Spatially-Explicit Life Cycle Assessment of Sun-to-Wheels

Dec 26, 2012 - Department of Civil and Transport Engineering, Norwegian University of Science and Technology, 7491 Trondheim, Norway. Environ. Sci...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/est

Spatially-Explicit Life Cycle Assessment of Sun-to-Wheels Transportation Pathways in the U.S. Roland Geyer,*,† David Stoms,† and James Kallaos‡ †

Bren School of Environmental Science and Management, University of California, Santa Barbara, California 93106-5131, United States ‡ Department of Civil and Transport Engineering, Norwegian University of Science and Technology, 7491 Trondheim, Norway S Supporting Information *

ABSTRACT: Growth in biofuel production, which is meant to reduce greenhouse gas (GHG) emissions and fossil energy demand, is increasingly seen as a threat to food supply and natural habitats. Using photovoltaics (PV) to directly convert solar radiation into electricity for battery electric vehicles (BEVs) is an alternative to photosynthesis, which suffers from a very low energy conversion efficiency. Assessments need to be spatially explicit, since solar insolation and crop yields vary widely between locations. This paper therefore compares direct land use, life cycle GHG emissions and fossil fuel requirements of five different sun-to-wheels conversion pathways for every county in the contiguous U.S.: Ethanol from corn or switchgrass for internal combustion vehicles (ICVs), electricity from corn or switchgrass for BEVs, and PV electricity for BEVs. Even the most land-use efficient biomassbased pathway (i.e., switchgrass bioelectricity in U.S. counties with hypothetical crop yields of over 24 tonnes/ha) requires 29 times more land than the PV-based alternative in the same locations. PV BEV systems also have the lowest life cycle GHG emissions throughout the U.S. and the lowest fossil fuel inputs, except for locations with hypothetical switchgrass yields of 16 or more tonnes/ha. Including indirect land use effects further strengthens the case for PV.

1. INTRODUCTION Transitioning transportation from petroleum to renewable energy can help mitigate climate change, navigate peak oil, and increase energy security.1,2 Transportation currently accounts for 70% of U.S. petroleum consumption and 35% of energyrelated U.S. CO2 emissions.3 Two trends have recently been unfolding. Between 2000 and 2010, the U.S. increased fuel ethanol output from 6 to 50 billion liters per year and overtook Brazil as the world’s largest producer.4 In 2010, fuel ethanol consumed 40% of U.S. corn production.4,5 At the same time there is increasing electrification of vehicle technology and infrastructure, which started in 1996 with General Motors’ EV1 and has recently been reinvigorated with the launch of several plug-in hybrid and battery electric vehicles (PHEV, BEV), as well as growing numbers of vehicle recharging stations.6 There is an emerging scientific consensus that corn ethanol offers at best modest reductions in Greenhouse Gas (GHG) emissions, while the reduction potential of cellulosic ethanol is large.7−10 Some studies pointed out that converting energy crops into electricity for BEVs rather than ethanol for internal combustion vehicles (ICVs) would further reduce GHG emissions and use agricultural land more efficiently,11,12 while the limitations of such attributional comparisons are being stressed by others.13 Biofuels for ICVs and bioelectricity for BEVs use photosynthesis to convert solar radiation into transportation services, © 2012 American Chemical Society

that is, they are sun-to-wheels transportation pathways. While photosynthesis has a theoretical maximum energy conversion efficiency of 33%, the overall conversion efficiency of sunlight into terrestrial biomass is typically below 1%, regardless of crop type and growing conditions.14,15 Therefore any biomass-based energy pathway is very land-use-intensive. As a result, biomassbased transportation pathways are increasingly seen as a threat to food supply and natural habitats.10,16−20 A third type of sun-to-wheels pathway is the use of photovoltaics (PV) to convert sunlight directly into electricity for BEVs. In 2008, the U.S. ranked only fourth in terms of cumulative installed PV capacity, after Germany, Spain, and Japan, despite its outstanding solar resources and the fact that it has many times the land area of the three frontrunners. The cumulative installed PV capacity in 2008 was 5.3 GW in Germany, 3.4 GW in Spain, 2.1 GW in Japan, and only 1.1 GW in the U.S.A.21 In contrast to this, the average annual solar insolation in much of the contiguous U.S. is the same, or better, than that of Spain, the country with the best photovoltaic solar resources in Europe.21 Germany’s solar resource, on the other Received: Revised: Accepted: Published: 1170

July 24, 2012 November 26, 2012 December 26, 2012 December 26, 2012 dx.doi.org/10.1021/es302959h | Environ. Sci. Technol. 2013, 47, 1170−1176

Environmental Science & Technology

Article

yield maps shows that that the conversion efficiency from sunlight to calorific content of the corn ranges between 0.1% and 0.3%. Switchgrass is not currently harvested in sufficient volume to be surveyed by the USDA-NASS so a predictive model was required to estimate the geographic distribution of yield. Researchers at Oak Ridge National Laboratory (ORNL) recently developed an empirical model to predict switchgrass yield.24 Parameters were estimated from yield data from field trials across the country and associated environmental data. The predictive equations were then applied using information on 30 year average temperature and precipitation data generated by the Parameter-elevation Regressions on Independent Slopes Model (PRISM) to produce maps of predicted yield for lowland and upland cultivars at 400 m resolution. Based on the modest response of yield to nitrogen fertilization, the model assumed an application of 100 kg N/ha. This rate of fertilization was therefore used in our study for calculating fossil fuel inputs and GHG emissions. N fertilizer input rates are hypothetical, since there is currently no commercial production of switchgrass. Values used in existing literature vary from 50 to 150 kg/ha. Depending on the chosen input rate and based on a N2O−N release rate of 1.5%, N fertilizer is responsible for 42−71% of total fuel cycle GHG emissions of switchgrass ethanol.7−10 There is some evidence that the N2O−N release rate might be considerably higher than 1.5%.25 Yields in some areas were extrapolated beyond the climatic conditions of the field trials (e.g., colder winters in the Rocky Mountains, drier conditions west of the Rockies). The researchers assumed no artificial irrigation in the predictions, so the yields are quite low in the arid western states. Other scientists have predicted that parts of the west would be suitable if irrigated.26 Predicted yields represent the 90th quantile assuming average weather, best management practices, and optimal fertilization. Thus the ORNL results represent an optimistic forecast for switchgrass yields. ORNL subsequently averaged the grid cell values for the highest yielding cultivar by county in the data provided to us. On the basis of these yield estimates, the conversion efficiency of sunlight to switchgrass biomass varies between 0.1% and 0.7%. 2.2. Fuel Cycle. GHG and fossil fuel data for the production of corn and switchgrass and their conversion to ethanol are based on the EBAMM Model,7 which has been combined with the crop yield maps and updated with data from version 1.8c.0 of the GREET model and other recent literature.8,9 It is assumed that the NCV of corn and switchgrass is 18 MJ per kg, and that 2.53 kg of corn and 2.62 kg of switchgrass are required to produce 1 L of ethanol with 21.2 MJ NCV. Energy consumption and GHG emission values of the biorefineries include coproduction credits and in- and outbound logistics. The crop-to-electricity conversion model assumes that half of the biomass is converted in biomass boilers and the other half is cocombusted with coal to generate electricity.11 Inventory models for both product systems are based on Ecoinvent data and reports.27,28 A biomass-toelectricity conversion efficiency of 32% has been used, and an electricity transmission and distribution efficiency of 92%.2 The PV system life cycle is based on 2005 technology and production data. The life cycle inventory for the PV system is based on Ecoinvent and Fthenakis et al.29,30 The CdTe module is assumed to have a 9% conversion efficiency and an initial

hand, has about the same range as Alaska’s. There is a vast untapped photovoltaic potential in the U.S., some of which could be used to generate sun-to-wheels transportation services. Existing environmental assessments of biofuels and photovoltaic energy pathways use average biomass and PV yields,7−12,16,19,20,22 even though these yields vary widely between geographical locations. Spatially explicit assessments are more informative, since pathway performance depends on location, and land use decisions are always local by nature. This article presents life cycle assessments of five different sun-towheels conversion pathways for every county in the contiguous U.S: Ethanol from corn or switchgrass for ICVs, bioelectricity from corn or switchgrass for BEVs, and PV electricity for BEVs using cadmium telluride (CdTe) solar cells. The assessments include the production and use of the transportation energy (the fuel cycle) and the life cycle of the vehicle.

2. METHODS AND DATA The functional unit of the assessment is defined as 100 km driven in a compact passenger vehicle during one year. Three environmental indicators are calculated for each county of the contiguous U.S.: The first is the land area required for the corn and switchgrass fields or the PV installation, which is called direct land use and measured in m2/100 km driven. The second indicator is the total global warming potential from the vehicle and fuel life cycles, measured in kg CO2 equiv/100 km driven. The third indicator is the total fossil fuel consumption from the vehicle and fuel life cycles, measured in MJ of net calorific value (NCV) per 100 km driven. The system boundary includes vehicle production, use, and end-of-life management, as well as fuel production and use. In the case of PV electricity, the fuel cycle consists of production, use, and end-of-life management of the PV system. Many scientists believe that the conversion of existing agricultural land from food and feed to fuel crops will displace the original use to other locations, causing indirect land use change (iLUC). Size and effects of iLUC are highly uncertain and have therefore only been calculated as a range for four exemplary counties. 2.1. Spatially Explicit Data. Three spatial data sets were used to account for the spatial variability of biomass and PV yields. Color-coded maps of the three can be found in the Supporting Information (SI). A GIS layer of solar insolation was obtained from the National Renewable Energy Laboratory (NREL). A satellite radiation model used hourly radiance images from geostationary weather satellites, daily snow cover data, and monthly averages of atmospheric water vapor, trace gases, and the amount of aerosols in the atmosphere to calculate the hourly total insolation (sun and sky) falling on a horizontal surface.23 Horizontal insolation was then converted for collectors tilted at the angle of latitude to reflect standard practice. The NREL model calculated insolation at 10 km resolution. For this study, grid cell values were averaged within counties. The United States Department of Agriculture National Agricultural Statistics Service (USDA-NASS) compiles data on acreage and yields of major crop types from sampling surveys of farmers and reports the estimates annually at the county level.5 Data for corn yield (reported in bushels per acre) were downloaded for the years 2005−2009. To smooth out interannual variability caused either by weather or reporting problems, the five years of yield data were averaged and converted to kg/ha. Years with zero yield were ignored in computing averages. Combining the solar insolation and corn 1171

dx.doi.org/10.1021/es302959h | Environ. Sci. Technol. 2013, 47, 1170−1176

Environmental Science & Technology

Article

Table 1. Characteristics and Environmental Indicators for Four Illustrative U.S. Counties Corn Belt: Delaware, IA area (in ha) 150 000 insolation (in kWh/m2day) 4.46 corn yield (in kg/ha) 10 965 switchgrass yield (in kg/ha) 13 460 land use (in m2/100 km driven) corn EtOH ICV [iLUC range] 27.70 + [6.93−22.16] switchgrass EtOH ICV [iLUC range] 23.41 + [0−6.79] corn electricity BEV [iLUC range] 11.15 + [2.79−8.92] switchgrass electricity BEV [iLUC range] 9.08 + [0−2.63] CdTe PV BEV 0.18 GHG emissions (in kg CO2 equiv/100 km driven) corn EtOH ICV [iLUC GHG range] 21.4 + [6.9−22.2] switchgrass EtOH ICV [iLUC GHG range] 7.7 + [0−6.5] corn electricity BEV [iLUC GHG range] 8.6 + [2.8−8.9] switchgrass electricity BEV [iLUC GHG range] 6.9 + [0−2.6] CdTe PV BEV 5.9 fossil fuel consumption (in MJ/100 km driven) Corn EtOH ICV 223 Switchgrass EtOH ICV 73 Corn electricity BEV 89 Switchgrass electricity BEV 79 CdTe PV BEV 76

Switchgrass Belt: Monroe, TN

Solar Belt: Maricopa, AZ

Northern U.S.A.: Rolette, ND

169 000 4.69 7550 24 557

2 389 100 6.50 11 643 3188

243 300 4.42 4406 3059

40.23 + [10.06−32.19] 12.83 + [0−3.72] 16.20 + [4.05−12.96] 4.98 + [0−1.44] 0.17

26.09 98.83 10.50 38.36 0.12

68.94 + [17.24−55.15] 103.00 + [0−29.87] 27.75 + [6.94−22.20] 39.97 + [0−11.59] 0.18

24.8 + [10.1−32.2] 6.1 + [0−3.6] 9.9 + [4.0−13.0] 6.3 + [0−1.4] 5.9

20.9 + [6.5−20.9] 19.2 + [0−27.6] 8.4 + [2.6−8.4] 11.4 + [0−11.1] 5.7

32.5 19.8 13.1 11.6 5.9

244 63 98 75 76

220 143 88 106 73

292 147 117 107 76

+ + + +

[6.52−20.87] [0−28.66] [2.63−8.40] [0−11.12]

+ + + +

[17.2−55.2] [0−28.8] [6.9−22.2] [0−11.6]

depend on location and are therefore constant for all counties. The complete set of data used for the vehicle cycle models can be found in the SI. 2.4. Limitations. Fuel crop production on fertile lands seems likely to induce indirect land use change, which is estimated to lead to significant amounts of GHG emissions.35 To illustrate its potential significance, we calculated plausible iLUC impact ranges for four exemplary counties, following the methodology of Plevin et al.36 Values for iLUC effects could simply be added to the results for all counties. However, to emphasize the significance of the impacts from direct land use alone, effects from iLUC have been omitted in all result maps, which thus show conservative land use and GHG emission values for fuel crop production. In the direct land use comparisons we only account for the land area required to grow the energy crop or accommodate the PV system. This seems to be a reasonable cutoff, since a recent study concluded that land needed at other life cycle stages, for example, for ethanol and bioelectricity plants or PV manufacturing, is small.17

performance ratio (PR) of 0.8. It is further assumed that every year of its 30 lifetime the PR decreases by 1%, which results in a lifetime average PR of 0.684.29,30 The balance of system (BOS) is assumed to be similar to that of a 3 kW peak slanted-roof installation, which may generate overestimates for larger-sized field plants since these are likely to require fewer BOS components per panel.29 The complete set of data used for the fuel cycle models can be found in the SI. 2.3. Vehicle Cycle. GHG emissions and fossil fuel consumption throughout the vehicle life cycles are mostly based on the models from Samaras and Meisterling.31 This means that economic input−output life cycle assessment (EIOLCA) is used to derive energy and GHG values for the production of a compact ICV and data on 2005 lithium ion (Liion) battery technology is added to model PHEVs of equivalent size. The resulting energy and GHG values are 102,000 MJ and 8,500 kg CO2eq per compact ICV, and 1700 MJ and 120 kg CO2 equiv/kWh of Li-ion battery. This data is in good agreement with other vehicle LCAs, such as GREET 2.7 and Notter et al.32,33 These comparisons also show that EIOLCA and process-based models of compact vehicles yield comparable results. A 150 km-range BEV model is derived from Samaras and Meisterling by increasing the battery size in their PHEV model. This may overestimate GHG emissions and fossil fuel consumption of BEV production since we merely added the battery to an ICV and did not deduct the internal combustion engine or related components. The compact ICV requires 255 MJ NCV of liquid fuel per 100 km, for example, eight liters of gasoline or 10.6 L of pure ethanol, which means that gasoline and ethanol have the same calorific fuel economy. An equivalent BEV is assumed to require 64.8 MJ or 18 kWh of electric charge per 100 km.11,31,34 Together with the maximum range of 150 km and a maximum depth of discharge (DOD) of 0.8, the BEV energy demand translates into a required battery size of 33.75 kWh. The life cycle mileage of both vehicles is assumed to be 240 000 km. Results for the vehicle cycle do not

3. RESULTS AND DISCUSSION For all counties of the contiguous U.S., direct land use, life cycle GHG emissions and life cycle fossil fuel consumption of the five sun-to-wheels pathways are calculated. Color-coded maps of all results are available in the SI. In the following sections the most important results are presented and discussed. 3.1. Direct Land Use. Direct land use, expressed in m2/100 km driven, is not only a measure of the land use efficiency of the sun-to-wheels systems but also of the energy conversion efficiency from solar resource to final energy service. Bioethanol and bioelectricity transportation pathways both start with converting sunlight into biomass, in this study corn or switchgrass. As mentioned earlier, the vast majority of the solar radiation is not converted into biomass energy, which makes this step the bottleneck of the conversion pathway. The 1172

dx.doi.org/10.1021/es302959h | Environ. Sci. Technol. 2013, 47, 1170−1176

Environmental Science & Technology

Article

Figure 1. Ratio of direct land use for switchgrass bioelectricity (BE) + BEV over PV + BEV.

biomass is then either converted into ethanol, which contains 45−47% of its calorific energy or into electricity containing around 30% of the biomass energy. Measured in MJ/km, the battery-to-wheel conversion efficiency of a BEV is about 4 times the tank-to-wheel conversion efficiency of an equivalent ICV.11,31,34 The overall result is that, independent of location or crop type, the bioelectricity BEV system requires around 1− 0.45/(0.3 × 4) = 60% less crop land than the bioethanol ICV system (see Table 1). Nevertheless, both sun-to-wheels systems suffer from the poor sun-to-biomass conversion efficiency. One way to address this conversion bottleneck created by photosynthesis is to use photovoltaics instead. While locations with high corn yields exist throughout the entire U.S., the Corn Belt (represented by Delaware County, IA) has the highest sun-to-biomass conversion efficiencies for corn. For switchgrass, the highest energy conversion and yield estimates are predicted for Tennessee, the border region between Arkansas and Missouri, and neighboring regions, which we will call the Switchgrass Belt (represented by Monroe County, TN). Solar radiation is highest in Arizona, New Mexico, the southern parts of California and Nevada, and some neighboring regions, which we will call the Solar Belt (represented by Maricopa County, AZ). Rolette County, ND, is representative of regions with low solar radiation and low crop yields. Anywhere in the contiguous U.S. the land required for the PV BEV system is orders of magnitudes less than for any of the biomass pathways. The most land-efficient biomass pathway is switchgrass-based electricity for BEVs in the Switchgrass Belt, where the yield map shows potential harvests of 24−25 t of switchgrass per year. Whether such yields are commercially achievable is not clear, yet even in these areas the

PV BEV pathway improves the land use efficiency of the bioelectricity pathway by at least a factor of 29, for example, from around 5 for switchgrass to 0.17 m2/100km for PV in Monroe County, TN (see Table 1 and Figure 1). The factor jumps to 75 or more when PV BEV is compared with a switchgrass ethanol pathway in the Switchgrass Belt. Predictably, the corn-based pathway with the smallest land use is cornbased electricity for BEVs in the Corn Belt. Here, the PV BEV pathway improves land use efficiency by at least a factor of 51, e.g. from around 9 to 0.18 m2/100km in Delaware County, IA (see Table 1 and Figure 1). If compared to corn ethanol in the Corn Belt, the factor increases to 154 or more. The land use improvements are even more dramatic in the northern USA or in the Solar Belt (see Table 1 and Figure 1). In other words, no matter where in the contiguous U.S., a sun-to-wheels pathway based on CdTe PV can generate the same amount of passenger transport as any biomass-based pathway with 3% of the land or less. The PV land use numbers are conservative, since many PV cells have conversion efficiencies above 9%, and performance ratio (PR) degradation can be reduced through PV system maintenance. Across the contiguous U.S., land use figures for biomass pathways range from 5 to 100 m2/100 km and beyond. Land requirements for PV BEV systems, on the other hand, only vary from 0.12 to 0.18 m2/100 km. This very attractive feature of PV-based sun-to-wheels pathways is mainly because photovoltaic energy conversion mostly depends on groundlevel solar insolation, with other climate variables such as rainfall or temperature having only minor impact. PV also does not compete with food supply and natural habitat as it does not require fertile agricultural soil but is equally suitable for 1173

dx.doi.org/10.1021/es302959h | Environ. Sci. Technol. 2013, 47, 1170−1176

Environmental Science & Technology

Article

Figure 2. Ratio of life cycle GHG emissions (excluding indirect land use change) for switchgrass bioelectricity (BE) + BEV over PV + BEV.

emissions of both switchgrass-based systems are only slightly higher than the PV BEV pathway. In the PV BEV system, the PV life cycle makes up only 8− 12% of total GHG emissions, since emissions from BEV production alone are 5.2 kg CO2 equiv/100 km. Compared to gasoline ICVs, the relative contributions of fuel and vehicle cycle are reversed. The carbon intensity of PV electricity is already so low that further GHG reductions are more likely to be found in vehicle production.37 Fossil fuel inputs do not entirely mirror the GHG results because of noncarbon GHG emissions and varying carbon intensities of fossil fuels. For the data in Table 1, the correlation coefficient between the two indicators is 0.97. The most significant difference is the fact that no single sun-to-wheels pathway has the lowest life cycle fossil fuel requirements everywhere. In much of the Southeast and the southern part of the Midwest, where possible switchgrass yields of 16 tons/ha or more are predicted, fossil fuel consumption of the switchgrass ethanol ICV pathway is up to 16% lower than PV BEV and switchgrass electricity BEV pathways. In the same regions, the fossil fuel requirements for switchgrass electricity BEV and PV BEV systems are virtually equal. In areas where switchgrass yield estimates are below 16 tons/ha, mostly the northern and western parts of the U.S., the PV BEV baseline consumes the least fossil fuel per kilometer driven. Over 80% of the difference between GHG emission and fossil fuel consumption results is due to N2O emissions from nitrogen fertilizer use. The remaining discrepancy is caused by the fact that our BEV production data has slightly lower GHG emissions per MJ fossil energy input than the ICV production data. 3.3. Discussion. To estimate the implications of our results for the U.S. as a whole we derive average values for insolation

marginal, barren, or developed land areas, including road sides, parking lots, and rooftops. 3.2. Life Cycle GHG Emissions and Fossil Fuel Requirements. Table 1 also shows the life cycle GHG emissions and fossil fuel requirements of the five sun-to-wheels pathways for the four representative counties mentioned earlier. A useful benchmark for these results is an equivalent compact ICV driven with gasoline, which has life cycle GHG emissions of 27.5 kg CO2 equiv/100 km and life cycle fossil energy consumption of 343 MJ/100 km. For both measures, slightly less than 15%, 3.5 kg CO2 equiv/100 km and 43 MJ/100 km, come from vehicle production. With the exception of corn ethanol, all sun-to-wheels systems have at least some geographical regions where life cycle GHG emissions without considering iLUC are 70−80% lower than the gasoline ICV baseline. The lowest life cycle GHG emissions of the cornbased bioelectricity pathway range from 8 and 10 kg CO2 equiv/100 km; mostly in the Corn Belt, but also in various other regions. iLUC may increase GHG emissions for corn by 30−170% (see Table 1). The lowest life cycle GHG emissions of the switchgrass-based pathways are in the eastern and midwestern U.S., where they range from 6 to 8 kg CO2 equiv/ 100 km. In these high yield locations, switchgrass-based ethanol and bioelectricity pathways have roughly the same GHG emissions. iLUC could potentially more than double the emissions for switchgrass (see Table 1). Throughout the contiguous U.S., the PV BEV system has the lowest life cycle GHG emissions of all five studied sun-towheels pathways. Virtually everywhere it is 6 kg CO2 equiv/100 km or less, and in many areas it offers a substantial reduction from the second best pathway (see Table 1 and Figure 2). In the high yield locations of the Switchgrass Belt, life cycle GHG 1174

dx.doi.org/10.1021/es302959h | Environ. Sci. Technol. 2013, 47, 1170−1176

Environmental Science & Technology

Article

Figure 3. Direct land use, life cycle GHG emissions (excluding indirect land use change), and life cycle fossil fuel requirements to generate the transportation services provided by 17.8 × 1012 MJ NCV of gasoline, the amount used in transportation in the U.S. in 2009.

and ICV production yields more realistic BEV-based estimates than the ones given in Jacobson.22 Vehicles powered with switchgrass electricity or ethanol come second and third with 0.46 and 0.48 billion tons of CO2 equiv, yet these numbers do not include any GHG emissions from indirect land use change. The three sun-to-wheels pathways with the lowest fossil fuel requirements are switchgrass ethanol for ICVs, switchgrass electricity for BEVs, and PV electricity for BEVs, with 4.7, 5.4, and 5.2 trillion MJ. For both BEV-based pathways, over 85% of fossil fuels are consumed during vehicle production. Of all studied sun-to-wheels systems, corn ethanol for ICVs has by far the highest land requirements, GHG emissions, and fossil fuel requirements. While other environmental concerns, such as water use, ecosystem services, and human and eco-toxicity should also be assessed carefully, land use efficiency, climate change impacts, and fossil fuel consumption are of paramount importance. Thanks to the federal tax credit, PV-powered BEVs are already cost-competitive, given that the batteries do not have to be replaced. At $4/gallon gasoline and $0.20/KWh PV electricity, the nondiscounted sum of vehicle plus fuel cost is around $36 000 for both a typical compact ICV and a Nissan-Leaf-type BEV. Assuming $3.75/gallon E85, a typical flex fuel compact ICV comes in at $43 000 (see SI for details). Assuming that the economics of PV and BEV technology will further improve and issues of material availability, and electricity transmission and storage can be resolved, PV offers land-efficient and low-carbon sun-to-wheels transportation. Unlike fuel crops, PV electricity does not have to compete with food production and biodiversity for fertile land and could potentially replace all gasoline used in U.S. transportation.

and biomass yields. A simple average of the equal-area grid cell values from the insolation map yields 5.055 kWh/day·m2. According to the USDA, the national average corn yield for the period 2005−2009 was 9616 kg/ha·year. For switchgrass, an area-weighted average of all potential county-level yields, ignoring counties with yields under 8 tons/ha, results in 17 757 kg/ha·year. Around 1.1 million ha of CdTe PV systems with a relatively low panel conversion efficiency of 9% would generate enough electricity to replace all gasoline consumed in U.S. transportation in 2009 (see Figure 3).38 This is 45% of the area of Maricopa County, AZ, or 37% of the area of all four counties from Table 1. The four biomass-based pathways would require 48−220 million ha or 40−200 times more land than the studied PV BEV system. This is a very important observation as signs increase that energy crops compete with food and feed for fertile lands. For reference, in the 2007 Census of Agriculture the U.S. had 165 million ha of cropland. Converting the entire U.S. corn crop of 2009 into ethanol would have displaced a mere 15.7% of 2009 U.S. gasoline consumption.4,5 Our averaged U.S.-wide land use requirements are in general agreement with results from earlier, more generic and nonspatial studies.16,22 However, direct land use requirements show large variations between different biomass-based pathways and different U.S. counties, ranging from a minimum of 5 m2/100 km to 100 m2/100 km and beyond. This indicates that many areas in the contiguous U.S. may not be particularly suited for biobased sun-to-wheels pathways, at least in terms of land use efficiency. The land use requirements for the PV-based sun-to-wheels system, on the other hand, are less than 0.2 m2/ 100km virtually everywhere, which makes this a land-useefficient option throughout the United States and not just in the Solar Belt, that is, the Southwest. Figure 3 shows that, relative to the gasoline baseline, the PV and switchgrass scenarios would also reduce associated GHG emissions and fossil fuel consumption from production and use of vehicles and fuels by 75−80% relative to gasoline ICVs. The PV-based pathway would reduce life cycle GHG emissions, including vehicle production, by almost 80%, from 1.92 to 0.41 billion tons of CO2 equiv. Accounting for the GHGs of BEV



ASSOCIATED CONTENT

S Supporting Information *

Descriptions and color-coded maps of the spatially explicit data, data for and models of the fuel and vehicle cycles, as well as color-coded maps of all indicator results. This material is available free of charge via the Internet at http://pubs.acs.org. 1175

dx.doi.org/10.1021/es302959h | Environ. Sci. Technol. 2013, 47, 1170−1176

Environmental Science & Technology



Article

(18) Delucchi, M. A. Impacts of biofuels on climate change, water use, and land use. Ann. N.Y. Acad. Sci. 2010, 1195, 28−45. Fenner, R. Comparative analysis of environmental impacts of maize-biogas and photovoltaics on a land use basis. Sol. Energy 2010, 84, 1255−1263. (19) Graebig, M.; Bringezu, S.; Fenner, R. Comparative analysis of environmental impacts of maize-biogas and photovoltaics on a land use basis. Sol. Energy 2010, 84, 1255−1263. (20) Campbell, J. E.; Block, E. Land-use and alternative bioenergy pathways for waste biomass. Environ. Sci. Technol. 2010, 44, 8665− 8669. (21) NREL. Figure 1.1 & Figure 3.2. In 2008 Solar Technologies Market Report; DOE/GO-102010-2867; National Renewable Energy Laboratory: Golden, CO, 2010. (22) Jacobson, M. Z. Review of solutions to global warming, air pollution, and energy security. Energy Environ. Sci. 2009, 2, 148−173. (23) Perez, R.; Ineichen, P.; Moore, K.; Kmiecik, M.; Chain, C.; George, R.; Vignola, F. A new operational model for satellite-derived irradiances: Description and validation. Sol. Energy 2002, 73, 307−317. (24) Wullschleger, S. D.; Davis, E. B.; Borsuk, M. E.; Gunderson, C. A.; Lynd, L. R. Biomass production in switchgrass across the United States: Database description and determinants of yield. Agron. J. 2010, 102, 1158−1168. (25) Crutzen, P. J.; Mosier, A. R.; Smith, K. A.; Winiwarter, W. N2O release from agro-biofuel production negates global warming reduction by replacing fossil fuels. Atmos. Chem. Phys. 2008, 8, 389−395. (26) Barney, J. N.; DiTomaso, J. M. Bioclimatic predictions of habitat suitability for the biofuel switchgrass in North America under current and future climate scenarios. Biomass Bioenergy 2010, 34, 124−133. (27) Dones, R.; Bauer, C.; Röder, A.; Kohle, Report No. 6-VI; Ecoinvent: Dübendorf, Switzerland, 2007. (28) Heck, T.; Wärme-Kraft-Kopplung, Report No. 6-XIV; Ecoinvent: Dübendorf, Switzerland, 2007. (29) Jungbluth, N.; Tuchschmid, M.; Photovoltaics, Report No. 6-XII; Ecoinvent: Dübendorf, Switzerland, 2007. (30) Fthenakis, V. M.; Kim, H. C.; Alsema, E. Emissions from photovoltaic life cycles. Environ. Sci. Technol. 2008, 42, 2168−2174. (31) Samaras, C.; Meisterling, K. Life cycle assessment of greenhouse gas emissions from plug-in hybrid vehicles: Implications for policy. Environ. Sci. Technol. 2008, 42, 3170−3176. (32) Burnham, A.; Wang, M.; Wu, Y. Development and Applications of GREET 2.7The Transportation Vehicle-Cycle Model, ANL/ESD/06-5; Argonne National Laboratory: Lemont, IL, 2006. (33) Notter, D. A.; Gauch, M.; Widmer, R.; Wäger, P.; Stamp, A.; Zah, R.; Althaus, H.-J. Contribution of Li-ion batteries to the environmental impact of electric vehicles. Environ. Sci. Technol. 2010, 44, 6550−6556. (34) EPRI. Environmental Assessment of Plug-In Hybrid Electric Vehicles. Vol. 1: Nationwide Greenhouse Gas Emissions, No. 1015325; Electric Power, Research Institute: Palo Alto, CA, 2007. (35) Searchinger, T.; Heimlick, R.; Houghton, R. A.; Dong, F.; Elobeid, A.; Fabiosa, J.; Tokgoz, S.; Hayes, D.; Yu, T.-H. Use of U.S. croplands for biofuels increases greenhouse gases through emissions from land-use change. Science 2008, 319, 1238−1240. (36) Plevin, R. J.; O’Hare, M.; Jones, A. D.; Torn, M. S.; Gibbs, H. K. Greenhouse gas emissions from biofuels’ indirect land use change are uncertain but may be much greater than previously estimated. Environ. Sci. Technol. 2010, 44, 8015−8021. (37) Geyer, R. Parametric assessment of climate change impacts of automotive material substitution. Environ. Sci. Technol. 2008, 42, 6973−6979. (38) EIA. Table 3.8c. In February 2011 Monthly Energy Review, DOE/ EIA-0035(2011/02); Energy Information Administration: Washington, DC, 2011.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 1 805 893 7234. Fax: 1 805 893 6113. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank S. Gaines, D. Tilman, and the two reviewers for their valuable comments. This research was made possible through support from the National Science Foundation (CBET0932369).



REFERENCES

(1) Goldemberg, J. Ethanol for a sustainable energy future. Science 2007, 315, 808−810. (2) MacKay, D. J. C. Sustainable EnergyWithout the Hot Air; UIT: Cambridge, U.K., 2009. (3) EIA. Figure 2.0 & Tables 11.1 & 11.3d. In Annual Energy Review 2010; DOE/EIA-0384(2010); Energy Information Administration: Washington, DC, 2011. (4) EIA. Table 10.3. In Annual Energy Review 2010; DOE/EIA0384(2010); Energy Information Administration: Washington, DC, 2011. (5) USDA. National Agricultural Statistics Service; U. S. Department of Agriculture: Washington, DC, 2012; www.nass.usda.gov/Data_and_ Statistics. (6) Electrification Coalition. Electrification Roadmap; Electrification Coalition: Washington, DC, 2009. (7) Farrell, A. E.; Plevin, R. J.; Turner, B. T.; Jones, A. D.; O'Hare, M.; Kammen, D. M. Ethanol can contribute to energy and environmental goals. Science 2006, 311, 506−508. (8) Bai, Y.; Luo, L.; van der Voet, E. Life cycle assessment of switchgrass-derived ethanol as transport fuel. Int. J. Life Cycle Assess. 2010, 15, 468−477. (9) ANL. GREET 1.8c.0 Fuel-Cycle Model; Argonne National Laboratory: Lemont, IL, March 2009; www.transportation.anl.gov/ modeling_simulation/GREET/index.html. (10) Hill, J.; Nelson, E.; Tilman, D.; Polasky, S.; Tiffany, D. Environmental, economic, and energetic costs and benefits of biodiesel and ethanol biofuels. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 11206− 11210. (11) Campbell, J. E.; Lobell, D. B.; Field, C. B. Greater transportation energy and GHG offsets from bioelectricity than ethanol. Science 2009, 324, 1055−1057. (12) Hedegaard, K.; Thyø, K. A.; Wenzel, H. Life cycle assessment of an advanced bioethanol technology in the perspective of constrained biomass availability. Environ. Sci. Technol. 2008, 42, 7992−7999. (13) Lemoine, D. M.; Plevin, R. J.; Cohn, A. S.; Jones, A. D.; Brandt, A. R.; Vergara, S. E.; Kammen, D. M. The climate impacts of bioenergy systems depend on market and regulatory contexts. Environ. Sci. Technol. 2010, 44, 7347−7350. (14) Walker, D. A. Biofuels, facts, fantasy, and feasibility. J. Appl. Phycol. 2009, 21, 509−517. (15) Blankenship, R. E.; Tiede, D. M.; Barber, J.; Brudvig, G. W.; Fleming, G.; Ghirardi, M.; Gunner, M. R.; Junge, W.; Kramer, D. M.; Melis, A.; Moore, T. A.; Moser, C. C.; Nocera, D. G.; Nozik, A. J.; Ort, D. R.; Parson, W. W.; Prince, R. C. Comparing photosynthetic and photovoltaic efficiencies and recognizing the potential for improvement. Science 2011, 332, 805−809. (16) McDonald, R. I.; Fargione, J.; Kiesecker, J.; Miller, W. M.; Powell, J. Energy sprawl or energy efficiency: Climate policy impacts on natural habitat for the United States of America. PLoS One 2009, 4, e6802. (17) Fthenakis, V.; Kim, H. C. Land use and electricity generation: A life-cycle analysis. Renewable Sustainable Energy Rev. 2009, 13, 1465− 1474. 1176

dx.doi.org/10.1021/es302959h | Environ. Sci. Technol. 2013, 47, 1170−1176