fuel cycle assessment of ethanol produced at cilion’s

17
1 Fuel Cycle Assessment of Ethanol Produced at Cilion’s Keyes Ethanol Plant Prepared by Richard Plevin Energy and Resources Group UC Berkeley September 6, 2007

Upload: others

Post on 11-Apr-2022

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

1

Fuel Cycle Assessment of Ethanol Produced at Cilion’s Keyes Ethanol Plant

Prepared by Richard Plevin

Energy and Resources Group UC Berkeley

September 6, 2007

Page 2: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

2

Contents

1. Introduction and Summary of Results..............................................................................3 2. Well-to-Tank Analysis ......................................................................................................4

2.1. Approach..................................................................................................................4 2.2. GREET Customizations..........................................................................................5 2.3. Using the Spreadsheet .............................................................................................5 2.4. Energy Inputs for Cilion .........................................................................................6 2.5. Corn farming in GREET.........................................................................................8 2.6. Shipping of Distillers Grains ..................................................................................9 2.7. Electricity...............................................................................................................10 2.8. Ethanol Denaturant................................................................................................11 2.9. California Reformulated Gasoline .......................................................................12

3. Results ..............................................................................................................................12 3.1. Greenhouse Gas Emissions...................................................................................12 3.2. Criteria Pollutant Emissions .................................................................................14 3.3. Net Energy Balance...............................................................................................15 3.4. Petroleum Displacement .......................................................................................16

4. Conclusion .......................................................................................................................16 5. Acknowledgements .........................................................................................................17 6. Works Cited .....................................................................................................................17

Tables and Figures

Table 1. Relative Global Warming Impact and Fossil Energy Value for Cilion and Baseline Ethanol......................................................................................................3

Table 2. Performance data for Keyes plant provided by Cilion............................................7 Table 3. Values customized in GREET ..................................................................................7 Table 4. Mass of dry and wet distillers grains per bushel corn and per gallon ethanol .....10 Table 5. Corn transport energy in GREET ...........................................................................10 Table 6. Shipping assumptions for WDG and DDGS .........................................................10 Table 7. Electricity generation resource mix in the MROW eGRID region. .....................10 Table 8. Comparison of life cycle GHG emissions for CAMX (California), MROW, and

US average. (g CO2eq/kWh) ................................................................................11 Table 9. Life cycle results for Cilion Keyes ethanol compared to standard GREET

ethanol and CARFG..............................................................................................13 Table 10. Life cycle GHG emissions for Cilion ethanol compared to average Midwest

ethanol, and to California reformulated gasoline (CARFG). (g CO2eq/MJ).....14 Table 11. Life Cycle Emissions of Criteria Pollutants for Cilion Ethanol compared to

baseline Midwest natural-gas, dry-mill corn ethanol and CARFG. (g/GJ) .......14 Table 12. Net Energy Balance of Cilion Ethanol .................................................................16

Figure 1. Total Life Cycle Emissions of Criteria Pollutants for Cilion Ethanol, Baseline Ethanol, and CA reformulated gasoline.................................................................4

Figure 2. Excel’s macro selection panel .................................................................................6 Figure 3. eGRID regions defined by the US EPA................................................................11

Page 3: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

3

1. Introduction and Summary of Results

This study examines the life cycle emissions of greenhouse gases and criteria pollutants and the net energy balance from the production, distribution, and combustion of ethanol produced in Cilion’s ethanol plant in Keyes, California. The Cilion process involves the production of ethanol from corn using a dry mill and distributing byproduct distillers grains to local cattle operations. The distillers grains are shipped without drying, which represents an energy savings for the plant. The analysis was performed using the GREET model1 from Argonne National Laboratory with customizations based on performance data provided by Cilion. For comparison, we show the results for Cilion’s ethanol against California reformulated gasoline (CARGF) and natural-gas fired, dry-mill corn ethanol produced in the Midwest, as estimated by an unmodified version of GREET. In summary:

1. Cilion ethanol offers significantly reduced life cycle global warming emissions through cogeneration of electricity and avoidance of drying distillers grains. Table 1 shows that Cilion ethanol offers an addition 29% reduction in global warming intensity (GWI) from average corn ethanol produced in a natural gas fired dry mill, and a 46% reduction in GWI versus California Reformulated Gasoline (CARFG) without oxygenate.

2. For the same reasons, Cilion ethanol offers lower total emissions of all criteria pollutants tracked in GREET versus the same Midwest ethanol (Figure 1). However, the production of either ethanol results in more emissions of all tracked criteria pollutants than does the production of CARFG. This study did not examine the differential emissions in the combustion phase.

Table 1. Relative Global Warming Impact and Fossil Energy Value for Cilion and Baseline Ethanol

Cilion (Avg US Corn) NG-fired Dry mill GWI reduction from MW EtOH 29% 0% GWI reduction from CARFG 46% 25% Fossil Energy Value* 15.5 9.6 Fossil Energy Ratio* 2.6 1.6 *Fossil Energy Value and Fossil Energy Ratio are defined in section 3.3,

1 The June 21, 2007 release of GREET 1.7 was used for this analysis. A copy of the spreadsheet can also be

downloaded directly from http://plevin.berkeley.edu/tools/greet/.

Page 4: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

4

Figure 1. Total Life Cycle Emissions of Criteria Pollutants for Cilion Ethanol, Baseline Ethanol, and CA reformulated gasoline.

2. Well-to-Tank Analysis

2.1. Approach

Energy inputs and GHG emissions were calculated to compare Cilion ethanol against other ethanol production options and conventional RFG. These energy impacts are typically determined on a well-to-tank (WTT) basis using fuel cycle energy models such as GREET. These models track the various feedstock production, transport, refining, and fuel marketing steps and present the results on a WTT basis. The analysis shown here was performed using the GREET model with assumptions that reflect Cilion’s process conditions. The model developed for this analysis calculates several individual components of the corn based ethanol production and combines these to produce the WTT result. The components for ethanol production include:

• Corn production including credit for distillers grains • Ethanol production including plant energy consumption, imported power, and ethanol

distribution • Denaturant

Page 5: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

5

2.2. GREET Customizations

The standard GREET model is parameterized to model average fuel pathways rather than specific facilities. The model inputs represent an average performance for corn and ethanol production and average transport modes and distances for shipping inputs and final products. The model can, however, be parameterized to estimate energy use and emissions for a single facility. Some inputs (e.g. total energy use in an ethanol plant, percentage of total energy from electricity) must be calculated outside of GREET to provide data in the format the model is designed to handle. Other inputs must merely be changed to better reflect specific performance and plant location. In some cases, values must be calculated outside of GREET since it represents only one pathway for each fuel type (e.g. electricity) when more than one pathway is employed in a single production process. For example, Midwest electricity is used for corn farming, but California electricity is used at the ethanol plant. Rather than directly modify cells in the GREET spreadsheet, we developed a Visual Basic macro that reads a specification (in another Excel spreadsheet) of which cells to modify, and updates the specified GREET workbook accordingly. This approach has several advantages over directly modifying GREET:

1. The changes to GREET are explicit and transparent, including documentation and supporting calculations. The modifications are easily alterable and reproducible.

2. The changes can be applied to a “virgin” copy of GREET, ensuring that the only changes are those defined in the specification spreadsheet. The GREET workbook can be “read-only” so that inadvertent or undocumented changes cannot corrupt the results.

3. The modifications can more easily be applied to new releases of GREET. If the location of target cells changes, the specification spreadsheet is easily updated, although in practice, most cell locations are unchanged from release to release.

2.3. Using the Spreadsheet

The analysis workbook Cilion_Fuelcycle_v13.xls (delivered with this report) contains all the data, computations, and macros used to produce this report. These instructions are provided to allow Cilion to re-run the analysis using modified assumptions.

NOTE: A copy of GREET must be open when using the analysis spreadsheet. The analysis workbook identifies the current version of GREET at the top of the “Settings” worksheet. The analysis has been run with the June 21, 2007 release of GREET 1.7, which can be downloaded from http://Plevin.berkeley.edu/tools/greet/, or downloaded and extracted from a release of GREET from Argonne National Laboratory as described on the above web page. The GREET workbook must be identified on the Settings sheet of the analysis workbook. (The file is currently named greet1.7_20070621.xls.)

To customize GREET as specified on the settings page, run the macro PokeCells:

1. From the Tools menu, select Macros->Macros…

2. Choose Sheet1.PokeCells from the list of macros. (See Figure 2.)

Page 6: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

6

3. Click the Run button.

The macros applies all the listed changes to GREET, selects GREET’s Fuel_Prod_TS sheet (required by GREET to apply the updates), and recomputes the results. Re-run the macro after changing values on the Settings worksheet so the updates are applied to GREET.

Figure 2. Excel’s macro selection panel

Results are shown in the worksheet WTT Results in the analysis workbook. More information about running the analysis can be found in that workbooks Intro worksheet.

2.4. Energy Inputs for Cilion

The key factors that distinguish the Cilion process from other corn to ethanol production processes includes:

• Location of plant and transportation distance for corn

• Energy inputs for plant

• Plant type

• Plant fuel

• Thermal and electrical energy consumption Since the Cilion plant produces wet distillers grains, the energy inputs are lower than those for conventional corn to ethanol plants. The plant also produces electricity on-site, which results in a lower WTT energy than using imported electric power.

Table 2 shows the performance data values provided by Cilion for the Keyes plant, based on engineering estimates. Table 3 lists the customizations to GREET 1.7 applied for this analysis. (Note that GREET does not account for trucking distillers grains. See section 2.6 for an explanation of how this was handled in the present report.)

Page 7: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

7

Table 2. Performance data for Keyes plant provided by Cilion

Item Value Plant capacity (MMgpy) 55 Plant type Dry-mill Energy source NG Thermal energy (Btu/gal, HHV) 23,400* Electricity (kWh/gal) 0.0 Corn rail distance (mi) 1500 Corn truck-to-plant distance (mi) 0 Percent denaturant 4.76% Denaturant type Natural Gasoline Denatured EtOH yield (gal/bu) 2.8 Percent of DDGS dried 0% Percent dry matter in DGs 35% Trucking distance for WDGs (mi) 30

*Based on .234 Therms where 1 Therm = 100,000 BTU, HHV

Table 3. Values customized in GREET

Target Sheet Cell Value Category Description Inputs B4 2010 Target Year Year of analysis. Left at 2010 to be able

to set values w/o interpolation Fuel_Prod_TS BH345 0.6% Feedstock Production Oil fraction of MROW grid Fuel_Prod_TS BI345 1.8% NG fraction of MROW grid Fuel_Prod_TS BJ345 74.6% Coal fraction of MROW grid Fuel_Prod_TS BK345 16.0% Nuclear fraction of MROW grid Fuel_Prod_TS BL345 0.8% Biomass fraction of MROW grid Inputs C309 4 Selects “User Defined grid” for

stationary electricity use Inputs AU191 1500 Feedstock Transport Google maps shows the driving distance

between Lincoln, NE and Keyes, CA as 1600 mi. Direct distance is about 1300 mi. We'll assume rail is somewhat more direct than car at 1500 mi.

Inputs AU193 0 No truck transport required between rail and Keyes plant.

Inputs AU189 0 No barge transport between NE and CA Fuel_Prod_TS D277 2.67 Fuel Production Anhydrous ethanol yield. Fuel_Prod_TS L277 21,122 Ethanol plant energy use (LHV).

Electricity is added later. Fuel_Prod_TS D291 100% Fraction of corn ethanol plants from dry

mills. Set to 100% to ignore wet mills. Fuel_Prod_TS T277 0% Fraction of dry mills using coal. Set to

0% to model a NG-fired plant. EtOH K117 0.0% Fraction of total dry mill energy from

electricity. Natural Gas fraction is calculated in GREET as (1 - electricity).

Inputs BA189 0 Ethanol Distribution Barge (miles one way)* Inputs BA191 800 Rail (miles one way)*

Page 8: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

8

Inputs BA192 80 Truck1 to Bulk Terminals (miles one way)*

Inputs BA193 30 Truck2 for distribution to refueling station (miles one way)*

T&D FZ87 0% Distribution from plant (Fuel Share)

Barge: % Diesel, rest residual oil*

T&D GA87 20% Pipeline: % Diesel* T&D GA88 50% Pipeline: % Residual Oil* T&D GA89 24% Pipeline: % Natural Gas (rest is

electricity)* T&D GB87 100% Rail: % Diesel (rest is electricity)* T&D GC87 100% Truck1: % Diesel (rest is ethanol)* T&D GD87 100% Truck2: % Diesel (rest is ethanol)* *From modifications to GREET made for California context under AB1007 Full Fuel Cycle Analysis.

2.5. Corn farming in GREET

The assumptions about on-farm energy use for corn production are exogenous to GREET and were last documented for GREET version 1.5 (1999), yet the values in GREET 1.7 are considerably different. These assumptions include both the total on-farm energy consumption per bushel of corn, and the allocation of that total between various energy types: diesel, gasoline, natural gas, LPG, and electricity. Based on discussions with (GREET author) Michael Wang, I have attempted to reconstruct the calculation resulting in the numbers used in GREET 1.7, but have failed to duplicate his results to my satisfaction.2 Besides the lack of documentation, there are several issues with the handling of corn farming energy in GREET 1.7. The analysis here represents a best estimate at applying the most recently published data to the GREET framework for the energy inputs to corn production.

1. Data vintage The assumptions in GREET are based on a report published by the USDA in 2002 (Shapouri, Duffield et al. 2002; Wang 2007). This report utilized the most recent data available in 2002 from the USDA’s survey of corn farmers, which was conducted in 1996. In 2004, the USDA updated its estimate of the 2001 net energy balance of corn ethanol based on the newly available 2001 agricultural survey (Shapouri, Duffield et al. 2004). GREET 1.7, however, remains based on the 1996 data, now questionably representative of current energy practices.

2. Projected energy use GREET uses a set of "time series" tables to define values for parameters that are expected to change over time. Corn farming energy is one such parameter, showing a progressive decline through the years. The rationale behind the projected values is undocumented. The allocation of this energy between energy types is constant in GREET (not subject to time series), which is equivalent to assuming that energy use will decline by equal proportions across all fuel types.

2 Detailed calculations are provided in the Corn worksheet of the delivered Excel workbook.

Page 9: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

9

3. Replicability Two data values from the underlying USDA farming survey that are recorded as costs must be converted to quantities of energy and allocated across the types of energy tracked in GREET. These parameters are Custom Work and Input Hauling. The prior includes any hired farming services such as planting, fertilizing, harvesting, or drying the corn. The latter involves the transportation of agricultural chemicals. Neither the 1999 GREET documentation, nor the USDA studies document these conversions. For this report, we rely on a well-regarded study of corn ethanol of the same era which does detail the assumptions regarding energy use for custom work (Graboski 2002). These values do not match the default values in GREET and represent our best estimate of the use of the USDA data.

The Corn worksheet of the delivered spreadsheet computes the corn farming energy based on the two cited USDA studies, producing results in GREET format. While the version based on the 2002 paper (1996 data) comes reasonably close to the assumptions in GREET 1.7, the more current (2001) data from the 2004 paper show a sharp decline in total energy: 15,600 Btu/bushel average in 2001 versus 22,900 Btu/bushel average in 1996. Note that the default GREET assumption projected for 2010 is 22,500 Btu/bushel. The bottom line is that data limitations prevent a reliable and comparable estimate of Cilion’s ethanol using the assumptions about corn farming in the states from which the company sources its corn. Comparing historical survey data from six or eleven years ago with projections three years out would not yield a meaningful result.

2.6. Shipping of Distillers Grains

The GREET model does not account for the emissions or energy used in the shipping of coproducts. Presumably these are considered comparable to shipping the caloric equivalent in whole corn or soybeans, and therefore there is not much difference from the base case. We compared the energy use for shipping DDGS versus WDG to understand how this process difference affects the life cycle assessment. Based on standard assumptions about DDGS yield (18 lbs/bushel @ 10% moisture) we compute the mass of WDG produced per gallon, for Cilion’s stated 35% dry matter. As shown in Table 4, we estimate 16.5 lbs of WDG are produced per gallon of ethanol. Table 5 shows the energy per ton-mile and per lb-mile estimated in GREET for shipping corn from the field to stacks. (We use the emissions from shipping corn as a proxy for those from shipping an equal mass of distillers grains.) Table 6 shows our assumptions for shipping of distillers grains. We assume the WDG is trucked 30 miles, with no rail transport. We compare this to shipping DDGS 10 miles by truck and 750 miles by rail. As rail is much more energy efficient, and DDGS is lighter, the energy required (in diesel fuel) for these two alternatives is comparable.3

3 These assumptions can be adjusted in the accompanying spreadsheet to compare alternative shipping scenarios.

Page 10: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

10

Table 4. Mass of dry and wet distillers grains per bushel corn and per gallon ethanol

Dry matter (wt %)

Weight (lb/bu)

Solids (lb/bu)

Water (lb/bu)

Weight (lb/gal)

Solids (lb/gal)

Water (lb/gal)

DDGS 90% 18.0 16.2 1.8 6.4 5.8 0.6 WDG 35% 46.3 16.2 30.1 16.5 5.8 10.7 Note: WDG mass is computed from assumptions about standard DDGS yield and Cilion's stated WDG composition.

Table 5. Corn transport energy in GREET

Truck Rail Btu/ton-mile Btu/lb-mile Btu/ton-mile Btu/lb-mile

4398 2.20 370 0.19

Table 6. Shipping assumptions for WDG and DDGS

Truck miles Rail miles Btu/gal WDG transport 30 0 1091 DDGS transport 10 750 1033

Since the analyses we wish to use for comparison have not modeled the shipping of DDGS, and since the diesel energy use is similar, we believe it is reasonable to ignore shipping of WDG in the present analysis as well. However, a more complete analysis (especially of local and regional pollutants) would consider the differences between these scenarios. In addition, the shipping of water in WDG reduces somewhat the pumping energy for delivery of water to the feedlot, which should be accounted for as a credit in the WDG case.

2.7. Electricity

Electric power is a component of corn and conventional ethanol production. Typical modern ethanol plants consume 0.75 to 1 kWh of power per gallon of ethanol produced. By default, GREET assumes that average US electricity is used for all processes. GREET allows customization of the resource mix in the electricity grid, but the chosen grid definition is used for all phases of the life cycle. In Cilion’s case, the corn is produced in the Nebraska, Minnesota, and Iowa, all of which are located in the MROW region, as shown in Figure 3. Electricity in this region is heavily coal-dominated, with GHG emissions about 20% higher per kWh than the national average.4 Table 7 shows the mix of resources used to generate electricity in these three states. Table 8 shows the life cycle greenhouse gas emissions for the CAMX (California) and MROW regions, and for the US. Table 7. Electricity generation resource mix in the MROW eGRID region.

Source: EPA eGRID2006 V2.1 (based on 2004 data), downloadable from http://www.epa.gov/cleanenergy/egrid/

4 Based on the author’s calculations in developing the BEACCON model, available at LifeCycleAssociates.com.

MROW eGRID Region % of generation Oil 0.6% Natural Gas 1.8% Coal 74.6% Nuclear 16.0% Biomass 0.8%

Page 11: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

11

Table 8. Comparison of life cycle GHG emissions for CAMX (California), MROW, and US average. (g CO2eq/kWh)

eGRID region CAMX MROW US average Upstream 46 42 46 Combustion 417 899 730 Total 464 941 776 Source: EPA eGRID2006 and GREET 1.7.

To address this limitation, we configured GREET to use the MROW regional grid, while setting to zero the amount of electricity used in Cilion’s ethanol plant. We then used the emissions factors for marginal electricity as defined in the AB1007 report (i.e. natural gas combined-cycle generation), combining these results to produce the total emissions for Cilion’s ethanol. [An update from Cilion indicated that the Keyes plant produces 100% of it’s required electricity, so the calculation of plant electricity is moot. However, we left this in the model and here in the documentation to allow Cilion to explore the effects of purchasing CA grid electricity for use in its plants.]

Figure 3. eGRID regions defined by the US EPA.

2.8. Ethanol Denaturant

Ethanol plants add gasoline to the anhydrous ethanol to produce denatured ethanol, which is considered the final output of the ethanol plant. Cilion uses natural gasoline as denaturant. The energy and GHG analysis here therefore includes this specific denaturant in the ethanol. GREET does not explicitly model denaturant. Rather, the stated percentage of denaturant added to ethanol is used to compute the percentage of anhydrous ethanol in the resulting fuel mix. The non-ethanol portion of the fuel is assumed to be uniform, i.e. if blending denatured ethanol into

Page 12: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

12

CARFG, the denaturant is implicitly treated as CARFG. We correct for this in our model by using GREET’s results for pure anhydrous as used in an ethanol fuel cell vehicle and adding in the correct percentage of denaturant in our post-processor. However, since GREET does not model natural gasoline, we use GREET’s results for naptha as the nearest available proxy. The maximum denaturant allowed in California is 5 parts to 100 parts ethanol. This translates to a volumetric limit of 5/105, or 4.76% denaturant. The comment on the GREET input sheet identifies this correctly: “Share of gasoline by volume added in EtOH as denaturant” (emphasis added), but the value is used incorrectly in the calculation of the fraction of anhydrous in low-level blends of ethanol, i.e. it’s treated as if there is 5% denaturant. Granted, this is a minor difference, but worth noting since we correct this in our calculations for Cilion ethanol. Note that the “standard” GREET Midwest NG-fired ethanol was modeled using the same approach to denaturant to allow for a more appropriate comparison with Cilion ethanol.

2.9. California Reformulated Gasoline

We modeled CARFG without oxygenate in GREET by setting the oxygenate to “5 – No oxygenate” in the Inputs sheet.

3. Results

The fuel cycle energy inputs, GHG emissions, and critieria pollutant emissions are shown here. The results include the upstream or WTT portion plus the fuel itself and are expressed Joules per Joule of fuel (for energy), or as grams per MJ of fuel (for emissions.) GREET produces an accounting of the life cycle emissions of greenhouse gases and criteria pollutants, as well an accounting of total fossil energy and total petroleum energy used in the production and use of different fuels. Table 9 shows the complete result vectors for Cilion Keyes ethanol under various assumptions about corn farming energy, compared to baseline natural-gas fired dry-mill ethanol in GREET 1.7, and compared to baseline CA reformulated gasoline (without ethanol added.) The results are examined in detail below.

3.1. Greenhouse Gas Emissions

GREET tracks the life cycle emissions of carbon dioxide (CO2), carbon monoxide (CO), methane (CH4), nitrous oxide (N2O) and volatile organic compounds (VOC). In the case of CO and VOCs, GREET computes the fraction of carbon in each that is converted to CO2. The total GHG emissions balance is computed as the sum of the three primary GHGs (CO2, CH4, and N2O), weighted by their respective 100-year global warming potential (GWP) values as defined by the IPCC. The 100-year GWP values are estimates of the relative climate forcing effect of a given mass of gas compared to the same mass of CO2. The values for N2O and CH4 are 296 and 23, respectively. That is, each gram of N2O is equivalent to 296 grams of CO2 in terms of climate forcing over a 100-year horizon. Note that the choice of time horizon is fairly arbitrary, and the changes in GWP values are non-linear with the value of time horizon, so this choice materially affects the outcome. The use of 100-year GWPs appears to be the most common, thus their use in GREET and in this study.

Page 13: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

13

Table 9. Life cycle results for Cilion Keyes ethanol compared to standard GREET ethanol and CARFG

Parameter Units Cilion

Avg US

Corn

NG-fired

dry mill

CARFG

w/o EtOHFuel Cycle Energy J/J 1.09 1.38 0.22

WTT Efficiency % 0.48 42.0% 81.7%

Fossil Fuels J/J 0.44 0.71 0.22

Coal J/J 0.04 0.10 0.04

Natural Gas J/J 0.34 0.52 0.08

Petroleum J/J 0.06 0.09 0.10

CO2 g/MJ -36.06 -17.21 16.75

CH4 g/MJ 0.07 0.11 0.10

N2O g/MJ 0.04 0.04 0.00

GHGs (weighted) g/MJ -21.85 -2.01 19.15

VOC (total) g/GJ 37.50 41.30 24.87

CO (total) g/GJ 33.45 42.94 12.12

NOx (total) g/GJ 95.88 126.74 35.01

PM10 (total) g/GJ 20.55 31.91 8.26

PM2.5 (total) g/GJ 6.86 10.36 2.94

SOx (total) g/GJ 36.72 60.72 20.40

VOC (urban) g/GJ 13.05 13.22 15.69

CO (urban) g/GJ 0.72 1.52 6.08

NOx (urban) g/GJ 2.91 5.88 13.95

PM10 Exh (urban) g/GJ 0.16 0.32 1.15

PM2.5 Exh (urban) g/GJ 0.10 0.20 0.81

SOx (urban) g/GJ 2.13 6.19 10.81

GHG LUC Delta (for case 2 - 1) g/MJ 0.07 0.07 0.00

Net GHG, No LUC g/MJ 49.3 69.1 92.1

Fossil Energy Value MJ/L 15.5 9.6

Fossil Energy Ratio J/J 2.6 1.6

GWI reduction from MW EtOH 29% 0%

GWI reduction from CARFG 46% 25%

Standard GREET

Finally, it is important to note that these three gases do not account for the entire climate impact of biofuels production. Other factors, such as indirect GHGs (i.e. gases which enable the conversion of other gases to GHGs); biogeophysical factors such as changes in albedo (reflectance), evapotranspiration; or market mediated effects such as global land use conversion are not accounted for in GREET or in this study. This is important because regulations such as California’s Low Carbon Fuel Standard may eventually take these additional factors into consideration. GREET 1.7 does include a small charge for domestic land use conversion due to the expansion of corn production due to increased demand from ethanol markets. The factor built into GREET, however, is very outdated. It is based on a 1999 economic analysis by the USDA considering the expansion of U.S. ethanol capacity from 1.5 billion to 3 billion gallons per year (Wang 2007). It is beyond the scope of this study to compute an updated emissions factor for land use conversion.

Page 14: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

14

Table 10 summarizes the GWI of corn ethanol produced in Cilion’s Keyes facility compared with ethanol produced in the Midwest based on standard GREET corn farming energy assumptions. Note that the Standard GREET ethanol does not include shipping to California. That is, this compares Cilion ethanol produced in California to Midwest ethanol shipped the standard GREET shipping distance. Using the default treatment of corn farming energy in GREET 1.7, the GWI of Cilion ethanol is 29% lower than that of the baseline natural-gas fired, dry-mill corn ethanol produced in the Midwest, and 46% lower than that of California RFG.

Table 10. Life cycle GHG emissions for Cilion ethanol compared to average Midwest ethanol, and to California reformulated gasoline (CARFG). (g CO2eq/MJ)

Cilion EtOH GREET corn

farming energy

Standard GREET NG-gas fired

EtOH

Standard GREET CARFG (without

ethanol) GWI 49.3 69.1 92.1

3.2. Criteria Pollutant Emissions

GREET tracks the following criteria pollutants: oxides of nitrogen (NOx), oxides of sulfur (SOx), carbon monoxide (CO), volatile organic compounds (VOC), sub-10 micron particulate matter (PM10), and sub-2.5 micron particulate matter (PM2.5). The result for each is reported below. GREET tracks total emissions and for each stage in the life cycle designates the “urban share” of the emissions, meaning the fraction of the total that are emitted in urban environments, which is related to US smog. However, GREET does not attempt to track where urban emissions occur, so neither the total emissions nor the urban emissions is useful in any regulatory sense. Cilion ethanol has lower emissions of all criteria pollutants tracked in GREET than the baseline Midwest dry-mill ethanol—for all three corn energy values. This is surely due to the use of coal-heavy electricity in the standard Midwest case. Even so, Cilion ethanol has higher emissions of all of the criteria pollutants than GREET’s baseline CARFG.

Table 11. Life Cycle Emissions of Criteria Pollutants for Cilion Ethanol compared to baseline Midwest natural-gas, dry-mill corn ethanol and CARFG. (g/GJ)

Pollutant Cilion Avg US Corn

NG-fired dry mill

CARFG w/o EtOH

VOC 37.50 41.30 24.87 CO 33.45 42.94 12.12 NOx 95.88 126.74 35.01 PM10 20.55 31.91 8.26 PM2.5 6.86 10.36 2.94 SOx 36.72 60.72 20.40

Page 15: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

15

3.3. Net Energy Balance

Although widely cited, Net Energy Balance and Net Energy Ratio are not particularly useful metrics for understanding the value or performance of an ethanol production system (Farrell, Plevin et al. 2006; Dale 2007).5 Petroleum intensity and global warming intensity are far more relevant to the primary concerns which biofuels are generally considered to address. However, as interested parties may request Cilion’s net energy balance, the calculation may be of interest. GREET calculates the life cycle use of energy in the production of different fuels, differentiated by energy source. Specifically, it tracks: total energy (fossil plus renewable), total fossil energy, and fractions of energy from each of petroleum, natural gas, and coal. We report each of these results below. Table 12 shows the total energy and the consumption of specific energy resources per unit of each fuel produced. Cilion ethanol uses less total energy than baseline Midwest corn ethanol, but all ethanol uses much more energy than does CARFG. Each Joule of Cilion ethanol requires 0.44 Joules of fossil energy, mostly from natural gas (0.34 J), with much smaller contributions from coal (0.04 J) and petroleum (0.06 J). In fact, less petroleum is consumed to produce a Joule of Cilion ethanol than is consumed to produce a Joule of CARFG. As non-fossil energy inputs are ignored, the measurement generally called “net energy value” is better termed fossil energy value, or FEV (Farrell, Plevin et al. 2006). Following (Spitzley and Keoleian 2005) we define FEV as:

FEV = Eout - (FF + PF)

where Eout is the energy content in the delivered fuel (net of coproduct credits), FF is primary fossil energy in feedstocks (including denaturant), and PF is the primary fossil input energy in non-feedstocks. Using standard GREET corn farming assumptions, the coproduct fossil energy credit for displaced whole corn and soybean meal is 3.3 MJ per liter of ethanol. The lower heating value (LHV) of Cilion’s denatured ethanol is 21.8 MJ/L, so Eout is 21.8 + 3.3 or 25.1 MJ/L. The factor PF is computed by simply scaling up the J/J factor for Fossil Fuels given in Table 12 by the LHV of ethanol. Cilion ethanol has a Fossil Energy Value of 14.6 MJ/L, which compares quite favorably to the default NG-fired dry-mill corn ethanol in GREET, with FEV = 8.7 MJ/L. Net energy balance can also be expressed as a ratio of output energy divided by input energy. Again, as only fossil energy inputs are counted, this quantity is more accurately termed Fossil Energy Ratio, or FER. We define FER as: FER = Eout / (FF + PF) 5 This issue is discussed at length in a Science article on ethanol (and the supporting online materials) published by

my colleagues and myself in 2006, available at http://rael.berkeley.edu/ebamm.

Page 16: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

16

Note that the ratio suffers the same conceptual challenges as FEV in terms of adding dissimilar energy types, while in addition, FER varies depending on whether one subtracts coproduct energy from the denominator or adds it to the numerator. In this report, we have added the coproduct energy to the numerator.

Table 12. Net Energy Balance of Cilion Ethanol

Units Cilion Avg US Corn

NG-fired dry mill

CARFG w/o EtOH

Fuel Cycle Energy J/J 1.09 1.38 0.22 WTT Efficiency % 0.48 42.0% 81.7% Fossil Fuels J/J 0.44 0.71 0.22 Coal J/J 0.04 0.10 0.04 Natural Gas J/J 0.34 0.52 0.08 Petroleum J/J 0.06 0.09 0.10 Fossil Energy Value MJ/L 15.5 9.6 Fossil Energy Ratio J/J 2.6 1.6

3.4. Petroleum Displacement

California reformulated gasoline is currently mandated to contain 5.7% ethanol by volume. Therefore, unless and until the state raises this blend level, ethanol produced and sold in-state by Cilion would mostly displace ethanol imported from the Midwest (90%) and Brazil (10%); it would not displace additional CARFG. To estimate the actual quantity and type of fuel displaced by an additional ethanol plant in California would require an economic model of fuel, food, and feed markets, which is beyond the scope of this paper. We can, however, compare the amount of petroleum required to produce a liter each of CARFG and Cilion Ethanol. While it would be incorrect to characterize the difference between these as the net displacement, the difference does provide some demonstrate the much lower petroleum requirements for producing ethanol. As shown in Table 12, producing a Joule of CARFG consumes 0.10 Joules of petroleum. Therefore, a Joule of CARFG requires 1.10 Joules of petroleum for each Joule delivered. This represents the fuel itself plus the petroleum consumed in its production. The production of each Joule of Cilion’s denatured ethanol consumes 0.06 Joules of petroleum energy over the life cycle, and directly includes 0.07 Joules of natural gasoline, which is 100% petroleum-based, for a total of 0.13 Joules of petroleum per Joule of denatured ethanol. The ethanol therefore requires only 0.13/1.10 or 12% as much petroleum per Joule of pure CARFG. Thus a gasoline gallon equivalent of denatured ethanol used in place of CARFG (pure, with no ethanol) would save 88% of the petroleum.

4. Conclusion

Energy inputs and emissions from Cilion ethanol are lower on all measures than those of NG-fired dry-mill corn ethanol produced in the Midwest. This achievement is due primarily to two

Page 17: Fuel Cycle Assessment of Ethanol Produced at Cilion’s

17

factors: (1) the avoidance of drying distillers grains, and (2) much cleaner (self-produced) electricity.6

5. Acknowledgements

Thanks to Stefan Unnasch and Brent Riffel of Life Cycle Associates for their suggestions and improvements to the model, and helpful comments on the report.

6. Works Cited

Dale, B. E. (2007). "Thinking clearly about biofuels: ending the irrelevant 'net energy' debate and developing better performance metrics for alternative fuels." Biofuels, Bioproducts and Biorefining 1(1): 14-17.

Farrell, A. E., R. J. Plevin, et al. (2006). "Energy returns on ethanol production - Response." Science 312(5781): 1747-1748.

Farrell, A. E., R. J. Plevin, et al. (2006). "Ethanol Can Contribute to Energy and Environmental Goals." Science 311: 506-508.

Graboski, M. S. (2002). Fossil Energy Use in the Manufacture of Corn Ethanol, National Corn Growers Association.

Shapouri, H., J. A. Duffield, et al. (2004). The 2001 Net Energy Balance of Corn-Ethanol. Arlington, VA. Shapouri, H., J. A. Duffield, et al. (2002). The Energy Balance of Corn Ethanol: An Update. AER-814.

Washington, DC, US Department of Agriculture, Available at http://www.usda.gov/oce/oepnu/aer-814.pdf.

Spitzley, D. V. and G. A. Keoleian (2005). Life Cycle Environmental and Economic Assessment of Willow Biomass Electricity: A Comparison with Other Renewable and Non-Renewable Sources. Ann Arbor, University of Michigan.

Unnasch, S. (2007). Full Fuel Cycle Assessment, Well to Tank Energy Inputs, Emissions, and Water Impacts, California Energy Commission, Available at http://www.energy.ca.gov/ab1007/documents/.

Wang, M. (2007). Corn farming energy in GREET 1.7. (Pers. comm.). R. J. Plevin. Wang, M. (2007). Land use change in GREET. (Pers. comm.). R. J. Plevin.

6 It isn’t clear how ethanol produced under similar conditions in the Midwest and shipped to California would

compare to Cilion ethanol produced with corn shipped to California. Assuming whole corn was being shipped to California to feed cattle, about 1/3 of the feed value of the whole corn would remain available in the form of WDG, thereby requiring replacement of only 2/3 of the original quantity of corn imported. On the other hand, the WDGs might displace DDGS shipped from the Midwest, which would have a different energy / emissions profile.