Science of the Total Environment 408 (2010) 5010–5018

Contents lists available at ScienceDirect

Science of the Total Environment

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c i t o t e n v

Greenhouse gases emissions and energy use of wheat grain-based bioethanol

fuel blends
C.C.O. Scacchi a,⁎, S. González-García b, S. Caserini a, L. Rigamonti a
a
Politecnico di Milano, DIIAR-Environmental Section, Piazza L. Da Vinci 32, 20133 Milan, Italy
b
Department of Chemical Engineering, School of Engineering, University of Santiago de Compostela, 15782-Santiago de Compostela, Spain

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

Article history: This study focuses on the potential energetic and environmental impacts associated with the production of
Received 10 March 2010 wheat grain-based bioethanol in Lombardia (Italy), with a “seed-to-wheel” approach (i.e. taking into account
Received in revised form 9 July 2010 the production and use phase). Greenhouse gas emissions (GHGs) were estimated through the CML 2
Accepted 17 July 2010
baseline 2000 methodology counting the CO2 equivalent emissions, while the energy flow indicator was
Available online 7 August 2010
estimated using the Ecoindicator 95 methodology. The impact of the different phases involved in the
Keywords:
production and use of bioethanol have been analysed: the agricultural production of wheat grain, its
Bioethanol transformation into bioethanol, the production of gasoline and the use of 5 different blends (from pure
Greenhouse gas emissions (GHGs) gasoline to pure ethanol).
Life cycle assessment (LCA) The results show that ethanol fuel, used in the form of blends in gasoline, can help reduce energy use and
Lombardia GHGs. In particular, the use of pure ethanol was found to be the best alternative presenting the lowest GHGs
Wheat (saving about 32% of CO2eq emissions in comparison to gasoline) and the minor energy use (63% saving).
Biofuel Differences between low-ethanol blends and gasoline are minimal and dependent on the specific fuel
consumption of the vehicle. The sensitivity analysis performed to test the robustness of results through the
change of some basic assumptions (specific fuel consumption, N2O emissions from agricultural phase,
allocation method) shows the sensitivity of GHGs saving to the adopted allocation method.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction
Nowadays society faces important decisions regarding climate
change mitigation with concerns to potential accumulation in
atmosphere of greenhouse gas emissions (GHGs). In the European
Union (EU), transport is responsible for an estimated 21% of all GHGs
that are contributing to global warming, and the percentage is rising
(EC, 2006). The use of biofuel is one of many strategies proposed by
the European Commission to reduce GHGs in this sector. In order to
improve the usage and production of biofuels, the Renewable
Transportation Directive was adopted in 2003 with the aim of
increasing biofuel share in transportation fuels to 2% in 2005 and
5.75% in 2010 (EPC, 2003).
Major studies show that most biofuels produce lower levels of GHGs
than fossil fuels on a life cycle basis (Johansson et al., 1993; Edmonds
et al., 1996; Nakicenovic et al., 2000; Edwards et al., 2007). Nevertheless,
according to Zah et al. (2007), the environmental impacts of biofuels are
variable and depend very much on location and production methods
and consequently a case-by-case assessment may be required.
This work aims thus to evaluate the GHGs and energy use of
bioethanol produced from wheat in Lombardia (northern Italy) using
the life cycle assessment (LCA) methodology. Although many studies
have been performed concerning this topic (as reviewed by Von
Blottnitz and Curran, 2007), the originality of this is the specific
geographical localization analysed (Lombardia), together with the
“seed-to-wheel” approach which considers the entire life cycle of the
five ethanol blends analysed. The Lombardia region is characterised
by high production of wheat (Fig. 1; Istat, 2007) and 200 km
proximity of recently designed refinery plants of grain-based ethanol.
These considerations, along with the remarked objective of EU in the
biofuel issue (EPC, 2009) and the great increase of CO2 emissions from
transport (+41% between 1990 and 2005 in Lombardia, Caserini,
2008), were the basis of this study in relation to wheat grain-based
bioethanol production and use.
2. Goals and scope definition
2.1. Objectives

⁎ Corresponding author. Tel.: + 39 0223996430; fax: + 39 0223996499. The main objective of this study has been the evaluation of the
E-mail address: costanza.scacchi@libero.it (C.C.O. Scacchi). impact of both production and use of bioethanol, comparing different

0048-9697/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.scitotenv.2010.07.046
 C.C.O. Scacchi et al. / Science of the Total Environment 408 (2010) 5010–5018 5011

Fig. 1. Land dedicated at wheat cultivation and specific production in Lombardia (Northern Italy). Data source: Istat, 2007.

ethanol blends with consideration to total energy use (all energy
input including fossil and non-fossil based energy) and environmental
impact in terms of CO2 equivalent emissions.
Five different blends were considered:
- E0: pure gasoline;
- E5: 5% ethanol and 95% gasoline;
- E10: 10% ethanol and 90% gasoline;
- E85: 85% ethanol and 15% gasoline;
- E100: pure ethanol.
LCA methodology was used to estimate the impacts of these
products, following the ISO standards (ISO 14040 and 14044, 2006). In
particular the impact due to GHGs associated with the different blends
were estimated through the CML 2 baseline 2000 methodology
(Heijungs et al., 1992; Guinée et al., 2001) whereas energy flow
indicator was performed with the Ecoindicator 95 methodology
(Goedkoop, 1995).
CML 2 baseline is an update of the CML method, issued in 1992 by
the “Centre of Environmental Science” of Leiden University. It contains
ten impact categories defined for the midpoint approach (PRé
Consultants, 2008). The characterisation model used for the definition
of the characterisation factors of “global warming (GWP100)”
category was developed by the Intergovernmental Panel on Climate
Change (IPCC, 2001): these factors are defined as Global Warming
Potential for a time horizon of 100 years (GWP 100) and are
expressed in kgCO2/kgemission.
The Ecoindicator 95 method was developed under the Dutch
NOH (National Reuse of Waste Research) program by PRé-
consultants. This method is implemented in the software used for
this study (i.e. SimaPro) with eleven impact categories character-
ised by a damage oriented approach (PRé Consultants, 2008).
“Energy resources” category refers to the extraction and produc-
tion of fuels and to energy generation, accounting for energy
efficiency.
2.2. Functional unit
The functional unit (FU) is the quantified performance of a product
system used in reference in a LCA study (ISO 14040, 2006). In this
assessment, the FU assumed is the amount of fuel required for 1 km
distance: it was assumed that, whether used neat or in blends, the fuel
consumption on energy basis would remain the same as for the base fuel
(Edwards et al., 2007). Thus, gasoline consumption was set equal to
6.95 l per 100 km whereas that of pure ethanol was set equal to 10.5 l
per 100 km.
2.3. Description of the system under study
Fig. 2 shows the flowchart representing the system analysed,
which can be broken up in the following subsystems:
• Agricultural Subsystem (S1), which deals with wheat-growing.
Wheat is cultivated in a farming system using the most common
farming practices in Lombardia (intensive and non-irrigated culture
from October/November to June/July). Rainwater was considered to
be an adequate supply of water for wheat cultivation. This
subsystem includes all agricultural field operations for wheat
cultivation (ploughing, harrowing, fertilizers and pesticide spread-
ing, sowing and harvesting), based on a survey of collected data
including questionnaires filled by farmers and manuals of farming
techniques related to northern Italy territory. The analysis included
wheat seeds, tractors, machineries, fertilizers and pesticide produc-
tion as well as their transportation to the farm gate. The amount of
agricultural machinery required for a specific work process was
obtained by multiplying the weight of the utilised machinery by the
time required for the field work over the lifetime of the machinery
(Nemecek et al., 2004). Energy carrier (mainly diesel) production
and transportation were included, too.
• Refinery Subsystem (S2), which deals with the production of
bioethanol. Harvested grain is transferred to the refinery plant
whereby it is transformed into bioethanol through fermentation.
Distillation processes which consume electricity and steam are
included. Production and use of chemicals is accounted for, too. The
transport of grain from farm to refinery plant is delivered by a 40 t
lorry covering a distance of 200 km (one-way trip, from Lombardy
to Venice), while transport of chemical compounds is carried by a
32 t lorry covering a distance of 110 km (Bernesson et al., 2006).
During the fermentation process 1.01 kgCO2/kgethanol are generated
and released to air (Lechón et al., 2005).
• Gasoline Production Subsystem (S3), which describes the produc-
tion and distribution of fossil gasoline. It includes oil field
exploration, crude oil production, long distance transportation, oil
refining and regional distribution. Emission of air and waterborne
pollutants, waste production, and requirements of energy and
working material have all been considered and included in the
subsystem boundaries (Dones et al., 2004).
• Fuel-use Subsystem (S4), which describes the production and use of
the five blends. Emissions due to the combustion of these different
blends in a flexi fuel vehicle was calculated according to the quantity
of bioethanol added to gasoline, emission factors found in literature
(Table 1) and specific fuel consumption. Emissions were referred to
the same reference vehicle (a passenger car, 1.6 cm3), whose
manufacture, servicing and disposal were excluded being the
5012 C.C.O. Scacchi et al. / Science of the Total Environment 408 (2010) 5010–5018

Fig. 2. Systems under study subdivided in the four subsystems analysed: system boundaries and process chain.

same in all the five options. The transport of ethanol and gasoline
from production plant to petrol station is detailed in Fig. 3. The
preparation of E85 and E100 was considered to take place in the
bioethanol refinery plant.
Carbon emissions originated by direct and indirect land-use
change (Börjesson, 2009; Searchinger et al., 2008; Fargione et al.,
2008) were not included in the analysis. As in Jury et al. (2010), it was
assumed that the land has already been cultivated with the same
agricultural practices since 20 years and therefore the carbon soil
content has become stable, i.e. on a yearly basis the carbon uptake
offsets the carbon emissions (West and Marland, 2003). Moreover,
the consequences of the displacement of food production, i.e. the
need to assure the same food production in an alternative way, which
would be considered in a consequential LCA, are not addressed in this
study. Despite the relevance that environmental impacts caused by
additional production or importation of food may have, consequen-
tial LCA is still an emerging and not well established research topic,
and therefore the attributional approach was preferred.
2.4. Allocation of by-products
Most industrial and agricultural processes are multifunctional. In
particular many feedstocks for biofuels are either co-produced with
other products or are by-products of other production processes. The
 C.C.O. Scacchi et al. / Science of the Total Environment 408 (2010) 5010–5018 5013

Table 1
Data sources.

Data from literature Field data

S1: Agricultural phase

Farming techniques – Farmer's survey
Specific production Istat (2007), CeSPra (2004) Farmer's survey
Fertilizers/chemical products Althaus et al. (2004), Scacchi (2008), Brentrup et al. (2000), Audsley et al. (1997) Farmer's survey
Diesel ENAMA, 2005, DM 26/2/2002; Dones et al. (2004) Farmer's survey
Tractors and machinery Nemecek et al. (2004) Farmer's survey
Seeds Nemecek et al. (2004), Scacchi (2008) –
Transport Spielmann et al. (2004) –
S2: Refinery phase
Plant design and structure – Plant data
Specific production – Plant data
Chemical reagents Bernesson et al. (2006), Althaus et al. (2004), ETH-ESU (1996) –
Steam Airfeen et al. (2007), Dones et al. (2004) –
Electricity Dones et al. (2004) Plant data
CO2 emission Lechón et al. (2005)
Transport Bernesson et al. (2006), Dones et al. (2004), Spielmann et al. (2004) Italian maps (http://maps.google.it/)
S3: Gasoline production
Field exploration, production, Dones et al. (2004), Jungbluth (2004)
transportation, refining
S4: Use phase
Fuels characteristics Lechón et al. (2005), Edwards et al. (2007)
Low sulphur petrol ETH-ESU (1996)
Transport Bernesson et al. (2006), Dones et al. (2004), Spielmann et al. (2004) Italian maps (http://maps.google.it/)
Emissions from combustion Malça et al. (2005), IDIADA (2003a,b), Camaleño Simón (2007), Lechón et al. (2005)

international standard on LCA (ISO 14040, 2006) states that allocation
should be avoided where possible by sub-division or system boundary
expansion. If this is not possible, allocation should be undertaken
using causal relationships based on economic or physical properties of
the co-products.
In this study, S1 and S2 generate by-products which are wheat straw
and distilled dried grain with soluble (DDGS) respectively. These by-
products can be used as an energy resource or as animal food according
to their protein content (Edwards et al., 2007). Nevertheless, the use of
these by-products was not considered in the analysis and allocation was
made on a mass basis, following previous studies (Poitrat et al., 1998;
Ecobilan, 2002). In particular, in S1, burdens were divided equally
between grain and straw (50% to straw and 50% to grain), and the same
was made in S2 (50% to DDGS and 50% to ethanol). A sensitivity analysis,
focusing on GHG emissions, was then performed to evaluate how the
results can change using an allocation method based on economic
considerations.
3. Life cycle inventory (LCI)
This life cycle assessment was developed considering the specific
characteristics of the production of wheat in Lombardia and of the
transformation of feedstock into bioethanol that takes place near
Venice (Veneto). The data necessary for the study were collected in
different ways from different sources, as shown in Table 1.
In order to better represent the real case, real data coming from the
site under study has been used (e.g. to model the farming techniques
system and the refinery plant of bioethanol, for type of machinery and
chemicals reagents use); where these types of data were not available,
they have been integrated with literature values concerning the same
system (e.g. specific production of cereals, agricultural fuel consump-
tion and seed used in the farming system), or similar systems (for
reagents consumption in the refinery plants). For Life Cycle of
materials, the SimaPro database has been used (e.g. Ecoinvent data
for chemicals, machinery and seed used in agricultural phase).
In particular, to quantify the inputs and outputs associated with
the agricultural field activities for the production of grain (subsystem
1) several surveys were designed for Italian farmers. Moreover, these
data were completed with the literature. For example, Nitrogen and
Phosphorous based emissions due to the application of fertilizers
were calculated according to Brentrup et al. (2000) and Audsley et al.
(1997), whereas energy related emissions derived from the use of
agricultural machinery and tractors were taken from Nemecek et al.
(2004). Table 2 shows the inventory data for subsystem 1.
Regarding S2, handled inventory data were supplied by the
refinery plant by means of surveys and corresponding to on-site
measurements. They were completed with bibliography resources,
too (among others, Airfeen et al., 2007; Bernesson et al., 2006). In
particular, CO2 emissions released during the fermentation process
were taken from Lechón et al. (2005). Table 3 shows the inventory
data for the subsystem 2.
Subsystem 3 (gasoline production) is not specific for this study and
so inventory data for it were obtained from Ecoinvent database
(Dones et al., 2004) and other bibliographic sources (Table 1).

Fig. 3. Fuel distribution patterns (distances in km, way to; Lechón et al., 2005): in the refinery plant ethanol is blended with gasoline to form E85; E100 and E85 are sent to
distribution centre where gasoline is added to form E5 and E10 blends. All the blends are sent to petrol station by lorry.
5014 C.C.O. Scacchi et al. / Science of the Total Environment 408 (2010) 5010–5018

Table 2 fuel. N2O emissions were assumed equal to 2.5 10− 5 kg/km for all
Inventory data (per kg of grain) for subsystem 1: agricultural phase. blends (IDIADA, 2003a,b; Camaleño Simón, 2007), whereas methane
Input from Technosphere emission factor was considered equal to 3.2·10− 6 kg/km for E0, E5
Materials and fuels and E10, and 4.8·10− 6 kg/km for E85 and E100 (IDIADA, 2003a,b;
1. Diesel 3.23·10− 2 kg Lechón et al., 2005; Camaleño Simón, 2007).
2. Urea 2.54·10− 2 kg
3. Triple superphosphate 1.04·10− 2 kg
4. Pesticide unspecified 1.73·10− 4 kg 4. Data consistency analysis
5. Seeds 3.66·10− 2 kg
6. Tractors 4.46 kg In the present study a large amount of data has been used,
7. Plough 2.16 kg
referring where possible to on-field data and real scale case. This type
8. Rotary Harrow 3.29 kg
9. Fertilize spreader 0.48 kg
of data has been completed with those from a conventional database
10. Field sprayer 0.24 kg (e.g. Ecoinvent).
11. Sower 0.97 kg To assess the quality of all these data, a score system for quality
11. Harvester 6.30 kg indicators has been used: this system, developed by Weidema (1997),
Transport
associates a score, from 1 to 5, to each quality indicator to provide an
1. Van 5.52·10− 4 tkm
2. Barge 2.69·10− 2 tkm overview with regard to data quality (Table 5).
3. Rail 3.57·10− 3 tkm The results of the analysis on data used in the present study are
4. Lorry, 28t 3.57·10− 3 tkm reported in Table 6. Data collected for the study achieve the higher
Input from Nature score, because they are directly related to the system under study,
1. Arable soil, not irrigated 1.75·10− 4 Ha
while those coming from database (i.e. those for the LCI of diesel,
Output to Technosphere machinery, etc.) present worst values, since they represent average/
Products estimated values, often for different areas and period.
1. Grain 1.00 kg
2. Straw 1.00 kg
5. Results
Output to nature
Emissions to air Fig. 4 shows the results of the LCA of the production and use of the
1. NH3 4.62·10− 3 kg five blends in terms of energy consumption and greenhouse gas
2. N2O 8.51·10− 4 kg emissions.
3. CO2 1.01·10− 1 kg
4. CH4 4.18·10− 6 kg
The results indicate that ethanol blends (E5, E10, E85 and E100)
5. CO 2.1·10− 4 kg allow reducing GHG emissions and energy consumption compared
Emissions to water with fossil gasoline (E0). The highest environmental impact is in fact
1. NO3− 5.23·10− 3 kg associated with E0 blend whose life cycle causes an emission of 214 g
2. P 2.49·10− 4 kg
CO2eq per km. The blend which implies the minor impact is E100 with
70 g CO2eq per km saved compared with E0, while the E85 causes an
impact only 2% higher than those of E100. The difference between
Finally, in subsystem 4, to calculate the emissions associated with low-ethanol content blends and fossil gasoline are minimal (E5 and
biofuel blends combustion in the reference vehicle, emission factors E10 avoid about 9 g CO2eq per km).
available in literature were used. The most impacting GHGs in this The different contributes to the emissions of CO2 eq. associated
phase are CO2, CH4, and N2O (Malça et al., 2005) and their emissions with the different subsystems are detailed in Fig. 5. The impact due to
are listed in Table 4. We specify that ethanol was considered “carbon the transport of the blends (Fig. 3) is taken out from the impact of the
neutral” because the amount of carbon dioxide released in the subsystems 4, to understand its magnitude. The results show that in
combustion step is the same as the amount stored during the growing the LCA of pure gasoline and low-ethanol content blends the most
phase of the wheat. Therefore the carbon dioxide emissions counted significant impact is that associated with the phase of use (S4),
come from the combustion of only the fossil fraction contained in the whereas for E85 and E100, ethanol production phase, which includes
cultivation and refinery, causes the highest impact.
Energy use associated with the different subsystem is presented in
Table 3 Fig. 6. In this case subsystem 4 includes only the impact due to the
Inventory data (per kg of ethanol) for subsystem 2: refinery phase. transport of the blends as their combustion do not cause energy
Input from Technosphere consumption. E0 blend implies the highest energy use, but the
Materials and fuels difference with that of low-ethanol blends (E5 and E10) is very small
1.Grain 3.50 kg (b5%). The lowest energy use is associated with E100. In fact, the
2. Yeast 9.00·10− 5 kg
reduction of fossil oil extraction, which is an energy intensive activity,
3. Phosphoric acid 1.20·10− 4 kg
4. Sulphuric acid 2.00·10− 3 kg enables the reduction of energy use in the ethanol containing fuels,
5. Urea 4.60·10− 4 kg which is proportionate to the biofuel content, as it can be seen in
Energy Fig. 6.
1. Electricity 4.25·10− 4 MWh
2. Steam 3.13 kg
Transport
6. Sensitivity analysis
1. Lorry 40t 7.95·10− 2 tkm
2. Lorry 32t 3.88·10− 5 tkm In the sensitivity analysis (S.A.) three cases have been evaluated
and compared with Base Case:
Output to Technosphere
1. Ethanol 1.00 kg
2. DDGS 1.00 kg 1. Base Case: fuel specific consumption linearly growing with ethanol
content; N2O emissions from S1 due to the application of fertilizers
Output to nature equal to 1,25% of N applied as fertilizer (Brentrup et al., 2000);
Emissions to air allocation of environmental charge based on the mass value of co-
1. CO2 1.01 kg
products;
 C.C.O. Scacchi et al. / Science of the Total Environment 408 (2010) 5010–5018 5015

Table 4
Inventory data (per km) for subsystem 4: final use phase.

E0 E5 E10 E85 E100 Unit

Input from Technosphere

Materials and fuelsa
1. Gasoline 5.23·10− 2 5.09·10− 2 4.94·10− 2 1.12·10− 2 0.00 kg
2. Ethanol 0.00 2.83·10− 3 5.80·10− 3 6.73·10− 2 8.34·10− 2 kg
Total fuel amount 6.95·10− 2 7.13·10− 2 7.31·10− 2 9.97·10− 2 10.5·10− 2 L
Transport
1. Oil pipeline 1.05·10− 4 1.02·10− 4 9.89·10− 5 0.00 0.00 tkm
2. Lorry 40 t 2.14·10− 2 2.64·10− 2 2.70·10− 2 3.71·10− 2 3.88·10− 2 tkm

Output to Technosphere
Emissions to air
1. CO2 1.67·10− 1 1.54·10− 1 1.51·10− 1 2.09·10− 2 0.00 kg
2. CH4 3.20·10− 6 3.20·10− 6 3.20·10− 6 4.80·10− 6 4.80·10− 6 kg
3. N2O 2.50·10− 5 2.50·10− 5 2.50·10− 5 2.50·10− 5 2.50·10− 5 kg
a
Fuel consumption was considered proportionally growing with ethanol content; gasoline specific consumption is equal to 6.95 l/100 km and pure ethanol to 10.5 l/100 km
(Edwards et al., 2007).

2. S.A.1: fuel specific consumption assumed equal for low-ethanol
blends (E0, E5 and E10) as assessed in other studies (Fu et al., 2003;
EC, 2005); N2O emissions and allocation method as in Base Case;
3. S.A. 2 a,b,c: N2O emissions from S1 equal to 2%, 3%, 5% of N applied
as fertilizer (Crutzen et al., 2008); fuel specific consumption and
allocation method as in Base Case;
4. S.A. 3: allocation of environmental charge based on the economic
value of by-products i.e. 73% to grain and 27% to straw, and 84% to
ethanol and 16% to DDGS (Istat, 2007; plant data; Edwards et al.,
2007; Ethanol statistics, 2008); fuel specific consumption and N2O
emissions as in Base Case.
The results of the sensitivity analysis are outlined in Fig. 7 for GHG
emissions. The most influential parameter is the allocation method
used in the analysis: in the S.A.3, in fact, the environmental impact of
ethanol blends E85 and E100 is higher than that of fossil gasoline and
the other ethanol blends (E5 and E10) allow only a slight saving in
GHGs compared with E0.
The study's findings are sensitive to N2O emission factor, too, due
to the high global warming potential of this gas (Heijungs et al., 1992;
Guinée et al., 2001). The rich-ethanol content blends still remain the
preferable, as in the Base Case; nevertheless in S.A.2c E100 allows to
save only 13 gCO2eq/km compared with E0, whereas in the Base Case
this saving was equal to 70 gCO2eq/km. However, Crutzen's
hypothesis on N2O emissions are plausible but seem to be based on
some inappropriate assumptions that overestimate the emission
(Rauh, 2007) and thus this topic is still under discussion.
Finally, the hypotheses made about the fuel specific consumption
change the results but only to a little extent (reduction in GHGs of 2%
and 4%, compared to the Base Case, for E5 and E10 respectively).
7. Discussion
This study confirms that bioethanol in comparison with gasoline
allows the reduction of both GHGs and energy consumption on a life
cycle basis, in agreement with other studies (Beer et al., 2002; Fu et al.,
2003). As assessed in previous studies (Fu et al., 2003; Lechón et al.,
2005), the difference between low-ethanol blends (E5 and E10) and
gasoline are minimal and dependent on the specific fuel consumption
of the vehicle. Also the difference between E100 and E85 is small due
to the high impact associated with the production of ethanol (Nguyen
et al., 2007; Nguyen and Gheewala, 2008).
Since most data used have been collected on field, they often
represent an average value without carrying information about their
distribution, and not allowing us to carry out a comprehensive
uncertainty analysis. In order to have at least an evaluation of outcomes,
results of this LCA study have been compared with other available

Table 5
Pedigree matrix with data quality indicators (Weidema, 1997).

Score Reliability Completeness Correlation

Temporal Geographical Technological

1 Validated-data based on Representative data from a
measurements sufficient sample of sites over
an adequate period to
level out normal fluctuations
2 Validated data partially based Representative data from a
on assumption or non-validated smaller number of sites but for
data based on measurements adequate periods
3 Non-validated data partly Representative data from an
based on assumption adequate number of sites but for
shorter periods
4 Qualified estimate (e.g. by Representative data but from a smaller
industrial expert) number of sites and shot periods or
incomplete data from an adequate
number of sites and periods
5 Non-qualified estimate Representativeness is unknown or
incomplete data from a smaller number
of sites and/or from shorter periods
Less than 3 years of Data from area under study
distance to year study
Less than 6 years of Average data from larger
distance to year study area in which the are under
study is included
Less than 10 years of Data from area with similar
distance to year study poduction conditions
Less than 15 years of Data from area with slightly
distance to year study similar poduction conditions
More than 15 years of Data from unknown area or
distance to year study area with very different
production conditions
Data from enterprises, process
and materials under study
Data from process and materials
under study but form different
enterprises
Data from process and materials
under study but form different
technology
Data on related process and
materials but same technology
Data on related process or
materials but different
technology
5016 C.C.O. Scacchi et al. / Science of the Total Environment 408 (2010) 5010–5018

Table 6
Pedigree matrix for data used in the study.

Reliability Completeness Correlation

Temporal Geographical Technological

S1: Agricultural phase

Farming techniques 1 2 1 1 1
Specific production 1 1 1 2 2
Diesel consumption 4 1 2 2 2
Diesel LCI 3 3 3 2 3
Machinery/tractors type 1 2 1 1 1
Machinery/tractors LCI 1 1 4 3 2
Chemicals reagents consumption 1 2 1 2 2
Chemicals reagents LCI 1 2 1 3 2
Direct emissions from reagents used 2 1 4 2 4
Seed use 4 1 1 3 2
Seed LCI 3 2 1 1 1
S2: Refinery phase
Plant design and structure 1 1 1 1 1
Specific production 1 1 1 1 1
Chemical reagents consumption 1 2 2 3 3
Chemical reagents LCI 1 2 1 3 2
Steam consumption 1 2 2 3 3
Steam LCI 1 2 3 3 5
Electricty consumption 4 2 1 1 1
Electricity LCI 1 1 3 2 2
CO2 direct emissions 1 2 3 3 2
S3: Gasoline production phase
Oil LCI 3 1 3 2 5
S4: Use phase
Fuel blends characteristics 2 2 3 3 2
Low sulphur petrol 1 1 4 3 3
Transport and blending 2 2 2 1 2
Direct emissions 3 4 3 3 2

datasets from literature in which the GHG emissions have been
estimated using the same “seed-to-wheel” approach (Table 7); for the
sake of simplicity, just the pure ethanol blend (E100) has been
compared. For evaluation purposes, figures are expressed in terms of
gCO2eq/kggasoline and gCO2eq/kgbioethanol. Density, calorific value and
average consumption of bioethanol are the same of those used in the
present LCA (Scacchi, 2008). In terms of avoided emissions, the value
found in this study is similar to those in compared literature: bioethanol
allows to save 55% of GHGs compared to fossil gasoline, versus an
average value of 65% met in the literature.
8. Conclusions and recommendations
This study considers the case of the Lombardia region, but results
can be applied to Northern Italy or other areas that have similar
climatic conditions and farming techniques. A land that has already
been cultivated with the same agricultural practices for many years
has been considered, thus Carbon emissions originated by direct and
indirect land-use change were not included in the analysis.
Based on the assumption and data sources applied to this LCA study,
ethanol-based fuels may help to reduce GHG emissions and energy
consumption in high ethanol blends (E85 and E100), which allow a
saving of 67 and 70 gCO2eq/km and 1.6 and 1.9 MJ/km respectively in
comparison with pure gasoline. The use of E5 and E10 fuels may save
energy and GHGs, but only to a small degree and dependent on the
specific consumption of ethanol blending (i.e. the amount of fuel
required to run the reference distance, expressed in L/km). The most
interesting solution seems to be pure ethanol, which allows maximum
energy and GHG savings.
The most influential phase in GHG emissions is the final use of fuel,
whereas in energy use is the gasoline production.
LCA of wheat-based ethanol turns out to be highly sensitive to the
allocation method used, which can be based on mass value of by-
products or on economic considerations. Other LCA studies developed

Fig. 4. Relative greenhouse gas emissions and energy use of the compared ethanol blends, with E0 being the baseline (Index = 100).
 C.C.O. Scacchi et al. / Science of the Total Environment 408 (2010) 5010–5018 5017

Table 7
Literature results for the LCA of the production and use of gasoline and bioethanol
compared with present study values, expressed in terms of greenhouse gases emissions
per kg of fuel (gCO2eq/kgfuel); all studies refer to wheat-based ethanol, except from Wu
et al. (2006), which assessed corn grain-based ethanol.

Source Gasoline Bioethanol GHGs emissions saved

Edwards et al. (2007) 3078 1309 1768

Wu et al. (2006) 4066 1364 2702
Galbraight et al. (2006) 3475 1802 1673
SenterNovem (2005) 3999 1806 2193
EC (2005) 4401 1780 2621
ADEME/DIREM, 2002 3650 922 2728
Present study 4089 1845 2244

Fig. 5. Contribution analysis for the global warming potential impact category (values
given g CO2eq. per km).
Fig. 6. Contribution analysis for the energy use category (values given in MJ per km).
on biofuels present the same sensitivity (Beer et al., 2002; Lechón
et al., 2005; Malça and Freire, 2006) and it is important to recognize
that there is no single allocation procedure which is appropriate for all
biofuel processes (Mortimer et al., 2003). The study's findings are also
sensitive to N2O emission factor associated with the application of
fertilizers in the wheat production.
Acknowledgments
This work has been supported by the Erasmus Exchange Program
in cooperation with the University of Santiago de Compostela and
Politecnico di Milano, and by the “Kyoto Project”, funded by Lombardy
Region and Lombardy Foundation for the Environment.
References
ADEME/DIREM (Agence de l'Environment et de la Maítrise de l'Energie – Direction des
Ressources Energétiques et Minérals). Energy and greenhouse gas balances of
biofuels and production chains in France: Paris; 2002.
Airfeen N, Wang R, Kookos IK, Webb C, Koutinas AA. Process design and optimization of
novel wheat-based continuous bioethanol production system. Biotechnol Prog
2007;23:1394–403.
Althaus HJ, Chudacoff M, Hellweg S, Hischier R, Jungbluth N, Osses M, et al. Life cycle
inventories of chemicals; 2004. Ecoinvent report No. 8. Swiss Centre for Life Cycle
Inventories: Dübendorf.
Audsley E, Alber S, Clift R, Cowell S, Crettaz P, Gaillard G, et al. Harmonisation of
environmental life cycle assessment for agriculture. Final Report AIR3-CT94-2028;
1997.
Beer T, Grant T, Williams D, Watson H. Fuel-cycle greenhouse gas emissions from
alternative fuels in Australian heavy vehicles. Atmos Environ 2002;36:753–63.
Bernesson S, Nilsson D, Hansson P. A limited LCA comparing large- and small-scale
production of ethanol for heavy engines under Swedish conditions. Biomass
Bioenerg 2006;30:46–57.
Börjesson P. Good or bad bioethanol from a greenhouse gas perspective — what
determines this? Appl Energy 2009;85:589–94.
Brentrup F, Küsters J, Lammel J, Kuhlmann H. Methods to estimates on-field nitrogen
emissions from crop production as an input to LCA studies in the agricultural sector.
Int J Life Cycle Ass 2000;6:349–57.
Camaleño Simón MC. Inventario des emisiones procedentes del transporte por
carretera en Castilla y León, 2002–2005. Doctoral Thesis (in Spanish); 2007.
Caserini S. Greenhouse gas emissions in LombardyChapter 3 in: Progetto Kyoto
Lombardia, Fondazione Lombardia per l'Ambiente Ed., Milano; 2008. (in Italian).
Centro Studi per una nuova etica economica (CeSpra). Scheda Frumento (triticum),
versione giugno 2004. Available online at http://cespra.provincia.venezia.it/sp_cg.
htm.
Crutzen PJ, Mosier AR, Smith KA, Winiwarter W. N2O release from agro-biofuel
production negates global warming reduction by replacing fossil fuels. Atmos
Chem Phys 2008;8:389:395.
Decreto Ministeriale (DM) 26/2/2002. Determinazione dei consumi medi dei prodotti
petroliferi impiegati in lavori agricoli, orticoli, in allevamento, nella selvicoltura e
piscicoltura e nelle coltivazioni sotto serra ai fini dell'applicazione delle aliquote o
dell'esenzione dell'accisa (in Italian).
Dones R, Bauer C, Bolliger R, Burger B, Faist Emmenegger M, Frischknecht R, et al. Life
cycle inventories of energy systems: results for current systems in Switzerland and
other UCTE countriesEcoinvent report No. 5. Paul Scherrer Institut Villigen, Swiss
Centre for Life Cycle Inventories: Dübendorf; 2004.

Fig. 7. Sensitive analysis (S.A.) results for the greenhouse gas emissions: S.A.1: equal fuel specific consumption for low-ethanol blends; S.A.2: N2O emissions from S1 equal to 2%–3%–
5% (respectively in case a–b–c) of N applied as fertilizer; S.A.3: allocation based on economic value of by-products.
5018 C.C.O. Scacchi et al. / Science of the Total Environment 408 (2010) 5010–5018
Ecobilan. Energy and greenhouse gas balances of biofuels' production chain in France.
Report, ADEME: France; 2002.
Edmonds JA, Wise MA, Sands RD, Brown RA, Kheshgi H. Agriculture, land use, and
commercial biomass energy: a preliminary integrated analysis of the potential role
of biomass energy for reducing future greenhouse related emissions. Washington,
US: Pacific Northwest National Laboratory; 1996.
Edwards R, Larivé J-F, Mahieu V, Rouveirolles P. Well-to-wheels analysis of future
automotive fuels and powertrains on the European context. Eucar Concawe JRC,
Well-to-Wheels Report; 2007.
Ente per la Meccanizzazione Agricola (ENAMA). Prontuario dei consumi di carburante
per l'impiego agevolato in agricoltura (in Italian); 2005.
Ethanol Statistics. United States ethanol and commodity prices; 2008. Available
online at http://www.ethanolstatistics.com/Commodity_Prices/US_Ethanol_and_
Commodity_Prices.aspx.
ETH-ESU. SimaPro Database Manual. The ETH-ESU 96 libraries; 1996. Available online
at: http://www.pre.nl/download/manuals/DatabaseManualETH-ESU96.pdf.
European Commission (EC. Shift Gear to biofuels: results and recommendations from
the VIEWLS project: Brussels; 2005.
European Commission (EC. Communication from the commission, an EU strategy for
biofuels. COM(2006)34 final: Brussels; 2006.
European Parliament and Council (EPC. Directive on the promotion of the use of energy
from renewable sources and amending repealing Directives 2001/77/EC and
2003/30/EC; 2009. Available on line at http://europa.eu.int.
European Parliament and Council (EPC). Directive on the promotion of the use
of biofuels or other fuels for transport, 2003/30/EC; 2003. Available online at
http://europa.eu.int.
Fargione J, Hill J, Tilman D, Polasky S, Hawthorne P. Land clearing and the biofuel carbon
debt. Science 2008;29:1235–8.
Fu GZ, Chan AW, Minns DE. Life cycle assessment of bio-ethanol derived from cellulose.
Int J Life Cycle Ass 2003;8:137–41.
Galbraight D, Smith P, Mortimer N, Stewart R, Hobson M, McPherson G, et al. Review of
Greenhouse Gas Life Cycle emissions, air pollution impacts and economic of
biomass production and consumption in Scotland. Final Report Seerad Project
FF/05/08; 2006.
Goedkoop M. The Ecoindicator 95 weighting method for environmental effects that
damage ecosystems or human health on a European scale — contains 100 indicators
for important materials and processes; 1995.
Guinée JB, Gorreé M, Heijungs R, Huppes G, de Koning A, van Oers L, et al. Life cycle
assessment. An operational guide to the ISO standards. Leiden University: Centre of
Environmental Science; 2001.
Heijungs R, Guineé JB, Huppes G, Lankreijer RM, Udo de Haes HA, Ansems AMM, et al.
Environmental life cycle assessment of products. Guide. NOH Report 9266. Leiden
University: Centre of Environmental Science; 1992.
IDIADA Automotive Technology. Technical Report: comparison of fuel tank evaporative
emissions from E0 an E5 petrol; 2003a.
IDIADA Automotive Technology. Technical Report: comparison of vehicle emissions at
European Union annual average temperatures from E0 an E5 petrol; 2003b.
International Standard Organization (ISO) 14040. Environmental management – life
cycle assessment – principles and framework; 2006.
International Standard Organization (ISO) 14044. Environmental management – life
cycle assessment – requirements and guidelines; 2006.
IPCC. Climate change 2001: the scientific basis. Cambridge University Press; 2001.
, Istituto Italiano di Statistica (Istat. Agriculture database; 2007. Available online at
http://www.istat.it.
Johansson TB, Kelly H, Reddy AKN, Williams RH. A renewable-intensive global energy
scenario (Appendix in chapter 1)Renewable energy: sources for fuels and
electricity, Washington, USA; 1993. p. 1071–143.
Jungbluth N. Erdöl. In: achbilanzen von Energiesystemen: Grundlagen für den
ökologischen Vergleich von energiesystemen und den Einbezug von Energiesyste-
men in Ökobilanzen für die Schweiz (Ed. Dones R.). Final report ecoinvent 2000 No.
6-IV, Paul Scherrer Institut Villigen, Swiss Centre for Life Cycle Inventories,
Dübendorf, CH; 2004.
Jury C, Benetto E, Koster D, Schmitt B, Welfring J. Life Cycle Assessment of biogas
production by monofermentation of energy crops and injection into the natural gas
grid. Biomass Bioenergy 2010;34:54–66.
Lechón Y., Cabal H., Lago C., de la Rúa C., Sáez M.R., Fernández M. Análisis del ciclo de vida de
combustibles alternativos para el transporte. Fase I: Análisis del ciclo de vida comparativo
del etanol de cereales y de la gasolina. Centro de Investigaciones energéticas
medioambientales y tecnológicas (Ciemat): Madrid, Spain (in Spanish); 2005.
Malça J, Freire F. Renewability and life-cycle energy efficiency of bioethanol and bio-
ethyl tertiary butyl ether (bioETBE): assessing the implications of allocation.
Energy 2006;31:3362–80.
Malça J, Rozakis S, Freire F. Bioethanol replacing gasoline: greenhouse gas emissions
reduction, life-cycle energy saving and economic aspects2nd International
Conference of Life Cycle Management: Barcelona; 2005.
Mortimer N, Cormack P, Elsayed M, Horne R. Evaluation of the comparative energy,
global warming and social costs and benefits of biodiesel. Report, Resource
Research Unit: Sheffield Hallam University, UK; 2003.
Nakicenovic N, Alcamo J, Davis G, de Vries B, Fenhann J, Graffin S, et al. Emission
scenarios, a special report of Working Group III of the Intergovernmental Panel on
Climate Change (IPCC): Cambridge; 2000.
Nemecek T., Heil A., Huguenin O., Meier S., Erzinger S., Blaser S. et al. Life Cycle
Inventories of Agricultural Production Systems. Final report ecoinvent 2000 No. 15.
Agroscope FAL Reckenholz and FAT Taenikon, Swiss Centre for Life Cycle
Inventories: Dübendorf, CH; 2004.
Nguyen TLT, Gheewala SH. Life Cycle Assessment of fuel ethanol from Cassava in
Thailand. Int J Life Cycle Ass 2008;13:147–54.
Nguyen TLT, Gheewala SH, Garivait S. Fossil energy savings and GHG mitigation
potentials of ethanol as a gasoline substitute in Thailand. Energy Policy 2007;35:
5195–205.
Poitrat E, Leviel B, Vergé C, Gosse G. Total costs and benefits of biomass in selected
regions of the European Union (Biocosts). Report to the European Commission on
the Case Study “ETBE from sugar beet under French conditions”; 1998.
PRé Consultants, SimaPro Database Manual – Methods library. Available online at
http://www.pre.nl; 2008.
Rauh S. Interactive comment on “N2O release from agro-biofuel production negates
global warming reduction by replacing fossil fuels” by P.J. Crutzen et al. Atmos
Chem Phys Discuss 2007;7:S4616–9.
Scacchi C. Riduzioni delle emissioni di gas serra dall'utilizzo di biocarburanti: il caso
della Lombardia. Politecnico di Milano, Degree thesis (in Italian): Milano; 2008.
Searchinger T, Heimlich R, Houghton RA, Dong F, Elobeid A, Fabiosa J. Use of U.S.
croplands for biofuels increases greenhouse gases through emissions from land-use
change. Science 2008;319:1238–40.
SenterNovem (Agency of Dutch Ministry of Economic Affairs. Participative LCA on biofuels.
Report 2GAVE-05.08. Climate Neutral Gaseous and Liquid Energy Carriers; 2005.
Spielmann M., Kägi T., Stadler P., Tietje O. Life cycle inventories of transport services.
ecoinvent report No. 14. Swiss Centre for Life Cycle Inventories: Dübendorf; 2004.
Von Blottnitz H, Curran M. A review of assessments conducted on bio-ethanol as a
transportation fuel from a net energy, greenhouse gas, and environmental life cycle
perspective. J Cleaner Prod 2007;16:1943–8.
Weidema PB. Environmental assessment of products. Helsinki, Finland: The Finnish
association of graduate engineers; 1997.
West TO, Marland G. Net carbon flux from agriculture: carbon emissions, carbon
sequestration, crop yield, and land-use change. Biogeochemistry 2003;63:73–83.
Wu M., Wang M., Huo H. Fuel-cycle Assessment of selected bioethanol production
pathways in the United States. ANL/ESD/06-7; 2006.
Zah R, Hischier R, Gauch M, Lehmann M, Böni H, Wäger P. Life Cycle Assessment of
energy products: environmental Impact Assessment of biofuels. Bundessamt für
Energie, Bundessamt für Umwelt, Bundessamt für Landwirtschaft: Bern; 2007.