The Effects of Air Flow Rates, Secondary Air Inlet ... - ACS Publications

Aug 5, 2016 - Department of Mechanical Engineering, Gonzaga University, 502 E. Boone ... Factor(E) Ventures, 430 N. College Ave., Fort Collins, Colora...
0 downloads 0 Views 26MB Size
Subscriber access provided by Northern Illinois University

Article

The effects of air flow rates, secondary air inlet geometry, fuel type, and operating mode on the performance of gasifier cookstoves Jessica Tryner, James Tillotson, Marc E. Baumgardner, Jeffrey Mohr, Morgan DeFoort, and Anthony J Marchese Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b00440 • Publication Date (Web): 05 Aug 2016 Downloaded from http://pubs.acs.org on August 5, 2016

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 34

Environmental Science & Technology

The effects of air flow rates, secondary air inlet geometry, fuel type, and operating mode on the performance of gasifier cookstoves Jessica Tryner,† James W. Tillotson,† Marc E. Baumgardner,‡ Jeffrey T. Mohr,† Morgan W. DeFoort,¶ and Anthony J. Marchese∗,† 1

†Department of Mechanical Engineering, Colorado State University, 1374 Campus Delivery, Fort Collins, CO 80523-1374 ‡Department of Mechanical Engineering, Gonzaga University, 502 E. Boone Ave., AD Box 26, Spokane, WA 99258 ¶Factor(E) Ventures, 430 N. College Ave., Fort Collins, CO 80524 E-mail: [email protected] Phone: (970) 491-2328. Fax: (970) 491-3827

2

Abstract

3

Development of biomass cookstoves that reduce emissions of CO and PM2.5 by more

4

than 50% and 95%, respectively, compared to a three-stone fire has been promoted as

5

part of efforts to reduce exposure to household air pollution (HAP) among people that

6

cook with solid fuels. Gasifier cookstoves have attracted interest because some have

7

been shown to emit less CO and PM2.5 than other designs. A laboratory test bed and

8

new test procedure were used to investigate the influence of air flow rates, stove geom-

9

etry, fuel type, and operating mode on gasifier cookstove performance. Power output,

10

CO emissions, PM2.5 emissions, fuel consumption rates, producer gas composition,

1 ACS Paragon Plus Environment

Environmental Science & Technology

11

and fuel bed temperatures were measured. The test bed emitted < 41 mg·MJ−1 d PM2.5

12

and < 8 g·MJ−1 d CO when operating normally with certain prepared fuels, but order

13

of magnitude increases in emission factors were observed for other fuels and during

14

refueling. Changes in operating mode and fuel type also affected the composition of

15

the producer gas entering the secondary combustion zone. Overall, the results suggest

16

that the effects of fuel type and operator behavior on emissions need to be considered,

17

in addition to cookstove design, as part of efforts to reduce exposure to HAP.

18

Introduction

19

An estimated 2.8 billion people worldwide rely on solid fuels for cooking 1 and, as a re-

20

sult, face a range of health risks associated with household air pollution (HAP). 2–4 Emis-

21

sions targets designed to reduce HAP have been established for solid-fuel cookstoves, 5,6

22

and questions remain as to whether efforts to reduce HAP should focus on improving ac-

23

cess to clean fuels (e.g., LPG and electricity), disseminating improved solid-fuel cookstoves,

24

or both. 6–8 Development of “advanced combustion stoves” 8 that result in large reductions

25

in carbon monoxide (CO) and particulate matter (PM) emissions compared to a baseline

26

three-stone fire has been encouraged, and several discussions 8–10 have hypothesized that top-

27

lit up draft (TLUD) gasifier cookstoves have the potential to achieve the low emissions that

28

are sought. Cookstoves that utilize the TLUD design, 11,12 in which solid fuel is gasified and

29

the resulting producer gas burns in a secondary combustion zone, have attracted interest

30

because they often emit less CO and PM than other biomass cookstove designs in both lab-

31

oratory 13–15 and field 16–18 settings. TLUDs have been marketed and studied extensively in

32

India 10,16–22 and China, 23,24 but there are also reports of TLUDs being marketed elsewhere

33

in Asia 12,25–27 as well as in Africa, Central America, and South America. 12,26

34

A TLUD is designed to operate as inverted downdraft gasifier. 11 In a stratified downdraft

35

gasifier, primary air first passes through unreacted solid biomass before oxidizing the volatiles

36

released from the biomass in the flaming pyrolysis zone. The heat released by oxidation 2 ACS Paragon Plus Environment

Page 2 of 34

Page 3 of 34

Environmental Science & Technology

37

reactions drives the pyrolysis and drying of the upstream biomass. The products leaving the

38

flaming pyrolysis zone are reduced when they pass through the downstream char layer. 28

39

As a result of the presence of oxidizer and high temperatures in the flaming pyrolysis zone,

40

most of the tars released during pyrolysis are consumed and the final tar concentration in

41

the producer gas is low. 29 In a conventional updraft gasifier, on the other hand, primary air

42

reacts with char in the oxidation zone at the bottom of the fuel bed. Once the oxygen in

43

the primary air has been consumed, products of the oxidation reactions are reduced as they

44

pass through the remainder of the char layer. Hot gases leaving the char layer then pyrolyze

45

and dry the unreacted solid fuel as they flow up through the fuel bed. 28 Because the gases

46

and tars released during pyrolysis only pass through low-temperature sections of the fuel

47

bed where no oxidizer is present, the final tar concentration in the producer gas is high. 29

48

Despite the low emissions frequently observed with the TLUD design, some TLUDs

49

have been shown to emit more CO and PM2.5 than a three-stone fire. 15 In a recent study,

50

emissions from natural-draft TLUDs were found to be affected by fuel type and, in some

51

cases, dominated by large, transient increases in emission rates. 30 Increases in emission rates

52

have been observed when stoves transitioned from gasifying volatile biomass to burning the

53

char left behind by the gasification process 21,30–34 and when batch-fed stoves were refueled

54

by adding volatile fuel on top of a hot char bed. 30,35 Emissions associated with refueling are

55

important because real-world users have been found to refuel TLUDs in this manner. 27

56

Given the promising yet variable nature of TLUD performance, questions remain as to:

57

(1) whether or not TLUDs can meet the Tier 4 performance targets outlined in the ISO

58

IWA tiers 5 and (2) what must be done, in terms of cookstove design and dissemination, to

59

achieve consistently low emissions. Accordingly, the objectives of the present study were

60

to characterize how stove design parameters and fuel properties, as well as refueling and

61

transition to char burnout, affect TLUD operation and performance.

62

This study expands upon previous work in which the effects of fuel type, 21,30,32 fuel mois-

63

ture content, 24,36 fuel particle size, 32,36–38 primary air flow rate, 31,33,34,39 ratio of secondary

3 ACS Paragon Plus Environment

Environmental Science & Technology

64

to primary air flow, 21,33,39 secondary air inlet area, 21,40 chimney sections, 38 and operating

65

mode 23,32 on gasifier cookstove performance were investigated experimentally. A comprehen-

66

sive evaluation of performance, including measurements of CO emissions, PM2.5 emissions,

67

useful power output, producer gas composition, and fuel bed temperatures, was completed.

68

As noted previously, existing test protocols such as the Water Boiling Test 41 (WBT)

69

are not well suited for batch-fed cookstove designs 42 and may fail to capture the range

70

of operating modes and emission rates observed in the field. 23,27 Consequently, a new test

71

procedure was developed to evaluate the range of performance that can be observed with

72

batch-fed TLUDs in various modes of operation including refueling and transition to char

73

burnout.

74

Experimental

75

A description of the test bed used to vary cookstove design parameters is provided below,

76

followed by descriptions of the test matrices, the test procedure used to evaluate performance,

77

and the equations used to calculate fuel consumption rate, fuel gas production rate, and

78

global equivalence ratio in the secondary combustion zone.

79

Modular Test Bed

80

The modular TLUD test bed, shown in Figure 1, was constructed from stainless steel (SS)

81

and is similar to those used in previous studies. 31,32,39 Primary and secondary air flows were

82

driven by compressed air, which was regulated to 276 kPa, and flow rates were controlled

83

using rotameters. Secondary air was preheated by two electric heaters connected to PID

84

controllers. These features allowed the primary air flow rate, secondary air flow rate, and

85

secondary air temperatures to be set to known values so that: (1) the influence of these

86

parameters on performance could be investigated and (2) the influence of other design pa-

87

rameters could be investigated independently of the air flow rates.

4 ACS Paragon Plus Environment

Page 4 of 34

Page 5 of 34

Environmental Science & Technology

88

The ring through which secondary air entered the combustion chamber was removable

89

so that the inlet geometry could be varied. Secondary air inlets that featured 32 holes

90

ranging from 2 to 10 mm in diameter were tested. Inlets that featured 4-mm-diameter holes

91

positioned at angles to induce swirl in the horizontal plane or direct the secondary air down

92

towards the fuel bed were also tested (see Figure S1, SI Section S1.1). During relevant

93

experiments, the arms that supported the pot were adjusted to change the distance between

94

the top of the test bed and the pot, constrictions intended to recirculate gases inside the

95

stove were inserted before and after the secondary air inlet (see Figure S2, SI Section S1.1),

96

and an insulated chimney with a diameter of 108 mm and a height of 100 mm 43 was installed

97

downstream of the secondary air inlet. Another assembly allowed 10% of the secondary air

98

to be injected earlier in the fuel chamber through six 2-mm-diameter holes spaced equally

99

around the circumference (see Figure S3, SI Section S1.1). Figure 1: A photograph and a cross-section of the modular TLUD test bed with key features labeled.

100

The test bed was mounted on a digital balance that recorded the mass of the system at

101

a rate of 1 Hz. Thermocouples installed along the height of the fuel chamber tracked the

102

progression of the primary combustion zone. Gas entering the secondary combustion zone

103

was sampled for compositional analysis via a probe installed between the fuel bed and the

104

secondary air inlet. Several additional thermocouples were installed throughout the test bed,

105

and temperature data from all thermocouples were logged at 1 Hz. For information on the

106

thermocouples used and their locations, see Figure S4 and Table S1, SI Section S1.1.

107

Test Matrix

108

Experiments were conducted using the two test matrices shown in Table 1. For each test, one

109

parameter was varied and all other parameters were held constant at the default values shown

110

in bold. Together, the parameters shown in bold represent the “baseline”configuration. For

5 ACS Paragon Plus Environment

Environmental Science & Technology

111

more information on the baseline geometry and justifications for the default values see SI

112

Section S1.2. Table 1: The two test matrices for which experiments were completed. Baseline configuration data are indicated in bold. Matrix 1 — Secondary air delivery parameters Parameter Values tested Parameter Values tested Secondary to 2 Secondary air 4-mm-dia. holes primary air 3 opening 6-mm-dia. holes flow ratio 4 8-mm-dia. holes (mass basis) 5 10-mm-dia. holes Secondary air 0◦ Secondary air 0◦ swirl angle 15◦ downward 10◦ ◦ 30 angle 20◦ 45◦ 30◦ Pot gap 15 mm Constriction None 30 mm location Before secondary air inlet, 45 mm 2.50:4.25 ratio After secondary air inlet, 2.50:4.25 ratio ◦ Secondary air 100 C temperature 150 ◦ C 200 ◦ C 250 ◦ C 300 ◦ C Matrix 2 Primary air flow rates, fuel properties, and secondary air delivery parameters Parameter Values tested Parameter Values tested Primary air 15 Fuel moisture 0 flow rate 20 content 7 −1 (g·min ) 25 (weight % on 15 30 wet basis) 25 Fuel type Corn cob chips Fuel bulk 174 Eucalyptus chips density (corn 137 Douglas fir chips cobs) (kg·m−3 ) 126 Lodgepole pine pellets Secondary air 4-mm-dia. holes delivery 2-mm-dia. holes parameters Insulated chimney Early secondary air injection

113

For the first test matrix, CO emissions, useful power output, fuel consumption rate, and

6 ACS Paragon Plus Environment

Page 6 of 34

Page 7 of 34

Environmental Science & Technology

114

mass loss rate were measured while secondary air delivery parameters were varied. Douglas

115

fir (Pseudotsuga menziesii ) chips with a moisture content of 7% on a wet basis and a bulk

116

density of 156 kg·m−3 (95% CI = 155–157) were used as the fuel and the primary air flow

117

rate was set to 25 g·min−1 .

118

For the second test matrix, CO and PM2.5 emissions, useful power output, producer gas

119

composition, fuel consumption rate, and mass loss rate were measured while primary air

120

flow rate, fuel properties, and secondary air delivery parameters were varied. The Douglas

121

fir chips used as the default fuel had a bulk density of 160 kg·m−3 (95% CI = 159–161)

122

and were more uniform in size and shape than the chips used in the first matrix (see Figure

123

S5, SI Section S1.2). The “corn cob chips”and “174 kg·m−3 bulk density”cases shown in

124

italics in Table 1 were identical. A complete list of test cases, and test parameter values for

125

each case, can be found in Tables S2 and S3, SI Section S1.2. For detailed fuel properties,

126

including the results of proximate and ultimate analyses, biochemical compositions, heating

127

values, densities, porosities, and moisture contents, see Tables S5 and S6, SI Section S1.2.

128

Test Procedure

129

The test procedure that was developed to capture three characteristic operating modes (nor-

130

mal, post-refueling, and char burnout) is represented graphically in Figure 2. During the first

131

phase (normal operation), the test bed started at room temperature and the fuel chamber

132

was filled with 1.5 L of biomass. To ignite the fuel bed, 50 mL of biomass was wetted with 2

133

mL of kerosene, placed on top of the fuel bed, and lit with a match. A SS pot containing 2.5

134

L of water at 15 ± 2 ◦ C was placed on the pot support, and emissions were measured while

135

the water was heated to 90 ◦ C. This volume of water (2.5 L) was selected to ensure that

136

a pot of water could be boiled using the lower bulk density fuels. This phase was similar

137

to the cold start phase of the WBT 41 and Emissions and Performance Test Protocol. 44,45

138

As shown in previous studies 27,30 and the results below, the emissions measured during this

139

phase represented the best-case scenario for high-power performance. 7 ACS Paragon Plus Environment

Environmental Science & Technology

Figure 2: Plot of typical CO emissions (mg·s−1 ) and water temperature (◦ C) as a function of time for the three phase test procedure (normal operation, post-refueling, and char burnout). The CO emission rates shown on the left y-axis are representative of those measured for the baseline test case in Matrix 2.

140

When the water temperature reached 90 ◦ C and the first phase ended, the pot of water

141

was removed and the test bed was allowed to operate normally until the secondary flame

142

extinguished. The latter event signaled that the first batch of fuel had been consumed. For

143

the second phase (refueling), 1.0 L of biomass fuel was added to the fuel chamber on top of

144

the hot char bed left behind by the first batch of fuel. The test bed was re-lit using 50 mL of

145

biomass wetted with 2 mL of kerosene. Emissions and power-output measurements resumed

146

when a pot containing 2.5 L of water at 15 ± 2 ◦ C was placed on the pot support and

147

continued until the water temperature reached 90 ◦ C or the secondary flame extinguished.

148

As shown in previous studies 27,30 and the results below, emissions measured during this

149

phase typically represented the worst-case scenario for high-power performance.

150

The third phase started when the secondary flame extinguished. During this phase, the

151

pot of water was left on the pot support and the char remaining from the two fuel batches

152

was allowed to burn while emissions were measured for 20 minutes.

153

The producer gas entering the secondary combustion zone was sampled for 30–60 seconds

154

starting at ten minutes after the start of Phase 1, five minutes after the start of Phase 2,

155

and five minutes after the start of Phase 3. The gas composition was analyzed using a gas

156

chromatograph with a thermal conductivity detector. For more information on the fume

157

hood in which experiments were conducted, the instrumentation used to measure emissions,

158

as well as producer gas sampling and analysis, see SI Section S1.3.

159

During tests with Lodgepole pine pellet fuel, two modifications were made to Phases 1

160

and 2 of the procedure: (1) 2.5 L of water was boiled twice (i.e., 5 L of water was boiled using

161

two identical pots) and (2) the producer gas was sampled 30 minutes after the start of Phase

162

1 and 15 minutes after the start of Phase 2. These modifications, and the justifications for

8 ACS Paragon Plus Environment

Page 8 of 34

Page 9 of 34

163

Environmental Science & Technology

them, are discussed in SI Section S1.3.

164

The number of replicates collected for each data point is listed in Table S4, SI Section

165

S1.2. A recent study suggests that 10 or more replicate tests may be necessary for the width

166

of the confidence intervals on PM2.5 emissions measurements to asymptote to a constant

167

value. 46 In this study, 3-4 replicates were collected for most data points to allow a large

168

number of cases to be screened and 90% confidence intervals were calculated for the results.

169

Calculations

170

The average useful power outputs and dry fuel consumption rates for Phases 1 and 2 were

171

calculated using equations similar to those used by Huangfu et al. 24 . Average useful power

172

output was calculated using Equation 1: Pusef ul =

mH2O,i cH2O (Tf − Ti ) + hf g,H2O (mH2O,i − mH2O,f ) tf − ti

(1)

173 174

where mH2O,i and mH2O,f were the masses of the water in the pot at the start and end of

175

the phase, respectively, in g, cH2O the specific heat of water in J·g−1 ·K−1 , Tf the temperature

176

of the water in the pot at the end of the phase in ◦ C, Ti the temperature of the water at the

177

start of the phase in ◦ C, hf g,H2O the heat of vaporization of water in J·g−1 , tf the time at

178

which the water in the pot reached Tf , in s, and ti the time at which the pot of water was

179

placed on the stove in s.

180

The average rate of dry solid biomass fuel consumption was calculated using Equation 2: m ˙ dry solid f uel =

msolid f uel (1 − M Cf uel ) te − ti

(2)

181 182

where msolid f uel was the mass of fuel added to the stove at the start of the phase in g,

183

M Cf uel the moisture content of the fuel as weight % on a wet basis, ti the time at which

184

the stove was lit in s, and te the time at which the secondary flame extinguished at the end

185

of the phase in s. The solid fuel consumption rate was calculated as shown in Equation 2, 9 ACS Paragon Plus Environment

Environmental Science & Technology

Page 10 of 34

dm dt

 , because

186

and not simply set equal to the mass loss rate recorded by the digital balance

187

some mass of solid char was left behind after the biomass had devolatilized. In this manner,

188

the remaining char was treated as a product of the gasification process.

189

190

The mass flow rate of producer gas entering the secondary combustion zone was calculated by applying the conservation of mass to a control volume drawn around the fuel/char bed: dm =m ˙ primary air − m ˙ f uel gas dt

191

where

dm dt

(3)

was the rate of change of the mass of the fuel/char bed recorded by the digital

192

balance in g·min−1 , m ˙ primary air the primary air flow rate in g·min−1 , and m ˙ f uel gas the mass

193

flow rate of gas leaving the fuel bed and entering the secondary combustion zone in g·min−1 .

194

The stoichiometric air-to-fuel ratio in the secondary combustion zone was calculated from

195

the gas composition and fuel gas flow rate using Equation 4:

(m ˙ air /m ˙ f uel )stoic

Mair = × 4.76 × Mgas



1 1 7 xCO + xH2 + 2xCH4 + xC2 + 5xC3 2 2 2

 (4)

196

where Mair and Mgas were the molecular weights of air and the fuel gas, respectively,

197

in g·mol−1 , and x the mole fraction of each component in the fuel gas. Mole fractions of

198

hydrocarbons larger than C3 were negligible. The small concentration of O2 in the gas

199

samples was neglected.

200

201

Global equivalence ratios in the fuel bed and secondary combustion zone were calculated using Equation 5: φ=

(m ˙ air /m ˙ f uel )stoic (m ˙ air /m ˙ f uel )

(5)

202

where m ˙ air was the mass flow rate of primary or secondary air in g·min−1 and m ˙ f uel the

203

fuel consumption rate calculated using Equation 2 or the fuel gas flow rate calculated using

204

Equation 3 in g·min−1 .

205

The equations used to calculate commonly-reported CO and PM2.5 emissions metrics 10 ACS Paragon Plus Environment

Page 11 of 34

Environmental Science & Technology

206

are provided in SI Section S1.4. Briefly, CO emissions were calculated by integrating the

207

concentration of CO measured in the fume hood exhaust at 1 Hz using NDIR spectroscopy.

208

Total PM2.5 emissions were calculated, assuming isokinetic sampling, from the mass of PM2.5

209

collected on a PTFE filter.

210

Results and discussion

211

A complete set of experimental results is provided in SI Section S2. The parameters that

212

had the greatest influence on performance are discussed below.

213

Effects of air flow rates, fuel moisture, and secondary air delivery

214

parameters

215

Useful power output increased linearly (r2 ≥ 0.99) with primary air flow for the range of flow

216

rates tested (see Figure 3). The global equivalence ratio in the fuel bed remained between

217

3 and 5 when the primary air flow rate was varied, and the average dry fuel consumption

218

rate increased linearly with primary air flow rate (see Figure S21, SI Section S2.2). The

219

composition of the producer gas entering the secondary combustion zone during Phase 1 was

220

not affected by primary air flow rate (see Figure S27, SI Section S2.3). The producer gas

221

consisted primarily of H2 , CO, CH4 , CO2 , and N2 , and the mole fraction of hydrocarbons

222

larger than CH4 was less than 2% in all samples collected during the second test matrix. Figure 3: Effects of primary air flow rate and fuel moisture content on average useful power output during normal operation (Phase 1, green triangles) and post-refueling (Phase 2, gold circles). Black symbols refer to the baseline configuration (Douglas fir chip fuel, 7% moisture content, 25 g·min−1 primary air, 75 g·min−1 secondary air, 200 ◦ C secondary air temperature, 4-mm-dia. secondary air holes, no swirl, no downward angle, 15 mm pot gap, no flow constriction, no other features). Error bars represent 90% confidence intervals. Data points have been offset slightly on either side of the values listed in Table 1 to improve the readability of the plot.

223

The average high-power PM2.5 emissions measured during normal operation remained 11 ACS Paragon Plus Environment

Environmental Science & Technology

224

between 21 mg·MJd −1 and 42 mg·MJd −1 when the primary air flow rate was varied (see

225

Figure S18, SI Section S2.2). The secondary flame was difficult to re-light after refueling

226

when the primary air flow rate was low and, as a result, the variability in Phase 2 PM2.5

227

emissions was high for the test cases with primary air flow rates of 15 g·min−1 and 20 g·min−1 .

228

Higher primary air flow rates may have made the test bed easier to re-light by causing fuel

229

gas production to resume more quickly. The concentration of CO in the producer gas during

230

Phase 2 increased with primary air flow rate (see Figure S27, SI Section S2.3).

231

Useful power output and average dry fuel consumption rate decreased linearly with in-

232

creasing fuel moisture content (see Figure 3 and Figure S21, SI Section S2.2). Fuel moisture

233

content did not affect CO or PM2.5 emissions during normal operation (Phase 1), but Phase

234

2 (post-refueling) emissions increased with increasing moisture content (see Figure S18, SI

235

Section S2.2). Similar to the test cases with low primary air flow rates, the secondary com-

236

bustion zone was difficult to re-light after refueling for higher moisture contents (15% and

237

25%), and relighting difficulties likely contributed to the higher Phase 2 emissions that were

238

measured for the higher moisture content fuels. Higher moisture contents may have made

239

the test bed difficult to re-light by preventing fuel gas production from resuming quickly.

240

A minimum in high-power (Phase 1 and 2) CO emissions was observed for secondary to

241

primary air flow ratios of 3:1 and 4:1 (see Figure 4). This result is in agreement with previous

242

studies 21,39 that recommended a secondary to primary air flow ratio of approximately 3:1

243

for high-power operation. Figure 4: Effects of secondary to primary air flow ratio and secondary air opening size (secondary air velocity) on average high-power CO emissions. Average values that were calculated from only two replicates are marked with a ‘+’. Error bars represent 90% confidence intervals for markers with no ‘+’ and the total range of the measurements for markers with a ‘+’. Data points have been offset slightly on either side of the values listed in Table 1 to improve the readability of the plot.

244

Higher secondary air velocities resulted in lower CO emissions (see Figure 4). These

245

higher velocities most likely led to better mixing between the secondary air and the fuel 12 ACS Paragon Plus Environment

Page 12 of 34

Page 13 of 34

Environmental Science & Technology

246

gas. This result stands in contrast to previous work 21 in which CO emissions decreased

247

with increasing secondary air inlet area; however, the total secondary air flow rate was not

248

decoupled from the inlet area in that study. 40 The 2-mm-diameter secondary air inlet holes

249

tested in the second matrix resulted in average Phase 1 high-power CO and PM2.5 emissions

250

that were only 0.9 g·MJd −1 and 8 mg·MJd −1 lower, respectively, than those for the test case

251

with 4-mm-diameter holes (see Figure S20, SI Section S2.2). This result may indicate

252

diminishing returns associated with further increases in secondary air velocity. Results may

253

be different for a natural draft or fan-powered stove. Primary and secondary air flow rates

254

were fixed in the modular test bed. In a natural draft or fan-powered TLUD, the secondary

255

air inlet area and/or the presence of a chimney section would be expected to affect the

256

secondary air flow rate.

257

Varying the other secondary air delivery parameters investigated in the first and second

258

test matrices did not improve performance compared to the baseline configuration (see SI

259

Sections S2.1 and S2.2). For example, the 15◦ swirl angle had a negligible effect on

260

emissions, and the 30◦ and 45◦ swirl angles resulted in higher Phase 1 CO emissions relative

261

to the baseline case (see Figure S8, SI Section S2.1). Photographs of the secondary

262

combustion zone for the 15◦ , 30◦ , and 45◦ swirl test cases are shown in Figure S16, SI

263

Section S2.1. In the photograph of the 15◦ case, the secondary flames are attached to

264

the air inlets and have a bluish color. In contrast, the flames are not attached, are more

265

luminous, and rise further out of the top of the stove in the 30◦ and 45◦ degree cases. The

266

higher luminosity suggests that PM2.5 emissions were also higher in these two cases. The

267

larger swirl angles appear to have promoted soot formation by increasing the residence time in

268

fuel-rich combustion zones rather than enhancing fuel/air mixing. In addition, impingement

269

of the flame on the cold cooking surface in the 30◦ and 45◦ swirl cases would be expected to

270

quench the reactions that oxidize CO and PM.

13 ACS Paragon Plus Environment

Environmental Science & Technology

271

Effects of fuel type

272

Changes in fuel type sometimes resulted in order of magnitude changes in the average high-

273

power PM2.5 emissions measured during normal operation (see Figure 5). The test bed

274

exhibited the highest emissions when fueled with corn cobs and the lowest emissions when

275

fueled with Lodgepole pine pellets. The producer gas composition during normal operation

276

(Phase 1) was also affected by fuel type (see Figure 6). Differences in producer gas composi-

277

tion were not attributed to differences in elemental composition between the fuels. Ultimate

278

analysis revealed that C:H:N:O ratios for the fuels were almost identical (see Table S5, SI

279

Section S1.2). The differences in producer gas composition and emissions may have re-

280

sulted from differences in fuel particle size, density, biochemical composition, and inorganic

281

constituents affecting yields of light gases and tar from the gasification process. 36,47–50 Figure 5: Effects of varying fuel type on high-power CO and PM2.5 emissions during normal operation (Phase 1) and post-refueling (Phase 2). Gray symbols correspond to the test case with the highest bulk density corn cobs (174 kg·m−3 ). Error bars represent 90% confidence intervals. The data points representing the Phase 1 and Phase 2 results have been offset to improve the readability of the plot.

Figure 6: Effects of fuel type on producer gas composition for normal operation (Phase 1), post-refueling (Phase 2), and char burnout (Phase 3). Error bars represent 90% confidence intervals.

282

When the Lodgepole pine pellet fuel was used, average H2 concentrations in the producer

283

gas were 103% and 23% higher during Phase 1 and Phase 2, respectively, compared to the

284

baseline (Douglas fir chips). Varunkumar et al. 32 also reported that pellet fuel resulted in

285

higher H2 concentrations in the producer gas compared to wood chip fuel. The maximum fuel

286

bed temperatures measured during Phase 1 were highest when the Lodgepole pine pellet fuel

287

was used (see Figure S37, SI Section S2.3), and higher temperatures would be expected

288

to result in increased conversion of tars to light gases. 36,51,52 In addition, the Lodgepole

289

pine pellets were 6.4 mm in diameter, whereas pieces of the chipped fuels were ≤ 3-mm14 ACS Paragon Plus Environment

Page 14 of 34

Page 15 of 34

Environmental Science & Technology

290

thick on average. The larger particle thickness may have resulted in a longer residence time

291

for primary pyrolysis products to undergo secondary reactions inside the particle and, as a

292

result, increased conversion of tars to light gases. 47

293

When the chipped corn cob fuel was used, the average H2 and CO2 concentrations were

294

64% lower and 43% higher, respectively, during Phase 1 compared to the baseline. The max-

295

imum fuel bed temperatures measured during Phase 1 were lowest when the corn cob fuel

296

was used (see Figure S37, SI Section S2.3), and lower temperatures would be expected to

297

result in reduced conversion of tars to light gases. 36,51,52 The corn cobs had a lower lignin con-

298

tent and higher hemicellulose content compared to the wood fuels. Hemicellulose pyrolyzes

299

more rapidly and over a narrower temperature range compared to lignin. 48,53 Yang et al. 48

300

observed a higher concentration of H2 in the gaseous products of lignin pyrolysis and a higher

301

concentration of CO2 in the gaseous products of hemicellulose pyrolysis. Worasuwannarak

302

et al. 49 observed higher tar yields from pyrolysis of corn cobs than from pyrolysis of rice

303

husk and rice straw.

304

Higher tar yields could contribute to higher PM2.5 emissions. Primary and secondary

305

tars released during pyrolysis of the fuel can undergo further reaction in high-temperature

306

zones to form PAHs, 54 which are precursors to soot, and condensible organics in the ex-

307

haust can condense onto existing particles or nucleate to form new particles. 55,56 Tar yields

308

were not quantified in these experiments. The manner in which fuel properties affect tar

309

yields, producer gas composition, and overall emissions from the stove are worthy of further

310

investigation.

311

Effects of operating mode

312

Phase 2 emissions where higher than Phase 1 emissions for all test cases. The concentrations

313

of H2 , CO, and CH4 in the producer gas were highest during Phase 1 (normal operation)

314

(see Figure 6). For the baseline case (Douglas fir chips), the average H2 , CO, and CH4 con-

315

centrations were 61%, 36%, and 23% lower, respectively, during Phase 2 (post-refueling) and 15 ACS Paragon Plus Environment

Environmental Science & Technology

316

93%, 81%, and 81% lower, respectively, during Phase 3 (char burnout). The concentration

317

of O2 /Ar in the producer gas was low during Phases 1 and 2, but was 3.3× the average of

318

the Phase 1 and Phase 2 values during Phase 3. The concentrations of H2 , CO, and CO2

319

observed during normal operation (Phase 1) and char burnout (Phase 3) were similar to

320

those reported by Varunkumar et al. 32 .

321

The temperatures measured in the fuel chamber revealed that gasification proceeded from

322

the top of the biomass fuel bed to the bottom during Phase 1 and from the bottom to the

323

top during Phase 2 (see Figure 7). During Phase 1, the temperatures in the top of the fuel

324

bed increased first, followed by the temperatures in the middle and bottom. This progression

325

indicates that the stove operated as an inverted downdraft gasifier (i.e., as a TLUD) during

326

Phase 1. After the first batch of fuel had completely gasified, the temperature of the bottom

327

thermocouples remained high. After the second batch of fuel was added on top of the

328

remaining hot char, the temperatures in the middle of the fuel bed increased first, followed

329

by the temperatures in the top and at the gas sampling height. This progression indicates

330

that the stove operated as a conventional updraft gasifier during Phase 2. Conventional

331

updraft gasifier operation would be expected to result in high tar concentrations in the

332

producer gas. 29 Figure 7: Temporal variation in temperature (◦ C) at four different vertical locations in the combustion chamber for the baseline case. See Figure 1 for thermocouple locations. The primary combustion zone progressed from the top to the bottom of the fuel bed during Phase 1, and from the bottom to the top of the fuel bed during Phase 2.

333

Together, the emissions, producer gas composition, and fuel bed temperature measure-

334

ments reveal that TLUDs only really operate as TLUDs, and exhibit the associated low

335

emissions, while the initial batch of biomass fuel is gasifying. The proportion of the mass

336

collected on the Phase 2 filter samples that is attributable to the transient emissions observed

337

immediately after re-lighting vs. steady state post-refueling operation is unknown. The de-

338

gree to which conventional updraft gasification influences steady state PM2.5 emissions, if at

16 ACS Paragon Plus Environment

Page 16 of 34

Page 17 of 34

Environmental Science & Technology

339

all, can be investigated by collecting separate filter samples for transient and steady state

340

post-refueling operation in future experiments.

341

Tier 4 high-power emissions were readily achievable when the baseline configuration was

342

fueled with Douglas fir wood chips or Lodgepole pine pellets and operating normally (during

343

Phase 1) (see Figure 5). However, the same stove exhibited Tier 0–3 emissions during normal

344

operation with a low-quality fuel (corn cobs) and post-refueling (during Phase 2). These

345

results suggest that the effects of fuel type and operator behavior on emissions from biomass

346

cookstoves need to be considered, in addition to cookstove design, as part of efforts to reduce

347

exposure to household air pollution resulting from solid fuel combustion. Locally available

348

fuels and user behavior will influence performance in the field, and improving user access to

349

high-quality prepared biomass fuels, such as pellet fuels, may be necessary to achieve Tier 4

350

emissions reliably.

351

Acknowledgement

352

The authors acknowledge support from U.S. Department of Energy award number DE-

353

EE0006086.

354

Supporting Information Available

355

Additional information on the experimental methods and a complete set of results graphs

356

are available in the Supporting Information.

357

358

This material is available free of charge via the Internet at http://pubs.acs.org/.

References

359

(1) Bonjour, S.; Adair-Rohani, H.; Wolf, J.; Bruce, N. G.; Mehta, S.; Pr¨ uss-Ust¨ un, A.;

360

Lahiff, M.; Rehfuess, E. A.; Mishra, V.; Smith, K. R. Solid Fuel Use for Household 17 ACS Paragon Plus Environment

Environmental Science & Technology

Page 18 of 34

361

Cooking: Country and Regional Estimates for 1980–2010. Environ. Health Perspect.

362

2013, 121, 784–790, DOI: 10.1289/ehp.1205987.

363

(2) Bruce, N.; Rehfuess, E.; Mehta, S.; Hutton, G.; Smith, K. Indoor Air Pollution. In

364

Disease Control Priorities in Developing Countries, 2nd ed.; Jamison, D. T., Bre-

365

man, J. G., Measham, A. R., Alleyne, G., Claeson, M., Evans, D. B., Jha, P., Mills, A.,

366

Musgrove, P., Eds.; The World Bank: Washington, DC, 2006; Chapter 42, pp 793–815.

367

(3) Lim, S. S. et al. A comparative risk assessment of burden of disease and injury at-

368

tributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a system-

369

atic analysis for the Global Burden of Disease Study 2010. Lancet 2012, 380, 2224 –

370

2260, DOI: 10.1016/S0140-6736(12)61766-8.

371

(4) Forouzanfar, M. H. et al. Global, regional, and national comparative risk assessment of

372

79 behavioural, environmental and occupational, and metabolic risks or clusters of risks

373

in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease

374

Study 2013. Lancet 2015, 386, 2287–2323, DOI: 10.1016/S0140-6736(15)00128-2.

375

(5) ISO

International

Workshop

on

Clean

and

Efficient

Cookstoves.

2012;

376

http://www.pciaonline.org/files/ISO-IWA-Cookstoves.pdf (accessed Feb 25, 2014).

377

(6) World Health Organization, WHO guidelines for indoor air quality: household fuel

378

combustion; WHO Document Production Services: Geneva, Switzerland, 2014.

379

(7) Subramanian, M. Deadly dinners. Nature 2014, 509, 548–551, DOI: 10.1038/509548a.

380

(8) Simon, G. L.; Bailis, R.; Baumgartner, J.; Hyman, J.; Laurent, A. Current debates and

381

future research needs in the clean cookstove sector. Energy Sustainable Dev. 2014, 20,

382

49–57, DOI: 10.1016/j.esd.2014.02.006.

383

(9) Smith, K. R.; Balakrishnan, K. Mitigating climate, meeting MDGs, and moderating

18 ACS Paragon Plus Environment

Page 19 of 34

Environmental Science & Technology

384

chronic disease: the health co-benefits landscape. In Commonwealth Health Ministers’

385

Update 2009 ; Commonwealth Secretariat: London, 2009; Chapter 4, pp 59–65.

386

387

388

389

390

391

392

(10) Smith, K. R.; Dutta, K. “Cooking with Gas”. Energy Sustainable Dev. 2011, 15, 115– 116, DOI: 10.1016/j.esd.2011.05.001. (11) Reed, T. B.; Larson, R. A wood-gas stove for developing countries. Energy Sustainable Dev. 1996, 3, 34–37, DOI: 10.1016/S0973-0826(08)60589-X. (12) Anderson, P. S.; Reed, T. B.; Wever, P. W. Micro-Gasification: What it is and why it works. Boiling Point 2007, 53, 35–37. (13) Jetter, J. J.;

Kariher, P. Solid-fuel household cook stoves:

Characterization

393

of performance and emissions. Biomass Bioenergy 2009, 33, 294–305, DOI:

394

10.1016/j.biombioe.2008.05.014.

395

(14) MacCarty, N.; Still, D.; Ogle, D. Fuel use and emissions performance of fifty cooking

396

stoves in the laboratory and related benchmarks of performance. Energy Sustainable

397

Dev. 2010, 14, 161–171, DOI: 10.1016/j.esd.2010.06.002.

398

(15) Jetter, J.; Zhao, Y.; Smith, K. R.; Khan, B.; Yelverton, T.; DeCarlo, P.; Hays, M. D.

399

Pollutant Emissions and Energy Efficiency under Controlled Conditions for Household

400

Biomass Cookstoves and Implications for Metrics Useful in Setting International Test

401

Standards. Environ. Sci. Technol. 2012, 46, 10827–10834, DOI: 10.1021/es301693f.

402

(16) Kar, A.; Rehman, I. H.; Burney, J.; Puppala, S. P.; Suresh, R.; Singh, L.; Singh, V. K.;

403

Ahmed, T.; Ramanathan, N.; Ramanathan, V. Real-Time Assessment of Black Carbon

404

Pollution in Indian Households Due to Traditional and Improved Biomass Cookstoves.

405

Environ. Sci. Technol. 2012, 46, 2993–3000, DOI: 10.1021/es203388g.

406

(17) Singh, S.; Gupta, G. P.; Kumar, B.; Kulshrestha, U. C. Comparative study of indoor air

19 ACS Paragon Plus Environment

Environmental Science & Technology

407

pollution using traditional and improved cooking stoves in rural households of Northern

408

India. Energy Sustainable Dev. 2014, 19, 1–6, DOI: 10.1016/j.esd.2014.01.007.

409

(18) Sambandam, S.; Balakrishnan, K.; Ghosh, S.; Sadasivam, A.; Madhav, S.; Ra-

410

masamy, R.; Samanta, M.; Mukhopadhyay, K.; Rehman, H.; Ramanathan, V. Can

411

Currently Available Advanced Combustion Biomass Cook-Stoves Provide Health Rel-

412

evant Exposure Reductions? Results from Initial Assessment of Select Commercial

413

Models in India. EcoHealth 2015, 12, 25–41, DOI: 10.1007/s10393-014-0976-1.

414

(19) Thurber, M. C.; Phadke, H.; Nagavarapu, S.; Shrimali, G.; Zerriffi, H. ‘Oorja’ in India:

415

Assessing a large-scale commercial distribution of advanced biomass stoves to house-

416

holds. Energy Sustainable Dev. 2014, 19, 138–150, DOI: 10.1016/j.esd.2014.01.002.

417

(20) Panwar, N. L.;

Rathore, N. S. Design and performance evaluation of a

418

5 kW producer gas stove. Biomass Bioenergy 2008, 32, 1349–1352, DOI:

419

10.1016/j.biombioe.2008.04.007.

420

(21) Mukunda, H. S.; Dasappa, S.; Paul, P. J.; Rajan, N. K. S.; Yagnaraman, M.; Ku-

421

mar, D. R.; Deogaonkar, M. Gasifier stoves — science, technology and field outreach.

422

Curr. Sci. 2010, 98, 627–638.

423

(22) Johnson, M. A.; Pilco, V.; Torres, R.; Joshi, S.; Shrestha, R. M.; Yagnaraman, M.;

424

Lam, N. L.; Doroski, B.; Mitchell, J.; Canuz, E.; Pennise, D. Impacts on household fuel

425

consumption from biomass stove programs in India, Nepal, and Peru. Energy Sustain-

426

able Dev. 2013, 17, 403–411, DOI: 10.1016/j.esd.2013.04.004.

427

(23) Carter, E. M.; Shan, M.; Yang, X.; Li, J.; Baumgartner, J. Pollutant Emissions and

428

Energy Efficiency of Chinese Gasifier Cooking Stoves and Implications for Future Inter-

429

vention Studies. Environ. Sci. Technol. 2014, 48, 6461–6467, DOI: 10.1021/es405723w.

430

(24) Huangfu, Y.; Li, H.; Chen, X.; Xue, C.; Chen, C.; Liu, G. Effects of moisture content in

20 ACS Paragon Plus Environment

Page 20 of 34

Page 21 of 34

Environmental Science & Technology

431

fuel on thermal performance and emission of biomass semi-gasified cookstove. Energy

432

Sustainable Dev. 2014, 21, 60–65, DOI: 10.1016/j.esd.2014.05.007.

433

(25) Belonio, A. T. Rice Husk Gas Stove Handbook ; Appropriate Technology Center, De-

434

partment of Agricultural Engineering and Environmental Management, College of Agri-

435

culture, Central Philippine University: Iloilo City, Philippines, 2005.

436

(26) Roth, C. Micro-gasification: cooking with gas from dry biomass, 2nd ed.; Deutsche

437

Gesellschaft f¨ ur Internationale Zusammenarbeit (GIZ) GmbH: Eschborn, Germany,

438

2014.

439

(27) Lobscheid, A. B.; Lodoysamba, S.; Maddalena, R. L.; Dale, L. L. A pilot study of

440

traditional and energy-efficient stove use in Ulaanbaatar, Mongolia. Boiling Point 2015,

441

66, 40–43.

442

(28) Reed, T. B.; Das, A. Handbook of Biomass Downdraft Gasifier Engine Systems;

443

SERI/SP-271-3022; Solar Energy Research Institute, U.S. Government Printing Of-

444

fice: Washington, DC, 1988; pp 21–47.

445

(29) Milne, T. A.; Evans, R. J.; Abatzoglou, N. Biomass Gasifier “Tars”: Their Nature,

446

Formation, and Conversion; NREL/TP-570-25357; National Renewable Energy Labo-

447

ratory, National Technical Information Service (NTIS): Springfield, VA, 1998.

448

(30) Tryner, J.; Willson, B. D.; Marchese, A. J. The effects of fuel type and stove design

449

on emissions and efficiency of natural-draft semi-gasifier biomass cookstoves. Energy

450

Sustainable Dev. 2014, 23, 99–109, DOI: 10.1016/j.esd.2014.07.009.

451

(31) Varunkumar, S.; Rajan, N. K. S.; Mukunda, H. S. Experimental and computational

452

studies on a gasifier based stove. Energy Convers. Manage. 2011, 53, 135–141, DOI:

453

10.1016/j.enconman.2011.08.022.

21 ACS Paragon Plus Environment

Environmental Science & Technology

454

(32) Varunkumar, S.; Rajan, N. K. S.; Mukunda, H. S. Single Particle and Packed Bed

455

Combustion in Modern Gasifier Stoves — Density Effects. Combust. Sci. Technol. 2011,

456

183, 1147–1163, DOI: 10.1080/00102202.2011.576658.

457

(33) Kirch, T.; Birzer, C. H.; Medwell, P. R.; Holden, L. The role of primary and secondary

458

air on wood combustion in cookstoves. Int. J. Sustainable Energy 2016, 1–10, DOI:

459

10.1080/14786451.2016.1166110.

460

(34) Kirch, T.; Medwell, P. R.; Birzer, C. H. Natural draft and forced primary air combustion

461

properties of a top-lit up-draft research furnace. Biomass Bioenergy 2016, 91, 108–115,

462

DOI: 10.1016/j.biombioe.2016.05.003.

463

(35) Maddalena, R. L.; Lunden, M. M.; Wilson, D. L.; Ceballos, C.; Kirchstetter, T. W.;

464

Slack, J. L.; Dale, L. L. Quantifying Space Heating Stove Emissions Related to Dif-

465

ferent Use Pattern in Mongolia. Energy and Environ. Res. 2014, 4, 147–157, DOI:

466

10.5539/eer.v4n3p147.

467

(36) James R., A. M.; Yuan, W.; Boyette, M. D. The Effect of Biomass Physical

468

Properties on Top-Lit Updraft Gasification of Woodchips. Energies 2016, 9, DOI:

469

10.3390/en9040283, 283.

470

(37) Saravanakumar, A.; Haridasan, T. M.; Reed, T. B.; Bai, R. K. Experimental investiga-

471

tion and modeling study of long stick wood gasification in a top lit updraft fixed bed

472

gasifier. Fuel 2007, 86, 2846–2856, DOI: 10.1016/j.fuel.2007.03.028.

473

(38) Birzer, C.; Medwell, P.; Wilkey, J.; West, T.; Higgins, M.; MacFarlane, G.; Read, M.

474

An analysis of combustion from a top-lit up-draft (TLUD) cookstove. J. Humanitarian

475

Eng. 2013, 2, 1–8.

476

(39) Reed, T. B.; Anselmo, E.; Kircher, K. Testing & Modeling the Wood-Gas Turbo Stove.

477

In Progress in Thermochemical Biomass Conversion; Bridgwater, A., Ed.; Blackwell

478

Science Ltd: Oxford, U.K., 2008; pp 693–704. 22 ACS Paragon Plus Environment

Page 22 of 34

Page 23 of 34

Environmental Science & Technology

479

(40) Report on tests on reverse downdraft stoves (REDS) for stove applications - 2003 to

480

2008. 2014; http://cgpl.iisc.ernet.in/site/Portals/0/Publications/Report2004-2008.pdf

481

(accessed Apr 16, 2016).

482

483

(41) The Water Boiling Test:

Version 4.2.3. 2014; http://cleancookstoves.org/binary-

data/DOCUMENT/file/000/000/399-1.pdf (accessed Jan 4, 2016).

484

(42) Kshirsagar, M. P.; Kalamkar, V. R. A comprehensive review on biomass cookstoves

485

and a systematic approach for modern cookstove design. Renewable Sustainable Energy

486

Rev. 2014, 30, 580–603, DOI: 10.1016/j.rser.2013.10.039.

487

488

(43) Anderson, P. S. Construction Plans for the “Champion-2008” TLUD Gasifier Cookstove. 2009; http://www.drtlud.com/?resource=prt09320 (accessed Jun 21, 2015).

489

(44) DeFoort, M.; L’Orange, C.; Kreutzer, C.; Lorenz, N.; Kamping, W.; Alders, J.

490

Stove Manufacturers Emissions and Performance Test Protocol (EPTP); Engines

491

and Energy Conversion Laboratory, Colorado State University:

492

CO, 2010;

493

1.pdf (accessed Jan 4, 2016).

Fort Collins,

http://cleancookstoves.org/binary-data/DOCUMENT/file/000/000/73-

494

(45) L’Orange, C.; DeFoort, M.; Willson, B. Influence of testing parameters on biomass stove

495

performance and development of an improved testing protocol. Energy Sustainable Dev.

496

2012, 16, 3–12, DOI: 10.1016/j.esd.2011.10.008.

497

(46) Wang, Y.; Sohn, M. D.; Wang, Y.; Lask, K. M.; Kirchstetter, T. W.; Gadgil, A. J.

498

How many replicate tests are needed to test cookstove performance and emissions?—

499

Three is not always adequate. Energy Sustainable Dev. 2014, 20, 21–29, DOI:

500

10.1016/j.esd.2014.02.002.

501

502

(47) Bryden, K. M.; Ragland, K. W.; Rutland, C. J. Modeling thermally thick pyrolysis of wood. Biomass Bioenergy 2002, 22, 41–53, DOI: 10.1016/S0961-9534(01)00060-5.

23 ACS Paragon Plus Environment

Environmental Science & Technology

503

(48) Yang, H.; Yan, R.; Chen, H.; Lee, D. H.; Zheng, C. Characteristics of hemicellulose, cel-

504

lulose and lignin pyrolysis. Fuel 2007, 86, 1781–1788, DOI: 10.1016/j.fuel.2006.12.013.

505

(49) Worasuwannarak, N.; Sonobe, T.; Tanthapanichakoon, W. Pyrolysis behaviors of rice

506

straw, rice husk, and corncob by TG-MS technique. J. Anal. Appl. Pyrolysis 2007, 78,

507

265–271, DOI: 10.1016/j.jaap.2006.08.002.

508

509

(50) Jensen, A.; Dam-Johansen, K. TG-FTIR Study of the Influence of Potassium Chloride on Wheat Straw Pyrolysis. Energy Fuels 1998, 12, 929–938, DOI: 10.1021/ef980008i.

510

(51) Neves, D.; Thunman, H.; Matos, A.; Tarelho, L.; G´omez-Barea, A. Characterization

511

and prediction of biomass pyrolysis products. Prog. Energy Combust. Sci. 2011, 37,

512

611–630, DOI: 10.1016/j.pecs.2011.01.001.

513

(52) Faravelli, T.; Frassoldati, A.; Barker Hemings, E.; Ranzi, E. Multistep Kinetic Model

514

of Biomass Pyrolysis. In Cleaner Combustion: Developing Detailed Chemical Kinetic

515

Models; Battin-Leclerc, F., Simmie, J. M., Blurock, E., Eds.; Green Energy and Tech-

516

nology; Springer-Verlag: London, 2013; Chapter 6, pp 111–139, DOI: 10.1007/978-1-

517

4471-5307-8 5.

518

(53) Liu, Q.; Zhong, Z.; Wang, S.; Luo, Z. Interactions of biomass components during

519

pyrolysis: A TG-FTIR study. J. Anal. Appl. Pyrolysis 2011, 90, 213–218, DOI:

520

10.1016/j.jaap.2010.12.009.

521

(54) Evans, R. J.; Milne, T. A. Chemistry of Tar Formation and Maturation in the Thermo-

522

chemical Conversion of Biomass. In Developments in Thermochemical Biomass Con-

523

version; Bridgwater, A. V., Boocock, D. G. B., Eds.; Springer Science+Business Media:

524

Dordrecht, Netherlands, 1997; Vol. 2; pp 803–816.

525

(55) Gustafsson, E.; Lin, L.; Strand, M. Characterization of particulate matter in the hot

526

product gas from atmospheric fluidized bed biomass gasifiers. Biomass Bioenergy 2011,

527

35, S71–S78, DOI: 10.1016/j.biombioe.2011.02.053. 24 ACS Paragon Plus Environment

Page 24 of 34

Page 25 of 34

Environmental Science & Technology

528

(56) Morgalla, M.; Lin, L.; Seemann, M.; Strand, M. Characterization of particulate matter

529

formed during wood pellet gasification in an indirect bubbling fluidized bed gasifier

530

using aerosol measurement techniques. Fuel Process. Technol. 2015, 138, 578–587, DOI:

531

10.1016/j.fuproc.2015.06.041.

25 ACS Paragon Plus Environment

Environmental Science & Technology

532

Graphical TOC Entry

533

26 ACS Paragon Plus Environment

Page 26 of 34

PageEnvironmental 27 of 34 Science & Technology

ACS Paragon Plus Environment

Environmental Science & Technology Potzcontainingz2.5Lzwater Potzsupportzarms Secondaryzairzinletzring Secondaryzairzheater Thermocouples

Digitalzbalance

Gaszsamplingzprobe

Thermocoupleszfor trackingzprimary combustionzzone

Secondaryzairzflowzcontrol Primaryzairzflowzcontrol Temperaturezcontrollers forzsecondaryzair

Primaryzairzinletzring

ACS Paragon Plus Environment

Primaryzairzinlet

Figure 1

Page 28 of 34

Page 29 of 34

Environmental Science & Technology

ACS Paragon Plus Environment

Figure 2

Environmental Science & Technology

ACS Paragon Plus Environment

Figure 3

Page 30 of 34

a. . -di -dia

Phase)1 Phase)2

mm

16

Tier)1

12

Tier)2 Tier)3 Tier)4

8 4

4-

Tier01

16 12

Tier02 Tier03 Tier04

8 4

ACS Paragon Plus Environment 2

3

4

5

Secondary)to)primary)air)flow)ratio

0

Figure 4

.

dia

Phase01 Phase02

20 Carbon0monoxide0(g/MJ d)

20 Carbon)monoxide)(g/MJ d)

a.

24

24

0

-di

m m Environmental Science & Technology -m m 6-m 10 8-m

Page 31 of 34

1

2

3

4

Secondary0air0velocity0(m/s)

5

24

Environmental Science & Technology Phasex1 20

Phase 2

103

16

Tierx1

12

PM2.5 )mg/MJ d)

Carbonxmonoxidex6g/MJ dP

Page Phase 32 1 of 34

Phasex2

Tierx2 Tierx3 Tierx4

8

TierP1 Tier 2 Tier 3

102 Tier 4

4 101 0

lets hips hips hips xpel Paragon obxc odxc xwoodxc xpineACS c o x n w le xfir tusx Cor gepo glas alyp Lod Dou Euc

Fuelxtype

ps

ips

ps

ets

i ch chi pell b ch Plus Environment ood ood n co pine fir w Cor tus w ole

Figure 5

alyp

Euc

glas

Dou

gep

Lod

Fuel type

Environmental Science & Technology Phase 1

30

80 25

60

20

50 15

40 30

10

20 5

10 CO

CH 4

O2/Ar

CO 2

N2

Phase 2

30

90 80

25

70 60

20

50 15

40 30

10

Mole % (N 2)

Mole % (H 2, CO, CH 4, O 2/Ar, CO 2)

Mole % (N 2)

70

H2

20 5

10 H2

Mole % (H 2, CO, CH 4, O 2/Ar, CO 2)

90

CO

CH 4

20

CO 2

N2

Phase 3

30 25

O2/Ar

90 80

Corn cob chips Eucalyptus chips Douglas fir chips Lodgepole pine pellets

70 60 50

15

40 30

10

20 5

10 H2

CO

CH 4

O2/Ar

ACS Paragon Plus Environment

Figure 6

CO 2

N2

Mole % (N 2)

Mole % (H 2, CO, CH 4, O 2/Ar, CO 2)

Page 33 of 34

End fuel batch 1

Temperature (°C)

1000

End fuel batch 2

Environmental Science & Technology

Page 34 of 34 Gas Top Middle Bottom

800 600 400 200

ACS Paragon Plus Environment 0 0

500

Figure 7

1000

1500

2000

2500

Time since start of test (s)

3000

3500

4000