DEVELOPMENT, AND DISSEMMINATION
By
Samuel F. Baldwin
Princeton University
Support for the publication of
this volume was provided by
the Directorate General
for Development Cooperation
Ministry of Foreign Affairs
Government of the Netherlands
VITA
1600 Wilson Boulevard, Suite 500
Arlington, Virgnia 22209 USA
Tel: 703/276-1800 . Fax: 703/243-1865
Internet: pr-info@vita.org
Center For Energy and Environmental Studies
Princeton University
Princeton, New Jersey 08544 USA
Biomass Stoves
Copyright [sup.c] 1987 Volunteers in Technical Assistance
All rights reserved. No part of this publication may be produced or transmitted
in any form or by any means, electronic or mechanical, including photocopy,
recording, or any information storage and retrieval system without the written
permission of the publisher.
Manufactured in the United States of America.
Published by
VITA
1600 Wilson Boulevard, Suite 500
Arlington, Virgnia 22209 USA
Library of Congress Cataloging-in-Publication Data
Baldwin, Samuel F., 1952-
Biomass stoves.
Bibliography: p.
Includes index.
1. Biomass stoves--Design and construction.
2. Biomass energy--Developing countries. 3. Fuelwood--
Conservation--Developing countries. I. Title
TH7436.5.B35 1987 683'.88 87-6107
ISBN 0-86619-274-3
To my sister, Hannah
ACKNOWLEDGEMENTS
The work presented in this volume began in West Africa, under the auspices
of a long-term project implemented by Volunteers in Technical Assistance
(VITA) and the Comite Permanent Inter-etats de Lutte Contre la Secheresse
dans le Sahel (CILSS). Since then, numerous people and organizations have
assisted at every step in its development. Many of the contributors have
been carefully noted in the detailed references and so will not be
repeated here. However, special thanks are due the following:
For financial support while in Africa: United States Agency for International
Development and IBM-Europe.
For institutional support while in Africa: CILSS, Ouagadougou; l'Institut
Burkinabe de l'Energie (IBE), Ouagadougou; Mission Forestiere Allemand
(MFA), Ouagadougou; Laboratoire d'Energie Solaire (LESO), Bamako; Centre
des Etudes et des Recherches des Energies Renouvelables (CERER), Dakar;
Association Bois de Feu, Marseille; Association pour le Developpement des
Energies Renouvelables en Mauritanie (ADEREM), Nouakchott; Church World
Service (CWS), Niamey; United States Agency for International Development
(USAID); and United States Peace Corps.
For partial financial support in the U.S.: World Resources Institute and
the Rockefeller Brothers Foundation, The Hewlett Foundation, the Center
for Energy and Environmental Studies of Princeton University, and VITA.
For illustrations and graphics assistance: Ellen Thomson, Thomas O.
Agans, and Mike Freeman.
For editorial and production assistance: Julie Berman, Margaret Crouch,
Juleann Fallgatter, Maria Garth, and Jim Steward of VITA.
For review comments and suggestions: Alfredo Behrens, Margaret Crouch,
Gautam Dutt, Eric Larson, Cliff Hurvich, Eric Hyman, Willett Kempton,
Robert Morgan, H.S. Mukunda, Tom Norton, Kirk Smith, Bob Williams, and
Timothy Wood.
For providing optical scanning equipment: Charles Creesy of Princeton
University.
For preparation and publication support: The Hewlett Foundation, the
Center for Energy and Environmental Studies, and VITA.
Simply listing those who have helped, however, does not adequately
describe the critical role that so many have played in this work. The
original improved stoves project with CILSS began in 1980 when IBM-Europe
approached VITA with a request to design a program with CILSS for the
research and development of improved stoves as a way to combat deforestation.
USAID later provided funds to keep this program going. It was the
foresight and unwavering support of these two organizations -- the aid
agency and the corporation -- that allowed this work to take place at all.
Timothy Wood was the first Technical Coordinator of the VITA/CILSS
improved stove project and it was his fine work in organizing many of the
national projects and in beginning the development of fired clay stoves
that, in large part, paved the way for the work described here.
Following my arrival in West Africa as the second Technical Coordinator,
the work described was made possible only through assistance far above and
beyond the call of duty by: Issoufou Ouedraogo, Georges Yameogo, Frederic
Yerbanga, and Stephen and Cornelia Sepp in Burkina Faso; Yaya Sidibe,
Cheick Sanogo, and Terry Hart in Mali; Massaer Gueye, Lamine Diop, and
Susan Farnsworth in Senegal; Ralph Royer in Niger; Bill Phelan in Mauritania;
and above all, Moulaye Diallo of CILSS and Sylvain Strasfogel of
Association Bois de Feu. At the same time, I received superb support from
Paula Gubbins and Juleann Fallgatter at VITA headquarters. Many, many
others also helped significantly and to them I must apologize for not
specifically citing their names here.
With my return to the United States I have continued to receive invaluable
assistance from many sources. Among those listed above, special thanks
are due Margaret Crouch, Gautam Dutt, Eric Larson, and Ellen Thomson. In
particular, Margaret and Gautam have provided countless hours of editorial
and production assistance, and unflagging support in this long endeavor.
To all of these people I give a heartfelt thanks. Those mistakes that
remain in the text are mine alone and somehow remain despite all the
patient editorial assistance that I have received. Similarly, several
illustrations of lower quality remain -- they are due to my shaky hand and
somehow remain despite the professional assistance available to me. I
hope the reader will understand the underlying themes of this work despite
these shortcomings.
I would also like to thank my sister, Hannah, for first making me aware of
the problems in developing countries. This book is testimony to the
profound impact a simple trip to visit her in Senegal in 1972 has had on
my career.
Finally, I would like to thank my wife, Emory, for her love, patience, and
understanding during the long months while what was intended to be a 50-page
technical report turned into a 300-page book.
Sam Baldwin
November 1986
TABLE OF CONTENTS
Acknowledgements
Table of Contents
I. INTRODUCTION AND OVERVIEW
II. FUELWOOD, CHARCOAL, DEFORESTATION, AND STOVES
Fuelwood
Charcoal
Environmental Impacts
Economics and Policy Options
III. STOVE DESIGN
Conduction
Convection
Radiation
Combustion
Other Aspects of Stove Efficiency
IV. STOVE CONSTRUCTION
Construction Options
Template Design: Cylindrical Stoves
Metal Stove Production
Fired Clay Stove Production
V. STOVE TESTING
Laboratory Tests
Controlled Cooking Tests
Production Tests
Field Tests
Marketing Tests
VI. CHARCOAL FUELED SYSTEMS
Charcoal Stoves
High Temperature Furnaces
APPENDIXES
A. Conduction
B. Convection
C. Radiation
D. Combustion
E. Heat Exchangers
F. Financial Analysis
G. Statistical Methods
H. Testing Equipment
I. Units and Conversions
J. Institutions
NOTES, REFERENCES, AND FURTHER READING
INDEX
CHAPTER I
INTRODUCTION AND OVERVIEW
Developing countries are now suffering serious and increasingly rapid
deforestation. In addition to environmental degradation, loss of forest
cover removes the wood energy resources on which traditional rural
economies are based. In response to the increasingly serious shortages,
programs to conserve fuelwood supply and to expand fuelwood production
have multiplied, but have frequently been ineffective due to a lack of
understanding of the economic, political, social, and technical complexities
of these problems.
The primary intent of this book is to resolve some of the technical
problems of conserving fuelwood supply(1). This is done by using the
principles of modern engineering heat transfer to redesign traditional
energy technologies. As shown, this unlikely marriage of the modern and
the traditional is a powerful tool for improving the lives of the Third
World's poor.
The book is divided into two parts, the text and the technical appendixes.
The text is written for generalists who need a qualitative yet detailed
understanding of stove design and testing. The appendixes are written for
specialists who need an introduction to the application of the principles
of combustion and heat transfer to stove design. The two parts are combined
into a single volume so as to emphasize the importance of technical
analysis to stove design, development, and dissemination. In brief, the
contents are as follows.
______________________
(1) A companion volume discusses policy aspects of using biomass energy
resources for rural development (1). Stove program planning and implementation
are discussed at length in reference (2).
Chapter II, Fuelwood, Charcoal, and Deforestation, reviews the role of
fuelwood in traditional societies, and the environmental, economic, and
policy considerations of increasing deforestation and worsening fuelwood
shortages. Although fuelwood demand is not a primary cause of deforestation
on the global scale, it can significantly increase pressures on
forest resources locally, particularly around urban areas in arid regions
where the fuelwood demand is large and the biomass productivity of the
land is small. In turn, deforestation places an enormous financial and
physical burden on hundreds of millions of people in developing countries
as they struggle to obtain vital supplies of fuel with which to cook their
food and heat their homes.
Responses to these problems might include tree planting programs, improved
land management, or the importation of fossil fuels for cooking. All of
these may be important components of any long-term strategy to meet the
energy needs of developing countries (1). Yet in many rural and urban
areas such programs cannot be implemented quickly enough or are too
expensive to overcome the rapidly growing fuelwood deficits.
Improving the energy efficiency of biomass burning stoves potentially
offers a highly cost-effective alternative for easing the burden of buying
fuel by urban poor and collecting fuel by rural poor. Better stoves also
promise important health benefits to their users by reducing smoke
emissions. Finally, stoves may ease pressures on forests as well as help
maintain long-term soil productivity by reducing the need to burn crop
residues and dung.
Chapter III, Stove Design, discusses the technical aspects of combustion
and heat transfer as applied to improving biomass burning cookstoves(2). The
following points are emphasized:
o Conduction processes in the stove require the stove to be as lightweight
as possible to minimize stored heat in the walls and, where
possible, to be lined with lightweight, high temperature insulants to
reduce heat loss to the outside. Their light weight and easy transportability
allow centralized mass production with distribution through
existing commercial channels or decentralized mass production with
distribution by informal sector artisans.
______________________
(2) "Biomass" as used in this book refers to raw or unprocessed biomass
fuels such as wood, agricultural wastes, or dung. In contrast, fuels such
as charcoal, ethanol, methanol and others that are derived from raw
biomass are termed "processed biomass" fuels.
"Cookstoves" (or simply "stoves") refers primarily to stoves designed for
heating water. Uses could include domestic, restaurant, or institutional
scale cooking (boiling) or hot water heating; commercial and industrial
uses such as beer brewing, cloth dyeing, or food processing (boiling); and
others. It does not refer to stoves for frying foods or to woodburning
ovens, nor does it apply to space heating stoves, although many of the
same considerations will generally be applicable.
Introduction
o Convection processes in the stove require very precise control over the
stove dimensions and precise matching of the stove to the pot. The
high degree of precision needed necessitates mass production based on
standard templates.
Thus, because of fundamental principles of heat transfer, site-built or
massive stoves are unlikely to show acceptable performance; mass produced
lightweight stoves with carefully optimized and controlled dimensions are
much preferred.
In addition, combustion and radiation heat transfer processes are discussed
in Chapter III and opportunities are presented for further research to
increase efficiency and reduce emissions.
Chapter IV, Stove Construction, applies the technical findings of Chapter
III to the practical aspects of actual stove construction. Template design
and step by step production are described in detail for several metal and
fired clay stoves recently developed and now being disseminated in West
Africa. Additionally, suggestions are made for a variety of other stove
configurations that may better suit conditions in other areas.
In Chapter V, Stove Testing, step-by-step procedures are recommended for
testing stove prototypes and establishing a rudimentary stove industry. In
brief, laboratory and controlled cooking tests are used to select particularly
promising prototypes. From these tests, standard templates are
developed that conform to the local pot sizes and shapes. A production
test is then run producing 50, 100, or more stoves for each of the most
popular pot sizes. During this production test, a detailed analysis is
performed of the costs, the problems encountered, and potential improvements
in the production method.
Some of the stoves produced are distributed on a short-term, temporary
basis to selected families for field testing to determine both their
acceptability and their actual performance.
Another portion of those stoves is put on display in local commercial
outlets and sold on a commission basis. Such simultaneous marketing may
allow some indirect feedback on how neighbors of the selected families
perceive the stoves' potential. Marketing techniques such as radio and
newspaper advertising, billboards and other publicity, and public demonstrations
at social centers, schools, religious centers, and elsewhere
should also be attempted. As interest develops, the stove promoter can
gradually withdraw, leaving the stove producer in direct contact with the
various commercial outlets. If interest does not develop, modifications
will be necessarily based on the field and market surveys and any other
information that is available.
It must be emphasized that detailed, methodical testing of prototype
stoves; careful financial and statistical analysis of the results; and use
of these results to improve subsequent prototypes is crucial if improved
stoves are to be disseminated successfully and widely. In some areas the
testing prescriptions provided will need to be modified; in other areas
they will need to be completely reworked. But everywhere, careful,
methodical testing and use of the results are crucial to understanding and
overcoming obstacles to good stove performance and acceptability.
Chapter VI briefly examines improvements in Charcoal Fueled Systems such
as stoves and high temperature furnaces that may save large amounts of
fuelwood when developed.
Technical Appendixes document the text in detail and provide the technical
reader the foundation for more detailed understanding. Topics discussed
include conductive, convective, and radiative heat transfer processes;
principles of combustion; air to air heat exchanger design; and techniques
for financial and statistical analysis of test data. Analytical and
numerical solutions to heat transfer equations are described in detail and
the results are presented in the text. Extensive references are noted for
those who wish to do more detailed work and a list of institutions is
provided for contact with ongoing programs.
The specific technologies discussed in this book are by no means finalized:
rather they are beginnings. Each has certain advantages, such as
fuel efficiency or safety, compared to traditional forms, but also brings
with it certain disadvantages such as reduced flexibility or increased
cost. Whether or not the improved technology is adopted in any area will
depend on the local fuel supply, the local economy, and a host of other
factors. Further, the response will be dynamic, changing as conditions
change. As biomass energy resources decrease, however, the demand for
more fuel efficient technologies must grow. Adaptation and further
development of the technologies described here can provide the vital
energy services needed by the world's poor in an increasingly resource
limited world.
Similarly, this book is by no means a completed study but rather is an
introduction to the application of modern scientific analysis to traditional
technologies. In the examples discussed below, when modern engineering
heat transfer is applied to traditional energy technologies, new
technologies are developed with enormous potential to improve the lives of
the world's poor. Combined with modern mass production techniques that can
carry the fruits of a single dedicated engineering effort to the entire
world, this enormous potential can be realized. There is not time to
waste.
CHAPTER II
FUELWOOD, CHARCOAL, DEFORESTATION, AND STOVES(1)
Ever since people learned to control fire they have been actively deforesting
their environment, initially using fire to aid in the hunt and
later to clear land for agriculture. Tierra del Fuego or "Land of Fire"
was so named by Magellan in 1520 because of the numerous fires he saw
there set by indigenous South Americans. Tropical savannahs and temperate
grasslands are, in large part, a consequence of such repeated burnings.
An estimated half of the world's deserts were similarly created (1).
Recorded history has numerous examples of such deforestation. Crete, once
heavily forested, suffered severe wood shortages by 1700 BC due to the
demands of a growing population. Cyprus provided the bronze needed by the
ancient Greeks for weaponry. Wood shortages are a likely cause for the reduction
in bronze smelting there by 1300 BC which forced rationing on the
Greek mainland and weakened the Mycenaeans to outside attack. Aristotle
and Plato both documented the destruction of forests in Greece and the
consequences. The Romans were forced to import wood from North Africa,
France, and Spain to keep their industries, public baths, and military
operational. England suffered severe deforestation in many areas during
her early industrial period -- citizens even rioted over rising wood
prices -- until the transition to coal was made (2,3).
Today, the world's forests face unprecedented pressures. While potentially
a renewable resource, forests are disappearing faster than they are being
replaced. The United Nations Food and Agriculture Organization estimates
that forests are being lost to agriculture, grazing, commercial timber,
uncontrolled burning, fuelwood, and other factors at a rate of more than
11 million hectares per year, with 90% of the cleared land never replanted
(4,5).
_____________________
(1) The author would like to acknowledge the assistance of Timothy Wood
in preparing portions of this chapter.
As forests disappear, the financial and physical burden of obtaining wood
fuel for cooking and space heating increases for the world's poor. In
response, many turn to crop wastes and dung as an alternative, but one
that has potentially serious consequences for future soil fertility (6,7).
This is not a small or isolated problem. Nearly two million metric tons
(tonnes) of wood, charcoal, crop wastes, and dung are burned daily in
developing countries, or approximately one kilogram each day for every
man, woman, and child. Although the energy obtained represents only about
10% of the energy consumed worldwide, it is over half the energy consumed
in some 50 to 60 developing countries and is as much as 95% of the
domestic energy used there (6-9).
Biomass fuels thus play a critical role in the economies of the developing
countries. In this chapter the supply and demand of these fuels, their
production and economics, and the environmental consequences of their use
are reviewed in detail. Although the extensive statistics presented are
themselves unemotional, one cannot be unemotional about the awesome toll
on human well-being that they represent. The high cost of fuelwood
represents food, medicine, and clothing that the urban poor must forego.
The long distances walked and heavy loads carried by the rural poor
foraging for fuel represent time and labor better spent growing food or
producing goods for sale in village markets. The large amounts of smoke
emitted by traditional stoves represent the discomfort and disease that
this smoke can cause the user. Only in such a broad context can the full
impact of traditional fuels and stoves on human life and well-being be
appreciated.
FUELWOOD
The total global annual growth of forest biomass has been variously
estimated to be about 50 times annual wood consumption and five times
total annual energy consumption including fossil fuels (Note 142)(2) (10).
Despite the large average global supply, there are acute and growing
shortages of fuelwood regionally and locally. Some regions, such as Asia,
have very little per capita forest growing stock (Note 143). Within
regions, some countries are well endowed with biomass energy resources,
and others have totally inadequate supplies, (Table 1); and within
countries themselves, there are similar local abundances and shortages.
Zaire, for example, consumes only 2% of its sustainable yield of forest
biomass but has serious deforestation around Kinshasa (12).
In areas where forest resources cannot meet the demand, crop residues and
animal dung are marginally sufficient substitutes at best. In Bangladesh,
for example, crop residues and animal dung can supply about 300 watts per
capita (Table 1). This is barely enough to meet minimum needs.
_______________________
(2) So as to not overburden the text yet still provide the reader with
detailed information, a number of Tables are given as Notes beginning on
page 251.
Estimates such as these are, of course, only very crude approximations.
As these traditional fuels do not usually move through monitored commercial
markets, estimates of their production and use can only be made by
detailed measurements at the locale in question. Further, there is
considerable confusion in the literature over the units used to measure a
given quantity. For example, foresters generally use volumetric units to
measure wood but sometimes fail to specify whether it is in units of solid
cubic meters or stacked cubic meters (steres). Nor is the species and
density specified. Note (144) gives very rough equivalences between the
two volumetric units for different classes of harvested wood. Similarly,
charcoal is usually measured by volume, but its energy content is determined
by its mass, which in turn is determined by the species from which
it was carbonized (14), the temperatures at which it was carbonized, i.e.,
its residual volatile content (15), and its packing density.
When estimates of energy content are based on weight, the preferred
method, it is similarly vital to know the moisture content of the fuel and
whether the weight is on a wet or dry basis (see Chapter III).
Estimating biomass energy resources should therefore be done by direct
measurement. Forest resources can be measured by estimating standing
volumes or by cutting an area and making a direct weight or volume
measurement (16-19). Crop residues from the same species can vary widely
by soil type and rainfall as shown in Note (145) and similarly should be
directly weighed. Growth rates can be estimated by numerous repetitions
of such measurements on comparable, adjacent samples over a period of
time. Finally, where animal dung is, or could be, used as an energy
resource, it, too, should be measured directly. Estimates of dung
production rates are given in Note (146). Calorific values for a number
of different biomass fuels are given in Appendix D.
Biomass energy resources have been estimated for a variety of local,
national, and regional cases as described in references (4,7,9,13,20-28).
Fuelwood Demand
Numerous estimates of biomass fuel demand have been made on the local,
national, and regional scale (29-59). The rate of energy use by the
typical villager is usually in the range of 200-500 watts per person and
can vary dramatically with the season, climate, and general availability
of various fuels. Energy survey results are given for nearly 40 towns and
villages in Note (147). Much of this energy is used for domestic cooking
(Tables 2,3,6) and these values are much higher than the amounts of energy
used in developed countries for cooking (Table 4). This is due to the
inefficiency of traditional fuels and stove technologies as well as
changes in diet and lifestyle that are possible with higher incomes.
Globally, biomass fuels are the principal source of cooking energy for
most developing countries (Table 5). Additionally, they provide energy
for household needs such as heating bath water, ironing, and other uses.
Though perhaps atypical, 60% of domestic wood consumption in Bangalore,
India, is used to heat bath water (45).
Although their principal use in developing countries is domestic, biomass
also fuels much of the industry. As seen in Tables 7 and 8, biomass fuels
two-thirds of Kenyan industry and commerce and it is used for such things
as beer brewing, blacksmithing, crop drying, and pottery firing.
(**) Provided by 111 bullocks, 143 cows, 93 calves, 113 buffalo and 489
sheep and goats.
Reference (50)
Estimates of the energy intensity of commercial uses vary widely, but all
indicate substantial amounts of fuelwood used and often at very low
efficiencies. One stacked cubic meter of wood, for example, is required
to cure 7-12 kg of tobacco leaf. The efficiency of tobacco drying barns
in Tanzania has been estimated to be as low as 0.5% (49). Tobacco curing
uses 11% of all fuelwood in Ilocos Norte, Philippines and 17% of the
national energy budget in Malawi (34,39,47,56,59).
Tea processing requires roughly 9.5 GJ or 500 kg of dry wood to produce 30
kg of dry tea leaves from 150 kg of green leaves (45,47). Fish smoking/
drying is variously estimated to require from 0.25 kg (39) to 3 kg (40) of
fuelwood per kilogram of fish dried (47,59). Brickworks require roughly
one stacked cubic meter of fuelwood to fire 20-25 pots (39) or 1000 bricks
(59). In Bangalore, dyeing a tonne of yarn requires some 8.3 tonnes of
fuelwood; bakeries use 0.58 kg of fuelwood per kilogram of traditional
bread produced (45). In Tanzania, beer brewing requires a stacked cubic
meter to produce 180 liters (59), and the brewing industry in Ouagadougou
uses 14% of the total fuelwood used (60). Other major users include
institutional kitchens, wood processing (45), and sugar production, for
which the bagasse itself is used. Overall, biomass fuels supply up to 40%
of the industrial energy used in Indonesia, 28% in Thailand, 17% in
Brazil, and similarly large fractions in many other countries (9)(3).
_________________________
(3)A variety of units, GJ (giga-joules), kg., [m.sup.3] , tonnes, etc. , are
used here to correspond to the literature rather than using a single set
of units -- preferably GJ and watts. Conversion tables for all these
units are given in Appendix I, approximate stacking factors for wood and
charcoal are given in Notes (144,149), and calorific values are given in
Appendix D. The author regrets the inconvenience.
Biomass fuels are crucial to the economies of most developing countries.
Note (148) lists 60 countries in which biomass fuels provide 30-95% of the
total energy used. The energy these fuels provide, however, is only a
fraction of that used by fossil fuel based economies (8,31). In the
developed world, average per capita energy use is about 6 kW while in
Africa and Asia it is barely one tenth of this (8); in North America,
energy use is over 10 kW, while in Africa it is about 450 W (8,31).
With these rates of biomass energy use and supply there is a serious and
growing shortage of fuelwood in many areas. The UNFAO has estimated that
the number of people suffering an acute shortage of fuelwood will increase
from about 100 million in 1980 to over 350 million in the year 2000 (Table 9).
Such shortages increase costs for urban dwellers, lengthen foraging
for fuel by rural dwellers, and rob the soil of nutrients as people switch
to crop wastes and dung.
CHARCOAL
Charcoal is produced by heating wood in the absence of oxygen until many
of its organic components gasify, leaving behind a black porous high
carbon residue. The charcoal thus produced retains the same shape as the
original wood but is typically just one fifth the weight, one half the
volume, and one third the original energy content. A more precise
relationship is given in Note (149).
The charcoal has a calorific value of 31-35 MJ/kg, depending on its
remaining volatile content, compared to 18-19 MJ/kg for oven-dry wood.
Table D-2 illustrates how the temperature history of the carbonization
process affects the volatile content and calorific value of the resulting
charcoal.
There are two different classes of carbonization equipment, kilns and
retorts. Kilns burn part of the wood charge being carbonized to provide
the heat necessary for the carbonization process. Retorts use a separate
fuel source to provide heat and thus can conserve the higher quality
product being carbonized by using a lower quality fuel such as twigs and
branches for the heating. An extensive review is given in reference (156).
The most widespread system used in the developing world is a kiln made of
earth. In this case the wood is stacked compactly either in a pit or on
the flat ground, covered with straw or other vegetation, and, finally,
buried under a layer of soil. It is ignited with burning embers introduced
at one or more points at the bottom of the stack. The task of the
charcoal-maker throughout the ensuing "burn" is to open and close a
succession of vent holes in the soil layer to draw the fire evenly around
the wood stack, heating the wood while burning as little of it as possible.
Other systems in use include brick kilns, which are used extensively
in Brazil (66,67).
The size of the kiln can be as much as 200 stere (68) and the energy
efficiency of the conversion process is variously given as 15% in Tanzania
(47), 24% in Kenya with an additional loss of 5% of the charcoal itself
during distribution (24), 29% in Senegal (69) and Ethiopia (70), and over
50% in Brazil with brick kilns (67). Advanced retorts are claimed to be
capable of achieving 72% energy efficiencies in converting wood to charcoal
if there is complete recovery of all of the gaseous by-products (67).
The large variation in reported kiln efficiencies may be due in part to
confusion about units -- energy, weight, or volume, and wet or dry basis.
When side-by-side tests are done, energy efficiencies are typically in the
30-60% range as indicated in Table 11 (71,72). The relative economic
performance of a few types of kilns is given in Table 12. The poor economics
of the earthen kiln listed in Table 12 may be due to the very small
size studied. Others have found traditional earthen kilns to have fairly
high performance and a good financial return with relatively little labor
(71). Their disadvantages, however, include a variable yield and quality,
slow burns, and seasonal availability (not during the rainy season). No
matter what system is used, however, producing charcoal results in a very
large net energy loss. In terms of conserving forest resources, it is
always better to use wood rather than first converting it to charcoal.
Charcoal Transport
It has been frequently argued that it is cheaper and more efficient to
transport charcoal than wood because of its higher energy content per unit
mass. As shown below, however, the amount of energy, whether in the form
of wood or charcoal, that can be carried per truckload is about the same.
As transport costs are primarily due to vehicle depreciation and maintenance,
the cost of hauling wood or charcoal is about the same per unit
of energy carried (150).
By assuming transport costs at a fixed US$0.10 per metric ton-kilometer,
Earl found that it was cheaper to transport energy in the form of charcoal
than in the form of wood for distances greater than 82 km (13). Chauvin
similarly used a fixed cost per ton-km. in his analysis of the economics
of transporting charcoal from the Ivory Coast to Burkina Faso by rail (60)
Expressing transport costs in terms of ton-km's is a standard practice in
aggregated transportation statistics, but is not applicable in this
situation. Most of the energy is used to move the vehicle itself, to
overcome wind resistance, internal friction and so forth. Thus, an empty
truck uses nearly as much energy as one that is full. A linear regression
on data presented in reference (73) shows that the energy intensity of
transport by tractor-trailers in the USA is related approximately to the
payload for the range 8-25 metric tons by the equation
E = 23.6/M + 0.476
where E is the energy intensity in MJ per metric ton-km the load is moved,
and M is the mass of the load in metric tons. Transport is more often
limited by volume than by weight and this is particularly true in the
developing world where vehicles are usually filled to overflowing. In
this case of volume limited transport, Table 13, 13% more energy can be
transported per truckload of wood than of charcoal at a cost of a 21%
increase in fuel use.
Fuel costs, however, are only a small part of the total transport costs
and at least in some cases, do not substantially increase even on unimproved
roads (74). Maintenance and repair of vehicles is a large factor
(74) and vehicle depreciation and labor are even larger (75).
US$0.40/man-hr.
Reference (72). Also see (72) for data on 12 other types of kilns.
b) From Table 21.
c) Estimated from reference (75) data on tire depreciation and
vehicle repair charges assuming that these costs increase proportionately
to the total vehicle weight.
When these costs are considered, Table 14, the cost of hauling energy,
whether in the form of wood or charcoal, is virtually identical. In
practice, factors such as vehicle size, labor and fuel costs, part-load or
back-haul of goods, and many others will complicate this analysis.
When production costs are included, charcoal is more expensive than
fuelwood. These costs are reflected in their relative prices: the price
per GJ of charcoal is typically twice that of fuelwood (76).
Charcoal Demand
Despite its higher price, charcoal is a very popular fuel, particularly in
urban areas where people have a cash income. According to a 1970 report
from Thailand, 90% of the wood cut for urban markets was converted into
charcoal (34). In Tanzania that figure is 76%, with 10-15% of all wood
cut converted to charcoal (40,59). In Senegal, 15% of all wood cut is
converted to charcoal for Dakar alone, transported to Dakar from as far as
600 km away, and used there by 90% of the households at a rate of 100
kg/person-year (77,78). In Kenya, 35% of the wood cut is converted to
charcoal (24).
Although traditional charcoal stoves have an efficiency (15-25%) somewhat
higher than the open wood fire (15-19%), this does not compensate for the
drastic energy loss in the initial conversion from wood (79,80).
There are a variety of reasons for this popularity despite high cost and
energy inefficiency. Unlike some wood species that must be used within as
little as a month of drying to avoid significant losses to termites,
charcoal is impervious to insect attack (21). It can, therefore, be
prepared far in advance of, for example, the rainy season when other fuels
are unavailable. Even more important is that charcoal is a very convenient
fuel to use. Charcoal is nearly smokeless. Cooking can be done indoors
in relative comfort without blackening the walls with soot. Metal pots
stay relatively clean, and there is no smoke irritation to eyes or lungs.
Although there can be a high output of dangerous carbon monoxide, which is
a health hazard in poorly ventilated kitchens, this does not cause as
obvious discomfort to the user. Additionally, once it is lit, a charcoal
fire needs little further attention from the cook, while a wood fire
requires frequent adjusting of the fuel.
The willingness of urban dwellers to purchase expensive charcoal should
thus encourage designers of improved stoves who are attempting to eliminate
smoke, ease the drudgery of cooking, and further reduce fuel costs.
At the same time, it should serve as a warning to those who pay attention
only to fuel efficiency.
Charcoal is also extensively used commercially. In Brazil, some 19
million cubic meters of charcoal were used during 1983 to produce pig
iron, 2.5 million were used to produce cement, and 600,000 were used for
metallurgy. Overall, about 18% of the energy used in the Brazilian steel
industry is from charcoal. About 17% of this charcoal was generated from
plantations (43,67,82).
Large amounts of charcoal are traded internationally as well. In 1981,
Indonesia, Thailand, and the Philippines each exported 44-49 thousand
tonnes of charcoal. Large importers include Japan, with 52,000 tonnes,
and Hong Kong, with 23,000 tonnes (65).
ENVIRONMENTAL IMPACTS
There is now rapid and increasing deforestation around the world. The
UNFAO (5,83) has estimated total annual global deforestation at about 11.3
million hectares (Table 15). Others have estimated it to be as high as 20
million hectares and more per year (7). Among the causes are the following.
Shifting agriculture damages or destroys about 0.6% of tropical
forestland annually and accounts for some 70% of forest loss in Africa
(84). Opening pastureland to grow beef for export annually clears some 2
million hectares per year in Latin America (85-87). Commercial timber
operations clear roughly 0.2% of tropical forestland annually (84), and
timber access roads open the areas to farmers leading to additional
degradation (87). The Ivory Coast, for example, is losing some 6.5% of its
forests annually (5,83). Finally, uncontrolled burning is believed
responsible for the creation of much of the world's savannah and grassland
(1,88,89). Such brushfires in the African grasslands burn more than 80
million tons of forage annually, cause volatilization of organic nitrogen,
and allow excessive leaching of valuable salts (90). This may be particularly
damaging in much of the Sahel where growth is already strongly
limited by the small available quantities of nitrogen and phosphorus (91).
The use of fuelwood increases pressures on forest biomass and can lead to
local deforestation (12,88), particularly in arid regions around urban
areas where demand is high and biomass growth rates are low. Generally,
rural subsistence farmers cause relatively little damage to forests as
they take only small limbs, etc., and these often from hedgerows or from
near their farmlands. For example, in Kenya, trees outside the forest
supply half the wood demand (37); in Thailand in 1972, 57% of the wood
consumed came from outside the forests (40). In contrast, commercial
fuelwood and charcoal operations, even relatively small-scale ones, cut
whole trees and can damage or destroy large areas of forest.
Among the potential impacts of deforestation are erosion, flooding,
climatic changes, desertification, and fuelwood shortages (92-94). Essentially
no soil or rainfall is lost from naturally forested areas. However,
when tree cover is removed, massive amounts of soil can be washed away as
the rainfall flows across the surface. Measurements in Tanzania indicated
that up to half the rainfall was lost as run-off from bare fallow (3.5[degrees]
slope), carrying some 70 tonnes/ha of soil with it (95). Similar impacts
have been noted elsewhere (5,81,87,88,96,97).
Erosion chokes downstream waterways and reservoirs with silt, making them
even less capable of handling the increased volumes of water running
directly off the watersheds (2,7). In 1982, flood and erosion damage due
to clearing India's forests was estimated to total $20 billion over the
previous 20 years. This estimate included loss of top soil, loss of
property to floods, and shortened reservoir lifetimes (5). Other estimates
place the direct costs of repairing flood damage at more than $250
million per year (98). A general review of this problem in India is given
in reference (99).
As two-thirds of all rainfall is generated from moisture pumped back into
the atmosphere by vegetation, deforestation may cause serious climatic
change (1,100). The surface reflectance is also changed and may affect
climate (1). With no shading, soil temperatures rise dramatically and can
greatly reduce the vital biological activity in the soil (87,101).
Following deforestation, overgrazing and trampling can quickly destroy the
grass layer. Without the protection of ground cover, the soil receives
the full force of pounding raindrops, bringing clay particles to the
surface and causing surface hardening and sealing that seeds cannot
penetrate (102,103). The end result is often desertification. During the
past fifty years, an estimated 65 million hectares of once productive land
have thus been lost to desert along the southern edge of the Sahara alone
(104,105). Additional data for Africa are given in references (90,106).
As forest resources are lost, whether to agriculture, timber, brush fires,
or as fuelwood, villagers are increasingly forced to use lower quality
fuels such as crop wastes and dung to meet their minimum needs for cooking
and other purposes. Globally, an estimated 150 to 400 million tonnes of
cow dung are now burned annually. The burning of each tonne of dung
wastes enough nutrients potentially to produce an additional 50 kg of
grain. The cow dung now burned in India wastes nutrients equal to more
than one-third of the chemical fertilizer used (7).
Increasing use of agricultural residues for fuel may cause serious damage
to soils. Organic matter in soils provides most of the nitrogen and sulfur
and as much as half the phosphorus needed by plants. It increases the
cation exchange capacity of the soil, binding important minerals such as
magnesium, calcium, potassium and ammonium that would otherwise be leached
away. It buffers the pH of soils, and it improves the water retention and
other physical characteristics (151).
Reference (31)
The destruction of forests may also have serious consequences in terms of
loss of genetic resources, loss of potential new medical products, and
others. These are reviewed in reference (5).
The burning of biomass fuels has serious environmental impacts due to the
smoke released (107-112). Although there have been numerous anecdotal
accounts of ill health associated with indoor biomass combustion, only
recently have systematic scientific studies of the problem begun (112).
Results to date indicate that in village homes, indoor concentration of
carbon monoxide, particulates, and hydrocarbons can be 10-100 and more
times higher than World Health Organization (WHO) Standards (111).
Further, cooks using traditional biomass burning stoves can be exposed to
far more carbon monoxide, formaldehyde, carcinogenic benzo(a)pyrene, and
other toxic and carcinogenic compounds than even heavy cigarette smokers.
From this it is expected that smoke is a significant factor in ill-health
in developing countries. The diseases implicated range from bronchiolitis
and bronchopneumonia to chronic cor pulmonale to various forms of cancer
(110,111). Indeed, the WHO now cites respiratory disease as the largest
bsex21.gif (600x600)
cause of mortality in developing countries (112). Table 16 lists air
pollution emission factors for a variety of fuels and combustion systems.
Reducing and controlling exposure to biomass fuel emissions must be a
primary consideration in any stove program. Further information is
available from the East-West Center (Appendix J).
ECONOMICS AND POLICY OPTIONS
The growing fuelwood shortage has a variety of economic impacts on both
rural and urban dwellers, the rural labor force, and the national economy.
For the rural subsistence dweller, depletion of local fuelwood resources
means ever longer foraging times. There are numerous estimates of these
times ranging as high as 200-300 person days per year per household in
Nepal or 7% of all labor (22,46,98) and similarly high labor rates in
Tanzania (59) and other countries (99). Approximate correlations relating
foraging distance to the local population density are easily developed by
equating the average consumption by a population to the area required to
provide a sustained yield, as shown in note (114). A second example is
given in reference (115). In arid regions with a low biomass growth rate a
village of as few as 500-1000 people can use up all the fuelwood within a
walking distance. Foraging is also heavy work; in Burkina Faso, typical
headloads weigh 27 kg (113).
When wood becomes scarce, crop wastes and dung are the villagers' only
alternative; there is no cash for commercial fuels, nor do the long-term
environmental costs of using agricultural wastes outweigh their immediate
value as fuel. In India, it has been estimated that a tonne of cow dung
applied to the fields would result in increased grain production worth
US$8, but if burned would eliminate the need for firewood worth $27 in the
market (116,117). Some have argued that due to the relatively low efficiency
of cow- dung in providing nutrients such as nitrogen, phosphorus,
potassium, and zinc to the soil in a useable form, it makes better sense
to burn it (117). This, however, ignores other important contributions of
organic materials to soil fertility (151).
With a high market value for biomass fuels, the poor and landless are
sometimes denied access to their traditional fuel sources (118). It has
even been reported that farm laborers in Haryana, India, formerly paid
cash wages, are sometimes instead paid crop residues to be used for fuel
(99) -- fuel they previously received free.
In contrast, urban dwellers often have no choice but to purchase their
fuel. Again, there are numerous estimates of the financial burden this
imposes ranging up to as high as 30% of total family income in Ouagadougou
(34), to 40% in Tanzania (39), to nearly half in Bujumbura, Burundi (36).
During the 1970s the cost of wood and charcoal increased at a rate of 1-2%
per year faster than other goods (76). Due to their rapid price escalation
during the 1970s, fossil fuels are often not viable alternatives. In
Malawi, the use of kerosene declined 24% between 1973 and 1976, allegedly
due to higher prices (34). Others have noted similar impacts (71).
The use of traditional fuels is important in stimulating the rural
economy. The value of fuelwood and charcoal exceeds 10% of the Gross
Domestic Product in countries such as Burkina Faso, Ethiopia, and Rwanda,
and exceeds 5% in Liberia, Indonesia, Zaire, Mali, and Haiti (76). This
pumps large amounts of cash into the rural economy and provides much
needed employment to rural dwellers (Table 17). To supply Ouagadougou with
wood during 1975, for example, required some 325,000 person-days of labor
and generated over $500,000 in income directly and an additional $2.5
million in income through transport and distribution (34). In Uganda, an
estimated 16 tonnes of charcoal are produced per person-year (13). Other
estimates are given in Table 18 and references (71,72). In many countries,
people in the poorest areas, where conditions do not permit
expansion of crop or animal production and the natural woody vegetation is
the only resource, depend heavily on sales of firewood for their income
(34,99). Whatever program is put in place to meet the fuelwood shortage,
it will be necessary to take the employment impacts into account.
Alternatives
To meet the growing fuelwood shortage (Table 9), governments could import
fossil fuels as a substitute; plant fast-growing trees and improve the
management of existing forests; and develop more fuel efficient stoves and
other woodburning equipment, among other actions.
If every person now using fuelwood switched to petroleum based fuels, the
additional consumption would be just 3.5% of 1983 world oil output. The
cost of kerosene and liquified petroleum gas (LPG) for all household needs
would be 15% of total merchandise exports or less for Kenya, Thailand,
Zimbabwe, and many other countries. Importing fuels for cooking may then
be an important response in such areas (152).
In contrast, for Niger, Burundi, and others, a switch to petroleum fuels
for household energy needs would absorb almost all merchandise export
earnings (152). Efforts to stimulate use of butane gas through subsidies
have begun in West Africa but have proven to be a heavy financial burden
(34,119). There is also evidence that such subsidies benefit the wealthy
far more than the poor. In West Sumatra in 1976, the poorest 40% of the
population used only 20% of the kerosene even though it was heavily
subsidized (58). Yet without such subsidies, petroleum fuels are beyond
the reach of the poor. In these areas, other actions are needed.
As a second response, plantations of fast-growing tree species can be
developed to provide fuel (123-126). Extensive data on species, their
growth patterns, and their uses are given in references (5,12,102,123,124)
Donor agencies are now spending some $100 million per year on forestry
projects (116), and additional large funding is provided by the national
governments themselves. The U.N., however, has estimated that $1 billion
per year is needed to meet the minimum needs of the year 2000 when a
shortage of about 1 billion cubic meters per year is expected with no
intervention (6). To keep this sum in perspective, however, it must be
compared to the $130 billion per year needed for all energy sector
development in developing countries (154).
Plantations can provide rural employment (115) of some 150-500 person-days/hectare
during the first three years and almost twice that amount
during harvesting (127). Additionally, plantations and planting trees
generally can provide very important environmental benefits. Among these
are stabilizing and protecting soils from wind and water erosion, providing
protection to birds (which may eat crop-destroying insects -- or
the crops themselves) and other animals, and providing important soil
nutrients. These are reviewed in (155).
Monocropping plantations, however, ignore the many traditional non-fuel
uses of forests such as food, fiber, medicines, and others (128). Some
fast-growing species such as Eucalyptus, though productive and hardy, may
also deplete ground water supplies and soils, be inedible as livestock
fodder, and impede neighboring crop growth (5,99). For other species,
however, interplanting with crops can be valuable. Acacia albida can
increase yields of millet and sorghum by up to 3-4 times by fixing nitrogen
and by pumping other nutrients from deep within the soil. Additionally
it provides large amounts of cattle fodder during the dry season
(102). Other valuable species include the Tamarisk, used in southern Iran
to control salinity (129).
Some countries have begun to develop substantial plantations. Brazil, for
example, has successfully planted 5 million hectares, mostly fast-growing
Eucalyptus, for fuel and pulp since 1970 (67). In contrast, in Tanzania
an estimated 200,000 hectares of plantation were needed in 1983 to meet
the country's needs, but only 7300 were to be planted (47). Substantial
progress is being made, despite sometimes high costs -- over $1000 per
hectare in places, yields that have sometimes been far below expectations
(127,130), and numerous other problems (5,99,116,125,131,132,155). In
parts of Kenya, for example, individual woodlots are now being established
bsex25.gif (600x600)
widely (140). In Table 19 several fossil and renewable fuels are compared
on the basis of their cost and the performance of the stoves used with
them. As seen there, fuelwood is far less expensive than petroleum based
fuels or other renewable energy options. Although this cost advantage
will decrease in arid regions, it will likely still be significant.
Village woodlots may further reduce the cost of fuelwood (Note 157-C).
Thus, wood will be a primary energy source in developing countries for the
foreseeable future.
As a third response, improving the efficiency with which biomass fuels are
used could greatly extend forest resources and at a very low cost. In
this case, the cost advantage of wood as a cooking fuel becomes even more
apparent (Table 19). The importance of the results shown in Table 19
cannot be overemphasized. No other energy resource comes close to the
cost advantage of wood used in fuel efficient stoves. Certainly, as
incomes rise the cleanliness and convenience of higher quality fuels such
as kerosene, LPG, or ethanol will be gladly paid for; but this is not now
a viable option for many of the world's poor. Thus, a significant effort
must be focused on the development of stoves that burn wood, but do so
cleanly and safely, with high efficiency, and that are easily controlled.
The cost of saving energy by using an improved stove can also be compared
to the cost of producing fuelwood. A typical household of eight people
who use fuelwood for cooking on a traditional stove (thermal efficiency of
17%) at a rate of 300 watts/person will consume about 150 GJ of energy in
a two-year period. Alternatively, if this same household did their
cooking on two $3 improved channel-type woodstoves, which have observed
fuel savings of 30-40% in the field (thermal efficiency of 30%, Chapter
V), they would only consume 90-105 GJ over the two-year life of these
stoves. The energy savings would be achieved at a cost of just $0.10-0.13/GJ
-- a factor of 10 less than the cost of plantation produced
fuelwood (Table 19). The energy needed to produce these stoves does not
change this result. Currently, 0.022-0.027 GJ/kg is needed to produce
steel from raw ore and new industrial processes could reduce this to
0.009-0.012 GJ/kg (136). A typical stove might use 2-3 kg of steel and
thus require 0.1 GJ to produce while saving 25 GJ or more over its
lifetime.
Comparing these options in this manner is not intended to argue that
improved stoves are a substitute for planting trees. Both are needed now
and both are important components of any longer-term energy strategy.
The cost of providing such fuel efficient stoves to every family on earth
now using biomass fuels for cooking would be less than a typical 1 GW
nuclear power plant, yet save some 10-20 times as much energy each year as
the reactor would produce during its entire lifetime (153). The design,
production, and dissemination of low-cost, fuel efficient biomass stoves
and other technologies are the subjects of the following chapters.
CHAPTER III
STOVE DESIGN
In this chapter the basic physical principles of combustion and heat
transfer will be applied to the design of cookstoves burning raw biomass
fuels such as wood and agricultural wastes and guidelines for improving
their efficiency will be developed. These guidelines form the basis for
the development of highly fuel efficient stoves. These are, however,
guidelines only. To determine accurately the effects on performance of
various design modifications and to optimize a design requires painstaking
testing as described in Chapter V. The actual combustion and heat transfer
processes occurring in a stove are too complicated, too highly interdependent,
and too variable to model and predict easily. Testing is a must.
To begin understanding how to improve the performance of a stove, both the
theoretical limits as well as the current practical limits to stove
performance must be understood. The theoretical limits are examined first.
Consider, for example, cooking rice or porridge. As shown in Table 1,
heating the appropriate amounts of dry grain and water to boiling and
inducing the necessary chemical reactions requires, in this ideal case,
the equivalent of about 18 grams of wood per kilogram of cooked food. Yet,
controlled cooking tests with the open fire have required some 268 grams
of wood per kilogram of food cooked and even improved metal stoves have
used some 160 grams -- nine times the theoretical requirement. (Chapter V
and reference 2).
To determine where the rest of this energy is lost requires detailed
experimental work, including monitoring stove wall temperatures, flue gas
temperatures and volumes, and emissions, and has only been done in a few
bse1x290.gif (600x600)
special cases (3-5). Some of these are sketched in Figure 1 below.
(**) For wood with a calorific value of 18 MJ/kg.
References (1,3).
From these heat balances, several observations can be made.
o Generally the largest loss, 14-42% of the input energy, is by beat
conduction into and through the walls. In massive stoves
bse1c290.gif (486x486)
bse1bx29.gif (486x486)
stove (Figure lb) it is conducted through and lost from the outside
bse1a.gif (388x432)
surface.
o The loss of energy in hot flue gas accounts for some 22-39% of the
total input to the woodstove. The energy efficiency of a stove can be
dramatically increased by making use of the energy in this hot flue gas
through improved convective heat transfer to the pot.
o Although not explicitly detailed in Figure 1a, in open fires radiant
bse1a.gif (353x437)
heat transfer is the mechanism for two-thirds of the heat transfer to
the pot and cannot be greatly increased (7).
o The energy losses due to incomplete combustion are relatively small,
typically less than 8% of the input energy. The greater problem with
incomplete combustion is the emission of poisonous carbon monoxide and
hydrocarbons -- many of which are toxic, even carcinogenic (8).
o Typically half the energy entering the pot is lost in the form of steam
bse1a290.gif (281x432)
losses also occur in getting that energy into the pot. Eliminating this
steam loss by more carefully controlling the fire could, in principle,
reduce total energy use by half. Similarly, convective heat losses from
the surface of the pot are quite important (Figure 1d). For typical pot
bse1dx30.gif (437x486)
loss rates of 700 W/[m.sup.2] (42,43), a 28-cm-diameter cylindrical pot with
10-cm exposed to ambient air will lose energy at the rate of 100 W.
Over an hour, this is energetically equivalent to 20 grams of wood.
FIGURE 1: Heat Balances In Cooking Stoves
Figure 1a: Traditional Open Fire
Final Energy Balance:
Gains:
8% absorbed by water and food
Losses:
10% lost by evaporation from pot
82% lost to environment
Reference (6)
Figure 1b: Two pot uninsulated metal
wood stove with chimney.
Final Energy Balance:
Gains:
17.6% absorbed by first pot
10.3% absorbed by second pot
the fraction lost by evaporation
from pots is unknown
Losses:
2% absorbed by stove body
40.4% lost by convection and radiation
from stove body
22.2% lost as thermal energy in
flue gases
7.8% lost due to incomplete combustion
Reference (5)
Figure 1c: Two pot massive wood
stove with chimney.
Final Energy Balance:
Gains:
11.8% absorbed by first pot
3.6% absorbed by second pot
Losses:
29.2% absorbed by stove body
1.9% lost by convection and radiation
from stove body
39.0% lost as thermal energy in
flue gases
2.7% lost due to incomplete combustion
11.8% unaccounted for
Reference (5)
Figure 1d: Three pot mass wood
stove with chimney.
Final Energy Balance:
Gains:
6% absorbed by water and food
Losses:
4% lost by evaporation from pots
2.1% lost from pot surfaces
13.9% absorbed by stove body
30.2% lost as thermal energy in
flue gases
1.1% lost as carbon monoxide
1.9% lost to evaporate moisture in
fuel
5.9% lost as latent heat of vaporization
of water produced
by combustion
11.% lost as charcoal residue
Reference (3)
Figure 1e: Thai charcoal stove.
Final Energy Balance:
Gains:
3.1% absorbed by water and food
Losses:
4.6% lost by evaporation from pot
0.2% lost by convection and
radiation from pot lid
13.0% absorbed by stove body
1.3% lost by convection and radiation
from stove body
2.1% lost as thermal energy in
flue gases
0.7% lost as carbon monoxide due
to incomplete combustion
75.% lost in the conversion of
wood to charcoal
Reference (4)
Improving the fuel efficiency of a stove thus requires attention to a
number of different factors. Among these are:
Combustion Efficiency: so that as much of the energy stored in the combustible
as possible is released as heat.
Heat Transfer Efficiency: so that as much of the heat generated as
possible is actually transferred to the contents of the pot. This
includes conductive, convective, and radiative heat transfer processes.
Control Efficiency: so that only as much heat as is needed to cook the
food is generated.
Pot Efficiency: so that as much of the heat that reaches the contents
of the pot as possible remains there to cook the food.
Cooking Process Efficiency: so that as little energy as possible is
used to cause the physico-chemical changes ocurring in cooking food.
The combustion and heat transfer efficiencies are often combined for
convenience and are then termed the thermal efficiency of the stove. When
they are also combined with the control efficiency, the three together are
termed the stove efficiency. Different tests measure different combinations
of these factors. High power water boiling tests, for example,
measure the thermal efficiency. High/low power water boiling tests and
controlled cooking tests are two different methods of measuring the stove
efficiency.
The heat transfer efficiency will be discussed first in terms of the
conductive, convective, and radiative processes going on in and around the
stove. These processes are sketched in Figure 2. The other aspects of
bse2x32.gif (600x600)
efficiency will be discussed in turn. The appendixes document the text in
detail and provide extensive references for further reading.
CONDUCTION
The temperature of a solid, liquid, or gas is a measure of how rapidly the
atoms and molecules within it are moving: the faster they are moving the
hotter the substance is. In gases and liquids, conductive heat transfer
occurs when high velocity molecules randomly collide with slower molecules,
giving up some of their energy. In this way, heat is gradually
transferred from higher temperature regions to those at lower temperatures.
Because of their low density and the consequent low collision rate
between molecules, gases have a low thermal conductivity. High quality
insulators take advantage of this by trapping millions of miniscule air
pockets in a matrix of (very porous or spongy) material: most of such
insulators is in fact air. The solid material is there only to hold the
air in place -- to prevent currents of air that would increase the heat
transfer rate. Thus, such insulators lose some of their insulating value
if they are compressed, which reduces the size of the air pockets, or get
wet, which fills the air pockets with higher conductivity water.
Reference (9). A more complete table is given in Appendix A.
In a solid, heat is conducted as more rapidly vibrating atoms excite and
speed up the vibration rate of more slowly moving neighbors. Additionally,
in metals heat is conducted as free electrons with a high velocity move
from regions at a high temperature into regions at a lower temperature
where they collide with and excite atoms. In general, heat conduction by
such electrons is much more effective than that by adjacent atoms exciting
each other. For this reason, metals (which conduct electricity) have much
higher thermal conductivities than electrically insulating solids.
A brief table of thermal conductivities and other factors is presented in
Table 2 above. The points just made about the low conductivity of gases,
the high conductivity of metals, and quality insulators being mostly air
(notice the low density) can be clearly seen in this table.
Calculating Thermal Conductivity
bse3x33.gif (317x317)
The thermal conductivity of an object can
be expressed approximately by the equation
kA([T.sub.1] - [T.sub.2])
Q= --------------------------- (1)
s
where Q is the rate of heat transfer, k is
the thermal conductivity of the material,
A is the area, s is the thickness of the
object across which heat is being conducted,
and ([T.sub.1-[T.sub.2]) is the temperature difference
between the hot and cold sides. Thus, we see that if the plate is
large and thin (A/s large) the rate of heat tranfer will be large. If the
plate is small in area and thick, more like a rod (A/s small), the rate of
heat transfer will be small. The heat transfer also varies directly with
the thermal conductivity and the temperature difference across the object
(Appendix A).
However, using this equation alone for the heat transfer across a stove
wall would lead to values that are many times too large. The heat transfer
into and out of an object depends on the conductivities to and from the
surfaces as well as the conductivity within the object itself (Appendix
A). In some cases, dirt or oxide layers may reduce the heat transfer
across the surface; in other cases, the air at the surface itself significantly
reduces the heat transfer. Taking this into account then gives
A([T.sub.1] - [T.sub.2])
Q = ------------------------
1 s 1
- + - + -
[h.sub.1] k [h.sub.2] (2)
where [h.sub.1] and [h.sub.2] are the inner and outer surface heat transfer coefficients
(Appendix B). Typical values for h are 5 W/[m.sup.2][degrees]C in still air to over 15
W/[m.sup.2][degrees]C in a moderate 3 m/s wind. The inverse values 1/h and s/k are the
thermal resistances to heat transfer. Typical values of the thermal
resistances (s/k) for different stove walls are 0.0000286 [m.sup.2][degrees]C/W for 1-mm-thick
steel, 0.04 [m.sup.2][degrees]C/W for 2-cm-thick fired clay, and 0.10 [m.sup.2][degrees]C/W for a
10-cm-thick concrete wall. In contrast, the thermal resistance of the air
at the surface of the stove wall (1/h) is 0.2 [m.sup.2][degrees]C/W for still air and
0.0667 [m.sup.2][degrees]C/W for a 3 m/s wind. These values must then be doubled to
account for both the inside and outside surfaces.
Thus, it is the surface resistance, not the resistance to heat transfer of
the material itself, that primarily determines the rate of heat loss
through the stove wall. This is true until very low conductivity (high
thermal resistance) materials such as fiberglass insulation are used.
Fiberglass, for example, has a thermal resistance (1/k) typically about 25
m[degrees]C/W or, for a 4-cm-thick lining, a total resistance (s/k) of about I
[m.sup.2][degrees]C/W. In this case the insulation, not the resistance of the surface
air layers, is the primary determinant of the stove's rate of heat loss.
The steady state rate of heat loss through a metal stove wall can now be
crudely estimated. If the wall has an area of 1mx0.2m-0.2[m.sup.2], a temperature
difference of 500[degrees]C between the inside and outside, and is in still air
(.2)(500)
Q= ------------------------ = 250 watts
(.2) + (0.0000286) + (.2)
If the resistance of the surface boundary layer of air had been ignored, a
rate of heat loss 14,000 times greater would have been calculated -- an
absurdly large value.
Conductive heat transfer also carries heat through the pot to its contents.
High conductivity aluminum pots can save energy compared to clay
pots because they more readily conduct the heat of the fire to the food.
At the same time, however, aluminum pots will suffer greater heat loss
than clay pots from the warm interior to the portions of the exterior exposed
to cold ambient air. These portions of the pot could be insulated to
reduce this heat loss. The overall heat transfer coefficient of aluminum
pots has been estimated to be about 18 W/[m.sup.2][degrees]C compared to 9.7 W/[m.sup.2][degrees]C for
clay pots (3,10). In controlled cooking tests with aluminum pots, fuel
savings were about 45% (3) compared to using clay pots. Coating aluminum
pots with mud to protect their shine, or allowing a thick layer of soot to
build up on the outside reduce the pots' energy efficiency and should be
discouraged. In addition to their high performance and ease of use cooks
prefer aluminum pots because, unlike traditional fired clay pots, they
won't break. In a very few years the production and use of aluminum pots
has spread widely in many developing countries.
Calculating Thermal Storage
Another factor of importance in conductive heat transfer calculations is
the ability of a material to store thermal energy, measured as its
specific heat. The specific heat of a material is the amount of energy
required to raise the temperature of 1 kg of its mass by 1[degrees]C. For a given
object, the change in the total heat stored is then given by
dE - [MC.sub.p](dT) (3)
where M is the object's mass, [C.sub.p] is its specific heat, and (dT) is its
change in temperature. Thus, if the wall of a 3 kg metal stove increases
by 380[degrees]C during use, the change in energy stored in its wall is
dE = (3kg)(480Ws/kg[degrees]C)(380[degrees]C) = 547200 Ws or 547.2 kJ
Thus, the thermal conductivity carries thermal energy through a material;
the specific heat and mass of an object store this heat energy. The
larger the mass and specific heat of an object the more energy it can
store for a given change in temperature. Thus a thermally massive (large
[MC.sub.p]) object warms up slowly; a thermally lightweight (small [MC.sub.p]) object
will warm rapidly. This is called the thermal inertia of an object and is
an important design parameter in stoves.
Wall Loss Calculations
Reducing the heat loss into and through the stove walls to the outside
requires a detailed analysis of the conduction process, which is presented
in Appendix A. In reviewing these calculations, it is important to note
first that they are based on a particular assumed combustion chamber
geometry and heat flux from the fire. Because of this, the values listed
below are in watts, degrees, etc., rather than in dimensionless units.
Second, for simplicity and convenience the calculations were done assuming
that the fire is kept at a single power level all the time. Thus, the
results listed are intermediate between those observed in practice for the
high power boiling phase and the low power simmering phase due to the
assumed values for the heat fluxes. Although the values given are shifted
by these factors, they nevertheless show trends that will remain the same
for any combustion chamber.
When cooking begins, the walls of the stove are cold. With time they warm
up at a rate determined by their mass and specific heat as discussed
above. Lightweight walls have a low thermal inertia and warm quickly.
Thick, heavy walls warm more slowly. Heat loss from the combustion chamber
is determined by how quickly these walls warm and subsequently how much
heat the wall loses from its outside surface. This is shown clearly in
bse4x37.gif (600x600)
Figure 4, where the thicker the wall the more slowly it warms.
Although a thick wall of dense high specific heat material may have
slightly lower heat loss than a thinner wall after several hours (See
Appendix A), it takes many hours more for the eventual lower heat loss of
the thick wall to compensate for its much greater absorption of heat to
warm up to this state. Thus, it is always preferable to make the solid
(non-insulator) portion of the wall as thin and light as possible.
Additionally, the use of lightweight insulants such as fiberglass or
bse4bx37.gif (486x486)
double wall construction can dramatically lower heat loss (Figure 4B).
Materials such as sand-clay or concrete, which have a high specific heat
and density, and which must be formed in thick sections to be sufficiently
strong to support a pot or resist the fire, should therefore be avoided.
Heat Recuperation
It has frequently been argued that the large amounts of heat absorbed by
the walls of a massive stove should be utilized by either extinguishing
the fire early and using this heat to complete cooking or by later using
it to heat water. Water heating tests on hot massive stoves, however, have
shown that only 0.6-1.3% of the energy released by the fire, of which
perhaps one-third was stored in the massive wall, could be recuperated -- heating
the water by typically 18-19[degrees]C (2). What is often thought to be
heating or cooking by heat recuperation is actually done by the remaining
coals of the fire.
That heat recuperation from massive walls is so difficult can be easily
understood by considering the following. First, heat conduction through
the wall is slow (Appendix A) so that little energy can be transported to
the pot directly. Second, air is a relatively good insulator. Thus, little
heat can be carried from the wall into the air space inside the stove and
then to the pot. Third, both of these heat paths are further slowed by the
relatively small temperature difference between the wall and the pot. The
low temperature of the wall also reduces the radiant transfer to the pot.
Finally, the heat stored in the wall tends to equilibrate within the wall
and then leak to the outside. The result of all these processes is shown
bse6x39.gif (600x600)
in Figure 6 and agrees very well with the experimental data cited above.
Rather than depending on low efficiency massive stoves (Table V-1) for
cooking and then attempting to recuperate heat for hot water, such water
heating can be much more efficiently done directly with a high performance
stove. Further, it can then be done when needed rather than being tied to
the cooking schedule. Similarly, using stored heat to complete cooking is
an extremely inefficient technique compared to using a high efficiency
lightweight stove and possibly a "haybox" cooker (discussed below under
OTHER ASPECTS).
Heat recuperation is clearly desirable, however, when it can be done
efficiently, cost effectively, and without excessively interfering with
the primary purpose of the device. For example, heating water by heat
recuperation might be efficiently done by forming the wall of a high
performance metal stove itself into a water tank. Heat that would otherwise
be lost into and through the wall would then instead be directly
absorbed by the water. Whether or not the lower average combustion
chamber temperatures would significantly reduce the pot heating efficiency
or interfere with combustion would need to be tested.
Thus, lightweight walls have the intrinsic potential for much higher
performance than massive walls due to their lower thermal inertia. This
does not, however, necessarily mean that a lightweight stove will automatically
save energy or that a massive stove cannot. For a lightweight
stove to save energy its heat loss to the exterior must also be minimized
and the convective and radiant heat transfer to its pot must be optimized.
Conversely, massive stoves can and sometimes do save energy despite their
large wall losses. Such stoves can save energy if the convective and
radiative heat transfer to the pot is carefully optimized.
Reducing Wall Losses
If a lightweight single wall (metal) stove is heavily tarnished and sooted
bse5x39.gif (600x600)
on the outside its exterior heat loss can be quite large (Figure 5). This
heat loss is due to the emission of radiant energy (see Appendix C) and
can be reduced by chemically or mechanically polishing or coating the
exterior surface to leave a bright metallic finish. Although such a finish
may have commercial appeal, its effectiveness in reducing heat loss will
last only so long as it is kept relatively clean and free of soot and
rust, etc. It should be noted that most paints, even white paint, will
actually increase the radiant heat loss from a stove and should be
avoided; to decrease radiant heat loss, the surface must be metallic.
Lighweight single wall stoves are easy to construct, are low cost, and
have relatively high performance when convective heat transfer is optimized.
However, during use they can be quite hot on the outside and can
bsex40.gif (600x600)
burn the user as well as be uncomfortable to use (Table 3). To reduce heat
loss and thus reduce this hazard, either double wall construction and/or
lightweight insulants such as fiberglass or vermiculite can be used.
Double wall construction with metal alone can significantly reduce heat
loss (Figure 5), user discomfort, and the hazard of burns (Table 3). The
double wall serves two functions in reducing heat loss. First, the dead
air space between the two walls is a moderately good insulator. It should
be noted, however, that increasing the thickness of this dead air space
does not improve its insulating value. This is due to the convection
currents, which flow more freely the larger the space, carrying heat from
one wall to the other. Second, the inner wall acts as a radiation shield
between the fire and the outer wall. Both of these factors can be seen in
Figure 5. There, the emissivity or, more accurately, the radiant coupling
between the inner and outer walls is the prime determinant of heat loss.
The exterior surface emissivity is less important due to the lower temperature
of that wall. As the temperature of the exterior wall increases due
to greater radiant heat transfer from inner to outer wall ([[epsilon].sub.i] increasing)
the exterior emissivity, [[epsilon].sub.e], becomes more important (Appendix C).
In practice there are several potential difficulties:
o Although it is preferable to minimize radiant coupling between the two
walls by giving them a bright, long-lasting metallic finish, they will
tend to rust, tarnish, and soot over time. Keeping them clean would be
difficult. Even in the worst case ([[epsilon].sub.1] = .9, [[epsilon].sub.e] .9), however, the double
wall still performs better than the best ([epsilon].sub.e] = .9) single metal wall.
o The dead air space is a good insulator on its own, but attaching the
inner wall to the outer will tend to short circuit its insulating value
due to the high thermal conductivity of metal. It is necessary that the
two walls together be mechanically rigid, but they should not easily
conduct heat from one to the other. This might be done by using nonmetallic
spacers or fasteners, or tack welding the walls together at
selected points. Long continuous welds should be avoided if possible.
o The insulating value of the dead air space is reduced if air is allowed
to flow through. Thus, the dead air space should be closed at the top.
Double wall metal stoves are now being developed and commercialized in
Botswana (11,12) and Guinea (13).
Better yet is to use a high quality insulant such as fiberglass or
vermiculite with the double wall to hold it in place and protect it. As
seen in Figure 5, layers of insulation as thin as a few millimeters are
effective in reducing heat loss. Such stoves have been tested in Mali
(14). Other lightweight insulants worth investigating include wood ash,
diatomaceous earth, and, possibly, chemically treated (to reduce its
flammability) straw or charcoal among others (see Table A-1).
Just as insulated walls reduce the exterior temperatures (Table 3), they
increase the inner wall temperature. This can increase heat transfer to
the pot by convective heat transfer, by radiative heat transfer from the
inner wall surface, and possibly by improving the quality of combustion.
CONVECTION
Convective heat transfer occurs when a gas or liquid is forced or flows
naturally into a region at a different temperature and then exchanges heat
energy by conduction - - by the interaction of individual particles. It is
by convective heat transfer that the hot gas leaving the fire heats the
pot, or that the wind cools a hot stove. In open fires and many traditional
bse1x290.gif (600x600)
stoves much of the heating potential of this gas is lost (Figure 1).
Increasing convective heat transfer to the pot is the single most
important way to increase the thermal efficiency of a woodburning stove.
Increasing Convective Heat Transfer
In general, convective heat transfer is given empirically by the equation:
Q = hA([T.sub.1]-[T.sub.2]) (4)
For the case of a pot being heated by hot gas leaving the fire, Q is the
heat transferred from the gas to the pot, h is the convective heat
transfer coefficient, A is the area of the pot across which the heat
exchange takes place, and ([T.sub.1]-[T.sub.2]) is the temperature difference between
the hot gas and the pot.
To increase the heat transfer Q to the pot there are then, in principle,
three things one can do. First, the temperature [T.sub.1] of the hot gas can be
increased. This can be done only by closing the stove and controlling the
amount of outside air that enters. This is often impractical as it
requires manipulating a door on the wood entry, prevents easy visual monitoring
of fire, and usually requires cutting the wood into small pieces so
that the door can be closed behind them. Further, the user must consistently
close the door. Stoves with enclosed fireboxes are, however, being
developed and disseminated in India (15-18). If successful on a large
scale, this is an important innovation.
Second, as much of the area A of the pot should be exposed to the hot gas
as possible. This is very important. The pot supports, for example, must
be strong enough to support the pot but should be kept small in area so as
not to screen the hot gas from the pot. The gas should be allowed to rise
up around the pot and contact its entire surface.
Third, the convective heat transfer coefficient h should be increased.
This can be done by increasing the velocity of the hot gas as it flows
past the pot.
In convective heat transfer, the primary resistance to heat flow is not
within the solid object (unless it is a very good insulator), nor within
the flowing hot gas. Instead, the primary resistance is in the "surface
boundary layer" of very slowly moving gas immediately adjacent to a wall.
Far from a wall, gas flows freely and readily carries heat with it. As the
pot wall is approached, friction between the pot and the gas prevents the
gas from flowing easily, Within this region, heat transfer is primarily by
conduction and, as previously noted, the conductivity of gases is quite
low. It is this surface boundary layer of stagnant gas that primarily
limits heat transfer from the flowing hot gas to the pot.
To improve the thermal efficiency of a stove, the thermal resistance of
this boundary layer must be reduced. This can be accomplished by (among
others) increasing the flow velocity of the hot gas over the surface of
the pot. This rapid flow helps "peel" away some of this surface boundary
layer and, thinner, the boundary layer of stagnant gas then offers less
bse7x43.gif (600x600)
resistance to conductive heat transfer across it to the pot (Figure 7).
Fundamental Stove Types
Efforts to improve convective heat transfer have resulted in three
fundamental types of biomass stoves, which will be generically termed
multipot, channel, and nozzle (Figure 8). In each of these, the flow
bse8x44.gif (600x600)
velocity of the hot gas over the pot is increased by narrowing the
channel(1) gap through which the gas must flow past the pot. (Because the
volume of hot gas flowing past any point is constant, its flow velocity
through a narrow gap must be faster than through a wider one). This,
however, results in a serious handicap inherent in any improved stove
program. As these channel gaps must be precise to within a few millimeters
to be effective, stove and pot dimensions must correspond and be precisely
determined - - greatly complicating both production and dissemination.
Multipot stoves heat two or more pots from a single fire. In principle,
this increases the pot surface area exposed to the fire and hot gas and
raises the thermal efficiency. In practice, however, it is difficult if
not impossible to individually control the heat input to each of the pots
(see OTHER ASPECTS). The resulting stove efficiency is then usually lower
than channel or prototype nozzle stoves now under development.
Channel stoves increase the pot area exposed to the hot gas by forcing the
gas over as much of the surface of a single pot as practicable. Radiant
transfer is maximized by placing the pot close to the firebed yet without
excessively interfering with the combustion. Channel stoves offer higher
______________________
(1) The channel dimensions are called "length" for the direction of gas
flow, "width" for the circumference of the pot or stove, and "gap" for the
space between the pot and stove walls.
efficiencies, better control, and lower cost than most multipot stoves.
Emissions from channel stoves, however, are often no less than from
multipot stoves and in some cases may be worse.
The development of nozzle type stoves has only recently begun (18,19), yet
they appear to offer considerable promise. As for channel stoves, nozzle
stoves have a single pot, the entire surface of which is exposed to the
f ire and hot gas. Similarly, as for both channel and multipot stoves,
nozzle stoves increase the velocity of the hot gases flowing past the pot
by forcing them through a narrow channel. Additionally, the large height
and the narrowing throat of the nozzle stove's combustion chamber accelerate
the gases to a higher velocity before they contact the pot. This is
done, however, at the expense of reduced radiant transfer.
Prototype nozzle stoves have achieved efficiencies of 43% in laboratory
tests (18,19), comparable to the best multipot stoves (15-17) and channel
stoves (14). Further, because the shape of the combustion chamber improves
combustion, nozzle stoves have much lower emissions than other types.
Recent tests of nozzle stoves have shown emissions of carbon monoxide (CO)
to be just 5-6 ppm at peak power and of soot, less than 2.5 mg/[m.sup.3] (18,19).
These are far less than the open fire. By comparison, typical emissions
from kerosene stoves at peak power are 25 ppm of CO and 0.2 mg/[m.sup.3] of soot.
Current prototypes, however, suffer the severe handicap of accepting only
very small pieces of biomass. Whether or not this can be overcome remains
to be seen(2).
______________________
(2) For further information, readers should contact H.S. Mukunda and U.
Shrinivasa at ASTRA (See Appendix J).
Modeling Convective Heat Transfer
Understanding convective heat transfer underpins all efforts to improve
the efficiency of biomass burning stoves. A detailed empirical model of
convective heat transfer in channel stoves is developed in Appendix B;
references to an empirical model of multipot stoves are also provided
there. Numerical analysis of convective heat transfer in channel and
nozzle stoves is now underway by the author and will be presented elsewhere.
Because channel stoves generally have much better performance than
multipot stoves and because they are more fully developed and tested than
nozzle stoves, critical elements in their design will be presented here.
The empirical model of convective heat transfer in channel stoves developed
in Appendix B provides considerable insight into their performance
and limitations. This model is not sufficiently precise to be used to
predict the absolute quantitative performance of a real stove -- that can
only be done by detailed testing as discussed in Chapter V. Nevertheless,
the model is useful in illustrating general trends in the performance of
this type of stove and its sensitivity to dimensional changes.
From the above discussion of convective heat transfer and surface boundary
layers one expects narrower channels to have higher rates of heat transfer
to the walls. This is clearly seen in the model predictions presented in
Figure 9. In fact, the channel efficiency, defined as the fraction of
energy in the hot gas entering the channel that is transferred to the pot,
is extremely sensitive to changes in the channel gap. For a 10-cm-long
channel, the channel efficiency drops from 46% for an 8-mm gap to 26% for
a 10-mm gap. Thus the stove and pot dimensions must be very precisely
controlled. Multipot and nozzle stove performance is similarly sensitive
to the channel gap.
The lower efficiency of wide channel gaps can be partially compensated for
bse9x46.gif (600x600)
by making the channel longer (Figure 9) or by closing the combustion
chamber to control excess air and thus raising the average gas temperatures
(Appendix B). However, closing the firebox is often not practical,
as discussed below under Radiation, and longer channels can seldom fully
bse9bx46.gif (486x486)
compensate (Figures 9,11). As seen in Figure 9B, additional channel length
bse9x460.gif (600x600)