Article pubs.acs.org/est
Life Cycle Greenhouse Gas Emissions of Sugar Cane Renewable Jet Fuel Marcelo Moreira,†,‡ Angelo C. Gurgel,§ and Joaquim E. A. Seabra*,‡ †
Agroicone, Av. General Furtado do Nascimento, 740 cj. 8, São Paulo, SP Brazil 05465-070 Faculdade de Engenharia Mecânica, UNICAMP (University of Campinas), Cidade Universitária “Zeferino Vaz”, Rua Mendeleyev 200, Campinas, SP Brazil 13083-860 § São Paulo School of Economics, Fundaçaõ Getulio Vargas, Rua Itapeva 474, São Paulo, SP Brazil 01332-000 ‡
ABSTRACT: This study evaluated the life cycle GHG emissions of a renewable jet fuel produced from sugar cane in Brazil under a consequential approach. The analysis included the direct and indirect emissions associated with sugar cane production and fuel processing, distribution, and use for a projected 2020 scenario. The CA-GREET model was used as the basic analytical tool, while Land Use Change (LUC) emissions were estimated employing the GTAP-BIO-ADV and AEZ-EF models. Feedstock production and LUC impacts were evaluated as the main sources of emissions, respectively estimated as 14.6 and 12 g CO2eq/MJ of biofuel in the base case. However, the renewable jet fuel would strongly benefit from bagasse and trashbased cogeneration, which would enable a net life cycle emission of 8.5 g CO2eq/MJ of biofuel in the base case, whereas Monte Carlo results indicate 21 ± 11 g CO2eq/MJ. Besides the major influence of the electricity surplus, the sensitivity analysis showed that the cropland−pasture yield elasticity and the choice of the land use factor employed to sugar cane are relevant parameters for the biofuel life cycle performance. Uncertainties about these estimations exist, especially because the study relies on projected performances, and further studies about LUC are also needed to improve the knowledge about their contribution to the renewable jet fuel life cycle.
■
INTRODUCTION The global concern about climate change, associated with the world dependence on fossil fuels, price volatility, and uncertainties of oil supply, has motivated a growing interest for renewable energy sources, more particularly in the form of biofuels.1 This can be extended to the aviation sector, which presents few alternatives to conventional fuels. According to the IPCC,2 estimates of CO2 emissions from global aviation increased by a factor of about 1.5, from 330 MtCO2/yr in 1990 to 480 MtCO2/yr in 2000, and accounted for about 2% of total anthropogenic CO2 emissions. As aviation CO2 emissions are projected to continue to grow strongly, the use of biofuels has been encouraged, among other options, as a way to reduce the impacts on climate.3,4 The air transportation industry aims to reduce its CO2 emissions by half until 2050.5 Some of this should be achieved through technology improvements in air traffic management, aircraft design, and engine efficiency, but alternative fuels are being considered as the main component of this effort. As this goal will be hardly achieved without the use of biofuels, several jet fuel companies are investing in the development of proper biomassbased fuels and investigating their environmental attributes. Given the potential benefits provided by biofuels, several countries are implementing policies to foster their wider use in order to improve the sustainability of the transport sector. The greenhouse gases (GHG) emissions in the production, transportation, and final use of a biofuel, and their relative © 2014 American Chemical Society
magnitude compared to the emissions of the fossil counterpart are one of the most important parameters for its acceptance by regulators.6,7 Methodologies applied for emissions evaluation are not yet sufficiently well established to reach consensus, but local regulatory agencies are trying to advance in defining rules to be used. This is promoting a great effort to improve the knowledge, leading to much sounder regulations in the future. So far, two approaches to calculate direct and indirect GHG emissions related to biofuels production have been used for regulatory purposes in the U.S.: one developed by the California Air Resources Board (CARB)8 for the Low Carbon Fuel Standard (LCFS), and the U.S. Environmental Protection Agency (EPA)9 approach used for the Renewable Fuel Standard (RFS2). In Europe, the Directive on the promotion of use of energy from renewable sources lays down a methodology to calculate the GHG emission savings from the use of biofuels and provides default savings that can be adopted in case the emissions from LUC are equal to or less than zero.10 All these methods shall evolve in the near future, as well as the criteria for biofuels acceptance. Further, the biofuel’s production parameters will also change in the next years, since the advanced options are at initial stages of development, starting now the first Received: Revised: Accepted: Published: 14756
October 10, 2013 November 5, 2014 November 6, 2014 November 24, 2014 dx.doi.org/10.1021/es503217g | Environ. Sci. Technol. 2014, 48, 14756−14763
Environmental Science & Technology
Article
commercial production. Therefore, on several occasions the evaluation of a biofuel will require analyses for projected scenarios, based on expected technical developments for the next years. This study evaluated the life cycle GHG emissions (carbon footprint) of a renewable jet fuel produced from sugar cane in Brazil through Amyris’ proprietary technology platform.11 In June 2012, the renewable jet fuel was tested in a demonstration flight between the cities of Campinas and Rio de Janeiro. The technology is now employed at commercial scale at a facility in Brazil and remains under development and testing at lab and pilot scales. The performance parameters assumed in this study refer to the conditions projected for a 2020 technological scenario, considering the technology advances at commercial scale expected by Amyris. The analysis was based on the consequential approach adopted by the CARB for the LCFS. Emissions related to sugar cane production in Brazil were based on the values given in the CARB report12 for the base case, while LUC emissions were recalculated to better represent the expansion of the renewable fuel in Brazil. The possibility to clearly separate the direct and indirect effects is seen as an important advantage of the CARB methodology for the purpose of this study. These emissions were added to the estimated emissions from sugar cane processing, fuel production, and fuel use, leading to the biofuel life cycle emissions.
However, as the pathway for the renewable jet fuel production is not incorporated in the CA-GREET model, external calculations were performed to estimate the life cycle GHG emissions, though still relying on parameters given by the CA-GREET model. Because detailed values for sugar cane production in Brazil are available in the Staff Report,8,12 which already account for the direct effects of sugar cane production, the emissions per tonne of sugar cane were maintained in the base case for the purpose of this analysis. The mechanized harvest scenario was the selected pathway for this study because it is the current practice of the new sugar cane mills, as well as the expected conditions of practically all sugar cane areas in 2020. Additionally, the scenario considered in this study assumes the collection of sugar cane straw (trash), so the emissions from sugar cane production and transport were added to the emissions related to trash transportation, using the transportation parameters provided in the GREET model. The LUC emissions were also estimated using the same methodology employed in CARB8 and the later revised version.15 It combines an economic general equilibrium model with a GHG emissions model. The economic model (GTAP-BIO-ADV)16,17 is able to project LUC at the global level from the expansion of some specific biofuel type, considering the domestic and international market mediated effects.18,19 The GHG emissions model is known as the Agro-Ecological Zone Emission Factor (AEZ-EF) model,20 which transforms the GTAP-BIO-ADV outputs (LUC in hectares) into GHG emissions. More details about the LUC assessment are provided below. Product System and Data Set. This analysis involved the evaluation of a renewable jet fuel made from sugar cane using Amyris’ proprietary technology platform,21 which is intended to be compliant with Jet A/A-1 fuel specifications when blended with fossil fuel (blends up to 10% of the renewable fuel meet the updated ASTM D7566 for Jet A/A-1).22 Farnesane is the renewable hydrocarbon molecule that will compose this alternative fuel. The production pathway comprises the fermentation of plant sugars into farnesene (a 15-carbon long-chain branched hydrocarbon) using specialized yeasts developed by Amyris. The farnesene forms a separate phase on the top of the fermentation broth, hence facilitating subsequent recovery and purification. The farnesene produced at the sugar cane processing plant is then transported to a hydrogenation plant where it is converted to farnesane, a hydrocarbon molecule. The farnesane is then transported to the airport fuel tanks in Brazil to be blended with regular jet fuel. Figure 1 shows
■
MATERIALS AND METHODS Methodological Approach. The main goal of this study was to assess the life cycle GHG emissions (in g CO2eq/MJ) of a renewable jet fuel under a consequential approach,13 using a similar method employed by the CARB staff in the regulatory process of the LCFS.8,12 Similarly, a well-to-tank (WTT) assessment was developed including all steps from feedstock production up to final product distribution. The study included the direct and indirect emissions associated with sugar cane production, fuel processing, and fuel distribution for a projected 2020 scenario. Those results were added to the use phase of the biofuel, burning at the aircraft’s engine, leading to the total “cradle-to-grave” results. The life cycle emissions were assessed using the CA-GREET model as the basic analytical tool, with the same methodological approach described in CARB.12 This model is an adapted version of the “Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation” (GREET) model,14 version 1.8b, which was modified to better represent California-specific conditions, parameters, and data. The major changes incorporated into the CA-GREET model can be found in CARB.8,12
Figure 1. Life cycle stages and assumptions/sources (shown in parentheses). 14757
dx.doi.org/10.1021/es503217g | Environ. Sci. Technol. 2014, 48, 14756−14763
Environmental Science & Technology
Article
Figure 2. Schematic representation of the farnesene production stage based on syrup fermentation.
SOx, NMVOC, particulates, and other trace components including hazardous air pollutants. Little or no N2O emissions occur from modern gas turbines.26 Methane (CH4) may be emitted by gas turbines during idle and by older technology engines, but recent data suggest that little or no CH4 is emitted by modern engines.25 In this study, CO2 was treated as biogenic and disregarded from the analysis, so only the emissions of N2O and CH4 were considered in the GHG emissions accounting. Because no specific emission factors are available for farnesane, IPCC’s Tier 1 default values were used in this analysis. LUC Assessment. As described above, the emissions from LUC were estimated using the GTAP-BIO-ADV and AEZ-EF models. The AEZ-EF model considers not only the above and below ground carbon stocks but also assumptions about carbon loss considering mode of conversion, management practices, number and types of species, carbon in harvested wood products and char, and foregone sequestration. The combination of GTAP-BIO-ADV and AEZ-EF models works as follows. An increment of volume of some biofuel type in some specific country is applied as a shock in the GTAP model considering the economic environment of year 2004. The GTAP model calculates all the LUC (direct and indirect) necessary to accommodate such increment in 19 regions of the world (each region has at least one AEZ) representing the global economy. These changes are passed to the AEZ-EF model, which calculates the total global emissions from LUC.27 Finally, the total emissions from the AEZ-EF model together with the volume of fuel and its characteristics are used to calculate the LUC factor (direct + indirect LUC) in g CO2eq/MJ over a 30-year amortization period. In this study we simulated the impact of the production of 150 million liters of farnesane in Brazil produced from sugar cane, considering a conversion factor of 20.09 tonnes of sugar cane to produce one cubic meter of farnesane. The GTAP-BIO-ADV and AEZ-EF models have their own databases, but some changes were implemented to better reflect the particular case of the sugar cane-based farnesane produced in Brazil. First, it was necessary to simulate a more appropriate shock in the GTAP model which could better represent the likely production of Amyris farnesane for renewable jet fuel in 2020 (150 million liters). Because the GTAP model does not have a built-in farnesane sector, the 150 million liters were converted, as a surrogate, into an equivalent amount of sugar cane (20.09 tonnes of sugar cane per m3 of farnesane), in order to perform the shock in the GTAP-BIO-ADV model. Regarding the GTAP model, the endogenous “yield elasticity related to cropland−pasture” is a central assumption concerning the potential for pasture intensification and, as consequence,
a schematic representation of the life cycle stages under study and the main assumptions. The provision of sugar cane sugars for fermentation comprises cane juice concentration up to a 65 Brix syrup to form the fermentation must (Figure 2). The yeasts employed in the fermentation are recovered and dried to form a valueadded coproduct (33% protein content), and the liquid effluent (vinasse) is sent to the sugar cane field for fertirrigation. Table 1 summarizes the main parameters of the mass and energy Table 1. Parameters of Farnesene Production Stage (Per Tonne of Farnesene) parameter
unit
value
sugar cane t chemicals nutrients kg sodium hydroxide (50%) kg antifoam and demulsifier kg other chemicals kg Outputs (Coproducts) electricity surplus kWh yeast kg
27.2
Inputs
56.1 10.3 0.8 1.6 2598 139
balances projected for the 2020 scenario, according to the targets established by Amyris specialists. As for the background life cycle inventories, parameters were retrieved respectively from the CA-GREET model and the Ecoinvent database.23 In the projected scenario the whole energy required for farnesene production would be supplied by the bagasse-fueled cogeneration plant. Additionally, it was assumed that 40% of the sugar cane trash produced in the field would be used as supplementary fuel to bagasse. For the cogeneration plant, a similar configuration of the current high pressure systems was adopted,24 which are equipping all new sugar cane mills in Brazil and several retrofitted units. For sugar cane preparation and crushing, completely electrified systems were considered. In the hydrogenation plant, H2 provided from natural gas reforming is employed for the hydrogenation of farnesene into farnesane. The excess H2 supplied to the farnesene hydrogenation process is subsequently combusted in other processes providing additional energy. Once converted, the farnesane-based drop-in fuel is ready for final transport/distribution and use. At the use phase, GHG emissions arise from the combustion of the renewable jet fuel and they should therefore be accounted for in the life cycle assessment. As reported by IPCC,25 aircraft engine emissions are roughly composed of about 70% CO2, a little less than 30% H2O, and less than 1% each of NOx, CO, 14758
dx.doi.org/10.1021/es503217g | Environ. Sci. Technol. 2014, 48, 14756−14763
Environmental Science & Technology
Article
Treatment of Coproducts. In line with CARB,12 the substitution approach was the method adopted to deal with the coproducts, i.e. the emissions of producing an equivalent amount of coproduct were subtracted from the life cycle inventory of the renewable jet fuel. The electricity surplus coproduced at the farnesene production stage is the main coproduct in the chain, and the appropriate emission factor for crediting was calculated based on the CA-GREET model and CARB.12 The marginal Brazilian electricity (natural gas) was the assumed electricity mix displaced by the bagasse-fired exported electricity produced at the sugar cane mill.12 Dried yeasts are also a coproduct of farnesene production. The same substitution approach was applied, and emissions credits were assigned to farnesene. The emissions credits were estimated assuming that soy meal would be the alternative feedstuff product, for which an equivalence ratio was calculated based on the protein content of each product. Parameters for the soy meal were retrieved from the CA-GREET model. Because direct and indirect LUC emissions for soybeans are not accounted for, emissions credits may be therefore underestimated. At the hydrogenation stage, the excess H2 coming from the hydrogenation reactor is combusted, which avoids the use of other fuels. Here it was assumed that the consumption of an energy equivalent amount of natural gas would be avoided, and the associated emissions credits (evaluated using the CA-GRRET model) were assigned to farnesane. It must be noted, however, that such practice is dependent on the process
mitigation of the possible indirect effects of crop expansion on forestlands. This is due to the fact that such parameter drives the changes in the yields of the land category cropland−pasture because of the increases in the rent of this type of land. Its default value in GTAP-BIO-ADV is 0.4 in U.S. and 0.2 in Brazil, which means that, under a 1% increase in the cropland−pasture land rent in the U.S., a 0.4% increase in its yields should be observed. The U.S. livestock sector is already much more intensified than that in Brazil, suggesting that Brazil has higher potential to increase yields than the U.S. Productivity gains were responsible for 79% of growth in beef production in Brazil.28 As the average stocking rate in the country is still relatively low (1.08 head per hectare),29 there is room for high gains in productivity due to intensification practices. Another fact contributing to intensification practices in the livestock sector under higher cattle and beef prices is the increase in monitoring and enforcement of environmental laws, which reduces the scope for expansion of cattle production at the extensive margin. Although Brazil has a much higher tendency to intensify pastureland than the U.S., there is no final recommendation regarding the most appropriate values for the endogenous yield elasticity related to cropland−pasture. We opted to keep the original elasticities from Tyner et al.16 and show results from different alternatives in the sensitivity analysis. Higher cropland−pasture yield elasticity in Brazil should increase livestock intensification and decrease emissions from LUC. The AEZ-EF model relies on IPCC 2006 guidelines25 to calculate regional soil carbon changes when land use is changed. According to this method, the carbon stock change is estimated using reference carbon stocks and three stock change factors (land-use, management regime, and input of organic matter). The original AEZ-EF model considers cropland as one single worldwide category, using the land-use and management factors representing (a) long-term cultivation, (b) full tillage, and (c) medium input. Even though this representation can be assumed as the best approximation for annual crops like corn, it is inappropriate for sugar cane. Differently from annual crops, sugar cane is harvested every year but is replanted only after 5 or 6 years. Harris et al.29 and Marelli et al.30 stated that it is important to distinguish sugar cane from annual crops when calculating land-use related GHG emissions for biofuels and considered sugar cane as a perennial crop. Above- and below-ground carbon stocks for sugar cane are significantly higher than those for annual crops, and similar to those of perennial crops.31−33 It is therefore more appropriate to consider sugar cane as a perennial crop than a long-term cultivated crop within the IPCC guidelines. With respect to management factors, green harvesting of sugar cane (mechanical harvest without preharvest field burning) is the dominant harvesting practice today.34,35 Preharvest burning of sugar cane fields is set to phase out in the vast majority of the sugar cane fields in Brazil by 2017 and restricted by the national sugar cane agroecological zoning.36,37 Unburnt sugar cane harvesting leaves greater soil coverage after harvesting than the traditional burning practice, leading to a mean annual C accumulation of 1.5 Mg/ha·year at 30 cm depth.38 Following IPCC guidelines, “Reduced till” is therefore more appropriate to represent sugar cane management, rather than “full till”. Because the specific case of a sugar cane-based renewable jet fuel is under analysis, specific sugar cane parameters (for land use and tillage emission factors) have been used wherever possible to identify sugar cane expansion instead of generic cropland expansion.
Table 2. Parameters for the Monte Carlo Analysis parameter LUC factor agrochemicals production and use impacts sugar cane farming emissionsb sugar cane transportation emissions farnesene yield electricity surplus farnesene transportation distance
distribution
units
valuesa
lognormal normal
g CO2eq/MJ (2.27; 0.63) kg CO2eq/t cane (15.7; 5.5)
normal
kg CO2eq/t cane (22.4; 5.8)
normal
kg CO2eq/t cane (2.6; 0.9)
minimum extreme L/t cane minimum extreme kWh/t cane normal km
(44.9; 1.3) (96; 11) (150; 30)
a
Normal distribution (mean; standard deviation), log-normal (log mean; log standard deviation), minimum extreme (likeliest; scale). b Include emissions from fuels used in sugar cane farming and from residues applied to the soil.
Table 3. Renewable Jet Fuel Life Cycle Emissions life cycle stage sugar cane production land use change (luc) farnesene production chemicals biomass combustion electricity surplus yeast farnesene transport hydrogenation fuel transport fuel use net emissions 14759
GHG emissions (g CO2eq/MJ) 14.6 12.0 2.8 3.5 −33.9 −0.6 0.3 9.0 0.3 0.6 8.5
dx.doi.org/10.1021/es503217g | Environ. Sci. Technol. 2014, 48, 14756−14763
Environmental Science & Technology
Article
Figure 3. (a) LUC and (b) LUC emissions due to farnesane production, by GTAP regions. F = forest; P = pasture; C = cropland; CP = cropland− pasture.
the literature39,40 and the range of LUC estimations from this work.
configuration of the hydrogenation plant, which might be set to recycle the hydrogen excess (thereby reducing the input of natural gas), instead of using it as fuel. Sensitivity and Uncertainty Analyses. Uncertainties about the life cycle performance of the renewable jet fuel exist, especially because the study relies on projected conditions for future scenarios. A sensitivity analysis was developed to show the impact of variations of the main parameters on the net life cycle emissions, including different production volumes and its effect on LUC values. Additionally, the uncertainties of the model were assessed through a Monte Carlo analysis using 5000 trials. Uncertainty distributions were generated for seven parameters (Table 2) based on information from
■
RESULTS AND DISCUSSION On the basis of the CA-GREET default values (apart from the LUC impacts), the production and use of agrochemicals is the leading source of GHG emissions associated with sugar cane production and transport. Assuming the mechanized harvest practice, these sources account for more than two-thirds of the total 24.2 kg CO2eq emitted per tonne of sugar cane available at the mill gate (base case scenario, excluding LUC), although it must be mentioned that compared to the Center-South averages reported in Seabra et al.,39 the overall emissions are optimistic. 14760
dx.doi.org/10.1021/es503217g | Environ. Sci. Technol. 2014, 48, 14756−14763
Environmental Science & Technology
Article
Figure 4. Sensitivity analysis of the renewable jet fuel life cycle emissions.
the typical range of values reported in the literature for sugar cane cultivation.39,40 Different volumes for the projected production were also tested, especially concerning the impact that higher expansion could have on land use emissions. As shown in Figure 5,
Those emissions are completely offset by the avoided emissions due to the coproduction of electricity at the farnesene production stage (∼34 g CO2eq/MJ), which leads to a net life cycle emissions of 8.5 g CO2eq/MJ of jet fuel, including LUC impacts (Table 3). Even though the coproduction of yeast also leads to emissions avoidance, its contribution is very small. Following the LUC modeling described above, the production of 150 million liters of renewable jet fuel would require 3.014 million metric tonnes of sugar cane, which would lead to an expansion of the sugar cane area of 37.6 thousand hectares in Brazil, while reducing all other crop areas. Worldwide, total direct and indirect emissions would add up to 1.8 × 106 t CO2eq, hence producing a LUC factor of 12 g CO2/MJ for a 30-year amortization period. World LUC by GTAP regions and type of land conversion is presented in Figure 3a, and Figure 3b presents the respective emissions. After feedstock production and LUC impacts, emissions from the hydrogenation step are the most relevant in the renewable jet fuel life cycle. The CO2 emitted from natural gas reforming represents an important GHG source, but some part of those emissions is offset by the displacement of natural gas due to the combustion of the H2 excess. The environmental burden associated with farnesene and, later, farnesane transport is relatively small since short transportation distances are involved. It was assumed in this study that the whole field-to-tank biofuel chain would occur within the Brazilian territory, more specifically within the State of São Paulo; but in case longer distances are involved, possibly comprising overseas exports, more significant impacts may then be expected. As discussed above, the electricity surplus from farnesene production plays a special role in the biofuel life cycle and any variation in power generation leads to a substantial impact on the net emissions. The tornado chart of Figure 4 shows that if the electricity surplus is close to the current levels of new sugar cane mills (i.e., equivalent to 50 kWh/t cane instead of 96 kWh/t), the biofuel life cycle emissions jump to 24.7 g CO2eq/MJ. Parameters affecting the LUC results are also critical for the overall performance of the renewable jet fuel, as captured in the sensitivity analysis with respect to the definition of sugar cane as a perennial or long-term cultivated crop (which affects the soil carbon stocks) and the crop−pasture elasticity (which affects the land substitution pattern). Farnesene yield impacts both LUC and direct fuel production emissions, but within a narrower range. Significant sensitivity was also verified with respect to the emissions from sugar cane production, considering
Figure 5. LUC factor for different production scenarios.
although the estimated emissions increase when more enthusiastic scenarios are considered, the LUC results have a small response to the size of the production once additional land use emissions are compensated by the additional amount of energy produced. Only the most unrealistic scenario of 7 billion liters (close to the total jet fuel consumption in Brazil in 2011) presents significant deviations from the projected base case scenario, with a LUC factor of 17.2 g CO2eq/MJ. As expected, the estimation of the renewable jet fuel life cycle performance is highly uncertain (Figure 6), while the results for the base case are rather optimistic. The results of the Monte Carlo analysis indicate life cycle emissions of 21 ± 11 g CO2eq/MJ (mean ± SD), with substantial influence from the LUC factor. The second most influential parameter on the variations is the electricity exports, which are basically affected by the energy self-consumption in farnesene production. The distributions assumed for emissions from sugar cane farming and the impacts from agrochemicals production and use have practically the same effect on the variations, while sugar cane and farnesene transportation have minor influence. Despite those uncertainties, it is possible to say that the renewable jet fuel presents a substantial potential to mitigate the GHG emissions of the aviation sector considering that the life cycle emissions of its fossil counterpart usually lies within the 80−95 g CO2eq/MJ range.23,41 However, in order to have 14761
dx.doi.org/10.1021/es503217g | Environ. Sci. Technol. 2014, 48, 14756−14763
Environmental Science & Technology
Article
(9) U.S. EPA. Renewable Fuel Standard Program (RFS2) Regulatory Impact Analysis; Assessment and Standards Division, Office of Transportation and Air Quality, U.S. Environmental Protection Agency: Washington, DC, February, 2010. (10) EC. Directive 2009/28/EC of the European Parliament and of the Council of 23 April 2009 on the promotion of the use of energy from renewable sources and amending and subsequently repealing Directives 2001/77/EC and 2003/ 30/EC, 2009. (11) Amyris. http://www.amyris.com/ (September 2012). (12) CARB. Staff Report: Detailed California-Modified GREET Pathway for Brazilian Sugar Cane Ethanol; California Air Resources Board: Sacramento, CA, July 20, 2009. (13) Finnveden, G.; Hauschild, M.; Ekvall, T.; Guinee, J.; Heijungs, R.; Hellweg, S.; Koehler, A.; Pennington, D.; Suh, S. Recent developments in Life Cycle Assessment. J. Environ. Manage. 2009, 91 (1), 1−21. (14) The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) Model, v 1.8c.0; Argonne National Laboratory: Argonne, IL, 2009. (15) CARB. Low Carbon Fuel Standard 2011 Program Review Report; California Air Resources Board: Sacramento, CA, September 12, 2011. (16) Tyner, W. E.; Taheripour, F.; Golub, A. A. Calculation of Indirect Land Use Change (ILUC) Values for Low Carbon Fuel Standard (LCFS) Fuel Pathways; CARB and California Environmental Protection Agency: Sacramento, CA, 2011; https://www.gtap.agecon.purdue. edu/resources/download/5629.pdf. (17) Hertel, T. Global Trade Analysis: Modeling and Applications; Cambridge University Press: Cambridge, 1997. (18) Hertel, T.; Golub, A.; Jones, A.; O’Hare, M.; Plevin, R.; Kammen, D. Effects of US maize ethanol on global land use and greenhouse gas emissions: Estimating market-mediated responses. Bioscience 2010, 60 (3), 223−231. (19) Hertel, T.; Tyner, W.; Birur, D. The global impacts of biofuel mandates. Energy J. 2010, 31 (1), 75−100. (20) Plevin, R. J.; Gibbs, H. K.; Duffy, J.; Yeh, S. Agro-ecological Zone Emission Factor Model; University of California: Berkeley, CA; University of Wisconsin: Madison, WI; California Air Resources Board: Sacramento, CA, September 12, 2011. (21) Gardner, T. S.; Hawkins, K. M.; Meadows, A. L.; Tsong, A. E.; Tsegaye, Y. Production of acetyl-coenzyme a derived isoprenoids. U.S. Patent US8415136 B1, 2013; Accessed via Google Patents. (22) Amyris. Total and Amyris Renewable Jet Fuel Ready for Use in Commercial Aviation (press release); 2014. http://www.amyris.com/ News/398/Total-and-Amyris-Renewable-Jet-Fuel-Ready-for-Use-inCommercial-Aviation. (23) Ecoinvent Centre. http://www.ecoinvent.ch. (24) Seabra, J.; Macedo, I. Comparative analysis for power generation and ethanol production from sugarcane residual biomass in Brazil. Energy Policy 2011, 39 (1), 421−428. (25) IPCC. 2006 IPCC Guidelines for National Greenhouse Gas Inventories; IGES: Hayama, 2006. (26) IPCC. Aviation and the Global Atmosphere; Cambridge University Press: Cambridge, England, 1999. (27) Gibbs, H. K.; Yui, S. New Spatially-Explicit Estimates of Soil and Biomass Carbono Stocks by GTAP Region and AEZ; University of Wisconsin: Madison, WI; University of California: Davis, CA, 2011. (28) Martha, G.; Alves, E.; Contini, E. Land-saving approaches and beef production growth in Brazil. Agric. Syst. 2012, 110, 173−177. (29) Harris, N.; Grimland, S.; Brown, S. Land Use Change and Emission Factors: Updates since the RFS Proposed Rule; Report submitted to EPA, available at Docket ID #EPA-HQ-OAR-20050161-3163, www.regulations.gov. 2009 (30) Marelli, L.; Ramos, F.; Hiederer, R.; Koeble, R. Estimate of GHG Emissions from Global Land Use Change Scenarios; JRC, IE, IES: Ispra, Italy, 2011. (31) Lisboa, C.; Butterbach-Bahl, K.; Mauder, M.; Kiese, R. Bioethanol production from sugarcane and emissions of greenhouse gases - Known and unknowns. Global Change Biol. Bioenergy 2011, 3 (4), 277−292.
Figure 6. Monte Carlo results for the net life cycle emissions of the renewable jet fuel.
a better indication of the overall mitigation potential, the market impacts of the substitution of the fossil jet fuel by the renewable jet fuel must be considered, which could reduce the price of the fossil fuel and slightly decrease the demand for the biofuel.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected];
[email protected]; tel.: +55 19 3521 3284. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We thankfully acknowledge the financial support from IDB, Boeing, and Embraer, and the technical support from Amyris. We are also grateful to the anonymous reviewers for their valuable comments on the manuscript.
■
REFERENCES
(1) ElMekawy, A.; Diels, L.; De Wever, H.; Pant, D. Valorization of cereal based biorefinery byproducts: Reality and expectations. Environ. Sci. Technol. 2013, 47 (16), 9014−9027. (2) IPCC. Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change; Cambridge University Press: Cambridge, United Kingdom, 2007. (3) Sgouridis, S.; Bonnefoy, P.; Hansman, R. Air transportation in a carbon constrained world: Long-term dynamics of policies and strategies for mitigating the carbon footprint of commercial aviation. Transp. Res., Part A: Policy and Practice 2011, 45 (10), 1077−1091. (4) Marsh, G. Biofuels: Aviation alternative? Renewable Energy Focus 2008, July/August, 48−51. (5) ICAO. ICAO Environmental Report 2010: Aviation and Climate Change; International Civil Aviation Organization: Montreal, 2010; p 260. (6) Khatiwada, D.; Seabra, J.; Silveira, S.; Walter, A. Accounting greenhouse gas emissions in the lifecycle of Brazilian sugarcane bioethanol: Methodological references in European and American regulations. Energy Policy 2012, 47, 384−397. (7) Walter, A.; Dolzan, P.; Quilodran, O.; de Oliveira, J.; da Silva, C.; Piacente, F.; Segerstedt, A. Sustainability assessment of bio-ethanol production in Brazil considering land use change, GHG emissions and socio-economic aspects. Energy Policy 2011, 39 (10), 5703−5716. (8) CARB. Staff Report: Proposed Regulation to Implement the Low Carbon Fuel Standard; California Air Resources Board: Sacramento, CA, March 5, 2009. 14762
dx.doi.org/10.1021/es503217g | Environ. Sci. Technol. 2014, 48, 14756−14763
Environmental Science & Technology
Article
(32) Galdos, M.; Cerri, C.; Lal, R.; Bernoux, M.; Feigl, B.; Cerri, C. Net greenhouse gas fluxes in Brazilian ethanol production systems. Global Change Biol. Bioenergy 2010, 2 (1), 37−44. (33) Amaral, W. A. N.; Marinho, J. P.; Tarasantchi, R.; Beber, A.; Guiliani, E. Environmental sustainability of sugarcane ethanol in Brazil. In Sugarcane Ethanol: Contributions to Climate Change Mitigation and the Environment; Zuurbier, P., van de Vooren, J., Eds.; Wageningen Academic Publishers: Wageningen, 2008. (34) Novaes, M. R.; Rudorff, B. F. T.; Almeida, C. M.; Aguiar, D. A. Analise espacial da reduçaõ da queima na colheita da cana-de-açuć ar. Eng. Agric.́ 2011, 31 (3), 572−583. (35) Aguiar, D.; Rudorff, B.; Silva, W.; Adami, M.; Mello, M. Remote sensing images in support of environmental protocol: Monitoring the sugarcane harvest in Sao Paulo State, Brazil. Remote Sens. 2011, 3 (12), 2682−2703. (36) Agroenviromental protocol. www.ambiente.sp.gov.br/ etanolverde/protocolo-agroambiental/o-protocolo/ (September 2012). (37) Embrapa. Zoneamento agroecológico da cana-de-açuć ar - ZAE Cana; 2009. (38) Cerri, C.; Galdos, M.; Maia, S.; Bernoux, M.; Feigl, B.; Powlson, D.; Cerri, C. Effect of sugarcane harvesting systems on soil carbon stocks in Brazil: An examination of existing data. Eur. J. Soil Sci. 2011, 62 (1), 23−28. (39) Seabra, J.; Macedo, I.; Chum, H.; Faroni, C.; Sarto, C. Life cycle assessment of Brazilian sugarcane products: GHG emissions and energy use. Biofuels, Bioprod. Biorefin. 2011, 5 (5), 519−532. (40) Macedo, I.; Seabra, J.; Silva, J. Green house gases emissions in the production and use of ethanol from sugarcane in Brazil: The 2005/ 2006 averages and a prediction for 2020. BiomassBioenergy 2008, 32 (7), 582−595. (41) The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) Model, GREET1_2013; Argonne National Laboratory: Argonne, IL, 2013.
14763
dx.doi.org/10.1021/es503217g | Environ. Sci. Technol. 2014, 48, 14756−14763