Subscriber access provided by UNIV OF LOUISIANA
Energy and the Environment
Aerosol Optical Properties and Climate Implications of Emissions from Traditional and Improved Cookstoves Georges Saliba, R Subramanian, Kelsey Bilsback, Christian L'Orange, John Volckens, Michael Johnson, and Allen L. Robinson Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b05434 • Publication Date (Web): 30 Oct 2018 Downloaded from http://pubs.acs.org on November 1, 2018
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 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 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.
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 33
Environmental Science & Technology
1
Aerosol Optical Properties and Climate Implications of Emissions from Traditional and
2
Improved Cookstoves
3
Georges Saliba1‡, R. Subramanian1, Kelsey Bilsback2, Christian L’Orange2, John Volckens2,
4
Michael Johnson3, Allen L. Robinson1*
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
1Carnegie
Mellon University, 5000 Forbes Avenue, Pittsburgh, PA, 15213 State University, 430 North College Avenue, Fort Collins, CO, 80524 3Berkeley Air Monitoring Group, 1900 Addison Street, Berkeley, CA, 94704 2Colorado
‡Now
at Scripps Institution of Oceanography, University of California San Diego, 9500 Gilman Dr, La Jolla, CA, 92093 *Corresponding author: Prof. Allen L. Robinson Dept. of Mechanical Engineering Carnegie Mellon University 5000 Forbes Avenue Scaife Hall 401 Pittsburgh, PA 15213 Tel. 412-268-3657 Fax. 412-268-3348 Email:
[email protected] 22 23 24 25 26 27 28 29
ACS Paragon Plus Environment
Environmental Science & Technology
30
TOC/ABSTRACT ART
31 32
Abstract
33
Cookstove emissions are a major global source of black carbon but their impact on climate is
34
uncertain because of limited understanding of their optical properties. We measured optical
35
properties of fresh aerosol emissions from 32 different stove/fuel combinations, ranging from
36
simple open fires to high-performing forced-draft stoves. Stoves were tested in the laboratory using
37
the firepower sweep protocol, which measures emissions across the entire range of functional
38
firepower. There is large variability in measured optical properties across the entire range of
39
firepower. This variability is strongly correlated with black carbon-to-particulate matter mass ratio
40
(BC/PM). In comparison, stove type, fuel, and operational metrics were poor predictors of optical
41
properties. We developed parametrizations of the mass absorption cross-section, the absorption
42
angstrom exponent, and the single scattering albedo of fresh emissions as a function of BC/PM.
43
These parametrizations, derived from laboratory data, also reproduce previously reported field
44
measurements of optical properties of real-world cooking emissions. We combined our new
45
parametrizations of intensive optical properties with published emissions data to estimate the direct
46
radiative effect of emissions for different stove technologies. Our data suggest that so-called
ACS Paragon Plus Environment
Page 2 of 33
Page 3 of 33
Environmental Science & Technology
47
“improved” stove reduce CO2 equivalent emission (i.e., climate benefits) by 20% – 30% compared
48
to traditional stoves.
49
Introduction
50
Aerosol emissions contribute to climate change by absorbing and scattering outgoing and
51
incoming radiation.1 Cookstoves are important global sources of light-absorbing aerosol such as
52
black carbon (BC)2 and brown carbon (BrC).3–6 Indoor pollution from cookstove emissions
53
contribute to more than four million annual deaths worldwide.7 To reduce climate, health, and
54
environmental impacts, so-called “improved” stoves are being promoted as cleaner alternatives to
55
traditional technologies. Improved stove designs range from adding a combustion chamber (e.g.
56
natural-draft rocket-style stove) to including an electric fan (e.g. forced-draft stove). These changes
57
can improve combustion efficiency and heat transfer to the cook-piece and therefore reduce
58
emissions. Field studies indicate that improved stoves have modestly lower particulate matter (PM)
59
mass emissions compared to traditional stoves;8–11 however, improved stoves can also emit higher
60
BC fractions due to their altered combustion conditions.9,12–14 Therefore, the large-scale
61
deployment of improved stoves could have important climate implications.
62
Predicting the future climate impacts of newer stove technologies requires accurate emission
63
inventories and correct representation of the optical properties of their emissions.2 The limited
64
published data reveal significant variability in the optical properties of stove emissions.9,15,16 For
65
example, emissions from traditional stoves tested in Honduras had a factor of two greater
66
absorption (per BC mass) compared to stoves tested in Malawi.9,15 The causes of these differences
67
are not well understood; they are likely due to the complex interplay between optical properties
68
and stove type/fuel/fire conditions. Misrepresenting optical properties of stoves emissions can lead
69
to a dramatically different predictions of climate impacts. For example, assuming a complete
ACS Paragon Plus Environment
Environmental Science & Technology
70
internal mixture between BC and non-BC material and treating the organics as “brown” can change
71
the direct radiative forcing of biomass burning and stove emissions from negative (net cooling) to
72
positive (net heating) compared to a baseline scenario of a complete external mixture with non-
73
absorbing organic material.17,18
74
Previous studies have tried to develop parametrizations to explain the variability in optical
75
properties of cookstove emissions. One approach is to use the modified combustion efficiency
76
(MCE, ratio of background corrected CO2 emissions to background corrected CO+CO2
77
emissions), which is a measure the completeness of combustion for converting carbonaceous fuel
78
into CO2. This approach has been used to parametrize aerosol absorption and scattering from
79
cookstove9 and biomass burning emissions19–21 with moderate to poor correlations. Another
80
parameter of potential importance is the ratio of black carbon mass-to-total particulate matter mass
81
(BC/PM). BC/PM can indicate coating thickness around the BC particles, which can directly affect
82
their radiative properties.22 Several studies report good correlation between the BC/PM and optical
83
properties such as the mass absorption cross section (MACPM, absorption per unit PM mass), the
84
absorption angstrom exponent (AAE) and the single scattering albedo (SSA, ratio of scattering to
85
extinction) for cookstove9,15 and biomass/biofuel burning emissions.6,19,20,23,24 However, these
86
studies only considered a limited number of stove technologies from specific field locations.
87
Although comprehensive parametrizations of aerosol optical properties exist for biomass burning
88
emissions,19 we are not aware of similar parametrizations for cookstove emissions.
89
In this study, we report measurements of aerosol intensive optical properties (MACBC, SSA,
90
and AAE) of emissions from 18 different cookstoves operating on two different fuels (wood and
91
charcoal). All the stoves were tested in the laboratory using the firepower sweep protocol, which
92
systematically spans the range of functional stove firepowers.25 Using our dataset of 32 stove/fuel
ACS Paragon Plus Environment
Page 4 of 33
Page 5 of 33
Environmental Science & Technology
93
combinations, we investigate the role of stove technology, fuel, and operational metrics (firepower,
94
MCE, and BC/PM) on measured optical properties. We develop parametrizations for these optical
95
properties as a function of BC/PM. We then combine our new parametrizations with field emission
96
data to estimate the direct radiative forcing from emissions of different stove technologies.
97
Methods
98
Stoves, fuels, and the firepower sweep protocol
99
We measured emissions from 18 different stoves operated in the laboratory using the firepower
100
sweep protocol including: three-stone fire, Justa, sunken pot, clay/metal traditional stoves,
101
metal/ceramic charcoal stoves, rocket-style stoves, natural- and forced-draft gasifiers. Compared
102
to previous studies9,10,15,16 reporting optical properties from stove emissions, we tested a much
103
wider range of stove technologies. The stoves tested span the ISO total emission’s tiers of
104
performance,26 from the highest emitting Tier 0 stoves to cleaner Tier 3 stoves. We did not test
105
any Tier 4 stoves, the lowest emitting category. For discussion, the stoves are separated into two
106
major categories: wood and charcoal stoves. These two categories contained multiple sub-
107
categories based on emission certifications. Wood stoves were separated into three sub-categories:
108
traditional (Tier 0), rocket-style (Tier 2), and gasifier (Tier 3) stoves. Similarly, charcoal stoves
109
were segregated into three sub-categories: metal (Tier 0), ceramic (Tier 0), and improved (Tier 2)
110
stoves.
111
The stoves were tested using different fuels to investigate the effects of fuel type on optical
112
properties. We burnt two different charcoal fuels: coconut and lump-hardwood, and four different
113
wood fuels: red oak, Douglas fir, Eucalyptus, and wood pellets. In total, 32 individual tests were
114
performed; they are detailed in Table S1 in the supporting information (SI).
ACS Paragon Plus Environment
Environmental Science & Technology
115
To investigate the role of operating conditions, each stove was tested using the firepower sweep
116
protocol, which consists of a startup phase followed by systematic changes in firepower and finally
117
a shutdown phase.25 Firepower was changed by controlling the rate at which dried wood sticks
118
(dimensions: 1.5 cm x 1.5 cm x 12 cm or 3 cm x 3 cm x 12 cm) with no bark were added to, or
119
removed from, the combustion zone. Carbon and moisture contents of the fuels are listed in Table
120
S2. The firepower sweep protocol has 5 to 7 intermediate fire power steps or holds that span the
121
full range of functional firepower that is feasible for a given stove design. Each firepower step
122
lasts five- to ten-minutes (to collect adequate mass on a filter for gravimetric analysis and to obtain
123
a few SMPS scans) during which there is quasi-steady operating conditions. There are transient
124
operations between each firepower step as the fuel loading is adjusted. Finally, the protocol
125
includes startup and shutdown phases, which can contribute significantly to BC emissions.8,27 Each
126
firepower step is intended to represent different cooking practices (e.g., high firepower is often
127
used for boiling water while low firepower is common during simmering). This protocol allows
128
us to systematically investigate emissions over a wide range of operating conditions for the same
129
stove (versus simply integrating the data over the entire test).
130
Experimental setup
131
The experimental setup is shown in Figure S1. The entire emissions were captured and diluted
132
inside a constant flow hood. At the hood outlet, a sample of diluted exhaust was isokinetically
133
drawn through a cyclone that provided an aerodynamic cut at 2.5 µm and then into a suite of gas
134
and particle instruments. Gas measurements included CO and CO2 (Siemens ULTRAMAT gas
135
analyzer). Refractory BC (rBC) particle mass concentrations and size distributions were measured
136
using a single-particle soot photometer (SP2 [DMT]). The SP2 uses a 1064 nm laser to incandesce
137
rBC particles which is calibrated to rBC mass.28,29 The rBC emission factors measured by the SP2
ACS Paragon Plus Environment
Page 6 of 33
Page 7 of 33
Environmental Science & Technology
138
were linearly correlated (Figure S2, R2=0.8) with BC emission factors measured by the micro-soot
139
sensor (MSS) over a 4-fold increase in rBC emission factors. Moreover, the residual data did not
140
exhibit a curvature indicating no concentration-based bias. Therefore, the SP2 appears subject to
141
minimal coincidence errors. If these errors were significant, we would expect a deviation from the
142
linear behavior at high rBC emission factors. Particle light-scattering and absorption coefficients
143
were measured using two photo-acoustic extinctiometers (PAX [DMT]) operating at 405 and 532
144
nm. All coefficients measured by the PAX were above detection limits (~5 Mm-1). Particle size
145
distributions were measured using a Scanning Mobility Particle Sizer (SMPS [TSI Inc.]). Prior to
146
sampling with the particle-phase instruments, the emissions were diluted by an additional factor
147
of 99:1 using two Dekati diluters in series (Figure S1). The SP2 was calibrated using size-selected
148
fullerene soot particles.30 The PAXs were calibrated using PSL spheres for scattering and fullerene
149
soot for absorption, following the manufacturer’s recommendation. Real-time data were collected
150
continuously during the entire teste; these data were then averaged on a one-minute time basis for
151
analysis.
152
Teflon filters were also collected at the hood outlet to determine gravimetric PM mass
153
emissions. Separate filters were collected during each firepower step, startup and shutdown periods
154
but not during the transient periods between firepower steps. Filters were equilibrated for 24 hours
155
before weighing in a temperature and relative humidity-controlled environment.
156
Stove operational metrics and aerosol optical properties
157
We examined the influence of three operational metrics on aerosol optical properties: the
158
firepower, the MCE, and the BC/PM. Firepower and MCE are measures of the combustion
159
conditions. MCE is the ratio of background corrected CO2 concentrations to background corrected
160
CO+CO2 concentrations (MCE=ΔCO2/(ΔCO+ΔCO2)). The background corrections for the CO
ACS Paragon Plus Environment
Environmental Science & Technology
Page 8 of 33
161
and CO2 signals were negligible. Firepower is the fuel energy content (efuel) multiplied by the fuel
162
feed rate. Fuel feed rate is calculated using emission rates of CO (ERCO) and CO2 (ERCO2) and a
163
carbon fraction (fc) for each fuel as given in Table S2.
164
𝐹𝑖𝑟𝑒𝑝𝑜𝑤𝑒𝑟 = 𝑒𝑓𝑢𝑒𝑙.1/𝑓𝑐. 𝐸𝑅𝐶𝑂.
165
Where MC, MCO, and MCO2, are the molar masses for carbon, CO, and CO2 respectively. Emission
166
rates for CO and CO2 are determined using the measured CO and CO2 concentrations and the total
167
hood and instrument flow rates. Measured energy densities of the fuels (efuel) were: 16.6 MJ/kg for
168
coconut, 18.6 MJ/kg for Douglas Fir, 19.4 MJ/kg for Eucalyptus, and 28.6 MJ/kg for lump
169
hardwood. The BC/PM is a commonly measured characteristic of aerosol emissions and may
170
indicate coating around BC particles which can influence aerosol optical properties. The BC/PM
171
was calculated using filter-based gravimetric PM mass (collected for each firepower step, for
172
startup and shutdown period) and SP2-based rBC (averaged over the same periods as the filters).
(
( ) + 𝐸𝑅 .( )) 𝑀𝐶
𝑀𝐶𝑂
𝑀𝐶 𝐶𝑂2 𝑀𝐶𝑂 2
(1)
173
Averages (one-minute and for each firepower hold, startup, and shutdown periods) of the BC-
174
mass absorption cross-sections (MACBC) were calculated by dividing the absorption coefficients
175
measured by the PAX (babs) by rBC mass measured using the SP2 (MACBC=babs/rBC). MAC per
176
total PM mass were calculated for each filter sample as: MACPM=babs/PM. Mass scattering cross-
177
sections (MSCPM) were calculated for each filter sample by dividing the PAX measured scattering
178
coefficients (bsca) by gravimetric PM mass (MSCPM =bsca/PM). The absorption angstrom exponent
179
(AAE, wavelength dependence of absorption) was calculated using the PAX absorption
180
532 measurements at 405 (𝑏405 𝑎𝑏𝑠 ) and 532 nm (𝑏𝑎𝑏𝑠 ), as shown in Equation 2.
ACS Paragon Plus Environment
Page 9 of 33
Environmental Science & Technology
log 181
𝐴𝐴𝐸 = ―
( ) 𝑏405 𝑎𝑏𝑠 𝑏532 𝑎𝑏𝑠
(2)
( )
405 log 532
182
The single scattering albedo (SSA) is the ratio of scattered to extinguished light (sum of
183
scattering and absorption) that is incident upon a particle. SSA is an estimate of the overall
184
radiative impact from emissions (SSA>0.85 is usually associated with net cooling and SSA
