Subscriber access provided by NEW YORK UNIV
Article
Estimation of organic and elemental carbon emitted from wood burning in traditional and improved cookstoves using controlled cooking test Pooja Suresh Arora, and Suresh Jain Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/es504012v • Publication Date (Web): 17 Feb 2015 Downloaded from http://pubs.acs.org on February 18, 2015
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 26
Environmental Science & Technology
Graphical abstract
ACS Paragon Plus Environment
Environmental Science & Technology
1 2
Estimation of organic and elemental carbon emitted from wood burning in traditional and improved cookstoves using controlled cooking test
3
Pooja Aroraa, Suresh Jaina, b1 a
4 5 6 7 8
Department of Natural Resources, TERI University, Delhi, 10, Institutional Area, Vasant Kunj, New Delhi 110070, India b Department of Energy and Environment, TERI University, Delhi, 10, Institutional Area, Vasant Kunj, New Delhi 110070, India
9
Abstract
10
Emission of various climate and health related pollutant species from solid biomass
11
burning in traditional cookstoves is a global concern. Improved cookstoves serve as a possible
12
solution to mitigate the associated impacts. However, there is a need to intensify the efforts in
13
order to increase the data availability and promote revision of existing metrics of cookstove
14
testing. In this study, the effect of different phases of a cooking cycle of Northern India on
15
emission factors of OC and EC (char and soot) was assessed for four cookstoves (advanced,
16
improved and traditional) using Acacia nilotica. Lowest EFs for OC (0.04 g/MJ) and EC (0.02
17
g/MJ) were observed in case of the forced draft cookstove; while the traditional and natural
18
draft top feed cookstove emitted the highest OC (0.07 g/MJ) and EC (0.09 g/MJ),
19
respectively. Variation in terms of EFs for OC and EC (char & soot) within the cooking cycle
20
was also found to be significant.
21
Key words: Biomass energy; Controlled cooking test; Black carbon; Regional cooking cycle;
22
Regression analysis
23 24
Introduction
25
In India, ~66% percent of the households depend on solid biomass fuels such as wood,
26
cow dung and crop residues in order to fulfil their cooking energy needs.1 Combustion of these
27
solid biomass fuels in inefficient traditional cookstoves leads to deterioration of indoor air
28
quality due to emission of both particulate matter (PM) and gaseous pollutants. A characteristic 1
Corresponding author, Department of Natural Resources, TERI University, 10, Institutional Area, Vasant Kunj, New Delhi-110070, India; Tel: +91-11-2612 2222; Fax: +91-11-2612 2874; Email:
[email protected];
[email protected] 1 ACS Paragon Plus Environment
Page 2 of 26
Page 3 of 26
Environmental Science & Technology
29
of carbonaceous part (organic and black carbon) of PM has a strong bearing on the atmospheric
30
radiative balance. Black carbon (BC) or elemental carbon (EC) content of PM has been
31
identified as the major precursor of climate change due to its high intensity radiative forcing
32
despite its short residence time in the atmosphere.2 Organic carbon (OC), on the other hand, has
33
been linked to light scattering properties3 leading to cooling effect in earth’s atmosphere. OC
34
(also known as ‘brown carbon’) has recently been found to be a significant source of light
35
absorption activity in the atmosphere.4,5 Apart from impact at the global level, chemical
36
composition of PM is also responsible for altering the environmental conditions at regional and
37
local levels, which primarily include visibility reduction and changed rainfall patterns in areas
38
where biomass burning is a common practice.6,7 Chemical characteristics of PM emitted from
39
inefficient cookstoves can also be linked to adverse effects on human health because of their
40
toxicity.8-10
41
Chemical composition of PM is highly dependent on the type of combustion device, fuel
42
properties and type of activity leading to variation in emissions.11 Field studies have a definite
43
advantage over laboratory studies when the primary aim of the study is to build emission
44
inventories and assess the impacts on human health. There have been a few field studies, which
45
focus on quantification of emissions of carbonaceous aerosols from cookstoves under actual
46
operating conditions.12,13 However, due to high variation within the data, it is difficult to
47
ascertain the variables responsible for a typical performance of a cookstove.14 One such variable
48
is the “burn cycle” or ‘cooking cycle”, which has been found to influence cookstove
49
performance.15-17 The effect of different cooking cycles can be captured using a laboratory based
50
cooking test by controlling the fuel characteristics such as fuel size and moisture, which has also
51
been found to affect the cookstove performance in a few studies.18-21 The laboratory based
52
cooking tests, such as the controlled cooking test (CCT), can serve as a standard protocol to
53
assess the performance of cookstoves under the influence of different cooking cycles. To the best
54
of the author’s knowledge, no attempt has been made to use CCT as a test in order to identify
55
and assess variation in chemical composition of PM with different cooking cycles associated
56
with diverse food items consumed and the sequence in which they are cooked in different
57
regions. Additionally, the present study has also quantified emission of char and soot-EC from
58
biomass burning in cookstoves. Quantification of these light absorbing chemical species is
59
important since both char and soot-EC act as light absorbing fraction of PM; however, soot-EC 2 ACS Paragon Plus Environment
Environmental Science & Technology
60
has a high light absorption potential compared to char-EC. Char-EC is generated during low
61
temperature combustion while formation of soot-EC takes place under high temperature
62
combustion conditions. Therefore, their quantification can also be used to identify dominant
63
combustion conditions in a particular cooking activity in context of cookstoves.22 The studies
64
conducted on carbonaceous aerosols emitted from cookstoves so far, have not considered char
65
and soot-EC as a measure to predict the dominant combustion conditions in a particular cooking
66
activity. Therefore, the aim of the present study is to quantify the OC and EC content of the PM
67
and further analyse the char and soot EC fractions and their relationship with different
68
combustion conditions in different cookstoves by using the cooking cycle of Northern India.
69
However, the scope of the study is limited to three cookstoves selected under the improved and
70
advanced category, which were readily available in India. Nevertheless, the distinctiveness in
71
designs adopted in the selected cookstoves provides, to some extent, useful set of data on effect
72
of some specific variables on the cookstove performance.
73
Materials and methods
74
Survey
75
A survey was conducted in two villages, Thalan village in district Uttarkashi and Aam
76
Kheri village in Roorkee lying in the plains of Uttarakhand. The households were selected using
77
stratified random sampling with 50 households in each village. The survey was conducted using
78
a semi-structured questionnaire, which comprised of questions related to primary food type, its
79
quantity, frequency and sequence of cooking. However, tasks other than cooking food were not
80
considered during the study. The data collected from the survey was used to standardize the
81
amount and type of each food item cooked in a single meal.
82
Traditional cookstove
83
Traditional (TR) cookstove used in the study was constructed in the laboratory with
84
dimensions and material same as those observed on the field. The cookstove was made up of six
85
bricks covered with mixture of clay, hay and cow dung. The cookstove was constructed on a
86
thick metal sheet to allow ease of movement under the hood during testing. The walls of the
87
cookstove were ~6 cm thick and the height, width and length of the combustion chamber was
88
~22, ~14 and 18 cm, respectively, giving it a U-shaped top view. The cookstove design was
89
similar to the traditional cookstove used by Kar et al.23 in field based studies conducted in
90
Northern India. The cookstove was recoated with clay slurry at the completion of each meal 3 ACS Paragon Plus Environment
Page 4 of 26
Page 5 of 26
Environmental Science & Technology
91
preparation. The traditional cookstove design selected for baseline study holds a great
92
importance in order to assess the actual improvement in a particular study area. For example, the
93
single pot traditional cookstove design in Maharashtra (Western region of India) as reported by
94
Dutta et al. 24 has a V-shaped combustion chamber, which might perform differently from the
95
traditional cookstoves used in Northern India.
96
Improved cookstoves
97
The improved cookstoves (ICSs) were selected for the study among some of the popular
98
designs available in the Indian markets. ICSs can be placed in three common categories, i.e.,
99
basic, intermediate and advanced depending on the technology used to upgrade its performance
100
as described in a report by Dalberg Global Development Advisors.25 The forced draft cookstove
101
(PF) used in this study is manufactured by Philips and can be placed under advanced cookstove
102
category. The PF (Model-4012) cookstove works as a “quasi-gasifier” cookstove. The forced air
103
supply in this cookstove (both primary and secondary) allows adequate mixing of air and
104
combustible gases leading to efficient combustion of fuel.26,27 The other ICSs were; a top feed
105
natural draft cookstove (PN) manufactured by Philips and a front feed natural draft (EN)
106
cookstove manufactured by Envirofit International Ltd, Colorado, USA. The PN (Model-
107
HD4008) and EN (Model-G3300) cookstoves can be considered as an intermediate technology
108
where the former is also a “quasi-gasifier” but with natural air convection and latter is a rocket
109
cookstove. Rocket cookstoves utilize “L” shaped structure in order to enhance the combustion
110
process. The photographs of the cookstoves can be accessed in the research article by Arora et
111
al.16
112
Fuel used
113
Acacia nilotica (Keeker) wood was used in the present study owing to its availability at
114
the test location. Keeker was also found to be one of the commonly used wood species for
115
cooking purposes in the surveyed areas. Since the aim of the study was to assess the relative
116
performance of different cookstoves with a common cooking cycle of the Uttarakhand region,
117
the use of different wood types could be overlooked. Still, the emission factors (EFs) generated
118
in the present study will not be specific to Uttarakhand region in terms of fuel practice and may
119
vary with other wood species used in the study area. In the study region, wood was collected
120
from the closest forest area and gathered wood consisted of a combination of thin and thick
121
stems, which were used for ignition and high power cooking, respectively. A similar pattern was 4 ACS Paragon Plus Environment
Environmental Science & Technology
122
followed in the laboratory experiments where wood was procured from a common source of fuel
123
and only thick branches were used for each cooking task. However, the wood logs were cut into
124
pieces according to the recommended size for the two top feed cookstoves. In case of PF and PN
125
cookstoves, wood size was kept 2×3×10 cm while it was 1.5×1.5×25 cm for EN cookstove. In
126
case of TR cookstove, the wood pieces were also cut into a uniform size of 3×3×25 cm (without
127
bark). The wood moisture during the testing period was 10-18%, which is also the case in field
128
during actual cooking. The calorific values for wood and charcoal were calculated using bomb
129
calorimeter and were found to be ~21 MJ/kg and ~32 MJ/kg respectively.
130
Test protocols
131
The CCT Version 2.0 developed under Household Energy and Health Programme, Shell
132
Foundation (28) was used as a test method to assess the cookstove performance in terms of mass
133
emission factors (EFs) of OC, EC and PM. The EFs were calculated using the total fuel energy
134
consumed for specified cooking task. CCT is a laboratory based international protocol, which is
135
used to compare cookstoves in terms of fuel savings, and involves actual cooking of a meal
136
dominant in the study area. The meal was prepared by a female cook who was a resident of the
137
same region. The experiments were started after adequate training of the female cook with the
138
new ICSs. All relevant parameters were measured during the test in order to calculate OC/EC
139
EFs. During each test, sampling was started as soon as the vessel was put over the cookstove and
140
start time was recorded, which also included the ignition phase emissions. The wood chips were
141
prepared from the same lot of wood used as kindling fuel, in order to maintain consistency. At
142
the end of each cooking phases, sampling was terminated followed by weighing of cooked food
143
and wood left, and the end time was recorded. The charcoal formed was also weighed along with
144
the cookstove and was used for next cooking phase. The graphic representation of steps carried
145
out in CCT is presented in Figure S1 of SI.
146
The Water Boiling Test (WBT) Version 4.1.229 was also used to compare performance of
147
cookstoves with the results obtained from CCT. WBT is a simulation of actual cooking cycles
148
designed principally to study the effect of design on the performance of cookstoves (for more
149
details related to the protocols, please refer to research article by Arora et al.21
150
Sampling set-up
151
The experiments were performed at TERI University’s testing facility located in New
152
Delhi. The testing facility consists of an extraction hood and a dilution tunnel with location 5 ACS Paragon Plus Environment
Page 6 of 26
Page 7 of 26
Environmental Science & Technology
153
provided for PM sampling. Size of the hood was large enough to accommodate the cookstoves
154
manufactured for domestic cooking purposes. The PM sampling point was kept ~9 times the duct
155
diameters from the closest point of disturbance and the PM deposited on the filter (Whatman, 47
156
mm quartz Q-MA, with pore size 0.2 µm) consisted of total particulate matter (TPM). The
157
details of the experimental setup and sampling procedure have been mentioned in detail by Arora
158
et al.16, 17 The schematic diagram of the experimental set-up can be referred in the research article
159
by Arora et al.16
160
OC/EC analysis
161
PM samples collected on quartz microfiber filters (pore size, 0.2 µm) were preserved at a
162
temperature of around -20ºC until the OC/EC analysis was carried out. The filter samples for
163
correction of OC artifacts were collected for few experiments; however, the values were not
164
reduced from the actual concentrations. This was because of the variation of data observed in gas
165
phase OC artifacts, which might result in over or underestimation of the actual OC concentration
166
in the rest of the experiments for which the artifact samples were not collected.
167
The analysis was conducted with the help of thermal-optical method using EC/OC
168
analyser (DRI, Model 2001) using the IMPROVE-TOR carbon analysis.30 The PM mass
169
deposited on the quartz filter is analysed using a 0.5 cm2 section of the filter. In the thermal-
170
optical analyser, the filter punch is first heated in an inert atmosphere (100% Helium, He) which
171
converts the OC fraction into carbon dioxide (CO2). The EC fraction is converted to CO2 in an
172
oxidizing atmosphere (98% He/ 2% Oxygen, O2). The concentration of CO2 produced during
173
both the conditions is analysed by reducing it to methane, which is eventually measured on a
174
flame ionization detector (FID). Transmittance of laser light (633 nm) through the filter is
175
measured to set the split between OC and EC. Further, the actual EC is calculated by correcting
176
for the pyrolysed carbon formed during charring of OC in the first phase. EC data was also used
177
to calculate the char and soot-EC concentrations using EC1, EC2 and EC3, which are produced
178
at a temperature of 550, 700 and 800˚C respectively in 2% O2 and 98% He atmosphere during
179
analysis. Char-EC was calculated by using EC1 minus pyrolysed carbon (PC) and soot-EC using
180
sum of EC2 and EC3. However, limited studies are available in the literature where TOR method
181
has been employed for estimation of char and soot-EC, and the method is still under scientific
182
scrutiny.31
183
Statistical analysis 6 ACS Paragon Plus Environment
Environmental Science & Technology
184
The experimental design for the study consisted of two factors i.e. cookstove types (four
185
levels - TR, PF, PN and EN) and cooking cycle (three levels – pulse (legumes), rice and roti
186
making). Both single and two-way analysis of variance (ANOVA) with replicates were used to
187
confirm significant (α = 0.05) difference in average values of OC and EC EFs and interaction
188
between cookstoves and burn cycle. Further Student’s-test was conducted in order to identify
189
pairs with statistically significant difference in average OC and EC EFs with the help of
190
Bonferroni correction method, where α=α/c (c is the number of paired means compared).32
191
The relationship was established using the coefficient of determination (R2) and
192
Pearson’s correlation coefficients, which confirm that variation in PM EFs was associated to the
193
change in OC and EC EFs.
194
Results and discussion
195
Cooking cycles – Survey results
196
The data collected through the survey conducted in the villages of Uttarakhand, showed
197
that ~64% of the households show two prominent sequences in a single meal i.e. pulse-rice-roti
198
(Indian wheat bread) or pulse-vegetable-roti. Rice and vegetable were found to be cooked
199
alternatively in a single meal throughout the week. Roti's were cooked mostly at the end because
200
of its tendency to loose moisture. Therefore, in the present study the cooking sequence followed
201
an order in which pulse was cooked in the first place, followed by rice and roti, for all the
202
experiments. The vegetable preparation was omitted from the cooking cycle because of the
203
possible interference in the OC quantification in PM due to use of cooking oil. The survey results
204
also showed that the average quantity of pulse, rice and wheat flour consumed in a single meal
205
was around 250 ± 20 g, 300 ± 34 g and 500 ± 23 g, respectively for an average household size of
206
five members. Detailed survey results are available in the study conducted by Arora et al.16
207
Variation in emission factors for OC and EC
208
The variation in EFs for OC and EC during different phases of cooking cycle and their
209
average in the four cookstoves during CCT are presented in Figure 1 and Table 1, respectively.
210
The average EF for OC was found to be the highest in TR cookstove followed by EN, PN and PF
211
cookstoves and difference in average OC EF between TR, and three cookstoves was found to be
212
statistically significant (p