United States EPA/600/R-00/052
Environmental Protection
Agency June 2000
&EPA Research and
Development
GREENHOUSE GASES FROM
SMALL-SCALE COMBUSTION DEVICES
IN DEVELOPING COUNTRIES: PHASE IIA
Household Stoves in India
Prepared for
Office of Air and Radiation
Prepared by
National Risk Management
Research Laboratory
Research Triangle Park, NC 27711
-------
FOREWORD
The U. S. Environmental Protection Agency is charged by Congress with pro-
tecting the Nation's land, air, and water resources. Under a mandate of national
environmental laws, the Agency strives to formulate and implement actions lead-
ing to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research
program is providing data and technical support for solving environmental pro-
blems today and building a science knowledge base necessary to manage our eco-
logical resources wisely, understand how pollutants affect our health, and pre-
vent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for
investigation of technological and management approaches for reducing risks
from threats to human health and the environment. The focus of the Laboratory's
research program is on methods for the prevention and control of pollution to air,
land, water, and subsurface resources, protection of water quality in public water
systems; remediation of contaminated sites and-groundwater; and prevention and
control of indoor air pollution. The goal of this research effort is to catalyze
development and implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed by EPA to
support regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regulations
and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by EPA's Office of Re-
search and Development to assist the user community and to link researchers
with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
EPA REVIEW NOTICE
This report has been peer and administratively reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
-------
EPA/600/R-00/052
June 2000
GREENHOUSE GASES FROM SMALL-SCALE COMBUSTION DEVICES IN
DEVELOPING COUNTRIES
Phase Ha
Household Stoves in India
by
Kirk R. Smith
Environment Program, East-West Center, Honolulu, HI 96848-1601 and
Environmental Health Sciences, University of California, Berkeley, CA
94720-7360
R. Uma, V.V.N. Kishore, K. Lata, V. Joshi
Tata Energy Research Institute, New Delhi 110003, India
Junfeng Zhang
Environment Program, East-West Center, Honolulu, HI 96848-1601 and
Environmental and Occupational Health Sciences Institute
Piscataway, NJ 08854
R.A. Rasmussen
Oregon Graduate Institute of Science and Technology
Beaverton, OR 97006
M.A.K. Khalil
Portland State University
Portland, OR 97207
EPA Cooperative Agreement CR820243-01
with the East-West Center, Honolulu, HI
EPA Project Officer: Susan A. Thorneloe
Atmospheric Protection Branch
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
-------
FOREWORD
Early in the 1990s, a pilot study was conducted in Manila, Philippines, to measure the
concentrations of a range of greenhouse gases from small-scale cookstoves burning biomass,
charcoal, kerosene and liquefied petroleum gas (Smith et a/., 1992; 1993). Based on intriguing
results, a more comprehensive study to characterize the emissions of non-CC>2 gases and other
pollutants from cookstoves using different solid, liquid, and gaseous fuels was undertaken in
China and India under a project organized by East-West Center (EWC) and funded by the US
Environmental Protection Agency (USEPA). The study focuses on more than two dozen of the
most common fuel/stove combinations in each nation. Since these countries contain more than
half of all stoves in developing countries, the stoves in this study represent a large fraction of the
combinations in use world-wide. In this report we describe the methodology and results of the
study undertaken in India. The monitoring took place in a simulated kitchen built at the Gual
Pahari Campus of the Tata Energy Research Institute (TERI), just outside New Delhi.
Laboratory analyses took place at TERI and at the Oregon Graduate Institute of Science and
Technology (OGIST).
ABSTRACT
This report presents a database containing a systematic set of measurements of the CO2,
CO, CH/j, TNMOC, N2O, SO2, NO2, and TSP emissions from the most common combustion
devices in the world, household stoves in developing countries. A number of different stoves
using 8 biomass fuels, kerosene, LPG, and biogas were examined - a total of 28 fuel/stove
combinations. Since fuel and stove parameters were monitored as well, the database also allows
examination of the trade-off of emissions per unit fuel mass, fuel energy, and delivered energy as
well as construction of complete carbon balances. Confirming the preliminary results in the
Manila pilot study, the database shows that solid biomass fuels are typically burned with
substantial production of PIC (products of incomplete combustion). In addition, as has often
been shown in the past, biomass stoves usually have substantially lower thermal efficiencies than
those using liquid and gaseous fuel. As a result, the emissions of CO2 and PIC per unit delivered
energy are considerably greater in the biomass stoves. In general, the ranking follows what has
been called the "energy ladder" from lower to higher quality fuels, i.e., emissions decrease and
efficiencies increase in the following order: dung-crop residues-wood-kerosene-gas. There are
variations, however, depending on specific stove designs.
-------
CONTENTS
Page
FOREWORD ii
ABSTRACT ii
LIST OF FIGURES iv
LIST OF TABLES v
GLOSSARY vi
ACKNOWLEDGEMENTS vii
I. Introduction and Summary 1
II. Conclusions and Recommendations 3
III. Methods 7
IV. Results 19
V. Discussion: National GHG Inventory and Fuel/ Stove Comparisons 41
VI. References 47
Appendix A: Simulated Rural Kitchen (SRK) 53
Appendix B: Details of Stoves Tested 59
Appendix C: Measurement Technologies 65
Appendix D: Calculation Procedures 68
Appendix E: Fuel Analyses 70
Appendix F: Measured Fluegas Concentrations 71
Appendix G: Error Analysis 82
Appendix H: Estimation of Indian Household Fuel Consumption 83
in
-------
LIST OF FIGURES
No. Title
1. GWC-full per MJ delivered: Mean values for each fuel 5
2. GWC-basic per MJ delivered: Mean values for each fuel 6
3a. Instant carbon balance: eucalyptus in improved vented ceramic
stove 17
3b. Carbon balance of char combustion after primary combustion 18
4. Power input for various fuel/stove combinations 24
5. ESI and instant combustion and heat transfer: along the household
energy ladder 27
6. Power input vs. efficiency 28
7. Major efficiencies and ESI by stove type: unprocessed biomass fuels 29
8. CC>2 emission factors: per MJ delivered to the pot 36
9. CO emission factors: per MJ delivered to the pot 37
10. CH4 emission factors: per MJ delivered to the pot 38
11. TNMOC emission factors: per MJ delivered to the pot 39
12. GWC-full per MJ delivered: along energy ladder 45
13. GWC-basic per MJ delivered: along energy ladder 46
A-l. Simulated rural kitchen (view from above) 55
A-2. Simulated rural kitchen (section A-A') 56
A-3. Simulated rural kitchen (section B-B') 57
A-4. Hood arrangement for stove with flue 58
B-l Photographs of the stoves tested in the study 61
B-2. Diagram of the traditional mud stove 62
B-3. Diagram of the three-rock stove 62
B-4. Diagram of the hara stove 63
B-5. Diagram of the angethi stove 63
B-6. Diagram of the kerosene pressure stove 64
F-l. Regression Analysis for CO2 : (TERI vs. OGIST) 75
F-2. Regression Analysis for CO : (TERI vs. OGIST) 76
F-3. Regression Analysis for CH4 : (TERI vs. OGIST) 77
IV
-------
LIST OF TABLES
No. Title Pa£
1. Fuel/stove combinations for gaseous and liquid fuels 11
2. Fuel/stove combinations for solid fuels 11
3. Instant emission ratios and nominal combustion efficiencies
(NCE) for all tests 20
4. Power input and thermal efficiency for gaseous and liquid fuels 22
5. Power input and thermal efficiency for solid fuels 23
6. Gross instant and ultimate carbon balances 30
7. Ultimate emissions by fuel mass on a pollutant mass basis (g/kg)
and on a carbon mass basis (g-C/kg) 32
8. Ultimate emission factors of pollutant mass by fuel energy content
(g/MJ) and delivered energy to pot (g/MJ-del) 33
9. Comparison of emission factors (g/kg) by fuel mass with results
from other studies 35
10. IPCC default (uncontrolled) emission factors for residential fuel
combustion (g/kg) 35
11. Coefficients of variation (COV) for measurements for 3 tests of each
fuel-stove combination 40
12. Weighted average emission factors and GHG emission from major
fuel/stove combinations in India (1990-91) 42
13. Inventory of GHG emissions from India (1990-91) 43
C-l. Analytic instruments used 65
E-l. Fuel chemical composition, moisture content, and net energy 70
F-l. Concentration of TSP and carbon as TSP 72
F-2. Concentrations of CO2, CO, and CH4 (ppm) in fluegas and indoor
background air (analyzed in TERI laboratory) 73
F-3. Concentrations of CO2, CO, CH4, TNMOC, and N2O (ppm) in fluegas
samples (analyzed by OGIST) 74
F-4. Comparison of TERI and OGIST CO2, CO, and CH4
concentrations (ppm) 78
F-5. Corrected fluegas and indoor concentrations (ppm) and resulting
net values for all relative fuel/ stove combinations 79
F-6. Background and concentrations of SO2 and NOX (ppb) 81
G-l. Error analysis 82
H-l. State list of rural households, penetration of improved stoves,
and biomass fuel consumption 85
H-2. Fuel consumption by stove type in India (million tons/year) 88
-------
GLOSSARY
Acacia
BIS
COV
EFbc
EFd
EFe
EFm
Emission ratio
EPA
ESI
Eucal
EWC
GHG
Gross carbon balance
GWC
GWPi
Kara
HTE
imet
Instant emissions
IPCC
IREP
ivc
ivm
Kero-pres
Kero-wick
KVIC
LPG
MJd
MNES
NCAEC
tree used as source of woodfuel in tests
Bureau of Indian Standards
coefficient of variation = (standard deviation)/(mean)
emission factor per burn cycle experiment
emission factor per MJ delivered to cooking pot
emission factor per unit net energy (MJ) of fuel
emission factor per unit mass (kg) of fuel
EFbcmolecular ratio of emitted specie (e.g., CO) to emitted CO2
U.S. Environmental Protection Agency
Environmental Stove Index
Eucalyptus, tree used as source of woodfuel in tests
East-West Center, Honolulu, HI
greenhouse gas (in this report: CO2 CH4, N2O, CO, TNMOC)
distribution of fuel carbon into gases, ash, char, and aerosol
global warming commitment = sum over i of GHG;*GWP;
global warming potentials in kg C as CO2 per kg C in GHG (20-
year time horizon)
CO2= 1.0, by definition
CO = 4.5 (IPCC, 1990)
CH4 = 22.6 (IPCC, 1995)
TNMOC = 12 (IPCC, 1990)
N2O = 290 (IPCC, 1995), on a molar basis with CO2
In the renewable case, 1.0 is subtracted from each (except N2O) to
account for the recycling of C as CO2 in photosynthesis.
Basic set - those with specified GWP in IPCC (1995)
Full set - those with specified GWP in IPCC (1990, 1995)
traditional unvented mud stove for use with dung
heat transfer efficiency = r|/NCE
improved metal stove (unvented)
from combustion of original fuel, with char left unburned
Intergovernmental Panel on Climate Change
Integrated Rural Energy Planing Programme
improved vented ceramic stove
improved vented mud stove
pumped kerosene stove (unvented)
simple wick kerosene stove (unvented)
Khadi and Village Industries Commission
liquefied petroleum gas contained in pressurized cylinders: butane
and propane
megajoule delivered to the cooking pot
Ministry of Non-Conventional Energy Sources
National Council for Applied Economic Research
VI
-------
NCE
OGIST
PIC
REDB
ren
SRK
TERI
3-R
Tg
tm
TNMOC
Tons
TSP
Ultimate emissions
nominal combustion efficiency = fraction of airborne carbon
emissions released as CC>2= 1/(1+K) see Eq. 2
Oregon Graduate Institute of Science and Technology, Beaverton
airborne products of incomplete combustion (CO, CH4, TNMOC,
TSP)
Rural Energy Database
renewable, as in GWC (ren)
simulated rural kitchen
Tata Energy Research Institute, New Delhi
traditional 3-rock stove (unvented)
teragram = 1012 g = one million tons
traditional mud stove (unvented)
total non-methane organic compounds (molecular weight taken as
18/carbon atom)
metric tons
Total Suspended Particulates
instant emissions plus emissions from burning leftover char
overall energy efficiency of a stove (Appendix D)
ACKNOWLEDGEMENTS
We wish particularly to thank A.G. Rao and S. Mande of TERI for assistance with sampling and
laboratory analysis as well as the staff members of the TERI Gual Pahari Campus workshop. In
addition, we appreciate the advice and support of R.K. Pachauri, Director of TERI. At OGIST,
we appreciate the help of D. Stearns, and at the University of California, David Pennise and
Sharon Gorman.
Vll
-------
I: INTRODUCTION AND SUMMARY
Household stoves, although individually small, are numerous and thus have the potential
to contribute significantly to inventories of greenhouse gases (GHG), particularly in those many
developing countries where household use is a significant fraction of total fuel use. In addition,
the simple stoves in common use in such countries do not obtain high combustion efficiency,
thereby emitting a substantial amount of fuel carbon as products of incomplete combustion (PIC)
- such as carbon monoxide (CO), methane (CH4 ), and total non-methane organic compounds
(TNMOC) - as well as carbon dioxide (CO2). This is true for fossil fuels, such as coal and
kerosene, but is particularly important for unprocessed biomass fuels (animal dung, crop
residues, and wood), which make up the bulk of household fuel use in developing countries.
Many greenhouse analyses of human fuel use assume that renewably harvested biomass fuels do
not contribute to global warming, i.e., have no global warming commitment (GWC), because the
released carbon is entirely recycled through photosynthesis in growing biomass that replaces the
burned biomass. Even under renewable harvesting, however, the gases released as PIC
contribute to global warming because of higher radiative forcing per carbon atom than CO2
(Hayes and Smith, 1994). Thus, such fuels have the potential to produce net GWC even when
grown renewably.
It is estimated that biomass combustion contributes as much as 20-50 percent of global GHG
emissions (Crutzen and Andreae, 1990; IPCC 1990). Though the major fraction of the emissions
is from large-scale open combustion associated with permanent deforestation, savannah fires, and
crop residues, combustion in small-scale devices such as cookstoves and space-heating stoves
also releases a significant amount of GHG. A more accurate estimation of emissions from
biomass combustion would require an inventory for GHG from different types of biomass
combustion as well as better estimates of amount of biomass burnt.
The emissions of non-CO2 greenhouse gases from small-scale combustion of biomass are not
well characterized (Levine 1996), but are known to be different from open large-scale
combustion, such as forest and savannah burning, which have been the focus of more research.
Emissions from other fuels as commonly used in developing-country households are also not
well known. Therefore, extensive measurements of emission factors for GHG from a range of
fuels and combustion devices would lead to removing some of the uncertainty in the estimates of
total emissions from biomass combustion and also will provide a baseline database to understand
the potential for reduction in GHG emissions due to various mitigation measures, such as fuel
switching, in the household sector.
A pilot study was conducted in Manila, Philippines to measure the concentrations of a range of
GHG from small-scale cookstoves burning biomass, charcoal, kerosene and liquefied petroleum
gas (LPG) (Smith et al. 1992; 1993). The results indicate that the emission factors for CH4, CO,
and TNMOC from the combustion of wood and charcoal in cookstoves are high. In the case of
wood combustion, the analysis also revealed that, the global warming commitment (GWC) of the
non-CO2 GHG - CO, CH4 , and TNMOC - may in some circumstances rival or exceed that from
CO2 itself. In addition, the study seemed to indicate that in some instances substitution of
-------
biomass by fossil fuels, such as kerosene and gas, could be considered as means to lower GWC,
even when the biomass fuel is harvested renewably. If verified, these would have important
implications in setting energy and global-warming policies.
To explore these tentative findings further, a series of more detailed measurements were
undertaken in India. A total of 28 fuel/stove combinations in common Indian use were
successfully tested for a range of GHG and other emissions while simultaneously being
monitored for fuel, thermal efficiency, and other parameters.
-------
II. CONCLUSIONS AND RECOMMENDATIONS
This database contains a systematic set of measurements of the CC>2, CO, CH/j, TNMOC,
N2O, 862, NO2, and TSP emissions from the most common combustion devices in the world,
household stoves in developing countries. A number of different stoves using 8 biomass fuels,
kerosene, LPG, and biogas were examined - a total of 28 fuel/stove combinations. Since fuel
and stove parameters were monitored as well, the database also allows examination of the trade-
offs of emissions per unit fuel mass, fuel energy, and delivered energy as well as construction of
complete carbon and mass balances.
Confirming the preliminary results in the Manila pilot study (Smith et a/., 1992, 1993), the
database shows that solid biomass fuels are typically burned with substantial production of PIC
(products of incomplete combustion). Some fuel/stove combinations diverted more than 20% of
the fuel carbon into PIC. No biomass stove produced less than 5%. In addition, as has often
been shown in the past, biomass stoves usually have substantially lower thermal efficiencies than
those using liquid and gaseous fuel. As a result, the total CO2 and PIC emissions per unit
delivered energy are substantially greater in the biomass stoves. In general, the ranking follows
what has been called the "energy ladder" from lower to higher quality fuels, i.e., emissions
decrease and efficiencies increase in the following order: dung-crop residues-wood-kerosene-
gas. There are important variations, however, depending on the specific stove designs.
The global warming commitment (GWC) of the fuel/stove combinations depends on which PIC
gases are included in the calculations and whether the biomass fuels are considered to be
renewably harvested. (Crop residues, dung, and biogas - which is made from dung - are assumed
always to be renewable; LPG and kerosene are always non-renewable.) In the non-renewable
case, because of their low efficiencies and high PIC emissions, all biomass stoves produce
substantially more total GWC per unit delivered energy than the kerosene and LPG stoves, of
which LPG is best. If GWC from only CO2 , CH4, and N2O are considered (Basic GHG Set), a
few of the crop residue and dung stoves are comparable to kerosene. In the renewable basic set,
about half the biomass fuel/stove combinations produce less GWC than kerosene. If the GWP
of all PIC are included (Full = Basic set plus CO and TNMOC)1, a few wood and rootfuel stoves
are comparable to kerosene, but no others. Interestingly, however, biogas is by far the best of all,
with only some 10% of LPG GWC and more than a factor of 100 less than the most GWC-
intensive solid biomass fuel/stove combinations.
For a complete analysis, the GWC of the rest of the fuel cycles should be included as well. The
fossil fuels, for example, will have GHG releases at the oil well, refinery, and transport stages of
the fuel cycle (Schlamadinger, et al. 1997). Biogas will lose some of its apparent lead because of
CtL; leaks from the digester and pipelines, although preliminary measurements indicate that these
are relatively small (Khalil etal., 1990). Charcoal's GWC will rise dramatically because of the
inefficient operation of most charcoal kilns (Smith et al., 1999). Nevertheless, it is clear that the
database confirms some of the preliminary counter-intuitive conclusions of the Manila pilot
1 There is disagreement, however, about the appropriate mean GWP values of CO and hydrocarbons to use for such
calculations because of geographic and seasonal variations (IPCC, 1995). Here we apply those published in IPCC
(1990).
-------
study, i.e., that in some circumstances a switch from solid biomass fuels, even if renewably
harvested, to kerosene or LPG can be recommended for the purpose of reducing GHG emissions.
One surprising result, however, is that LPG is only marginally superior to kerosene. The
remarkable performance of biogas is because it is the only fuel tested here that is favored with
both the high thermal and combustion efficiency of gaseous fuel along with the advantages of
renewability. As such, it foreshadows the large potential for liquid and gaseous fuels made
from biomass to substantially reduce the GWC and health-damaging emissions from household
use of unprocessed biomass.
Figures 1 and 2 summarize the results aggregated by fuel and divided according to type of
analysis (renewable/nonrenewable; Basic/Full GHG). Note the strong performance of kerosene
and LPG when the full set of GHG is used and that even in the renewable case wood has only a
relatively modest advantage over fossil fuels using the basic GHG set. The strikingly superior
performance of biogas is seen in all cases. All these results, of course, represent the means for
the particular mix of stoves tested for each fuel in this study, which does not necessarily
represent the mix in the country as a whole.
Three main conclusions can be drawn:
Even if renewably harvested, biomass fuel cycles are not GHG neutral because of their
substantial production of PIC.
To be nearly GHG neutral, not only must biomass fuel cycles be based on renewable
harvesting, they must have close to 100% combustion efficiency, which most do not in their
current configurations in India.
In the processed form of biogas, however, biomass seems to offer the opportunity of providing
a renewable source of household energy with extremely low GWC because of its double blessing
of being gaseous when burned and renewable when harvested.
Compared to the default emission factor values recommended by the IPCC (1997) for residential
"oil" and natural gas, our results for kerosene and LPG are substantially higher for CO, TNMOC,
and N2O, but similar for CH4. The IPCC values for biomass fuels are generally within the range
we found for the different biomass-stove combinations.
From these measurements it seems that CH4 emissions from biomass combustion in India may be
about 1.9 Tg (million ton). It is thought that Indian biomass stoves represent about 27% of the
global total (UNDP, 1997). Thus, if the distribution of stove types globally is similar to India's,
it could be expected that biomass stoves produce globally about 7.1 Tg of CH4 annually. This is
approximately 7% of total methane emissions from all global activities related to fossil fuel
harvesting and use (Houghton et a/., 1996).
-------
Figure 1. GWC-full per MJ Delivered
Mean Values for Each Fuel
Biogas
LPG
Kerosene
Wood
Root
Crop Residues
Dung
Grams Carbon as CO2
10 100
1000
T
^Renewable (except for Kerosene and LPG)
^Nonrenewable Wood and Root
Full GWC = CO2, CH4, N2O, CO, TNMOC
-------
Figure 2. GWC-basic per MJ Delivered
Mean Values for Each Fuel
1
Biogas
LPG
Kerosene
Wood
Root
Crop Residues
Dung
Grams Carbon as CO2
10 100
1000
\
Renewable (except for Kerosene and LPG)|
^Nonrenewable Wood and Root
[Basic GWC = CO2, CH4, N2O|
-------
III. METHODS
This study was designed to measure the emission factors of greenhouse gases from
household cooking stoves in India and conduct a preliminary estimate of total national emissions
from such sources. The specific objectives are to:
choose commonly used fuel/stove combinations in India that represent all major fuel
types;
determine the energy content and chemical composition of all chosen fuels;
collect samples of gaseous emissions following a sampling procedure that represents
operating conditions in the field;
analyze these samples in the laboratory for estimating concentrations of CC>2, CO, CH/i,
N2O, TNMOC;
measure the concentrations of other important pollutants including total suspended
particulates (TSP), sulfur dioxide (802) and nitrogen dioxide (NO2)
measure thermal parameters such as burn rate and determine over-all thermal efficiency
of each fuel/stove combination;
based on existing data sources, estimate the annual consumption of cooking fuels in
different regions of India and
estimate national GHG inventory for Indian cookstoves.
To accomplish these objectives, the following approach was taken;
A. Experimental Design
Cooking is not a continuous process and practices vary in different parts of the nation as
to the breakdown between high-power, low-power, and other phases. Unlike gaseous fuels the
emission characteristics for solid fuels vary at different times during the burn. Hence it is
necessary to choose a burn cycle that is reasonably close to the common cooking practice in the
field. For the present study the "water boiling test," a procedure developed as a standard
international method to compare the efficiencies of different stoves was used with slight
modification (VITA 1985). The water boiling test is a relatively short, simple simulation of
common cooking procedure in which a standard quantity of water is used to simulate food. The
test includes "high power" and "low power" phases. The high power phase involves heating the
standard quantity of water from the ambient temperature to boiling temperature as rapidly as
possible. The low power phase follows in which the power is reduced to the lowest level needed
to keep the water simmering. This procedure has the added advantage of enabling simultaneous
measurement of emissions and efficiency. The burncycle ranged from 30 to 45 minutes for most
fuel/stove combinations.
All stoves were placed under a hood and gas samples were collected through a probe placed
inside the hood exhaust duct. The hood method (sometimes called the "direct" method) has
been used in studies of unvented cookstoves and kerosene space heaters. (Davidson et al. 1987;
Lionel et al. 1986; Ballard and Jawurek 1996). Tedlar bags were used to collect the emissions
from fire start to fire extinction. In a second Tedlar bag, background air during non-cooking
times was also collected.
-------
A pilot study was carried out with wood fuel in a traditional stove to finalize the protocol. Hood
and background samples were analyzed in TERI and OGIST laboratories and the results were
compared. Main phase experiments were started after satisfactory conclusions had been obtained
from the pilot phase. During the main phase three burncycle experiments were conducted for
each fuel/stove combination. A total of 28 fuel/stove combinations were tested.
All experiments were carried out in a simulated rural kitchen (SRK) constructed in the Gual
Pahari campus of TERI. The design of the kitchen was based on an earlier facility used to test
the thermal performance and emission characteristics of cookstoves (Ahuja et al. 1987).
Although the earlier study used mudwalls and a thatched roof, the current kitchen is constructed
with brick masonry coated with cement and tiled roof. The cement coating was given to avoid
the resuspension of particles from wall. The facility is located in a rural environment where there
are no nearby pollution sources. The ventilation conditions of the simulated kitchen can be
adjusted by the researchers. The emissions were captured by a hood through which a fixed
airflow rate was maintained by an electrical blower. The stoves, whether fitted with a chimney
or not, were placed so that the exhaust gases were entirely captured by this hood. A detailed
description of the simulated rural kitchen and hood system is given in Appendix A.
B. Fuels
A wide range of fuels is used for household cooking in India. The last National Census
(1991) found the following household distribution:
Animal Dung: 15%
Wood and crop residues: 62%
Charcoal: 0.8%
Coal: 3.5%
Kerosene: 7.2%
LPG (liquid petroleum gas): 7.9%
Biogas: 0.5%
Electricity and other: 3.2%
with large differences among regions and between rural and urban settings. (Detailed and more
recent estimates are presented in Section V and Appendix G.) Here, 11 typical fuels covering the
entire spectrum were chosen for testing:
Eucalyptus (safedd). Eucalyptus trees are largely grown in farm forestry (trees with crops) and
along road and railway lines. The Ministry of Environment and Forestry promotes eucalyptus
since it has a good commercial value, is easily grown in any area, and is not browsed by animals.
Because of its high calorific value, it is preferred for cooking. Eucalyptus trees are mostly grown
in the Indian states of Punjab, Haryana, Uttar Pradesh, Karnataka and Maharashtra.
Acacia (keekar\ Acacia is a small tree grown mainly in barren land and roadsides. These trees
are common in all parts of India and are mainly used as a fuel.
-------
Root fuel (Calligonium poligonidus). In some parts of Rajasthan state (where the forest cover is
minimal and the soil is dry) people use the root portion of the plant as a fuel. This plant is a fast-
growing bush-type plant and its root burns like wood.
Charcoal. When wood is burnt in the absence of air (this is usually done slowly in underground
or other semi airtight conditions), the volatile content in the biomass will be greatly reduced
leaving a solid with about twice the energy density of the wood. The resulting product is known
as charcoal. In India about three-quarters of the charcoal produced is used in small-scale
industries such as jewelry making, laundries (in traditional ironing machine), silk reeling units
and bakeries. Only about one-quarter is used for cooking. Here we bought in a Delhi market
low-quality charcoal of the type used in households.
Charbriquette. The waste carbon material remaining in the gasifier after the biomass gasification
is briquetted into charbriquettes. The charbriquettes for this study came from a gasifier using
wood.
Dungcakes. At 15% of households, cakes made mainly from the dung of cattle, buffalo, or
camels are used as major fuel. They are mainly used in rural areas and among poor groups in
cities. The dung (cattle waste) is mixed with a bit of crop residue and sundried. Dung cakes are
commonly used in all parts of the country except the Northeastern states. Haryana and Utter
Pradesh have the greatest use of dung as a fuel (Joshi and Sinha 1993).
Mustard stalk and rice (paddy) straw. Crop residues are also used by about 15% of households
nationwide. They are the plant materials left in the field after the main crop product has been
extracted and can be in the form of straw, stalk, husk, or fibrous material. The type of crop
residues available for fuel varies as the type of crops grown in the region. Other common crop
residues used as fuel are cotton stalk, jute stalk, tobacco stalk, wheat straw, and pulse stalk.
Kerosene, a middle distillate from petroleum refining, is mainly used in cities where about 25%
of the population relies on it (Census of India 1991).
Liquid Petroleum Gas (LPG) is marketed by Indian Oil Corporation and Bharat Petroleum under
the names of "Indane" and "Bharat" in 14.2 kg cylinders. It typically consists of about 80%
butane and 20% propane.
Biogas is a versatile gas used for cooking and lighting. Biogas is a relatively clean gaseous fuel
produced mainly from cattle dung and other animal waste in anaerobic digesters. It typically
consists of about 60 % methane, 30 % CO2 and 2 % H2 with traces of ammonia, nitrogen, and
hydrogen sulfide. Widespread dissemination of biogas plants began in 1981 through the
National Project on Biogas development (Ramana 1991). Since several animals are needed to
supply for each biogas plant, biogas stoves are mainly found in rural areas where, overall,
somewhat more than 1% have such devices.
C. Stoves
-------
Here is a brief description of all the stoves tested. Details of each with drawings are
found in Appendix B. Note that only the two marked "vented" are equipped with chimneys.
Traditional mud stove (-tm). This is a simple "U" shaped heavy stove for a single pot made by
households with locally available clay and coated with cowdung clay mixture.
Three-rock arrangement (3-R). Rural people with nomadic tendencies and people who live in
pavements with no permanent shelter arrange three stones or bricks for cooking and heating
purposes. This is a simple open fire cooking arrangement. No special skill or investment cost is
involved in constructing, operating and maintaining them. The pot hole size can also be varied
by adjusting the stones.
Improved Metal (imet) This is a portable metal non-chimney woodstove with a single pothole
developed in 1983 by Central Power Research Institute (CPRI), Bangalore, India. In 1991, the
stove was brought under Indian standards (BIS 1991).
Improved Vented Mud (ivm) This is a two-pot cookstove with chimney, called the Nada chulha.
A tunnel connects the fire box to the second pot hole and to a chimney. Since two pot holes are
provided two things can be cooked on it at the same time with only one fire.
Improved Vented Ceramic (ivc). This is also a two-pot cookstove with chimney. Made of a
ceramic lining with mud coating, this stove was developed at the Central Glass and Ceramic
Research Institute, Khirja, Uttar Pradesh, which is one of the Technical Back-up Units of the
national improved stove program.
Kara. This is a traditionally designed earthen pot for burning dung cakes and used mainly for
slow heating of milk over three to four hours such that, without boiling, the cream of the milk
separates as a thick layer at the surface. It is also used for cooking fodder.
Angethi (used for charcoal and charbriquette). This is a portable stove fabricated with a
galvanized iron bucket, mud/concrete, and grate. The fuel has to be fed above the grate by lifting
the pot in a batch operation.
Kerosene wick (kero-wick). The model used in the study was developed by Indian Oil
Corporation and marketed from 1977 under the brand name of "NUTAN."
Kerosene pressure (kero-pres) This single-burner pump-type kerosene stove is among the less
expensive versions available.
LPG stove. LPG stoves are commonly used by urban families. There are two types of LPG
stoves, with single and double burners, for household cooking. The stove tested in the present
study is a single-burner model with standards specified by Indian standards (BIS, 1978).
10
-------
Biogas stove. A two-burner model was used for study, but only one burner was operated during
the test.
D. Fuel/Stove Combinations
Since emissions and efficiency are functions of both fuel and stove (as well as cooking
technique and other factors), it is most appropriate to discuss our results by "fuel/stove
combination." The 28 fuel/stove combinations successfully tested are shown in Tables 1-2.
Note that several stoves were used with the same biomass fuels: traditional mud, three-rock,
improved metal, improved mud with chimney, and improved ceramic with chimney.
Table 1. Fuel/stove combinations for gaseous and liquid fuels
Fuel
LPG
Biogas
Kerosene
Stove
Burner Pressure Wick
0
o
0 0
Table 2. Fuel/stove combinations for solid fuel (all unvented, unless stated otherwise)
Fuel
Abbreviation =
Charcoal
Charbriquette
Eucalyptus
Acacia
Root fuel
Mustard stalk
Rice straw
Dungcakes
Stove
Angethi Traditional Improved Improved
Mud Metal Vented
Mud
tm imet ivm
0
o
o o
000
000
o o o
o o
0 0
Improved 3 -rock Kara
Vented
Ceramic
ivc 3-R
o o
0 0
o
0 0
11
-------
E. Sample Collection and Parameters Measured (details in Appendix C):
In each experiment emission gases and indoor air samples were collected in the flue gas
stream, which was kept at a constant flow rate by a blower (Appendix A). Emission samples
were taken under near isokinetic conditions through a probe in the hood connected to a low-
volume air sampler at a constant flowrate (about 2 1/min) through a filter and into a Tedlar bag.
Indoor background samples were collected at stove mouth height near the door using the same
arrangement. Ambient measurements (outdoor and indoor) were also done during non-cooking
hours. Ambient outdoor samples were collected at a height of 8 feet (2.5 m).
Time, temperature, and the weight of water, fuel, and char were recorded at the beginning and
end of the high and low cooking phases. For gaseous fuels, the volume of gas consumed was
recorded during each experiment. Fuel calorific values and moisture content were also analyzed
to calculate overall thermal efficiency. (See Appendices C-F.)
Fuel, ash, and char samples were analyzed for carbon, sulfur, ash and nitrogen contents. Air
samples were analyzed for carbon dioxide (62), carbon monoxide (CO), methane (CH4) total
non-methane hydrocarbon, sulfur dioxide (802), and nitrogen dioxide (NC^). TSP was
determined by subtracting the pre- and post-weights of the filters. One filter from each fuel/stove
combination was analyzed for carbon content.
One emission gas sample for each fuel/stove combination was placed in a 850-ml stainless-steel
canister and sent to OGIST for gas analysis, which in addition to the above gases included N2O
and hydrocarbon speciation. For each fuel, one canister was filled in duplicate through an
ascarite trap (to reduce N2O artifacts in the canister).
F. Careful efforts were made to maintain the following Quality Control Plan.
Six pilot-phase experiments were run to develop the protocols and become familiar with
the system operation.
For each fuel/stove combination, one or two preliminary experiments were conducted to
standardize the burncycle and minimize the natural viability due to differences in
operator behavior (a parameter not studied in these experiments). Prior to the three
planned tests for each fuel/stove combination, trial runs were conducted until a
satisfactory method precision was obtained. Results from these replicate samples were
< 20% RSD.
Each solid fuel to be tested was procured in one lot, sun-dried, and wrapped in plastic
sheets to avoid any change in moisture content.
Wood and root fuels were chopped into pieces of same length and width before packing.
Dungcakes used in all fuel/stove combinations were made by the same person using the
same ratio of dung and crop residue.
12
-------
After each experiment, the doors and windows were opened. Exhaust fan and side fans
were switched on to clean the room properly.
Char and ash remaining in each experiment were covered with aluminum foil and
labeled for carbon analysis.
Tedlar bags and Teflon tubing used in each experiment were flushed adequately with
compressed clean air for cleaning.
Tedlar bags and Teflon tubing used for low-grade fuels such as solid biomass fuels were
not used again.
After each fuel/stove combination was tested, the probe and the hood were cleaned with
a vacuum cleaner.
A mixture of calibration gases sent from EWC to TERI and OGIST was used to
calibrate the TERI GC.
Leak-proof tested and certified canisters were filled with duplicate samples and sent to
OGIST for further analysis of gaseous emissions. OGIST values were compared with
TERI values and in cases where there were many deviations (>20%) the experiments
were repeated.
The pumps used for collection of aerosol samples were calibrated with a bubble tube
before and after each experiment.
Filters used for TSP measurements were weighed at least twice. If the difference was
more than 0.005 milligram in the two weighings, the balance was calibrated and the
filter was weighed again.
Blank filters were weighed and treated in the same fashion; approximately one blank for
20 samples was used.
After post weighing, the filter cassettes were sealed for carbon content analysis.
The spectrophotometer used for SO2 and NO2 analysis was calibrated carefully and
checked with standards after each set of analyses (See Appendix C).
G. Emission Factors
Since each experiment was done while performing the standardized cooking test
(Appendix C), the total emissions measured are those of the standard cooking task, which
consists of heating 2.2 kg of water from ambient temperature to boiling, followed by simmering
(Ahuja et a/., 1987). Here we break down the emission calculations into two parts. The first,
called "instant emissions," addresses the emissions during a particular test. The rate of these
emissions is appropriate for estimating indoor or local concentrations. The second, called
"ultimate emissions," is an estimate of the ultimate emissions in typical household conditions in
India from a unit of fuel and are most appropriate for determining greenhouse-gas inventories
from fuel demand. The two types of emissions differ only for some of the solid fuels. The
calculation of each differs solely in the way the remaining partly charred fuel is handled.
G.I. Instant Emissions: The carbon balance method (Smith etal. 1992; 1993) is used to
calculate these emission factors. During combustion, fuel carbon (FC) is mainly converted to the
gases, carbon dioxide (CO2), carbon monoxide (CO), methane (CH4), and total non-methane
13
-------
organic compounds (TNMOC). Some is diverted into airborne aerosol (TSP) and bottom ash or
remains as the partially burned material, char. Since we are focusing on the emission factors for
airborne components, we subtract char and ash carbon from the fuel used. This also fits with
actual practice, in that householders usually save unburned char for later use, e.g., at the next
meal. To accurately track all the carbon, it is also necessary to account for the small amount of
kerosene used to start the solid fuel stoves, which is done to attain more uniformity during the
often-smoky first period of combustion and also is common practice in many households.
On a carbon basis,
FC = CO+CH4+TNMOC+CO2+TSP (1)
r\
FC = [(Fuel consumed x carbon fraction) + (Kerosene, if any x carbon fraction)]-
[(Char produced x carbon fraction) + (Ash produced x carbon fraction)]
CO2 = FC - (CO+ CH4+TNMOC+TSP) (2)
Dividing by CO2
1 = FC/CO2-(CO+CH4+TNMOC+TSP)/CO2 (3)
or
1 = (FC/CO2)-K
K = is the sum of emission ratios to CO2 = (CO+ CH4+TNMOC+TSP)/CO2
Emission factors per burn cycle experiment = EFbc (g/burncycle).
CO2 as g carbon = FC/(1+K) (4)
CO as g carbon = (emission ratio for CO) x CO2 as g carbon (5)
CH4 as g carbon = (emission ratio for CH4) x CO2 as g carbon (6)
TNMOC as g carbon = (emission ratio for TNMOC) x CO2 as g carbon (7)
(we assume that the equivalent molecular weight of TNMOC is 18 per carbon atom)
The emission factor for TSP carbon is calculated
TSPC = (TSP/CO2 ratio) x CO2 (8)
This is converted to TSP mass:
TSPm = TSPc/Measured carbon fraction in the TSP (9)
Since it has no carbon, N2O is not included in the carbon balance equation. Its emission factor
can be calculated as
2 Used in small quantities to initiate burning in some of the solid fuels.
14
-------
N2O (g) = (N2O emission as molar ratio to 62)
x CO2 as g carbon) x 3.67 (10)
Since the molecular weight of N2O is 3.67 times heavier than the atomic weight of carbon.
The emission factors above are calculated for one burn-cycle experiment. The emission factor
per unit fuel:
EFm =(EFbc)/mass of fuel used in experiment (11)
where (EFm) is expressed as g/kg of dry fuel.
The emission factor per unit net fuel energy content (g/MJ) is found as
EFe = EFm/(energy content of fuel (MJ/kg)) (12)
The emission factor per unit delivered energy (g/MJd) is
EFt =EFe/T| (13)
where T| is the thermal efficiency of the stove (Appendix D).
G.2. Ultimate Emissions: The instant emissions calculated above are specific to the conditions
of the tests, but need modification in some cases to reflect actual field conditions. This is
because of the diversion of a significant amount of fuel carbon into production of low-quality
charcoal in the root and wood stoves. In households, of course, this charcoal is usually not
wasted, being either left in the stove to be burned along with fresh fuel at the next meal or
extracted and stored for later use to cook a meal entirely with charcoal fuel. Both practices are
common in India, but we have no data indicating the actual percentage breakdown. Thus, the
inherent assumption in the analysis of Section G.I that the charcoal carbon does not enter the
atmosphere is not valid.
Figure 3a shows a typical result for a wood-fired stove in this study, in this case Eucalyptus in
the improved vented ceramic (ivc) stove, a stove that tends to produce high charcoal yields.
Note that the kilogram of wood produces 161 g of charcoal containing 130 g or 29% of the
original carbon. The results shown are from the instant analysis. Since this charcoal would be
burned eventually in field conditions, however, these numbers cannot be used directly to
calculate ultimate emissions. To handle this situation, we also measured the emissions of the
kind of low-quality charcoal produced in such stoves. Figure 3b shows the additional emissions
that would result from burning the 161g of charcoal produced from the original wood in Figure
3a. Note that the remaining char produced in this case contains less than 0.4% of the original
carbon (1.6 g) in material that is only 20% carbon, i.e., too poor to be attractive as fuel. It seems
justifiable, therefore, to consider this as the solid carbon that becomes part of the disposed ash
and char and is thus sequestered from the atmosphere, if not permanently, at least for long
periods.
15
-------
The ultimate emissions per kilogram of wood in this case, therefore, are the total of those shown
in Figures 3a&b. Note that compared to instant emissions alone all the major emissions
increase by roughly the same amount as the fraction of charcoal carbon compared to the fuel
carbon, i.e. 20-30%, except for CO, which nearly doubles. The larger increase for CO reflects
the dominance of char burning compared to flaming combustion because of charcoal's low
volatile content compared to wood.
In a similar fashion, the ultimate K-factor is somewhat different from what is found by instant
analysis alone. Both types are reported here, therefore.
In reporting emissions per unit fuel energy, it is simply necessary to divide the ultimate emissions
per kilogram by the original fuel's lower heating value in megajoules (MJ/kg), as in Eq. 12. In
reporting emissions per unit delivered energy, however, it is necessary to consider what stove
efficiency (r|) to apply. There are two major options:
A. Use the energy efficiency measured in the primary stove (the one using the original solid fuel)
for the entire process; or
B. Use the energy efficiency measured in the primary stove only for the fuel consumed in the
process shown in Figure 3a and apply the efficiency measured in the charcoal stove
(Angethi) for the remaining consumed in the process of Figure 3b.
We have chosen the first option, which basically assumes that most of the produced char will be
used in the original stove and not saved for later use in a special charcoal stove (Eq. 13). Since
the measured efficiency (18%) of the charcoal-using Angethi is within the range for stoves using
wood (17-29%) and rootfuel (14-23%), and only a fraction of the carbon is converted to charcoal,
the difference in estimated ultimate emissions per MJ delivered energy between the two options
is not large in any case.
16
-------
Fig. 3 a. Instant Carbon Balance:
Eucalyptus in Improved Vented Ceramic Stove
Instant k-factor = 0.095
Wood: 1.0 kg
454 g Carbon
I
C02 Carbon: PIC Carbon: Char/Ash: 161 g
295.5 g CO: 18.5 g 130.2 g Carbon
CH4: 2.8
TNMOC: 5.2 g
TSP Carbon:
1.7g
17
-------
Figure 3b. Carbon balance of char combustion after primary
combustion. Ultimate k-factor= 0.124 (processes
in 3a and 3b)
Char: 161 g
Carbon: 130 g
C02 Carbon:
107g
PIC Carbon:
CO: 19 g
CH4: 0.95 g
TNMOC: 1.1 g
TSP Carbon:
0.32g
I
Char/Ash: 7.6 g
Carbon: 1.6 g
18
-------
IV. RESULTS
We successfully tested 28 fuel/stove combinations, three times each. The methods and
results of primary measurements are found in Appendices E & F. Here we derive instant
emissions ratios and K-factors, power levels, efficiencies, carbon balances and ultimate K-
factors, and emission factors.
A. Emission Ratios
Gross and net concentrations of pollutants in the fluegas of fuel/stove combinations are
presented in Appendix F along with a discussion of the cross-laboratory comparison for quality
control the resulting corrections applied to the data. Table 3 shows the resulting instant ratios to
CC>2. Also shown are the instant K-values.
According to the Indian standard for domestic LPG stoves, the limit for CO/CO2 emission ratio is
0.02 (BIS, 1984). This ratio provides a simply measured indicator of combustion quality and this
limit is thought to keep the risk of acute CO poisoning to acceptable levels. In our experiments,
the mean CO/CO2 ratios for biogas, LPG, and kerosene wick stoves are below this limit. The
ratios for all biofuels and charcoal are much higher than this value. The highest CO/CO2 ratio is
found for charcoal. These are the same results as found in the Manila pilot study (Smith et al.
1992; 1993).
The CO emission ratio for wood varied from 0.03 to 0.17. The higher emission ratio 0.17 was
recorded for wood in the improved mud stove. The CO emission ratios for the two wood species
in traditional mud and three-rock stove are between 0.03 and 0.04. Hao et al. (1990) reported the
CO emission ratio for wood stoves as 0.06 for open combustion over a range of biomass types.
This discrepancy may be due to the difference in measurement techniques, particularly in that
Hao et al. were not able to monitor all carbon outputs, which would tend to inflate the apparent
CO emission ratios.
The range of CO emission ratios (0.14-0.16) for the improved vented mud stove (ivm) is much
higher than the CO emission ratio for some of improved mud stoves (between 0.04 and 0.07)
reported in FAO (1993); whereas the range of CO emission ratios for wood fuel in the improved
vented ceramic stove (ivc) is within this range (0.03-0.6). The CO emission ratio for wood in
the improved unvented metal stove (imet), is the same (0.04) as given in FAO (1993). Clearly,
because of the large differences that occur with changes in design, more effort is needed to
identify exactly which aspects of stove design affect these ratios.
The CO emission ratios for dungcake and crop residues are higher than the ratios for wood fuel
in all types of stoves tested. This is similar to the findings of the earlier study by FAO. Except
for dungcake, all other tested fuels produced a CO ratio higher in the ivm stove. In general, our
N2O/CO2 ratios are lower than the 0.007 quoted by Crutzen and Andreae (1990), who, however,
did not monitor small-scale combustion devices directly.
19
-------
Table 3. Instant emission ratios and nominal combustion efficiencies (NCE) for all tests.
(K = sum of ratios of all carbon in all airborne products of incomplete combustion to carbon in
C02)
Fuel-Stove
Gas
LPG
LPG
LPG
Biogas
Biogas
Biogas
Kerosene
kero-pres
kero-pres
kero-pres
kero-wick
kero-wick
kero-wick
Charfuel
Charcoal
Charcoal
Charcoal
Charbriq
Charbriq
Charbriq
Wood
Acacia-imet
Acacia-imet
Acacia-imet
Acacia-ivc
Acacia-ivc
Acacia-ivc
Acacia-ivm
Acacia-ivm
Acacia-ivm
Acacia-3R
Acacia-3R
Acacia-3R
Acacia-tm
Acacia-tm
Acacia-tm
C0/C02
6.3 0 E-3
9.34E-3
7.24E-3
2.05E-3
3.00E-3
1.34E-3
0.0350
0.0380
0.0267
6.69E-3
0.0109
0.0100
0.197
0.201
0.143
0.135
0.103
0.121
0.0465
0.0409
0.0393
0.0232
0.0236
0.0392
0.152
0.131
0.142
0.0359
0.0342
0.0387
0.0397
0.0288
0.0351
CH4/C02
1.27E-5
1.21E-4
5.72E-6
3.46E-4
0.00524
2.02E-4
0.00120
0.00107
7.40E-4
1.20E-4
4.09E-4
2.59E-4
0.0128
0.00680
0.00762
0.00749
0.00562
0.0146
0.00968
0.00784
0.00626
0.00741
0.00356
0.00575
0.0290
0.0346
0.0374
0.0174
0.0211
0.0286
0.0103
0.00598
0.00590
TNMOC/CO2
0.0186
0.0156
0.0105
4.22E-4
0.00207
3.97E-4
0.0125
0.0180
0.0174
0.0122
0.0131
0.0108
0.00938
0.0131
0.00949
0.0301
0.0268
0.0174
0.0169
0.0174
0.0245
0.0361
0.0305
0.0290
0.0362
0.0297
0.0288
0.0209
0.0163
0.0209
0.0128
0.0161
0.0154
TSP/CO2
5.77E-4
5.46E-4
7.10E-4
3.73E-4
0.00146
4.05E-5
6.12E-4
1.05E-3
9.67E-4
9.06E-4
2.67E-4
4.63E-4
0.00318
0.00474
0.00151
0.00516
0.00373
0.00105
0.0122
0.00700
0.0175
0.0145
0.0129
0.0115
0.0158
0.00959
0.0108
0.00483
0.00440
0.00823
0.00111
0.00235
0.00258
K-Instant NCE= ,
0.0255
0.0256
0.0185
0.00319
0.0118
0.00198
0.0494
0.0581
0.0459
0.0109
0.0246
0.0215
0.222
0.226
0.162
0.177
0.139
0.154
0.0853
0.0731
0.0875
0.0813
0.0706
0.0855
0.233
0.205
0.219
0.0791
0.0759
0.0965
0.0639
0.0533
0.059
l/(l+k)
0.975
0.975
0.982
0.997
0.988
0.998
0.953
0.945
0.956
0.981
0.976
0.979
0.818
0.816
0.861
0.849
0.878
0.867
0.921
0.932
0.920
0.925
0.934
0.921
0.811
0.830
0.820
0.927
0.929
0.912
0.940
0.949
0.944
(continued)
20
-------
Table 3 (continued)
Fuel-Stove
Eucal-imet
Eucal-imet
Eucal-imet
Eucal-ivc
Eucal-ivc
Eucal-ivc
Eucal-ivm
Eucal-ivm
Eucal-ivm
Eucal-3R
Eucal-3R
Eucal-3R
Rootfuel
root-ivm
root-ivm
root-ivm
root-imet
root-imet
root-imet
root-tm
root-tm
root-tm
Crop Residues
must-ivm
must-ivm
must-ivm
must-ivc
must-ivc
must-ivc
must-imet
must-imet
must-imet
must-tm
must-tm
must-tm
rice-ivm
rice-ivm
rice-ivm
rice-tm
rice-tm
rice-tm
CO/CO2
0.0356
0.0543
0.0525
0.0638
0.0907
0.0358
0.166
0.144
0.156
0.0316
0.0401
0.0281
0.0370
0.0439
0.0494
0.0416
0.0642
0.0475
0.0246
0.0205
0.0474
0.158
0.0972
0.158
0.0505
0.0889
0.0928
0.0558
0.0945
0.0469
0.0762
0.108
0.0555
0.288
0.0921
0.117
0.0865
0.0785
0.0448
CH4/CO2
0.00289
0.00967
0.00772
0.0169
0.00265
0.00924
0.0298
0.0233
0.0419
0.00300
0.00627
0.00322
0.00314
0.00599
0.00738
0.00331
0.00629
0.00550
0.0239
0.00250
0.0320
0.0421
0.111
0.0423
0.00646
0.0140
0.0148
0.00731
0.0122
0.00425
0.0199
0.0204
0.00830
0.00916
0.0111
0.0151
0.0126
0.0224
0.00584
TNMOC/CO2
0.0439
0.0284
0.0175
0.0388
0.0133
0.00162
0.0632
0.0451
0.0884
0.0117
0.0168
0.0113
0.0367
0.0308
0.0251
0.00744
0.0285
0.0163
0.0252
0.0268
0.0205
0.0614
0.0790
0.0517
0.0333
0.0883
0.0543
0.0273
0.0348
0.00744
0.0335
0.00730
0.00732
0.0200
0.0200
0.0200
0.0192
0.0246
0.0189
TSP/CO2
0.00789
0.00547
0.00365
0.00711
0.00691
0.00358
0.00977
0.00487
0.00996
0.00207
0.00164
0.00204
0.0143
0.00487
0.00557
0.00307
0.00202
0.00169
0.00320
0.000615
0.00221
0.0136
0.0119
0.0126
0.00831
0.0205
0.0129
0.00791
0.00338
0.00670
0.00163
0.00196
0.00175
0.0590
0.105
0.0113
0.00221
0.00298
0.00286
K-Imtant NCE= ,
0.090
0.098
0.081
0.127
0.114
0.050
0.269
0.218
0.296
0.048
0.065
0.045
0.091
0.086
0.087
0.055
0.101
0.071
0.077
0.050
0.102
0.275
0.299
0.265
0.099
0.212
0.175
0.098
0.145
0.065
0.131
0.138
0.073
0.376
0.228
0.164
0.121
0.129
0.072
l/(J+k)
0.917
0.911
0.925
0.888
0.898
0.952
0.788
0.821
0.771
0.954
0.939
0.957
0.917
0.921
0.920
0.947
0.908
0.934
0.929
0.952
0.907
0.784
0.770
0.791
0.910
0.825
0.851
0.910
0.873
0.939
0.884
0.879
0.932
0.727
0.814
0.859
0.892
0.886
0.932
(continued)
21
-------
Table 3 (continued)
Fuel-Stove
Dung
dung-ivc
dung-ivc
dung-ivc
dung-tm
dung-tm
dung-tm
dung-ivm
dung-ivm
dung-ivm
dung-hara
dung-hara
dung-hara
C0/C02
0.0367
0.0696
0.0377
0.0709
0.0835
0.0737
0.0362
0.0607
0.0383
0.132
0.0987
0.0720
CH4/CO2
0.00740
0.0148
0.00646
0.0128
0.0187
0.0145
0.00693
0.0140
0.00457
0.123
0.0226
0.0128
TNMOC/CO2
0.0653
0.0935
0.0646
0.0483
0.0450
0.0410
0.0589
0.0804
0.0645
0.0551
0.0736
0.0466
TSP/CO2
0.00622
0.00959
0.00591
0.00703
0.00409
0.00627
0.00508
0.00702
0.00496
0.00181
0.00249
0.00190
K-Instant NCE= ,
0.116
0.188
0.115
0.139
0.151
0.136
0.107
0.162
0.112
0.311
0.197
0.133
l/(l+k)
0.896
0.842
0.897
0.878
0.869
0.881
0.903
0.861
0.899
0.763
0.835
0.882
B. Power and Thermal Efficiency
Thermal performance measured as power input and overall thermal efficiency (T|) of various
stove fuel combinations tested were calculated according to the methodology described in
Appendix D. We did not attempt to change the power in different experiments except those due
to interventions in the fire to ensure a steady flame. The power input and efficiency values for
three experiments for each fuel/stove combination were averaged and given in Tables 4 and 5.
The tables show that the power input of the stoves tested ranged from 1.3 kW for kerosene wick
stove to 7.6 kW for mustard stalk in traditional stoves. The average power inputs for the stoves
burning gaseous and liquid fuels were low, 1.3- 1.7 kW. For solid fuels the power inputs varied
from 1.6 kW for char briquettes in Angethi to 7.6 kW for mustard stalks in traditional stoves.
Compared with the improved stoves, the traditional stove had high power in all of the fuel
categories. Among various fuels tested the power-input increases from gaseous fuel and kerosene
to wood, and charcoal to dung cake to crop residues (Figure 4), generally in line with the energy
ladder framework (Smith 1990; OTA 1992).
Table 4. Power input and thermal efficiency for gaseous and liquid fuels
Fuel/stove
LPG
Biogas
Kerosene/wick
Kerosene/pressure
Power kW
1.6±(0.1)
1.4 ±(0.1)
1.3 ±(0.1)
1.7 ±(0.1)
Efficiency % (rtf
53. 6 ±(2.2)
57.3 ±(0.5)
50.0 ±(6.7)
47.0 ±(2.2)
22
-------
Table 5. Power input and thermal efficiency for solid fuels
Fuel-stove
Acacia-ivc
Eucal-ivc
Acacia-imet
Acacia-ivm
Root-imet
Eucal-ivm
Must-imet
Eucal-imet
Root-ivm
Must-ivc
Acacia-tm
Acacia-3 rock
Eucal-3 rock
Charcoal
Eucal-tm
Charbriquette
Root-tm
Must-ivm
Dung-ivc
Must-tm
Rice-ivm
Dung-ivm
Rice-tm
Dung-tm
Dung-hara
Power kW
2.5 ± (0.2)
2.5 ±(0.1)
2.4 ±(0.6)
3.1 ±(0.2)
3.4 ±(0.5)
3.9 ±(0.5)
5.8 ± (0.2)
3. 5 ±(0.3)
2.8 ±(0.5)
4.9 ±(0.4)
4.1 ±(0.2)
2.9 ±(0.2)
4.6 ±(0.1)
2.6 ± (0.2)
4.1 ±(0.0)
1.6 ±(0.3)
4.7 ±(0.9)
6.1 ±(1.2)
4.0 ±(0.1)
7.6 ±(1.0)
4.8 ± (0.4)
3.9 ±(0.1)
6.6 ± (0.2)
4.1 ±(0.5)
6.4 ± (0.6)
Efficiency % (r\)
29.0 ±(1.9)
28.7 ±(1.0)
25. 7 ±(2.5)
23. 5 ±(2.2)
22.8 ±(1.2)
22.0 ±(1.8)
21.7±(1.6)
21.4±(1.8)
19.7±(1.3)
18. 5 ±(0.8)
18.2 ±(0.6)
18.1±(0.6)
17.7±(0.3)
17.5 ±(2.7)
16.7 ±(0.7)
16.4 ±(0.5)
14.2 ±(1.8)
13. 5 ±(0.5)
12.8 ±(1.0)
12.4 ±(1.0)
10.9 ±(1.0)
10.0 ±(0.2)
9.8±(1.1)
9.4 ±(0.6)
8.2 ±(1.3)
(Standard Deviation of three tests shown)
23
-------
Figure 4. Power input for various fuel/stove combinations
9 -j-,.
8 --
7 --
^ 6 --
t 5"
HH
o3 4 --
O o
p , J
2 --
1 --
0 -_
_l 1 1 1 j_
f .
_l 1 1 (_
_l 1 1 (_
.* .^o
Fuel/stove
-------
The average thermal efficiency (T|) of the biogas stove (57.3%) is the highest among all stoves
tested. Khadi and Village Industries Commission (KVIC) and Bureau of Indian Standards (BIS)
recommend that the efficiency of domestic biogas burner should not be less than 55%. A report
of KVIC states that a thermal efficiency of 59.5% could be obtained for the corresponding power
of 1.61kW (Kishore and Dhingra 1990), quite close to our average efficiency of 57.3% for the
corresponding power of 1.59 kW. The average efficiency of the LPG stove is 53.6%, which is
less than the BIS specification of 60% (BIS-4246 1984). The kerosene wick stove had the
efficiency of 50% and the average efficiency of kerosene pressure stove was 47%. The efficiency
of the kerosene wick stove is less than the efficiency of 57% reported previously (TERI1987).
In addition, previous studies have sometimes found that the pressure stove is more efficient,
unlike our finding.3
The efficiency of Angethi (17.5%) with charcoal is comparable to that (15.3%) quoted by Wazir
(1981). The average efficiency of traditional stoves with various biomass fuels varied from 9.4
to 18.2%, being low for dungcake and high for wood. Wazir (1981) reported the efficiencies of
the traditional stove vary from 5 to 20%. George (1997) found the efficiency of traditional mud
stove to average 17.9%. The average efficiency of the 3-rock stove was also about 18% which is
within the efficiency range (12-24%) reported in TERI (1987).
The efficiencies of the improved stoves were higher than that of the traditional and 3-rock stoves.
The improved vented ceramic (ivc) had high efficiency for all fuels except crop residues. The
average efficiencies of the improved vented mud stove (ivm - Nada chulha) across fuels varied
from 10% to 23.5%, which is compatible with the range reported by Pal and Joshi (1989) of
10.8% to 19.6%. Our measurements using wood fuels in the improved unvented stove (Priyagni
- imet) of 21.4 & 25.7% are compatible with the 26% reported by FAO (1993). Among various
fuels, dungcake had the lowest efficiency in all stoves, being lowest of all in the Kara stove
(8.2%).
Tables 4 and 5 show that the overall thermal efficiency (T|) increases by moving up the energy
ladder from dungcake to crop residue to wood to kerosene to gas. This pattern is similar to the
typical energy ladder of South Asia discussed by Smith et al. (1994).
Overall stove thermal efficiency was determined by the method outlined in Appendix D, i.e.
dividing the calorific value of the fuel used in a test run into the heat absorbed by the water in the
pot during the same run. It is a linear combination of two internal efficiencies:
3 It is useful to note in this context, however, that the standard deviation of the kero-wick stove efficiencies was high
in our experiments (COV = 13%, Table 3), indicating no statistically significant difference between the two kerosene
stoves in overall efficiency (r|).
25
-------
T|=NCE*HTE (14)
NCE (nominal combustion efficiency) is the percentage of the chemical energy in the fuel that is
actually released and is defined here as the percentage of airborne fuel carbon released as CO2
NCE = 1/(K+1) - see Equations (1-3) (15)
Instant NCEs are shown in the last column of Table 3. HTE (nominal heat transfer efficiency) is
the percentage of heat released by combustion that is absorbed by the water in the pot. This was
not measured directly in our experiments and is determined using Equation 14, since both NCE
and r| are available from the tests.
From an environmental point of view, the two most important parameters are 1/(1-NCE) which is
a direct indicator of how much PIC pollution is released and T| which indicates the amount of fuel
used. To ease comparisons, we will frequently summarize our main results by fuel/stove
combination using the ranking derived by application of an Environmental Stove Index (ESI) that
is composed of these two parameters:
ESI = ln[ri/(l-NCE)] (16)
As shown in Figure 5, HTE and NCE each trends downward with ESI, although the differences
between stove designs cause some deviations.
The average overall efficiency of fuel/stove combinations decreases with increasing average
power levels in a nonlinear way (Figure 6). Biogas, LPG, and kerosene stoves burned at low
power with high efficiencies, the reverse of dungcake and crop residues.
The relative performance of stove types is shown in Figure 7. Note the relatively good
performance of the improved metal stove (imet) compared to the other two improved stoves.
The other two, however, are vented, which would presumably reduce indoor pollution levels. It
is interesting also that the simplest stove in the world, the three-rock stove (3R) is a better
performer than most of the improved stoves tested.
C. Carbon Balances
Table 6 shows the gross carbon balances per unit fuel carbon of each fuel/stove combination.
The first columns are for instant combustion, as in Figure 3a. The second set of column show
the ultimate values, which represent the total of processes in Figures 3a and 3b. The two are
the same for kerosene and gaseous fuels because they produce no char and the same for dung and
crop residues because they produce char of too low quality to burn. Also shown are the ultimate
K-factors and NCEs.
26
-------
:igure 5. ESI and Instant Combustion and Heat Transfer Efficiencies
Along the Household Energy Ladder
Combustion and Heat Transfer Efficiencies (%)
C
Biogas
LPG
Kero-wick
Kero-pres
Root-imet
Root-ivm
Root-tm
Acacia-imet
Eucal-ivc
Acacia-ivc
Eucal-3R
Eucal-imet
Acacia tm
Acacia-3R
Acacia-ivm
Eucal-ivm
Must-imet
Must-ivc
Must-tm
Rice-tm
Must-ivm
Rice-ivm
Dung-ivc
Dung-ivm
Dung-tm
Dung-Hara
0
) 20 40 60 80 100
I I I I
o ^-0 ^
O ET A
I A m A
4 \^ IL
/A
f
A
* m A
T f 4
~^+ f A
«C 1 >
> ff <
f ja >
*^ J* ^
^t J2' A""^
4 CI A^
O JZ\ A
^^<( 0^^^ J^
j*^ r^ ^^
-------
60
50
Si 40 +
o
= 30 +
'o
£ 20
LU
10
0
Figure 6. Power Input Vs Efficiency
0246
Power input (kW)
28
-------
Figure 7. Major Efficiencies and ESI by Stove Type
Unprocessed Biomass Fuels
Overall, Combustion, and Heat Transfer Efficiencies (%)
0 20 40 60 80 100
Root-imet
Acacia-imet
Must-imet
Eucal-imet
Eucal-3R
Acacia-3R
Eucal-ivc
Acacia-ivc
Must-ivc
Dung-ivc
Acacia-tm
Root-tm
Must-tm
Rice-tm
Dung-tm
Root-ivm
Acacia-ivm
Eucal-ivm
Dung-ivm
Must-ivm
Rice-ivm
Dung-Hara
I I I I
; ^ / i
- E> m A
ft Ef £
~~ IT "P A
^ii JD ~\
M&^ 0^ A^^
m ET i
/ HA
ia JL
IZI ^
JZl ^
L
^x J* ^*
^r^f izr^^ JT^'
^^^^4 m JL^^
^^ .El ^>A
^N ° ^^ AT^^'
mV El 1
- 150 n A
i
0.1 1 10
Environmental Stove Index
H Environmental Stove Index (lower axis) oNominal Heat Transfer Efficiency
A Nominal Combustion Efficiency Overall Efficiency = n
29
-------
L.J
o
Table 6. Gross instant and ultimate carbon balances; grams carbon based on 1.0 kilogram fuel input. (See Figure 3.)
measures are the same except for wood and root fuels. Ultimate K-factors, nominal combustion efficiencies (NCEs),
efficiencies (HTEs) are also shown.
The two
and heat transfer
Fuel/stove
LPG
Biogas
Kero-pressure
Kero-wick
Charcoal
Charbriq
Eucal-imet
Eucal-ivm
Eucal-ivc
Eucal-3R
Acacia-tm
Acacia-imet
Acacia-ivm
Acacia-ivc
Acacia-3R
Root-tm
Root-imet
Root-ivm
Must-tm
Must-imet
Must-ivm
Must-ivc
Rice-tm
Rice-ivm
Dung-tm
Dung-ivm
Dung-ivc
Dung-hara
Fuel
860
396
843
843
800
503
454
454
454
454
418
418
418
418
418
518
518
518
421
421
421
421
381
381
334
334
334
334
Char/ash
0
0
0
0
9.93
0.601
76.9
157
130
98.9
130
102
169
189
120
56.4
74.5
110
26.2
15.0
48.2
62.0
49.2
46.0
14.4
7.07
9.56
12.9
Instant Ultimate K-factor NCE HTE
CO2 PIC TSP Char/ash CO2 PIC TSP
841.4 19.0 0.514 0.0231 0.978 0.548
393.8 1.97 0.247 0.00562 0.995 0.577
802.6 40.2 0.699 0.0510 0.951 0.494
825.5 17.7 0.449 0.0220 0.978 0.511
657.5 131 2.05 0.202 0.831 0.210
434.7 66.4 1.43 0.156 0.861 0.190
345.8 29.1 1.96 0.954 409.0 41.7 2.16 0.107 0.902 0.237
236.2 59.4 1.90 1.94 364.9 85.0 2.32 0.239 0.807 0.273
295.9 26.5 1.71 1.62 402.9 47.8 2.09 0.124 0.889 0.323
337.6 17.1 0.644 1.23 418.9 33.2 0.936 0.0815 0.924 0.191
272.7 15.4 0.558 1.61 379.5 36.6 0.888 0.0988 0.910 0.200
291.3 20.3 3.54 1.26 375.1 37.0 3.84 0.109 0.902 0.285
204.8 42.4 2.56 2.09 343.6 70.0 3.02 0.212 0.824 0.285
213.0 14.0 2.78 2.34 368.0 44.9 3.35 0.131 0.884 0.328
276.0 21.6 1.63 1.49 374.8 41.2 1.97 0.115 0.896 0.202
428.7 31.8 0.857 0.699 475.1 41.1 1.02 0.0886 0.917 0.155
412.4 30.3 0.912 0.924 473.6 42.5 1.12 0.0921 0.915 0.249
376.1 30.0 3.09 1.36 466.3 47.9 3.36 0.110 0.921 0.219
355.1 39.4 0.631 0.113 0.898 0.138
368.7 35.3 2.22 0.102 0.907 0.239
291.5 77.6 3.71 0.279 0.781 0.173
309.5 45.3 4.26 0.160 0.861 0.215
300.3 31.2 0.802 0.106 0.903 0.108
268.1 51.8 14.9 0.249 0.769 0.136
280.1 38.1 1.61 0.142 0.822 0.107
290.5 35.2 1.63 0.126 0.887 0.113
285.3 37.4 2.03 0.138 0.877 0.146
265.6 55.0 0.545 0.209 0.824 0.099
-------
D. Ultimate Emission Factors
Emission factors were estimated separately for the three experiments in each fuel/stove
combination and the results expressed as an average of the three experiments done for each.
Three types of ultimate emission factors are presented here:4
Emission factors per kilogram fuel in pollutant mass (Efm): Table 7
Emission factors per kilogram fuel in pollutant carbon mass (Efm): Table 7
Emission factors per MJ net energy in fuel (EFe): Table 8
Emission factors per MJ delivered energy (EFd): Table 8
is based on 1 .0 MJ delivered to the pot and thus takes into account the energy efficiency of
the stove. Although there is obviously much variation throughout the nation, 1 .0 MJ delivered
represents a typical amount of energy used to cook a household meal.
The appropriate type of emission factor to use depends on the policy question being asked. Here,
we start with a discussion of emissions factors per unit fuel mass.
The CC>2 emission factor by fuel mass is high for LPG due to the high carbon content in the fuel
(about 86%) and good combustion efficiency of the stove, which lead to high CO2 and less PIC
(products of incomplete combustion - CO, CH/i, TNMOC).
The CO emission factor is high for charcoal (275 g/kg) and low for biogas (2 g/kg), reflecting
relative NCEs. CO emission factors for eucalyptus varies from 26-85 g/kg, with those from the
three-rock stove being at the low end. For rootfuel and rice straw, the emission factors for
improved stoves are also higher than the traditional stoves, a finding consistent with Ahuja et al.
(1987). Increased emission factors for "improved" stoves is consistent with previous evidence
that design changes directed at improving efficiency can actually increase emission factors for
many pollutants (TERI 1985). This is because they generally work to increase NTE, but in the
process lower NCE.
CFLj emission factors are low for gases and kerosene, but quite high for crop residues in
improved stoves. Among the three improved stoves, in most of the cases the emission factor is
high for the ivm stoves and lower for ivc stoves. Comparatively, the efficiency is higher in ivc,
which may be due to the ceramic lining and the firebox design that helps in proper airflow and in
turn enhances NCE.
4 As discussed in Appendix F, because of canister shipping problems, no N2O data are available for rootfuel and
dung. Consequently, we have estimated the N2O emissions by extrapolation from the measured wood and crop
residue emissions and relative N content in the fuels, as explained in the footnotes to Table 7.
31
-------
Table 7. Ultimate emissions by fuel mass on a pollutant mass basis (g/kg) and on a carbon mass basis (g-C/kg)
Fuel-Stove
Biogas
LPG
Kero-wick
Kero-pres
Root-imet
Acacia-imet
Eucal-ivc
Acacia-ivc
Must-imet
Eucal-3R
Eucal-imet
Acacia~tm
Root-ivm
Acacia-3R
Root-tm
Must-ivc
Acacia-ivm
Must-tm
Charbriq
Eucal-ivm
Dung-ivc
Charcoal
Rice-tm
Dung-ivm
Dung-tm
Must-ivm
Rice-ivm
Dung-hara
K-factor
0.0056
0.023
0.022
0.051
0.092
0.109
0.124
0.131
0.102
0.082
0.107
0.099
0.110
0.115
0.089
0.160
0.212
0.113
0.156
0.239
0.138
0.202
0.106
0.126
0.142
0.279
0.249
0.209
CO2
1444
3085
3027
2943
1737
1373
1477
1349
1352
1536
1500
1391
1710
1374
1742
1135
1260
1302
1594
1338
1046
2411
1101
1065
1027
1069
983.0
974.0
CO
1.950
14.93
17.65
62.10
74.68
63.61
87.96
79.04
55.97
60.15
64.71
66.47
75.89
64.70
49.98
55.34
125.8
65.57
120.6
139.1
31.62
275.1
48.70
30.31
49.58
94.10
101.0
61.39
By Pollutant Mass (g/kg) By Pollutant Carbon Mass (g-C/kg)
CH4 TNMOC N2O TSP CO2 CO CH4 TNMOC N2O TSP
1.005 0.5670 0.0950 0.5250 393.8 0.8357 0.7538 0.3780 0.0605 0.2470
0.0500 18.78 0.1470 0.5140 841.4 6.399 0.0375 12.52 0.0935 0.5140
0.2880 14.86 0.0790 0.5160 825.5 7.564 0.2160 9.907 0.0503 0.4490
1.071 19.20 0.1020 0.7010 802.6 26.61 0.8033 12.80 0.0649 0.6990
3.501 11.77 0.4764 1.176 473.6 32.01 2.626 7.847 0.3032 1.123
4.111 9.777 0.2765 3.811 374.5 27.26 3.083 6.518 0.1760 3.839
5.051 9.436 0.1722 2.107 402.9 37.70 3.788 6.290 0.1096 2.088
3.422 12.621 0.2048 3.320 368.0 33.88 2.566 8.414 0.1303 3.349
3.840 12.65 0.1620 2.224 368.7 23.99 2.880 8.433 0.1031 2.224
2.833 7.982 0.0728 0.9416 418.9 25.78 2.125 5.321 0.0463 0.9358
3.883 16.60 0.1922 2.463 409.0 27.73 2.912 11.06 0.1223 2.156
3.936 7.762 0.0921 1.038 379.5 28.49 2.952 5.174 0.0586 0.8880
3.864 18.76 0.4470 3.969 466.3 32.52 2.898 12.50 0.2845 3.364
9.399 9.653 0.1782 2.054 374.8 27.73 7.049 6.435 0.1134 1.974
11.69 16.30 0.4890 1.040 475.1 21.42 8.766 10.87 0.3112 1.021
4.792 26.92 0.1770 4.251 309.5 23.72 3.594 17.95 0.1126 4.258
10.79 11.94 0.1929 3.001 343.6 53.92 8.093 7.961 0.1227 3.022
7.580 8.487 0.0490 0.6310 355.1 28.10 5.685 5.658 0.0312 0.6310
5.335 16.13 0.1590 2.859 434.7 51.68 4.001 10.75 0.1012 1.431
11.45 25.13 0.1592 2.532 364.9 59.63 8.589 16.75 0.1013 2.324
3.580 31.68 0.3140 2.050 285.3 13.55 2.685 21.12 0.1998 2.032
7.906 10.48 0.2410 2.375 657.5 117.9 5.930 6.987 0.1534 2.049
5.390 9.390 0.2200 0.8050 300.3 20.87 4.043 6.260 0.1400 0.8020
3.250 29.49 0.3190 1.645 290.5 12.99 2.438 19.66 0.2030 1.631
5.700 18.81 0.3080 2.210 280.1 21.25 4.275 12.54 0.1960 1.609
24.92 27.87 0.1830 3.702 291.5 40.33 18.69 18.58 0.1165 3.707
4.240 8.036 0.1970 15.47 268.1 43.29 3.180 5.357 0.1254 14.85
17.56 23.22 0.2920 0.5500 265.6 26.31 13.17 15.48 0.1858 0.5450
*For those fuel-stove combinations where N2O measurements are missing, the emission ratios were extrapolated from those for the
same fuel or the fuel with a similar nitrogen content.
-------
Table 8. Ultimate emission factors of pollutant mass by fuel energy content (g/MJ) and delivered energy to pot (g/MJ-del)
Fuel-Stove
Biogas
LPG
Kero-wick
Kero-pres
Root-imet
Acacia-imet
Eucal-ivc
Acacia-ivc
Must-imet
Eucal-3R
Eucal-imet
Acacia~tm
Root-ivm
Acacia-3R
Root-tm
Must-ivc
Acacia-ivm
Must-tm
Charbriq
Eucal-ivm
Dung-ivc
Charcoal
Rice-tm
Dung-ivm
Dung-tm
Must-ivm
Rice-ivm
Dung-hara
Energy
(kJ/kg)
17710
45840
43120
43120
15480
15100
15330
15100
16530
15330
15330
15100
15480
15100
15480
16530
15100
16530
15930
15330
11760
25720
13030
11760
11760
16530
13030
11760
Overall
Eff=ri
0.574
0.536
0.500
0.470
0.228
0.257
0.287
0.290
0.217
0.177
0.214
0.182
0.197
0.181
0.142
0.185
0.235
0.124
0.164
0.220
0.128
0.175
0.098
0.100
0.094
0.135
0.109
0.082
C02
81.54
67.30
70.20
68.25
112.2
90.95
96.37
89.36
81.79
100.2
97.83
92.15
110.4
91.01
112.5
68.66
83.43
78.77
100.1
87.28
88.95
93.74
84.50
90.56
87.33
64.67
75.44
82.82
By Fuel Energy (g/MJ)
CO CH4 TNMOC
0.1101
0.3257
0.4093
1.440
4.824
4.213
5.738
5.235
3.386
3.924
4.221
4.402
4.902
4.285
3.229
3.348
8.331
3.967
7.570
9.076
2.689
10.70
3.738
2.577
4.216
5.693
7.751
5.220
0.0567
0.00109
0.0067
0.0248
0.2262
0.2723
0.3295
0.2266
0.2323
0.1848
0.2533
0.2606
0.2496
0.6224
0.7550
0.2899
0.7146
0.4586
0.3349
0.7471
0.3044
0.3074
0.4137
0.2764
0.4847
1.508
0.3254
1.493
0.0320 0
0.4097 0
0.3446 0
0.4453 0
0.7604
0.6475
0.6155
0.8358
0.7653 0
0.5207 0
1.083
0.5140 0
1.212
0.6392
1.053
1.629
0.7908
0.5134 0
1.013
1.639
2.694
0.4075
0.7206
2.507
1.599
1.686
0.6167
1.974
N2O TSP
.00536 0.0296
.00321 0.0112
.00183 0.0120
.00237 0.0163
0.0308 0.0760
0.0183 0.2524
0.0112 0.1374
0.0136 0.2199
.00980 0.1345
.00475 0.0614
0.0125 0.1607
.00610 0.0687
0.0289 0.2564
0.0118 0.1360
0.0316 0.0672
0.0107 0.2572
0.0128 0.1988
.00296 0.0382
0.0100 0.1795
0.0104 0.1652
0.0267 0.1743
0.0094 0.0923
0.0169 0.0618
0.0271 0.1399
0.0262 0.1879
0.0111 0.2240
0.0151 1.187
0.0248 0.0468
C02
142.0
125.6
140.4
145.2
492.0
353.9
335.8
308.2
376.9
566.1
457.2
506.3
560.6
502.8
792.5
371.2
355.0
635.2
610.1
396.7
694.9
535.7
862.2
905.6
929.0
479.0
692.1
1010
By Delivered Energy (g/MJ-del)
CO CH4 TNMOC N2O TSP
0.1918
0.6076
0.8186
3.064
21.16
16.39
19.99
18.05
15.60
22.17
19.72
24.19
24.89
23.67
22.74
18.10
35.45
31.99
46.16
41.26
21.01
61.13
38.14
25.77
44.85
42.17
71.11
63.66
0.0989 0
0.00203 0
0.0134 0
0.0528 0
0.9920
1.059
1.148
0.7814
1.071
1.044
1.184
1.432
1.267
3.439
5.317
1.567
3.041
3.698
2.042
3.396
2.378
1.756
4.221
2.764
5.156
11.17
2.985
18.21
.0558 0.00935 0.0516
.7643 0.00598 0.0209
.6892 0.00366 0.0239
.9474 0.00503 0.0346
3.335 0.1350 0.3332
2.519 0.0713 0.9820
2.145 0.0391 0.4788
2.882 0.0468 0.7581
3.527 0.0452 0.6200
2.942 0.0268 0.3470
5.059 0.0586 0.7509
2.824 0.0335 0.3776
6.151 0.1466 1.301
3.532 0.0652 0.7515
7.415 0.2225 0.4733
8.803 0.0579 1.390
3.365 0.0543 0.8457
4.141 0.0239 0.3078
6.174 0.0609 1.094
7.452 0.0472 0.7507
21.05 0.2086 1.362
2.328 0.0535 0.5277
7.354 0.1723 0.6304
25.07 0.2713 1.399
17.02 0.2786 1.999
12.49 0.0820 1.659
5.658 0.1387 10.89
24.08 0.3028 0.5704
-------
The average emission factors (EFm) for various fuel/stove combinations are compared with other
reported values in Table 9. It shows that the CO2, CO and CH4 emission factors for LPG are
comparable to the emission factors for LPG found in Manila Pilot study. But the TNMOC
emission factor (19 g/kg) is much higher than reported in the Manila study. For kerosene wick
the CO2, TNMOC emission factors are close to the Manila study results. But CO and CFLj
emission factors are less than the Manila study results.
The CO emission factor for the kerosene wick stove is even less than that reported by TERI
(1987). For charcoal the CO2, CO, & CFL; emission factors of the present study are comparable
to the Manila study results, but TNMOC is higher. For fuelwood, the CO emission factors are
lower than the CO emission factor 100 g/kg reported in the Manila study, but fall in the range of
13-68 reported by TERI (1987) and the range 17-130 reported by Smith (1987). CO emission
factor for dungcake and crop residues are within the range reported by TERI (1987).
Figures 8, 9,10, and 11 show the emission factors by delivered energy (EFt) for CO2, CO, CFL;,
and TNMOC for various fuel/stove tested. Note the general agreement with the energy ladder
framework (Smith 1990; OTA, 1992); i.e., that efficiency increases and emissions per meal
decrease along a spectrum from solid to liquid to gaseous fuels.
E. Comparison with IPCC Default Emission Factors
Table 10 shows the default emission factors recommended by the IPCC (1997) for residential
fuel use. As can be seen by comparison with Table 7, the IPCC values generally lie within the
range of values found for various biomass-stove combinations in India. Compared to those for
kerosene and LPG, however, the IPCC values for "oil" and natural gas, however, are
substantially lower for CO, TNMOC, and N2O, although being similar for methane. These
differences indicate that the IPCC values are probably not suitable for use with these cooking
fuels, at least under Indian conditions.
F. Variation
To give an idea of the statistical variation, the COV for all Efm over the three
separate test runs, are presented in Table 11 (an error analysis is presented in Appendix G).
Here are comments by pollutant:
CO2 emissions show little variation across all fuel/stove combinations tested, i.e., COV < 0.1.
CO emissions exhibit intermediate levels of variation, i.e. 0.1>COV<0.4.
CFL; emissions show high COV (1.5) for the two gas stoves, probably because measured
fluegas concentrations were near background levels and the equipment detection limits.
Dung-hara exhibited a high COV (1.1) because one run had a particularly high level. All
other fuel/stove combinations exhibit COV < 0.8, with most <0.5.
TNMOC emissions all have COV < 1.0 with many < 0.3.
N2O emissions exhibit four COV above 1.0 with most of the rest between 0.5 and 1.0.
TSP emissions for biogas and charbriquette were above 1.0, but most others were below 0.5.
34
-------
Table 9. Comparisons of emission factors (g/kg) by fuel mass with results from other studies
Fuel-stove
LPG
Kero-wick
Charcoal
Acacia-imet
Acacia-tm
Must-imet
Dung-ivm
CO2
3085
3027
2411
1373
1391
1352
1065
This
CO
15
18
275
64
66
56
30
Study
CH4
0.05
0.3
7.9
4.1
3.9
3.8
3.3
TNMOC
18.8
14.8
10.5
9.8
7.8
12.7
29.5
N2O
0.15
0.08
0.24
0.28
0.09
0.16
0.32
Manila Pilot
CO2 CO
3110
3030
2740
1560
24
38
230
99
Study Results (1)
CH4 TNMOC N2O
0.04
1
8
8
3 0.03
11 0.05
4 0.04
12 0.06
TERI (2)
CO
33-93
24-39
13-68
76-114
26-67
Other (3)
CO
17-130
Source: 1 Smith etal., 1992
2TERI, 1987
3 Smith, 1987
Table 10. IPCC default (uncontrolled)
Gas1
Oil2
Wood
Charcoal
Dung/ Agricultural Wastes 3
CO
2
0.9
80
200
68
emission factors for residential fuel combustion (g/kg)
CH4
0.2
0.4
5
6
4
TNMOC
0.2
0.2
9
3
8
N2O
0.005
0.03
0.06
0.03
0.05
1 Determined using the IPCC emission factors given for "Natural Gas" and the net calorific value given for "LPG"
Determined using the IPCC emission factors given for "Oil" and the net calorific value given for "Other Kerosene"
3 Determined using the IPCC emission factors given for "Other Biomass and Wastes" and the average of the net calorific values given
for "Dung" and "Agricultural Waste"
Source: IPCC, 1997
-------
Figure 8. Carbon Dioxide Emission Factors
PerMJ Delivered to the Pot
Biogas
LPG
Kero-wick
Kero-pres
Acacia-ivc
Acacia-imet
Eucal-ivc
Eucal-3R
Acacia-tm
Eucal-imet
Acacia-3R
Acacia-ivm
Eucal-ivm
Root-imet
Root-ivm
Root-tm
Must-imet
Must-ivc
Must-tm
Rice-tm
Rice-ivm
Must-ivm
Dung-ivc
Dung-ivm
Dung-tm
Dung-flara
Charcoal
Charbriq
200
Grams PerMJ Delivered
400 600 800
1000
1200
36
-------
Figure 9. Carbon Monoxide Emission Factors
PerMJ Delivered to the Pot
Biogas
LPG
Kero-wick
Kero-pres
Acacia-ivc
Acacia-imet
Eucal-ivc
Eucal-3R
Acacia-tm
Eucal-imet
Acacia-3R
Acacia-ivm
Eucal-ivm
Root-imet
Root-ivm
Root-tm
Must-imet
Must-ivc
Must-tm
Rice-tm
Rice-ivm
Must-ivm
Dung-ivc
Dung-ivm
Dung-tm
Dung-Hara
Charcoal
Charbriq
Grams PerMJ Delivered
20 40 60
80
37
-------
Figure 10. Methane Emission Factors
PerMJ Delivered to the Pot
0.001
Biogas
Kero-wick
Kero-pres
Acacia-ivc
Acacia-imet
Eucal-ivc
Eucal-3R
Acaciatm
Eucal-imet
Acacia-3R
Acacia-ivm
Eucal-ivm
Root-imet
Root-ivm
Root-tm
Must-imet
Must-ivc
Must-tm
Rice-tm
Rice-ivm
Must-ivm
Dung-ivc
Dung-ivm
Dung-tm
Dung-flara
Charcoal
Charbriq
0.01
Grams PerMJ Delivered
0.1 1
10
100
38
-------
Figure 11. TNMOC Emission Factors
PerMJ Delivered to the Pot
0.01
Biogas
Kero-wick
Kero-pres
Acacia-ivc
Acacia-imet
Eucal-ivc
Eucal-3R
Acaciatm
Eucal-imet
Acacia-3R
Acacia-ivm
Eucal-ivm
Root-imet
Root-ivm
Root-tm
Must-imet
Must-ivc
Must-tm
Rice-tm
Rice-ivm
Must-ivm
Dung-ivc
Dung-ivm
Dung-tm
Dung-flara
Charcoal
Charbriq
Grams PerMJ Delivered
0.1 1 10
100
39
-------
Table 11. Coefficients of variation (COV) for measurements for 3 tests of each fuel-stove
combination
Biogas
LPG
Kero-wick
Kero-pressure
Root-imet
Acacia-imet
Eucal-ivc
Acacia-ivc
Must-imet
Eucal-3R
Eucal-imet
Acacia-tm
Root-ivm
Acacia-3R
Root-tm
Must-ivc
Acacia-ivm
Must-tm
Charbriq
Eucal-ivm
Dung-ivc
Charcoal
Rice-tm
Dung-ivm
Dung-tm
Must-ivm
Rice-ivm
Dung-Hara
C02
0.017
0.052
0.068
0.046
0.042
0.19
0.036
0.055
0.019
0.062
0.076
0.029
0.087
0.034
0.11
0.049
0.055
0.059
0.076
0.12
0.087
0.12
0.10
0.009
0.013
0.046
0.062
0.077
CO
0.41
0.15
0.30
0.14
0.27
0.28
0.41
0.27
0.40
0.13
0.26
0.14
0.18
0.10
0.55
0.30
0.13
0.36
0.21
0.051
0.35
0.21
0.24
0.30
0.09
0.29
0.59
0.22
CH4
1.49
1.47
0.60
0.22
0.34
0.40
0.73
0.33
0.53
0.37
0.56
0.33
0.42
0.29
0.81
0.39
0.11
0.42
0.51
0.24
0.44
0.47
0.59
0.57
0.20
0.57
0.32
1.10
TNMOC
1.01
0.28
0.13
0.19
0.64
0.074
1.05
0.13
0.62
0.17
0.43
0.14
0.31
0.16
0.083
0.51
0.18
0.89
0.31
0.27
0.20
0.092
0.15
0.16
0.09
0.17
0.062
0.24
TSP
1.26
0.18
0.65
0.28
0.33
0.36
0.31
0.15
0.38
0.18
0.18
0.10
0.53
0.36
0.68
0.48
0.32
0.15
1.18
0.11
0.26
0.38
0.22
0.20
0.16
0.091
0.81
0.21
40
-------
V. DISCUSSION: National GHG Inventory and Fuel/Stove Comparisons
A number of analyses can be done with the database developed in this study. In Section I
(Introduction and Summary) we showed comparisons of global warming implications by fuel.
Below we examine briefly two additional issues: national GHG inventory and fuel/stove
comparisons.
A. Indian GHG Inventory
To determine the inventory of GHG emissions from cookstoves, an accurate fuel use
estimation is needed. The details of our estimation are presented in Appendix H.
The estimated emission factors for various fuel/stove combinations were averaged and used to
determine the GHG inventory. We tested two types of improved stoves improved mud and
improved mud with ceramic coating. At present in India, the latter are not widely disseminated.
Thus we have taken the weighted average of the improved mud: improved mud with ceramic
coating at the ratio of 90:10 as the emission factor for improved stoves.
Similarly for wood species in traditional stove we have taken the weighted average of wood in
traditional mud and 3-rock in the ratio of 90:10. The results from the two wood species measured
here were averaged. We tested two kinds of crop residues: mustard stalk and rice straw. In most
of India, only stalk variety is used as a fuel and straw is mainly used as cattle fodder. So it is
assumed that all crop residues are stalk variety in the emission calculations. The weighed
average emission factors and estimated greenhouses emissions from various stove fuel
combinations used in India are given in the Table 12.
The estimates of GHG emissions summarized by fuel are summarized in Table 13 where it can
be seen that by far the highest emissions from Indian households are from biomass burning
stoves. The estimates were compared with the earlier reported values. Mehra and Damodaran
(1993) quoted that the GHG emissions from biomass burning for the year 1989-90 were as 554,
35.22, 2.02 and 0.018 Tg/y for CO2, CO, CH4, and N2O respectively. These estimates include
the emissions from biomass combustion in other sectors such as small industry and forest fires.
But the CH4 estimate in the present study is similar to this earlier estimate. The N2O emission
estimates are same as the values reported by Mehra and Damodaran (1993). Methane emissions
from biomass combustion in India during 1990 were also estimated by Mitra and Bhattacharya
(1998) using IPCC default emission factors of 1.4 Tg/year (plus about 0.1 Tg from charcoal
production), which are lower than estimated here because of their use of IPCC default fuel-use
factors rather than results of actual energy surveys done in India.
41
-------
Table 12. Weighed average emission factors and GHG emissions from major fuel/stove combinations in India (1990-91)
Fuel/stove
Gas
Biogas
LPG
Kerosene
Wick
Pressure
Fuel wood
Traditional mud
Improved mud
Improved metal
Crop residues
Traditional mud
Improved mud
Improved metal
Dung cake
Traditional mud
Kara
Improved mud
Charcoal
Angethi
C02
1444
3085
3027
2943
1397
1980
1437
1302
1076
1352
1027
974
1063
2411
Emission factor
CO CH4
2
15
18
62
66
128
64
66
90
56
50
61
31
275
1
0.05
0.3
1
4
13
4
7.6
23
3.8
6
18
3
8
(g/kg)
TNMOC
0.6
18.8
14.9
19
8
24
13
8.5
27.8
12.7
18.8
23.2
29.8
10.5
N2O
0.09
0.15
0.08
0.1
0.09
0.28
0.24
0.05
0.18
0.16
0.31
0.29
0.32
0.24
C02
0.962
6.479
6.538
5.356
270.3
12.6
1.02
76.3
3.1
0.41
32.5
18.6
2.9
1.2
GHG
CO
0.001
0.032
0.039
0.113
12.8
0.81
0.045
3.89
0.26
0.017
1.58
1.17
0.09
0.14
emissions
CH4
0.001
0.0001
0.0006
0.002
0.77
0.084
0.003
0.445
0.067
0.001
0.190
0.344
0.008
0.004
(Tg/y)
TNMOC
0.0004
0.039
0.032
0.035
1.55
0.16
0.009
0.498
0.081
0.004
0.595
0.444
0.083
0.005
N2O
0.00006
0.0003
0.0002
0.0002
0.018
0.002
0.0002
0.003
0.001
0.00005
0.010
0.006
0.001
0.0001
-------
Table 13. Inventory of GHG emissions from India (1990-91)
Fuel
Biofuels
LPG
Kerosene
Biogas
CO 2 (Tg/y)
418.9
6.48
11.9
0.962
CO (Tg/y)
20.74
0.0315
0.152
0.001
CH4 (Tg/y)
1.92
0.0001
0.0025
0.0007
TNMOC (Tg/y)
3.41
0.0395
0.068
0.0004
N20 (Tg/y)
0.033
0.0003
0.0004
0.00006
B. Fuel/Stove Comparisons
The data developed in this study can be used to evaluate the global warming commitment (GWC)
of the different fuel/stove combinations and thus calculate the global warming implications of
policies to promote or discourage particular combinations. To calculate GWC, however, it is
necessary to make two choices:
whether to assume renewable or non-renewable harvesting of biomass fuels. If renewably
harvested, then the carbon dioxide in the biomass fuels is completely recycled and there is no net
increase in GWC from CO2. The GWC from the PIC, however, which are higher than CO2 per
carbon atom, must still be considered. In non-renewable harvesting, all the carbon in biomass is
a net addition to the atmosphere, as for fossil fuels. Here we examine both options. Note that
crop residues, dung, and biogas are assumed to always derive from renewable harvesting and the
LPG and kerosene are always non-renewable. It is only wood, root, and char fuels that vary.
whether to include the global warming commitments from CO and TNMOC, which are not as
well characterized as those from CO2, CH4 and N2O (TPCC, 1995). Here, we term GWC from
CO2, CH4, and N2O as GWC(basic) and that from CO2 plus CH4, CO, TNMOC, and N2O as
GWC(full).
With these considerations in mind,
GWC (global warming commitment) = sum over i of GHG;*GWP;
(17)
where GHG; is the gas of concern and GWP; is the global warming potential of that
particular GHG (total warming per molecule compared to CO2). See Glossary for the
particular GWP; used in this report.
Figure 12 shows the ranking of GWC(ren) and GWC(non-ren) using the full set of GHG. Note
that all biomass fuels, except biogas, have substantially higher GWC(non-ren) per standard meal
than any of the fossil fuels tested. This is because of the low combustion and thermal
43
-------
efficiencies of biomass stoves, even improved ones, compared to the liquid and gaseous fuels. In
the case of GWC(ren), a few of the wood and root stoves are comparable to the kerosene stoves,
and two wood stoves (Acacia-ivc and Eucal-imet) actually do better than LPG.
Figure 13 shows the same calculations using only the basic set of GHG. In this case, several of
the dung and crop residue stoves are comparable to kerosene and LPG for GWC(non-ren). In the
case of GWC(ren), however, 15 of the biomass stoves have comparable or lower GWCs than the
fossil-fuel stoves.
Although it is not the purpose here to provide detailed evaluation of individual stove types, it is
useful to note the relatively poor overall performance of the improved vented mud stove (ivm).
With both crop residues and both wood species tested, ivm was the worst performer among all
stoves. The reason can be gleaned from Figures 5 and 7, which show that with all these fuels,
the superior HTE of the ivm stoves was overwhelmed by decreased NCE, resulting in high GWC
per delivered energy even though fuel use was generally lower, as shown in Table 5. This
counter-intuitive result, i.e., that improvements in stoves that result in higher fuel efficiency can
still lead to greater emissions per unit delivered energy, is consistent with previous studies
(Smith, 1995).
44
-------
Figure 12. GWC-full per MJ Delivered
Along Energy Ladder
Grams Carbon as CO2
1
Biogas
LPG
Kero-wick
Kero-pres
Acacia-ivc
Acacia-imet
Eucal-ivc
Eucal-3R
Acacia-tm
Eucal-imet
Acacia-3R
Acacia-ivm
Eucal-ivm
Root-imet
Root-ivm
Root-tm
Must-imet
Must-ivc
Must-tm
Rice-tm
Rice-ivm
Must-ivm
Dung-ivc
Dung-ivm
Dung-tm
Dung-nara
10 100 1000
I i
\
V\
; '<°,
\
i *
^KfcMiewable (exuepllui Keiusene and Lrtj)
^Nonrenewable Wood and Root
=ull GWC = CO2, CH4, N2O, CO, TNMOC|
45
-------
Figure 13. GWC-basic per MJ Delivered
Along Energy Ladder
Grams Carbon as CO2
1
Biogas
LTG
Kero-wick
Kero-pres
Acacia-ivc
Acacia-imet
Acacia tm
Eucal-imet
Acacia-3R
Acacia-ivm
Eucal-ivm
Root-imet
Root-ivm
Root-tm
Must-imet
Must-ivc
Must-tm
Rice-tm
Rice-ivm
Must-ivm
Dung-ivc
Dung-ivm
Duna-tm
Dung-Hara
10 100 1000
: m
- 1
VA
N ?
* .V,
a
! %D
^Renewable (except for Kerosene and LP»j)
^Nonrenewable Wood and Root
|Basic GWC = CO2, CH4, N2O|
46
-------
VI. REFERENCES
Ahuja DR, Joshi V, Smith KR, and Venkataraman C. 1987
Thermal performance and emission characteristics of unvented biomass-burning
cookstoves: A proposed standard method for evaluation
Biomass 12: 247-270.
Ballard-Tremeer G and Jawurek HH. 1996
Comparison of five rural, wood-burning cooking devices: efficiencies and emissions
Biomass andBioenergy 11 (5): 419-430.
BIS. 1970
Methods for Measurement of Air Pollution: Part II Sulfur Dioxide
[IS: 5182 (part II)- 1969]
New Delhi: Bureau of Indian Standards. 11 pp.
BIS. 1975
Methods for Measurement of Air Pollution: Part VI Nitrogen Oxides
[IS: 5182 (part II)-1975]
New Delhi: Bureau of Indian Standards. 6 pp.
BIS. 1978
Specifications for Liquefied Petroleum Gases
[IS: 4576-1978]
New Delhi: Bureau of Indian Standards. 8 pp.
BIS. 1984
Specification for Domestic Gas Stoves for Use with Liquefied Petroleum Gases
[IS: 4246-1984]
New Delhi: Bureau of Indian Standards.
BIS. 1987
Indian Standard Methods of Tests For Coal and Coke (Part I: Proximate Analysis)
[IS: 1350-1987]
New Delhi: Bureau of Indian Standards.
BIS. 1991
Indian Standard Methods of Tests for Solid Biomass Chulha
[IS: 13152-1991]
New Delhi: Bureau of Indian Standards.
Census of India. 1991
Final Population Totals: Brief Analysis of Primary Census Abstract
New Delhi: Registrar General and Census Commissioner, Ministry of Home Affairs.
47
-------
Crutzen PJ and Andreae MO. 1990
Biomass burning in the tropics: Impact on atmospheric chemistry and biochemical cycles
Science 250: 1669-1678.
Davidson CI, Borrazzo JE, and Hendrickson CT. 1987
Pollutant emission factors for gas stoves: a literature survey
Research Triangle Park, NC, U.S. EPA. [Report EPA-600/9-87-005 (NTIS PB 87-171328)].
FAO. 1993
Indian improved cookstoves: A compendium food and agricultural organization of the
United Nations. P. 80 [Table No. 1.2]
Bangkok: Food and Agricultural Organization. 104 pp.
FAO. 1994
FAO Year Book: Forest Products 1981-1992 P 27
Rome: Food and Agricultural Organizations of the United Nations [FAO Forestry series No. 27].
George R. 1997
Commercialization of technology for domestic cooking applications
In Biomass energy systems, edited by P Venkata Ramana and S N Srinivas
New Delhi: Tata Energy Research Institute. 478 pp.
[Proceedings of the International Conference, New Delhi, 26-27 Feb. 1996, organized by Tata
Energy Research Institute].
HaoWMetal.1990
Biomass burning: an important source of atmospheric CO, COi, and hydrocarbon
Proceedings of the 1990 Chairman Conference on Global biomass burning,
Cambridge, MA: MIT Press.
Hayes P and Smith KR. Eds. 1994
The Global Greenhouse Regime: Who Pays?
London: Earthscan.
Houghton JT, etal (eds.) 1996
Climate Change 1995: The Science
United Kingdom: Cambridge University Press.
IPCC (Intergovernmental Panel on Climate Change). 1990
Climate Change: The IPCC Scientific Assessment
United Kingdom: Cambridge University Press.
IPCC. 1995
Climate Change 1994: Radiative Forcing of Climate Change
United Kingdom: Cambridge University Press.
48
-------
IPCC. 1997
Guidelines for National Greenhouse Gas Inventories: Reference Manual (rev. 1996)
Vol 3, United Kingdom: Blacknell.
Johnson RL, Shah JJ, Gary RA, and Huntzicker JJ. 1981.
An automated thermal-optical method for the analysis of carbonaceous
aerosol
ES Macias ES and Hopke PK, eds., ACS Symposium Series No. 167
Atmospheric Aerosol: Source/Air Quality Relationships.
Joshi V and Sinha C S. 1993.
Energy demand in the rural domestic sector.
UrjaBharati 3(3): 20-28.
Khalil MAK, Rasmussen RA, Wang MX, and Ren L. 1990
Sources of methane in China: Rice fields, biogas pits, cattle, urban areas, and wetlands
83r Annual Meeting, Air and Waste Management Association, 90-139.5, June.
Kishore VVN and Dhingra S. 1990
A gas regulator for fixed dome biogas plants
Changing Villages 9(1): 32-39.
Kishore VVN and Joshi V. 1995
Greenhouse gas emission from cookstoves
Energy Environment Monitor 11(5): 161-166.
Levine JS., ed. 1996
Biomass Burning and Global Change, Vol. 1 & 2
Cambridge, MA: MIT Press.
Lionel T, Martin RJ, and Brown NJ. 1986
A comparative study of combustion in kerosene heaters
Environment Science and Technology 20: 78-85.
Mehra M and Damodaran M. 1993
Anthropogenic emissions of greenhouse gases in India (1989-90)
Climatic change Agenda: An Indian perspective, Amrita N Achanta (ed.)
New Delhi: Tata Energy Research Institute.
Mitra AP and Bhattacharya S 1998
Greenhouse Gas Emissions in India for 1990 Using IPCC Standard Methodology
Scientific Report #11: New Delhi: Centre on Global Change, National Physical Laboratory.
49
-------
MOF. 1992
Economic Survey 1993-1994.
New Delhi: Ministry of Finance, Government of India Press.
MoPNG. 1993
Indian Petroleum & Natural Gas Statistics 1991-1992 P. 51
New Delhi: Ministry of Petroleum and Natural Gas. 218 pp.
National Council of Applied Economic Research (NCAER). 1985
Domestic Fuel Survey with Special Reference to Kerosene
New Delhi: National Council of Applied Economic Research.
NCAER. 1992
Evaluation survey of household biogas plants setup during the 7th five year plan (vol. I)
New Delhi: National Council of Applied Economic Research, pp. 243.
Office of Technology Assessment (OTA) 1992
Fueling Development: Energy Technologies for Developing Countries
OTA-E-516
Washington, DC: USGPO No. 052-003-01279-1.
Pal RC and Joshi V. 1989.
Improved cook stoves for household energy management - a case study
Productivity 30(1): 53-59.
RamanaPV. 1991.
Biogas Programme in India
TERI Information Digest on Energy, 1(3): 1- 12.
Rasmussen RA and Khalil MAK. 1981
Atmospheric methane trend: trends and seasonal cycles
J. Geophysical Research 86:5172-5178.
Rasmussen RA, Khalil MAK, and Chang JS. 1982
Atmospheric trace gases over China
Environ. Sci. & Technol 16: 124-126.
SalariyaKS. 1983
Energy Conservation in Domestic Combustion, p. 22 [Table 218]
Ludhiana: Punjab Agricultural University, pp. 69.
Schlamadinger B, et al. 1997
Towards a standard methodology for greenhouse gas balances of bioenergy systems in
comparison with fossil energy systems
Biomass & Bioenergy 13(6): 359-375.
50
-------
Smith KR. 1987
Biofuels, Air Pollution, and Health
New York: Plenum., pp. 452.
Smith KR. 1990
Indoor air pollution and the risk transition
in H. Kasuga, ed., Indoor Air Quality,
Springer-Verlag, Berlin, pp. 448-456.
Smith KR. 1995
Health, energy, and greenhouse-gas impacts of biomass combustion
Energy for Sustain. Develop 1(4): 23-29.
Smith KR, Rasmussen RA, Manegdeg F, and Apte M. 1992
Greenhouse Gases from Small-Scale Combustion in Developing Countries: A Pilot Study in
Manila
Research Triangle Park, NC: U.S. Environmental Protection Agency [EPA/600/R-92-005 (NTIS
PB92-139369)] pp. 67.
Smith KR, Khalil MAK, Rasmussen RA, Apte M, and Manegdeg, F. 1993
Greenhouse gases from biomass and fossil fuel stoves in developing countries:
A Manila pilot study
Chemosphere 26(1-4): 479-505.
Smith KR, Apte MG, Yuring M, Wongselziarttirat W, and Kulkarni A. 1994
Air pollution and the energy ladder in Asian cities
Energy 19(5): 587-600.
Smith, KR, Pennise DM, Khummongkol P, Chaiwong V, Ritgeen K, Zhang J, Panyathanya W,
Rasmussen RA, and Khalil MAK, 1999
Greenhouse Gases from Small-Scale Combustion in Developing Countries: Charcoal-
making Kilns In Thailand,
Research Triangle Park, NC: U.S. Environmental Protection Agency [EPA/600/R-99-109 (NTIS
PB 2000-102245)].
Tata Energy Research Institute (TERI) 1985
Cookstove technology, p 57-61.
Bombay: [now in New Delhi] Tata Energy Research Institute, pp.157.
TERI. 1987
Evaluation of Performance of Cookstoves with Regard to Thermal Efficiency and
Emissions from Combustion.
New Delhi: Tata Energy Research Institute.
51
-------
TERI. 1997
TEDDY (TERI Energy Data: Directory and Yearbook 1997/98)
New Delhi: Tata Energy Research Institute.
UNDP. 1997
Energy After Rio: Energy for Sustainable Development
New York, NY: United Nations Development Programme.
USEPA. 1993
Compendium Method TO-12a: Method for the Determination of Non-Methane Organic
Compounds (NMOC) in Ambient Air Using Cryogenic Preconcentration and Direct Flame
lonization Detection (PDFID).
Office of Research and Development
Washington, DC: U.S. Environmental Protection Agency.
USEPA. 1997
Compendium Method TO-14a: Determination of Volatile Organic Compounds (VOCs) in
Ambient Air Using Specially Prepared Canisters with Subsequent Analysis by Gas
Chromatography
Office of Research and Development
Washington, DC: U.S. Environmental Protection Agency.
VITA. 1985
Testing the Efficiency of Wood-burning Cookstoves: International Standards
Arlington, VA: Volunteers in Technical Assistance, Inc.
WazirS. 1981
Evaluation of Chulas
Bombay: Indian Institute of Technology. 60 pp. [Master of Technology thesis in Mechanical
Engineering].
WHO. 1984
Evaluation of Exposures to Airborne Particles in the Work Environment
Offset Pub. 80
Geneva: World Health Organization.
Zhang J and Smith KR. 1999
Emissions of carbonyl compounds from various cookstoves in China
Environmental Science & Technology 33(14): 2311-2320.
52
-------
Appendix A. Description of the Simulated Rural Kitchen (SRK)
The SRK is shown in Figures A-l - A-2. At 8 feet x 8 feet (244 cm x 244 cm) with the height
of the roof being 9 feet (275 cm) on one side sloping down to 8.5 feet (259 cm) on the other side,
the kitchen has a volume of 16 m3.
1) SRK details: A 6.5 feet x 3 feet (198 cm x 92 cm) door was fixed in the south wall for
entering the kitchen. There are three windows measuring of size 3 ft x 2 ft (92 cm x 62 cm)
fitted about 3 feet (92 cm) above the ground level. There is no window in the wall where the
door is fixed. There are four rectangular ventilators of size 2 ft x 1 ft (61cm x 31cm) fitted in
four walls. Out of these, two were placed in the bottom 1.0 ft above the ground level (BVi &
BV2) and the other two (TVi & TV2) were placed in the top (about 2.5 ft below the roof). In
addition to these rectangular ventilators, five circular ventilators (C V) with a diameter of 9 inches
(23 cm) are provided, out of which four were situated about 1.5 ft below the roof and one was
placed 3 inches (8 cm) above the ground level. The windows and ventilators were provided
primarily to vary the ventilation conditions if desired.
The entire laboratory was surrounded by an outer boundary wall with floor dimension (457 cm x
457 cm) of 15 feet x 15 feet and a height of 10 feet (305 cm). The function of the outer
enclosure is to reduce the wind effects and to keep uniform ventilation conditions in the hut
throughout the experiment. To reduce wind effects, the windows, ventilators, and door fitted in
the outer boundary wall were closed during all experiments. Between runs, however, they were
opened to facilitate comfort and to help bring indoor concentrations down to ambient levels.
A hood arrangement with an adjustable vertical height mechanism was set up on the one side of
the kitchen for collection of emissions gases. Also two wooden platforms of the size of 3 ft x 3ft
(92cm x 92cm) were fitted on two walls for keeping emissions gas collection bags (Tedlar bags).
One platform was fixed near the hood arrangement at a height of 3.5 ft (107 cm) from the ground
level. This was used to keep the Tedlar bag and sampler used for emissions gas collection.
Another platform was fixed near the door at a height of 2 ft (61cm) from the ground level. This
was used to keep the Tedlar bag and sampler used for simultaneous collection of indoor
background air. These two wooden platforms can be folded up and latched with the help of a
locking arrangement provided in the walls.
2) Hood arrangement for stoves without flue (chimney): The hood was designed so that it
collects a fairly high proportion of the emission gases, while not interfering in any way with the
normal combustion of the stove. Also the sample collected should represent the whole of the
combustion gases and not those from one particular point.
A hood consists of a skirt portion, 4"x 4" duct (10 cm x 10 cm), 6"x 6" (15 cm xl5 cm) duct and
an exit pipe. The skirt portion consists of 2 metal frames made up of 'L' section angles. One
frame is rectangular in shape with the size of 3 feet x 2.5 feet (91 cm x 75 cm). Size of another
metal frame is 4"x 4"(10 cm x 10 cm). These two frames were connected to each other by four
angles. The structure was covered with metal sheet. This gave the structure of convergent duct.
The top portion of the skirt (10 cm x 10 cm metal frame) was connected to 10 cm x 10 cm duct
-------
which was overlapped by 15 cm x 15 cm duct in a telescopic arrangement. The gap between the
two ducts was stuffed tightly with glass wool to prevent leakage.
The 15 cm x 15 cm duct was suitably bent and taken outside the kitchen wall through the circular
ventilator fitted on the kitchen wall. This was further connected to the outer wall with 23 cm
diameter circular PVC exit pipe. The exit pipe ends on the outer wall and an exhaust fan was
fitted at the end in the outer wall. During all experiments, the fan was run at a constant speed to
facilitate mixing and to maintain the constant flue flow rate needed for the carbon balance
method.
For stoves without flue, 1.5 feet (45 cm) table was used to place the stove. Asbestos sheet was
placed on the top of the table to withstand the high temperature. The height of the hood
arrangement was adjusted according to the height of the stove and vessel. The hood was fixed in
the metal rods fitted in the table with the help of screws. The gap between the hood and mouth of
the vessel was kept between 1.5 to 2 inches (4 -5 cm) to read the temperature in the
thermometer. A stainless steel monitoring probe was placed in the 10 cm x 10 cm duct of the
hood to collect samples. A thermocouple was also set near the probe to measure emission gas
temperature at the point of collection. Figure A-3 shows the hood arrangement for a stove
without flue.
3) Hood arrangement for stove with flue (chimney). The hood arrangement was modified
slightly to test stoves with flues (Figure A-4). The height of the hood was raised to its maximum
level (about 240 cm from the ground level) by reducing the length of the two ducts. The stove
was placed on the ground, with its chimney ending under the hood. A monitoring probe was
placed into the 23 cm PVC pipe that penetrated the inner and outer walls as shown.
54
-------
Figure A-l. Simulated rural kitchen (view from above)
4.6 m
B-
0.6 m 0.6 m
1.2 m
W1
TV1
\JV2
D1
Outer enclosure
'20
3V2 0.6 m
W3
A-
BV1
0.9 m
D2
D1 & D2=Door=2 m xO.9 m
wall thickness=2.54 cm
BV1 & BV2=Bottom ventilators=0.6 xO.3, 0.3 m
TV1 & TV2=Top ventilators=0.6 xO.3, 0.2 m
55
-------
Figure A-2. Simulated rural kitchen (section A-A')
274cm
ted I ar bag
MAX
-L , I
J , I .
'"-L , IL
-L , I
-L , I
-L , I
-L , I
-L , I
22.9cn
sampler \
3'x3- \\
15.2cmx15.2cm
TSP filter
10.2cmx10.2crr
259cm
56
-------
Figure A-3. Simulated rural kitchen (section B-B')
305
57
-------
Figure A-4. Hood arrangement for stove with flue
tsp filter
58
-------
Appendix B: Details of Stoves Tested
Traditional mud stove (TM) The wall thickness of the stove is about 3 cm. The height of the fire
box (from the bottom of the stove to bottom of the pot) is about 18 cm. Fuelwood, crop residues
and dung cakes are commonly used in this stove. A diagram is shown in Figure B-2 and a
photograph in Figure B-la.
Three-rock arrangement (3-rock). To represent the three-rock arrangement, three bricks (6 cm x
22 cm x 11 cm) were arranged at approximately 120° to one another. The pot hole size was fixed
as 190 mm diameter to keep 20 cm diameter pot. The stove can accommodate pots of 18-30 cm
in diameter. Figure B-3 shows the arrangement (see the photograph in Figure B-lc).
Improved Metal (EVIet). The stove is cylindrically shaped with metal stands. The top a circular
metal sheet is provided with a hole in the center and slots. A metal grate is provided at the
bottom for airflow and to ensure smooth combustion. The stove can accommodate pots of 18-30
cm in diameter. The stove is specifically suitable for fuelwood and twigs. The stove is
commercially available in the names of Priagni and Vishal. About 5 million stoves have been
disseminated in all parts of the country. Figure B-ld shows a photograph of a typical version.
Improved Vented Mud (FVM) The stove is constructed with sundried prefabricated clay slabs
(chapris). The slabs are made with a mixture of good clay and fibrous material such as chopped
crop residues. Because of this the slab becomes strong and does not crack on drying. The stove
consists of firebox, two potholes, connecting tunnel and chimney. The height from the firebox
floor to the lower edge of the cooking hole is about 18 cm. The height of the tunnel from the
ground level is about 2" (5 cm) at chimney and firebox ends. Whereas in the middle (at second
pot hole) the height of the tunnel from the ground level is about 4.5"(11 cm). This rise helps in
the maximum utilization of heat to the second pot. 3 "(8 cm) inner diameter cement pipe is used
as a chimney. Damper is provided between the second pothole and chimney to control the draft.
The whole surface of the stove is coated with clay, dung and crop residue mixture. Fuels such as
fuel wood, crop residues, and dungcakes can be used in this stove. The stove is mainly used in
rural areas of India (see Figure B-lf).
Improved Vented Ceramic (IVC). This stove is commonly called "Sugam" The stove is same as
IVM except the most critical four parts (two fireboxes, tunnel, and chimney) are made of
ceramic. The ceramic lining helps in heat retention, which helps improve combustion and
increases the efficiency of the stove. Presently the stove is disseminated in the villages of Uttar
Pradesh.
Kara This dung-burning stove is widely used in villages of Haryana, Uttar Pradesh, Punjab and
some parts of rural Rajasthan, Bihar and Madhya Pradesh. There are two designs of the Kara:
One is portable, but heavy, and made of a mixture of mud, clay, and crop residue. The other is
made of a similar mixture, but fixed in the ground. The portable version was chosen for the
study and is shown in Figures B-lb and B-4.
Angethi. This bucket stove has a 23 cm top diameter; 12 cm bottom diameter, and a height of 17
cm. It is divided into two halves by a grate and the inner wall of the bucket is coated with
59
-------
mud/concrete. There is a small air vent below the grate and three projections above the bucket to
form the pot seat. Charcoal, coal, and coke are the major fuels burned in this stove. For the
present study, charcoal and charbriquettes were tested. A diagram of the Angethi is given in
Figure B-le and a photograph in Figure B-5.
Kerosene wick. The weight of the empty stove is about 2.6 kg. The stove consists of fuel tank,
burner assembly and load bearing assembly. The fuel tank capacity of the stove is 2 liters. The
fuel tank is fitted with filter cap assembly, a kerosene level indicator (float) to indicate the level
of kerosene in the tank, and a wick control lever designed for raising/lowering the wicks to
control the intensity of the flame. The burner assembly consists of 10 wicks and inner and outer
sleeves. The space between the two sleeves is designed to supply more pre-heated air to ensure
better combustion. An insulated triple wall outer burner casing is provided to minimize the heat
loss. At the top of the burner assembly a load-bearing assembly (26.5 cm) is placed to provide
the platform for vessel. An optional triangular pan support is also provided to place small
utensils. The stove is used in all parts of India especially in urban areas (see Figure B-lh).
Kerosene pressure. The major units of the stove are fuel container, roarer type burner, and a top
ring. The fuel container is made up of brass sheet with a capacity of 2 liters. The fuel container
is fitted with a hand-operated pump, pressure release screw, and fuel filler cap assembly. The
pressure release screw is for releasing the container pressure quickly and safely. By decreasing
the pressure the flame can be adjusted. The fuel container is fitted with a socket and a spirit cup.
The fuel container rests on metallic legs, which are extended up to the top ring. The burner
assembly consists of a nipple, burner, and a flame ring. The top ring (21 cm diameter) is placed
on top of the burner assembly. Figure B-lg shows a diagram of the kerosene pressure stove. A
schematic is given in Figure B-6.
LPG stove. The stove is made up of stainless steel body for use with liquefied petroleum gases
sold in refillable tanks at 2.5-3.4 kPa (kN/m2) pressure. A tap is provided in the stove to control
the pressure. If the tap is turned "full on" the intensity of the flame is high. A detachable metal
frame is provided to support the pan. The stove is connected to the gas cylinder with rubber
tubing. A detachable regulator is provided at the end of the tube to connect to the cylinder. There
is a key in the regulator to control the supply of the gas from cylinder to the stove.
Biogas stove. There is a tap in the stove to control the intensity of the flame. The circular burner
has three rows of 4.7 mm holes as follows:
Pitch Hole
Diameter (mm) No. of holes
Inner row 40 6
Middle row 57 6
Outer row 72 23
60
-------
a. Traditional mud stove
p». - -.- ,,, --.,. ..-. . mm
-iii-js&KWs *"==>::"":v=-.r.-.. B|
b. Kara
I
I
c. Three-rock
d . Improved metal
e. Angethi
f. Improved vented mud
g. Kerosene pressure
h. Kerosene wick
Figure B-l (a-h). Photographs of the stoves tested in the study
61
-------
Figure B-2. Diagram of the traditional mud stove.
11 cm
Figure B-3. Diagram of the three-rock stove.
62
-------
, %
X A A A. ,.x.
X X X x X x
' v' V V V y'
X '* x
X X V V V >
V" V ''.' v ' ^ '
/ » / *
XA.X,
Cooking material
Dungcake
-Ground level
Figure B-4. Diagram of the hara stove.
Handle
Mud coating
Grate
Air vent
Figure B-5. Diagram of the Angethi stove.
63
-------
Burner
Oil cap
Oilfiller
Fuel container
Spirit cup
Burner socket
- Nipple
Pressure release screw
Pump
Figure B-6. Diagram of the kerosene pressure stove.
64
-------
Appendix C: Measurement Techniques
Analytic instruments used in this study are listed in Table C-l. Principles involved in the
measurement of moisture content, calorific value, total suspended particulates, sulfur dioxide,
nitrogen dioxide and GC analysis are given below.
Table C-l. Analytic instruments used
Instrument
Flow rate (l/m) Make
2
3
4
Air sampler
- SKC 224 43 X
- SKC 224 PC XR
- Gilian
- Casella AS 808
Gas Chromatograph
AIMIL-NUCON Series 5700
Spectrophotometer
UV-VIS Spectrophotometer 119
Bomb Calorimeter
Muffle Furnace
0-4.0
0-4.0
0-4.0
0-20
SKC, USA
SKC, USA
USA
UK
NUCON Engineers, India
Systronics INDIA
Toshniwal Instruments,
India
India
Moisture content (wet basis). To determine the moisture content of any fuel it is necessary that it
should be of small particle size. The wood was sawed to make sawdust in such a way that the
whole area, including cell wall, was included. About five pieces of the fuel samples taken from
different places were sawed and the sawdust obtained were mixed properly and used for moisture
content measurement. These steps were all carried out in triplicate.
A known quantity of sample was taken in a crucible and kept in an oven maintained at 105 °C till
the weight stabilizes. The weight loss was measured and the moisture content of the sample was
estimated as follows.
Moisture Content (M.C.) =
j -Wc
x 100
Wi = initial weight of sample
Wf = final weight of sample
Wc = weight of crucible
65
-------
Calorific value. Calorific value (energy content) of a fuel was determined by calorimetry.
Benzoic acid was used to standardize the bomb calorimeter. One gram of sample was taken in a
crucible and made into a pallet and the initial weight was noted. It was placed in the bomb,
which was pressurized to 18 atm of oxygen. The bomb was placed in a vessel containing a
measured quantity of water. The ignition circuit was connected and the water temperature noted.
After ignition the temperature rise was noted every minute till a constant temperature was
recorded. The pressure was released and the length of unburned fuse wire was measured. The
calorific value was calculated as:
(tcxw)-(m + n)
weight of sample (g)
tc = temperature rise ( C)
w = apparent heat capacity by benzoic acid (J)
m = calorific value of thread (J)
n = calorific value of Ni chrome ignition wire (J)
The apparent heat capacity by benzoic acid (w), calorific value of thread (m), and the calorific
value of Ni chrome ignition wire were provided by the instrument supplier.
TSP Measurement. Quartz fiber filters of 37 mm diameter (Pallflex Products Co., Putnam, CT,
USA) were used for Total Suspended Parti culate (TSP) measurements. The flow rate of the
sampling pump was adjusted to fill an 80-liter Tedlar bag throughout a burn cycle. The flow
pumps were calibrated before and after measurements using the soap bubble method (WHO,
1984). TSP was calculated from the filter weight difference and volume of air sampled.
Quartz fiber filters were conditioned by heating at 800 °C for 2 hours and then placed in a
desiccator for at least 24 hours before weighing. The filters were carefully placed in the filter
holders and used for sample collection. After sampling, the filters were taken out of the holder
and placed in a petri dish, desiccated for 24 hours and weighed. The net increase in the weight
of the filter after sampling was divided by the total flow to determine the concentration. One
filter from each fuel/stove combination was analyzed for carbon content.
Carbon contents of TSP collected on quartz fiber filters were measured using a thermal-optical carbon
analysis technique (Johnson et a/., 1981) at Sunset Laboratory, Forest Grove, OR, U.S.A.
Sulfur dioxide. The West and Gake method (BIS 1970) was followed to estimate sulfur dioxide in
emission gas and indoor background samples. The air samples were bubbled through the
absorbing media containing sodium tetrachloromercurate at a constant flow rate (1.5-2.0 1/m)
during the entire burncycle experiment. The non-volatile dichlorosulphitomercurate ion formed in
this process was reacted with acid bleached pararosaniline and formaldehyde to form a complex
ion, the absorbance of which was read spectrophotometrically at 560 nm. The corresponding SO2
66
-------
concentration was measured by comparing the absorbance with a standard graph developed with
known concentrations of 862. Sodium metabisulphite solution was used as a standard solution
for calibration (1 ml of 0.01 N metabi sulphite solution contains 320 |ig of 802).
Nitrogen oxides. Nitrogen oxides were measured as nitrogen dioxide by a modified Jacob and
Hochhier method (BIS 1975). Emissions and indoor background samples were bubbled through
an absorbing media containing sodium hydroxide - sodium arsenite solution to form a stable
solution of sodium nitrate. The nitrate ion produced during sampling was reacted with
phosphoric acid, sulphanilamide andN-(l-napthyl)-ethylenediamine dihydrochloride to form an
azo dye. The absorbance of the azo dye was read in spectrophotometer at 550 nm and the
corresponding concentration was estimated using a standard graph made with known NC>2
concentration. Sodium nitrate solution was used as a standard for NO2 calibration.
GC analysis. A gas chromatograph (GC) was set up to analyze background samples and samples taken
out of the filled Tedlar bags for CO2, CO, CH/i, and TNMHC. A system of GC-flame ionization detector
(FID) - methanizer was employed for analysis of CC>2, CO, and CH/i. In this system, a Carbonsphere-
packed column was used to separate these three compounds. The separated CO and CO2 were
converted by hydrogen at 375 °C in a nickel catalytic device (the methanizer) to Ctfy which was then
determined by the FID. TNMHC was measured by subtracting CH4 from the total hydrocarbon (THC)
which was determined using a FID and a blank GC column (the air peak was corrected). All GCs were
calibrated daily with locally made standards and periodically checked with a standard gas mixture of
CO2, CO, CFLjprepared by Scott Specialty Gases, Inc., Plumsteadville, PA, U.S.A. The agreement
between the locally made standards and US made standards was within ± 4%.
The filled canisters were shipped back to Oregon Graduate Institute of Science and Technology (OGIST)
to be analyzed mainly for hydrocarbon speciation. Up to 70 individual hydrocarbons were determined by
using the procedure established as EPA Compendium Method TO-14a (U.S. EPA , 1997), a method that
uses the GC to separate hydrocarbon species and uses the FID to determine the compounds (Rasmussen
and Khalil, 1981; Rasmussen et al., 1982; USEPA, 1993). These canister samples were also analyzed
for CO2, CO, and CFfy using similar analytical procedures to those used in the local laboratories and for
non-methane organic compounds (NMOC) using EPA Method TO-12a (USEPA, 1993). This provides
data for inter-laboratory comparison.
Two or more injections were made for each sample to ensure a RSD < 10%. Calibration curves
for all measured compounds were made daily and had linear regression R2 > 0.99. Results
obtained by the local GC analyses were compared with results of canister samples analyzed by
OGIST. When the measured concentrations were close to the method detection limit, the
agreement appeared poorest. The method detection limit, reported by the TERI laboratory, was 1
ppm for CO, CH4, and THC. The flue gas and background CO2 concentrations were much
higher than the CO2 detection limit.
67
-------
Appendix D: Calculation Procedures
Based on the measurements, power and thermal efficiency were estimated to check the thermal
performance of the stove.
1) Thermal efficiency
Thermal efficiency is the product of combustion efficiency and heat transfer efficiency.
Combustion efficiency measures the extent of which the chemical energy in wood is converted
into heat and subsequently used to evaporate water in the vessel. Heat transfer efficiency
indicates what fraction of the heat produced is actually transferred to the vessel and water. The
amount of heat used to evaporate water is considered as useful heat input to the vessel since the
primary interest is to compare stoves rather than cooking efficiency for any given stock. The
burn rate and net corrected calorific value of fuel are used in the calculation of thermal
efficiency. The equation for thermal efficiency calculation is given below (Ahuja et al., 1987).
T| (%) = {[W; * a * (Tf - T;) + (Wi - Wf )] * L / (F* t*Hw)} * 100
T| = efficiency (%)
W; = initial weight of water (kg)
a = specific heat of water (kJ/deg-kg)
Tf = temperature final (°C)
T; = temperature initial (°C)
Wf = final weight of water (kg)
L = latent heat of vaporization for water (kJ/kg)
F = burn rate (kg/h)
t = duration (hour)
Hw = net calorific value of main fuel (kJ/kg)
2) Burn rate
The burn rate is corrected for the amount of kerosene used as a lighter, the charcoal remaining
and the moisture content of the fuel wood. The burn rates for crop residues and dungcake
combustion are similarly calculated by replacing Ww and Hw by their appropriate values for the
two fuels. The burn rate calculation for kerosene stoves is more straightforward - weight of
kerosene consumed divided by experimental time.
lOOxW,,, W,H WH
100 + M H H
F = burn rate (kg/h)
t = duration of the experiment (hour)
Ww = weight of wood (kg)
= weight of kerosene (kg)
= calorific value of kerosene (kJ/kg)
68
-------
Wc = Weight of charcoal (kg)
Hc = Calorific value of charcoal (kJ/kg")
M = Moisture content of wood (%)
Hw = Calorific value of wood (kJ/kg)
3) Power
Power refers to the rate at which the energy is used. The power (kW) is calculated as follows:
Power (kW) = FX Hw x 1/860
F = burn rate (kg/h)
Hw = calorific value of main fuel (kJ/kg)
69
-------
Appendix E: Fuel Analyses
Solid fuels and kerosene were analyzed for carbon, ash, sulfur, nitrogen and hydrogen content
using standard methods (BIS 1987). For biogas the energy, carbon, and hydrogen content were
estimated from the gas analysis by GC/TCD method. For LPG, the energy content was given by
"BHARAT Petroleum Co." The chemical composition, moisture content and net (low heating
value) energy of the fuels using the methods in Appendix C are given in Table E-l.
Table E-l. Fuel chemical composition, moisture content, and net energy
Fuel
LPG
Biogas
Kerosene
Eucalyptus
Acacia
Root fuel
Charcoal
Char-
briquette
Mustard
straw
Rice straw
Dung cake
Moisture
content (%)
-
-
-
6.1
6.5
5.7
1.7
7.2
5.9
8.8
7.3
Net Energy
(kJ/kg)
45837
17707
(kJ/M3)1
43116
15333
15099
15480
25715
15928
16531
13027
11763
Carbon
86.0
39.6
84.3
45.4
41.8
51.8
80.0
50.3
42.1
38.1
33.4
Nitrogen
0.02
0.14
0.35
1.18
0.69
0.25
0.36
0.40
0.90
Ash
0.0
0.4
2.89
7.0
7.4
40.0
2.7
15.6
52.2
H2
6.5
14.2
6.4
6.3
4.5
1.8
3.2
6.3
6.2
3.9
Sulfur
0.04
0.02
0.01
0.08
0.06
0.05
0.01
0.05
0.07
1 standard temperature and pressure
The measurements are generally similar to those published for these fuels (Smith, 1987).
Dungcakes stand out because they have low carbon content, low net energy, and high ash
content. Ash content of 52% for dung cakes is higher than the earlier reported ash content of
about 15-20% and 31% (Smith 1987, Salariya 1983). The ash content in dungcake and char
briquettes is much higher than wood and root. This may be due to the presence of more dirt
particles in these fuels. The ash content of rice straw is close to the reported value of 15.5 %
(Salariya 1983).
70
-------
Appendix F: Measured Fluegas Concentrations
A. Total Suspended Particulates (TSP)
The net TSP concentrations (flue gas- indoor) for various fuel/stove tested are given in the Table
F-l. The increases in the TSP concentration for various fuel/stove were biogas-LPG- kerosene-
charcoal- rootfuel- dungcake- wood- crop residues. Also shown are the results of the carbon
analyses.
B. Gases
The TERI concentrations of CO2, CO, and CFLj in flue gas and indoor background samples for
three experiments are averaged in Table F-2. One of the three flue gas samples for each
fuel/stove combination was collected in stainless steel canisters and analyzed at OGIST for CO,
CO2, CH4 and TNMOC, as shown in Table F-3. TERI values were plotted against OGIST
r\
values in Figures F-l - F-3. The R values for the three regression analyses were all above 0.80.
Based the OGIST laboratory's extensive experience in GC analysis, we considered it as the
reference. Using x variable(m) and intercept(c), TERI CO2, CO, and CH4 values for each
experiments were corrected. For example, based on Figure F-l (CO), [OGIST data] = 8.32 +
0.52[TERI data]. The corrected values were reported here and used for emission factor
calculations.
Among 28 fuel/stove combinations, canisters for seven stove fuel combinations were opened by
Indian Customs during shipment. During the pilot phase experiments with Eucal-tm, CO2
calibration was not stabilized due to improper conditioning of the column. So TERI values for
the pilot phase experiments were not considered for comparison. Due to the GC problem during
the experiments with Rice-tm and Dung-ivm, TERI values for those experiments were not
reliable. For the rest of the experiments, TERI values were compared with OGIST results and
given in Table F-4. The corrected concentrations, net of background, shown in Table F-5, were
used for estimating the emission factors and emissions inventory.
The net concentrations (fluegas minus indoor) of SO2 and NOx (measured as NO2) for the
fuel/stove tested are given in Table F-6, which reveals that for SO2 the difference between flue
gas and indoor is marginal (less than Ippb) for LPG, Biogas, charcoal and charbriquette. For
crop residues the average net concentrations of SO2 vary from 0.7 to 2.9 ppb in different stoves.
For wood fuels the range for SO2 concentration is 1.2 to 6.3 ppb and for dungcakes the values
range from 0.3 to 6.3ppb.
Among the various fuel/stove tested, the net NO2 concentration is high for LPG(11 ppb). For wood
fuels the net NO2 concentrations vary from 1 to 4 ppb. For crop residues and dungcakes the net NO2
concentration did not exceed 5 ppb. The low NO2 emissions for biofuel are probably due to lower
combustion temperatures than the liquid and gaseous fuels, which are premixed with air before
combustion.
71
-------
Table F-l. Concentration of TSP and Carbon as TSP. Net = flue level minus background.
Standard deviations shown.
Fuel/Stove
LPG*
Biogas
Kerosene/wick
Kerosene/pressure
Charbriquette
Charcoal
Eucal-tm*
Eucal-3 rock*
Eucal-imet
Eucal-ivm
Eucal-ivc
Acacia-tm
Acacia-3 rock
Acacia-imet
Acacia-ivm*
Acacia-ivc*
Root-tm
Root-imet
Root-ivm
Mustard-tm
Mustard-imet
Mustard-ivm
Mustard-ivc
Rice-tm*
Rice-ivm*
Dung-tm
Dung-hara
Dung-ivm*
Dung-ivc*
TSP (mg/m3 ) in
Flue gas
0.68±(0.13)
0.80 ±(0.33)
0.82 ± (0.22)
1.06 ±(0.19)
3.54 ±(0.23)
2.02 ±(0.70)
3. 10 ±(0.04)
4.19 ±(0.56)
4.24 ±(1.06)
4.76 ±(0.41)
3.51 ±(0.58)
3.67 ±(0.12)
3.38 ±(0.63)
3.87 ±(0.80)
4.73 ± (0.75)
5.00 ±(0.19)
2.93 ±(1.63)
3.29 ±(0.63)
2.43 ± (0.84)
4.09 ±(0.12)
4.68 ±(1.07)
6.49 ±(1.45)
7.26 ±(0.31)
6.53 ±(0.73)
6.60 ±(1.20)
5.03 ±(0.78)
2.96 ±(0.09)
4.05 ± (0.24)
4.61 ±(0.32)
TSP (mg/m3)
Background
Level
0.36 ±(0.15)
0.55 ±(0.14)
0.36 ±(0.12)
0.58 ±(0.15)
0.67 ± (0.49)
0.53 ±(0.21)
0.26 ±(0.11)
0.57 ±(0.28)
0.87 ±(0.48)
0.42 ±(0.09)
0.36 ±(0.09)
3.57 ±(0.21)
0.42 ±(0.14)
0.35 ± (0.26)
0.42 ±(0.11)
0.35 ±(0.11)
0.61 ±(0.15)
0.35 ±(0.13)
0.41 ±(0.09)
0.55 ±(0.18)
0.64 ± (0.27)
0.75 ± (0.09)
0.57 ±(0.15)
1.28 ±(0.61)
0.57 ±(0.12)
0.98 ±(0.16)
0.58 ±(0.13)
0.28 ± (0.02)
0.26 ±(0.08)
TSP (mg/m3)
Net Cone, in
Flue
0.32 ±(0.14)
0.25 ±(0.21)
0.46 ±(0.32)
0.48 ±(0.05)
2.87 ±(0.36)
1.49 ±(1.00)
2.84 ±(0.13)
3.62 ±(0.58)
3.37 ±(0.59)
4.34 ±(0.38)
3. 15 ±(0.57)
3.09 ±(0.17)
2.96 ±(0.67)
3.52 ±(1.05)
4.32 ±(0.64)
4.66 ±(0.17)
2.32 ±(1.49)
2.94 ±(0.58)
2.02 ±(0.88)
3.54 ±(0.10)
4.04 ±(0.85)
5.74 ±(1.37)
6.69 ±(0.26)
5.25 ±(0.79)
6.02 ±(1.17)
4.05 ± (0.66)
2.38 ±(0.16)
3.77 ±(0.26)
4.35 ± (0.25)
Net Cone, of
Carbon as TSP
(mg/m3)
0.32 ±(0.14)
0.12 ±(0.09)
0.41 ±(0.24)
0.48 ±(0.05)
2.18 ±(1.00)
1.27 ±(0.49)
2.84 ±(0.13)
3.62 ±(0.58)
2.83 ±(0.21)
3.99 ±(0.99)
3. 15 ±(0.57)
2.54 ±(0.80)
2.83 ±(0.81)
3.52 ±(1.05)
4.32 ±(0.64)
4.66 ±(0.17)
2.32 ±(1.49)
2.88 ±(0.61)
1.64 ±(0.76)
3.54 ±(0.10)
4.04 ±(0.85)
5.74 ±(1.37)
6.60 ±(0.26)
5.25 ±(0.79)
6.02 ±(1.17)
2.99 ±(0.78)
2.38 ±(0.16)
3.77 ±(0.26)
4.35 ± (0.25)
* The carbon content value greater than the TSP value was considered as 100% carbon.
72
-------
Table F-2. Concentrations of CO2, CO, and CH4 (ppm) in fluegas and indoor background air
(analyzed in TERI Laboratory).
Fuel/Stove
LPG
Biogas
Kerosene/wick
Kerosene/pressure
Charbriquette
Charcoal
Eucal-3 rock
Eucal-imet
Eucal-ivm
Eucal-ivc
Acacia-tm
Acacia-3 rock
Acacia-imet
Acacia-ivm
Acacia-ivc
Root-tm
Root-imet
Root-ivm
Mustard-tm
Mustard-imet
Mustard-ivm
Mustard-ivc
Rice-tm
Rice-ivm
Dung-tm
Dung-hara
Dung-ivm
Dung-ivc
C02
flue gas indoor
2249±(465) 779±(24)
1509±(636) 404±(77)
2789±(303) 665±(93)
2149±(267) 518±(114)
2622±(736) 459±(79)
1566±(241) 349±(91)
5824±(231) 419±(26)
2020±(655) 412±(100)
1722±(400) 510±(18)
2293±(913) 577±(115)
4306±(515) 557±(79)
1897±(213) 469±(10)
1330±(175) 466±(33)
1603±(125) 552±(15)
1563±(170) 525±(60)
3801±(445) 553±(30)
4289±(128) 392±(51)
1124±(83) 516±(54)
6165±(646) 418±(51)
3107±(230) 800±(794)
1860±(279) 567±(38)
2257±(699) 658±(63)
6251±(834) 583±(39)
1123±(631) 530±(50)
2048±(131) 564±(35)
3677±(389) 333±(47)
2312±(313) 372±(63)
2181±(368) 389±(53)
CO
flue gas indoor
15±(5.0) bdl
2.7±(1.9) bdl
26.1±(6.7) bdl
73±(21) bdl
375.8±(128) 31.0±(17.8)
327±(86) 17.0±(11.8)
250.1±(38) 9.0±(3.5)
111.0±(62) 3.9±(1.8)
322.4±(23) 22.8±(4.3)
134.7±(4.0) 8.0±(4.8)
182.8±(51) 7.3±(4.4)
82.2±(8.9) 13.2±(2.6)
50.7±(13.4) 1.5±(2.5)
220.1±(14) 22.4±(3.0)
45.9±(17.8) 5.3±(1.3)
148.1±(86) 7.8±(1.2)
284±(146) 6.3±(7.3)
45.5±(10.6) 9.5±(3.2)
613.4±(153) 16.4±(0.3)
238.1±(240) 1.8±(1.1)
271.6±(105) 25.3±(10)
164.8±(38) 13.2±(3.1)
542.5±(165) 17.5±(3.8)
129±(94) 17.5±(8.9)
158.8±(14) 7.5±(5.0)
456.6±(155) 5.8±(4.2)
136±(18.3) 22.4±(4.7)
132.1±(19) 22.7±(6.4)
CH4
flue gas indoor
1.6±(0.3) 1.2±(0.1)
3.3±(1.3) 3.8±(0.8)
2.3±(0.3) 1.2±(0.5)
3.7±(0.8) 1.8±(0.2)
28.0±(21.2) 3.6±(1.4)
15.7±(6.2) 3.1±(0.5)
27.4±(9.6) 2.6±(0.4)
15.0±(9.1) 1.4±(0.09)
54.8±(11.0) 4.9±(0.3)
22.2±(15.0) 2.9±(1.2)
35.0±(14.9) 3.4±(0.2)
39.4±(9.9) 4.1±(1.4)
10.0±(3.0) 2.2±(0.3)
47±(10.9) 7.3±(2.0)
9.2±(0.6) 2.9±(1.3)
77.3±(64.1) 2.5±(0.4)
25.3±(13.2) 1.9±(0.3)
6.9±(2.1) 3.0±(0.2)
105.7±(44) 3.1±(1.5)
29.7±(23.5) 1.6±(0.2)
92.3±(33.2) 5.8±(1.3)
22.1±(5.6) 3.5±(0.4)
84.3±(39.0) 3.0±(0.3)
15.5±(9.8) 8.3±(8.9)
28.5±(5.2) 3.2±(0.8)
206.1±(245) 2.4±(0.2)
24.3±(7.4) 6.9±(0.2)
23.1±(7.4) 5.2±(1.2)
73
-------
Table F-3. Concentrations of CO2, CO, CH4, TNMOC, andN2O (ppm) in fluegas samples
(analyzed by OGIST). Blanks indicate missing values.
Fuel/Stove
LPG
Biogas
Kerosene-
pressure
Charbriquette
Charcoal
Eucal-tm exl
Eucal-tm ex2
Eucal-tm ex3
Eucal-3 rock
Eucal-imet
Eucal-ivm
Eucal-ivc
Acacia-tm
Acacia-3 rock
Acacia-imet
Acacia-ivm
Acacia-ivc
Root-tm
Root-imet
Root-ivm
Mustard-tm
Mustard-imet
Mustard-ivm
Mustard-ivc
Rice-tm
Rice-ivm
Dung-tm
Dung-hara
Dung-ivm
Dung-ivc
C02 CO
874 8
1435 1
1355 47
2902 318
1576 192
3870 163
4310 149
3762 182
3300 112
2131 102
canister opened
canister opened
3314 139
1254 73
9939 47
1174 2
690 15
canister opened
canister opened
984 20
5461 340
canister opened
1583 150
1333 80
3408 329
744 43
1556 146
canister opened
6386 35
1127 85
CH4
2
3
3
26
9
2
2
27
17
16
on the way
on the way
25
13
10
22
5
on the way
on the way
5
10
on the way
35
14
36
8
23
on the way
9
18
TNMOC
11
0
6
25
6
42
45
55
24
21
27
14
8
13
58
33
N2O
0.362
0.354
0.354
0.649
0.392
0.755
0.447
0.454
0.588
0.388
0.468
74
-------
Figure F-l. Regression analysis for CC>2 (TERI vs. OGIST)
8000
7000 - -
6000 - -
5000 - -
o
o.
I 4000 - -
Q.
£
O 3000 - -
2000 - -
1000 --
Carbon Dioxide
y = 0.7093x+18.322
R2 = 0.8232
1000 2000 3000 4000 5000 6000 7000 8000
C02 in ppm (TERI)
75
-------
Figure F-2. Regression analysis for CO (TERI vs. OGIST)
Carbon Monoxide
700
100
200 300 400 500
CO in ppm (TERI)
600 700
76
-------
Figure F-3. Regression analysis for CFLi (TERI vs. OGIST)
160
Methane
140-
120 --
(0
7n 100-
I 80
Q.
o
60-
40 -
20 -
y = 0.6394x-1.117
R2 = 0.8225
20
40
60 80 100
CH4 in ppm (TERI)
120 140
160
77
-------
Table F-4. Comparison of TERI and OGIST CO2, CO, and CH4 concentrations (ppm)
Fuel/stove
Biogas
LPG
Kerosene-pressure
Charbriquette
Charcoal
Eucal-3rock
Eucal-imet
Acacia-tm
Acacia-3rock
Acacia-imet
Acacia-ivm
Acacia-ivc
Root-ivm
Mustard-tm
Mustard-ivm
Mustard-ivc
Rice-tm
Dung-tm
Dung-ivc
TERI
1960
2116
1911
3458
1304
5794
1829
4175
1785
1380
1458
1722
1220
6716
1918
2996
6772
1995
2352
C02
OGIST
1435
874
1355
2902
1576
3300
2131
3314
1254
993
1174
690
984
5461
1583
1333
3408
1556
1127
TERI
1.2
16.6
71.8
522.3
290.6
214.2
112.5
178
82.7
56.1
210
66
57.6
655
314
173
392.7
166.8
116.7
CO
OGIST
0.7
8.1
47.0
317.5
191.9
111.9
101.5
139
72.6
47.2
91.8
15.4
20.1
340
150
79.8
328.8
145.8
85.1
TERI
2.5
1.9
3.4
52.5
10.6
22.1
18.1
26.8
45.6
10.7
35.0
10.0
9.2
142
68.5
20.4
43.5
32.8
18.3
CH4
OGIST
3.1
2.1
3.2
25.6
8.5
17.2
16.0
24.5
13.3
9.7
22.2
4.9
5.4
106
35.4
13.9
35.6
22.9
18.1
78
-------
Table F-5. Corrected fluegas and indoor concentrations (ppm) and resulting net values for all
fuel/stove combinations.
fuel/stove
LPG
Biogas
Kerosene-press
Kerosene-wick
Charbriquette
Charcoal
Eucal-3R
Eucal-imet
Eucal-ivm
Eucal-ivc
Acacia-tm
Acacia-3R
Acacia-imet
Acacia-ivm
Acacia-ivc
C02(f)c
1339
1517
1979
1283
570
1407
1746
1505
1372
1936
1816
2235
1486
2470
1673
1162
941
1278
4324
3998
4129
1067
1314
1967
1324
1561
1327
1449
1114
2366
3475
2979
2762
1268
1282
1536
1062
821
995
1050
1206
1203
998
1238
1140
C02(I)c
549
575
580
249
358
298
308
470
372
455
444
563
291
402
331
202
255
330
296
309
333
389
266
267
390
364
377
393
364
518
432
453
347
341
355
348
371
343
324
407
396
418
352
435
376
CO2(N)c
790
942
1399
1034
212
1109
1438
1036
1000
1481
1372
1671
1195
2069
1343
960
686
948
4028
3689
3796
678
1048
1700
934
1197
950
1056
750
1848
3043
2526
2414
927
927
1188
691
479
671
643
809
785
646
803
764
C0(f)c
12
15
17
9
7
8
57
34
45
17
22
23
177
283
157
202
161
147
137
160
120
32
66
98
171
192
169
80
76
78
132
101
78
45
50
55
39
25
36
118
132
121
24
42
28
CO(I)c
1
1
1
1
1
1
1
1
1
1
1
1
16
34
19
13
23
12
10
11
13
8
9
9
16
19
21
13
8
12
11
13
8
12
15
14
7
7
9
20
17
18
9
10
10
CO(N)c
5
9
10
2
1
1
50
28
38
10
15
17
161
250
138
189
138
136
127
149
107
24
57
89
155
173
149
67
68
66
121
89
70
33
36
41
32
19
27
98
115
103
15
31
18
CH4(f)c
0.0
0.1
0.0
2.0
1.6
0.7
1.8
0.8
1.1
0.2
0.6
0.4
9.4
32.4
8.5
13.3
5.7
7.7
12.6
23.5
13.0
2.0
10.4
13.1
29.7
30.0
42.1
19.5
2.1
17.7
32.3
16.1
15.6
16.8
28.1
27.4
6.9
3.2
5.8
21.6
35.3
30.0
5.0
5.0
4.4
CH4(I)c
0.0
0.0
0.0
1.6
0.5
0.5
0.1
0.1
0.0
0.0
0.0
0.0
0.4
2.2
1.0
1.1
1.0
0.5
0.5
0.3
0.8
0.0
0.3
0.0
1.8
2.1
2.3
1.6
0.1
0.6
0.9
1.1
1.1
0.6
1.5
2.3
0.2
0.2
0.5
2.9
5.0
2.9
0.2
0.4
1.7
CH4(N)c
0.0
0.1
0.0
0.4
1.1
0.2
1.7
0.8
1.1
0.2
0.6
0.4
8.9
30.2
7.5
12.3
4.7
7.2
12.1
23.1
12.2
2.0
10.1
13.1
27.9
27.9
39.8
17.9
2.0
17.1
31.4
14.9
14.4
16.1
26.6
25.0
6.7
3.0
5.3
18.7
30.3
27.1
4.8
4.6
2.7
TNMOC TSP
14.7
14.7
14.7
0.4
0.4
0.4
18.0
18.0
18.0
18.0
18.0
18.0
36.0
36.0
36.0
9.0
9.0
9.0
47.0
62.0
43.0
29.8
29.8
29.8
59.0
54.0
84.0
41.0
10.0
3.0
38.9
38.9
38.9
19.4
19.4
19.4
11.7
11.7
11.7
23.3
23.3
23.3
T3 1
23.3
T3 1
23.3
T3 1
23.3
0.5
0.5
0.9
0.4
0.3
0.0
0.9
1.1
1.0
1.3
0.4
0.8
6.2
2.2
5.0
3.1
3.3
1.4
8.3
6.1
7.8
5.4
5.7
6.2
9.1
5.8
9.5
7.5
5.2
6.6
3.4
6.5
5.7
4.5
7.6
5.2
8.5
8.4
4.7
10.1
8.8
7.5
9.4
9.2
9.9
(continued)
79
-------
Table F-5 (continued)
fuel/stove
Root-tm
Root-imet
Root-ivm
Mustard-tm
Mustard-imet
Mustard-ivm
Mustard-ivc
Rice-tm
Rice-ivm
Dung-tm
Dung-hara
Dung-ivm
Dung-ivc
C02(f)c
2590
3074
2477
2081
3224
3876
779
778
881
4783
3887
4507
1363
1194
4107
1377
1120
1509
2142
1156
1556
4766
4823
3771
617
495
1325
1402
1432
1575
2824
2743
2310
1811
1757
1402
1741
1685
1263
CO2(I)c
411
385
427
332
286
262
426
357
362
341
323
271
263
253
1234
448
399
406
431
510
506
451
438
397
423
353
399
430
387
430
285
252
218
249
331
258
333
260
280
CO2(N)c
2179
2689
2050
1748
2938
3615
354
422
519
4442
3564
4236
1100
940
2872
928
721
1103
1711
645
1050
4315
4385
3373
194
143
927
972
1044
1145
2539
2491
2092
1562
1426
1145
1408
1425
984
C0(j)c
65
138
53
85
146
239
27
29
37
354
400
242
68
51
280
173
87
192
98
72
112
389
215
279
77
24
124
82
95
95
342
186
218
77
70
89
73
69
89
CO(I)c
11
10
11
13
7
7
13
10
12
15
15
7
7
7
8
26
17
17
12
15
15
15
18
14
21
11
16
13
8
11
8
7
11
20
16
20
21
15
20
CO(N)c
54
127
42
73
139
232
13
19
26
338
385
235
61
44
271
147
70
175
86
57
97
373
197
265
56
13
109
69
87
84
334
179
207
57
55
69
52
54
68
CH4(f)c
52.4
87.0
5.4
6.1
16.4
22.7
2.0
3.2
4.8
89.7
74.4
35.1
8.0
4.0
35.2
42.7
82.1
48.9
11.9
10.1
17.0
55.2
26.7
76.4
10.9
2.4
14.9
13.4
19.9
18.0
311.8
32.2
47.9
14.0
9.9
19.3
13.2
10.6
17.1
CH4(i)c
0.4
0.8
0.2
0.3
0.2
0.0
0.9
0.7
0.9
1.3
1.6
0.0
0.0
0.0
0.1
3.5
2.1
2.2
0.8
1.1
1.4
0.8
1.1
0.7
9.1
0.8
0.9
1.0
0.4
1.4
0.4
0.3
0.5
3.2
3.4
3.3
2.8
1.4
2.6
CH4(N)c
52.0
86.2
5.2
5.8
16.2
22.7
1.1
2.5
3.8
88.5
72.9
35.1
8.0
4.0
35.1
39.1
80.0
46.6
11.1
9.0
15.6
54.4
25.6
75.7
1.8
1.6
14.0
12.5
19.5
16.6
311.3
31.9
47.3
10.8
6.5
16.0
10.4
9.2
14.5
TNMOC
55.0
55.0
55.0
13.0
48.0
103.0
55.0
55.0
55.0
149.0
26.0
31.0
30.0
7.0
100.0
57.0
57.0
57.0
57.0
57.0
57.0
83.0
83.0
83.0
0.0
0.0
0.0
47.0
47.0
47.0
140.0
116.0
154.0
92.0
92.0
92.0
92.0
92.0
92.0
TSP
7.0
6.0
1.3
5.4
5.0
7.3
5.1
2.1
2.9
7.2
7.0
7.4
8.7
6.3
9.7
12.6
8.6
14.0
14.2
13.2
13.5
9.5
12.6
10.1
11.5
15.0
10.4
6.8
4.3
7.2
3.4
6.5
5.7
4.5
7.6
5.2
9.4
9.2
9.9
Note: CO2(f)c = Corrected concentration of CO2 in the flue gas
CC>2(I)c = Corrected concentration of CC>2 in the indoor
CC>2(N)c = Net concentration of CC>2 in flue gas
CO(f)c = Corrected concentration of CO in the flue gas
CO(I)c = Corrected concentration of CO in the indoor
CO(N)c = Net concentration of CO in flue gas
CH4(f)c = Corrected concentration of CKt in the flue gas
CH4(I)c = Corrected concentration of CKt in the indoor
CH4(N)c = Net concentration of CH4 in flue gas
TNMOC = Total non methane organic carbon
TSP = carbon as total suspended particles
80
-------
Table F-6. Background and concentrations of SC>2 and NOX (ppb)
Fuel/stove
LPG
Biogas
Kerosene-wick
Kerosene-
pressure
Charcoal
Charbriquette
Eucal-tm
Eucal-3 rock
Eucal-imet
Eucal-ivm
Eucal-ivc
Acacia-tm
Acacia-3 rock
Acacia-imet
Acacia-ivm
Acacia-ivc
Root-tm
Root-imet
Root-ivm
Mustard-tm
Mustard-imet
Mustard-ivm
Mustard-ivc
Rice-tm
Rice-ivm
Dung-tm
Dung-hara
Dung-ivm
Dung-ivc
Flue SO 2
5.3±(2.0)
6.7±(1.5)
6.0±(1.0)
7.0±(1.0)
4.5±(0.2)
4.0±(1.8)
6.1±(0.4)
5.3±(2.1)
8.3±(2.3)
8.5±(0.8)
1.4±(1.0)
6.3±(1.5)
5.3±(0.6)
6.7±(1.1)
13±(2.1)
6.7±(1.1)
5.3±(0.6)
5.7±(0.6)
5.0±(0.0)
4.4±(4.0)
7.0±(3.0)
6.7±(1.5)
8.3±(1.5)
4.5±(0.6)
5.7±(0.6)
3.2±(0.4)
4.0±(0.0)
8.7±(1.5)
10.3±(1.5)
Background
SO2
4.6±(2.0)
6.0±(1.0)
4.7±(1.2)
5.0±(0.0)
3.6±(0.45)
3.8±(1.7)
4.0±(0.6)
4.3±(2.1)
7.0±(1.7)
4.8±(0.3)
1.4±(1.0)
4.7±(2.1)
4.2±(0.8)
5.0±(1.0)
6.8±(2.6)
5.3±(0.2)
3.0±(1.8)
5.0±(0.9)
4.3±(0.2)
3.1±(2.8)
5.3±(2.1)
4.5±(0.7)
5.4±(2.0)
3.8±(0.6)
3.8±(0.8)
2.9±(0.3)
3.3±(0.3)
4.6±(2.1)
6.1±(0.7)
Net SO 2
0.7±(0.3)
0.7±(0.6)
1.3±(0.6)
2.0±(1.0)
0.9±(0.6)
0.2±(0.1)
2.1±(0.6)
1.1±(0.1)
1.3±(0.6)
3.7±(0.5)
1.4±(1.0)
1.7±(0.6)
1.2±(0.3)
1.7±(0.6)
6.3±(1.0)
1.4±(1.0)
2.3±(0.8)
0.7±(0.3)
0.7±(0.2)
1.3±(0.6)
1.7±(1.1)
2.2±(1.2)
2.9±(0.5)
0.7±(0.3)
1.9±(0.7)
0.3±(0.2)
0.7±(0.3)
4.1±(1.5)
4.2±(1.0)
Flue NOX
30.0±(2.7)
20.0±(2.0)
19.0±(1.7)
18.0±(8.0)
21.0±(4.0)
22.0±(2.0)
14.0±(0.5)
19.0±(2.0)
20.0±(1.5)
17.0±(3.1)
16.0±(1.0)
19.0±(1.5)
18.0±(2.0)
18.0±(2.7)
14.0±(2.8)
14.0±(0.6)
16.0±(0.6)
17.0±(1.7)
14.0±(0.6)
20.0±(2.0)
20.0±(1.6)
17.0±(4.4)
16.0±(1.7)
14.0±(2.6)
13.0±(1.5)
14.0±(1.5)
13.0±(0.6)
12.0±(1.0)
12.0±(2.0)
Background
NOX
19.0±(1.0)
18.0±(2.6)
18.0±(1.3)
16.0±(8.0)
14.0±(0.4)
12±(0.7)
12.2±(0.5)
18.0±(2.0)
17.0±(1.1)
14.1±(3.0)
12.0±(1.3)
17.0±(2.0)
14.0±(4.0)
17.0±(3.2)
10.0±(0.7)
11.0±(1.5)
14.0±(0.0)
16.0±(1.8)
13.0±(0.7)
18.0±(2.0)
16.8±(0.6)
12.0±(1.5)
11.0±(0.6)
12.0±(1.9)
11.0±(0.5)
13.0±(1.7)
12.0±(0.8)
10.0±(1.1)
10.0±(1.5)
NetNOx
11.0±(3.5)
2.0±(0.6)
1.0±(0.8)
2.0±(0.2)
7.0±(3.6)
10±(2.5)
1.8±(0.5)
1.0±(0.6)
3.0±(2.0)
3.0±(1.5)
4.0±(1.9)
2.0±(1.0)
4.0±(2.5)
1.0±(0.6)
4.0±(2.2)
4.0±(1.9)
2.0±(0.6)
1.0±(0.5)
3.0±(1.3)
2.0±(2.0)
3.0±(1.1)
5.0±(2.9)
5.0±(1.1)
2.0±(0.8)
2.0±(1.0)
1.0±(0.5)
1.0±(0.8)
2.0±(0.5)
2.0±(0.7)
Standard deviations are given in parentheses.
81
-------
Appendix G: Error Analysis
Since the carbon balance method relies on ratios to CO2, the sensitivity of the calculated
emission factors to potential errors in measured fluegas concentrations is not directly obvious.
The error analysis in Table G-l shows typical percentage changes in calculated emission factors
(including K and NCE) as a function of hypothetical 10% errors in the measured fluegas
concentrations for each of the major airborne species. Note that the emission factors for any one
species are quite insensitive to errors in any of the other gases except CC>2. Because the
calculation depends on ratios, of course, there is also little sensitivity to problems that affect
entire samples, such as leakage of ambient air into the sample container during sampling, storage,
or GC injection.
Table G-l. Error Analysis
A 10% change in:
C02
CO
CH4
TNMOC
TSP
Gives
K
12
6
1
O
0.6
this % change
NCE CO2
1
0.7
0.2
0.4
0.1
1
0.7
0.2
0.4
0.1
in final
CO
11
11
0.2
0.4
0.1
emission estimates:
CH4 TNMOC
11
0.7
11
0.4
0.1
11
0.7
0.2
9
0.1
TSP
11
0.7
0.2
0.4
10
82
-------
Appendix H: Estimation of Indian Household Fuel Consumption
The limitations in available estimates are listed below.
In India a wide variety of fuels such as liquid petroleum gas (LPG), kerosene, biogas, coal,
coke, charcoal, fuelwood, dungcakes, rootfuel and crop residues (mustard stalks, jute stalks,
cotton stalks, rice straw, etc.) are used for cooking purposes.
A variety of improved stoves such as metal stove, mud stove (with single pot, two pots) and
ceramic stoves are now in use through efforts of the Ministry of Non-conventional Energy
Sources (MNES).
The life of the improved stoves is limited (not more than 2 years). So the number of
improved stoves in working condition is far less, but by an uncertain number, than the total
disseminated to date.
In India due to the large variation in the agricultural climatic conditions and life style, the
types of crop produced also vary from region to region. Depending on the type of crops
produced, the crop residues used as fuel also vary.
There is a considerable variation in the types of food cooked, cooking practices etc. For
example the stove known as Kara, employed for simmering milk and fodder preparation,
consumes large quantities of dungcake as a fuel and is common in northern states of India.
This stove is not in use in southern region.
Energy consumption levels also vary for different agricultural climatic regions (397-1393
useful kcal/person-day). The biofuel consumption database for India which was made
based on the rural energy surveys is found to be quite inadequate. There is a wide variation
in the existing rural energy database of India (Joshi and Sinha 1993).
Keeping in mind these limitations, we attempted to estimate the amount of fuel used in India for
the year 1990/91.
Biofuel estimation. Large amounts of biofuels are used in rural areas. Three different sources of
biofuel consumption estimates for rural India are available. They are:
Rural Energy Database (REDE). REDB is based on the analysis of data compiled for 638
villages in 17 states spread over 14 agricultural climatic regions and covering 39000 households.
Integrated Rural Energy Planning Programme (IREP). IREP database, compiled by the
Planning Commission, Government of India, is based on block level surveys covering nearly 250
blocks. (Blocks are the local administrative subdivision under the district. Each block consists of
groups of villages.)
National Council for Applied Economic Research (NCAER) database. The NCAER data are
based on surveys conducted in 7500 households (in rural areas) selected from 600 villages in 300
districts.
Among these three estimates, the REDB estimates are on the higher side and
NCAER estimates are on the lower side. So we have used IREP estimates in our fuel use
estimation of rural India even though IREP database has the following uncertainties.
83
-------
1. The IREP estimate of crop residues for West Bengal is zero, whereas it is known that crop
residues are used extensively in the state.
2. There are no estimates for Goa.
3. There are no data for dungcake and crop residues for the northeastern states.
Steps involved in biofuel consumption estimation by stove type for rural India (see Figure H-l).
The state biofuel figures given by IREP estimates are divided by the total number of
households in different states of rural India to get the per household consumption.
The state data distribution of improved stoves till March 1991 was collected from MNES.
From the total number of improved cookstoves installed, the number of improved
cookstoves in working condition was calculated based on the assumption that only 60%
are functional.
In the improved stoves, 10% of the improved cookstoves are assumed to be improved
metal and 90% of the improved cookstoves are mud stoves.
The remaining households are assumed to be using traditional stoves.
It is estimated that there is only one stove in use in each household.
It is assumed that each stove consumes the three biofuels in the same proportion given by
IREP.
For biofuel consumption in traditional stoves the number of stoves are multiplied by the
household consumption of biofuels.
The total biofuel in improved cookstoves biofuel consumption is estimated by multiplying
the household consumption by 0.80 assuming that the improved cookstoves save 20% fuel
consumption and further multiplied by the total number of improved stoves working.
The number of rural households, improved stoves and biofuel consumption in each stove in rural
India for the year 90/91 are given in Table H-l.
84
-------
Table H-l. State list of rural households, penetration of improved stoves, and biomass fuel consumption
State/Union
territories
Andhra
Pradesh
Arunachal
Pradesh
Assam
Bihar
Goa
Gujarat
Haryana
Himalchal
Pradesh
Karnataka
Kerala
Madhya
Pradesh
Maharashtra
Manipur
Meghalaya
Mizoram
Nagaland
Orissa
No. of rural
households
9579605
129956
3265110
10682935
120758
4289530
1825870
830856
4920170
3908425
8243710
8410655
203193
256914
60348
168918
4773275
No. of
improved
stoves installed
until 31.Mar.91
762598
6042
101357
509946
48429
534200
569136
360886
470886
220333
845023
662639
21576
10200
7694
7000
339528
No. of improved stoves
working
Metal Mud Total
stoves stoves
45756 411803 457559
363 3263 3626
6081 54733 60814
30597 275371 305968
2906 26152 29058
32052 288468 320520
34148 307333 341481
21653 194878 216531
28313 254818 283132
13220 118908 132200
50701 456312 507014
39758 357825 397583
1295 11651 12946
612 5508 6120
462 4155 4616
420 3780 4200
20372 183345 203717
No. of
traditional
stoves
8817007
123914
3163753
10172989
72329
3755330
1256735
469971
4448285
3688092
7398687
7748016
181617
246714
52654
161918
4133747
Total consumption of
biofuels (million tons/year)
Fuel- Dung- Crop
wood cake residues
10.8 2.9 3.6
0.5 0.0 0.0
12.3 0.0 0.0
26.9 9.9 13.0
0.0 0.0 0.0
9.1 2.2 3.0
1.7 2.9 4.3
3.3 0.4 0.2
8.3 1.8 3.2
10.0 0.0 1.6
13.1 1.8 1.5
16.0 6.7 5.8
0.8 0.0 0.0
0.9 0.0 0.0
0.2 0.0 0.0
0.6 0.0 0.0
11.2 0.6 0.4
Per household consumption
of biofuels (tons/year)
Fuel- Dung- Crop
wood cake residues
1.13 0.30 0.38
3.85 0.00 0.00
3.77 0.00 0.00
2.52 0.93 1.22
0.00 0.00 0.00
2.12 0.51 1.70
0.93 1.59 2.36
3.97 0.48 0.24
1.69 0.37 0.65
2.56 0.00 0.41
1.59 0.22 0.18
1.90 0.80 0.69
3.94 0.00 0.00
3.50 0.00 0.00
3.31 0.00 0.00
355 0.00 0.00
2.50 0.13 0.09
oo
(Jl
(Continued)
-------
Table H-l (continued)
State/Union
territories
Punjab
Rajasthan
Sikkim
Tamil Nadu
Tripura
Uttar Pradesh
West Bengal
Union
Territories
TOTAL
No. of rural
households
2257090
5441095
67318
8027750
430649
16784590
8384490
285878
103049088
No. of
improved
stoves installed
until 31.Mar.91
515796
1080764
18597
742420
5971
1209179
317179
151096
9519475
No. of improved stoves
working
Metal Mud Total
stoves stoves
30948 278530 309478
64846 583613 648458
1116 10042 11158
44545 400907 445452
358 3224 3583
72551 652957 725507
19031 171277 190307
9066 81592 90658
571169 5140517 5711685
No. of
traditional
stoves
1741295
4360331
48721
7285330
424678
15575411
8067311
134782
93529617
Total consumption of
biofuels (million tons/year)
Fuel- Dung- Crop
wood cake residues
1.9 3.4 5.0
4.3 2.1 0.8
0.2 0.0 0.0
8.5 2.0 2.5
1.4 0.0 0.0
21.9 17.2 17.3
4.4 0.0 0.0
0.4 0.3 0.4
168.7 54.2 62.6
Per household consumption
of biofuels (tons/year)
Fuel- Dung- Crop
wood cake residues
0.84 1.51 2.22
0.79 0.39 0.15
2.97 0.00 0.00
1.06 0.25 0.31
3.26 10.00 0.00
1.30 1.02 1.03
0.52 0.00 0.00
1.40 1.05 1.40
54.96 9.55 12.03
oo
ON
-------
The 1991 census found that only 30% of the urban population use biofuels, which we assume to
be nearly all fuelwood in traditional stoves with a consumption norm of 1 kg/person-day. The
total urban consumption of fuelwood is thus calculated to be 23.8 million ton/year.
Charcoal consumption in cookstoves for the year 1990/91 is calculated from the charcoal
production data. FAO (1994) reported that in India about 2 million ton of charcoal was produced
in the year 1991. Most of the charcoal was used for small-scale industries such as bakeries,
laundries, silk re-reeling, jewelry making, etc. so it is assumed only 25% was used in cookstoves.
Biogas consumption is estimated from the number of family biogas plant installed. Up to
1990/91, 1.4 million family type biogas plants were installed (TERI1997). The plant capacity is
2m3/day. The NCAER survey indicates that only 66% of the biogas plants installed are in
working condition (NCAER 1992). Based on the assumption that 66% of the installed biogas
plants produce biogas with a 70% of the plant capacity, 666 million m3 of biogas was consumed
in India.
Commercial Fuels (LPG, kerosene). Commercial fuels such as LPG and kerosene are used by
30% of the population, mainly in urban areas. For the year 1990-91 total LPG consumption was
2.4/5 million ton. Out of which, 78.4% (1.894 million ton) was used for domestic purpose
(MoPNG, 1993).
In 1990/91, kerosene consumption was 8.4 million ton/year, but it is unclear what fraction was
used for cooking. In 1991, 60% of the kerosene was used in rural sector (MoF 1992), where
most is used for lighting. NCAER (1985) indicated a cooking: lighting ratio of 0.186:1 in rural
areas and 3.46:1 for urban areas. Kishore and Joshi (1995) reported that the predominant use of
kerosene for lighting in rural areas and for cooking in urban areas continues. It is thus estimated
that 3.98 million ton of kerosene was used for cooking during 1991, of which 29% is used in
rural areas. In the absence of data on how much is used in each kind of stove, it is assumed that
in urban area 60% of the kerosene is used in wick stoves and 40% in pressure stoves. The
reverse percentages are assumed for rural areas.
The estimated fuel consumption by stove in India for 1990/91 is in Table H-2.
87
-------
Table H-2. Fuel consumption by stove type in India (million tons/year)
Stove
Traditional mud
(tm)
Improved
metal (imet)
Improved
mud (ivm)
Kara
Angethi
Kerosene-
pressure
Kerosene-
wick
LPG
Biogas
Total
Fuel- Dung Crop Charcoal Kerosene LPG Biogas
wood cake residues (million m )
193.4 31.6 58.6
0.71 0.3
6.36 2.77 2.9
19.1
0.5
1.82
2.16
2.1
666
200.5 53.5 61.8 0.5 3.98 2.1 666
------- |