Comparative assessment of selected sugarcane biorefinery-centered systems in Brazil: A multi-criteria method based on sustainability indicators

Accepted Manuscript Comparative assessment of selected sugarcane biorefinery-centered systems in Brazil: A multi-criteria method based on sustainability indicators Edgard Gnansounou, Catarina M. Alves, Elia Ruiz Pachón, Pavel Vaskan PII: DOI: Reference:

S0960-8524(17)31090-8 http://dx.doi.org/10.1016/j.biortech.2017.07.004 BITE 18423

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

5 May 2017 30 June 2017 1 July 2017

Please cite this article as: Gnansounou, E., Alves, C.M., Ruiz Pachón, E., Vaskan, P., Comparative assessment of selected sugarcane biorefinery-centered systems in Brazil: A multi-criteria method based on sustainability indicators, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.07.004

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Comparative assessment of selected sugarcane biorefinery-centered systems in Brazil:

2

A multi-criteria method based on sustainability indicators

3

Edgard Gnansounou *, Catarina M. Alves, Elia Ruiz Pachón, Pavel Vaskan

4

Bioenergy and Energy Planning Research Group, EPFL, Switzerland

5

* Corresponding author at: Bioenergy and Energy Planning Research Group, GC

6

A3 424 (Bâtiment GC), ENAC INTER GR-GN, EPFL, Station 18, CH-1015 Lausanne,

7

Switzerland. Tel.: +41 216930627. E-mail address: [email protected].

8 9

Abstract

10

This work proposes a new sustainability assessment framework aiming to compare

11

selected options of biorefineries subject to provide the same services to a community. At

12

this end, a concept of biorefinery-centered system helps to develop a joint resources and

13

policy-oriented comparison. When an option of biorefinery cannot provide the given

14

amounts of certain services from its own production, it complements its supply portfolio by

15

purchasing the lacking amounts from complementary and conventional production systems.

16

The proposed sustainability assessment framework includes a multi-criteria method used to

17

compare the selected biorefinery options resulting in identifying their respective

18

weaknesses and strengths against categories of criteria. Finally, the methodology helps

19

finding the non-dominated option. Application to selected sugarcane-based biorefineries

20

shows promising results that match with those obtained in a previous work. However, the

21

new methodology allows extended and richer analyses.

22

Keywords

23

Multi-criteria method, sugarcane biorefinery, sustainability, LCA

24

1. Introduction

25

Two main goals drive the development of biorefineries: substituting fossil-based products

26

for bio-based products and valorising the whole biomass feedstocks into value added 1

27

products. In that sense, biorefineries must be conceptually defined considering not only the

28

conversion part, but also the whole value chain, from biomass production, collection and

29

procurement to the delivery of services such as fuels for mobility, electricity services, sugar,

30

chemical and materials for industrial processes. The concept of biorefinery systems

31

emerged from that way of thinking and was first used to couple the type of biomass to the

32

conversion technologies. Kamm and Kamm (2004) emphasized three biorefinery systems:

33

Whole Crop Biorefinery, Green Biorefinery and Lignocellulosic Feedstock Biorefinery. This

34

concept was improved in the framework of the International Energy Agency (IEA) Task 42,

35

by defining a classification of biorefinery that considers features such as platforms, final bio-

36

based products, feedstocks and bio-processes (de Jong and van Ree, 2009; de Jong and

37

Jungmeier, 2015). The work of IEA was extended to other aspects such as sustainability

38

and flexibility (Gnansounou and Pandey, 2017).

39

The sustainability performance must be a core concern of biorefinery systems since the

40

sustainable valorisation of biomass is the backbone of that concept. Several works were

41

devoted to biorefineries from the sustainability point of view. Azapagic (2014) identified

42

several environmental issues of biorefineries’ sustainability, as greenhouse gas emissions,

43

biodiversity, land use change, water use, and other environmental effects such as local

44

emissions of pollutants. Whereas, the author set out relevant economic considerations such

45

as cost of feedstocks, capital costs, biofuels and coproducts costs. Finally, for social

46

considerations, the issues selected were jobs, regional development, health, human and

47

labour rights, land availability, food prices and intergenerational matters. Once these issues

48

are identified, there is a need for an integrated assessment including the different

49

sustainability dimensions. Despite the large number of valuable feasibility and

50

environmental assessments of biorefineries in literature, the majority of the studies and

51

methods for assessing sustainability are confined to the traditional techno-economic (e.g.

2

52

net present value) and environmental (e.g. life cycle assessment) evaluations. Even though

53

some present more than one dimension, the conclusions lack in integrating the different

54

dimensions of sustainability. Furthermore, from a societal perspective, biomass refining

55

may impact significantly the resources supply systems, energy security and rural economic

56

development (Lynd et al., 2005). Still, the investigation of the social aspects is recurrently

57

neglected since the social issues are typically harder to quantify given its dynamics and

58

strong context dependency. The exclusion of social metrics and the lack of a multi-criteria

59

analysis can negatively impact the project stakeholders’ confidence, including investors and

60

communities. Hence, an integrated multi-criteria comparative analysis adds value to the

61

evaluation of the biorefinery sustainability. In this line, Schaidle et al. (2011) used an

62

Analytic Hierarchy Process (AHP) multi-criteria method to compare three biorefineries from

63

sustainability point of view. The biorefineries studied were the following: grain ethanol,

64

cellulosic ethanol and Fischer-Tropsch diesel. The authors used various metrics for

65

modelling the sustainability. The environmental metrics were energy demand, greenhouse

66

gas emissions, SOx and NOx emissions, eutrophication potential and water use. Moreover,

67

the return on investment was selected as economic metric and four metrics for the social

68

dimension of sustainability were taken: job creation, food price, health effects and location

69

of pollutant emissions. The results of the comparison depend on the weights assigned to

70

each sustainability pillar (environment, economy and society). The main drawback of the

71

current assessment is related to the comparison of the biorefineries based on one mega

72

joule (MJ), with no consideration to the scale effect. Also, the context and the set of

73

services may be different from one biorefinery to the other, which makes it difficult to make

74

a straight comparison. Moreover, Sacramento-Rivero (2012) presented a multi-criteria

75

methodology for biorefineries at the conceptual design stage which relies sustainability

76

scales. The indicators results were converted using a normalization formula that returns

3

77

values from zero to infinity. The two boundaries were zero as the theoretical ideal value

78

(“ideal sustainability”) and the unity as the critical value (any value greater than one is

79

considered “unsustainable”). Both boundaries values were defined based on the EU

80

Renewable Energy Directive targets (Sacramento-Rivero, 2012). Later, this method was

81

adapted and implemented by Sacramento-Rivero et al. (2016) in order to assess multi-

82

product biorefineries from switchgrass. The highlights of that work are the incorporation of

83

critical values for defining the sustainability scales, the adaptation to the local context and

84

the graphical interpretation of the systems performance for each indicator. Nevertheless, no

85

integrated method comprising all the indicators has been defined, making it difficult to

86

perform systems comparison. Moreover, Gnansounou et al. (2015) compared four 1G2G

87

biorefineries in the context of Brazil, based on a fixed amount of sugarcane 13000 tons/day

88

(wet basis). The economic metric was the prospective economic performance, whereas five

89

environmental metrics were estimated through a life cycle assessment (LCA): climate

90

change, fossil depletion, human toxicity, freshwater ecotoxicity and freshwater

91

eutrophication. However, neither any social metric was considered nor was any multi-

92

criteria comparison performed. With regards to the environmental metrics, the authors

93

distinguished between two types of comparison: one based on the absolute performance

94

and another based on relative performance that is the difference between the performance

95

of the reference system and the system under consideration. The latter comparison is

96

policy-oriented. The results of these two types of comparison may be different which

97

reveals the importance to clarify the meaning and orientation of the comparison. Thus, the

98

main objective of this paper is to develop a joint service and policy-oriented comparison

99

method for biorefinery systems. The method is applied to the four biorefinery scenarios

100

earlier described by Gnansounou et al. (2015). The economic and environmental figures of

101

the scenarios are converted into sustainability metrics. Moreover, the novel method covers

4

102

social metrics and allows for an integrated multi-criteria analysis of the systems which

103

provide equal services to the community. The authors propose a strong contextualization

104

which relies on local figures and cultural values.

105

2. Materials and methods

106

The methodology developed in this work is composed of five modules (Fig. 1). The first

107

module determines the context, the choice of cultural values and the criteria for the

108

selection of the biorefinery systems. Module 2 features the biorefineries modelling, mass

109

and energy balances needed to evaluate the performance matrix of the scenarios against

110

the criteria. Module 3 evaluates the performance matrix and Module 4 compares the

111

biorefinery systems. Finally, Module 5 uses other relevant criteria for an in-depth

112

assessment of the non-dominated biorefinery systems.

113

2.1.

114

2.1.1. Context definition

115

The aim and the context of the comparative assessment would dictate the selection of

116

biorefineries to be compared. Comparison of biorefineries can be driven by design

117

purposes, with the aim to select the most efficient route for given bio-products and

118

feedstock. For instance, the context could be comparing different conversion routes for a

119

given amount of a defined feedstock, or different mixes of feedstocks that are available in

120

various amounts, with the aim of producing the same amount of bio-products. Depending

121

on the context, the types of biorefineries that would be relevantly selected and compared

122

would be quite different. In the case of Schaidle et al. (2011), few elements of context are

123

defined by the authors such as the products and the type of conversion. However, in that

124

paper, other elements are lacking such as the type of biomass for the cellulosic ethanol and

125

Fischer-Tropsch diesel. In general, the context definition may include the main goal of the

126

biorefinery concepts, the type of feedstock, the size of the plant that depends on the

Context definition, choice of cultural values and biorefineries selection

5

127

available feedstock, the distance of collection, the scale economy, the existing and

128

perspectives of the bio-products’ market (Luo et al., 2010; Cherubini and Jungmeier, 2010;

129

FitzPatrick et al., 2010). The context in Gnansounou et al. (2015) was the comparison of

130

several routes of valorisation of sugarcane into bioethanol, sugar, C molasses, surplus

131

electricity and syrup for animal feed through technology and product lines integration. The

132

four selected options in that case were designed accordingly. That context explained why

133

each of them was assumed to process the same amount of sugarcane. The results of the

134

comparison would change if the context was to restraint the biorefinery options to provide

135

the same amounts of services to the community.

136

2.1.2. Choice of cultural values

137

Another aspect of the first module is the selection of the cultural values. As far as multi-

138

criteria comparison is concerned, the different criteria may be weighted according to the

139

choice of cultural values. In most cases, the authors choose various weights vectors that

140

represent diverse preferences. The issue about how to weight the criteria pertains to social

141

sciences. However, the meaning of weighting criteria in the decision of selecting options of

142

biorefineries is not obvious. Its relevance may depend on several factors including context,

143

types of decision makers and decision process. For example, if the decision process is the

144

search of consensus among a group of stakeholders, the meaning and use of weighting

145

would be different compared to the case of one exclusive decision maker or a deliberation

146

based on community vote. In general, even in the case of one exclusive decision maker,

147

social acceptance of biorefineries has a significant importance for siting issue for instance.

148

Rejection as well as long lasting judicial processes that would result in cost increase and

149

economic unfeasibility are two main reasons that could justify growing attention paid to

150

social acceptance. Few authors analyzed factors that influence social acceptance or

151

rejection of sustainable energy facilities (Wüstenhagen et al., 2007; Gupta et al., 2011;

6

152

Huijts et al., 2012). However, the most consistent theoretical investigation was undertaken

153

in the framework of the Theory of Cultural Values (TCV) that led to sets of archetypes

154

based on attitude categories. Schwarz (1999) investigated some implications of TCV with

155

respect to work and distinguished seven-contrasted value types: Conservatism - Intellectual

156

autonomy - Affective autonomy - Hierarchy - Egalitarianism - Mastery - Harmony. He

157

validated these values types using data from 49 countries. Hofstetter et al. (2000) also used

158

TCV and proposed three archetypes in the framework of LCA: Hierarchist - Egalitarianist -

159

Individualist. LCA considers these archetypes to derive assumptions regarding aspects

160

such as perception of time and resources. However, when applying to the issue of

161

comparison of biorefineries systems based on sustainability criteria, it is not straightforward

162

to interpret these archetypes. Should they characterize the dominant attitudes in the

163

community or the attitudes profile of the facility owner? The answer to this question should

164

be elucidated in the first stage of the LCA where the goal and scope are defined.

165

Nevertheless, when the methodology includes a multi-criteria analysis in addition to LCA,

166

the context and values system must be defined as a whole for the sake of consistency. This

167

paper considers the context of one exclusive decision maker whose attitude type is

168

egalitarianism. Therein, he is sensitive to global issues such as global warming and

169

depletion of abiotic resources, social issues such as job creation and services provided to

170

the community. He considers that long term should dominate short term and gives a

171

preference to prevention instead of pure adaptation. His conception of sustainability gives

172

more weights to global environmental and social impacts than to economic profitability but

173

remains sensitive to economic feasibility and risk.

174

2.1.3. Biorefineries selection

175

In line with these cultural values, the four biorefinery systems under comparison are not

176

only subject to use the same amount of sugarcane, but also to provide equal amount of

7

177

services. However, depending on their technological assets, they must complement their

178

own production by purchasing and delivering to the community less sustainable products

179

listed in the reference system. This defines the concept of biorefinery-centered system as a

180

joint resource and policy-oriented system. The “Ethanol Distillery Only Fuel” (ED OF)

181

biorefinery produces the maximum amount of ethanol and some electricity. However, its

182

portfolio will be complemented by purchasing conventional sugar, animal feed and

183

electricity. The “Ethanol Distillery Fuel and Feed” (ED FF) produces the maximum amount

184

of syrup, some ethanol and some electricity. The lack of products is supplemented with

185

conventional electricity, sugar, animal feed and gasoline. The “Sugar Mill” biorefineries (SM)

186

produce the maximum amount of sugar and C molasses. SM OF sells to the community the

187

maximum amount of surplus electricity but has to purchase animal feed and ethanol. As the

188

ED FF system, SM FF produces the maximum syrup but must purchase the lacking amount

189

of both electricity and ethanol. The complementary services (CS) are gasoline from oil

190

refining industry (mobility services), sugar and molasses from conventional sugarcane mills

191

(food and feed services), and electricity from the grid (electricity services).

192

2.2.

193

The process was simulated in Aspen Plus v8.6. It was split in several areas where the main

194

biomass transformations take place. Convergence of all process required to simulate the

195

entire plant was successfully completed, closing the energy and mass balance. The

196

simulation results are used for economic and environmental assessment. Because the

197

biorefinery-centered systems must provide the same amount of services, the CS are

198

calculated based on the mass and energy balances of the sugarcane biorefinery. The

199

selected sustainability metrics are calculated based on the mass/energy balances and CS.

Process design, flow sheeting, mass and energy balances

8

200

2.3.

Sustainability metrics

201

The criteria used to compare and analyze the selected biorefineries were designed based

202

on the social, economic and environmental components of sustainability, choice of cultural

203

values, life cycle and need to compare the four biorefinery options. Hence, the current multi-

204

criteria comparison does not cover indicators showing no significant difference among the

205

options. As an example, relevant indicators proposed by Sadhukhan et al. (2014) that

206

address social issues related to governance, labour, human rights, health and safety, have

207

not been incorporated in this study as they are expected not to change from one scenario to

208

the other.

209

2.3.1. Socioeconomic indicators

210

(i) Cost of services. Since each biorefinery center has different design and costs, the

211

current indicator aims to answer: how expensive is the provision of a determined type and

212

quantity service in each biorefinery-centered system, i.e., what is the total cost of services

213

provision in each system? The cost of services indicator figures the cost of services

214

produced by the biorefinery center and the CS costs. In order to allocate the biorefinery

215

operation and capital expenditures to the different products, the authors have adopted a

216

value-based approach that is an economic value analysis consisting on the calculation of

217

the streams value (one by one) starting from the final product streams. The value(s) of the

218

input stream(s) of the operation unit correspond to the sum of the value(s) of the output

219

stream(s), subtracted by the costs of auxiliary raw materials, utilities and annualized capital

220

cost of the operation unit. Such method has been described in detail by Gnansounou et al.

221

(2015). The costs of feedstocks, supplies and annuity related to the direct installed capital

222

costs of each process unit were allocated based on the value-based analysis. Regarding

223

the allocation of the cogeneration costs to electricity and steam, one has initially fixed the

224

steam value as per the cost of conventional steam production from natural gas – 21.9

9

225

US$/Gcal (USDE, 2003; 2012). Once the steam cost was fixed, the electricity cost share

226

could be estimated. The electricity cost should then be shared by the electricity consumed

227

and the surplus electricity. The costs associated to the electricity consumed at the plant,

228

steam and cooling water were allocated to the biorefinery products by using a simplified

229

economic allocation based on the product revenue share over the total revenues. Such

230

allocation was also applied to allocate other costs that are not necessarily directly

231

associated to a process stage, e.g. fixed operating costs, storage costs, and annuity costs

232

related to warehouse, site development, field expenses, construction fees and project

233

contingency. Furthermore, the complementary costs were estimated based on the services

234

purchased from the surrounding complementary industries.

235

(ii) Cross-subsidisation 1G/1G2G. It refers to the strategy of reducing the benefit provided

236

by a product (1G ethanol in this work) to subsidize the loss of revenue due to pricing

237

another product below its production cost (2G ethanol). The level of cross-subsidisation of

238

2G ethanol against 1G ethanol in every system is analysed since a unique selling price is

239

attributed to the 1G2G ethanol product. Then the cross-subsidisation 1G/1G2G indicator

240

answers the question: Which is the percentage of subsidisation of 1G2G ethanol by 1G

241

ethanol? The cross-subsidisation 1G/1G2G indicator is based on Equation 1, where M is

242

the profit margin for ethanol and subscript x may refer to both 1G or 1G2G (Equation 1-a).

243

The closer

244

production of 1G ethanol), the cross-subsidisation parameter becomes lower. Therefore, a

245

lower cross-subsidisation (%) indicates reduced differences between 1G and 2G ethanol

246

production cost and, consequently, lower dependence on the 1G product.

gets to

(cost of production of 1G2G ethanol close to cost of

247

(1)

248

(1-a)

10

249

(iii) Sensitivity of the biorefinery owner to price volatility. The current indicator was

250

designed to assess the sensitivity of the biorefinery owner to price volatility. It is a weighted

251

sum of the product price volatilities where the weight of each price volatility is the elasticity

252

of the net present value to that price.

253

(iv) Energy security. This indicator is defined as the uninterrupted availability of energy

254

source at an affordable price (IEA, 2016). The lack of energy security due to sudden

255

changes within the energy supply-demand balance leads to negative impacts on social and

256

economic welfare. Bioenergy can enhance energy security by reducing dependency to

257

volatile supplied energy expenditures and shifting the consumption towards more stable

258

local energy supply (Dale et al., 2013). The considered biorefinery-centered systems

259

provide renewable energy to the local communities in the form of electricity. Thus, the

260

Energy Security indicator aims to reply to the question: what is the contribution of the

261

biorefinery system to the local energy security levels? The indicator is based on the surplus

262

electricity produced by the biorefinery center.

263

(v) Employment creation. Dale et al. (2013) have identified employment as one of the

264

most relevant indicator of socioeconomic sustainability. The biorefinery system is envisaged

265

to be placed in rural areas close to the sugarcane production fields. Thus, besides

266

enhancing energy security, an increase in sugarcane production and implementation of new

267

processing industries are expected to contribute to economic progress and social

268

development through job creation. This indicator evaluates the social investment created by

269

the biorefinery-centered system through employment creation. Employment creation was

270

estimated for the different sections: sugarcane plantation, biorefinery center, operation at

271

the conventional sugar and oil refining industries. The number of jobs associated to the

272

sugarcane plantation is expected to be the largest in the supply chain.

11

273

2.3.2. Environmental indicators

274

(i) Climate change. Climate change refers to a change in the state of the climate conditions

275

due to natural variability or as a result of human activity that persists in an extended time

276

(IPCC, 2007). Human activities contribute to climate change through the emission of

277

greenhouse gases (GHG) to the atmosphere (McBride et al., 2011). These GHGs are

278

weighted by their global warming potentials for a time horizon of 100 years, measured in kg

279

CO2 equivalent per year (IPCC, 2007; SimaPro, 2016). Previous work reported that climate

280

change impact had been reduced in 56-71% from the reference system (gasoline) and the

281

biofuels (Gnansounou et al., 2015). Therefore, significant differences are expected among

282

the current scenarios given the variations in the gasoline and biofuel supply. The climate

283

change indicator was estimated based on LCA principles, which allows for the quantification

284

of the GHG emissions along the whole life cycle of the products, raw materials and wastes.

285

The LCA methodology applied in this work is consistent with Gnansounou et al. (2015).

286

(ii) Fossil fuel depletion. The fossil fuel depletion quantifies the contribution of each

287

system to the global depletion of fossil fuel reserves. The fossil fuel depletion impact relates

288

to the amount of extracted fossil fuel, expressed in kg of oil equivalent per year (Simapro,

289

2016). This indicator was estimated based on LCA methodology, such as the previous

290

indicator. Lower fossil depletion levels indicate reduced impacts on the natural reserves.

291

(iii) Freshwater eutrophication. Water quality and quantity reflect the diversity of

292

conditions and land practices occurring upstream, as well as past events. The properties of

293

the freshwater in streams draining from bioenergy croplands will have an impact in the

294

ecosystems within and downstream of these streams (McBride et al., 2011). Thus, the

295

levels of freshwater eutrophication was considered as an environmental indicator.

296

Freshwater eutrophication is generally related to the environmental persistence of the

297

emissions of phosphorus containing nutrients in the water. Typical sources of P are

12

298

agricultural fertilizers. Freshwater eutrophication is expressed in kg of P equivalent per year

299

and is calculated based on the LCA methodology. The more intense the land practices, the

300

greater the impacts on freshwater eutrophication due to the use of agrochemicals.

301

(iv) Freshwater consumption. It concerns the preservation of water natural reserves. This

302

indicator presents the amount of freshwater required in the biorefinery center, conventional

303

sugar mill (sugar and sugarcane molasses complementary supply) and oil refining industry

304

(fuel complementary supply). The biorefinery-center freshwater requirements are estimated

305

based on the process simulation results. The water balance depends on the heating and

306

cooling utilities, liquefaction and washing requirements and recycle water collected from the

307

wastewater treatment unit. The make-up water required by the plant is the evaluable

308

parameter for each biorefinery center.

309

(v) Use of chemicals. This indicator is defined with the objective to highlight the difference

310

of chemicals consumption in the different biorefinery schemes. Processes free of synthetic

311

chemicals lead to greener industries and have a direct impact in the sustainability of fuels

312

supply chain. From the process simulation and mass balances, different amount of

313

chemicals are used in OF and in FF schemes due to the use of Dilute Acid (DA) and Liquid

314

Hot Water (LHW) pretreatment techniques. The current indicator aims to quantify the impact

315

of the utilisation of acids in the total chemicals requirements of the biorefinery plant.

316

2.4.

317

When comparing alternatives subject to multiple criteria, several methods can be used such

318

as multiple objectives or multi-attributes (Keeney and Raiffa, 1993), and Multi-Criteria

319

Decision Analysis (MCDA). Although these terms are often used interchangeably, the term

320

MCDA is restricted in this research work to a family of methods that do not necessarily

321

optimize. Indeed, these methods could allow qualitative criteria and incomparability

322

between certain alternatives. Few examples of MCDA are “Analytic Hierarchy Process

Multi-criteria comparison

13

323

(AHP)” (Saaty, 1980, 1994), “Elimination et Choix Traduisant la Réalité (ELECTRE)” (Roy,

324

2016) and “Preference Ranking Organisation Method for Enrichment of Evaluations

325

(PROMETHEE)” (Brans, 2016). The multi-criteria comparison module of the methodology

326

uses PROMETHEE to compare the four biorefineries based on the sustainability criteria

327

and weights chosen according to the cultural values. The algorithm of comparison builds on

328

outranking relations (Roy, 1991; Brans and Vincke, 1985). PROMETHEE I allows

329

incomparability between some alternatives, meaning that the alternatives are ranked based

330

on a partial pre-order relation; the result is a partial ranking with a list of non-dominated

331

alternatives. Conversely, PROMETHEE II uses a total pre-order allowing total ranking by

332

transitivity closure.

333

2.5.

334

Once the best or non-dominated biorefineries are found from the fourth module, an in-depth

335

assessment can be completed that would include an optimized design aiming to maximize

336

the production of services and a refined characterization of these biorefineries with regard

337

to a more extensive set of sustainability metrics.

338

3. Results and discussion

339

The methodology above presented (Fig. 1 Modules 1-4) has been implemented in the case

340

study. Results are presented and discussed in the current section. The application of the

341

Module 5 to the best ranked biorefinery will be presented in another paper.

342

3.1.

343

The methodology proposed in section 2 was applied to the four sugarcane biorefinery-

344

centered systems located in Sao Paulo state region in Brazil. They comprise the production

345

of bioethanol, electricity, raw sugar and animal feed based on pentose sugars and

346

molasses (Fig. 2). A sugarcane-based biorefinery is the center of a system operating for

347

7200 working hours per year. Depending on the option, it includes an ethanol distillery or a

In-depth analysis of the non-dominated biorefineries

Context of the case study

14

348

sugar mill, a bioethanol production plant and a cogeneration unit. Sugarcane processing

349

capacity of the biorefinery centers is fixed, leading to different product capacities. Since

350

each biorefinery system must provide the same services (mobility, electricity, sugar and

351

feed) to the local community, the ones that cannot be supplied by the biorefinery center will

352

be provided by complementary systems, as schematized in Fig. 2. Thus, the definition of

353

the maximum services provided by the system is based on the maximum production among

354

the biorefinery centers. Moreover, the lifetime of the biorefinery-based projects is 35 years

355

(2016-2050). The first phase is envisaged to be ready in a 5 years timespan (2016-2020)

356

and comprises scope definition, conceptual design, sustainability assessment and basic

357

engineering. The second phase of the project involves the detailed engineering and it is

358

assumed to last 2 years (2021-2022). The third phase is the plant construction and

359

integration within the existing facilities, which is expected to be completed within 3 years

360

(2023-2025). Finally, the plant will start operating in 2026 and will run in full operation mode

361

for 25 years (2026-2050). Currently, the project is in phase 2, in which the biorefinery

362

conceptual design is performed based on literature and information from partners. The

363

sugarcane mill design was based on existing plant data assuming no changes in

364

technology. The second generation part (sugarcane bagasse process) was developed

365

based on Humbird et al. (2011) and the IBUS process (Larsen et al., 2008, 2012). NREL

366

presents no introduction of new technologies in 8 years (2007-2014). So, no changes are

367

assumed until the end of the detailed engineering stage. Moreover, IBUS process is

368

running at INBICON demo plant with the technology proposed in 2008, therefore, it is

369

reasonable to assume IBUS data for the current study.

370

3.2.

371 372

Description of the biorefinery scenarios

Four sugarcane biorefineries for bioethanol, sugar, animal feed and electricity production were designed based on sugar mill and ethanol distilleries. Aspen Plus simulation results

15

373

previously published by Gnansounou et al. (2015) were used in this work. The process

374

operation units are described in the following paragraphs. For representation purposes, the

375

units have been aggregated in blocks. The block diagrams are shown in Fig. 3.

376

The factory is fed by 13000 tons/day (wet basis) of sugarcane (similar for the four

377

configurations). Subsequent to the sugarcane feedstock handling, the extraction of sugars

378

is carried out in the mill where sugarcane juice and bagasse are separated. After that, the

379

clarification and purification of the extracted juice take place. In the ED scenarios, the

380

purified juice is conveyed to the fermentation area, whereas in the SM scenarios a multi-

381

effect evaporator concentrates the juice. In the SM scenarios, the major part of the

382

concentrated juice is funnelled to the crystallisation for sugar production, and only the

383

remaining 2% is fed to the fermentation in order to increase the ethanol fermentation yield.

384

The mixture of crystals (mainly sucrose) is separated from the syrup fraction (B and C

385

molasses). Overall, more than 72% (w/w) of the initial sugar is recovered as raw sugar, the

386

main product of the SM scenarios. C molasses are sold for animal feed, while B molasses

387

are destined to ethanol fermentation. The above described operations are represented in

388

Fig. 3 by the blocks A1 (ED OF), B1 (ED FF), C1 (SM OF) and D1 (SM FF). The energy

389

requirements of these blocks impact largely the overall economy of the biorefineries. Fig. 3

390

shows that C1 and D1 present higher heat requirements than A1 and B1 (71 Gcal/h vs.

391

41.8 Gcal/h), due to the evaporation unit for sugar production. The electricity required in

392

these four blocks is similar (17.9 MW), given the common high demand of the milling.

393

Furthermore, 60% (w/w) of the available sugarcane bagasse is processed in the

394

pretreatment unit, while the remaining part of the bagasse is sent to the cogeneration.

395

Hence, 2184 ton/day (wet basis) of sugarcane bagasse enter the feed handling area to be

396

transported and prepared for pretreatment, where the lignocellulosic biomass is degraded

397

into simple sugars. Two pretreatment techniques were considered depending on the

16

398

desirable products for the biorefinery: LHW for the FF scenarios and DA for the OF

399

scenarios (Gnansounou et al., 2015). After the LHW pretreatment, the hydrolysate slurry is

400

separated into liquid and solid fractions. The liquid fraction (containing C5 sugars) is

401

evaporated to produce C5 syrup. The C5 syrup product is ready to be used for animal feed,

402

whereas the solid fraction is conveyed to enzymatic hydrolysis (saccharification) to obtain

403

C6 sugars. In case of DA pretreatment, the whole pretreated hydrolysate slurry is

404

transferred to the saccharification. In all the configurations, the saccharification is performed

405

using cellulases produced in-situ and a 90% conversion of non-soluble sugars to glucose is

406

assumed. The saccharified slurry is latterly mixed with the juice and cooled down for

407

fermentation. In the fermentation reactor, sugars are converted to bioethanol. The

408

fermentation broth (including lignin and other solids) is fed to a distillation unit composed by

409

two columns. Further, the ethanol stream is conveyed to a molecular sieve adsorption for

410

further dehydration to 99.4 % (v/v). In the ED OF scenario, all the sugars are converted to

411

fuel, which makes it the largest ethanol producer (Fig. 3). The bottoms from the beer

412

distillation are filtered and the solid residues are separated from the thin stillage. The thin

413

stillage is conducted to the wastewater treatment area (WWT). The biogas and sludge

414

produced are used to generate heat and power in the cogeneration unit. The blocks A2 (ED

415

OF), B2 (ED FF), C2 (SM OF) and D2 (SM FF) in Fig. 3 include: feed handling,

416

pretreatment, C5 syrup production (FF scenarios only), saccharification, fermentation,

417

distillation, dehydration and wastewater treatment units. The amount of output residues of

418

B2 and D2 is larger than the ones of A2 and C2 since the LHW pretreatment is less efficient

419

than DA. In turn, the biogas production in the FF scenarios is lower than in the OF

420

scenarios since a liquid fraction has been extracted for C5 syrup production.

421 422

Finally, the solid residues, biogas, sludge, 40% of sugarcane bagasse and 50% (w/w) of the GHR are fed to the combined heat and power cogeneration unit (the so-called A3, B3,

17

423

C3 and D3 blocks in Fig. 3). The remaining GHR are left in the plantation field for ecologic

424

reasons. The cogeneration unit must produce the total heat and power that is required to

425

supply the plant. After satisfying the plant’s demand, the surplus electrical energy is sold to

426

the grid. Regarding the overall electricity requirements, OF scenarios need more electricity

427

than the FF scenarios. This is because in OF scenarios the liquid fraction remains in the

428

process after pretreatment, leading to larger mass flows to be pumped. Furthermore, the

429

heat duty requirements for the FF scenarios are higher than for the OF scenarios due to the

430

larger steam consumption in the LHW pretreatment and C5 syrup production units. Since

431

the cogeneration plant is designed to provide the exact heat duty demand, the lower the

432

heat duty requirements, the higher is the electricity generated. That explains why SM OF

433

presents the largest surplus electricity among the four scenarios.

434

3.3.

435

Cost of services (i1). The total cost of services indicator is presented in Table 1 for the

436

different scenarios. SM OF biorefinery-centered system provides the services at the lowest

437

costs. The total production costs of the SM biorefinery center services are lower than the

438

ED ones. However, SM scenarios present larger CS costs. SM OF system presents a good

439

compromise in terms of costs in comparison with SM FF one, due to the reduced CS costs

440

(gasoline and electricity). In spite of presenting higher costs in the biorefinery center, the ED

441

OF system is the second cheapest option since it implies lower complementary costs.

442

Cross-subsidisation 1G/1G2G (i2). The results show that there is an economy of scale

443

within the ED scenarios – the larger the ethanol production capacity, the lower the ethanol

444

production cost. Therefore, ED OF scheme presents lower ethanol production cost per unit

445

leading to lower cross-subsidisation in comparison to ED FF (Table 1). On the other hand,

446

the allocation process in the SM scenarios attributes biorefinery costs to sugar, leading to

447

lower ethanol unit cost in comparison with such cost in ED schemes. Since the allocation

Evaluation of the socioeconomic indicators

18

448

methods are based on market value of products, the scenarios providing more co-products

449

present a greater costs distribution. Accordingly, SM FF presents the lowest cross-

450

subsidisation level of all systems. Sensitivity of the biorefinery owner to price volatility

451

(i3). Table 1 depicts the final economic sensitivity for the different systems. It can be seen

452

that less profitable scenarios (with lower NPV) present higher sensitivity: SM (vs. ED) and

453

scenarios FF (vs. OF). ED OF is the scenario presenting the lowest sensitivity to product

454

price fluctuations due to the reduced number of co-products and the lower elasticity (higher

455

sales and NPV). Oppositely, SM FF is the most sensitive due to the high elasticity to

456

product prices variations (reduced NPV) and the larger number of products. Energy

457

security (i4). As shown in Table 1, SM scenarios provide higher surplus of electricity than

458

the ED scenarios do. Consequently, SM-based systems display greater contribution to the

459

energy security levels. Employment creation (i5). Currently, 0.36 employees per 1000 tons

460

of sugarcane per year are required at the plantation fields, according to Costa & Guilhoto

461

(2011). The jobs created at the biorefinery sugarcane milling block and CS sugar mill were

462

estimated based on the sugarcane processing sector employment capacity of Sao Paulo.

463

Moraes et al. (2007) projected 75300 employees in the sugar production industry of Sao

464

Paulo state by 2020/21 for a total sugarcane production of 544 million tons. One estimates

465

0.14 jobs created per 1000 tons of sugarcane processed. Furthermore, the number of

466

employees required in the rest of the biorefinery plant was considered to be similar for all

467

the scenarios and it was taken from NREL (Humbird et al., 2011). Moreover, 80% of the oil

468

refining products in Brazil are produced by Petrobras, 96 082 formal of jobs are associated

469

with the oil refining chain based on the company employment data (Petrobras, 2016; ANP,

470

2015). The number of jobs was allocated to gasoline product. The allocation factor is

471

proposed as the ratio between the total volume of gasoline produced and the total volume

472

of petroleum consumed in Brazil in one year: 30 078 550 m3 and 122 263 477 m3,

19

473

respectively (Sindicom, 2016). Considering such allocation factor (0.246), 23 638 jobs are

474

associated with gasoline, i.e., 0.00079 jobs/m3 gasoline per year. The total employment

475

creation results are presented in Table 1. Job creation is mainly influenced by the system

476

sugarcane requirements. Moreover, the largest number of jobs is created at the sugarcane

477

plantation and transportation from the field to the plant. In ED systems the sugar fractions

478

are used for fuel production, so, complementary sugar and molasses require additional

479

sugarcane feedstock leading to a higher job demand in the plantation in ED systems. Same

480

reason for the higher employment creation of OF against FF systems. The oil refining

481

industry employs less people than the sugarcane industry.

482

3.4.

483

Climate change (i6). As shown in Fig. 4-a, the largest climate change impacts are

484

associated with the mobility services supply. The biofuel product impacts are estimated

485

from 0.60 kg CO2 eq. /L (SM FF) to 0.75 kg CO2 eq. /L (ED FF), lower than the

486

complementary fossil-based fuel (gasoline) climate change impact of 3.02 kg CO2 eq. /L.

487

Therefore, the scenarios with higher supply of gasoline (larger CS requirements) contribute

488

to higher climate change impact. The complementary sugar has a considerable impact on

489

the ED schemes emissions. The complementary sugar emissions per unit are higher than

490

the biorefinery sugar emissions– 0.45 kg CO2 eq. /kg vs. 0.31 kg CO2 eq. /kg. Likewise, the

491

supply of electricity from the grid has an impact of 0.21 kg CO2 eq. /kWh, much higher than

492

the impact of the electricity produced in the biorefinery scenarios, estimated to be in the

493

range 0.04-0.08 kg CO2 eq. /kWh. Fossil fuel depletion (i7). SM scenarios require more

494

complementary gasoline than ED scenarios do, leading to higher fossil fuel depletion (Fig.

495

4-b). Gasoline has an impact of 1.01 kg oil eq. /L, which corresponds to approximately four

496

times the biofuel impacts – estimated as 0.21-0.25 kg oil eq. /L, depending on the scenario.

497

ED schemes levels of fossil fuel depletion are largely associated to the consumption of

Evaluation of the environmental indicators

20

498

gasoline for the production of the blend. Food and feed services production require

499

agricultural and other activities subjected to fossil fuels, logically leading to impacts on fossil

500

fuel depletion. LCA shows that the biorefinery sugar represents half of the impact in

501

comparison with the conventional sugar – 0.08 kg oil eq. /kg vs. 0.17 kg oil eq. /kg.

502

Moreover, C molasses and C5 syrup biorefinery products present far lower impacts than the

503

reference cane molasses in Brazil do – 0.01 and 0.06 kg oil eq. /kg TSS vs. 0.18 kg oil eq.

504

/kg TSS. The lower impacts of the biorefinery products in comparison with the impacts of

505

conventional products are explained by reduced allocation factors for sugar and molasses

506

products due to value-based allocation procedure applied throughout the whole biorefinery.

507

In addition, the reference mill uses sulphur dioxide to lighten the colour of the sugarcane

508

juice while the biorefinery employs only lime for purification purposes. The biorefinery

509

electricity represents half of the impact compared to the grid electricity – 0.02 kg oil eq.

510

/kWh vs. 0.04 kg oil eq. /kWh. This is due to the fact that the electricity matrix of Brazil

511

includes 10% of fossil fuels. Freshwater eutrophication (i8). ED scenarios present the

512

largest freshwater eutrophication impact due to the bio-based production of fuels and the

513

consequent higher sugarcane requirements, leading to larger consumption of fertilizers

514

(Fig. 4-c). One cubic meter of biofuel represents an impact of 0.11 kg of phosphate

515

equivalent, which corresponds to double of gasoline impact. Sugar represents an impact of

516

0.09 kg P eq. /kg if it is produced within the biorefinery and 0.17 kg P eq. /kg if it is coming

517

from a conventional sugar mill. Similarly, the impacts of C molasses and C5 syrup are

518

approximately 0.01 kg P eq. /kg TSS, while reference cane molasses represent 0.18 kg P

519

eq. /kg TSS. Finally, 1 MWh of electricity supplied corresponds to an impact of 0.01 and

520

0.05 kg P eq. if it is from the biorefinery or from the grid, respectively. Freshwater

521

consumption (i9). In a Brazilian conventional sugar mill, approximately 10% of the water

522

required by the plant is freshwater, while the rest of the water demand is continually

21

523

recycled (Saad et al., 2010). About 0.71 m3 of freshwater is required per ton of sugarcane

524

processed in a conventional sugarcane mill. The consumption of freshwater was calculated

525

for the amount of sugarcane required to produce the complementary sugars. Concerning

526

the oil refining industry, Nacheva et al. (2011) reports an average consumption of

527

freshwater of 1.5 m3/ton of raw petroleum processed. An oil refining platform presents a

528

wide range of products besides gasoline. The consumption of freshwater associated with

529

gasoline production was estimated to be 1.53 m3 of freshwater/ton of gasoline using

530

allocation based on production volumes and data from ANP (2015) and Sindicom (2016).

531

Large differences are observed between ED and SM scenarios (Table 1). SM scenarios

532

demand higher quantities of freshwater than the ED ones do, due to the water evaporation

533

for the production of thick juice. In the ED scenarios, the whole sugarcane juice is sent to

534

the fermentation, useful for achieving the adequate dilutions along the process. In ED OF,

535

the continuous feedstock water input stream is sufficient to operate the plant while in ED FF

536

there is the need for supplying extra freshwater due to the water losses throughout the

537

three-effect evaporator for C5 syrup production. Concerning the CS from the conventional

538

sugar mills, higher quantities of complementary sugar and molasses are provided to the ED

539

systems, meaning larger freshwater requirements at the mills. Moreover, the oil refining

540

process is more optimized in terms of water consumption. Thus, the oil industry freshwater

541

requirements represent only 4-11% of the total needs. Use of chemicals (i10). As depicted

542

in Table 1, the OF scenarios have higher consumption of chemicals due to the pretreatment

543

technique selected. OF scenarios lignocellulosic biomass pretreatment is conducted with

544

DA hydrolysis technique, efficient for hemicellulose solubilisation, but which creates harsh

545

acid conditions. Thus, DA is followed by an alkaline treatment with ammonia in order to

546

increase the pH of the hydrolysate stream, which facilitates the solubilization of the lignin

547

fraction. The addition of ammonia leads to the formation of SOX. Furthermore, OF scenarios

22

548

require the use of considerable amounts of NaOH in order to rise the pH of the wastewater

549

streams. Finally, lime is required in large quantities in the cogeneration unit of the OF

550

biorefineries in order to remove from flue gas the SOX formed during pretreatment. In the

551

FF scenarios, the chemical free LHW technique was chosen since C5 syrup is an end-

552

product for food applications. Such technique has no neutralisation requirements and low

553

formation of degradation products, leading to null consumption of ammonia and sodium

554

hydroxide. Still, in FF scenarios there is a small amount of lime required for removing the

555

SOX gases that are released during the combustion of biomass. Note that a fraction of the

556

SOX in the OF schemes derives also from the combustion camera, still, this amount is

557

reduced in comparison with to the SOX coming from the pretreatment unit.

558

3.5.

559

The selected biorefinery systems were proved to be economically feasible based on

560

techno-economic assessment. PROMETHEE II has been applied to obtain the outranking

561

of the four biorefinery systems based on the defined sustainability metrics. The base-case

562

indicator weights have been defined based on an egalitarianist attitude, as per section

563

2.1.2. A larger relative importance has been attributed to the indicators i3, i4, i5, i6, i7 and i8,

564

which are related to risk aversion (i3), social responsibility (i4 and i5) and global

565

environmental concerns (i6, i7 and i8) in a long-term time perception (12% each). Moreover,

566

an egalitarian stakeholder gives a lower priority to the local environmental indicators i9 and

567

i10 (8% each), and to the two economic indicators i1 and i2 (6% each). The results of the

568

multi-criteria assessment show that ED OF system presents the highest sustainability

569

performance of all schemes (Fig. 5). ED OF comprises the largest bioethanol production

570

among the scenarios and smallest gasoline requirements, leading to advantages in terms of

571

costs and emissions (Fig. 5). Additionally, in ED OF complementary food and feed services

572

are entirely provided by conventional sugar mills in Sao Paulo, bringing a clear advantage

Multi-criteria comparison and discussion

23

573

in terms of extra employment creation. ED FF is ranked second in the sustainability

574

performance outranking (Fig. 5). Introducing the manufacturing of C5 syrup along with

575

bioethanol leads to higher costs of production, lower profitability and larger economic

576

sensitivity. The production cost of bioethanol in ED FF increases compared to ED OF,

577

which is justified by the ED OF economies of scale. Thus, ED FF presents also higher

578

ethanol cross-subsidisation. As explained before, the ED FF cogeneration unit generates

579

more steam leading to a lower electricity production when compared with ED OF.

580

Therefore, ED FF scenario has lower electricity surplus and reduced contribution to energy

581

security compared to ED OF scenario. Moreover, employment creation levels are

582

significantly influenced by the sugarcane production scale. Thus, once C5-based feed is

583

produced in the ED FF biorefinery, no additional sugarcane is required to produce the

584

complementary feed and, therefore, employment creation is reduced. In comparison to ED

585

OF, larger climate change and fossil fuel depletion impacts are verified in ED FF due to the

586

larger requirements for complementary fossil-based mobility services. In addition, ED FF

587

needs larger amounts of freshwater in order to compensate the water lost in the C5 syrup

588

production unit. Moreover, the study shows that both SM systems are less sustainable than

589

the ED systems, being ranked at 3rd and 4th places (Fig. 5). Besides exhibiting comparable

590

cost of services and lower cross-subsidisation, both SM schemes present higher sensitivity

591

to product price volatility than ED schemes do, due to the larger number of products.

592

Furthermore, while in ED systems the entire food services are supplied by conventional

593

sugar mills, in SM systems sugar is produced within the biorefinery, reducing the need for

594

extra sugarcane which negatively affects employment creation. Additionally, both SM

595

systems present lower biofuel production capacities, implying higher gasoline consumption

596

and, therefore, larger climate change and fossil fuel depletion impacts. Within the SM

597

systems, SM FF presents the most unfavourable indicator results, revealing to be the least

24

598

sustainable scenario. Fig. 5 shows that the ED FF and SM OF schemes are placed in the

599

middle of the rank between the contrasting configurations providing two and five products

600

(ED OF and SM FF, respectively). One analysed the influence of a weighting profile

601

variation on the sustainability performance ranking. A sensitivity analysis was performed by

602

implementing variations in the base-case indicator weights, remaining in an egalitarianist

603

stakeholder profile (represented by the vertical segments in Fig. 5). First of all, it has been

604

shown that the first (ED OF) and last (SM FF) positions remain unaltered. However, the two

605

intermediate biorefinery schemes can switch positions in the ranking. In fact, once a slightly

606

higher priority is given to the economic indicators i1 and i2 (>7%) ED FF becomes

607

dominated by SM OF. This is justified by the fact that the latter has larger economic

608

performance (reduced cost of services and ethanol cross-subsidisation).

609

4. Conclusions

610

Four sugarcane-based biorefineries-centered systems were compared from socioeconomic

611

and environmental criteria through a novel methodology that includes a multi-criteria

612

analysis method. The biorefineries systems are subject to use the same amount of

613

sugarcane and provide the same quantity of services to the local community. The ED OF

614

biorefinery system produces the highest amount of bioethanol and shows the highest

615

sustainability performance. Compared to the previous study, the new methodology values

616

more the environmental performance of the ED OF biorefinery mainly due to the

617

complementary fossil fuel required by the three other biorefinery systems.

618

Acknowledgments

619

This work was co-funded by the Swiss National Foundation in the framework of

620

ENERCHEMS project. Data from the European project ProEthanol2G were used as well.

25

621

Appendix A. Supplementary data

622

Supplementary material associated with this article can be found, in the online version, at

623

introduce link. Appendix A comprises data on the services provided, allocation factors;

624

biorefinery costs; assumptions and sustainability metrics intermediate results

625

(supplementary file 1); and multi-criteria analysis (supplementary file 2).

626

References

627

1. ANP, 2015. Brazilian Statistical Yearbook of Petroleum, Natural Gas and Biofuels 2015.

628

Brazil's National Agency of Petroleum, Natural Gas and Biofuels (ANP) (in Portuguese).

629

2. Azapagic, A., 2014. Sustainability considerations for integrated biorefineries. Trends in

630

Biotechnol. 32, 1-4.

631

3. Brans, J.P., De Smet, Y., 2016. PROMETHEE Methods, in: Greco, S. et al. (Eds.).

632

Multiple Criteria Decision Analysis: State of the Art Surveys. Springer, 2nd Ed., pp. 187-219.

633

4. Brans, J.P., Vincke, P., 1985. A Preference Ranking Organization Method: The

634

PROMETHEE Method for Multiple Criteria Decision-Making. Manag. Science. 31, 647-656.

635

5. Cherubini, F., Jungmeier, G., 2010. LCA of a biorefinery concept producing bioethanol,

636

bioenergy, and chemicals from switchgrass. Int. J. Life Cycle Assess. 15, 53-66.

637

6. Costa, C.C., Guilhoto, J.J.M., 2011. Expected growth of sugarcane industry and impact

638

on the Brazilian economy: 2015 and 2020. SSRN Electronic Journal.

639

7. Dale, V. H., Efroymson, R. A., Kline, K. L., Langholtz, M. H., leiby, P. N., Oladosu, G. A.,

640

Davis, M.R., Downing, M. E., Hilliard, M. R., 2013. Indicators for assessing socioeconomic

641

sustainability of bioenergy systems: A short list of practical measures. Oak Ridge National

642

Laboratory. Ecol. Indic. 26, 87-102.

643

8. De Jong, E., Jungmeier, G., 2015. Biorefinery concept in comparison to petrochemical

644

refineries, in: Pandey, A., Höfer, R., Taherzadeh, M., Nampoothiri, M., Larroche, C., (Eds),

645

Ind. Biorefineries and White Biotechnology. Elsevier, pp. 3-33.

646

9. De Jong, E., van Ree, R., 2009. Adding value to the sustainable utilization of Biomass.

647

IEA Bioenergy T.42.

648

10. FitzPatrick, M., Champagne, P., Cunningham, M.F. and Whitney, R.A., 2010. A

649

biorefinery processing perspective: treatment of lignocellulosic materials for the production

650

of value-added products. Bioresour. Technol. 101, 8915-8922.

651

11. Gnansounou, E., Dauriat, A., 2010. Technoeconomic Analysis of Lignocellulosic

652

Ethanol. Bioresour. Technol. 101, 4980-4991. 26

653

12. Gnansounou, E., Pandey, A., 2017. Classification of biorefineries taking into account

654

sustainability potentials and flexibilities, in: Life Cycle Assessment of Biorefineries. Elsevier.

655

13. Gnansounou, E., Vaskan, P., Ruiz Pachón, E., 2015. Comparative techno-economic

656

assessment and LCA of selected integrated sugarcane-based biorefineries. Bioresour.

657

Technol. 196, 364-375.

658

14. Gupta, N., Fischer, A.R.H., Frewer, L.J., 2011. Socio-psychological determinants of

659

public acceptance of technologies: A review. Public Underst. Sci. 21, 782-795.

660

15. Hofstetter, P., Baumgartner, T., Scholz, R.W., 2000. Modelling the Valuesphere and

661

Ecosphere: Integrating the Decision Maker Perspectives into LCA. Int. J. LCA, 5, 161-175.

662

16. Huijts N.M.A., Molin, E.J.E., Steg, L., 2012. Psychological factors influencing

663

sustainable energy technology acceptance. A review-based comprehensive framework.

664

Renew. Sustainable Energy Rev. 16, 525-531.

665

17. Humbird, D., Davis, R., Tao, L., Kinchin, C., Hsu, D., Aden, A., Schoen, P., Lukas, J.,

666

Olthof, B., Worley, M., Sexton, D., Dudgeon, D., 2011. Process Design and Economics for

667

Biochemical Conversion of Lignocellulosic Biomass to Ethanol – Dilute-Acid Pretreatment

668

and Enzymatic Hydrolysis of Corn Stover. NREL.

669

18. IEA, 2016. Energy security. International Energy Agency (IEA). Access 19-08-2016.

670

.

671

19. Kamm B., Kamm M., 2004. Biorefinery-Systems. Chem. Biochem. Eng. Q.18, 1-6.

672

20. Keeney, R.L., Raiffa, H., 1993. Decisions with Multiple Objectives – preferences and

673

value tradeoffs. Cambridge University Press, Cambridge and New York.

674

21. Larsen, J., Østergaard Petersen, M., Thirup, L., Wen Li, H., Krogh Iversen, F., 2008.

675

The IBUS process – lignocellulosic bioethanol close to a commercial reality. Chem. Eng.

676

Technol. 31, 765–772.

677

22. Luo, L., van der Voet, E., Huppes, G., 2010. Biorefining of lignocellulosic feedstock –

678

Technical, economic and environmental considerations. Bioresour. Technol.101, 5023-

679

5032.

680

23. Lynd, L.R., Wyman, C., Laser, M., Johnson, D., Landucci, R., 2005. Strategic

681

Biorefinery Analysis: Analysis of Biorefineries, 2002. NREL.

682

24. IPCC, 2007: Climate Change 2007: Synthesis Report. Contribution of Working Groups

683

I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate

684

Change. Core Writing Team, Pachauri, R.K and Reisinger, A. (Eds.). IPCC, Switzerland.

685

25. Larsen, J., Haven, M.Ø., Thirup, L., 2012. Inbicon makes lignocellulosic ethanol a

686

commercial reality. Biomass Bioenergy. 46, 36–45. 27

687

26. McBride, A. C., Dale, V. H., Baskaran, L. M., Downing, M. E., Eaton, L. M.,

688

Efroymson, R. A., Garten, C. T., Kline, K. L., Jager, H. L., Mulholland, P. J., Parish, E. S.,

689

Schweizer, P. E., Storey, J.M., 2011. Indicators to support environmental sustainability of

690

bioenergy systems. Oak Ridge National Laboratory. Ecol. Indic. 11, 1277-1289.

691

27. Moraes, M.A.F.D., 2007. The labor market of sugarcane agroindustry: challenges and

692

opportunities. Econ. Aplic. 11, 605-619 (in Portuguese).

693

28. Nacheva, P. M., 2011. Water Management in the Petroleum Refining Industry.

694

Mexican Institute of Water Technology. Intechopen.

695

29. Petrobras, 2016. Petrobras. Access 10-07-2016. < http://www.petrobras.com.br/pt/>.

696

30. Roy B., 1991. The outranking approach and the foundations of ELECTRE methods:

697

Theory and Decision. 31, 49-73.

698

31. Roy, B., 2016. Paradigms and challenges, in: Greco, S., et al. (Eds.), Multiple Criteria

699

Decision Analysis: State of the Art Surveys. Springer, Ed. 2, pp. 19-39.

700

32. Saad, A.M., Signoretti, R. P., Silveira, F. T., Verza, S. S., Mazucatto, J.R., Nakahodo,

701

L. N., Machado, R. E., Marcelino, F. A., 2010. Environmental Impact Report: scale-up of the

702

sugarcane plant Usina Açucareira São Manoel S.A.

703

33. Saaty T.L., 1980. The analytic hierarchy process. New-York, NY: Mc Graw-Hill.

704

34. Saaty T.L., 2008. Decision making with the analytic hierarchy process. Int. J. Serv.

705

Sci. 1, 83-98.

706

35. Sadhukhan, J., Ng, K. S., Martinez-Hernandez, E., 2014. Biorefineries and Chemical

707

Processes: Design, Integration and Sustainability Analysis. Wiley, Chichester, UK.

708

36. Sacramento-Rivero, J. C., 2012. A methodology for evaluating the sustainability of

709

biorefineries: framework and indicators. Biofuels, Bioprod. Bioref. 6, 32-44.

710

37. Sacramento-Rivero, J. C., Navarro-Pineda, F., Vilchiz-Bravo, L. E., 2016. Evaluating

711

the sustainability of biorefineries at the conceptual design stage. Chem. Eng. Res. Design.

712

107, 167-180.

713

38. Schaidle. J. A., Moline, C.J., Savage, P.E., 2011. Biorefinery sustainability

714

assessment. Environ. Prog. Sustain. Energy. 30, 743-753.

715

39. Schwarz S.H., 1999. A theory of cultural values and some implications for work. Appl.

716

Psychol. Int. Rev. 48, 23-47.

717

40. SimaPro, 2016. SimaPro Database Manual – Methods library. April 2016. Access 18-

718

08-2016. .

719

41. Sindicom, 2016. Volume of refined petroleum distillate per Unit of Federation and

720

origin from 2000 to 2016. National Syndicate of Fuel&Lubricant Distributors (in Portuguese). 28

721

42. USDE, 2003. How to calculate the true cost of steam. U.S. Department of Energy.

722

Industrial Technologies Program.

723

43. USDE, 2012. Benchmark the Fuel Cost of Steam Generation. U.S. Dep. Energy.

724

44. Wüstenhagen, R., Wolinski, M., Bürer, M.J., 2007. Social acceptance of renewable

725

energy innovation. An introduction to the concept. Energy Policy. 35, 2683-2691.

726

Captions of figures and tables

727

Fig. 1: Integrated Assessment Methodology.

728

Fig. 2: Sugarcane-based biorefinery centered system. Biorefinery center products,

729

intermediate products, complementary systems and their respective source.

730

Fig. 3: Sugarcane-based biorefineries block diagrams. (a) Ethanol Distillery Only Fuel

731

(ED OF); (b) Ethanol Distillery Fuel and Feed (ED FF); (c) Sugar Mill Only Fuel (SM OF);

732

(d) Sugar Mill Fuel and Feed (SM FF). Operation units in the blocks A1/B1 - feedstock

733

handling, juice extraction, clarification and purification; C1/D1 - feedstock handling, juice

734

extraction, clarification, purification, evaporation and crystallisation. A2/C2 - feed handling,

735

pretreatment, saccharification, fermentation, distillation, dehydration and wastewater

736

treatment; B2/D2 - feed handling, pretreatment, C5 syrup production, saccharification,

737

fermentation, distillation, dehydration and wastewater treatment; A3/B3/C3/D3 - heat and

738

power cogeneration plant.

739

Fig. 4: Life cycle assessment (LCA) as per the individual services. (a) Climate change;

740

(b) Fossil fuel depletion; (c) Freshwater eutrophication. BS – Biorefinery service; CS –

741

Complementary service.

742

Fig. 5: Sustainability performance ranking of the four sugarcane biorefinery-centered

743

systems in the context of Brazil. Ranking position of the scenario () and sensitivity

744

analysis by implementing variations in the base-case weights (vertical segment).

745

Table 1: Sustainability indicators results. Socioeconomic and environmental indicators.

29

746

Tables

747

Table 1

Indicator Socioeconomic i1 Cost of services i2 Cross-subsidisation 1G/1G2G i3 Sensitivity to price volatility i4 Energy security i5 Employment creation Environmental i6 Climate change i7 Fossil fuel depletion i8 Freshwater eutrophication i9 Freshwater consumption i10 Use of chemicals

Units

Target

ED OF

ED FF

SM OF

SM FF

M US $/yr Fraction Fraction MW Jobs

Min Min Min Max Max

421 0.54 0.45 13.5 3883

443 0.67 0.52 7.3 3734

411 0.46 0.54 55.6 2391

447 0.52 0.77 26.6 2241

kton CO2 eq./yr kton oil eq./yr ton P eq./yr 1000 m3/yr kton/yr

Min Min Min Min Min

675 226 156 2188 22.6

762 251 150 2817 0.2

1029 337 74 3336 22.5

1150 368 76 3204 0.1

748

30

749

Figures

750 751

Fig. 1

31

752 753

Fig. 2

32

754 755

Fig. 3

33

756

757

758 759

Fig. 4

34

760 761

Fig. 5

762

35

763 764 765 766 767 768 769 770 771

Highlights    

A new sustainability assessment framework to compare biorefinery systems is proposed. The multi-criteria method relies on sustainability indicators and cultural values. The selected criteria must comprise the most contrasting features among systems. The system with the highest bioethanol production capacity is the most sustainable.

772 773

36