Life-Cycle Assessment of Biodiesel Produced from Grease Trap Waste

Subscriber access provided by SAN DIEGO STATE UNIV

Article

Life Cycle Assessment of Biodiesel Produced from Grease Trap Waste Megan E. Hums, Richard Cairncross, and Sabrina Spatari Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b02667 • Publication Date (Web): 26 Jan 2016 Downloaded from http://pubs.acs.org on February 10, 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 27

Environmental Science & Technology

ACS Paragon Plus Environment


1

Life Cycle Assessment of Biodiesel Produced from

2

Grease Trap Waste

3

Megan E. Hums1, Richard A. Cairncross1*, & Sabrina Spatari2 1

4 2

5

Page 2 of 27

Chemical and Biological Chemical Engineering, Drexel University

Civil, Architectural, and Environmental Engineering, Drexel University

6

KEYWORDS: Grease Trap Waste, Fats, Oils, and Greases (FOG), Biodiesel, attributional and

7

consequential LCA, Renewable Energy, GHG emissions

8

ABSTRACT: Grease trap waste (GTW) is a low-quality waste material with variable lipid

9

content that is an untapped resource for producing biodiesel. Compared to conventional biodiesel

10

feedstocks, GTW requires different and additional processing steps for biodiesel production due

11

to its heterogeneous composition, high acidity, and high sulfur content. Life cycle assessment

12

(LCA) is used to quantify greenhouse gas emissions, fossil energy demand, and criteria air

13

pollutant emissions for the GTW-biodiesel process where the sensitivity to lipid concentration in

14

GTW is analyzed using Monte Carlo simulation. The life cycle environmental performance of

15

GTW-biodiesel is compared to that of current GTW disposal, the soybean-biodiesel process, and

16

low sulfur diesel (LSD). The disposal of the water and solid wastes produced from separating

17

lipids from GTW has a high contribution to the environmental impacts; however, the impacts of

18

these processed wastes are part of the current disposal practice for GTW and could be excluded

19

with consequential LCA system boundaries. At lipid concentrations greater than 10%, most of

20

the environmental metrics studied are lower than those of LSD and comparable to soybean-

21

biodiesel.

22

INTRODUCTION


1

Page 3 of 27


23

Biodiesel is a renewable fuel that can be produced from a variety of feedstocks including

24

vegetable oils, animal fats, waste greases, and algal oil. Commercial biodiesel is produced

25

mostly from refined plant oils1. However, vegetable oils are an expensive feedstock and account

26

for 70-88% of biodiesel production cost2. The use of waste greases, such as brown grease

27

extracted from grease trap waste (GTW), as biodiesel feedstock offers a low-cost alternative to

28

refined vegetable oils3. Additionally, recovery of brown grease into a value-added product

29

represents an opportunity to “recapture” and recycle waste streams from the food industry.

30

GTW is an effluent from commercial kitchen wastewater that is collected in grease

31

interceptors to avoid sewer blockages4. GTW is a combination of fats, oils, and greases (FOG),

32

water, and solids with highly variable composition3, 5. FOG, or brown grease, is the lipid portion

33

of GTW that can be extracted and converted into biodiesel. The quantity of lipids in GTW varies

34

depending on the source and GTW management practice3, 6, and GTW generation ranges from

35

1,406-11,000 kg/yr/restaurant with a range of 0.1-40% lipid content3, 5, 7. In the United States, an

36

estimated 1.8 billion kg/yr of lipids could be recovered from GTW4 which could produce about

37

1.3 billion kg of biodiesel/yr assuming conversion data from the process model used in this

38

paper. The variability of the generated GTW amount and lipid content make GTW a complex

39

feedstock for biodiesel production as compared to conventional biodiesel feedstocks.

40

GTW disposal methods vary based on location and municipal regulations3. Common practices

41

are to dispose of GTW at a landfill, incinerator, or anaerobic digester8. The degradation of

42

organic material in a landfill emits methane gas (biogas) which is a more potent greenhouse gas

43

than carbon dioxide9, 10 and is also accounted with fossil GHG emissions unlike biogenic CO2;

44

the collection and use of this biogas can benefit waste disposal facilities. Landfill gas collection

45

and anaerobic digestion offer ways to reduce the emission of the methane gas by flaring or co-


2


Page 4 of 27

46

generating heat and electricity; converting the methane to biogenic carbon dioxide thus considers

47

the carbon to be short-lived unlike fossil CO2 a emission11. Recent laboratory and LCA research

48

has focused on anaerobic co-digestion of GTW with residual biosolids such as sewage sludge8, 12-

49

14

50

wastewater has been reported to produce more usable energy than anaerobic digestion alone15.

51

Also, the variability of the feedstock could affect the microbiological activity in the anaerobic

52

digester which is sensitive to changes in feedstock composition particularly long-chain fatty

53

acids15. Separating lipids from GTW for biodiesel production has several potential benefits for

54

waste management facilities including reducing the volume of GTW that is processed for

55

disposal and replacing petroleum diesel combustion by a renewable fuel.

; however, extracting the lipids for biodiesel production and anaerobic digestion of the

56

Life Cycle Assessment (LCA) has been used to estimate the life cycle impact assessment

57

(LCIA) metrics for biodiesel produced from a variety of feedstocks. Dufour and Iribarren

58

performed LCA on biodiesel production from inedible and low-quality biodiesel feedstocks such

59

as sewage sludge and used vegetable oil. They showed that the production of biodiesel from

60

used vegetable oil and from sewage sludge reduced greenhouse gas (GHG) emissions by 79.7%

61

and 24.5%, respectively as compared to low sulfur diesel16. The sewage sludge GHG reduction

62

is smaller because 10,000 kg of sewage sludge was needed to be processed to produce 1000 kg

63

of biodiesel whereas only 1205 kg of waste vegetable oils is needed for 1000 kg of biodiesel16.

64

The variability of lipid concentration in sewage sludge is similar to that of GTW; the lower the

65

lipid content, the greater volume of starting waste material is needed to produce the same amount

66

of fuel.

67

greenhouse gas emissions of GTW-biodiesel processing and performed Monte Carlo simulation

68

for a sensitivity analysis. They found that the GTW-biodiesel process had potentially lower

Tu and McDonnell recently published an analysis of the life cycle energy and


3

Page 5 of 27


69

greenhouse gas emissions and energy usage than conventional fuels primarily when anaerobic

70

digestion was used for waste disposal17. However, this analysis relies on literature for the GTW-

71

biodiesel production and excludes the biodiesel purification step necessary for compliance with

72

ASTM-grade biodiesel.

73

Life cycle assessment (LCA) is a systematic framework for examining the implications of

74

products, processes, and activities, using specific metrics through life cycle impact assessment

75

that approximate environmental damages18. This research focuses on the production of biodiesel

76

from GTW utilizing a process model created from laboratory data from our recent reactor and

77

purification research. The LCA includes the entire life cycle of the fuel from the collection of

78

the GTW feedstock to the combustion of the biodiesel in a vehicle. This paper also includes a

79

parametric study on lipid content of the GTW to analyze the 100-year global warming potential,

80

fossil cumulative energy demand, and selected air pollution emissions associated with the

81

combustion of the fuel.

82

RESEARCH SCOPE AND METHODS

83

Goal and Scope

84

LCA was used to evaluate the energy and selected LCIA metrics of producing and combusting

85

1 MJ of biodiesel (the functional unit) from GTW. The LCIA metrics analyzed were midpoint

86

life cycle impact assessment metrics of 100-year global warming potential (GWP100)11, fossil

87

cumulative energy demand (CEDfossil)19, and the criteria air pollutant emissions20:

88

monoxide (CO), mono-nitrogen oxides (NOx), sulfur oxides (SOx), and particulate matter (PM).

carbon

89

LCA can be performed using an attributional or a consequential framework. Attributional

90

LCA is used to determine the total emissions from the process21, 22 whereas consequential LCA is

91

used for analyzing the change in emissions which is due to a change in process for handling


4


Page 6 of 27

92

GTW21, 22. In this paper, an attributional LCA of the GTW-biodiesel process was first used to

93

determine the LCIA metrics of the entire biodiesel production process including the

94

transportation of GTW from restaurants to the grease hauler’s aggregation location (transfer

95

station), separation of GTW lipids, disposal of GTW wastewater and waste solids, conversion of

96

lipids into fuel, and operation of fuel in a vehicle. Despite lipid separation, there is still a large

97

volume of wastewater and solids that need to be disposed of, which leads to high greenhouse gas

98

emission17; therefore, a consequential LCA was also used to examine the GWP100 of the GTW-

99

biodiesel process and to compare it with current disposal of the same amount of GTW (lipids

100

not separated for biodiesel production). GTW disposal at a landfill was chosen; other waste

101

disposal options such as incineration and anaerobic digestion are outside the scope of this paper

102

and will be addressed in a future study. Lastly, the sensitivity of LCA impacts due to the

103

variability of lipid content in the GTW-biodiesel process was evaluated using a Monte Carlo

104

simulation.

105

This study utilized various tools for the LCA to determine the LCIA metrics of the fuel

106

processes. SimaPro823 and the EcoInvent database20 were used to analyze the impacts for the

107

GTW-biodiesel production except for the natural gas used for steam production. GREET-201424

108

was used to determine the life cycle impacts of natural gas used for steam production, the

109

soybean-biodiesel process, the LSD process, and vehicle operation because GREET is specific

110

towards the production and combustion of these fuels in the United States. Oracle Crystal ball25

111

was used to run a Monte Carlo simulation for the sensitivity of lipid content on the LCIA

112

metrics.

113 114


5

Page 7 of 27

115


System Boundary

116

Three fuel production processes and one comparative process were studied for the production

117

of biodiesel from GTW (Figure 1), current GTW disposal, biodiesel from soybeans, and LSD.

118

The system boundary of the life cycle model of the fuel production process included three stages:

119

1) Pre-treatment, 2) fuel production, and 3) vehicle operation. A full system boundary figure is

120

shown in the supporting information (SI).

121

GTW-Biodiesel Process Description

122

A process model was created using laboratory data26, 27 and unit operation material balances

123

from design projects28-30 to estimate energy and material requirements. The process model

124

simulated a GTW-biodiesel plant with a capacity for producing 840 L/day of biodiesel and

125

analyzed the material and energy requirements for the GTW-biodiesel process. The model

126

included the extraction of grease lipids from GTW, conversion of lipids into biodiesel with

127

methanol recycling, washing of crude biodiesel, and purification using vacuum distillation (see

128

SI for more details).

129

GTW Transportation to Transfer Station

130

The GTW was delivered to the transfer station in a 16 metric-ton truck with a round-trip

131

transportation distance of 286 km using data collected from routes traveled by a grease hauler

132

collecting GTW31.

133

Pre-treatment

134

The GTW pre-treatment stage included two sub stages: oil extraction (separation of lipids) and

135

waste management (WM).

136

Oil Extraction


6


Page 8 of 27

137

The oil extraction separated the lipids from the remaining GTW (floating solids, wastewater,

138

and sediments) with heating to liquefy the lipids for faster separation. The volumetric balance of

139

the GTW lipids was varied from 1-40%, wastewater 25-64%, and the floating solids and

140

sediments were kept constant at 10% and 25%, respectively.

141

Pre-treatment Waste Management

142

The waste management included the transportation and treatment of GTW wastewater (GTW-

143

WW) at a wastewater treatment plant as well as the transportation and disposal of GTW waste

144

solids (GTW-WS) at a landfill. The transportation distance of both materials was 50 km in a 16

145

metric-ton truck.

146

The GHG emissions associated with the landfill was estimated using landfill data of food waste

147

from the EPA. Food waste data were used for the GTW-WS because the GTW comes from

148

kitchen waste32. The landfill gas emitted was analyzed using two methods: A) Flaring and B)

149

Co-generation for electricity and heat production. The co-generation products were treated as

150

avoided emissions of electricity and natural gas.

151 152

Fuel Production The fuel production stage contained four sub-stages:

conversion, purification, waste

153

management, and service station. A co-product “bio-bunker” was produced which is similar to a

154

heavy fuel oil.

155

Conversion

156

The lipids were reacted using a bubble column reactor that has been developed by researchers

157

at Drexel University26 and was inspired by experiments done by Kocsisová et al.33. The lipids

158

contained 97% free fatty acids which was the typical amount found at the time of creating the

159

process model. Oleic acid was used to represent the free fatty acids because it is the most


7

Page 9 of 27


160

prevalent fatty acid in waste oils8, 34. Transesterification was not included due to the high level

161

of free fatty acids. The esterification was performed at atmospheric pressure and 120 ⁰C for 2

162

hrs utilizing 0.5% w/w sulfuric acid as catalyst and 4.5 molar ratio of methanol to lipids.

163

Methanol was recovered and recycled to the reactor using a partial condenser and a distillation

164

column.

165

Purification

166

The crude biodiesel was first neutralized and water washed and then distilled in a short-path

167

evaporator for further purification and sulfur reduction.

GTW-biodiesel has a high sulfur

168

concentration because the sulfur content of GTW lipids is 200 PPM on average. Vacuum

169

distillation is necessary to reduce the sulfur concentration35 to meet the 15 PPM sulfur

170

specifications for on-road biodiesel36. The evaporator was run at 0.1 bar and 260 °C. This

171

condition is conservative compared to more recent experimental data (performed after this

172

analysis) operating at 1 mbar and 190 °C with optimization ongoing. Therefore, the energy

173

demand for the evaporator in this paper is likely higher than necessary. This stage is essential for

174

sulfur reduction and was not considered by Tu and McDonnell17.

175

The high-boiling point material remaining after distillation was described as “bio-bunker.”

176

Because allocation methods can change the life cycle impact of a process37 and the

177

mass/volume/energy of bio-bunker compared to biodiesel is small, no allocation method was

178

applied to the LCA. Thus, the GTW-biodiesel LCIA metrics estimated in this analysis are

179

conservative. Potentially, future work could assign an allocation to the bio-bunker, or it could be

180

evaluated as a substitute for industrial use of natural gas or of heavy fuel oil.

181

Fuel Conversion Waste Management


8


182

The wastewater produced from the methanol recovery and water washing was transported 50

183

km by a 7.5 metric-ton truck for treatment at a wastewater treatment plant.

184

Service Station

185

The purified biodiesel was transported 100 km by a 7.5 metric-ton truck to the service station.

186

Service station operations were evaluated according to SimaPro8.

187

Vehicle Operation

188

Page 10 of 27

The use and end-of-life of biodiesel was combustion in a vehicle. The CO2 credit for biodiesel

189

was represented in the fuel’s combustion.

The CO2 produced from biogenic sources was

190

considered zero because of the recent sequestration of carbon from the atmosphere as opposed to

191

that of petroleum fuels38. Biodiesel from oleic acid is treated as methyl oleate, which consists of

192

19 carbons: 18 renewable, biogenic carbon atoms from the lipids and one non-renewable, non-

193

biogenic carbon from the methanol. Therefore, the biodiesel combustion emissions from non-

194

renewable sources are 1/19th of the total CO2 combustion emissions.

195

The vehicle operation of the GTW and soybean biodiesel were evaluated in the same way

196

using GREET-201424 vehicle combustion data combined with biodiesel emission data reported

197

in a review by the EPA. The EPA performed a study and fit an emissions curve to data from a

198

compilation of literature reports on the engine combustion of biodiesel for CO, NOx, and PM

199

which was then represented as a percent change from LSD39. The sulfur content of the fuel was

200

used to determine SOx emissions during vehicle operation.

201

Comparative Scenarios

202

The current GTW disposal represented the transportation of GTW to the transfer station,

203

gravity settling of the GTW where GTW-WS and lipids were dewatered and sent to the landfill

204

while the GTW-WW was sent to wastewater treatment plants. The LCIA metrics associated


9

Page 11 of 27


205

with the transportation, wastewater treatment, and landfill were treated the similarly to the GTW-

206

biodiesel process description. It was assumed that the emissions associated with the GTW

207

gravity settling are negligible.

208

Soybean-biodiesel and low sulfur diesel life cycle assessments were reported using the

209

GREET-2014 fuel processes24. Details on these processes can be found in the SI.

210

Life Cycle Inventory (LCI)

211

A process-based LCI model was developed from sequential material and energy balances

212

following ISO methods18. Data were used to create an inventory of the materials and utilities

213

required for producing 1 MJ of biodiesel from GTW, shown in the SI.

214

Uncertainty/Model Fitting

215

A model was developed to analyze and test trends in the LCIA metrics of producing biodiesel

216

from GTW with varying lipid contents shown in equation 1. The derivation of the equation is

217

found in the SI.

218 219

= ∆ + ∆ + ∆ (1)

Where,

220

Îi = LCIA metric intensity of process stage i per unit mass of input (PT = pre-treatment, FP = fuel

221

production, and VO = vehicle operation: combustion emissions)

222

∆ĤC = heat of combustion of biodiesel (lower heating value, MJ/kg)

223

Φ = yield of fuel production process

224

x = lipid content

225

EBiodiesel = energy content of biodiesel produced


10


Page 12 of 27

226

This equation predicts that the total LCIA metrics are linearly proportional to the reciprocal of

227

the GTW lipid content, 1/x. Linear regression of Equation 1 to the LCIA metrics versus 1/x was

228

used to estimate the slope and intercept.

229

Sensitivity to GTW Composition and Monte Carlo Simulation

230

Monte Carlo simulation was used to test the effects of lipid variability on GWP100 and other

231

LCIA metric metrics. A lognormal distribution function was fit to lipid percentages that were

232

found during a longitudinal study of GTW composition31. LCIA metrics were found by utilizing

233

the fitted equation described in the previous section where lipid content, x, was varied using the

234

distribution curve. Oracle Crystal Ball was used to determine the distribution curve of the lipids

235

running Monte Carlo in 5000 trials.

236

The GTW-biodiesel process is also sensitive to the FFA content of the GTW lipids. The

237

process model assumed a 97% FFA content representative of laboratory data collected at the

238

time of its development. A scenario of using a lower FFA content was examined and is shown in

239

the SI. Tu and McDonnell also performed a sensitivity analysis on lipid content, FFA content,

240

and other anaerobic digestion conditions17.

241

emissions was due to changes in lipid concentration; therefore, the Monte Carlo analysis in this

242

paper is only applied to lipid content sensitivity.

243

RESULTS & DISCUSSION

They found the greatest variability for GHG

244

The parametric life cycle results represent the LCIA metrics for 1 MJ of biodiesel produced

245

from GTW with a lipid content of 2-40%. The GWP100, CEDfossil, and selected air emissions for

246

the GTW-biodiesel process were studied and compared to soybean-biodiesel and LSD and are

247

shown in detail in the SI.

248

100-year Global Warming Potential


11

Page 13 of 27

249


Attributional LCA Approach for GTW-Biodiesel

250

The life cycle GWP100 of GTW-biodiesel by process stage, with lipid content ranging from 2-

251

40% is evaluated in two scenarios for treatment of the landfill gases from disposal of the GTW-

252

WS, flaring (Figure 2A) and co-generation of heat and electricity (Figure 2B). The impact due

253

to the waste management of the pre-treatment is presented separately from the rest of the pre-

254

treatment (steam production and electricity) because of its large contribution to the emissions.

255

For lipid contents less than 10%, the GWP100 of the GTW-biodiesel process is dominated by the

256

emissions from delivery of the GTW to the transfer station and the pre-treatment waste

257

management (pre-treatment WM; the transportation and treatment of GTW-WW and GTW-WS).

258

As lipid content decreases, there is an increased amount of GTW transported to the transfer

259

station and an increase in waste sent for disposal which contributes to the higher GWP100. The

260

impact from the pre-treatment without WM and the fuel process stages are all dominated by

261

combustion of natural gas for steam production; as the lipid content increases, there is less

262

natural gas consumed to produce steam to separate the lipids, which leads to a lower GWP100.

263

The difference in the total GWP100 between the flaring and co-generation scenarios is smaller

264

at higher lipid contents. The flaring scenario has 11% higher total GWP100 compared to co-

265

generation (difference of 3 g-CO2-eq/MJ-fuel) at 40% lipid content. At 2% lipid content, the

266

flaring scenario has 28% higher total GWP100 (difference of 64 g-CO2-eq/MJ-fuel). The benefits

267

of co-generation are the avoided GWP100 from the electricity and natural gas (shown as a credit -

268

negative contribution to GWP100 in Figure 2B) which causes lower GWP100 for the co-generation

269

scenario.

270

The curves in Figure 2 are a correlation based on a linear regression of Equation 1 to the

271

GWP100 results (equation shown in each graph). In the correlation, the constant (~18 g-CO2-


12


Page 14 of 27

272

eq/MJ-fuel) represents the GWP100 of the fuel production and vehicle operation which is

273

independent of lipid content because the feedstock volume is adjusted to have the lipid volume

274

needed to produce 1 MJ of biodiesel. When the lipid content is below 5.5% and 3.7% in the

275

flaring and co-generation scenarios, respectively, the net GWP100 for GTW is higher than that for

276

LSD (93-gCO2-eq/MJ-fuel). The flaring and co-generation scenarios results show that for lipid

277

contents above 6%, without accounting for fuel offsets, the net GWP100 of the proposed GTW-

278

biodiesel process are lower than the LSD.

279

Consequential LCA Approach for GTW-Biodiesel Process v. Current GTW Disposal

280

When accounting for the avoided emissions for replacing current GTW disposal with the

281

proposed GTW-biodiesel, the consequence or change leads to reduced LCIA metrics for all lipid

282

contents of GTW studied. In Figure 3, the GWP100 associated with current GTW disposal and

283

offset LSD consumption are treated as avoided emissions. The avoided emissions are treated as

284

a credit because both offset LSD and current GTW disposal are not needed when the GTW-

285

biodiesel is implemented. In Figure 3B, the co-generated heat and electricity associated with

286

biodiesel production (teal) is treated as a credit; however, the co-generated heat and electricity

287

associated with the current GTW disposal (teal stripes) is treated as a penalty – the difference

288

between these two contributions in Figure 3B is the net change in avoided co-generation utilities

289

when converting from current GTW disposal to the proposed GTW-biodiesel process. The net

290

difference between the co-generated avoided utility emissions in the current GTW disposal and

291

GTW-biodiesel process avoided utilities is equal to the co-generation impacts associated with the

292

lipids (5.5 g-CO2eq/MJ-fuel avoided utility). Tables for the current GTW disposal and the

293

biodiesel production are in Table SI-S6 and Table SI-S7 in the SI.


13

Page 15 of 27


294

In both waste solid treatment scenarios, the total GWP100 associated with the GTW-biodiesel

295

process is always higher than the current GTW disposal because of additional GWP100 associated

296

with the biodiesel conversion and purification processes. However, the net GWP100 is negative

297

for both scenarios and all lipid contents when accounting for avoided emissions of LSD and

298

current GTW disposal.

299

In Figure 3, at low lipid contents, the largest impacts are due to waste management: positive

300

contribution of pre-treatment waste management in the GTW-biodiesel process (orange with

301

blue dots) and negative contribution of avoided waste management of the current GTW disposal

302

(orange hashed). The difference in impacts between current GTW-disposal (avoided) and pre-

303

treatment waste management is due to the lipids that are removed from GTW during pre-

304

treatment and is independent of lipid content for the chosen functional unit (1 MJ biodiesel

305

corresponds to a constant amount of lipids). The GWP100 difference between the current GTW

306

disposal and the waste disposal from biodiesel production is 7.7 and 4.5 g-CO2-eq/MJ-fuel for

307

flaring and co-generation, respectively. The co-generation has a smaller difference because of

308

the reduction in avoided electricity and heat when the lipids are removed from the waste

309

treatment (-5.5 g CO2-eq/MJ-fuel).

310

The removal of the lipids and production of biodiesel results in avoiding LSD (93 g-CO2-

311

eq/MJ-fuel). In the attributional LCA analysis of Figure 2, the GWP100 for the GTW-biodiesel

312

process is lower than GWP100 of LSD for lipid contents above 5%.

313

consequential analysis of Figure 3, which represents replacing current GTW disposal and LSD

314

use with the GTW-biodiesel process, the net emissions (black bar) are negative for all lipid

315

contents studied.

However in the


14


Page 16 of 27

316

The GTW-biodiesel process has a better net GWP100 making fuel production more favorable

317

than current GTW disposal. A new system boundary is proposed to omit the impacts associated

318

with the current GTW disposal. For the remaining LCIA metrics we include the “without waste

319

management” scenario (w/o WM) where the pre-treatment WM and the transportation of GTW

320

to the transfer station are omitted in the LCA of the GTW-biodiesel process which better

321

represents the effect of implementing a GTW-biodiesel process in the current waste management

322

system.

323

Sensitivity to GTW Composition and Monte Carlo Simulation

324

The yield of biodiesel produced from GTW is sensitive to the composition of GTW – both the

325

lipid content of GTW and the percent FFA of the lipids extracted from GTW. The variability of

326

percent FFA in lipids extracted from GTW is relatively small (70% to 98%) compared to the

327

variability in lipid content (0.15-65%); so the variability in percent FFA has a smaller effect on

328

the LCA impacts17. The SI includes a table showing the sensitivity of LCA impacts to FFA

329

content; decreasing FFA from 97% to 80% causes a 10%-20% increase in GWP100 for the

330

consequential LCA depending upon the lipid content. Future research will include sensitivity

331

analysis for FFA content and incorporate a two-step biodiesel conversion process for both FFA

332

and acyl glycerides.

333

The Monte Carlo analysis results presented in Figure 4 correspond to three scenarios: 1)

334

attributional LCA with landfill gas flaring, 2) attributional LCA with landfill gas co-generation,

335

and 3) LCA omitting impacts that are from GTW waste management (impacts associated with

336

GTW transportation to the transfer station and waste solid and wastewater disposal). To evaluate

337

the sensitivity that lipid content has on the LCIA metrics, a Monte Carlo simulation was

338

performed using two experimental lognormal distributions of lipid contents for GTW: the results


15

Page 17 of 27


339

in Figure 4A and 4B correspond to a lipid content distribution for raw GTW (named Raw GTW)

340

with median and mean lipid contents of 2% and 5%, respectively and the results in Figure 4C

341

and 4D correspond to a lipid content distribution for partially-dewatered GTW (named

342

Dewatered GTW) with median and mean lipid contents of 27% and 29%, respectively31. Both

343

lipid distributions were truncated between 0.15%-65%. Box plots of the normalized results are

344

shown in Figure 4; each LCIA metric was normalized by dividing the impact by that of LSD. In

345

Figure 4, the LCIA metric of LSD equals one (red lines), and the green lines represent the

346

impacts of soybean biodiesel.

347

statistical results (mean, median, standard deviation, 10 and 90 percentiles, etc. are shown in the

348

SI).

The lipid content distribution functions and a table of the

349

For all but one of the LCIA metrics in Figure 4, the flaring scenario has the highest median

350

value and largest range, and the without waste management scenario has the lowest median value

351

and smallest range. The magnitudes of the LCIA metrics for the raw GTW distribution (Figure

352

4A and 4B) are significantly larger than for the dewatered GTW distribution (Figure 4B and 4C)

353

because raw GTW has a much larger volume and produces more wastes that are landfilled than

354

dewatered GTW. Compared to flaring, the co-generation scenario reduces in median impact by

355

1% for CO, 25% for GWP100 , 25% for NOx, 42% for CEDFossil, and 66% for PM.

356

The SOx emissions in Figure 4 are negative for the landfilling with co-generation scenario due

357

to offset electricity from co-generation, and raw GTW has more negative SOx emissions than

358

dewatered GTW.

359

sulfuric acid used in the phosphorous fertilizer. In this analysis the electricity is generated

360

primarily from coal (46%) and other fossil sources (18%) which contribute to high SOx

361

emissions. GTW with a lower lipid content produces a higher volume of waste solids that is sent

Also, the soybean-biodiesel SOx emissions are higher than LSD due to


16


Page 18 of 27

362

to the landfill, degrades to methane, generates electricity that offset grid demand and reduces

363

SOx emissions. This is a case of a co-product (electricity) producing a benefit in one LCIA

364

metric that favors the co-product over the product. A similar model sensitivity was observed for

365

ethanol produced from corn stover, where the authors found improved LCIA results at lower

366

ethanol yields40. However, as noted by the authors, an ethanol biorefinery would never operate

367

at a lower yield in order to optimize offsets from its co-products40. Similarly, a GTW-biodiesel

368

conversion process would be optimized for higher biodiesel yield for economic benefits.

369

Adding a dewatering step dramatically reduces the LCIA metrics because the higher lipid

370

content of dewatered GTW (shown in Figure 4C and 4D) requires less pre-treatment process

371

energy and contains a lower volume of solid wastes for disposal.

372

dewatering reduces the median value of the LCIA metrics by about 80%. In co-generation, the

373

median values of all LCIA metrics (with the exception of SOx) are reduced 55-80% between raw

374

GTW and dewatered GTW.

375

increase in lipid content in the dewatered GTW distribution, there is less solid waste sent to the

376

landfill resulting in lower electricity credits than in the raw GTW scenario. In general, the

377

environmental burden of producing GTW-biodiesel is highly dependent on the lipid content of

378

the GTW; if the lipid content is below 10%, LCIA metrics increase hyperbolically (specifically

379

shown for GWP100 in Figure 2 and in model equation 1). This result suggests that a grease

380

dewatering process should be employed to concentrate the lipids prior to heating.

For flaring scenarios,

Alternatively, SOx shows a slight increase because with the

381

The LCIA metrics for the GTW-biodiesel process without GTW waste management are similar

382

to that of soybean-biodiesel. In the all disposal scenarios with dewatered GTW, the GWP100 and

383

CEDFossil results for GTW-biodiesel are also comparable to published LCA results on biodiesel

384

produced from other waste materials and inedible feedstocks such as jatropha. In this paper, the


17

Page 19 of 27


385

average GWP100 of biodiesel produced from dewatered GTW for flare, co-gen, and w/o WM

386

scenarios are 37, 31, and 22 g-CO2-eq/MJ-fuel. Published GWP100 for waste cooking oil,

387

sewage sludge, and jatropha biodiesels are approximately 18, 20 and 35 g-CO2-eq/MJ-fuel16,41.

388

In this paper, the average CEDFossil for the GTW flare, co-gen, and w/o WM are 0.41, 0.34, and

389

0.33 MJ/MJ-fuel.

390

biodiesels are approximately 0.2, 0.8 and 0.6 MJ/MJ-fuel, respectively16,41. The GWP100 and

391

CEDFossil for biodiesel produced from GTW were determined by Tu and McDonnell. GWP100 for

392

scenarios with and without anaerobic digestion are approximately 12 and 40 g-CO2-eq/MJ-fuel,

393

respectively and the CEDFossil are approximately 0.3 and 0.6 MJ/MJ-fuel, respectively17. More

394

information on the comparison to Tu and McDonnell is in the SI.

Published CEDFossil for waste cooking oil, sewage sludge, and jatropha

395

There is a trade-off between using vegetable oil such as soybean oil versus GTW lipids as

396

feedstocks for biodiesel production. For vegetable oils, the LCIA metrics are largely due to the

397

pre-treatment (e.g. soybean grain to soybean oil) and the conversion of vegetable oils to biodiesel

398

is a low energy, low material process. In contrast, the LCIA metrics from the GTW-biodiesel

399

process are primarily due to the fuel production because of the high energy required for reaction

400

and purification into ASTM-grade biodiesel or the waste treatment (depending on the system

401

boundary).

402

Because GTW-biodiesel is produced from a waste source, it is important not only that GTW-

403

biodiesel have environmental impacts comparable to or better than soybean-biodiesel and LSD,

404

but also that it should be a better alternative to current GTW disposal techniques.

405

consequential LCA boundary demonstrates that while producing GTW-biodiesel increases GHG

406

emissions in the waste management system by 13-43 g-CO2-eq/MJ-fuel, it reduces GHG

The


18


Page 20 of 27

407

emissions by 20-75% (Figure 3) when displacing LSD in the fully expanded system with flaring

408

or co-generating electricity for every MJ of biodiesel produced.

409

ASSOCIATED CONTENT

410

Supporting Information

411

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

412

11 figures, 37 tables, and 11 equations. The content of this document includes: full system

413

boundaries, process model description and full life cycle inventory, in-depth landfill gas analysis,

414

all selected LCIA metric results for full range of lipid contents, effect of FFA content in GTW

415

lipids, and statistical results for Monte Carlo simulations.

416

Corresponding Author

417

*E-mail: [email protected], Phone: 215-895-2230

418

Notes

419

The authors declare no competing financial interest.

420

Funding Sources

421

The funding for this project comes from the EPA P3 Design Award—SU-83352401, GAANN

422

RETAIN—Award No. P200A100117, and WERF Research Grant—U3R13, Extraction of Lipids

423

from Wastewater to Produce Biofuels.

424

REFERENCES

425 426 427 428 429 430 431 432 433

1. Eisentraut, A. Sustainable Production of Second-Generation Biofuels; International Energy Agency: France, 2010. 2. Haas, M. J.; McAloon, A. J.; Yee, W. C.; Foglia, T. A., A process model to estimate biodiesel production costs. Bioresource technology 2005, 97, (4), 671-8. 3. Wiltsee, G., Waste Grease Resources in 30 US Metropolitan Areas. In BioEnergy '98: Expanding BioEnergy Partnerships, Great Lakes Regional Biomass Energy Program: Madison, Wisconsin, 1998. 4. Ragauskas, A. M. E.; Pu, Y.; Ragauskas, A. J., Biodiesel from grease interceptor to gas tank. Energy Science & Engineering 2013, 1, (1), 42-52.


19

Page 21 of 27

434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478


5. Ward, P. M., Brown and black grease suitability for incorporation into feeds and suitability for biofuels. Journal of food protection 2012, 75, (4), 731-7. 6. Austic, G., Feasibility Study: Evaluating the profitability of a trap effluent dewatering facility in the Raleigh area. In LLC, P. B., Ed. ECO Collections: Raleigh, NC, 2010. 7. Canakci, M., The potential of restaurant waste lipids as biodiesel feedstocks. Bioresource technology 2007, 98, (1), 183-90. 8. Long, J. H.; Aziz, T. N.; Reyes, F. L. d. l.; Ducoste, J. J., Anaerobic co-digestion of fat, oil, and grease (FOG): A review of gas production and process limitations. Process Safety and Environmental Protection 2012, 90, (3), 231-245. 9. Landfilling; EPA: 2012; pp 1-22. 10. Sundqvist, J. O. Life cycles assessments and solid waste; Swedish Environmental Research Institute: Stockholm, Sweden, 1999. 11. Foster, P. V.; Ramaswamy, P.; Artaxo, T.; Berntsen, R.; Betts, D. W.; Fahey, J.; Haywood, J.; Lean, J.; Lowe, D. C.; Myhre, G.; Nganga, J.; Prinn, R.; Raga, G.; Schulz, M.; Van Dorland, R. 2007: Changes In Atmospheric Constituents and in Radiative Forcing; Cambridge University Press: Cambridge, United Kingdom and New York, NY, USA, 2007. 12. Silvestre, G.; Rodriguez-Abalde, A.; Fernandez, B.; Flotats, X.; Bonmati, A., Biomass adaptation over anaerobic co-digestion of sewage sludge and trapped grease waste. Bioresource technology 2011, 102, (13), 6830-6. 13. Razaviarani, V.; Buchanan, I. D.; Malik, S.; Katalambula, H., Pilot-scale anaerobic codigestion of municipal wastewater sludge with restaurant grease trap waste. Journal of environmental management 2013, 123, 26-33. 14. Gough, H. L.; Nelsen, D.; Muller, C.; Ferguson, J., Enhanced Methane Generation During Thermophilic Co-Digestion of Confectionary Waste and Grease-Trap Fats and Oils with Municipal Wastewater Sludge. Water Environment Research 2013, 85, (2), 175-183. 15. Lopez, R. J.; Higgins, S. R.; Pagaling, E.; Yan, T.; Cooney, M. J., High rate anaerobic digestion of wastewater separated from grease trap waste. Renewable Energy 2014, 62, 234-242. 16. Dufour, J.; Iribarren, D., Life cycle assessment of biodiesel production from free fatty acid-rich wastes. Renewable Energy 2012, 38, (1), 155-162. 17. Tu, Q.; McDonnell, B. E., Monte Carlo analysis of life cycle energy consumption and greenhouse gas (GHG) emission for biodiesel production from trap grease. Journal of Cleaner Production 2015. 18. Environmental management -- Life cycle assessment -- Principles and framework. In International Organization and Standardization: 2006; Vol. ISO 14040: 2006, pp 1-20. 19. Frischnect, R.; Jungbluth, N.; Althaus, H.; Bauer, C.; Doka, G.; Dones, R.; Hischier, R.; Hellweg, S.; Humbert, S.; Köllner, T.; Loerincik, Y.; Margni, M.; Nemecek, T. Implementation of Life Cycle Impact Assessment Methods; Swiss Centre for Life Cycle Inventories: Dübendorf, 2007. 20. Jungbluth, N.; Chudacoff, M.; Dauriat, A.; Dinkel, F.; Doka, G.; Faist Emmenegger, M.; Gnansounou, E.; Kljun, N.; Schleiss, K.; Spielmann, M.; Stettler, C.; Sutter, J. Life Cycle Inventories of Bioenergy Data v2.0 (2007); Swiss Centre for Life Cycle Inventories: Dübendorf, CH, 2007. 21. Rehl, T.; Lansche, J.; Müller, J., Life cycle assessment of energy generation from biogas—Attributional vs. consequential approach. Renewable and Sustainable Energy Reviews 2012, 16, (6), 3766-3775.


20


479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522

Page 22 of 27

22. Brander, M.; Tipper, R.; Hutchinson, C.; Davis, G. Consequential and Attributional Approaches to LCA: a Guide to Policy Makers with Specific Reference to Greenhouse Gas LCA of Biofuels; Ecometrica Press: 2008. 23. SimaPro 8.0.3.14 PhD; PRé Consultants The Netherlands, 2014. 24. Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model, Argonne National Laboratory: 2014. 25. Oracle Crystal Ball, 11.1.2.4.000; Oracle 2014. 26. Stacy, C. J.; Melick, C. A.; Cairncross, R. A., Esterification of free fatty acids to fatty acid alkyl esters in a bubble column reactor for use as biodiesel. Fuel Processing Technology 2014, 124, (0), 70-77. 27. Mohammed, M. Mathematical Modeling of a Two-Phase Bubble-Column Reactor for Biodiesel Production from Alternative Feedstocks. Drexel University, Philadelphia, PA, 2011. 28. Lam, A.; Matthew, S.; Melick, C.; Mohammed, M. Technoeconomic assessment of biodiesel production from alternative feedstocks via three processes; Unpublished Senior Design Report. Drexel University: Philadelphia, PA, 2010. 29. Haas, F. M.; Sanchez, J.; Letterle, K. Design of a Waste Cooking Oil Upgrader for BioFuel Processing: Trap Grease Pretreatment; Unpublished Senior Design Report. Drexel University: Philadelphia, PA, 2005. 30. Bucher, L.; DeVitis, D.; Morris, M.; Wallowitch, G. Technoeconomic feasibility study of a brown grease to biodiesel process; Unpublished Senior Design Report. Drexel University: Philadelphia, PA, 2014. 31. Cairncross, R. A.; Olson, M. S.; Spatari, S., Extraction of Lipids from Wastewater to Produce Biofuels: Statistical Variability of Grease Composition and Quantity. In Unpublished Interim Report for WERF, Drexel University: Philadelphia, PA, 2015. 32. Eleazer, W. E.; William S. Odle, I.; Wang, Y.-S.; Barlaz, M. A., Biodegradability of Municipal Solid Waste Components in Laboratory-Scale Landfills. Environmental Science and Technology 1997, 31, (3), 911-917. 33. Kocsisova, T.; Cvengros, J.; Lutisan, J., High-temperature esterification of fatty acids with methanol at ambient pressure. European Journal of Lipid Science and Technology 2005, 107, (2), 87-92. 34. Canakci, M.; Van Gerpen, J., Biodiesel Production from Oils and Fats with High Free Fatty Acids. Transactions of the ASAE 2001, 44, (6), 1429-1436. 35. Gardner, E. R.; Shang, Y.; Yuan, Y.; Gray, D. M. D., Producing Biodiesel from Fat, Oil and Greases (FOG) and Other Waste Material at Wastewater Treatment Plants. In Proceedings of the Water Environment Federation, Water Environment Federation: 2013. 36. Standard Specification for Biodiesel Fuel Blend Stock (B100) for Middle Distillate Fuels. In ASTM D6751-15, ASTM International: West Conshohocken, PA, 2015. 37. Lopez, D. E.; Mullins, J. C.; Bruce, D. A., Energy Life Cycle Assessment for the Production of Biodiesel from Renedered Lipids in the United States. Industrial & Engineering Chemical Research 2010, 49, 2419-2432. 38. Goedkoop, M.; Oele, M.; de Schryver, A.; Vieira, M. SimaPro Database Manual Methods Library; PRé Consultants: the Netherlands, 2008. 39. A Comprehnsive Analysis of Biodiesel Impacts on Exhaust Emissions; EPA420-P-02-001; Assessment and Standards Division, Office of Transportation and Air Quality: 2002.


21

Page 23 of 27

523 524 525 526 527


40. Spatari, S.; MacLean, H. L., Characterizing Model Uncertainties in the Life Cycle of Lignocellulose-Based Ethanol Fuels. Environmental Science and Technology 2010, 44, 87738780. 41. Kumar, S.; Singh, J.; Nanoti, S. M.; Garg, M. O., A comprehensive life cycle assessment (LCA) of Jatropha biodiesel production in India. Bioresource technology 2012, 110, 723-729.

528


22


Transportation Material Flow

Materials/Utilities GTW Delivery

Select Emissions

GTW-BIODIESEL

(1) PRE-TREATMENT

(2) FUEL PRODUCTION (a) Conversion

(a) Oil Extraction Lipids

GTW-Waste Solids

(i) Reaction

(b) Waste Management (i) WS Transportation

(ii) WS Treatment

(iii) WW Transportation

(iv) WW Treatment

(ii) Methanol Recovery

(3) VEHICLE OPERATION

(b) Purification (i) Washing

(ii) Distillation

Production Wastewater

(c) Waste Management (i) WW Transportation

(ii) WW Treatment

Biobunker Biodiesel (d) Service Station (i) Biodiesel Transportation

(ii) Operation

Biodiesel

GTW-Wastewater

Page 24 of 27

529 530 531 532 533

Figure 1. System boundary for the GTW-biodiesel process. Each of the three main stages include the material and energy inputs and emission outputs for 1) Pre-treatment (orange), 2) Fuel production (yellow), and 3) Vehicle operation (gray). Some process stages have a sub-stage marked with letters a-d and some sub-stages have individual steps marked i-iv.

534


23

Page 25 of 27


300

300

200

250 200 150

100

100

50

50

0

0

-50

-50

40%

30%

-150 20%

-150 10%

-100 7%

-100

Flaring: Biodiesel from GTW with Starting Lipid Content 300

300

250

250

200

150

1 = 2.8 + 18 % ' 2 = 1.0

200

100

150

100

100

50

50

0

0

-50

-50

40%

-150 20%

-150 10%

-100 7%

-100 2% 3% 4% 5%

GWP100 g-CO2-eq/MJ-fuel

B)

1 = 4.1 + 18 % ' 2 = 1.0

100

150

2% 3% 4% 5%

GWP100 g-CO2-eq/MJ-Fuel

250

30%

A)

Co-Generation: Biodiesel from GTW with Starting Lipid Content

535 536 537 538 539 540 541 542

GTW Delivery to Transfer Station

Fuel Production

Avoided Natural Gas

Pre-Treatment WM

Vehicle Operation

Total

Pre-Treatment w/o WM

Avoided Electricity

Theoretical Total

Figure 2. Attributional LCA approach for the parametric study on the affect of lipid content on the total GWP100 for GTW-biodiesel process for A) Flared Landfill Gas and B) Co-Generation of Landfill Gas. The stacked bars represent GTW-biodiesel stages: delivery of GTW to transfer station (red), pre-treatment WM (orange with blue dots), pre-treatment without WM (orange), fuel production (yellow), vehicle operation (gray), avoided electricity production from cogeneration (light green), and avoided natural gas from co-generation (teal). The total GWP100 and modeled curve (black line) are also shown.

543


24


450

450

350

350

250

250

150

150

50

50

-50

-50

-350

-450

-450

40%

-350

30%

-250

20%

-250

10%

-150

7%

-150

2% 3% 4% 5%

GWP100, g-CO2-eq/MJ-fuel

A)

Page 26 of 27

Flaring: Biodiesel from GTW with Starting Lipid Content 450

350

350

250

250

150

150

50

50

-50

-50

-350

-450

-450

40%

-350

30%

-250

20%

-250

10%

-150

7%

-150

2% 3% 4% 5%

GWP100, g-CO2-eq/MJ-fuel

B) 450

Co-Generation: Biodiesel from GTW with Starting Lipid Content Current GTW Delivery to Transfer Station

Displaced Current Processes

544 545 546 547 548 549 550 551 552 553

GTW Delivery to Transfer Station

Current GTW Disposal

Pre-Treatment WM

Current GTW Co-Gen Avoided Utilities

Biodiesel Co-Gen Avoided Utilities

LSD

GTW-Biodiesel Rest of Process

Proposed GTW-Biodiesel Process

Total

Figure 3. Consequential LCA approach to compare GTW-biodiesel production to current GTW disposal. GWP100 shown for GTW-biodiesel process for A) Flared Landfill Gas and B) Cogeneration of Landfill Gas. The lipid content of the GTW was varied from 2-40%. The colored negative bars represent avoided impacts including current GTW transportation (red striped), current GTW disposal (orange hashed) and avoided impacts due to Co-generation (electricity and natural gas, blue striped) and LSD (purple). The positive bars represent the GTW transportation (red), GTW-biodiesel process (green) and the no longer avoided impacts (electricity and natural gas) from the current GTW disposal (teal). The total emissions (black bar) represent the difference between total biodiesel process and avoided emissions.


25

Page 27 of 27


C) Low Sulfur Diesel Soybean Biodiesel

8

1.0 ImpactGTW/ImpactLSD

6 4 2

CED

0.4 0.2

GWP100

CO

Raw GTW Distribution

15 0 -15 -30

CG

W/O

F

W/O

CO

3.0 ImpactGTW/ImpactLSD

30

CED

Dewatered GTW Distribution

D)

45

CG

F

W/O

F

CG

W/O

CG

F

W/O

F

CG

CG

W/O

F

GWP100

2.0 1.0 0.0 -1.0

NOx

SOx

Raw GTW Distribution

PM

NOx

W/O

CG

F

W/O

CG

W/O

CG

F

W/O

CG

F

W/O

F

CG

PM

F

-2.0

-45

W/O

ImpactGTW/ImpactLSD

0.6

0.0

0

B)

0.8

F

ImpactGTW/ImpactLSD

10

CG

A)

SOx

Dewatered GTW Distribution

554 555 556 557 558 559 560 561 562 563 564 565 566

Figure 4. Monte Carlo analysis of the sensitivity of six LCIA metrics for several GTW waste management scenarios and two distributions of lipid content. All impacts are normalized to the corresponding impact for low sulfur diesel (LSD). Panels A and C display 100-year Global Warming Potential (GWP100), Fossil Cumulative Energy Demand (CED), and carbon monoxide (CO), and panels B and D display Particulate Matter (PM), mono-nitrogen oxides (NOx), and sulfur oxides (SOx) emissions. Panels A and B display results based on lipid content distributions in raw GTW, and panels C and D display results for dewatered GTW. Scenarios compared are landfill gas flare (F), landfill gas co-generation (CG) and without waste management (W/O). The line in the middle of each box represents the median, the upper half of the box represents the 3rd quartile, and the lower half of the box represents the 2nd quartile. The positive and negative error bars represent the 90% and 10% percentile intervals. The green line represents soybean-biodiesel and red line represents LSD.


26

Life-Cycle Assessment of Biodiesel Produced from Grease Trap Waste

Recommend Documents