energy and environment scenarios for senegal

Natural Resources Forum 1994 18 (1) 3 1 4 7

Energy and environment scenarios for Senegal

Michael Lazarus, Souleymane Diallo and Youba Sokona

In this paper, the energy and environmental dimensions of several proposed energy strategies for Senegal are explored. An analytical framework to compare the energy and measurable environmental impacts of a set of scenarios is developed, and the limitations of the quantitative approach are discussed. It is found that policies to promote substitution of liquid petroleum gas (LPG) for charcoal use in households may actually reduce greenhouse gas emissions, while also improving more important near-term environmental problems. Substitution of LPG for charcoal would not necessarily lead to a significant increase in Senegal's oil import bill, since other petroleum product usage wi f f continue to dominate. Despite past industrial sector initiatives, considerable potential for energy eficiency investment remains, and presents additional opportunities for minimizing environmental impacts.

African nations face increasing environmental problems related to energy use and production. Land clearing for agriculture and energy requirements can lead to significant local environmental impacts and contribute to both local and global climate change. The rapid expansion of urban areas is changing energy use patterns, as more people enter the cash economy and commercial fuels begin to displace traditional biomass fuels. These energy transitions have significant impacts on air, water and land resources.

Opportunities currently exist to minimize the long- term economic and social costs of energy use and production. To this end, the Stockholm Environment Institute (SEI) and Environment and Development in the Third World (ENDA-TM) have initiated a project to build institutional capacity for integrated energy+nvi- ronment planning in Africa.' This paper presents the results of preliminary energy and environment scenarios for Senegal that comprise the initial phase of this project.

Michael Lazarus is with the Stockholm Environment Institute - Boston (SEI-B), at Tellus Institute, I 1 Arlington Street, Boston, MA 021 16-341 I , USA. Souleymane Diallo and Youba Sokona are with the Energy Programme of Environment and Development in the Third World (ENDA- TM), 54 rue Carnot, Dakar, Senegal.

The authors wish to thank the United Nations Environment Programme and the International Development Research Center for their financial support, and Libasse Ba, Dominique Revet and David von Hippel, for their assistance.

Energy and environment issues in Senegal Senegal faces issues that are unique to the Sahel region of West Africa: highly vulnerable semi-arid ecosystems, relatively poor endowments of energy resources and, as a result, high electricity prices. I t also confronts two energy issues which are common to most African countries: heavy dependence upon imported petroleum and high levels of biomass energy use with a declining wood resource base.

Together, oil (0.80 million tonnes oil equivalent), wood (0.83 Mtoe) and other biomass (0.15 Mtoe) accounted for 98% of Senegal's 1988 primary energy supply [ I I ] . Between 1980 and 1988 oil imports accounted for 17% to 28% of total imports, and from 26% to 59% of total non-energy export receipts. On a per capita basis, Senegal's annual primary commercial energy consumption of about 5 GJ (0.1 toe) is less than one-tenth the world average of 57 GJ (1.4 toe). However, it is slightly higher than the comparable figure for most non-oil exporting sub-Saharan African countries [ 33 1.

Comparing across sectors in Figure 1 , transport accounted for the largest share of 1988 non-biomass energy consumption, followed by industry (over half as fuel oil) and households (half electricity, half LPG and kerosene). Because Senegal is a major hub for West African air traffic, air transport requirements account for ' Cofunded by the United Nations Environment Programme, the International Development Research Centre, SEI and ENDA-TM, the project involves the participation of the Senegalese Ministry of Industry and Artisans and the Ministry of Economy, Finance and Planning.

0165-0203/94/01003 1- 17 0 1994 United Nations (text) and Butterworth-Heinemann Ltd (design and typography) 3 1

Energy and environment in Senegal: M . Lazai~us et a1

increasingly precarious, particularly around Dakar [24]. Other Industry Indoor air pollution from household use of biomass

fuels may also pose serious health risks. Proposed Fisheries hydroelectric development presents potential health

risks and threatens traditional agricultural production systems in flooded and downstream areas, as well as dis- rupting areas in the path of proposed long distance, high voltage transmission lines.

Furthermore, with considerable land area near sea level and a dry Sahelian climate, Senegal is highly vulnerable to the potential effects of global climate change. Saltwater intrusion into aquifers is already a problem in many areas. A significant rise in sea level could result in severe damage to many key areas, including the biologi- cally rich island ecosystems of the Saloum region,

2 % 12%

5 y o

Household 55%

Transport 26%

Figure 1. Final energy consumption by sector, 1988

over 40% of transport sector energy consumption and the dominance of the transport sector in the national energy balance. Fisheries, an important source of foreign exchange earnings, account for almost 10% of total commercial energy consumption. However, when biomass fuels - firewood, charcoal and agricultural residues - are included, households account for over half of total energy use. Charcoal, used largely for cooking, domi- nates urban household fuel consumption, while firewood accounts for almost all rural energy use.

One recent trend is worth emphasizing. Households consume over 90% of LPG in Senegal, and LPG use has grown very rapidly in recent years. While LPG consumption grew only 7% from 1985 to 1987, in July 1987, increased subsidies resulted in a 30% decline in LPG prices for small consumers. As illustrated in Figure 2, from 1987 to 1988, LPG consumption grew by nearly 50%. Meanwhile, overall consumption of petroleum products stayed relatively constant from 1985 to 1988; LPG still only accounts for 3% of the total.

Senegal’s most widespread and immediately apparent energy related environmental problems are the localized scarcity of biomass fuels and the land degradation resulting from agricultural expansion and the charcoal trade. Indeed, environmental discussions in Senegal are dominated by concern for the declining forest cover, soil erosion, local climate changes, and interactions with the agricultural and pastoral activities of rural people. Other environmental issues have only begun to enter into energy planning discussions. Some observers note that ‘with regard to the questions of energy efficiency and of environmental impact, they are I ) raised only by energy users, and 2 ) for the time being, relatively, not to say totally, absent from the debate’ ([ 131, p 13). At the same time, the Senegal government’s report to UNCED notes that uncontrolled industrialization and various pollutants (eg auto emissions) are rendering living conditions

important coastal tourist areas and low lying rice fields in Senegal’s richest farming area, the Casamance. On a per capita basis, Senegal emits approximately one-quar- ter the world average greenhouse gas emissions from all sources: its per capita fossil fuel CO, emissions are one- tenth the global average [28]. As a country of about 7 million inhabitants, Senegal’s contribution to the global increase in heat trapping gases would thus appear very small indeed. None the less, several political and economic factors suggest the importance of a closer scruti- ny of greenhouse gas emissions and the options for reducing them.

An integrated approach Three basic principles underlie integrated energy+mvi- ronment analysis. First, the analysis considers all fuels and technologies - both supply side and demand side - on an equal footing. Second, the ultimate goal is the provi- sion of end-use services and amenities (hot water, lighting, transport etc) rather than fuels and electricity alone, at the lowest social cost. Finally, the analysis incorporates the economic externalities - most notably environmental and equity impacts - absent from a traditional cost-benefit analysis based on market prices alone. The necessary longer-term planning horizon (eg longer than 10 years), stretches beyond, but is not isolated from, the short-term issues that often dominate the planning realm.

A recent study for the USA, America’s Energy Choices, provides a concrete example of a thorough energy-environment analysis along these lines [30]. For the US study, the Long-Range Energy Alternative Planning (LEAP) system provided the organizing analytical framework. Together with its associated environmental database (EDB), the LEAP system also provided the analytical tool for creating our energy and emission scenarios for

’Together. LEAP/EDB comprise a computerized modelling system designed to explore alternative energy futures, along with their princi- pal environniental impacts. As a flexible, model building tool. model

Nutural Resources Forum 1994 Volume I8 Nuniher- I 32

Energy and environment in Senegal: M . Lazarus et al

30

I987 Price Subsidy 20

Thousand TOE

1975 1976 1977 1978 1979 1980

Figure 2. Household LPG consumption trends.

Given the complexities involved in full assessment of environmental impacts across a wide range of energy activities, the focus of this paper is limited to a few scenarios, representing a handful of energy policies and a limited set of emission and impact categories. Data availability and uncertainty dictated the design of a simple model of Senegal’s energy system. Recent surveys by ENDA [6] and the World Bank [31] enabled a more disaggregated model for the household sector (ie by end use) than for other energy consuming sectors. LEAP was used to model the operation of each of Senegal’s major energy extraction and transformation (electricity, charcoal and refinery) industries.

Using this model, a reference case was prepared, which provides the background for the analysis of policy scenarios over a time horizon of 1988 to 2005 (Table 1). Assumptions regarding demographic and economic growth were drawn from indicative estimates developed by a team of researchers and managers con- tributing to the government’s Plan d’orientation 1989- 1995. These figures reflect continued rapid urbanization and growth rates for individual economic subsectors that range from over 3% per year for fishing, services and other industries, to 2% per year for extractive industries (phosphates and salt), to no growth in the troubled veg- etable oil industry. The saturation of household electri-

(Contittued} relationships and detail can be tailored to the local dynamics and data constraints of individual applications. EDB con- tains a large existing database of coefficients describing air, water, solid waste, and occupational health and safety effects, derived from 70 literature sources of international origin. LEAP has been developed by SEI and the United Nations Environment Programme. For a description of a early version of the LEAP system, see Ref. 23.

1981 I982 1983 1984 1985 1986 1987 1988

cal appliances was assumed to nearly double during the 17 year time period considered here. Older oil fired electric capacity is gradually replaced by new higher efficiency diesel base load and peaking stations, according to the reference plan of the state electric utility (SEN- ELEC). The reference case also assumes continued operation of the national refinery (with its questionable cost-effectiveness) and no additional penetration of improved technologies (ie static or ‘frozen’ efficiency).

Selected energy results of the reference case are shown in Figures 3 and 4. From 1988 to 2005 total final energy consumption grows steadily at a pace of 3.0% per year, while primary energy increases somewhat more slowly at 2.6% per year. This difference is largely due to the improvement of electric sector thermal efficiency from a system average of 27% to 38% as the inefficient older generating units are replaced. As the result of increased urbanization, electricity, charcoal and LPG consumption grow more rapidly (3.5% to 3.6% per year) than firewood (2.6% per year). From 1988 to 2005, overall oil imports increase by 56% and wood requirements for charcoal production increase by 80% to over 2.3 million cubic metres per year. Not surprisingly, the reference case presents continued burdens on both national accounts and natural resources.

Against the backdrop of this reference case, four preliminary policy scenarios are presented. Each scenario builds upon its predecessor, with the order based on the judgement of study participants regarding the cost and likelihood of implementation. Scenario A, butanization, considers the continued promotion of an urban transition from charcoal to LPG (butane) cooking. Scenario B,

Nurural Resources Forum 1994 Volume 18 Number I 33

w

P

Tabl

e 1.

Env

iron

men

tal a

nd o

ther

ext

erna

litie

s as

soci

ated

wit

h Se

nega

l sce

nario

s.

Nat

ural

ec

osys

tem

s

Indo

or a

ir po

llutio

n

Loca

l air

pollu

tion

Aci

d pr

ecip

itatio

n

Glo

bal

war

min

g

Sce

nari

o C

:

use

effic

ienc

y

Issu

e In

dica

tors

(onl

y ite

ms

in

Ref

eren

ce ca

se

Sce

nari

o A

: S

cena

rio

B ita

lics

anal

ysed

her

e)

buta

niza

tion

impr

oved

im

prov

ed e

nd-

biom

ass

effic

ienc

y

Impa

cts

rela

tive

to r

efer

ence

case

(nu

mbe

rs s

how

diff

eren

ces

in 2

005)

C

urnu

loriv

e lu

nd c

leur

ed (1

988-

420

000

to 1

200

000

hec

tare

s cl

eare

d 34

% re

duct

ion

in c

umul

ativ

e la

nd 3

7% re

duct

ion

in c

umul

ativ

e la

nd c

lear

ine

Land

deg

rada

tion

2005

) for

chur

coul

pro

duct

ion

Soil

eros

ion

Bio

dive

rsity

habi

tat l

oss

Hou

seho

ld e

mis

sion

s of

C

O. N

or, H

C, S

O,,.

TSP

Inte

rior a

rchi

tect

ure/

vent

ilaio

n

Em

issi

ons of

nitro

gen

o.vi

des.

ozon

e, H

C. s

ulph

ur d

io.\-

ide,

ui

r to.r

ics

Sul

phur

dio

xide

N

itrog

en o

.tide

s

Gre

enho

use

gus

emis

sion

s:

curh

on d

io.ri

de (C

O,)

.. fo

r cha

rcoa

l pro

duct

ion

1988

-200

5 cl

earin

g

Con

tinue

d ha

bita

t los

s an

d re

duct

ion

in b

iodi

vers

ity li

kely

as c

harc

oal

dem

and

grow

s an

d la

nd c

lear

ing

cont

inue

s

No

maj

or c

hang

es

Poss

ible

bio

dive

rsity

and

hab

itat b

enef

its d

ue to

redu

ced

land

cle

arin

g

Hou

seho

ld se

ctor

em

issi

ons d

own

42

4%

(dep

endi

ng on

pol

luta

nt)

Mos

t em

issi

ons

up 6

C-70

%,

2005

, exc

ept S

O,,

NO

, 19

88-

Emiss

ions

dow

n 10

% (N

o,),

30%

A

ppro

xim

atel

y sa

me

Emis

sion

s dow

n 20

%

(CO

), 60

% (T

SP, H

C)

as A

(P

b, S

O,,

NO

,), 3

0%

(CO

), 60

% (T

SP, H

C)

SO,,

dow

n 20

%, N

O, u

p 14

0%

1988

-200

5, d

ue to

cha

nges

in

elec

tric f

uels

See a

bove

Tota

l GH

G e

mis

sion

s up

60%

by 2

005

(GW

P, 1

992

IPC

C 1

00 y

ear

as A

30

%

GH

G e

mis

sion

s dow

n 20

%

App

roxi

mat

ely

sam

e G

HG

em

issi

ons d

own

wla

tile

hyd

rocu

rbok

(C

H,.

HC

) in

tegr

atio

n va

lues

) or

her G

HG

s (N

20. C

O, e

re)

-.

5 W

ater

use

and

BO

D, C

OD

, sus

pend

ed so

lids

5 po

llutio

n ch

emic

als,

toxi

cs, t

herm

al d

is-

char

ges,

silt,

wat

er a

vaila

bilit

y a

2

Emis

sion

s and

impa

cts h

ighl

y si

te sp

ecifi

c, n

ot e

stim

ated

Sce

nari

o D

: hy

dro

deve

lopm

ent

Sam

e as

B.

In a

dditi

on, F

elou

dam

will

H

ood

area

s in

nei

ghbo

urin

g M

ali;

dow

nstre

am a

reas

in

Sene

gal s

ubje

ct to

alte

red

How

and

silta

tion

patte

rns

from

hyd

ro o

pera

tions

Po

ssib

le b

enef

its d

ue to

re

duce

d la

nd c

lear

ing;

po

tent

ial h

arm

to fi

sh p

opu-

la

tions

from

hyd

ro op

erat

ions

App

roxi

mat

ely

sam

e as A

Emis

sion

s do

wn

20%

(Pb)

, 30

% (S

O,.

CO

), 60

% (T

SP,

NO

,, H

C)

GH

G e

mis

sion

s dow

n 40

%

(6%

low

er th

an 1

988)

Sene

gal R

iver

aff

ecte

d by

op

erat

ion

of h

ydro

faci

litie

s

6

cont

inue

d on p

age

35

2 21

L 1 f 4

C

iD

a - 7

E'

3 4

4

&

\o

\o

4

Tab

le 1

cont

inue

d.

3 G Is

sue

2 c co In

dica

tors

(onl

y ite

ms

in

italic

s ana

lyse

d he

re)

Ref

eren

ce c

ase

c

men

tal i

ssue

s not

co

nsid

ered

Empl

oym

ent

Econ

omic

de

velo

pmen

t

Rur

al w

elfa

re

Agr

icul

tura

l pr

oduc

tivity

Solid

and

toxi

c w

aste

s, oz

one

depl

etio

n et

c

Jobs

crea

tedn

ost

GD

P N

atur

al r

esou

rce

acco

unts

Effe

cts o

n lo

cal f

arm

ing

syst

ems

and

subs

iste

nce

activ

ities

Lo

cal f

irew

ood

avai

labi

lity

Soil

and

nutri

ent l

oss

Mic

mlim

ate

chan

ges

Con

tinue

d de

grad

atio

n w

ith o

n go

ing

char

coal

mak

ing

activ

ities

If su

gges

tions

that

red

uced

bio

mas

s co

ver l

eads

to d

etrim

enta

l mic

re

clim

ate c

hang

es (w

ind

eros

ion,

m

oist

ure

loss

) are

cor

rect

, then

co

ntin

ued l

and

clea

ring

coul

d de

grad

e fa

nnin

g co

nditi

ons

Scen

ario

D:

Scen

ario

A:

Scen

ario

B

Scen

ario

C:

buta

niza

tion

impr

oved

im

prov

ed e

nd-

hydr

o de

velo

pmen

t bi

omas

s us

e ef

ficie

ncy

effic

ienc

y St

udie

s ind

icat

e po

tent

ial

incr

ease

in w

ater

-bor

ne

dise

ases

Shor

t-ter

m: M

ore j

obs c

ould

be

lost

in c

harc

oal i

ndus

try th

an g

aine

d th

roug

h in

crea

sed

LPG

supp

ly a

nd e

quip

men

t. Po

sitiv

e im

pact

s on

othe

r rur

al a

ctiv

ities

cou

ld o

ffse

t oth

er lo

sses

. Im

pact

s of o

ther

pol

icie

s (eg

impr

oved

faci

litie

s),

diff

icul

t to j

udge

Mac

roec

onom

ic ef

fect

s of s

ubsi

dy

App

roxi

mat

ely

Sam

e as

B, w

ith

Incr

ease

d de

bt fr

om ex

pen-

lik

ely

to b

e ne

gativ

e, b

ut m

ay b

e si

ve h

ydro

proj

ect a

nd tr

ans-

of

fset

in p

art b

y en

viro

nmen

tal

mis

sion

line

mus

t be

com

be

nefit

s re

duct

ion

in o

il pa

red

with

add

ition

al 2

0%

redu

ctio

n in

tota

l oil imports

Like

ly b

enef

its, s

ince

char

coal

So

me a

dditi

onal

Sa

me a

s B

Sa

me a

s B, e

xcep

t pot

entia

l in

dust

ry ca

n be

dis

rupt

ive t

o ru

ral

disr

uptio

n to

loca

l com

mun

- ac

tiviti

es a

nd w

ell b

eing

(ben

efits

iti

es in

Sen

egal

Riv

er B

asin

, m

ostly

to u

rban

are

as)

due t

o up

stre

am h

ydro

pro

ject

s R

educ

ed ch

arco

al h

arve

stin

g co

uld

Som

e add

ition

al

Sam

e as

B Sa

me b

enef

its as

B, b

ut p

oten

tial

thus

lead

to im

prov

ed ag

ricul

tura

l ha

rm to

agr

icul

ture

in S

eneg

al

Riv

er B

asin

. (A

void

ed n

utrie

nt

prod

uctiv

ity

depo

sitio

n)

sam

e as A

m

acro

econ

omic

be

nefit

s of 2

0%

impo

m

impr

ovem

ents

, re

lativ

e to

A

impr

ovem

ents

re

lativ

e to

A

-. 3

Energy and environment in Senegal: M. Laiaru.s et a1

I800

1600

1400

I200

1000 Thousand

TOE

800

600

400

200

0

I988 1990 I995 2005 Figure 3. Final energy consumption: reference case." LPG

Period 1988-90 not to scale.

Steam

I

2500

2000

1500 Thousand TOE

I000

500

0

1988 1990 1995

Figure 4. it Period 1988-90 not to scale.

Primary energy 1988-2005: reference case."

2000 2005

36

Energy and environment in Senegal: M. Luzur.us et al

convenient four-burner stoves that use the 12kg and larger containers. The relatively high availability of the subsidized small (2.7 kg) containers and associated stoves helps poorer urban households overcome the obstacle of ‘lumpy’ payments often cited as a major obstacle to the purchase of gas and electricity. Poor households typically buy small amounts of charcoal or wood on a daily basis [IS]. At only CFA325 per 2.7 kg bottle (US$1.2 at CFA270 = US$I), LPG becomes a manageable expense to many poorer urban households, who may none the less still purchase charcoal (at approximately CFA65 or US$0.24 per kilogram, or US$O.18 per typical meal, see Table 4), when cash is in particularly short supply.

Recent urban household energy surveys by Tibesar and White 1291, the World Bank [31] and ENDA [6] provide somewhat conflicting analyses and interpreta- tions of subsidy impact and future policy prospects in Senegal. While Tibesar and White suggested in 1985 that relying on price to reduce much charcoal consumption will not succeed, LPG use and appliance ownership have increased substantially among urban households since the government reintroduced subsidies in 1987. In contrast, the more recent 1989 ENDA survey suggests that fuel consumption is indeed price elastic, as we might expect.

Due to large uncertainties in actual (rather than reported) charcoal consumption, i t is difficult to establish precisely to what extent this increased LPG use has replaced charcoal consumption, but survey analysts estimate that each additional kilogram of LPG consumption displaces from 1.6 to 2 kg of charcoal.

Future use of energy for cooking is projected on the basis of current observed patterns of mixed fuel usage to reflect the actual experience of households that have switched to LPG, rather than assumption of a strict ratio for the substitution of charcoal by LPG (either the above ratio or a 1 : I useful energy basis). Recent urban surveys have grouped households into categories of increasing transition toward cleaner, more convenient cooking fuels. Based on these, i t is projected that in the major urban areas that already use substantial levels of LPG (Dakar, Thies, St Louis), mean household fuel use in 2000 will approach the relatively equal levels of charcoal and LPG use found in the largest grouping (46%) of the ENDA survey, with LPG dominating afterwards. Although such a transition may appear optimistic, comparison of the recent survey results suggests that such rapid changes are not ~nprecedented.~ In other urban areas, a slower transition is assumed.

Although this approach has certain advantages, it potentially obscures the implicit differences in income and lifestyle among groups in different fuel categories. It could be that households currently using gas with charcoal have different cooking and eating patterns than households using charcoal alone, as a function of disposable income and status for which changes are obviously much harder to implement than a butanization policy.

improved biomass efficiency, looks at improved stoves and kilns, coupled with the butanization strategy of Scenario A. Scenario C adds improved commercial energy efficiency for electricity and fossil fuel end-uses. Scenario D, hydro, assumes the completion of plans to develop hydroelectric resources jointly with neighbour- ing countries.

Scenario A : butanization

The promotion of modern fuels to substitute for traditional fuels in household end uses such as cooking has been a common strategy among many developing countries, most commonly implemented in the form of fuel and/or equipment subsidies. These subsidies have been justified as both an environmental benefit (avoiding deforestation) and a basic needs support for lower income groups. However, poorly designed subsidies achieve neither ~ b j e c t i v e . ~

Since the mid- 1970s the Senegalese government has used subsidies and promotion campaigns to encourage the transition from charcoal to LPG. In 1974, the government began a programme of equipment subsidies, but switched to fuel subsidies in 1977. These early efforts appear to have benefited middle class rather than poor households, with little impact on overall charcoal demand [29]. One survey [29] conducted in the mid- 1980s even suggested that demand for charcoal is inelastic with respect to the relative prices of charcoal and LPG. It thereby questioned the efficacy of either an LPG subsidy or a charcoal tax.

More recent LPG subsidies and promotions appear to be more successful. Over 60% of all households in five major cities now own LPG stoves [14]. In a 1989 survey [7], LPG was the primary household fuel in 47% of Dakar households, compared with 24% in a 1987 survey [31]. Over 75% of Dakar households now own stoves designed for the subsidized, smaller 2.7 and 6 kg LPG bottles, while only 14% own the unsubsidized but more

For example, Pitt [21] found that kerosene subsidies in Indonesia in the 1970s disproportionately benefited wealthier urban households and through econometric analysis, he ‘conclusively rejects the deforestation argument’, by showing that the cross-price elasticity of firewood demand with respect to kerosene price is very small (p 215). The very costly and broad kerosene subsidy was available to all consumers, regardless of income or location; in any case, kerosene is highly transportable and tradable, rendering fuel subsidies difficult to target. At the same time, low cost kerosene could stimulate additional lighting use rather than substitution for wood for cooking, achieving an additional social policy objective of educational and well being benefits. Fuel subsidies have also been challenged because they can pose a disincentive to energy efficiency and can inhibit economic growth [ 161. In general, a more economically ‘rational’ approach to addressing the environmental externalities of wood fuel related land degradation would be a tax on wood fuel production or consumption. Such an approach has been suggested for Senegal in the form of increased fees on wood harvesting by charcoal makers 1311; however, difficulties in administering and enforcing such fees present formida- ble obstacles.

Natural Resources Forum 1994 Volume 18 Number I 37

Energy and environment in Senegal: M . Lazai~us et al

I t is unclear whether subsidies must continue for this transition to occur. Increasing the charcoal price could achieve the same goal, but perhaps at greater hardship to poorer urban dwellers and with significant political obstacles hindering implementation. Improving the reli- ability of LPG supply might increase LPG use at a lower marginal cost to the government than subsidies alone. In addition, further reduction of LPG equipment costs should be investigated. Rapid household switching to LPG has taken place in a number of developing countries during the 1980s [ 181. While declining LPG prices relative to charcoal and rising household incomes were also important factors, consumer surveys cited the decrease in LPG equipment prices as more important.

Scenario B: improved biomass eficiency

Many efforts to develop and disseminate improved wood and charcoal stoves and kilns in Senegal have met with limited success [9, 25, 311. Therefore, conservative estimates of achievable savings potential have been developed.’ The shift away from charcoal to LPG further reduces these savings since fewer old charcoal stoves and kilns exist to replace. But this does not imply a dismissal of stove programme effectiveness. Several countries, such as Kenya and Tanzania, have implemented effective stove programmes, and a project is now under way to adapt a Kenyan jiko stove to Senegalese conditions.

Scenario C: impimwd electricity and fossil fuel end-use eflciency

This scenario reflects the potential for cost-effective demand-side, efficiency improvements to reduce both energy costs and environmental impacts. Senegal is the site of one of the most intensive industrial sector efficiency projects in Africa to date. Industrial establish- ments comprising over 60% of industrial sector energy use have been audited, indicating a potential for cost- effective energy savings of 14%. Companies have implemented, or plan to implement, many of the low initial cost measures; however, most companies have yet to undertake higher cost options which require financing

These assumptions include a 15% penetration of charcoal stoves that save, on average, 40%. leading to a savings of only 6% by 2005; and a 20% penetration of improved charcoal making, leading to funher savings of about 6%. ‘The World Bank helped to set up a line of credit available to Senegalese industries for the capital investment needed for major equipment purchases. Few companies applied and some observers have suggested that the credit mechanism was flawed, since companies were wary of the required government involvement in screening applications [ZO]. The Senegal Bureau des Economies d’Energie has sought alternative means of assisting conservation investment financing.

for new equipment purchases [20]. While opinions vary on the obstacles to additional improvement, the economic potential remains, and new financing or incentive approaches may be required to achieve them.6 It is assumed that between 1988 and 1995, overall industrial end-use efficiency will improve by lo%, with 1% improvement per year thereafter.

For other sectors, because no similar extensive efficiency studies exist, evidence from studies in other countries was used. For instance, in the service sector, where the government predicts over 3% GDP growth per year, commercial building studies in other tropical countries have shown substantial potential for savings, particularly with improved lighting and cooling technologies. Most African countries have not yet implemented equipment standards, building codes, shared savings programmes and other measures increasingly common in industrialized countries. With aggressive programmes aimed at all major service sector end uses, it is estimated that energy efficiency could improve at an annual rate of 2.5% [ 171.

For transport, the major oil consuming sector of the economy, there are many opportunities for improving efficiency, from ‘feebates’ (using taxes on gas guzzlers to subsidize efficient vehicles) to mandatory maintenance of older vehicles. Given the lack of local data on transportation energy-use patterns, rough estimates are used, based on assumptions for efficiency improvement potential for all developing countries [ 171. Equipment standards and reduced import duties for high-efficiency household appliances could also reduce the 6% growth in household electricity use which, as projected in the reference case, would increase the household share of total electricity use from 20% to 29% by 2005.

Scenario D: hydroelectric additions

This scenario assumes the successful completion of projects to import hydropower from the first projects of the Senegal River Development Organization: the Manantali and Felou dams in Mali. Although Senegal, Mauritania and Mali have already jointly constructed the Manantali dam, political obstacles have prevented them from reaching a full cost and capacity sharing agree- ment, installing over 200MW of turbines, and building the required long-distance transmission lines. Since the recent reopening of the Senegal-Mauritania border, intercountry negotiations appear close to resolution, and there are strong indications that the hydroelectric potential will soon be tapped. However, the high cost of the long high-voltage transmission line to the main load centres in coastal regions must be balanced against the costs and benefits of alternative strategies. In 1989 a SENELEC analysis indicated that a purely thermal expansion of the system would probably result in lower

38 Nanrr-ul Resour-crs Forum 1994 Volume 18 Number- I

Energy and environment in Senegal: M . Lazarus et ul

~

3000

2500

2000

TRANSPORT

F D Air. Water. Soil

SCALE ISSUES

Indoor/On-Site

Local

Regional

I500 Thousand TOE

1000

500

0

Reference Case Scenario A Scenario B Scenario C Scenario D

Figure 5 Primary energy, 2005.

CAUSES

Infrastructure

Economic

Agricultural Policies

Financial Constrsints

POVORV

Land Tenure I

-- ENVIRONMENTAL

EMISSION/INSULT

Facility Construction Air &Water Emissions

Fuel Extraction Solid Waste

Refining& Conversion

Fuel Combustion Poplit ion Displacement

Land Clearing

Figure 6 Pathways from energy to environmental impacts.

overall costs. In addition, improved end-use efficiency would delay the need for either new thermal or hydro capacity, yielding even lower overall costs. For this scenario, the assumption is made that hydro capacity will come on line as planned, with Senegal’s capacity share at 45%, with 90MW available in 1996 (Manantali) and 47 MW in 1998 (Felou).

Projected energy requirements As shown in Figure 5, each successive scenario reduces future national primary energy requirements. The results are most marked for scenarios A and C. Relative to the reference case in 2005, the butanization scenario (A) reduces the wood requirements for charcoal production by almost 80%, while increasing LPG consumption by 115%. Relative to 1988 levels, wood requirements for charcoal production decline by almost 70%, from over 900000 tonnes in 1988 to 300000 tonnes in 2005. Urban firewood use declines from about 17 000 tonnes in 1988 to less than 2000 tonnes in 2005. Total national

Hydro

by Firewood

w Wood for Charcoal

0 Oil Imports

G1 Vegetal Waste

Vapor

RESPONSE/

DAMAGE

Ecosystem

Human Health

Aesthetics

Land Depradation

Biodiversity

Economic Harm

oil imports increase by less than 5% (from 1.25 to 1.31 Mtoe) relative to the reference case in 2005. Significant reductions in total primary energy requirements are achieved (from 2.83 to 2.34Mtoe) as the substantial losses due to inefficient charcoal production and use are greatly reduced.

Scenario B, with more efficient charcoal kilns and stoves, could also reduce these losses. However, two factors, the conservative penetration rates and decreased savings from the switch to LPG, cause scenario B to produce comparatively small energy savings.

With the achievement of improved efficiency poten- tials in most demand sectors, combined with butanization, by 2005, scenario C demonstrates significant decreases in final consumption of nearly every fuel compared with the reference case: electricity (19%), all petroleum products including LPG (8%), charcoal (80%) and firewood (4%). When translated into primary energy, the reduction in fuel requirements for power generation leads to a total reduction in oil requirements of 12% relative to the reference case in 2005. This more


Energy and environment in Senegal: M . Luzarus et ul

than offsets the increased petroleum product consumption of increased butanization. Oil consumption for 2005 decreases by 31% in scenario D; new hydroelectric capacity displaces 130MW of oil fired generation and provides 78% of total electricity generation. Although still preliminary, these results indicate that potential savings in both oil and charcoal consumption could accrue from a combined strategy of butanization, improved efficiency and hydro development.

Environmental analysis The scientific literature abounds with approaches to analysing the social, health and ecological effects of various energy technologies [3, 10, 321. Figure 6 represents many of the dimensions of these complex and interrelated issues. The focus here is on the link between energy activities and environmental emissions. However, other relevant factors cannot be overlooked. Underlying dynamics and causal factors related to development and socioeconomic relations can either foster or undermine otherwise well intended policies. The transport of emissions can lead to transboundary problems (such as acid precipitation and global warming) that cannot be resolved solely on a national level. There are also broader issues of the distribution of benefits and damages. For example, wood fuel consumption casts an ‘urban shadow’ over rural areas that bear the brunt of the impacts, while these new areas benefit mini- mally from loss of wood resources.

Comparative quantitative environmental analysis presents two considerable challenges: the consideration of difficult to quantify impacts and the comparison of seemingly incommensurate impacts (eg habitat loss versus human health impacts). In addition, there are considerable uncertainties in identifying and generaliz- ing relationships between emissions, impacts and actual resulting damages. In particular, for energy sources with extensive land-use impacts, such as wood and hydro, site specific factors influence the outcome so significantly that generalized models are nearly impossible. The danger of biasing energy choices towards those options whose environmental impacts are most difficult to assess or quantify - ‘confusing the countable with the things that count’ - must be avoided.

At the same time, increasing quantification, even monetization, of environmental externalities is proceed- ing at a rapid pace. In the USA 29 states now incorpo- rate environmental externality costs in electric sector planning 1191. Internalizing environmental costs has been termed ‘the wave of the future’, with 85 pollution taxes already in place in 1989 in OECD countries ([19], p 190). So-called market based initiatives are rapidly spreading worldwide, although in Africa they remain relatively rare. Forestry levies targeted toward reducing

negative environmental impacts of wood fuel harvesting are a possible exception, although their efficacy is questionable because of enforcement difficulties.

While simplifying the complex web of human and environmental interactions, an approach has been sought that local planners can implement within local institutional and data constraints. As an initial step, a limited set of land clearing and emission factor estimates was developed, covering all major energy production, conversion and consumption processes in Senegal. Most emission factors were derived from the environmental database (EDB). Since few emission tests or measurements have been conducted in Africa, OECD data were used for closely related technologies. As described below, for transport and household sectors, emission factors were used from Asia, which shares with Africa the characteristics of an older, less maintained vehicle stock and the use of small household stoves fired by traditional biomass fuels.

Given the paucity of available data, coefficients for items such as soil erosion, direct health and safety, solid waste, and water effluent emissions from energy processes were not included in this analysis. The analysis was limited to the following 10 air emission categories: CO, from fossil fuels, net biogenic CO, from biomass fuels, carbon monoxide (CO), methane (CH,), other volatile hydrocarbons (HC or VOCs), nitrogen oxide (NOI), nitrous oxide (N,O), sulphur oxides (SOr), total suspended particulates (TSP) and lead (Pb). These pollutants can contribute to global climate change (CO,, CH,, N,O, CO, HC and NO^,), local air quality (NOK, SOx, HC, Pb, TSP), indoor air quality (NO-,., SO,r, HC, TSP), and acid precipitation (NO,l., SO,). Thus, they provide indicators for four potentially important environmental issues.

The uncertainties associated with emission factors vary greatly among pollutants. Estimates of fossil fuel CO, emission factors primarily depend on fuel carbon content, and thus estimates usually vary relatively little, particularly for fossil fuels. Other emission factors are typically based on the results of a relatively small number of tests of fuel use by particular types of equipment. Judicious assumptions were required to apply these factors to Senegalese conditions. Increased emissions testing in both the developing and industrialized countries, with centralized international reporting of results, would help greatly in reducing uncertainty.

The extensive use of traditional cooking fuels and equipment may pose considerable health risks, particularly to women. For example, very high concentrations of carbon monoxide (940ppm) have been reported in households in Lagos, Nigeria [I]. Lacking emission measurements for the commonly used Senegalese stoves under local cooking conditions, average emissions were estimated based on an average of available test data for

40 Natural Resources Forum 1994 Volume I8 Number I

Energy and environment in Senegal: M. Lazarus et a1

Biomass energy use, land degradation and net CO, emissions The role of biomass fuel consumption in land degradation, rural welfare and greenhouse gas emissions remains highly site specific and difficult to quantify. There are even questions as to whether biomass energy use poses serious environmental consequences in Senegal and other heavily wood fuel dependent developing countries. While many studies and observers are often quick to say yes, the answer is not necessarily evi- dent. For many African countries, simple calculations made in the 1970s and early 1980s showed that the demand for firewood and charcoal exceeded estimated annual growth and wood stocks, and would inevitably lead to permanent deforestation of large areas, to short- falls or gaps in supply, and eventually to the total harvest of available wood. The spectre of massive deforestation and resulting desertification of semi-arid regions was invoked, and subsequently challenged [ 151.

In most cases, it has become increasingly clear that primary direct causes of land degradation are the expansion of cultivated areas, overgrazing, and, in some areas, logging for commercial wood exports. Other underlying factors include land tenure, expanding populations and low agricultural productivity. In many cases, the increasing and marginalized populations gather wood fuel supplies from land cleared for agriculture and graz- ing. Does continuing reliance on traditional biomass resources lead to land degradation? Where would cooking and heating fuels come from if such land clearing did not occur? And how do these dynamics change with expanding population densities? Globally, i t has been estimated that biomass energy use causes one-eighth of observed global deforestation, with the rest attributable to logging, agricultural land clearing and road building [27]. For Brazil, Poole and Moreira [22] suggest that firewood production for residential and agricultural uses does not contribute significantly to deforestation. In con-

Asian stoves closest in fuel and design. The results of recent tests of Asian stoves were incorporated, which indicate emissions of previously unmeasured greenhouse gases (CH,, N 2 0 etc) could be very high [27].

Transport sector emissions are the major source of increasing levels of urban air pollution in Senegal, and Africa generally. In Asian, Latin American and OECD cities, motor vehicles typically account for about 90% of CO emissions, and often a majority of HC and NO,, emissions [12]. Sub-Saharan Africa, with 10% of global population, accounts for only 2% of the global stock of 470 million vehicles. While not yet approaching the severity found in major Asian and Latin America cities, urban air quality problems related to transport emissions are of increasing concern in many African cities. High levels of CO and SO, have been measured in Ibadan City, Nigeria, and haze and eye irritation are indicative of high levels of photochemical smog on major transportation routes in Lagos [I ] . Continued use of leaded gasoline poses health risks, particularly to small children who breathe the higher lead concentrations found near tailpipe height.

Transport sector emissions depend on a variety of factors: vehicle type, emission controls, file1 characteristics, maintenance level, fleet age and driving conditions. Current 'best guess' estimates were developed for Senegal by averaging the emission characteristics of 1985 European vehicles and those of average Indian vehicles [S] to reflect maintenance and age characteristics more typical of an African country. For rail, water and air transport, data derived directly from US studies were used. Similarly, the emission characteristics of rep- resentative generic, US technologies (without emission controls') were used for the industrial and service sectors, and for most energy sector activities (electricity generation, refining, oil and gas production, fuel losses and charcoal kilns).

~ ~ ~

Table 2. Land clearing estimates for Senegal.

Total forest clearing (hectareslyear) land areaa Attributable to Source

200 000 1.3 All causes Senegal government, 1984 as cited in Ref 25

60 000 0.4 Agricultural clearing Ref 24

50 000 0.3 All wood fuels -

50000 0.3 Charcoal production Ref 2 18-33 000 0.1-0.2 Charcoal production Ref 26

% of natural

b

Including steppe (3 200000 ha), dry savanna (4400000 ha), wooded savanna (6700000 ha) and forest (400000 ha). Senegal national report, as cited in Environnemenr et diveloppemenr soufenu dans la rkgion SoudunoSahelien, Synthese des Rapports Nationaux,

October I99 I .

Nurural Uesources Forum 1994 Volume 18 Number I 41

Energy and environment in Senegal: M . Lazarus et a1

trast, they assume that 50% of total charcoal production leads to deforestation and thus to net C02 emissions. Unfortunately, the term deforestation is often unclearly defined: it is used here, as implied in the numbers above, to mean a non-regenerative reduction in biomass stocks.

Land clearing alone does not necessarily imply permanent deforestation. Under certain conditions - ade- quate soil moisture, limited soil erosion, presence of seeds or coppices and limited land pressures - forests can regenerate within one to two human generations. A review of studies [25] of forests previously cleared for wood fuel production in Senegal and Nigeria suggests a range of impacts under actual conditions. Hosier [14] found that among recently harvested charcoal production areas in Tanzania, biomass cover showed signs of recovery. Similarly, Chidumayo [8] found little evidence of land degradation due to wood fuel harvesting

for urban charcoal supply in Zambia. However, the question of whether harvested lands will return to earlier levels of biomass stocks remains open. Harvesting methods, soil and nutrient loss, and most importantly, post- harvest land use and management determine whether full regeneration will occur.

Therefore, at the significant rates of land clearing in Senegal shown in Table 2, it is unclear to what extent competing land pressures will inhibit regeneration, and lead to long-term land degradation, habitat loss and net reductions in biomass stocks. Estimates of wood fuel related land clearing, largely by the charcoal industry, range from 18 000 to 50000 hectares per year. The government estimate (50 000 hectares per year) corresponds to the clearing of an estimated 900000 tonnes of wood annually, at average stock removal of 18 tonnes per hectare. For charcoal production, it has been assumed that 50% of this biomass does not regenerate subsequent

Table 3.1988 emission estimates by sector.a

Supply/transformation sectors Crude oil production Natural gas production Charcoal production Natural gas transmission and

distribution loss Refinery Electricity (SENELEC + self-

generated Diesels Gas turbines (distillate) Gas turbines (natural gas) Steam boilers (fuel oil) Bagasse

Total supply/transformation

Demand sectors Industry Fisheries Transpon Household Other

Total demand

Total system

CO, C02:net Fossil biogenicb COh HC CH, PbC NO, N,Od SO: TSP (1000 t) (1000 t) (1000 t) (1000 t) (1000 t) (t) (1000 t) (t) (1000 t) (1000 t)

0 1 0 0

699 31 36 9 2 40

I l l I 1

0 I 0 1 I

57 0 0 1 0 0 49 0 0 0 0 0 15 0 0

740 0 0 0 2 I0 I 0 0 0 3

973 699 32 38 10 0 6 I I 44

313 0 0 I 3 0 I64 8 3 0 1 3 0 0 878 20 3 0 16 5 10 I I 105 179 156 9 12 I 81 I 12

2 0

1462 179 I85 16 12 16 8 94 5 13

2435 878 217 53 22 17 15 94 16 57

Blank entries indicate that no values have been entered (inadequate data or not an applicable category); zero values indicate a value of less than

All charcoal related CO, emissions reported under charcoal production, while household line reflects all firewood-related emissions. The net bio- 0.5.

genic emissions reflects the assumption 50%-90% of the carbon emitted, as described in the text, is either balanced by wood regrowth or is not attributable to other factors such as land clearing primarily for non-energy purposes. These factors were also applied to the carbon emitted as CO from biomass fuel consumption.

N,O estimates were included only for recent household measurements [27] and are thus useful only for comparisons within this sector. Lead and sulphur emissions will be proportional to their content in fuels used in Senegal. Senegal purchases and produces petroleum products with

a wide variation in sulphur content (heavy fuel oil sulphur content ranged from 0. I % to 3.3% from 1986 to 1991); thus standard assumptions were made about sulphur content.

42 Natural Resources Forum 1994 Volume 18 Number I

Energy and environment in Senegal: M. Lazar-us et a1

Although rural dwellers consume far more firewood than charcoal, they generally gather firewood as twigs and dead wood rather than from live trees [25, 311. Therefore, at present, rural firewood consumption contributes relatively little to permanent land degradation and to net carbon emissions. For the scenario analysis, it is assumed that 10% of firewood harvested does not regenerate, an estimate that lies between the assumptions for Brazil (20%) and an assumption of no impact (O%), which seems unlikely. With increasing population, however, the impact of rural firewood demand could increase. During the time horizon of this study (to 2005), this 10% assumption could prove too conservative. As with the charcoal assumption above, uncertainty will be reflected in future sensitivity analyses.

to harvest. Although similar to the estimate for Brazil above, this assumption may appear high in light of the findings noted above regarding regeneration potential, but low with respect to other assumptions for Senegal [2]. While it may overstate the permanent impacts of charcoal production, this assumption reflects a high concern for the other immediate impacts on rural well being and reduced ecological diversity. If pressures of global and local climate change continue to mount along with other land pressures, this assumption could indeed prove low.

For charcoal production, ecosystem impacts are only part of the environmental picture. Charcoal production affects villagers most strongly. About half of village women surveyed blamed charcoal producers for wood scarcity [25 ] . They also blamed charcoal makers for the disappearance of game species, destruction of fodder, water conflicts and social problems. Some researchers have found that in areas where charcoal production takes place, villagers experience fuel wood scarcity that either was not present before or is not found in other vil- lages where charcoal makers are absent. While stopping charcoal production in Senegal would certainly not halt land clearing, and the precise nature of damage to rural environments of this industry is not fully understood, a pervasive impression is that 'charcoal demand is a key cause of the decline of Senegal's ecology' ([31], p 32).

Environmental results The emission assumptions result in the national sectoral profile for 1988 are shown in Table 3. While further analysis might enable a better assessment of the impact of these emissions, particularly their spatially dependent relationships to indoor and ambient urban concentrations, this table none the less provides a starting point for general comparison among sectors and emission categories.

1988 1990 1995

+ C02: Fossil Fuel ( I 0,OOOT)

- - CO (1 OOOT)

--A- - Pb (0.IT)

C02: Net Biogenic (I0,OOOT)

U SOx(100T)

4 TSP(IOO0T)

--+-- HC(1000T)

CH4(1000T)

-0- NOx(I0OOT)

2000 2005

Figure 7. a Period 1988-90 not to scale.

Reference case emission estimates.a

Naturul Resources Forum 1994 Volume 18 Number I 43

Energy and environment in Senegal: M . Lazat~us et a1

Scenario A Scenario B Scenario D

10%

0 Yo

- 10%

-20%

-30%

-40%

-50%

-60%

-70%

C02: Fossil Fuel

[7 C02: Net Biogenic

co

i-1 HC

CH4

Pb

NOx

sox

TSP

Figure 8. Emission changes relative to reference case in 2005.

The three major categories of net CO, emissions are oil fired electric generation, transport and charcoal production, each accounting for over 20% of the national total. The biomass harvesting assumptions noted earlier significantly influence the results. If all wood was assumed to be unsustainably harvested, with no regeneration, the net biogenic CO, would exceed the contribution of all fossil fuels, and would be over 3.5 times greater than current estimates. As calculated here, biomass contributes 27% of total CO, emissions, and 33% of total global warming potential (GWP), using the IPCC’s 1992 100 year integration estimates.

In addition to its other local environmental impacts, charcoal production could produce over 70% of particulate emissions, over 40% of methane, and over 60% of other hydrocarbon emissions. Because these emissions occur over a dispersed area, their contribution to local air quality, local climate and visibility may be small, and they should be compared with emissions from other forms of rural biomass burning, such as brush and crop waste fires. Household consumption of biomass fuels is the largest energy related contributor to methane and carbon monoxide emissions, and the second major contributor to particulate and other hydrocarbon emissions after charcoal production. Particulate, carbon monoxide and hydrocarbon emissions might provide a useful, if

very generalized, indicator of health related indoor air quality issues. The transport sector accounts for the largest share of current nitrogen oxide emissions, a share likely to increase with urbanization and increased traffic congestion. Fuel oil use in electric and industrial steam boilers is responsible for most sulphur oxide emissions.

As illustrated in Figure 7, under the reference case, emissions increase for all categories except sulphur oxides. This decrease in sulphur oxides reflects assumptions regarding the use of lower sulphur fuels with newer electric generating units. Total annual CO, emissions in 2005 increase by 1.9Mt, an increase of 59% over 1988 levels. The share of net biogenic emissions remains relatively constant at about 30% of the total throughout the period.

The policy scenarios result in significant decreases in all emission categories by 2005, as shown in Figure 8. The butanization scenario (A) reduces total CO, emissions by 0.9 Mt relative to the reference case in 2005, cutting in half the projected increase in reference case CO, emissions. The decrease in net biogenic CO, of almost 70% shown above, far more than offsets the increased CO emissions from fossil fuels by about 4%. At the same time, total emissions of CO, HC, CH, and TSP decrease from 28% to 56% relative to the reference case, as two major sources of these emissions, charcoal

44 Narurul Resources Forum 1994 Volume 18 Number I

Energy and environment in Senegal: M. Lazarus el a/

I800

I700

1600

I500

Global Warming Potential

(1000 Tonnes C Equivalent)

1400

1300

I200

1100

1000

Reference Case HH Switching from Charcoal to LPG (Scen A)

Improved End-Use

\ New Hydro Capacity (Scen D)

I985 I990 1995 2000 2005

Figure 9. Global warming potential for Senegal scenarios (1992 IPCC 100 year integration).

production and use, are greatly curtailed. Scenario C results in an additional decrease of 0.6Mt

in fossil fuel CO, relative to the reference case, leading to only a 14% increase in total CO, emissions over the 17 year study period. In addition, all other categories of emissions further decrease relative to the reference case. By avoiding oil fired power plant emissions, hydro additions in Scenario D result in significant decreases in NO,, SOx and CO, emissions. Total CO, emissions in the year 2005 decrease an additional 10.7 Mt to 3.1 Mt,

an overall decline of 8% from 1988 levels. If Senegal pursued the policy options in these scenar-

ios it would improve many categories of environmental emissions and impacts relative, not only to the reference case, but also to current conditions. If the strategies out- lined above are economically and socially desirable, and achievable, they could represent no or low regret options for abating greenhouse gas emissions. If successful, these combined policies could stabilize CO, and overall GWP at 1988 levels by 2005, as Figure 9 shows.

Table 4. Comparative emissions and fuel costs for cooking a typical Senegalese dish.

Fuel use Recent fuel Fuel cost Emissions per dish per dish price per dish CO, CWP ( W a (CFA/kg)' (CFA) (kdh (kgCOp4'

Charcoal Standard (Malgache) 0.77 64 49 3.48 6.83 Improved (Saakanal) 0.53 64 34 2.39 4.70

LPG (3-6 kg bottles) Standard 0.36 120 43 1.12 1.14 Improved 0.25 I20 30 0.78 0.79

Ref 1 1, Ref 31. Emissions are for the full fuel cycle, including those from charcoal production and petroleum refining. CO, equivalent values were calculated using the IPCC 1992 values for 100 year integration.


Energy and environment in Senegal: M . Lazarus et a1

Although reducing greenhouse gas emissions is not necessarily a policy objective most immediately relevant to a small developing country, this potential is relevant from an international policy perspective. These results confirm that reducing emissions in developing countries could prove less expensive to implement than many in industrialized countries. This suggests possibilities for mutually beneficial assistance or transfer payments, if needed, to overcome foreign exchange constraints or other obstacles to policy implementation.

In the case of Senegal, we can look more specifically at the cost of greenhouse gas emission abatement associated with butanization, a policy with both local and global benefits. Table 4 indicates the fuel use and emissions associated with cooking a typical meal in Senegal derived from field surveys using different stoves and fuels. Because of the inefficiency of charcoal production and use, CO, emissions are about three times higher than either LPG or kerosene production and use.’ In terms of GWP, the contrast is even more dramatic: emissions from charcoal are up to six times higher. Even if all wood destined for charcoal production were harvested on a renewable, fully regenerative basis, overall greenhouse gas emissions would still probably be higher than from LPG refining and use.

Assuming that a CFA60/1 kg LPG subsidy is required to overcome barriers to switching from charcoal to LPG, this policy costs about US$30 per tonne of CO, reduced, and around US$15 per tonne of CO, equivalent GWP.8 This cost compares favourably with a recently proposed European Community carbon/energy tax of US$22 per tonne by the year 2000. LPG subsidy may not necessarily be the best approach to encouraging substitution (indeed, it may be a relatively high cost option), but it has proven politically acceptable in practice.

Conclusions The results of the preliminary scenarios indicate that LPG substitution policies could substantially reduce greenhouse gas emissions, while improving more important near-term environmental problems in an African country such as Senegal. Given their potential as low- cost contributors to reducing greenhouse gas emissions,

’This calculation uses the previous assumption that SO% of charcoal production involves non-regenerative harvesting. The ‘break-even’ level at which LPG and charcoal CO, emissions would be equal is about 16%.

This calculation assumes that: (a) fuel switching would not occur in the absence of the subsidy; (b) switching occurs between the most commonly available charcoal and small LPG equipment (switching between the most efficient of each would have the same effect, at lower total cost): (c) the 50% non-renewable fraction stays constant. It does not include the cost of end-use equipment, which needs to be added, but is unlikely to dramatically change the overall economics as fuel costs dominate. The current subsidy is targeted to low- and middle-income families by providing the subsidy only for small stoves.

these policies may deserve additional attention and support from aid and funding sources. The effects on oil imports and resulting foreign exchange requirements of increasing LPG use in households is limited to roughly 5% of future oil consumption, which represents the end- use fuel requirements for household cooking.

A set of combined improved efficiency efforts for Senegal could reduce emissions and indicators for several environmental impacts, relative not only to the future reference case but to current levels. Even when combined with butanization, improved efficiency efforts could reduce oil imports by up to 12% relative to reference case levels in 2005, despite a dramatic increase in LPG consumption. Thus, the dual objectives of Senegal’s energy policy are achieved. Although an extensive industrial audit programme and several improved stove and kiln efforts have already been attempted, additional efficiency programmes targeted at household, services and transport sectors are needed to achieve the significant potential. In particular, integrated resource planning efforts in the electric sector could be modelled after the promising efforts currently under way in Thailand [7] and elsewhere. Given the small relative size of the Senegalese economy, regional West African import policy, standards and equipment testing efforts should also be contemplated.

Ideally, the simplified environmental assessment approach used here should be expanded from the treat- ment of aggregate measures of emissions and land clearing to actual damage analysis. In addition, more emphasis should be given to less easily quantified impacts, such as hydroelectric land use and other impacts. Omitting such impacts risks giving an implicit advantage to resources and technologies whose impacts are the most difficult to measure. While risk and economic analyses can help further translate various energy scenarios into estimates for ecological, human health and welfare, and employment, balancing these benefits and damages impacts is inherently the domain of a nation’s social values. As shown in Table 1 , a matrix of potential impacts across scenarios and social, environmental and economic issues of concern provides a first step toward a comprehensive overview to assist decision making. The next step involves placing values or relative weights on each type of impact.

Efforts to make these values and weights explicit in the policy making process will help to ensure that energy choices do not ignore important environmental considerations, and will provide a more transparent and publicly accessible basis for decisions. To this end, several methods have been explored for valuing externalities and implementing decision frameworks [4]. Practical experience with these comparative approaches is relatively limited, however. Perhaps the most extensive experience is the recent efforts to internalize air emission impacts in the US electric sector planning

46 Noturul Resources Forum 1994 Volume 18 Numhcr I

Energy and environment in Senegal: M. Lazarus et a1 use in Tanzania: prices, substitutes and poverty, Energy Policy, Vol 21, No 5, May 1993, pp 454-473. M. Kosmo, Money to Burn? The High Costs of Energy Subsidies, World Resources Institute, Washington DC, October 1987. M. Lazarus, L. Greber, J. Hall et al, Towards a Fossil Free Future: The Nest Energy Transition, Greenpeace International, Amsterdam, 1993. G. Leach, ‘The energy transition’, Energy Policy, Vol 20, No 2, February 1992, pp 116-123. R. Ottinger, ‘Consideration of environmental externality costs in electric utility resource selections and regulation’, in E. Vine, D. Crawley and P. Centolella, eds, Energy Efftciency and the Environment: Forging the Link, American Council for an Energy-Efficient Economy, Washington, DC, 199 I . M. Philips, The Least Cos? Energy Path fo r Developing Countries: Energy Efftcient Investments for the Multilateral Development Banks, International Institute for Energy Conservation, Washington, DC, September 1991. M.M. Pitt, ‘Equity, externalities, and energy subsidies: the case of kerosene in Indonesia’, Journal of Development Studies, Vol 17, 1983. A. Poole and J. Moreira, ‘CO, emission abatement in Brazil’, in P. Hayes and K. Smith, eds, Global Greenhouse Regime: Who Pays.?, United Nations University PressEarthscan Press, Tokyo, 1993. P. Raskin, ‘Integrated energy planning in developing countries: the role of computer systems’, AMBIO, Vol 4, Nos 4-5, 1985. RCpublique du SCnCgal, Vers un de‘veloppement durable, National Report to UNCED, Rio de Janeiro, June 1992. J.C. Ribot, Markets, States, and Environmental Policy: The Political Economy of Charcoal in Senegal, PhD Thesis, University of California, Berkeley, 1990. J.C. Ribot, ‘Forestry policy and charcoal production in Senegal’, Energy Policy, Vol 21, N o 5, May 1993,

K.R. Smith, M.A.K. Khalil, R.A. Rasmussen et ul. ‘Greenhouse gases from biomass and fossil fuel stoves in developing countries: a Manila pilot study’, Chemosphere, Vol26, Nos 1 4 , 1993, pp 479-505. S. Subak, P. Raskin and D. von Hippel, National Greenhouse Gas Accounts: Current Anthropogenic Sources and Sinks, Stockholm Environment Institute, 1992. A. Tibesar and R. White, ‘Pricing policy and household energy use in Dakar, Senegal’, The Journal of Developing Areas, Vol25, October 1990, pp 3 3 4 8 . Union of Concerned Scientists, Natural Resources Defense Council, Alliance to Save Energy, American Council for an Energy Efficient Economy, Tellus Institute, America’s Energy Choices, Union of Concerned Scientists, Cambridge, MA, 1991. World Bank, Urban Household Energy Strategy, ESMAP, 1989. World Health Organization, Management and Control of the Environment, Report WHO/PEP/89, Geneva, 1989. World Resources Institute, World Resources 1992-93, Oxford University Press, New York, 1992.

pp 559-585.

process. These efforts have generally used monetary cost values (eg planning adders) for individual pollutants, rather than alternative ranking scoring systems. Can such efforts be extended across the full set of energy resources? Can these methods be usefully applied in smaller developing countries such as Senegal? Or will this ‘wave of the future’ affect larger electric systems alone?

References I

2

3

4

5

6

7

8

9

10

1 1

12

13

14

15

0. Adegbulugbe, ‘Energy-environmental issues in Nigeria’, International Energy Workshop, Harvard University, Cambridge, MA, June 1992. S. Amous, De‘gradution de I’environnement au Se‘ne‘gal: le role de I’e‘nergie, Programme Energy-Environnement en Afrique, 1992. Asian Development Bank, Environmental Considerations in Energy Development, Energy and Industry Department, May 199 1. S. Rernow, B. Biewald and D. Marron, ‘Environmental externalities measurement: quantification, valuation, and monetization’, in 0. Hohmeyer and R. Ottinger, eds, External Environmental Costs of Electric Power, Springer-Verlag, Berlin, 199 1. R.K. Bose and V. Srinivasachary, Policies to Reduce Energy Use and Environmental Emissions in the Transport Sector: A Case of Delhi Ciry, Tata Energy Research Institute, New Delhi, 1992. D. Cavard, Y. Sokona and L. Ba, La Consommation dknergie des menages a Dakar: changements de struc- tures, e‘volution des e‘quipements, et des comporrements des consommateurs, COPEDEPEENDA-TM, 1991. M. Cherniack and P. du Pont, Demand Side Management for Thailand’s Electric Power System: Five-Year Master Plan, International Institute for Energy Conservation, Bangkok, 1991. E. Chidumayo, Zambian charcoal production: miombo woodland recovery’, Energy Policy, Vol 21, No 5 , May 1993, pp 586-597. D.R. Diour, Se‘ne‘gal: srrategie de I’e‘nergie dnmestique, SENELEC, June 199 I . P. Ehrlich, A. Ehrlich and J. Holdren, Ecoscience, Freeman, 1977. Environment and Development in the Third World (ENDA-TM), L’Energie au Se‘ne‘gal, Edition 1990, Programme Energie. A. Faiz, ‘Automotive emissions in developing countries - relative implications for global warming, acidification and urban air quality’, Conference on Transportation and Global Climate Change, Pacific Grove, CA, August 1991. J. Girod, J.P. Thomas, Y. Sokona and S. Diallo, Les Pe$ormances du systPme e‘lectrique en Afrique de L’Ouest: Burkina Faso, Cote D’lvoire, Mali, Niger et Senegal, ENDA-TM, April, 199 1. R. Hosier, ‘Charcoal production and environmental degradation: environmental history, selective harvesting and post-harvest management’, Energy Policy, Vol 2 I , No 5, May 1993, pp 491-509. R. Hosier and W. Kipondya, ‘Urban household energy

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

Narural Resources Forum 1994 Volume 18 Number I 47

energy and environment scenarios for senegal

Documents