Bioresource Technology 135 (2013) 490–499

Contents lists available at SciVerse ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Potential of bioethanol as a chemical building block for biorefineries:

Preliminary sustainability assessment of 12 bioethanol-based products
John A. Posada ⇑, Akshay D. Patel, Alexander Roes, Kornelis Blok, André P.C. Faaij, Martin K. Patel
Energy & Resources, Copernicus Institute of Sustainable Development, Utrecht University, Budapestlaan 6, NL-3584 CD Utrecht, The Netherlands

h i g h l i g h t s

" Potential of bioethanol as raw material for biorefineries (to bulk chemicals).
" Multi-criteria early-stage sustainability assessment method for bio-based process.
" Screening and selection of bioethanol derivatives based on sustainability.
" Bioethanol derivatives categorization: favorable, promising and unfavorable.
" Sensitivity, scenarios and uncertainty analyses were performed.

a r t i c l e i n f o a b s t r a c t

Article history: The aim of this study is to present and apply a quick screening method and to identify the most promising
Available online 24 September 2012 bioethanol derivatives using an early-stage sustainability assessment method that compares a bioetha-
nol-based conversion route to its respective petrochemical counterpart. The method combines, by means
Keywords: of a multi-criteria approach, quantitative and qualitative proxy indicators describing economic, environ-
Bioethanol derivatives mental, health and safety and operational aspects. Of twelve derivatives considered, five were catego-
Biorefinery analysis rized as favorable (diethyl ether, 1,3-butadiene, ethyl acetate, propylene and ethylene), two as
Biorefinery systems
promising (acetaldehyde and ethylene oxide) and five as unfavorable derivatives (acetic acid, n-butanol,
Ethanol conversion
Sustainability assessment
isobutylene, hydrogen and acetone) for an integrated biorefinery concept.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction
Numerous policies supporting the production and use of trans-
port biofuels have globally been established in the last decade
(Sorda et al., 2010; Lamers et al., 2011) in order to promote energy
independence and mitigate negative environmental impacts
caused by the use of conventional fossil fuels (e.g., gasoline, diesel,
etc.). As a consequence, bioethanol has become the largest biofuel
with an estimated worldwide production of above 100 billion liters
in 2011 and with an expected production increase of around 3–7%
p.a. in the years 2012–2015 (OECD-FAO, 2012). The rapid develop-
ment of the bioethanol market has been accompanied by a growing
interest in its use as renewable feedstock for the manufacture of
bio-based chemicals (Weusthuis et al., 2011; Rass-Hansen et al.,
2007). Moreover, bioethanol was recently identified as one of the
potential top bio-based raw materials for the chemical industry
(Bozell and Petersen, 2010). Significant technological advances
have also been achieved on the catalytic conversion of bioethanol
(e.g., dehydration, dehydrogenation, oxidation, reforming, gasifica-
⇑ Corresponding author. Tel.: +31 30 253 5516; fax: +31 30 253 7601.
E-mail address: J.A.PosadaDuque@uu.nl (J.A. Posada).
tion, decomposition, coupling, etc. (see Table 1)) and many impor-
tant chemicals have successfully been synthesized; some examples
are: ethylene, propylene, 1,3-butadiene, iso-butylene, hydrogen,
acetaldehyde, ethylene oxide, n-butanol, acetic acid, ethyl acetate,
acetone and dimethyl ether. This outlook of growing production
and technological advances for its conversion to bulk chemicals to-
gether with the promise of new technologies for cheap cellulosic
bioethanol, could work as solid pillars to create competitive biore-
fining systems based on bioethanol as chemical building block.
Given the large number of options for bioethanol conversion,
the identification of its most promising derivatives from a sustain-
ability point of view is by no means a trivial task. This type of re-
search question has traditionally been addressed by in-depth
analysis based on conceptual process design (e.g., flowsheeting
and process simulation with Aspen Plus and subsequent eco-
nomic/environmental assessment (Posada et al., 2012)). But such
approaches are very time-intensive and require detailed data re-
lated to downstream processing which are not available at early
design phases (e.g., lab or pilot scale). This is even a more demand-
ing task in the case of integrated biorefineries (i.e., the combined
sustainable production of bio-based chemicals, biofuels, bio-based
polymers, pharmaceuticals, food and/or feed; adapted from

0960-8524/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biortech.2012.09.058
 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499 491

Table 1
Collection data of the considered reactions.

Product Mass and Temp Involved Reference Conventional process Market Main uses Commercial price
(molar) yieldsa (°C) reactionsa (million m.t.) (€/tonne)b
Ethylene 0.60 (0.99) 450 c1, c7, c15, Bi et al. (2010) Steam cracking of 136.2 Raw material 1100–1350
c16 naphtha
Propylene 0.32 (0.35) 550 c1, c2, c3 Song et al. (2010) Steam cracking of 51.7 Raw material 1120–1370
naphtha
1,3- 0.51 (0.44) 350 c7, c13, c14 Ohnishi et al. (1985) Steam cracking of 12.2 Raw material 1845–2255
Butadiene naphtha
IsoButylene 0.34 (0.28) 450 c1, c5, c7 Sun et al. (2011) Steam cracking of 24.4 Raw material 520–640
naphtha
Hydrogen 0.21 (4.79) 450 c6, c17, c18 Liu et al. (2010) Methane steam 54.1 Chemical agent 1745–2135
reforming
Acetaldehyde 0.75 (0.78) 230 c7, c18 Chen et al. (2007) Oxidation of ethylene 3.5 Raw matrial 745–910
Ethylene 0.92 (0.96) 325 c1, c4, Lippits and Nieu- Oxidation of ethylene 20.4 Raw material 1235–1510
oxide wenhuys (2010)
n-Butanol 0.21 (0.13) 400 c1, c5, c7, c9, Tsuchida et al. (2006)) Propylene 3.1 Raw material 920–1125
hydroformilation and solvent
Acetic acid 1.25 (0.96) 150 c10, c18 Jørgensen et al. (2007) Methanol 11.3 Raw material 765–935
carbonylation and solvent
Ethyl acetate 0.63 (0.33) 260 c7, c11, Santacesaria et al. Esterif. of acetic acid 1.7 Solvent and 1010–1230
(2012) with ethanol coating agent
Acetone 0.57 (0.46) 400 c1, c12 Nakajima et al. (1989) Cumene oxidation 6.2 Raw material 920–1125
(Hock process) and solvent
Diethyl ether 0.59 (0.37) 350 c1, c15, Ali et al. (2010) Direct hydration of 14.4 Solvent 1730–2115
ethylene
a
See Fig. 1.
b
For 2008–2011.

Cherubini and Strømman (2011)) due to the technological com-

plexity related to each processing step (e.g., collection and pre-
treatment of biomass, production and separation of precursors,
conversion of precursors to chemical building blocks and produc-
tion of secondary chemicals/products (Cherubini and Strømman,
2011)). Since both biorefinery concepts and bioethanol conversion
alternatives are still under development, it is a challenge to per-
form a preliminarily sustainability assessment of these conversion
systems with the limited information available.1 Against this
background, one of the objectives of this paper is to identify those
bioethanol derivatives that offer largest benefits in sustainability
terms. The other main goal is to explore whether the applied meth-
odology is more suitable to pinpoint the conversion steps that are
particularly attractive for future biorefinery systems. To this end,
* Reactions γ16, γ 17 and γ 18 refer to the production of: ethane, methane and carbon dioxide,
various bioethanol conversion alternatives were compared from a respectively.
sustainability point of view (i.e., covering economic, environmental,
and health and safety aspects) with respect to their petrochemical Fig. 1. Derivatives of bioethanol by catalytic conversion.
counterparts and the plausibility of the results was checked by con-
ducting sensitivity, scenarios and uncertainty analyses. This study
focuses exclusively on the chemical conversion of bioethanol, there-
by excluding other strategies for improved sustainability such as
Screening of either synthesis pathways (i.e., from different raw
choice of crop, cultivation method or location.
materials to one specific product) or conversion alternatives (i.e.,
from one specific raw material to different products) is one of
the first steps related to the design of new chemical conversion
2. Methodology
routes. The task was successfully addressed by Sugiyama et al.
(2008) who developed a multi-objective decision framework
After a comprehensive literature survey on bioethanol conver-
which allows to screen processes based on information that is
sion (more than 200 papers), 12 final derivatives were identified
available at early design stages. The inputs required for the deci-
as candidates to be analyzed as shown in Fig. 1. The highest re-
sion framework were: reaction mass balances, prices of raw mate-
ported yields and other important information related to the reac-
rials and products, life-cycle environmental impacts represented
tive systems and markets are shown in Table 1.
by cumulative energy demand (CED) and greenhouse gas (GHG)
emissions of the feedstocks, physical–chemical properties of
reactants and products, and hazards. This framework was later
1
Most of the sustainability assessment approaches and indicators have been modified by Patel et al. (2012) in order to better account for
developed for bio-energy systems. However, these methods could in principle be circumstances related to the production of bio-based chemicals.
extended to bio-based products. For instance, this is case of the 24 GBEP (Global
Bioenergy Partnership) sustainability indicators for bioenergy. But this type of
This goal was achieved by including features such as the need for
approaches require in general detailed information which is not available at early biomass pretreatment, distribution of environmental burdens by
design phases. product allocation, number of co-products, risk aspects and
492 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499
EC: Economic constraint; EI: Environmental impact of raw materials; PCEI: Process costs and environmental impacts; EHSI:
Environmental-Health-Safety index; RA: Risks aspects.
Fig. 2. Relationship between the indicators and both traditional economics and contributions to environmental impacts.
process comparison to the petrochemical counterpart. The adapted
method integrates five sustainability indicators namely: economic
constraint (EC), environmental impact of raw materials (EI),
process costs and environmental impact (PCEI), Environmental-
Health-Safety index (EHSI) and risk aspects (RA). These sustainabil-
ity indicators were combined to a single score index by means of
weighting factors allowing a quick comparison between the
process alternatives. The first four indicators were based on the
decision framework developed by Sugiyama et al. (2008) while
the last one was proposed by Patel et al. (2012). All of these com-
plementary indicators are calculated using quantitative informa-
tion (i.e., related to raw materials and process conditions) and
qualitative indicators that reflect the sustainability of the process.
For instance, aspects of classical economics and contributions to
environmental impacts are covered by the five indicators as shown
in Fig. 2, thereby avoiding overlap and double counting. A descrip-
tion of these indicators, the respective weighting factors and the
way to both calculate and integrate them in a single score index
is provided below.
Economic constraint (EC) was defined as the ratio of the total
costs of raw materials going into the process related to the value
of all the marketable products and co-products leaving the process,
as represented by Eq. (1) in Table 2. This index is the simplest
(however incomplete) approach to evaluate the economic viability
of a chemical process which is also the primary practical concern of
any new project. A lower ratio means a higher economic potential
because the raw materials represent a smaller share of the value of
all the marketable products.
In the case of multi-product processes economic allocation (EA)
was applied instead of mass allocation, in order to both ensure a
fairer assessment in line with the product values and avoid over-
assignment of the overall process costs to a low-value by-product.
The allocation factor was calculated according to Eq. (1.1) in Table
2.
Environmental impact of raw materials (EI) were determined
based on two parameters: CED and GHG (CO2 eq.) emissions as
shown in Eq. (2) in Table 2. CED represents the total energy
requirements from cradle to factory gate thereby covering total
fossil and renewable energy inputs for feedstocks production.
GHG emissions are primarily associated to the use of fossil
resources for feedstocks production. The values used include fossil
carbon embedded in the product, following a cradle-to-grave ap-
proach. This approach was applied based on the assumption that
the embedded carbon would be released at a later point in time
by either waste incineration or by the action of micro-organisms
in the case of organic chemicals.
The indicator process costs and environmental impacts (PCEI),
an extended version of the indicator energy loss index (ELI) devel-
oped by Sugiyama et al. (2008), represents the expected costs and
process energy use next to the emissions associated to the reaction
and separation section. The PCEI (see Eq. (3) in (Table 2)) combines
7 different categories, namely: presence of water at reactor outlet,
molar concentration of the main product at reactor outlet, mini-
mum difference of boiling point between the main product and
the substances at the reactor outlet, inherent reaction mass loss
(mass loss index), reaction enthalpy (heating or cooling duty),
number of co-products and pre-treatment of feedstock (as ex-
plained by Patel et al. (2012), they added the latter two indicators
to the set developed by Sugiyama et al. (2008). These categories
were scored between 0 and 1 for low and high impacts, respec-
tively and they were used to ‘‘mimic’’ the energy use and the re-
lated emissions of the process without developing a detailed
model based on engineering principles (e.g., flowsheeting with As-
pen Plus). For the reaction enthalpy category an additional scale
between 1 (high) and 0 (low energy recovered) was integrated
for high-temperature heat recovered from exothermic reactions.
The scaling of the seven used categories is shown in Table 2 (Eqs.
(3.1)–(3.7)). For processes with multiple reaction steps, a separate
PCEI must be calculated for each reaction and separation step, gen-
erating an aggregated PCEI score for the whole process (Patel et al.,
2012).
Environmental-Health-Safety index (EHSI) represents a proxy
measure of the EHS characteristics of a chemical process. This indi-
cator considers three categories and ten subcategories for a chem-
ical process. Persistency (half-life in water), air hazard (index value
of chronic toxicity), water hazard (L(E)C50 aquatic, R-codes) and so-
lid waste (based on substance class) were included in the environ-
ment category, while irritation (EU-class, R-codes, LD50dermal) and
chronic toxicity (EU-class, GK, R-codes) were covered by the health
category. Mobility (Partial pressure, Boiling point), fire/explosion
 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499 493

Table 2
System of equations and weighting factors.

Indicator Equation Eq. Internal Global

No. weight factor weight factor
Pr RM RM
Pp
Economic constraint EC ¼ i¼1 mi C i = mP C P
j¼1 j j
(1) 1.0 0.3
P P
Pp P P
AF n ¼ mn C n = j¼1 mj C j (1.1) NA NA

Environmental impact of EIRM ¼ CEDn þ GHGn (2) 1.0 0.2

P
Raw material CEDn ¼ AF n ð1=mPn Þ ri¼1 mRM RM
i CEDi
(2.1) 0.5
Pr
GHGn ¼ AF n ð1=mn Þ i¼1 mi GHGRM
P RM
i
(2.2) 0.5
PPCEIc
Processing cost and PCEI ¼ i¼1 IWF i  PCEIi (3) 1.0 0.2
environmental impacts PCEI1 ¼ 0:0; if water is NOT present (3.1) 0.143
0:5; if water IS present
1:0; if water must be distilled
PCEI2 ¼ 1  ð1=2Þðlog5 ð100  C n ÞÞ (3.2) 0.143
PCEI3 ¼ 1  ð1=2Þðlog2 ðDT bp =5ÞÞ (3.3) 0.143
PCEI4 ¼ ð1=2Þðlog10 MLI þ 1Þ (3.4) 0.143
PCEI5 ¼ ðjDH0Rxn j  100Þ=200; if : DH0Rxn P 0 (3.5) 0.143
or if : DH0Rxn < 0 and T R < 200O C
þð100  jDH0Rxn jÞ=200; when : DH0Rxn < 0and T R > 200O C
PCEI6 ¼ 0:015  N 2cp þ 0:28  N cp  0:25 (3.6) 0.143
PCEI7 ¼ 0; if feedstock pretreatment is NOT required (3.7) 0.143
1; if feedstock pretreatment IS required
Hazard Index EHSI ¼ AF n  ðIWF EH  EH þ IWF HH  HH þ IWF SH  SHÞ (4) 1.0 0.2
P P P P
EH ¼ ECat i ðZ  maxðmFi ÞIECat Þ þ Ecat j ðmOut ECat (4.1) 0.4
F
i j Ij Þ
P
HH ¼ HCat maxðmi Ii ÞUN HCat (4.2) 0.2
i
P
SH ¼ SCat maxðmaxðmFi ISCat ÞÞ (4.3) 0.4
i
F i

Risk aspects RA ¼ IWF GFA  GFA þ IWF LFP  LFP þ IWF MS  MS þ IWF CCI  CCI þ IWF IB  IB (5) 1.0 0.1
GFA ¼ 0:0; for Large scale availabilityðCommodity chemical or fuelÞ: (5.1) 0.25
0:5; for Potential for near term bulk availability:
1:0; for Conceptual feedstockðNeeds fundamental developmentÞ:
LFP ¼ 0:0; for Feedstock is locally available in bulk quantities: (5.2) 0.15
0:5; for Feedstock available in other parts of the world in free and open markets:
1:0; for Feedstock primarily available in regulated markets with limited global market access:
MS ¼ 0:00; for Existing bulk chemical=fuel market: (5.3) 0.25
0:33; for Existing commodityðex:Lacticacid; levulinic acidÞ:
0:66; for Near term bulk chemical=fuel market potential:
1:00; for Long term market potential; possibly accelerated by interesting properties:
CCI ¼ 0:00; Process can be integrated or retrofitted into existing processing infrastructure; (5.4) 0.20
and the existing target product enters existing processing and supply chains:
0:33; New processing plants required based on known technologies;
and the existing target product enters existing processing and supply chains:
0:66; New processing plants required based on known technologies;
and new target product which would need new processing and supply chains:
1:00; New greenfield process plants built with new technologies;
and new target product which would need new processing and supply chains:
IB; (5.5) 0.15 or 0.15
Chemicals;
Functional groups;
¼ 0:0; Between2and4functional groups:Platform molecule:
¼ 0:5; More than 4 functional groups:Difficult platform molecule to work with:
¼ 1:0; One functional group:Limited potential for platform chemical:
Retention of raw material functionality;
¼ 0:0; Complete functionality is preserved:
¼ 0:5; Limited modification of functionality:
¼ 1:0; All functionality stripped off :
or
Fuels;
Energy density;
¼ 0:0; High energy density; more than or equivalent to gasoline
¼ 0:5; Energy content80  90%that of gasoline:
¼ 1:0; Energy content below80%of gasoline:
Engine compativility;
¼ 0:0; Perfectly compatible:Gasoline=Diesel equivalent:
¼ 0:5; Potential for use in existing engines in mixture with gasoline:
¼ 1:0; Engine modification necessary for use:

Note: AF: allocation factor, C: commercial price or cost (€/kg), c: concentration (mol/mol), CCI: compatibility with current infrastructure, CED: cumulative energy demand (MJ/
kg), EC: economic constraint, ECat: specific category of environmental hazards, EH: environmental hazards, EHSI: Environmental-Health-Safety Index, EIRM: environmental
impact of raw materials, GFA: global feedstock availability, GHG: greenhouse gas (kg CO2 eq/kg), HCat: specific category of health hazards, HH: health hazards, I: index value of
one specific component for either category (EH, HH or SH), IB: inherent benefits, IWF: internal weight factor, LFP: local feedstock potential, m: mass flow (kg/kg product),
L(E)C50 aquatic: aquatic lethal or effect concentration using daphnia magna, MLI: mass loss index, calculated as the ratio of the total mass of all components in the reactor outlet
except for the main- and co-products, to the mass of main and co-products from the reaction, MS: market size, Ncp: number of co-products, PCEI: process costs and
environmental impacts (1 to 7, see methodology), PCEIc: specific category of process costs and environmental impacts, p: number of products, r: number of raw materials, RA:
risk aspects, SCat: specific category of safety hazards, SH: safety hazards, TR: temperature of reaction (°C), Z: fraction of mass emitted to the environment in case of an accident
from the maximum mass present in the overall process (=0.1), DH0rxn: standard enthalpy of reaction (kJ/mol), DTbp = smallest absolute difference between the boiling point of
the product and another substance that has to be separated from this product (°C). Sub-indexes: i, j, n: counter for species i, j and main product; cp: number of co-products. Super
indexes: F: refers to a any internal stream in the process, Out: refers the stream leaving the process (contains main product and co-products), P: refers to the main stream leaving
the process (contains the main product), RM: refers to the stream feeding the process (contains the raw materials), UN: refers to a unit mass stream (i.e., 1 kg of each substance).
494 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499
(Flash point, R-codes), reaction/decomposition (NFPA reactivity, R-
codes) and acute toxicity (IDLH, EU-class, GK, R-codes) were incor-
porated in the safety category. For each subcategory an index value
ranging from 0 to 1 was assigned to each chemical present in the
system based on mass flows and specific indices as shown by
Eqs. (4.1)–(4.3) in Table 2 (a detailed description is given by Sugiy-
ama (2007)). Here, economic allocation was again applied to dis-
tribute the process hazards among the main product and co-
products.
Risks aspects (RA) represents an indicator based on both exter-
nal economic aspects and technical aspects related to the process,
which can play a crucial role in the practical implementation of
new bio-based processes (Patel et al., 2012). Five categories were
considered (i.e., global feedstock availability, local feedstock poten-
tial, market size, compatibility with current infrastructure and
Inherent benefits (chemicals or fuels)) and assessed based on qual-
itative phrases as shown by Eqs. (5.1)–(5.5) in Table 2.
Aggregation of these five indicators into a single evaluation in-
dex was performed by normalizing the various scores and using
weighting factors as shown in Table 2. Selection of these weighting
factors were based on expert opinions (Patel et al., 2012) and sup-
ported by the following reasoning: economic feasibility is the first
requirement to implement a process on a commercial scale; long-
term sustainability should be complemented by minimization of
environmental impacts; short-term or immediate hazards associ-
ated with the process should be considered; and risk aspects could
potentially be crucial for decision makers. Therefore, cost-related
aspects (i.e., raw materials cost, products price and processing
costs) accounted for 40% of the total score (30% from EC and 10%
from PCEI) while environmental impacts had an effective overall
weight of 30% (20% from EI2 and 10% from PCEI). EHSI which repre-
sented 20% of the final score has been described as extremely impor-
tant especially in a social context (Patel et al., 2012). Only 10% was
assigned to risk aspects due to the high uncertainty in quantifying
the effects of these parameters.
Each one of the five indicators was normalized by the maximum
(i.e., the worst) score of the two processes under comparison (i.e.,
bio-based vs. petrochemical). The maximum possible normalized
value is 1 and this was obtained for the processing option with a
higher indicator score (less attractive) while the alternative pro-
cess received accordingly a lower normalized value (more attrac-
tive). The normalized values for each indicator were added up
using the weighting factors presented in Table 2 and as result a sin-
gle index (named total score) was obtained for both processes, the
bio-based and the petrochemical. The ratio of these two scores of-
fered a final direct comparison. An index ratio <1 indicates that the
bio-based process can provide benefits over the petrochemical
counterpart in terms of sustainability, while the opposite is true
for an index ratio >1. In addition, specific advantages can be iden-
tified by a direct comparison of a contribution analysis on the five
indicators.
Additionally, sensitivity, scenario and uncertainty analyses
were performed in order to account for variation of data inputs
that change over time (e.g., yields, prices, CED, GHG emissions,
etc.) or with people’s opinions (i.e., weighting factors). The sensi-
tivity analysis was conducted by varying both the yield of the main
product i.e., reduction by 20% and theoretical yield, and the com-
mercial price of bioethanol. For the scenario analysis four different
countries/regions were compared, i.e., EU, USA, Brazil and China,
using country-specific data for CED (Jungbluth et al., 2007), GHG
emissions (Jungbluth et al., 2007) and prices of bioethanol (based
on historical data from ICISpricing (2011)). These three parameters
2
An equal contribution to the costs and the environmental impacts (i.e., 10% each)
is here assumed.
vary significantly for each region due not only to local conditions
but also to the different feedstocks (Hoefnagels et al., 2010). Thus,
bioethanol obtained from sugar beets, corn, sugarcane and sweet
sorghum were considered for EU, USA, Brazil and China respec-
tively. In addition, two hypothetical scenarios of integrated biore-
fineries (i.e., the bioethanol production and subsequent conversion
processes are coupled) were studied for today and for 2020. The
bioethanol production costs in each country/region, previously re-
ported by Balat and Balat (2009), Crago et al. (2010), Dias et al.
(2011), Li and Chan-Halbrendt (2009), were updated using the
Consumer Price Index (CPI) (El-Halwagi, 2011). For integrated bior-
efineries in 2020, the bioethanol production costs were estimated
by Hettinga et al. (2009), van den Wall Bake et al. (2009) and
Junginger et al. (2010) based on experience curves for the entire
production chain (i.e., feedstock costs and industrial production
costs). Low and high prices of bioethanol in 2020 were estimated
considering a low growth rate (5% growth/year) and a high growth
rate (8% growth/year) respectively, with a learning rate of 30% (or a
progress ratio of 70%).
Finally, the variability related to the subjective choice of the
weighting factors was examined with an uncertainty analysis
using the Monte Carlo technique. For this analysis, 10,000 different
randomly generated weighting sets were specified for each prod-
uct within established upper and lower limits. While the upper
limits for EC, CED, GHG, PC&EI, EHSI and RA were set at 0.60,
0.30, 0.30, 0.35, 0.30 and 0.25 respectively, the lower limits were
0.25, 0.05, 0.05, 0.15, 0.05 and 0.05 respectively.
3. Results and discussions
The comparative sustainability assessment of the catalytic syn-
thesis of 12 products from bioethanol (see Table 1 and Fig. 1) was
performed and their integrated scores are shown in Fig. 3. In five
cases, the EC values were lower for the bioethanol-based processes
than for the petrochemical ones, suggesting a potentially attractive
economic margin for these processes. On the other hand, while
CED was higher for all the bio-based products (with the exception
of ethyl acetate), a good potential performance on GHG was no-
ticed in 10 cases. Regarding PCEI, advantageous results were ob-
tained for seven bio-based alternatives compared to their
petrochemical counterpart. In the cases of EHSI and RA, the bioeth-
anol-based synthesis presented better results for three and five
products respectively.
As described above, the index ratio (i.e., total score of bio-based
process/total score of petrochemical process) allows a straight
comparison of both alternatives. Fig. 4 shows the index ratio for
each product with the corresponding sensitivity analyses on yields
and prices. Considering that an index ratio lower than ‘1’ indicates
potential economic and/or environmental benefits for the bio-
based alternative compared to the petrochemical process, three
main groups of cases were distinguished. The first group corre-
sponds to these processes with a significant sustainability poten-
tial for products synthesis from bioethanol, i.e., when the index
ratio is clearly lower than ‘1’ (i.e., <0.9). The second group relates
to processes with an index ratio around ‘1’ (i.e., 0.9–1.2), in which
current advantages are not substantial, but where further techno-
logical improvements or lower ethanol prices could lead to a better
performance for bio-based alternatives. The last group includes the
products with an index ratio higher than ‘1’ (i.e., >1.2), which seem
to offer less benefits than petrochemical production.
From a sustainability point of view, 1,3-butadiene and diethyl
ether were the most favorable compounds to be produced as
compared to the petrochemical route (see Group I in Table 3). These
results were confirmed when the sensitivity analysis was applied to
both yields and bioethanol price. Lower limits in Fig. 4 for yield
 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499
Fig. 3. Total scores and detailed indicators for the sustainability assessment of bioethanol derivatives.
sensitivity correspond to theoretical maximum yield (favorable con-
ditions) while higher limits were for a 20% decrease in the current
bioethanol conversion yield (unfavorable conditions). Upper and
lower limits in Fig. 4 for the cost sensitivity (i.e., unfavorable and
favorable conditions respectively) were represented by a range of
bioethanol price of ±20%. The results of the sensitivity analysis in
Fig. 4 show that the index ratios were very close to those obtained
for theoretical maximum yields which indicated a very limited
scope for further improvement by advanced catalysts. Small differ-
ences were also obtained when the bioethanol price was reduced
by 20%. More importantly, the index ratio of these two products
was always below ‘1’, even when unfavorable conditions were as-
sumed in the sensitivity analysis. Benefits can be achieved by pro-
ducing these two compounds through bioethanol-based process
for three main reasons: (i) because 1,3-butadiene and diethyl ether
are traditionally obtained as high value co-products from petro-
chemical processes (i.e. ethylene and ethanol production respec-
tively), (ii) since the conversion process is more selective for these
two bio-based compounds, lower values of PCEI were found com-
pared to the petrochemical processes indicating potentially lower
processing requirements; and (iii) GHG emissions of the raw mate-
rial due to these bio-based processes are around 50% lower.
Production of ethylene, propylene, ethyl acetate, acetaldehyde
and ethylene oxide, from bioethanol (see Group II in Table 3) must
be analyzed in more detail because their index ratios were very
close to ‘1’ and, in most of the cases, the bandwidth obtained from
the sensitivity analysis contained the value ‘1’. Results obtained for
this second group of bioethanol derivatives indicated that, under
the analyzed conditions, the bio-based processes do not necessar-
ily offer evident benefits with respect to their petrochemical equiv-
alents. But considering that some aspects such as technological
advances in processing, higher prices of petrochemical raw materi-
als and lower bioethanol production costs are very plausible in the
medium term, the index ratios of these five products could signif-
icantly be reduced. Thus, based on this screening method, they
were considered as promising candidates and a subsequent more
detailed analysis is recommended to determine their potential
benefits.
On the other hand, iso-butylene, hydrogen, n-butanol, acetic
acid and acetone production from bioethanol had index ratios
495
much higher than ‘1’ (see Group III in Table 3), suggesting that
these alternatives might not be more sustainable than their petro-
chemical counterparts. In all five cases, the index ratios were above
1.2, even when favorable conditions were considered during the
sensitivity analysis. The low performance of these bioethanol-
based processes are mainly because their ECs were significantly
higher compared to the petrochemical processes. The EC and EI
values could be reduced by considering the use of hydrous ethanol
as raw material because it has a lower price and less life-cycle
environmental impacts (represented by CED and GHG). But only
3 (i.e., hydrogen, isobutylene and acetone) out of the 12 analyzed
products can be synthesized in presence of water. These three
products were in turn found to be the worst cases of the unfavor-
able derivatives group. Although their index ratios were reduced
(by 7.2%, 9.4% and 10.6% respectively) when hydrous ethanol was
considered as raw material, the new values were still above 1.2,
indicating that these products remain in the category of unfavor-
able derivatives. Therefore, these products were discarded as
potential derivatives from bioethanol and their production was
not recommended through the bioethanol-based catalytic route.
Instead, alternative raw materials or processes (e.g., fermentation
or digestion) are recommended for further analysis, unless the
process data of the cases studied can be clearly improved.
The results discussed above were calculated using European
(EU) conditions for: CED, GHG and commercial price of bioethanol.
These parameters were changed in order to analyze the potential
benefits of these bioethanol derivatives in three additional regions:
USA, Brazil and China. The index ratios improved slightly (i.e., they
were reduced) for USA and China, while in the case of Brazil the
reductions were more substantial, i.e. between 7% and 15% (see
Fig. 5a). Although the price of bioethanol in the USA is around
15–20% lower than in China, the index ratios in both cases were
quite similar because of the lower GHG emissions assigned to its
production in China (Jungbluth et al., 2007).
In the case of Brazil a clear potential to develop the industrial pro-
duction of bioethanol-derived chemicals was found since the index
ratios of seven derivatives (i.e., 1,3-butadiene, diethyl ether, ethyl-
ene, propylene, ethyl acetate, acetaldehyde and ethylene oxide)
were below ‘1’; this was the case only for three derivatives in the
EU. The index ratio of these seven derivatives were below ‘1’ even
496 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499
Fig. 4. Index ratio and sensitivity analyses on yields and prices.
Table 3
Categorization of bioethanol derivatives.
GI: Group I, most favorable compounds to be produced from bioethanol.
GII: Group II, promising candidates.
GIII: Group III, unfavorable derivatives from bioethanol.
when unfavorable conditions were considered for the ethanol pro-
duction in Brazil. Thus, this group of products could be considered
as the most favorable compounds to be produced from bioethanol
under Brazilian conditions. For USA and China, sensitivity analyses
under unfavorable conditions (results not shown) led to five favor-
able compounds to be produced from bioethanol: 1,3-butadiene,
diethyl ether, ethylene, propylene and ethyl acetate.
The results discussed so far were based on the assumption that
bioethanol is directly bought from today’s market. In other words,
the concept of a non-integrated biorefinery was assumed. Two fur-
ther scenarios were analyzed: (i) integrated biorefineries today
(i.e., bioethanol is produced and converted at the same facility)
and (ii) integrated biorefineries in 2020 (van den Wall Bake et al.,
2009; Hettinga et al., 2009; Junginger et al., 2010). In the latter
case, two conditions were analyzed: low and high expected pro-
duction cost of bioethanol. The index ratios for these integrated
scenarios, under EU conditions, are shown in Fig. 5b. The results
show that the number of favorable compounds to be produced in
the EU increases from 2 to 7 when the integrated biorefinery was
considered and two further chemicals (i.e., isobutylene and n-buta-
nol) showed slight advantages. In addition, hydrogen and acetic
acid showed potential advantages if the bioethanol production cost
were significantly reduced. Acetone was the only compound that
remained unattractive compared to its petrochemical counterpart
under all considered scenarios. This result suggested that future re-
search should aim at finding alternative routes for its production.
The results were similar for USA, Brazil and China, however with
slightly lower index ratios in all cases (results not shown).
The sensitivity of the results to the weighting factors were stud-
ied by means of an uncertainty analysis using the Monte-Carlo tech-
nique for the EU case. An index ratio distribution was obtained, for
each bioethanol derivative, considering 10,000 possible combina-
tions for the weighting factors. Fig. 6a. shows the results of the
uncertainty analysis, under EU conditions, in which the mean value
as well as the 25% quartile and the 75% quartile are displayed. Abso-
lute differences between the originally obtained index ratio and the
mean value calculated by the uncertainty analysis did not exceed 5%.
These small changes show the reliability of the results and support
further decision making. For instance, the first originally obtained
index ratios and the mean value from the uncertainty analysis were
ranked and compared respectively. Although small differences were
observed on the ranking order, three main groups with the same
components in both cases were identified: (i) most favorable com-
pounds to be produced from bioethanol, (ii) promising candidates
and (iii) unfavorable derivatives from bioethanol. The number of
components belonging to each group changed mainly depending
on whether the biorefinery was integrated or not, but it also changed
for each country/region as shows Table 3.
 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499 497

Fig. 5. Scenario analysis: (a) country/region and (b) integrated-, non-integrated- and 2020- biorefineries.

Fig. 6. (a) Index ratio from the uncertainty analysis and (b) potential beneficial impact of bioethanol derivatives.

In addition, a contribution analysis was performed in order to
check the influence of the environmental and economic indicators
on the final results. When only indicators related to economics (i.e.,
EC and PCEI) were taken into account, four bioethanol derivatives
were identified as favorable compounds and six derivatives as
unfavorable (see Table 4). In the case of environmental indicators
(i.e., EI and PCEI), the same four bioethanol derivatives were cate-
gorized as favorable and another four as unfavorable as shown in
Table 4. Considering both economic and environmental indicators
(i.e., EC, PCEI and EI), only three bioethanol derivatives were iden-
tified as favorable compounds and five as unfavorable. Finally, only
two derivatives were found to be favorable based on all five indica-
tors, which shows that hazard and risk aspects should additionally
be taken into account in order to determine the real potential of a
bio-based product.
The PEP Report 235 (Process Economics Program Report 235,
2007) assessed the production of ethylene, acetic acid and ethyl
acetate from bioethanol based on techno-economic criteria. For
the case of ethylene production, the bio-based process was found
to be cost-competitive with the steam cracking at small plant
capacities, but it was also suggested that significant economic
improvements could be achieved by using the integrated facility
498 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499
Table 4
Contribution analysis to the categorization of bioethanol derivatives.
GI: Group I, most favorable compounds to be produced from bioethanol.
GII: Group II, promising candidates.
GIII: Group III, unfavorable derivatives from bioethanol.
concept. This result is pretty much in concordance with those ob-
tained here, specially for the case of integrated biorefineries where
the index ratio was reduced by 12–19% depending on the specific
region. A similar conclusion was given in the case of ethyl acetate
where the bio-based process was very economically competitive
compared to the conventional esterification process. But for the
acetic acid production from ethanol, the PEP report pointed out
that both capital and production costs of the ethanol-based route
were much higher than the methanol carbonylation process. These
two conclusions are also in line with the main results here ob-
tained, since ethyl acetate was categorized as promising derivative
(group II) where a main contribution to the low index ratio came
from the EC. On the other hand, acetic acid was found as the worst
case with EC being one of the main contributors to the high index
ratio (see Fig. 1).
The analysis presented above refers to one tonne of product as
functional unit, while not factoring in the difference in size of the
world market for the products studied (see Table 1). To do so, we
multiplied the worldwide market of each product by its potential
benefits previously presented. We conducted this calculation for
non-integrated biorefineries using the average index ratio from
the four countries/regions as shown in Eq. (6).
GPBI ¼ ð1  AIRÞ  NGMS
GBPI is the global potential beneficial impact, AIR is the average
index ratio (for the four regions) and NGMS is the normalized glo-
bal market size (dimensionless) for each bioethanol derivative.
Since the first factor indicates potential sustainability benefits of
the bioethanol-based process compared to the petrochemical
counterpart, the GPBI parameter is an indicator of the integral ben-
efits that could be obtained if each derivative were produced from
bioethanol.
The maximum global potential beneficial impacts of bioethanol
derivatives could be obtained for large-scale production of ethyl-
ene (except for EU conditions) and diethyl ether as shown in
Fig. 6b. Significant worldwide benefits are also possible for propyl-
ene (except for EU conditions) and 1,3-butadiene. In contrast espe-
cially, large-scale production of bioethanol-based hydrogen might
lead to adverse effects in terms of sustainability. For this reason,
it is recommended to analyze other alternative ways to produce
hydrogen from renewable resources.
4. Conclusions
The applied method proved to be a promising approach for sus-
tainability assessment and screening of new alternative process
routes typically encountered in novel biorefinery concepts. This
can hence be applied as a first step prior to in-depth analysis of
the most attractive options. Here, attractive opportunities from a
sustainability point of view were found for producing bio-based
chemicals from bioethanol. 1,3-Butadiene and diethyl ether are
the most promising derivatives followed conservatively by ethyl-
ene, propylene, acetaldehyde, ethylene-oxide and ethyl acetate.
On the other hand, iso-butylene, hydrogen, n-butanol, acetic acid
and especially acetone do not seem to represent attractive options
for bioethanol conversion.
Acknowledgements
The financial support from the Netherlands’ Organisation for
Scientific Research (NWO, Nederlandse Organisatie voor Wet-
enschappelijk Onderzoek) in the context of the China-Netherlands
Joint Scientific Thematic Research Programme (JSTP Project
700.10.703, Biorefineries for China and Europe? The Road to Sus-
tainability) is gratefully acknowledged.
References
Ali, T.T., Al-Thabaiti, S.A., Alyoubi, A.O., Mokhtar, M., 2010. Copper substituted
heteropolyacid catalysts for the selective dehydration of ethanol. Journal of
Alloys and Compounds 496, 553–559.
Balat, M., Balat, H., 2009. Recent trends in global production and utilization of bio-
ethanol fuel. Applied Energy 86, 2273–2282.
Bi, J., Guo, X., Liu, M., Wang, X., 2010. High effective dehydration of bio-ethanol into
ethylene over nanoscale HZSM-5 zeolite catalysts. Catalysis Today 149, 143–
147.
ð6Þ Bozell, J.J., Petersen, G.R., 2010. Technology development for the production of
biobased products from biorefinery carbohydrates – the US Department of
Energy’s ‘‘Top 10’’ revisited. Green Chemistry 12, 539–554.
Chen, J., Tang, X., Liu, J., Zhan, E., Li, J., Huang, X., Shen, W., 2007. Synthesis and
characterization of Ag-Hollandite nanofibers and its catalytic application in
ethanol oxidation. Chemistry of Materials 19, 4292–4299.
Cherubini, F., Strømman, A.H., 2011. Chemicals from lignocellulosic biomass:
opportunities, perspectives, and potential of biorefinery systems. Biofuels,
Bioproducts and Biorefining 5, 548–561.
Crago, C.L., Khanna, M., Barton, J., Giuliani, E., Amaral, W., 2010. Competitiveness of
Brazilian sugar cane ethanol compared to US corn ethanol. Energy Policy 38,
7404–7415.
Dias, M.O.S., Cunha, M.P., Jesus, C.D.F., Rocha, G.J.M., Pradella, J.G.C., Rossell, C.E.V.,
Filho, R.M., Bonomi, A., 2011. Second generation ethanol in Brazil: can it
compete with electricity production? Bioresource Technology 102, 8964–8971.
El-Halwagi, M.M., 2011. Sustainable Design Through Process Integration:
Fundamentals and Applications to Industrial Pollution Prevention, Resource
Conservation, and Profitability Enhancement. Elsevier, pp. 448.
Hettinga, W.G., Junginger, H.M., Dekker, S.C., Hoogwijk, M., McAloon, A.J., Hicks,
K.B., 2009. Understanding the reductions in US corn ethanol production costs:
an experience curve approach. Energy Policy (37), 190–203.
Hoefnagels, R., Smeets, E., Faaij, A., 2010. Greenhouse gas footprints of different
biofuel production systems. Renewable and Sustainable Energy Reviews 14,
1661–1694.
ICISpricing, 2011. <http://www.ICISpricing.com>.
Jørgensen, B., Christiansen, S.E., Thomsen, M.L.D., Christensen, C.H., 2007. Aerobic
oxidation of aqueous ethanol using heterogeneous gold catalysts: efficient
routes to acetic acid and ethyl acetate. Journal of Catalysis 251, 332–337.
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., 2007.
Life cycle inventories of bioenergy. Ecoinvent Report No. 17. Swiss Centre for
Life Cycle Inventories, Dübendorf, CH.
Junginger, M., van Wilfried, S., Faaij, A., 2010. Technological Learning in the Energy
Sector, first ed. Edward Elgar Publishing Limited, Cheltenham, UK.
Lamers, P., Hamelinck, C., Junginger, M., Faaij, A., 2011. International bioenergy
trade – a review of past developments in the liquid biofuel market. Renewable
and Sustainable Energy Reviews 15, 2655–2676.
Li, S.-Z., Chan-Halbrendt, C., 2009. Ethanol production in (the) People’s Republic of
China: potential and technologies. Applied Energy 86, S162–S169.
 J.A. Posada et al. / Bioresource Technology 135 (2013) 490–499
Lippits, M.J., Nieuwenhuys, B.E., 2010. Direct conversion of ethanol into ethylene
oxide on copper and silver nanoparticles: effect of addition of CeOx and Li2O.
Catalysis Today 154, 127–132.
Liu, J.-Y., Lee, C.-C., Wang, C.-H., Yeh, C.-T., Wang, C.-B., 2010. Application
of nickel–lanthanum composite oxide on the steam reforming of
ethanol to produce hydrogen. International Journal of Hydrogen
Energy 35, 4069–4075.
Nakajima, T., Tanabe, K., Yamaguchi, T., Matsuzaki, I., Mishima, S., 1989. Conversion
of ethanol to acetone over zinc oxide–calcium oxide catalyst: optimization of
catalyst preparation and reaction conditions and deduction of reaction
mechanism. Applied Catalysis 52, 237–248.
OECD-FAO, 2012. Agricultural Outlook 2011–2020, <http://stats.oecd.org/
Index.aspx>.
Ohnishi, R., Akimoto, T., Tanabe, K., 1985. Pronounced catalytic activity and
selectivity of MgO–SiOrNa20 for synthesis of buta-1,3-diene from ethanol.
Journal of the Chemical Society, Chemical Communications, 1613–1614.
Patel, A.D., Meesters, K., den Uil, H., de Jong, E., Blok, K., Patel, M.K., 2012.
Sustainability assessment of novel chemical processes at early-stage:
application to biobased processes. Energy and Environmental Science 5,
8430–8444.
Posada, J.A., Rincón, L.E., Cardona, C.A., 2012. Design and analysis of biorefineries
based on raw glycerol: addressing the glycerol problem. Bioresource
Technology 111, 282–293.
Process Economics Program Report 235, 2007. Chemicals from Ethanol. <http://
www.ihs.com/products/chemical/technology/pep/index.aspx>.
Rass-Hansen, J., Falsig, H., Jørgensen, B., Christensen, C.H., 2007. Perspective
bioethanol: fuel or feedstock? Journal of Chemical Technology and
Biotechnology 82, 329–333.
499
Santacesaria, E., Carotenuto, G., Tesser, R., Di Serio, M., 2012. Ethanol
dehydrogenation to ethyl acetate by using copper and copper chromite
catalysts. Chemical Engineering Journal 179, 209–220.
Song, Z., Takahashi, A., Nakamura, I., Fujitani, T., 2010. Phosphorus-modified ZSM-5
for conversion of ethanol to propylene. Applied Catalysis A: General 384, 201–
205.
Sorda, G., Banse, M., Kemfert, C., 2010. An overview of biofuel policies across the
world. Energy Policy 38, 6977–6988.
Sugiyama, H., 2007. Decision-Making Framework for Chemical Process Design
Including Different Stages of Environmental, Health and Safety (EHS)
Assessment. Dissertation ETH Nr 17186. ETH Zurich, 2007. At <http://e-
collection.ethbib.ethz.ch/ecol-pool/diss/fulltext/eth17186.pdf>.
Sugiyama, H., Fischer, U., Hungerbuhler, K., Hirao, M., 2008. Decision framework for
chemical process design including different stages of environmental, health,
and safety assessment. AIChE Journal 54, 1037–1053.
Sun, J., Zhu, K., Gao, F., Wang, C., Liu, J., Peden, C.H.F., Wang, Y., 2011. Direct
conversion of bio-ethanol to isobutene on nanosized ZnxZryOz mixed oxides
with balanced acid-base sites. Journal of the American Chemical Society 133,
11096–11099.
Tsuchida, T., Sakuma, S., Takeguchi, T., Ueda, W., 2006. Direct synthesis of n-butanol
from ethanol over nonstoichiometric hydroxyapatite. Industrial & Engineering
Chemistry Research 45, 8634–8642.
van den Wall Bake, J.D., Junginger, M., Faaij, A., Poot, T., Walter, A., 2009. Explaining
the experience curve: cost reductions of Brazilian ethanol from sugarcane.
Biomass and Bioenergy (33), 644–658.
Weusthuis, R.A., Aarts, J.M.M.J.G., Sanders, J.P.M., 2011. From biofuel to bioproduct:
is bioethanol a suitable fermentation feedstock for synthesis of bulk chemicals?
Biofuels, Bioproducts and Biorefining (5), 486–494.