Sunteți pe pagina 1din 10

Applied Energy 88 (2011) 20882097

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Life-cycle greenhouse gas emissions and energy balances of sugarcane ethanol


production in Mexico
Carlos A. Garca a,, Alfredo Fuentes b,g, Anna Hennecke c,d, Enrique Riegelhaupt e, Fabio Manzini f,
Omar Masera b
a
Posgrado en Ingeniera (Energa), Centro de Investigacin en Energa, Universidad Nacional Autnoma de Mxico, Privada Xochicalco S/N, Colonia Centro, Temixco,
Morelos 62580, Mexico
b
Centro de Investigaciones en Ecosistemas, Universidad Nacional Autnoma de Mxico, Antigua Carretera a Ptzcuaro No. 8701 Colonia Ex-Hacienda de San Jos de la Huerta,
58190 Morelia, Michoacn, Mexico
c
IFEU Institute for Energy and Environmental Research Heidelberg GmbH, Wilckensstr. 3, 69120 Heidelberg, Germany
d
Center for Development Research (ZEF), University of Bonn, Walter-Flex-Str. 3, 53113 Bonn, Germany
e
Red Mexicana de Bioenerga A.C. Av. San Jos del Cerrito, 400-51 Col. El Pueblito CP 58341 Morelia, Mich., Mexico
f
Centro de Investigacin en Energa, Universidad Nacional Autnoma de Mexico, Privada Xochicalco S/N, Colonia Centro, Temixco, Morelos 62580, Mexico
g
Posgrado en Ingeniera (Energa), Facultad de Ingeniera, Universidad Nacional Autnoma de Mxic, Av. Universidad No. 3000. 04510, Mexico D.F.

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

Article history: The purpose of this work was to estimate GHG emissions and energy balances for the future expansion of
Received 13 August 2010 sugarcane ethanol fuel production in Mexico with one current and four possible future modalities. We
Received in revised form 23 December 2010 used the life cycle methodology that is recommended by the European Renewable Energy Directive
Accepted 25 December 2010
(RED), which distinguished the following ve system phases: direct Land Use Change (LUC); crop produc-
Available online 26 January 2011
tion; biomass transport to industry; industrial processing; and ethanol transport to admixture plants. Key
variables affecting total GHG emissions and fossil energy used in ethanol production were LUC emissions,
Keywords:
crop fertilization rates, the proportion of sugarcane areas that are burned to facilitate harvest, fossil fuels
Sugarcane ethanol
GHG emissions
used in the industrial phase, and the method for allocation of emissions to co-products. The lower emis-
Energy balance sions and higher energy ratios that were observed in the present Brazilian case were mainly due to the
Life cycle assessment lesser amount of fertilizers applied, also were due to the shorter distance of sugarcane transport, and to
Biofuel the smaller proportion of sugarcane areas that were burned to facilitate manual harvest. The resulting
modality with the lowest emissions of equivalent carbon dioxide (CO2e) was ethanol produced from
direct juice and generating surplus electricity with 36.8 kgCO2e/GJethanol. This was achieved using bagasse
as the only fuel source to satisfy industrial phase needs for electricity and steam. Mexican emissions were
higher than those calculated for Brazil (27.5 kgCO2e/GJethanol) among all modalities. The Mexican modality
with the highest ratio of renewable/fossil energy was also ethanol from sugarcane juice generating sur-
plus electricity with 4.8 GJethanol/GJfossil.
2010 Elsevier Ltd. All rights reserved.

1. Introduction feedstocks. Despite that not all authors agree with this statement
in our work we assumed a Global Warming Potential of zero for
The climate change problem coupled with declining oil and gas CO2 from biofuels according to the RED [1].
reserves has led to the development of energy sources to minimize Bioethanol produced by fermentation and distillation of sugary
greenhouse gas (GHG) emissions and expand energy supplies from or starchy feedstocks is of particular importance. Global ethanol
solar, wind, hydraulic, geothermal and bioenergy sources. output grew from 17,000 million liters (ML) in 2000 to 65,614 ML
The bioenergy sources used most frequently in the transport in 2008 [2] and might provide up to 7% of the global energy used
sector are liquid biofuels that are either pure or mixed with con- in the transport sector by 2030 [3]. Sugarcane is one of the most
ventional, oil-derived fuels. Biofuels are a renewable energy source widely used feedstocks for ethanol, especially in Brazil [4] where
because they are obtained from crops and other biomass sources. It it constituted 37% of global ethanol production in 2008 [2].
is common practice in most LCA to count CO2 from biomass as neu- Mexico has a long tradition of sugarcane cropping that was
tral because CO2 from biofuel combustion was previously seques- introduced by the Spaniards in 1524. The Mexican sugar industry
tered from the atmosphere during the growth of plants used as currently generates 450 thousand direct jobs and 2.2 million
indirect jobs, comprising 2.5% of the Gross Internal Product of the
Corresponding author. Tel.: +52 55 5622 9704; fax: +52 55 5622 9791. industrial sector [5]. In the 2008/2009 milling season, 663,057
E-mail address: cagab@cie.unam.mx (C.A. Garca). hectares of sugarcane were harvested and 42.5 million tonnes

0306-2619/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apenergy.2010.12.072
C.A. Garca et al. / Applied Energy 88 (2011) 20882097 2089

(Mt) of sugarcane were processed by sugar mills resulting in 4.9 Mt replanting. Medium and high yields may vary between 70 and
of sugar. In addition, 1.5 Mt of molasses were produced, of which 93 t/ha year, depending on fertilization and irrigation. Soil tillage
0.05 Mt (3.6%) were used to produce ethanol with an output of is fully mechanized, but harvest is mostly manual (95%). This last
14.5 ML [6]. While most of the remaining molasses was exported, practice makes it necessary to burn sugarcane straw (cane tops
mainly to the US, it could alternatively be used as an ethanol feed- and leaves) in the elds before and after harvesting to ease the cut-
stock that might yield up to 400 ML of biofuel. ters work and to clear residues. The bagasse content in milled sug-
A recent study estimated that 2.9 million hectares are poten- arcane is not affected by eld burning, since bagasse is the ber
tially suitable for sugarcane cultivation in Mexico and could yield that always remains within the sugarcane stalks. Truck loading is
up to 16,240 ML of ethanol per year, which would be similar to mechanized.
the present Brazilian output [7]. This study assessed the technical
potential for new sugarcane plantations under the following condi- 2.2. Industrial processing
tions and assumptions that were deemed necessary to ensure sus-
tainability and to avoid fuel versus food conicts: (a) adequate Ethanol can be produced directly from sugarcane juice or from
rainfall ranges; (b) low terrain slopes; (c) long frost-free periods; molasses produced during the sugar making process (Fig. 1). Sugar
(d) nature protection area and cultivated area exclusions; (e) pres- production starts with sugarcane milling to extract juice, and this
ent land uses (only natural and cultivated grasslands were in- juice is claried, ltered and concentrated. The concentrated juice
cluded); and (f) high to medium potential yields. is then crystallized, and sugar crystals are separated from molasses
Debate has re-emerged in Mexico as to the rationale for promot- by centrifugation. This process may be repeated up to three times.
ing ethanol use in the transport sector to achieve GHG mitigation, Products of the rst crystallization/centrifugation stage are termed
energy diversication, and increased employment in industrial A Sugar and A molasses. Second stage products are termed B
and rural sectors. A biofuels Promotion and Development Law was sugar and B molasses. Third stage products are termed C sugar
passed in 2008 [8] to establish ethanol as an oxygenation additive and C or nal molasses, and these C products do not contain
in gasoline. As a result of statutory provisions, a program for biofuels recoverable sucrose but still have 50% fermentable sugars [23]. Fi-
introduction was drafted under the framework of the Intersectorial nal molasses are fermented by yeast cultures, and the wine that is
Bioenergy Strategy. The Program foresees the adoption of sustain- obtained is distilled to obtain ethanol.
ability criteria for ethanol production, and the balance of GHG emis- An alternative process that allows for higher molasses quanti-
sions and fossil energy are important aspects of the program [9]. ties to be fermented involves starting fermentation using B
The balance of GHG emissions and fossil energy in the sugar- molasses. In this case, only two crystallizations are undertaken to
cane-sourced ethanol life-cycle is important because they evaluate obtain A and B sugars, while B molasses are derived to pro-
GHG emissions and primary fossil fuel consumption along the en- duce ethanol instead of C sugar and C molasses, with lesser
tire ethanol production. Thus, these variables determine sugarcane amounts of sugar obtained.
ethanol effective GHG emissions mitigation potential and fossil Hydrated ethanol of 96% purity must be dehydrated to 99.8%
fuel substitution level. purity to allow for use as a gasoline additive.
Despite the rapid growth of global ethanol production, its ben-
ets in mitigating GHG emissions and substituting for fossil fuels 2.3. Possible modalities of ethanol production in Mexico
continue to be questioned. Several life cycle studies focusing on
GHG emissions and energy balances have been conducted, and Five modalities of ethanol production in Mexico were analyzed
these studies have arrived at different and sometimes contradic- (Table 1).
tory conclusions [1019]. One explanation for this disparity is that
different methodologies and assumptions have been used, such as I. Ethanol from C molasses with bagasse and fuel oil (EMF).
different system boundaries, fertilization levels, emission factors, Medium average crop yields and corresponding inputs levels
co-product inclusions and exclusions, and diverse methods for are assumed. The sugar mill produces A, B and C sug-
allocating emissions to co-products. ars plus nal molasses. The production involves burning all
To date, no evaluations of life-cycle GHG and energy balances for the bagasse and some fuel oil to produce steam and a part
ethanol fuel produced in Mexico have been published. This data gap of the needed electricity, while the remaining energy needs
is mainly due to the lack of reliable and systematic statistical data on are imported from the National Power Grid (NPG). This
inputs used in agricultural and industrial phases. However, efforts modality is actually applied by 38 of 54 sugar mills in Mex-
have been undertaken to produce initial estimates based on data ico [24] that have not attained energy self-sufciency and
from other publications, default values, and limited eld data [Mario use from 1 to 22 liters of fuel oil per ton of sugarcane pro-
Molina Center, pers.com.] [2022]. The purpose of this work was to cessed. Our analysis assumed that a distillery was annexed
determine GHG emission and energy balances based on eld data to to a sugar mill and operated using fuel oil as its only energy
allow for a better evaluation of environmental and energetic rami- source.
cations of ethanol fuel production in Mexico. In addition, this re- II. Ethanol from B molasses with fuel oil (EMBF). Medium
search has the following additional aims: to contribute to the average crop yields are assumed. The mill produces A
debate on environmental impacts of ethanol fuel production; to as- and B sugars plus B molasses. Boilers use bagasse and
sess the effects of different methodologies on Life cycle assessment fuel oil as supplementary fuels, and a part of the electricity
(LCA) calculations; to highlight the inuence of local conditions on is imported from the NPG. As in modality I, we assumed that
these results; and to inform decision making processes in Mexico. a distillery was annexed to the mill that used fuel oil as its
only energy source.
III. Ethanol from C molasses without fuel oil and with bagasse
2. Sugarcane production and processing as the only fuel (EMB). Medium average sugarcane yields are
assumed. The mill produces A, B, and C sugars, plus
2.1. Agricultural production nal molasses for ethanol production. Boilers use only
bagasse for steam and power production, which is consistent
Visits were made to two sugar mills and one ethanol distillery with practices in about 17 mills in Mexico. However, part of
where sugarcane is cultivated in ve-year cycles followed by the electricity is purchased from the NPG. The distillerys
2090 C.A. Garca et al. / Applied Energy 88 (2011) 20882097

Fig. 1. Industrial processes of ethanol production using sugarcane.

Table 1
Possible modalities for ethanol production in Mexico.

Modality Crop yield (t/ Ethanol Sugar output (kg/t Energy source for sugar Energy source for Electricity to/from Ethanol output (L/t
ha) source cane) mill distillery grid cane)
EMF 70 Final 112a Bagasse + fuel oil Fuel oil IMP 8.8a
molasses
EMBF 70 B molasses 92a Bagasse + fuel oil Fuel oil IMP 17.1a
EMB 70 Final 120c Bagasse Bagasse IMP 7.9c
molasses
EDJ 70 Direct juice Bagasse 83.2b
EDJE 70 Direct juice Bagasse EXP 83.2b

EMF ethanol from C molasses using bagasse and fuel oil.


EMBF ethanol from B molasses using bagasse and fuel oil.
EMB ethanol from C molasses using bagasse.
EDJ ethanol from direct juice using bagasse and without surplus electricity.
EDJE ethanol from direct juice using bagasse and generating surplus electricity.
a
Source: [23].
b
Source: Based on [16].
c
Field data, La Gloria sugar mill and distillery.

boiler is also bagasse-operated. This modality corresponds to 3. Methodology of LCA


one actual case in Mexico.
IV. Ethanol from direct juice without surplus electricity (EDJ). We used the methodology recommended by the Renewable En-
Medium average sugarcane yields and an autonomous dis- ergy Directive (RED) of the European Union Parliament to perform
tillery are assumed whereby only ethanol is produced. LCAs for the ve modalities of interest [1]. System boundaries were
Bagasse is the only energy source with no electricity surplus. dened and followed by a detailed inventory of inputs and outputs
This is a case for which inputs and outputs at the industrial that were registered in an especially designed data base. While
phase were estimated on the basis of Brazilian references. most data were obtained by eld observations, some of the indus-
V. Ethanol from direct juice using bagasse as a fuel and gener- trial phase data were taken from other references.
ating surplus electricity (EDJE). This modality is basically the GHG emissions and fossil energy balances were calculated with
same as modality IV. The difference between this modality a modied computer model originally developed by the Institute
and modality IV is that surplus electricity is exported to for Energy and Environment Heidelberg (IFEU). This is a spread
the national grid because more efcient, high-pressure boil- sheet model based in the RED methodology and previously used
ers are used. in other LCAs [25]. Emission and energy allocations were calcu-
lated according to the energy content of products and co-products
In all modalities since both sugar extraction and ethanol pro- for modalities that combine the production of sugar and ethanol.
duction are fairly mature technologies, no signicant increases For modalities where only ethanol is produced (autonomous dis-
are foreseen regarding industrial yields of sugar and ethanol. tilleries), emission credits were allocated if surplus electricity is
C.A. Garca et al. / Applied Energy 88 (2011) 20882097 2091

produced and exported to the grid. Emissions due to land use 498 gCO2e/kW h. We also determined an average energy efciency
changes were calculated for each modality to explore effects of of 40.7% from the weighted mean of all fossil fuel power plants.
expanding sugarcane cultivation to well-suited lands. Emission
and fossil energy balances were also calculated for a Brazilian case 3.3. Assumptions about land use change (LUC)
on the basis of data published by Macedo [16] and following the
RED methodology. This measurement allowed us to compare re- Every biofuel production process starts at crop settlement and
sults under the same system boundary denitions and with identi- thereby implies a change in land use. LUCs result in altered carbon
cal emission factors. stocks in soil and vegetation. Emissions by direct LUC are due to vari-
ations in AGB carbon content (e.g., stems, branches, leaves, and lit-
ter), BGB carbon content (e.g., roots and rhizomes), and OCS. Our
3.1. System boundaries denition
methodology assumed that carbon lost through LUC is emitted as
CO2. A strong negative input in GHG balances occurs when high car-
The following system boundaries dened processes that were
bon lands are cultivated as is the case when forested lands are con-
included from the analysis: Land Use Change (LUC); sugarcane cul-
verted to sugarcane elds. This substantial LUC results in negative
tivation and harvest; sugarcane transportation to ethanol plants;
overall CO2 balances for ethanol production [3032].
industrial processing; and end product transportation to admix-
The impact of sugarcane expansion to lands suitable for diverse
ture facilities.
crops was evaluated under each Mexican production modality.
Emissions from sugarcane cultivation and harvest include the
Expansion to degraded lands was not considered because analyzed
following: volatilization of nitrogen fertilizers; emissions and fossil
production systems considered medium sugarcane yields that can-
energy from fertilizers1 and pesticide production; fossil fuels used
not be achieved on degraded lands without restoration and
by agricultural machinery; and direct emissions from sugarcane
irrigation.
burning before and after harvest. Emission and fossil energy use
AGB values were estimated for diverse vegetation types.
from industrial process included those from boiler fossil fuels and
According to the National Inventory of GHG Emissions INEGEI
grid-imported electricity. Emissions and energy use from transporta-
[33], the following vegetation types occur in Mexican climatic
tion, including transporting sugarcane to mills and ethanol to admix-
zones with good aptitude for sugarcane cultivation: grasslands;
ture facilities, occurs from fossil fuel use by trucks. Emissions from
low and medium tropical forests; high tropical forests, brushlands;
land use changes result from the reduction of carbon content in
and cultivated areas. We used the average carbon content (tC/ha)
above ground biomass (AGB), below ground biomass (BGB), and or-
in AGB and OCS reported by INEGEI (Table 3). Data on BGB were
ganic carbon in soil (OCS) in areas needed to expand sugarcane crop-
also derived from INEGEI. While there is uncertainty surrounding
ping. GHG emissions considered in the analysis were CO2, methane
the OCS values, they were the best data available for Mexican soils.
(CH4) and nitrous oxide (N2O) with global warming potentials of 1,
LUC emissions were annualized over a 20-year period [1].
23, and 296, respectively [26].
Other GHG emissions from indirect LUC occurred when biofuel
The following emissions and energy sources were not consid-
crop production was expanded to areas that had been previously
ered in this study as they are not included in the RED methodol-
cultivated, resulting in crop displacements to previously unculti-
ogy: machinery, equipment and facility construction; human
vated lands. These indirect emissions cannot be attributed to spe-
work and seed.
cic production processes because they depend on market
Spreadsheets were used for data registration and to run a model
demand, relative prices, and substitute availability. Indirect LUC
to calculate emission and energy balances. Data were gathered
is a complex process that is not fully understood by the scientic
in situ by the authors. The calculation model developed by the
community and so is not included in this study.
Institute for Energy and Environment Heidelberg (IFEU) was mod-
ied and adapted to t conditions in Mexico. A module was also
3.4. Allocation
added to calculate fossil energy balances.
One or more additional products besides ethanol or biodiesel
3.2. Data are obtained in many biofuel production systems. It is necessary
to decide the share of emissions and energy that corresponds to
Original data were obtained from the Motzorongo sugar mill each additional co-product. Allocation of energy and/or emissions
and La Gloria distillery, which are located in Veracruz, Mexico. Sug- can be determined by economic value, co-product mass, energy
arcane growers, rural workers, truck and loader operators, mill content, or substitution.
workers, technicians, engineers and plant managers were inter- Economic valuation considers the amount and market price of
viewed. Data on energy inputs in sugarcane cultivation, sugarcane products and co-products and is based on the assumption that
loading and transportation were also collected (Table 2). Emissions market prices are the driver of the production process. One disad-
associated with nitrogen fertilizer volatilization were calculated vantage of this approach is that prices are not constant over time.
according to Intergovernmental Panel on Climate Change (IPCC) In addition, allocations to biofuels would be strongly inuenced by
recommendations [27]. Industrial phase inputs and outputs were price variations in co-products markets [34,35]. Finally, subsidies
also collected during the eld visits, although some data points toward fuels and co-products might distort relative prices [17,35].
were taken from the literature. Values to capture transportation Allocation by mass and energy contents account for physical
to admixture plants were estimated based on the average distance properties. Mass content accounts for the relative masses of biofu-
from mills to existing gasoline storage and mixing plants. els and co-products, and energy content accounts for the energy
To assess GHG emissions from NPG electricity imports, average content value in biofuels and co-products. The energy content con-
emission factors were calculated using data from the National Energy sideration is adopted by RED and its main advantage is that heating
Balance 2007 [28]. A matrix was constructed that contained the share values are constant, easily determined and comparable to alloca-
of each primary energy source. After applying emission factors from tions by substitution. A possible disadvantage of this allocation is
the IPCC [29], we calculated an average emission factor of that a given co-product may have a high caloric content but a
low market price.
1
Emission factors assumed for fertilizers were 6.41 kgCO2e/kg N; 1.18 kgCO2e2e/kg In substitution allocations or system expansion, the biofuel is
P2O5; 0.66 kgCO2e2e/kg K2O; 0.29 kgCO2e2e/kg CaO. considered the only product, but emissions or energy substituted
2092 C.A. Garca et al. / Applied Energy 88 (2011) 20882097

Table 2
Data used for LCA calculations (eld-collected unless otherwise indicated).

EMF EMBF EMB EDJ EDJE Brazil EDJEa


Cultivation
Nitrogen (kg/ha yr) 126 126 126 126 126 80e
Phosphate (P2O5; kg/ha yr) 42 42 42 42 42 45e
Potassium oxide (K2O; kg/ha yr) 150 150 150 150 150 114.6e
CaO (kg/ha yr) 3000 3000 3000 3000 3000 1900
Pesticides (kg/ha yr) 2 2 2 2 2 3
Diesel (kg/ha yr) 109 109 109 109 109 164
Irrigation (MJ/ha yr)
Yield (t cane/ha yr) 70 70 70 70 70 87.1
Industrial phase
Fuel oil (kg/t cane) 10.7b 7.4b
Electricity buying (kWh/t cane) 7.6d 9.0d 7.5
Surplus electricity (kWh/t cane) 8.8c 9.2f
Factory yield (L ethanol/t cane) 8.8b 17.1b 7.9 83.2c 83.2c 86.3
Transport
Biomass transport (km, round trip) 87.2 87.2 87.2 87.2 87.2 46
Admixture transport (km) 300 300 300 300 300 300g

EMF ethanol from C molasses using bagasse and fuel oil.


EMBF ethanol from B molasses using bagasse and fuel oil.
EMB ethanol from C molasses using bagasse.
EDJ ethanol from direct juice using bagasse and without surplus electricity.
EDJE ethanol from direct juice using bagasse and generating surplus electricity.
Brazil EDJE Brazilian ethanol from direct juice using bagasse as a fuel and generating surplus electricity.
a
Taken from Macedo [16].
b
Taken from [23].
c
Based on Brazilian conditions and study calculations.
d
Assumptions based on [23] and study calculations.
e
Average values based on data from [16].
f
Electricity mix for Brazil 180 g CO2e/kW h.
g
Assumed to be the same as Mexico.

Table 3
Carbon content values used for LUC calculations.

Grassland Dry foresta Rainforestb


Previous land cover
Total t C/ha 95.4 89.3 148.6
In Biomass above and below ground t C/ha 34.2 21.6 50.5
In Soil t C/ha 61.2 67.7 98.1
After LUC to sugarcane
Total t C/ha 69.4 69.4 69.4
In Biomass above and below ground t C/ha 12.6 12.6 12.6
In Soil t C/ha 56.8 56.8 56.8

Source: study calculations based on data from INEGEI [33].


LUC land use change.
a
Includes low and mid tropical forests plus brush lands.
b
High tropical forests.

by co-products are deducted. This procedure is recommended by ucts to be compared in a meaningful manner, recognizing that
the International Organization for Standardization (ISO) [36,37]. the alternatives must be compared on the basis of the equivalent
However, it may be difcult to apply substitution in many cases function which they perform or provide [18]. In this work for
because one co-product can be utilized in more than one form GHG emissions we have chosen the functional unit kgCO2e/GJethanol
and a choice has to be made about the type of substitution. In addi- in equivalence to the established by the RED [1]. For the energy
tion, data may not be available on life cycle emissions and substi- balance we assumed the relation GJethanol/GJfossil.
tuted product energy values.
We used allocation by energy content for this study. The
adopted Low Heating Values (LHVs) were as follows: 26.8 MJ/kg 4. Results
for ethanol and 16.5 MJ/kg for sugar. The avoided GHG emissions
associated with the excess of electricity were taken to be equal 4.1. GHG emissions
to the amount of GHG that would be emitted when an equal
amount of electricity was generated by the national grid. Results of our GHG emission balances are expressed in CO2e
emission units per ethanol energy unit (kgCO2e/GJethanol). Fig. 2
3.5. Functional unit presents gures for the Brazilian case to allow for direct compari-
son against Mexican modalities. Fig. 2 also provides emissions of
The functional unit is a measure of the performance of a system. the reference fossil fuel proposed by the RED (83.8 kgCO2e/GJethanol)
It is very important because enables results for alternative prod- to ease the assessment of mitigation potentials.
C.A. Garca et al. / Applied Energy 88 (2011) 20882097 2093

Fig. 2. GHG emissions without LUC.

It was necessary to include an ethanol transportation phase in sugarcane crops. Under this hypothesis, all modalities generate
the Brazilian case to have consistent system boundaries with Mex- emissions between 47% and 267% above the fossil reference. How-
ico. Ethanol emissions in Brazil are 26.6 kgCO2e/GJethanol without ever, Johnson [7] estimates that it is possible to expand sugarcane
considering transportation to admixture plants. When transporta- cultivation on 2.9 million hectares of actual grasslands, both native
tion similar to that of Mexico is included, the ethanol emissions in- and cultivated. GHG emissions are much lower when sugarcane
crease to 27.5 kgCO2e/GJethanol. None of the modalities analyzed in expansions occur on grasslands, such as in the EMB, EDJ, and EDJE
Mexico achieve higher mitigation than the Brazilian case. modalities (Fig. 3).
The emission value for the reference fossil fuel was taken from
[1] because reliable values for life cycle emissions of gasoline in 4.2. Energy balance
Mexico do not exist at present.
GHG emissions are higher because of LUC when sugarcane cul- Primary energy balances can be expressed as the ratio between
tivation expands to new growing areas. This increase in GHG emis- primary fossil energy inputs to renewable energy output (Fig. 4).
sions is especially important when rainforest is converted to Only ethanol is produced under the various Mexican modalities,

Fig. 3. GHG emissions with LUC.


2094 C.A. Garca et al. / Applied Energy 88 (2011) 20882097

Fig. 4. Energy balances for ethanol production under Mexican and Brazilian modalities.

whereas ethanol and renewable electricity are obtained under Bra- CH4 and N2O. This eld burning contributed 29% of emissions un-
zilian and EDJE modalities. This presentation of results allows us to der EMF, EMBF and EMB cases and 31% of GHG emissions under
show energy consumption at each production stage. Negative val- EMB, EDJ and EDJE cases. This condition assumed that 95% of the
ues during the industrial phase of Mexican EDJE and Brazilian EDJE sugarcane elds were burned in Mexico. About 69% of the area
modalities are due to electricity exports from mills to the energy was burned in Brazil [16], and this source represented 35% of
grid. GHG emissions that corresponded to the agricultural phase. The
A different way to express energy balance results involves using third greatest contributor to GHG emissions was N2O from soils de-
the ratio of renewable energy output to fossil energy input (Table rived from nitrogen fertilization. Assuming that 1% of the nitrogen
4). A ratio of one means that no energy gain is achieved by the eth- content in fertilizers is volatilized as N2O, as documented by the
anol production because the renewable energy obtained is equal to IPCC, this source of GHG represents between 11% and 14% of agri-
the fossil energy consumed. Ratios less than one mean that ethanol cultural emissions in Mexico and 11% in Brazil. Although these per-
production provides less renewable fossil energy than the fossil en- centages are similar among Mexico and Brazil, the absolute
ergy consumed. Ratio values greater than one mean that energy emission levels are much lower in Brazil because overall emissions
gains are achieved in the process. are lower and less fertilizer is used. The relative high weight of fer-
tilizer-derived emissions has already been discussed in other stud-
ies [34,3840]. However, robust debate continues because recent
5. Discussion
research suggests that up to 5% of all nitrogen in fertilizers could
be emitted as N2O [41]. With this higher level, emissions in the to-
5.1. GHG emissions
tal agricultural phase would increase from 28.4 to 44.8 kgCO2e/
GJethanol in EMF and EMB cases, 30.748.5 kgCO2e/GJethanol in EMBF,
5.1.1. Agricultural phase
from 29.5 to 44.6 in EDJ and EDJE, and 21.631.5 in the Brazilian
Agricultural phase emissions were the most important in
case.
modalities EMB, EDJ, EDJE and Brazilian EDJE where they ac-
Emissions from diesel used by agricultural machinery represent
counted for 5681% of total emissions. However, agricultural phase
10% under modalities EMF and EMBF, 9% under modalities EMB,
emissions decreased to 2528% under modalities EMF and EMBF
EDJ and EDJE, and 16% under the Brazilian EDJE case.
where the industrial phase was the dominant GHG emissions
source.
Variables of substantial impact on GHG emissions were identi- 5.1.2. Industrial phase
ed. The variable with the greatest impact on GHG emissions in- GHG emissions from the industrial phase accounted for 70% and
volved the production of fertilizers, accounting for 3746% of 67% of total emissions under modalities EMF and EMBF, respec-
GHG emissions under Mexican modalities and 38% of GHG emis- tively. Conversely, GHG emissions from the industrial phase ac-
sions in Brazil. In both cases, most of these values corresponded counted for 33%, 7% and 3% of the GHG emissions under
to N2O emissions during fertilizer manufacture. The variable with modalities EMB, EDJ and EDJE, respectively. The difference be-
the second highest GHG emissions involved burning sugarcane tween modalities was due to the fuel used at mills. High emissions
elds to ease harvest and clear residues, resulting in emissions of corresponded to mills where low-pressure, low-efciency boilers
C.A. Garca et al. / Applied Energy 88 (2011) 20882097 2095

Table 4 Table 5
Energy balances of modalities for ethanol production in Mexico and Brazil, expressed CO2e emissions for different allocation methods with sugar as a co-product (kgCO2e/
as GJethanol/GJfossil. GJethanol).

EMF EMBF EMB EDJ EDJE Brazil EDJE Energy allocation Economic allocation
GJethanol/GJFossil 1.1 1.2 4.4 4.7 4.8 8.4 EMF 112.1 123.4
EMBF 109 118.2
EMF ethanol from C molasses using bagasse and fuel oil. EMB 50.6 59.3
EMBF ethanol from B molasses using bagasse and fuel oil.
EMB ethanol from C molasses using bagasse. EMF ethanol from C molasses using bagasse and fuel oil.
EDJ ethanol from direct juice using bagasse and without surplus electricity. EMBF ethanol from B molasses using bagasse and fuel oil.
EDJE ethanol from direct juice using bagasse and generating surplus electricity. EMB ethanol from C molasses using bagasse.
Brazil EDJE Brazilian ethanol from direct juice using bagasse as a fuel and gen-
erating surplus electricity.

tion values under all modalities, including 10% more under modal-
existed that did not allow for the generation of power surpluses ities EMF, 8% more under modality EMBF, and 17% more under
and required both bagasse and fuel oil as supplementary fuels for modality EMB (Table 5). Higher ethanol emissions have been re-
process heat generation (EMF and EMBF). ported by other authors when the economic value method was
Low emissions corresponded to mills that only used bagasse as chosen [17,34,35]. A 30% increase in the price of ethanol with sugar
fuel (EDJ, EDJE and Brazil). The modality EDJE had the lowest emis- prices kept constant resulted in an increase of GHG emissions allo-
sions in the industrial phase because surplus electricity was gener- cated to ethanol of 21% for EMF, 18% for EMBF, and 38% for EMB.
ated and delivered to the grid thereby avoiding emissions from This relationship shows the importance of the method for alloca-
power utilities. Assuming that high-pressure boilers are used, sur- tion, and its effect on the assessment of GHG mitigation when eth-
plus electricity of 8.8 kW h/tcane can be obtained, resulting in total anol is used in the transportation sector.
emissions of 36.8 kg CO2e/GJethanol. under modality EDJE. For higher
efciency boilers, a surplus of 50 kW h/tcane can be achieved, 5.2. Energy balance
resulting in total emissions of 24.9 kgCO2e/GJethanol. These gures
show advantages of the efcient use of residual biomass as a pro- 5.2.1. Agricultural phase
cess fuel as seen in Brazil [4,34,42]. Results of this study show that the main energy used corre-
sponded to industrial and agricultural phases. In modalities EMB,
5.1.3. Transportation phase EDJ and EDJE, the agricultural phase represents from 52% to 81%
Transportation of sugarcane from elds to mills has a share of of all the energy used. These results are similar to the Brazilian case
415% of the overall GHG emissions. In all Mexican modalities, where energy use in the agricultural phase was 83%. For modalities
similar results of about 5 kgCO2e/GJethanol were determined. The EMF and EMBF, the energy used in the agricultural phase was 0.12
geographical distance from elds to mills was the main factor in and 0.13 GJfossil/GJethanol. These results were similar to modality
similar transportation outcomes, and this condition was similar EMB with energy use values of 0.13 GJfossil/GJethanol. However,
in all Mexican cases. In addition, average travel distances were EMF and EMBF values represent only 1315% of the total energy
higher in Mexico than in Brazil, generating more emissions under used because fueloil is also used in the industrial phase. The pri-
this phase. mary energy used in the agricultural phase of the Brazilian case
Transportation from distilleries to admixture facilities gener- was lower than in Mexican modalities because lesser amounts of
ates emissions of little signicance. In Mexican modalities, its fertilizer are applied to Brazilian crops, resulting in lower energy
share was between 0.8% and 2.4% of total emissions. Doubling ratios.
the distance from mills to admixture plants would increase this
share to 1.64.8%. Thus, only long distances would produce signif- 5.2.2. Industrial phase
icant increases in emissions under this phase. For example, using Bagasse use as an energy source in the industrial phase was a
data from the California-GREET model [43] and assuming importa- major factor in the balance of primary energy as reported in other
tion of ethanol from Brazil to the US, GHG emissions reach 17% works [4,44]. This was easily seen in modalities EMF and EMBF
from transportation alone because the distance from distilleries where the industrial phase represented 82% of the fossil energy
to admixture plants is over 12,000 km. used over the entire life cycle because fuel oil is needed as supple-
mentary fuel in boilers. Under modality EMB, there was also sub-
5.1.4. Land use change stantial fossil fuel energy use because of the need to import
Emissions due to LUC are of great signicance when crop electricity from the national grid.
expansion is assumed under Mexican modalities. GHG emissions
under the three types of previous land use considered possible 5.2.3. Transport phase
for sugarcane crop expansion: grasslands, dry forest, and rainfor- Sugarcane transportation from elds to mills accounted for 3
est; results in very different emissions, ranging from a net GHG 17% of the primary fossil energy demand under Mexican modali-
reduction of 24% under modality EDJE on tropical dry forest lands ties. In the Brazilian case, fossil fuel demand and resulting GHG
to a net GHG increase of 267% under modality EMBF on rainforest emissions were lower than in Mexico because growing areas are
relative to reference fossil fuel emissions. This wide disparity in closer to industrial processing plants and high tonnage trucks
emissions clearly shows that the expansion of sugarcane cultiva- make the Brazilian transportation system more efcient. The trans-
tion to lands of high carbon content would actually increase GHG portation of ethanol to admixture plants was of marginal signi-
emissions as already concluded by other authors [3032]. cance in Mexico, representing 13.7% of all fossil energy demand.

5.1.5. Allocation 6. Conclusion


Allocation values by energy content were 26.8 MJ/kg for ethanol
and 16.5 MJ/kg for sugar, while by economic value the estimated Three modalities for ethanol production in Mexico showed low-
allocation values were 0.6 USD/kg for ethanol and 0.286 USD/kg er GHG emissions than the reference fossil fuel. The best modality
for sugar. Allocation by economic value resulted in higher alloca- was EDJE with 36.8 kgCO2e/GJethanol., followed by EDJ with
2096 C.A. Garca et al. / Applied Energy 88 (2011) 20882097

38.4 kgCO2e/GJethanol. None of the Mexican modalities achieved [5] PRONAC (Programa Nacional de la Agroindustria de la Caa de Azcar),
Secretara de Agricultura, Ganadera, Desarrollo Rural, Pesca y Alimentacin
lower emissions than the Brazilian case (27.5 kgCO2e/GJethanol) this
(SAGARPA), Mxico; 2007.
is due to the following reasons: because less fertilizer is applied to [6] UNC (Unin Nacional de Caeros); 2009. <http://www.caneros.org.mx/>.
sugarcane in Brazil than in Mexico, due to the smaller proportion of [7] Johnson T, Alatorre C, Romo Z, Liu F. (Mxico: Estudio para la Disminucin de
sugarcane areas that were burned to facilitate manual harvest, due Emisiones de Carbono, MEDEC). The International Bank for Reconstruction
and Development, World Bank; 2009.
to the sorter distance of sugarcane transport, and because fossil [8] LPDB (Ley de Promocin y Desarrollo de los Bioenergticos), Diario Ocial de la
fuel energy is not used in the industrial phase. Emissions due to Federacin, February, Mxico; 2008.
LUC associated with the expansion of crop areas are the main con- [9] SENER (Secretara de Energa), Programa de Introduccin de Bioenergticos.
Mxico; 2008.
tributors to the total GHG emissions in Mexico, especially when [10] Shapouri H, Dufeld J, Wang M. The energy balance of corn ethanol: an update.
this expansion occurs on lands of high carbon stock such as tropical Agricultural economics report no. 813, USDA; 2002.
rain forests. [11] Patzek T. Thermodynamics of the corn-ethanol biofuel cycle. Crit Rev Plant Sci
2004;23(6):51967.
Sugarcane eld burning to ease manual harvesting is another [12] Pimentel, Patzek. Ethanol production using corn, switchgrass, and wood;
important source of emissions, suggesting that mechanized har- biodiesel production using soybean and sunower. Nat Resour Rev
vest without previous burning is preferable. However, mechanized 2005;14:6576.
[13] IFEU, Quirin M, Grtner SO, Pehnt M, Reinhardt GA. CO2 mitigation through
harvesting increase the amount of diesel fuel needed for harvest biofuels in the transport sector. Status and perspectives, Germany; 2004.
operations. Other important sources of emissions are related to [14] Dias de Oliveira M, Vaughan Jr M, Rykiel E. Ethanol as a fuel: energy, carbon
N2O volatilization from fertilizer applications and production. En- dioxide balances, and ecological footprint. Bioscience 2005;55:50893.
[15] E Farrell A, Pelvin RJ, Turner BT, Jones AD, OHare M, Kammen DM. Ethanol can
hanced efciency of fertilization and increased biofertilizers to
contribute to energy and environmental goals. Science 2006;311:5068.
substitute for chemical nitrogen fertilizers would help reduce [16] Macedo I, Seabra J, Silva J. Green house gases emissions in the production and
emissions. Diesel fuel use had an important share in emissions as use of ethanol from sugarcane in Brazil: the 2005/2006 averages and a
shown in the Brazilian case where diesel consumption was high prediction for 2020. Biomass Bioenergy 2008;32:58295.
[17] Gnansounou E, Duriart A, Villegas J, Panichelli L. Life cycle assessment of biofuels:
because of higher degrees of harvest mechanization. energy and greenhouse gas balances. Bioresour Technol 2009;100:491930.
The efcient use of bagasse as a fuel in the industrial phase was [18] Elsayed MA, Matthews R, Mortimer ND. Carbon and energy balances for a
a major factor in achieving emission reductions. Results clearly range of biofuels options. Project No. B/B6/00784/REP URN 03/836. Project
carried out as part of the DTI Sustainable Energy Programme. <http://
show that modalities demanded fuel oil as a supplementary fuel www.rms.lv/bionett/Files/File/BioDE2003101%20DTI%20biodiesel%20LCA%
to generate high GHG emissions. In the transportation phase, the 20study.pdf>.
most important factor was the distance from elds to mills, fol- [19] Thamsiriroj T, Murphy JD. Is it better to import palm oil from Thailand to
produce biodiesel in Ireland than to produce biodiesel from indigenous Irish
lowed by distance from mills to admixture plants. Thus, more com- rape seed? Appl Energy 2009;86:595604.
pact growing areas of sugarcane around distilleries and higher [20] INE (Instituto Nacional de Ecologa). Direccin de Economa Ambiental.
efciency of transportation vehicles would help reduce emissions Propuesta de certicacin como instrumento para regular la produccin y
uso de biocombustibles en Mxico. Propuesta de Consultora presentado por C.
from the transportation phase. A. Corporacin de Servicios Profesionales, Mxico; 2008.
The energy balance for all modalities was positive, resulting in [21] INE. Direccin de Economa Ambiental. Anlisis de tres casos especcos para
energy ratios greater than one. This implies that more ethanol en- produccin de biodiesel en Mxico a partir de soya, utilizando el modelo INE
para evaluacin energtica y ambiental. Propuesta de Consultora presentado
ergy was obtained than fossil energy consumed, and all modalities
por C. A. Corporacin de Servicios Profesionales, Mxico; 2008.
evaluated were energetically cost-efcient. Modalities with the [22] INE. Direccin de Economa Ambiental. La certicacin como instrumento para
highest energy ratio in Mexico were EDJE (4.8 GJethanol/GJfossil) fol- regular la produccin de los biocombustibles en Mxico. Propuesta de
lowed by EDJ (4.7 GJethanol/GJfossil), both of which were lower than Consultora presentado por C. A. Corporacin de Servicios Profesionales,
Mxico; 2008.
the Brazilian case (8.4 GJethanol/GJfossil). The main components of [23] SENER, BID, GTZ. Potenciales y Viabilidad del uso de Bioetanol y Biodiesel para
primary fossil energy consumption were as follows: fertilizer pro- el Transporte en Mxico, Mxico; 2006.
duction and diesel oil use by machinery in the agricultural phase; [24] Enriquez PM. Prospectiva de la Cogeneracin de la Agroindustria de la Caa de
azcar en Mxico. Presentation in: VI Reunin Nacional RED Mexicana de
fuel oil use in the industrial phase; and diesel oil use by sugarcane Bioenerga y Simposio Internacional de Proyectos Bioenergticos, Mxico;
and ethanol trucks in the transportation phase. Energy ratios in 2009.
Mexico could be improved by implementing the following: [25] Fehrenbach H, Giegrich J, Reinhardt G, Schmitz J, Sayer U, Gretz M, et al.
Criteria for a sustainable use of bioenergy on a global scale. Environmental
Enhancing fertilizer efciency; limiting fuel usage to bagasse in research of the federal ministry of the environment, nature conservation and
the industrial phase; compacting sugarcane growing areas; and nuclear safety. Research report 206 41 112 UBA-FB 001176/E, Germany; 2008.
increasing the efciency of transportation systems. <http://www.ifeu.de/nachhaltigkeit/pdf/IFEU_Report_criteria_for_sustainable_
bioenergy_UBA.pdf>.
[26] IPCC, Third assessment report climate change 2001 complete online versions.
<http://www.grida.no/publications/other/ipcc_tar/?src=/climate/ipcc_tar/wg1/
Acknowledgement
248.htm>.
[27] IPCC, Guidelines for national greenhouse gas inventories, vol. 4. Energy.
This study was supported by: Deutsche Gesellschaft fuer Agriculture, forestry and other land use; 2006.
Technische Zusammenarbeit (GTZ) GmbH, and projects Fondo [28] SENER. Balance Nacional de Energa 2007. Secretara de Energa. Subsecretara
de Planeacin Energtica y Desarrollo Tecnolgico. Direccin General de
Sectorial CONACYT-SENER-Sustentabilidad Energetica No. 117808 Informacin y Estudios Energticos, Mxico; 2008.
and PAPIIT-UNAM No. IN118208. [29] IPCC, Guidelines for national greenhouse gas inventories, vol. 2. Energy; 2006.
[30] Searchinger T, Heimlich R, Houghton RA, Dong F, Elobeid A, Fabiosa J, et al. Use
of US croplands for biofuel increases greenhouse gases through emissions from
References land use change. Science 2008;319:123840.
[31] Fargione J, Hill J, Tilman D, Polasky S, Hawthorne P. Land clearing and the
biofuel carbon debt. Science 2008;319:12358.
[1] EUC (Commission of the European Communities). DIRECTIVE 2009/28/EC OF
[32] Silalertruksa T, Gheewala SH, Sagisaka M. Impacts of Thai bio-ethanol policy
THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 23 April 2009 on the
target on land use and greenhouse gas emissions. Appl Energy 2009;86:S1707.
promotion of the use of energy from renewable sources and amending and
[33] INEGEI. Inventario Nacional de Emisiones de Gases de Efecto Invernadero
subsequently repealing Directives 2001/77/EC and 2003/30/EC. <http://eur-
19902002. Secretara de Medio Ambiente y Recursos Naturales. Instituto
lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:140:0016:0062:EN:PDF>.
Nacional de Ecologa, Mxico; 2006.
[2] RFA (Renewable Fuels Association); 2009. <http://www.ethanolrfa.org/
[34] Brjesson P. Good and bad bioethanol from a greehouse gas perspective
industry/statistics/>.
what determines this? Appl Energy 2009;86:58994.
[3] C Escobar J, Lora ES, Venturini OJ, Yez EE, Castillo EF, Almazan O, et al.
[35] Reijnders L, Huijbregts MAJ. Biofuels for Road Transport. A Seed to Wheel
Biofuels: environment, technology and food security. Renew Sust Energy Rev
Perspective. Springer; 2009.
2009;13:127587.
[36] ISO. ISO Norm 14040:2006 and 140444:2006. Life cycle assessment: principles
[4] Goldemberg J, Teixeira CS, Guardabassi P. The sustainability of ethanol
and framework. Environmental Management; 2006.
production from sugarcane. Energy Policy 2008;36:208697.
C.A. Garca et al. / Applied Energy 88 (2011) 20882097 2097

[37] ISO. ISO Norm 140444:2006. Life cycle assessment: requirements and [41] Crutzen PJ, Mosier AR, Winiwarter KASmithW. N2O release from agro-biofuel
guidelines. Environmental Management; 2006. production negates global warming reduction by replacing fossil fuels. Atmos
[38] E4Tec, The RSB GHG accounting scheme. Feasibility of a meta-methodology Chem Phys Discuss 2007;7:11191205.
and way forward. Version 4.1; 2009. <http://cgse.ep.ch/webdav/site/cgse/ [42] Nguyen TLT, Hermansen JE, Sagisaka M. Fossil energy savings potential of
shared/Biofuels/Documents%20and%20Resources/09-10-08_E4Tech%20 Report% sugar cane bio-energy systems. Appl Energy 2009;86:S1329.
20GHG%20Accounting_V4%201_08October 09.pdf>. [43] California Air Resources Board. CARB. Detailed California-Modied GREET
[39] Winrock International. The Impact of Expanding Biofuel Production on GHG Pathway for Brazilian Sugarcane Ethanol. Sacramento, CA. California Air
emissions. White Paper #1: Accessing and interpreting existing data; 2009. Resources Board: 54. 246; 2009.
<http://www.globalbioenergy.org/uploads/media/0904_Winrock_International_-_ [44] Nguyen TLT, Gheewala SH, Garivait S. Full chain energy analysis of fuel ethanol
White_paper__1_GHG_implications_ biofuel.pdf>. from cane molasses in Thailand. Appl Energy 2008;85:72234.
[40] Achten WMJ, Almeida J, Fobelets V, Bolle E, Mathijs E, Singh VP, et al. Life cycle
assessment of jatropha biodiesel as transportation fuel in rural India. Appl
Energy 2010;87:365260.