 189.
TGD. fiencm me Economics of BioEnergy in Developing COuIttrieS,
in H. Egneus et al. (eds.), Bioenergy 84, Volume V: Bioenergy in Developing
Countries (Amsterdam: Elsevier, 1985). It is estimated that 90 percent of all rural households collect all their woo@ IOpercent purchase some of their
wood at $0.50m3 or $0.04/GJ. Urban households buy their wood at a cost of $0.12/GJ. In contras~ plantation-derived fuelwood can cost $1.50 to
$2.00/GJ. A farmer could plant trees, but the loss of 0.4 hectare of farmland reduces maize production by a total of $125 and profit by $30. In contras~
trees produced on 0.4 hectare will be worth $6 in 7 years.
~p.N. Bradley, N. Chava.ngi, and A. Van GeIder, Development Research and Energy Planning in Kenya, -1o, VO1. 14, NOS. 4-5, 1985, PP.
228-236.
Chapter 3
Energy Services in
Contents
Page
I ntroducti on
. . * *, .,.*,.,. . . 47
The Resi denti al /Commerci al Sector
... ... ... **. ***. .*, .., ... ... ,.. ,. *Q,**,+**,***+
50
Cooking ... ... ... *.. ... ... ... ... ... ... *.. ... ... ...
+
..
$
*** ***. ..
c
**.
++q4 @
**
4
****
51
Li ghti ng ..*. *.. ... ... ..*. ... ... ... ... ... ... ... ... ... .**
+
*.. .*. ..*
+ q
**e.***.****
55
Space Condi ti oni ng, Refri gerati on, and Other Appl i ances
. .........*..*..,.,..***., 58
The I ndustri al Sector ... ... ... ... ... ... ... ... *.. ... ... ... .**. *.. *.** .**.
+4@+
.* **a,
62
and Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Process Heat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., * . . . . . . . . * * * . * . . * . . * * . * * *,*.,,...
63
Mechani cal Dri ve ... ... ... ... ... ... ... ... *.. ... .
o
. ... ***. *.** .**. **. ..
a
*v*.***
67
Barri ers to Effi ci ency I mprovements
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
71
The Tracti on (Agri cul tural ) and Transportati on Sectors
. .......**,.,*,*...**.**..*,. 72
Tracti on and Agri cul ture ... ... ... ... ... ... ... ... ... ...
e
#. *.*. *.** ..
+
* .
<
*
c
. .
+
. .
o 72
Transportati on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
76
Concl usi on ... ... ... ... ... ... ... ... ... *.. ... ... ..*. ... ..
$
***
+
*.** **** ..
<
* ****be@*,
80
Appendi x 3-A: Energy Bal ances for Sel ected Devel opi ng Countri es . . . . . . . . . . . . . . . . . . 81
Appendi x 3-B: Sources for Tabl e 3-9
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Box
. . . . . . . . . * . . . * * * * . , . . * . . . , . . , . . Pl ow q . . . . . .
Figures
Figure Page
3-L Per-Capi ta Energy Use i n Vi l l age Househol ds i n Devel opi ng Countri es . . . . . . . . . . . . . . . . . 50
3-2. Choi ce of Cooki ng Fuel by I ncome for Fi ve Medi um-Si zed Towns i n Kenya ... ... ......52
3-3. Representati ve Effi ci enci es and Di rect Capi tal Costs for Vari ous Stoves . . . . . . . . . . . . . . . . 53
3-4. Di rect Energy Use for Cooki ng i n West Java, I ndonesi a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3-5. Dai l y Load Profi l es for Cooki ng Energy, Pondi cherry,I ndi a, 1980 . . . . . . . . . . . . . . . . . . . . . 55
3-6. Li ght Output and Effi ci ency of Vari ous Li ghti ng Technoi ogi es . . . . . . . . . . . . . . . . . . . . . . . . . 56
3-7. Costs of Vari ous Li ghti ng Technol ogi es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3-8. Changes i n Capaci ty and Type of I nstal l ed El ectri c Li ghti ng Per Househol d
Wi th I ncome Level South Bombay, I ndi a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3-9. Househol d El ectri ci ty Use for Li ghti ng v. Househol d I ncome, i n Brazi l . . . . . . . . . . . . . . . . . 59
3-10. E1ectri c Appl i ance Ownershi p i n Urban Mal aysi a by I ncome Group, 1980 .. .. ... ... .....60
3-11. El ectri c Appl i ance Ownershi p i n Urban Java,1988 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3-12. Refri gerator Ownershi p i n Bei ji ng, Chi na, 1981-87 ....*.,. . . . . . . . . . . . . . . . . . . ...***... 60
3-13. Reducti on i n the Real Cost of Refri gerators Over Ti me i n the Uni ted States . . . . . . . . . . . . . 60
3-14. Per-Capi ta Steel Consumpti on v. GNP for Vari ous Countri es . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3-15. Per -Capi ta Cement Pr oducti on v. GNP for Var i ous Countr i es ..?,.... . . . . . . . . . . . . . . . . . . 66
3-16. Effi ci ency of El ectri c Motors i n the Uni ted States, Brazi l , and I ndi a . . . . . . . . . . . . . . . . . . . . 71
3-17. Energy Losses i n an Exampl e El ectri c Motor-Dri ven Pumpi ng System
i n the Uni ted States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3-18. Use of Agri cul tural Pumpsets i n I ndi a 1950-90 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3-19. Passenger and Frei ght Transport v. GNP for Sel ected Countri es, 1%0-81 . . . . . . . . . . . . . . . 77
Tables
Table Page
3-1. Total Del i vered Energy by Sector, i n Sel ected Regi ons of the Worl d, 1985 ........ .....49
3-2. Del i vered Energy Per Capi ta by Sector i n Sel ected Regi ons, 1985 . . . . . . . . . . . . . . . . . . . . . 49
3-3. Per Capi ta Energy Use by Servi ce i n Sel ected Countri es . . . . . . . . . . . . . . . . . +....,,.. . . . 49
3-4. Pri nci pal Cooki ng Fuel s Used by the Worl d Popul ati on, 1976 . . . . . . . . . . . . . . . . . . . . . . . . 51
3-5. Typi cal Ai r Pol l uti on Emi ssi ons From Vari ous Cooki ng Fuel s . . . . . . . . . . . . . . . . . . . . . . . . 52
3-6. Resi denti al Nonheati ng El ectri ci ty I ntensi ty i n Sel ected Countri es, 1970 and 1986 ... ...61
3-7. Kenyan Nati onal Ener gy Use by Fuel , 1980 . . . . . . . . ..,.,...* . . . . . . . . . . . . . . . . . . . . . . . 63
3-8. Annual Consumpti on of Fuel wood and Charcoal i n Kenya by Rural Cottage I ndustri es,
(GJ/Capi ta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3-9. Effi ci ency of Fuel Use i n Tradi ti onal (Devel opi ng Countri es)andModern
(I ndustri al Countri es) Commerci al and I ndustri al Operati ons . . . . . . . . . . . . . . . . . .,,..,.*, 65
3-10 Energy Consumpti on by Chi nese I ndustry, 1980 . . . . . . . . . . . . . . . .,,,,... . . . . . . . . . . . . . 65
3-11. Average Energy I ntensi ti es of Bui l di ng Materi al s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3-12. Energy I ntensi ti es of End Products Usi ng Al ternati ve Bui l di ng Materi al s ... ... ... ......67
3-13. Average Dai l y Househol d Consumpti on of Water, Gujurat, I ndi a . . . . . . . . . . . . . . . . . . . . . . 68
3-14. I ndustri al El ectri ci ty End Use i n Brazi l , 1984 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3-15. Projected El ectri ci ty Consumpti on i n I ndi a by Sector and End Use, 1990...............70
3-16. Agri cul tural I ndi cators for Sel ected Countri es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..+. 74
3-17. Passenger Fl eet Si ze and Growth i n Sel ected Countri es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3-18. Energy Effi ci ency of Trucks i n Sel ected Countri es . . . . . . . . . . . . . . . . . . . . . . . . . . ...,.*.,. 79
3-19. Compari son of Rai l Systems i n Chi na, I ndi a, and the Uni ted States . . . . . . . . . . . . . . . . . . . . 80
3A-1. Brazi l : Energy Suppl i es and Servi ces, 1987, Exajoul es and Percent of
Nati onal Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3A-2. Chi na: Energy Suppl i es and Servi ces, 1987, Exajoul es and Percent of
Nati onal Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3A-3. I ndi a: Energy Suppl i es and Servi ces, 1985, Exajoul es and Percent of
Nati onal Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3A-4. Kenya: Energy Suppl i es and Servi ces, 1980, Petajoul es and Percent of
Nati onal Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3A-5. Tai wan: Energy Suppl i es and Servi ces, 1987, Petajoul es and Percent of
Nati onal Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3A-6. Uni ted States: Energy Suppl i es and Servi ces, 1985, Exajoul es and Percent of
Nati onal Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Chapter 3
Energy Services in Developing Countries
Energy use i n the worl ds devel opi ng countri es i s
i ncreasi ng rapi dl y. I n 1960, devel opi ng countri es
consumed just 15 percent of the worl ds commerci al
fuel s; by 1985, thei r share had i ncreased to about 24
per cent.
1
I ncl udi ng tradi ti onal fuel s, the devel opi ng
country share of worl d energy consumpti on was
about 28 percent i n 1986. Devel opi ng countri es
mi ght consume as much commerci al energy as
todays i ndustri al countri es by earl y i n the next
century.
2
Factors dri vi ng thi s rapi d i ncrease i n
energy use i ncl ude popul ati on growth, economi c
growth, and i ncreasi ng urbani zati on (see ch. 2).
Even wi th thi s rapi d growth, overal l per-capi ta
energy consumpti on rates i n devel opi ng countri es i n
2025 woul d be just one-fi fth that of the Uni ted States
i n 1987.
The purpose of thi s chapter i s to exami ne how
energy i s used i n devel opi ng countri es. As expl ai ned
i n chapter 1, the focus of thi s report i s on the servi ces
energy provi des rather than the amount of energy
consumed. The reason for thi s approach i s si mpl e:
energy i s not used for i ts own sake, but rather for the
servi ces i t makes possi bl e. For exampl e, wood mi ght
be burned to cook food, heat water, warm a house on
a wi nter eveni ng, heat an i ndustri al boi l er, or to
provi de other servi ces.
3
Si mi l arl y, di esel and gaso-
l i ne are used pri mari l y to provi de transportati on
servi ces.
There may be many di fferent means of provi di ng
a desi red servi ce, each wi th i ts own costs and
benefi ts. Transportati on, for exampl e, mi ght be
provi ded by bi cycl es, motorcycl es, cars, buses, l i ght
rai l , or ai rcraft. The consumer chooses among these
accordi ng to such cri teri a as cost, comfort, conven-
i ence, speed, and aestheti cs. Wi thi n these consumer
constrai nts, a more effi ci ent car maybe preferabl e to
an i ncrease i n refi nery capaci ty i n order to reduce
capi tal and/or operati ng costs or because of i ts
envi ronmental benefi ts. Thus, i n addi ti on to engi -
neeri ng and economi cs, energy anal yses shoul d al so
consi der soci al , cul tural , and i nsti tuti onal factors.
Such factors are more readi l y i ncl uded i n a servi ces
framework than i n a conventi onal energy suppl y
anal ysi s.
The amount of energy consumed i n the mai n end
use markets-resi denti al and commerci al , i ndustry,
and transportati on-i s exami ned fi rst; then the
major servi ces provi ded by energy are exami ned
wi thi n each end-use market. I n the househol d sector,
the servi ces exami ned are cooki ng,
4
l i ghti ng, space
condi ti oni ng, and refri gerati on; i n i ndustry, process
heat and motor dri ve; i n agri cul ture, i rri gati on and
tracti on; and, fi nal l y, transportati on. These servi ces
are chosen on the basi s of thei r current or l i kel y
future l evel s of energy consumpti on or thei r soci al
and economi c i mpacts.
Wi thi n thi s servi ces framework, changes i n en-
ergy use are traced from tradi ti onal rural areas to
thei r modern urban counterparts. The progressi on
from the tradi ti onal rural to the modern urban
i l l umi nates wel l the wi de range of technol ogi es now
bei ng used i n the devel opi ng countri es and the
dynami cs of how energy use can be expected to
change i n the future.
Energy use i n tradi ti onal rural vi l l ages refl ects a
much di fferent set of consi derati ons from that of the
modern urban economy. Fi rst, tradi ti onal energy use
i s part of a compl ex and i nterdependent bi ol ogi cal
system, rather than bei ng based on fossi l fuel s. The
bi omass that i s used for fuel i s part of a system that
provi des food for humans, fodder for ani mal s,
constructi on materi al s, fi ber for ropes, and even
tradi ti onal medi ci nes. Si mi l arl y, the bul l ock that
pul l s a pl ow al so provi des mi l k, meat, l eather, and
dung for ferti l i zer or fuel .
Second, peopl e i n tradi ti onal economi es careful l y
assess thei r choi ces and make compl ex tradeoffs
IJay~t Sathaye, Andre Ghirar@ and Lee Schipper, ~ergy D~din Developing Countries: A Sectoral Analysis of Recent Trends, Annuu2
Rewew of Energy, vol. 12, 1987, p. 253.
T.S. Environmental Protection Agency, Policy Options for Stabilizing Global Climate, vol. 2, February 1989, p. VII-30. Rapidly Char@ng
World Scenario.
% some cases, particularly in the industriahxd countries, wood might be burned in a fweplace simply for aesthetic reasons.
dWa@rh=@ is ~ ~ ~w~nt residen~mmm~~ energy servim tit is often S- to coo- b t- of the tdlrlologies wed. It k nO~
however, explicitly considered here.
-47
between the numerous pressures they face i n day-to-
day survi val , at a l evel sel dom seen i n the modern
economy. Gatheri ng fuel , for exampl e, i s not free: i t
costs ti me and personal energy that must be bal anced
agai nst al l the other demands that one faces,
parti cul arl y duri ng the agri cul tural season when
l abor demand i s at i ts peak. There are al so compl ex
tradeoffs i nvol ved i n gai ni ng access to fuel s on
common l ands or on pri vatel y owned l and.
Thi rd, al though peopl e i n rural areas may use
energy i neffi ci entl y i n compari son to what i s possi -
bl e wi th modern commerci al technol ogi es, they use
energy rather effi ci entl y and wi sel y gi ven the
constrai nts on thei r resources, technol ogy, and
capi tal s They have l i ttl e choi ce i n thi s i f they are to
survi ve on thei r meager resources. Rather than
maxi mi zi ng producti on, as i s done i n modern
i ndustri al soci ety, tradi ti onal peopl es focus on
mi ni mi zi ng ri sk i n the face of the vagari es of drought
and other natural di sasters.
The effi ci ency and producti vi ty of tradi ti onal
energy technol ogi es i n devel opi ng countri es can be
si gni fi cantl y i mproved. To do so effecti vel y, how-
ever, wi l l requi re an understandi ng of the compl ex
l i nkages of vi l l age l i fe. I n general , vi l l age popul a-
ti ons operate rati onal l y wi thi n thei r framework;
6
change then requi res that the framework be changed
through the i ntroducti on of external i nputs
fi nanci al , manageri al , materi al , and techni cal . The
l ack of success of many devel opment programs can
be attri buted i n part to a fai l ure to recogni ze the
rati onal i ty of rural l i festyl es and the need to address
the over al l fr amewor k i n whi ch vi l l ager s oper ate.
For the devel opi ng countri es as a whol e, the
resi denti al /commerci al and i ndustri al sectors consti -
tute the l argest end use energy markets, together
accounti ng for 85 percent of the energy used by fi nal
consumers when tradi ti onal fuel s are i ncl uded.
Transportati on accounts for the remai ni ng 15 per-
cent. There are, however, consi derabl e di fferences
among devel opi ng nati ons.
7
The resi denti al /com-
mer ci al sector accounts for a par ti cul ar l y hi gh shar e
of energy use i n Afri can countri es (mostl y i n the
form of bi omass fuel s for cooki ng), whi l e i ndustrys
share i s qui te l ow. Transportati on accounts for an
excepti onal l y hi gh share of the total i n Lati n
Ameri ca, whereas i ts share i n I ndi a and Chi na i s l ow.
Tabl es 3-1,3-2, and 3-3 provi de sectoral and energy
servi ce breakdowns for the devel opi ng countri es;
fi gure 3-1 shows per-capi ta energy use i n rural
househol ds as determi ned by vi l l age surveys i n
Afri ca, Asi a, and Lati n Ameri ca. Resi denti al cook-
i ng and i ndustri al process heat account for al most
two-thi rds of al l the energy used i n the devel opi ng
worl d. About 40 percent of al l energy consumed i n
provi di ng these servi ces i n devel opi ng countri es, or
wel l over a quarter of the total energy consumed i n
devel opi ng countri es, i s used i n I ndi a and Chi na.
Cooki ng i s the si ngl e l argest energy use i n many
devel opi ng countri es. There i s a wel l -establ i shed
transi ti on i n cooki ng fuel s associ ated wi th hi gher
i ncomes, i mproved suppl y avai l abi l i ty, and urbani -
zati on. I n rural areas, and i n poor urban househol ds,
tradi ti onal fuel s (wood, crop wastes, and dung) are
used i n si mpl e stoves. I n more affl uent househol ds,
peopl e swi tch to modern stoves and cl ean, conven-
i ent fuel s such as kerosene, Li quefi ed Petrol eum
Gases (LPG), and el ectri ci ty. Because wood stoves
are rel ati vel y i neffi ci ent, househol ds that use kero-
sene or LPG can consume si gni fi cantl y l ess energy
for cooki ng than those usi ng wood and charcoal .
Li ghti ng technol ogi es fol l ow a si mi l ar technol og-
i cal progressi on, from candl es or l i ght from wood
frees i n some rural areas, to kerosene and butane
l amps, to el ectri ci ty, whi ch i s a hi ghl y pri zed energy
servi ce. El ectri ci ty use for l i ghti ng ri ses rapi dl y wi th
househol d i ncome.
Rel ati vel y l i ttl e energy i s used for resi denti al
space cool i ng i n devel opi ng countri es. Space cool -
i ng i s becomi ng si gni fi cant i n commerci al and
government bui l di ngs, however, and energy use for
space cool i ng i s l i kel y to grow rapi dl y i n the future.
sNotable examples of studies on this topic include: N.H. Ravindranath et al.,
An Indian Village Agricultural Ecosystem-Case Study of Ungra
Village, Part I: Main Observations,Biomass, vol. 1, No. 1, September 1981, pp. 61-76; Amulya Kumar N. Reddy, An Indian Village Agricultural
lkosyste~e Study of Ungra Village, Part II: Discussion Biomuss, vol. 1, No. 1, September 1981, pp. 77-88; M.B. Coughenour et al.,
EnergyExtraction and Use in a Nomadic Pastoral Ecosystew Science, vol. 230, No. 4726, Nov. 8, 1985, pp. 619-625.
Ibid.
%nergyuse at the village level is fairly similar in both quantity and source (biomass), and in application (cooking, subsistence agriculture) throughout
the world. Energy use by the economically well off is also reasombly similar throughout the industrial countries as well as among the urban elite in
developing countries. The large differences in energy use between countries are due primarily to the relative numbers of villagers and economictdly well
off inthepopulatiow the form and quantity of energy use by those who are making the transition between these two extremes; and the development path
being followed.
Chapter 3--Energy Services in Developing Countries q 49
Table 3-lTotal Delivered Energy by Sector, in Selected Regions of the World, 1985 (Exajoules)
Residential/commercial Industry Transport Total
Commercial Traditional Commercial Traditional Commercial Traditional Commercial Traditional Total
Region fuels fuels
a
fuels fuels
a
fuels fuels
a
fuels fuels
a
energy
Africa . . . . . . . . . . . . . . 1.0 4.0 2.0 0.2 1.5 4.4 4.1 8.5
Latin America . . . . . . . 2.3 2.6 4.1 0.8 3.8 10.1 3.4 13.5
India and China . . . . . 7.3 4.7 13.0 0.2 2.0 22.2 4.8 27.1
Other Asia . . . . . . . . . . 1.9 3.2 4.0 0.4 1.9 7.8 3.6 11.3
United States . . . . . . . 16.8 16.4 18.6 51.8 51.8
-Not available or not applicable.
a these estimates of traditional fuels are lower than those generally observed in field studies. See figure 3-1, app. 3-A, and ch. 4.
b this is delivered energy and does not include conversion losses.
NOTES: 1 exajoule (10
18
Joules) equals 0.9478 Quads. The residential and commercial sector also includes others (e.g., public services, etc.) that do not fit
in industry or transport. Traditional fuels such as wood are included under commercial fuels for the United States. These figures do not include
conversion losses (from fuel to electricity, in refineries, etc).
SOURCE: International Energy Agency (lE/l), Wbt# Energy Sfatisfics and 8a/ances 1971-1987 (Paris: OECD, 1989); IEA, Energy 6a/ances of OECD
Countries f970-1985(Paris: OECD, 1987); and IEA, Energy 6a/ances of Developing Counties 1971-1982 (Paris: OECD, 1984).
Table 3-2Delivered Energy Per Capita by Sector in Selected Regions, 1985
(gigajoules) (Includes traditional fuels)
Region Residential/commercial Industry Transport Total
Africa . . . . . . . . . . . . . . . . . 11.8 5.2 3.5 20.5
Latin America . . . . . . . . . . 12.7 12.5 9.7 34.9
India and China . . . . . . . . 6.7 7.3 1.1 15.1
Other Asia . . . . . . . . . . . . . 7.2 6.2 2.7 16.1
United States . . . . . . . . . . 69.8 68.5 77.5 215.8
NOTE: These estimates do not include conversion losses in the energy sector and underestimate the quantity of
traditional fuels used compared to that observed in field studies. See app. 3-A for better estimates of traditional
fuel use and for sectoral energy use including conversion losses.
SOURCE: Derived from table 3-1.
Table 3-3-Per Capita Energy Use by Service in Selected Countries (Gigajoules)
Brazil China India Kenya Taiwan U.S.A.
Residential . . . . . . . . . . . . . . . . . . 6.2 11.7 5.5 16.9 8.9 64.9
cooking . . . . . . . . . . . . . . . . . . . 5.3 8.5 5.0 16.4 4.7 3.5
lighting . . . . . . . . . . . . . . . . . . . 0.3 0.4 0.5 0.5 0.7
appliances . . . . . . . . . . . . . . . . 0.6
0.05
3.1 13.0
a
Commercial . . . . . . . . . . . . . . . . . 1.5 0.7 0.26 0.4 4.2 45.2
cooling . . . . . . . . . . . . . . . . . . . 0.4
0.13 0.24 1.9

lighting . . . . . . . . . . . . . . . . . . .
0.5
0.05 0.16 0.8 7.2

appliances . . . . . . . . . . . . . . . . 0.6
0.07
1.5
.
Industrial . . . . . . . . . . . . . . . . . . . . 19.4 13.8 4.1 4.8 39.2 94.1
process heat . . . . . . . . . . . . . . 17.5 10.2 2.7
55.8
motor drive . . . . . . . . . . . . . . .
1.6 3.6 1.3 20.4

lighting . . . . . . . . . . . . . . . . . . .

0.1
0.05

Transport . . . . . . . . . . . . . . . . . . . 13.3 1.2 1.3 2.7 11.5 80.8
road . . . . . . . . . . . . . . . . . . . . . . 12.0 0.2 0.8 1.8 10.1 66.7
rail . . . . . . . . . . . . . . . . . . . . . . 0.2 0.7 0.4 0.2 0.1 2.0
air . . . . . . . . . . . . . . . . . . . . . . . 0.7
0.1 0.7 0.7 11.3

Agriculture . . . . . . . . . . . . . . . . . . 2.1 1.8 0.6 0.5 2.6 2.5
Total . . . . . . . . . . . . . . . . . . . 43.4 27.0 11.7 25.6 67.7 288.0
Not available or not applicable.
a
This is the combined total for applicances and lighting.
NOTE: These estimates include the upstream conversion losses in the energy sector, such as the loss in going from fuel to electricity or crude to refined
petroleum products. This in in contrast to tables 3-1 and 3-2 where energy sector conversion losses were not included.
SOURCE: Adapted from app. 3-A, tables 1 through 6.
Figure 3-lPer-Capita Energy Use in Village
Households in Developing Countries
Per-capita energy use, GJ/year
3 5 ~ - - - -
3 0 -
2 5 -
2 0 -
15
10
5
I
o -
1 Ii I l l I l l I l l l d l l l l
Asia Africa Latin
America
This figure shows per-capita use of biomass fuels as found in
household energy surveys in Asia, Africa, and Latin America. The
observed energy use depends on such factors as fuel availability,
climate, diet, income, and other factors. The generally higher
levels of energy use in Africa and Latin America reflect, in large
part, the greater availability and accessibility of wood and other
biomass fuels. The sporadic peaks in energy use shown in the
figure are typically for villages in colder, more mountainous
regions.
SOURCES: D.O. Hall, G.W. Barnard, and P.A. Moss, Biomass for f%erav
in the Developing Countries (Oxford: Pergamon Press, 19~~,
pp. 212; World Bank, Bolivia: Issues and Options in the
Energy Sector, UNDP/WB Energy Sector Assessment Pro-
gram, Rpt. 4213-60, April 1983; J.S. Singh, U. Pandey, and
A.K. Tiwti, Man and Forests: A Central Himalayan Case
Study, AMBIO, VOI. 12, No. 2, 1984, pp. 80-87; Issoufou
Boureimaand Gilles De Chambre, Rapport sur/Eva/uation du
Programme Foyers Ame/lores (Niamey, Niger: Association
des Femmes du Niger and Church World Service, November
1982).
Al so, el ectri c appl i ances are qui ckl y penetrati ng the
resi denti al sector. Many of these ai r condi ti oners and
appl i ances, notabl y refri gerators, have l ow effi ci en-
ci es. These end uses are havi ng strong i mpacts on the
el ectri c power i nfrastructure.
Many commerci al and i ndustri al processes re-
qui re process heat, rangi ng from the l ow-tem-
perature heat provi ded by bi omass used to dry food
i n cottage i ndustri es to the hi gh-temperature proc-
esses used i n the l arge-scal e steel and cement
i ndustri es. Wi th some excepti ons, the effi ci enci es of
these processes are typi cal l y much l ower than those
found i n i ndustri al i zed countri es.
Much of the popul ati on i n devel opi ng countri es
depend for thei r mechani cal work i n both i ndustry
and agri cul ture on human or ani mal muscl e, wi th
l ow effi ci enci es and power outputs that seri ousl y
l i mi t producti vi ty. The effi ci enci es of modern di esel
and el ectri c motors are si gni fi cantl y l ower i n devel -
opi ng countri es than i n the i ndustri al i zed countri es
as wel l .
As i n other sectors, there i s a transi ti on i n
transportati on technol ogi es. Wal ki ng and use of
domesti cated ani mal s are the domi nant transport
technol ogi es i n poorer and rural areas. The next step
up i s bi cycl es, and then the i nternal combusti on
engi ne. Transport servi ces i n the devel opi ng worl d,
as i n the i ndustri al worl d, are based l argel y on
hi ghways. I n the devel opi ng countri es, however,
frei ght rather than passenger traffi c i s the most
i mportant transport acti vi ty i n terms of energy
consumpti on.
The Residential/Commercial Sector
8
Energy use i n the resi denti al /commerci al sector of
devel opi ng countri es typi cal l y accounts for about 30
percent of commerci al energy use and two-thi rds or
more of tradi ti onal fuel use (see app. 3-A). Cooki ng
i s by far the l argest use of fuel i n rural areas; i n urban
and more devel oped areas, l i ghti ng and appl i ances
(refri gerators and el ectri c fans, for exampl e) are al so
l arge energy users. Ai r condi ti oni ng i s l i kel y to
become i mportant i n the future i n resi dences and i s
al ready wi del y used i n commerci al , i nsti tuti onal ,
and government bui l di ngs i n devel opi ng countri es.
The average energy effi ci ency of the most com-
mon cooki ng, l i ghti ng, and appl i ance technol ogi es
i n use i n devel opi ng countri es today can be i m-
proved dramati cal l y,
9
but usual l y at a si gni fi cant
addi ti onal capi tal cost to the consumer. Neverthe-
l ess, the advantages of these more modern technol -
ogi es--conveni ence, comfort, effecti veness-are i n-
centi ve enough for consumers to make the i nvest-
ment where the technol ogi es and the necessary fuel
suppl i es are avai l abl e, affordabl e,
10
and reasonabl y
r el i abl e.
%this analysis, the residential/cmmnercial sector includes other energy uses such as public buildings not included in the industrial and transportation
sectors.
% can be accomplished by changing both the mix of technology (e.g., shifting users from Iow-efllciency wood stoves to high-efficiency LPG
stoves) and by improving the individual technologies themselves (e.g., moving toward higher efilciency refrigerators).
IOApp~priate f~cial mechatims y h etid
Table 3-4-Principal Cooking Fuels Used by the World Population, 1976
Percent of people using fuel
Fossil Dung and
Region energy
a
Fuelwood crop waste
Africa South of Sahara . . . . . . . . . . . . . . . . . . . . . . .
India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Rest of South Asia . . . . . . . . . . . . . . . . . . . . . . . . .
East Asia, developing Pacific . . . . . . . . . . . . . . . . .
Asia centrally planned economies . . . . . . . . . . . . .
Middle East, North Africa. . . . . . . . . . . . . . . . . . . . .
Latin America and Caribbean . . . . . . . . . . . . . . . . .
North America, OECD Pacific. . . . . . . . . . . . . . . . .
Western Europe . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Europe, centrally planned economies . . . . . . . . . .
Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
10
12
36
22
53
71
100
100
100
47
63
47
46
41
51
17
26
0
0
0
33
27
43
42
23
27
30
3
0
0
0
20
a Includes electric cooking.
SOURCE: Adapted from David Hughart, Prospects fbr Traditbua/ and fVon-Convenfiorm/ Energy Soumes in
Developing Countries, Worfd Bank staff working paper No. 346,132 pp., July 1979.
Cooking
ll
The most i mportant si ngl e energy servi ce i n many
devel opi ng countri es i s cooki ng. I n rural areas of
devel opi ng countri es, tradi ti onal fuel s--wood, crop
wastes, and dung-are used for cooki ng; i n many
urban areas, charcoal i s al so used. More than hal f of
the worl ds peopl e depend on these crude fuel s for
thei r cooki ng and other energy needs (see tabl e
3-4).
12
Hi gher i ncomes and rel i abl e fuel suppl i es
enabl e peopl e to swi tch to modern stoves and cl ean
fuel s such as kerosene, LPG, and el ectri ci ty.
Tradi ti onal Fuel s
Tradi ti onal fuel s are predomi nant i n rural areas
because they can be gathered at no fi nanci al cost and
used i n very si mpl e stovesas si mpl e as an open
free. At the nati onal l evel , the use of bi omass for fuel
reduces expensi ve energy i mports. These are sub-
stanti al benefi ts.
Use of tradi ti onal fuel s al so exacts substanti al
costs. Large amounts of l abor are expended to gather
these fuel s i n rural areas, and a si gni fi cant porti on of
househol d i ncome i s spent for them i n poor urban
areas (see ch. 2). Cooki ng wi th tradi ti onal fuel s i s
awkward and ti me-consumi n g. Unl i ke modern gas
or el ectri c stoves, stoves that use tradi ti onal bi omass
fuel s must be constantl y tended to mai ntai n an
adequate fl ame. Thi s demands a l arge share of
womens ti me i n devel opi ng countri es-averagi ng
perhaps 3 to 5 hours per day
13
and i nterferes wi th
other acti vi ti es.
Cooki ng wi th tradi ti onal fuel s i s al so usual l y
unpl easant and unheal thy due to the l arge amount of
noxi ous smoke emi tted (see tabl e 3-5). Measure-
ments of i ndoor concentrati ons i n homes i n devel op-
i ng countri es have found l evel s of carbon monoxi de,
parti cul ate, and hydrocarbons 10 to 100 ti mes
hi gher than Worl d Heal th Organi zati on standards.
Cooks can be exposed to as much or more carbon
monoxi de, formal dehyde, benzo(a)pyrene, and other
toxi ns and carci nogens as heavy ci garette smokers.
14
Smoke from cooki ng stoves i s therefore thought to
be a si gni fi cant factor i n i l l -heal th i n devel opi ng
llAItiou@ tie ~scussion here focuses on household cooking, the same considerations apply to commercial and hstilut.iOMl set-s.
12Hea~g water for ba~ ~d cl~~ and kiling water for &inking are implicitly included in the discussion here, u the teehOIO@es uSed are
often the same for the lower and middle income groups in developing countries, and separation of energy use for these purposes is difikult,
13Ric~d Morse et~.,
OrganiAngCurrentInformation forRuralEnergyand Development Planning, M. NurulIslamj Richard Morse, andM. Hadi
Soesastro (eds.), Rural Energy to MeetDevelopment Neea%: Asian Village Approaches (Boulder, CO: Westview Press, 1984), table 7, p. 498.
1a~ R. Smi@ BioM~~ Fuels, Air Pollution, a~Health: A Global Review (New York NY: plenum pr~s, 1987)
Table 3-5-Typical Air Pollution Emissions From Various Cooking Fuels
Efficiency
Grams per gigajoule of delivered energy
a
Fuel (percent)
TSP s o, NOx HC GO
Wood (tropical) . . . . . . . . . . . . . . . 15 3,800 250 300 3,200 34,000
Cow dung (Hawaiian) . . . . . . . . . 15 10,000 3,200
.
44,000
Coal (Indian) . . . . . . . . . . . . . . . . .
20 280 2,200 460 2,200 27,000

Coconut husk . . . . . . . . . . . . . . . . 15 17,000

54,000
Natural gas . . . . . . . . . . . . . . . . . . 60 0.7
13 7 330
a TSP, total Suspended particulates; SO
2
, sulfur dioxide; NO
X
, nitrogen oxides; HC, hydrocarbons; CO, carbon monoxide.
SOURCE: Adapted from Kirk R. Smith, Biomass Fuels, Air Po//ution, and Hea/th:A G/oba/ Review (New York, NY: Plenum Press, 1987).
countri es. The di seases i mpl i cated i ncl ude severe
eye i rri tati on, respi ratory di seases, and cancer.
15
Fi nal l y, al though the expansi on of agri cul tural
and grazi ng l ands and commerci al l oggi ng are the
most i mportant causes of deforestati on gl obal l y, the
use of wood for fuel may al so contri bute to deforest-
ati on i n some l ocal areas, parti cul arl y where the
popul ati on densi ty i s hi gh and the cl i mate i s dry such
as the West Afri can Sahel (see ch. 5).
The Transi ti on to Modern Stoves
and Cl ean Fuel s
Peopl e are general l y observed to make the transi -
ti on to modern, effi ci ent stoves and cl ean fuel s as
soon as they are avai l abl e and affordabl e (see fi gure
3-2).
16
These technol ogi es are preferred for thei r
conveni ence, comfort, cl eanl i ness, ease of opera-
ti on, speed, and other attri butes.
There i s a natural progressi on i n effi ci ency, cost,
and performance as consumers shi ft from wood
stoves to charcoal , kerosene, LPG or gas, and
el ectri c stoves (see fi gure 3-3). I mproved wood and
charcoal stoves have al so begun to fi l l a potenti al l y
i mportant ni che between tradi ti onal wood or char-
coal stoves and modern kerosene or gas stoves.
Cul tural factors are often ci ted as a barri er to the
adopti on of i mproved bi omass stoves and fuel s.
Figure 3-2-Choice of Cooking Fuel by Income for Five
Medium-Sized Towns in Kenya
Percent of income group using fuel
100 r

1
i
80,
>
i .- -
60
[
\
40
/
/
r
20
o L . . - L ( - ~
Low Middle High
Income group
- - Wood + Ch a r c o a l + Ke r o s e n e
+
Gas
-
E l e c t r i c i t y
Many households use more than one fuel depending on the
particular food cooked and the supply and cost of fuel. Note the
shift in fuel choice from wood to charcoal and kerosene, and then
from charcoal and kerosene to gas and electricity.
SOURCE: John Soussan, Fuel Transitions Within Households, Discus-
sion paper No. 35, Walter Elkan et al. (eds.), Transtions
Between TracWonaland Commeraal Energy in the Third kldkl
(Guiidford, Surrey, United Kingdom: Surrey Energy Economics
Center, University of Surrey, January 1987).
Al though cul tural factors may pl ay a rol e i n choi ces
of stoves or fuel s, i t i s hardl y a domi nant one, as
evi denced by the wi de vari ety of stoves and fuel s
that have al ready been adopted across the ful l range
of cl ass, cul tural , and i ncome groups i n devel opi ng
countri es. More typi cal l y, the reason that vari ous
stoves have not been adopted by targeted groups i n
IsAlthough the limited data available linkins human
exposure to the smoke from wood fwes to lung cancer is still ambiguous (but may indicate
anomalously low cancer rates), there is now evidence of excess lung cancer among cooks using certain types of coal irt China. Overall, the World Health
Organization now cites respiratory disease fmm all causes as the leading cause of mortality in developing countries. See Kirk R. Smi@ PAHand the
Household Cook in Developing Countries: The Lung Cancer Anomaly, paper presented at the Symposium on Polynuclear Aromatic Hydrocarbons
in the Workplace, Intermtional Chemical Congress of Paci.t3c Basin Societies, Honolulu, HI, December 1984, to be published in M. Cooke and A.J.
Dennis (eds.), Polynuclear Aromutic Hydrocarbons: Formation, Metabolism and Measurement (Columbus, OH: Batelle Press); J.L. Mumford et al.,
Lung Cancer and Indoor Air Pollution inXuan Wei, chin% Science, vol. 235, Jan. 9, 1987, pp. 217-220; H.W. de Koning, K.R. SmitlL and J.M. LasL
Biomass Fuel Combustion and Heal@ Bulletin of the World Health Organization (EFP/84.64).
Icrhis ~msition is complex ~d not yet well unde~tood. Factors that affect ahouseholds shift to modern stoves ~d fuels include household income
and fuel-producing assets (land, trees, animals, etc.); reliability of access to modem fuels; relative cost of traditional and modern fuels and stoves; level
of education of the head of household; cooking habits; division of labor and control of fwces within the household; and the relative performance of
the stoves and fuels available.
Figure 3-3-Representative Efficiencies and Direct Capital Costs for Various Stoves
Stove efficiency (%)
Capital cost ($)
.
7 0
r
30
20
10
8 0
6 0
/
I
/ /
,
I 4 0
20
0
I I I I
o
Agricultural Traditional Improved Kerosene Electric
waste in charcoal charcoal
pressure hot plate
traditional q stoves q stoves q stoves
q tovea
Animal Tradition al Improved Kerosene
dung In
LPG
wood WOO d W i c k
traditional
stoves
stoves stoves
q tovea
q stoves
The range of performance both in the laboratory and in the field is much larger than that suggested by
this figure and is affected by such factors as the size of the stove and pot, the climate (wind), the quality
of the fuel used, the care with which the stove is operated, the type of cooking done, and many other
factors. The type of material that the pot is made of is also a significant factor: aluminum pots are almost
twice as efficient as traditional clay pots due to their better conduction of heat. Although the efficiency
of improved charcoal stoves is shown as slightly higher than improved wood stoves-the case today for
the simplest uninsulated metal woodstoves--the potential performance of wood stoves is higher than
that for charcoal stoves.
SOURCE: Samuel F. Baldwin, ~omass Stoves: Engineering Design, Development, and Dissemination (Arlington, VA:
VITA, 1986); and OTA estimates.
the devel opi ng countri es i s that they si mpl y have not
worked wel l .
17
The transi ti on to modern stoves and fuel s i s often
sharpl y constrai ned due to thei r hi gher capi tal costs
(fi gure 3-3) and uncertai nty i n the suppl y of fuel . I n
Col ombo, Sri Lanka, for exampl e, the cost of
converti ng to LPG i n 1983 was equi val ent to 1
months i ncome for 70 percent of the popul ati on and
5 months i ncome for the poorest 20 percent.
18
Yet
cooki ng wi th gas can be the l owest cost al ternati ve
when both capi tal and operati ng costs are i ncl uded.
I n Rai pur, I ndi a, the cost of cooki ng wi th LPG i s l ess
than that for wood for househol d di scount rates of 30
percent or l ess; yet many househol ds conti nue to use
wood, presumabl y because effecti ve househol d di s-
count r ates ar e hi gher
19
(the capi tal cost of gas stoves
was ci ted as a major reason for the fai l ure to swi tch
from wood to LPG for cooki ng).
Because of the hi gh cost of LPG cooki ng, char-
coal and kerosene are wi del y used as an i ntermedi ate
step i n the transi ti on from wood to gas stoves.
Charcoal i s very popul ar i n some urban areas. For
exampl e, i t i s the fuel of choi ce i n urban Kenya (see
fi gure 3-2) and Senegal -whi ch have a tradi ti on of
charcoal producti on and use remai ni ng from the
hi stori cal Saharan trade caravans.
20
Consumer s
ITSamuel F. Baldw@ BioMss Stoves: Engineenng Dewgn, Development, and Dissermnation (Adingtoq V!: VITA, 1986); Sam B~dti et ~.,
Improved Woodburning Cookstoves: Signs of Success, AMBZO, vol. 14, No. 4-5, 1985.
lg&r~d hack Household Energy in South An-a (New YorlG NY: Elsevier Applied Science, 1987).
lgJ. Dunkerley et al., Consumption of Fuelwood and Other Household Cooking Fuels in Indian Cities, Energy Policy, January/February 1990, pp.
92-99. Discount rates are a measure of the time value that households place on their available cash income.
World BardG Energy I@wtrnen4 Review of Household Energy Issues in Afiicz draft report, May 1987, p. 3, p. 1.6.
prefer charcoal to wood because i t gi ves off l ess
smoke,
21
bl ackens pots l ess, requi res l i ttl e tendi ng of
the fi re, and i n some areas costs l ess.
22
At the
nati onal l evel , however, cooki ng wi th charcoal
consumes far more forest resources than cooki ng
di rectl y wi th wood, due to the l ow energy effi ci ency
of converti ng wood to charcoal -typi cal l y just 40 to
60 percent and often much l ower.
23
Kerosene i s usual l y the next step up i n the pro-
gressi on of cooki ng fuel s. I n many areas, kerosene
pri ces-often subsi di zed or freed by the govern-
mentform a reasonabl y effecti ve cap on the pri ce
of wood and charcoal .
24
Consumers swi tch between
these fuel s accordi ng to pri ce and avai l abi l i ty.
LPG or natural gas i s often the fi nal step i n the
progressi on i n cooki ng fuel s. LPG i s wi del y used by
hi gher i ncome groups i n many urban areas, and
natural gas i s wi del y used where i t i s avai l abl e. I n
Dhaka, Bangl adesh, for exampl e, over 50 percent of
the urban popul ati on use natural gas; l ess than 10
percent use kerosene; and none use charcoal .
25
I n
some cases, el ectri ci ty i s al so used for cooki ng by
the hi ghest i ncome groups.
As househol ds make the transi ti on from wood to
modern fuel s, overal l energy use for cooki ng can
vary dramati cal l y, dependi ng on the choi ce of
technol ogy and the si tuati on i n whi ch i t i s used.
Total househol d energy use for cooki ng wi th kero-
sene (see fi gure 3-4) or wi th LPG can be si gni fi -
cantl y l ess
26
than that for wood or charcoal , due to
the hi gher effi ci ency of kerosene and gas stoves.
27
Total househol d energy use for cooki ng wi th kero-
sene or LPG i s al so si gni fi cantl y l ess than for
cooki ng wi th charcoal or (non-hydro) el ectri ci ty,
due to the l ow conversi on effi ci ency of wood to
charcoal and of fuel to el ectri ci ty.
28
The transi ti on to modern stoves and fuel s thus
offers users many benefi ts-reduced ti me, l abor,
and possi bl y fuel use for cooki ng, and reduced l ocal
ai r pol l uti on. Means of l oweri ng capi tal and operat-
i ng costs and ensuri ng the rel i abi l i ty of suppl y are
needed i f the poor are to gai n access to these cl ean,
hi gh-effi ci ency technol ogi es. At the nati onal l evel ,
the transi ti on to modern stoves and fuel s coul d
i mprove the l ocal envi ronment
29
and si gni fi cantl y
reduce bi omass energy consumpti on for cooki ng; to
real i ze these benefi ts, however, coul d i mpose a
substanti al fi nanci al burden on poor nati ons.
A l arge-scal e transi ti on to LPG woul d requi re a
si gni fi cant i nvestment i n both capi tal equi pment and
ongoi ng fuel costs. opti mi sti cal l y assumi ng that the
cost of LPG systems woul d average $10 per capi ta,
the i nvestment woul d be roughl y 3.5 percent of GNP
and 20 percent of the val ue added i n manufacturi ng
zlck@~ stov= Ca however, give off hi@ levels of carbon monoxide-a serious health hazard in inad~uately ventilated kitchens--but Ws dws
not cause as much obvious discomfort to the user as the smoke from a wood fire.
22Douglas F. Barnes, World Bar& Household Energy Unit Industry and Energy Departmen4 Undemamiing Fuelwood prices ~ Developing
Nations, Oct. 31, 1989, table 1. Conversion to dollazs per unit of energy was done using 30 MJ/kg for charcoal, and using 700 kg per cubic meter
multiplied by 16 MJ/kg for wood with typical moisture contents observed in the market.
~~eenergyefflciencyof the conversion process is variously given as 15 percerlt in ~
a, 24percent in Kenya withan additional loss of 5 percent
of the charcoal itself during distribution% 29 percent in Senegal and Ethiopia, and over 50 percent in Brazil with brick kilns. Advanced retorts are claimed
to be capable of achieving 72 percent energy efficiencies in converting wood to charcoal if there is complete recovery of all the gaseous byproducts.
See E. Uhi@ Preliminary Charcoal Survey in Ethiopia, U.N. Economic Commission for Afriw FAO Forest Industries Advisory for Africa, Dec.
M75-1 122, 1975,30 pp.; M.J. and M.L. Luhanga, Energy Demand Structures in Rural Tanzurua, Department of Electrical Engineering, University of
Dar-es-Salaam, Tanzzmiq 1984; Phil OKeefe, Paul Rask@ and Steve Bemow, Energy and Development in Kenya: Opportunities and Constraints
(Sweden: Beijen Institute, 1984); G.E. KarclL Carbonization: Final Technical Report of Forest Energy Specialist, UNFAO, SEN/78/002, 1980.; T.S.
Wood, Report on Domestic Energy Usefor Cooking (Energy Assessment Missio~ Ethiopia) (Washingto~DC: World Ba~ 1983), p. 33; FLORAS~
Man-h-fade Forestsfor Wood and Charcoal in Brazil (Mi_nas Gerais, Brazil: Florestal Acesita, S.A., Belo Horizonte, October 1983), p. 53.
24Doug~ F, Barnes,
Understanding Fuelwood Prices in Developing Nations, op. cit., footnote 22.
~MJ. ~or, Fuel ~kets in UrbanB~ladm~ WorldDevelopment, VO1. 14, No. 7, pp. *65-872.
XNot ~ tie potenti~ efficiencyg~ of LPG may bere~dby the vex-y poor. For example, one-third to One-halfof the poor fi~lside s~nty tow
of Rio de Janeiro own just one LPG bottle. To avoid the risk of running out of gas and having no substitute, many households exchange their gas bottles
before they are completely empty. See Alfredo Behrens, Household Energy Consumption in Rio De Janeiro Shanty Towns (Rio de Janerio, Brazil:
Colegio da America Latina, 1985).
zT~practice, howevw, the ~v~s~~ LPG are not quite as large as would be expected from the higher efficiency ~dbetter con~l of these stoves.
This may be due, in part, to less precise control of the stove; to taking advantage of greater useful energy; and other factors. See Kevin B. Fitzgerald,
Douglas Barnes, and Gordon McGranaha~ Interl%el Substitution and Changes in the Way Households Use Encxgy: The Case of Cooking and Lighting
Behavior in Urban Java, U.N. Working Paper on Interfuel Substitution Analysis, June 13, 1990.
2SO~er facton tit ~wthouwhold enmm use for cooking ~clude tie size of ~ehousehold, tie die~ ad tie amount of processed or prepaZedfOOdS
eaten.
2!qt ~gh~ however, incmme global carbon dioxide emissiom.
Chapter 3-Energy Services in Developing Countries .55
Figure 3-4Direct Energy Use for Cooking in West
Java, Indonesia
Per-capita energy use, MJ/day
3 5 ~ -
I
5
I
o L T - - - 1 !
Low Lower middle Upper middle High
Income group

W o o d
~ Kerosene
This figure compares energy use in households using only wood
with that in households using only kerosene within the same
income class. Households using kerosene consume roughly half
as much energy as households using wood.
SOURCE: M. Hadi Soesastro, Policy Analysis of Rural Household Energy
Needs in West Java, M. Nurul Islam, Richard Morse, and M.
Hadi Soesastro (eds.), f?ura/ Energy to Meet Development
Me&: Asian Village Approaches (Boulder, CO: Westview
Press, 1984).
for the nearl y three bi l l i on peopl e i n the l owest
i ncome countri es.
30
The LPG used
31
woul d be
equi val ent to one-fourth of the total commerci al
energy consumpti on today by these countri es and
woul d be a si gni fi cant fracti on of thei r export
earni ngs.
32
Si gni fi cant economi c growth i s needed i f
these costs are to be absorbed.
Costs woul d be even hi gher i f el ectri ci ty were
used for cooki ng. Di rect capi tal costs for el ectri c
burners typi cal l y approach $100 per househol d or
more. Moreover, at the nati onal l evel , the capi tal
cost of i nstal l i ng generati on, transmi ssi on, and
di stri buti on equi pment to power el ectri c burners i s
much greater, perhaps several thousand dol l ars per
househol d.
33
I f rel ati vel y few househol ds are usi ng
el ectri ci ty for cooking, these hi gh capi tal costs are
parti al l y offset by the numerous other uses for
Figure 3-5-Daily Load Profiles for Cooking Energy,
P o n d i c h e r r y , I n d i a , 1 9 8 0
Per-capita power demand for cooking, watts
1 , 2 0 0 -
1,000
/ n
800-
600-
400-
200-
0
,
~ J \
r T T
~
T
- r ~ r T -
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Hour of day
This figure illustrates the highly peaked power demand for cooking
energy as measured in a village survey.
SOURCE: C.L. Gupta, K. Usha Rae, and V.A. Vasudevaraju, Domestie
Energy Consumption in India (Pondicherry Region), Energy,
vol. 5, pp. 1213-1222.
el ectri c power throughout the day. I f a si gni fi cant
fracti on of househol ds swi tch to el ectri ci ty for
cooki ng, however, the hi ghl y peaked energy demand
for cooki ng (see fi gure 3-5) wi l l overwhel m other
basel oad appl i cati ons, and these costs must i ncreas-
i ngl y be assi gned to cooki ng al one.
Lighting
34
Li ghti ng accounts for onl y a smal l fracti on of total
nati onal energy use i n both devel opi ng and i ndus-
tri al countri es. I n Kenya, for exampl e, just 1.7 per-
cent of nati onal energy use i s for domesti c l i ghti ng
(app. 3-A). Li ghti ng does, however, account for a
si gni fi cant fracti on of total el ectri ci ty use, and the
el ectri ci ty sector i s very capi tal i ntensi ve (see ch. 4).
Despi te i ts rel ati vel y l ow energy use, l i ghti ng
meri ts parti cul ar attenti on as i t pl ays a very i mpor-
~orld Brmlq World Development Reporr, 1989 (New York NY: Oxford University Press, 1989), tables 1 and 6.
slAssuming aper-eapitapower rate for cooking with LPG systems of 100 watts. This is comparable to that seen in the United States and about twice
that seen in European countries. It is likely that people in developing counties would continue to eat less processed food, less restaurant food, and
probably more grains and so would continue to use somewhat more fuel than is used in households in the industrialized countries. Energy use rates for
household cooking in different countries are given in K. Krishna Prasad, Cooking Energy, workshop on end-use focused global energy strategy,
Princeton University, Princetoq NJ, Apr. 21-29, 1982.
szWorldB@ WorZdDeve/op~ntReport, 1989, op. cit., footnote 30, table 5. Kilograms of oil e@wdent hive been converted m energy at42 ~/k8.
33Ass@ a @powm de~d of 2 kw. A pe~power &ud i s assumed here rather than an average power as for the w Cue above, b-use
electric power systems cannot easily store power and must be able to meet peak demands.
~~cip~ sowes forthe~omtionin this ~tion are Robert vander Piss and A.B. de Graaf, World BaM A Comptisonof Lamps for Domestic
Lighting in Developing Countries,energy series paper No. 6, June 1988; and Robert van der Plas, World Bam Domestic Lighting, Energy Sector
Management and Assessments, Industry and Energy Department, working paper No. WPS 68, November 1988.
tant soci al rol e i n domesti c l i fe and i n commerce and
i ndustry, maki ng acti vi ti es possi bl e at ni ght or
where natural l i ghti ng i s i nadequate. As rural
i ncomes i ncrease, or as peopl e move to urban areas
and gai n greater access to modern fuel s and el ectri c-
i ty, l i ghti ng servi ces and the energy used to provi de
them i ncrease dramati cal l y.
Li ghti ng technol ogi es fol l ow a fai rl y cl ear tech-
nol ogi cal progressi on i n performance, effi ci ency,
and cost (see fi gure 3-6). Consumers choi ces of
l i ghti ng technol ogi es l argel y fol l ow the same pro-
gressi on as househol d i ncomes i ncrease and as
el ectri ci ty becomes avai l abl e.
I n tradi ti onal rural areas, peopl e are often l i mi ted
to the l i ght avai l abl e from wood fi res, frequentl y
obtai ned i n conjuncti on wi th cooki ng. Kerosene
wi ck l amps are usual l y the fi rst step up i n the
progressi on. These may be as si mpl e as a wi ck i n a
jar of kerosene, or as compl ex as a hurri cane l amp
wi th a gl ass chi mney .35 Gl ass chi mney l amps
general l y provi de more l i ght and at a hi gher effi -
ci ency than open wi ck l amps. Gl ass chi mney l amps
al so cost sl i ghtl y more-a few dol l ars-and use
somewhat more fuel . These addi ti onal costs can be
a substanti al bar r i er to thei r use i n r ur al ar eas. For
exampl e, a survey of si x vi l l ages i n Bangal ore, I ndi a,
found that three-fourths of the househol ds used
si mpl e open-wi ck l amps, and onl y one-fourth used
l amps wi th gl ass chi mneys.
36
The l i ght provi ded by wood fi res, candl es, or
kerosene wi ck l amps i s suffi ci ent to fi nd ones way,
but i s general l y i nadequate for tasks such as readi ng
or fi ne work. Usi ng two l amps doubl es the cost, but
does not come cl ose to provi di ng adequate l i ght to
work by. Thus, the poorest househol ds tend to use
just one l amp. Weal thi er househol ds may add an
addi ti onal l amp or two for other rooms i n the house
or move up to a kerosene mantl e l i ght; however, the
amount of kerosene used per househol d does not
general l y i ncrease i n proporti on wi th i ncome. As a
resul t, the amount of kerosene used for l i ghti ng i s
si mi l ar (wi thi n a factor of two or so) across di fferent
Figure 3-6-Light Output and Efficiency of Various
Lighting Technologies
Flux (lumens) Device efficiency (lumens/watt)
1,000
I
, , + 100
800
L
A
10
800
400
I
Wood Candle Wick Mantle lncandes- Floures-
Li ghti ng technol ogy
cent cent
.

Fl ux ( I umens)
+ Devi ce e f f i c i e n c y
Includes the candle, kerosene wick lamp, kerosene mantle lamp,
60-watt incandescent lamp, and 22-watt standard fluorescent
lamp. No value is given for a wood fire, as its light output depends
on size and other factors. The light output of candles and
kerosene lamps are similarly highly variable; the values listed are
representative. Only the efficiency of the device (plus ballast) itself
is considered. System efficiencies-induding refinery losses in
kerosene production and generation, transmission, and distribu-
tion Iosses for electricity--will be considered in a later report of this
OTA study.
SOURCES: Robert van der Plas, Wrid Bank, Domestic Lighting, Energy
sector Management and Assessments, Industry and Energy
Department, working paper No. WPS 68, November 1988.
Van der Plas cites the efficiency of electricity production as 30
percent, but this factor is apparently not taken into aooount in
the incandescent light effideney figure of 12 lm/W. See, for
example, Samuei Berman, Energy and Lighting, David
Hafemeister, Henry Kelly, and Barbara Lsvi (ads.), Energy
Soutces: Conservation and Renewable (New York, NY:
American Institute of Physics, 1985). Berman gives the output
of a 1OO-W incandescent as 1,600 lumens and a 50-W
fluorescent as 3,300 lumens. The efficiencies shown here are
slightly iowereorresponding to the lower, assumed wattage of
the light. See also Terry McGowan, Energy-Effiiient Light-
ing, Thomas B. Johansson, Birgit Bodlund, and Robert H.
Williams (eds.), E/ectbcify: Eftkient En#Jse and New Gener-
ation Ttinologie and Their Planning Impkations (Lund,
Sweden: Lund University Press, 1989).
i ncome groups and i n di fferent regi ons of the
worl d.
37
Despi te the drawbacks of kerosene wi ck l amps,
they are a predomi nant technol ogy i n poor rural and
urban areas. Al though thei r l i ght output i s l ow, the
capi tal and operati ng costs of kerosene wi ck l amps
are al so l ow (see fi gure 3-7). Further, kerosene can
be purchased i n smal l quanti ti es as fami l y fi nances
qslhe light given off by wick lsmps &pends on a host of factors, including size, condition of the wick (unraveled or uneven), ~d the ~Ount of soot
deposited on the glass chimuey (if present).
~AS~, Rur~ Energy Consumption Patterns: A Field Study, Bangalore, ~dti, 1981.
qT~r~d~achad ~cia GoWe% World B- Hou~hold Energy Handwk, t~hni~ paper, No. CT, 1987; sul i ~ si wati bau, Rura/Energy
In Fiji (Ottaw& Canada: IntemationaJ Development Research Center, 1981); Girja Sharan (cd.), Energy Use in Rura2 Gujurat (New Dethi: Oxford and
IBHPublishing Co., 1987). The Gujurat study found that one of the few variables affecting kerosene use was the number of rooms per household. Even
this, however, was a relatively weak relationship.
Figure 3-7--Costs of Various Lighting Technologies
Total cost to operate ($/hr)
~
0.08
1 r- l
New installation
0.06 -
0.04 -
0.02 -
. 1
Marginal cost
Wood Candle Wick Mantle Incandescent Flourescent
Lighting technology
Cost per light output ($/lumen)

0 0 0 E - 0 3
r
\
- -
- - - -
< \ New installations
1.000 -04
4
w
1. 000 - 05
I
\ . \
1
x.
Marginal cost
1 .000 -06 --r T -r -- -T
Wood Candle Wick Mantle Incandescent Fluorescent
Lighting technology
(A) Direct costs to the consumer of operating various lighting technologies per hour of service. (B) Direct costs to the consumer of various
lighting technologies per unit of light output. The costs shown include only cash expenditures; they do not include labor costs for maintaining
kerosene lamps, etc. The high value for electric lights shows the effect of applying all the grid connection charges to a single light
corresponding to the situation faced by the poor rural household that will initially use but one or two lights. The low value for electric lights
ignores the cost of grid connection charges, corresponding to the marginal cost of adding additional lights after being connected to the grid.
The assumed discount rate is 10 percent. In practice, individuals in both the developing and industrialized world tend to apply much higher
discount rates when making investment decisions in energy-conserving technologies. Rates observed in the United States are typically
in the range of 40 to 80 percent. Similarly, high effective discount rates have been observed in developing countries. If higher effective
discount rates are applied, the higher capital costs of kerosene mantle lamps and, especially, electric grid connections will tend to present
more of a barrier to investment.
SOURCES: Derived from figures 3-6 and 3-7A. See also: Harry Chernoff, individual Purchase Criteria for Energy-Related Durabtes: The Misuse of Life Cycle
Cost, Energy Journal, vol. 4, No. 4, October 1983, PP. 81-86; David French, The Economies of Renewable Enerav Svstems for DeveloDina
. . ,
Countries, Washington, DC, June 1979.
. .
permit. Kerosene wi ck l amps are thus wel l matched
to the real i ty of rural l i fe i n devel opi ng countri es,
where capi tal and resources are sharpl y l i mi ted.
Wood fi res and/or kerosene wi ck l amps are the
pri mary sources of l i ght for more than two bi l l i on
peopl e worl dwi de.
Next i n the progressi on are butane or pressuri zed
kerosene mantl e l amps. These are much l i ke the gas
l amps used for campi ng i n the Uni ted States. Mantl e
l amps gi ve substanti al l y more l i ght and are more
effi ci ent than wi ck l amps; they al so cost more to
purchase and operate, tend to be hot and noi sy, and
can cause consi der abl e gl ar e.
Fi nal l y, i n contrast to kerosene l amps or other
nonel ectri c l i ghti ng technol ogi es, el ectri c l i ghti ng i s
cl ean, rel ati vel y safe, easy to operate, effi ci ent, and
provi des hi gh-qual i ty l i ght. Peopl e i n rural areas and
smal l towns of devel opi ng countri es pl ace el ectri c
l i ghti ng hi gh on thei r l i st of desi red energy servi ces.
For exampl e, a survey of 320 househol ds i n several
vi l l ages and smal l towns of Ni geri a found that 90
percent ranked el ectri ci ty--pri mari l y for l i ghti ng
as thei r top choi ce i n desi red energy servi ces.
38
Even where el ectri c l i ghti ng i s avai l abl e, how-
ever, the hi gh cost of connecti ng to the el ectri c gri d
creates a substanti al barri er for poor fami l i es that use
onl y one or a few l i ghtbul bs (see fi gure 3-7),
39
and
thi s substanti al l y sl ows penetrati on. A study i n
Gujurat, I ndi a, found that 10 years after vi l l ages had
gai ned access to the el ectri c gri d, l ess than a thi rd of
the househol ds had connected; thi s i ncreased to
about two-thi rds after 20 years.
40
Uncertai n el ectri c
suppl y i n many devel opi ng countri es-i ncl udi ng
bl ackouts and brownouts-al so tends to di scourage
potenti al users and forces those who have connected
to the gri d to si mul taneousl y mai ntai n al ternati ve
kerosene l i ghti ng systems.
Sslljwwd I. Onyebuchi,
Analysis of Rural Energy Choices in Nigeria, Natural Resources Forum, vol. 12, No. 2, 1988, pp. 181-186.
s~i~ 3-7 tends t. ~ders~te tie barrier tit grid commtion costs present to people in poor rural SKXM. The perceived ~d us@Y the r~ costs
to fmncecormection charges areoftenmuchhigher in developing countries than the 10 percent disemmt rate assumed for this figure. Using more realistic
effeetivediscount rates of 50 percent the cost of eleetric Iighting-ifthe villager could raise the money at sil-peroperating hour would rise from $O.07
to $0.34, compared to $0.02 for a kerosene mantle light and $0,01 for a kerosene wick lamp. The choice of kerosene wick or mantle lamps is thus logical
given the f~ncial constraints that the poor face.
4oGya s~m (~.), Energy Use in Rural Gujurat, Op. cit., footnote 37.
33-718 0 - 90 - 3
El ectri ci ty use for l i ghti ng ri ses rapi dl y wi th
househol d i ncome. For exampl e, i n South Bombay,
I ndi a, rates of househol d el ectri ci ty use duri ng the
eveni ng vari ed from 93 watts for the l owest i ncome
group to 365 watts for the hi ghest i ncome group.
4l
The choi ce of el ectri c l i ghti ng technol ogy al so
vari es as i ncomes i ncrease. Low-i ncome househol ds
i n South Bombay i nstal l ed more conventi onal fl uo-
rescent l i ghts-despi te thei r hi gher capi tal cost
and operated them more i ntensi vel y due to thei r
l ower operati ng costs. As i ncomes i ncreased, house-
hol ds shi fted away from the harsh l i ght of conven-
ti onal fl uorescent to the more natural l i ght of
i ncandescent (see fi gure 3-8).
42
As i ncomes i ncrease wi th economi c devel opment,
househol ds begi n to buy other appl i ances-radi os,
TVs, fans, refri gerators, and ai r condi ti oners. El ec-
tri ci ty use for l i ghti ng usual l y conti nues to i ncrease,
but i t becomes onl y a smal l fracti on of total
resi denti al el ectri ci ty use (see fi gure 3-9). El ectri ci ty
use for l i ghti ng i n the commerci al and servi ce
sectors al so grows rapi dl y as the economy expands.
The demand for l i ghti ng has al so conti nued to
i ncrease i n the i ndustri al i zed countri es over the past
30 years as i ncomes have i ncreased. Today, the
average rate of l i ghti ng use ranges from roughl y 20
to 100 mi l l i on l umen-hours per capi ta per year
(Ml mhr/cap-yr) i n the i ndustri al countri es.
43
I n
compari son, annual househol d l i ght producti on i n
South Bombay vari es wi th househol d i ncome from
.
44
l i ght producti on i n the
about 1 to 3 Ml mhr/cap-yr,
commerci al sector mi ght doubl e these numbers.
Thi s i s equi val ent to a per-capi ta consumpti on l evel
that i s onl y 10 to 30 percent of the l owest l evel s
among i ndustri al i zed countri es.
I f l i ghti ng servi ces equal to hal f the mi ni mum
l evel observed i n the i ndustri al i zed countri es-10
Ml mhr/cap-yr--are to be provi ded i n devel opi ng
countri es, then per-capi ta demand for l i ghti ng el ec-
tri ci ty wi l l be about 500 (kWh) ki l owatthours per
Figure 3-3-Changes in Capacity and Type of Installed
Electric Lighting Per Household With Income Level in
South Bombay, India
Installed wattage, kW
Percent fluorescent
1.6
I I
50
1. 4-
1. 2-
1 -
0 . 8 -
0 . 6 -
0 . 4 -
0 . 2 -
I
20
10
I
0 ! I 1 0
RS 0-1000 RS 1000-2500 RS 2500-4000
RS 4000-
Income group

Installed wattage
+ percent fluorescent
lnstalled wattage per household and the fraction of installed
wattage that is fluorescent (the remainder is incandescent) is
shown versus household income in rupees. The intensity of use
of this installed wattage varied with the type of lighting and the
household income. The lowest income group used 80 percent of
their installed capacity of fluorescent and 45 percent of their
incandescent during the evening. The highest income group
used just 25 percent of their installed capacity of both fluorescents
and incandescents during the evening.
SOURCE: Aehok Gadgil and Bhaskar Natarajsn, Impact of Soeio-
Eeonomic and Arehiteetural Factors on Peak Eleetrieity De-
mand: A Case Study of South Bombay, Energy, vol. 14, No. 4,
1969, pp. 229-236.
year. Thi s i s equi val ent to an eveni ng power demand
of perhaps 150 watts per capi ta.
45
I f that l evel of
eveni ng demand occurred at the uti l i ty system peak
l oad, as i s typi cal i n devel opi ng countri es, then the
capi tal cost to provi de el ectri ci ty for l i ghti ng woul d
be roughl y $300 per person.
46
Space Conditioning, Refrigeration, and
Other Appliances
Space Condi ti oni ng
Heati ng resi denti al or commerci al bui l di ngs wi l l
never bean i mportant energy servi ce i n the majori ty
of devel opi ng countri es si nce most have tropi cal
cl i mates. Space heati ng wi l l be i mportant i n some
dlc~c~t~~m data in Ashok Gadgil and Bhaslmr Nawajn,
Impact of Socio-Economic and Architectural Factors on Peak Electricity Demand:
A Case Study of South Bombay, Energy, vol. 14, No. 4, 1989, pp. 229-236. The lowest income group uses 80 percent of their installed 71 watts of
fluorescent and 45 percent of their installed 81 watts of incandescent; the highest income group uses 25 percent of their installed 1,460 watts of
fluoreseents and incandescent.
42~s p~culw case con~asts ~th the more typic~ si~tio~ as discussed for cooking, where the poor ae P*CUkWIY sensitive to f~t costs.
43
Terry McGowaW Energy-Efficient Lighting, in Thomas B. Johansso~ Birgit Bodlund, and Robert H. Williams (eds.), Electncity: E@cient
En&Use and New Generation Technologies, and Their Planning Implications (Lund, Sweden: Lund University Press, 1988).
~Calcfited from Gad@ and Natarajan, op. cit., fOOtIIOte 41.
45~s ~swes tit li~t Ouqut is my split ~~=n in~ndes~nt and fluoreswnt fighfig, ~d tit the de~nd is h the s hews of the evening.
~Assuming an installed capital cost for the system of $2,000 per kW of generating capacity.
Chapter 3-Energy Services in Developing Countries . 59
Figure 3-9-Household Electricity Use for Lighting
v. Household Income, in Brazil
Electricity use, Brazil (kWh/household-month)
4 0 0
T - -
300-
200-
100-
/ /
o
1 I I 1 ~
MWU O-2 MWU 2-5 MWU 5-10 MWU 10-20 MWU 20
income group

Tot al el ect ri ci t y + Lighting electricity
This graph shows that electricity use for lighting continues to grow
with income even in a relatively properous developing country
such as Brazil. Lighting electricity is, however, only a small fraction
of total household electricity use in this case. MWU are minimum
wage units.
SOURCE: Aehok Gadgil and Gilberto De Martino Jannuzzi, Conservation
Potential of Compact fluorescent Lamps in Inc#a and Brazil,
Lawrence Berkeley Laboratory and Universidade Estadual de
Carnpinas (Brazil: June 23, 1989).
areas, however, such as mountai nous regi ons and
hi gh-l ati tude areas l i ke northern Chi na. Bei ji ng, for
exampl e, has about the same annual average l ow
temperature as Chi cago. Nearl y 20 percent of
Chi nas total annual coal consumpti on and 5 percent
of i ts annual bi omass consumpti on are used for
space heati ng (app. 3-A).
47
I n Chi na, resi dences rarel y have any i nsul ati on
and often have l arge gaps around doors and wi n-
dews.
48
I ndoor temperatures i n these homes are
control l ed not by a thermostat or by comfort
requi rements, but by fuel suppl y-and fuel , though
cheap, i s scarce. I n Kezuo county, Northeast Chi na,
for exampl e, average i ndoor temperatures are at the
freezi ng poi nt duri ng the wi nter, compared to
average outdoor temperatures of3C to-5 C wi th
l ows of 25 C.
49
Addi ti ons to coal suppl y, more
effi ci ent stoves, or better wal l i nsul ati on woul d thus
resul t mai nl y i n comfort i mprovements but not i n
energy savi ngs.
Si mi l arl y, al though many devel opi ng countri es
have hot cl i mates,
50
l i ttl e energy i s used at present
for space cool i ng i n devel opi ng countri es. Tradi -
ti onal bui l di ng desi gns somewhat moderate the
extremes i n temperature through natural venti l ati on
and other techni ques that make use of l ocal materi al s
and do not requi re addi ti onal energy i nputs.
51
I ncreasi ng urbani zati on and the use of commerci al
bui l di ng materi al s, however, have made these tradi -
ti onal practi ces l ess practi cal and l ess popul ar.
Acti ve space venti l ati on by el ectri c fans has become
popul ar i n many ar eas wher e ther e i s r el i abl e el ectr i c
servi ce and costs are affordabl e. For exampl e,
el ectri c fan ownershi p i n Bei ji ng, Chi na, jumped
from 47 percent of househol ds i n 1981 to 77 percent
i n 1984.
52
Ai r condi ti oni ng i n resi dences i s a l uxury i tem
found onl y i n the hi ghest i ncome househol ds i n
devel opi ng countri es (see fi gure 3-10).
53
I n contrast,
60 percent of al l homes i n the Uni ted States-nearl y
al l who need i thave ai r condi ti oners.
54
A substan-
ti al proporti on of commerci al , i nsti tuti onal , and
government bui l di ngs i n devel opi ng countri es are
ai r condi ti oned.
Ai r condi ti oni ng systems i n devel opi ng countri es
are al so often l ess effi ci ent than those i n i ndustri al -
i zed countri es. Bui l di ngs usual l y are poorl y i nsu-
l ated, wi th l arge amounts of ai r i nfi l trati on; and ai r
condi ti oners are general l y l ess effi ci ent than those i n
the west and are poorl y mai ntai ned and control l ed.
TVac~v sti, es wern, contractor report prepared for the Office of Technology Assessment 1990.
4SRo&~M. w~wtm, Energy-ConservationS tin&r& for Buildings inC~ Energy, vol. 13, N0. 3,1988, pp.265-274; RobertM. W~
and Chang Song-ying, Energy Conservation in Chinese Housing, Energy Policy, vol. 15, No. 2, pp. 158-168.
@World B@ <*-: Co~~-~velR~ ~ergy Assessmen~: A Joint Study of ~~ and -se -, Activity Completion Report No.
101/89, May 1989.
~~ 50 of tie ~orld~~ hott~t ~ties ~ in tie devdopingworld. me hottest is Djibouti, wi& SII av~ge annual high temperature of 113 ~. NO!M
of the 50 coldest cities is in the developing world. See V. Showers, World Facts andFigures (New Yo& NY: John Wiley& Sons, 1979).
51LfiJ =Yuq ~~Tm&tio~Housing: A solution to Hopelessness in & ~world: me Malaysian ~pl e, T&&OZOg@ VO1. 18, No. 1,
1988, pp. 16-23; Mehdi N. Bahadori, Passive Cooling Systems inhmiankchitecture, Scient@cAmerican, vol. 238,1978, pp. 144154; R,K. Hill,
Utilization of Solar Energy For an Improved Environment Within Housing For the Humid Tropics (VictoriA Australia CSIRO, 1974).
52J. Sa~Ye, A. G~~ and L. sc~PPerj Energy
Demand in Developing Countries: A Sectoral Analysis of Recent Trends, Annual Review of
Energy, vol. 12, 1987, pp. 253-281.
53J ayant Sathaye and Stephen Meyers,
Energy Use in Cities of the Developing Countries,Annual Review of Energy, 1985, vol. 10, pp. 109-133.
54~erm ~omtion A*~tion, HoWng Characteristics 1984, DoE~-0314(84)(WSS~C)U DC: U.S. GOV
ernment Printing Office,
October 1986), p. 5.
Figure 3-10-Electric Appliance Ownership in Urban
Malaysia by Income Group, 1980
Penetration by income group, percent
100
J
80-
80-
<
40-
20-
0+
Bottom 3% Lower 20% Middle 28% Middle 29% Upper 15% Top 5%
i nc ome group

Ref r i ger at or s
+ Air conditioners + Water heaters
This figure shows the rapid penetration of refrigerators, air
conditioners, and water heaters as household incomes rise. The
incomes (percentage of households) are in ascending order:
150-299 Malaysian dollars per month (M$/month) (3 percent); M$
300-599 (20 percent); M$ 600-999 (28 percent); M$ 1,000-1,999
(29 percent); M$ 2,000-4,999 (15 percent); and M$ 5,000+ (5
percent).
SOURCE: Jayant Sathaye and Stephen Meyers, Energy Use in Cities of
the Developing Countries, Annua/ Review of Enemy, vol. 10,
1965, pp. 109-133.
Figure 3-12Refrigerator Ownership in
Beijing, China, 1981-1987
Percentage of households
70
6 0 -
5 0 -
4 0 -
3 0 -
2 0 -
1 0 -
0 c 1 !
1980 1982 1984 1986 1988
This figure shows the rapid penetration of refrigerators into the
household sector over just a 6-year period.
SOURCE: Stephen Meyers and Jayant Sathaye, Electricity Use in the
Developing Countries: Changes since 1970, EneWy, vol. 14,
No. 8, 19S9, pp. 435-441, table 6.
Figure 3-1 lElectric Appliance Ownership
in Urban Java, 1988
Penetration by income group, percent
, , 2 0 ~ -
~
Low Middle High
Income group
- L i g h t s + I r o n + B&W/ Col or TV
-
F a n
+
R e f r i g e r a t o r + Wat er pump
This figure shows the rapid penetration and relative importance
within household purchasing patterns of lights, TVs, irons, fans,
refrigerators, and water pumps. Income groups (share of house-
holds) in ascending order are: less than 75 (Rupees)/month (24
percent), 75-120 Rp/month (22 percent), 121-185 Rp/month (21
percent), 186-295 Rp/month (14 percent), and greater than 295
Rp/month (9 percent).
SOURCE: ke Schipper and Stephen Meyers, Improving Applianoe
Efficiency in Indonesia, Energy Po/icyforthcoming.
Figure 3-13-Reduction in the Real Cost of
Refrigerators Over Time in the United States
Real cost of refrigerators, 1989=1.00
7
1
J
6 -
5 -
4 -
3 -
2 -
1 -
0 -
0
k
1940
1950
1 i 1 1
50 100 150 200 250
Cumulative production volume (millions)
Over the past 40 years, the real price of refrigerators has dropped
by almost a factor of 5. For developing countries, such price
reductions would allow households to invest in refrigerators at a
much earlier point in time than was the case for the United States
and other industrialized countries at a similar level of develop-
ment.
SOURCES: Rick Bahr, Bureau of Labor Statistics, U.S. Department of
Labor, personal communication, July 9, 1990 (CP1/
refrigerators); John Chirichiello, National Science Foundation
SRS Computer Bulletin Board, personal communmtion, July
6, 1990 (GNP deflator, 1953-19S9); U.S. Department of
Commerce, Bureau of the Census, Historical Statistics of the
United States: Colonial Times to 1970 (Washington, DO: U.S.
Government Printing Office, 1975), p. E1-12 (GNP deflator,
1935-53).
Chapter 3--Energy Services in Developing Countries . 61
Table 3-6-Residential Nonheating Electricity
Intensity in Selected Countries, 1970 and 1986
(kilowatthours per capita)
Country 1 9 7 0 1 9 8 6
India . . . . . . . . . . . . . . . . . . . . . . .
Indonesia . . . . . . . . . . . . . . . . . . .
Pakistan . . . . . . . . . . . . . . . . . . . .
Philippines . . . . . . . . . . . . . . . . . .
Thailand . . . . . . . . . . . . . . . . . . . .
Malaysia . . .
:
. . . . . . . . . . . . . . . .
Mexico . . . . . . . . . . . . . . . . . . . . .
South Korea . . . . . . . . . . . . . . . . .
Brazil . . . . . . . . . . . . . . . . . . . . . . .
Argentina . . . . . . . . . . . . . . . . . . .
Venezuela . . . . . . . . . . . . . . . . . .
Taiwan . . . . . . . . . . . . . . . . . . . . .
Japan . . . . . . . . . . . . . . . . . . . . . .
West Germany . . . . . . . . . . . . . . .
United States . . . . . . . . . . . . . . . .
7
8
10
3 4
17
37
75
2 5
9 0
210
158
163
25
3 3
5 9
78
1 1 0
184
190
248
261
307
422
557
975
1,210
3,050
SOURCE: Stephen Meyers and Jayant Sathaye, Electric.Hy Use in the
Developing Countries:Changes Since 1970; Energy, VOI.14,
No. 8, 1989, pp.435-441.
The potenti al for i ncreased energy use for space
cool i ng i s very l arge. The Uni ted States now uses
about 1,400 kWh of el ectri ci ty per person per year
for space cool i ng.
55
I f I ndi a used thi s much el ectri c-
i ty per person for space cool i ng, i ts total annual
el ectri ci ty generati on woul d have to i ncrease to
more than fi ve ti mes present l evel s.
56
The hotter
cl i mate of I ndi a coul d i ncrease these requi rements
sti l l more.
Refri gerators and Other Appl i ances
El ectri ci ty-usi ng appl i ances--refri gerators, tel e-
vi si ons, washi ng machi nes, etc.are rapi dl y pene-
trati ng the resi denti al sector of devel opi ng countri es.
Factors contri buti ng to thi s expl osi ve growth i n-
cl ude ur bani zati on, i ncr easi ng el ectr i fi cati on i n r ur al
areas, economi c growth, i mproved access to appl i -
ances, and decreasi ng real costs of appl i ances
whi ch make them affordabl e to a broader segment of
the popul ati on than ever before. Factors l i mi ti ng
appl i ance penetrati on i ncl ude the l ack of el ectri c
servi ce, parti cul arl y i n rural areas. I n Brazi l , for
exampl e, 90 percent of urban househol ds but onl y 24
percent of rural househol ds have el ectri c servi ce.
57
The rapi dl y i ncreasi ng use of househol d appl i -
ances i n the devel opi ng countri es pl aces addi ti onal
demand on el ectri c power i nfrastructures that are
typi cal l y al ready short of capaci ty. Further, much of
the resi denti al demand comes at peak ti mes. A
revi ew of 13 of the l argest devel opi ng countri es for
the peri od 1970-86 found that the growth rate of
el ectri ci ty consumpti on was hi ghest i n the resi den-
ti al sectoraveragi ng 9.9 percent annual l y, com-
pared to 8.3 percent annual growth i n the i ndustri al
sector .
58
Tabl e 3-6 shows el ectri ci ty i ntensi ty for the
resi denti al sector i n sel ected devel opi ng and i ndus-
tri al i zed countri es. Even the most advanced devel -
opi ng countri es use, on average, just a smal l fracti on
of the el ectri ci ty consumed by Ameri cans. El ectri c-
i ty consumpti on by the economi cal l y wel l off i n
devel opi ng countri es, however, di ffers l i ttl e from
that found i n the Uni ted States or Europe.
Li ghts are usual l y the fi rst appl i ance i nstal l ed
when a househol d gets el ectri c servi ce. Acqui si ti on
of other appl i ances vari es by househol d i ncome and
regi on (see fi gures 3-10 and 3-1 1). I n I ndi a, fans are
typi cal l y among the fi rst appl i ances acqui red, fol -
l owed by tel evi si ons and refri gerators. I n Brazi l ,
even rel ati vel y poorer, newl y el ectri fi ed househol ds
often have tel evi si ons and refri gerators, as these
appl i ances are comparati vel y i nexpensi ve and are
avai l abl e secondhand.
59
Refri gerator ownershi p i s at present qui te l ow i n
most devel opi ng countri es. I n Chi na, for exampl e,
l ess than 1 percent of househol ds have refri gerators,
al though refri gerator ownershi p has been growi ng
rapi dl y i n Bei ji ng (see fi gure 3-12). I n Brazi l , 63
percent of househol ds have refri gerators.
a
I n con-
trast, i n the Uni ted States, 99.7 percent of house-
hol ds have refri gerators.
G1
55GSS Rese~h IIIstitute, Stiategic ~ysis and Energy Forecasting Divisio~ Baseline Projection Data Book (WaShin@OQ DC: GSS Research
Institute, 1988), pp. 37, 120.
56TStS Energy ResWchrnsti@te, TERIEnergy Data Directory and Yearbook (TEDDY) 1988 (NCW DelM India: ISW, p. 73.
57Gilberto De l&u-tino J ~uzzi, Residenti Energy Demand in Brazil by IIKome CISSSeS,
Energy Policy, vol. 17, No. 3, p. 256.
58Stephen Meyemand Jayat Satiye, E1ectrici~USe intheDeve@ingCOun~~: Changes Since 1970, Energy, VO1. 14, No. 8,1989, PP. 435-1.
59A. Ga@il and G. De Martin
. 0 Jannuzzi, Conservation Potem.al of Compact Fluorescent Lumps in India andBrazil, LBL2721O (Berkeley, CA:
Lawrence Berkeley Laboratory, July 1989), p. 5.
@Howard S. Geller, Electricity Conservation in Brazil: Status Report and Analysis, contractor report prepared for the Ofilce of Technology
Assessment, March 1990, p. 17.
GIEner~ ~o~tion Administration Housing Characteristics 1984, Op. Cit., fOOhlOte 5A, p. Is.
62 q Energy in Developing countries
The refri gerators used i n devel opi ng countri es are
typi cal l y hal f the si ze of Ameri can refri gerators, or
smal l er. They are al so much l ess effi ci ent than the
best refri gerators now commerci al l y avai l abl e (the
average refri gerator used i n the Uni ted States i s
si mi l arl y much l ess effi ci ent than the best avai l abl e).
I n I ndonesi a, most refri gerators are assembl ed l o-
cal l y from i mported components and, i n general , do
not take advantage of proven energy effi ci ency
features such as rotary compressors and i ncreased
i nsul ati on.
62
The effi ci ency of Brazi l i an refri gera-
tors i s bei ng i mproved--el ectri ci ty consumpti on by
the average new model was reduced by 13 percent
between 1986 and 1989--but they are unabl e to
make use of the very effi ci ent motor-compressors
(whi ch Brazi l manufactures and exports), as these
uni ts cannot tol erate the vol tage fl uctuati ons found
i n Brazi l .
63
Advances i n materi al s and manufacturi ng tech-
ni ques, coupl ed wi th a growi ng secondhand market,
are forci ng down the fi rst cost of refri gerators and
other appl i ances. The real cost of new refri gerators
i n the Uni ted States, for exampl e, has pl ummeted b y
a factor of nearl y 5 over the past 40 years (see fi gure
3-13). Thi s trend shoul d make many househol d
appl i ances affordabl e to a much l arger share of
devel opi ng country popul ati ons than was the case
for todays i ndustri al i zed countri es at a comparabl e
l evel of devel opmenta generati on or more ago. As
a resul t, energy use coul d i ncrease si gni fi cantl y
above the hi stori cal record i n the near to mi d-term.
For exampl e, the average new refri gerator i n the
Uni ted States uses about 1,000 kWh of el ectri ci ty
per year.
64
I f every househol d i n Chi na had a
U.S.-styl e refri gerator, an addi ti onal 200,000 gi ga-
watthours (GWh) of el ectri ci ty per year-or the
output of about 50 ful l -si ze coal -burni ng power
pl antswoul d be requi red, at a cost for the power
pl ants al one of about $100 bi l l i on.
65
The Industrial Sector
The i ndustri al sector typi cal l y consumes 40 to 60
percent of total commerci al fossi l energy i n devel op-
i ng countri es (see tabl e 3-l );
66
i t al so makes heavy
use of tradi ti onal bi omass fuel s-often traded i n
commerci al markets. The pri mary energy servi ces
requi red by i ndustry are process heat and mechani cal
dri ve. These servi ces wi l l be treated generi cal l y
here; i n a l ater report of thi s OTA study, they wi l l be
exami ned as speci fi c parts of i ntegrated i ndustri al
pr ocesses.
Fi rms i n the i ndustri al sector of devel opi ng
countri es today vary wi del y i n si ze and sophi sti ca-
ti on. Atone end of the spectrum are smal l tradi ti onal
fi rms that use rel ati vel y energy-i neffi ci ent and
l ow-producti vi ty manufacturi ng technol ogi es.
67
At
the other end are l arge, modern fi rms, often wi th
mul ti nati onal parent compani es, that have worl d-
cl ass manufacturi ng capabi l i ti es.
Manufacturi ng operati ons typi cal l y fal l i nto three
broad si ze categori es-househol d or cottage, smal l
workshops and factori es, and l arge-scal e i ndustry.
Over ti me, a few smal l er compani es tend to grow
i nto l arge ones as the transport i nfrastructure i m-
proves and i ncomes ri se, i ncreasi ng the si ze of
markets and provi di ng economi es of scal e that turn
the advantage to l arger fi rms.
68
I n many devel opi ng countri es, one-hal f to three-
quarters of manufacturi ng empl oyment i s i n house-
hol d-scal e establ i shments, wi th the remai nder di -
vi ded between medi um and l arge operati ons. Much
of the empl oyment i n the smal l tradi ti onal (and
l argel y rural ) househol d i ndustri es i s seasonal l abor
avai l abl e duri ng the nonagri cul tural l y acti ve ti mes
of year. Typi cal l y one-fourth to one-thi rd of rural
62~e Schippm, ~~Effl~mt Household El~~ci~ use in Indonesi~ Lawrence Berkeley Laboratory, draft repo~ Jan- 1989, P. 3) ~tion o
n
Conservation Potential.
G3Howmd S. Geller, Electnci~ Conservation in Brazil: Status Report and Analysis, Op. Cit., footnOte @P. 29.
MHow~d S. Gefler, Residential Equipment Efficiency,
contractor report prepared for the Off& of Technology Assessmen4 May 1988.1990,
NAECA standard.
6SA~~_: 5 ~ople perhouaehold, 45 percent load factor, no @a
nsmission and distribution losses, and a capital cost of $2,000 per kW of installed
capacity.
66J. Satiye, A. G~~di, L. Sc~pWr, ~~er. De~nd in Developing co~&ies: A secto~ A@@ of Recent Trends, op. cit., foo~ote 52,
table 5.
67~thou@ they me s~l ~d often use fitde modern technology or me~ods, these ~ufacturing enterprises me not inefficient h SOme respects.
High tmnsportandmarketingc osts and small market size might greatly raise the cost to larger, modern fixms if they should try to enter these small village
markets, making them the higher cost producers.
68De~s ~derson, World B@ S~ ~us~ in ~velop~gcountries: some Issues, SW WOIIC@ Paper No. 518, 1982.
Table 3-7-Kenyan National Energy Use by Fuel, 1980 (percent of total)
a
Commercial
Biomass fuels
Energy service fuels
Wood Charcoal Other Total
Household . . . . . . . . . . . . . . . . . . 2.9 46.3 6.6 2.7 58.5
Cooking/heating . . . . . . . . . . . . 1. 0 46.3 6.1 2.7
Lighting . . . . . . . . . . . . . . . . . .
.
1. 7

Other . . . . . . . . . . . . . . . . . . . . .
.
0.2
0.5

Industry . . . . . . . . . . . . . . . . . . . . . 8.6 14.5 1.0 24.1
Large . . . . . . . . . . . . . . . . . . . .
8.6 5.3 0.3

Informal urban . . . . . . . . . . . . .

0.1 0.6
Informal rural . . . . . . . . . . . . . .

9.1 0.1
.
Commerce . . . . . . . . . . . . . . . . . . 0.6 0.5 0.1
1. 2
Transportation . . . . . . . . . . . . . . . 13.7

13.7
Agriculture . . . . . . . . . . . . . . . . . . 2.5

2.5
Total . . . . . . . . . . . . . . . . . . . 28.4 61.3 7.6 2.7 100.0
a Total national energy consumption .332 million gigajoules; per capita power consumption=658 watts.
SOURCE: Phil OKeefe, Paul Raskin, and Steve Bernow(eds.), Energy andDeve/opmentin Kenya: Qmortunities and Constraints fUddevalia. Sweden: tiijer
Institute and Scandinavian Institute of African St~ies, 19&).
nonfarm empl oyment i s i n manufacturi ng.@ Thi s i s
an i mportant source of i ncome and empl oyment for
the rural and poor urban sectors.
70
Process Heat
Many commerci al and i ndustri al processes re-
qui re heatrangi ng from the l ow-temperature heat
used to dry food by cottage i ndustry to the hi gh-
temperature processes used by l arge i ndustri es to
produce steel and cement. The effi ci enci es of these
processes are typi cal l y much l ower than those found
i n i ndustri al i zed countri es.
Tradi ti onal Process Heat Technol ogi es
Bi omass i s used extensi vel y i n both tradi ti onal
rural and more modern i ndustry i n devel opi ng
countri es. I n Kenya, for exampl e, l arge i ndustry
accounts for about 8.6 percent of nati onal energy use
i n the form of commerci al fuel s, and 5.6 percent of
total nati onal energy consumpti on i n the form of
. s , .
bi omass (wood and charcoal ). I nformal rural and
urban i ndustri es use l i ttl e or no commerci al fuel , but
they account for about 10 percent of total nati onal
energy use i n the form of bi omass (see tabl e 3-7).
Rural appl i cati ons i ncl ude beer brewi ng, bl ack-
smi thi ng, crop dryi ng, and pottery fi ri ng (see tabl e
3-8).
Esti mates of the use of bi omass energy for
i ndustri al processes are si mi l arl y hi gh el sewhere.
Tobacco curi ng uses 11 percent of al l fuel wood i n
I l ocos Norte, Phi l i ppi nes, and represents 17 percent
of the nati onal energy budget i n Mal awi .
71
I n
I ndonesi a, the bri ck, ti l e, and l i me i ndustry con-
sumes roughl y 2.5 percent of nati onal energy use.
72
Beer brewi ng uses 14 percent of the total fuel wood
consumed i n Ouagadougou, Burki na Faso.
73
Over -
al l , bi omass fuel s suppl y up to 40 percent of the
i ndustri al energy used i n I ndonesi a, 28 percent i n
Thai l and, 17 percent i n Brazi l , and si mi l arl y l arge
fracti ons i n many other countri es.
74
@?DennisAnderson and MsrkI-eiserso~ RurslNonfsrmErnployment in Developing Countries, Economic DevelopmentandCultural Change,voL
28, No. 2, 1980, p. 245, table A2, cited in Donald W. Jones, Energy Requirements for Rural Development (Oak Ridge, TN: Oak Ridge National
Laboratory, June 1988).
%inyinnachuta and S.V. Sethuraman (eds.), Rural Small-Scale In&sm.es andEmployment in Aj%ica andAsia (Geneva: International Labor Office,
19s4).
TIE.L. H= The Demand for Woodfuels by Cottage Industries in theprovinee of Ilocos Norte, Philippines, Energy, vol. 9, pp. 1-13, 1984; E.M.
Mnzavz VillageIndustries vs. Savannah Forests, UNASYLVA, vol. 33, No. 131,1981, pp. 24-29; E.M. Mnzmq Fuelwood and Charcoal in Afriw
W. Paley, P. ChsrtierandD.O. Hall (eds.),Ener~fiornBiomass @mdon: Applied Science Publishers, Ltd., 1980); M.J. Mwandosyasnd M.L. Luhan~
Energy DemandStructures in RuraZ Tanzania (PrincetoU NJ: Center for Energy and Environmental Studies, Princeton University, and Dar-Es-S_
mnzania: Department of Electrical Engineering, University of Dar-Es-Salaam, 1984).
TWorldB@ Energy Sector Management Assistance program,
Indonesiz Energy Effieiemy Improvement in the BnclG Tile and Lime Industries
on Javz March 1987.
TsHenri ~u~ men ~ ~~ City RUIIS Out of Fue~ UNASIZVA, VO1. 33, No. 133, 1981, pp. 11-20.
TAJoy Dunkerley et al., Energy Strategies for Developing Countries (Baltimore, MD: JOhIIS HOPkiDS UtiversiV Prms, 1981), P. 265.
Table 3-8-Annual Consumption of Fuelwood
and Charcoal in Kenya by Rural Cottage Industries,
GJ/Capita
Industry Fuelwood Charcoal
a
Brewing . . . . . . . . . . . . . . . . . . . . . 1.07
Construction wood . . . . . . . . . . . 0.50
Butchery . . . . . . . . . . . . . . . . . . . . 0. 24
Restaurants . . . . . . . . . . . . . . . . . 0.17
Baking . . . . . . . . . . . . . . . . . . . . . . 0. 13
Brick firing . . . . . . . . . . . . . . . . . . 0.06
Blacksmithing . . . . . . . . . . . . . . .
Crop drying . . . . . . . . . . . . . . . . . . 0.04
Tobacco curing . . . . . . . . . . . . . . 0.04
Fish curing . . . . . . . . . . . . . . . . . . 0. 02
Total . . . . . . . . . . . . . . . . . . . . . 2.27
.
0. 06
0. 04
0.06
0.16
a This does not include the losses in converting wood to charcoal.
SOURCE: Phi10Keefe,PaulRaskin,and Steve Bernow(eds.) Energyand
Development in Kenya: @ortunities and Constraints (Udde-
valla, Sweden: Beijer Institute and Scandinavian Institute of
African Studies, 19S4).
The effi ci ency wi th whi ch these tasks axe done
can be qui te l ow (see tabl e 3-9). On cl ose exami na-
ti on, however, the performance of tradi ti onal bi omass-
fuel ed technol ogi es i s often found to be careful l y
opti mi zed i n terms of effi ci ency, capi tal , and l abor,
gi ven exi sti ng materi al s and technol ogi cal con-
strai nts. An exampl e of thi s i s the tradi ti onal bri ck
ki l n i n Sudan, whi ch hol ds as many as 100,000
bri cks at a ti me and gai ns economi es through si ze
and other desi gn factors. To i mprove the perform-
ance of these technol ogi es usual l y requi res the i nput
of modem materi al s and technol ogi es, i ncl udi ng
modern means of measuri ng effi ci enci es.
Modern Large-Scal e I ndustry
Modern l arge-scal e i ndustri es i n devel opi ng coun-
tri es are model ed after thei r counterparts i n i ndustri -
al i zed countri es, but they are often operated at
si gni fi cantl y l ower effi ci enci es. A few energy-
i ntensi ve materi al s-steel , cement, chemi cal s (espe-
ci al l y ferti l i zer), and paper-account for much of the
energy used by i ndustry (see tabl e 3-10). The total
energy used to produce these materi al s wi l l i ncrease
rapi dl y as devel opi ng countri es bui l d thei r nati onal
i nfrastructures.
Steel I n the OECD countri es, the steel i ndustry
typi cal l y consumes about one-fi fth of the energy
used i n the i ndustri al sector.
75
Devel opi ng countri es
such as Chi na, I ndi a, and Brazi l devote a si mi l ar
share-18 percent, 23 percent, and 20 percent,
respecti vel y
76
--of i ndustri al commerci al energy con-
sumpti on to steel producti on. The top 10 producers
account for about 90 percent of the crude steel made
i n the devel opi ng worl d; many other devel opi ng
countri es produce l i ttl e or no steel .
Per-capi ta steel consumpti on i ncreases rapi dl y as
nati onal i nfrastructures are bui l t, and then tends to
saturate the market and l evel off at hi gher i ncome
l evel s
77
(see fi gure 3-14). A si mi l ar trend has been
found for a wi de vari ety of materi al s.
78
Si mpl y put,
there i s a l i mi t to the number of steel -i ntensi ve cars,
refri gerators, washi ng machi nes, bui l di ngs, bri dges,
pi pel i nes, etc., a person needs. Eventual l y, con-
sumpti on l evel s tend to pl ateau at repl acement
l evel s. When these wants for basi c materi al s are
ful fi l l ed, peopl e tend to spend i ncremental i ncome
on hi gher val ue-added materi al s-such as those
wi th a hi gh-qual i ty fi ni sh--or on l ess materi al -
i ntensi ve but hi gher val ue-added consumer goods.
The l evel of per-capi ta steel consumpti on needed
to pr ovi de a gi ven ser vi ce has al so been r educed over
ti me through a vari ety of technol ogi cal i mprove-
ments, i ncl udi ng hi gher wei ght-to-strength steel
al l oys, more effi ci ent motors and engi nes, better
desi gn, and the substi tuti on of al ternati ve products
such as hi gh-performance pl asti cs. For exampl e, the
tensi l e strength of steel i ncreased fourfol d between
1910 and 1980.
79
Over al l steel pr oducti on has been i ncr easi ng by a
l i ttl e over 7 percent per year i n the devel opi ng
countri es, whi l e remai ni ng rel ati vel y constant i n the
i ndustri al i zed countri es. At current rates, steel pro-
ducti on by devel opi ng countri es wi l l overtake that i n
the i ndustri al i zed countri es earl y i n the next century.
The energy effi ci ency of steel producti on i n the
devel opi ng countri es vari es wi del y. I n some cases,
i t has si gni fi cantl y l agged that of the i ndustri al i zed
75-W Y. IMeunier and Oscar de Bruyn KOPS,
Energy Efficiency in the Steel Industry With Emphasis on Developing Countries, World Bank
technical paper, No. 22, 1984.
T%id.
Per-capita steel eonsumption inenmses approximately linearly with per-capita income up to several thousand dollars.
T8Ro&fiH. Willi=, fi~Do~WU ~d~H. Ross, c~teri~s, ~uence, and~d~tri~~ergyuse, AnnM/R~iW ofEnergy, VO1. 12,1987,
pp. 99-144.
T~ono~c Codssion for EUIOW, Evolution of the Specific Consumption of Steel (New York m: Ufited Nations) 19~).
Table 3-9-Efficiency of Fuel Use In Traditional (Developing Countries) and Modern
(Industrial Countries) Commercial and Industrial Operations
Estimated efficiency of Estimated efficiency of
traditional technology
modern technology in U.S.
Activity Location
(percent) (percent)
Cooking
Beer brewing
Tobacco drying
Tea drying
Baking
Fish smoking
Brick firing
Foundry work
West Africa
Burkina Faso
Burkina Faso
Tanzania
Tanzania
Sudan
India
Guatemala
Tanzania
Sudan
India
Uganda
Indonesia
15-19
15-17
0.3-0.7
0.5
2.9
12-19
16
3
2-3
8-16
6.4
5-10
3
50-60
79
0.6
43
6-11
40
NOTE: Not applicable or not available.
SOURCE: For complete list of sources, see app. 3-B.
Table 3-10-Energy Consumption by
Chinese industry, 1980
Final energy use
Sector Exajoules Percent
Basic metals (iron and steel) . . . . . . . . . . 2.38 25.7
Chemicals (fertilizer) . . . . . . . . . . . . . . . . . 2.23 24.1
Building materials (cement, brick tile) . . . 1.44 15.6
Machine building . . . . . . . . . . . . . . . . . . . . 0.82 8.8
Textiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.64 6.9
Food, beverages, tobacco . . . . . . . . . . . . . 0.38 4.1
Pulp and paper. . . . . . . . . . . . . . . . . . . . . . 0.25 2.7
Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.12 12.1
Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.26 100.0
SOURCE: World Bank, China: The Energy Sector(Washington, DC: 1985).
countri es. I ntegrated steel pl ants i n I ndi a and Chi na
currentl y use, on average, 45 to 53 gi gajoul es (GJ)
per ton of crude steel produced; i ntegrated steel
pl ants i n the Uni ted States and Japan use hal f as
much energy .80 Some devel opi ng countri es have
made si gni fi cant stri des to reduce energy use i n steel
producti on. The Brazi l i ans, for exampl e, cut energy
consumpti on from 34 GJ to 27 GJ per ton of crude
steel between 1975 and 1979,
81
and the South
Korean steel i ndustry i s among the most effi ci ent i n
the worl d.
CementThe cement i ndustry typi cal l y con-
sumes 2 to 6 percent, and someti mes more, of the
commerci al energy used i n devel opi ng countri es.
The use of cement i s expected to i ncrease rapi dl y as
nati onal i nfrastructures of roads, bri dges, bui l di ngs,
etc., are bui l t. I n general , per-capi ta consumpti on of
cement i ncreases approxi matel y l i nearl y wi th i n-
come up to several thousand dol l ars, and then
saturates and l evel s off at hi gher i ncomes (see fi gure
3-15). Despi te the energy i ntensi ty of cement
producti on, i t i s one of the l east energy-i ntensi ve
constructi on materi al s when i n i ts fi nal form of
concrete/aggregate (see tabl es 3-11 and 3-12).
The val ue of cement i s qui te l ow compared to i ts
wei ght. Because of thi s and because the raw
materi al s for cementl i mestone, vari ous cl ay mi n-
eral s, and si l i ca sand-are wi del y avai l abl e, cement
i s usual l y produced rel ati vel y near i ts poi nt of use.
I n the Uni ted States, the maxi mum range for truck
shi pments of cement i s about 300 km. I n devel opi ng
countri es, where the transport i nfrastructure i s l ess
wel l devel oped, economi cal transport di stances are
often l ess. I n Chi na, for exampl e, 150 to 200 km i s
the typi cal l i mi t of transport; i f transport over l onger
di stance i s needed, the constructi on of a new cement
pl ant i n the l ocal area wi l l be consi dered.
82
Thus, as
a resul t of i nadequate transport i nfrastructures,
cement pl ants are often smal l and rel ati vel y i neffi -
ci ent.
The energy requi red to produce cement vari es
wi del y wi th the type of producti on process, qual i ty
W-CS Y. Metier and Oscar de BNp Kops, op. cit., footnote 75; Sven ~tow,
Energy Considerations of Classical and New Iron- and
Steel-Making Technology, Energy, vol. 12, No. 10/11, 1987, pp. 1153-1168.
BIMaUIim Y. Metier and Oscsr de Bruyn Kops, op. cit., footnote 75.
8% ~p~, ~mat ~d~~ in - Rock Products, February 1985, P. 32.
Figure 3-14-Per-Capita Steel Consumption v. GNP for Various Countries
Per capita steel consumption, kilogram
10004
100
10
J
s
q
s
9
. .
q
.
. 8
:
q ,m .
9
. 9
q
.
.
q
. . .
s
q m
9
m
.
q
9
- w .
- 9 .
q m .
.
.
q
q q
s
q q
q
q m
. 9
1
I
I I 1
0 5 10 15 20
GNP (thousand 1984 $U.S.)
The saturation of the steel market at higher income levels is readily seen in linear or logarithmic plots.
It is shown herein a semi-log plot so as to better display both low-end and high-end data. Each data point
represents a country,
SOURCE: United Nations, Statist&/ Yewbook 1985/86 (New York, NY: 1988), pp. 550-552, table 130.
Figure 3-15-Per-Capita Cement Production v. GNP for Various Countries
Per-capita cement production, kilogram
10000 j
1
.
1000
.
.
.
m
. .
m
-
.
.
m
. . q
.
.
.
.
i
1 .
10
/
.
:.
q
q9
s
.
.
1
I
I 1 I
o 5 10 15 2 0
Per-capita GNP (thousands, 1984 $U.S.)
The saturation of the cement market at higher income levels is readily seen in linear or logarithmic plots.
It is shown herein a semi-log plot so as to better display both low-end and high-end data Each data point
represents a country.
SOURCE: United Nations, Stafisticd Ye@ook 198548, (New York, NY: 1988), pp. 524-526, table 116.
Table 3-1 lAverage Energy Intensities of Building
Materials (megajoules per kilogram)
Table 3-12Energy Intensities of End Products
Using Alternative Building Materials
(megajoules per square meter)
Material Energy intensity
Concrete aggregate . . . . . . . . . . . . . . . . . . . 0.18
Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.80
Brick and tile. . . . . . . . . . . . . . . . . . . . . . . . . 3.7
Cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.9
Plate glass . . . . . . . . . . . . . . . . . . . . . . . . . . 25.0
Steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28.0
SOURCE: Mogens H. Fog and I@hore L. Nadkarni, World Bank, Energy
Efficiency and Fuel Substitution in the Cement Industry With
Emphasis on Developing Countries, teehnieal paper No. 17,
1983.
of raw materi al s, pl ant management and operati ng
condi ti ons, and other factors. The performance of
cement pl ants i n devel opi ng countri es al so vari es
wi del y and i s di ffi cul t to characteri ze si mpl y. Many
pl ants approach the effi ci ency of those i n the
i ndustri al i zed countri es, dependi ng on when they
were bui l t and the condi ti ons under whi ch they are
operated. Others show si gni fi cant i neffi ci enci es-
usi ng 25 to 50 percent more energy than effi ci ent
pl ants of the same type and wi th the same qual i ty of
raw materi al s i nput.
83
Mechanical Drive
Tradi ti onal Dri ve Power
The producti vi ty of peopl e i n many rural and poor
urban areas of devel opi ng countri es i s now l i mi ted
by thei r r el i ance on human and ani mal muscl e power
for water pumpi ng, grai n gri ndi ng, agri cul tural
acti vi ti es, transportati on, and smal l i ndustry. When
onl y muscl e power i s avai l abl e, many hours can be
spent si mpl y on enabl i ng acti vi ti es, such as
haul i ng water or gri ndi ng grai n, rather than on
di rectl y economi cal l y producti ve acti vi ti es. Produc-
ti ve acti vi ti es themsel ves are sharpl y l i mi ted by the
effi ci ency and total output of muscl e power. I f the
producti vi ty of peopl e i n rural areas of devel opi ng
Structure Concrete Steel Asphalt Brick
Building wall . . . . . . . . 400 600
Bridge (per m
2
) . . . . . . 4,000 8,000
Roadway (per m
2
).... 800 3,000
Not applicable or not available.
SOURCE: Mogens H. Fog and Kishore L. Nadkarni, Wotid Bank Energy
Effieieney and Fuel Substitution in the Cement Industry With
Emphasis on Developing Countries, T=hnieal Paper No. 17,
1988.
countri es i s to be i ncreased, modem motor dri ve
technol ogi es and supporti ng i nfrastructures must be
made avai l abl e at affordabl e costs. As these technol -
ogi es are adopted, energy use-especi al l y el ectri ci ty
wi l l i ncrease rapi dl y.
A persons power output and energy effi ci ency
are l ow. The basal metabol i sm of a person i s about
100 watts; for each uni t of work output, an addi ti onal
4 to 5 uni ts of food energy must be consumed.
84
Worki ng 8 hours a day at a rate of 50 watts of output,
a person consumes about 15 megajoul es (MJ) of
energy and produces 1.5 MJ of work output, for a
dai l y (24-hour) average effi ci ency of 10 percent.
85
Much of the l abor expended i n devel opi ng coun-
tri es i s not di rectl y producti ve, but i s i nstead for
enabl i ng acti vi ti es-that i s, domesti c chores.
Haul i ng water from the vi l l age wel l can take 0.5 to
3 hours per househol d each day, wi th a correspond-
i ng energy i nput i n the form of food of 0.3 to 3.0
MJ.
86
The poorest househol ds must often go further
and thus have l ess avai l abl e ti me to haul water,
resul ti ng i n much l ower water usage even wi th
greater effort (see tabl e 3-13). Water coul d i nstead
be pumped by a motor and pi ped to the home usi ng
just 3 to 5 percent as much energy .87 For el ectri ci ty
pri ced at $0.10 per kWh, the di rect energy cost for
83Mogem H. Fog and~shore L.IWUUEi@ EnergyE@a.en~ andFuel Substitution in the CementIndus~ Wi#EWhasis on D~eloping Countries
(Washington, DC: World Banlq 1983), see figure 5-1, p. 39.
S4w. ~mdmq ~fl= R~~~h Group, ~t~ti o~ Devel opment Re~~h cen~, ottsw~ _ 1S Them a Vicious CyCle Of h3W F@
Energy Intake and Low Human Output? July 1984 (Mimeo); ChristopherHum4 Energy Research Group, International Development Researeh Center,
Ottawz Canada, Human and Animal Energy in Transition: The Changing Role of Metabolized Energy in Economic Development June 19S4
(Mimeo); Roger Revelle, EnergyUseInRural India, Science, vol. 192, June 4, 1976, pp. %9-975.
gsFor a co~~r~, see G.M.O. Maloiy et d.,
Energetic Cost of Carrying bads: Have African Women Discovered an Economic Way? Nature,
vol. 319, Feb. 20, 1986, pp. 668-669.
86~ ~ s~dy of Gu@t, ~fiq tie time rw~~ to fe~h water was found to vary ffom 0.5 hour to more * 3 ho~ per day, wi~ ~ en= we of
100 to 800 kcal/day. Household washing takes 4.5 to 6.3 hours per week and is as strenuous as hauling water. Girja Sharan (etI.), Energy Use in Rural
Gujurat, op. cit., footnote 37.
871t ~ often ~~ tithe sw~ ~taaction provid~by activities such ss fomging for fielwood, ti~ water, grti @.id@, ~d OtherS h SIl
important element of village life and should not be tampered with naively. One notes, however, that village women spend 10 to 12 hours per day in such
activities. Surely they would not object to such social interaction while having a leisurely eup of tea instead.
Table 3-13-Average Daily Household Consumption
of Water, Gujurat, India
Consumption
class (liters per day)
Landless . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Less than 2 ha.. . . . . . . . . . . . . . . . . . . . . . . 126
Less than 2-4 ha.. . . . . . . . . . . . . . . . . . . . . 134
Less than 4-10 ha.. . . . . . . . . . . . . . . . . . . . 161
More than 10 ha. . . . . . . . . . . . . . . . . . . . . . 256
SOURCE: Girja Sharan (cd.), Energy Usein Rura/ Gu@t(New DelhL
Oxford and BH Publishing Co., 1987).
the typi cal 1.5 hours spent haul i ng water woul d be
just one-fi fth of a penny ($0.002). Thus, l ack of
access to capi tal has si gni fi cant i mpacts on l abor and
energy use.
Si mi l arl y, i n Afri ca, to pound mai ze or mi l l et by
hand can take 1 to 2 person-hours per day per
househol d.
88
Thi s requi res perhaps 1 MJ of energy
(at 300 watts of i nput). A typi cal motor-dri ven mi l l
can do the same job i n a mi nute or l ess, wi th an
energy expendi ture of l ess than 0.2 MJ--or 0.05
kWh. Thi s i s l ess than one-hal f of a penny ($0.005)
worth of di rect energy (at $0.10 per kWh) for 1 to 2
hours worth of hard l abor. The capi tal costs i n these
cases, of course, are a seri ous barri er to i nvestment;
but wi th the ti me saved, the person mi ght have done
somethi ng more producti ve, such as make handi -
crafts for market.
The advantages of mechani cal processi ng of
grai ns has l ed to a rapi d transi ti on i n many parts of
the worl d. I n Java, I ndonesi a, for exampl e, the
fracti on of ri ce processed by hand dropped from
perhaps 80 percent to l ess than 40 percent between
1971 and 1973.
89
Thi s freed many women from the
chore of gri ndi ng grai n; i t al so cost many of the
poorest househol ds an i mportant source of i ncome
earned by hand poundi ng ri ce for weal thi er house-
hol ds.
90
The i ntroducti on of mechani cal ri ce mi l l i ng
i n Bangl adesh i n the earl y 1980s was esti mated to
di spl ace an addi ti onal 100,000 or more poor women
per year from thei r tradi ti onal part-ti me empl oyment
at hand poundi ng ri ce. For the poorest, l andl ess
women, thi s represented roughl y hal f of thei r annual
i ncome and 15 percent of fami l y i ncome.
91
The power output and effi ci ency of draft ani mal s
are si mi l arl y l i mi ted i n performi ng typi cal farm
tasks. A typi cal 500-kg ox or buffal o has a basal
metabol i c rate of about 1,000 watts.
92
Average net
output over a 6-hour worki ng day i s typi cal l y 250
watts, and the net effi ci ency whi l e worki ng i s 29 to
39 percent, whi ch drops to about 10 percent over the
24-hour worki ng day. A typi cal draft ani mal mi ght
work just 40 days per year as many of the jobs
formerl y done by draft ani mal s-pumpi ng water,
crushi ng sugar cane, haul i ng goods to market-have
al ready been taken over by modern motor-dri ven
equi pment. At such a l ow rate of usage, the
effi ci ency of a draft ani mal i s 2 percent or l ess on an
annual basi s. These effi ci enci es are rai sed somewhat
when the val ue of the ani mal s dung, mi l k, meat, and
l eather i s i ncl uded.
93
The l ow power output and effi ci ency of a draft
ani mal severel y restri cts the potenti al work that can
gsfi~hu P@@, Yves Bigo~ ~d Hans P. Binswanger, Agricultural Mechanization and the Evolution of Farming systems in Sub-i$aharan ~ca
(Baltimore and Londom Johns Hopkins University Press for the World BanlG 1987); Mead T. C@ The Economic Activities of Children in a Village
in Bangladeak Population andDevelopmentReview, vol. 3, No. 3, September 1977, pp. 201-227; A.S. B- ChoosingTeehniques: Handpounding
V. Machine-Milling of Rice: An Indian Case, O#ord Economic Papers, vol. 17, No. 1, March 1965, pp. 147-157; Margaret Haswea Energy for
Subsistence (Lmdorx MacMillan Press, Ltd., 1981).
S9C. Pet~T~~,choi& of f&hniwe ~M~_ onJ av~ ml K. Eicherand J o~ M. S-Q (~.),Ag~c~~aZD~e/op~~ in the Third
World (Baltimore, MD: Johns Hopkins University Press, 1984), pp. 278-288. See also A.S. B- Choosing Techniques: Handpounding V.
Machine-Miliing of Rice: An Indian Case, op. cit., footnote 88.
~illiamL. Collier et al., A Commen6 in Carl K. Eicher and John M. Staatz (eds.), Agricuhura/ Development in the Third WorM (Baltimore,
MD: Johns Hopkins University Press, 1984).
91Glo~L. Scottand ww~, WOrldB@ CThe ~actof 1~~olo~choice OnR~ women inBang~&~ problems ati oppo~ties,
Staff Working Paper No. 731, 1985.
~Pe@r~~~U~dAnthonySmi@ ABetterB~tof B~~ NeWS~entist, Apr. 21, 1988, pp. 49-53. oxen~d~~~~ zjoul ~ ~l nl Ck!r
travekxlperkg of body weight (2 J/m&g). A . srumal weighing 500 kg and wdki.ng at 1 m/s will use an extra kW, approximately doubling its resting
metabolic rate. Most agricultural animals move at 0.6 to 1.1 m/s. . Aurnala use more energy for carrying loads than they do for carrying their own weight
ranging from 2.6 to 4.2 J/m/kg. See alao A.R. Rae, Bioenergetics of Bullock Power,Energy, vol. 9, No. 6, 1984; N.H. Ravindranathet al., AnIndian
Village Agricultural Ikosyste~Study of Ungra Village. Part I. Main Observations, op. ci~, footnote 5; Amulya Kumar N. Reddy, AnIndian
VillageAgrieuitural EcosystemCase Study ofUngraVillage. PartII. Discussion op. cit., footnote5. They estimstetheefficieneyof anhdianbulloek
as 8.7 pereent when working full time, or if working just 20 days per year as observed, the bullock would have an overall efficiency of 0.5 pereent.
93N.H. Rafi&~th~d H.N. q%Bioms Based E!ZUXgy System fora South Indian Village, Biomass, vol. 9, No. 3, 1986, pp. 215-233.
Draft animal efficiency is 3.5 Perce@ including nitrogen in manure for fertiker. Without nitroge~ the efficiency is 2.0 pereent.
Table 3-14-industrial Electricity End Use in Brazil, 1984
Percent of total
Fraction of subsector total for each end use (percent)
industrial electricity Process Direct Electro-
Industry consumption Motor heat heat chemical Light Other
Nonferrous metals . . . . . . . . . . . . . . . . . . . .
Iron and steel . . . . . . . . . . . . . . . . . . . . . . . .
Chemicals . . . . . . . . . . . . . . . . . . . . . . . . . . .
Food and beverage . . . . . . . . . . . . . . . . . . .
Paper and pulp . ....... . . . . . . . . . . . . . . .
Mining/pelletization . . . . . . . . . . . . . . . . . . .
Textiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Steel alloys . . . . . . . . . . . . . . . . . . . . . . . . . .
Ceramics . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cement . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Total
a
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
20.9
12.4
11.9
9.0
6.5
5.6
5.3
4.8
3.9
2.7
17.0
100.0
32
1
79
6
87
50
89
7
65
91
76
49
1
5
78
8
.
4
2
10
35
98
4
16
2
49
1
92
34
6
16
32
32
.
.
1
1
3
1
3
1
5
1
1
3
5
2
3
1
.
1
1

a Total industrial electricity use was 105 terawatthours.
SOURCE: Howard S. Geller, Electricity Conservation in Brazil: Status Report and Analysis, contractor report prepared for the Office of Technology
Assessment, March 1990.
be done. To i rri gate a l -hectare ri ce crop, for
exampl e, requi res the work output of two bul l ocks,
whi ch i n turn requi re the fodder produced from 2
hectares of crop.
94
By hi msel f, the i ndi vi dual farmer
coul d not, however, pump thi s much water by hand
i n an enti r e year .
Modern Dri ve Technol ogi es
El ectri c motor dri ve consumes an esti mated 58 to
68 percent of the el ectri ci ty used i n the Uni ted
States, and an even hi gher percentage i n the i ndus-
tri al sector al one.
95
Motor dri ve i s si mi l arl y i mpor-
tant i n devel opi ng countri es (see tabl es 3-14 and
3-15). El ectri c motors are the workhorses of modern
i ndustri al soci ety. They run home refri gerators;
dri ve offi ce ai r condi ti oners; power i ndustri al pumps,
fans, and compressors; and keep ci ty water suppl i es
fl owi ng.
The effi ci ency, conveni ence, and hi gh degree of
control of el ectri c motors provi de dramati c effi -
ci ency and producti vi ty i mprovements i n i ndustry.
%
Thi s l ed to a rapi d transi ti on i n the i ndustri al i zed
countries from water- and steam-powered dri ve to
el ectri c dri ve i n the earl y 1900s;
97
and the el ectri c-
i ty i ntensi ty of i ndustry conti nues to i ncrease today
i n i ndustri al i zed as wel l as devel opi ng countri es.
The effi ci ency of el ectri c motors i s general l y
fai rl y hi gh i n the i ndustri al i zed countri es, but can be
si gni fi cantl y l ower i n devel opi ng countri es due to
the use of l ower qual i ty materi al s for constructi on
and i mproper techni ques for mai ntenance, repai r,
and rewi nd.
98
Fi gure 3-16 compares the effi ci ency of
el ectri c motors i n Brazi l , I ndi a, and the Uni ted
States.
Hi gher effi ci ency motors are someti mes readi l y
avai l abl e i n devel opi ng countri es but cannot be used
because of the poor qual i ty of the el ectri c power
avai l abl e. I n Brazi l , for exampl e, the l argest manu-
facturer of smal l motors exports more effi ci ent
model s than those sol d at home. These hi gh-
effi ci ency motors
99
cannot be used i n Brazi l due to
the excessi ve vari ati on i n the power l i ne vol tage.
%eoffrey Barnard and Lars Kristoffersou Agricultural Residues as Fuel in the Third World (London: Earthscanj 1985).
qss~uel F. Baldw@ Energy-Efticient Eleetric Motor Drive Systems,
in Electricity: E@cient E&-Use and New Generation Technologies and
Their Planning Zmp2ications (Lun& Swedem Lund University Press, 1989).
MSsmuelF.B~dw@ me Materia Revolution and 13nergy-~lcient Electric Motor Drive SyStCmS,
Annual Review ofEnergy, vol. 13,1988, pp.
67-94; W.D. Devine, Jr., Historical Perspective on Electrification in Manufacturing, S. Schurr and S. Sonenblum (e&.), Electricio Use: Productive
E~ciency andEconomic Growth (Palo Alto, CA: Electric Power Research Institute, 1986).
~S~uel F. Baldw@ The Materials Revolution and Energy Eft3cient Motor Drive SySt-,
Annual Review of Energy, vol. 13, 1988, p. 67-94.
9$S~UelF. B~dwinsnd~e Finlay, Princeton TJniversity, Center for Energy and fivironnlental StUdieS,
Enexgy-EfflcientElectric Motor Drive
Systems: A Field Study of the Jamaican Sugar Industry, working paper, No. 94, February 1988. In particular, when motors are rewound they are
sometimes simply put onanopenfn to burn the insulation off the windings rather than in temperature-controlled ovens, This can damage the insulation
between the core laminations and lead to greater losses.
99Ihe efllciency of these motors is equivalent to the standard efilciency in the industrialized countries.
Table 3-15-Projected Electricity Consumption in India by Sector and End Use, 1990
(percent of total national electricity use)
Industrial process Space conditioning
Motor Process Cooling/
Appliances
Sector Total
a
drive Electrolysis heat Lighting ventilation Heating Refrigeration Other Miscellaneous
Residential . . . . . . . . . . . . . . . . . . 13.0 4.2 3.5 1.5 1.0 2.9
Urban . . . . . . . . . . . . . . . . . . . .
10.4 2.9 2.9 1.2 1.0 2.4

Rural . . . . . . . . . . . . . . . . . . . . . 2.6
1.3 0.5 0.3 0.5

Commercial . . . . . . . . . . . . . . . . .
11.2 4.8 1.6 1.5 0.4 0.8 2.1

Agriculture . . . . . . . . . . . . . . . . . . 18.4 18.4
Industrial . . . . . . . . . . . . . . . . . . . . 54.8

33.4 10.8
5.5 5.1
Primary metals
b
... . . . . . . . . .

17.2

6.4 6.9 3.0 0.9
Chemicals . . . . . . . . . . . . . . . .

13.8

8.8 3.6 0.1 1.3
Textiles . . . . . . . . . . . . . . . . . . .

10.2 7.8

0.4 2.1
Coal, cement . . . . . . . . . . . . . . 6.8

5.8

0.5 0.4
Secondary metals
c
. . . . . . . . . 3.4

1.5 0.2 1.4 0.2
Paper .. .. ... ... ... .... . . 3.4

3.0

0.1 0.3
Railway traction . . . . . . . . . . . . . . 2.6

2.6

Total . . . . . . . . . . . . . . . . . . . . . . . 100.0 54.4 10.8 5.5 14.5 5.1 1.5 1.9 1.8 5.0
Motor drive . . . . . . . . . . . . . . . . . . 61.4 54.4 5.1
1.9
Not available or nonapplicable.
a Total national consumption is projected to be 249.1 terawatthours in 1990.
b Aluminum, nonferrous, iron, and steel.
C lron andsteel.
SOURCE: Ahmad Faruqui, Greg Wilder, and Susan Shaffer, Application of Demand-Side Management (DSM) To Relieve Electricity Shortages in India, contractor report prepared for the Office
of Technology Assessment, April 1990.
Figure 3-16-Efficiency of Electric Motors in the United States, Brazil, and India
Motor efficiency, percent
95
90
85
80
75
70
65
I
d
A
I I I I I I I I I
I
1 10
Electric motor

Brazi l standard eff. + Brazi l hi gh
100
size, HP
eff.
~ India manufacturer
~ India field test
+
USA hi gh ef f .
This figure shows the efficiency for motors in Brazil, India, and (high-efficiency only) the United States.
Note the large difference in motor efficiency as measured in field tests and as cited by manufacturers
in India.
SOURCES: Un/tsd Btates: John C. Andreas, Energy-EMcient E/ectric Motors (New York, NY: Marcel Dekker, 1982).
Brezik Howard S. Geller, ElectricityConservation in Brazil: Status Report and Anaiysis,contractor report
prepared for the Office of Technology Assessment, March 1990; Irxfk: S. Anand, and V.S. Kothari,
Characterization of Electric Motors in Industry and Enetgy Conservation Potential in India (New Delhi,
India: Tata Energy Research Institute, no date).
Thi s fi rm has al so devel oped motors wi th effi ci en-
ci es comparabl e to the hi ghest performance motors
i n i ndustri al i zed countri es.
l 00
Al though the effi ci ency of el ectri c motors them-
sel ves can be qui te hi gh, the effi ci ency of the overal l
system i s general l y l ow. For exampl e, the conver-
si on of coal to el ectri ci ty typi cal l y resul ts i n the l oss
of two-thi rds of the i nput coal energy. There are
addi ti onal l osses throughout the system, wi th the
resul ti ng net output as l ow as 5 percent of the i nput
energy (see fi gure 3-17). Si gni fi cant energy savi ngs
are possi bl e through the use of better technol ogi es
and better control strategi es throughout the system.
Barriers to Efficiency Improvements
A number of factors l i mi t the effi ci ency, produc-
ti vi ty, and performance of i ndustri al operati ons i n
devel opi ng countri es: pl ants that are too smal l to be
effi ci ent; technol ogi es that are of l ow qual i ty and
often obsol ete; raw materi al s that are of l ow qual i ty;
i nadequate nati onal i nfrastructures; l ack of forei gn
exchange to purchase cri ti cal components not avai l -
abl e l ocal l y; and a l ack of ski l l ed techni ci ans,
engi neers, and managers.
The average U.S. paper mi l l , for exampl e, has an
annual capaci ty of 100,000 tons, whereas i n Lati n
Ameri ca, Afri ca, and Asi a (except Japan), the
average capaci ti es are 18,000,9,000, and 5,000 tons,
respecti vel y. These smal l er scal es can l ead to
si gni fi cant i neffi ci enci es. Studi es i ndi cate that a
paper mi l l wi th an annual capaci ty of 30,000 tons
can consume from 30 percent to as much as 100
percent more energy/steam respecti vel y per uni t
output than a mi l l wi th a capaci ty of 150,000 tons.
l O1
I n addi ti on, a vari ety of energy-conservi ng technol o-
gi es, such as waste heat recovery systems and
l~owmd S. ~ller, Electricity Conservation in Brazil: Status Report and -ysk,
contractor report prepared for the OftIce of Technology
Assessment, March 1990.
IOIA@ew J, Ew@, ~~~~lciency tithe IMp and Paper Industry with Emphasis on Developing Countries, World Bti techni~ PaPr, No.
M, Washin@oIL ~, 1985, p. 45.
Figure 3-17Energy Losses in an Example Electric
Motor-Driven Pumping System in the United States
Fraction of input coal energy remaining
1 0 0
80
r
6 0 -
40-
20-
- Generation (33%)
Transmission and distribution (90%)
~
Motor (91%)
/
Shaft coupling (98%)
Pump 77% Throttle valve 66%
/
L ,
L .
-Piping (
o i I 7 7 7 r
0 1 2 3 4 5 6 7 8
Point along the pumping system
b)
35%)
This figure shows the useful energy remaining at each stage of a
pumping system. The values in parentheses are the efficiencies of
the particular device at each stage.
SOURCE: Samuel F. Baldwin, Enwgy-E#kxent E/ectrk Motor Drive Sys-
tems, in Electricity: Efficient End-Use and New Gene@ion
T&nologies and Their Planning Implications (Lund, Sweden:
Lund University Press, 1989).
cogenerati on systems, become fi nanci al l y l ess at-
tr acti ve or even uneconomi cal at smal l er scal es.
The raw materi al s avai l abl e to i ndustri es i n
devel opi ng countri es are often of l ow qual i ty. For
exampl e, coal resources i n I ndi a are poor, provi di ng
bl ast-furnace coke wi th an ash content that typi cal l y
ranges from 21 to 27 percent. Thi s l owers the energy
effi ci ency of the steel maki ng process as wel l as
potenti al l y i nterferi ng wi th steel producti on.
102
I nadequate nati onal i nfrastructures al so reduce
effi ci ency and producti vi ty. Frequent el ectri c power
brownouts or bl ackouts are parti cul arl y damagi ng.
I n Ghana, for exampl e, the GI HOC Bri ck and Ti l e
Co. had 152 hours of el ectri ci ty outages i n 1986.
When an outage occurs, the fuel oi l feed to the ki l n
burners i s cut off and the fi re must be stoked wi th
wood. Thi s i s a haphazard process and si gni fi cantl y
reduces the qual i ty of the fri ed bri cks.
103
The l ack of forei gn exchange to buy spare parts
can al so be a seri ous handi cap. Thi s has been an
i mportant factor i n the decl i ne of the Tanzani an
cement i ndustry, whi ch operated at just 22 percent of
rated capaci ty i n 1984.
104
Assi stance may be useful at several l evel s. The
effi ci ency and producti vi ty of tradi ti onal rural i n-
dustri es mi ght be si gni fi cantl y i ncreased i n a cost-
effecti ve reamer wi th the i ntroducti on of a l i mi ted
set of modern technol ogi es and management tool s.
To do thi s, however, i s extremel y di ffi cul t due to the
smal l and scattered nature of tradi ti onal rural i ndus-
tri es and the l arge extensi on effort needed to reach
i t. Large i ndustry i n devel opi ng countri es has many
of the same needs-techni cal , manageri al , and
fi nanci al assi stance--but can be reached more
readi l y. I nternati onal ai d agenci es and a few non-
governmental organi zati ons are provi di ng such as-
si stance to the extent that thei r funds al l ow.
Dramati c i mprovements i n the energy effi ci ency
and producti vi ty of basi c materi al s processi ng tech-
nol ogi es are al so sti l l possi bl e-even beyond the
l evel s currentl y achi eved i n the i ndustri al i zed coun-
tri es. Research i s goi ng on i n thi s area, but much
mor e coul d be done. Much of the cur r ent r esear ch i s
focused on hi gher val ue-added speci al ty materi al s
and hi gh-grade fi ni shi ng rather than on pri mary
processi ng.
105
Devel opi ng countri es, however, have
l i ttl e capi tal to undertake the research needed to
i mprove thei r i ndustri al processes, and few i nterna-
ti onal agenci es support research of thi s ki nd.
The Traction (Agricultural) and
Transportation Sectors
Traction and Agriculture
Agri cul ture entai l s a seri es of operati ons: soi l
preparati on, sowi ng, weedi ng, harvesti ng, and post-
harvest storage. I n tradi ti onal agri cul ture, these
operati ons are performed by manual l abor wi th, i n
some cases, assi stance from ani mal s. As agri cul ture
i ncreases i n scal e and i s commerci al i zed, many of
these operati ons are done by machi nes.
Agri cul ture i s by far the l argest empl oyer and
goods-produci ng economi c sector i n the poorer
devel opi ng countri es. I n sub-Saharan Afri ca, for
exampl e, 75 percent of the work force i s engaged i n
lo?Bnergyancl Environmental &UdySiS, k.,
Conserv@ Process HeatinPrimary Industries of India and - contractor report prepared for the
~ce of Technology lkW3SSlne13~ April 1990.
~os~p~orld B* ~eW S=tor -gemmt ~sis~ce fio~ G-: EQergy ~o-tion in b III&w_@ Seetor, JWW 1988.
IWD.F. Stew@~B,M~@ StrategiesforMeetingThnzama
sFuture C!ementNeeds, NaturulResourcesFonun, November 1989, pp. 294-302.
IOSRO~ H. WilliamS, prince~n University, personal communication Feb. 1, 1989.
agriculture, compared to just 2 percent i n the Uni ted
States. Agri cul ture al so provi des a si gni fi cant frac-
ti on of GDP i n devel opi ng countri es--one-thi rd of
GDP for the nearl y 3 bi l l i on peopl e i n l ow-i ncome
countri es
106
(see tabl e 3-16).
Agri cul ture i n the l owest i ncome countri es i s
l argel y by smal l , fami l y farms usi ng human and
ani mal power and organi c ferti l i zer wi th l i ttl e access
to or knowl edge of modern i nputs such as chemi cal
ferti l i zers, hybri d seeds, or mechani cal dri ve. Ethi o-
pi ans, for exampl e, use on average just 4 kg of
chemi cal ferti l i zer per hectare of cropl and, whi l e the
Engl i sh use 368 (see tabl e 3-16). Low soi l ferti l i ty
and i nadequate or i rregul ar rai nfal l sharpl y l i mi t the
producti vi ty of l ow-i nput farms i n devel opi ng coun-
tr i es.
There i s a general trend toward l arger farms,
, and greater use of commer- gr eater mechani zati on
ci al i nputs i n many devel opi ng countri es, resul ti ng
i n greater producti vi ty but at the cost of greater
di rect and i ndi rect energy i nputs. I ndi a, for exampl e,
nearl y doubl ed i ts i rri gated area between 1950 and
1984 i n order to reduce i ts vul nerabi l i ty to poor
monsoons.
107
I ncreased i rri gati on and use of hi gh-
yi el d vari ety crops have contri buted to i ncreases i n
both absol ute and per-capi ta agri cul tural produc-
ti on.
108
Chi na has si mi l arl y moved toward greater mecha-
ni zati on and use of modern i nputs. Agri cul ture i n
Chi na i s sharpl y constrai ned by l and avai l abi l i ty
onl y about 10 percent of the l and can support
crops-yet per-capi ta producti on i ncreased by 18
percent from 1979 to 1983 wi th l i ttl e i ncrease i n
cul ti vated area.
l 09
I mproved water control and
di stri buti on, i ncreased use of tractors and ferti l i zers,
and the adopti on of new crop vari eti es contri buted to
thi s achi evement.
110
Tradi ti onal Shi fti ng Agri cul ture
Tradi ti onal shi fti ng agri cul ture begi ns wi th forest-
fal l ow systems, i n whi ch pl ots of forest l and are
cl eared and cul ti vated for a few years and then l eft
fal l ow for 20 year s or mor e. Cl ear i ng by fi r e r equi r es
l i ttl e l abor, and stumps are l eft for rapi d regrowth
duri ng fal l ow. Because the ground underneath tree
cover i s soft, no further l abor i s requi red before
sowi ng, and because the forest cover has l ong
suppressed weeds, few seeds remai n and l i ttl e
weedi ng i s needed. Such burni ng does, however,
effecti vel y l ead to very l arge agri cul tural energy
i ntensi ti es due to the l arge amount of forest cover
that i s burned off.
l l l
Wi th i ncreasi ng popul ati on densi ty, the fal l ow
peri od becomes shorter. As a resul t, regrowth duri ng
the fal l ow peri od i s reduced to bush, and fi nal l y to
grass. Si nce fi re does not ki l l roots, extensi ve hoei ng
and weedi ng become necessary. I nputs of organi c
ferti l i zer are needed to mai ntai n soi l ferti l i ty, and
there i s a shi ft from si mpl e addi ti on of organi c
materi al to more compl ex comporti ng and manuri ng
techni ques. Further i ncreases i n popul ati on l ead to
annual cul ti vati on and eventual l y mul ti pl e croppi ng.
(I n the humi d tropi cs, however, soi l s tend to be poor
and easi l y eroded and l eached, and the potenti al for
conti nuous cul ti vati on i s l i mi ted.
112
) As the need for
hoei ng and weedi ng i ncreases i t becomes advanta-
geous to go to the extra effort of destumpi ng the l and
and obtai ni ng, trai ni ng, and mai ntai ni ng ani mal s or
mechani cal agri cul tural technol ogy.
113
There are a number of potenti al advantages
associ ated wi th the use of ani mal or mechani cal
tracti on for agri cul ture. Properl y done, ti l l age i m-
proves the condi ti on of the soi l for crop growth
i ncreasi ng porosi ty, aerati on, root penetrati on, and
water i nfi l trati on whi l e reduci ng evaporati on. Ex-
l~world B@ Wor&f Development Repo~ 1990 (lNew York NY: Oxford university -ss, 19$0.
107~ ~e~ Re~h ~titute, T.EMEnergy Data Directo~ and Yearbook (T~DY) 1988, Op. Cit., fOOtiOte 56, P. 128.
108Fo~ ~ Agri~@e (lrg-tion of the United Nations,
The State of Food and Agriculture, 1984, Rome, Italy, 1985, p. 137.
l@Ibid.
11OC. Howe, Chinas Economy (New YorlL NY: Basic Books, kc., 1978).
111A. T- RUIIIW, why Shifting Cultivators Keep Shifting: Understanding Farmer
Deeision-Making in Traditional Agmfor@ry Systems,
Community Forestry: Some Aspects, UNDP THA/81~ (Bangko~ Tbadan d: Environment and Policy Institute, East-West Center, Honolulw and
UNFAO Regional OffIce for Asia and the Paeif3c, 1984).
llz~abhu w, Yves Bigot ~d ~ p. Binswqer, Agncu[~raIMec~~mtion ad the Evol~on of Fu~ng sytie~ in Sub-Saharan A&ca,
op. cit., footnote 88.
llq~bhuping~i et& ibid. me e=ct ~itionpo~t fim~d to m ~d then~ctor t~ology@, of co-, depend olln~
us factors,
including how difficult the soil is to work the vslue of @ me@ hides, and other serviees provided by cattle; the use of manure; the cost of tmining
-~
and main ; the cost of destumping and otherwise preparing fields and weeding; the length of time that animals w be * the risk of
disease such as trypauosomiasis (transmitted by the tsetse fly); and many others.
Table 3-16--Agricultural Indicators for Selected Countries
GNP/Cap Agricultural GDP as Agricultural employment as Percent of farms larger Fertilizer
Crop yields (kg/ha)
Country (1987) percent of total GDP percent of total employment than 5 ha (kg/ha) Cereal Roots/tubers
Ethiopia. . . . . . . . . . . . . . . 130 45
4 4 1,081 2,827
Zambia . . . . . . . . . . . . . . . 250 13

14 1,747 3,687
India . . . . . . . . . . . . . . . . . 300 30 61 9 43 1,590 14,268
China . . . . . . . . . . . . . . . . . 290 26 57
176 3,891 15,614

Brazil . . . . . . . . . . . . . . . . . 2,020 10 36 63 35 1,719 12,072
UK ...0 . . . . . . . . . . . . . . . 10,420 2 2 83 368 6,081 36,072
U.S.A. . . . . . . . . . . . . . . . . 18,530 2 2 90 101 4,618 31,215
SOURCES: World Bank Wbdd Deve/oprnent F@or_t 1989(New York, NY: Oxford University Press, 19S9).
Food and Agriculture Organization of the United Nations, The State of Food andAgtiu/ture 1984 (Rome: FAO, 1965), pp. 163-165, 17S-160.
Tata Energy Research Institute, TER/ Energy Data Dhctoryand Yearbook 7988(N6w Delhi, India: Tata Energy Research Institute, 1969), p. 123.
World Resources Institute, WwkfResources f988-89(New Yo~ NY: Oxford University Press, 19S6), pp. 272-277.
Box 3-AThe One-Ox Plow
1
Many farmers i n devel opi ng countri es are unabl e to support the two draft ani mal s needed to pul l a tr adi ti onal
pl ow. Al though hal f of the househol ds i n Bangl adesh keep cattl e, onl y a quarter have two or more. I n Ethi opi a, onl y
a thi rd of the farmers own two draft ani mal s-and many of these are l ost i n the peri odi c droughts. At peak cul ti vati on
ti mes, these farmers must then rent or borrow a second ani mal and maybe del ayed i n pl anti ng thei r crops, whi ch
depend cri ti cal l y on catchi ng the sparse and i rregul ar rains in a ti mel y manner-both for maki ng maxi mum use of
the ni trogen rel eased wi th the first rains (see ch. 2) and for reachi ng maturi ty wi th the l ast rai ns.
Researchers at the I nternati onal Li vestock Center for Afri ca (I LCA) i n Addi s Ababa, Ethi opi a, responded to
thi s si tuati on by redesi gni ng the traditional doubl e yoke for a si ngl e ox: experi ments showed that one ox coul d pul l
wi th 70 per cent of the for ce of two.
I n the fi el d, however, Ethi opi an farmers qui ckl y converted the one-ox pl ows back i nto the tradi ti onal two-ox
form. On exami nati on , researchers found that the tradi ti onal two-ox form had a number of advantages. Farm oxen
were not as wel l fed nor as strong as those whi ch had been tested at the I LCA headquarters and coul d not pul l as
hard; and two oxen were abl e to steady each other when one stumbled. The ri gi d coupl i ng of the tradi ti onal two-ox
pl ow al so enabl ed the farmer to steer the oxen and to shift some of the wei ght of the pl ow to the oxen during a turn.
In contrast, the si ngl e ox yoke used a fl exi bl e rope harness whi ch reduced the farmers abi l i ty to steer the ani mal s
and forced him to carry the ful l wei ght of the pl ow when turni ng. The one ox pl ow al so had a skid to regulate the
depth of the cut: it broke easi l y but coul d not be repai red by the farmers themsel ves.
Further, where the qual i ty of the feed is very poor-a common si tuati on i n many tropi cal areas-worki ng
ani mal s ar e unabl e to compensate for thei r energy expenditure by eati ng more and consequentl y l ose wei ght. A
worki ng ani mal al so has a 10 percent hi gher basal metabol i c rate than a nonworki ng ani mal -requi ri ng more food
just for mai ntenance. I n thi s case, i t may be better to use two oxen to do what l i ttl e they can wi thout l osi ng too
much wei ght r ather than to have one ox whi ch soon becomes exhausted beyond recovery.
As one researcher at the I LCA noted, I t mi ght have occurred to us that i f Ethi opi an farmers hadnt i nvented
somethi ng as si mpl e as the one-ox pl ow i n 3,000 years of agri cul ture, they probabl y had reasons.
Some have si mi l arl y thought that the same ani mal mi ght be used to provi de both l abor and mi l k. Experi ments
i n Costa Ri ca showed that cows coul d, i n fact, provi de bothi f fed adequatel y. Tropi cal pastures, however, are not
adequate. To provi de the ani mal a suffi ci ent di et for such a hi gh rate of energy expendi ture requi red concentrated
feed suppl ements such as grai n. Thi s coul d create a direct confl i ct over food between draft cows and peopl e i n many
parts of the worl d.
l l XbontMaCKenzi e, Ethiopiaswdto the plough, New Scientist, Oct. 1, 1987, pp. 52-55; Peter Lawrence and Anthony Stit4 A
Better Beast of Burde~ New Scientist, Apr. 21,1988, pp. 49-53; A.K.M. Abdul Quaderand K. Ikhtyar Ornar, Commonwealth Science Council,
Resources and Energy Potentials in Rural Banglade@ technical publication series No. 191, Imndo~ 1986.
peri ments show that yi el ds can be i ncreased by Peasant farmers have responded to thei r often
pl owi ng.
114
I n practice, however, little increase is
di ffi cul t ci rcumstances i n vari ed waysboth l ogi cal
observed as farmers tend i nstead to focus on
(see box 3-A) and frequentl y i ngeni ous. For exam-
i ncreasi ng cul ti vated area
115
or on saving l abor,
pi e, around 1925-1930, ani mal tracti on began to be
rather than i mprovi ng the qual i ty of thei r ti l l age. I n
used i n northwestern and central Senegal and
West Afri ca, the soi l s are so hard they often cannot
northern Ni geri a for peanut cul ti vati on. The l i ght,
be pl owed (wi thout damage to equi pment) unti l the
sandy soi l s of Senegal do not require pl owi ng, and
as the growi ng season i s so short, rapi d pl anti ng of
rai ns begi n, but then any del ay reduces the avai l abl e
peanuts whi l e the soi l i s moi st i s essenti al . Conse-
growi ng ti me and ri sks a shortage of water when
quentl y, seeders are used by the peasants so that
pl ants reach maturi ty.
116
l arger areas can be cul ti vated wi thi n the avai l abl e
Ildprabhu ping~i, yves Bigot, and Hans P. Binswimger, Agncultural Mechanization and the Evolution ofFarmang SySte?nS in Sub-Saharan&Iea,
op. cit., footnote 88; Peter Munzinger, Anznud Traction in Afica (l%chbom, West G~y: GTZ, 1987), p. 279.
lls~seneg~ tie av
erage exp
ansi
on
ofagric~~ area by the introduction of draft animals to smallholders is 1(K) to 160percenG in W the av~ge
expansion is 150 to 200 percent. Peter Munzinger, Animal Traction in Aflca, op. ci~ footnote 114, p. 287.
116~bhu Pingali et al., op. Cit., footnote 88.
ti me. Horses are used i nstead of oxen, si nce the
greater power of oxen i s not needed (there i s no
pl owi ng) and horses are faster, further i ncreasi ng the
pl anted area. I n Ni geri a, where peanuts are grown i n
mi d-sl ope regi ons on soi l s hi ghl y suscepti bl e to
erosi on, ox-drawn ri dgers are used to control the
er osi on.
117
Modern Commerci al Agri cul ture
As popul ati on grows agri cul tural producti vi ti es
must be rai sed. Modern i nputs are needed to achi eve
thi s. Commerci al fuel i nputs to agri cul ture i ncl ude
mechani zed l and preparati on, mechani zed i rri ga-
ti on, and syntheti c ferti l i zers.
118
The degree of mechani zati on vari es wi del y, but
general l y i ncreases wi th per-capi ta i ncome. Mecha-
ni zati on reduces the ti me and l abor needed for
prepari ng, pl anti ng, and harvesti ng crops. I n favora-
bl e areas, i t al so ai ds doubl e croppi ng. The tractors
themsel ves come i n many formsi n Chi na the most
popul ar i s probabl y the Worker-Peasant, a 7-hp
garden tractor. I n I ndi a, where the number of tractors
al most doubl ed from 1972 to 1977,
119
the most
popul ar i s a 30-hp di esel .
I rri gati on i s most commonl y done wi th ei ther
el ectri c motor or di esel dri ven pumps. El ectri c
pumps are qui te rel i abl e (al though subject to i nter-
rupti ons i n the el ectri c power gri d) and conveni ent,
and are often the l owest cost al ternati ve. Di esel -
el ectri c pumpi ng systems, i n whi ch di esel generators
produce el ectri ci ty that i s then used to dri ve el ectri c
pumps, and di rect di esel and gasol i ne-powered
pumps are more often used where no el ectri c gri d i s
avai l abl e. These are much l ess mechani cal l y de-
pendabl e than el ectri c pumps.
I n Chi na, i rri gati on i s a si gni fi cant consumer of
el ectri ci ty. I t i s esti mated that 70 percent of the
el ectri ci ty consumed i n rural areas i s for i rri gati on,
wi th the remai nder used for food processi ng, vari ous
rural i ndustri es, and l i ghti ng.
120
I n I ndi a, the number
of el ectri c pump sets for i rri gati on has grown rapi dl y
(see fi gure 3-18), and the el ectri ci ty consumpti on for
these pump sets has gone from 4,470 GWh i n
Figure 3-18-Use of Agricultural Pumpsets in India,
1950-1990
Number of pumps by type (millions)
7
I
1950 1960 1970 1980 1990

El ect ri c pump set s + Diesel pump sets
+ Animal-powered pump
SOURCE: Tata Energy Research Institute, TER/ EiieWy Data Directory
and Yearbook (TE/2DYjl 7988 (New Delhi, India: 1989), footnote
56.
1970-71 to 23,420 GWh i n 1985-86. The number of
di esel pump sets has al so grown, but they sti l l are
fewer i n number than the el ectri c uni ts.
121
Transportation
The transportati on sector accounts for a quarter or
more of total commerci al energy use i n most
devel opi ng countri es-I ndi a and Chi na bei ng the
most notabl e excepti ons (see tabl e 3-l ). Most of thi s
transport energy i s from oi l . Energy use for transpor-
tati on i n the devel opi ng worl d i s expected to grow
rapi dl y i n the future, as i ncreasi ng urbani zati on and
i ncomes (see fi gure 3-19) l ead to i ncreased demand
for transportati on servi ces. Thi s wi l l i ncrease the
outfl ow of scarce forei gn exchange for the oi l -
i mporti ng countri es, and wi l l al so requi re consi dera-
bl e i nvestment i n roads and rel ated i nfrastructure.
Transportati on can be provi ded by ai r, rai l , road,
or water. I nmost of the devel opi ng worl d, as wel l as
i n the i ndustri al i zed worl d, road technol ogi es pro-
vi de most transport servi ces. Notabl e excepti ons are
I ndi a and Chi na, whi ch have l arge rai l networks.
I n rural and poor areas of the devel opi ng worl d,
wal ki ng i s the pri nci pal transport technol ogy.
The advantages of wal ki ng are many-i t requi res no
capi tal i nvestment, i t i s not restri cted to roads, and
lly~ido
118IMs excludes emrgy used for ferlher
production and that used for crop preparation both of which we include under industrial energy use.
11- _ ~-h Institute, TEDDY 1988, op. cit., footnote 56, p. 137.
1%. Howe, Chims Economy, op. cit., footnote 110, p. **.
121~ 1983~, fore-le, Mm we= about 5.3 fion electric p~pSe& ~d 3.1 won ~esel pump sets ill hldi~ SOIIM Of thCSC dieSCl UllitS WerC
used as backups for the electric units. lMaEnergy Research Institute, T~DY1988, op. cit, footnote 56, p. 135.
Figure 3-19-Passenger and Freight Transport v. GNP for Selected Countries, 1960-81
Passenger traffic and GNP per capita
10,000
r
1
I I I
I I I 1
I
I 1 I
I I 1 1
U.S.A.
/B
dia
//
Japan
1
1
4
t 1 I I I 1 1
100 1,000 10,000
GNP per capita (1979-1981 constant use)
Total freight and GNP for selected countries, 1980-1981
1,000
100
10
1-
1 I 1
1 I I
U.S.S.R.
I I I 1 1 1 , I I I I I I 1 1
10 100 1,000
I
GNP (BIL.US$)
This figure shows how passenger and freight transport energy use have increased with GNP for seven countries. The individual data points
are for specific years.
SOURCE: J. Venny and L. Uy, Transport in China, World Bank staff working paper, No. 723, Washington, DC, 1985.
i t requi res no fossi l fuel s. On the other hand, i t i s
sl ow, ti ri ng, and requi res energy i n the form of food.
Ani mal technol ogi es, such as bul l ock carts, are
someti mes faster, have a much greater frei ght
capaci ty, and i nvol ve l ess work for peopl e. Capi tal
i nvestment i n the bul l ock and cart i s requi red,
however, as wel l as an operati onal cost for feed.
These two technol ogi es-wal ki ng and domesti cated
ani mal s-are the pri nci pal means of transport i n
many poorer and rural areas, parti cul arl y i n Afri ca
and Asi a.
Bi cycl es are a popul ar transport technol ogy,
especi al l y i n Chi na, where from 50 to 90 percent of
urban vehi cl e tri ps are made by thi s mode.
122
The
frost cost of a bi cycl e can be a barri er-a new bi ke
costs the equi val ent of 7 to 8 months wages i n
Tanzani a, for exampl e-but i n some areas bi cycl es
can be bought on credi t. I n I ndi a, government
empl oyees are enti tl ed to l oans for vehi cl e purchase,
whi ch can be used to buy a bi cycl e. Bi cycl es work
wel l i n congested urban areas, where they have some
advantages over pri vate automobi l es-they are eas-
i er to park and store, l ess expensi ve to own and
operate, and do not contri bute to ai r pol l uti on. Thei r
range and frei ght capaci ty, however, are l i mi ted.
The technol ogi cal l eap to the i nternal combusti on
engi ne al l ows for much hi gher speeds, l onger
di stances, l arger frei ght capaci ty, and greater com-
fort. The di sadvantages of the i nternal combusti on
engi ne are technol ogi cal compl exi ty, movement
l argel y constrai ned to roads, hi gh fi rst cost and
operati ng cost, and envi ronmental damage due to
fossi l fuel burni ng. There are al so secondary effects,
such as i njury and death due to acci dents and l and
use for roads and parki ng. Despi te i ts di sadvantages,
the i nternal combusti on engi ne i s the domi nant
transport technol ogy i n the i ndustri al i zed and devel -
opi ng worl d, and i ts use i s growi ng rapi dl y.
Passenger Road Transport Technol ogi es
Mechani zed passenger road transport i n the devel -
opi ng worl d i s performed by a wi de range of
technol ogi es, i ncl udi ng mopeds, pri vate autos, and
buses. Devel opi ng countri es have onl y about 1
percent as many autos per personas does the Uni ted
States (see tabl e 3-17), but thei r automobi l e fl eets
are growi ng rapi dl y. Further, the scrappage rate (the
fracti on of vehi cl es reti red each year) i s very l ow i n
devel opi ng countri es, due to the hi gh val ue pl aced
Table 3-17Passenger Fleet Size and Growth
Average annual
Automobiles per growth in automobile
Country 1,000 people, 1986 fleet size, 1982-86
China . . . . . . . . . . . . . .
India . . . . . . . . . . . . . .
Kenya . . . . . . . . . . . . .
Thailand . . . . . . . . . . . .
Brazil . . . . . . . . . . . . . .
Japan . . . . . . . . . . . . .
West Germany . . . . . .
United States . . . . . . .
0.7
1.8
8.9
21.9
87.0
234.0
444.8
673.4
41.6
8.2
3.2
8.8
8.9
3.0
3.3
2.4
SOURCES: Fleet size and growth from Energy and Environmental Anaiy-
sis, Inc., Poliq Options for Improving Transportation Energy
Efficiency in Developing Countries, contractor report pre-
pared for the Office of Technology Assessment, March 1990.
Population from World Bank, IWw/dDeve/oprnentl %porf 1989
(New York, NY: Oxford University Press, 1989). Figures forthe
United States include both autos and Iighttrucks; data arefmm
Oak Ridge National Laboratory, Transportation Energy Data
8ook, Edition 10, ORNL-6565 (Oak Ridge, TN: Oak Ridge
NationaJ Laboratory, September 1989).
on any vehi cl e that runs. Therefore, choi ces made
now as to the energy effi ci ency of new vehi cl es i n
devel opi ng countri es are doubl y i mportant-these
vehi cl es wi l l soon be the majori ty of the fl eet and
they wi l l be on the road a l ong ti me.
Al though thei r techni cal effi ci ency (vehi cl e ki l o-
meters travel ed per l i ter of fuel consumed) i s l ower,
vehi cl es i n devel opi ng countri es average a much
hi gher l oad factor (persons per vehi cl e) than those i n
i ndustri al i zed countri es. Buses are chroni cal l y over-
l oaded, and mopeds and motorcycl es desi gned for
one often carry two or more. Shared ri de technol o-
gi es, such as ji tneys, are commonl y fi l l ed beyond
rated capaci ty. Thi s i ncreases the effi ci ency of the
transportati on system i -n terms of passenger-
ki l ometers per l i ter of fuel consumed, but reduces
safety and comfort.
The vehi cl es themsel ves are often l ess energy-
effi ci ent than those found i n the i ndustri al i zed
worl d. They are often based on desi gns that empha-
si ze sturdi ness and dependabi l i ty under adverse
condi ti ons (poor roads, chroni c overl oadi ng, l i ttl e
mai ntenance) over energy effi ci ency.
Frei ght Road Transport Technol ogi es
Road frei ght movement i n the devel opi ng worl d
i s provi ded mostl y by di esel trucks (wi th the
excepti on of Chi na where much of the truck fl eet
uses gasol i ne) and these trucks account for over hal f
Izzworld B@ cGridlock Weary, Some lbrn to Pedal power,
The Urban Edge, vol. 14, No. 2, March 1990.
Table 3-18-Energy Efficiency of Trucks in Selected Countries
Energy consumption
Country/ Capacity (megajoules per metric ton
region Truck name (metric tons) per kilometer)
OECD Mercedes Benz 1217 (1979) 7.0 1.0
OECD Man-VW 9136 (1980) 5.9 1.0
India TATA 1201 SE/42 5.0 2.1
India Ashok Leyland Beaver 7.5 1.6
China Jiefang CA-10B 4.0 2.3
China Dongfeng EQ140 5.0 1.8
NOTE: OECD and Indian trucks use diesel, Chinese trucks use gasoline.
SOURCE: J. Yenny and L. Uy, World Bank, Transport in China, staff working paper No. 723, 19S5, p. 70.
the energy used for road transport i n the devel opi ng
wor l d.
123
The movement of hei ght i s requi red for
most economi c acti vi ty, and i n many devel opi ng
countri es the pri ces of di esel fuel are kept l ower than
gasol i ne pri ces. I n the Uni ted States, for exampl e,
gasol i ne and di esel pri ces at the pump are al most the
same, whi l e i n I ndi a di esel i s sl i ghtl y l ess than hal f
the pri ce of gasol i ne.
124
Trucks i n the devel opi ng worl d are rel ati vel y
i neffi ci ent, requi ri ng 1.5 to 2.5 ti mes as much energy
to move one ton of frei ght one ki l ometer as
comparabl e trucks i n the OECD countri es (see tabl e
3-18). I n devel opi ng countri es, however, trucks
must cope wi th more di ffi cul t operati ng condi ti ons:
the roads are typi cal l y congested and poorl y mai n-
tai ned, aggravati ng techni cal i neffi ci ency and accel -
erati ng wear.
Rai l Technol ogi es
Rai l roads are si gni fi cant provi ders of transport
servi ces onl y i n I ndi a and Chi na, and i n these two
countri es the rai l share of total transport i s decl i ni ng
rapi dl y due to the much faster growth of road
transport. I n Chi na, for exampl e, the share of
passenger traffi c usi ng rai l ways dropped from 69
percent i n 1965 to 48 percent i n 1987. Rai l way
frei ght transport shows the same trend of decreasi ng
rel ati ve use.
125
Si mi l arl y, I ndi a shows a mode shi ft
toward roads and away from rai l for both passenger
and frei ght traffi c.
126
Despi te these modal shi fts, the rai l systems i n
both countri es sti l l account for si gni fi cant energy
use. Chi na and I ndi a have extensi ve rai l networks
that consume, respecti vel y, 72 percent and 29
percent of transportati on energy (see app. 3-A). The
I ndi an rai l system, al though i n rel ati ve decl i ne, sti l l
carri es a si gni fi cant amount of frei ght and passen-
gers (see tabl e 3-19), usi ng a mi x of steam (bei ng
phased out), di esel , and el ectri c l ocomoti ves.
I mpl i cati ons for Energy Demand
Road transport-pri vate autos for passengers and
trucks for frei ght--has become the domi nant mode
of transportati on i n devel opi ng countri es. I ncreases
i n popul ati on, i ncome, and auto ownershi p rates
(autos per person) combi ne to yi el d a rapi d i ncrease
i n the number of pri vate vehi cl es. I ncreasi ng urbani -
zati on l eads to greater congesti on, whi ch reduces the
effi ci ency of pri vate vehi cl es. Urbani zati on, eco-
nomi c growth, and i ndustri al i zati on requi re l arge
i ncreases i n frei ght movement, as producers move
farther from markets. The net effect of these factors
wi l l be an i ncrease i n the energy needed to provi de
transportati on servi ces.
I mprovements i n the energy effi ci ency of devel -
opi ng worl d transport systems can be made i n
several areas. Road-goi ng vehi cl es i n the devel opi ng
worl d are l ess energy-effi ci ent than comparabl e
vehi cl es i n the i ndustri al i zed worl d, suggesti ng that
effi ci ency gai ns can be made i n the vehi cl es
lmTmcIKs account for 50 to 75 pmcent of meqly consum
ed for road transport in the developing worl~ compared to 30 to 35 percent for many
industrialized countries. Clell G. Harrsl, Meeting the Transportation Aspirations of Developing Countries: Energy and Environmental Effeets,
Proceedings of the Energy and Environment in the 21st Century Conference (Csmbndge, MA: Massachusetts Institute of Technology, March 1990).
l~fierW~omtionA*~tioQInter~tio~/Energy A~/ 1988, DOE/EIA-0219(88) (Wash@toQ ~: U.S. ~v~nm~t~ting Office,
1989).
12sp. Kuirun and S. Guojie,
Overview of Transport Development in Chimq papex presented at the New Energy Technologies Transportation and
Development Workshop, Ottawa Canadz September 1989.
126JOY
Dunkerley, Irving Hock Charu Gadhdq K@ rh~ Energy and Transport-IheIndian Experience, PaciicandAsianJournal ofEnergy,
1987, pp. 1-12.
Table 3-19-Comparison of Rail Systems in China,
India, and the United States
United
China India States
Length of rail network (km) . . . . . . 53, 000 62, 000 235, 000
Rail energy use (percent of
total transport energy use) . . . . 51 27 3
Percent of freight traffic carried
by rail . . . . . . . . . . . . . . . . . . . . . . 45 47 30
Percent of passenger traffic
carried by H-l . . . . . . . . . . . . . 55 22 1
SOURCES: P. Kuirun and S. Guojie, Overview of Transport Development
in China, paper presented at the New Energy Technologies
Transportation and Development Workshop, Ottawa, Canada,
September 1989; Tata Energy Research Institute, TER/
Energy Data Directory and Yearbook (TEDDY) 1988 (New
Delhi, India: 1989); International Energy Agency, I#tld En-
ergy Statistics and Ba/ances: 198 f-87 (Paris: OECD, 1989);
Oak Ridge National Laboratory, TranspMdion Energy Data
Book: E&ion 70, ORNL-8585, 1989; Association of American
Railroads, Railroad Facts 7989 (Washington, DC: American
Association of Railroads, 1990).
themsel ves. I mprovements i n the transportati on
i nfrastructure, such as i mproved roads and reduced
congesti on, can al so i ncrease energy effi ci ency.
Mode choi ces, such as a movement away from
pri vate autos and motorcycl es to buses and bi cycl es,
can hel p. Of course, al l these opti ons have benefi ts
and costs: these wi l l be expl ored l ater i n thi s OTA
study. The i mportant concl usi ons for thi s secti on,
however, are that the demand for transportati on
servi ces i s i ncreasi ng rapi dl y, technol ogi es i n use
today are not as effi ci ent as they coul d be, and the
energy i mpacts of technol ogy choi ces made today
wi l l be fel t far i nto the future.
Conclusion
This survey of energy servi ces and how they are
provi ded i n devel opi ng countri es reveal s three
common characteri sti cs. Fi rst, each servi ce i s pro-
vi ded by a wi de range of technol ogi es and fuel s.
Cooki ng i s provi ded by technol ogi es rangi ng from
open frees to mi crowave ovens, wi th a l arge number
of possi bi l i ti es i n between. The range of passenger
transport servi ces i s si mi l arl y wi de, varyi ng from
foot to jet passenger ai rpl anes. Second, there i s i n
al most al l cases a wel l -establ i shed transi ti on be-
tween technol ogi es, dependi ng on two mai n factors--
i ncome and avai l abi l i ty of fuel suppl i es. Thi rd, the
servi ces are currentl y bei ng provi ded by technol o-
gi es whose effi ci ency coul d usual l y be si gni fi cantl y
i mpr oved.
The fol l owi ng chapter wi l l exami ne how the
many types of energy used i n devel opi ng countri es
fossi l fuel s, el ectri ci ty, and bi omass fuel s--are
provi ded, i ncl udi ng domesti c producti on, i mported
suppl i es, the energy di stri buti on system, and the
energy conversi on sector.
Chapter 3--Energy Services in Developing Countries .81
Appendix 3-AEnergy Balances for
Selected Developing Countries
The fol l owi ng energy bal ances begi n wi th I nterna-
ti onal Energy Agency (I EA)
127
energy producti on, trade,
and stock change total s for commerci al fuel s. Thi s
provi des a common framework for eval uati ng i ndi vi dual
countri es and for compari ng di fferent countri es. Bi omass
fuel s for the tradi ti onal sector are not i ncl uded i n the
country-speci fi c I EA data and so are separatel y added
based on country speci fi c fi el d survey data. The year
chosen for each country i s determi ned pri mari l y by the
year for whi ch the bi omass energy data i s avai l abl e.
I n contrast to the I EA procedure, energy suppl y
producti on, conversi on, and transformati on l osses are not
separatel y tal l i ed i n the energy bal ances presented here.
I nstead, these l osses are carri ed forward i nto the sectoral
breakdowns i n proporti on to the I EA sectoral breakdown
of energy use. Thi s more accuratel y i ndi cates sectoral
ener gy usage by showi ng the l osses i ncur r ed i n pr ovi di ng
energy to each sector.
El ectri ci ty i s i ni ti al l y di vi ded i nto two categori es i n the
fol l owi ng energy bal ances: nonthermal and thermal .
Nonthermal el ectri ci ty i s gi ven i n terms of the el ectri c
power output--the joul e equi val ent of kWh. Thermal
el ectri ci ty i s gi ven i n terms of thermal energy i nput; l osses
i ncur r ed i n gener ati on, tr ansmi ssi on, and di str i buti on ar e
kept i n the total . Nonthermal and thermal el ectri ci ty
quanti ti es, therefore, can not be di rectl y compared.
The I EA conventi on for el ectri ci ty producti on di vi des
the hydroel ectri c output i n kWh by 0.385 i n order to make
hydroel ectri c power appear to be on the same thermal
equi val ent i nput basi s as ther mal l y gener ated el ectr i ci ty
when l i sted on the basi s of fossi l fuel i nput. The I EA
subsequentl y mul ti pl i es the sum of hydroel ectri c ther-
mal equi val ent and thermal el ectri c i nputs by 0.385 to
get an el ectri c power output i n kWh. Thermal and
hydroel ectri c thermal equi val ent l osses are l umped
together as an energy producti on l oss.
Thi s conventi on of thermal equi val ents l eads to a
l arge mi srepresentati on i n the energy bal ances for hydro-
ri ch countri es such as Brazi l . The procedure used here
avoi ds the I EA conventi on of assi gni ng a thermal
equi val ent for hydroel ectri c or other nonthermal power. I t
al so carri es the l osses i n thermal generati ng pl ants
through to the end-use sectors as noted above. At the
sectoral l evel , the thermal and nonthermal el ectri ci ty are
added together di rectl y to i ndi cate the average amount of
energy, i ncl udi ng fossi l fuel , used by each sector. These
fi gures are shown i n brackets to denote that the fi gure i s
a sum of nonthermal el ectri ci ty output and thermal energy
i nput. Thi s procedure l owers the energy suppl y total s
compared to those usual l y found i n the l i terature.
Percentage breakdowns by end-use sector are based on
the I EA data; percentage breakdowns by energy servi ce
wi thi n end-use sectors are based on country-speci fi c
surveys as noted. The end servi ce breakdowns are the best
esti mates that OTA coul d make gi ven the poor qual i ty and
pauci ty of avai l abl e data These breakdowns are provi ded
here onl y as an i ndi cati on of the rel ati ve i mportance of
sel ected energy servi ces; they shoul d not be construed to
be a preci se quanti tati ve measure of the energy consumed
i n del i veri ng these servi ces or to be a preci se l i sti ng of
energy servi ces and thei r i nterrel ati onshi ps. Some i mpor-
tant energy servi ces and fuel mi xes are overl ooked i n
many of the avai l abl e ener gy ser vi ce br eakdowns. For
exampl e, l i ghti ng and the use of tradi ti onal fuel s are
l argel y l eft out of the i ndustri al sector. I n addi ti on, a
number of i mportant energy servi ces are general l y l eft out
of the breakdowns: an exampl e mi ght be the use of
ani mal s for tracti on i n agri cul ture and the use of crop
resi dues to feed them.
lz7~te~o~ mum Agency, WortiE~rgy Statistics and Balances: 1971-87 @tiS: OECD, 1989).
Table 3A-lBrazil: Energy Supplies and Services, 1987 Exajoules (10
18
Joules= 0.9478 Quad)
and Percent of National Total
Fossil fuels Eleotricity Biomass
Cane/ Wood/
Energy supplies coal Oil Gas Nonthermal Thermal alcohol Charcoal Other Total
Production . . . . . . . . . . . . . . .
Trade/stock change . . . . . . . .
Electric generation . . . . . . . . .
Nonenergy . . . . . . . . . . . . . . .
Total energy . . . . . . . . . . . . . .
Percent of total . . . . . . . . . . . .
Energy services
Residential . . . . . . . . . . . . . . .
Residential . . . . . . . . . . . . . . .
Cooking/heating . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Refrigeration . . . . . . . . . . . .
Television . . . . . . . . . . . . . .
Air conditioning . . . . . . . . .
Other . . . . . . . . . . . . . . . . . .
Commercial
c
. . . . . . . . . . . . .
Commercial . . . . . . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Refrigeration . . . . . . . . . . . .
Other . . . . . . . . . . . . . . . . . .
Industrial . . . . . . . . . . . . . . . .
Industrial . . . . . . . . . . . . . . . . .
Motor drive . . . . . . . . . . . . .
Process heat . . . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Electrochemical . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Road . . . . . . . . . . . . . . . . . .
Rail . . . . . . . . . . . . . . . . . . .
Air . . . . . . . . . . . . . . . . . . . .
Other . . . . . . . . . . . . . . . . . .
Agriculture . . . . . . . . . . . . . .
Agriculture . . . . . . . . . . . . . . .
0.124
0.315
-0.044
0. 40
6.6%
0. 40
6.58%
6 . 5 %
0. 00
( O*
0 *
1.55
1.040
-0.073
-0.167
2.349
38.6%
0.202
3.3%
3.3%
0.026
0.42%
0.42%
0.56
9.22%
9.22%
1.417
23.3%
20.4%
0.43%
1.74%
0.77%
0.146
2.4%
0.112
0.112
1.8%
0.007
0.1%
O.1%
0.003
0.05%
0.05%
0.10
1.65%
.
1.65%
0.681
0.023
0.117
( 0 . 8 2 1 )
.
(0.169)
2.8%
0.77%
3
0.67%
0.90%
0.1 7%
0.09%
0.21 %
(0.159)
2.62%
0.20%
1.1570
0.45%
0.52%
0.29%
(0.46)
7.57%
3.7170
3. 18%
o.15?40
0.53%
(0.005)
0.08%
0.08%
0.026
0.43%
0.91
-0.02
0.89
14.6%
0.44
7.24%
.
7.24%
0.44
7.24%
7.24%
1.39
1.39
22.9%
0. 48
7.9%
7.9%
0.018
0.3%
0.3%
0.75
12.3%
12.3%
0.126
2.1 %
0.13
-0.01
0.12
2.0%
.
.
4.90
1.37
0.00
-0.19
6.08
100YO
0.86
14.1%
1 2.1%
0.7%
0.9%
0.2%
0.170
0.2%
0.21
3.5%
1 .O%
1.270
0.5%
0.5%
0.3%
2.71
44.6%
3.7%
40.2%
0.2%
0.5%
1.86
30.6%
27.6%
0.5%
1 .7%
0.8%
0.30
4.9%
() data in parentheses is sum of nonthermal energy output and thermal energy input.
a The use of bagasse for energy production(cogeneration)isdivided proportionately betweenindustrialpprocess heat and road transport. Electricity generation
within the cane industry is not given separately.
b This is mostly for water heating(10 TWh). Only O.5 TWh were for cooking.
c
Excludes public buildings.
SOURCES: Adapted from international Energy Agency, Wodc/Energy Statistics and Balances: 1971 -87(Pans: OECD, 1989); and Brazilian Ministry of Mines
and Energy, Natina/ Energy Ba/ance for Brazi7 1988 (Brasilia, 1988), provided by Howard S. Geller, American Counal for an Energy Efficient
Economy, Washington, DC, and Sao Paulo, personal communication, Mar. 8,1990.
Table 3A-2-China: Energy Supplies and Services, 1987 Exajoules (10
18
Joules = 0.9478 Quad)
Fossil fuels Electricity Biomass fuels
Energy supplies coal Oil Gas Nonthermal Thermal Crop Wood Dung Biogas Total
0.15
0.15
0.5%
0.15
0.5%
0.4%
0.1%
0.03
0.03
0.1%
0.03
0.1%
0.1%
31.0
-2.2
-0.15
28.7
100%
12.4
43.%
31 .4%
10.3%
1.370
0.77
2.7%
14.7
51 .3%
37.8%
13.5%
1.24
4.3%
0.8%
2.4%
1.270
1.92
6.7%
- . . . .- .- - -- - -- - -- - . - -
Production . . . . . . . . . .
Trade/stock change . .
Electric generation . . .
Nonenergy . . . . . . . . . .
Total energy . . . . . . . .
percent of total . . . . . .
18.15
-0.80
-3.15
5.32
-1.40
-0.59
-0.15
3.18
11.170
0.12
0.4%
0.4%
0.27
0.9%
1.77
6.2%
3.4%
2.8%
0.57
2.0%
0.8%
1 .2%
0.44
1.570
0.46
-0.02
0.33
0.00
3.76
( 4 . 1 0 )
14.3%
3.4
3.2
3.2
1 1.2%
3.2
1 1.2%
10.5%
0.7%
.
14.20
49.5%
0.44
1.570
3.4
11.870
Energy services
Residential . . . . . . . . .
Residential . . . . . . . . .
Cooking . . . . . . . . . .
Space heating . . . . .
Lighting . . . . . . . . . .
5.23
18.2%
8.7%
9.5%
0.07
0.2%
0.2%
(0.27)
0.94%
3.3
11 .5%
11.5%
0.94%
Commercial/public . . .
Commercial/public . . .
0.31
1.1%
9.41
32.8%
32.8%
(0.19)
0.66%
(3.06)
10.7YO
0.00
0.37
1 .3%
1 .3%
0. 00
0.1
0.3%
0.3%
Industrial . . . . . . . . . . .
Industrial . . . . . . . . . . .
Process heat. . . . . .
Mechanical drive . .
10.7%
(0.08)
0.285
Transport. . . . . . . . . . .
Transport. . . . . . . . . . .
Road . . . . . . . . . . . .
Rail . . . . . . . . . . . . .
Air . . . . . . . . . . . . . .
Other . . . . . . . . . . . .
0.59
2.1 %
2.1 %
0.00
0.28%
0.00
Agriculture . . . . . . . . . .
Agriculture . . . . . . . . . .
0.98
3.4%
(0.50)
1 .7%

( ) data in parentheses is sum of nonthermal energy output and thermal energy input.
SOURCE: Adapted from International Energy Agenoy, 144w# Energy Sfafisfks and Ba/ances: 1971-87 (Pwis: OECD, 1989); and Vadav 8mil, Ohinas
Energy, contractor report prepared for the Office of Technology Assessment, 1990.
Table 3A-3-lndia: Energy Supplies and Services, 1985 Exajoules (10
18
Joules = 0.9478 Quad)
Fossil fuels Electricity Biomass
Energy supplies coal Oil Gas Nonthermal Thermal Wood Dung Crop Total
Production . . . . . . . . . . . . . . .
Nonenergy . . . . . . . . . . . . . . .
Total energy . . . . . . . . . . . . . .
percent of total . . . . . . . . . . . .
Energy services
Residential . . . . . . . . . . . . . . .
Residential . . . . . . . . . . . . . . .
Cooking/water heating . . . .
Lighting . . . . . . . . . . . . . . .
Appliances . . . . . . . . . . . . .
Commercial/pubiic . . . . . . . . .
Commercial/pubiic . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Appliances . . . . . . . . . . . . .
industrial . . . . . . . . . . . . . . . . .
industrial . . . . . . . . . . . . . . . . .
Process Heat . . . . . . . . . . .
Motor Drive . . . . . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Appliances . . . . . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Road . . . . . . . . . . . . . . . . . .
Rail . . . . . . . . . . . . . . . . . . .
Air . . . . . . . . . . . .. .:.....
Agriculture
a
. . . . . . . . . . . . . . .
Agriculture . . . . . . . . . . . . . . .
Motor Drive.... . . . . . . . . .
Traction . . . . . . . . . . . . . . .
3.16
0.03
-1.29
1.29
0.48
-0.11
-0.09
1.57
18%
0.39
4.4%
1.9%
2.1%
0.03
0.3%
0.3%
0.40
4.5%
3.1%
0.8%
0.77
8.7%
7.1%
0.7%
0.8%
0.15
1.7%
1.0%
0.7%
0.17 0.20
0.00
1.45
0.87
1.2
1.2
13%
1.2
14%
14%
.
1.6
1.6
18%
1.6
18%
18%
q
*
*
8.5
0.5
-0.09
8.9
100%
4.2
47%
4370
3.9%
0.4%
0.20
2.2%
1.1%
0.4%
0.6%
3.1
35%
23%
10.8%
0.5%
0.5%
1.0
11.8940
7.1%
3.8%
0.8%
0.43
4.8%
4.l%
o.7%
-0.05
0.12
1.3%

(1.65)/19%
0.87
10%
1.90
21%
0.06
o.7%
0.6%
0.007
0.0870
0.08%
0.78
8.8%
8.8%
1.8%
o.4%
0.02
0.2%
0.2?!
(0.09)
.
1.0%
0.4%
0.6%
0.05
0.6%
0.6%
1.61
18%
18%
0.113
1.3%
1.3%
(0.97)
10.9%
0.04
0.5%
0.5%
10.0%
0.5%
0.5%

0.23
2.6%
(0.04)
0.5%
0.5%
2.6%

0.003 (0.28)
3.1%
3.1%
0.03%

Not available or not a aplicable.
q Small.
a baseline data from the international Energy Agency for petroleum use in agriculture have been modified to correspond better with TERl energy data.
SOURCE: Adapted from international Energy Agency, 144rkj Enqgy Statistics and Ba/ances: 197f-87(Paris: OECD, 1989); Tata Energy Researoh
Institute, TEDDY op. cit., footnote 58; and Aehok Desai, contractor report prepared for the Office of T6chnoiogy Assessment and personai
communication.
Chapter 3-Energy Services in Developing Countries q 85
Table 3A-4-Kenya: Energy Supplies and Services, 1980, Petajoules (10
15
Joules) and Percent of National Total
Fossil fuels Electricity Biomass fuels
Energy supplies coal Oil Gas Nonthermal Thermal Wood
a
Residues Total
Production . . . . . . . . . . . . . . .
Nonenergy . . . . . . . . . . . . . . .
Total energy . . . . . . . . . . . . . .
percent of total . . . . . . . . . . . .
Energy services
Residential . . . . . . . . . . . . . . .
Residential . . . . . . . . . . . . . . .
Cooking . . . . . . . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Commercial . . . . . . . . . . . . . .
Commercial . . . . . . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Industrial . . . . . . . . . . . . . . . . .
industrial . . . . . . . . . . . . . . . . .
Informal . . . . . . . . . . . . . . .
Formal. . . . . . . . . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Road . . . . . . . . . . . . . . . . . .
Rail . . . . . . . . . . . . . . . . . . .
Air . . . . . . . . . . . . . . . . . . . .
Agriculture . . . . . . . . . . . . . . .
Agriculture . . . . . . . . . . . . . . .
0.42
0.42
0.1%
0.42
0.1%
.
0.1%
78
-6.1
-2.3
70
17.1%
6.84
1.67%
0.53%
1.14%
0.75
0.18%
0.05%
0.13%
16.5
4.o%
4.0%
43.3
10.6%
6.9%
0.6%
2.8%
6.6
1.6%
3.82
0.44
6.1
320
9.3 333
79
0.0
-2.3
410
100%
320
78%
9.3
2.3%
(2.19)
0.53%
0.32%.
0.21%
(1.92)
0.47%
253
62%
62%.
9.3
2.3%
2.3%.
271
67%
65%
2%
6.2
1.5%
0.9%
0.6%

3.5
0.85%
0.85%

0.47%
56
13.6
9.4%
4.2%
76
18.5%
9.4%
9.1%
43.1
10.6%
6.9%
0.6%
2.8%
7.7
1 . %
(3.23)
0.79%
0.79%

(1.06)
0.26%
a Includes both commmercial and noncommercial
uses of wood;does not
include wood used as a feedstock or as a construction material..Alsoincludes charcoal
that is produced from wood.This conversion takes roughly 110PJ of wood and converts it into about 27 PJ of charcoal, of which about 1.3 PJ is lost during
distribution.
SOURCE: Adapted from International Energy Agency, Mwk/EnergySt#Mics and Bakmces: 1971 -87(Paris: OECD, 1989); and Energy and Development
in Kenya, Eds. Phil OKeefe, Paul Raskin, and Steve 13ernow, Beijer Institute, Royal Swedish Academy of sciences, Stockholm, Sweden, 1984.
Table 3A-5-Taiwan: Energy Supplies and Services, 1987, Petajoules (101
15
Joules) and Percent of National Total
Energy supplies coal Oil Gas Nonthermal Thermal wood Total
Production . . . . . . . . . . . . . . .
Nonenergy . . . . . . . . . . . . . . .
Total energy . . . . . . . . . . . . . .
percent of total . . . . . . . . . . . .
Energy services
Residential . . . . . . . . . . . . . . .
Residential . . . . . . . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Refrigeration . . . . . . . . . . . .
Television . . . . . . . . . . . . . .
Fans/air conditioning . . . . .
Commercial/public . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Other Appliances . . . . . . . .
Industrial . . . . . . . . . . . . . . . . .
Industrial . . . . . . . . . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Road . . . . . . . . . . . . . . . . . .
Rail . . . . . . . . . . . . . . . . . . .
Air . . . . . . . . . . . . . . . . . . . .
Other . . . . . . . . . . . . . . . . . .
Agriculture . . . . . . . . . . . . . . .
Agriculture . . . . . . . . . . . . . . .
38.9
363
-220
.
182
13.7%
0.14
q YO
0.12
q YO
q VO
180
13.6%
5.4
758
-58.3
-46.8
658
49.6%
46.3
3.5%
3.5%
31.6
2.4%
2.4%
314
23.7%
222
16.7%
15.O%
0.1%
1.0%
0.6%
37.3
2.8%
42.3
42.3
3.2%
19.6
1.5%
1.5%
4.0
0.3%
0.3%
18.5
1.4%
145
278
( 4 2 3 )
31.9%
(87.5)/6.6%
0.3%
I.O%
1.9%
1.2%
1.5%
(&.22
0.1%
1.2%
1.1%
1.2%
(255)
19.2%
(2.7)
0.2%
O.1%
q
q
(13.0)/1.0%
.
21
.
21
1.6%
21
1.6%
1.6%
0.3
q %
253
1121
0
-47
1326
100.%
175
13.2%
6.9%
1.0%
1.9%
1.2%
1.5%
83.2
6.3%
2.8%
1.2%
1.1%
1.2%
767.5
57.9%
225
17.O%
15.0%
0.2%
1.0%
0.6%
50.3
3.8%
q Small.
aCharcoal is includedunder wood. The charcoal conversion effiency is assumed to be a relatively high level of50 percent by energy.
SOURCE: Adapted From International Energy Agency, 14bti Energy Stafisk and Ba/ances: 197f-87(Paris: OECD, 19S9); and personal communication,
Dr. Gwo-Tzeng, Energy Research Group and Institute of Traffic and Transportation, National Chiao Tung University, Taipei, Taiwan.
Table 3A-6-United States: Energy Supplies and Services, 1985,
a
Exajoules (10
18
Joules =0.9478 Quad)
b
Space conditioning . . . . . .
Water heating . . . . . . . . . .
Lighting . . . . . . . . . . . . . . .
Industrial . . . . . . . . . . . . . . . . .
Industrial . . . . . . . . . . . . . . . . .
Heat
f
. . . . . . . . . . . . . . . . . .
Motor Drive . . . . . . . . . . . . .
Off-Highway transport . . . .
Transport . . . . . . . . . . . . . . . .
Transport . . . . . . . . . . . . . . . .
Road . . . . . . . . . . . . . . . . . .
Rail . . . . . . . . . . . . . . . . . . .
Air . . . . . . . . . . . . . . . . . . . .
Other . . . . . . . . . . . . . . . . . .
19.6
-1.8
-14.8
3.0
4.4%
0.1
O.1%
0.1
0.2%
2.6
3.8%
3.8%
0.9%
20.9
8.9
-1.1
-1.8
26.9
39%
1.5
2.2%
2.0%
0.2%
1.2
1 .7%
1.6%
0.1%
4.3
6.2%
(11.8%)9
19.3
28.%
23.1%
0.7%
3.9%0
0.3%
0.6
0.9%
1.2
-3.0
14.7
21.3%
5.3
7.7%
5.3%
1.8%
0.5%
0.1%
2.9
4.2%
2.9%
0.1%
Energy supplies coal Oil Gas Nonthermal Thermal Wood

a
Total
Production . . . . . . . . . . . . . . . 16.5 2.53
c
2.80 62.3
Nonenergy . . . . . . . . . . . . . . .
Total energy . . . . . . . . . . . . . .
percent of total . . . . . . . . . . . .
Energy services
Residential . . . . . . . . . . . . . . .
Residential . . . . . . . . . . . . . . .
Space conditioning . . . . . .
Water heating . . . . . . . . . .
Cooking . . . . . . . . . . . . . . .
Refrig/Freezers . . . . . . . . .
Lighting/Other
e
. . . . . . . . . .
6.5
9.4%
0.06
d
18.9
( 2 1 . 5 )
31.2%
(7.6)
11.%
4.3%
1.5%
0.7%
1.6%
2.8%
(6.6)
9.6%
6.1%
0.1%
2.5%
(7.3)
10.6%
1.2%
7.1%
(0.03)/*%
q %
-0.02
2.8
4.1%
1.0
1.5%
1.5%
1.8
2.6%
2.6%
8.4
-1.8
68.9
100%
Agriculture
h
. . . . . . . . . . . . . . .
Agriculture . . . . . . . . . . . . . . .
()data in parentheses is sum of nonthermal energy output and thermal energy input.
Small.
aNote that the lEA data used as a framework for this energy balance differs slightly from official U.S. energy statistics. See, for example, Energy lnformation
Administration, Month/yGwrgy Review, January 1990.
b split 58 percent Nuclear and 42 percent Hydroelectric and other (geothermal, solar, etc.).
d Imports of electricity into the United States are primarily hydroelectric based power from Canada.
e lncludes clothes washers and dryers, dishwashers, lighting, and miscellaneous.
f lncludes fuels USed for cogeneration applications.
g0il and gas applications are combined here.
h This does not include indirect inputs (fertilizer, etc.).
15.5
22.5%
13.1%
3.5%
1 .2%
1.6%
2.9%
10.8
15.7%
10.6%
0.3%
2.5%
22.5
32.7%
19.4%
7.1 %
0.9%
19.3
28%
23.1 %
0.7%
3.9%
0.3%
0.6
0.9%
SOURCE: Adapted from International Energy Agency, WbddEnergy Statistics and6a/ances: 1971 -87( Paris: OECD, 19S9); and Paui D. Holtberg, Thomas
J. Woods, Marie L. Lihn and Nancy C. McCabe, Baseline %ojection Data Book: 1989 GR/ Baseline Projection of U.S. Energy Supp/yandDernand
to 2010 (Washington, DC: Gas Research institute, 1990).
Appendix 3-BSources for Table 3-9
As used here, effi ci ency i s approxi matel y the fi rst-l aw
effi ci ency; that i s, the total heat transfer to the materi al
bei ng processed di vi ded by the heat i nput to the system.
Thi s i s to be contrasted wi th the second-l aw effi ci ency,
whi ch compares the heat transfer achi eved i n the process
wi th the maxi mum possi bl e heat that coul d be transferred
for the same purpose by any system usi ng the same energy
i nput. For more i nformati on, see Efficient Use of Energy,
AIP Conference Proceedi ngs, No. 25, Ameri can I nsti tute
of Physi cs, New York 1975.
Cooking: The pri mary source for tradi ti onal and
modern cooki ng technol ogi es i s Samuel F. Bal dwi n,
Biomass Stoves: Engineering Design, Development, and
Dissemination, op. ci t., footnote 17. An enormous range
of effi ci enci es have been reported for the open fi re,
rangi ng horn 2 to 30 percent. A l i st of some 45 reports on
tradi ti onal cooki ng technol ogy effi ci enci es was devel -
oped by Jas Gi l l i n 1981 and i s ci ted i n K. Kri shna Prasad,
Woodburning Stoves: Their Technology, Economics, and
Deployment (Geneva: 1983). I nter nati onal Labor Or gani -
zati on, Worl d Empl oyment Programrne Research. Most
of these reports, however, do not ci te a source, nor any
detai l s as to how such an effi ci ency fi gure was deri ved.
Traditional Beer Brewing: Data are from Frans
Sul i l atu, I mproved Beer Cookers I n Burki na Faso,
Boiling Point, No. 10, pp. 10-14, August 1986. Thi s i s the
thermal effi ci ency of heati ng the brew to boi l i ng, not for
the enti re brewi ng process. I n Burl &a Faso, West Afri ca,
tradi ti onal dol o (beer) cookers usi ng 80-l i ter cl ay jars
have thermal effi ci enci es of 17 percent wi th a power
output of 80 kW for a cooker wi th four jars arranged i n a
square, and 15 percent at 35 kW wi th the jars arranged i n
a l i ne wi th fi res between adjacent jars. Si mpl e i mprove-
ments i n stove desi gn and the use of al umi num pots r ai sed
effi ci enci es up to as hi gh as 53 percent.
The efficiency of the brewing process can also b e
determi ned by cal cul ati ng the total amount of useful
energy provi ded. Mnzava, ci ted bel ow, has esti mated that
0.12 to 0.5 m
3
of fuel wood are needed to brew 100 l i ters.
Assumi ng that 1 m
3
of stacked fuel wood wei ghs 500 kg
and has an energy content of 16 M.J/kg for a total of 8 GJ;
and assumi ng that the energy i s used to bri ng the water to
a boi l once to steri l i ze i t and then to mai ntai n i ts
temperature at a l ow l evel whi l e i t ferments; the energy
requi red to heat 100 l i ters of water to a boi l from ambi ent
(20 C) i s 100(80)4.186=14 MJ. Thi s gi ves an effi ci ency
r ange of 0.3 to 0.7 per cent for the beer -br ewi ng pr ocess.
Thi s very l ow effi ci ency compared to the thermal
effi ci ency found for dol o cookers i n Burki na Faso maybe
due to the l ong, l ow temperature heati ng needed to
concentrate the mash as wel l as for other operati ons. See:
E.M. Mnzav~ Fuel wood and Charcoal I n Afri ca, W.
Pal z, P. Charti er, and D.O. Hal l (eds.), Energy From
Biomass, 1st E.C. Conference on Bi omass, Bri ghton, East
Sussex (London: Appl i ed Sci ence Publ i shers, Ltd., 1980).
Modern Beer Brewing: Data are from Bernard B.
Hamel et al ., Energy Anal ysi s of 108 I ndustri al Proc-
esses, 1980. The fi gure of 79 percent i s theoveral l boi l er
effi ci ency for a modern brewery. Thi s boi l er provi des
process steam, hot water, and other heati ng servi ces. The
effi ci ency of the cooki ng process i s somewhat reduced
from thi s l evel , but no separate esti mate was avai l abl e.
The val ue of 6 percent i s based on a total energy
demand for a brewery found by Harnel et al ., of 1,439 Btu
per pound of beer produced or, equi val entl y, 3.35 MJ/kg
compared to the energy requi red to heat the brew to
boi l i ng a si ngl e ti me to steri l i ze i tas i n the compari son
made for the tradi ti onal technol ogy.
Tobacco Drying: The estimate of tobacco dryi ng
effi ci ency i s from M.J. Mwandosya and M.L. Luhang~
Energy Use Patterns In Tanzania, Center for Ener ~ and
Envi ronmental Studi es, Pri nceton Uni versi ty, Report No.
180, February 1985; and M.J. Mwandosya and M.L.
Luhang~ Energy Demand Structures i n Rural Tanza-
ni a, Department of El ectri cal Engi neeri ng, Uni versi ty of
Dar-es-Sal aam, Tanzani a, 1984.
Traditional Tea Drying: Tea dryi ng i s based on the
data i n Mwandosya and Luhang~ l i sted above. They
esti mate that 150 kg. of green tea requhes 9.4 GJ of
fuel wood, resul ti ng i n 30 kg. of dri ed tea. To evaporate
120 kg of water requi res 2,260( 120)kJ = 271 MJ of
ener gy, for an effi ci ency of 2.9 per cent.
Traditional Baking: For tradi ti onal bakeri es, Ahmed
and El amgzoub found 0.5 to 0.8 kg of wood used per kg
of fl our. Typi cal rati os for bread are 720 g fl our, 500 g
l i qui d, and 50 g sugar i nput per kg of bread output. Wi th
speci fi c heats of 1.8 kJ/kg C for fl our, 4.186 kJ/kg C for
water, and arbi trari l y assumi ng 4.0 kJ/kg C for sugm, and
noti ng that approxi matel y hal f the water evaporates, the
rest remai ni ng i n the bread (Gel l er); then the energy
requi red to bake 1 kg of fl our i nto bread at 190 C i s:
(1.0)(1.8)(170) + (0.7)(4.186)(80) + (0.35)(2260) +
(0.35)(4.186)(90) + (0.09)(4.0)(170)= 1.5 MJ, wher e i t
was assumed that the speci fi c heat of the water remai ni ng
i n the bread, 0.35 kg, remai ned 4.186 and the chemi cal
reacti ons and heati ng of the vapori zed steam were
i gnored. By compari son, 0.5 to 0.8 kg wood have an
energy content of 8 to 13 MJ. Abdel Sal aam Ahmed and
El Shei kh El amgzoub, Survey of Fuelwood Consumption
in Khartoum Province Industries (Khartoum, Sudan:
Nati onal Energy Admi ni strati on, Mi ni stry of Energy and
Mi ni ng for the Energy Research Counci l , Sudan Renew-
abl e Energy Project, Apri l 1985. Howard S. Gel l er and
Gautam S. Dutt, Measuri ng Cooki ng Fuel Economy,
Wood Fuel Surveys (Rome: 1983). Food and Agri cul ture
Organi zati on of the Uni ted Nati ons, GCP/I NT/365/SWE.
Reddy and Reddy found that 0.583 kg wood were used
to cook 1 kg of mai da, correspondi ng to an effi ci ency of
16 percent. Amul ya Kumar N. Reddy and B. Sudhakar
Reddy, Energy Use i n a Strati fkd Soci ety: Case Study
of Fi rewood, i n Bangal ore, Economic and Political
Weekly (India), vol . 18, No. 41, Oct. 8, 1983.
Shi rey and Sel ker l i st the effi ci enci es of a number of
tradi ti onal and modem ovens used i n a vari ety of
countri es. Ovens i n Somal i a Sudan, Guatemal a, Zi m-
babwe, and Sri Lanka have typi cal measured effi ci enci es
of 1 to 3 kg wood per kg fl our, gi vi ng effi ci enci es, as
cal cul ated above, of 3 to 8 percent. I n contrast, an
i mproved wood-fi red Somal i oven i s ci ted as usi ng 0.16
kg of wood to cook 1 kg of fl our i nto bread-an effi ci ency
of 58 percent; and modem natural gas ovens are l i sted as
baki ng 360 kg of fl our i nto bread usi ng 1 GJ of energy
-an effi ci ency of 54 percent. E. Shi rey and J. Sel ker,
Bread Ovens, Boiling Point, No. 10, pp. 18-21,1986.
Modern Bakeries: Ho, Wi jeysundera, and Chou found
fi rst-l aw effi ci enci es for a modem i ndustri al bakery i n
Si ngapore to be 43 percent for the enti re process,
i ncl udi ng preparati on of the dough. Second-l aw effi ci en-
ci es were al so cal cul ated and found to be 15.5 percent.
J.C. Ho, N.E. Wi jeysundera, and S.K. Chou, Energy
Anal ysi s Appl i ed to Food Processi ng, Energy VOL 11,
No. 9, 1986, pp. 887-892.
Fish Smoking: Mwal yosi esti mates that smoki ng 1 kg
of fresh fi sh requi res 4 to 5 kg dry wood. I f 70 percent of
the fi sh i s assumed to be water, then i t requi res (2,260
kJ/kg)(O.7 kg)= 1.6 MJ to evaporate the water compared
to (4 to 5 kg)(16 MJ/kg) = 64 to 80 MJ of wood to
accompl i sh the task for an effi ci ency of 2.0 to 2.5 percent.
Raphael B. Mwal yosi , Management of the Mtera Reser-
voi r i n Tanzani a, AMBIO, vol. 15, No. 1, 1986, pp.
30-33.
Traditional Brick Firing: Schmi tt esti mates 1.36 MJ
of energy i s requi red per kg of bri ck produced i n order to
evaporate moi sture from the raw bri ck (after dryi ng i n the
sun) and heat i t to a fi ri ng temperature of 850 C, and an
addi ti onal 0.2 to 0.4 MJ/kg i s needed for the chemi cal
reacti ons. Based on observati ons at si x si tes, an average
of 2.5 MJ fuel wood and other organi c matter were used
per kg of bri ck produced, for an effi ci ency of (0.2 to
0.4)/2.5 = 8 to 16 percent. I t shoul d be noted that these
resul ts were for very l arge ki l ns, fi ri ng typi cal l y 100,000
bri cks at a ti me. Kl aus Schmi tt and Werner Si emers,
Energy From Agricultural Residues and Energy Utiliza-
tion In Small Scale Industries In The Sudun, Secti on 5.4,
Bri ck Ki l ns (Gotti ngen, Sweden: for the Nati onal En-
33-718 0 - 90 - 4
ergy Admi ni strati on of Sudan, Khartoum, September
1985).
Gandhi found an effi ci ency of 6.4 percent for bri ck
ki l ns i n I ndi a, r epr esenti ng the i r r ever si bl e r eacti ons that
take pl ace duri ng fi ri ng. Theoveral l heat ba.l ancefound by
Gandhi fora Bul l strench was: energy i n= 3.88 MJ +0.29
MJ i n car bon i n bri ck energy out i s 61.4 percent i n dry
exhaust; 16.9 percent i n moi sture i n exhaust; 6.4 percent
i n i rreversi bl e reacti ons; 4.0 percent i n heat l oss of CO;
0.3 percent i n carbon i n ash; and other heat l osses (by
di fference) of 11 percent-presumabl y, much of thi s l oss
was through the ki l n wal l s. Other types of ki l ns requi re
horn 2 to 18 MJ/bri ck for fi ri ng. Wi th an average bri ck
si ze of 108 i n3 or 108(16.387) cm3 and an average bri ck
densi ty of 1,800 kg/m3, thi s gi ves an energy requi rement
of 2 to 18 MJ/3.18 kg or 0.637 to 5.7 MJ/kg. Suni ta
Gandhi , The Bri ck I ndustry i n I ndi a: Energy Use,
Tmdi ti on and Devel opment, Ph.D. Thesi s, Tri ni ty
Col l ege, Cambri dge, October 1986.
The bri ck and ti l e i ndustry i n Uganda uses 0.5 to 1.8
stacked cubi c meters of wood per metri c ton of bri ck
produce~ with 7,650 MJ/m3 for eucal yptus at 510
kg/stacked m
3
to gi ve, at best, 3,800 MJ per metri c ton of
bri ck output. Potenti al energy savi ngs of 35 percent may
be possi bl e si mpl y wi th better fi i g techni ques and ki l n
constructi on, and by the i ntroducti on of smal l cavi ti es and
organi c materi al s i nto the bri ck to reduce mass and
i mprove the uni formi ty of fi ri ng. Usi ng the fi gures for
Sudan, thi s gi ves an effi ci ency of about 5 to 10 percent
when assumi ng the chemi cal reacti ons need 0.2 to 0.4
MJ/kg; usi ng the fi gures for I ndi a thi s gi ves an effi ci ency
of about 2 percent. Uganda: Energy Effi ci ency I mprove-
ment i n the Bri ck and Ti l e I ndustry, Worl d Bank/UNDP
Energy Sector Management Assi stance Program, March
1989.
Modern Brtck Industry: Assumi n g the same r ange as
for Sudan, that i rreversi bl e chemi cal reacti ons for the
process are 0.2 to 0.4 MJ/kg of fi red bri c~ a modem bri ck
factory has an effi ci ency of 6-11%. The rel ati vel y hi gh
observed effi ci ency of the tradi ti onal process rel ati ve to
modem ki l ns i s l argel y due to substanti al underfi ri ng i n
tradi ti onal ki l ns and correspondi ng l ow-qual i ty product.
Cal cul ated from Bernard B. Hamel et al . Energy
Anal ysi s of 108 I ndustri al Processes, op. ci t.
Traditional Foundry Work: In I ndonesi a, an esti mated
1 kg of charcoal i s used per kg of al umi num mel ted and
cast i nto pots. From the CRC Handbook of Chemistry and
Physics, the mel ti ng poi nt of al umi num i s 933 OK and i ts
speci fi c heat vari es l i nearl y wi th temperature from CP=0.9
kJ/kgCat300K to 1.19kJ/kgCat933K. The energy
needed to heat i t to i ts mel ti ng poi nt i s then gi ven by
MCPDT=658 kJ/kg. To mel t the al umi num requi res an
adchti onal 398 kJ/kg (CRC Handbook). The total process
then r equi r es 1,056 kJ/kg. Char coal has a cal ori fi c val ue Modern Foundry work: Fi gure of 40 percent i s from
of about 33 M.T/kg. The process i s therefore about 3 Bernard B. Hamel et al ., Ener~ Anal ysi s of 108
percent effi ci ent. Worl d Ba~ Indonesia: Issues and
I ndustri al Processes, op. ci t. above, p. 282.
Options in the Energy Sector, UNDP/Worl d Bank Energy
Sector Assessment Program Repor& No. 3543 -I ND, No-
vember 1981.
kJ/kg.
Bank,
UNDP/World
calorific
Hamel
'Energy
Chapter 4
Energy Supplies in the
Developing World
.
Contents
Page
Introduction and Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
The Overal l Pri mary Fuel Mi x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Coal ... ... ..*. .., ,.. .*. ... **. ... ... ... ,., ... ..*,
+
. * * , ,
o
, *.** **, ,
8
** **, *
4
*
+ o
. * * ,
o a 95
Oil ... ..*. **. *.. ... ... ... ... ... ... *.. ... ... ... ... ... .*** .*
aa o
* * * . * * * * * * . * +.*.+..*.
96
Refineries and Di stri buti on ... .*+. *.. ... ... ... ..*. ... ..*. ... ... ... .***
+
*
@q
. .. .$++
98
Natural Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Bi omass Fuel s ... .*. ... ... .,. .*. .,. ... ... ..
+ee
**
+o
**. ,
+o
*. *., .**, ,,
+
, ,*, .**,
e
*,
$
100
Bi omass Resource Base ... ... ... ... ... ... *.. ... ... *.. *.. ... *.. .*. .**.
o
***+4.*.. 101
Rural Bi omass Markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Commerci al Bi omass Markets ... ... ... ... ... ... ... ... ... ... ,
4
. ,
@
. *.. .,
q+
****c+* 101
charcoal ... ..*. ... ... ... ... .*. ... ... ..** ... ... ... ..** ... ..*. *
+
*
4
*.
+
. ***.
+
.*.
e
102
Bi omass Pri ci ng ... ... ... ... ... *.. ... ... ... ... ... ... ... ... ..*. .
@
**
Go
. *******e* 102
El ectri ci ty ... ... ... ... ... ..*. ... ... ... ... ... ... ..
+
. ... ... ...
+
*
+
. *
@
** *.**
e<
*
aoo
. *e
103
El ectr i ci ty Gener ati on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
El ectri ci ty Transmi ssi on and Di stri buti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
I nsti tuti onal I ssues ..*. *.. ... ... ... ... ... ... ... *.. ... ... ..*
o
. *.*. .
+
. * e . **+**..+@
105
El ectri ci ty Consumpti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
outl ook for I mprovements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Box
Box
4-A. Refi nery Technol ogy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figures
q 106
. . . . . . . . . . 106
Page
. * *.*...... 99
Page
4-1. Energy Suppl y Mi x i n the Devel opi ng Worl d and i n the Uni ted States, 1987 . . . . . 94
4-2. Energy Suppl y Mi x for Sel ected Regi ons of the Devel opi ng Worl d, 1987
95
4-3. Pri macy Energy Producti on, Consumpti on, and Exports i n the
.. +.*.*.
Devel opi ng Worl d, 1985 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
95
4-4. Oi l Consumpti on by End Use i n the Devel opi ng Worl d and i n the
Uni ted States, 1985 ... ... ... *.. ... ... ... ... ... ... ..
+
* ... *.. **. ..*
+
.
q
* *Q**+*.+,
97
4-5. El ectri ci ty Generati on by Fuel Type i n Sel ected Regi ons of the
Devel opi ng Worl d, 1987 ... ... *.. ... ... ... ... ... .*. ..
+
. **
c a
* . . . * * * * * *+*.**.*+ 103
4-6. El ectri ci ty Consumpti on by Sector i n Sel ected Regi ons of the
Devel opi ng Worl d, 1985 *.. ... ... ... ... *.. ... ... .*** ... .
e
. ... .*. *
+
*
+
... ..
m
*** 106
Sector ,
Tabl es
Table
4-1. Energy Suppl y Mi x, 1987 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4-2. Oi l Consumpti on by End-Use Sector, 1985 . . . ., * * *..*.....
4-3. Bi omass Energy Resources i n Sel ected Devel opi ng Countri es
4-4. El ectri ci ty Generati on by Fuel , 1987 . . . . . . . . . . . . . . . . . . . . . . .
4-5. El ectri ci ty Consumpti on by 1985 . . . . . . . . . . . . . . . . . . .
Chapter 4
Energy Supplies in the Developing World
Thi s chapter provi des an overvi ew of the energy
supply sector i n the devel opi ng worl dthe proc-
esses and technol ogi es by whi ch energy i s produced,
converted from one form i nto another, and del i vered
to users. The energy suppl y sector i s cri ti cal for
economi c devel opment for two reasons: fi rst, eco-
nomi c growth depends on the avai l abi l i ty of rel i abl e
sources of energy; and second, the energy suppl y
sector absorbs a l arge share of i nvestment--over 40
percent of the total publ i c i nvestment i n some
devel opi ng countri es,
l
and about 15 percent of
forei gn assi stance.
2
The energy suppl y sector al so
rel i es heavi l y on other resources that can be scarce
i n the devel opi ng worl d, i ncl udi ng ski l l ed l abor and
management, water, and l and.
The energy suppl y sector i s cri ti cal to economi c
devel opment. Economi c growth depends on the
avai l abi l i ty of rel i abl e sources of energy, and the
energy suppl y sector absorbs a l arge share of publ i c
and forei gn ai d i nvestment.
The devel opi ng countri es use a wi de range of
energy sources. Coal i s the l argest pri mary energy
source i n devel opi ng countri es, due mai nl y to the
coal -based energy sectors of Chi na and I ndi a, the
two l argest energy consumers i n the devel opi ng
worl d. Excl udi ng Chi na and I ndi a, oi l and el ectri ci ty
are the mai nstays of commerci al energy suppl i es i n
devel opi ng countri es. I n rural and poor urban areas,
tradi ti onal bi omass fuel s are the pri mary energy
sour ce.
Al though the devel opi ng worl d as a whol e i s a net
oi l exporter, the exports ori gi nate from a rel ati vel y
smal l number of countri es. Most of the countri es i n
the devel opi ng worl d depend heavi l y on i mported
oi l , and these i mports consume a l arge fracti on of
export earni ngs.
The commerci al energy suppl y system--coal , oi l
and gas, and el ectri ci ty-requi res l arge amounts of
forei gn exchange, ski l l ed l abor, and trai ned manage-
ment. I t i s characteri zed by state ownershi p, i n
contrast to suppl i es of tradi ti onal fuel s, whi ch are
l argel y i n pri vate hands.
Commerci al energy suppl i es i n many countri es
ar e unr el i abl e and of poor qual i ty due to oper ati onal
i neffi ci enci es, rapi d i ncreases i n energy demand,
probl ems of reachi ng di spersed popul ati ons served
by i nadequate transport systems, and i nappropri ate
pri ci ng and al l ocati on systems.
Operati onal effi ci enci es i n the energy i ndustry are
somewhat l ower i n many devel opi ng countri es than
i n the i ndustri al countri es. Thi s record of poor
performance often refl ects the ol der equi pment and
di ffi cul t condi ti ons under whi ch energy faci l i ti es
operate. The exi stence of such di fferences between
operati onal effi ci enci es i n reasonabl y standardi zed
operati ons suggests that i mprovements are possi bl e
i f some of these obstacl es can be overcome.
General i zati ons about energy suppl i es i n devel op-
i ng countri es obscure the heterogenei ty of the
devel opi ng worl d. Performance standards vary con-
si derabl y between countri es, as do prospects for
expandi ng energy suppl i es i n the future.
The Overall Primary Fuel Mix
Accordi ng to the I nternati onal Energy Agency
(I EA), coal suppl i ed 35 percent of the devel opi ng
worl ds pri mary energy
3
i n 1987 (tabl e 4-1 and
fi gure 4-l ), fol l owed by oi l (31 percent), bi omass
fuel s (19 percent), pri mary el ectri ci ty (mostl y hydro-
power) (8 percent), and natural gas (7 percent). Some
anal ysts bel i eve i nstead that bi omass i s i n fact the
l argest source of energy, suppl yi ng up to one-thi rd
of pri mary energy i n the devel opi ng worl d.
IM. M~inghe, Elecwic Power Econonucs (London: Buttenvorths, 1990.
World Banlq AnnzualReport 1989 (M%hin@oq DC: 1989). Data include ordy International Development Authority (IDA) andrnternationaI Bank
for Reconstruction and Development (lBRD) lending in fiscal year 1989.
3- energy refers to fuels in the~ mW state> before they are processed into forms suitable for use by fti conmm
ers. Primary fuels include
coal, o~ gas, biomass, and electricity generated from nuclear, hydro, geothermal, and solar sources. Fi.oa.l energy, suitable for end-use COnsumptiom
includes electricity generated from fossil fuels as well as primary electricity. For countries with fossil fuel electricity generation facilities, the amount
of electricity in the fti energy mix is therefore higher than in the primary energy mix, and the amounts of fossil fuels are lower by the amounts used
to generate electricity. Electricity generated fi-om fossil fuels is not included in primary energy in order to avoid double counting.
93
Table 4-lEnergy Supply Mix, 1987 (percent)
Rest of Total
developing developing United
Fuel China India Brazil world world States
coal . . . . . . . . . . . . . . . . . . . . . 70 38 6 17 35 24
Oil ........ ... . . . . . . . . . . . 17 22 38 41 31 41
Natural gas . . . . . . . . . . . . . . . 2 3 2 12 7 22
Other . . . . . . . . . . . . . . . . . . . 4 5 26 8 8 9
Biomass fuels
b
. . . . . . . . . . . . 7 33 28 22 19 4
Total (percent) . . . . . . . . . . 100 100 100 100 100 100
Total (exajoules) . . . . . . . . 26.7 9.6 6.9 42.0 85.2 77.9
a lncludes hydropower, nuclear, geothermal, solar, and electricity produced from biomass. Converted at heat equivalent.
b Note that these estimates for the share of energy supplied by biomass are lower than that indicated by detailed field surveys. See, for example, figure 3-1
and app. 3-A.
NOTE: Total may not add to 100 percent due to rounding.
SOURCE: International Energy Agency, 14kwki&rergy Statistics andl?ahnces 1971 -1987 (Pans: OECD, 1989); IEA, Energy Ba/ances of OECD Countries
1987-1988 (Paris: OECD, 1990).
Figure 4-lEnergy Supply Mix in the Developing World and in the United States, 1987
(exajoules)
Coal
Other +
Natural gas
6 . 8
6. 0
Developing world
Oil
26.4
,i Bi omass
16.2
Oi l
31.9
ass
1
gas
Ot her * 17.1
7.1
Un i t e d S t a t e s
Includes hydropower, geothermal, other renewable, and nuclear converted at heat equivalent. Biomass numbers
may be underestimates, see text, and alternative source, figure 1-2.
SOURCE: International Energy Agency, Wor/dEnergy Statistics and Ba/ances 1971- 1987 (Parfs: OECD, 1989); IEA,
Energy Ba/ances of OECD Countries 1987-1988 (Paris: OECD, 1990).
The rel ati ve shares of these energy sources i n the
overal l energy suppl y mi x vary si gni fi cantl y across
di fferent regi ons and countri es, due i n part to
unequal endowments of energy resources. Coal
suppl i es about hal f of the energy requi rements for
devel opi ng countri es i n Asi a, due l argel y to hi gh
l evel s of coal consumpti on i n Chi na and I ndi a
4
(tabl e
4-1 and fi gure 4-2). Oi l i s the major source of
commerci al pri mary energy for most countri es of the
devel opi ng worl d, I ndi a and Chi na bei ng the notabl e
excepti ons. Natural gas suppl i es a rel ati vel y smal l
fracti on of energy i n the devel opi ng worl d, al though
i n countri es wi th wel l -devel oped resources, gas
often represents an i mportant source of energy (i n
Mexi co, for exampl e, natural gas suppl i es 20 percent
of the pri mary energy consumed).
s
Bi omass fuel s
are a si gni fi cant energy source throughout the
devel opi ng worl d, parti cul arl y i n rural areas and i n
the poorest countri es.
Overal l , the devel opi ng worl d produces more
energy than i t consumes, and si gni fi cant amounts of
both oi l and gas are exported from devel opi ng
countri es (fi gure 4-3). There are, however, major
di spari ti es among countri es: onl y a few devel opi ng
countri es export energy, and most i mport over 50
percent of the commerci al energy they consume.
d~temtio~ Energy Ag~~ @A), W~rldEnergy Statistics and Balances 1971-1987 (Ptis: OECD, 1989), ~kg 1987 dti. ~s includes ~
estimates for biomass consumption and converts electricity to energy units at the heat equivalent,
SIbid.
Chapter 4-Energy Supplies in the Developing World q 95
Figure 4-2Energy Supply Mix for Selected Regions
of the Developing World, 1987
50
40
30
20
10
0
Exajoules
n
,,, ;, ,. ,., :.,:,
~ Bi o ma s s
= O t h e r *
~ Nat ural gas
~ O i l
!ZZZ2 Coal
China India Brazil Rest of
developing world
Includes hydropower, geothermal, other renewable, and nuclear con-
verted at best equivalent. Biomass numbers maybe underestimates, see
text.
SOURCE: International Energy Ageney, Wor/d Energy Statistks and
Ba/ances 1971 -1987 (Pans: OECD, 1989).
Coal
Coal producti on and consumpti on i n the devel op-
i ng worl d are concentrated i n a rel ati vel y smal l
number of countri es. Chi na, I ndi a, South Afri ca,
South Korea, and North Korea are responsi bl e for 96
percent of the total coal producti on i n the devel opi ng
worl d. Chi na al one accounts for 65 percent of
devel opi ng worl d coal producti on and 27 percent of
worl d coal producti on.
6
The l argest consumer of coal i n the devel opi ng
worl d i s i ndustry, whi ch accounts for over 40
percent of total devel opi ng worl d coal use.
7
The bul k
of the i ndustri al coal use i s i n Chi na. Other si g-
ni fi cant coal -consumi ng sectors are el ectri c uti l i ti es,
transportati on (coal -burni ng l ocomoti ves), and the
resi denti al sector i n Chi na, where coal i s used for
cooki ng and space heati ng.
Rates of coal producti on are growi ng rapi dl y.
Hard coal producti on i n Asi a grew at an average
40
30
20
10
0
Figure 4-3-Primary Energy Production,
Consumption, and Exports in the
Developing World, 1985
Total production (exajoules)
I I
Coal Oil Natural gas
~ Net export s
ISI!II! Consumption
SOURCE: International Energy Agency
p
Worid Energy Statistics and
Ba/ances 1971 -1987 (Paris: OECD, 1989).
annual rate of 5.7 percent from 1980 to 1987, and
l i gni te producti on grew at 8.9 percent duri ng the
same peri od.
9
Thi s growth i s expected to conti nue.
Coal qual i ty i n the l ess devel oped worl d vari es
wi del y, both wi thi n and among countri es.
10
I n I ndi a
and Chi na, most coal has a rel ati vel y l ow sul fur
content but a hi gh ash content.
l l
There are few
washi ng and sorti ng processes at coal mi nes, so the
qual i ty of coal suppl i es i s unpredi ctabl e. I n Chi na,
for exampl e, l ess than one-fi fth of al l coal mi ned i s
cl eaned before combusti on. The rest i s used raw,
l i mi ti ng the effi ci ency of combusti on.
The coal i ndustry accounts for about 16 percent of
total commerci al energy i nvestment requi rements i n
l ow-i ncome devel opi ng countri es.
12
I n I ndi a and
Chi na, the l argest coal users, domesti c sources
fi nance most of thi s i nvestment.
Government-owned enti ti es are responsi bl e for
most coal mi ni ng, transport, and di stri buti on i n
devel opi ng countri es. I n Chi na, the Mi ni stry of Coal
6rbid., p. 59, 1987 production of hard cod o~Yo
mid., 1985 data. The United States, in contrasl uses most of its coal for electricity generation.
8A~ ~hom fi fiWe 43, net ~o~ expom me fisi@lc~~ so for the developing world as a whole COd production is the same as COal CO~
ption.
Some individual developing countries do trade in coal-e.g., South Africa is a coal exporter and the Repubtic of Korea is a coal importer.
%?A, op. cit., footnote 4.
l~e ~po~t at~butes of ~~ p~ ~ ener= demity (~ic~y mem~ed in J o~eS per Mogr~or Btu ~r Poud), S@IU content and mh
content.
llTa~ J ?nerw Research ~ti~te, T~ Energy D~t~ D~~ecr~~ u~~ yeu~~~~~ ]$)SS (New De~ ~dia: 1989); Vaclav Smil, chhiis Energy,
contractor report prepared for the OffIce of Technology Assessment, 1990.
Izworld B@ The Energy Transition i
n
D~el~ping Counmes (w~hingtoq w: 1983), p. 68. For period 1982-92.
I ndustry control s about 600 of Chi nas 20,000
mi nes, grouped under 84 Coal Mi ne Admi ni strati ons
or Coal I ndustry Compani es. The remai ni ng, mostl y
smal l er mi nes ar e r un thr ough l ocal gover nments at
the provi nci al , county, or prefecture l evel , or as
col l ecti ve townshi p and vi l l age enterpri ses. I n I ndi a,
the government-owned Coal I ndi a Corp. accounts
for 87 percent of producti on. Coal I ndi a, wi th
675,000 empl oyees, cl ai ms to be the worl ds l argest
si ngl e corporate empl oyer.
13
The al l ocati on of sup-
pl i es i s al so i n the hands of a government enti ty.
Al though coal mi ni ng technol ogi es i n the devel -
opi ng worl d are di verse, the major coal producer
-Chi na-rel i es heavi l y on manual l abor.
14
About
two-thi rds of the extracti on from l arge mi nes i n
Chi na depends on manual l abor, as does vi rtual l y al l
producti on from l ocal l y run mi nes and smal l pri vate
pi ts. Not surpri si ngl y, l abor producti vi ti es are very
l ow, averagi ng l ess than 1 ton per mi ner per shi ft.
Thi s rate i s si gni fi cantl y l ower than current U.S.
rates, whi ch average about 10 tons per mi ner per
shi ft.15
Al though Chi na has consi der abl e exper i ence wi th
a wi de vari ety of advanced underground mi ni ng
techni ques and has the abi l i ty to produce most of the
machi nery requi red, the country does not have
suffi ci ent capi tal or techni cal experti se to moderni ze
i ts coal i ndustry compl etel y. However, surface
extracti on methods, whi ch can be l ess expensi ve, are
bei ng used at many newl y devel oped si tes. Fi ve
l arge pi ts are now under devel opment i n Chi na wi th
a total i ni ti al capaci ty of 50 mi l l i on tons per year.
The l argest of these shoul d eventual l y produce up to
60 mi l l i on tons per year.
Transportati on requi rements often l i mi t coal pro-
ducti on. I n Chi na, coal accounts for 40 percent of al l
frei ght movement, most of whi ch i s by rai l . Chi nas
al ready overl oaded transport system i s struggl i ng to
keep up wi th i ts growi ng coal producti on. I ncom-
pl ete and poorl y confi gured networks, backups at
mode transfer poi nts, and breakdowns al l contri bute
to the unrel i abi l i ty of the coal transport system. As
a resul t, coal -usi ng i ndustri es must stockpi l e up to a
years suppl y, or turn to other fuel s. Si mi l ar prob-
l ems occur i n I ndi a.
l 6
Oil
Petrol eum products are easy to transport and
versati l e i n use i n al l sectors and at al l scal es of
operati on; consequentl y, they pl ay an i mportant rol e
i n the energy sectors of devel opi ng countri es. These
attr i butes l ed to an aver age annual gr owth r ate of 4.5
percent for oi l consumpti on i n the devel opi ng worl d
from 1971 through 1987.
17
Oi l consumpti on i s
expected to conti nue ri si ng by about 3 percent per
year, thereby doubl i ng between 1985 and 2010.
18
More than one-thi rd of the oi l consumed i n the
devel opi ng worl d i s used for transportati on (fi gure
4-4 and tabl e 4-2). The share of oi l used for
transportati on vari es from 13 percent i n Chi na to 42
percent i n Lati n Ameri ca--consi derabl y l ower than
the 62 percent share i n the Uni ted States. The
devel opi ng worl d, compared to the Uni ted States,
uses proporti onal l y more oi l for el ectri ci ty genera-
ti on and for i ndustry. The enti re devel opi ng worl d
consumes about 25 percent l ess oi l than the Uni ted
States al one.
The bul k of devel opi ng worl d oi l producti on i s
concentrated i n a few countri es14 devel opi ng
countri es account for over 90 percent of devel opi ng
worl d oi l producti on.
19
Al though the devel opi ng
worl d as a whol e i s a net oi l exporter, the exports
ori gi nate from a rel ati vel y smal l number of coun-
tri es.
20
Most devel opi ng countri es depend heavi l y
on oi l i mports. More than hal f of the l ow- and l ower
mi ddl e-i ncome countri es i mport 90 percent or more
lSIDEA,~c., cle~co~ Technologies for Developing Countries,
contractor mportprepared for the Office of Technology Assessme@ May 1990.
ldThis dismssion is dram from Vaclav Smil, Chinas Energy, op. cit., footnote 11.
IsRelative t. ~ Ufitd Smtes, la~r is ch~~r ~d mec~~tion is more expensiv~o one wo~d expt greater U* of ktbor klpllk ~d ksS
mechanization in China than in the United States.
16For emple, one tex~e ~~ ~ ~d~ is ~nve~ from cod ~ ri~ hm~ u a boiler fiel ~~ of tie extrew unreliability Of Cod SUpplieS (V.
Kothari, consultan6 Isotem Services, New Delhi, I.m@ personal communicatio~ April 1990).
IT~ op. cit., footnote 4.
18~m s. he ad ~ sc~tte~olmr, lnter~fio~l ~~~~gywo~~s@: Ovemiew of poll Re~o~es &do Alto, CA: StdOrd UIdverSity, Jldy
1989) Intermtional Energy Project.
19These me, h order of d~~ing oil product i on M~ico, - veneme~ ~don~@Nigeri& Alge~ Liby& Egypt, Brr@ tid~ hfdayS@
Argen@ Colombb and Angola. IE~ op. cit., footnote 4.
~ive countries-Mexico, Nigek Venezuel~ Liby~ and Indonesia-account for over 60 percent of LDC oil exports. IE& op. cit., footnote 4.
Chapter 4--Energy Supplies in the Developing World q 97
Figure 4-4-Oil Consumption by End Use in the Developing World and in the United States, 1985 (exajoules)
Industry
5 6
ity
Transportation
18.5
Transportation
8.4
Transformation*
De v e l opi ng wor l d
- , nd- t r y
4.2
J
Electricity y
1.2
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
::::::::::::. Other**
. . . . . . . . . . . . .
. . . . . . . . .
. . . . . . . .
. . . . . . .
.: :::.
6.1
. . .
Transformatlon*
1.2
Un i t e d S t a t e s
*Refinery use and losses.
q *Residential, agricultural, and nonenergy.
SOURCE: International Energy Agency, Wodd Energy Statistics and6a/ances 1971 -1987 (Paris: OECD, 1989) IEA, Energy /3a/ances of OECD Countries
1970/1985 (Pans: OECD, 1987).
Table 4-2-Oil Consumption by End-Use Sector, 1985 (percent)
Total
Latin developing
Sector Asia
a
China Africa America world United States
Electricity . . . . . . . . . . . . .
Transformation
b
. . . . . . . .
Industry . . . . . . . . . . . . . . .
Transportation . . . . . . . . .
Other
c
. . . . . . . . . . . . . . . .
Nonenergy
d
. . . . . . . . . . .
Total (percent) . . . . . . .
Total (exajoules) . . . . .
12
10
24
35
16
3
100
7.5
15
10
40
13
19
4
100
3.9
13
5
18
41
17
6
100
3.5
12
10
16
42
16
4
100
9.1
13
10
23
35
16
4
100
24.0
4
4
14
62
11
6
100
29.9
a excluding China.
transformation includes losses, refinery use, and statistical differences.
c Other is largely residential and agriculture.
d Non-energy use includes waxes, asphalt, and lubricants.
NOTE: Totals may not add to 100 percent due to rounding.
SOURCE: International Energy Agency, Wonkf Energy Statistics and Baiances 1971 -1987 (Paris: OECD, 1989); IEA, Energy Balances of OECD Countries
1970/1985 (Paris: OECD, 1987).
of thei r commerci al energy; al most al l of the i mports
are i n the form of oi l (see app. 1A). By compari son,
the Uni ted States i mports about 17 percent (net) of
i ts energy, of whi ch 90 percent i s i n the form of oi l
or oi l products.
21
The conti nued expansi on Of oi l
consumpti on i mposes an economi c burden on devel -
opi ng countri es, ei ther i n the form of forei gn
exchange for i mported oi l or for i nvestment i n oi l
expl orati on and devel opment. I nvestment i n the oi l
suppl y i nfrastructure i s comparabl e wi th that i n the
el ectri ci ty sector, each accounti ng for about 40
percent of total commerci al energy i nvestment i n the
devel opi ng worl d.
22
I n most devel opi ng counti es, natural resources
such as oi l and gas are consi dered state property.
Ownershi p or other economi c ri ghts to these natural
resources are not bought and sol d wi th the surface
ri ghts (as i n the Uni ted States), but are reserved for
the state to expl oi t or to contract out to thi rd parti es.
I nmost oi l -exporti ng devel opi ng countri es, supreme
authori ty over oi l and gas i s l egal l y vested i n a
central government enti ty-(e.g., the Mi ni stry of Oi l
i n Egypt or the Mi ni stry of Petrol eum and Mi neral s
i n Oman), whi ch del egates sectoral oversi ght and
operati onal responsi bi l i ty to a nati onal oi l company
(e.g., the Egypti an General Petrol eum Corp.). I n
Zl~erW ~o~tion Adrninis~tion (HA), Monthly Energy Review December 1989, DOE/EIA-0035(89/12) (wa$~o~ ~: Us. Governrn@
Printing 0t31ce, February 1990).
~World Banlq op. cit., footnote 12, p. 68.
practi ce, however, there are wi de di spari ti es among
di fferent countri es i n the actual rel ati onshi ps among
executi ve power, sectoral mi ni stri es, and nati onal
operati ng enti ti es. I n oi l -exporti ng countri es, such as
Argenti na and Al geri a, the nati onal operati ng com-
pani es have more power than the mi ni stri es they
depend on, due to the compani es autonomous
sources of revenue from si gni fi cant oi l producti on.
I n contrast, i n oi l -i mporti ng countri es, such as
Sudan, the nati onal oi l compani es have l i ttl e power.
Onl y the weal thi er oi l -exporti ng devel opi ng coun-
tri es i nvest much of thei r own money i n expl orati on
and devel opment. Al l the poorer ones, and the
majori ty of the others, rel y on the i nternati onal oi l
i ndustry for most of the requi red capi tal and
techni cal experti se. The enabl i ng mechani sm
under whi ch forei gn oi l compani es i nvest i n devel -
opi ng countri es vari es wi del y from country to
country, but certai n aspects are wi despread. Gener-
al l y, the mi ni stry i n charge of the sector authori zes
the nati onal oi l company to negoti ate expl orati on
ri ghts wi th forei gn compani es. Thi s negoti ati on may
requi re the formati on of a formal joi nt venture wi th
the nati onal company. More typi cal l y, however, a
l oose rel ati onshi p i s formed i n whi ch the two
associ ates remai n enti rel y separate, wi th the forei gn
party providing al l the capi tal and most of the
techni cal control of operati ons. Once a deal i s
concl uded, i t i s rati fi ed at the central government
l evel .
The fi scal arrangements made under these ena-
bl i ng mechani sms have i n the past been i nfl exi bl e,
provi di ng for the same overal l rate of government
take on al l si zes of fi el ds. These arrangements are
sati sfactory for the mi ddl e range of expected re-
serves and costs. When oi l fi el ds are l arger than
expected, however, the government fai l s to capture
a reasonabl e share of the profi ts on the petrol eum
bei ng produced. When fi el ds turn out smal l er then
expected-a typi cal occurrence i n many devel opi ng
countri es-the government share precl udes eco-
nomi c devel opment by the forei gn operator. Thi s
di scourages the devel opment of smal l fi el ds. A
further di scouragement to forei gn oi l compani es for
the devel opment of oi l fi el ds for domesti c consump-
ti on rather than export i s uncertai nty over the
avai l abi l i ty of forei gn exchange for the remi ttance of
pr ofi ts.
Despi te these di sadvantages, U.S. oi l compani es
i n recent years have i ncreased thei r i nvestments
outsi de North Ameri ca, parti cul arl y i n devel opi ng
countri es. The prospect of addi ng to corporate
reserves at hal f the cost of domesti c expl orati on
23
apparentl y more than compensates for the uncer-
tai nty of forei gn i nvestment.
Refineries and Distribution
Al though the worl ds oi l refi ni ng capaci ty i s
concentrated i n the i ndustri al i zed countri es, many of
the devel opi ng countri es, i ncl udi ng oi l -i mporti ng
countri es, have consi derabl e domesti c refi ni ng ca-
paci ty. Over hal f of the countri es i n Afri ca have
refi neri es,
24
many of whi ch are hydroski mmers (see
box 4-A). I n the devel opi ng worl d, refi neri es are
usual l y owned by the government, whi ch acts
through a nati onal oi l company, al though forei gn
compani es someti mes manage and operate the
refi neri es.
The refi nery product mi x i n devel opi ng countri es
di ffers si gni fi cantl y from that i n the Uni ted States,
due i n part to the i nfl exi bi l i ty of the hydroski mmi ng
technol ogy. I n devel opi ng countri es, between 60
and 70 percent of refi nery output i s di esel and
resi dual , compared wi th 30 percent i n the Uni ted
States. On the other hand, gasol i ne accounts for
about hal f of U.S. producti on, compared wi th about
20 percent i n devel opi ng countri es.
25
I n general ,
refi neri es i n devel opi ng countri es have hi gher pro-
ducti on costs than those i n the i ndustri al i zed worl d:
aver age r efi ner y oper ati ng costs i n Afr i ca ar e $2 per
barrel , compared to $0.75 per barrel i n the rest of the
wor l d.
26
Refi neri es i n the devel opi ng worl d al so
suffer from l arge l osses due to the use of ol d,
i neffi ci ent technol ogy, as wel l as poor mai ntenance.
Refi nery l osses, whi ch shoul d not exceed 1 percent
23~e ~o~t of fim~g ~ -l of ~fl ~ tie Ufitti s~es is $6, comp~d ~~ $3 in tie rest of tie world (Solomon Brothers Inc., Proved Petroleum
Reserves of 30 Large Energy Companies 1980~7 (New Yorlq NY: 1988), p. 14.
~United Nations, 1986 Energy Statistics Yearbook (New York, NY: 1988).
~Ibid.
ZW. Gorton, OiI and Gas Development in Third World Countries,draft contractor report prepared for the Office of Technology Assessment April
1990.
Box 4-ARefinery Technology
Crude oi l i s a mi xture of hydrocarbons. I n the
si mpl est refi ni ng process, pri mary di sti l l ati on (al so
known as hydroski mmi ng), crude oi l i s si mpl y
heated. Thi s causes the l i ghtest hydrocarbons, such
as gasol i ne and l i qui d petrol eum gas, to boi l off
fi rst. As the temperature i ncreases, di fferent prod-
ucts (such as kerosene and di esel oi l ) boi l off. The
resi dual , known as resi dual fuel oi l , remai ns. Thi s
rel ati vel y si mpl e process does not al l ow for much
fl exi bi l i ty i n the mi x of yi el ded products. For
exampl e, 18 to 23 percent gasol i ne and 30 to 55
percent resi dual fuel oi l are typi cal yi el ds from
pri mary di sti l l ati on.
l
Secondary conversi on, whi ch i ncl udes fl ui d cata-
l yti c cracki ng (FCC) and hydrocracki ng, uses hi gh
temperatures to crack l arge mol ecul es i nto
smal l er ones. Thi s process al l ows for greater
fl exi bi l i ty i n product mi x, and i s often used to
i ncrease the proporti ons of l i ghter hydrocarbons,
such as gasol i ne and LPG, whi l e reduci ng the
proporti on of resi dual fuel oi l . For exampl e, hydro-
cracki ng al l ows for up to 86 percent (by wei ght)
gasol i ne producti on, compared wi th 18 to 23
percent for pri mary di sti l l ati on.
2
However, second-
ary cracki ng i s rel ati vel y expensi ve and compl ex.
IL.Wijefleke and A. Ody, World Refiiery I.ndustry-rhe
Need for Restructuring, World Bank Technical Paper No. 32,
Washington DC, 1984, p. 32a.
21bid, p. 33a.
i n a properl y mai ntai ned and operated refi nery, often
exceed 2 or even 4 per cent i n devel opi ng countr i es .27
Duri ng the 1970s and 1980s, both the structure
and the l evel of petrol eum product pri ces changed.
As the pri ces of petrol eum products i ncreased, coal ,
gas, and hydroel ectri ci ty were substi tuted for resi d-
ual fuel oi l i n el ectri ci ty generati on, l eadi ng to a
rel ati ve decrease i n the demand for resi dual fuel oi l .
When the gap between gasol i ne and di esel pri ces
wi dened (due to di esel subsi di es and/or gasol i ne
taxes), consumers swi tched to di esel cars. As a
resul t, the structure of demand for petrol eum prod-
ucts changed i n many countri es; the demand for the
mi ddl e di sti l l ates i ncreased rel ati ve to the demand
for gasol i ne at the top end and resi dual fuel at the
bottom. Devel opi ng country refi neri es, whi ch typi -
cal l y do not have secondary conversi on technol ogy,
coul d not adjust to these changes. As a resul t, some
of thei r surpl us petrol eum products had to be
exported, often at di stress pri ces, whi l e other prod-
ucts had to be i mported.
28
Because of these devel opments i n the i nterna-
ti onal petrol eum market, several devel opi ng coun-
tri es are produci ng refi ned petrol eum products at
costs hi gher than those prevai l i ng on the i nterna-
ti onal market. Up unti l recentl y, for exampl e,
Li beri a had onl y one refi nery, whi ch was poorl y
mai ntai ned, i neffi ci ent, and i n need of upgradi ng.
An economi c anal ysi s of thi s refi nery showed that
i mporti ng the refi ned petrol eum products di rectl y
and shutti ng down the i neffi ci ent domesti c refi nery
woul d resul t i n net savi ngs of $15 mi l l i on to $20
mi l l i on per year (U.S. dol l ars), equi val ent to a gai n
of about 2 percent i n Li beri as gross domesti c
product (GDP) .29 Several other Afri can countri es are
i n si mi l ar si tuati ons.
The di stri buti on system for petrol eum products
pl ays a key rol e i n determi ni ng whether or not the
economy has suffi ci ent and dependabl e fuel sup-
pl i es. Because l arge porti ons of the popul ati on are
scattered i n rural areas, and transport i nfrastructures
are frequentl y i nadequate, the di stri buti on systems
i n devel opi ng countri es are often unrel i abl e. I n
addi ti on, pri ce control s, suppl y monopol i es, rati on-
i ng systems, and requi rements for uni form pri ci ng i n
al l areas may further l i mi t the dependabi l i ty of
petrol eum i n devel opi ng countri es.
Natural Gas
Natural gas pl ays a rel ati vel y mi nor rol e i n the
energy suppl y system i n most devel opi ng countri es,
suppl yi ng onl y 7 percent of total energy use i n the
devel opi ng worl d (see tabl e 4-l ). Thi s contrasts
sharpl y wi th i ts rol e i n the i ndustri al i zed worl d. I n
the Uni ted States, for exampl e, natural gas suppl i es
22 percent of domesti c energy consumpti on. How-
ever, gas i s an i mportant source of energy for a smal l
number of devel opi ng countri es. Fi ve countri es
Mexi co, Argenti na, Venezuel a, Chi na, and Al geri a
zTRefinery losses re~t from lage, evaporation, and spills; they are dkdnct frOIII Own use co
IISUInptiO& which is Oti iIlttdiOlldly COIISUnled
in the refinery process for hea~ transpo~ and drive. See T. Gorton, op. cit., footnote 26.
XL. Wijetileke and A. ody, World Refinery Industry-The Need for Restructuring,
World Bank Technical Paper No. 32, Washington+ DC, 1984.
~. Wilbanks and S. Wright Energy for Development: ORNL Returnsto the l%ird World, ORNL Review, No. 3, 1988.
account for 58 percent of devel opi ng worl d natural
gas consumpti on. Natural gas consumpti on i n the
devel opi ng worl d i s concentrated i n i ndustry, where
i t i s used both as a fuel and as a feedstock. Natural
gas i s al so used for el ectri ci ty generati on.
Di scovered gas reserves i n many devel opi ng
countri es are not devel oped. The current producti on
l evel s from devel oped reserves are onl y 16 percent
of the l evel that current proven reserves coul d
sustai n.
30
Thus, many devel opi ng countri es i mport
l arge quanti ti es of crude or fuel oi l whi l e possessi ng
reserves of natural gas that coul d serve more
economi cal l y and wi th l ess harm to the envi ron-
ment.
31
The reasons for the rel ati ve underuti l i zati on of
natural gas i n the devel opi ng worl d l i e more i n
i nsti tuti onal than i n techni cal or fi nanci al con-
strai nts. Al though gas i s associ ated wi th oi l expl ora-
ti on and devel opment and therefore shares many of
the same probl ems, i t has addi ti onal di ffi cul ti es of i ts
own. I n the expl orati on phase, due to the fi scal or
contractual terms under whi ch gas i s di scovered, gas
di scoveri es are often treated as dry hol es by oi l
expl orati on compani es.
32
I n the devel opment phase,
gas requi res heavy front-end capi tal i nvestments
(pi pel i nes from produci ng to consumi ng regi ons, or,
i n the case of overseas trade, costl y faci l i ti es and
tankers to l i quefy and transport the gas) and l ong-
term agreements between suppl i ers and consumers.
And the hi gh cost of bui l di ng di stri buti on systems i n
ci ti es deters the devel opment of domesti c markets
for natural gas.
The speci fi c requi rements of gas devel opment are
parti cul arl y di ffi cul t for devel opi ng countri es, where
fi nanci ng for l arge capi tal -i ntensi ve projects i s hard
to fi nd, and where the mai n purchaser for the gas i s
the state. When the devel oper i s a pri vate forei gn
corporati on, the probl ems are compl i cated by uncer-
tai nty that suffi ci ent forei gn exchange wi l l be
avai l abl e for the i nvestor to repatri ate profi ts. Thi s
l ast di ffi cul ty i s especi al l y acute i n the case of hi ghl y
i ndebted devel opi ng countri es where the bul k of
scarce hard currency i s ear mar ked i n advance for
payments on debt. Unl i ke the el ectri ci ty sector,
whi ch shares many of these characteri sti cs, devel op-
ment of natural gas reserves has general l y not been
fi nanced by donor agenci es.
Biomass Fuels
Bi omass fuel s are an i mportant source of energy
i n the devel opi ng countri es, suppl yi ng over three-
fourths of the total energy consumed i n al most al l of
the l ower i ncome devel opi ng countri es.
33
The con-
tri buti on of bi omass fuel s to total energy suppl i es i n
the enti re devel opi ng worl d i s uncl ear. Bi omass fuel
consumpti on i s di ffi cul t to measure, as much of i t
never enters a commerci al market. As shown i n tabl e
4-1, bi omass fuel s suppl y about 19 percent of total
energy accordi ng to the I nternati onal Energy Agency.
Other researchers, however, esti mate thi s number at
33 to 35 percent.
34
Bi omass fuel s are the domi nant energy source i n
rural areas, and they are al so wi del y used i n poorer
urban areas, as wel l as i n some l arge-scal e i ndustri al
processes. These tradi ti onal fuel s consi st of
wood (fi rewood and charcoal ), dung (from cattl e and
other ani mal s), and crop wastes (e.g., wheat, ri ce
straw, and sugar cane bagasse). Wood i s the most
wi del y used and preferred fuel due to i ts superi or
combusti on characteri sti cs. I ts share of total bi omass
energy suppl y vari es wi del y among the devel opi ng
countri es, accordi ng to regi on and agri cul tural and
forestry resource base. I n many regi ons of Afri ca and
Lati n Ameri ca (wi th the i mportant excepti on of
Brazi l , where bagasse i s al so used), wood i s the
pri mary form of bi omass energy used. I n Asi an
countri es, wood remai ns the domi nant bi omass fuel
(accounti ng for one-hal f of al l bi omass consumed i n
World B* op. cit., foomote 12, p. 36.
SITO mation @t tie SUb-S_ ~can countries with undeveloped gas fields: Mozambique, Ethiopia, SOmal@ hh&g~Ca, Cote dIvohe,
Equatorial Guinez Suda~ Senegal, Tanzani A and Namibia.
3ZBY one es~te, about ~ of tie na~~ gas pr~uc~ in the developing world is flared or othenvke witskd. me Compmable n-r for *-
Europe is 2 percent. A. Mashayekhi, Natural Gas Supply and Deman din IAXS Developed Countries, Annual Review of Energy, vol. 13, 1988, pp.
119-129.
33~e role of bioms fiel~ fi tie tom ~tim ~pply ~~es ~eady ~o% co~tries, show@ genefiy m inv~se relationship with GNPper capita.
For example, Ethiopia (GNP percapita= $U.S. 130) meets 92 percent of its energy needs with biomass fuels, while Argentina (GNPper capita= $U.S.
2,390) meets only 7 percent of its needs with biomass fuels.
3lK. Smiti, me Biofiel T~itio~ Pacific andAsian Journul ofEnergy, 1987, pp. 13-32; P. OKeefe, J. SOUSS@l, B. MWM1OW, ~d D. SP~ce*
Wood Energy in Eastern and Southern Afric~ Annual Review of Energy, vol. 14, 1989, pp. 445-468.
Chapter 4--Energy Supplies in the Developing World .101
China, and three-quarters i n I ndi a), but crop wastes
and ani mal dung al so pl ay a si gni fi cant rol e.
35
Biomass Resource Base
For a vari ety of reasons, the fuel wood suppl y base
i s shri nki ng rapi dl y. Thi s coul d have seri ous i mpacts
on the popul ati ons that depend on forests for fuel ,
food, fi ber, fodder, and other needs.
The total gl obal annual growth of forest bi omass
i s subject to great uncertai nty, but has been esti -
mated to be about 50 ti mes annual wood consump-
ti on or fi ve ti mes total annual energy consumpti on,
i ncl udi ng fossi l fuel s. Despi te thi s apparentl y l arge
average gl obal suppl y, there are acute and growi ng
shortages of fuel wood both l ocal l y and regi onal l y.
Some regi ons, such as Asi a, have very l i ttl e forest
stock per capi ta (tabl e 4-3). Wi thi n regi ons, some
countri es are wel l endowed wi th bi omass energy
resources, whi l e others have total l y i nadequate
suppl i es; and wi thi n countri es themsel ves, l ocal
abundances and shortages are common. Zai re, for
exampl e, consumes onl y 2 percent of i ts sustai nabl e
yi el d of forest bi omass but has seri ous deforestati on
around Ki nshasa.
36
Despi te the uncertai nti es regardi ng rates of bi o-
mass energy use and suppl y, i t i s cl ear that the
popul ati ons affected by fuel wood shortages are
i ncreasi ng. The Uni ted Nati ons Food and Agri cul -
ture Organi zati on (UNFAO) has esti mated that the
number of peopl e sufferi ng acute shortages of
fuel wood wi l l i ncrease from about 100 mi l l i on i n
1980 to over 350 mi l l i on i n the year 2000. Such
shortages i ncrease fuel costs for urban dwel l ers,
l engthen the ti me spent foragi ng for fuel by rural
dwel l ers, and rob the soi l of nutri ents as peopl e
swi tch from wood to crop wastes and dung (al though
the i mpact of thi s nutri ent l oss maybe l i mi ted except
i n the much l onger term).
Rural Biomass Markets
Much of the bi omass fuel suppl y i n devel opi ng
countri es-especi al l y twi gs, branches, dung, and
crop wastesi s gathered l ocal l y and used by fami l y
members wi thout enteri ng commerci al i zed markets.
Table 4-3-Biomass Energy Resources in Selected
Sustainable energy yield
(GJ per capita per year
Crop Animal
Country Wood residues dung
Congo . . . . . . . . . . . . . 570
Brazil . . . . . . . . . . . . . . 350
Zaire . . . . . . . . . . . . . . 135
Argentina . . . . . . . . . . . 123
Thailand . . . . . . . . . . . . 37
Nepal . . . . . . . . . . . . . . 21
Burkina Faso. . . . . . . . 10
India . . . . . . . . . . . . . . 7
Bangladesh . . . . . . . . . 2
China . . . . . . . . . . . . . .
1
8
1
25
9
7
5
5
4
7
16
1
40
4
13
7
6
5
3
data not available or not applicable
SOURCES: G. Barnard and L. Kristoffereon, AgfkWtura/ ReskJues as Fuel
in the ZWd Wbrki (London: Earthsoan, 1985); D. Hall, G.
Barnard, and P. Moee, Biomass for Enetgy in the Developing
Countriee, Pergamon Preee, 1982. R. Moea and W. Morgan,
Fuelwood and Rural Energy Producbon and Supply in the
Humid Tropics, (Dublin, Ireland: Tyoooiy International Publish-
ing Ltd, 1981).
These suppl i es are gathered free of charge (i f the
consi derabl e cost of the l abor used i n gatheri ng i s not
i ncl uded) from fi el ds, hedgerows, gardens, and
nearby forest l ands. I n some cases, however, the
poor may have to pay wi th l abor servi ces for the
pri vi l ege of gatheri ng bi omass fuel s from pri vatel y
owned l and.
Commercial Biomass Markets
Bi omass fuel s, notabl y l ogs and charcoal , are al so
traded i n commerci al markets far from thei r ori gi n i n
government and pri vate forests, farms, or pl anta-
ti ons. Low-i ncome urban househol ds and smal l
commerci al enterpri ses use the bul k of these fuel s.
I n some cases, however, bi omass fuel s are used for
advanced i ndustri al appl i cati ons, as i n the case of
charcoal for i ron smel ti ng i n Brazi l . I n such cases,
the i ndustri al users often organi ze the bi omass fuel
suppl i es.
Unl i ke other forms of energy, suppl i es of com-
merci al i zed bi omass fuel s are l argel y i n the hands of
the pri vate sector. Much of the fuel wood may be
grown on pri vatel y owned l and,
37
and the transport
and di stri buti on charnel s for commerci al i zed bi o-
Sscrop wastes account for one-hal.fof total traditional energy supply in China ~d just OV= 10 ~ent in mdia. ~
dung accounts for about 20
pexcent of traditional fuel use in India, PakistaxL and Bangladesh but under 2 percent in China.
~R. MOSS and W. Morgan, Fuelwood and Rural Energy Production and Supply in the Humid fiopics (Dubti b~d: UCOOIY ~~tiod
Publishing Ltd., 1981).
svksev~~~citiw, forexarnple, government lands were found to provide less than IOpereent of total fuelwood supplies (M. -J. mdey,
and A. Reddy, Fuelwood Use in the Cities of the Developing World: Two Case Studies From Ire@ NaruraZResources Foq vol. 9, No. 3, 1985).
mass, fuel s are typi cal l y i n pri vate hands, as are
charcoal ki l ns.
38
On the other hand, the fuel wood trade i s often
subject to government regul ati on, wi th stri ct rul es
about cutti ng trees i n government forests and even
on pri vate l ands. Al though i t i s bel i eved that such
regul ati ons are not stri ctl y enforced, often because
of the di ffi cul ty of enforcement, proscri pti ons agai nst
cutti ng trees can di scourage the devel opment of
l ong-term suppl i es, as farmers and others are unwi l l -
i ng to i nvest i n tree pl anti ng for fuel wood i f they
have no assurance that they can harvest the trees at
maturi ty. Fuel wood pri ces may al so be subject to
pri ce control s (i n Senegal , for exampl e, charcoal
pri ces are control l ed by the government). And i n
some cases, governments may pl ay a rol e i n the
di stri buti on system as wel l . Compared wi th com-
merci al fuel s, however, the bi omass trade i s rel a-
ti vel y unregul ated.
The transport of wood and charcoal to urban areas
i s carri ed out i n a vari ety of ways. I n I ndi a, poor
women carry head l oads of fuel to urban markets; i n
Ni ger, camel s carry fuel i nto the capi tal ci ty of
Ni amey; and el sewhere fuel i s carri ed by bi cycl e,
ani mal cart, moped, and other means. I n hi gher
i ncome areas, trucks or trai ns carry the bul k of the
fuel .
Charcoal
In rural areas, the cutti ng of fuel wood and i ts
conversi on to charcoal i s a major source of i ncome
and nonagri cul tural empl oyment. Charcoal i s made
by stacki ng the wood, coveri ng i t wi th a l ayer of di rt,
and l etti ng i t burn wi th a l i mi ted suppl y of ai r. The
effi ci ency of converti ng wood to charcoal i n these
si mpl e earthen ki l ns i s qui te l ow, typi cal l y rangi ng
from 40 to 60 percent.
39
I f a capi tal i nvestment i s
made, rangi ng from a few hundred dol l ars for si mpl e
modi fi cati ons to tradi ti onal ki l ns
40
to $100,000 or
mor e for a moder n conti nuous r etor t, hi gher ener gy
effi ci enci es can be achi eved.
Al though i t i s wi del y bel i eved that charcoal i s
cheaper to transport than wood due to i ts hi gher
energy content by wei ght, detai l ed studi es have
found that the transportati on costs for wood and
charcoal are about the same.
41
The hi gher energy
content of charcoal per uni t wei ght i s counter-
bal anced by i ts l ower wei ght per uni t vol ume.
Despi te i ts hi gher pri ce, charcoal i s wi del y used i n
some countri es, parti cul arl y i n urban areas where
peopl e have cash i ncomes. A 1970 report from
Thai l and, for exampl e, i ndi cated that 90 percent of
the wood cut for urban markets was converted to
char coal .
42
I t has several i mportant advantages over
wood. Charcoal i s i mpervi ous to i nsect attack,
unl i ke some wood speci es that must be used wi thi n
as l i ttl e as a month of dryi ng to avoi d si gni fi cant
l osses to ter mi tes.
43
AS i t i s nearl y smokel ess,
charcoal cooki ng can be done i ndoors i n rel ati ve
comfort wi thout bl ackeni ng wal l s or metal pots wi th
soot. I n addi ti on, charcoal causes l i ttl e smoke
i rri tati on to eyes or l ungs. Al though i t can emi t l arge
amounts of dangerous carbon monoxi de and other
pol l utants, whi ch i s a heal th hazard i n poorl y
venti l ated ki tchens, charcoal causes l i ttl e obvi ous
di scomfort to the user. Addi ti onal l y, once l i t,
charcoal fi res need l i ttl e attenti on from the cook,
whereas wood fi res requi re frequent adjusti ng of the
fuel .
Biomass Pricing
When peopl e move from rural to urban areas i n
devel opi ng countri es, they typi cal l y conti nue to
fol l ow tradi ti onal patterns of bi omass fuel use. I n
contrast to the l abor-i ntensi ve col l ecti on of bi omass
fuel s i n rural areas, however, the urban poor often
have no choi ce but to pur chase fuel wood or char coal
i n commerci al markets. I n Tanzani a, the cost of
purchasi ng these fuel s reportedl y ranges as hi gh as
38M. ~, J . Dunkerley, and A. Reddy, Ibid.
sgc~rcoal Production Improvement for Rural Development in Ttiiland @an@o& ~
d: Royal Thai Government and U.S. Agency for
International Development, 1984); D. Earl CharcoalProduction, Sudan Renewable Energy Project, Energy Research Council, USAID, Report No. 002,
Khartoum, Su&rL February 1984; J. Wartluft and S. White, Comparing Simple Charcoal Proaluxion Technologies for the Caribbean (Arlirigto!L VA:
Volunteers in Technical Assistance, 1984).
~. ChristopherseU G. Karc4 and J . Seve, Production and Transportation of Fuelwood and Charcoal From Wood Surplus to Deilcit Regions in
Niger: Technical and Economic Feasibility (Washington DC: Energy/Development International, March 1988).
41T. Wood and S. Baldwiq Fuelwood and Charcoal Use in Developing Countries, Annual Review of Energy, vol. 10, 1985, pp. 407-429.
4ZJ0 hold, ctWood Energy and Rural Communities,
Natural Resources Forum, vol. 3, 1979, pp. 229-252.
43s~on~ono~ and Bent SoreWW A Na~ ~=m Smdy ~ T~fi: me ~ of B~dilya Vtige, NamraZResource Forum, VO1. 8, No.
1, 1984, pp. 51-62.
Chapter 4--Energy Supplies in the Developing World . 103
40 percent of the i ncome of poor fami l i es.
44
Mor e
typi cal l y, energy accounts for 5 to 10 percent of the
expendi tures of poor househol ds.
45
Bi omass fuel pri ces i n urban markets often ri se
rapi dl y as wood resources are seri ousl y depl eted,
and then cl osel y fol l ow fossi l fuel markets. Bi omass
costs cannot ri se very far above the cost of an
equi val ent amount of useful energy from kerosene or
LPG, as users can and wi l l then swi tch fuel s.
%
Fami l i es that purchase modern stoves and fuel s,
however, rarel y di scard the ol der stoves. Mai ntai n-
i ng both technol ogi es al l ows easy and fl exi bl e
swi tchi ng between fuel s i n response to avai l abi l i ty
and pri ce. Fol l owi ng the 1973 and 1979 oi l pri ce
i ncreases, for exampl e, many peopl e swi tched back
to wood and charcoal for thei r cooki ng needs. I n
Mal awi the use of kerosene, pri mari l y for cooki ng
and l i ghti ng, decl i ned by 24 percent between 1973
and 1976.
47
Electricity
El ectri fi cati on pl ays a central rol e i n promoti ng
economi c and soci al devel opment i n any nati on. At
the same ti me, the el ectri ci ty sector consumes l arge
amounts of economi c, soci al , and envi ronmental
resources. Accordi ngl y, the el ectri c power sector
recei ves si gni fi cant attenti on and resources from
both devel opi ng country governments and i nterna-
ti onal devel opment agenci es. For exampl e, the
Worl d Bank di rects over 80 percent of i ts energy
l endi ng to the el ectri ci ty sector.
Al though el ectri ci ty accounts for l ess than 9
percent of the energy used by consumers i n devel op-
i ng countri es,
48
el ectri ci ty producti on i n the devel -
opi ng worl d i s i ncreasi ng rapi dl y, at an average
annual rate of 7.6 percent.
49
However, thi s rapi d
growth sti l l l eaves the devel opi ng worl d at a far
l ower l evel of el ectri ci ty producti on than the i ndus-
tri al i zed worl d: average annual el ectri ci ty produc-
ti on i n the devel opi ng worl d i s about 520 ki l owatt-
Figure 4-5--Electricity Generation by Fuel Type in
Selected Regions of the Developing World, 1987
Thousand GWh
600
500
400
300
200
100
0
m Hydr o/ ot her *
= N u c l e a r
m Nat ural Gas
U O i l
- C o a l
Africa Latin America India China
Includes hydropower, geothermal, and other renewable.
SOURCE: International Energy Ageney, Wor/d Energy Statistics and
Balances 1971-1987 (Paris: OECD, 1989).
hours (kWh) per capi ta,
50
as compared wi th the U.S.
fi gure of 10,500 kWh per capi ta.
51
Electricity Generation
In the devel opi ng worl d, as i n the Uni ted States,
a vari ety of fuel s and technol ogi es are used for
el ectri ci ty generati on (see fi gure 4-5 and tabl e 4-4).
Coal use for el ectri ci ty producti on i n the devel op-
i ng worl d i s concentrated, l argel y i n countri es wi th
si zabl e domesti c coal reserves (I ndi a, Chi na, and
South Afri ca). Si mi l arl y, natural gas generati ng
capaci ty i s found mai nl y i n the few countri es where
natural gas i s produced domesti cal l y. I n other
countri es, oi l i s often used for el ectri ci ty generati on.
Hydroel ectri c faci l i ti es range from mi crohydro-
power stati ons wi th l ess than 0.1 megawatt (MW) of
capaci ty to l arge-scal e hydropower pl ants such as
the 12,600 MW I tai pu faci l i ty i n Brazi l . For coun-
tri es wi th hydroel ectri c potenti al , hydropower offers
an i ndi genous source of el ectri ci ty generati on wi th
very l ow operati ng costs, al though the capi tal costs
44E.M. MIWW4 village Industries vs. Savannah Forests,
UNASYLVA, vol. 33, No. 131, 1981, Pp. 24-29.
ds&r~d ~achand -ia ~wen, Household Energy Handboolq World B* TKhnicd Paper No. 67, 1987, p. 50.
~Doug~ F. Barnes,
Understanding Fuelwood Prices in Developing Nations, World BanlL Household Energy Unit, Industry and Energy
Departmen~ Oct. 31, 1989.
47J . -old,Wood Energy and Rural Communities,
Natural Resources Forurn, vol. 3, 1979, pp. 229-252.
~IEA, op. cit., footnote 4, pp. 112, 120, 124, 128.
d~or the period 1971-87. IEA, op. cit., fOO~Ote 4.
~.S. Agency forrnternational Development PowerShortages in Developing Countries: Magnitude, Impacts, Solutions andtheRole of the Private
Sector (Washingto~ DC: (Mllce of Ermgy, U.S. AID, March 1988), p. 2.
51E~er~ ~o-tion A_s~tio~ Ann~Z Energy Re~ew, DoE/EIA-oq&$(gg) (washingto~ DC: U.S. Govermnent Prhting ~1~, 1989).
Table 4-4-Electricity Generation by Fuel, 1987 (percent by kWh delivered)
Fuel Africa Latin America China India United States
coal . . . . . . . . . . . . . . . . . . . . . . . . 52 3 68 66 57
oi l . . . . . . . . . . . . . . . . . . . . . . . . . 14 24 12 8 5
Natural gas.... . . . . . . . . . . . . . . 14 8
*
1 11
Nuclear . . . . . . . . . . . . . . . . . . . . . 3 1 0 2 18
Hydro/other
a
. . . . . . . . . . . . . . . . . 18 63 20 23 10
Total (percent) . . . . . . . . . . . . . 100 100 100 100 100
Total(GWh) . . . . . . . . . . . . . . . 283,340 520,290 497,320 217,500 2,732,530
Less than l%.
a lncludes hydropower, geothermal, and other renewables.
NOTE: Does not include heat losses. Totals may not add to IOO percent due to rounding.
SOURCE: International Enerav Aaencv, Wm/d,Enenw Statistics and Baiances 1971-1987 (Paris: OECD, 1989): EA. Eneruv Ba/ances of OECD Countries
. , . -.
1 9 8 7 - 1 9 8 8 (Paris:OE~D, l@O).
are hi gh. I n Lati n Ameri ca, hydropower pl ays a l arge
rol e, suppl yi ng al most two-thi rds of total el ectri c-
i ty.
52
More than 90 percent of Brazi l s total el ectri c-
i ty suppl y comes from hydropower.
53
I n Afri ca,
al most hal f of the el ectri ci ty generated outsi de of
South Afri ca comes from hydropower,
54
and more
than two-thi rds of the countri es i n Afri ca produce
el ectri ci ty from hydroel ectri c pl ants.
55
Onl y ei ght devel opi ng countri es were generati ng
el ectri ci ty from nucl ear powerpl ants as of l ate
1989,
56
al though several more have pl ants pl anned
or under constructi on.
57
Nucl ear power generati on
i nvol ves hi gh capi tal costs, very l arge scal es of
operati on, and consi derabl e techni cal experti se. I n
two upper i ncome devel opi ng countri es, Tai wan and
Korea, nucl ear power provi des over 40 percent of
total el ectri ci ty generati on. I n other countri es, how-
ever, there have been some di ffi cul ti es wi th nucl ear
power--e.g., two of I ndi as three operati ng nucl ear
pl ants have suffered hi gh outage rates,
58
and thei r
fourth pl ant has had constructi on del ays of 7 years.
A number of al ternati ve generati on technol ogi es
have al so been used i n devel opi ng countri es, i ncl ud-
i ng cogenerati on, geothermal , conversi on of sol ar
energy to el ectri ci ty, wi nd-dri ven generators, and
the burni ng of waste materi al . Al though these
sources contri bute rel ati vel y smal l amounts to total
el ectri ci ty suppl i es i n the devel opi ng worl d, there i s
strong pol i cy i nterest i n expandi ng the use of these
al ternati ve generati ng technol ogi es.
The operati ng effi ci enci es of el ectri ci ty gener-
ati ng pl ants are general l y l ower i n devel opi ng
countri es than i n the Uni ted States.
59
Al though many
factors affect power pl ant effi ci ency (notabl y i nput
fuel qual i ty), the use of l ess effi ci ent, poorl y
mai ntai ned technol ogi es contri butes to the l ow
effi ci enci es of devel opi ng worl d powerpl ants.
Electricity Transmission and Distribution
Transmi ssi on and di stri buti on systems i n the
devel opi ng worl d have rel ati vel y hi gh l osses. A
recent survey of 76 devel opi ng countri es found that,
i n one-hal f of the countri es surveyed, transmi ssi on
and di stri buti on l osses (as a share of total genera-
ti on) exceeded 15 percent, compared wi th typi cal
l osses of l ess than 10 percent i n the i ndustri al i zed
countri es.
60
These l osses i ncl ude both techni cal
l osses and unmetered consumpti on (theft).
521E~ op. cit., footnote 4.
%id.
~lJnit~ Nations, op. cit., footnote 24.
ss~id.
56These ~ ~m~ Br@ ~di% Mefico, p~s~q Republic of Kor~ South Afli@ ~d ~WSn. United Nations, ~erW @lorstion ~
Development Trends in Developing Countries, Report of the Secretary-Genera May 14, 1990.
57~gen~ B- _ ~~ M~co, ~~ ~d R~ublic of Korea all had nuclear powerplants under Construction ~ 1989. rbid.
sgTa@ mer~ Research Institute, op. cit., footnote 11.
5~~t~ Natiom Cotierence o
n
T-e ~d Development ~c~), T!chno/ogy Policy in the Energy ~ec~or. ]SSUSS, Scope and options f o r
Developing Counzi+es, UNCTADfIT~, June 15, 1989.
%bid.
Institutional Issues
Governments i n both the devel opi ng and i ndustri -
al i zed worl d general l y have taken l eadi ng rol es i n
di recti ng the devel opment and operati on of the
el ectri c power sector; thi s refl ects both the i mpor-
tance of el ectri c power i n meeti ng economi c and
soci al objecti ves and the hi gh cost of el ectri c power
systems.
Because el ectri ci ty i s consi dered an i mportant
tool wi thi n a broader nati onal devel opment strat-
egy
61
, developing countries often subsi di ze el ec-
tri ci ty pri ces. I n Paki stan, for exampl e, 60 percent of
the cost of el ectri ci ty i s subsi di zed; i n I ndi a, the
fi gure i s 20 percent.
62
Many countri es have subsi -
di es for el ectri ci ty i n sel ected sectors (e.g., agri cul -
ture and resi denti al ), refl ecti ng ei ther the pol i ti cal
i nfl uence of the subsi di zed sectors or government
i nterest i n promoti ng certai n economi c or soci al
ends. Al though el ectri ci ty pri ces nearl y doubl ed
between 1980-81 and 1986-87 i n I ndi a, for exampl e,
current pri ces are sti l l far l ower than suppl y costs for
resi denti al and agri cul tural consumers.
63
There may be many reasons for subsi di zi ng
el ectri ci ty pri ces i n di fferent sectors. For exampl e,
rural el ectri fi cati on i s promoted as a means to reduce
mi grati on to ci ti es by i mprovi ng economi c opportu-
ni ti es and l i festyl es i n rural areas. Supporters argue
that these subsi di es benefi t soci ety as a whol e and
not just rural consumers. Others rogue that there i s
no concl usi ve evi dence that r ur al el ectr i fi cati on can
actual l y produce thi s benefi t.
64
Rural el ectri fi cati on
i s al so seen as a powerful i nstrument at the di sposal
of central governments to foster pol i ti cal stabi l i ty i n
rural areas, al though agai n the evi dence i s mi xed.
65
Subsi di zed el ectri ci ty pri ces al so al l ow el ectri ci ty
to compete wi th subsi di zed pri ces for al ternati ve
fuel s. Si mi l arl y, when pri ces of farm products are
control l ed and kept arti fi ci al l y l ow, el ectri ci ty pri ces
(and the pri ces of other i nputs such as ferti l i zers) are
often subsi di zed to keep farms operati ng.
66
Subsi di zed el ectri ci ty can have negati ve i mpacts,
i ncl udi ng contri buti ng to power shortages, si nce
consumers tend to use more subsi di zed el ectri ci ty
than they woul d otherwi se. Moreover, pri ce subsi -
di es keep power company revenues at l evel s i nade-
quate for devel opi ng addi ti onal suppl i es and even
for mai ntai ni ng exi sti ng faci l i ti es effi ci entl y. Hi gher
pri ces, on the other hand, l i mi t suppl i es to hi gher
i ncome groups, an outcome that may confl i ct wi th
the soci al goal s of el ectri fi cati on.
Capi tal spendi ng on el ectri ci ty systems i n the
devel opi ng worl d i s currentl y esti mated at $50
bi l l i on to $60 bi l l i on annual l y.
67
Even at that hi gh
l evel of expendi ture, i nvestment i s expected to be
i nadequate to meet demand. The Uni ted States
Agency for I nternati onal Devel opment (U.S. AI D)
has esti mated that meeti ng the growi ng demand for
el ectri ci ty wi l l requi re capi tal i nvestment of around
$125 bi l l i on per year over the next two decades.
68
Thi s enormous capi tal mobi l i zati on requi rement
represents a l arge fracti on of both total economi c
acti vi ty and total gross domesti c i nvestment i n the
devel opi ng worl d. The total economi c output of al l
l ower- and mi ddl e-i ncome countri es as measured by
GDP was $2,716 bi l l i on i n 1987, wi th total gross
domesti c i nvestment of $662 bi l l i on.
69
Much of the
capi tal costs of el ectri ci ty pl ants must be pai d i n
forei gn exchange,
70
l eadi ng to bal ance of payment
probl ems and compoundi ng the probl ems of hi gh
operati ng and capi tal costs i n the el ectri c sector.
61ufit~ N8ti~ns ce~tm for Human Setiements @bitat), Guidelines for the ~j anni ng of Rural Settzemnts andrnfim~tire: Electrij?catioA
Methodology (Nairobi, Kenya: United Nations, 1985), p. 43.
6~.s. Agacy for ~ternatio~ Development, Op. Cit., footnote 50, P. 26.
63A. FW~ et ~e, Application of De~nd~ide ~a~ge~nt (DS~) to Relieve Elecmci~ sho~ges i n I ndi a, ~cont r act or I epol l pI epaI ed fO1
the Otllce of Technology Assessrnen~ April 1990, p. 59.
~For ~ discmsion of ~ese ismes, s= D. Balyles, E1ecm~ ~owerfor Rural Growth @o~der, C(): Westview Press, 1987), pp. 109-118.
~United Nations Center for Human
Settlements (Habitat), op. cit., footnote 61.
66Mo~ Munasin@e, Rural E/ecmfication for Development (Bouldex, CO: Weswiew pr~s, 1987), P. 247.
67u.s. Agency for International Development, op. cit., fOOlnOte 50, P. iv.
68fiid., p. 25. A Sli@fly lower es~te ($60 t. $100 billion) from tie World B@is gi ven i xI A. ~UI C~l ~dRo Saundm, Fi MI I chg Of the ~lgy
Sector in Developing Countries, World Bank, Industry and Energy Department Worldng Paper Energy Series, Paper No. 14, April 1989.
@World BanlG World Development Report 1989 (New Yorb NY: Oxford University Press, 1989), tables 3 and 9.
W.S. AID estimates that around 45 percent of capital investment will be in foreign exchange. U.S. Agency for International Development footnote
50, p. 25.
33-718 0 - 90 - 5
Even wi th the enormous fi nanci al , techni cal , and
i nsti tuti onal effort targeted to the devel opment of the
el ectri ci ty sector over the past decades, capaci ty sti l l
i s often i nsuffi ci ent to provi de rel i abl e, hi gh-qual i ty
power i n devel opi ng countri es. Outages are common
i n many countri es. For exampl e, Bombay experi -
enced 1,000 outages annual l y over a recent 5-year
peri od.
71
Even when power i s avai l abl e, vol tage
fl uctuati ons are often extreme, restri cti ng the use of
some types of equi pment. The el ectroni c ci rcui ts of
todays compact fl uorescent l i ght bul bs, for exam-
pl e, do not tol erate wi de vol tage fl uctuati ons;
72
and
computer operati ons are di srupted by outages. As a
resul t, i n many countri es, poor power qual i ty and
l ack of rel i abi l i ty undermi ne the economi c benefi ts
of el ectri c power.
Low rel i abi l i ty resul ts i n formi dabl e l osses i n
economi c producti vi ty. Load sheddi ng i n I ndi a i s
esti mated to cost the equi val ent of 1 to 3 percent of
GDP annual l y.
73
Accuratel y esti mati ng the produc-
ti vi ty l ost when exi sti ng equi pment cannot be
operated due to power outages i s di ffi cul t, and
esti mati ng producti vi ty l ost as i ndustry forgoes the
purchase and use of new el ectri c equi pment i s even
more uncertai n; however, the i mpacts may be qui te
l arge.
I roni cal l y, a few devel opi ng countri es suffer from
an excess of el ectri c capaci ty. For exampl e, i t has
been esti mated that seven East Afri can countri es
have approxi matel y 7,000 MW of excess generati ng
capaci ty (i .e., capaci ty over and above what i s
needed for rel i abl e system operati on) .74 Thi s si tua-
ti on resul ts from the l umpi ness of el ectri c gener-
ati ng faci l i ti es, especi al l y hydropower. Putti ng a
l arge new generati ng faci l i ty i n servi ce before the
domesti c l oad can absorb the new suppl i es resul ts i n
overcapaci ty, and ti es up scarce capi tal .
75
Figure 4-6-Electricity Consumption by Sector in
Selected Regions of the Developing World, 1985
Thousand GWh
: T
= O t h e r
= Commer ci al / publ i c
m Resi dent i al
= Agr i cul t ur al
= I ndust r i al
China India Brazil
SOURCE: International Energy Agency, Wotld Energy Statistics and
Bahrxws 1971 -1987 (Pans: OECD, 1989).
Electricity Consumption
I ndustry consumes most of the devel opi ng worl ds
el ectri ci ty (see fi gure 4-6 and tabl e 4-5). I n contrast,
el ectri ci ty use i n the Uni ted States i s di vi ded among
the i ndustri al , resi denti al , and publ i c servi ce/
commerci al sectors (see tabl e 4-5). Chi na, the l argest
el ectri ci ty user i n the devel opi ng worl d,
76
uses 75
percent of i ts el ectri ci ty i n i ndustry. Si mi l arl y, I ndi a
and Brazi l use over hal f thei r el ectri ci ty i n i ndustry.
Agri cul ture uses l arge amounts of el ectri ci ty for
pumpi ng i n I ndi a and Chi na, whi l e resi denti al
l i ghti ng and appl i ances use l arge amounts of el ec-
tri ci ty i n Brazi l .
Outlook for Improvements
The mix of energy suppl i es vari es wi del y i n the
devel opi ng worl dfrom Chi nas heavy rel i ance on
coal i n the i ndustri al and resi denti al sectors to
Brazi l s extensi ve use of hydropower-based el ectri c-
i ty. Despi te the di versi ty of sources, however,
TITa@fierWRe~Wch ~ti~te, Two Strategie~forE/ectric ~~~velingforz~ia, p- I FM Repofi (New Delhi, India: 1987), p. 6., as Citd
inJ. VanDomele~ Power to Spare: The WorldBank andE2ecmcify Conservation (Washington, DC: World Wildlife Fund and Consemation Foundation
1988) p. 4.
Tz~WenceBe&eley~borat~V, Energy Te~hnologyforDeveloping Countries: Zssuesfor the u.S.Natio~lEnergyS~ategy, LBL-28N7 @erkeley,
CA: December 1989).
T3U.S. Agency for International Development, op. cit., footnote 50, P. 21.
TAI. H~e, ~er~ Efficiency in Developing Countries,
inM. Munasinghe and R. Saunders (eds.), Energy Issues in the Developing World, World
Bank Industry and Energy Department Working Paper, Energy Series Paper No. 1 (Washington DC: World Banlq 1988).
T5Ex~ess ~waci~ cm ~so 1e~ t. distofiom ~ Pficing ~d de~nd. For e~ple, the ~ge SU@US ~pacity in Brazil when Itaipu came On-he kd
the electricity authorities to offer industry exceptionally strong incentives to buy electricity. Electricity was so cheap that it was used by industrial
customers virtually as a boiler fuel. Within a few years, however, load growth in other seetors reduced surplus capacity, but the industrial users were
by then reluctant to give up the highly favorable rates.
Tb@er one-fo~ of the elmtrici~ in the developing world k ~d iII China.
Chapter 4--Energy Supplies in the Developing World .107
Table 4-5-Electricity Consumption by Sector, 1985 (percent)
End-use sector China India Brazil United States
Industry . . . . . . . . . . . . . . . . . . . . . 75
Agriculture . . . . . . . . . . . . . . . . . . 12
Residential . . . . . . . . . . . . . . . . . . 7
Public service/commercial . . . . . 5
Railroads . . . . . . . . . . . . . . . . . . . 2
Other/unspecified . . . . . . . . . . . . . q
Total (percent) . . . . . . . . . . . . . 100
Total generation (GWh) . . . . . 410,700
59
17
12
8
2
1
100
188,500
58
3
20
20
1
*
100
192,700
33
1
35
31
*
o
100
2,621,900
q Less than 1 percent.
NOTES: Totals may not add to 100 percent due to rounding. Brazil, China, and India account for 48 percent of
developing world electricity consumption.
SOURCES: Tata Energy Research Institute, Tf3?/EnergyfJata Directoryzmd Yearbook 1988(New Delhi, India: 1989);
IEA, World Energy Statktics and 6a/ances 197W1985 (Paris: OECD, 1987); E/4, Energy Ba/ances of
OECD Counties 1987- 1988( Pans: OECD, 1990).
several i mportant characteri sti cs of the energy sup-
pl y sector can be i denti fi ed:
q
q
q
The technol ogi es i n use are typi cal l y ol der, l ess
effi ci ent, and l ess sophi sti cated than compara-
bl e technol ogi es i n the i ndustri al i zed countri es.
The el ectr i ci ty sector accounts for a l ar ge shar e
of forei gn exchange resources. Rapi d growth i n
el ectri ci ty demand and the hi gh capi tal requi re-
ments of the sector suggest that the gap be-
tween needed and avai l abl e capi tal for el ec-
tri ci ty system expansi on wi l l wi den. Therefore
the el ectri ci ty systems i n many countri es coul d
conti nue to be rel ati vel y undependabl e, i neffi -
ci ent, and technol ogi cal l y outdated.
The publ i c sector pl ays a domi nant rol e i n most
aspects of energy suppl y, wi th the excepti on of
bi omass.
These characteri sti cs of the energy suppl y sector
i n the devel opi ng worl d suggest a number of
producti ve opportuni ti es for i mprovi ng the effi -
ci ency of the energy suppl y system, recogni zi ng that
there are wi de vari ati ons among devel opi ng coun-
tri es, and that many characteri sti cs that appear
i neffi ci ent or undesi rabl e when vi ewed from the
current perspecti ve of hi ghl y devel oped nati ons may
i n fact represent rati onal choi ces gi ven prevai l i ng
soci al , economi c, and techni cal condi ti ons.
A number of i nsti tuti onal , technol ogi cal , and
engi neeri ng opti ons can be consi dered for i mprovi ng
the extracti on, processi ng, and conversi on of energy
suppl i es. I n oi l and gas devel opment, for exampl e,
opti ons i ncl ude both the depl oyment of new technol -
ogi es, such as hori zontal dri l l i ng, and the devel op-
ment of i nnovati ve fi nanci ng mechani sms. Opti ons
for coal i ncl ude technol ogi es such as washi ng and
screeni ng, as wel l as strategi es for mi nemouth
el ectri ci ty generati on to rel i eve pressure on transport
systems. Si mi l arl y, i nsti tuti onal i ssues are cri ti cal i n
any di scussi on of i mprovi ng the sustai nabi l i ty of
bi omass resources.
Both devel opi ng country governments and i nter-
nati onal devel opment agenci es al ready pay much
attenti on to opportuni ti es i n the el ectri ci ty sector.
Technol ogi cal opportuni ti es range from i ndustri al
cogenerati on, to upgrades of transmi ssi on and di stri -
buti on system effi ci enci es, to the use of more
effi ci ent consumer appl i ances. I nsti tuti onal opportu-
ni ti es i ncl ude the contri buti on of nonuti l i ty genera-
tors to el ectri ci ty networks.
Al though technol ogy can do much to i mprove the
energy suppl y sector, other factors al so affect i ts
operati on: fi nanci al i ssues, such as subsi di es for
el ectri ci ty pri ces or the hi gh cost of natural gas
transportati on; i nsti tuti onal and management i ssues,
i ncl udi ng shortages of trai ned personnel ; and the
i ncenti ve structure, notabl y the domi nant rol e pl ayed
by government i n fossi l fuel expl orati on and del i v-
ery systems, al l strongl y i nfl uence system opera-
ti ons, management, and deci si onmaki ng i n the
energy suppl y sector.
Chapter 5
Energy and the Environment
in Developing Countries
Page
I ntroducti on and Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
The Rural Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Causes of Envi ronmental Degradati on q . . . . . * . . . . . . . . . . . . . * . * . * * + . . . * . * *****..** 113
Envi ronmental I mpacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
The Urban Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Causes of Envi ronmental Degradati on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Envi ronmental I mpacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
Energy Producti on and Conversi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
coal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Oi l and Gas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Hydroel ectri ci ty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Bi omass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Nucl ear Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Sol ar, Wi nd, and Other Renewabl e Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Greenhouse Gases and Gl obal Cl i mate Change . * . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Government Pol i ci es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Figures
Figure Page
5-1. Suspended Parti cul ate Matter Level s i n Sel ected Ci ti es, 1980-84 . . . . . . . . . . . . . . . 125
5-2. Sul fur Di oxi de Level s i n Sel ected Ci ti es, 1980-84 ., * * . . . . . . . . . . . . . . . . . . . . . . . . . 126
Tables
Table
Page
5-1. Causes and Consequences of Envi ronmental Degradati on i n Rural Areas . . . . . . . . 112
5-2. Esti mates of Land Use Changes i n Cl osed and Open Tropi cal Forests,
ci rca 1980 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
5-3. Changes i n Land Area Uses * . * . . . . . . . . . . . . . *, * . . . . * * . * *,, . . * . . . . * *, + *****.... 117
5-4. Sedi mentati on Rates of Some Reservoi rs i n I ndi a
q ........*,***..*.......,*,.* 121
Chapter 5
Energy and the Environment in Developing Countries
Many devel opi ng countri es are experi enci ng si g-
ni fi cant envi ronmental degradati on. I n rural areas,
popul ati on pressure and l ow agri cul tural producti ve-
ti es are among the factors forci ng peopl e i nto
margi nal and ecol ogi cal l y fragi l e l ands. Woodl ands
are bei ng cl eared for cropl and and pastures and are
bei ng commerci al l y l ogged. The use of bi omass for
fuel or fodder pl aces further demands on woodl ands
and grassl ands, parti cul arl y i n ari d regi ons wi th hi gh
popul ati on densi ti es. Farmi ng, ranchi ng, l oggi ng,
and the use of bi omass fuel s are al l necessary i f the
peopl e dependent on these resources are to survi ve.
But these vari ous pressures can al so have negati ve
i mpacts: destructi on of tropi cal forests and bi odi ver-
si ty; deserti fi cati on; soi l erosi on and i ncreased
downstream fl oodi ng and si l tati on; and ai r pol l uti on-
l ocal , regi onal , and gl obal .
I n many urban areas of devel opi ng countri es,
rapi d popul ati on growth, i nadequate i nfrastructure,
and economi c and i ndustri al growth wi th mi ni mal or
i nadequatel y enforced envi ronmental control s have
l ed to hi gh l evel s of pol l uti on. Level s of sul fur
di oxi de, parti cul ate, ground-l evel ozone, and ni tro-
gen oxi des often exceed those i n i ndustri al i zed
countri es. Major sources i ncl ude el ectri ci ty genera-
ti on, transportati on, and i ndustri al producti on. Greater
use of fossi l fuel s i n the modern, pri mari l y urban,
sector can al so l ead to envi ronmental degradati on
and pol l uti on i n the rural areas where these fuel s are
extracted from the ground and transported to the
ci ti es, and where hydroel ectri c faci l i ti es are si ted.
Many have vi ewed envi ronmental costs--degra-
dati on and pol l uti on of the natural resource base-as
the pri ce that must be pai d i n order to devel op
economi cal l y.
l
I ncreasi ngl y, however, others argue
that envi ronmental protecti on and economi c devel -
opment are ti ghtl y i nterconnected and mutual l y
suppor ti ve.
2
The l andl ess peasants who mi grate to
fragi l e watersheds i n order to feed thei r fami l i es, for
exampl e, cl ear l and that i s hi ghl y suscepti bl e to
erosi on. Thi s can l ead to seri ous soi l l oss and
downstream fl oodi ng and si l tati on-ul ti matel y re-
duci ng the producti vi ty of thei r l and as wel l as that
of l and downstream. I n turn, thi s can exacerbate thei r ,
and others poverty. Thi s may be parti cul arl y
si gni fi cant i n devel opi ng countri es because thei r
economi es are so heavi l y dependent on agri cul ture.
For the 2.8 bi l l i on peopl e i n the l owest i ncome
countri es, agri cul ture accounted for ful l y 31 percent
of gross domesti c product (GDP) i n 1987 and an
even greater porti on of empl oyment.
3
Economi c
anal yses that i ncl ude envi ronmental i mpacts are
i mprovi ng knowl edge of thi s i nterdependency and
may assi st pol i cymakers i n more wi sel y maki ng
these tradeoffs.
Energy producti on and use contri bute to envi ron-
mental degradati on i n devel opi ng countri es. Other
contri buti ng factors i ncl ude popul ati on growth,
i nequi tabl e l and tenure, unsustai nabl e agri cul tural
and forestry practi ces, i ndustri al i zati on, and govern-
ment pol i ci es. I n order to understand the rol e of
energy i n overal l envi ronmental degradati on, i t i s
therefore necessary to i ncl ude these other factors i n
the anal ysi s.
Energy, used wi sel y, mi ght al so potenti al l y pro-
vi de several i mportant envi ronmental benefi ts i n
devel opi ng countri es. Greater energy i nputs i nto
agri cul ture i n the form of tracti ve power, ferti l i zer,
and i rri gati on, for exampl e, can substanti al l y i m-
prove agri cul tural producti vi ti es where soi l s and
cl i mates are appropri ate, and mi ght hel p sl ow the
expansi on of agri cul tural l ands necessary to feed a
burgeoni ng popul ati on. (At the same ti me, however,
moder n agr i cul tur e mi ght al so cause envi r onmental
damage: by overuse of pesti ci des, herbi ci des, and
ferti l i zers; by waterl oggi ng and sal i ni zi ng i rri gated
Iclem TiSde~, ~fSuS~ble Deve@=~t: D~e~ persp~tives of ~olo@ts -d &onomists, ad Relev~ce to - s, Wor/d Developnwnt,
vol. 16, No. 3, 1988, pp. 373-384.
world Commission on Environment and Development, Our Common Future (New York NY: Oxford University Press, 1987).
Sworld Ba~ Wor/d Develop~nt Repo~ 1989 (New York, NY: Oxford University Press, 1989), indicator @bles 1 and 3.
111
Table 5-l-Causes and Consequences of Environmental Degradation in Rural Areas
Desertification
Air pollution
Commercial agriculture, ranching
Consequences Direct cause Underlying cause
Deforestation Shifting agriculture Population growth
Loss of biodiversity Permanent agriculture Poverty
Soil erosion Permanent pasture Lack of land tenure
Flooding Low-level agriculture inputs
Mechanization of agriculture and/or the consolida-
tion of agricultural lands
Commercial logging Destructive logging, lack of forest management and
protection, poor reforestation
Increased access to forests along logging roads for
farmers and ranchers
Production for export markets
Fiscal policies and legislation, in part to promote
exports of primary products due to need for
foreign exchange to service debt.
Inappropriate economic valuations of natural re-
sources and biodiversity
Use of Biomass Fuels Inefficient use of fuelwood; overcutting of fuelwood
resources
Use of forest biomass for fodder Shortages and lack of alternative sources of fodder
Agricultural expansion onto fragile lands Population growth
Overgrazing Poverty
Burning of grasslands Lack of land tenure
Low-level agriculture and/or the consolidation of
agriculture lands
Use of biomass fuels Inefficient use of fuelwood; overcutting of fuelwood
resources
Climate change Various; not well understood
Use of biomass fuels Lack of access to higher quality fuels and stoves
Slash and burn agriculture Population growth
Burning of grasslands Poverty
Lack of land tenure
Low-level agricultural inputs
Salinization and water-logging Poor planning and management Cheap or free water contributing to inefficiency
of irrigated lands Inadequate investment in infrastructure
SOURCE: Office of Technology Assessment, 1990.
l ands; and by use of these techni ques under i nappro-
pri ate soi l and cl i mati c condi ti ons).
4
Energy-effi ci ent technol ogi es may offer the po-
tenti al of si mul taneousl y reduci ng the total cost of
del i veri ng energy servi ces and cutti ng pol l uti on.
New i ndustri al processes may i ncrease producti vi ty
and l ower costs whi l e reduci ng hazardous wastes.
Much research and devel opment work remai ns to be
done to bri ng many of these i nnovati ons to frui ti on.
A l ater report from thi s OTA study wi l l exami ne the
extent to whi ch i mproved technol ogi es for energy
producti on and use can contri bute to economi c
devel opment and envi ronmental protecti on; under
what ci rcumstances energy technol ogy i nnovati ons
mi ght ease the di ffi cul t tradeoffs between economi c
growth and the envi ronment; and pol i cy i ssues
associ ated wi th accel erati ng the i ntroducti on of
i mproved energy technol ogi es that mi ni mi ze envi -
ronmental degradati on.
Thi s chapter traces the causes and consequences
of envi ronmental degradati on i n devel opi ng coun-
tri es, begi nni ng i n rural areas and fol l owi ng them
through to urban areas, modern i ndustry, and the use
of fossi l fuel s. Thi s organi zati on has several advan-
tages over the more conventi onal catal ogi ng of
envi ronmental probl ems by thei r i mpacts on l and,
water, and ai r. Fi rst, i t hel ps capture the dynami c
causes of envi ronmental degradati on i n devel opi ng
4u.s. Co=ss, ~lce of Twhnolow Assessment Enhancing Agriculture in Afica: A Role for U.S. Development Assistance, OT..-35656
W-O% ~: U.S. Government mt@ mice, September 1988). Some note, however, that even steep or acid-infertile lands can be productive
over long periods as shown by the centuries of terraced rice farming in Asia or continuous sugar-cane cropping in the Dominican Republic. See Rieardo
Radulovich A View on Tropical Deforestatio~ Nature, vol. 346, July 19, 1990, p. 214.
Chapter 5-Energy and the Environment in Developing Countri es q 113
countri es i n the vari ous contexts of the rural struggl e
for survi val by the l andl ess peasant, or of the urban
effort to devel op modern i ndustry i n an economy
that has l i mi ted capi tal , technol ogy, and ski l l ed
human resources. Second, i t hi ghl i ghts the di ffer-
ences i n the causes and consequences of envi ron-
mental degradati on between rural and urban areas.
Thi rd, i t hel ps i l l ustrate some of the compl ex
l i nkages between these envi ronmental probl ems that
make effecti ve responses so di ffi cul t. I f economi c
needs are to be met whi l e si mul taneousl y mai ntai n-
i ng envi ronmental qual i ty, these dynami cs and
compl ex i nterconnecti ons must be understood and
responded to appropri atel y.
The Rural Sector
Many devel opi ng countri es are sufferi ng si gni fi -
cant envi ronmental degradati on i n rural areas, i n-
cl udi ng: rapi d devegetati on/deforestati on and the
l oss of i rrepl aceabl e bi odi versi ty; deserti fi cati on;
erosi on of crop and pasturel and; watershed degrada-
ti on, si l tati on of downstream waterways, down-
stream fl oodi ng, and other i mpacts on water qual i ty
and suppl i es; and l ocal and regi onal ai r pol l uti on. I n
turn, these envi ronmental i mpacts may damage or
destroy farm and range l and and force more peopl e
i nto watersheds and other ecol ogi cal l y fragi l e l ands--
potenti al l y creati ng a vi ci ous cycl e of degradati on.
Cl eari ng and burni ng woodl ands and grassl ands for
agri cul ture or pasture al so contri bute to the gl obal
i ncrease i n atmospheri c greenhouse gases.
The pri nci pal di rect causes of these forms of
envi ronmental degradati on are shi fti ng cul ti vati on,
conversi on of forest l ands to permanent pasture and
agri cul ture, commerci al l oggi ng, overgrazi ng, i nap-
propri ate management of i rri gated l and, and the use
of bi omass for fuel . These factors are often i ntercon-
nected. Commerci al l oggi ng of cl osed tropi cal
forests, for exampl e, opens new areas wi th roads and
parti al cl eari ngs, whi ch enabl es settl ers to fol l ow,
converti ng forests to pasture and cropl and.
Underl yi ng these causes are factors such as:
popul ati on growth; poverty; the l ack of access to
modern energy-i ntensi ve i nputs for agri cul ture (me-
chani cal tracti on, i rri gati on, ferti l i zers, etc.); the l ack
of secure l and tenure for many peopl e; and govern-
ment pri ci ng, tax, and other pol i ci es that may
encour age mi suse of natur al r esour ces. These causes
and consequences are l i sted i n tabl e 5-1.
The extent and rate of envi ronmental degradati on
are di ffi cul t to quanti fy accuratel y under the best of
ci rcumstances, even when rel i abl e data are avai l -
abl e. As i t i s, esti mates vary wi del y on the basi s of
the underl yi ng defi ni ti ons of what consti tutes eco-
l ogi cal vul nerabi l i ty or degradati on; the methodol o-
gi es used for the anal ysi s; and the assumpti ons used
to go from l ocal measures of degradati on to gl obal
extrapol ati ons.
The depi cti on of envi ronmental degradati on al so
depends on what measure i s used for compari son.
Local degradati on may be severe and rapi d. I rrespec-
ti ve of the hardshi p thi s pl aces on peopl e i n the
i mmedi ate regi on and the need for l ocal responses,
i t does not necessari l y transl ate i nto correspondi ng
probl ems at the regi onal or gl obal scal e. There i s
consi derabl e vari ati on i n form and degree of envi -
ronmental degradati on from one regi on to another.
Consequentl y, an attempt i s made bel ow to pl ace
esti mates of the rate of l ocal envi ronmental degrada-
ti on i nto the gl obal context.
Causes of Environmental Degradation
Agri cul ture
As the popul ati ons of devel opi ng countri es grow,
the demands on the l and for food, fuel , and fodder
i ncrease accordi ngl y. Farmers then face three basi c
choi ces: they can mi ne the l andtaki ng more out
of i t than they put i n-unti l the l and i s exhausted;
they can mi grate to new l ands; or they can i ncrease
the l evel of (capi tal -, energy-, and l abor-i ntensi ve)
agri cul tural i nputs-mechani cal tracti on, ferti l i zer,
and i rri gati oni nto the l and i n order to rai se yi el ds.
These mi ght al so i ncl ude hi gher i nputs of i nforma-
ti on and management as mi ght be the case for
i ntercroppi ng, agroforestry, i ntegrated crop-l i ve-
stock, or other sophi sti cated agri cul tural systems.
s
Mining the LandMining the l and usual l y
takes the form of shortened fal l ow peri ods-l eavi ng
i nadequate ti me for the natural regenerati on of
soi l -as popul ati on pressures mount. I t i s obvi ousl y
a temporary sol uti on, but one often resorted to by
those wi thout access to the modern agri cul tural
5u.s. COnmSS, O&l ce of TeChnOIO~ Assessmen~ Enhancing Agriculture in Af+ca: A Role for U.S. Development Assistance, OTA-F-356
(TVashingto& DC: Government Printing Office, September 1988).
i nputs needed to rai se crop yi el ds.
6
Decl i ni ng yi el ds
from such mi ni ng are seen i n a number of l ocal
and regi onal areas, parti cul arl y i n Afri ca.
7
Mi grati on-Many peopl e mi grate to new l ands or
to urban areas. I n addi ti on to popul ati on pressures or
soi l exhausti on, factors forci ng peopl e to mi grate
from l ong-establ i shed farmi ng areas to new l ands
i ncl ude the l ow producti vi ty of tradi ti onal agri cul -
ture; i nequi ti es i n l and tenure for many subsi stence
farmers; drought or other di sasters; and, i n some
areas, mechani zati on of agri cul ture and/or consol i -
dati on of agri cul tural l ands. I n many cases, currentl y
farmed areas have been subdi vi ded among succes-
si ve generati ons to the poi nt that the l andhol di ngs
for those who remai n are, or wi l l soon become,
margi nal . I n Rwanda, for exampl e, the average
smal l hol der had 1.2 hectares i n 1984 and, by
tradi ti on, woul d di vi de i t equal l y among hi s average
of four sonsl eavi ng them 0.3 hectares each. I f the
same trend conti nues, the fol l owi ng generati on wi l l
have l ess than 0.1 hectare each.
8
Esti mates of the
number of l andl ess or near-l andl ess (wi th too l i ttl e
l and to subsi st) i n devel opi ng countri es range as hi gh
as 1 bi l l i on peopl e, most of them i n Asi a.
9
Wage-payi ng jobs are scarce for those who are
forced to mi grate. Land-anywhere they can get
i tfor subsi stence agri cul ture and fuel wood i s often
thei r onl y means of survi val . I ncreasi ngl y, however,
avai l abl e l ands are remote, onl y margi nal l y produc-
ti ve, or ecol ogi cal l y fragi l e--on upl and regi ons that
are easi l y eroded when groundcover i s removed, on
ari d or semi -ari d l ands, or i n forested areas of hi gh
bi ol ogi cal di versi ty (but whi ch may have poor soi l s).
As many as 370 mi l l i on peopl e i n devel opi ng
countries may l i ve i n r ur al ar eas that ar e ecol ogi cal l y
vul ner abl e.
10
To generate good yi el ds on a sustai n-
abl e basi s on these l ands often requi res l arger i nputs
of l abor and/or capi tal and technol ogy than the l ands
l eft behi nd-i nputs to whi ch these peopl e sel dom
have access.
l l
Thi s maybe parti cul arl y true i n newl y
opened areas where i nfrastructure (i ncl udi ng access
to extensi on efforts) i s especi al l y weak. The i mmi -
grants i nto these areas may be unfami l i ar wi th the
di fferent agri cul tural techni ques appropri ate (sus-
tai nabl e) to these new l ands and resources.
I n many regi ons, shi fti ng agri cul ture i s i ni ti al l y
12 (Si ng agri cul -
practi ced by those who mi grate.
ture i s al so practi ced tradi ti onal l y by l ong-term
resi dents i n many areas and i s a sustai nabl e form of
agri cul ture i f fal l ow peri ods are suffi ci entl y l ong.)
Shi fti ng agri cul ture begi ns wi th forest-fal l ow sys-
tems i n whi ch smal l patches of l and are cl eared and
cul ti vated for a few years and then l eft fal l ow for as
l ong as two to three decades (see ch. 3). Thi s remai ns
an i mportant form of agri cul ture i n west Afri ca,
southeast Asi an hi l l communi ti es, parts of South
Ameri ca,
13
and el sewhere.
14
Thi s form of shi fti ng
or sl ash and burn-agri cul ture i s bel i eved by many
to be the most i mportant cause of secondary
15
forest
destructi on and to be roughl y comparabl e to com-
merci al l oggi ng i n i ts i mpact on pri mary forest (tabl e
5-2).
16
Shi fti ng cul ti vati on consumes enormous
amounts of bi omass energy i n the process of cl eari ng
the forest.
17
Modern Agriculture-Those farmers wi th access
to good soi l s and water resources and modern
agri cul tural i nputs can i ncrease the yi el ds of thei r
cropl ands. Modern agri cul tural practi ces i n devel op-
6p~r soilq~~ or fid~~te Waterresources may also be signitlca.nt constraints on raising crop yields even with high levek Of agrictimd @u@.
~.S. Congress, OffIce of Technology Assessment op. cit., footnote 5, pp. 63 ff.; United Nations Food and Agriculture Organizatiom Afican
Agriculture: The Next 25 Years, Annex II, The Land Resource Base (Rome, Italy: United Nations, 1986).
gufited Nations Population Fund,
The State of World Population 1990, New Yorlq 1990.
%. Jeffrey Leonard, Enviromnentand the Poor: Deve20pment Strategiesfor a Common Ageti (New Brunswick NJ: Transaction Books, 1989).
l%id.
1lu.s. CO~SS, office of Technology Assessment, op. cit., footnote 5.
lz~~~tive]y those who migrate may continue their previous pattern of agricuhur~ftenW
rmanent agriculture--rather than adopting shifting
agriculture techniques as traditional in the new area.
13Norm~ My~s, D@ore~tation Rates i
n
TroPi~OIFore~ts and Their Cli~tic Implications @ndon: Friends of the ~ Limited, 1989).
14u.s. Conwss, Oflce of T~~ology Assessment, c~nging By DegreeS: ~feps TO Reduce &ee~house Gzses, forthcodlg.
15 Sewn@ f
ore
st is ~twhich~b=lo=~ ~the p~t~dthen~ow~ toregrow, or~oth~~b~nsi~lc~fly affected byhurnanactivity.
IGSee alSO Julia C. Allen and Douglas F. B~=,
The Causes of Deforestation in Developing Countries, Annals of the Association ofAmerican
Geographers, vol. 75, No. 2, 1985, pp. 163-184.
IT~R. Smi@ ~~~e Biofuel T~sitioU pacific a~Asian Jour~l of Energy, 1987, pp. 1s-s2; T- ~, H~
Ecology Research on
Tropical Agroecosystems in SoutheastAs~ Singapore Journal of Tropical Geography, vol. 3, No. 1, 1982. Some of the biomass ash generated serves
as nutrients for the crops subsequently planted.
Table 5-2Estimates of Land Use Changes in Closed and Open Tropical Forests, circa 1980 (million hectares per year)
Closed forest Open forest
Seiler and
Crutzen
Food and
Agricultural
Seiler and
Crutzen
Food and
Agricultural
Land use change Low High Organization Lanly Myers Low High Organization Lanly
Primary forest to:
Shifting cultivation . . . . . . . . . . . 2.6 3.6 3.2 3.4 1.9

1.2 1.7
Permanent pasture ., ,, . . . . . . 1.6 1.4 1.1 2.5 0.6 1.7 2.1 0.8 1.3
Permanent agriculture . . . . . . . 0.3 2.2 1.1 2.3 0.0 0.2 3.0 0.7 0.8
Logged forest . . . . . . . . . . . . . . 4.6 3.7 4.5

Logged forest to:
Permanent pasture . . . . . . . . . .
1.1
0.6

Permanent agriculture . . . . . . .
1.0
3.9

Secondary forest to: 14.9 40.0 18.5 22.0 3.4 6.9 21.9 11.4 18.6
Shifting cultivation . . . . . . . . . . .
Permanent pasture ...,. . . . . . 0.5 1.5 0.1
3.3 1.0 1.0

Permanent agriculture . . . . . . . 0.6 0.8 0.1
6.8 0.2 1.4

NOTE: The data in this table maybe significantly inaccurate. The table is presented here only to provide a general indication of the rates and causes of deforestation. The Food and Agricultural
Organization (FAO) and Lanly data are essentially the same (Lanly is the principal author of the FAO study), based primarily on official government statistics, and maybe underestimates
of the rate of deforestation. A more recent review by Myers, for example, indicates substantially higher rates.
1
The UNFAO currently has underway a more detailed study that may resolve
some of these large discrepancies.
I Norman ers, f)efofestatinRatesin Tropica/ Forests and Their Clirnatk hnpkations (London: Friends of the EWh, kwmber 19s9).
SOURCE: R.P. Detwiler and Charles A.S. Hall, Tropical Forests and the Global Carbon Cycle, Science, vol. 239, Jan. 1, 1988, pp. 42-47, citing Seiler and Cmtzen, Food and Agricultural
Organization, Lanly, and Myers.
i ng countri es use, to varyi ng degrees, i mproved
pl ant speci es, syntheti c ferti l i zers, pesti ci des and
herbi ci des, i rri gati on, and mechani zed operati ons to
generate hi gher crop yi el ds. These requi re hi gh
l evel s of capi tal and energy i nputs.
Modern agri cul ture has, wi th mi xed success, pro-
vi ded envi ronmental , soci al , and economi c benefi ts.
I t has moderated cropl and expansi on i nto ecol og-
i cal l y fragi l e or parti cul arl y val uabl e l ands through
technol ogi cal advance (the green revol uti on) and
energy i nputs i n the form of ferti l i zer, i rri gati on, and
mechani cal operati ons. Over the 20-year peri od
from 1965 to 1985, cropped areas i ncreased by onl y
14 percent, 35 percent, and 4 percent i n Afri ca, South
Ameri ca, and Asi a whi l e thei r popul ati ons i ncreased
by roughl y 75 percent, 60 percent, and 50 percent
respecti vel y (tabl e 5-3). Modern agri cul ture has al so
rai sed the personal i ncomes of many farmers; and i t
has contri buted to nati onal economi c growth i n
many countri es, especi al l y i n Asi a.
Modern agri cul ture has al so had seri ous short-
comi ngs. I t has i ncreased economi c i nequi ti es be-
tween those farmers who have suffi ci ent l and and
access to capi tal and other i nputs necessary for
hi gh-yi el d agri cul ture and those farmers who do not
have such resources, and i t has di spl aced l aborers i n
many cases. I t has caused envi ronmental damage
through the mi suse of ferti l i zers, pesti ci des, and
herbi ci des. I nadequate i nvestment and poor man-
agement have l ed to waterl oggi ng and sal i ni zati on of
val uabl e i rri gated l ands. Fi nal l y, modern agri cul -
tur al techni ques r equi r e dr amati c i ncr eases i n com-
merci al energy use.
18 Concer n over envi r onmental
i mpacts and hi gh dependence on purchased i nputs
has l ed to consi derabl e i nterest i n farmi ng systems
that depend more on resources i nternal to the
farmi ng system and l ess on external purchases.
These agri cul tural and agroforestry systems tend to
be very i nformati on and management-i ntensi ve.
19
Irrigated LandsI rri gati on i s an i mportant el e-
ment i n modern agri cul ture. I t frees the farmer from
dependence on i rregul ar rai ns and rai ses yi el ds,
al l owi ng doubl e- and even tri pl e-croppi ng. Some
160 mi l l i on hectares of l and i n devel opi ng countri es
ar e i r r i gated. I n Asi a al one, 100 mi l l i on hectar es ar e
i rri gated, and thi s l and produces roughl y 60 percent
of the regi ons food on just 45 percent of i ts cropped
ar ea.
20
I n I ndi a, more than 6 mi l l i on el ectri c and 3
mi l l i on di esel pump sets have been depl oyed (see ch.
3 ) , consumi ng nearl y 2,000 GWh of el ectri ci ty and
3 mi l l i on tons of di esel fuel i n 1985.
21
I nadequate i nvestment and poor management
have resul ted i n vari ous degrees of sal i ni zati on
and/or waterl oggi ng of i rri gated l ands i n many
countri es. For exampl e, by one esti mate 75 percent
of Paki stans i rri gated l and suffers sal i ni zati on
and/or waterl oggi ng, wi th correspondi ng reducti ons
i n crop yi el ds.
22
Some 20 mi l l i on hectares-roughl y
hal f--of I ndi as i rri gated cropl ands have suffi ci ent
sal t bui l dup to reduce producti vi ty; another 7 mi l l i on
hectares of l and i n I ndi a now l i e unused due to
excessi ve sal t.
23
(Si mi l ar probl ems affl i ct the Uni ted
States, where 20 to 25 percent of the 20 mi l l i on
hectares of i rri gated l ands are affected by sal i ni za-
ti on.
24
) Recl amati on i s possi bl e through i mprove-
ments i n canal s and other i nfrastructure to reduce
l eakage and by provi di ng drai nage from the fi el ds,
but i t i s expensi ve.
The technol ogy to prevent or mi ni mi ze sal i ni za-
ti on and waterl oggi ng has been avai l abl e si nce the
IsNote ~tthe ener~ intensity of shifting+r slash and burn-cultivation may be significantly higher than commercial agrkxdture when the energy
consumed by burning off the standing biomass is taken into account. However, commercial fuels are little used, if at all, in shifthg agriculture.
lgAdvisoV co~tte on the s~el, Agrofores@ in the ~e~~ African Sahel, Bo~d on Science ~d Techno@y for kternational Development
National Research Council (Washington DC: National Academy Press, 1983, 1984); U.S. Congress, Office of Technology Assessment, op. cit., footnote
5; Clive A. Edwards et al., Sustainable Agricultural Systems (Ankeny, Iow& Soil and Water Consemation Society, 1990); Robert Winterbottom and
Peter E Hazlewood, Agroforestry and Sustainable Development: Making The CormectioQ AWBZO, vol. 16, No. 2-3, 1987, pp. 1(W11O; C. Okail
and J.E. Sumberg, Sheep and Goats, Men and Women: Household Relations and Small Ruminan
t Production in Southwest Nige@ Understandzng
Aficas Rural Households and Farming Systems, Joyce Lewinger Moock (cd.) (Boulder, CO: Westview Press, 1986).
mMon~We Yude@ Sus~ble ~d Equitable Development in krigilted WvkonmentS,
Environment and the Poor: Development Strategies
for a ConvnonAgendiz, Jeffrey Leonard (cd.) (New BrunswIckj NJ: Transaction Books, 1989).
ZIAshok De@ fiergyB~nc~ for~di~ 1985-86, contractor report prepared for the Office of Technology Assessment, 1990. w is @vd~t
to 125,000 GJ.
~udehna~ op. cit., footnote 20. Total cmp areas from World Resources Institute, World Resources 1990-91 (New York NY: Oxford University
Press, 1990). For other estimates, see C.E. Whitman et al. (eds.), RainfedAgricu/ture in the Near East Region, proceedings of the worlmhop at Amman
Jorda~ Jan. 18-23, 1986, USDA and USAID, p. 92.
~Yude~~ op. cit., footnote 20. Total crop areas from World Resources 1990 -91, op. Cit., fOOblOk 22.
~World ReSo~eS Institute, World Resources 1987 (New York NY: Basic BOOkS, 1987), p. 280.
Table 5-3-Changes in Land Area Uses
Cropland area Permanent pasture Forest/woodlands Other land
1964-66 to 1964-66 to 1964-66 to 1964-66 to
Total Population 1985-87 1983-85 1985-87 1983-85 1985-87 1983-85 1985-87 1983-85
area density, 1989 percent percent percent percent percent percent percent percent
(million (people per of total change of total change of total change of total change
Region hectares) thousand ha) land area in cropland land area in pasture land area in forest land area in other uses
Africa . . . . . . . . . 2,965 212 6.2 14.0 26.7 0.5 23.2 -8.6 43.9 3.6
N. America . . . . 2,139 197 12.8 7.7 17.2 2.4 32.0 -5.9 38.0 1.5
S. America . . . . 1,753 166 8.0 35.2 27.0 9.4 51.6 -7.3 13.4 -1.4
Asia . . . . . . . . . . 2,679 1,139 16.8 4.2 25.3 -3.0 20.2 -4.6 37.7 -0.7
Europe . . . . . . . 473 1,050 29.6 5.0 17.8 -4.9 33.2 7.1 19.4 3.9
SOURCE: World Resources Institute, 14btfd Resources 1988-89 (New York, NY: Bask Books, 19SS) table 16.1; World Resources institute, Work/ Resources 1990-91 (New YorlG NY: Oxford
University Press, 1990), table 17.1.
1950s, yet desi gners frequentl y fai l to i ncorporate
these i mprovements i n thei r projects. Many factors
contri bute to thi s fai l ure. One factor may be that
desi gners tend to be overl y opti mi sti c i n thei r i ni ti al
apprai sal s: they assume that waterl oggi ng and sal -
i ni zati on wi l l not occur for 20 to 30 years, and that
drai nage systems and other capi tal -i ntensi ve i m-
provements wi l l not be necessary before then. Thi s
l owers the apparent costs of thei r projects and may
i mprove the chances that the projects wi l l be
approved.
25
Pasture-The openi ng up of l ands for pasture i s
a si gni fi cant cause of deforestati on (tabl e 5-2)
parti cul arl y i n Central and South Ameri ca where
grazi ng l ands have i ncreased by more than 9 percent
duri ng the past 20 years (tabl e 5-3).
Overgrazi ng may al so pl ay a si gni fi cant rol e i n the
deserti fi cati on of semi -ari d regi ons.
26
Overgrazi ng
and trampl i ng can qui ckl y destroy the grass l ayer.
Wi thout the protecti on of ground cover, topsoi l can
be washed or bl own away, l oweri ng ferti l i ty. I n
some areas, the ful l force of rai n on the soi l can bri ng
cl ay parti cl es to the surface and cause surface
hardeni ng and seal i ng that seeds cannot penetrate.
27
The end resul t of such processes can be deserti fi ca-
ti on.
28
Herders burn grassl ands to encourage new
growth; numerous studi es have shown thi s new
growth to be parti cul arl y good forage for thei r
herds.
29
More general l y, repeated burni ng i s be-
l i eved to hel p create and mai ntai n much of the
worl ds savannah and grassl and.
30
Such brushfi res
i n the Afri can grassl ands may burn as much as 2
bi l l i on tons of bi omass annual l y, cause vol ati l i zati on
of organi c ni trogen and sul fur, and al l ow excessi ve
l eachi ng of other val uabl e nutri ents.
31
Thi s maybe
parti cul arl y damagi ng i n much of the Sahel , where
growth i s al ready strongl y l i mi ted by the l ack of
nutri ents.
32
Thus, brushfi res hel p the herder feed hi s
ani mal s i n the near term but, i n the l onger term,
l ower soi l ferti l i ty, and ki l l brush and trees that hol d
the soi l s and pul l nutri ents up from deep i n the
gr ound.
33
Brush and grassl and fi res may al so be
si gni fi cant contri butors to regi onal ai r pol l uti on and
may contri bute modestl y to the gl obal i ncrease i n
greenhouse gases.
34
More stabl e suppl i es of forage mi ght reduce the
need of herders to mai ntai n l arge numbers of ani mal s
i n order to ensure the survi val of a few through
peri ods of drought. Hi gher qual i ty forage (hi gher
protei n content) woul d reduce the need to burn
grassl ands. I nputs of capi tal - and energy-i ntensi ve
ferti l i zer, i ncreased suppl i es of water, and mechani -
cal harvesti ng of the forage (or even grai n crops)
when i ts protei n content i s at a maxi mum-as i s
common i n i ndustri al countri es-mi ght ai d i n achi ev-
i ng both of these goal s.
Commercial Loggi ng--Commerci al l oggi ng i m-
pacts perhaps 3 to 5 mi l l i on hectares of pri mary
tropi cal forest annual l y (tabl e 5-2). I n many areas,
onl y the hi ghest grade l ogs are removed from the
forest. But for every tree removed, roughl y 5 to 10
other trees are destroyed.
35
Commer ci al l oggi ng al so
devel ops roads that al l ow settl ers access to forested
regi ons, where they can cl ear the forests for farms or
r anches.
Biomass Fuel sBi omass-wood, crop resi dues,
and ani mal dung-i s the pri mary fuel for peopl e i n
25yu&.~n, Op. Ci t . , fOO~Ote 20.
xwil~ H. Schlesinger et al.,
Biological Feedbacks in Global Desetilcatioq Science, vol. 247, Mar. 2, 1990, pp. 1043-1048.
ZTNatio~Re=hCoUcil, Board On science andT~~ologyfiD~elopmen~Agf~~~~e~~ ~~ ~~e West@icansaheJ (waShh@O~Dc: National
Academy of Sciences, 1983); Georges Novikoff and Mohamed Skouri, Balancing Development and Conservation in Pre-Saharan Iimish AMBZO,
vol. 10, NOS. 2-3, 1981, pp. 135-141.
~Ufited Natiom Food and Agric~~ ~ga~tioq African Agncu/~re: The ~~t 25 years, hex n, me Land Resource Base, Op. Cit.,
footnote 7.
mJ. Dirck Stryker, Technology, Human Pnxsure, and Ecology in the Arid and Semi-Arid Tropics, in H. Jeffrey I.scmard (cd.), op. cit., footnote 9.
~Cml Sagm, ow~ B. T~~ ~d James B. pollac~ ~~opogefic ~b~oages and the ~sclimate, science, vol. ZM, 1979, p p .
1363-1368; Daniel FiruL Land Use and Abuse in the East African Region, AMBZO, vol. 12, No. 6, 1983, pp. 296-301; DJ. Pratt and M.D. Gwynne
(eds.), RangelandiUanagement and Ecology in East A&-ca (Huntington NY: Robert E. Kreiger Publishing Co., 1977).
slNatio~Re~emhCowcil,BOmdon Science ~d T~~ology ~Developrnen~E~vi~o~~~ra/ Change in the WestAfi-can SaheZ(Washingto~ DC:
National Academy of Sciences, 1983); World Resources Institute, World Resources 1990-91 (New Yorkj NY: Oxford University Press, 1990).
32H. Bre~n ~d C-T. dewi~ $~gel~d productivity ~d &ploitation iII he s~e~ Science, VO1. 221, 1983, pp. 1341-1347.
sss~ker, op. cit., footnote 29.
~World Reso~ces mti~te, World Resources 1990-91 (New York NY: Oxford University press, 19$0.
ssFred pe~e, Hit and RW in saraw~ ~ew Scientist, my 12, 1990, pp. %9.
Chapter 5--Energy and the Environment in Developing Countries q 119
rural areas, and i n many regi ons i t i s al so i mportant
for the urban poor (see ch. 3). Unti l recentl y, i t was
wi del y bel i eved that many devel opi ng countri es
were on the verge of bei ng transformed i nto treel ess
deserts or mountai n wastel ands due to the demand
for fuel wood. These predi cti ons were l argel y based
on the gap model , whi ch assumed a freed
per-capi ta demand for fuel wood and a growi ng
popul ati on whi l e at the same ti me the forest base was
decl i ni ng due to agri cul tural expansi on and over-
cutti ng of the forest.
36
I n more recent studi es, however, i t has been found
that the use of fuel wood i s hi ghl y el asti c accordi ng
to i ts avai l abi l i ty and the l abor requi red to col l ect i t
or, i f traded, i ts pri ce. When fuel wood i s di ffi cul t to
obtai n by foragi ng, peopl e qui ckl y resort to l ower
qual i ty wood, agri cul tural resi dues, or dung. More
general l y, rural subsi stence farmers cause rel ati vel y
l i ttl e damage to forests, as they usual l y take onl y
deadwood or smal l l i mbs. They do not have the tool s
to cut down l arge trees. Much of the wood they
collect is from hedger ows or other sour ces near thei r
farms. I n Kenya, for exampl e, trees outsi de the forest
suppl y hal f the fuel wood demand.
37
And i n West
Java, one study found that three-fourths of al l the
fuel col l ected came from wi thi n fami l y courtyards
and gardens, and two-thi rds of thi s fuel was branches
and twi gs .38
I n contrast to rural foragers, commerci al i zed
fuel wood and charcoal operati ons to suppl y urban
househol ds, commerci al faci l i ti es and i ndustri al
operati ons often cut whol e trees and can damage or
destroy forested areas. The i mpact of commerci al
demands for fuel wood are l i mi ted, however, as users
wi l l swi tch to fossi l fuel s when fuel wood becomes
scarce and pri ces cl i mb (ch. 3).
39
The extent of the
damage to the forest resul ti ng from commerci al i zed
fuel demand wi l l then depend on the di stance
between the user and the forest, the si ze of the forest,
the si ze of the demand, the rate of regrowth, and
other factors. The use of bi omass for fuel i s not
usual l y a pri nci pal cause of deforestati on. I t does,
however, add addi ti onal pressure on forest re-
sources. I n ari d or semi -ari d regi ons, where forest
growth i s sl ow and there i s a hi gh popul ati on densi ty
or a concentrated urban demand for fuel wood, such
as the Afri can Sahel , the use of bi omass fuel s can
contri bute si gni fi cantl y to l ocal deforestati on.
40
Si mi l arl y, the use of bi omass fuel s can l ead to l ocal
deforestati on i n some cases where there i s a concen-
trated commerci al or i ndustri al demand.
Environmental Impacts
The envi ronmental i mpacts of agriculture, r anch-
i ng, l umberi ng, and the use of bi omass for fuel
i ncl ude deforestati on, deserti fi cati on, soi l erosi on,
fl oodi ng, pesti ci de and ferti l i zer runoff from crop-
l and, and ai r pol l uti on from bi omass fuel s.
Defor estati on
41
The forests of the devel opi ng worl d provi de a
number of resources and benefi ts, i ncl udi ng food,
fuel , fodder, fi ber, ti mber, and medi ci nes. Forests are
a vi tal ecol ogi cal resource, protecti ng soi l s, mai n-
tai ni ng di verse pl ant and ani mal l i fe, regul ati ng the
fl ow of water, and pl ayi ng an i mportant rol e i n the
gl obal carbon cycl e. Forests may al so strongl y
i nfl uence regi onal cl i mates.
42
The l oss of tropi cal
forests not onl y threatens to depri ve the worl d of
val uabl e resources on whi ch i t currentl y depends,
but al so to forecl ose opportuni ti es to di scover new
36@r~dF01ey, DiscussionPaper on De~nd_gemen~*proceedings of the ESMAP titern~d SouthemAtiicaHousehold ~~gyplx
Seminar, Harare Zimbabwe, Feb. 1-5, 1988, UNDP/World Bank Energy Sector Management Assistance Program, Activity Completion Report No.
085/88.
ST~ OKeefe, Fuel for the people: Fuelwood in the Third WorlL AiUBZO, VO1. 12, 1983, W. 21-26.
38M, ~di Swsasw, policy ~ysis of R~ Household Energy N~& in West Jav~ Rural Energy to Meet Develop~ntNeeds: Asian Village
Approaches, M. Nund Is~ Richard Morse, and M. Hadi Soesastro (eds.) (Boulder, CO: WestView Press, 1984), p. 114.
s~ug~ F. Barnes,
Understanding Fuelwood Prices in Developing Nations, draft, Household Energy Uni4 Industry and Energy Departrnen$
World B@ Washington DC, Oct. 31, 1989.
%ouglas F. Barnes, World B* Population Growth, Wood Fuels, and Resource problems in Sub-Saharan Africa, Industry and Energy
Department Working Paper No. 26, March 199Q R. Moss and W. Morga~ Fuelwood and Rural Energy Production and Supply in the Hunud Tropics
(Dubl@ Ireland: ~cooly International Publishing Ltd., 1981); Finq op. cit., footnote 30; Dennis Anderson and Robert Fishwic~ World B@
Fuelwood Comumption and Deforestation in African Countries, World Bank Staff Working Paper, No. 704, 1984.
alForde@~ rWiews of defores~tion and po~~ re~mes to i~ see: U.S. Congress, ~lce of Technology Assessm@ Changing By Degrees: Steps
To Reduce Greenhouse Gases, forthcom@ U.S. Congress, Offke of Technology Assessment, Technologies to Sustain Tropical Forest Resources,
OTA-F-214 (Springfield, VA: National Technical Information Service, March 1984); and Robert Winterbotto~ Taking Stock: The Tropical Forestry
Action Plan After Five Years, World Resources Institute, Washington, DC, June 1990.
42J. Shul@ C. Nobre, and P. Sellers, Amazon Deforestation and Climate Change, Science, vol. 247, Mar. 16, 1990, pp. 1322-1325.
potenti al sources of weal th and sci enti fi c knowl -
edge. On the other hand, tropi cal forests offer an
economi c opportuni ty that the poor i n devel opi ng
countri es wi sh to sei ze as qui ckl y as possi bl e.
43
Governments may al so want to col oni ze forest l ands
i n or der to establ i sh mor e cl ear l y thei r l egal cl ai m to
nati onal terri tory.
The rate of gl obal deforestati on-from al l causes--
i s hi ghl y uncertai n, wi th esti mated rates i n the earl y
1980s rangi ng from roughl y 0.5 percent
44
to 1.8
per cent
45
annual l y. These rates appear to be accel er-
ati ng due to i ncreasi ng popul ati on and other pres-
sures noted above.
The UNFAO 1990 Forest Resources Assessment
estimates the current annual gl obal deforestati on
rate as 1.2 percent-doubl e thei r esti mate for
1980.
46
Forested areas and deforestati on rates vary
wi del y between countri es. I ndonesi a and Brazi l have
huge areas of cl osed forests (100 and 350 mi l l i on
hectares, respecti vel y) and annual rates of deforesta-
ti on of perhaps 0.4 to 1.4 percent and 0.5 to 2.3
percent, respecti vel y. Aggregate fi gures, however,
tend to obscure severe deforestati on occurri ng i n
certai n regi ons. I vory Coast, for exampl e, has just 16
mi l l i on hectares of forest remai ni ng, and suffers an
annual deforestati on rate vari ousl y esti mated be-
tween 6.5 and 15.6 percent. A number of other
countri es l i e between these extremes, wi th forested
areas of 5 to 50 mi l l i on hectares and deforestati on
rates vari ousl y esti mated i n the range of 2 to 8
percent annual l y .47 At these rates, thei r cl osed
forests coul d di sappear i n a few decades.
Deser ti fi cati on
Deserti fi cati on can resul t from a vari ety of factors,
dependi ng on the regi on, i ncl udi ng l ong-term cl i -
mate trends, overgrazi ng, poor farmi ng practi ces,
and deforestati on.
48
Al though anecdotal evi dence
i ndi cates that dryl ands i n many regi ons are becom-
i ng decerti fi ed at an i ncreasi ng rate, there i s l i ttl e
rel i abl e data to support the case. The Gl obal
Assessment of Soi l Degradati on, i ni ti ated i n l ate
1987 by the Uni ted Nati ons Envi ronmental Program
and the I nternati onal Soi l Reference and I nformati on
Centre i n the Netherl ands, shoul d provi de some of
these data.
I mpacts of Deforestati on and Deserti fi cati on
Among the potenti al i mpacts of deforestati on and
deserti fi cati on are soi l erosi on and degradati on,
fuel wood and fodder shortages, i ncreased fl oodi ng,
mi crocl i mati c changes, and l oss of bi odi versi ty.
49
Soi l sLi ttl e soi l i s l ost from forests or grass-
l ands. When vegetati on i s removed, massi ve amounts
of soi l can be washed away as rai nwater fl ows across
the surface. Measurements i n Tanzani a i ndi cated
that up to hal f the rai nfal l was l ost as runoff from
bare fal l ow (3.5 sl ope), carryi ng wi th i t some 70
tons of soi l per hectare.
50
Si mi l ar i mpacts have been
noted el sewhere.
51
Wi th no shadi ng, soi l tempera-
tures ri se dramati cal l y and can greatl y reduce the
vi tal bi ol ogi cal acti vi ty i n the soi l .
52
Loss of tree
cover al so al l ows hi gher average wi nd vel oci ti es
(and soi l erosi on) and, combi ned wi th the reduced
soi l moi sture content, can l ower crop yi el ds.
53
AsMcmdo ~ulovich A View on Tropical Deforestation,
Nature, vol. 346, July 19, 1990, p. 214.
44Jean.pa~ My, Tropical Forest ReSOWCeS,
Forestry Paper No. 30, United Nations Food and Agriculture Organization, Rome 1982.
45Mye~, op. Cit., foo~ote 13-
i
World Deforestation Increases at Quicker Rate Than Expected, MultinationalEnvironmental Outlook, Aug. 21, 1990, p. 134.
47Myem, op. cit., fOOtnOte 13) P.
2
*
4
0
48G0 No~ofi, ~ Desefimtionby ~er@g,~B]O, vol. 12, No. 2, 1983, pp. l@-105; H-F. ~rey iuidI-IusseinYussuf, P~todism~d
Desert Encroachment in NorthernKeny~ AMBZO, vol. 10, Nos. 2-3,1981, pp. 131-134; National Research Council, Environmental Change in the West
@can Sahel, op. cit., footnote 31; D. kdersom and R. Fishwiclq Fuelwood Consumption and Deforestation in Mrican Countries, World Ba~
Staff Working Paper No. 704, 1984; Schlesinger et al., op. cit., footnote 26.
A~mwa~~~~ et~., ~e~pactof -of a Tropic~For@ i.ll~done@AJfBz(J, VO1. 10, NOS. 2-3,1981, pp. 115-1 19; L@erR. Bm~
World Population Grow@ Soil Erosiom and Food Security, Science, vol. 214, 1981, pp. 995-1002; Alain Grainger, Deser@cation (londom
-C% 1984), p. 94.
50phil O*K&fe, ~~rhe CauKs, co~~~ms ad Remedies of Sofl Erosion ~ K~y~ AMBZO, VO1. 12, No. 6, 1983, pp. 302-305.
51u.s. CoweSs, ~lce of T~~oloW Ass=ment, p+ cit., fm~o~ 41; F~ op. cit., footnote 3Q Ene~ Sakfi and Peter B. Vose, Depletion of
Tropical Rainforests; AMBIO, vol. 12, No. 2, 1983, pp. 67-71, Vaclav Smil, Deforestation in China, AMBIO, vol. 12, No. 5, 1983, pp. 226-231;
Nigel J.H. Smi@ Colonization Lessons from a Tropical Foresg Science, vol. 214, 1981, pp. 755-761; A.H. Gentry and J. @ez-Par~
Deforestation and Increased Flooding of the Upper AmazoU Science, vol. 210, 1980, p. 1354.
52s~ti ~d Vose, op. cit., foo~ote 51; He~ ~sso, J- ~uis Gui~ume~ and l@cohrI I@l@. md Use Problems fi a Tropical Fores4
AMBZO, vol. 10, No. 2-3, 1981.
ssDe~s ~erso~ Declining Tree Stocks in African Countries,
World Development, vol. 14, No. 7, 1986, pp. 853-863.
Chapter 5--Energy and the Environment in Developing Countries . 121
Fuel and FodderAs forests and grassl ands
di sappear, rural peopl e are i ncreasi ngl y forced to
rel y on agri cul tural resi dues and dung for thei r
cooki ng and heati ng needs, and on crop resi dues for
ani mal fodder. The fai l ure to return organi c materi -
al s to the soi l can have si gni fi cant envi ronmental
i mpacts, even i f these i mpacts are l onger term and
more subtl e than i s someti mes suggested. Peopl e i n
many areas al ready di vert organi c resi dues to other
uses, often wi th l i ttl e apparent near-term effect on
yi el ds. For exampl e, crop resi dues such as mi l l et or
sorghum stal ks tend to be poor ferti l i zers and are
di ffi cul t to recycl e; they are often burned i n the
fi el ds to prevent them from harbori ng crop pests.
Si mi l arl y, dung qui ckl y l oses i ts ni trogen and much
of i ts effecti veness as a ferti l i zer when l eft l yi ng i n
the sun, as is comrnon.
54
In areas with poor soi l s
and/or hi gh rai nfal l s that qui ckl y l each nutri ents out,
however, crop yi el ds may drop qui ckl y i f resi dues
are not returned to the soi l .
I n the l onger term, the l oss of organi c materi al can
reduce the producti vi ty of even the hi ghest qual i ty
soi l s. Organi c matter i n soi l s provi des i mportant
nutri ents needed by pl ants; i t hel ps the soi l bi nd
i mportant mi neral s-e. g., magnesi um, cal ci um, and
potassi um-that woul d otherwi se be l eached away;
i t buffers the aci di ty of the soi l ; and i t i mproves
water retenti on and other physi cal characteri sti cs.
55
WaferWhen the natural water regul ati on sys-
tem provi ded by forests and grassl ands i s removed,
stream fl ows tend to become more errati c, wi th
reduced fl ows duri ng dry seasons and worse fl oods
i n the wet season. Thi s can i nterfere wi th agri cul ture,
fi shi ng, and darns and can threaten i nhabi tants.
Table 5-4-Sedimentation Rates of Some
Reservoirs in India
Lifetime (years)
Reservoir Planned Revised
Bhakra . . . . . . . . . . . . . 88 47
Maithon . . . . . . . . . . . . . 246 24
Hirakund . . . . . . . . . . . . 100 35
Ram Ganga . . . . . . . . . 185 48
SOURCE: Kunwar Jalees, Loss of Productive Soil in India, /nterndona/
Journal of Envfronmentai Studies, vol. 24,1985, pp. 245-250.
Eroded soi l s choke downstream waterways and
reservoi rs, reduci ng thei r abi l i ty to handl e the
i ncreased vol umes of water runni ng di rectl y off the
water sheds.
56
Over the past 10 years, the area
annual l y fl ooded i n I ndi a has i ncreased by 18
percent.
57
Some observers attri bute thi s i ncreased
fl oodi ng to the cl eari ng of regi onal forests. Fl ood
and erosi on damage due to the cl eari ng of I ndi as
forests has been esti mated at $20 bi l l i on for the
peri od from roughl y 1960 to 1980, i ncl udi ng l oss of
topsoi l , l oss of property to fl oods, and shortened
reservoi r l i feti mes (tabl e 5-4).
58
Other esti mates
pl ace the di rect costs of repai ri ng fl ood damage i n
I ndi a at more than $250 mi l l i on per year.
59
ClimateIn some regi ons, a si gni fi cant porti on
of the rai nfal l i s generated from moi sture pumped
back i nto the atmosphere by vegetati on. Removal of
thi s vegetati on may then contri bute to cl i mati c
change i n the regi on.
60
The surface refl ectance i s
changed and may l i kewi se affect cl i mate.
61
To the
extent that the l ocal cl i mate changes due to the l oss
of vegetati on, i t may become more di ffi cul t to
reverse the process and restore grassl ands and
WGoCo Agg-~, J~&cio~U= of ~g in The l%ird World, Energy, vol. 14, No. 6, 19*9, PP. 349-352
SsGeoffr~ Barnard and IAWS KristofersoU Agncultural Residues as Fuel in The Third WorZd (London: -CmW*S.
MWp. fiho~~sing Ground: EnViro~ntal Stress and World Food Prospects (New York NY: W.W. Norton& CO., IW@.
57u.s. Conwss, GflIce of Technology Assessment, op. cit., footnote Al.
sg~id. Amoreg~m~reviewof this problem inlnd~ is given in The State ofIndiasEnvironment 1984-84; The SecondCitizensReport (New Delhi:
Center for Science and Environment).
sgJohn Speazs, Preserving Watershed Environment.s, UM4SYLVA, VOI. 34, No. 137, 1982, pp. 10-14.
@Sag~ et al., op. cit., footnote 30; J. Shukls and Y. Mintz. Inthence of Land-Surface Evspotranspiration on the Earths Climate, Science, VO1.
215, 1982, pp. 1498-1501.
slRobert S. ~del, Mechanisms Governing the Cliwte of the Sahel:A Survey ofRecent Modelling and Observational Studies, OECD/~S/Club
du Sahel, Report Sahel D(84)252, October 1984. See also Sagan et al., op. cit., footnote 30, estimates of slbedo changes are given in Gregory Morris,
Environmental Impacts of Bioenergy Developmen~ Biomass Energy Systems: Building Blocks for Sustainable Agriculture, Jan. 29-Feb. 1, 1985,
Airlie House, Airlie, VA, The Rockefeller Brothers Fund and World Resources Institute, Wsshingto~ DC.
forests to somethi ng approachi ng thei r ori gi nal form
and extent.
62
Bi odi versi ty
63
--Pl ants i n tropi cal rai n forests
have evol ved a parti cul arl y ri ch array of potenti al l y
useful chemi cal s, yet fewer than 1 percent of al l
tropi cal pl ant speci es have so far been screened for
bi ochemi cal of use to man.
64
A number of val uabl e
medi cal products have al ready been devel oped from
tropi cal pl ants, i ncl udi ng hi ghl y effecti ve drugs for
l eukemi a and hypertensi on, muscl e rel axants, and
other s.
65
Many other aestheti c, sci enti fi c, and com-
merci al benefi ts may be real i zed from tropi cal forest
products. Tropi cal deforestati on and the l oss of
speci es woul d forecl ose many of these opportuni -
ti es.
66
Al though they have recei ved much l ess attenti on,
pl ants i n ari d and semi -ari d regi ons si mi l arl y offer
many potenti al benefi ts. These areas harbor a
number of wi l d rel ati ves to our basi c crops-beans,
potatoes, wheats, barl eys, mi l l ets, sorghums, and
many othersand are sources for geneti c resi stance
to drought, heat, pests, and di sease. The di verse
geneti c resources of wi l d vari eti es can be an
i mportant i nput i nto the hi gher yi el di ng but geneti -
cal l y narrow vari eti es used i n i ntensi ve agri cul ture.
Many val uabl e geneti c strai ns and speci es may beat
ri sk i n these ari d and semi -ari d regi ons.
67
Deforestati on can resul t i n a di rect confl i ct
between the survi val of speci es and the l i vel i hoods
of peopl e who expl oi t the l and. Such a confl i ct
recentl y erupted i n the Uni ted States over the
endangered northern spotted owl .
68
Ther e ar e si mi l ar
confl i cts i n many devel opi ng countri es. Tropi cal
rai n forests contai n at l east hal f and perhaps as many
as 90 percent or more of the speci es found on earth,
yet cover just 7 percent of the total l and area. No one
knows preci sel y the share of speci es hel d by tropi cal
rai n forests, because fewer than 2 mi l l i on speci es
have been offi ci al l y cl assi fi ed out of a total number
that i s vari ousl y esti mated to range from l ess than 5
to more than 50 mi l l i on.
69
Many countri es are taki ng
steps to protect thei r bi odi versi ty, and a few are
real i zi ng some success. Protecti on of the brow-
antl ered deer i n northeast I ndi a, for exampl e, i s
hel pi ng i t make a comeback from near exti ncti on.
70
There are concerns i n devel opi ng countri es,
however, that the costs of conservi ng bi odi versi ty
i n jobs and/or access to l and and resources for the
poor, securi ty enforcement, admi ni strati on, and the
mi strust generated as governments deny thei r own
peopl e access to much-needed resources-wi l l be
born by the devel opi ng countri es, but the benefi ts
wi l l be l argel y captured by commerci al i nterests i n
i ndustri al i zed countri es.
71
Ai r Pol l uti on
The burni ng of bi omass generates l arge amounts
of ai r pol l uti on i n devel opi ng countri es. Food i s
typi cal l y cooked over an open fi re or a poorl y vented
stove-exposi ng women and chi l dren, as wel l as
other members of the fami l y, to hi gh l evel s of toxi c
smoke (ch. 3). Si mi l arl y, i n col der cl i mates, homes
i n rural areas are often heated by open fi res, further
i ncreasi ng exposures to toxi c smoke.
72
One i mpact
of thi s i ndoor ai r pol l uti on i s exacerbati on of one of
the most deadl y cl asses of i nfecti ous i l l ness, acute
respi ratory i nfecti ons i n chi l dren.
The use of bi omass for fuel , cl eari ng forest l and
for agri cul ture, and burni ng grassl ands to generate
GZJ. Shuklz C. Nobre, P. Sellers,
AmazonDeforestation and Climate Change Science, vol. 247, Mar. 16, 1990, pp. 1322-1325. H. Schlesinger et
al., op. cit., footnote 26.
GsFor a complete review of this complex issue, see U.S. Congress, Offke of Technology Assessmen4 Technology to Maintain BioZogicaZDivertify,
OZ&F-330 (Springfield, VA: National Technical Information Service, March 1987).
~Nefl A. Belso~ Tropical Deforestation and the Response of the United SGXfX COWWN,
Georgetown International Environmental Law Review,
vol. 2, Fall 1989; U.S. Congress, Office of Technology Assessmen~ op. cit., footnote 41.
fiu.s. CO~SS, tilce of Technology Assessment, op. cit., f@@e 41.
66~w~d (). Wilsou ~~~ats t. Biodiv~i@,~> Scientific American, &@e~r 1989, pp. 108-1 16; w~ter V. Reid md Kenton R. Miller, Keeping
Options Alive: The Scientific Basis for Conserving Biodiversity (Washingto% DC: World Resources Institute, October 1989).
GTG~Nab~~ HowAreTropi~ Defores~on and Desertiflcation Affecting Plant Genetic Resources,
Annals ofEarth, vol. 4, No. 1, pp. 21-22,
1986; Paul Raebur% Seeds of Despair, Zssues in Science and Technology, Winter 1989-90, pp. 71-76.
6sLibrq of Congress, Congressional Research !kviw,
Economic Impacts of Rotecting the Northern Spotted Owl, 90-74 ~ Mar. 5, 1990.
@Robert M. May, How Many Species Are There on Ear@ Science, Sept. 16, 1988, pp. 1441-1449.
Sanjoy Hamrika, A Deer Comes Back From the Brink in India, New York Times, July 31, 1990.
71~c~o ~ulovicm A View on Tropic~ Defo~s@tio~ Na~e, VOI. 346, J~y 19, 1~, p. 214.
72~ R. s~~ Bio@el~, Ai
r
Pollution, andHealth: A Glo~lRe~w mew York ~: plen~ press, 1987).
fresh forage for l i vestock al l generate l arge amounts
of smoke that contri butes to regi onal ai r pol l uti on.
These acti vi ti es al so pump greenhouse gases i nto the
atmosphere, potenti al l y contri buti ng to gl obal cl i -
mate change.
73 This is discussed in much more de-
tai l i n the forthcomi ng OTA report, Changing By
Degrees: Steps To Reduce Greenhouse Gases.
The Urban Sector
Causes of Environmental Degradation
Urban areas of devel opi ng countri es are growi ng
rapi dl y, i n l arge part due to mi grati on from rural
areas. For the l ow-i ncome countri es, urban popul a-
ti on jumped from 17 percent of total popul ati on i n
1965 to 30 percent i n 1987. I n 1960, there were 59
cities with more than 500,000 persons i n devel opi ng
countri es; by 1980 that number had grown to 165
ci ti es.
74
I n the earl y 1990s, ci ti es i n devel opi ng
countr i es wi l l need to absor b mor e than 100 mi l l i on
addi ti onal peopl e annual l y.
75
Thi s trend has i mpor-
tant i mpl i cati ons for both energy use and energy-
rel ated envi ronmental i mpacts.
Urbani zati on and moderni zati on can provi de many
desi rabl e benefi ts-i mproved standards of l i vi ng,
i ncreased opportuni ti es for educati on and empl oy-
ment, and greater i nsul ati on from the vagari es of
drought endured by those i n rural areas. (They al so
reduce pressures on some aspects of the rural
envi ronment.) They al so have costs, i ncl udi ng
potenti al damage to the envi ronment.
Urbani zati on changes the consumpti on patterns
of goods and energy-rel ated servi ces (ch. 3). House-
hol ds make a transi ti on from bi omass to commerci al
fuel s for cooki ng and other domesti c servi ces.
Demand for consumer goods, notabl y el ectri c appl i -
ances, i ncreases. The economi c base changes from
agri cul ture and smal l rural i ndustry to l arger manu-
facturi ng and servi ces, wi th a correspondi ng change
i n the demand for commerci al energy. The need for
publ i c and personal transport grows as empl oyment
shi fts from agri cul ture or rural i ndustry l ocated
wi thi n wal ki ng di stance of resi dences to l arge
i ndustry and commerce l ocated further away. The
hi gh concentrati on of peopl e requi res the transport
of food, fuel , and other materi al s from l ong di s-
tances, as wel l as effecti ve management of wastes.
These changes have si gni fi cant i mpacts on energy
use and on the envi ronment.
Resi denti al /Commerci al Sector
The transi ti on from bi omass to cl ean commerci al
fuel s such as l i quefi ed petrol eum gas (LPG) and
el ectri ci ty for resi denti al cooki ng general l y reduces
the total amount of ai r pol l uti on emi tted and l argel y
shi fts that whi ch i s emi tted from the househol d to
di stant refi neri es and el ectri ci ty generati on pl ants
(ch. 3).
76
I n some countri es, however, cl ean com-
merci al fuel s are not wi del y avai l abl e or remai n too
expensi ve. Chi na, for exampl e, uses about one-thi rd
of i ts coal i n resi dences, of whi ch nearl y hal f i s used
for cooki ng (app. 3-A). Due i n part to heavy
resi denti al coal use, some northern Chi nese ci ti es,
such as Bei ji ng and Ti anji n, have very hi gh sul fur
di oxi de concentrati ons.
Li ghts, refri gerators, ai r condi ti oners, and other
el ectri cal appl i ances i n the resi denti al /commerci al
sector provi de hi ghl y desi rabl e servi ces and are
penetrati ng urban areas much more rapi dl y than
rural areas (ch. 3). These appl i ances consume l arge
amounts of el ectri ci ty, however, the generati on of
whi ch can cause si gni fi cant envi ronmental damage
i f not properl y control l ed. Refri gerators and ai r
condi ti oners al so use chl orofl uorocarbons (CFCs),
whi ch have al ready damaged the earths protecti ve
ozone l ayer and are potent greenhouse gases.
I ndustri al Sector
I ndustry provi des many goods that contri bute to
our materi al comfort and wel l -bei ng. I ndustry can
damage the envi ronment through a vari ety of
acti vi ti es, however, i f they are mi smanaged or
i nadequatel y control l ed. These i ncl ude: pl aci ng
heavy demands on natural resources as feedstocks or
other i nputs; i ntensi vel y usi ng el ectri ci ty for me-
chani cal dri ve and other needs; burni ng fossi l fuel s
TqR.p. D~~~ ~d -1= A-s. ~, cTropic~ For~ts ad he Glob~ &bon ~cle, science, VO1. 239, J ~. 1, lgs& pp. 4247; Ritid A.
Houghto~ TheFutureRoleof Tropical Forests inliffecting the Carbon Dioxide Concentration of the Atmosphere, AMBlO, vol. 19, No. 4, July 1990,
pp. 204-209.
Tdworld B@ World D#e/op~nt Repo~ 1989 (New York NY: OxfOrd Ufivemity press, 1989).
TsWor/dDoezopmntReportJ990, op. cit.; ~d,icator~bles 1 ~d 31, me mb~pop~tio~l percent of thedevelopingcountrytoti of 3.95 billion
people-1.6 billion people and is growing at 6.9 percent annually.
76~~ ~efer~ t. ~~n monoxide, p~c~t~, - diofide, fi~gen oxides, and o~er such che~~ ~ies. It does not refer to carbon dioxide,
which may be increased by the use of nonbiomass fuels but not by biomass if the biomass is being produced on a sustainable basis.
for process heat; and generati ng hazardous wastes
that may be di scharged i nto l andfi l l s, water systems,
or the ai r. Each i ndustry has a di fferent mi x of these
acti vi ti es, and each acti vi ty has a di fferent set of
envi ronmental i mpacts, dependi ng on the use of
envi ronmental control s and other means of mi ti ga-
ti on.
I ndustri al pol l uti on can pose a parti cul arl y severe
heal th hazard i n devel opi ng countri es when i ndus-
tri es are establ i shed near exi sti ng resi denti al areas,
or when resi denti al areas bui l dup around them. Thi s
has l ed to tragedi es such as Bhopal , as wel l as seri ous
systemi c pol l uti on such as found i n Cubatao, Brazi l ,
i n the 1970s and earl y 1980s.
77
Often, the l owest
i ncome peopl e are most seri ousl y affected. Thi s can
l ead to the si tuati on i n whi ch the poor face al l of thei r
tradi ti onal ri sks to heal th-i nfecti ous di sease, hun-
ger and mal nutri ti on, ai r pol l uti on from bi omass
fuel s-and at the same ti me face even greater than
normal modem ri sks-such as exposure to hazard-
ous wastes and toxi c ai r pol l utants.
78
Transportati on Sector
The transportati on sector i s the l argest contri butor
to ai r pol l uti on i n many ci ti es. For exampl e, i n
I ndi an ci ti es, gasol i ne-fuel ed vehi cl esmostl y two
and three wheel ersare responsi bl e for 85 percent
of carbon monoxi de and 35 to 65 percent of
hydrocarbons i n the ai r from fossi l fuel s.
79
Di esel
vehi cl esbuses and trucks-are responsi bl e for
over 90 percent of ni trogen oxi de (NO
X
) emi ssi ons
i n urban I ndi a.
80
Use of emi ssi on control devi ces and
engi ne modi fi cati ons have si gni fi cantl y reduced
exhaust emi ssi ons on newer vehi cl es i n the Uni ted
States, but these pol l uti on control s are usual l y not
standard on vehi cl es i n devel opi ng countri es. Runoff
from roads al so contri butes to water pol l uti on.
I n each of these sectors, there are often di ffi cul t
tradeoffs between l onger term envi ronmental i m-
acts and i mmedi ate fi nanci al costs to consumers.
Many devel opi ng countri es al so have shortages of
ski l l ed techni cal manpower to i mpl ement mi ti gati on
efforts. The extent to whi ch technol ogi cal i nnova-
ti ons or other advances mi ght ease these tradeoffs
wi l l be exami ned i n a l ater report of thi s OTA study.
Environmental Impacts
Ai r
Ai r qual i ty i n many of the devel opi ng worl ds
ci ti es i s poor, and has been deteri orati ng. The Uni ted
Nati ons Envi ronment Program (UNEP) esti mates
that up to one-hal f of the worl ds urban popul ati on,
i ncl udi ng resi dents of many i ndustri al i zed countri es,
l i ve i n ar eas wi th mar gi nal or unacceptabl e l evel s of
sul fur di oxi de (S0
2
) i n the ai r. Concentrati ons of
SO
2
(see fi gure 5-1) and concentrati ons of total
suspended parti cul ate (see fi gure 5-2) i n major
ci ti es i n the devel opi ng worl d are consi derabl y
above Worl d Heal th Organi zati on gui del i nes. Photo-
chemi cal smog has become a recurrent seasonal
probl em i n many l arge tropi cal and subtropi cal
ci ti es.
The sources of these pol l utants vary. Coal -used
for el ectri ci ty generati on, i ndustri al power and
process heat, and domesti c heati ng and cooki ng
(Chi na)--pri mari l y emi ts sul fur di oxi de, parti cu-
l ate, and ni trogen oxi des. The combusti on of oi l or
gas i n stati onary sources, such as el ectri c generati ng
uni ts, emi ts many of the same pol l utants-ni trogen
oxi des, hydrocarbons, parti cul ate, and sul fur di ox-
i de--but i n much di fferent proporti ons.
81
Ni trogen
oxi de, a major i ngredi ent of urban smog, i s the most
harmful pol l utant rel eased on a l arge scal e when
el ectri ci ty i s generated from oi l or gas. Mobi l e
sources--cars, trucks, two and three wheel ers, and
buses-rel ease l arge amounts of carbon monoxi de,
NO=, hydrocarbons, and parti cul ate and are typi -
cal l y the l argest source of these pol l utants i n urban
ar eas. These mobi l e sour ces ar e al so major contr i bu-
tors of l ead pol l uti on due to the use of l ead as an
octane booster i n gasol i ne. Burni ng fossi l fuel s
unavoi dabl y generates carbon di oxi de, the most
i mportant greenhouse gas. These pol l utants (except
carbon di oxi de (C0
2
)) can damage crops, forests,
and structures and can aggravate human heal th
pr obl ems.
~world Resources Institute, World Resources 1990-91 (New York NY: Oxford University hess, W90), P. 41.
Tsk the more gene~ C&, mere is a shift in the ~es of risk that people are exposed to from the traditional to b modem. S* fik R. Smi@ me
Risk Transitio~ Znternatiomd Environmental &airs, vol. 2, No. 3, in press.
~ta Energy Research Institute, T~Energy Data Directory and Yearbook 1988 (New Deti India: 1989), p. 250.
%id.
81 Mitre Corp., Healthand Envinmmental Effeets of Oil and Gas Technologies: Researeh Needs, report to the Federal Interagency Committee on
the Health and Environmental Effects of Energy Technologies, July 1981.
Figure 5-1-SuIfur Dioxide Levels in Selected Cities, 1980-84
3
1
Concentration (ug/m
3
)
10
/ \
100
I I I I I I I I I I I I I I I I I I I
I I I
Key
1. Milan
2.. Shenyang 3.Tehran
4. Seoul
5. Rio de Janeiro
6. Sao Paulo
7. Xian
8. Paris
9. Beijing
10. Madrid
11. Manila
12. Guangzhou
13. Glasgow
14. Frankfurt
15. Zagreb
16. Santiago
17. Brussels
18. Calcutta
19. London
20. New York City
21. Shanghai
22.Hong Kong 23. Dublin
24. St. Louis
25. Medellin
26. Montreal
27. New Delhi
28. Warsaw
29. Athens
30. Wroclaw
31. Tokyo
32. Caracas
33. Osaka
34. Hamilton
35. Amsterdam
36. Copenhagen
37. Bombay
38. Christchurch
39. Sydney
40. Lisbon
41. Helsinki
42. Munich
43. Kuala Lumpur
44. Houston
45. Chicago
46. Bangkok
47. Toronto
48. Vancouver
49. Bucharest
50. Tel Aviv
51. Cali
52. Auckland
53. Melbourne
54. Craiova
I I
[ 1.
annual averages
I
I 12.
~
I I 3.
I I 14.
4
[ I
I 5.
Combined site
I I 6.
I
I
I 7.
average 1980-84
I
I 8.
[
I
19.
I I I 10.
I I
I 11.
I
I 12.
I I 13.
I
14.
I
[ 15.
I
[ 16.
I 17
:
1
I
18.
I I
I 19.
I
[ 20.
I
I
21.
I I
22.
I
I ~ 23.
24.
I
25.
I
126.
I I
127,
I
]28.
I
]29.
\ , ,1
30.
31.
132.
I I
[ 33.
[
] 34
1 35.
I I
36.
[
] 37.
1
] 36.
1
I
I 39.
I
I
I
~42.
43.
I I
47.
I
] 48.
I
] 49.
150.
I I
51.
I
I
52.
I
J 53.
I
I
I 54.
a Note logarithmic scale.
SOURCE: Wortd Health Organization and United Nations Environment Programme, G/o&/ Po//ution md/fea/th (Imndon: Yale University Press, 1987), figure
2, p. 5. (Logarithmic scale.)
Figure 5-2-Suspended Particulate Matter Levels in Selected Cities, 1980-84
3
10 ~ 1 0 0
1,000
I I I I I I I I I I I I I I I I I I
Key
1. Kuwait
2. Shenyang
3. Xian
4. New Delhi
5. Beijing
6. Calcutta
7. Tehran
8. Jakarta
9. Shanghai
10. Guangzhou
11. Illigan City
12. Bangkok
13. Bombay
14. Kuala Lumpur
15. Zagreb
16. Rio de Janeiro
17. Bucharest
18. Accra
19. Lisbon
20. Manila
21. Chicago
22. Caracas
23. Birmingham
24. Helsinki
25. Hamilton
26. Sydney
27. Houston
28. Craiova
29. Toronto
30. Melbourne
31. Medellin
32. Chattanooga
33. Fairfield
34. Montreal
35. Vancouver
36. New York City
37. Tokyo
38. Osaka
39. Cali
40. Copenhagen
41. Frankfurt
I 1
J 1.
annual averages
I
2.
I I
I I
3.
I I 1
I I
4.
4 1 I
5.
Combined site I I
16.
average 1980.84
I
I
7.
I I
16.
I I
I 9.
[
I
10.
I 11.
[
I
12.
I
13.
I
I
14.
I
15.
1!
16.
I
17. . , *
/ ,
1
16.
19.
I
20.
II
21.
I l l
22.
I
I
I
I I
25.
I
26.
I I
127.
[
26.
129.
I I
33.
I
34.
1 I
35.
] 36.
I I
37.
38.
39.
I I
140.
1 1 1
41.
a Note logarithmic scale.
SOURCE: Wortd Health Organization and United Nations Environment Programme, Globa/PolMion andlfedth (Imndon: Yale University Press, 1987), figure
3, p. 6. (kgarithmie scale.)
Efforts to reduce emi ssi ons from coal burni ng bers) control s sul fur emi ssi ons by capturi ng the
usual l y i nvol ve removi ng harmful materi al from sul fur i n the form of l i qui d sl udges. El ectrostati c
exhaust gases after combusti on. Pol l uti on control s,
preci pi tators are wi del y used i n the devel opi ng
such as el ectrostati c preci pi tators or baghouses to worl d, but FGD systems are not common.
82
Thi s
control parti cul ate, capture pol l utants from exhaust may be due i n part to the hi gh costs of such systems.
gases, but they create sol i d wastes i n the process. One study i n I ndi a found that the addi ti on of an FGD
Fl ue gas desul furi zati on (FGD) equi pment (scrub- system woul d add about 15 percent to the cost of
8~~rmtio~~velopment ~d lhm~ As=~tm, ~c.~
Clean Coal Technologies for Developing Countries, contractor report prepared for the
Mice of Technology Assessment Msy 1990, p. 25.
el ectri ci ty .83 Perhaps as si gni fi cant i s that I ndi an and
Chi nese coal s have rel ati vel y l ow sul fur contents
typi cal l y 0.5 to 2.0 percent compared to 3 to 6
percent for Eastern European and some U.S. Coal s.
84
Adopti on of more effecti ve pol l uti on control
technol ogi es al ready common i n the i ndustri al i zed
worl d and i mprovements i n energy effi ci enci es
mi ght si gni fi cantl y i mprove ai r qual i ty i n many
ci ti es of the devel opi ng worl d.
Water
The pri mary urban water pol l uti on probl em i s the
di scharge of untreated or mi ni mal l y treated human
waste i nto water systems (thi s i s al so a si gni fi cant
probl em i n rural areas). As of the mi d-1980s, onl y
about hal f the devel opi ng worl ds urban popul ati on
had adequate sewage faci l i ti es, and perhaps three-
fourths had access to safe
dri nki ng water. Thi s
si tuati on contri butes to the growth and spread of
pathogeni c vi ruses, bacteri a, and protozoa and i s
di rectl y or i ndi rectl y responsi bl e for perhaps three-
fourths of al l i l l ness and 80 percent of al l chi l d
deaths.
85
I ndustri al wastes al so contri bute si gni fi cantl y to
water pol l uti on i n some areas. These range from
agri cul tural i ndustry wastes to heavy metal s (mer-
cury, cadmi um, l ead, and chromi um) and chl ori n-
ated hydrocarbons-to name onl y a few. These
pol l utants can have seri ous i mpacts on water re-
sources al l the way downstream to the ocean. Many
coastal areas are al so bei ng affected.
86
A recent
study found that Worl d Heal th Organi zati on (WHO)
standards for heavy metal s-cadmi um, l ead, and
mercury-were exceeded i n one-thi rd to three-
fourths of the fi sh and shel l fi sh sampl es taken from
Jakarta Bay, I ndonesi a.
87
Mitigation strategi es that
mi ght be appl i ed i ncl ude treatment of i ndustri al
effl uents and reducti on of the generati on of hazard-
ous materi al s themsel ves through i mprovements i n
manufacturi ng processes.
88
Land
The major causes of soi l pol l uti on i n urban areas
are the di sposal of sol i d wastes and the deposi ti on of
ai r pol l utants such as S0
2
and parti cul ate. The
focus here i s on sol i d wastes. Low-i ncome countri es
generate much l ess urban refuse (around 0.5 kg per
capi ta per day) than i ndustri al countri es (0.7 to 1.8
kg per capi ta per day).
89
I n l arge ci ti es, however,
even modest per-capi ta l evel s can l ead to l arge
amounts of waste generati on. Mexi co Ci ty i s esti -
mated to generate 11,000 tons a day. Most of thi s
waste ori gi nates from househol ds, l argel y i n the
form of vegetabl e and putresci bl e materi al s. There i s
al so a si gni fi cant component of hazardous wastes
(contai ni ng chemi cal s, heavy metal s, and/or radi o-
acti ve wastes) from i ndustri es.
I n the devel opi ng countri es, current procedures
for di sposi ng of such wastes are i nadequate. Much
househol d garbage i s not col l ected at al l , but l eft to
decompose. Hazardous wastes may be dumped on
uncontrol l ed muni ci pal l andfi l l s. Onl y a smal l part
of total wastes are di sposed of i n an envi ronmental l y
safe reamer. Nonethel ess, costs of even i nadequate
di sposal can amount to as much as 30 percent of
muni ci pal budgets.
90
Energy coul d, however, pl ay a part i n sol vi ng
these probl ems. Urban waste can be used as feed-
stock for cogenerati on pl ants. Process changes i n
energy-i ntensi ve i ndustri es such as steel , cement,
and ferti l i zer may wel l , by economi zi ng on the use
of al l materi al s, al so reduce hazardous wastes.
83
1bid., p. 25.
~Ibid., pp. 7, 8, 27.
gsworld R60mc~ ~sti~te, World Resources 1990-91 (New York NY: Oxford University Press, 1990); World Resources mtitute, Development
and Pollution in the Third Worl~ Tokyo Conference on the Global li!nvironment and Human Response, Sept. 11-14, 1989; World Health Organization
The International Drinkzng Water Supply and Sanitation Decade (Geneva: United Nations, 1986); World Health Organization and United Nations
Environment Programme, Global Pollution and Health (New Haveu CT: Yale University Press, 1987).
~United Nations Environment Programme,
The State of the Marine Environment, 1990.
gTWorld Reso~ces ~ti~te, Development and Pollution in the Thhd WOrlL op. Cit., footnote 85.
S8uos+ CoWss, ~lce of TW~olom ~sesmm~ Senous Red~tion of Hazar&us Waste For pollution prevention and lnd~t?ial Eficiency,
OTA-ITE-317 (Washingto~ DC: U.S. Government Printing Gfllce, September 1986). U.S. Congress, ~lce of Technology Assessmen~From Po/lution
to Prevention: A Progress Repon on Waste Reduction, OTA-ITE-347 (Wasbingto~ DC: U.S. Government Printing Office, June 1987).
89sm~ J. coin~~u, J7nviron~ntalMamgeme~ of ur~n Solid Wastes in Developing Counmes:A proje~ @fi, quoted in Development ~d
Pollution in the Third Worl~ prepared by The World Resources Institute for the To~o Conference on the Global Environment and Human Response,
Sept. 11-14, 1989 (Washingto~DC: World Resources Institute, July 1989).
World Resources Institute, World Resources 1988-89 (New York NY: Basic Books, Inc., 1988), p. 46.
Energy Production and Conversion
Energy pl ays a cri ti cal rol e i n fuel i ng the transi -
ti on from a tradi ti onal to a modern soci ety and thus
ai di ng economi c and soci al devel opment. I n turn,
thi s energy i s extracted from rural areas-ei ther as
fossi l fuel s or renewabl e energy-and can have
si gni fi cant i mpacts on the rural economy and
envi ronment. Energy thus provi des an i mportant
two-way l i nkage between rural and urban areas.
Used wi sel y, energy can provi de envi ronmental
benefi ts; i f mi sused i t can al so exact substanti al
envi ronmental costs to the l and, water, and ai r. The
many soci al , economi c, and envi ronmental benefi ts
have been revi ewed i n the precedi ng chapters; the
envi ronmental costs of extracti ng and transporti ng
the major energy resources used i n devel opi ng
countri es today--coal , oi l , gas, hydroel ectri ci ty, and
bi omass-are bri efl y revi ewed bel ow.
Coal
Coal has si gni fi cant envi ronmental effects through-
out the fuel cycl e.
91
The envi ronmental effects of
coal mi ni ng depend on the techni ques used. Mi ni ng
methods are sel ected accordi ng to the depth of the
coal , the thi ckness of the seams, and the avai l abi l i ty
of capi tal and equi pment. I n surface mi ni ng, topsoi l
and overburden are removed to expose the coal . The
mi ni ng process can di sturb surface l ands and waters,
and may al so contami nate or di sturb underground
aqui fers. I ncreased erosi on, downstream si l tati on,
and water contami nati on can fol l ow i f excavated
materi al i s not properl y managed duri ng mi ni ng.
Soi l producti vi ty and water resources can be de-
graded i f l ands are not recl ai med adequatel y. The
most severe i mpacts associ ated wi th underground
mi ni ng are surface subsi dence over mi ned-out areas,
di srupti on of aqui fers, and contami nati on of water
by aci d drai nage. Addi ti onal l y, dust and emi ssi ons
from coal mi ni ng, preparati on, and transport or
rel ated equi pment can contri bute to l ocal ai r pol l u-
ti on.
92
Oil and Gas
Oi l and gas producti on have si mi l ar envi ron-
mental i mpacts. Duri ng expl orati on and producti on,
the major envi ronmental concerns onshore i ncl ude
l and di sturbances, aqui fer contami nati on, l eaks and
spi l l s, and di sposal of l i qui d and sol i d wastes. For
offshore operati ons, concerns i ncl ude the i mpacts of
operati ons on fi sheri es and mari ne habi tats, l eaks
and spi l l s, and waste di sposal .
Large spi l l s of oi l or petrol eum products, whi ch
occasi onal l y occur duri ng producti on, storage, or
transportati on, are perhaps the most dramati c envi -
ronmental threat associ ated wi th oi l producti on.
Smal l l eaks and spi l l s are much more common,
however, and may have a greater overal l i mpact on
the envi ronment. Oi l spi l l s can poi son fi sh and
aquati c ani mal and pl ant l i fe. Addi ti onal l y, sedi -
ments can trap oi l , creati ng a l ong-term source of
pol l uti on. The wel l -publ i ci zed Al askan oi l spi l l of
1989 reveal ed the di ffi cul ti es i nherent i n cl eani ng up
l arge oi l spi l l s i n di ffi cul t ocean envi ronments.
93
Devel opi ng countri es have had thei r share of oi l
spi l l s too. The June 1979 bl owout of the I XTOC 1
producti on wel l i n the Mexi can Cari bbean was the
l argest oi l spi l l to date.
94
I t rel eased an esti mated 139
to 428 mi l l i on gal l ons of crude oi l i nto surroundi ng
waters, as much as 40 ti mes the amount spi l l ed by
the Exxon Val dez. Natural gas l eaks pose l i ttl e toxi c
threat to pl ants and ani mal s but can expl ode or cause
fi res. The pri mary consti tuent of natural gas
methane--i s al so a potent greenhouse gas.
Petrol eum refi ni ng generates hazardous l i qui d
and sol i d wastes, as wel l as ai r emi ssi ons. These
faci l i ti es have pol l uti on control requi rements si mi l ar
to many other l arge i ndustri al and chemi cal pl ants,
but l eaks of toxi c compounds occur neverthel ess.
Hydroelectricity
The major envi ronmental i mpacts of l arge hydro-
power producti on projects resul t from the i ni ti al
constructi on of the hydropower faci l i ti es, fi l l i ng the
reservoi rs, and changi ng ri ver fl ows. Thi s i s i n
91u.s. C~~B=S, OfiCe of T~~oloW As=ssment, The Dire~ Use of Coal, OTA-E86 (Sp~leI& VA: Natio@ Technic~ hlfo~tion SCIViCe,
April 1979), p. 186.
%bid.
93For a discussion of the technologies reklted to ~
spill cleanup, see U.S. Congress, Office of Technology Assessmen~ Coping With An Oiled
Sea: An Analysis of Oil Spill Response Technologies, OTA-BP-O-63 (WashingtorL DC: U.S. Government Printing Office, March 1990).
~~id. WY of tie lmge spills listed in the OTA background paper occumd in or near developing COuntrieS.
contrast wi th other energy sources, for whi ch major
i mpacts are spread over the enti re fuel cycl e.
Large dam constructi on often requi res the cl ear-
i ng of l ands for access routes and someti mes for
removal of constructi on materi al , wi th resul ti ng soi l
degradati on and erosi on. Fi l l i ng the reservoi r can
fl ood l arge tracts of l and, uprooti ng peopl e and
l eadi ng to l oss of forests, wi l dl i fe habi tat, and
speci es di versi ty. For exampl e the Akosombo reser-
voi r on the Vol ta i n Ghana, wi th a l and requi rement
of 8,730 km
2
, approaches the si ze of such smal l
countri es as Lebanon or Cyprus.
95
Some currentl y
proposed very l arge dam projects woul d i nundate
hi ghl y popul ated val l eys and requi re l arge-scal e
resettl ement of l ocal resi dents.
The Three Gorges Dam project under consi dera-
ti on for the Yangtze ri ver i n Chi na woul d be the
worl ds l argest hydroel ectri c project, generati ng the
equi val ent of approxi matel y one-fi fth the hydroel ec-
tri ci ty currentl y produced i n al l of Chi na each year,
as wel l as provi di ng fl ood control . Debate over thi s
project has conti nued for si x decades because i t
woul d entai l the resettl ement of approxi matel y 1
mi l l i on peopl e; because i t woul d deface one of
Chi nas most famous natural si tes, and because
cri ti cs bel i eve that power coul d be provi ded more
cheapl y wi th smal l er thermal and hydroel ectri c
pl ants or through conservati on.
96
Dams di srupt the natural fl ow of ri vers. Changed
fl ows can erode ri verbeds, al ter fl ood patterns, harm
aquati c ecosystems, and i nterrupt the spawni ng and
mi gratory patterns of fi sh and other speci es. The
i ntroducti on of a new l ake can affect water tabl es and
groundwater fl ows and i nterfere wi th the necessary
fl ow of nutri ents, and may i nduce mi crocl i mati c
changes i n humi di ty, cl oud cover, and rai nfal l . I n
addi ti on, i f the vegetati on i s not cl eared from the
reservoi r area before fl oodi ng, rotti ng organi c matter
rel eases si gni fi cant amounts of C0
2
and methane,
two potent greenhouse gases.
Hydropower faci l i ti es, i n turn, are vul nerabl e to
envi ronmental degradati on. Heavi er than expected
si l tati on of reservoi rs from deforested and/or de-
graded l ands upstream can reduce the l i fespan of
hydroel ectri c projects (see tabl e 5-4).
97
The Hi rakud
reservoi r i n I ndi a, for exampl e, was expected to be
producti ve for about 110 years, but now has an
esti mated producti ve l i feti me of 35 years.
98
I n Costa
Ri ca, excess sedi mentati on from soi l eroded from
steep sl opes pl anted wi th coffee trees over the past
20 years has caused esti mated l osses of $133 mi l l i on
to $274 mi l l i on at the Cachi hydroel ectri c stati on.
99
Dams and the i rri gati on schemes dependent on
these dams have al so contri buted to i ncreased
i nci dence of debi l i tati ng di seases, such as schi s-
tosomi asi s.
l 00
Smal l er hydroel ectri c systems and
better management may mi ni mi ze some of the above
adverse consequences.
Biomass
Bi omass fuel s, whi ch are cri ti cal to the rural and
poor urban sectors of devel opi ng countri es (chs. 2,
3, and 4), can have a vari ety of envi ronmental
i mpacts. These are di scussed above and are cl osel y
i ntertwi ned wi th deforestati on, deserti fi cati on, and
rural ai r qual i ty. A l ater report from thi s study wi l l
exami ne the envi ronmental effects of bi omass use i n
detai l , parti cul arl y i n the context of produci ng cl ean
bi omass fuel s (e.g., ethanol ) and usi ng them to fuel
modern equi pment.
Nuclear Energy
Nucl ear energy currentl y makes l i ttl e contri buti on
to the overal l energy requi rements of devel opi ng
countri es. Seven devel opi ng countri es produce ura-
ni um: South Afri ca, Ni ger, Gabon, I ndi a, Argenti na,
Brazi l , and Paki stan.
l O1
Ei ght devel opi ng countri es
had operati ng commerci al reactors as of l ate 1989
and several more had commerci al reactors under
g5R.S. P~day (cd.), Man-made Lakes and Human Health (pm- bo: University of Suriname, 1979).
wRobcrt Delfs, wealth and Woe: The LOW StfUggle tO ~eSS ~ Waters, Far Eastern Econonuc Review, vol. 147, Mar. 15, 1990, pp. 22-23;
She@ We In China, Dams Delay Spares a Valley for Now, New York Times, Apr. 18, 1989, p. Cl.
WU.S. Congress, Ofilce of Technology Assessment, op. Cit., footnote 41, p. 43.
Wbid., p. 43.
~avid Dudenhoefer, Forest Crisis Nears, The Tiw Times, vol. 34, Feb. 16, 1990.
l~fited Nations Food and Agricultural Organizatio~ op. cit., foo~ote 7, p. 69.
lolunited Nations, Energy Statistics Yearbook, 1986 (New York ~: 1988).
constructi on (ch. 4). A total of 28 devel opi ng
countri es had research reactors as of l ate 1988.
102
The conventi onal nucl ear fuel cycl e i ncl udes
ur ani um mi ni ng and processi ng, fuel fabri cati on,
el ectri ci ty generati on, and radi oacti ve waste di s-
posal . Each of these steps has the potenti al to rel ease
varyi ng amounts of toxi c and/or radi oacti ve materi -
al s to the envi ronment. Rel eases usual l y take the
form of smal l l eakages but have, on rare occasi on,
al so resul ted i n catastrophes such as Chernobyl . On
the other hand, the nucl ear power opti on rel eases
l i ttl e carbon di oxi de or other greenhouse gases to the
atmosphere, nor does i t emi t much sul fur di oxi de,
ni trogen oxi des, or other ai r pol l utants.
103
These
tradeoffs pose di ffi cul t envi ronmental choi ces.
Solar, Wind, and Other Renewable Energy
Sol ar energy can be used to heat water or dry
crops, or can be turned di rectl y i nto el ectri ci ty by
photovol tai c cel l s. Wi nds can be harnessed for
pumpi ng water or generati ng el ectri ci ty. Though
hol di ng great promi se for the provi si on of decentral -
i zed forms of energy for remote areas, as yet these
sources provi de onl y smal l amounts of energy for
devel opi ng countri es.
Geothermal energy, though sti l l a very smal l part
of total energy suppl y i n devel opi ng countri es, i s
bei ng used i n several Lati n Ameri can and Asi an
devel opi ng countri es and i n Kenya. The envi ron-
mental probl ems associ ated wi th geothermal energy
producti on are hi ghl y dependent on the geochemi cal
characteri sti cs of each speci fi c si te and the extent to
whi ch good envi ronmental practi ce i s fol l owed.
Potenti al adverse envi ronmental i mpacts i ncl ude the
rel ease of gases (parti cul arl y CO
2
), contami nati on of
l ocal aqui fers by sal i ne (and someti mes toxi c)
geothermal fl ui ds i nto groundwater, subsi dence of
l and overl yi ng wel l s from whi ch geothermal fl ui ds
have been extracted, and the generati on of hi gh-
temperature l i qui d effl uents contai ni ng metal s and
di ssol ved sol i ds. Wi th exi sti ng economi cal l y proven
technol ogi es and good management, however, these
potenti al l y adverse i mpacts can be kept under
control .
Greenhouse Gases
Climate Change
and Global
The envi ronmental i mpacts descri bed above are
l argel y l i mi ted to the i ndi vi dual countri es con-
cerned. Some acti vi ti es-notabl y, the combusti on of
fossi l fuel s and deforestati on-can have a wi der
i mpact, i ncl udi ng i mpacts on the gl obal cl i mate. An
i nternati onal panel of sci enti fi c experts of the I n-
tergovernmental Panel on Cl i mate Change (I PCC)
104
recentl y concl uded that: emi ssi ons resul ti ng from
human acti vi ti es are substanti al l y i ncreasi ng the
atmospheri c concentrati ons of the greenhouse gases:
carbon di oxi de, methane, chl orofl uorocarbons (CFCs)
and ni trous oxi de. These i ncreases wi l l enhance the
natural greenhouse effect, resul ti ng, on average, i n
an addi ti onal warmi ng of the Earths surface. 105
Under the I PCC Busi ness as Usual scenari o,
emi ssi ons of carbon di oxi de are projected to grow
from approxi matel y 7 bi l l i on metri c tonnes of car-
bon i n 1985 to 11-15 bi l l i on metri c tomes i n
2025.
106
Methane emi ssi ons are projected to i n-
cr ease fr om 300 mi l l i on metr i c tonnes i n 1985 to 500
mi l l i on metri c tonnes by 2025. Based on current
model s and wi th these trends i n greenhouse gas
emi ssi ons conti nui ng, sci enti sts predi ct that gl obal
mean temperature wi l l i ncrease at a rate of about 0.3
C per decade duri ng the next century, a rate hi gher
than that seen over the past 10,000 years.
l 07
lo~orld Resources Institute, World ResoWces 1990-1991 (New York, NY: Oxford University Press, 1990). Installed capacities ~~dq~te
indicators by themselves, as capacity factors for nuclear, thermal, and hydroelectric plants vary considerably. Installed capacities and capacity factors
can be found in World Ba~ Summary Data Sheets of 1987 Power and Commercial Energy Statistics for 100 Developing Countries, Industry and
Energy Department Working Paper Energy Series No. 23, March 1990.
IosSome -bon dioxide may be released during *
g, trampo~ or other operations by fossil-fueled equipment and some maybe released in
fabricating the construction materials-particularly cement.
104~e ~te%ove-e~~ Panel on Cltite -eWC) is ~fitemtio~ ~force -ted by tie World Meteorological Organization ~d the
United Nations Environment Program. Initiated in 1987, this body focuses on analyziug current information available on climate change issues and
devising strategies to address climate change.
los~~rgovernmen~ Panel on Climate Change, Policymakers S~
of the Scientific Assessment of Climate Change: Report to WCC From
Working Group I, June 1990, p. i.
l~~tergovernmen~ Pmel on climate Change, Working Group III (Response StIlitCgiM WOIk.iIlg tiup), policyma.kers s~ of tie
Formulation of Response Strategies: Report Prepared for WCC, June 1990, p. iii.
loTIpcc Working Group I, op. cit., footnote 105, p. ii.
Chapter 5--Energy and the Environment in Developing Countries . 131
More than hal f of the warmi ng has been attri buted
to carbon di oxi de. Atmospheri c concentrati ons of
carbon di oxi de have i ncreased by about 25 percent
si nce prei ndustri al ti mes. Thi s ri se i s l argel y due to
carbon di oxi de emi ssi ons from the burni ng of fossi l
fuel s and from deforestati on.
I n 1985, accordi ng to esti mates for the I PCC
Worki ng Group I I I , devel opi ng countri es contri b-
uted about one-quarter (26 percent) of annual gl obal
energy sector CO
2
emi ssi ons;
108
three-fourths came
from the i ndustri al i zed market countri es and the
central l y pl anned European countri es (i ncl udi ng the
U.S.S.R.). By 2025, wi th expandi ng popul ati ons and
rapi dl y i ncreasi ng energy use, devel opi ng countri es
are projected to produce roughl y 44 percent of gl obal
energy sector CO
2
emi ssi ons.
Whi l e the CO
2
emi ssi ons from the energy sector
are fai rl y wel l known, there are l arge uncertai nti es
about the contri buti on of emi ssi ons from deforesta-
ti on and other l and use changes. Thi s makes i t
di ffi cul t to cal cul ate wi th confi dence the devel opi ng
country share of both total annual and cumul ati ve
emi ssi ons for CO
2
and other gases. Esti mates of the
CO
2
emi ssi ons from tropi cal deforestati on di ffer by
a factor of four.
109
By vari ous esti mates, deforesta-
ti on coul d be the source of between roughl y 7 to 35
percent of total annual CO
2
emi ssi ons. Devel opi ng
countri es are al so responsi bl e for at l east hal f of the
gl obal anthropogeni c generati on of two other i mpor-
tant greenhouse gases, methane and ni trogen oxi des.
I n addi ti on to i ncreases i n mean gl obal tempera-
ture, other effects expected to occur wi th gl obal
warmi ng woul d i ncl ude i ncreases i n sea l evel
110
and
shi fts i n regi onal temperature, wi nd, rai nfal l , and
storm patterns. These, i n turn, woul d submerge some
l ow-l yi ng coastal areas and wetl ands, threaten bui l d-
i ngs and other structures i n these areas, and i ncrease
the sal i ni ty of coastal aqui fers and estuari es. Such
changes coul d di srupt human communi ti es and
aquati c and terrestri al ecosystems and affect food
producti on and water avai l abi l i ty.
l l l
A number of
devel opi ng countri es may be especi al l y vul nerabl e
to these effects.
112
No one, however, can yet predi ct
the ti mi ng, severi ty, or extent of the potenti al
i mpacts wi th certai nty.
The I PCC sci enti sts al so noted that cl i mate
model s suggest that control l i ng emi ssi ons can sl ow
gl obal war mi ng
, perhaps from 0.3 to 0.1 C per
decade. Emi ssi on control strategi es that countri es
coul d consi der today i ncl ude i mproved energy
effi ci ency and cl eaner energy sources-strategi es
that may al so have economi c benefi ts. The expan-
si on of forested areas, i mproved l i vestock waste
management, al tered use and formul ati on of ferti l i z-
ers, and i mproved management of l andfi l l s and
wastewater treatment mi ght al so reduce or offset
emi ssi ons.
Achi evi ng meani ngful reducti ons i n emi ssi ons
wi l l requi re unprecedented l evel s of i nternati onal
cooperati on and must i ncl ude devel opi ng countri es.
I n addi ti on to the technol ogi cal chal l enges for the
energy, agri cul ture, and i ndustri al sectors, govern-
ments of the i ndustri al and devel opi ng countri es face
chal l enges i n i mprovi ng and expandi ng i nsti tuti onal
mechani sms for technol ogy transfer to devel opi ng
countri es for provi di ng vi tal energy servi ces whi l e
l i mi ti ng emi ssi ons. These i ssues are di scussed i n the
forthcomi ng OTA report Changing By Degrees:
Steps To Reduce Greenhouse Gases.
Government Policies
Responses to the deteri orati ng qual i ty of the
envi ronment i n devel opi ng countri es show a mi xed
pi cture. On the one hand, progress i s bei ng made i n
setti ng up the government apparatus of envi ron-
mental control , and concern and acti vi ty at the
108IPCC wor~g &oup III, op. ci~, footnote 106, p. 10, table 2.
WIPcc, Policymakers summary of the Formulation of Response Strategies: Report Repared for IPCC by Working Group III, June 1990, p. 5.
Estimates of COZ emissions from deforestation and land use changes (including wood fuel) were 0.6 to 2.5 PtC in 1980 according to a report prepared
for the IPCC Working Group 1. Scientilc Assessment of Climate Change: Peer Reviewed Assessment for WG1 Plenary Meeting, May 1990, Apr.
30, 1990, p. 1-9. More recent evidence from selected countries would indicate deforestation rates now are higher than they were in 1980. Estimates of
C0
2
emissions in 1987 were 5.7 +0.5 PtC from fossil fuel burning and total emissions were of 7 +1.1 PtC.
ll~e IPCC wor~g ~OUp pr~ct~ an avemge rate of global mean sea level rise of about 6 cm per decade over the next century, 20 cm by 2030,
and 65 cm by the end of the century with significant regional variations. This increase is prirnarily due to thermal expansion of the oceans and melting
of some land ice.
11 lrntergovernmenM Panel on Clhnate Change, Policymakers summary
of the Potential Impacts of Climate Change: Report from Working Group
II to the IPCC, May 1990, p. 8.
112J .D. Milliman et rd., Environmental and Economic Implications of Rising Sea Level and Subsiding Deltas: The Nile and Bengal Examples,
AMBIO, vol. 18, 1989, pp. 340-345.
grassroots l evel i s ri si ng rapi dl y. On the other hand,
i mpl ementati on of envi ronmental pol i ci es i s fre-
quentl y di ffi cul t to achi eve, and progress i s some-
ti mes i mpeded by uni ntended envi ronmental i m-
pacts of other pol i ci es.
The past two decades have seen consi derabl e
envi ronmental pol i cymaki ng acti vi ty i n the devel op-
i ng worl d, i ncl udi ng the establ i shment of nati onal
envi r onmental agenci es and or gani zati ons
113
and the
adopti on of l aws and practi ces desi gned to prevent
envi ronmental pol l uti on and degradati on or to pro-
tect uni que natural envi ronments and wi l dl i fe. Al l i n
al l , more than 100 devel opi ng countri es now have
governmental bodi es whose responsi bi l i ti es i ncl ude
envi ronmental protecti on.
114
However, onl y a l i mi ted
number of countr i es
115
have establ i shed i ndependent
envi ronmental agenci es i n thei r central govern-
ments. Most have i ncreased thei r envi ronmental
oversi ght capabi l i ti es by creati ng or expandi ng
secti ons or offi ces wi thi n envi ronment-rel ated gov-
ernment departments such as agri cul ture, energy,
fi sheri es, forestry, i rri gati on, natural resources, or
tour i sm.
The l arger devel opi ng nati ons wi th tradi ti ons of
extensi ve government pol i cymaki ng and i mpl emen-
tati on and a rel ati vel y advanced domesti c research
base have formul ated and adopted a vari ety of
envi ronmental protecti on l aws and standards, i n-
cl udi ng the requi rement of envi ronmental i mpact
statements for l arge projects. Devel opi ng countri es
wi th detai l ed nati onwi de ai r and water qual i ty
standards or wi th regul ati ons speci fyi ng the l evel s of
permi ssi bl e pesti ci de resi dues i n food are sti l l ,
however, i n the mi nori ty. Those that do have such
regul ati ons have usual l y based thei r l egi sl ati on on
the standards i n force i n the i ndustri al i zed nati ons;
consequentl y, thei r speci fi cati ons-at l east on paper
are usual l y as stri ct as and someti mes even stri cter
than, those of the i ndustri al countri es, al though
i mpl ementati on may l ag.
116
Some devel opi ng countri es have shown a si gni fi -
cant commi tment to protecti ng the envi ronment.
Chi nas most pol l uted coal -burni ng ci ti es have mean
sul fur di oxi de l evel s between 100 and 300 ug/m
3
,
about the same as i n Bri ti sh ci ti es pri or to the
i ntroducti on of Bri tai ns Cl ean Ai r Act i n the earl y
1950s. Chi na, wi th l evel s of i ncome much l ower
than the Uni ted Ki ngdom i n the earl y 1950s, has
al ready begun to respond to the probl em by i ntroduc-
i ng cl eaner urban fuel s and outfi tti ng coal -fi red
pl ants wi th effecti ve parti cul ate control s
117
.
Envi ronmental concerns i n devel opi ng countri es
have not been l i mi ted to domesti c i ssues. Devel op-
i ng countri es have al so parti ci pated i n i nternati onal
envi ronmental protecti on treati es. Most of the devel -
opi ng countri es bel ong to the Worl d Heri tage,
Endangered Speci es, and Law of the Sea conven-
ti ons. Ni ne devel opi ng countri es, i ncl udi ng Mexi co,
Egypt, Morocco, and Venezuel a, have al ready
si gned the Montreal protocol for protecti on of the
ozone l ayer.
Many devel opi ng countri es have al so experi enced
an i ncrease i n grassroots envi ronmental i sm i n addi -
ti on to thi s offi ci al acti vi ty. I n Ecuador, Col ombi a,
Guatemal a, and Brazi l movements organi zed by
ci ti zen groups have l ed to the banni ng of toxi c
pesti ci des and the i ni ti ati on of major urban cl ean-
ups.
118 Indigenous nongovernmental organi zati ons
(NGOs) have i ncreasi ngl y i nvol ved themsel ves i n
envi ronmental l y ori ented efforts. More than 600
Brazi l i an and I ndonesi an NGOs currentl y are work-
i ng on envi ronmental i ssues, as are several thousand
groups i n I ndi a.
119
113J efieyH.~~dandDavid Moreu, EmergenceofEnvironrnen~ ConcerninDeveloping Countries: Apolitical Perspective, StanfordJournul
of International Luw, Issue 2, Summex 1981.
l14B~~J . ~~che, world B- Environment and Natural Resources Management Institutions in Developing Countries, Mckground paper,
draft, Oct. 10, 1989.
llsBermu@ BW~ FZMO, Gaboq Ind~ Keny~ Omaq Papua New Guin% SenegiQ Singapore, Trinidad and Tobago, Venenek and *.
116The ~ew &ily avaage s~d~d for ~ient ~~ fioxide is stricter ixI each of the thr~ c~~s (m- ~llut~ SE@ lllb~ residential
locations, and industrial districts) tban the U.S. values; maximum nitrogen oxide levels permissible in residential areas are equivalent to the Japanese
standara and the maximum carbon monoxide levels allowed at any time am much lower than the l-hour averages in West Germany or the United States.
T.A. Siddiqi and C.X. Zhang, Ambient Air Quality Standard in Ch@ Environmental A4anugentent, vol. 8, 1984, pp. 473A79.
117Vaclav S@Environment in Developing COurltries,
contractor report prepared for the GffIce of Technology Assessment J une 1990.
118WorldR~o~c=~ti~~, DevelopmentandP ollution in the Third Worl& paper prepared for the 1bkyo Conference on the Global Mvironment
and Human Response, Sept. 11-14, 1989, J uly 1989, p. 33.
ll~id.
Despi te these achi evements, major probl ems re-
mai n. A fundamental constrai nt on envi ronmental
pol i cymaki ng i n al l but a handful of devel opi ng
countri es i s the absence of systemati c, i ntegrated
data on the current state of envi ronmental pol l uti on
and degradati on. Whi l e there i s someti mes a great
deal of si te-speci fi c i nformati on connected wi th
parti cul ar projects, vi rtual l y al l the devel opi ng
countri es l ack spati al l y representati ve networks for
moni tori ng l and, water, and ai r qual i ty or accuratel y
measuri ng the rates of deforestati on or changes i n
agri cul tural l and and soi l qual i ty. Chi na and I ndi a,
however, have accumul ated a fai rl y extensi ve i nfor-
mati on base through the combi ned acti vi ti es of
vari ous government departments and uni versi ty-
based researchers.
The i mpl ementati on and moni tori ng of envi ron-
mental pol i ci es and standards are al so di ffi cul t. I n
most countri es, the budgets of the envi ronmental
agenci es are a smal l fracti on of thei r mi ni mum
needs, and staffi ng i s i nadequate. Enforcement i s
often l ax; prosecuti ons are often costl y, uncertai n,
and sl ow; and the puni shments actual l y meted out
may be i nadequate deterrents.
120
Envi ronmental protecti on i s al so someti mes i m-
peded-i n both devel opi ng and i ndustri al i zed coun-
tri es-by government pol i ci es i ntended to promote
economi c growth, generate empl oyment, servi ce
forei gn debts, or meet other i mportant nati onal
needs. I n Brazi l , for exampl e, a vari ety of govern-
ment pol i ci es have encouraged rapi d devel opment
and thus rapi d deforestati on-of the Amazon. These
i ncl ude tax exempti ons, tax credi ts, subsi di es of
rural credi t, and l and acqui si ti on l aws (squatter
r i ghts).
121
These i nducements have i nadvertentl y en-
couraged a number of acti vi ti es that woul d not be
economi cal i n the absence of di rect or i ndi rect
government supports. For exampl e, the i mpl i ci t
government subsi dy for ranchi ng al one duri ng the
peri od 1975 to 1986 has been esti mated by the
Worl d Bank at more than $1 bi l l i on.
122
The Brazi l i an
government has recentl y begun to reverse many of
these pol i ci es as awareness of thei r costs and
envi ronmental i mpacts has grown.
Si mi l arl y, commerci al l oggi ng has al so been
encouraged by many governments through l ow
royal ti es and fees, reduced export taxes, and other
tax breaks for ti mber compani es and other domesti c
wood products i ndustri es.
123
Tax credi ts, i nvestment subsi di es, and other fi scal
supports may be necessary and justi fi ed to encour-
age needed devel opment, and such pol i ci es are
wi del y used i n devel opi ng and i ndustri al i zed coun-
tri es. The val ue of depl etabl e natural resources and
the costs of envi ronmental degradati on, however,
are often not properl y accounted for by these
fi nanci al i nstruments. For thi s reason, many econo-
mi sts advocate pol i ci es that more accuratel y i ncl ude
these costs than has someti mes been done i n the past.
Wi th proper val uati on of the natural resources,
envi ronmental i mpacts, and other costs and benefi ts,
such i nvestments mi ght be made more wi sel y.
124
Government pol i ci es i n some i ndustri al i zed na-
ti ons may al so i nadvertentl y affect the envi ronment
i n devel opi ng countri es. I n 1988, the European
Communi ty, the Uni ted States, and Japan provi ded
subsi di es to thei r agri cul tural sectors total i ng $97.5
bi l l i on, $67.2 bi l l i on, and $57.8 bi l l i on respecti vel y
a total of $222.5 bi l l i on.
125
Thi s i s nearl y one-thi rd of
Izocritiques of government enforcement of environmenttd laws in developing countries can be found in: R. Abmmsz and L. @tolano, 1987,
Environmental impact assessment in the Philippines: 1977-1985. Environmental Impact Assessment Review, vol. 7, pp. 293-310. C.M. Abraham and
A. Rosencranz, 1986, AnEvaluationof Pollution Control I@slationinIndi~ CoZumbia Journal ofEnvironmentaZ L.uw, vol. 11,1986, pp. 101-1 18;
H. Hae~ Conservation in Indonesia, AMBIO, vol. 17, 1988, pp. 218-222; A.D. Johns, Economic Development and Wildlife Conservation in
Brazilian Amazonia, AMBIO, vol. 17, 1988, pp. 302-306; J. May&, Environmental Legislation in Developing Countries: Some Parameters and
Constraints, Ecology Law Quarterly, vol. 12, 1985, pp. 997-1 (Y24; K. Ramalms hna, The Emergence of Environmental Law in the Developing
Countries, a Case Study of India, Eco20gy LawJ Quarterly, vol. 12, 1985, pp. 907-935; A.S. Tolentino, hgislative Response to Marine Threats in
the ASEAN Subregio& AMBIO, vol. 17, 1988, pp. 238-242.
lzl~e ~overment mspended provisions tit ~de C1
earing for ranches profitable and penalized owners of unimproved lands, but ranches are still
attractive as a speculative hedge against inflation. See U.S. Congress, Gfflce of Technology Assessmen4 op. cit., footnote 14; Climate Institute, Climate
News Around the Globe, CZimateAZert, vol. 2, No. 3, Fall 1989, p. 8; S.B. Hechg The Sacred Cow in the GnxmHell: Livestock and Forest Conversion
in the BrazilianAmazon, The Ecologist, vol. 19, No. 6, November/December 1989, pp. 229-234.
122~ p. BinswaWer, World B~, Fis~~ and ~g~ ~cent i ves wi t h Envi r Onmen~ Ef f ect s On t he B@f i hMZO~ A@cdhd ~d Rd
Development Department, Operational Policy Staff, May 1987; N. Myers, Deforestation Rates in Tropical Forests and Their Climatic Implications
(London: Friends of the Earth Limited, 1989).
123Robert Repetto, Deforestation in the Tropics, Scient#ic American, vol. 262, No. 4, April 1990, pp. 36-42.
12lRobertRepetto et~., WastingAssets:NaturalResources in theNationalIncomeAccounts (WaSh@tO~ DC: World Resources Institute, J~e 1989).
lzssfiw Auerbach, Statement Aimed at Farm Subsidies hnpaSSe,
Washington Post, July 12, 1990, p. A9.
the GDP of the 2.8 bi l l i on peopl e who l i ve i n
l ow-i ncome devel opi ng countri es, and i t i s compara-
bl e to the GDP of thei r enti re agri cul tural sector.
126
Some of these subsi di es may tend to hol d agri cul -
tural pri ces down i n devel opi ng countri es as wel l as
l i mi ti ng thei r export opportuni ti es.
127
I f thi s occurs
and reduces devel opi ng countri es earni ngs for thei r
agri cul tural output, they may have l ess i ncenti ve and
abi l i ty to i nvest i n agri cul tural research and devel op-
ment or i nfrastructure to achi eve hi gher yi el ds and
qual i ty. As noted above, l ow agri cul tural producti ve-
ti es (together wi th i nequi tabl e l and tenure, etc.)
contri bute to the expansi on of cropl ands i nto ecol -
ogi cal l y fragi l e areas. Si mi l arl y, i mport tari ffs to
protect domesti c ti mber i ndustri es i n i ndustri al
countri es may encourage i neffi ci ent harvesti ng i n
tropi cal forests because the ful l market val ue of the
resources cannot be obtai ned.
128
However wel l -i ntenti oned, the devel opment pol i -
ci es of bi l ateral devel opment agenci es and mul ti l at-
eral devel opment banks have al so someti mes di -
rectl y contri buted to envi ronmental degradati on due
to thei r emphasi s on l arge projects--transporta-
ti on i nfrastructures, hydroel ectri c faci l i ti es, and
i ndustri al compl exes.
129
Devel opment agenci es are,
however, becomi ng more sensi ti ve to these envi ron-
mental concerns.
130
The rol e of devel opment agen-
ci es i n promoti ng the adopti on of more envi ronmen-
tal l y sound energy technol ogi es wi l l be exami ned i n
a l ater report of thi s OTA study.
126worldB@ wor~DeYeJ op~~fRepo~ ~$lggI (w~~toq ~: 1989), indicator~bles 1 ~d q.Note ~t~s does not include the 1 billion people
in middle-income developing countries who had a 1987 GDP of nearly $2 trillion.
127Vmnon 0. Roningen and ~ven M. D~t, Econo~c I@ications ofAgriculturalpolicy R~o~s in Indusrnal Market Econo~es (wdlhlgtt)~
DC: U.S. Department of Agriculture, Economic Research Service, 1989).
128u.s. Conmess, Office of Technology Assessment Op. cit., footnote 14.
12~.s. Covess, ~lce of Tmhnology Assessmen4 op. cit., f~~ote 14; Bmce M. Rich The M~ti~ter~ Development Ban.ks, Environmental
Policy, and the United States, Ecology Law Quarterly, vol. 12, No. 4, 1985; P. Aufderheideand B. RiclL EnvironmentalR eform and the Multinational
Banks, World Policy Journal, Spring 1988, pp. 301-321; W. Reid, Sustainable Development: Lessons From Success, Environment, vol. 31, No.
4, May 1989, pp. 7-35,
130See, f
ore
-
p
i
e
, WorldBa~
{
AReviewof the Tr~~entofEnvironmenW As~ts of B@ Enmgy projects, ~d~~ and Energy Department
Working Paper, Energy Series Paper No. 24, March 1990.
Appendix A
Glossary of Energy Units
Appendix A
Glossary of Energy Units
British thermal unit (Btu)--The basi c uni t of energy i n
the Engl i sh system of uni ts. The energy requi red to
heat 1 pound of water 1 oF. One Btu i s about the
energy gi ven off by burni ng a si ngl e match ti p.
Hectare (ha)-An area of l and measuri ng 10,000 square
meters or equi val entl y 100 meters by 100 meters.
One hectare i s equal to about 2.5 acres.
Joul e (J)--A basi c uni t of energy i n the metri c system. A
joul e of energy i s very smal l (a si ngl e match ti p
gi ves off about 1,000 joul es), so joul es are often
mul ti pl i ed by orders often (10, 100, 1,000, etc.) to
yi el d useful uni ts:
Name Value Abbreviation Equivalent
kilojoule thousand ( 1 0
3
) joules kJ 0.95 Btu
megajoule million (10
6
) joules MJ
gigajoule billion (10
9
) joules GJ
terajoule 1 0
12
joules TJ
petajoule 1 0
15
joules PJ
exajoule 1 0
18
joules EJ 0.95 quads
Total energy production and consumption for a country is
typically given in exajoules; per capita energy consumption is
typically given in gigajoules (1 GJ is equivalent to the energy in
about 7.5 gallons of gasoline).
Kilogram (kg)--The basi c uni t of wei ght i n the metri c
system, equal to 2.2 pounds.
Kilowatt (kW)--The power (energy per uni t ti me) uni t i n
whi ch el ectri ci ty i s measured.
Kilowatthour (kWh)--The energy uni t i n whi ch el ec-
tri ci ty i s measured. A 100-watt l i ght bul b burni ng
for 10 hours consumes 100 W X 10 hours= 1,000
Wh = 1 kWh of energy.
Meter (m)--The basic unit of length in the metric system,
equal to 39.4 i nches.
Quad101
5
Btus. The Uni ted States currentl y consumes
about 80 quads (80X 10
15
Btus) of energy per year.
1 Quad i s equal to about 1.05 exajoul es (EJ).
Tonne (t)--Shor t for metric tonne, equal to 1,000
kilograms or about 2,200 pounds.
137

9118

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

9118

Загружено:

Авторское право:

Доступные форматы

Energy in Developing Countries

0.13 0.24 1.9

0.05 0.16 0.8 7.2

1.6 3.6 1.3 20.4

0.1 0.7 0.7 11.3

20 280 2,200 460 2,200 27,000

8.6 5.3 0.3

Not available or not applicable.

10.4 2.9 2.9 1.2 1.0 2.4

1.3 0.5 0.3 0.5

11.2 4.8 1.6 1.5 0.4 0.8 2.1

176 3,891 15,614

Not available or not applicable.

Energy supplies coal Oil Gas Nonthermal Thermal Wood

3.3 1.0 1.0

6.8 0.2 1.4

Вам также может понравиться