Docstoc

Sensitivity Analysis of Bioethanol LCA Models to Determine

Document Sample
Sensitivity Analysis of Bioethanol LCA Models to Determine Powered By Docstoc
					CHEMINFO




     Sensitivity Analysis of Bioethanol LCA Models
               to Determine Assumptions
        With the Greatest Influence on Outputs




                                        March 31, 2008




                                         Submitted By:

     2
(S&T) Consultants Inc.      Cheminfo Services Inc.      Dr. Heather MacLean     Fugacity Technology
11657 Summit Crescent     315 Renfrew Drive Suite 302   University of Toronto        Consulting
Delta, British Columbia        Markham, Ontario         Associate Professor,      37 Farlane Blvd.,
        V4E 2Z2                     L3R 9S7              Department of Civil      Ottawa, Ontario
 Phone: 604-590-5260         Phone: 905-944-1160             Engineering              K2E 5H3
  Fax: 604-661-2689            Fax: 905-944-1175        University of Toronto   Phone: 613-274-3376
                                                              M5S 1A4
                                                        Phone: 416-946-5056
     CHEMINFO




                                                     Table of Contents

1.      EXECUTIVE SUMMARY .................................................................................................. 1
     1.1        INTRODUCTION ................................................................................................................ 1
     1.2        OVERVIEW OF LIFE CYCLE ASSESSMENT ......................................................................... 2
     1.3        LCA CHALLENGES FOR BIOFUELS ................................................................................... 2
     1.4        LIFE CYCLE MODELS AVAILABLE INITIAL SCREENING .................................................... 3
     1.5        COMPARISON OF GREET AND GHGENIUS RESULTS ....................................................... 4
     1.6        SENSITIVITY ANALYSIS .................................................................................................... 7
     1.7        RECOMMENDATIONS FOR ENHANCING LCA CAPACITY IN CANADA ................................ 8
2.      INTRODUCTION............................................................................................................... 10
     2.1        BACKGROUND ................................................................................................................ 10
     2.2        PURPOSE OF THIS REPORT.............................................................................................. 12
     2.3        APPROACH AND METHODOLOGY ................................................................................... 13
     2.4        INTENDED USE OF REPORT............................................................................................. 14
     2.5        REST OF REPORT ............................................................................................................ 14
3.      THE ROLE OF MODELLING IN POLICY DEVELOPMENT................................... 16
     3.1   OVERVIEW OF CURRENT USES OF LIFE CYCLE ASSESSMENT ......................................... 16
     3.2   GOVERNMENT OF CANADA ENERGY, EMISSIONS AND ECONOMICS MODELLING
     ACTIVITIES ................................................................................................................................ 19
4.      LIFE CYCLE ANALYSES ................................................................................................ 25
     4.1        OVERVIEW ..................................................................................................................... 25
     4.2        ISO LIFE-CYCLE ASSESSMENT STANDARDS .................................................................. 27
     4.3        GOAL DEFINITION AND SCOPING ................................................................................... 34
     4.4        LIFE CYCLE INVENTORY ................................................................................................ 37
     4.5        LIFE CYCLE IMPACT ASSESSMENT ................................................................................. 40
     4.6        LIFE CYCLE INTERPRETATION........................................................................................ 44
     4.7        TYPES OF LCA MODELS ................................................................................................ 46
5.      LIFE CYCLE MODELS, INITIAL SCREENING ......................................................... 48
     5.1        OVERVIEW ..................................................................................................................... 48
     5.2        SCREENING ASSESSMENT OF AVAILABLE MODELS ........................................................ 49
6.      DETAILED ASSESSMENT OF LCA MODELS ............................................................ 65
     6.1        OVERVIEW ..................................................................................................................... 65
     6.2        EVALUATIONS OF MODELS FOR THE BIOETHANOL LIFE CYCLE ..................................... 67
     6.3        SUMMARY OF THE MODEL EVALUATIONS FOR ETHANOL .............................................. 84
7.      COMPARISON OF MODEL RESULTS......................................................................... 87
     7.1        INTRODUCTION .............................................................................................................. 87


                                                                      i
     CHEMINFO




     7.2        ANALYSIS OF REFERENCE FUEL RESULTS ...................................................................... 87
     7.3        ANALYSIS OF BIOFUEL RESULTS.................................................................................... 96
8.     SENSITIVITY ANALYSIS.............................................................................................. 120
     8.1        INTRODUCTION ............................................................................................................ 120
     8.2        PROCESS CONVERSION EFFICIENCIES........................................................................... 120
     8.3        PROCESS ENERGY REQUIREMENTS .............................................................................. 121
     8.4        EMISSION FACTORS...................................................................................................... 123
     8.5        LAND USE .................................................................................................................... 126
     8.6        FEEDSTOCK PRODUCTION ............................................................................................ 142
     8.7        CO-PRODUCTS ............................................................................................................. 144
     8.8        WATER, SOLID WASTE AND OTHER ENVIRONMENTAL IMPACTS ................................. 146
     8.9        UNCERTAINTY: MONTE CARLO ANALYSIS .................................................................. 146
9. RECOMMENDATIONS FOR ENHANCING LCA MODELLING CAPACITY IN
CANADA ................................................................................................................................... 152
     9.1        SUMMARY .................................................................................................................... 152
     9.2        ENHANCE GHGENIUS MODEL FOR CANADIAN CONDITIONS........................................ 152
     9.3        WORK IN PARTNERSHIP WITH GOVERNMENT DEPARTMENTS AND STAKEHOLDERS ..... 153
     9.4        CONTINUE TO SUPPORT TRANSPARENCY AND DOCUMENTATION ................................ 154
     9.5        REMAIN ENGAGED WITH INTERNATIONAL LCA MODELS AND RESULTS .................... 154
     9.6        DEVELOP LONG TERM VISION AND PATH FORWARD FOR CANADA'S LCA CAPACITY. 154
     9.7        EXPORT CANADIAN LCA CAPACITY ........................................................................... 155
10.         REFERENCES AND WEBSITES .............................................................................. 157
     10.1       WEBSITES USED TO ACCESS LCA MODELS, AND OTHERS .......................................... 161
11.         APPENDIX.................................................................................................................... 163
     11.1       SELECTED BIOETHANOL PAPERS.................................................................................. 163




                                                                     ii
   CHEMINFO




                                                                List of Tables
Table 1: Comparison of Lifecycle Gasoline Results with Different Models.................................................................5
Table 2: Comparison of Lifecycle E10 Results with Different Models ........................................................................6
Table 3: Comparison of Corn and Wheat Ethanol with Gasoline .................................................................................7
Table 4: Summary of Sensitivity Analysis ....................................................................................................................8
Table 5: Canadian Ethanol Capacity ...........................................................................................................................11
Table 6: ISO LCA Standards and Technical Reports..................................................................................................29
Table 7: Common Impact Categories ..........................................................................................................................42
Table 8: BEES Evaluation...........................................................................................................................................68
Table 9: BESS Evaluation ...........................................................................................................................................70
Table 10: EIO-LCA Evaluation...................................................................................................................................72
Table 11: GaBi Evaluation ..........................................................................................................................................74
Table 12: GEMIS Evaluation ......................................................................................................................................76
Table 13: GHGenius Evaluation..................................................................................................................................78
Table 14: GREET Evaluation......................................................................................................................................80
Table 15: LEM Evaluation ..........................................................................................................................................82
Table 16: SimaPro (v. 7.1 Analyst) Evaluation...........................................................................................................83
Table 17: Ethanol Model Evaluation Scores (Un-Weighted Scenario).......................................................................85
Table 18: Ethanol Model Evaluation Scores (Weighted Scenario) .............................................................................86
Table 19: Comparison of GHG Emissions for Gasoline .............................................................................................88
Table 20: Comparison of GHG Emissions for Gasoline Combustion.........................................................................91
Table 21: Comparison of Lifecycle Individual Air Contaminant Emissions for Gasoline Production .......................92
Table 22: Comparison of Individual Air Contaminant Emissions for Oil Production ................................................93
Table 23: Comparison of Individual Air Contaminant Emissions for Gasoline Refining...........................................93
Table 24: Comparison of Individual Air Contaminant Emissions for Gasoline Combustion .....................................94
Table 25: Energy Balance for Gasoline.......................................................................................................................94
Table 26: Comparison of Lifecycle Gasoline Results with Different Models.............................................................96
Table 27: Comparison of GHG Emissions for Corn Ethanol ......................................................................................99
Table 28: Comparison of GHG Emissions for Corn Production ...............................................................................100
Table 29: Comparison of Inputs for Corn Production ...............................................................................................101
Table 30: Comparison of Energy and GHG Intensities for Inputs ............................................................................102
Table 31: Comparison of Energy and GHG Intensities for 2 Models .......................................................................102
Table 32: Comparison of Land Use Emissions for Corn Production ........................................................................103
Table 33: Comparison of Energy and GHG Intensity for Ethanol Production..........................................................104
Table 34: DDGS Displacement Ratios ......................................................................................................................106
Table 35: DDGS GHG Credits..................................................................................................................................106
Table 36: Lifecycle GHG Emission Results for E10.................................................................................................109
Table 37: Lifecycle GHG Emission Results for E85.................................................................................................110
Table 38: Comparison of Lifecycle Air Emissions for Corn Ethanol Production .....................................................111
Table 39: Energy Balance for Corn Ethanol .............................................................................................................112
Table 40: Comparison of Lifecycle E10 Results with Different Models ..................................................................113
Table 41: Comparison of Lifecycle GHG Emissions- Gasoline, Corn and Wheat Ethanol (E100) ..........................114
Table 42: Comparison of Corn and Wheat Production Inputs ..................................................................................115
Table 43: Comparison of Land Use Emissions for Corn and Wheat Production ......................................................115
Table 44: DDGS GHG Credits..................................................................................................................................116
Table 45: Comparison of Air Emissions for Production of Gasoline, Corn and Wheat Ethanol ..............................117
Table 46: Comparison of Lifecycle Air Emissions- Gasoline, E10 Corn and Wheat Ethanol ..................................118
Table 47: Comparison of Energy Balance- Gasoline, Corn and Wheat Ethanol (E10).............................................119
Table 48: Comparison of Energy Balance- Gasoline, Corn and Future Corn Ethanol..............................................123
Table 49: Comparison of GHG Emissions- Impact of N2O Emission Factor ...........................................................124



                                                                               iii
   CHEMINFO




Table 50: Comparison of NOx Emissions- Impact of Combustion Emission Factor ................................................126
Table 51: Soil Carbon Changes under Improved Management.................................................................................130
Table 52: Soil Carbon Changes under Improved Management.................................................................................131
Table 53: Corn Supply and Demand .........................................................................................................................136
Table 54: Wheat Supply and Demand.......................................................................................................................138
Table 55: Wheat Characteristics................................................................................................................................139
Table 56: GHG Emissions Impact of Wheat Variety ................................................................................................140
Table 57: Arable Land for Rainfed Agriculture ........................................................................................................141
Table 58: Corn Crop Inputs.......................................................................................................................................143
Table 59: GHG Emissions Impact of Tillage Practice on Corn Emissions ...............................................................144
Table 60: Allocation Approaches ..............................................................................................................................145
Table 61: Alternative CO2 Approaches .....................................................................................................................145
Table 62: Comparison of Corn Ethanol GHG Emissions- Monte Carlo Simulation.................................................149
Table 63: Comparison of Wheat Ethanol GHG Emissions- Monte Carlo Simulation ..............................................151




                                                               List of Figures
Figure 1: Maple C Schematic ......................................................................................................................................21
Figure 2: Environment Canada's E3MC Modelling Framework .................................................................................22
Figure 3: Environment Canada's 3EMC Analytical Process for Estimating the Impacts of Emission Reduction
      Targets For Air Pollutants..................................................................................................................................23
Figure 4: Life Cycle Stages .........................................................................................................................................26
Figure 5: Phases of a LCA...........................................................................................................................................27
Figure 6: Process Flow Diagram .................................................................................................................................38
Figure 7: Life Cycle Interpretation Process.................................................................................................................45
Figure 8: Corn Ethanol Lifecycle Stages.....................................................................................................................97
Figure 9: Ethanol Production Process .........................................................................................................................98
Figure 10: Ethanol Plant Energy Consumption .........................................................................................................105
Figure 11: Impact of Yield on GHG Emissions ........................................................................................................121
Figure 12: Impact of Lower Energy Consumption on GHG Emissions....................................................................122
Figure 13: Impact of N2O Emission Factor for Fertilizer Application on Wheat Ethanol Emissions .......................125
Figure 14: Impact of Biological N2O Emissions on GHG Emissions .......................................................................127
Figure 15: Canadian Tillage Practices.......................................................................................................................129
Figure 16: Impact of Soil Carbon Change on Upstream Emissions ..........................................................................131
Figure 17: Agricultural Land Use in Canada.............................................................................................................133
Figure 18: World Corn Production and Yield ...........................................................................................................135
Figure 19: Rate of Change in Corn Yield..................................................................................................................137
Figure 20: World Wheat Production and Yield.........................................................................................................138
Figure 21: Results from Monte Carlo Simulation for Corn Ethanol .........................................................................148
Figure 22: Results from Monte Carlo Simulation for Wheat Ethanol .......................................................................150




                                                                               iv
CHEMINFO




                               1. Executive Summary

1.1 Introduction

Bioethanol presents an alternative transportation fuel to gasoline made from crude oil. 1 As a
result of the potential environmental, economic and social benefits associated with bioethanol
and other biofuels (e.g., biodiesel), the Government of Canada is committed to developing a
national renewable fuels strategy by targeting a five per cent (5%) renewable content in
Canadian gasoline 2 . In conjunction, federal government funds are being made available to
bolster the development of biofuels and other bioproducts. 3 As a consequence, it is expected that
the ethanol production capacity could triple by the end of the decade and continue to increase
thereafter. The rapid potential growth of Canadian ethanol production and use has resulted in
questions regarding the magnitude of the overall environmental benefits as well as drawbacks of
ethanol. Like all transportation fuels, the magnitude of the environmental attractiveness of
ethanol depends on its total life cycle footprint relative to the footprint(s) associated with
alternative fuels, notably gasoline from crude oil.

Environmental life cycle analysis (LCA) models are complex tools that can be applied to help
assess the relative attractiveness of transportation fuels and other products. LCA can be used to
inform government policy makers, industry, community stakeholders and other groups in making
more informed decisions for the Canadian and regional environment. However, there are a
variety of models available and results from these vary, mostly for valid reasons. This can reduce
confidence in LCA results for ethanol. In addition, the quality of conclusions drawn from LCA
models are influenced by the assumptions and data used in generating model results. As a result,
Environment Canada needed to develop a better understanding of LCA models available as well
as the underlying data, assumptions and associated calculations these tools use to generate
results.

The main overall purpose of this report is to provide an assessment of existing LCA models that
can be applied in determining the environmental footprint of bioethanol and competing fuels
(e.g., gasoline) in Canada. The assessment includes analysis to identify the key factors that
contribute to differences in the results from different models. This report also provides: an
analysis regarding the role of LCA in policy formulation; an overview of other modelling

1 In this report, "ethanol" and "bioethanol" are used interchangeably. Use of the word "ethanol" refers to bioethanol
unless otherwise specified. Ethanol can be commercially made through chemical synthesis (e.g., catalytic hydration
of ethylene) and via fermentation of sugars by yeast. There is no longer any commercial production of synthetic
ethanol from ethylene in Canada.
2 Canadian Renewable Fuels Association, July 2006. Canadian renewable fuels strategy. Prepared by (S&T)²
Consultants Inc. and Meyers Norris Penny LLP.
3 Government of Canada, December 20, 2006Canada's New Government Takes New Step to Protect the
Environment With Biofuels, http://www.agr.gc.ca/cb/index_e.php?s1=n&s2=2006&page=n61220


                                                        1
CHEMINFO




activities oriented to environmental policy development; an overview of what LCA is and how it
works; a brief description, assessment and availability (for the purposes and scope of this study)
of 37 LCA models, more detailed analysis of 9 models and selection of 2 models for detailed
analysis and comparison; sensitivity analysis using one model to identify the factors in the life
cycle that most strongly influence LCA results; and recommendations regarding the development
and enhancement of LCA modelling for Canada.


1.2 Overview of Life Cycle Assessment

Life cycle assessment is a "cradle-to-grave" (or “well to wheels”) approach for assessing
industrial systems and their products. "Cradle-to-grave" begins with the gathering of raw
materials from the earth to create the product and ends at the point when all materials are
returned to the earth. LCA evaluates all stages of a product's life from the perspective that they
are interdependent, meaning that one operation leads to the next. LCA enables the estimation of
the cumulative environmental impacts resulting from all stages in the product life cycle, often
including impacts not considered in more traditional analyses (e.g., raw material extraction,
material transportation, ultimate product disposal, etc.). By including the impacts throughout the
product life cycle, LCA provides a comprehensive view of the environmental aspects of the
product or process and a more accurate picture of the true environmental trade-offs in product
selection.

Specifically, LCA is a technique to assess the environmental aspects and potential impacts
associated with a product, process, or service, by:

 •   Compiling an inventory of relevant energy and material inputs and environmental releases;
 •   Evaluating the potential environmental impacts associated with identified inputs and
     releases;
 •   Interpreting the results to help make more informed decisions.

The term "life cycle" refers to the major activities in the course of the product's life span from its
manufacture, use, maintenance, and final disposal; including the raw material acquisition
required to manufacture the product.


1.3 LCA Challenges for Biofuels

Numerous LCAs of biofuels have been published (reviews include Fleming et al. 2006 and
Larson 2006). Most studies have followed ISO standards (ISO 2006) but a wide range of results
has often been reported for the same fuel pathway, sometimes even when holding temporal and
spatial considerations constant. The ranges in results may, in some cases, be attributed to actual
differences in the systems being modeled but are also due to differences in method interpretation,
assumptions and data issues.


                                                 2
CHEMINFO




Key issues in biofuel LCAs have been differing boundaries being adopted in studies (i.e., what
activities are included/excluded from the study), differences in data being collected and utilized,
and disparities in the treatment of co-products. Boundaries in prior LCAs have often differed due
to resource constraints. Data requirements in LCA are significant. Studies have not always used
up-to-date data or data that reflect the inputs in the relevant process under study (i.e., utilization
of electricity generation data for another jurisdiction rather than the one under study). There are
also gaps in scientific knowledge surrounding key variables. For example, these include
implications of land use change, N2O emissions related to feedstock production, and nutrient
depletion and erosion due to agricultural residue removal. Most biofuel production processes
produce several products [e.g., the production of distillers dried grain from the ethanol
production process]. Including co-product credits often significantly improves the environmental
performance of many fuel pathways. Utilization of different co-product methods, and in some
studies, ignoring co-products entirely, has had major impact on results of LCA studies (Kim and
Dale 2002, Larson 2006, Farrell et al. 2006).

1.4 Life Cycle Models Available Initial Screening

Literature and Internet searches were undertaken to identify models that could potentially
provide a life cycle analysis (LCA) for bioethanol. An extensive list of potential models was
identified, and 37 models were screened to determine which offered potential to evaluate
bioethanol. Proprietary models were not purchased, and models that required customization to
incorporate bioethanol or other transportation fuel capabilities were not favoured. The models
were initially reviewed to select those that offered the most potential to conduct a comparison
and sensitivity analysis under the context and main purpose of this study. Specifically, this
included selecting models that:
• were developed to address the biofuels life cycle (as opposed to addressing another life
    cycle, such as building materials);
• are expected to have the built-in capabilities to assess the biofuels life cycle (and do not
    require the user to assemble processes and data); and
• continue to be available to allow the consultants to conduct a sensitivity analysis under the
    context of this study.

The results of the screening process suggest that the following were the most capable of
conducting life cycle assessments of bioethanol under the context of this study:

   BEES                                              BESS
   EIO-LCA                                           GaBi
   GEMIS                                             GHGenius
   GREET                                             LEM
   SimaPro




                                                 3
CHEMINFO




An evaluation of the nine models identified as potentially offering the best options for assessing
the bioethanol life cycle are each described in this report, and evaluated on the basis of ten
evaluation criteria. The models were evaluated based on a review of materials available over the
Internet, and telephone consultations with individuals familiar with the model. In some cases
uncertainty exists over aspects of some models, for example their sophistication in terms of input
flexibility. In these cases, a scoring was assigned based on expectations. Any uncertainty is not
expected to significantly affect the selection, the scoring or the order that the relevant models are
ranked in terms of attractiveness.

The results of the evaluation found that the US-based GREET and Canadian-based GHGenius
models were favoured for this study. GREET and GHGenius have advantages and disadvantages
relative to other models and to each other. These models are tailored to specific transportation
fuel pathways but have less flexibility (than some other models) for creating new process chains
(unless the user is very experienced). The advantage of these models is that less skilled users can
utilize them to change the primary inputs for the pre-determined process chains and determine
the impact of those changes. The models are also focused primarily on energy balances, GHG
emissions, and air pollutant (i.e., criteria air contaminant-CAC) emissions and thus have a
narrower range of end-point outputs than some commercial models such as Gabi and SimaPro,
which require purchase.


1.5 Comparison of GREET and GHGenius Results

Detailed analysis and assessment of GREET and GHGenius model results, input databases,
parameters, as well as the model algorithms was carried out. Results from default baseline cases
of the models were first developed and compared, and then reasons for differences in the models
were investigated. This was followed by the analyses of the grain-based ethanol production
pathways in the models. GREET version 1.8 and GHGenius version 3.12, have been used for the
analysis of corn ethanol. In both cases, the models have been set to analyze the year 2007. A full
comparison between the models was not possible because of the limited pathways and
differences in geographic coverage. However, enough of a comparison was possible to get a
sense of the similarities and differences in the models.

GHGenius and GREET have a number of similarities in their basic structures and algorithms.
Both are energy-based models and both undertake a carbon balance approach to carbon dioxide
emissions. Criteria air contaminants (CAC) emissions for both models are based in part on US
EPA AP-42 emission factors, but in both cases there are modifiers applied to account for
emissions in a regulated environment.




                                                4
CHEMINFO




1.5.1 Comparison of Reference Gasoline Results

The lifecycle GHG emissions for the reference gasoline fuel are quite similar in the two models
but there are significant differences between the stages. GREET energy use data for crude oil
production are somewhat old and may not be reflective of the current situation in the United
States. It would be valuable to update this information to better align it with GHGenius, for
model comparison purposes. The default values are low and may somewhat underestimate the
energy requirements for crude oil production. The emissions for the refining portion are similar
in the two models. There is a difference in the emissions from fuel combustion that is driven by a
lower carbon monoxide emission rate in GREET compared to GHGenius. This results in higher
carbon dioxide emissions since both models use a carbon balance approach to calculating
emissions.

There are some differences in the individual air emissions for oil production and refining
between the two models. The GREET and GHGenius values have changed over time as more
current data reflecting changes in technologies, energy and emissions occur. More
documentation with details would be valuable to users of both models to better understand the
base estimates and changes. GHGenius utilizes different emission factors for Canada and the US
for oil refining and the results are slightly different as expected. Both models use US EPA AP-42
emission data in the calculations. AP 42 emission factors are used to calculate the uncontrolled
emissions in GHGenius. The GHGenius results have been compared to NPRI data in Canada and
NEI data in the United States so they should reflect air pollutant (i.e., criteria air contaminant) oil
refining emission rates. AP 42 emissions factors are also used in GREET, however, the
calibration process to arrive at the final controlled emission rate levels contained model is not
transparent.

    Table 1: Comparison of Lifecycle Gasoline Results with Different
                                Models

Parameter                     Units               GHGenius           GHGenius            GREET
                                                   Canada              US                 US
                       Joule consumed/joule
Energy                            produced                0.287              0.275              0.226
GHG                             g CO2eq/GJ               86,314             86,745             85,763
CO2                                   g/km                303.5              307.9              303.3
CH4                                   g/km                0.479              0.555              0.361
N2O                                   g/km                0.016              0.016              0.011
PM10                                  g/km                0.046              0.075              0.058
NOx                                   g/km                0.593              0.621              0.320
SOx                                   g/km                0.305              0.168              0.098
VOC                                   g/km                0.408              0.408              0.216
CO                                    g/km               10.957             10.976              2.929


                                                 5
CHEMINFO




1.5.2 Comparison of Corn Ethanol Findings

While the lifecycle GHG emissions for corn ethanol in GREET and GHGenius are generally
similar when the same region is modelled but there are differences in how those results are
arrived at. The basic processes that are modelled are similar in terms of process inputs but there
are differences in the approach, namely:

 •   The use of coal in some US ethanol plants has a significant impact on the GHG emissions.
 •   Energy use in ethanol plants is the primary driver of GHG emissions and it has been
     improving at a significant pace over the past 25 years, which makes up to date data on
     energy use very important for this pathway.
 •   GREET does not include the N2O emissions from the biological fixation of nitrogen
     whereas GHGenius does. The result is that the GHG emissions for producing soybeans are
     much higher in GHGenius than they are in GREET and this impacts the co-product credit
     for ethanol.
 •   The corn farming emissions in GREET are higher than in GHGenius due to the use of lime
     to adjust the acidity of soils in the US.
 •   There are some differences in land use emissions between the two models but a large
     portion of the difference is driven by the use of lime. The land use calculations in
     GHGenius are more complex than they are in GREET and have more capacity for
     sensitivity analysis.

  Table 2: Comparison of Lifecycle E10 Results with Different Models

Parameter                 Units               GHGenius          GHGenius             GREET
                                               Canada             US                  US
Energy (E100)      Joule consumed/joule             0.685            0.885               0.767
                               produced
GHG (E100)                  g CO2eq/GJ               39,454             56,990            64,508
CO2                                g/km               287.6              295.5             297.0
CH4                                g/km               0.456              0.538             0.357
N2O                                g/km               0.021              0.021             0.017
PM10                               g/km               0.048              0.079             0.066
NOx                                g/km               0.614              0.650             0.336
SOx                                g/km               0.295              0.177             0.110
VOC                                g/km               0.395              0.396             0.219
CO                                 g/km              10.635             10.655             2.764




                                               6
CHEMINFO




1.5.3 Comparison of Corn and Wheat Ethanol Results Using GHGenius

Wheat is not a significant biofuel crop in the United States and this is likely the reason that there
is no corresponding wheat ethanol pathway in GREET. Therefore a comparison and discussion
of the differences between corn and wheat ethanol was carried out using GHGenius results.

Both corn and wheat ethanol fuels offer advantages in terms of lower GHG emissions compared
to gasoline. The base case results indicate that corn ethanol GHG emissions are slightly lower
than wheat ethanol emissions. This is partly a function of the lower energy requirements for corn
ethanol. CAC emissions for ethanol can be higher or lower than the emissions from crude oil
derived gasoline depending on the contaminant. The lifecycle CAC emissions are quite similar
for the two ethanol sources.

     Table 3: Comparison of Corn and Wheat Ethanol with Gasoline

Parameter                   Units              Gasoline          Corn Ethanol Wheat Ethanol
                                               Crude Oil             E10          E10
                    Joule consumed/joule             0.287               0.685       0.7517
Energy (E100)                   produced
GHG (E100)                   g CO2eq/GJ               86,314              39,454             40,429
CO2                                 g/km               303.5               287.6              288.8
CH4                                 g/km               0.479               0.456              0.443
N2O                                 g/km               0.016               0.021              0.019
PM10                                g/km               0.046               0.048              0.049
NOx                                 g/km               0.593               0.614              0.587
SOx                                 g/km               0.305               0.295              0.299
VOC                                 g/km               0.408               0.395              0.395
CO                                  g/km              10.957              10.635             10.617



1.6 Sensitivity Analysis

The influence on the life cycle results resulting from changes in some of the key inventory data,
emission factors, allocation issues, future developments, and process changes were investigated.
The GHGenius model was selected for this work because it has both corn and wheat ethanol
pathways specific to Canada and it has features such as the sensitivity solver and the Monte
Carlo tool that allow for rapid investigation of some of these issues. A number of themes for
investigation were selected based on results from the previous section and input from
Environment Canada.




                                                7
CHEMINFO




                    Table 4: Summary of Sensitivity Analysis

Theme                                    Conclusions                 Comments / Examples
Process Conversion Efficiencies    A relatively small impact     Reducing conversion efficiency
                                    on the lifecycle results     from 100 to 95% reduces GHG
                                                                  emissions by about 500 g/GJ
Process Energy Requirements        A moderate impact on the       - 50% change in input values
                                       lifecycle results        results in - 30% impact on GHG
                                                                             emissions
Emission Factors                      N2O emissions from        Some offset in co-product credits
                                   nitrogen fixing crops is a    reduces the impact on lifecycle
                                    large uncertainty in the       emissions. Larger impact on
                                            process                      wheat than corn
Land Use                              Can be a very large         Probably less of an issue with
                                      component in some          Canadian produced feedstocks
                                           situations
Feedstock Production                Tillage practices have a    Indirect impacts on yield and soil
                                      small direct impact        carbon are probably larger than
                                                                           fuel savings
Co-Product Allocation               Can have a large impact      Significant differences between
                                          on results              approaches used in different
                                                                              models
Water, Other Environmental           North American LCA
Issues                                models focus on air
                                          emissions



1.7 Recommendations for Enhancing LCA Capacity in Canada

Life cycle analysis can be a valuable and power tool for informing policy-makers, stakeholders
and the public regarding the environmental emissions and potential impacts related to alternative
transportation fuels as well as other products. This report has identified many models that are
available for potential further development, as well as two models (GREET and GHGenius) that
had sufficient data and capabilities to support analysis for this study. One of the models –
GHGenius – has been developed in Canada and contains a robust set of data to support LCA
analysis for transportation fuels in Canada. The analysis identified various input data, boundary,
energy/emissions allocation and other issues that can arise with models and can lead to
differences in results and interpretation. Furthermore, Canada's LCA modelling capacity needs to
continue to improve to keep abreast of new data, additional pathways that could be analyzed, and



                                               8
CHEMINFO




new technologies that influence industrial performance, continual demands for up-to-date
documentation and increased demands for quality.

The following recommendations for enhancing Canada's LCA modelling capacity are provided
to Environment Canada for their consideration:

 •   Continue to work in partnership with government departments and stakeholders;
 •   Enhance GHGenius model for Canadian conditions.
 •   Continue to support enhancements in transparency and documentation;
 •   Remain engaged in understanding international LCA models and results;
 •   Develop long term vision and path forward for Canada's LCA capacity;
 •   Consider support for exporting Canadian LCA capacity;

Each of these suggestions are explained in Section 9 of the report.




                                               9
CHEMINFO




                                        2. Introduction
2.1 Background

Ethanol presents an alternative transportation fuel to gasoline. 4 Bioethanol-gasoline fuel blends
can offer lower air pollutant emissions for some contaminants versus gasoline use only.
Bioethanol, which is made from the fermentation of sugars contained in plants such as corn and
wheat, presents an option for Canadians to reduce greenhouse gas emissions from transportation
fuel use. The reason is that the carbon contained in the bioethanol is derived from plants that
have sequestered carbon dioxide from the atmosphere to make the sugars. Therefore, the carbon
dioxide emissions associated with bioethanol transportation fuel combustion are considered to be
neutral. There can also be economic benefits related to increased bioethanol production in
Canada. Bioethanol plants present new markets for farmers' crops, create construction and plant
operating jobs, and help diversify and strengthen rural economies. Furthermore, increased
domestic use of bioethanol can reduce Eastern and Central Canada's dependence on imported oil
used to make gasoline.

As a result of the potential environmental, economic and social benefits associated with greater
bioethanol production and use, the Government of Canada is committed to develop a national
renewable fuels strategy by targeting a five per cent (5%) renewable content in Canadian motor
fuels. The Federal Minister of the Environment has announced that the Government would
regulate an annual average renewable content of 5% in gasoline by 2010. In addition, the
Government intends to regulate a two per cent (2%) requirement for renewable content in diesel
fuel and heating oil by 2012. Federal and provincial government financial incentive programs
have been announced to help create new market opportunities for Canada's agricultural
producers. 5 , 6 Furthermore, bioethanol is being supported with Federal and Provincial tax
exemptions or tax credits ranging from 9 to 20¢ per litre.

There is an anticipated rapid increase in bioethanol production in Canada. With annual gasoline
consumption in Canada at approximately 40,000 million litres per year 7 , there would be a
requirement for 2,000 million litres per year of bioethanol to meet the 5% content requirement.

4
  In this report, "ethanol" and "bioethanol" are used interchangeably. Use of the word "ethanol" refers to bioethanol
unless otherwise specified. Ethanol can be commercially made through chemical synthesis (e.g., catalytic hydration
of ethylene) and via fermentation of sugars by yeast. There is no longer any commercial production of synthetic
ethanol from ethylene in Canada.
5
  Government of Canada, Canada's New Government Takes New Step to Protect the Environment With Biofuels,
December 20, 2006
http://www.agr.gc.ca/cb/index_e.php?s1=n&s2=2006&page=n61220
6
  Ontario Ministry of Food, Agriculture and Rural Affairs, Ontario Ethanol Growth Fund: Program Overview,
June, 2005
7
    Statistics Canada, Sales of fuel used for road motor vehicles, by province and territory,
http://www40.statcan.ca/l01/cst01/trade37a.htm?sdi=gasoline%20litres


                                                        10
CHEMINFO




Current installed capacity for bioethanol is approximately 950 million litres/year, 8 and an
additional 675 million litres per year of capacity is under construction for anticipated completion
in 2008. Half a dozen companies have announced their intention to install an additional
cumulative 1,500-1,600 million litres per year of bioethanol capacity.


                             Table 5: Canadian Ethanol Capacity9

    Company                                                     Location         Province           Capacity
                                                                                                (million litres/year)
                                  Existing Ethanol Plants
    Suncor Energy Inc.                       Sarnia           ON                                          210
    Greenfield Ethanol                       Chatham          ON                                         187 10
    Husky Energy Inc.                        Lloydminster     SK                                          130
    Husky Energy Inc.                        Minnedosa        MB                                          130
    Greenfield Ethanol                       Varennes         QC                                          120
    Blackstone Energy                        Collingwood      ON                                           52
    Permolex Ltd.                            Red Deer         AB                                           40
    NorAmera Bioenergy Corp.                 Weyburn          SK                                           25
    Greenfield Ethanol Inc.                  Tiverton         ON                                           23
    Tembec Inc.                              Temiscaming      QC                                          17 11
    Pound-Maker Agventures Ltd.              Lanigan          SK                                           12
    Iogen Corporation                        Ottawa           ON                                            4
    Total installed capacity                                                                              950
                             Ethanol Plants Under Construction*
    Greenfield Ethanol                       Johnstown        ON                                         200
    Greenfield Ethanol                       Hensall          ON                                         150
    Terra Grain Fuels Inc.                   Belle Plaine     SK                                         150
    IGPC Ethanol Inc.                        Aylmer           ON                                         150
    Northwest Terminals                      Unity            SK                                          25
    Total capacity under construction                                                                    675
Sources: S&T2 Consultants, Cheminfo Services Inc, Camford Information Services. Company representatives and
websites. * Note: All plants are scheduled for completion by 2009.



The rapid growth of Canadian bioethanol production and use has resulted in questions regarding
the magnitude of the overall environmental benefits as well as drawbacks of bioethanol. Like all
transportation fuels, the magnitude of the environmental attractiveness of bioethanol depends on
8
  A minor portion of this capacity is dedicated to non-fuel uses of ethanol (e.g., solvent use, etc.).
9
  Current as of March 1, 2007.
10
   About one third of capacity is for non-fuel applications.
11
   All non-fuel use. Product is not anhydrous.


                                                           11
CHEMINFO




its total life cycle footprint relative to the footprint(s) associated with alternative fuels, notably
gasoline. These footprints are complex to quantify since they involve both main products (i.e.
ethanol) and co-products that can be influenced by many factors, which are often regionally
specific. Environment Canada is seeking to address questions raised. At the same time, there is a
need to improve Environment Canada's capacity to identify, predict and assess the full range of
environmental, regulatory, technical, and scientific issues related to the anticipated accelerated
increase in production and use of transportation fuels derived from biomass in Canada.

Environmental life cycle analysis (LCA) models are tools that can be applied to help assess the
relative attractiveness of transportation fuels and other products. LCA can be used to guide
government policy makers, industry, community stakeholders and other groups in making more
informed decisions for the Canadian and regional environment by understanding how new fuels
compare to petroleum-derived gasoline, what the significant environmental issues are and in
what stage of the lifecycle they are released. There are a variety of LCA models available and
results from these vary, mostly for valid reasons. This can create confusion and reduce
confidence in the value of LCA results for bioethanol. In addition, the quality of conclusions
drawn from LCA models are influenced by the assumptions and data used in generating model
results.

As a result, Environment Canada needed to develop a better understanding of LCA models
available as well as the underlying data, assumptions and associated calculations these tools use
to generate results.


2.2 Purpose of This Report

The main overall purpose of this report is to provide Environment Canada with an assessment of
existing LCA models that can be applied in determining the environmental footprint of
bioethanol and competing fuels (e.g., gasoline) in Canada. The assessment includes analysis to
identify the key factors that contribute to differences in the results from different LCA models
applied to starch-based biofuel production. This report also provides:

 •   an analysis regarding the role of LCA in policy formulation;
 •   an overview of other modelling activities oriented to environmental policy development;
 •   an overview of what LCA is and how it works;
 •   a brief description, assessment and availability (for the purposes and scope of this study) of
     37 LCA models, more detailed analysis of 9 models and selection of 2 models for detailed
     analysis and comparison;
 •   sensitivity analysis using GHGenius to identify the factors in the corn and wheat derived
     ethanol life cycles that most strongly influence LCA results; and
 •   recommendations for further analysis, development and application of LCA models for
     Canada.



                                                 12
CHEMINFO




2.3 Approach and Methodology

The first step for this study was to identify, collect and review life cycle assessment models that
could encompass the environmental elements for biofuels. This step involved reviewing major
U.S. and European websites providing lists of environmental assessment tools. In addition, a
literature sources were also researched and reviewed to identify life cycle assessment tools that
were applied. Approximately 37 life cycle models were identified and reviewed. However, the
availability of the models and information on the models varied substantially. Some models are
fully available in the public domain free of charge, while others are only available through
purchase from the model developers. No models or database were purchased for review in this
study.

The 37 models identified in the first step were screened to select those suitable for this study
and/or that may be useful for further development. This screening process focused on identifying
models that were developed to address biofuel life cycles. Models that were preferred in the
selection process were those that had built-in capabilities to assess the biofuels life cycle (and
did not require the user to “build” models using processes), and continue to be available. Nine
models were identified as capable following this screening stage.

The capable models were reviewed in greater detail with the purpose of selecting a subset to
conduct model results, assumptions and data input comparison, as well as sensitivity analysis.
The sensitivity analysis had the aim of assessing which factors strongly influence results. The
models were evaluated on the basis of ten evaluation criteria. These criteria considered a wide
range of factors such as model cost, extent of biofuels pathways, life cycle stages, environmental
endpoints, transparency, data, and other features. GREET version 1.8 and GHGenius version
3.12 were the two models selected based on the evaluation criteria.

Baseline cases were established and results were generated from the two models. In both cases
the models were set to analyze the year 2007. Input databases, other parameters, as well as the
model algorithms for the models were analyzed to explain differences in the results. An analyses
of the pathways contained in the models was also conducted.

Sensitivity analysis was conducted with the GHGenius model to assess the impacts on the life
cycle results of uncertainty or changes in some of the key input data and assumptions. Process
conversion efficiencies, process energy requirements, emission factors, feedstock production,
other feedstocks, co-product allocation and other issues were examined in the sensitivity
analysis.




                                               13
CHEMINFO




2.4 Intended Use of Report

This report should be useful to Environment Canada and other experts involved with or have an
interest in the development of environmental policies related to bioethanol and other biofuels in
Canada. This would include Natural Resources Canada, Agriculture Canada, National Research
Council, provincial environment departments, biofuel industry associations (e.g., the Canadian
Renewable Fuels Association), industry participants (e.g., biofuel producers), as well as other
participants involved with various life cycle stages for biofuels and competing products (e.g.,
petroleum refining). The report provides analysis that should be valuable in expanding and/or
making improvements to Canada biofuels LCA capacity. LCA practitioners and those relying on
LCA results should find the report useful when considering how to conduct LCA analysis and
how to apply the results of LCA analysis to environmental decision making.

This report does not have the intent of making conclusions regarding the environmental
attractiveness of biofuels versus other transportation fuels. While the results presented in this
report may indicate reduced GHG and air emissions for certain biofuels the full environmental
impacts are not solely determined by releases to the environment. Local air and water quality,
biodiversity issues, health effects, and other considerations are involved in determining the best
fuel option.

Neither does this report have the intent of making definitive conclusions regarding which
existing LCA tools yield the best results. Several LCA tools were not available to the consultant
to analyze in detail. Also, the quality of underlying assumptions and data should ideally be
validated using facility-specific boundaries. LCA model results from even slightly different
boundaries or time periods may not reflect environmental performance for actual specific
regions, facilities, and exact time periods. It is costly and a challenge to maintain data in models
current. The analysis does identify the key factors that can lead to differences in results achieved
by different LCA models.


2.5 Rest of Report

This report includes 7 additional sections, References and an Appendix. Section 3: The Role of
LCA in Policy Development discusses how LCA is used in formulating environmental policies
and regulations, challenges and limits to LCA analysis, and provides a description of some of the
energy, emissions and economics modeling that is being carried out in Canada for application to
climate change and other environmental policy initiatives. Section 4: Life Cycle Analyses
provides a description of LCA modelling; ISO standards considerations; and many other
elements involved with proper life cycle analyses. Section 4 sets up Section 5: Life Cycle
Models, Initial Screening, which briefly analyses 37 LCA models to identify the most useful for
the purpose of this study and/or for potential further development in Canada. Section 6: Detailed
Assessment of LCA Models describes and assesses 9 models (selected from the 37) to identify the


                                                14
CHEMINFO




two best models that can be applied to conduct analysis to support this study. Section 7:
Comparison of Model Results compares the assumptions, data and results of the 2 selected
models, and serves to explain why LCA results from different models can differ. Section 8:
Sensitivity Analysis uses one of the models to conduct analysis of the impact on the life cycle
results related to uncertainty in some of the data and assumptions. Section 9: Recommendations
and Further Research provides suggestions for Environment Canada to consider in evolving and
improving Canada's LCA model capacity, as well as further research that may be useful to better
understand and interpret LCA data input, assumptions and interpret results. References, websites
researched are documented, and an Appendix summarizes some of the literature sources
reviewed.




                                              15
CHEMINFO




 3. The Role of Modelling in Policy Development

3.1 Overview of Current Uses of Life Cycle Assessment

To date LCA has been applied in evaluating the relative environmental performance of
alternative biofuel options with the primary aim of informing industry, government,
Environmental Non-governmental Organization (ENGO) and consumer decision-making.
Studies have been completed by LCA practitioners in consulting firms, academia, ENGOs,
industry, and government. The quality of the studies has varied but over the last decade, on
average, study quality has improved due to method development, data availability and higher
client expectations.

A few examples of uses of biofuels’ LCAs by various decision makers include the following.

   •   Industry: Through an examination of the results of a LCA of their biofuel production
       process, a producer may determine where in the process or supply chain an improvement
       could be made so as to lower their resource use or environmental discharges. The saying,
       “what is measured can be managed” is key. Quantifying the resource use/environmental
       discharges associated with the full life cycle of a biofuel allows industry to move forward
       toward managing these impacts.
   •   Government: As will be discussed in more detail below, LCAs of biofuels have been
       utilized for determining preferred biofuel pathways (feedstock/fuel production) for
       receiving government funding under biofuels’ expansion programs.
   •   ENGOs: These organizations have utilized LCAs of biofuels to support their positions in
       calling for increased attention to broad sustainability issues in expansion of biofuel
       production.
   •   Consumers: Results of biofuels’ LCAs have been presented by various organizations and
       utilized indirectly in advertising campaigns with the hope of influencing consumer choice
       with respect to fuel and vehicle options (e.g., purchase of a flexible fuel vehicle so as to
       have the potential to utilize a high level ethanol/gasoline (E85) blend).


3.1.1 Role of LCA in Public Policies/Regulations

Life cycle assessment’s role in public policy development to date has been focused on informing
public policy positions of industry (e.g., General Motors’ decision to support ethanol) and
government. In a limited set of cases, LCA has had a more direct role. For example, under the
US Renewable Fuel Standard resulting from the Energy Policy Act of 2005, some renewable
fuels (e.g., those from selected lignocellulosic feedstocks) that were expected to have lower life
cycle environmental impacts through a weighting system that “rewarded” such pathways. This


                                               16
CHEMINFO




and other similar programs, however, have not required detailed LCA. Generally, although LCA
has informed public policy positions it has not been the basis of public policies, in particular,
those that have binding targets directly related to the application of the LCA method.

This appears about to be changing. Over the past few years there have been several
announcements related to incorporating life cycle-based standards directly into climate change
regulations for transportation fuels. These regulatory initiatives include those covering all
transportation fuels in a particular jurisdiction, as well as the more numerous initiatives, which
are focused on biofuels. One of the most prominent initiatives is California’s Low Carbon Fuel
Standard (LCFS), which will consider all light-duty transportation fuels sold into State (State of
CA 2007). The United Kingdom’s Renewable Transportation Fuel Obligation Programme
(RTFO), the German Biofuels Ordinance, the European Union Fuels Directive, and the U.S.
Energy Independence and Security Act of 2007 all focus on biofuels. In Canada and the U.S.,
other federal, state and provincial governments have declared interest in adopting similar low
carbon fuel standards (e.g., British Columbia, Ontario, Minnesota, Massachusetts). The
programs are currently under development but they will require that the life cycle GHG
emissions associated with the production of relevant biofuels (and in some cases, other fuels) be
quantified. They will be the first regulations that will be based on systematic LCA.
The California LCFS and the UK RTFO, two of the more prominent initiatives, are described
briefly. On January 18, 2007, the State of California, through Executive Order S-1-07,
announced the intent to regulate a reduction of least 10% by 2020 in the life cycle carbon
intensity of transportation fuels sold in the State (State of CA 2007). Enforcement of the
standard will begin in 2010 while it will be fully in effect in 2020. It will complement other
policies related to vehicle and transportation system improvements. Under the LCFS fuel
providers (e.g., refineries, blenders, and importers) will be required to ensure that the mix of
fuels they sell into the California market meets, on average, a declining carbon intensity which is
expected to be based on estimates of carbon dioxide equivalent per energy unit of fuel on a life
cycle basis, adjusted for vehicle efficiency (Farrell and Sperling 2007). As noted above, the
California regulation applies to all fuels sold into the market, not just biofuels. This is in contrast
to the UK RTFO, which is focused exclusively on biofuels (UK DOT 2006). The RTFO will,
from April 2008, place an obligation on fuel suppliers to ensure that a certain percentage of their
aggregate sales are made up of biofuels. The effect of this will be to require 5% of all UK fuel to
come from a renewable source by 2010. The RTFO, like the LCFS, has reporting requirements
and methodologies for calculating life cycle GHG emissions but as well includes social and
environmental sustainability aspects, although these latter criteria will not be used in the issuing
of compliance certificates until the feasibility, accuracy, and efficiency of the reporting structure
are determined (UK DOT 2006).

The application of lifecycle analyses in a regulatory framework is not without its challenges. The
work underway in the EU, the UK, Germany, and the Netherlands to develop LCA criteria for
regulatory purposes differs significantly from more scientific LCA work. Good regulations are
generally simple regulations, so regulatory LCA work has been moving away from some of the
principles of ISO LCA work in order to simplify the process. These systems are adopting default


                                                 17
CHEMINFO




values that are deliberately conservative for many of the data inputs so that biofuel producers do
not have to invest in tracking and documenting the inputs through the lifecycle if they choose not
to. In other cases, some of these systems are developing co-product allocation systems that are
simple and are designed to try and ensure that there are no opportunities for making poor
decisions that would provide good results for one particular indicator (GHG emissions) at the
expense of another indicator (for example, land use). While they serve regulators’ needs, these
allocation schemes are not necessarily considered to be the most sound from an ISO perspective.

These simplified regulatory LCA frameworks, while providing the advantage of being simple
and possibly less expensive to utilize, will not produce the same results as well done, more
scientific LCAs. This will undoubtedly create some confusion for all stakeholders but more
importantly may result in missed opportunities to implement some attractive environmental
solutions.

A life cycle basis is important for informing environmental regulation because there can be very
different and significant impacts in various parts of the supply chain associated with biofuel
production. However, whether these regulations can achieve their intended objectives will
depend upon development and application of a robust LCA framework for biofuels and
successful implementation of the policy.

3.1.2 LCA Challenges for Biofuels

Numerous LCAs for bioethanol and other biofuels have been published (reviews include
Fleming et al. 2006 and Larson 2006). Most studies have followed ISO standards (ISO 2006) but
a wide range of results has often been reported for the same fuel pathway, sometimes even when
holding temporal and spatial considerations constant. The ranges in results may, in some cases,
be attributed to actual differences in the systems being modeled but are also due to differences in
method interpretation, assumptions and data issues.

Key issues in biofuels’ LCAs have been differing boundaries being adopted in studies (i.e., what
activities are included/excluded from the study), differences in data being collected and utilized,
and disparities in the treatment of co-products. In addition, LCAs, more generally (not solely
limited to those of biofuels) have often included limited or no analysis of uncertainty and
validation of model results. Boundaries in prior LCAs have often differed due to resource
constraints. Data requirements in LCA are significant. Studies have not always used up to date
data or data that reflect the inputs in the relevant process under study (i.e., utilization of
electricity generation data for another jurisdiction rather than the one under study). There are
also gaps in scientific knowledge surrounding key variables. For example, these include
implications of land use change, N2O emissions related to feedstock production, and nutrient
depletion and erosion due to agricultural residue removal. Utilization of different co-product
methods, and in some studies, ignoring co-products entirely, has had major impact on results of
LCA studies (Kim and Dale 2002, Larson 2006, Farrell et al. 2006).




                                               18
CHEMINFO




As explained further in Section 4, an LCA has several stages, the majority of LCA studies of
biofuels to date have focused on the life cycle emissions inventory analysis stage of LCA, with
fewer studies being carried through to environmental and health impact assessment. While this
may be considered a limitation, the conversion of inventory results to impact indicators, while
providing more information, as well introduces more uncertainty into the results. Understanding
the actual impacts on ecosystems and public health of biofuels (or any other product) is
challenging and a life cycle impact assessment provides only limited insights. Challenges to
understanding the actual impacts of biofuels include that LCAs most often do not contain
detailed site specific (temporal and spatial) emissions and exposure data needed for detailed risk
assessments. Detailed risk assessment is beyond the scope of what LCA can be practically
expected to accomplish (Matthews et al. 2002).

Life cycle assessment is a useful tool for comparing on a functional unit basis, the relative
environmental performance (based on a specific set of metrics) of different feedstock/fuel
pathways. However, LCA should be utilized along with other information in decision making
regarding biofuels. Decision-makers should be aware of both the strengths and limitations of
LCA. In order to more completely understand the implications on the environment (and
economy) of biofuel production (e.g., scale of production issues, impacts on ecosystem and
human health) LCA results should be augmented with those of other modeling systems
(potentially some of those described in Section 3.2), or perhaps, integrated modeling systems
could be developed in the future as well as decision makers’ good judgment.


3.2 Government of Canada Energy, Emissions and Economics
Modelling Activities

There are a number of models available to the Government of Canada and under development by
the Government of Canada to assist in policy development. Two modelling frameworks are
briefly described to provide a better understanding of the position of LCA models as well as the
consideration of future potential links between the models that may be considered in the path
forward. It is also useful to be aware of the basic energy and emission data sources that are used
to populate these models, since some of these data sources are also used in life cycle models and
analysis in Canada. Examples of common data sources are:

 •   Statistics Canada: Report on Energy Supply and Demand in Canada. Provides energy
     supply and demand by fuel for Canada, provinces/territories and major end-use sectors;
 •   Environment Canada: Canada’s Greenhouse Gas Inventory. Provides GHG emission
     factors and total GHG emissions in Canada by provinces/territories and major emitting
     sectors (Produced annually); and




                                               19
CHEMINFO




 •   Environment Canada: Canada's Criteria Air Contaminants Emissions Inventory. Provides
     total CAC emissions in Canada by province/territory and major emitting sectors (Produced
     annually). 12

3.2.1 Natural Resources Canada's MAPLE-C Model 13

Natural Resources Canada has been developing the Model to Analyze Policies Linked to Energy
in Canada (MAPLE-C) model. This model was applied in developing the NRCan's projections
for Canada’s Energy Outlook: The Reference Case 2006 are generated from the To develop
MAPLE-C, NRCan has spent the past three years modifying the US National Energy Modeling
System (NEMS) to reflect the Canadian economy and its provincial components. MAPLE-C is
maintained by the Analysis and Modelling Division of Natural Resources Canada.

MAPLE-C can be used to assess the energy and economic implications of new policy proposals.
In general, the historical data used for the projections are based on Statistics Canada’s annual
Report on Energy Supply and Demand in Canada. However, data are also taken from other
sources, such as the Environment Canada report, Canada’s Greenhouse Gas Inventory, to
estimate carbon dioxide emissions coefficients. The model has been developing its capabilities to
handle criteria air contaminants. For example, it recently developed capabilities to model NOx
and SO2 emissions in the electricity sector. 14

The projections in MAPLE-C are developed with the use of a market-based approach to energy
analysis. For each fuel and consuming sector, MAPLE-C balances energy supply and demand,
accounting for economic competition among the various energy sources. The current time
horizon of MAPLE-C is until 2020. In order to represent the regional differences in energy
markets, most of the modules of MAPLE-C function at the provincial/territorial level: the ten
provinces and one aggregated region for the three territories. Energy markets are represented by
end-use demand, oil, gas and coal supply, and electricity generation. Petroleum refining has
three regions: Western Canada, Ontario and Eastern Canada.

MAPLE-C is organized and implemented as a modular system. The modules represent each of
the energy supply markets, conversion sectors and end-use consuming sectors of the energy
system. MAPLE-C also includes a macroeconomic module. The primary flows of information
between each of these modules are the delivered prices of energy to the end-user and the
quantities consumed by product, region and sector. The delivered prices of energy encompass all
the activities necessary to produce, import and transport fuels to the end-user. The information
flows also include other data, such as economic activity, domestic production and exports. The
schematic of the model is shown below.
12
   Environment Canada, CAC Emissions Summaries, 2006. http://www.ec.gc.ca/pdb/cac/
13
   Natural Resources Canada, Analysis and Modelling Division, Canada's Energy Outlook: The Reference Case
2006, October 2006.
14
   SAIC, Cheminfo Services Inc., (2007), Operationalizing MAPLE-C With NOx and SO2 Emissions Capabilities
for Canada’s Electric Power Generation Sector, For Natural Resources Canada.


                                                  20
CHEMINFO




                                      Figure 1: Maple C Schematic




The model has relatively detailed modules for various life cycle stages for petroleum refining,
electricity, and transportation. However, detailed representations for biofuels production or use
do not exist in the model.

3.2.2 Environment Canada's E3MC Modelling Framework

Environment Canada is developing the E3MC model that will be used in assessing the impacts of
proposed emissions reduction targets for greenhouse gases and criteria air contaminants. The
model fully integrates energy, emissions, and economic factors. E3MC is being designed to
explicitly model energy, emissions and production for the major industrial sectors that are the
focus of recent Federal Regulatory Framework for Air Emissions 15 and the associated Clean Air
Regulatory Agenda (CARA). The major industrial sectors are: electricity produced by fossil
fuels, oil and gas, base metal smelters, iron and steel, some mining sectors, cement, pulp and
paper, aluminum, and chemicals production. Other sectors of the economy such as residential,
commercial, transportation and non-regulated industrial sectors are also encompassed by the
mode.

E3MC integrates other models into its analytical framework. For example, it uses Informetrica’s
TIM macroeconomic modelling capabilities and Environment Canada's Energy 2020 (E2020)
model for energy supply and demand analysis. E3MC's modelling framework can also be linked

15
     Government of Canada, Framework for Air Emissions, 2007


                                                      21
CHEMINFO




(external to the model) with air quality models and related ecosystem and human health impact
models. In this way the results of the model can be used to develop monetary costs and benefits
estimates of these impacts.

              Figure 2: Environment Canada's E3MC Modelling Framework
                                   Source: Environment Canada




                                              22
CHEMINFO




 Figure 3: Environment Canada's 3EMC Analytical Process for Estimating the Impacts of
                    Emission Reduction Targets For Air Pollutants
                                      Source: Environment Canada




Some additional features of E3MC are:

 •   uses Natural Resources Canada data as the starting point for energy demand. NRCan uses
     the MAPLE-C model for development of base case energy outlook (Canada's Energy
     Outlook, 2006) 16;
 •   input data and results can be specific to provinces and territories, and total Canada;
 •   conducts projections to the year 2075;
 •   explicitly covers some of the life cycle stages related to biofuels analysis at the total
     sectoral level and by fuel (electricity, agriculture, transportation, petroleum refining); and
 •   has explicit energy/fuel and GHG/CAC emissions for chemicals sector production (which
     would include biofuels production), but unlikely to have detailed bioethanol (or biodiesel)
     production and fuels consumption modelled.




16
  Natural Resources Canada, Analysis and Modelling Division, Canada's Energy Outlook: The Reference Case
2006, October 2006.


                                                  23
CHEMINFO




3.2.3 Other Models Used in Canada

There are other models available or under development in Canada. These are not LCA models,
but can use the same input databases as LCA models and may present useful future linkages with
LCA models for the purposes of biofuels policy development.

 •   Environment Canada's Energy 2020 model. Part of the E3MC framework. Analyses
     energy, GHG and CAC emissions, reduction technologies, costs and economics;
 •   Simon Fraser University: CIMS model. Simulation model that analyses energy, GHG and
     CAC emissions, reduction technologies, costs and economics;
 •   Environment Canada's Computable General Equilibrium (CGE) model. Under
     development. Being designed to incorporate energy, GHG and CAC emissions, and
     economic analysis, including trade;
 •   ICF's IPM model. Focused on electricity sector, analyses energy, GHG and CACs, mercury
     emissions, emission reduction technologies, and costs; and
 •   Haloa Inc: MARKAL model: Linear optimization model that analyses energy, GHG and
     CAC (for one province for some sectors) emissions, emission reduction technologies, costs
     and economics.

To calibrate these models to Canadian conditions, energy and production data are mostly taken
from Natural Resources Canada, Statistics Canada; and emission data from Environment
Canada. US EPA AP-42 emission factors may also used in these models.




                                             24
CHEMINFO




                            4. Life Cycle Analyses
4.1 Overview

As environmental awareness increases, governments, industries and businesses have started to
assess how their activities affect the environment. Society has become concerned about the
issues of natural resource depletion and environmental degradation. The environmental
performance of products and processes has become a key operational issue, which is why many
organizations are investigating ways to minimize their effects on the environment. Many have
found it advantageous to explore ways to improve their environmental performance, while
improving their efficiency, reducing costs and developing a “green marketing” advantage. One
such tool is called life cycle assessment (LCA). This concept considers the entire life cycle of a
product.

Life cycle assessment is a "cradle-to-grave" (or “well to wheels”) approach for assessing
industrial systems. "Cradle-to-grave" begins with the gathering of raw materials from the earth to
create the product and ends at the point when all materials are returned to the earth. LCA
evaluates all stages of a product's life from the perspective that they are interdependent, meaning
that one operation leads to the next. LCA enables the estimation of the cumulative environmental
impacts resulting from all stages in the product life cycle, often including impacts not considered
in more traditional analyses (e.g., raw material extraction, material transportation, ultimate
product disposal, etc.). By including the impacts throughout the product life cycle, LCA provides
a comprehensive view of the environmental aspects of the product or process and a more
accurate picture of the true environmental trade-offs in product selection.

Specifically, LCA is a technique to assess the environmental aspects and potential impacts
associated with a product, process, or service, by:

 •   Compiling an inventory of relevant energy and material inputs and environmental releases;
 •   Evaluating the potential environmental impacts associated with identified inputs and
     releases;
 •   Interpreting the results to help make more informed decisions.

The term "life cycle" refers to the major activities in the course of the product's life span from its
manufacture, use, maintenance, and final disposal; including the raw material acquisition
required to manufacture the product. The following figure illustrates the typical life cycle stages
that can be considered in an LCA and the quantified inputs and outputs.




                                                 25
CHEMINFO




                                  Figure 4: Life Cycle Stages




The LCA process is a systematic, iterative, phased approach and consists of four components:
goal definition and scoping, inventory analysis, impact assessment, and interpretation as
illustrated in the following figure:

 1. Goal Definition and Scoping - Define and describe the product, process or activity.
    Establish the context in which the assessment is to be made and identify the boundaries and
    environmental effects to be reviewed for the assessment.
 2. Inventory Analysis - Identify and quantify energy, water and materials usage and
    environmental releases (e.g., air emissions, solid waste disposal, wastewater discharge).
 3. Impact Assessment - Assess the human and ecological effects of energy, water, and material
    usage and the environmental releases identified in the inventory analysis.
 4. Interpretation - Evaluate the results of the inventory analysis and impact assessment to
    select the preferred product, process or service with a clear understanding of the uncertainty
    and the assumptions used to generate the results.




                                               26
CHEMINFO




                                  Figure 5: Phases of a LCA




4.2 ISO Life-Cycle Assessment Standards

The concept of life-cycle assessment emerged in the late 1980’s from competition among
manufacturers attempting to persuade users about the superiority of one product choice over
another. As more comparative studies were released with conflicting claims, it became evident
that different approaches were being taken related to the key elements in the LCA analysis:

•   boundary conditions (the “reach” or “extent” of the product system);
•   data sources (actual vs. modeled); and
•   definition of the functional unit.

In order to address these issues and to standardize LCA methodologies and streamline the
international marketplace, the International Standards Organization (ISO) has developed a series
of international LCA standards and technical reports under its ISO 14000 Environmental
Management series. In 1997-2000, ISO developed a set of four standards that established the
principles and framework for LCA (ISO 14040:1997) and the requirements for the different
phases of LCA (ISO 14041-14043). The main contribution of these ISO standards was the
establishment of the LCA framework that involves the four phases in an iterative process:


                                              27
CHEMINFO




     1.   Phase 1 - Goal and Scope Definition;
     2.   Phase 2 - Inventory Analysis;
     3.   Phase 3 - Impact Assessment; and
     4.   Phase 4 - Interpretation

By 2006, these LCA standards were consolidated and replaced by two current standards: one for
LCA principles (ISO 14040:2006); and one for LCA requirements and guidelines (ISO
14044:2006). Additionally, ISO has published guidance documents and technical reports (ISO
14047-14049) to help illustrate good practice in applying LCA concepts. The following table
summarizes the ISO standards and technical reports for Life-Cycle Assessment.

The ISO 14040:2006 standard describes the principles and framework for life cycle assessment
including: definition of the goal and scope of the LCA, the life cycle inventory analysis (LCI)
phase, the life cycle impact assessment (LCIA) phase, the life cycle interpretation phase,
reporting and critical review of the LCA, limitations of the LCA, the relationship between the
LCA phases, and conditions for use of value choices and optional elements. ISO 14040:2006
covers life cycle assessment (LCA) studies and life cycle inventory (LCI) studies. It does not
describe the LCA technique in detail, nor does it specify methodologies for the individual phases
of the LCA. The intended application of LCA or LCI results is considered during definition of
the goal and scope, but the application itself is outside the scope of this International Standard. 17




17
  International Standards Organization, ISO 14040:2006 - Environmental Management - Life cycle assessment -
Principles and framework, 2006; www.iso.org/iso/iso_catalogue (ISO 14040:2006).


                                                   28
CHEMINFO




               Table 6: ISO LCA Standards and Technical Reports

ISO Standard           Name*                  Description
Current Standards
ISO 14040:2006         Principles and         Provides the basic description and framework for LCA from which
                       Framework              the remaining LCA standards are based. This standard also defines
                                              the “comparative assertion” requirements, including critical review.
ISO 14044:2006         Requirements and       Provides guidance for:
                       Guidelines             • the establishment of goals, purpose, and scope of the LCA;
                                              • the process of quantifying and analyzing LCA inventories;
                                              • the assessment of life cycle impacts; and
                                              • the interpretation of LCA results.
ISO/TR 14047:2003      Technical Report       Provides illustrative examples on how to apply ISO 14042 (now
                                              ISO 14044) to life cycle impact assessment.
ISO/TS 14048:2002      LCA Data               Provides guidance on factors to consider when documenting LCA
                       Documentation          data.
                       Format
ISO/TR 14049:2000      Technical Report:      Provides illustrative examples on how to apply ISO 14041 (now
                                              ISO 14044) to goal and scope definition and inventory analysis.
Withdrawn Standards
ISO 14040:1997     Principles and           Establishes the basic description and framework for LCA from
                   Framework                which the remaining LCA standards are based.
                                            Withdrawn in 2006 and replaced by ISO 14040:2006
ISO 14041:1998        Goal and Scope        Establishes at the outset the goals, purpose, audience, scope, and
                      Definition and        stakeholders that will be impacted or influenced by the results. This
                      Inventory Analysis    information influences the actual conduct of the LCA study. The
                                            inventory analysis portion is where the resources and releases
                                            related to the product system are quantified.
                                            Withdrawn in 2006 and replaced by ISO 14044:2006
ISO 14042:2000        Life Cycle Impact     The phase of life cycle assessment aimed at understanding and
                      Assessment (LCIA) evaluating the magnitude and significance of the potential
                                            environmental impacts of a product system.
                                            Withdrawn in 2006 and replaced by ISO 14044:2006
ISO 14043:2000        Life Cycle            The interpretation phase of an LCA, where the significance and
                      Interpretation        relative contributions of the results are broken down and analyzed.
                                            Withdrawn in 2006 and replaced by ISO 14044:2006
Source: ISO Standards Online Catalogue www.iso.org
* Name is usually shown as: “Environmental Management - Life Cycle Assessment - Name”



The ISO 14044:2006 standard specifies requirements and provides guidelines for life cycle
assessment (LCA) including: definition of the goal and scope of the LCA, the life cycle
inventory analysis (LCI) phase, the life cycle impact assessment (LCIA) phase, the life cycle
interpretation phase, reporting and critical review of the LCA, limitations of the LCA,
relationship between the LCA phases, and conditions for use of value choices and optional




                                                       29
CHEMINFO




elements. ISO 14044:2006 covers life cycle assessment (LCA) studies and life cycle inventory
(LCI) studies. 18

4.2.1 ISO 14040 - General Principles and Framework

It is useful to consider seven basic principles in the design and development of life cycle
assessments as a measure of environmental performance. The seven principles outlined below
are the basis of ISO Standard 14040:2006:

•    Life Cycle Perspective (the entire stages of a product or service);
•    Environmental Focus (addresses environmental aspects);
•    Relative Approach and Functional Unit (analysis is relative to a functional unit);
•    Iterative Approach (phased approach with continuous improvement)
•    Transparency (clarity is key to properly interpret results)
•    Comprehensiveness (considers all attributes and aspects)
•    Priority of Scientific Approach (preference for scientific-based decisions)


4.2.1.1 Life Cycle Perspective

LCA considers the entire life cycle stages of a product or service, including: extraction and
acquisition of all relevant raw materials, energy inputs and outputs, material production and
manufacturing, use or delivery, end-of-life treatment, and disposal or recovery. This systematic
overview of the product “system” provides perspective on the potential differences in
environmental burden between life cycle stages or individual processes.


4.2.1.2 Environmental Focus

The primary focus of a LCA is on the environmental aspects and impacts of a product system.
Environmental aspects are elements of an activity, product, or service that cause or can cause an
environmental impact through interaction with the environment. Some examples of
environmental aspects are: air emissions, water consumption, releases to water, land
contamination, and use of natural resources. Economic and social aspects are typically outside
the scope of an LCA, although it is possible to model some of these elements. Other tools may
be combined with LCA for more extensive analysis.




18
  International Standards Organization, ISO 14044:2006 - Environmental Management - Life cycle assessment -
Requirements and guidelines, 2006; www.iso.org/iso/iso_catalogue (ISO 14044:2006).


                                                   30
CHEMINFO




4.2.1.3 Relative Approach and Functional Unit

LCA is a relative analytical approach, which is structured on the basis of a functional unit of
product or service. The functional unit defines what is being studied and the life cycle inventory
(LCI) is developed relative to one functional unit. An example of a functional unit is a light-duty
gasoline vehicle driving an average kilometre (with other details of time, geography, trip
characteristics, and potential fuels added). All subsequent analyses are then developed relative to
that functional unit since all inputs and outputs in the LCI and consequently the LCIA profile are
related to the functional unit.

An LCA does not attempt to develop an absolute inventory of environmental aspects (e.g. air
emissions inventory) integrated over an organizational unit, such as a nation, region, sector, or
technology group.


4.2.1.4 Iterative Approach

LCA is an iterative analytical approach. The individual phases of an LCA (Goal and Scope
Definition; Inventory Analysis; Impact Assessment; and Interpretation) are all influenced by, and
use the results from, the other phases. The iterative approach within and between phases
contributes to a more comprehensive analysis and higher quality results.


4.2.1.5 Transparency

The value of an LCA depends on the degree of transparency provided in the analysis (for
example: the system description, data sources, assumptions and key decisions). The principle of
transparency allows users to understand the inherent uncertainty is the analysis and properly
interpret the results.


4.2.1.6 Comprehensiveness

A well-designed LCA considers all stages of the product system (the “reach”) and all attributes
or aspects of the natural environment, human health, and resources. Tradeoffs between
alternative product system stages and between environmental aspects in different media can be
identified and assessed.




                                               31
CHEMINFO




4.2.1.7 Priority of Scientific Approach

It is preferable to make decisions from an LCA analysis based on technical or science reasoning,
rather than from social or economic sciences. Where scientific approaches cannot be established,
consensual international agreement (e.g. international conventions) can be used. The power of
the technical or scientific approach lies in the proper attribution of facts to sources and the
potential reproducibility of these facts under scientific conditions. While the scientific approach
is typically more objective than economic or social values, it does not preclude the use economic
or social values for informing LCA decisions.

4.2.2 Application of LCA to Product Comparisons 19

LCA can be an effective tool within organizations to improve environmental management or to
guide research activities because of its requirements for comprehensiveness and the iterative
approach. However, when LCA is used to make environmental claims disclosed to the public
about the performance of a product or service system as compared to alternatives (a
“comparative assertion”), the ISO 14044 standard requires that a more rigorous process be
followed in preparing the LCA. Some additional requirements for “comparative assertions”
include:

•    Data Quality - A high quality of data must be used in a LCA for comparative assertions.
     This includes addressing the following data elements:

            o   time-related coverage (comparable time effects: duration, diurnal, seasonal, etc.);
            o   geographical coverage (comparable geography: weather, terrain, systems, etc.);
            o   technology coverage (comparable technical effects: product life-cycle systems);
            o   data precision (e.g. number of decimal places);
            o   completeness (similar product system “reach”, scope of life-cycle stages);
            o   data representativeness (does modeled data truly reflect actual performance?); and
            o   methodology consistency and reproducibility (standard measurement tests, etc.).

•    Peer Review - The LCA must be peer reviewed by an expert panel in accordance with the
     “critical review process” as outlined in ISO 14040. A review by a single internal expert or
     external expert is not permitted for a “comparative assertion”.

•    Impact Assessment - An impact assessment is required that uses category indicators that are
     sufficiently comprehensive, internationally accepted, scientifically and technically valid, and
     environmentally relevant. Weighting must not be used.



19
  International Standards Organization, ISO 14040:2006 - Environmental Management - Life cycle assessment -
Principles and framework, 2006.


                                                   32
CHEMINFO




•    Comparable Systems - The LCA comparison must be performed on systems using the same
     functional unit and equivalent methodological considerations, such as performance, system
     boundaries, data quality, allocation procedures, decision rules on evaluating inputs and
     outputs, and impact assessment. Any differences between systems regarding these
     parameters must be identified.


4.2.3 Conclusions on LCA Application

The following conclusions about the ISO LCA standards have been drawn by Dr. James Fava,
founder of the SETAC LCA Advisory Group and head of the U.S. delegation in the development
of the ISO LCA standards. 20

•    The ISO LCA standards have established a worldwide set of rules to ensure that LCA studies
     are performed in a consistent, reproducible fashion. The standards provide a holistic way of
     thinking about product systems, a framework for analysis, and define the factors to consider
     in setting the goals and scope of the assessment, performing the inventory analysis,
     conducting an impact assessment, and how to interpret and communicate results.

•    The ISO peer review and criteria review process provides a system of checks and balances to
     ensure that LCA studies used for public policy and decision-making undergo additional
     review by independent and interested parties.

•    Practitioners should be able to demonstrate their knowledge of the requirements of the ISO
     LCA standards and that they have applied these requirements.

•    There is a learning curve in completing LCAs. A company’s first LCA study (either done
     internally or using external consultants) often takes more time and resources than expected.
     Subsequent studies usually become easier to complete.

•    Within the ISO LCA standards, there is sufficient flexibility to ensure that LCA studies can
     be completed on a number of applications, ranging from answers to questions on a select list
     of impact categories and/or life-cycle stages, to comprehensive studies supporting
     environmental claims.

•    Any LCA methodology used in the public context must have transparency, be publicly
     available, and must have undergone appropriate peer review.

•    The application of LCA internally within an organization to drive continuous improvement
     and innovation can achieve meaningful results but it must be consistently applied.

20
 Adapted from Fava, J., Can ISO Life Cycle Assessment Standards Provide Credibility for LCA? Building Design
& Construction, Nov. 2005, www.bdcnetwork.com


                                                    33
CHEMINFO




LCA studies can provide information on environmental tradeoffs and opportunities to improve a product
performance over its life cycle. However, complementary assessments, in particular those related to site-
specific environmental issues, are often necessary to provide a fuller understanding of absolute risks and
opportunities.


4.3 Goal Definition and Scoping

Goal definition and scoping is the first phase of the LCA process that defines the purpose and
method of including life cycle environmental impacts into the decision-making process. In this
phase, the following items must be determined: the type of information that is needed to add
value to the decision-making process, how accurate the results must be to add value, and how the
results should be interpreted and displayed in order to be meaningful and usable.

The goal definition and scoping of the LCA project will determine the time and resources
needed. The defined goal and scope will guide the entire process to ensure that the most
meaningful results are obtained. Every decision made throughout the goal definition and scoping
phase impacts either how the study will be conducted, or the relevance of the final results. The
following section identifies the decisions that must be made at the beginning of the LCA study
and the impact of these decisions on the LCA process.

The following six basic decisions should be made at the beginning of the LCA process to make
effective use of time and resources:

 •   Define the Goal of the Project.
 •   Determine What Type of Information Is Needed to Inform the Decision-Makers.
 •   Determine How the Data Should Be Organized and the Results Displayed.
 •   Determine What Will or Will Not Be Included in the LCA.
 •   Determine the Required Accuracy of Data.
 •   Determine Ground Rules for Performing the Work.

Each decision and its associated impact on the LCA process are explained below in further
detail.

4.3.1 Define the Goal of the Project

LCA is a versatile tool for quantifying the overall (cradle-to-grave) environmental impacts from
a product, process, or service. The primary goal is to choose the best product, process, or service
with the least effect on human health and the environment. There may also be secondary goals
for performing an LCA, such as:

 •   To prove one product is environmentally superior to a competitive product.


                                                   34
CHEMINFO




 •   To identify stages within the life cycle of a product or process where a reduction in
     resource use and emissions might be achieved.
 •   To determine the impacts to particular stakeholders or affected parties.
 •   To establish a baseline of information on a system's overall resource use, energy
     consumption, and environmental loadings.
 •   To help guide the development of new products, processes, or activities toward a net
     reduction of resource requirements and emissions.

4.3.2 Determine What Type of Information Is Needed to Inform the Decision-
   Makers

LCA can help answer a number of important questions. Identifying the questions that the
decision-makers care about will help define the study parameters. Some examples include:
 • What is the environmental impact to particular interested parties and stakeholders?
 • Which product or process causes the least environmental impact (quantifiably) overall or in
     each stage of its life cycle?
 • How will changes to the current product/process affect the environmental impacts across all
     life cycle stages?
 • Which technology or process causes the least amount of acid rain, smog formation, or
     damage to local trees (or any other impact category of concern)?
 • How can the process be changed to reduce a specific environmental impact of concern (e.g.,
     global warming)?

Once the appropriate questions are identified, it is important to determine the types of
information needed to answer the questions.

4.3.3 Determine How the Data Should Be Organized and the Results Displayed

LCA practitioners define how data should be organized in terms of a functional unit that
appropriately describes the function of the product/process being studied. Comparisons between
products/processes must be made on the basis of the same function, quantified by the same
functional unit. This ensures that the products/processes being compared are true substitutes for
each other. Careful selection of the functional unit to measure and display the LCA results will
improve the accuracy of the study and the usefulness of the results.

4.3.4 Determine What Will or Will Not Be Included in the LCA

As previously explained, an inventory analysis identifies and quantifies the environmental
releases of a product or process throughout its entire life cycle. Ideally, an LCA includes all four
stages of a product or process life cycle: raw material acquisition, manufacturing,
use/reuse/maintenance, and recycle/waste management.




                                                35
CHEMINFO




The four primary life cycle stages are explained in more detail below. These stages can be
expanded further to provide additional detail if necessary.

4.3.4.1 Raw Materials Acquisition

The life cycle of a product begins with the removal of raw materials and energy sources from the
earth. For instance, the harvesting of trees or the mining of non-renewable materials would be
considered raw materials acquisition. Transportation of these materials from the point of
acquisition to the point of processing is also included in this stage.

4.3.4.2 Manufacturing

During the manufacturing stage, raw materials are transformed into a product or package. The
product or package is then delivered to the consumer. The manufacturing stage consists of three
steps: materials manufacture, product fabrication, and filling/packaging/distribution.

Materials Manufacture

The materials manufacture step involves the activities that convert raw materials into a form that
can be used to fabricate a finished product.

Product Fabrication

The product fabrication step takes the manufactured material and processes it into a product that
is ready to be filled or packaged.

Filling/Packaging/Distribution

This step finalizes the products and prepares them for shipment. It includes all of the
manufacturing and transportation activities that are necessary to fill, package, and distribute a
finished product. Products are transported either to retail outlets or directly to the consumer. This
stage accounts for the environmental effects caused by the mode of transportation, such as
trucking and shipping.


4.3.4.3 Use/Reuse/Maintenance

This stage involves the consumer's actual use, reuse, and maintenance of the product. Once the
product is distributed to the consumer, all activities associated with the useful life of the product
are included in this stage. This includes energy demands and environmental wastes from both
product storage and consumption. The product or material may need to be reconditioned,
repaired or serviced so that it will maintain its performance. When the consumer no longer needs
the product, the product will be recycled or disposed.


                                                36
CHEMINFO




4.3.4.4 Recycle/Waste Management

The recycle/waste management stage includes the energy requirements and environmental
wastes associated with recycling and disposition of the product or material.


4.3.5 Determine the Required Accuracy of Data

The required level of data accuracy for the project depends on the use of the results and the
intended audience (i.e., will the results be used to support decision-making in an internal process
or a public forum?). For example, if the intent is to use the results in a public forum to support
product/process selection to a local community or regulator, then estimated data or best
engineering judgement for the primary material, energy, and waste streams may not be
sufficiently accurate to justify the final conclusions. In contrast, if the intent of performing the
LCA is for internal decision-making purposes only, then estimates and best engineering
judgement may be applied more frequently.

4.3.6 Determine Ground Rules for Performing the Work

Prior to moving on to the inventory analysis phase, it is important to define some of the logistical
procedures for the project.

 •   Documenting Assumptions - All assumptions or decisions made throughout the entire
     project must be reported along side the results of the LCA project.
 •   Quality Assurance Procedures - Quality assurance procedures are important to ensure that
     the goal and purpose for performing the LCA will be met at the conclusion of the project.
     The level of quality assurance procedures employed for the project depends on the available
     time and resources and how the results will be used.
 •   Reporting Requirements - Defining "up front" how the results should be documented and
     exactly what should be included in the final report helps to ensure that the final product
     meets the appropriate expectations. When reporting the results, or results of a particular
     LCA phase, it is important to thoroughly describe the methodology used in the analysis.


4.4 Life Cycle Inventory

A life cycle inventory is a process of quantifying energy and raw material requirements,
atmospheric emissions, waterborne emissions, solid wastes, and other releases for the entire life
cycle of a product, process, or activity.

In the life cycle inventory phase of an LCA, all relevant data is collected and organized. Without
an LCI, no basis exists to evaluate comparative environmental impacts or potential



                                                37
CHEMINFO




improvements. The level of accuracy and detail of the data collected is reflected throughout the
remainder of the LCA process.

An inventory analysis produces a list containing the quantities of pollutants released to the
environment and the amount of energy and material consumed. The results can be segregated by
life cycle stage, by media (air, water, land), by specific processes, or any combination thereof.

A life cycle inventory is usually completed used the following steps:
 • Develop a flow diagram of the processes being evaluated,
 • Develop a data collection plan,
 • Collect data,
 • Evaluate and report results.

Each step is summarized below.

4.4.1 Develop a Flow Diagram

A flow diagram is a tool to map the inputs and outputs to a process or system. The goal
definition and scoping phase establishes initial boundaries that define what is to be included in a
particular LCA; these are used as the system boundary for the flow diagram. Unit processes
inside of the system boundary link together to form a complete life cycle picture of the required
inputs and outputs (material and energy) to the system. A flow diagram for a complex process
may look like the following figure. The individual energy, materials flows, and emissions are
summed to provide the net flows in and out of the system boundary. The "system" or "system
boundary" can vary for every LCA project. This is one of the main reasons for the different
LCA results.
                                Figure 6: Process Flow Diagram




                                               38
CHEMINFO




4.4.2 Develop an LCI Data Collection Plan

As part of the goal definition and scoping phase, the required accuracy of data was determined.
When selecting sources for data to complete the life cycle inventory, an LCI data collection plan
ensures that the quality and accuracy of data meet the expectations of the decision-makers.

Key elements of a data collection plan include the following:
 • Defining data quality goals,
 • Identifying data sources and types,
 • Identifying data quality indicators.

Each element is described below.

 •   Define Data Quality Goals - Data quality goals provide a framework for balancing
     available time and resources against the quality of the data required to make a decision
     regarding overall environmental or human health impact.
 •   Identify Data Quality Indicators - Data quality indicators are benchmarks to which the
     collected data can be measured to determine if data quality requirements have been met.
 •   Identify Data Sources and Types - For each life cycle stage, unit process, or type of
     environmental release, specify the necessary data source and/or type required to provide
     sufficient accuracy and quality to meet the study's goals. Defining the required data sources
     and types prior to data collection helps to reduce costs and the time required to collect the
     data.

4.4.3 Collect Data

The flow diagram developed in Step 1 provides the road map for data to be collected. Step 2
specifies the required data sources, types, quality, accuracy, and collection methods. Step 3
consists of finding and filling in the flow diagram and worksheets with numerical data. This may
not be a simple task. Some data may be difficult or impossible to obtain, and the available data
may be difficult to convert to the functional unit needed. Therefore, the system boundaries or
data quality goals of the study may have to be refined based on data availability. This iterative
process is common for most LCA's.

Data collection efforts involve a combination of research, site-visits and direct contact with
experts, which generate large quantities of data. An electronic database or spreadsheet can be
useful to hold and manipulate the data. As an alternative to developing a computer model, it may
be more effective to buy a commercially available LCA software package.




                                               39
CHEMINFO




4.4.4 Evaluate and Document the LCI Results

Now that the data has been collected and organized into one format or another, the accuracy of
the results must be verified. The accuracy must be sufficient to support the purposes for
performing the LCA as defined in the goal and scope.

When documenting the results of the life cycle inventory, it is important to thoroughly describe
the methodology used in the analysis, define the systems analyzed and the boundaries that were
set, and all assumptions made in performing the inventory analysis.

The outcome of the inventory analysis is a list containing the quantities of pollutants released to
the environment and the amount of energy and materials consumed. The information can be
organized by life cycle stage, by media (air, water, land), by specific process, or any combination
thereof.

4.5 Life Cycle Impact Assessment

The Life Cycle Impact Assessment (LCIA) phase of an LCA is the evaluation of potential human
health and environmental impacts of the environmental resources and releases identified during
the life cycle inventory (LCI). A life cycle impact assessment attempts to establish a linkage
between the product or process and its potential environmental and health impacts.

The key concept in this component is that of stressors. A stressor is a set of conditions that may
lead to an impact. For example, if a product or process is emitting greenhouse gases, the increase
of greenhouse gases in the atmosphere may contribute to global warming. Processes that result in
the discharge of excess nutrients into bodies of water may lead to eutrophication. An LCIA
provides a systematic procedure for classifying and characterizing these types of environmental
effects.

Key Steps of a Life Cycle Impact Assessment

The following steps comprise a life cycle impact assessment.

 •   Selection and Definition of Impact Categories - identifying relevant environmental impact
     categories (e.g., global warming, acidification, terrestrial toxicity).
 •   Classification - assigning LCI results to the impact categories (e.g., classifying CO2
     emissions to global warming).
 •   Characterization - modeling LCI impacts within impact categories using science-based
     conversion factors. (e.g., modeling the potential impact of CO2 and methane on global
     warming).
 •   Evaluating and Reporting LCIA Results - gaining a better understanding of the reliability of
     the LCIA results.



                                               40
CHEMINFO




4.5.1 Select and Define Impact Categories

The first step in an LCIA is to select the impact categories that will be considered as part of the
overall LCA. This step should be completed as part of the initial goal and scope definition phase
to guide the LCI data collection process and requires reconsideration following the data
collection phase.

For an LCIA, impacts are defined as the consequences caused by the input and output streams of
a system on human health, plants and animals, or the future availability of natural resources.
Typically LCIA’s focus on the potential impacts to three main categories: human health,
ecological health, and resource depletion. The following table shows some of the more
commonly used impact categories.




                                               41
CHEMINFO




                              Table 7: Common Impact Categories

Impact             Scale      Relevant LCI Data                    Common           Description of
Category                      (i.e., classification)               Characterization Characterization
                                                                   Factor           Factor
Global Warming     Global     Carbon Dioxide (CO2)                 Global        Warming Converts LCI data to carbon
                              Methane (CH4)                        Potential             dioxide (CO2) equivalents
                              Nitrous Oxide (N2O)                                        Note:     global     warming
                              Perfluorocarbons (PFCs)                                    potentials can be 50, 100, or
                              Hydrofluorocarbons (HFCs)                                  500 year potentials.
                              Sulphur Hexafluoride (SF6)
                              Chlorofluorocarbons (CFCs)
                              Hydrochlorofluorocarbons
                              (HCFCs)
                              Methyl Bromide (CH3Br)
Stratospheric      Global     Chlorofluorocarbons (CFCs)           Ozone         Depleting Converts LCI data        to
Ozone Depletion               Hydrochlorofluorocarbons             Potential               trichlorofluoromethane
                              (HCFCs) Halons                                               (CFC-11) equivalents
                              Methyl Bromide (CH3Br)
Acidification      Regional   Sulphur Oxides (SOx)                 Acidification Potential   Converts LCI data to
                   Local      Nitrogen Oxides (NOx)                                          hydrogen     (H+) ion
                              Hydrochloric Acid (HCl)                                        equivalents.
                              Hydrofluoric Acid (HF)
                              Ammonia (NH3)
Eutrophication     Local      Phosphate (PO4)                      Eutrophication            Converts LCI    data to
                              Nitrogen Oxide (NO)                  Potential                 phosphate           (PO4)
                              Nitrogen Dioxide (NO2)                                         equivalents.
                              Nitrates
                              Ammonia (NH3)
Photochemical      Local      Non-methane hydrocarbon              Photochemical Oxidant Converts LCI data to ethane
Smog                          (NMHC)                               Creation Potential    (C2H6) equivalents.
Terrestrial        Local      Toxic chemicals with a reported      LC50                  Converts LC50 data to
Toxicity                      lethal concentration to rodents                            equivalents.
Aquatic Toxicity   Local      Toxic chemicals with a reported      LC50                  Converts LC50 data to
                              lethal concentration to fish                               equivalents.
Human Health       Global     Total releases to air, water, and    L50                   Converts LC50 data to
                   Regional   soil.                                                      equivalents.
                   Local
Resource           Global     Quantity of minerals used Quantity Resource      Depletion Converts LCI data to a ratio
Depletion          Regional   of fossil fuels used               Potential               of quantity of resource used
                   Local                                                                 versus quantity of resource
                                                                                         left in reserve.
Land Use           Global     Quantity disposed of in a landfill   Solid Waste           Converts mass of solid
                   Regional                                                              waste into volume using an
                   Local                                                                 estimated density.




                                                             42
CHEMINFO




4.5.2 Classification

The purpose of classification is to organize and possibly combine the LCI results into impact
categories. For LCI items that contribute to only one impact category, the procedure is a
straightforward assignment.

For LCI items that contribute to two or more different impact categories, a rule must be
established for classification. There are two ways of assigning LCI results to multiple impact
categories:
 • Allocate a representative portion of the LCI results to the impact categories to which they
     contribute. This is typically allowed in cases when the effects are dependent on each other.
 • Assign all LCI results to all impact categories to which they contribute. This is typically
     allowed when the effects are independent of each other.

4.5.3 Characterization

Impact characterization uses science-based conversion factors, called characterization factors, to
convert and combine the LCI results into representative indicators of impacts to human and
ecological health. Characterization factors also are commonly referred to as equivalency factors.
Characterization provides a way to directly compare the LCI results within each impact category.

Impact indicators are typically characterized using the following equation:

                  Inventory Data × Characterization Factor = Impact Indicators

For example, all greenhouse gases can be expressed in terms of carbon dioxide (CO2)
equivalents by multiplying the relevant LCI results by a CO2 characterization factor and then
combining the resulting impact indicators to provide an overall indicator of global warming
potential.

The key to impact characterization is using the appropriate characterization factor. For some
impact categories, such as global warming and ozone depletion, there is a consensus on
acceptable characterization factors. For other impact categories, such as resource depletion, a
consensus is still being developed.

4.5.4 Evaluate and Document the LCIA Results

When documenting the results of the life cycle impact assessment, thoroughly describe the
methodology used in the analysis, define the systems analyzed and the boundaries that were set,
and all assumptions made in performing the inventory analysis.




                                               43
CHEMINFO




4.6 Life Cycle Interpretation

Life cycle interpretation is a systematic technique to identify, quantify, check, and evaluate
information from the results of the life cycle inventory (LCI) and the life cycle impact
assessment (LCIA), and communicate them effectively. Life cycle interpretation is the last phase
of the LCA process.

The International Organization for Standardization (ISO) has defined the following two
objectives of life cycle interpretation:

 •   Analyze results, reach conclusions, explain limitations and provide recommendations based
     on the findings of the preceding phases of the LCA and to report the results of the life cycle
     interpretation in a transparent manner.
 •   Provide a readily understandable, complete, and consistent presentation of the results of an
     LCA study, in accordance with the goal and scope of the study.

Interpreting the results of an LCA is not as simple as “2 is better then 3, therefore Alternative A
is the best choice”! While conducting the LCI and LCIA, it is necessary to make assumptions,
engineering estimates, and decisions based on your values and the values of involved
stakeholders. Each of these decisions must be included and communicated within the final
results to clearly and comprehensively explain conclusions drawn from the data. In some cases,
it may not be possible to state that one alternative is better than the others because of the
uncertainty in the final results. This does not imply that efforts have been wasted. The LCA
process will still provide decision-makers with a very valuable and clearer understanding of the
environmental loadings and potential environmental and health impacts associated with each
alternative, where they occur (locally, regionally, or globally), and the relative magnitude of
each type of impact in comparison to each of the proposed alternatives included in the study.

The purpose of conducting an LCA is to better inform decision-makers by providing a particular
type of information (often unconsidered), with a life cycle perspective of environmental and
human health impacts associated with each product or process. However, LCA does not take into
account technical performance, cost, or political and social acceptance. LCA results should be
integrated with other business and public policy decision-making tools.

The following steps to conducting a life cycle interpretation are identified and discussed:

 •   Identify Significant Issues.
 •   Evaluate the Completeness, Sensitivity, and Consistency of the Data.
 •   Draw Conclusions and Recommendations.




                                                44
CHEMINFO




                          Figure 7: Life Cycle Interpretation Process




4.6.1 Identify Significant Issues

Determining significant issues of a product system may be simple or complex. Significant issues
can include:

 •   inventory parameters like energy use, emissions, waste, etc.
 •   impact category indicators like land use, acidification, eutrophication, etc.
 •   contributions of different life cycle stages to LCI or LCIA results such as individual unit
     processes or groups of processes (e.g., transportation, energy production).

4.6.2 Evaluate the Completeness, Sensitivity, and Consistency of the Data

The evaluation step of the interpretation phase establishes the confidence in and reliability of the
results of the LCA. This is accomplished by completing the following tasks to ensure that
products/processes are fairly compared:

 •   Completeness Check - examining the completeness of the study
 •   Sensitivity Check - assessing the sensitivity of the significant data elements that influence
     the results most greatly
 •   Consistency Check - evaluating the consistency used to set system boundaries, collect data,
     make assumptions, and allocate data to impact categories for each alternative.



                                                45
CHEMINFO




After completing steps 1 and 2, it has been determined that the results of the impact assessment
and the underlying inventory data are complete, comparable, and acceptable to draw conclusions
and make recommendations.

4.6.3 Draw Conclusions and Recommendations

The objective of this step is to interpret the results of the life cycle impact assessment (LCIA, not
the LCI) to determine which product/process has the overall least impact to human health and
the environment, and/or to one or more specific areas of concern as defined by the goal and
scope of the study.

4.6.4 Reporting the Results

The materials should be assembled into a comprehensive report documenting the study in a clear
and organized manner. This will help communicate the results of the assessment fairly,
completely, and accurately to others interested in the results. The report presents the results,
data, methods, assumptions and limitations in sufficient detail to allow the reader to comprehend
the complexities and trade-offs inherent in the LCA study.


4.7 Types of LCA Models

In the literature there appear to be two primary means of determining the emissions that are
embedded in energy production facilities: a process-chain analysis (PCA) and an input/output
analysis (IOA). The PCA calculates the energy embedded in and the emission-equivalents
caused by the production of materials used in the application. The IOA works with economic
sectors related to the manufacturing activities. The PCA approach requires some knowledge of
the materials included in the facility whereas the IOA only requires an understanding of the costs
of construction and the economic structure of the country or region where the construction is
occurring. Both approaches should yield similar results and they are discussed briefly in the
following sections.

4.7.1 Process Chain Analysis

The PCA looks at the materials (steel, concrete, plastics, etc.) and converts them, considering all
underlying production steps, into the corresponding amount of energy used and GHGs emitted.
Shortcomings of the PCA are that the method is intrinsically incomplete (some processes cannot
be expressed in an amount of material and are therefore likely to be overlooked) and that all
products, made from the same basic material, are dealt with in the same way. Materials produced
in different countries may also have different energy and materials flows making it sometimes
difficult to extrapolate results between regions.




                                                46
CHEMINFO




PCA basis is an inventory analysis using bottom-up data collection. It investigates the flow of
materials and energy in each production process. Each material or energy that forms the main
process is traced back through its initial extraction. It evaluates the embedded energy and the
embedded emissions caused by the material production. PCA considers all individual emission
points of GHG, and therefore requires careful analysis of all flow of energy and materials
associated with its links of production processes. Hence, emission factors of all energy types and
all materials required by all the process steps must be available. Data collection can be very time
consuming and complex.

For example, the PCA approach is the way that most pathways in the GHGenius model have
been developed. In GHGenius there is data on the energy and emissions associated with many of
the materials typically found in energy production facilities.

4.7.2 Input Output Analysis

The IOA divides a product into its economic components. Each input, which contributes to the
creation of the final product, is ascribed to an economic sector (machinery, electrical, services,
etc). For each sector, an average product is calculated, which is characterized by an amount of
energy needed and an amount of GHG emitted. The advantage of the IOA is that each input can
easily be expressed in an economic value. The main shortcoming of the IOA is that all products
are identified as an average product of the covering sector. A sector, however, contains many
products for which the ratio price/energy-input is not necessarily the same (e.g. the price
difference between a luxury vehicle and a sub-compact is much greater than the relative
difference in energy requirement, but both products belong to the same sector ‘vehicles’).
Another shortcoming is that the number of sectors may be limited.

The IOA also requires the relevant relationship between the economic value of the sector and the
energy and emissions attributed to the sector. These are not always available. New industrial
processes for which there is no historical data are difficult to assess with the input/output
analysis.




                                               47
CHEMINFO




           5. Life Cycle Models, Initial Screening

5.1 Overview

Literature and Internet searches were undertaken to identify models that could potentially
provide a life cycle analysis (LCA) for bioethanol. An extensive list of potential models was
identified, and these were screened to determine which offered potential to evaluate bioethanol.
The list of models identified is shown below.

•   ASPEN Plus                                •     GEMIS
•   BEES                                      •     GHGenius
•   BESS                                      •     GREET
•   The Boustead Model                        •     IVAM LCA
•   CMLCA                                     •     KCL-ECO
•   CUMPAN                                    •     LCAiT
•   DAYCENT                                   •     LCAPIX
•   Eco-Indicator 99                          •     LEM
•   Ecoinvent                                 •     MIET
•   ECO-it                                    •     PEMS
•   ECOPRO                                    •     PIA
•   EcoScan                                   •     Regis
•   EDIP                                      •     SimaPro
•   EIE-Athena                                •     SPINE
•   EIO-LCA                                   •     TEAM
•   EMPA                                      •     TRACI
•   EPS 2000                                  •     UMBERTO
•   EUKLID                                    •     WISARD
•   GaBi

These models were initially reviewed to select those that offered the most potential to conduct a
sensitivity analysis under the context of this engagement. Namely, this includes selecting models
that:

•   were developed to address the biofuels life cycle (as opposed to addressing another life
    cycle, such as building materials);
•   are expected to have the built-in capabilities to assess the biofuels life cycle (and do not
    require the user to assemble processes and data); and
•   continue to be available to allow the consultants to conduct a sensitivity analysis under the
    context of this study.




                                               48
CHEMINFO




The results of the screening process suggest that the following may be most capable of
conducting life cycle assessments of bioethanol under the context of this study:

     BEES                                                     BESS
     EIO-LCA                                                  GaBi
     GEMIS                                                    GHGenius
     GREET                                                    LEM
     SimaPro

These nine models are examined in further detail in a later section.


5.2 Screening Assessment of Available Models

The models were screened to identify their potential to address the ethanol life cycle. The
generic and available models were evaluated. Others may have developed or customized these
generic models to include modules for addressing the biofuels life cycle. Proprietary, customized
models were not included in the analysis.

5.2.1 ASPEN Plus

Aspen Plus is a process modeling tool for conceptual design, optimization, and performance
monitoring for the chemical, polymer, specialty chemical, metals and minerals, and coal power
industries. 21

It is not a model suitable for conducting a life cycle assessment of biofuels, and is not considered
further.

5.2.2 BEES

The BEES (Building for Environmental and Economic Sustainability) software was developed
by National Institute for Standards and Technology (NIST) Building and Fire Research
Laboratory. It was designed to help select cost-effective, environmentally-preferable building
products. BEES measures the environmental performance of building products by using the life-
cycle assessment approach specified in the ISO 14040 series of standards.

Version 4.0 of the Windows-based decision support software, aimed at designers, builders, and
product manufacturers, includes actual environmental and economic performance data for 230
building products. 22 BEES 4.0 runs on Windows 95, and beyond personal computers with at
least 60 MB of available disk space. A copy can be downloaded from the Internet. 23

21
   http://www.aspentech.com/products/aspen-plus.cfm
22
   http://www.bfrl.nist.gov/oae/software/bees/
23
   http://www.bfrl.nist.gov/oae/software/bees/registration.html


                                                         49
CHEMINFO




In support of the 2002 Farm Security and Rural Investment Act (P.L. 107-171), BEES was
adapted for application to biobased products (called BEES Please for USDA). The BEES
Environmental Performance Score combines product performance across all 12 environmental
impacts into a single score. These impacts are: 1. global warming; 2. acidification; 3.
eutrophication; 4. fossil fuel depletion; 5. indoor air quality; 6. habitat alteration; 7. water intake;
8. criteria air pollutants; 9. human health; 10. smog; 11. ozone depletion; and 12. ecological
toxicity. The lower the score, the better is the product’s overall environmental performance.

Give the adaptation to address biobased products, BEES Please for USDA may be suitable for
conducting a life cycle assessment of biofuels, and is examined in more detail in a later section. 24

5.2.3 BESS

BESS is the Biofuel Energy Systems Simulator developed at the University of Nebraska -
Lincoln. The BESS model is a software tool to calculate the energy efficiency, greenhouse gas
emissions, and natural resource requirements of corn–to-ethanol biofuel production systems. The
non-commercial version of BESS can be downloaded over the Internet for free. 25,26

The BESS model has four components: (1) crop production; (2) ethanol biorefinery; (3) cattle
feedlot; and (4) anaerobic digestion (optional). 27 The model does not include emissions from fuel
distribution or use. The model includes ethanol production from corn. The developers suggest
that the model will be extended to cover ethanol from corn stover and switchgrass for cellulosic
ethanol in the future.

The model provides outputs in terms of:

•    energy use;
•    greenhouse gas emissions - CO2, CH4, and N2O, as well as global warming potential based
     on those three greenhouse gases; and
•    environmental requirements - land, grain, water, and petroleum.

The model is populated with U.S. (average) data, and a regional analysis based on north-eastern
U.S. coal or natural gas inputs can be conducted.

BESS is a pre-built biofuels life cycle model that includes North American data, and as such is
considered further in a later section.

24
   Efforts were made to contact BEES Please for USDA to understand capabilities with respect to the biofuels life
cycle (telephone call January 8, 2008). No additional information on BEES Please for USDA was made available.
25
   http://www.bess.unl.edu/download/
26
   Several e-mail requests to the University of Nebraska - Lincoln requesting the price of the commercial version
have not been returned.
27
   University of Nebraska Lincoln (2007), BESS Biofuel Energy Systems Simulator Users Guide.


                                                      50
CHEMINFO




5.2.4 The Boustead Model

The Boustead Model is a computer modeling tool for lifecycle inventory calculations. 28 The
model is composed of three main groups of files: the program files, the core data files, and the
top data files.

A licence can be purchased to use the Boustead software and database. Licensees receive an off-
the-shelf boxed package containing the database and software on CD-ROM as well as an
operating manual, an Introduction to LCA book and a printed index of operations and conversion
factors.

In the supplied databases there are nearly 13,000 individual unit operations, covering a vast
number of materials processing and fuel production processes. However, a review of these
individual unit operations suggests that the processes do not cover biofuels (bioethanol or
biodiesel) production. 29

While this model could be used to build up a life cycle emissions model for biofuels, it does not
appear to come with those capabilities built in and is not considered further. 30

5.2.5 CENTURY and DAYCENT

The CENTURY agroecosystem model is the latest version of the soil organic model. This model
simulates carbon, nitrogen, phosphorus, and sulphur dynamics through an annual cycle over time
scales of centuries and millennia. CENTURY was especially developed to deal with a wide
range of cropping system rotations and tillage practices for system analysis of the effects of
management and global change on productivity and sustainability of agroecosystems. 31

Century 5 is a research version of the Century model that has not been released for public
download. Century 4.0 is a freely available version of the Century model that can be
downloaded. 32 The DAYCENT ecosystem model is a daily version of CENTURY.

It is not a model suitable for conducting a life cycle assessment of biofuels, and is not considered
further.




28
   http://www.boustead-consulting.co.uk/products.htm
29
   http://www.boustead-consulting.co.uk/download/Modelcontents.pdf
30
   Boustead Consulting was contacted on December 18 to confirm that the model does not have the built-in
capabilities to assess to address the biofuels life cycle. No response has been received.
31
   http://nrel.colostate.edu/projects/century5/reference/html/Century/overview.htm
32
   http://www.nrel.colostate.edu/projects/century/


                                                  51
CHEMINFO




5.2.6 CMLCA

CMLCA is an abbreviation of Chain Management by Life Cycle Assessment. It is a software
tool that is intended to support the technical steps of the life cycle assessment procedure (LCA).
A demonstration version with full capability is available. For commercial use, a fee for an
indefinite licence is due per copy installed. The fee is approximately C$1,500.

The model was prepared at Leiden University, the Netherlands. The latest version (Version 4.2)
was published in May 2004. The source website reports that “The present version is, like all
versions, an experimental release. It will almost surely contain errors ... The user is advised to
store data quite regularly, as the program may crash at unexpected points.” It also reports that
“CMLCA does not provide a flexible user interface. Exchange of data with other programs is
cumbersome. There is almost no graphical output. There is no online help, neither is there a
helpdesk. It may contain programming errors.” 33

A preliminary review of descriptions of the model suggest that it does not have the built-in
capabilities addressing the biofuels life cycle, but rather relies on custom builds. 34 CMLCA is
not considered further.

5.2.7 CUMPAN

The CUMPAN model was originally developed by the University of Hohenheim and released in
1996. It was later acquired by Debis Systemhaus Engineering AG (now majority owned by
Deutsch Telecom, along with Daimler Chrysler).

Little information is presently available describing CUMPAN. Only a German version was
reportedly available. 35 One report suggests that it could be licensed for about $8,400 Canadian in
2006, and that about 62 licenses were sold. 36

Extensive efforts were made without success to secure a copy of the CUMPAN and
corresponding documentation. 37 It is expected that the CUMPAN model has not been updated or
made available for several years, and it is not considered further for this study.




33
   http://www.leidenuniv.nl/cml/ssp/software/cmlca/faq.html
34
   http://www.leidenuniv.nl/cml/ssp/software/cmlca/faq.html
35
   Center for Clean Products and Clean Technologies, University of Tennessee (1996), Evaluation Of Life-Cycle
Assessment Tools.
36
   EU DG Environment (2005), Making Life Cycle Assessment Information and Interpretive Tools Available.
37
   Efforts were made to secure the CUMPAN model through extensive Internet searches, as well as telephone calls
to T-Systems and several e-mails to the CUMPAN prime contact (Horst Krasowski). In addition, a PowerPoint
presentation of CUMPAN capabilities was been requested by e-mail, nothing has been received to date (per
http://www.ivl.se/rapporter/pdf/B1390.pdf).


                                                     52
CHEMINFO




5.2.8 Eco-Indicator 99

The Eco-indicator 99 is a science based impact assessment method for LCA. It offers a way to
measure various environmental impacts, and shows a final result in a single score by PRé
Consultants. 38

The method is also the basis for the calculation of eco-indicator scores for materials and
processes. These scores can be used as a design for environment tool for designers and product
managers to improve products. The impact assessment method is widely used by life cycle
assessment practitioners around the world.

Eco-indicator 99 scores can be used to make your own environmental assessment of a product in
a matter of minutes (using over 200 predefined scores for commonly used materials and
processes).

Eco-Indicator 99 is an impact assessment method and not a model suitable for conducting a life
cycle assessment of biofuels, and is not considered further.

5.2.9 Ecoinvent

The Ecoinvent database v2.0 contains international industrial life cycle inventory data on energy
supply, resource extraction, material supply, chemicals, metals, agriculture, waste management
services, and transport services. 39

It is the product of the Swiss Centre for Life Cycle Inventories, a joint initiative of several
partners belonging to the Domain of the Swiss Federal Institutes of Technology (ETH) and
supported by different Swiss Federal Offices. The database is sold by vendors of life cycle
impact models that use Ecoinvent, for example GaBi and Sima Pro.

The database with its 4,000 life cycle inventory datasets and almost 300 impact assessment
indicators is available online. The model contains some data on biofuels, including a dataset on
ethanol from corn for the U.S. and biodiesel from soybeans for Brazil. 40

Ecoinvent is a database, and by itself is not suitable for conducting a life cycle assessment of
biofuels, and is not considered further.




38
   http://www.pre.nl/eco-indicator99/default.htm
39
   http://www.ecoinvent.org/
40
   http://www.ecoinvent.org/fileadmin/documents/en/ecoinventData-v2.0_Contents.pdf


                                                     53
CHEMINFO




5.2.10 ECO-it

ECO-it is a life cycle assessment software model offered by PRé Consultants. ECO-it calculates
the environmental load, and shows which parts of the product's life cycle contribute most. It is
described as a being a tool for product and packaging designers. 41

ECO-it comes with over 200 eco-indicator 99 scores for commonly used materials such as
metals, plastics, paper, board and glass, as well as production, transport, energy and waste
treatment processes (see Eco-Indicator 99). It calculates the environmental load of a product and
shows which parts of the product contribute most. A working, 10-day evaluation demonstration
program is available over the Internet. 42

Eco-it is a building block model, does not represent a constructed biofuels life cycle assessment
model, and is not considered further.

5.2.11 ECOPRO

EcoPro is no longer distributed, and is not considered further. 43

5.2.12 EcoScan

Ecoscan 3.0 analyzes the environmental impact and cost of products. The software tool can be
used by managers and engineers who implement EcoDesign in real life product development.

A current source of the EcoScan software was sought, but not found. It is expected that the
company offering EcoScan (TNO), no longer offers the product. A search of TNO’s website did
not find the product. 44

As it appears that EcoScan product is no longer available, it is not considered further.




41
   http://www.pre.nl/eco-it/eco-it.htm
42
   http://www.pre.nl/eco-it/download_eco-it.htm
43
   The Sinum Eco-Pro website (http://www.sinum.com/htdocs/e_software_ecopro.shtml) states “We apologise, but
the link you used is outdated and the distribution and development of EcoPro has been stopped a while ago.”
(November 19, 2007).
44
   http://www.tno.nl/index.cfm


                                                     54
CHEMINFO




5.2.13 EDIP

The EDIP Methodology (Environmental Design of Industrial Products) for life cycle assessment
was established during the 1990's. The methodology was mainly developed as a tool for product
development in Danish industry.

The EDIP PC-Tool was developed for the Danish EPA. The EDIP PC-Tool is a Windows
application and database that supports the LCA process carried out according to the EDIP
method. An updated beta version 3.0 was released in October 2002. The tool costs DKK 4375
(about C$900). 45 To carry out an LCA, detailed information on all the processes and materials
included in the life cycle of the product is needed. 46

The EDIP PC-Tool appears to be a building block model, does not represent a constructed
biofuels life cycle assessment model, and is not considered further.

5.2.14 EIO-LCA

The Economic Input Output-Life Cycle Assessment was developed by the Carnegie-Mellon
Green Design Institute. The model is based on 1997 U.S. economic data, and is available online
at no charge.

The EIO-LCA model allows for the estimation of some environmental impacts from producing a
certain dollar amount of any of 500 commodities or services in the United States. It provides
guidance on the relative impacts of different types of products, materials, services, or industries
with respect to resource use and emissions throughout the U.S. The impacts are from production
only. Impacts from use, waste disposal, etc. are not included.

The model provides estimates of economic activity, conventional air pollutants (SO2, CO, NOx,
VOCs, Lead, and PM10), greenhouse gases (CO2, CH4, N2O, CFCs), energy, toxic releases, and
employment. Data can be graphed and mapped.

EIO-LCA has components that may support life cycle assessment of biofuels, and will be
examined in more detail in a later section.

5.2.15 EMPA

EMPA is the Swiss materials science and technology research institution. It specializes in
applications, focused research and development, and provides high-level services in the field of
sustainable materials science and technology. EMPA has a Life Cycle Assessment and Modeling
Unit that analyses the effects of products, technical process and social activities on the
environment.
45
     http://glwww.mst.dk/indu/03040000.htm
46
     http://www.epa.gov/NRMRL/lcaccess/pdfs/appendices_lca101.pdf


                                                     55
CHEMINFO




EMPA is not a life cycle model for assessing biofuels, and is not considered further.


5.2.16 Environmental Impact Estimator

Environmental Impact Estimator can assess the environmental implications of industrial,
institutional, office, and both multi-unit and single family residential designs. The Estimator
incorporates the Athena Institute’s internationally recognized life cycle inventory databases,
covering more than 90 structural and envelope materials. It simulates over 1,000 different
assembly combinations and is capable of modeling 95% of the building stock in North
America. 47

This software is not designed to model fuel cycles, and is not considered further.

5.2.17 EPS 2000

The EPS (Environmental Priority Strategies) method was developed in Sweden. It calculates an
environmental load unit which essentially comes from taking LCI data, classifying it into five
categories known as safeguard areas (biodiversity, human health, production, resources, and
aesthetic values) and multiplying each item by unit effects expressed as monetary values. These
unit effects are estimates of a society’s willingness to pay to keep from causing damage to the
five safeguard areas. A final environmental load unit (ELU) is calculated and then equated to a
currency (e.g. 1 ELU = 1 ECU). 48

The EPS method does not represent a constructed biofuels life cycle assessment model, and is
not considered further.

5.2.18 EUKLID

Euklid was reportedly made available in 1996, primarily to assess the impacts of processes and
packaging. 49

It is expected that this model is no longer available, and it will not be considered further. 50




47
   http://www.athenasmi.ca/tools/docs/Athena_Institute-Software.pdf
48
   http://www.icmm.com/uploads/35Eco-Indices.pdf
49
   http://www.sematech.org/docubase/document/4238atr.pdf
50
   Attempts were made to find this model or descriptions of this model via an Internet search and through contact
with the Fraunhofer Institut (by telephone and e-mail).


                                                      56
CHEMINFO




5.2.19 GaBi

GaBi was developed by the German company PE Europe GmbH and IKP at the University of
Stuttgart. The first version of GaBi was developed about 15 years ago. Since then a wide range
of production companies has participated in the further development of the software. 51

GaBi is available in several versions, including GaBi Professional and GaBi Lite. GaBi costs
approximately C$10,000 for the professional version and core databases. Supplementary
databases cost more (some expected to cost in excess of C$10,000). 52

The Professional Version comes with an extensive core database, and the option to purchase
additional databases. The core database include German, European, and North American data.
Additional databases appear to provide information on the biofuels life cycle (notably Extension
Database XII). This information would allow the development of modules to estimate life cycle
impacts.

While GaBi does not appear to be Canadianized, and modules for biofuels life cycle impact
assessment are not built-in, it has some potential to assess biofuels life cycles and will be
examined in further detail in a later section.

5.2.20 GEMIS

GEMIS is a life-cycle analysis program and database for energy, material, and transport systems
- it is available freely at no cost (public domain). 53

The basic version 1.0 of the computer program GEMIS was developed in 1987-1989 as a tool for
the comparative assessment of environmental effects of energy by Öko-Institut and
Gesamthochschule Kassel (GhK). Since then, the model was continuously upgraded and
updated. GEMIS includes the total life-cycle in its calculation of impacts - i.e. fuel delivery,
materials used for construction, waste treatment, and transports/auxiliaries. The GEMIS database
offers information on fuels (including biofuels), processes, materials, and transport. It covers the
full product life-cycle and a wide range of impact categories (such as air pollutants, greenhouse
gases, and land use).

GEMIS may support or have components that support life cycle assessment of biofuels, and will
be examined in more detail in a later section.




51
   http://www.gabi-software.com/
52
   http://www.lca-center.dk/cms/site.asp?p=4633
53
   http://www.oeko.de/service/gemis/en/index.htm


                                                   57
CHEMINFO




5.2.21 GHGenius

Dr. Mark Delucchi developed the first version of his Lifecycle Emissions Model (LEM) during
the period of 1987-1993. Partial Canadianization of LEM was completed by Dr. Delucchi for
Natural Resources Canada in late 1998 through to March 1999. The partially Canadianized
version of the fuel cycle model was the basis for the development of GHGenius. The model was
used for a number of studies for Governments and Industry between 1999 and 2007. For each
study the data in the model was further refined for Canadian circumstances. 54

GHGenius focuses on the life cycle assessment (LCA) of current and future fuels for
transportation applications. All of the steps in the life cycle are included in the model from raw
material acquisition to end-use. The fuel cycle segments span feedstock production and recovery,
fertilizer manufacture, land use changes and cultivation associated with biomass derived fuels,
leaks and flaring associated with production of oil and gas, feedstock transport, fuel production
(as in production from raw materials), emissions displaced by co-products of alternative fuels,
fuel storage and distribution at all stages, fuel dispensing at the retail level, vehicle operation,
carbon in fuel from air, vehicle assembly and transport, and materials used in the vehicles. The
model includes pathways for ethanol and biodiesel production from various feedstocks.

GHGenius is capable of supporting a life cycle assessment of biofuels, and will be examined in
more detail in a later section.

5.2.22 GREET

Argonne National Laboratory, sponsored by the U.S. Department of Energy's Office of Energy
Efficiency and Renewable Energy, developed a life-cycle model called GREET (Greenhouse
gases, Regulated Emissions, and Energy use in Transportation).

The first version of GREET was released in 1996. Since then, Argonne has continued to update
and expand the model. The most recent GREET versions are GREET 1.8a version for fuel-cycle
analysis and GREET 2.8a version for vehicle-cycle analysis. The model is freely available over
the Internet. 55

Greet allows researchers and analysts to evaluate various vehicle and fuel combinations on a
fuel-cycle/vehicle-cycle basis. GREET includes more than 100 fuel production pathways for
energy feedstocks and more than 75 vehicle / fuel system configurations.

GREET takes into consideration various stages of corn, sugar cane, and cellulosic ethanol
production, as well as biodiesel from soybeans.


54
  http://www.ghgenius.ca/
55
  http://www.transportation.anl.gov/software/GREET/greet_1-8a_download_form.html
http://www.transportation.anl.gov/software/GREET/greet_2-8a_download_form.html


                                                   58
CHEMINFO




GREET is capable of supporting a life cycle assessment of biofuels, and will be examined in
more detail.

5.2.23 IVAM LCA

IVAM is an environmental research and consultancy institute affiliated to the University of
Amsterdam.

The IVAM LCA database is a compilation of several databases. It consists of over 1,300 unit
processes. The nomenclature of the database is adapted to the standard SimaPro database and the
Ecoinvent database for SimaPro. The database is extensive on the subjects of building and
construction sector, food production and waste management.

IVAM LCA is a database and not model focused on transportation fuels, and is not considered
further.

5.2.24 KCL-ECO

KCL is a privately-owned Finnish based research company serving the global paper and related
forest cluster industries as well as the end-users of paper and board products.

KCL EcoData is a continuously updated LCI database primarily intended for life-cycle inventory
calculations related to forest products. The model is specific to forest products, and does not
include capabilities for ethanol production.

KCL-ECO is not positioned to assess a biofuels life cycle, and it is not considered further.

5.2.25 LCAiT

THE LCAiT software has been developed by CIT Ekologik for the environmental assessment of
products and processes. 56

This model may no longer be available, and it is not considered further. 57

5.2.26 LCAPIX

The LCAPIX Module provides a stand alone software application, which can analyze processes
on a product basis, determine environmental load centres, and allow for development of a
comprehensive database. LCAPIX provides simple environmental comparisons of any product,


56
  http://ew.eea.europa.eu/Industry/Cleaner/Theme_2/b/Tools_for_analysis_and_evaluation/URL999072874/
57
  More information on LCAiT were not found, and the CIT Ekologik LCAiT website could not be accessed
(http://www.lcait.com).


                                                59
CHEMINFO




process or service. It is designed to help implement an effective environmental management
strategy (EMS), and ensure future efficiency and profitability.

A demo version can be downloaded free of charge. 58 A one time licensing fee of US$595 applies
for LCAPIX version 1.1 as a standalone application. A one time licensing fee of US$1995
applies for LCAPIX Client/Server Application.

An initial review of the LCAPIX manual suggests that this model is not built to address biofuels
life cycle impacts, and this model is not considered further. 59

5.2.27 LEM

The Lifecycle Emissions Model (LEM) was developed by Dr. Mark Delucchi.

The LEM model estimates energy use, criteria pollutant emissions, and CO2-equivalent
greenhouse-gas emissions from a variety of transportation and energy lifecycles. It includes a
wide range of modes of passenger and freight transport, electricity generation, heating and
cooking, and more. For transport modes, it represents the lifecycle of fuels, vehicles, materials,
and infrastructure. It energy use and all regulated air pollutants plus so-called greenhouse gases.
It includes input data for up to 20 countries, for the years 1970 to 2050, and is fully specified for
the U. S. 60

The model includes pathways for the production of biofuels (ethanol and biodiesel).

While LEM is not widely available, it appears to be capable of supporting a life cycle assessment
of biofuels, and will be examined in more detail in a later section.

5.2.28 MIET

MIET 3.0 is the successor of MIET 2.0, a Missing Inventory Estimation Tool for Life Cycle
Assessment (LCA).

MIET is a Microsoft Excel spreadsheet that enables LCA practitioners to estimate LCI of
missing flows that were truncated. MIET is based on U.S. input-output tables and environmental
data. MIET 3.0 includes 480 commodities and services, as well as about 1,200 different
environmental interventions including air, water, industrial and agricultural soil emissions, and
resource use by various industrial sectors.



58
   http://www.kmlmtd.com/demodld/index.html
59
   The user manual is available at http://www.kmlmtd.com/demodld/lcapix20_manual.pdf
60
   Delucchi, M. (2002), Overview of the Lifecycle Emissions Model
(http://www.its.ucdavis.edu/publications/2002/UCD-ITS-RR-02-02.pdf)


                                                    60
CHEMINFO




MIET 2.0 can be downloaded for free from the CML website after filling out a short
questionnaire. 61 MIET 3.0 is incorporated in the latest version of the SimaPro software of PRé
consultants and is available as stand-alone software package from Enviro Informatica under the
name CEDA 3.0 (Comprehensive Environmental Data Archive).

MIET 3.0 does not appear to have built-in capabilities to address the biofuels life cycle, 62 and is
not considered further.

5.2.29 PEMS

The PEMS software is no longer available, and is not considered further. 63

5.2.30 PIA

PIA, the Product Improvement Analysis, is sold by the Toegepaste Milieu Economie TME of
The Hague, Netherlands. A demonstration copy of the model is available at no cost over the
Internet. 64

PIA requires the user to identify and build processes, 65 and as such does not have built in
capabilities focusing on the biofuels life cycle, and is not considered further.

5.2.31 Regis

Regis is a company eco-audit software program, 66 and is not considered further.

5.2.32 SimaPro

SimaPro stands for “System for Integrated Environmental Assessment of Products”. 67 The
following inventory or LCA databases are included or available for SimaPro 7:

•    Ecoinvent v1.2              •   Danish I/O database          •   Data archive
•    ETH-ESU 96                  •   LCA food                     •   Dutch Concrete database
•    BUWAL 250                   •   Industry data                •   IVAM
•    Dutch I/O database          •   IDEMAT 2001                  •   FEFCO
•    US I/Odatabase              •   Franklin US LCI database     •   EuP Energy Products

The following impact assessment methods are included in all SimaPro versions:

61
   http://www.epa.gov/NRMRL/lcaccess/pdfs/appendices_lca101.pdf
62
   http://www.leidenuniv.nl/cml/ssp/software/miet/index.html
63
   E-mail correspondence with Pira International.
64
   http://www.tme.nu/english/index_uk.htm
65
   http://www.tme.nu/pdf/PIA-demo-instruction.pdf
66
   http://www.sinum.com/htdocs/e_software.shtml
67
   http://www.earthshift.com/simapro7.htm


                                                   61
CHEMINFO




•    Eco-indicator 99            •    EDIP/UMIP                    •   TRACI
•    Eco-indicator 95            •    EPS 2000                     •   Energy Demand
•    CML 92                      •    Ecopoints 97                 •   IPCC GHG Emission
•    CML 2 (2000)                •    Impact 2002+

SimaPro does not appear to have a built-in biofuels assessment module, however, this model will
be reviewed in further detail in a later section given the extensive databases and impact
assessment methods in contains.

5.2.33 SPINE

SPINE (Sustainable Product Information Network for the Environment) is the Swedish National
LCA database. It was developed by the Swedish Competence Centre for Environmental
Assessment of Product and Material System.

The database aims to develop a comprehensive knowledge base to support the industry needs in
the field of environmental assessment. The database includes Swedish data on: freight transport,
electricity, heat, fuels, chemicals, materials, etc. 68 The public version of the database contains
more than 500 data sets. The database can be accessed through their data store on the Internet
where data can be purchased. 69 The data in the database has been classified into three price
categories depending on the degree of documentation and level of aggregation at a range of 125
to 1,000 SEK (C$20 to C$160). 70

SPINE is a database and not a model for estimated the life cycle impacts of biofuels, and is not
considered further.

5.2.34 TEAM

TEAM (Tools for Environmental Management and Analysis) is Ecobilan’s Life Cycle
Assessment software. TEAM allows the user to build and use a large database and to model any
system representing the operations associated with products, processes and activities. 71

Several versions of TEAM are available. TEAM 4.0 discovery version is freely available as a
demonstration version on the Internet. An initial license costs 3,000 Euros (C$4,500).

A review of TEAM’s online demonstration suggests that the model 72 does not have built-in
capabilities for biofuels assessment, and as such TEAM is not considered further.

68
   http://buildlca.rmit.edu.au/downloads/BACKGROUNDREPORTFINAL.PDF
69
   http://deville.tep.chalmers.se/commdb/DbStart.htm
70
   http://faculty.washington.edu/cooperjs/Research/database%20projects.htm
71
   http://www.ecobalance.com/uk_team.php
72
   http://www.ecobalance.com/uk_team02.php


                                                    62
CHEMINFO




5.2.35 TRACI

The U.S. EPA developed TRACI, the Tool for the Reduction and Assessment of Chemical and
other environmental Impacts.

TRACI allows the examination of the potential for impacts associated with the raw material
usage and chemical releases resulting from the processes involved in producing a product.
TRACI allows the user to examine the potential for impacts for a single life cycle stage, or the
whole life cycle, and to compare the results between products or processes.

The purpose of TRACI is to allow a determination of priorities or a preliminary comparison of
two or more options on the basis of the following environmental impact categories: ozone
depletion, global warming, acidification, eutrophication, photochemical smog, human health
cancer, human health noncancer, human health criteria, ecotoxicity, fossil fuel use, land use, and
water use.

TRACI does not appear to have built-in capabilities for biofuels, and as such is not considered
further.

5.2.36 UMBERTO

UMBERTO was developed in cooperation between the Institut für Umweltinformatik Hamburg
and the Institut für Energie- und Umweltforschung Heidelberg. 73 The first version was released
in 1995 and version 5.0 is now available. Umberto Consult, the full version, costs 9,900 Euros
(C$15,000). 74 A 30-day trial version and other lower cost versions are also available.

Umberto helps prepare life cycle assessments. This is done through the creation of individual
projects. Each project is characterized by a freely definable and expandable list of products, raw
materials, pollutants, and forms of energy etc. Life cycle impacts are estimated based on these
projects. To facilitate this assessment, Umberto has a supporting module library that contains
data sets on generic upstream and downstream processes (including Ecoinvent data).

Umberto does not have built-in biofuels life cycle assessment modules, and as such is not
considered further.




73
     http://www.eco-shop.org/Resources/lcasoftwarereREVIEW.pdf
74
     http://www.umberto.de/en/product/prices/index.htm


                                                     63
CHEMINFO




5.2.37 WISARD

Ecobilan’s Waste Management Life Cycle Assessment software tool (WISARD) is a program to
assist the decision maker to evaluate alternative waste management scenarios. 75

This tool focus on alternative forms of waste management (including landfilling, incineration,
sorting/recycling, composting and anaerobic digestion), does not appear to provide life cycle
assessments for biofuels, and is not considered further.




75
     http://www.ecobalance.com/uk_wisard.php


                                               64
CHEMINFO




           6. Detailed Assessment of LCA Models

6.1 Overview

This section presents an evaluation of the nine models identified as potentially offering the best
options for assessing the bioethanol life cycle. The models are each described, and then
evaluated on the basis of ten evaluation criteria.

Also include are “guidelines” for scoring the models. These guidelines are to help focus on an
appropriate score but not to be overly prescriptive. The intention here is to broadly score models
higher if they have desirable attributes for the purpose of conducting a sensitivity analysis. In
this respect, model availability and affordability are important criteria.

1. Model Availability - To what degree is the model available to conduct sensitivity analysis.

   0 means the model is not available, 5 means parts of the model are available for
   demonstration purposes, and 10 means the model is widely available.

2. Model Affordability - To what degree is the model affordable for the purposes of conducting
   a sensitivity analysis.

   0 means the model is prohibitively expensive (ex., $25,000), 5 means parts of the model is
   expensive (ex., $5,000), and 10 means the model is free.

3. Ethanol Production Pathways - To what degree does the model have existing capabilities for
   estimating the impacts from various ethanol pathways.

   0 means the model doesn’t have a built-in ethanol pathway, 3 means it has one built-in
   pathway, 6 means it has 2 built-in pathways, 9 means it has 3 built-in pathways, and 10
   means it has more than 3 built-in pathways.

4. Life Cycle Stages - To what degree does the model cover all stages of the ethanol life cycle
   stages (raw material acquisition, fuel manufacturing, fuel distribution, fuel use, vehicle
   production, waste management, etc.).

   0 means the model has limited coverage (ex., only transportation emissions), and 10 means
   the model covers all stages.




                                               65
CHEMINFO




5. Input Flexibility - To what degree does the model have sophistication to address individual
   sources of environmental impacts.

   0 means the model has limited coverage of individual sources of environmental impacts, and
   10 means the model has extensive coverage of sources (such as including and allowing for
   variability in fertilizer application rates, distances that agricultural inputs are shipped to the
   plant, types of vehicles driven, etc.)

6. Environmental Media - To what degree does the model cover impacts to all receiving media
   (i.e., air, water, and land).

   0 means the model has limited coverage and 10 means the model covers air, water, and land.

7. Environmental Impact Assessment - To what degree does the model cover releases of
   environmental pollutants and uses of resources.

   0 means the model has limited coverage and 10 means the model covers extensive pollutants
   (with 2 points for covering each of releases of greenhouse gases, air pollutants, use of and
   releases to water, use of and releases to land, and energy use).

8. Transparency - To what degree are the data used by the model and calculation mechanisms
   of the model transparent.

   0 means the model has limited transparency and 10 means the model describes in detail data
   sources, boundaries, emission factors, methods, assumptions, etc.

9. Canadian Data - To what degree is the model available with Canadian data.

   0 means the model has no North American data, 5 means it contains North American data,
   and 10 means that the model is based entirely on Canadian data.

10. Other Features - To what extent are other attractive features available to support the model.

   0 means the model has no additional features, and 10 means that the model includes
   numerous features (such as emissions and energy metrics in output, forecasting, simulation,
   advanced graphical reporting, economic impact estimation, etc.).




                                                66
CHEMINFO




The models were evaluated based on a review of materials available over the Internet, and
telephone consultations with individuals familiar with the model. In some cases uncertainty
exists over aspects of some models, for example their sophistication in terms of input flexibility.
In these cases, a scoring was assigned based on expectations. Any uncertainty is not expected to
significantly affect the scoring or the order that the relevant models are ranked. Furthermore, the
intention of this exercise was to use a rationale approach for ranking models as suitable
candidates for the sensitivity analysis under the context of this study, i.e. not to declare that
“Model A scores 86.1 and “Model B scores 85.9”.


6.2 Evaluations of Models for the Bioethanol Life Cycle

6.2.1 BEES

Overview

The BEES (Building for Environmental and Economic Sustainability) software offers a
technique for selecting cost-effective, environmentally-preferable building products. 76 Version
4.0 of the Windows-based decision support software includes environmental and economic
performance data for 230 building products. It is freely available over the Internet. 77

BEES measures the environmental performance of building products by using the life-cycle
assessment approach specified in the ISO 14040 series of standards. All stages in the life of a
product are analyzed: raw material acquisition, manufacture, transportation, installation, use, and
recycling and waste management. Economic performance covers the costs of initial investment,
replacement, operation, maintenance and repair, and disposal. Environmental and economic
performances are combined into a single measure. The BEES model does not appear to have the
existing, built-in capabilities to address the ethanol life cycle.




76
   BEES Please for USDA has a cost of US$8,000 for the first product and US$4,000 for subsequent products
(http://www.bfrl.nist.gov/oae/software/bees/please/USDA/bees_please.html).
77
   http://www.bfrl.nist.gov/oae/software/bees/registration.html


                                                  67
CHEMINFO




In support of the 2002 Farm Security and Rural Investment Act (P.L. 107-171), BEES was
adapted for application to bio-based products. 78 The BEES Environmental Performance Score
combines product performance across all 12 environmental impacts into a single score. These
impacts are:

•    global warming                •     indoor air quality            •   human health
•    acidification                 •     habitat alteration            •   smog
•    eutrophication                •     water intake                  •   ozone depletion
•    fossil fuel depletion         •     criteria air pollutants       •   ecological toxicity

The BEES Environmental Performance Score indicates the share of annual per capita U.S.
environmental impacts attributable to the product. A score of 0.0130, for example, means the
production and consumption of a unit of the product is estimated to represent 0.0130 percent of
average annual U.S. per capita contributions to environmental impacts.

Evaluation

BEES scores 53 out of 100. Its relatively low score is largely the result of not having a built-in
ethanol pathway, not addressing different ethanol life cycle stages, and not having capabilities to
assess specific impact sources.


                                    Table 8: BEES Evaluation

                         Score
     Criteria                                                      Justification
                      (Unweighted)

Availability                        10    BEES is readily available
Cost                                10    Model is free
Pathways                             0    The model has no ethanol pathways
Stages                               0    The model looks at no ethanol life cycle stages
Flexibility                          0    The model looks at no specific ethanol impact sources
Media                                9    Model addresses all media
Pollutants                           8    Model has good coverage across pollutants
Transparency                         7    The model has some documentation
Data                                 5    The model is based on US and European data
Features                             4    The model has limited additional features

Total                               53



78
  BEES was adapted for application to bio-based products (called BEES for USDA). Efforts were made to contact
BEES staff to understand capabilities with respect to the biofuels life cycle. No additional information has been
provided, and as such the publicly available version of BEES was reviewed.


                                                      68
CHEMINFO




6.2.2 BESS

Overview

The BESS model is a software tool to calculate the energy efficiency, greenhouse gas (GHG)
emissions, and natural resource requirements of biofuel production systems.

The latest version of BESS (v. 2007.1.0) is freely available over the Internet for non-commercial
uses. 79,80

The present version of the model includes a corn-to-ethanol pathway. The developers suggest
that the model will be extended to cover ethanol from corn stover and switchgrass for cellulosic
ethanol in the future.

The model provides a “cradle-to-grave” analysis of the production life-cycle of biofuels from the
creation of material inputs to finished products. The model covers the production stage of the life
cycle only, and as such other stages (such as fuel distribution and use) are not included. The
model is populated with U.S. (average) data, and a regional analysis based on north-eastern U.S.
coal or natural gas inputs can be conducted.

Specific stages of the biofuels life cycle that can be modeled include: 81

•    crop production- using inputs for productivity, carbon sequestration, fertilizers, herbicides,
     and insecticides, water use, and fuel consumption;
•    biorefinery - using inputs for ethanol production, conversion rates, water use, co-product use,
     and energy (thermal and electricity) use; and
•    cattle feedlot - using inputs for transportation of co-products, cattle performance, and cattle
     diet.

The model provides outputs in terms of:

•    energy use;
•    greenhouse gas emissions - CO2, CH4, and N2O; and
•    environmental requirements - land use, grain consumption, and water use.




79
   http://www.bess.unl.edu/download/
80
   Several e-mail requests to the University of Nebraska - Lincoln requesting the price of the commercial version
have not been returned.
81
   University of Nebraska Lincoln (2007), BESS Biofuel Energy Systems Simulator Users Guide.


                                                      69
CHEMINFO




Evaluation

BESS scores 64 out of 100. Its score is influenced by the fact that there is a cost for the
commercial version, the model has one biofuels pathway, it does not include the full fuels life
cycle, and the model has limited coverage of pollutants.


                                       Table 9: BESS Evaluation

                            Score
       Criteria                                                   Justification
                         (Unweighted)

Availability                           10   BESS is readily available
Cost                                    5   There is a cost for the commercial version of BESS
Pathways                                3   The model has one built-in ethanol pathway
Stages                                  6   The model omits fuel distribution and use
Flexibility                             7   BESS offers flexibility in adjusting inputs
Media                                   8   Model addresses air, water, and land
Pollutants                              7   BESS covers three GHGs
Transparency                            8   The model appears reasonably well documented
Data                                    5   The model is based on US data
Features                                5   BESS has some additional features (ex., scenarios)

Total                                  64
Note: Several requests to obtain the price of the commercial version of BESS were made, but no response was
received. As such a score of 5 is assumed.

6.2.3 EIO-LCA

Overview

The Economic Input Output-Life Cycle Assessment (EIO-LCA) model traces out the various
economic transactions, resource requirements and environmental emissions require for a
particular product or service. The model is available for use over the Internet at no cost. 82




82
     The model is found at http://www.eiolca.net/use.html


                                                            70
CHEMINFO




The model provides estimates of the overall environmental impacts from producing a certain
dollar amount of any of 500 commodities or services in the United States. It provides rough
guidance on the relative impacts of different types of products, materials, services, or industries
with respect to resource use and emissions throughout the U.S.

The EIO-LCA model is based on the macro-economic principles of input-output analysis. The
model can be visualized as a set of large tables that represent the effects on other sectors as a
result of increased activity in another sector. Effects can be modeled as:

•   economic activity (direct economic impact, value added, and total economic impact);
•   energy (electricity, coal, natural gas, LPG, gasoline, distillate, kerosene, and jet fuel);
•   conventional air pollutants (SO2, CO, NOx, VOCs, lead, PM10);
•   greenhouse gases (CO2, CH4, N2O, CFCs, CO2E);
•   toxic releases (non-point, point to air, water, land, underground, and transfers); and
•   employment (employees).

Note that these impacts are for the production of goods, and not for use or disposal. As such,
there appear to be no means for estimating emissions from use of fuels for transportation.

The input/output matrix is the 1997 commodity/commodity input-output (IO) matrix of the US
economy as developed by the US Department of Commerce. Energy use comes from a variety of
sources, such as the 1998 Manufacturing Energy Consumption Survey (MECS) from the Energy
Information Administration and the 2000 Transportation Energy Data Book of the Department of
Energy.

Emissions data come from sources such as the U.S. EPA AIRS web site and the US EPA's 1995
toxics release inventory (TRI). Greenhouse Gas Emissions are calculated by emissions factors
from fuel use using U.S. EPA AP-42 emissions factors.

One of the approximately 500 commodities and services included in the model is “Other Basic
Chemical Manufacturing” (32519). This diverse grouping includes:

•   Gum and Wood Chemical Manufacturing;
•   Cyclic Crude and Intermediate Manufacturing;
•   Ethyl Alcohol Manufacturing; and
•   All Other Basic Organic Chemical Manufacturing.




                                                  71
CHEMINFO




“Ethyl Alcohol Manufacturing” is not disaggregated in any manner, including differentiation
based on feedstocks (ex., corn versus wheat). In addition, estimates of economic or
environmental implications of “Ethyl Alcohol Manufacturing” cannot be separated from “Other
Basic Chemical Manufacturing”.

Evaluation

EIO-LCA scores 59 of 100. It scored relatively low in terms of having existing ethanol
pathways, looking at the full life cycle, and flexibility.


                          Table 10: EIO-LCA Evaluation

                    Score
    Criteria                                            Justification
                 (Unweighted)

Availability                 10   Model is readily available
Cost                         10   Model is free
Pathways                      2   The model has a limited ethanol pathway
Stages                        3   The model focuses on the production stage
Flexibility                   0   The model looks at no specific ethanol impact sources
Media                         8   Model addresses all media
Pollutants                    8   Model has good coverage across pollutants
Transparency                  8   The model is well documented
Data                          5   The model is based on US and European data
Features                      5   The model has macro-economic capabilities

Total                        59




                                             72
CHEMINFO




6.2.4 GaBi

Overview

The software system GaBi is a tool for life-cycle assessments. GaBi The GaBi software is based
on a modular concept. By this means, plans, processes and flows and their functionalities
establish modular units. 83 These modular units must be built by the user, or purchased from the
vendor. GaBi (version 4.0) is commercially available from PE International with several package
options.

The price of the software including a database varies between about C$4,000 for the Lite
Version and C$10,000 for the Professional Version. A Demo Version is available for download
(with the capabilities for modeling the life cycle impacts of a photocopier).

The Professional Version comes with an extensive database, and the option to purchase
additional databases (the option to purchase additional databases is not available with the Lite
Version). The core database includes German, European, and North American data.

The model does not come with a built-in module for estimating the life cycle impacts of ethanol.
However, Extension Database XII (Renewable Raw Materials) comes with the data to model the
ethanol life cycle (from different feedstocks, including corn and wheat) and based on different
process elements. This database is based on German data. 84 None of the extension databases
provide information on the biodiesel life cycle. The reported price for Extension Database XII is
approximately C$3,000. 85

The model is capable of addressing a wide range of air pollutants, greenhouse gases, and land
use impacts across all stages of the life cycle. The model reportedly comes with extensive
documentation, though training is highly recommended. 86




83
   http://www.gabi-software.com/gabi/gabi-4/aboutgabi41
84
   Telephone consultation with PE International, January 16, 2008.
85
   http://www.gabi-software.com/gabi/databases1/extensiondatabases0
86
   Telephone consultation with PE International, January 16, 2008.


                                                     73
CHEMINFO




Evaluation

GaBi scores 65 of 100. The model scores lower in terms of cost, unavailability of built ethanol
pathways, and limited Canadian data.


                                    Table 11: GaBi Evaluation

                         Score
    Criteria                                                      Justification
                      (Unweighted)

Availability                        10    GaBi is commercially available
Cost                                 4    The cost would be approximately C$13,0001
Pathways                             0    Pathways can be built upon the available ethanol data
Stages                               8    GaBi considers most lifecycle stages
Flexibility                          7    The model can address different impact sources
Media                                8    The model provides good coverage across media
Pollutants                           9    The model has broad coverage of pollutants
Transparency                         8    GaBi is reportedly relatively transparent
Data                                 3    The data (for ethanol) is largely European
Features                             8    GaBi offers advanced features (like balancing, indicators)

Total                               65
1
     This estimate includes C$10,000 for the GaBi model and C$3,000 for Extension Database XII.
* Note that there are some data gaps with respect to the capabilities of GaBi in modeling the biofuels pathways,
life cycle stages, flexibility and pollutants. Scores have been inferred based on publicly available materials and
correspondence with the vendor.

6.2.5 GEMIS

Overview

GEMIS is the Global Emission Model for Integrated Systems. GEMIS is a life-cycle analysis
program and database for energy, material, and transport systems. GEMIS (v 4.4) is available at
no cost over the Internet.




                                                       74
CHEMINFO




The GEMIS model and database includes information on nearly 9,000 activities and processes.
The model is capable of estimating:

•   emissions to air for a wide range of pollutants (ex., SO2, NOx, particulates, CO, VOCs, NH3,
    PCDDs/PCDFs, metals, etc.);
•   greenhouse gases (CO2, CH4, and N2O);
•   residues and liquid effluents; and
•   energy use.

The GEMIS database offers information on:

•   fossil fuels (hard coal, lignite, natural gas, oil), renewables, nuclear, biomass (residuals, and
    wood from short-rotation forestry, Miscanthus, rape oil etc) and hydrogen (including fuel
    composition, and upstream data)
•   processes for electricity and heat (various powerplants, cogenerators, fuel cells, etc.)
•   materials: raw and base materials, and especially those for construction, and auxiliaries
    (including upstream processes)
•   transports: airplanes, bicycles, buses, cars, pipelines, ships, trains, trucks (for diesel,
    gasoline, electricity, and biofuels).

Some datapoints are available for ethanol from wheat, maize, sugar beet, sugar cane, and other
feeds (ex., sorghum, straw, etc.). Process data is available for fermentation, filling,
transportation, and co-generation. There appear to be significant differences in terms of the data
that are available for different process stages and feedstocks, and these differences are not
readily transparent.

The GEMIS database has been developed based on studies and reports, largely from Europe. A
small fraction (4 of 240) of the processes for ethanol are based on North American data. About
30% of the data is identified as “good” quality or better.

GEMIS offers a number of positive features, including advanced data filter, assessments of data
quality, process chain diagrams, and scenario modelling. The User Manual (originally written in
German) is ineffective, and there is no help available.

Evaluation

GEMIS scores 65 of 100. The model scores relatively low in terms of existing pathways,
coverage across media and pollutants (for ethanol processes), has limited Canadian data, and is
deficient in terms of support.




                                                75
CHEMINFO




                                  Table 12: GEMIS Evaluation

                         Score
    Criteria                                                       Justification
                      (Unweighted)

Availability                         10   GEMIS is available over the Internet
Cost                                 10   There is no charge to obtain and use the model
Pathways                              5   GEMIS appears to have some built pathways
Stages                                7   The model tends to cover most life cycle stages.
Flexibility                           5   The model is limited in the number of impact categories
Media                                 8   The model covers all media
Pollutants                            7   The model covers a wide range of pollutants
Transparency                          5   The model and user manual are not readily transparent
Data                                  3   Most of the ethanol data is largely European
Features                              5   The model has some features (but no support)

Total                                65
Note: The extent that GEMIS includes built-in pathways is not known given the limited documentation on the
model and unavailability of a help desk. A review of the user manual suggests that there are at least some complete
biofuels process chains. As such, a scoring of 5 is assumed for pathways.

6.2.6 GHGenius

Overview

GHGenius focuses on the life cycle assessment (LCA) of current and future fuels for
transportation applications. GHGenius (Version 3.11) is available at no cost over the Internet.

All of the steps in the life cycle are included in the model from raw material acquisition to end-
use. The model is capable of analyzing the emissions from conventional and alternative fuelled
internal combustion engines for light duty vehicles, for class 8 heavy-duty trucks, for urban
buses and for a combination of buses and trucks, for light duty battery powered electric vehicles,
and for fuel cell vehicles. There are currently approximately 200 vehicle, fuel and feedstock
combinations (pathways) possible with the model.




                                                       76
CHEMINFO




The model is capable of addressing:

•   ethanol from wood, grass, corn, sugar cane, sugar beets, and wheat; and
•   biodiesel from soybeans, canola, palm, tallow, yellow grease, and marine oils.

GHGenius focuses on estimating life cycle emissions for three impact categories; the primary
greenhouse gases, the criteria pollutants from combustion sources and the energy used. Like the
other models, GHGenius tends to focus on releases to air. The specific categories that are in the
model include:

•   Greenhouse Gases - carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O),
    chlorofluorocarbons (CFC-12), and hydrofluorocarbons (HFC-134a);
•   Other Air Contaminants - carbon monoxide (CO), nitrogen oxides (NOx), non-methane
    organic compounds (NMOCs), sulphur dioxide (SO2), and total particulate matter; and
•   Energy Use - total energy used per unit of energy produced for each stage of the fuel
    production steps, total fossil energy used per unit of energy produced for each stage of the
    fuel production steps, energy used per kilometre driven for the fuel used in light duty internal
    combustion engines, light duty fuel cell vehicles, heavy duty internal combustion engines,
    and heavy duty fuel cell vehicles, and the proportions of types of energy used for each stage
    of the fuel production cycle.

GHGenius can perform the LCA for specific regions (east, central or west) of Canada, the United
States and Mexico or for India as a whole. For Canada, it is also possible to model many of the
processes for the largest provinces. It is also possible for model regions of North America.
GHGenius can also predict emissions for past, present and future years through to 2050 using
historical data or correlations for changes in energy and process parameters with time that are
stored in the model.

GHGenius has data for Canada, the United States, Mexico and India for many of the steps in the
various fuel processes and it allows the user to provide data for some steps in the process. For
Canada, reports produced by Statistics Canada, Natural Resources Canada, Environment Canada
and the National Energy Board have been used as data sources. Industry associations such as the
Canadian Association of Petroleum Producers (CAPP) and the Canadian Gas Association (CGA)
have also been used as sources of data. The non-energy related process emissions in the model
are calculated based mostly on the US EPA AP-42 emission factors. The emissions from
vehicles for conventional fuels are derived from the Environment Canada model Mobile 6.2C.




                                                77
CHEMINFO




Evaluation

GHGenius scores 89 of 100. It scores relatively well for each of the evaluation criterion, but does
not cover releases to water and land.


                                 Table 13: GHGenius Evaluation

                           Score
       Criteria                                                    Justification
                        (Unweighted)

Availability                         10   GHGenius is available over the Internet
Cost                                 10   There is no charge to obtain and use the model
Pathways                             10   GHGenius has extensive fuel/vehicle pathways
Stages                                9   The model covers most life cycle stages.
Flexibility                           9   GHGenius has flexibility to address individual impacts
Media                                 7   The model focuses on releases to air
Pollutants                            6   The model covers GHGs, air pollutants, and energy
Transparency                          9   The model and data sources are well documented
Data                                 10   There is extensive Canadian data in the model
Features                              9   The model has strong non-core features

Total                                89


6.2.7 GREET

Overview

The GREET (Greenhouse gases, Regulated Emissions, and Energy use in Transportation) model
was developed by Argonne National Laboratory under the sponsorship of the U.S. Department of
Energy’s Office of Energy Efficiency and Renewable Energy. GREET allows researchers and
analysts to evaluate various vehicle and fuel combinations on a full fuel-cycle/vehicle-cycle
basis.

The first version of GREET was released in 1996. The most recent GREET versions are:

•      GREET 1.8a for fuel-cycle analysis; and
•      GREET 2.8a for vehicle-cycle analysis.

Both versions of the model are available free over the Internet as spreadsheet models in
Microsoft Excel. 87

87
     http://www.transportation.anl.gov/software/GREET/index.html


                                                       78
CHEMINFO




The model covers all stages of the fuel life cycle, from well-to-pump and pump-to-wheels,
including:

•   feedstock production, transportation, and storage;
•   fuel production, transportation, distribution, and storage,
•   vehicle operation, refuelling, fuel combustion/conversion, fuel evaporation, and tire/break
    wear.

In addition, GREET simulates vehicle-cycle energy use and emissions from material recovery to
vehicle disposal (raw material recovery, material processing and fabrication, vehicle component
production, vehicle assembly, and vehicle disposal and recycling).

The model includes:

•   emissions of greenhouse gases (CO2, CH4, and N2O);
•   emissions of six criteria pollutants (VOCs, CO, NOx, SOx, PM10, andPM2.5); and
•   energy use by fuel.

GREET includes more than 100 fuel production pathways and more than 70 vehicle/fuel
systems. The model includes pathways for ethanol from corn and cellulosic biomass and
biodiesel from soybeans. GREET considers such factors as agricultural chemical production and
transportation, corning farming, use of animal feed, and ethanol transportation and blending.

Data in the model are for the United States




                                              79
CHEMINFO




Evaluation

GREET scores 75 of 100. The model scores reasonably well for nearly all criteria, with the
exception of having limited Canadian data and focusing on releases to air.


                            Table 14: GREET Evaluation

                     Score
    Criteria                                             Justification
                  (Unweighted)

Availability                  10   The GREET model is available over the Internet
Cost                          10   There is no cost for the model
Pathways                       6   The model has several pathways for ethanol
Stages                         8   The model covers all the life cycle stages
Flexibility                    8   The model is flexible to address individual impact sources
Media                          7   GREET model focuses on releases to air
Pollutants                     6   The model covers GHGs, air pollutants, and energy
Transparency                   7   Data sources are not always explicit
Data                           5   GREET is based on US data.
Features                       8   Some additional features (like forecasting and graphing)

Total                         75


6.2.8 LEM

Overview

The Lifecycle Emissions Model (LEM) estimates energy use, criteria pollutant emissions, and
CO2-equivalent greenhouse-gas emissions from a variety of transportation and energy lifecycles.
The model was developed by Dr. Mark Delucchi of the University of California, Davis. The
LEM model is not publicly available.




                                               80
CHEMINFO




For motor vehicles, the LEM calculates lifecycle emissions for a variety of combinations of end-
use fuel, fuel feedstocks, and vehicle types. The model considers the following life cycle
stages: 88

•      lifecycle of fuels - feedstock production, feedstock transport, fuel production, fuel
       distribution and storage, dispensing of fuels, and fuel use;
•      lifecycle of vehicles - materials use, vehicle assembly, operation and maintenance, and the
       secondary fuel cycle; and
•      lifecycle of infrastructure - energy use and materials production.

The model includes eight classes of passenger vehicles and four freight transport modes. The
model includes gasoline, diesel, biodiesel, methanol, ethanol, methane, propane, hydrogen, and
electricity. Lifecycle impacts can be estimated for:

•      corn to ethanol;
•      switchgrass to ethanol;
•      wood to ethanol; and
•      soybeans to biodiesel.

The LEM estimates emissions of the following pollutants: carbon dioxide, sulphur dioxide,
methane, total particulate matter, nitrous oxide, particulate matter less than 10 microns, carbon
monoxide, chlorofluorocarbons, nitrogen oxides, hydrofluorocarbons, and nonmethane organic
compounds.

It includes input data for up to 20 countries, for the years 1970 to 2050, and is fully specified for
the U. S.

Evaluation

LEM scores 59 of 100. It scored low in terms of availability (and subsequently cost), Canadian
data, and its focus on releases to air.




88
     http://www.its.ucdavis.edu/publications/2002/UCD-ITS-RR-02-02.pdf


                                                      81
CHEMINFO




                                      Table 15: LEM Evaluation

                           Score
       Criteria                                                   Justification
                        (Unweighted)

Availability                            0    The LEM model is not available
Cost                                    0    The LEM model is not available
Pathways                                9    The model has extensive pathways
Stages                                  8    The model considers most stages of the life cycle
Flexibility                             8    The model is flexible to address individual impacts
Media                                   7    The LEM model focuses on emissions to air
Pollutants                              6    The LEM model has good coverage cross pollutants
Transparency                            8    The model is relatively transparent
Data                                    5    The LEM model includes U.S. and international data
Features                                8    Uncertain - Some scenario modeling, mapping, graphing

Total                                 59


6.2.9 SimaPro

Overview

SimaPro (version 7.1) provides a tool to collect, analyze, and monitor the environmental
performance of products and services. 89

The family of SimaPro products is commercially available from PRé Consultants. It comes in the
3 professional versions, that are available as stand alone or multi user network versions.

•      SimaPro Compact (~C$6,500) - The Compact version has all the functionality to perform a
       LCA study.

•      SimaPro Analyst (~C$10,000) - The Analyst version comes with advanced analytical
       features including parameter based scenario analysis and Monte Carlo analysis.

•      SimaPro Developer (~C$14,500)- The Developer version has advanced modeling capabilities
       (like direct Excel/ASP linking, COM interface, EcoSpold compatibility).




89
     http://www.pre.nl/simapro/default.htm


                                                        82
CHEMINFO




SimaPro (v. 7.1) comes inclusive of several inventory databases with thousands of processes
including: Ecoinvent, US Input Output database, Danish Input Output database, Dutch Input
Output database, LCA food database, ETH-ESU 96, BUWAL 250, IDEMAT 2001, Franklin US
LCI database, and some others. In addition, SimaPro comes with a range of impact assessment
methods, namely Eco-indicator 99, Eco-indicator 95, CML 92, CML 2 (2000), EDIP/UMIP, EPS
2000, Ecopoints 97, Impact 2002+, TRACI, EPD method, Cumulative Energy Demand, and
IPCC Greenhouse gas emissions.

The SimaPro products are based on developing a life cycle inventory and then conducting an
impact assessment. SimaPro does not have the built-in capabilities to conduct a ethanol life cycle
analysis. Rather, the product would have to be used to build up a life cycle inventory and then
conduct an impact assessment. 90 Some studies were found providing examples of these
capabilities. 91

Evaluation

SimaPro scores 64 of 100. The model scores low in terms of having no existing ethanol
pathways and Canadian data.


                     Table 16: SimaPro (v. 7.1 Analyst) Evaluation

                          Score
       Criteria                                                   Justification
                       (Unweighted)

Availability                        10    SimaPro is commercially available/
Cost                                 5    The Analyst version costs about C$10,000
Pathways                             0    The model has processes, but no built in pathways
Stages                               8    The model can consider the entire life cycle
Flexibility                          6    Uncertain
Media                                8    The model can cover all media
Pollutants                           9    The model can address a wide range of pollutants
Transparency                         7    The model is relatively transparent
Data                                 3    SimaPro has mainly international and U.S. data
Features                             8    Has advanced features (scenario / Monte Carlo analysis)

Total                               64
* Note that there are some data gaps with respect to the capabilities of GaBi in modeling the biofuels pathways,
life cycle stages, flexibility and pollutants. Scores have been inferred based on publicly available materials and
correspondence with the vendor.




90
     www.pre.nl/download/manuals/SimaPro7IntroductionToLCA.pdf
91
     www.infra.kth.se/fms/utbildning/lca/projects%202006/Group%2010%20(Biofuels%20in%20cars).pdf


                                                       83
CHEMINFO




6.3 Summary of the Model Evaluations for Ethanol

The exhibit on the following pages summarizes the model evaluations based on two scenarios:

•    Un-Weighted Scenario - No weights are applied to the evaluation criteria. Scores are simply
     added across each of the ten criterion to yield a total with a maximum of 100.

•    Weighted Scenario - The evaluation criteria are each weighted by a unique factor between 1
     and 10. 92 This means that some of the criteria have significantly higher weights than others.
     Scores are added across the weighted criteria to yield a total with a maximum of 550 (which
     are subsequently normalized to a total of 100).

For transparency and simplicity, the un-weighted scenario is recommended. Two models rank
significantly higher than the others under this scenario and indeed under the weighted scenario
as well. These models are GHGenius and GREET, both of which are available at no cost, and
have numerous built bioethanol pathways. As such it is recommended that these two models be
the focus of sensitivity analysis of the bioethanol models.

The GaBi and SimaPro LCA models have excellent databases (although mostly with European
data), a wide variety of environmental parameters, and are quite competent and flexible models.
This flexibility is both an advantage and a disadvantage. On the positive side, the flexibility
allows the skilled user to construct a very wide range of process configurations and thus
accurately model almost any energy or materials system imaginable. The disadvantage is that the
resulting output is very dependent on the skill of the user and how well the process chain is
constructed by the user. This has implications for comparison of results since two users may not
construct exactly the same process chain for the same system and thus have effectively different
system boundaries and could potentially produce very different results using the same model.

The GREET and GHGenius models, on the other hand, have been tailored to specific
transportation fuel pathways and have less flexibility for creating new process chains (unless the
user is very experienced). The advantage of these models is that less skilled users can utilize
them to change the primary inputs for the pre-determined process chains and determine the
impact of those changes. These models are also focused primarily on energy balances, GHG
emissions, and CAC emissions and thus have a narrower range of outputs than the commercial
models described above.



92
  These weightings can be changed, and any changes may affect the score obtained by each model and the ranking
of models.




                                                     84
       CHEMINFO




                        Table 17: Ethanol Model Evaluation Scores (Un-Weighted Scenario)

                                Max
         Criteria      Weight                BEES        BESS        EIO-LCA      GaBi        GEMIS GHGenius GREET       LEM        SimaPro
                                Score

   Model
 1
   Availability          1.00           10          10          10           10          10       10      10     10             0         10
   Model
 2
   Cost                  1.00           10          10           5           10           4       10     10      10             0          5
   Ethanol
 3
   Pathways              1.00           10           0           3            2           0        5     10          6          9          0
   Life Cycle
 4
   Stages                1.00           10           0           6            3           8        7      9          8          8          8
   Impact
 5
   Sources               1.00           10           0           7            0           7        5      9          8          8         6
   Environmental
 6
   Media                 1.00           10           9           8            8           8        8      7          7          7          8
   Environmental
 7
   Impact Assessment     1.00           10           8           7            8           9        7      6          6          6          9
   Transparency
 8
                         1.00           10           7           8            8           8        5      9          7          8          7
   Canadian
 9
   Data                  1.00           10           5           5            5           3        3     10          5          5          3
   Other
10
   Features              1.00           10           4           5            5           8        5      9          8         8          8

     Score                          100             53          64           59          65       65     89      75            59         64




                                                                        85
       CHEMINFO




                          Table 18: Ethanol Model Evaluation Scores (Weighted Scenario)
                                Max
         Criteria      Weight                BEES        BESS        EIO-LCA      GaBi        GEMIS GHGenius GREET       LEM        SimaPro
                                Score

   Model
 1
   Availability            10       100         100         100          100         100         100     100    100             0        100
   Model
 2
   Cost                     9           90          90          45           90          36       90     90      90             0         45
   Ethanol
 3
   Pathways                 7           70           0          21           14           0       35      70     42            63          0
   Life Cycle
 4
   Stages                   3           30           0          18            9          24       21     27      24            24         24
   Impact
 5
   Sources                  2           20           0          14            0          14       10     18      16            16         12
   Environmental
 6
   Media                    5           50          45          40           40          40       40     35      35            35         40
   Environmental
 7
   Impact Assessment        4           40          32          28           32          36       28      24     24            24         36
   Transparency
 8
                            6           60          42          48           48          48       30     54      42            48         42
   Canadian
 9
   Data                     8           80          40          40           40          24       24     80      40            40         24
   Other
10
   Features                 1           10           4           5            5           8        5      9          8          8          8

     Score                          550         353          359         378         330         383    507     421        258          331
     Normalize                      100        64.2         65.3         68.7        60.0        69.6   92.2    76.5       46.9         60.2




                                                                        86
CHEMINFO




                 7. Comparison of Model Results
7.1 Introduction

This section presents the results of more detailed analysis and assessment of the model results,
input databases, parameters as well as the model algorithms for the selected models, namely:
GREET and GHGenius. The results from the default baseline cases of the models are first
developed and compared, and then the reasons for any differences in the models are investigated.
This is followed by the analyses of the biofuel (corn derived bioethanol) pathways in the models.
GREET version 1.8 and GHGenius version 3.12, have been used for this analysis. In both cases
the models have been set to the year 2007. A full comparison between the models is not possible
because of the limited pathways and the limited geographic coverage in GREET but enough of a
comparison is possible to get a sense of the similarities and differences in the models.

GHGenius and GREET have a number of similarities in their basic structures and algorithms.
Both are energy based models and both undertake a carbon balance approach to carbon dioxide
measurements. Air pollutant or criteria air contaminant (CAC) emissions for both models are
based in part on US EPA AP-42 emission factors but in both cases there are modifiers applied to
account for emissions in a regulated environment.

In spite of the similar approaches the results from the two models can be quite different so in this
section of the report the results for the base cases are compared with the intent of determining
the underlying cause of the differences. GREET has more limited results output than GHGenius
so the presentation of the results is restricted to what is the easiest to extract from the GREET
model and the GHGenius lifecycle stage results are aggregated to include the same stages as
GREET reports.


7.2 Analysis of Reference Fuel Results

The reference conditions for gasoline, which is reference fuel typically used for ethanol LCA,
are critical for LCA of transportation fuels. Since most results are represented as a percentage
reduction in the popular press, the percentage reduction is strongly influenced by the emissions
for the reference fuel.

It is possible to provide a comparison of the default values for three cases with the two models,
GHGenius has information on emissions for Canada and the United States, and GREET can
provide emissions estimates for the United States only.

The GREET results have been adjusted to provide the emissions using Higher Heating Value
(HHV) basis since this is the only measure available in GHGenius. This is probably the more


                                                                                                 87
CHEMINFO




appropriate basis since all energy is sold on a HHV basis in North America. GREET also uses
the more recent IPCC Global Warming Potentials (GWP’s) of 23 (in contrast to the previous
value of 21) for methane and 296 (in contrast to 310) for nitrous oxide and the impact of this
change will be investigated.

7.2.1 Gasoline

The reference fuel for ethanol is gasoline produced from crude oil. The system boundaries for
the two models are broadly similar and encompass all emissions from the point of oil production
to end use. GHGenius does include an estimate of the energy and emissions associated with
drilling and maintaining the oil wells and this does not appear to be included in the GREET
calculations.


7.2.1.1 GHG Emissions

In the following table the GHG emissions for four cases are shown. The first three cases are the
GHGenius results, GHGenius Canada, GHGenius US, and GHGenius US with the latest GWP’s;
these are compared to the results from GREET. By looking at the three cases for GHGenius we
can start to see the impact of several of the assumptions on the overall emissions picture.


            Table 19: Comparison of GHG Emissions for Gasoline

Lifecycle Stage              GHGenius       GHGenius US        GHGenius US          GREET US
                                Canada
                             1990 GWP          1990 GWP         2002 GWP             2002 GWP
                                                     g CO2eq/GJ
Feedstock production              9,455             9,219            9,449                5,240
(i.e. crude oil prod’n)
Fuel production                  12,972             13,258            13,303             12,677
(i.e. crude oil
refining)
Combustion                       63,887             64,028            63,993             67,846
Total                            86,314             86,505            86,745             85,763

The differences in the results for the two models can be driven by a number of factors. The
databases in the models contain production data, technology shares, emission factors, yields,
energy consumption data, energy allocations, and emission control data, which are all important
elements for the models. All of these factors can influence the results.




                                                                                             88
CHEMINFO




7.2.1.1.1 Crude Oil Production

From the table it is apparent that there are large differences in terms of crude oil production and
transportation (referred to as feedstock production) and smaller differences in terms of refining
and combustion emissions.

The energy requirements for crude oil production in GHGenius for Canada are based on
information from Statistics Canada, Natural Resources Canada, and industry publications. The
US data in GHGenius is derived from US Census publications. More detail on the latest crude oil
information can be found in two GHGenius reports, 2007 Crude Oil GHGenius Update Report 93,
and US Data Update 94. The US database that Franklin Associates has developed for NREL for
SimaPro and other European LCA models also use the US Census information from 1997.
GHGenius uses 2002 US Census data.

GREET assumes that the energy required to produce conventional crude oil is 2% of the energy
in the crude oil and that the oil supply is 100% conventional. The 1999 documentation for
GREET 1.5 95 (the more recent documentation for versions 1.6 to 1.8 does not refer to this input
value) states that:

“On the basis of existing studies, GREET assumes an energy efficiency of 98% for petroleum
recovery”

The existing studies that were referred to were all published in the early 1990’s (including a very
early report by Delucchi) and thus the data is probably 20 years old at this point. The calculated
efficiency in GHGenius for US crude oil production is 93.5%. Changing the GREET crude oil
energy efficiency to 93.5% will increase the GHG emissions to 9,956 g/GJ, a value slightly
higher but much closer to GHGenius. There are also differences in the mix of energy used to
produce the crude oil with GREET assuming a more carbon intensive fuel mix.

The methane emissions from crude oil production (a measure of venting and flaring) in GREET
are 80.4 g/GJ and in GHGenius they are 106.5 g/GJ. This will also account for some of the
difference between the two models. The data on methane emissions in GREET appears to be
based on US EIA estimates from the early 1990’s. GHGenius also uses EIA data on methane
emissions but the data is from the mid 2000 period.

For this sector, the differences between Canada and the United States are not that large based on
the GHGenius results. The differences caused by the GWP assumptions are also not that large,

93
   2007 Crude Oil GHGenius Update Report. http://www.ghgenius.ca/reports/2007CrudeOilUpdateReport.pdf
94
   US data Update. (2007) http://www.ghgenius.ca/reports/USDataUpdate.pdf
95
        GREET          1.5       --      Transportation     Fuel-Cycle      Model      (August       1999).
http://www.transportation.anl.gov/software/GREET/pdfs/esd_39v1.pdf



                                                                                                        89
CHEMINFO




which shows that for this energy pathway the carbon content of the fuel is the most important
determinant and that the methane and nitrous oxide emissions are relatively small components of
the overall emissions.

7.2.1.1.2 Crude Oil Refining

The results for crude oil refining from the two models are similar and the approach taken by the
models is also quite similar although the actual calculation process in GHGenius is far more
complex than it is in GREET.

Both models start with the total energy consumption in the refineries as a function of the energy
in the crude oil and then allocate that energy to the refined petroleum products. In GREET these
values were originally estimated from the results from a few studies from the early 1990’s but
the values have been updated and changed slightly since then. In a 2004 96 paper authored by
Michael Wang (GREET model developer) and colleagues, a variety of approaches for energy
allocation in a petroleum refinery were discussed. The approach that is used in GREET was
described as “the energy-content-based allocation at the refinery level with rule-of-thumb
adjustment”.

In GHGenius, the same basic approach is applied but there are a number of detail differences.
Refinery energy use data for GHGenius for Canada are gathered from Statistics Canada and
Natural Resources Canada, and the Energy Information Administration (EIA) for the United
States data. The energy use in the refinery is also regionalized since different refineries have
different crude oil slates and process configurations. This energy use information is adjusted for
the crude oil density and sulphur content, and the sulphur content of the petroleum products.
This adjustment of energy for oil density and sulphur content provides more flexibility for
modelling different scenarios. This flexibility is not available in GREET. The allocation of the
refinery energy consumption to individual products is also undertaken on an estimated process
chain basis. GHGenius uses the average values for each specific region

The energy efficiency in GREET for conventional gasoline is 86%. GHGenius reports the
refinery energy use by product instead of efficiency. For gasoline in the United States the
equivalent value is 87.7% and for Canada it is 87.3%. GREET thus allocates slightly more
refining energy to gasoline than does GHGenius for this stage. The refining energy efficiency in
GREET does not appear to be based on the average values reported by the EIA.

The types of energy used in the refinery are calculated in GHGenius based on EIA data for the
United States and StatsCan data for Canada. The GREET data looks similar to the EIA data but
is rounded to even values.

96
  Allocation of Energy Use in Petroleum Refineries to Petroleum Products. 2004. Michael Wang, Hanjie Lee and
John Molburg. International Journal of Life Cycle Assessment 9 (1) 34 – 44 (2004).
http://www.transportation.anl.gov/software/GREET/pdfs/IJLCA-2004.pdf



                                                                                                         90
CHEMINFO




Oil refineries are very complex, producing multiple products and consisting of a variety of
individual process units. The allocation of energy and emissions to individual products is
therefore a very complex issue. A variety of approaches can be used to allocate energy and
emissions and some of these are described in Wang et al. (2004). In an ideal world a system
expansion would be undertaken using data from multiple refineries to determine the best
allocation but the information required for this exercise is not publicly available.


7.2.1.1.3 Gasoline Combustion

The gasoline combustion GHG emissions are relatively close for the two models as well. The
differences in GHGenius between Canada and the US are driven primarily by the different
compositional analysis for gasoline in Canada and the US. Canada has a lower aromatic content
and that leads to a lower carbon content in the fuel and lower CO2 emissions per unit of energy.
The impact is relatively small. Both models report the CO2 emissions by calculating the carbon
in the fuel and subtracting the carbon that is emitted as methane and VOC’s. These CAC
emissions can thus also impact on the GHG emissions calculated. The emissions of the three
main GHG gases for gasoline combustion from the two models are shown in the following table.


     Table 20: Comparison of GHG Emissions for Gasoline Combustion

Contaminant                GHGenius Canada           GHGenius US    GHGenius US                    GREET US
                                1990 GWP               1990 GWP       2002 GWP                     2002 GWP
                                                              g /GJ
Carbon Dioxide                           61,882           63,542         63,542                       67,162
Methane                                       7                 7             7                            3
Nitrous Oxide                                 4                 4             4                            2
GHG (CO2eq)                              63,887           65,506         63,993                       67,837

The carbon dioxide emissions difference between the US GHGenius and GREET models result
from small changes in the assumed carbon content, density and heating values between the
models. Most of the difference is caused by the different CO and VOC emissions. These will be
quantified in the following section.

The N2O emissions in GREET are a user input value and are derived from US EPA information.
The methane emissions are an average of values from MOBILE6.2 97 and the California
EMFAC 98 2002 model. In both cases the values are a user input and are static over time.
GHGenius methane emissions are based just on MOBILE6.2C but are adjusted based on the

97
     US EPA MOBILE6.2. http://www.epa.gov/otaq/m6.htm
98
     California Air Resources Board. EMFAC. http://www.arb.ca.gov/msei/onroad/latest_version.htm


                                                                                                         91
CHEMINFO




year. The N2O emissions in GHGenius are based on a similar assessment of available data but
with more emphasis on Canadian test data than the EPA used.


7.2.1.2 CAC Emissions

Both GREET and GHGenius calculate the lifecycle emissions of individual air contaminants
(both CACs and GHGs). The production emissions are considered separately from the use
emissions. The results for the three scenarios considered for GHG emissions are shown in the
following table.


      Table 21: Comparison of Lifecycle Individual Air Contaminant
                  Emissions for Gasoline Production

Air Contaminant              GHGenius Canada                   GHGenius US                    GREET US
                                                                g /GJ
VOC                                            37.2                    37.2                            23.6
Methane                                       128.7                   144.6                            94.1
CO                                             18.7                    24.1                            14.4
N2O                                             1.0                     1.0                             0.3
NOx                                            80.8                    88.4                            45.0
PM10                                            8.5                    16.4                             9.8
SOx                                            76.7                    39.8                            24.3

The emissions shown in the previous table are a combination of the emissions from the
combustion of fuels used in oil production and refining and the non-combustion process
emissions. The process emissions in GREET are based on data from the US EPA AP-42 but it is
not clear how improvements over time are built into the model. The emission factors in GREET
1.8 are not the same as those described in the manual accompanying version 1.5 of the model
and the manuals that describe the updates 1.6 99 and 1.7 100 do not provide additional information.

The emissions associated with the oil refinery in GHGenius have recently been updated. The
uncontrolled emissions are from AP-42 but the control factors that are applied have been
adjusted so that the actual emissions are closely aligned with those reported to NPRI in Canada
and the NEI (National Emissions Inventory) in the United States. This work has allowed the
model to be regionalized in both Canada and the United States. As can be seen from the

99
   Development and Use of GREET 1.6 Fuel-Cycle Model for Transportation Fuels and Vehicle Technologies (June
2001). http://www.transportation.anl.gov/pdfs/TA/153.pdf
100
      Operating Manual for GREET: Version 1.7 (November 2005, Revised February 2007).
http://www.transportation.anl.gov/pdfs/TA/353.pdf



                                                                                                         92
CHEMINFO




GHGenius columns in the table there is a significant difference between emissions in Canada
and the United States. These differences are driven by differences in refinery process
configuration, environment, and regulatory requirements.

The emissions shown in the previous table show that while there are some differences between
the two models, the results are of the same order of magnitude for each of the contaminants. To
try and determine where the differences are the following tables show the emissions for just the
crude oil production, or just the oil refining and downstream stages. The differences are larger
for oil production than for oil refining. This largely reflects the difference in the input data for oil
production between the two models.


Table 22: Comparison of Individual Air Contaminant Emissions for Oil
                            Production

Air Contaminant               GHGenius Canada                GHGenius US                   GREET US
                                                             g /GJ
VOC                                           8.2                     9.1                           3.1
Methane                                     106.5                  116.4                           80.4
CO                                            9.4                   12.7                              7
N2O                                           0.3                     0.3                           0.1
NOx                                            46                   54.1                           23.2
PM10                                          2.1                     9.3                           1.8
SOx                                          18.9                   10.6                            8.2


  Table 23: Comparison of Individual Air Contaminant Emissions for
                         Gasoline Refining

Air Contaminant               GHGenius Canada                GHGenius US                   GREET US
                                                             g /GJ
VOC                                          29.0                   28.1                           20.5
Methane                                      22.2                   28.2                           13.7
CO                                            9.3                   11.4                            7.4
N2O                                           0.7                     0.7                           0.2
NOx                                          34.8                   34.3                           21.8
PM10                                          6.4                     7.1                           8.0
SOx                                          57.8                   29.2                           16.1

The final table considers the emissions from the use of the fuel in the vehicle. The emissions are
shown on a g/km basis and the fuel economy for the two models has been set to be equal at 10.6


                                                                                                     93
CHEMINFO




l/100 km (22.2 mpg) the GREET default value. It is the large difference in CO emissions that
drives the difference in GHG emissions for the two models.

  Table 24: Comparison of Individual Air Contaminant Emissions for
                       Gasoline Combustion

Air Contaminant            GHGenius Canada             GHGenius US               GREET US
                                                       g /km
VOC                                     0.28                  0.28                      0.09
CO                                     10.89                 10.89                      2.87
NOx                                     0.30                  0.30                      0.15
PM10                                   0.015                 0.015                     0.005
SOx                                    0.023                 0.020                     0.004
Methane                                0.024                 0.024                     0.012
N2O                                    0.012                 0.012                     0.007
CO2                                    232.1                 237.6                     246.8

While both models have used MOBILE6.2 to derive the vehicle emissions, GHGenius uses the
Canadian version, which would have a different temperature profile as well as other changes.
GREET uses the US version of the model and then averages those results with the data from the
California model. It is not clear from the GREET documentation if the MOBILE data has been
converted from fleet average to a specific vehicle year.


7.2.1.3 Energy Balance

Both GREET and GHGenius report the energy consumed to make the various fuels in the model.
The results for conventional gasoline are summarized in the following table.


                       Table 25: Energy Balance for Gasoline

                           GHGenius Canada          GHGenius US                  GREET US
                                        Joules consumed/joule delivered
Feedstock production                 0.108                  0.095                      0.038
Fuel Production                      0.179                  0.180                      0.188
Total                                0.287                  0.275                      0.226

As noted previously, GREET assumes a much lower energy use in the production of crude oil.
There are some small differences between Canada and the United States according to GHGenius
for both the crude oil consumed and the refining stages.


                                                                                          94
CHEMINFO




7.2.1.4 Findings

The lifecycle GHG emissions for the reference gasoline are quite similar in the two models but
there are significant differences between the stages. GREET energy use data for crude oil
production may not be reflective of the current crude oil refined in the United States. It would be
valuable to update this information to better align it with GHGenius, for model comparison
purposes. The default values are low and may somewhat underestimate the energy requirements
for crude oil production.

The choice of GWP values between the older IPCC values and the newer values with higher
methane weighting and lower nitrous oxide weighting does not have a significant impact on the
lifecycle GHG emissions.

There are some differences in the individual air emissions for oil production and refining
between the two models. The GREET and GHGenius values have changed over time as more
current data reflecting changes in technologies, energy and emissions occur. More
documentation with details would be valuable to users of both models to better understand the
base estimates and changes. GHGenius utilizes different emission factors for Canada and the US
for oil refining and the results are slightly different as expected. Both models use US EPA AP-42
emission data in the calculations. AP 42 emission factors are used to calculate the uncontrolled
emissions in GHGenius. The GHGenius results have been compared to NPRI data in Canada and
NEI data in the United States so they should reflect CAC oil refining emission rates. AP-42
emissions factors are also used in GREET, however, the calibration process to arrive at the final
controlled emission rate levels contained model is not transparent.

The results are summarized in the following table. The emissions associated with producing the
vehicles are not included in this comparison.




                                                                                                95
CHEMINFO




   Table 26: Comparison of Lifecycle Gasoline Results with Different
                               Models

Parameter/                  Units             GHGenius         GHGenius           GREET
Contaminant                                    Canada            US                US
                     Joule consumed/joule
Energy                          produced              0.287            0.275            0.226
GHG                           g CO2eq/GJ             86,314           86,745           85,763
CO2                                 g/km              303.5            307.9            303.3
CH4                                 g/km              0.479            0.555            0.361
N2O                                 g/km              0.016            0.016            0.011
PM10                                g/km              0.046            0.075            0.058
NOx                                 g/km              0.593            0.621            0.320
SOx                                 g/km              0.305            0.168            0.098
VOC                                 g/km              0.408            0.408            0.216
CO                                  g/km             10.957           10.976            2.929



7.3 Analysis of Biofuel Results

GREET includes the corn ethanol pathway but not wheat ethanol so only corn ethanol can be
compared to the results for GHGenius. Later in this section the results from GHGenius for corn
ethanol will be compared to the results from wheat ethanol.

7.3.1 Corn Ethanol

Corn is the world’s dominant coarse grain with approximately 700 million tonnes produced each
year. About 40% of that is produced in the United States. Corn produces the highest grain yield
of all of the cereal crops and the kernels have a high starch content with corresponding low
protein levels making corn an ideal substrate for ethanol production.

The system boundaries for ethanol production are generally similar between the two models with
a few exceptions. GHGenius includes energy required to manufacture the farm equipment and
transportation equipment. GREET does not incorporate this energy requirement. The land use
calculations in GREET are different and less robust than they are in GHGenius. The co-product
credit allocations are similar in concept (displacement) in the two models but GHGenius
includes the capability of modelling carbon dioxide as a co-product, and includes a factor to
account for differences in methane emissions resulting from the animal feed on distillers dried
grains. In the following figure the lifecycle stages for corn ethanol are shown.



                                                                                            96
CHEMINFO




                           Figure 8: Corn Ethanol Lifecycle Stages




The typical corn ethanol production process is shown in the following figure. Corn enters the
facility and ethanol and distillers dried grains are produced as products. Energy is added to the
process in the forms of electricity and thermal energy. Water, enzymes, yeast and chemicals are
added at various points.




                                                                                              97
CHEMINFO




                           Figure 9: Ethanol Production Process




7.3.1.1 GHG Emissions

In the following table the lifecycle GHG emissions for corn ethanol are presented for four
scenarios. The first three cases are the GHGenius results, GHGenius Canada, GHGenius US, and
GHGenius US with the latest GWP’s; these are compared to the results from GREET. By
looking at the three cases for GHGenius we can start to see the impact of several of the
assumptions on the overall emissions picture.

The GREET model has been used to model dry mill ethanol plants. The default value assumes
that 80% of the plant process energy is natural gas and 20% is coal. The US version of
GHGenius has been set to use the same energy mix at the ethanol plant for comparison purposes.
The Canadian version assumes 100% natural gas process energy.




                                                                                           98
CHEMINFO




           Table 27: Comparison of GHG Emissions for Corn Ethanol

Lifecycle Stage                     GHGenius           GHGenius US            GHGenius US    GREET US
                                      Canada
                                    1990 GWP                1990 GWP        2002 GWP         2002 GWP
                                                                  g CO2eq/GJ
Feedstock                                                                                       24,446
production 101                            24,074                 24,590             23,873
Fuel production 102                       30,616                 48,107             48,384      39,412
Co-products                              -17,335                -17,689            -17,337           0
Combustion 103                             2,099                  2,188              2,070         650
Ethanol Total                             39,454                 57,196             56,990      64,508
Gasoline Total                            86,314                 86,505             86,745      85,763
% Reduction                                 54%                    34%                34%         25%


The co-product credits in GREET are included in the feedstock production emissions. The value
of the credit is 15,884 g/GJ. The derivation of this value will be discussed later.

The differences between Canada and the US corn ethanol that are calculated by GHGenius relate
primarily to different input values for energy used in the ethanol plants (due to the different
average age of the plants in the two countries) and the carbon intensity of the electricity and
thermal energy used in the ethanol plants. There are some differences created by the different
GWP’s but these are partially offset by changes in the co-product credits from the different
GWP’s.

The percent reductions in GHG emissions for ethanol compared to gasoline are also shown in
this table. The 25% value for GREET is better than often quoted 20% reduction but this is
because only dry mill plants are included in the comparison. Note that the lower emissions for
gasoline in GREET also impacts the calculation of percentage reduction.

While the lifecycle emissions are reasonably close for each of the models there are some
significant differences in how they are arrived at. There are significant differences in the
emissions for corn production (since GREET subtracts the co-product credit from corn
production) and the ethanol plants between the two models. These are discussed below by
looking at corn production, ethanol production, co-products, and fuel combustion.



101
    Includes the transportation of the corn from the farm to the ethanol plant.
102
    The cultivation of corn.
103
    Ethanol combustion.


                                                                                                   99
CHEMINFO




7.3.1.1.1 Corn Production

From both of the models it is possible to extract the GHG emissions from the production and
transportation of corn to the ethanol plant. For GREET it is possible to extract the information
before the application of the co-product credit to get a better comparison. These results are
shown in the following table.

      Table 28: Comparison of GHG Emissions for Corn Production

Lifecycle Stage                     GHGenius          GHGenius      GHGenius           GREET US
                                      Canada                 US             US
                                    1990 GWP          1990 GWP       2002 GWP           2002 GWP
                                                        g CO2eq/tonne corn
Feedstock transmission                   14,271           15,051         15,096             18,999
Feedstock production                     71,126           74,957         75,649             89,854
Land-use changes, cultivation            92,091           65,305         57,234            165,861
Fertilizer manufacture                   50,434           77,497         78,042            106,294
Total                                   227,922          232,810        226,021            381,008

The feedstock transmission (moving the corn from the farm to the ethanol plant) and feedstock
recovery (farming emissions) values are quite similar between the models, which indicates that
the overall energy requirements for those life stages are similar and that the emissions factors per
unit of energy used are essentially the same. GHGenius includes the energy required to produce
the farm tractors, which is not included in GREET. The transportation distance in GREET is 50
miles (80 km) and it is 72 km in GHGenius.

The farming energy in GHGenius does not change automatically when the region is changed so
the values here in the “US” model are really based on Ontario corn production. There are some
differences in production practices between Ontario and the US, particularly with respect to
tillage. The higher proportion of no-till and conservation tillage in Canada than the US would
result in less energy used on the farm, this is offset by the need to dry the corn more often in
Canada.

The difference in emissions between the Canada and US versions of GHGenius for fertilizer
manufacture and land use emissions is due to the assumption that some manure is used in
Canada but not in the US. This reduces the fertilizer energy but increases the N2O emissions
from application of the manure.




                                                                                                100
CHEMINFO




There are large percentage differences in the fertilizer manufacturing emissions and the land use
emissions between the models for the corn production processes. Each of these differences is
investigated below.

In the case of fertilizer manufacture the difference can be caused by the rate of fertilizer applied
or by the energy intensity of the fertilizer manufacture. The fertilizer rate assumptions in each
model are shown in the following table.

             Table 29: Comparison of Inputs for Corn Production

                                                  GHGenius US                              GREET
                                                           kg/tonne
Nitrogen                                                18.07                                  17.0
Potash                                                   7.85                                   7.0
Phosphorus                                                 5.9                                  6.0
Lime                                                       0.0                                 48.8
Sulphur                                                    0.0                                  0.0
Herbicides & pesticides                                  0.33                                  0.36
Seeds                                                    2.32                                   0.0

GREET does not include the energy for the production of the seeds and the estimates of the
fertilizer inputs are similar (except for lime) to those in GHGenius. The emissions from lime
application in GREET are included in the land use category and the emissions from lime
production in the fertilizer manufacture category.

In the following table the energy and emissions intensities of the various crop inputs are
summarized, again there are some small differences between the two models but none of the
values are out by an order of magnitude.




                                                                                                101
CHEMINFO




     Table 30: Comparison of Energy and GHG Intensities for Inputs

                GHGeniu       GHGenius          GREET       GHGeniu GHGenius               GREET
                s Canada             US                         s US   Canada
                               MJ /kg                                g CO2eq/kg
Nitrogen              45.0         51.9             50.8       2,671     3,446               3,317
Potash                 3.4          6.1              9.2         198       460                 689
Phosphorus             7.4          7.3             14.6         570       587               1,031
Lime                   7.0          1.7               7.0        549       142                 628
Sulphur                1.6          1.5              n.a.        117       121                 n.a.
Herbicides             244          244              300      16,230    16,752              22,174
& pesticides
Seeds                  5.8            5.8            n.a.          370           385            n.a

The difference in the fertilizer emissions between the two models is the sum of a number of
differences. The production of nitrogen fertilizer and potash in Canada is more efficient than it is
in the US based on industry surveys and this provides some advantage to Canadian production.
These differences are summarized in the following table for the US case. There has been some
rounding in the calculation of these individual values so the results are slightly different than in
Table 24.

   Table 31: Comparison of Energy and GHG Intensities for 2 Models

                             GHGenius US                                  GREET
               kg/tonne      g CO2eq/kg g CO2eq/tonne       kg/tonne            g g CO2eq/tonne
                                                 corn                    CO2eq/kg          corn
Nitrogen           18.07         3,446        62,269              17       3,317        56,389
Potash              7.85           460          3,611              7         689          4,823
Phosphorus           5.9           587          3,463              6       1,031          6,186
Lime                 0.0           142              0           48.8         628        30,646
Sulphur              0.0           121              0              0          n.a.            0
Herbicides          0.33        16,752          5,528           0.36      22,174          7,983
&
pesticides
Seeds               2.32           385              893            0           n.a               0
Total                                            75,764                                    106,027

The primary difference between the models related to agricultural inputs is the production and
application of lime. Not only does GREET model lime use on US corn fields (based on USDA
survey data) but the energy and emissions intensity of lime production is much higher than is


                                                                                                102
CHEMINFO




modelled in GHGenius. Based on input from the Ontario Ministry of Agriculture Food and Rural
Affairs no lime is used for corn production in Canada. This provides a significant GHG benefit.
The use of lime is obviously a function of the soil and general environmental quality. It is used
to adjust the pH to 6.5 to 7, a range that produces the greatest crop performance for both corn
and soybeans. Producers do not apply lime every year but only when required.

The energy required for lime production in GHGenius is much higher for Canada than it is for
the United States. The Canadian energy is derived from Statistics Canada data for this sector and
the US data is from the US Census. The Canadian data are quite close to the GREET values
however GREET’s value assumes that the energy requirements for producing lime are similar to
those of producing potash.

Lime can also be applied in a number of forms, for example as lime (CaO) or limestone (CaCO3)
and the energy requirements for producing these two forms are quite different. Most models take
the conservative approach.

Land use emissions can be caused by changes in soil carbon, methane and N2O emissions
resulting from tillage practices, both from fertilizer use and independent of fertilizer use, and
other changes in above ground biomass. Land use emissions in GREET are higher for corn than
they are in GHGenius. In GREET there are three components to land use emissions, soil carbon
changes, carbon dioxide from the neutralizing effects of lime, and N2O emissions from the
application of nitrogen fertilizer. These emissions are separated and compared to GHGenius land
use values in the following table.

  Table 32: Comparison of Land Use Emissions for Corn Production

                                                   GHGenius US                              GREET
                                                          g/tonne corn
Soil Carbon                                                147                                 7,678
N2O emissions                                          110,871                               136,714
Lime emissions                                                0                               21,469
Above ground biomass                                   -45,772                                     0
                                                        65,246                               165,861

In GHGenius it is assumed that 80% of the corn used for ethanol production is a result of higher
yields and 20% comes from land that was used previously to grow another crop. In the case of
the higher yield the change in soil carbon is modelled to be very small and is based on estimates
made a number of years ago by Agriculture Canada for soil carbon contents in Ontario. With the
level of no-till and reduced till corn produced in Ontario it is likely that soil carbon is increasing
in this region now. GHGenius also estimates the change in above ground biomass based on the
different crop patterns and crops. Since corn has such a high biomass yield growing more corn
stores carbon above the ground each year and this results in a credit to land use emissions.



                                                                                                  103
CHEMINFO




The differences in emissions due to N2O production arise from different methodologies utilized
in GREET and GHGenius. GREET uses an emission factor of 1.3% of synthetic nitrogen applied
plus the nitrogen in the residue returned to the soil. GHGenius uses a factor of 1.125% on both
the fertilizer and residue (but assumes more nitrogen in the residue), which was the default IPCC
factor in the 1996 guidelines. The new IPCC guideline 104 is 1.0%. GHGenius also considers
nitrogen leakage from the site and follows the IPCC guidelines in this regard.

7.3.1.1.2 Ethanol Production

The energy requirements and emissions associated with the production of ethanol from corn are
summarized in the following table. The energy requirements and emissions are quite close for
the ethanol production stage between the two models.

      Table 33: Comparison of Energy and GHG Intensity for Ethanol
                               Production

                          GHGenius Canada                 GHGenius US                         GREET
Direct Energy                      529 MJ/GJ                   680 MJ/GJ                    582 MJ/GJ
GHG Emissions             28,949 g CO2 eq/GJ          45,924 g CO2 eq/GJ           38,227 g CO2 eq/GJ

The energy use in the ethanol plants varies slightly between the models. In GHGenius the
Canadian data is based on plants built in Canada and assumed that there was some co-generation
of heat and power and all of the plants use natural gas. The electric power production in Canada
is also more efficient and lower carbon intensity than it is in the United States, also contributing
to better energy efficiency and lower GHG emissions. The GHGenius US uses older data without
co-generation. The GREET values are probably indicative of the average US fleet of plants.
Energy efficiency at ethanol plants has increased steadily over time as shown in the following
figure 105.




104
  2006 IPCC Guidelines for National Greenhouse Gas Inventories. Volume 4. Agriculture, Forestry and Other
Land Use. http://www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.htm
105
    Martin Junginger. Learning Curves for Biofuels. International Workshop on technology learning and
deployment.           IEA          headquarters,          Paris,     11-12         June          2007.
http://www.iea.org/Textbase/work/2007/learning/Junginger_biomass.pdf


                                                                                                     104
CHEMINFO




                        Figure 10: Ethanol Plant Energy Consumption




One difference between the models is that GHGenius calculates the energy and emission
requirements of the chemicals used in the ethanol production process. This source of emissions
is not included in GREET. For grain ethanol most of these emissions are quite small and don’t
contribute significantly to the lifecycle emissions.

7.3.1.1.3 Co-Products

Both GREET and GHGenius assume that distillers dried grain (DDG) from the ethanol process
displaces both corn and soybean meal from animal rations. The models each calculate the energy
and emissions associated with the production of corn and soymeal and subtract those emissions
from the other emissions that the model calculates for the corn ethanol lifecycle. In the case of
GHGenius the emission credit is transparent and provided in its own lifecycle stage and in the
case of GREET the credit is subtracted from the corn production emissions. It was noted earlier
that there was a difference in the value of the co-product credit in the two models and this is
investigated here.




                                                                                             105
CHEMINFO




GREET reduces the quantity of DDG that is available for co-product credits by 15.1% by
assuming that the availability of increased DDG product leads to new cattle production.
GHGenius has the capability of modelling this as well but no factor is applied. This concept of
discounting increased production is often found in economic models but it is extremely difficult
to verify the results. The concept should apply in both directions so if corn prices increase
because of ethanol production and this decreases cattle production then should the ethanol plants
receive a credit? In the following table the net displacement (after the GREET economic
adjustment factor) ratios from the two models are compared. GHGenius also provides a small
methane credit for the improved feed gain demonstrated with DDG.

                        Table 34: DDGS Displacement Ratios

                                                        GHGenius                          GREET
                                              kg/kg corn feedstock           kg/kg corn feedstock
Corn                                                         0.197                          0.243
Soybean Meal                                                 0.174                          0.186
Avoided methane emissions                     12.5 g CH4/kg DDG                                0.0

The displacement ratios in GREET were a result of a conference held between Argonne
researchers and several animal nutritionists with expertise in feeding DDG. The lower values
used in GHGenius resulted from a literature review looking for values to support the GREET
values. While it became clear that DDG did replace both corn and soybean meal in rations the
magnitude of the displacement found was typically lower than used in GREET. More discussion
of this issue and the avoided methane emissions are found in the 2005 GHGenius Ethanol
Update 106 report.

In the following table the GHG emissions credits for each of the components are compared in the
two models. A computational error was found in the GREET model that effectively increases the
corn credit by 16.7% by using the soybean bushel weight rather than the corn bushel weight in
the calculations. This effectively offsets the 15% discount given for economic factors.

                              Table 35: DDGS GHG Credits

                                                         GHGenius                       GREET
                                                        g CO2 eq/GJ                  g CO2 eq/GJ
Corn                                                          4,659                       11,448
Soybean Meal                                                 10,062                        4,435
Avoided methane emissions                                        0.0
Total                                                        14,721                       15,883

106
         (S&T)2      Consultants     Inc.,     2005,     Ethanol       GHG   Emissions     Update.
http://www.ghgenius.ca/reports/NRCanEthanolUpdate.pdf


                                                                                              106
CHEMINFO




While the final values of the co-products are similar there is a large variation in how the values
are derived. It was shown earlier that the emissions from corn production in GREET are higher
than they are in GHGenius due to the use of lime in the US and the fact that it is generally not
used in the Canadian corn belt. This factor plus the larger displacement ratio are the primary
reasons for the difference in the corn portion of the DDG credit. With the soybean meal, two
factors are involved, the first is that GREET assumes that no N2O is generated when nitrogen
fixing bacteria provide the nutrients for soybean production and the second is that GREET
allocates considerably more of the emissions to soybean oil than to meal compared to the default
approach in GHGenius.

There is some controversy and uncertainty with respect to the generation of N2O during the
nitrogen fixing process. Agriculture and Agri-Food Canada and Environment Canada have long
maintained that the emissions are much less than generated from an equivalent quantity of
synthetic fertilizer. There are other scientists who have provided data that suggests the emission
rates are similar. The IPCC, in the 2006 guidelines, have now changed their original position that
the emissions are the same to one where it states that there are some emissions but that the
emission rate may be less than the rate from synthetic fertilizers. GHGenius can model the
impact of this uncertainty quite easily and this issue will be investigated in the sensitivity section
that follows.

In the 2006 IPCC guidance document 107 the issue was summarized as follows.

        Biological nitrogen fixation has been removed as a direct source of N2O because of the
        lack of evidence of significant emissions arising from the fixation process itself (Rochette
        and Janzen, 2005). These authors concluded that the N2O emissions induced by the
        growth of legume crops/forages may be estimated solely as a function of the above-
        ground and below-ground nitrogen inputs from crop/forage residue (the nitrogen residue
        from forages is only accounted for during pasture renewal).

The Rochette and Janzen paper referenced by the IPCC reaches the following conclusion.

        In summary, there is little doubt that legumes can increase N2O emissions, during growth
        (Kilian and Werner 1996) and especially after harvest (Bremner et al. 1980; Larsson et
        al. 1998; Rochette et al. 2004) or plowdown (Wagner-Riddle et al. 1997; Baggs et al.
        2000; Millar et al. 2004). But field measurements indicate that much of this increase in
        emissions may be attributable to the N release from root exudates during the growing
        season and from decomposition of crop residues after harvest, rather than from
        biological N fixation per se. This implies that N2O emissions associated with the
        biological N fixation by legumes are smaller than previously estimated, and that, under

107
  2006 IPCC Guidelines for National Greenhouse Gas Inventories. Volume 4. Agriculture, Forestry and Other
Land Use. http://www.ipcc-nggip.iges.or.jp/public/2006gl/vol4.htm



                                                                                                     107
CHEMINFO




        field conditions, Rhizobia denitrification does not reduce significant amounts of nitrate
        or that N2O represents only a minor fraction of gaseous N products.

        Field flux measurements and process-level laboratory studies offer little support for the
        use of an emission factor for BNF 108 by legume crops equal to that for fertiliser N.
        Moreover, given the uncertainty regarding the direct role of Rhizobia in N2O emission
        under field conditions, the inclusion of this mechanism into the IPCC methodology is
        hard to justify. Consequently, we propose that:
            1. the biological fixation process itself be removed from the IPCC N2O inventory
                methodology;
            2. N2O emissions induced by the growth of legume crops be estimated solely as a
                function of crop residue decomposition using an estimate of above- and below-
                ground residue inputs, modified as necessary to reflect recent findings on N
                allocation.

One of the earlier papers produced in Canada (Rochette et al. (2004) 109) that was used to support
the “no emission position” significantly overestimated the nitrogen content of the residue
returned to the soil (a value of 3% rather than the typical actual nitrogen content of soybean
residue of 0.8% was used) and if the correct value was used then the paper supported the position
that the N2O emissions were similar to those of synthetic nitrogen.

This choice of emission factor is quite significant in GHGenius because it impacts not only the
emissions from growing soybeans but also from canola meal because of the system expansion
and from all of the other protein meals because of the displacement approach to co-product
allocation.


7.3.1.1.4 Fuel Use

Ethanol is used in blends with gasoline, either as E10 or as E85. The lifecycle emissions for
these E10 and E85 blends are summarized in the following two tables.




108
   BNF. Biologically fixed nitrogen.
109
   Rochette, P., Angers, D.A., Belanger, G., Chantigny, M.H., Prevost, D., Levesque, G. 2004. Emissions of N2O
from Alfalfa and Soybean Crops in Eastern Canada. Soil Science Society of America Journal. 68:493-506.


                                                                                                          108
CHEMINFO




                Table 36: Lifecycle GHG Emission Results for E10

                              GHGenius           GHGenius US             GHGenius US                GREET US
                                Canada
                              1990 GWP                1990 GWP                2002 GWP               2002 GWP
                                                               g/km
CO2                                  287.6                 295.5                    295.5                   297.0
CH4                                  0.456                 0.538                    0.538                   0.357
N2O                                  0.021                 0.021                    0.021                   0.017
E10 Total, CO2eq                     303.6                 313.4                    314.2                   310.3
Gasoline Total,                      318.8                 324.5                    325.4                   314.8
CO2eq
% Reduction                             4.7                     3.4                    3.4                     1.4

The relative percentage reductions in GHGenius (after adjustment for volume and energy) shown
in the above table are greater than are shown in Table 23 because a 1% greater energy efficiency
for the E10 blends is modelled. GREET assumes no difference in energy efficiency. Ethanol is
an oxygenated compound. As such it contains less energy than gasoline components that do not
contain oxygen. Ethanol has about 67% of the energy of gasoline per unit volume. Blends of
ethanol and gasoline have a poorer fuel economy on a volumetric basis since the fuel contains
less energy. This lower fuel economy has been demonstrated in a number of laboratory studies.
However the magnitude of the change in fuel economy is less than predicted by the change in
energy content.

The Auto/Oil Air Quality Improvement Research Program (Hochauser 1993 110 ) of the early
1990’s reported that the current vehicle fleet (1989 vehicles with emission control systems
similar to today’s vehicles) achieved a 1% better energy specific fuel economy when 10%
ethanol was added to gasoline. Ethanol blends were not tested in the older fleet, but methyl
tertiary butyl ether (MTBE) was tested, and it was found that better energy specific fuel
economy was found in the older fleet than in the current fleet.

Ethanol has a higher heat of vapourization, a higher specific energy ratio and produces more
moles of combustion products per mole of combustion air than gasoline (Owen 1990 111). These
three chemical characteristics probably account for the higher energy efficiency of ethanol
blended gasoline. For low level blends of less than 10% GHGenius scales the energy specific
fuel consumption in proportion to the ethanol content based on the results from the Auto/Oil

110
    Hochhauser, A.M., Benson, J.D., Burns, V.R., Gorse, R.A., Koehl, W.J., Painter, L.J., Reuter, R.M., Rutherford,
J.A. 1993. Fuel Composition Effects on Automotive Fuel Economy – Auto/Oil Air Quality Improvement Research
Program. SAE 930138.
111
    Owen, K., Coley, T. 1990. Automotive Fuels Handbook. Society of Automotive Engineers.



                                                                                                              109
CHEMINFO




study. For the 85% blends Wang (1999 112) reports a 5% better energy specific fuel economy and
that is modelled in GHGenius. Note that this provides less of a benefit than the 1% for E10 as
shown in the following table.

               Table 37: Lifecycle GHG Emission Results for E85

                             GHGenius          GHGenius US            GHGenius US              GREET US
                               Canada
                             1990 GWP               1990 GWP              2002 GWP               2002 GWP
                                                             g/km
CO2                                 149.2                190.5                   190.5                223.0
CH4                                 0.081                0.402                   0.402                0.380
N2O                                 0.069                0.076                   0.076                0.108
E85 total, CO2eq                    172.2                222.5                   222.2              263.793
Gasoline total,                     318.8                324.5                   325.4                314.8
CO2eq
% Reduction                          46.0                   31.4                  31.7                   16.2


7.3.1.2 CAC Emissions

Both GREET and GHGenius calculate the corn ethanol lifecycle emissions of individual air
emissions. The fuel production emissions are considered separately from the use emissions. The
results for the three scenarios considered for GHG emissions are shown in the following table.
From the analysis of the air emissions for gasoline and the GHG emissions for ethanol some
differences in the air emissions for ethanol can be expected.




112
  Wang, M., Saricks, C., Santini, D. 1999. Effects of Fuel Ethanol Use on Fuel-Cycle Energy and Greenhouse Gas
Emissions. ANL/ESD-38. Argonne National Laboratory.


                                                                                                          110
CHEMINFO




   Table 38: Comparison of Lifecycle Air Emissions for Corn Ethanol
                              Production

Contaminant               GHGenius Canada               GHGenius US                 GREET US
                                                         g /GJ
VOC                                         22                    24                          39
Methane                                    -16                    95                          99
CO                                         -42                   -37                          54
N2O                                         20                    23                          36
NOx                                        182                  223                          146
PM10                                        22                    38                          52
SOx                                         48                    82                          94

As with the case of the gasoline there are some large differences in the emissions of the
individual air contaminants that arise not only between the models but also in the same model
when different countries are analyzed. These emissions arise from the production of the corn and
from the ethanol production process. Most of the emissions will be from the combustion of some
fossil fuel, the production of electricity or the process related emissions.

Another feature apparent in the table is the fact that GHGenius can calculate negative emissions.
This is a function of how the allocation is done. In this particular case the energy required to
produce the soybeans (used in the calculation of the DDG credit) includes some gasoline use,
based on USDA data, whereas it has been assumed that no gasoline is used in the production of
corn, the common practice in Canada. In GREET some gasoline is used in corn production and
CO emissions from gasoline tractors and trucks are very high which creates the difference in
emissions between the two models for producing soybeans and the soybean meal and thus the
DDG credit.

A portion of the NOx emissions result from the application of fertilizer. Different conversion
factors are used in the two models, which results in higher emissions in GHGenius.

The reasons for the differences in VOC, methane and PM emissions between the models are not
readily apparent.


7.3.1.3 Energy Balance

Both GREET and GHGenius report the energy consumed to make the various fuels in the model.
The results for corn ethanol are summarized in the following table. This data is for the total
primary energy requirements so it includes the energy required to make the energy. This is why
the values differ between Canada and the United States, the different energy infrastructures have


                                                                                             111
CHEMINFO




different efficiencies, mostly related to the higher proportion of coal fired power production in
the US.

                   Table 39: Energy Balance for Corn Ethanol

                           GHGenius Canada           GHGenius US                    GREET US
                                         Joules consumed/joule delivered
Feedstock production                  0.193                  0.248                         0.186
Fuel Production                       0.547                  0.702                         0.581
Co-product credit                    -0.055                 -0.065                             0
Total                                 0.685                  0.885                         0.767

The energy balance of corn ethanol in GREET is more attractive than it is in GHGenius when the
US is modelled and the process is more energy efficient in the Canadian version of GHGenius.
The assumptions regarding plant process energy is different between the models and that
combined with the fundamental different energy infrastructures between the two countries
account for the differences.

7.3.1.4 Findings

While the lifecycle GHG emissions for corn ethanol in GREET and GHGenius are generally
similar when the same region is modelled there are differences in how those results are arrived
at. The basic processes that are modelled are similar in terms of process inputs but there are
differences in the approach, namely:

 •   GREET does not include the N2O emissions from the biological fixation of nitrogen
     whereas GHGenius does. The result is that the GHG emissions for producing soybeans are
     much higher in GHGenius than they are in GREET and this impacts the co-product credit
     for ethanol.
 •   The corn farming emissions in GREET are higher than in GHGenius due to the use of lime
     to adjust the acidity of soils in the US.
 •   There are some differences in land use emissions between the two models but a large
     portion of the difference is driven by the use of lime. The land use calculations in
     GHGenius are more complex than they are in GREET and have more capacity for
     sensitivity analysis.
 •   The use of coal in some US ethanol plants has a significant impact on the GHG emissions.
 •   Energy use in ethanol plants is the primary driver of GHG emissions and it has been
     improving at a significant pace over the past 25 years, which makes up to date data on
     energy use very important for this pathway.

The lifecycle results are summarized in the following table.



                                                                                             112
CHEMINFO




 Table 40: Comparison of Lifecycle E10 Results with Different Models

Contaminant                Units              GHGenius         GHGenius           GREET
                                               Canada            US                US
Energy (E100)       Joule consumed/joule            0.685           0.885             0.767
                                produced
GHG (E100)                   g CO2eq/GJ             39,454            56,990            64,508
CO2 (E10)                           g/km             287.6             295.5             297.0
CH4 (E10)                           g/km             0.456             0.538             0.357
N2O (E10)                           g/km             0.021             0.021             0.017
PM10 (E10)                          g/km             0.048             0.079             0.066
NOx (E10)                           g/km             0.614             0.650             0.336
SOx (E10)                           g/km             0.295             0.177             0.110
VOC (E10)                           g/km             0.395             0.396             0.219
CO (E10)                            g/km            10.635            10.655             2.764



7.3.2 Wheat Ethanol

The production of ethanol from wheat is not included in GREET. However, this is an important
feedstock for Canada so a comparison and discussion of the differences between corn and wheat
ethanol are presented here, based on GHGenius results.


7.3.2.1 GHG Emissions

Since only the results from GHGenius are being used to compare corn and wheat ethanol the
information from each portion of the lifecycle does not have to be aggregated to compare with
the GREET results. In the following table the lifecycle GHG emissions for the production of the
two types of grain-based ethanol are compared, assuming the fuels are both produced in Canada.
Baseline results for gasoline produced from crude oil are also provided in the table.




                                                                                           113
CHEMINFO




  Table 41: Comparison of Lifecycle GHG Emissions- Gasoline, Corn
                     and Wheat Ethanol (E100)

Fuel                                    Gasoline              Ethanol               Ethanol
Feedstock                               Crude oil              Corn                 Wheat
                                                            g CO2 eq/GJ
Fuel dispensing                                    137                 214                    214
Fuel distribution and storage                      559               1,452                  1,452
Fuel production                                 12,276              28,949                 31,880
Feedstock transmission                             926               1,507                    715
Feedstock recovery                               6,417               7,512                  6,382
Land-use changes, cultivation                        2               9,727                 18,774
Fertilizer manufacture                               0               5,327                 10,701
Gas leaks and flares                             2,109                   0                      0
CO2, H2S removed from NG                             0                   0                      0
Emissions displaced                                  0             -17,335                -31,786
Total                                           22,426              37,355                 38,334
Combustion Emissions                            64,293               2,099                  2,099
Grand Total                                     86,719              39,454                 40,429

Wheat has higher emissions associated with the ethanol plant and the feedstock production but
these emissions are offset by higher co-product credits due to the higher quantity and higher
protein of the wheat DDG compared to corn DDG. These issues are discussed further in the next
sections.

7.3.2.1.1 Wheat Production

The fertilizer and energy inputs for corn and wheat production in GHGenius for Canada are
summarized in the following table. It can be seen that there is a difference in nitrogen fertilizer
requirement and a much larger seed requirement for wheat. On the field energy side the need in
Canada to dry corn with some natural gas and propane creates a higher energy demand for corn.




                                                                                               114
CHEMINFO




       Table 42: Comparison of Corn and Wheat Production Inputs

                                                                   Corn                    Wheat
Nitrogen                      kg/tonne                            12.21                    19.31
Manure                        kg/tonne                             6.25                       5.0
Phosphorus                    kg/tonne                             5.91                    11.89
Potash                        kg/tonne                             7.85                     0.95
Lime                          kg/tonne                             0.00                     0.00
Sulphur                       kg/tonne                             0.00                     0.29
Pesticides                    kg/tonne                             0.33                     0.33
Seeds                         kg/tonne                             2.32                    46.73
Embedded Energy               MJ/tonne                              713                    1,312

Diesel                        L/tonne                             5.99                     10.60
Natural gas                   L/tonne                        10,907.91                      0.00
LPG                           L/tonne                             5.99                      0.00
Calculated                    kJ/tonne                        797,765                    409,728

The land use emissions for wheat are higher than they are for corn as shown in the following
table. This is again a function of the higher nitrogen application rates and the lower starch
content, which requires more feedstock per unit of ethanol produced.


   Table 43: Comparison of Land Use Emissions for Corn and Wheat
                             Production

                                                GHGenius Corn                   GHGenius Wheat
                                                           g/tonne
Soil Carbon                                               156                               -138
N2O emissions                                         137,527                            186,063
Above ground Biomass                                  -45,676                            -19,718
Total                                                  92,007                            166,207

There is the potential to lower the land use emissions for wheat. GHGenius currently uses the
same N2O emission factor for wheat and corn. Environment Canada, in developing Canada’s
national GHG emission inventory are using lower N2O emission rates for dry land farming,
which is the typical practice in western Canada. Some of the high starch, low protein feed wheats
that are being grown for fuel ethanol plants in western Canada also have lower nitrogen fertilizer
requirements and this would lower land use emissions.



                                                                                              115
CHEMINFO




7.3.2.1.2 Ethanol Production

The production process for wheat ethanol is essentially the same as it is for corn ethanol but the
different feedstock characteristics change the plant energy requirements. More energy is required
in the wheat process for the DDG dryers because more DDG is produced (and thus more water
must be evaporated) and the moisture content of the feed to the dryers tends to be slightly higher
due to the viscous nature of the wheat. In GHGenius the energy requirements for the wheat plant
are modelled as being 12% higher than the corn plant. Good operating data on the differences
between the two plants is not yet available but it should be available in the next year or so with a
couple of large modern wheat ethanol plants being started up in 2008.

7.3.2.1.3 Co-Product Allocation

The starch content of wheat is typically 6 to 10% less than it is for corn so the ethanol output
from a tonne of feedstock is also 6 to 10% less than it is for corn. This translates into increased
DDG production and the percent change in DDG production is about double that or 12 to 20%
higher (since the “missing” starch would have been converted to essentially equal amounts of
ethanol and CO2). The protein content of wheat DDG is also higher than corn DDG so it can
displace more soybean meal in livestock feed rations. It is also assumed that DDG displaces feed
wheat from animal diets rather than corn. The DDG credits for corn and wheat DDG are
summarized in the following table.


                             Table 44: DDGS GHG Credits

                                             GHGenius Corn                      GHGenius Wheat
                                               g CO2 eq/GJ                         g CO2 eq/GJ
Corn/Wheat                                           4,659                               9,241
Soybean Meal                                        10,062                              19,204
Avoided methane emissions                            2,652                               3,345
Total                                               17,373                              31,790


7.3.2.2 CAC Emissions

The individual air contaminants for the production of gasoline and corn and wheat ethanol are
summarized in the following table. This is for the production cycle only and does not include the
combustion emissions.




                                                                                                116
CHEMINFO




  Table 45: Comparison of Air Emissions for Production of Gasoline,
                     Corn and Wheat Ethanol

Fuel                                             Gasoline         Ethanol         Ethanol
Feedstock                                        Crude oil         Corn           Wheat
                                                                   g/GJ
Carbon dioxide (CO2)                                  19,404          31,470           36,149
Non methane organic compounds (NMOCs)                   37.2              22               21
Methane (CH4)                                          128.7             -16              -68
Carbon monoxide (CO)                                    18.7             -42             -104
Nitrous oxide (N2O)                                       1.0             20               12
Nitrogen oxides (NO2)                                   80.8             182               79
Sulphur oxides (SOx)                                    76.7              48               64
Particulate matter (PM)                                  8.5              22               25
HFC-134a (mg)                                             0.3              2                2
CO2-equivalent GHG emissions                          22,426          37,355           38,330

The emissions on a full lifecycle basis are summarized in the following table for gasoline and
E10. These emissions do not include the vehicle materials and assembly emissions. The
differences between the corn and wheat ethanol fuels are relatively small. The differences
compared to gasoline are also relatively small which is not surprising given than only 6.5% of
the energy is being displaced.




                                                                                          117
CHEMINFO




Table 46: Comparison of Lifecycle Air Emissions- Gasoline, E10 Corn
                        and Wheat Ethanol

                                                                 Ethanol         Ethanol
Fuel                                            Gasoline          (E10)           (E10)
Feedstock                                       Crude oil         Corn           Wheat
                                                                  g/km
Carbon dioxide (CO2)                                  303.5           287.6           288.8
Non methane organic compounds (NMOCs)                 0.408           0.395           0.395
Methane (CH4)                                         0.497           0.498           0.443
Carbon monoxide (CO)                                 10.957          10.635          10.617
Nitrous oxide (N2O)                                   0.016           0.021           0.019
Nitrogen oxides (NO2)                                 0.593           0.614           0.587
Sulphur oxides (SOx)                                  0.305           0.295           0.299
Particulate matter (PM)                               0.046           0.048           0.049
HFC-134a (mg)                                         0.003           0.003           0.003
CO2-equivalent GHG emissions                          318.8           303.6           304.4


7.3.2.3 Energy Balance
The energy balance results from GHGenius for gasoline and corn and wheat ethanol are
summarized in the following table. The differences between corn and wheat ethanol are all
expected based on the data and discussion presented in the earlier discussion concerning the
differences in the pathways.




                                                                                        118
CHEMINFO




 Table 47: Comparison of Energy Balance- Gasoline, Corn and Wheat
                           Ethanol (E10)

                                                                  Ethanol        Ethanol
Fuel                                               Gasoline        (E10)           (E10)
Feedstock                                          Crude oil       Corn           Wheat
                                                       Joules Consumed/Joule Produced
Fuel dispensing                                        0.0025         0.0038          0.0038
Fuel distribution, storage                             0.0053         0.0147          0.0147
Fuel production                                        0.1717         0.5290          0.5960
Feedstock transmission                                 0.0107         0.0141          0.0067
Feedstock recovery                                     0.0971         0.0953          0.0558
Ag. chemical manufacture                               0.0000         0.0839          0.1650
Co-product credits                                     0.0000        -0.0554         -0.0903
Total                                                  0.2872         0.6855          0.7517
Net Energy Ratio (J delivered/J consumed)              3.4813         1.4589          1.3303


7.3.2.4 Findings

There are some differences in the results between corn and wheat ethanol in individual stages of
the lifecycle but the differences cancel each other out in large part. There is greater uncertainty
with respect to some of the wheat ethanol numbers since large modern plants have not yet
operated in Canada. Some of these uncertainty factors will be evaluated in the next section.




                                                                                               119
CHEMINFO




                           8. Sensitivity Analysis

8.1 Introduction

In the previous section, the results for the two models that were identified that had North
American data and ethanol pathways were compared. The primary differences identified
between GREET and GHGenius were in five areas, the energy required to produce crude oil, the
allocation of energy and emissions to co-products both in petroleum refineries and ethanol
plants, the use of lime in corn production in the US, and the type of energy (coal vs. natural gas)
used in the ethanol plants, and land use emissions methodology. Other than these areas the
models had generally similar inventory data and similar approaches to modelling.

In this section, the impact on the life cycle results of changes in some of the key inventory data,
emission factors, allocation issues, future developments, and process changes are investigated.
The GHGenius model has been selected for this work because it has both corn and wheat ethanol
pathways specific to Canada and it has features such as the sensitivity solver and the Monte
Carlo tool that allow for rapid investigation of some of these issues.

A number of themes for investigation have been selected based on results from the previous
section and input from Environment Canada. Each of these themes is discussed in separate
sections below.


8.2 Process Conversion Efficiencies

When any new industry develops there is always some improvement in the efficiency of the
industry over time. In the initial stages of development conversion efficiencies can be low and
over time the industry typically develops new and innovative means of improving its
performance. The fuel ethanol industry in North America has grown over the past 25 years and
most of the issues with respect to the introduction of new technology are behind the industry.
Nevertheless the impact of process yield is investigated here for both corn and wheat ethanol
plants.

Ethanol yields in operating plants are typically 93 to 95% of the theoretical yield. The difference
being due to incomplete conversion of starch to fermentable sugars, residual sugars after
fermentation, growth of yeast mass instead of ethanol, and ethanol losses in fermentation and
distillation. The impact of increasing and decreasing the ethanol yield by 5% on the GHG
emissions of ethanol production are shown in the following figure.




                                                                                               120
CHEMINFO




                                             Figure 11: Impact of Yield on GHG Emissions

                               41,500

                               41,000
   GHG Emissions, g CO2eq/GJ




                               40,500

                               40,000

                               39,500
                               39,000

                               38,500

                               38,000
                               37,500
                               37,000

                               36,500
                                     340   350     360      370        380          390   400   410   420
                                                                  Ethanol Yield

                                                                  Corn            Wheat



These results are not what might be expected, as the yield increases the GHG emissions increase
slightly. This results from the decreasing value of the DDG co-product credit. These results
should be viewed with some caution since all that has been changed is the ethanol yield. In all
likelihood the energy requirements would also drop as the yield improved. What would be
needed to properly model the impact is a process model that is integrated into GHGenius.
Nevertheless this does illustrate the offsetting benefits that are a component of many of the
biofuel pathways, when emissions increase during one stage of the process they are often offset
in another stage.

8.3 Process Energy Requirements

It was shown earlier that the industry has made great progress in reducing the energy
requirements of ethanol plants. This advancement has been made possible not only by better
designs but also by better enzymes, better yeasts and a more thorough understanding of the
fermentation process. Many of these advancements have therefore benefited all plants and not
just new plants.

New corn ethanol plants are being built with process guarantees that use less than 9.2 MJ of
natural gas per litre and less than 0.20 kWh of electric power per litre. These plants are usually
operated with 10 to 20% less energy than the guarantee. The values that are in GHGenius for
2007 are 9.5 MJ/litre of ethanol for gas and 0.1 kWh/litre for electric power.




                                                                                                        121
CHEMINFO




Two new technologies are being introduced to the industry that have the potential to further
lower energy consumption. They are dryers that capture the latent heat of vapourization of the
evaporated water in the stack for use in the multiple effect evaporators and fractionation of the
grain prior to introduction into the process. The fractionation will reduce the non-fermentables in
the mash and the load on the drying system. Each of these systems has the potential to reduce the
energy consumption by about 1.7 MJ/litre of ethanol, thus with existing technology it may be
possible to build plants that use only 5.6 MJ of natural gas per litre, a 41% reduction from the
value currently modelled. The impact of changes of this magnitude on corn and wheat ethanol
plants is shown in the following figure.

                                  Figure 12: Impact of Lower Energy Consumption on GHG Emissions

                         45,000

                         40,000
   2eq /GJ




                         35,000

                         30,000
   GHG Emissions, g CO




                         25,000

                         20,000

                         15,000

                         10,000

                          5,000

                             0
                             0.400        0.500     0.600         0.700        0.800         0.900   1.000   1.100
                                                            Therm al Energy/Default Energy

                                                                     Corn        Wheat



If the new technologies of fractionation and high efficiency drying are implemented and the
plant energy consumption is reduced to one half of the default value in the model then this will
have a significant impact on the energy balance of corn ethanol as shown in the following table.
Note that the fuel production stage does not drop by 50% due to the electric power requirements
and all of the energy embedded in the process chemicals.




                                                                                                                 122
CHEMINFO




 Table 48: Comparison of Energy Balance- Gasoline, Corn and Future
                           Corn Ethanol

Fuel                                                Gasoline       Ethanol        Ethanol
Feedstock                                           Crude oil       Corn        Future Corn
                                                        Joules Consumed/Joule Produced
Fuel dispensing                                         0.0025         0.0038          0.0038
Fuel distribution, storage                              0.0053         0.0147          0.0147
Fuel production                                         0.1717         0.5290          0.3094
Feedstock transmission                                  0.0107         0.0141          0.0141
Feedstock recovery                                      0.0971         0.0953          0.0953
Ag. chemical manufacture                                0.0000         0.0839          0.0839
Co-product credits                                      0.0000        -0.0554         -0.0554
Total                                                   0.2872         0.6855          0.4658
Net Energy Ratio (J delivered/J consumed)               3.4813         1.4589          2.1469


8.4 Emission Factors

The impact of three emission factors in the ethanol lifecycle are investigated here. The first is the
N2O emission factor for biologically fixed nitrogen, the second is the N2O emission factor for
wheat production, and the third is the NOx emission factor for the combustion systems within
the plant.

Corn and wheat do not biologically fix nitrogen (BFN) but the DDG co-product does displace
soybean meal in animal feed rations and as noted earlier there is some uncertainty about the
quantity of N2O produced as nitrogen is fixed by bacteria in the soil. The default values in
GHGenius assume that N2O is emitted at the same rate as if synthetic fertilizer was applied. If
we assume that no N2O is emitted then the difference in GHG emissions can be seen in the
following table. This has a significant impact on the value of the co-product and the lifecycle
GHG emissions increase. While this would appear to be counter intuitive it results from the
dynamic co-product allocation within the canola and soybean systems in GHGenius and results
from more of the emissions being allocated to the oil and less to the meal in the system.




                                                                                                 123
CHEMINFO




      Table 49: Comparison of GHG Emissions- Impact of N2O Emission
                                 Factor

Fuel                                     Corn Ethanol               Wheat Ethanol
                                                  Without                    Without
Feedstock                          With BFN        BFN         With BFN        BFN
                                                       g CO2 eq/GJ
Fuel dispensing                            214           214           214           214
Fuel distribution and storage            1,452         1,452         1,452         1,452
Fuel production                         28,949        28,949        31,880        31,880
Feedstock transmission                   1,507         1,507           715           715
Feedstock recovery                       7,512         7,512         6,382         6,382
Land-use changes, cultivation            9,727         9,727        18,774        18,774
Fertilizer manufacture                   5,327         5,327        10,701        10,701
Gas leaks and flares                         0             0             0             0
CO2, H2S removed from NG                     0             0             0             0
Emissions displaced                    -17,335        -5,448       -31,786        -9,643
Total                                   37,355        49,242        38,334        60,476
Combustion Emissions                     2,099         2,099         2,099         2,099
Grand Total                             39,454        51,341        40,429        62,575

Agriculture and AgriFood Canada and Environment Canada have been working for a number of
years to improve the accuracy of the emission factors used in the calculation of the National
GHG Inventory. They are developing Canada specific factors rather than using the IPCC Tier 1
factors. In many cases these do not differ but in some cases there is a difference. One of the
differences is in the N2O emission factors for western Canada. The moisture that is available in
the system does have an impact on the N2O emission factors. There are other factors that can
also impact the results such as the cultivation practices. In the 2005 National Inventory 113 the
N2O emission factor used for western Canada ranged from 0.002 to 0.008 rather than the 0.0125
used in GHGenius. The impact of using factors from 0.002 to 0.0125 for wheat production are
shown in the following figure. The impact is quite significant and the default values in
GHGenius probably overestimate the GHG emissions for wheat ethanol. It is planned to look at
these factors in the next update of GHGenius.




113
      National Inventory Report, 1990-2005: Greenhouse Gas Sources         And   Sinks   In   Canada
http://www.ec.gc.ca/pdb/ghg/inventory_report/2005_report/tdm-toc_eng.cfm


                                                                                                124
CHEMINFO




Figure 13: Impact of N2O Emission Factor for Fertilizer Application on Wheat Ethanol
Emissions

                         45,000

                         40,000
   2eq /GJ




                         35,000

                         30,000
   GHG Emissions, g CO




                         25,000

                         20,000

                         15,000

                         10,000

                          5,000

                             0
                             0.000   0.002   0.004     0.006         0.008   0.010   0.012
                                                     N2O Em ission Factor




The third emission factor evaluated is the NOx emission factor for combustion sources at the
ethanol plant. In Table 41 it was shown that the lifecycle NOx emissions for corn ethanol were
about twice as high as they were for gasoline and that wheat ethanol emissions were similar. The
NOx emission factor in the model for natural gas combustion at the ethanol plants is about 56%
of the uncontrolled emission level but the NOx emissions are still 67 g/GJ. The use of low NOx
burners could reduce this emission to 30 to 35 g/GJ. The impact of this change on the lifecycle
emissions of NOx is shown in the following table. It can be seen that the production plant NOx
emissions are a relatively small portion of the lifecycle NOx emissions. The NOx from the
application of the fertilizer dominate the emissions.




                                                                                             125
CHEMINFO




    Table 50: Comparison of NOx Emissions- Impact of Combustion
                          Emission Factor
Fuel                               Gasoline                         Ethanol
Feedstock                                            Corn          Corn      Wheat          Wheat
                                                   Default      Lo NOx      Default        Lo NOx
                                                              g NOx/GJ
Fuel dispensing                           0.5          0.8          0.8         0.8             0.8
Fuel distribution and storage             1.8          3.5          3.5         3.5             3.5
Fuel production                          32.6         60.1         45.4        67.2            51.4
Feedstock transmission                   15.0          3.8          3.8         1.8             1.8
Feedstock recovery                       30.9         50.3         50.3        27.7            27.7
Land-use changes, cultivation             0.0        348.8        348.8       497.3           497.3
Fertilizer manufacture                    0.0         14.3         14.3        28.1            28.1
Gas leaks and flares                      0.0          0.0          0.0         0.0             0.0
CO2, H2S removed from NG                  0.0          0.0          0.0         0.0             0.0
Emissions displaced                       0.0       -300.0       -300.0      -547.3          -547.3
Total                                    80.8        181.7        166.9        79.0            63.3


8.5 Land Use
There are two aspects of land use that are investigated here; the direct land use emissions and the
indirect land use emissions. Direct land use emissions are those that arise from the production of
the actual feedstock used for biofuel production. Indirect emissions are those that potentially
arise from the production of a crop somewhere else in the world to replace the crop that was used
to make biofuels.

8.5.1 Direct Land Use Emissions

8.5.1.1 N2O Emission Factor

It was determined in the previous section that there was a significant difference in the calculation
of land use emissions between GHGenius and GREET and specifically with the emission factor
applied to N2O emissions from biologically fixed nitrogen. It was noted that this subject has been
controversial for quite some time, with some scientists (including some at Agriculture and
AgriFood Canada) insisting that there should be no N2O emissions from this source and others
(including, until recently IPCC) that insisted there are no differences between biological nitrogen
fixation and the application of synthetic nitrogen fertilizers.

This choice of emission factor is quite significant in GHGenius because it impacts not only the
emissions from growing soybeans but also from canola meal because of the system expansion



                                                                                                126
CHEMINFO




and from all of the other protein meals because of the displacement approach to co-product
allocation. In the following figure the GHG emissions for corn and wheat ethanol are shown as a
function of this choice of emission factor where one represents the same biological N fixation-
related emissions factor for N2O emissions as for synthetic fertilizer and zero represents no N2O
emissions. Values in between reflect some biological N fixation-related N2O emissions but at a
lower rate than those resulting from synthetic fertilizer.

                                     Figure 14: Impact of Biological N2O Emissions on GHG Emissions


                            70,000

                            60,000
      2eq /GJ




                            50,000
      GHG Emissions, g CO




                            40,000

                            30,000

                            20,000

                            10,000

                                0
                                 0.00     0.10   0.20   0.30   0.40      0.50      0.60    0.70   0.80   0.90   1.00
                                                               Biological N2O Em issions

                                                                      Wheat         Corn



More research on this topic is required in order to reach a consensus on these emissions. One of
the earlier papers produced in Canada (Rochette et al. (2004) 114) that was used to support the no
emission position significantly overestimated the nitrogen content of the residue returned to the
soil (a value of 3% rather than the typical actual nitrogen content of soybean residue of 0.8% was
used) and if the correct value was used then the paper supported the position that the N2O
emissions were similar to those of synthetic nitrogen.




114
   Rochette, P., Angers, D.A., Belanger, G., Chantigny, M.H., Prevost, D., Levesque, G. 2004. Emissions of N2O
from Alfalfa and Soybean Crops in Eastern Canada. Soil Science Society of America Journal. 68:493-506.


                                                                                                                   127
CHEMINFO




8.5.1.2 Soil Carbon

The calculation of differences in soil carbon between alternative land uses is quite complex. The
issue of changes in soil carbon can arise in both direct and indirect land use emission
calculations. Most of the papers on this subject use IPCC default values for soil carbon under
different types of cultivation and assume all of the carbon is lost between the current land use
and the future land use value. Other approaches have just assumed that 25% of the undisturbed
soil carbon is lost through cultivation. There is some debate if the values derived are realistic or
if they overstate the carbon loss.

There is no doubt that some cultivation practices can deplete soil carbon. Agricultural practices
in Canada and the United States in the early 1900’s undoubtedly reduced soil carbon, but as
awareness of the issue arose practices started to change. Reduced or no tillage practices are
being adopted throughout North America, inputs are generally increasing and many regions of
the continent that have been reducing soil carbon are starting to build it up again. Interestingly
no-till practices are more widely adopted in Canada than in the United States.

Canadian agricultural practices have been changing in recent years and Canada has become one
of the leading countries in the adoption of conservation and no-till management practices.
Statistics Canada 115 has reported tillage practices in Canada over the past 10 years as shown in
the following figure.




115
   StatsCan. 2007. Snapshot of Canadian Agriculture.
http://www.statcan.ca/english/agcensus2006/articles/snapshot.htm


                                                                                                128
CHEMINFO




                               Figure 15: Canadian Tillage Practices

                     60

                     50

                     40
        Percentage




                     30

                     20

                     10

                      0
                      1996                           2001                               2006
                                                     Year

                                    Conventional        Conservation         No Till




Conservation and no till practices provide many benefits including reduced herbicide and fuel
consumption, improved water efficiency and increased soil carbon content. These practices now
dominate in Canada and represent more than 70% of the acreage.

The Canadian situation is significantly different from the United States where more than 60% of
the corn and 70% of the wheat is still produced using conventional tillage and less than 20% of
the acreage is no-till (Cea Inc) 116.

GHGenius has a version of the IPCC soil carbon calculator included on sheet W. This tool
calculates changes in soil carbon content based on management practices including tillage
systems and input intensity. In the following table the changes in soil carbon content for a typical
dry Canadian soil that has been in long term cultivation with conventional tillage and medium
inputs are shown for the options available in the calculator. Land with higher moisture will have
higher rates of gain of soil carbon. Land that has less severe management would show a smaller
rate of gain. Note that the values returned by the calculator have the opposite sign as that
required in GHGenius for the calculations. The output from the IPCC tool is manually inputted
into the rest of the model.



116
   Cea Inc. 2007. Next Generation Feedstock Needs for Biorefineries.
http://www.innovation.gov.ab.ca/BioFeed/docs/Next_Generation_Feedstock.pdf


                                                                                                129
CHEMINFO




      Table 51: Soil Carbon Changes under Improved Management

                                       Reduced Tillage                 No Tillage
                                              Annual Carbon Change, kg-C/m2
Low Inputs                                             0.004                               0.014
Medium Inputs                                          0.017                               0.028
High Inputs without Manure                             0.028                               0.040
High Inputs with Manure                                0.072                               0.086


In Saskatchewan C-Green Aggregators Inc. are using soil carbon gains of 0.045 to 0.09 kg C/m2
for the annual rate of soil carbon increase in brown and black soils (0.90 to 1.8 kg C/m2 over 20
years). This is reported to have been verified by 20 years of research by Agriculture and Agri-
Food Canada at their Swift Current Research Station for all areas of Saskatchewan and it is in
line with the values derived from the IPCC tool within GHGenius.

The impact of increasing the soil carbon content due to improved tillage practices is shown in the
following figure. The rate of soil carbon change has been varied from 0.20 to -3.0 kg C/m2 (over
20 years and where – indicates a gain in soil carbon) for each of the two ethanol pathways.
Given the rate of adoption of reduced and no till management in Canada, the default soil carbon
values for direct land use emissions in the model may be too conservative, and for some of the
biofuel pathways this results in significant lifecycle emissions differences. Under high soil
carbon accumulation rates, the lifecycle GHG emissions can be almost zero for the wheat ethanol
pathways.




                                                                                              130
CHEMINFO




                                        Figure 16: Impact of Soil Carbon Change on Upstream Emissions

                                                                                                       45,000
                                                                                                       40,000
       Upstream Emissions (g/GJ)




                                                                                                       35,000
                                                                                                       30,000
                                                                                                       25,000
                                                                                                       20,000
                                                                                                       15,000
                                                                                                       10,000
                                                                                                        5,000
                                                                                                           0
                                   -0.350   -0.300   -0.250    -0.200     -0.150     -0.100   -0.050       0.000   0.050
                                                                Soil Carbon Loss (kg-C/m^2)

                                                                        Corn        Wheat



Another means of changing soil carbon is to change the level of fertilizer inputs. In the following
table the impact of changing form a low input regime to a higher input regime for Canada is
shown. The soil carbon changes from increased inputs are similar to those of improved tillage
practices but there will be offsetting emissions from nitrogen application and fertilizer
manufacture so this approach is not as fruitful as improving the tillage practice.


      Table 52: Soil Carbon Changes under Improved Management

                                                Medium Inputs       High Inputs without         High Inputs
                                                                          Manure                with Manure
                                                                  Annual Carbon Change, kg-C/m2
Full Tillage                                                  0.012               0.023                     0.065
Reduced Tillage                                               0.013               0.024                     0.067
No Tillage                                                    0.014               0.026                     0.072


Using the IPCC soil carbon tool it is possible to take non-degraded grassland in either warm and
moist, or tropical environments, and through the use of no till practices and high inputs with
manure hold the soil carbon constant and in some conditions increase the soil carbon. Since the
exercise of calculating direct or indirect land use emissions involves determining what might
happen in the future it is not possible with any degree of confidence to claim that a loss of 25%
(an estimate used in some recent papers) of the soil carbon is a more likely scenario than
increasing the soil carbon.


                                                                                                                           131
CHEMINFO




8.5.2 Indirect Land Use Emissions

Indirect land use emissions have recently become quite controversial for biofuels. The two recent
articles in Science 117 are typical of several papers and presentations made on the subject. The
papers calculate the change in above ground biomass and soil carbon for new land that is brought
into production to replace the feedstock that is diverted from its previous use to the production of
biofuels.

The issue of indirect land use is extremely complex. Most studies either assume that new land
must be brought into cultivation to meet the new demand or construct a scenario that increases
demand so quickly that new land becomes the only realistic option for increased biofuel
production. The calculation does require a look into the future to determine the amount of land
required, where this new land is located and what use it was in before being converted to
biofuels.

Since it is extremely difficult to predict the future, many of the papers on this subject choose to
illustrate a worst case scenario: if a worst case does not indicate a problem then no further
analysis is required but if it indicates an issue then further analysis is required. The worst case
involves assuming a large quantity of biofuels to be produced in a very short period of time so
that existing systems have difficulty in responding to the demand with normal or efficient
growth. The worst case also involves large soil carbon changes from cultivation even though
management practices exist to cultivate annual crops with no soil carbon losses and large losses
of above ground biomass.

Many of the papers on this subject therefore select tropical rain forests as the new land and
assume primitive agricultural practices that maximize soil loss in order to demonstrate the worst
case. Some of the papers also exclude the co-product issue (although this may be less important
for biodiesel than ethanol) and future yield increases.

In practice what should happen is that the increased demand for biofuels will lead to an increase
in demand for feedstock. This increased demand will increase the price of the feedstock and the
increased price should result in increased production.

There has not been a significant increase in cultivated land in Canada for more than 25 years as
shown in the following figure. The amount of cropland has increased in large part due to
decreases in the practice of summerfallow. New agricultural land in Canada is therefore not a
likely future scenario so an increase in domestic demand for a crop will need to be satisfied by

117
  Use of U.S. Croplands for Biofuels Increases Greenhouse Gases Through Emissions from Land Use Change.
Timothy Searchinger, Ralph Heimlich, R. A. Houghton, Fengxia Dong, Amani Elobeid, Jacinto Fabiosa, Simla
Tokgoz, Dermot Hayes, Tun-Hsiang Yu. Science. 29 February 2008: Vol. 319. no. 5867, pp. 1238 – 1240. And;
Land Clearing and the Biofuel Carbon Debt. Joseph Fargione, Jason Hill, David Tilman , Stephen Polasky, Peter
Hawthorne. Science. 29 February 2008:Vol. 319. no. 5867, pp. 1235 - 1238


                                                                                                            132
CHEMINFO




increased yield, crop shifting, or reduced exports. Reduced exports would require increased
production somewhere else in the world to meet existing demand. Some of the production may
be region – in proximity to the consuming market.

                                  Figure 17: Agricultural Land Use in Canada

                   80,000,000

                   70,000,000

                   60,000,000

                   50,000,000
        Hectares




                   40,000,000

                   30,000,000

                   20,000,000

                   10,000,000

                           0
                                1951 1956 1961 1966 1971 1976 1981 1986 1991 1996 2001 2006

                                      Total Farm Land      Land in Crops             Cultivated Land
                                      Summerfallow         Improved Pasture




8.5.2.1 Increased Production

Increased crop production can be met several ways, better management practices such as reduced
summerfallow, through increased yield on existing crop land, substitution of one use of existing
land for another, or new land being brought into production. It is very difficult to predict what
might happen in the future. There are a few papers that have looked at agricultural productivity
in developing countries and tried to analyze changes. One paper by Fulginiti and Perrin 118
considered productivity in 18 developing countries between 1961 and 1985 and concluded that
most of the change in agricultural productivity was due to changes in commercial inputs like
fertilizer and machinery. Land only accounted for about 5% of the productivity change. This
paper is not that current and it actually found that on average productivity declined over the
period but this was a time of declining prices and increase subsidies in the developed countries


118
   Fulginiti, L., Perrin, R. 1998. Agricultural Productivity in Developing Countries. Agricultural Economics 19
(1998) 45-51.


                                                                                                           133
CHEMINFO




so its conclusions may not be directly applicable to the current situation. Nevertheless 5% is
considerably different than the 100% considered in the Science papers.


8.5.2.1.1 Reduced Summerfallow

As shown above there has been a more or less continuous decrease in land summer fallowed in
Canada over the past 30 years. Summer fallow is the practice of allowing land to lie idle during
the growing season. Traditionally farmers used summerfallow one year as a risk-management
strategy to improve the chances of growing a crop the following year; during the fallow year,
farmers control weeds by either tilling the land or spraying herbicides.

Because moisture can accumulate in the soil during summerfallow, a certain dependence on the
practice developed, particularly in the areas of the Prairies with moisture deficit. In terms of CO2
production, the most significant effect of summerfallow is the increased rate of organic carbon
loss from the soil. The rate of decomposition of soil organic matter is accelerated in the moist,
warm surface soil of a summerfallowed field, resulting in the escape of mineralized carbon to the
atmosphere as CO2. Summerfallow, in conjunction with frequent tillage, also requires more
fossil fuel use than other crop production systems.

Over the long-term, the increased rate of soil organic matter decomposition coupled with no new
addition of plant residues in the summerfallow year, results in a net decrease in the amount of
organic matter stored in the soil. This means a net soil loss, not only of soil carbon, but also of
other nutrients associated with organic matter, particularly soil nitrogen. Soil nitrogen is a highly
mobile nutrient, subject to leaching and losses of nitrogen during fertilizer application and
volatilization of fertilizer nitrogen, both as nitrous oxide, can be significant.

Desjardins (2005) 119 suggested that the net impact of CO2 and nitrous oxide emissions from
summerfallow ranged from a loss of 12 to 24 g C/m2/year. This is a very significant emission
rate. Environment Canada in their 2004 National Inventory reported that N2O emissions from
summerfallow were similar to cropped land and that there was a reduction in soil organic
content. Since there is no crop produced during summerfallow it is difficult to calculate the
impact directly in GHGenius. The difference in N2O emissions between growing a crop on land
that would have been otherwise summerfallowed is zero. The crop should therefore have no
fertilizer related N2O emissions attributed to it.

Producing about 2.0 billion litres of renewable fuels in western Canada (the only region
practicing summerfallow) will require about 40% of the current summerfallow land. One of the
factors that has impacted the rate of reduction in summerfallow has been the poor economic
returns of crops grown in western Canada. Thus western Canada has the potential to increase
crop production through the continued reduction in summerfallow practices. Given the emissions
119
  Desjardins, R. 2005. Greenhouse Gas and Carbon Monitoring in the 21st Century. Presented at the BIOCAP
Canada, 1st National Conference, Ottawa, Ontario. Feb. 2005.


                                                                                                    134
CHEMINFO




of GHGs from summerfallow land the GHG emissions calculated here may be overstated since
the reference system (petroleum diesel fuel which would include summerfallow land) has
significant GHG emissions that are generally not included in any analyses.


8.5.2.1.2 Increased Yield

The alternative to increasing land in cultivation is to increase yield. Most papers do not consider
this an option, or only project yield increases at historical rates. This may not be an appropriate
assumption as there is a very large variation in yield of all crops around the world and in many
countries the yield is low because the non-subsidized price of the crop is less than the cost of
production when full fertilization is practiced. Higher prices will thus encourage more fertilizer
use and increased production far beyond the traditional rate of yield improvement. In the
following figures the top 20 producing countries for corn and wheat are shown along with the
crop production and yield with the data from the FAO database 120 for 2006. From both figures it
is apparent that yield is highly variable.

                                                  Figure 18: World Corn Production and Yield

                                 300,000,000                                                   10,000
                                                                                               9,000
                                 250,000,000
       Corn production, tonnes




                                                                                               8,000




                                                                                                        Corn Yield, kg/ha
                                                                                               7,000
                                 200,000,000
                                                                                               6,000
                                 150,000,000                                                   5,000
                                                                                               4,000
                                 100,000,000
                                                                                               3,000
                                                                                               2,000
                                  50,000,000
                                                                                               1,000
                                              0                                                0
                                                          ge dia




                                                          om d a




                                                    Ph kr ria
                                                            Br a




                                                         C ly
                                                                   a



                                                  So Hu nia

                                                           Af y




                                                 , R ilip ine
                                                         do ce




                                                          N ypt
                                                          Fr tina




                                                          et ia
                                                                   o
                                                            ex l




                                                               am
                                                         ub es
                                                            Eg a
                                                            C s

                                                          M i




                                                      Vi iop f
                                                         h ar
                                                                 n
                                                               az




                                                                si




                                                           h o
                                                                ic




                                                              Ita




                                                              ric
                                                                e




                                                        R na
                                                      I n an




                                                         U e
                                                              hi




                                                     ep n
                                                      Ar In



                                                            ne




                                                              a
                                                             at




                                                              a
                                                            ng




                                                         Et lic

                                                             N
                                                            ig
                                                             n




                                                            pi
                                                          St




                                                           a
                                          d




                                                     ut
                                        te
                                     ni
                                    U




                                               ia
                                             rb
                                          Se




                                                              Corn Production   Corn Yield




120
      http://faostat.fao.org/site/567/default.aspx


                                                                                                        135
CHEMINFO




Canada is a relatively small corn producer on a world scale. Note that the second and third
largest corn producers, China and Brazil, have much lower yields than the United States. China
uses about 85% of the nitrogen fertilizer per hectare that the US does and Brazil uses only 25%
of the nitrogen. In the case of Brazil this is due to poor economic returns that producers have
received from growing corn in an unsubsidized environment. Higher world prices can help to
address that and thus Brazil has the potential for very large increases in corn production without
any change in land area. Increasing the productivity of the top 20 corn producers to the US levels
has the potential to increase corn production by over 80% compared to 2006 levels.

Corn has been the primary feedstock for ethanol production in North America for the past 25
years. The rapid development of the ethanol industry in the United States over the past several
years has put increased attention on the ability of corn producers to meet demand. Canadian corn
production is only about 3% of US production and the supply and demand situation in Canada is
considerably different than it is in the United States as shown in the following table. The
Canadian data in the table is from Agriculture Canada, Statistics Canada with our estimates of
ethanol demand in Canada. Note that our estimate of ethanol demand in probably lower than the
Ag Canada estimate for 07/08 and this would impact on the level of imports reported for
2007/08. The US data is from the USDA. Note that the Canadian corn production is only about 3
to 4% of the size of the US production.

                        Table 53: Corn Supply and Demand

                             2004-2005           2005-2006          2006-2007       2007-2008 est
                                                        tonnes
Canada
Supply                        8,835,700            9,361000          8,990,000         11,000,000
Livestock demand              7,961,000           8,835,000          8,286,000          8,485,000
Fuel Ethanol                    375,000             375,000            975,000          1,250,000
Other Uses                    2,020,000           1,905,000          1,925,000          2,650,000
Imports                       2,419,000           1,905,000          2,100,000          1,700,000
Exports                         229,000             242,000            300,000            200,000
United States
Supply                      299,862,603        282,262,349        267,552,254         331,520,508
Livestock demand            156,371,302        156,310,349        146,031,746         146,031,746
Fuel Ethanol                 33,601,778         40,705,524         54,603,175          86,349,206
Other Uses                   34,107,937         34,501,333         34,321,778          34,725,079
Exports                      46,172,952         54,192,000         53,333,333          54,603,175

The National Corn Growers Association in the United States believes that the rate of change in
the annual corn yield is accelerating as shown in the following figure. Increased corn yield, if it
is driven by demand for ethanol will also alleviate the need for additional cropland being brought



                                                                                               136
CHEMINFO




into production. Based on recent yield trends the corn yield could reach 200 bu/acre by 2015,
one third higher than current levels.
                          Figure 19: Rate of Change in Corn Yield




Increased corn production in Canada will come through increased yield and possibly increased
substitution of other crops such as soybeans. There is a relatively small amount of corn grown in
western Canada where the low moisture levels and lower heat units limit the availability of land
suitable for corn production.
The following figure shows World wheat production and yields. Variations in crop yield can be
caused by seed variety, soil fertility, fertilizer application rates, moisture, growing season,
temperature, diseases, pests, and sunlight. Some of these factors are under the control of the
producer and choices made in the past with respect to some of these inputs may not reflect the
choices that would be made in the future, particularly when a fundamental shift in supply and
demand (and price) is made. Canadian wheat yield is quite low and reflects the dryland nature of
Canadian wheat production and the traditional focus on high protein milling varieties of wheat.




                                                                                             137
CHEMINFO




                          Figure 20: World Wheat Production and Yield

            120,000,000                                                            9.00
                                                                                   8.00
            100,000,000
                                                                                   7.00
             80,000,000                                                            6.00




                                                                                          Tonnes/ha
   Tonnes




                                                                                   5.00
             60,000,000
                                                                                   4.00
             40,000,000                                                            3.00
                                                                                   2.00
             20,000,000
                                                                                   1.00
                     0                                                             0.00
                               Fe Sta ia




                                   er da




                               m gd y
                         si i t e I n a




                                    Po ly
                                    kh ne




                               U oro d
                                   C ce




                                             t
                              Ka Uk ina




                                   ek co
                                     ki y




                                      Eg lia
                        n, ted Tu an




                                            n
                                    R m
                                    ge of
                                     ra s




                                     st n

                                          yp
                                    Fr ion




                           Is K rke
                                           n




                                  M lan
                                  P a an
                                          d




                                        Ita
                                  de te




                                         ta
                                  A u s ta
                                 G na
                                       an




                                 ic o
                                        hi




                                       ra




                                zb c
                                za rai
                                       st




                                       nt
                                Ar ep




                                      is
                                       t



                                      m
                                      C




                                     a




                             la in
                           an d
                      u s Un




                    Ira ni
                         U
                     R




                                      Wheat production      Wheat Yield




Wheat is the primary feedstock for ethanol production in western Canada. The supply and
demand situation is summarized in the following table. The fuel ethanol demand (from wheat)
for 2007-08 is our estimate. The fuel ethanol demand remains a relatively small portion of the
total demand for wheat.

                          Table 54: Wheat Supply and Demand

                               2004-2005        2005-2006          2006-2007   2007-2008 est
                                                       tonnes
Canada
Supply                        25,860,000       20,860,000         21,919,000      17,700,000
Livestock demand               3,884,000        4,605,000          4,225,000       4,245,000
Fuel ethanol                     110,000          110,000            410,000         900,000
Other Domestic Uses            3,980,000        3,663,000          3,376,000       3,179,000
Exports                       14,812,000       11,499,000         14,675,000      11,000,000
Imports                           14,000           23,000             25,000          24,000



                                                                                                      138
CHEMINFO




8.5.2.1.3 Crop Substitution

Some ethanol plants in western Canada have been switching to use a soft white wheat as their
feedstock instead of one of the other varieties such as Canadian Prairie Spring (CPS). The soft
white wheats (such as AC Andrew) provide a number of advantages including higher yield for
the producer and a higher starch content for the ethanol plant. This reduces the energy
requirements for the plant since less DDG is produced and higher ethanol contents can be
processed.

Along with a higher starch content, the crop has a lower protein level and this requires a lower
nitrogen fertilizer rate per unit of output and thus the producers input costs are also lower.

The characteristics of the two feedstocks of interest are summarized in the following table. These
values are the average values over two sites in 2005. The starch content can vary by +/- 2% and
the protein content by about +/- 0.5% between sites and between years.

                           Table 55: Wheat Characteristics

                                                           CPS                        AC Andrew
Starch, % wt dry basis                                       65                              69
Protein, % wt dry basis                                    12.4                            11.2

The default wheat ethanol cycle in the model was set up using CPS wheat as the input. For AC
Andrew it is expected that the per unit nitrogen requirement will be lower by the difference in
protein in the wheat compared to CPS wheat. That is instead of 20 kg/tonne of grain the nitrogen
fertilizer input would be 18 kg/tonne. The rest of the fertilizer inputs are not as energy intensive
and no changes will be made to them. The energy requirements of the ethanol plant are reduced
by 2 MJ/litre (53 litres of natural gas/litre) when the Andrew wheat is processed (rather than the
CPS) based on a process model simulation.




                                                                                                139
CHEMINFO




                   Table 56: GHG Emissions Impact of Wheat Variety

Feedstock                                                                CPS          AC Andrew
                                                                        g CO2 eq/GJ
Fuel dispensing                                                           214                 214
Fuel distribution and storage                                           1,452               1,452
Fuel production                                                        31,880              27,604
Feedstock transmission                                                    715                 689
Feedstock recovery                                                      6,382               6,149
Land-use changes, cultivation                                          18,774              17,001
Fertilizer manufacture                                                 10,701               9,705
Gas leaks and flares                                                        0                   0
CO2, H2S removed from NG                                                    0                   0
Emissions displaced                                                   -31,786             -29,488
Total                                                                  38,334              33,326

The reduction in GHG emissions is quite significant with the change in variety. There are also
supply benefits if the wheat yield is increased by 40% from the same land base. The increased
yield combined with the displacement of DDG and the availability of summerfallow land
presents a supply situation with wheat that could see considerably more wheat diverted to biofuel
production without any impact on exports and therefore no land use emissions for the cultivation
of additional crops.

8.5.2.2 New Land

Arable land requires sufficient moisture, the proper soil and climatic conditions in order to
produce a crop. All crops are also not suited for all arable land since they may require specific
climatic conditions or amounts of moisture. Quantifying the amount of arable land in the world
is not a simple task. The total land that has the appropriate soil and climatic conditions must be
discounted for protected areas (sensitive areas or land used for human settlements), for land that
cannot be physically accessed and other factors such as the use of irrigation. It is well known that
there are wide variations in the quality of data between developed and developing countries. One
estimate has been prepared by the FAO (World Soils Report) 121.




121
      FAO. 2000. World Soils Report. ftp://ftp.fao.org/agl/agll/docs/wsr.pdf


                                                                                                140
CHEMINFO




                   Table 57: Arable Land for Rainfed Agriculture

                                    Gross         Protected         Net           Actual         % of
                                   potential        land         potential        arable       Potential
                                    arable                        arable           land         arable
                                     land                          land                          land
                                                                                              actually in
                                                                                              use (1994)
                                                         (1000 ha)
Sub-Saharan Africa                  1,119,492         69,409 1,050,083            157,608               15
North Africa and Near East             50,017          5,202       44,815          71,580              160
North Asia, east of Urals             286,800         10,898      275,902         175,540               64
Asia and the Pacific                  812,551         69,879      742,672         477,706               64
South and Central America           1,048,071         68,125      979,946         143,352               15
North America                         463,966         32,478      431,488         233,276               54
Europe                                363,120         39,217      323,903         204,322               63
World                               4,144,017        295,208 3,848,809          1,463,384               38

This data set indicates that only about 38% of the potentially available arable land is currently
being utilized for crop production and thus some expansion of crops is potentially achievable.
Most of the arable land that is not in use is classified as permanent pasture in most land
inventories. It would therefore appear that some expansion of biofuel feedstock is potentially
possible with the conversion of this pastureland without considering converting forestland.

Some of the potential arable land will be required to support increasing populations, and
changing diets (due in part to increased wealth). For example, meat requires three to eight times
the equivalent weight of grain to produce a pound of food, and as more meat is included in diets
it places a strain on the grain supply systems.

Moreira (2004) 122 reports that a study carried out by IPCC considered that the total land crop
potential is 2.49 billion ha from which 0.90 billion ha was already in use for food production in
1990, and that an additional 0.42 billion ha will be required to feed the human population by the
year 2050. This still leaves 1.28 billion ha of land for extra biomass production, which could be
used for energy purposes. Moreira also notes that some of this land could be used for managed
forests for biomass production for energy production systems and these conversions could be
expected to have the opposite impact of deforestation, that is, they would result in above ground
carbon storage and soil carbon increases.

122
  Moreira, J., R. 2004. Global Biomass Energy Potential. Paper prepared for the Expert Workshop on Greenhouse
Gas Emissions and Abrupt Climate Change: Positive Options and Robust Policy, Paris 30 September – 1 October
2004. http://www.accstrategy.org/draftpapers/Moreira.doc



                                                                                                         141
CHEMINFO




It would appear that increased biofuel production is possible without significant deforestation
due to the availability of underutilized arable land. It is also likely that due to climatic conditions
forestland may not be suitable for biofuel crop production. Furthermore, second generation
biofuels may even result in a conversion of arable land to forestland with the accompanying
carbon sequestration benefits.

It should be noted that biofuels are not the only energy system that results in the potential for
carbon emissions through disturbed soils and loss of growing biomass. Coal mining, oil and gas
production all result in land use changes. Land use changes are calculated in the GHGenius
model for some of these other systems but they generally consider an average over all production
and not the incremental production caused by increased energy demand.


8.5.2.3 Above Ground Biomass

The worst case scenarios in the literature all assume that rain forests are cleared to provide
additional arable land. Notwithstanding that the most likely source of additional cropland is
converted grassland or pasture, it is possible that in some regions forests could be cleared to
provide additional cropland. In these cases, an historical store of carbon in the form of above
ground wood is removed from the ecosystem and that wood will either be used for fuel where the
carbon is immediately added to the atmosphere or used for other purposes such as lumber where
the wood will be used for some period of time and then at the end of its useful life be combusted
or allowed to decompose.

These forests do store significant amounts of carbon for relatively long periods of time. The
release of this carbon in a short period of time can have a very significant impact on the lifecycle
emissions of the some biofuel pathways even when it is amortized over long periods of time. The
real issue is how likely is it that this will happen in the future.


8.6 Feedstock Production

The input values for feedstock production in GHGenius are based on average values in Canada.
As such they represent a combination of cultivation practices. The Ontario Ministry of
Agriculture Food and Rural Affairs (OMAFRA) publishes a crop budget template for production
in that province. In the budget for corn 123 they provide input requirements for conventional,
minimum and no till practices. In the following table the differences in the three practices are
summarized.



123
      Field Crop Budgets. http://www.omafra.gov.on.ca/english/busdev/facts/pub60.htm#grain


                                                                                                   142
CHEMINFO




                               Table 58: Corn Crop Inputs

                              Conventional Till             Minimum Till                    No Till
Projected yield                        8.4 t/ha                  8.4 t/ha                   8.4 t/ha
Fuel                                  43.5 l/ha                 38.7 l/ha                   29 L/ha
Fuel                               5.2 l/tonne                4.6 l/tonne               3.46 l/tonne
Herbicide as % of                          100                        100                        122
control
Operating Labour                     $12.95/acre                $9.65/acre               $ 7.20/acre

There may be secondary impacts such as higher yield on no-till (after several years) and
increased soil carbon form reduced tillage but we will only consider the impact of fuel and
herbicide usage.

The diesel fuel usage for corn in GHGenius is 6.1 litres/tonne, which is slightly higher than
shown here but would also include the fuel used to move the crop from the field to the farm and
any custom work that is covered in other parts of the OMAFRA budget. The GHG emissions for
the three cases (where the conventional till is the default case in GHGenius) are shown in the
following table. The impact of these direct changes is quite small. Note a portion of the benefit is
offset by the reduced co-product credit for the more efficient corn production. It is likely that the
secondary impacts of higher yield and higher soil carbon content would have larger impacts on
GHG emissions than the direct impacts of reduced fuel consumption.




                                                                                                 143
CHEMINFO




       Table 59: GHG Emissions Impact of Tillage Practice on Corn
                             Emissions

                                     Conventional Till         Minimum Till           No Till Corn
Feedstock                                       Corn                   Corn
                                                             g CO2 eq/GJ
Fuel dispensing                                    214                  214                     214
Fuel distribution and storage                    1,452                1,452                   1,452
Fuel production                                 28,949               28,949                  28,949
Feedstock transmission                           1,507                1,507                   1,507
Feedstock recovery                               7,512                7,178                   6,569
Land-use changes, cultivation                    9,727                9,727                   9,727
Fertilizer manufacture                           5,327                5,327                   5,503
Gas leaks and flares                                 0                    0                       0
CO2, H2S removed from NG                             0                    0                       0
Emissions displaced                            -17,335              -17,269                 -17,183
Total                                           37,355               37,086                  36,739


8.7 Co-Products

GREET and GHGenius follow the same basic approach (the displacement method) for
accounting for ethanol production co-products. This is generally accepted to be the preferred
approach in scientific LCA work. Other approaches that have been used include allocating
emissions between the products based on mass, energy content, or economic value or trying to
determine the energy associated with specific parts of the process and associating that with one
or the other product.

There are advantages and disadvantages with each approach. Using economic value the co-
product allocation will change on a day by day basis as the market attributes different values to
the products. In addition should incentives be included in the value or not? The mass and energy
content approaches are simple and generally provide the highest co-product credits. The process
chain approach can still require some judgement since while the energy required for DDG drying
can be allocated to the DDG how does one allocate the cooking and distillation energy between
ethanol and DDG since both products move through that portion of the system? The
displacement method probably provides the best estimate but it requires some understanding of
the how the product is used and a forecast of what might be used if this product was not
available. Regulators find this aspect particularly difficult and are moving to other approaches
for their needs. In the following table estimates of the size of the co-product credits for corn and
wheat DDG using the various approaches are provided.



                                                                                                144
CHEMINFO




                           Table 60: Allocation Approaches

                                                    Corn DDG                        Wheat DDG
                                                % emissions allocated to co-product
Displacement (in GHGenius)                                 31.7                            45.3
Mass                                                         50                             56
Energy                                                       35                             50
Economic value                                               15                             21
Process energy                                               35                             40

Ethanol plants also produce carbon dioxide and there are at least three different ways this can be
treated in the models. In the first case the CO2 may just be vented to the atmosphere in which
case there are really no implications from a LCA perspective and that is the base situation that
has been modelled here. Some ethanol plants capture the CO2 and use it for industrial
applications. This gas is not sequestered as it eventually finds its way into the atmosphere but if
the gas displaces CO2 that was captured from a power plant or oil refinery (which are less energy
efficient sources) then there may be a small energy credit available to the ethanol plant. The third
approach would be to capture the CO2 and sequester it underground. Since this gas is not
normally counted in national inventories, as it is biogenic a credit can be generated for removing
this gas from the atmosphere on a permanent basis. A few studies have given very large credits
to CO2 on the basis that it displaces fuel that is combusted just to produce CO2 but this approach
to industrial gas production has not been practiced in North America for many years. In the
following table the GHG emissions for corn and wheat ethanol are summarized for each of the
three primary approaches.


                       Table 61: Alternative CO2 Approaches

                                                        Corn DDG                    Wheat DDG
                                              Production Lifecycle Emissions, g CO2 eq/GJ
No Treatment (GHGenius default)                             37,355                       38,334
Industrial gas                                              31,825                       32,810
Sequester                                                   11,484                       12,195

Both alternative approaches to the utilization of CO2 provide GHG emissions benefits. The
possibility of sequestering the gas is quite attractive as this is also one of the most efficient and
low cost options to carbon capture and sequestration.




                                                                                                 145
CHEMINFO




8.8 Water, Solid Waste and Other Environmental Impacts

GHGenius (and GREET) are not currently capable of conducting LCA related to water
consumption, competing uses of water, and pollutant releases to water. This is an area that could
be considered for further research and possible development. Consideration should be given
regarding alternative methods for managing water use and pollutant releases versus adopting a
life cycle analysis approach.

Data challenges with respect to developing an LCA approach include developing representative
water consumption data and associated water pollutant release data. These data sets are not as
well developed as are air emissions and energy data sets. Interpretation of results would also
require care, since the environmental effects are not easily normalized (e.g., as in greenhouse gas
emissions contributing to global warming potential). In addition, the results of water
consumption may need to be considered in context of regional water availability and competing
uses. For example, in eastern Canada, the importance of water consumption may be different
than on the prairies.

8.9 Uncertainty: Monte Carlo Analysis

The sensitivity of the GHG emissions performance with respect to various input values has been
discussed when each parameter is analyzed individually, but what is the impact of changing
multiple variables at one time? This multiple variable analysis can be evaluated using the Monte
Carlo simulator in GHGenius.

Monte Carlo simulations are a means of solving a numerical problem that cannot easily be
solved by means such as integral calculus or other numerical methods. A Monte Carlo simulation
involves using many iterations of random inputs to determine a set of outputs. The main strength
of using a Monte Carlo simulation is the ability to address the uncertainty of input values and to
determine the impact of this uncertainty on the output or results. Monte Carlo simulation is well
suited to computer solutions where multiple random numbers are generated (within user defined
boundaries) that are used as input values and their impact on the results is determined. A
limitation of Monte Carlo analysis is that one must have reasonable information about the
distributions of the variables to be examined which is not always the case for the fuels being
examined.

To do the Monte Carlo analysis the impact of variables will be analyzed in the lifecycle GHG
emissions for both corn and wheat ethanol will be analyzed. The following five variables will be
adjusted for the corn ethanol pathway:

   •   The N2O emission factor for the biological fixation of nitrogen will be varied from zero
       to 100% of the emission factor for synthetic nitrogen. It will be assumed that the random
       values will form a uniform distribution.


                                                                                               146
CHEMINFO




   •   The fuel used in the crop production will have a normal distribution with a mean value of
       6.1 litre/tonne (the default value) and a standard deviation of 1.5 litres/tonne.
   •   The thermal energy at the ethanol facility will have a mean value of 11.15 MJ of natural
       gas per litre of ethanol with a log normal distribution and a standard deviation of 2.
   •   The electrical energy at the ethanol facility will have a mean value of 0.12 kWh of power
       per litre of ethanol with a log normal distribution and a standard deviation of 0.02.
   •   The change in soil carbon due to the cultivation of corn will have a mean value of zero, a
       normal distribution and a standard deviation of 0.10 kg C/m2. This is a relatively small
       annual change.

These variables have been chosen for two reasons, the first is that there is some uncertainty with
respect to the appropriate values because of a lack of good commercial Canadian data or
uncertainty surrounding the underlying issue and because the values have been shown to have
some impact on the lifecycle emissions. The distributions were chosen to also reflect the degree
of uncertainty of the variable. Four thousand iterations were run using the input criteria shown
above and the output for the full lifecycle emissions were tracked. In the following figure the
distribution of the lifecycle GHG emissions per GJ of energy delivered resulting from the Monte
Carlo analysis is shown.




                                                                                              147
CHEMINFO




                                 Figure 21: Results from Monte Carlo Simulation for Corn Ethanol


                         6.0%


                         5.0%


                         4.0%
   Frequency of Values




                         3.0%


                         2.0%


                         1.0%


                         0.0%
                            27,089 30,721 34,352 37,984 41,615 45,247 48,878 52,510 56,141 59,772

                         -1.0%
                                                       GHG Emissions, g CO2eq/GJ



Not all of the stages of the lifecycle emissions changed as a result of the Monte Carlo simulation.
In the following table the default value for corn ethanol is compared to the mean values for the
simulation along with a few comments.




                                                                                                    148
CHEMINFO




 Table 62: Comparison of Corn Ethanol GHG Emissions- Monte Carlo
                            Simulation

Fuel                          Gasoline       Ethanol     Ethanol                       Comments
                                                          Monte
Feedstock                     Crude oil         Corn       Carlo
                                                         g CO2 eq./GJ
Fuel dispensing                    137           214        214                       No Change
Fuel distribution and                                                                 No Change
storage                            559         1,452       1,452
Fuel production                 12,276        28,949      28,943             Std Dev 4,208 g/GJ
Feedstock transmission             926         1,507       1,507                     No Change
Feedstock recovery               6,417         7,512       7,536               Std Dev 824 g/GJ
Land-use changes,                    2         9,628       9,642               Std Dev 767 g/GJ
cultivation
Fertilizer manufacture                0        5,327       5,327                      No Change
Gas leaks and flares              2,109            0           0                      No Change
CO2, H2S removed from                                                                 No Change
NG                                    0            0           0
Emissions displaced                                                   Std Dev 3,414 g/GJ. Mid
                                                                   Range of the MC sim is not
                                     0       -17,316     -11,385 the same as the default value
Total                           22,426        37,275      43,237           Std Dev 5,550 g/GJ

The exercise is also repeated for wheat ethanol. The following five variables will be adjusted for
this pathway:

            •   The N2O emission factor for the biological fixation of nitrogen will be varied
                from zero to 100% of the emission factor for synthetic nitrogen. It will be
                assumed that the random values will form a uniform distribution.
            •   The fuel used in the crop production will have a normal distribution about the
                mean (default value) of 10.83 litres/tonne. The standard deviation will be 2
                litres/tonne.
            •   The thermal energy at the ethanol facility will have a mean value of 320 litres of
                natural gas per litre of ethanol with a log normal distribution and a standard
                deviation of 30.
            •   The electrical energy at the ethanol facility will have a mean value of 0.34kWh of
                power per litre of ethanol with a log normal distribution and a standard deviation
                of 0.03.




                                                                                              149
CHEMINFO




                                •     The change in soil carbon due to the cultivation of wheat will have a mean value
                                      of zero, a normal distribution and a standard deviation of 0.10 kg C/m2. This is a
                                      relatively small annual change.

Four thousand iterations were run using the input criteria shown above and the output for the full
lifecycle emissions were tracked. In the following figure the percentage change in lifecycle GHG
emissions per GJ of fuel delivered are shown.

                                    Figure 22: Results from Monte Carlo Simulation for Wheat Ethanol

                         4.5%

                         4.0%

                         3.5%
   Frequency of Values




                         3.0%

                         2.5%

                         2.0%

                         1.5%

                         1.0%

                         0.5%

                         0.0%
                            32,087 35,619 39,152 42,684 46,217 49,749 53,282 56,814 60,347 63,879
                                                             GHG Emissions, g CO2eq/GJ



Not all of the stages of the lifecycle emissions changed as a result of the Monte Carlo simulation.
In the following table the default value for wheat ethanol is compared to the mean values for the
simulation along with a few comments.




                                                                                                                    150
CHEMINFO




Table 63: Comparison of Wheat Ethanol GHG Emissions- Monte Carlo
                           Simulation

Fuel                         Gasoline       Ethanol     Ethanol                       Comments
                                                         Wheat
                                Crude                    Monte
Feedstock                          oil       Wheat        Carlo
                                                        g CO2 eq./GJ
Fuel dispensing                    137          214         214                       No Change
Fuel distribution and                                                                 No Change
storage                            559        1,452       1,452
Fuel production                 12,276       31,880      32,294              Std Dev 1,618 g/GJ
Feedstock transmission             926          715         715                      No Change
Feedstock recovery               6,417        6,382       6,360              Std Dev 1,184 g/GJ
Land-use changes,                    2       18,774      16,727              Std Dev 2,428 g/GJ
cultivation
Fertilizer manufacture               0       10,701      10,701                       No Change
Gas leaks and flares             2,109            0           0                       No Change
CO2, H2S removed from                                                                 No Change
NG                                   0            0            0
Emissions displaced                  0      -31,786      -20,066     Std Dev 6,395 g/GJ. Mid
                                                                  Range of the MC sim is not
                                                                the same as the default value
Total                           22,426       38,334      48,398           Std Dev 6,826 g/GJ

The impact of N2O emissions on land use and co-product credits is much larger for wheat
ethanol than it was for corn ethanol as can be seen by the larger standard deviation for these two
categories. However, these two categories move in opposite directions and tend to cancel each
other out so that the standard deviation on the % reductions in GHG emissions is not
significantly different for the two fuels.




                                                                                              151
CHEMINFO




        9. Recommendations for Enhancing LCA
             Modelling Capacity in Canada

9.1 Summary

Life cycle analysis can be a valuable and power tool for informing policy-makers, stakeholders
and the public regarding the environmental loadings and potential impacts related to alternative
transportation fuels as well as other products. This report has identified many models that are
available for potential further development, as well as two models (GREET and GHGenius) that
had sufficient data and capabilities to support analysis for this study. One of the models –
GHGenius – has been developed in Canada and contains a robust set of data to support LCA
analysis for transportation fuels in Canada. The analysis identified various input data, boundary,
energy/emissions allocation and other issues that can arise with models and can lead to
differences in results and interpretation. To keep up with the rapid pace of biofuel development,
Canada's LCA modelling capacity needs to continue to keep abreast of new data, to analyze
relevant new biofuel pathways including new technologies that influence industrial performance,
and strive to meet the continual demands for up-to-date documentation and increased demands
for quality.

The following recommendations for enhancing Canada's LCA modelling capacity are provided
to Environment Canada for their consideration:

 •   Continue to work in partnership with government departments and stakeholders;
 •   Continue to support enhancements in transparency and documentation;
 •   Remain engaged in understanding international LCA models and results;
 •   Develop long term vision and path forward for Canada's LCA capacity;
 •   Consider support for exporting Canadian LCA capacity; and
 •   Enhance GHGenius model for Canadian conditions.


9.2 Enhance GHGenius Model for Canadian Conditions

It is suggested that consideration be given to incorporate water and land use, as well as pollutant
releases into existing model capacity in Canada (i.e., GHGenius). The incorporation of these
elements could follow a feasibility analysis to assess the available supporting datasets,
approach/methodologies, expected results along with costs. There may be challenges in
incorporating these elements (e.g., potential lack of representative water/land data, lack of
complete pollutant release inventories against which to calibrate), and normalization issues with



                                                                                               152
CHEMINFO




respect to environmental damages. However, some of these challenges needed to be overcome at
early stages of LCA related to air emissions.

Other ways to enhance GHGenius that could be considered are:

 •   incorporate additional modes of transportation (rail, marine, etc.) and segmented
     transportation fuel markets where biofuels could be used;
 •   assess how to leverage GHGenius to apply it for total national emissions inventory
     estimates and forecasting;
 •   add more pathways (e.g., in-situ oil sands mining, etc.);
 •   assess ways to incorporate damage functions for environmental

9.3 Work in           Partnership         with      Government           Departments          and
Stakeholders

It is recommended that Environment Canada, in collaboration with other federal government
departments, provincial governments as well as stakeholder organizations, continue to enhance
Canada's LCA modelling capacity. This will ensure that Canadian decision-makers have access
to the world's best possible models and correspondingly the best possible information on which
to base the important decisions related to the biomass-based transportation fuels, traditional fuels
and the environment. Environment Canada's government partners in this area would include, but
not necessarily be limited to:

 •   Natural Resources Canada;
 •   Agriculture and Agri-Food Canada;
 •   Transport Canada;
 •   National Research Council
 •   Industry Canada;
 •   Health Canada; and
 •   Environment and resource departments of respective provincial/territorial governments.

Industry stakeholders would include representatives from the biofuels industry (i.e., existing and
potential producers), the Canadian Renewable Fuel Association (CRFA), traditional fossil fuel
based suppliers (oil and gas and petroleum refining industry), the Canadian Association of
Petroleum Producers (CAPP), Canadian Petroleum Producers Institute (CPPI), Canadian
Chemical Producers Association (CCPA), and other industry groups. Agriculture sector
participants are also important stakeholders with respect to biofuels. Example of these
stakeholder groups would include, but not be limited to Canadian Canola Growers Association
(CCGA), Ontario Soybean Growers Association (OSGA), Western Canadian Wheat Growers
Association (WCWGA), Ontario Corn Growers Association (OCGA), and other similar
organizations. Environmental non-government organizations (ENGOs) and other civil society
organizations also important stakeholders.


                                                                                                153
CHEMINFO




9.4 Continue to Support Transparency and Documentation

The credibility of LCA models is largely based on their transparency and accompanying
documentation, which explains data sources, assumptions, algorithms and other elements. One
advantage of free publicly-available models is that they are transparent. However,
documentation for “free” models is an expensive and time-consuming undertaking. With free
models government and/or academic institutions are usually involved in supporting model
development as well as documentation requirements. Dedicated funds must be available to do
this. Environment Canada and its government partners need to take stock of transparency
requirements for Canadian and international models and encourage or support documentation
improvements. In particular, mechanisms may need to be identified and considered to encourage
transparency and documentation enhancements to international models, and in particular if these
models are yielding Canadian results or results that are indicated to be applicable to Canada.


9.5 Remain Engaged With International LCA Models and Results

In a world of biofuel trade, it is imperative for countries to understand one anothers’ LCA
results. The ease of publication and communication through the Internet, as well as
collaboration at industry conferences or other present the opportunity to exchange LCA findings
with a broad audience. Other countries’ results that are generated on biofuels may differ from
results generated in Canada to the degree that Canadian LCA findings are brought into question
for valid or invalid reasons. Canadian decision-makers need to keep abreast of the LCA field and
understand the models and their assumptions on which these models are based. This report has
served this to some degree, but additional results from other models are likely to come out in the
future. It is suggested that Environment Canada continue to explore its understanding of the
LCA models used by other countries.

One step to better understanding would be to conduct detailed analysis using the most commonly
used models (such as GaBi and SimaPro). Environment Canada should investigate their data and
features in detail, and assess their development in Canada.


9.6 Develop Long Term Vision and Path Forward for Canada's LCA
Capacity

It suggested that Environment Canada work with other government departments as well as
stakeholders to develop a long term common vision and craft the corresponding path forward for
LCA modelling in Canada. The long term vision may need to address current LCA limitations




                                                                                              154
CHEMINFO




and consider how LCA can be integrated with or systematically inform other Canadian
environmental policy modelling activities covering transportation fuels.

The long term path forward will need to be developed in the context of available financial
resources, expertise and driven by client requirements for LCA results. Further research is
suggested to better define the necessary amount of resources, available resources and the details
of LCA client requirements.

A plan for the path forward should also be developed in context of other modelling activities
being conducted or contemplated by other federal government departments. This would include
various groups within Environment Canada (e.g., Current Analysis and Modelling, Strategic
Policy Branch), as well as Agriculture and Agri-Food Canada, Health Canada, Transport
Canada, Natural Resources Canada and the National Research Council. Some of these
departments/groups are developing models that may be related to biofuels policy. While others
are supporting the development of new technologies or new biofuel pathways that need to be
assessed for their potential environmental performance from a lifecycle perspective. Common
awareness of the scope, approaches, methods, data sources and assumptions being applied by the
various groups would be beneficial for minimizing overlap and harmonizing various elements.
This would help ensure that the results from the various activities generate consistent results.


9.7 Export Canadian LCA Capacity

Developing and enhancing Canada's LCA biofuels modelling capacity may have benefits for
exporting Canada's expertise in this field to the international marketplace. Consideration could
be given by Environment Canada to assist in promotional or other development efforts to assist
countries or organizations that may need to use Canadian LCA modelling for their growing
biofuel industries. This may require promoting Environment Canada's and other Canadian
expertise not only in the field of LCA modelling, but also in the related fields of energy data
analysis, emission inventories, industrial process knowledge, environmental effects,
transportation fuels, computer modelling and policy development.

Helping other countries and international organizations to further develop their LCA modelling
capabilities, possibly with Canadian models and expertise would in turn help Canadian LCA
modelling capacity since the costs of making improvements and continuously updating Canadian
focused models capacity can be spread out over a larger base of potential clients. These clients
may also provide input and demands to improve modelling capacity over time.

One potential mechanism to foster Canadian LCA and related modelling activity is to cultivate
an LCA biofuels cluster of expertise. The field of life cycle analysis is multi-disciplinary (as
noted above). Members of the cluster could include academic, government, industry experts
(including consultants) and client groups. This might include 100 people, or so. This could be a
virtual cluster without specific geographic locus. Collaboration and communication would be


                                                                                             155
CHEMINFO




carried out among the group members through government facilitation and could include
workshops, small conferences, and other media. The exact role of government involvement
cannot be prescribed here, and it is assumed that it would need to evolve. This collaboration and
communication already takes place through various mechanisms (e.g., industry conferences,
journals, Internet), however, the intent for Environment Canada would be provide greater
support to accelerate learning and development of Canadian capabilities.




                                                                                             156
CHEMINFO




                      10. References and Websites
All websites were access between October 2007 and March 2008. Dates (year) of publishing for
some information sources were not available.

 •   (S&T)² Consultants Inc. and Meyers Norris Penny LLP, July 2006. Canadian Renewable Fuels Strategy. For
     Canadian Renewable Fuels Association
 •   (S&T)2 Consultants Inc., 2007, Crude Oil GHGenius Update Report.
     http://www.ghgenius.ca/reports/2007CrudeOilUpdateReport.pdf
 •   (S&T)2 Consultants Inc., 2007, GHGenius US Data Update,
     http://www.ghgenius.ca/reports/USDataUpdate.pdf
 •   (S&T)2         Consultants        Inc.,       Biodiesel        GHG                 Emissions           Update.
     http://www.ghgenius.ca/reports/NRCanbiodieselghgemissionsupdate.pdf
 •   Ahmed, I., Decker, J. & Morris, D. 1994. How much energy does it take to make a gallon of soydiesel?
     Institute for Local Self-Reliance. Minneapolis, Minnesota.
     http://www.carbohydrateeconomy.org/library/admin/uploadedfiles/How_Much_Energy_Does_It_Take_To_
     Make_A_Gallon_.pdf
 •   Australia Greenhouse Office. Comparison of Transport Fuels. Final Report. Life-cycle Emissions Analysis of
     Alternative Fuels for Heavy Vehicles.
 •   Beer, T., Grant, T, 2007, Life-cycle analysis of emissions from fuel ethanol and blends in Australian heavy
     and light vehicles, Journal of Cleaner Technology, Volume 15, Issue 8-9. 2007
 •   Blottnitz, H., Curran, MA., 2007, A review of assessments conducted on bio-ethanol as a transportation fuel
     from a net energy, greenhouse gas, and environmental life cycle perspective, Journal of Cleaner Technology,
     Volume 15, Issue 7. 2007
 •   California Air Resources Board. EMFAC. http://www.arb.ca.gov/msei/onroad/latest_version.htm
 •   Cea Inc. 2007. Next Generation Feedstock Needs for Biorefineries.
     http://www.innovation.gov.ab.ca/BioFeed/docs/Next_Generation_Feedstock.pdf
 •   Center for Clean Products and Clean Technologies, University of Tennessee (1996), Evaluation Of Life-Cycle
     Assessment Tools.
 •   Ceuterick D. and Spirinckx C., 1997, Comparative LCA of biodiesel and fossil diesel fuel, VITO report
 •   Delucchi, M. 2002. Overview of the Lifecycle Emissions Model
 •   Delucchi, Mark A. ITS-Davis. December 2003. APPENDIX A: Energy Use and Emissions from the Lifecycle
     of     Diesel-Like  Fuels    Derived     From   Biomass.   Publication No.  UCD-ITS-RR-03-17A.
     http://www.its.ucdavis.edu/publications/2003/UCD-ITS-RR-03-17A.pdf
 •   Desjardins, R. 2005. Greenhouse Gas and Carbon Monitoring in the 21st Century. Presented at the BIOCAP
     Canada, 1st National Conference, Ottawa, Ontario.
 •   Dornburg, V., Lewandowski, I., Patel M., (2003) Comparing the Land Requirements, Energy Savings, and
     Greenhouse Gas Emissions Reduction of Biobased Polymers and Bioenergy. An Analysis and System
     Extension of Life-Cycle Assessment Studies Journal of Industrial Ecology 7 (3-4) , 93–116



                                                                                                               157
CHEMINFO




•   Elsayed, M., Mathews, R., Mortimer, N. 2003. Carbon and Energy Balances for a Range of Biofuels Options.
•   Environment Canada. 2006. CAC Emissions Summaries. http://www.ec.gc.ca/pdb/cac/
•   Environment Canada. 2007. Toward an Updated Energy, Emissions and Economy Baseline For the
    Regulatory Framework for Air Emissions, Emissions Projection Working Group Meeting, November 7-8,
    2007
•   EU DG Environment. 2005. Making Life Cycle Assessment Information and Interpretive Tools Available.
•   European Union Joint Research Centre. December 2003. Well – to Wheels Analysis of Future Automotive
    Fuels and Powertrains in the European Context.
•   FAO. 2000. World Soils Report. ftp://ftp.fao.org/agl/agll/docs/wsr.pdf
•   Farrell, A. E., and Sperling, D. August 1, 2007. A Low-Carbon Fuel Standard for California. Part 1:
    Technical Analysis. Available at http://www.energy.ca.gov/low_carbon_fuel_standard/UC-1000-2007-002-
    PT1.PDF
•   Farrell, A.E., Plevin, R.J., Turner, B.T., Jones, A.D., O’Hare, M., Kammen, D., 2006. Ethanol can Contribute
    to Energy and Environmental Goals. Science, 311:506-508.
•   Fava, J., Nov. 2005. Can ISO Life Cycle Assessment Standards Provide Credibility for LCA? Building Design
    & Construction, www.bdcnetwork.com
•   Fleming, J.S., Habibi, S., MacLean, H.L. 2006. Investigating the Sustainability of Lignocellulose - Derived
    Light-Duty Vehicle Fuels through Life Cycle Analysis. Transportation Research Part D: Transport and
    Environment. 2006. 11. 146-159.
•   Fulginiti, L., Perrin, R. 1998. Agricultural Productivity in Developing Countries. Agricultural Economics 19
    (1998) 45-51.
•   Government of Canada. 2007. Regulatory Framework for Air Emissions.
•   Government of Canada. December 20, 2006. Canada's New Government Takes New Step to Protect the
    Environment With Biofuels, http://www.agr.gc.ca/cb/index_e.php?s1=n&s2=2006&page=n61220
•   Intergovernmental Panel on Climate Change, 2006, IPCC Guidelines for National Greenhouse Gas
    Inventories. Volume 4. Agriculture, Forestry and Other Land Use.              http://www.ipcc-
    nggip.iges.or.jp/public/2006gl/vol4.htm
•   International Standards Organization (ISO), 2006, ISO 14044:2006 - Environmental Management - Life cycle
    assessment - Requirements and Guidelines, www.iso.org/iso/iso_catalogue (ISO 14044:2006).
•   International Standards Organization (ISO), 2006. ISO 14041: Environmental management: Life cycle
    assessment—Goal and scope definition and inventory analysis. Geneva: ISO.
•   International Standards Organization (ISO). 2006. 14040:2006 - Environmental Management - Life cycle
    Assessment - Principles and Framework.
•   Joseph Fargione, Jason Hill, David Tilman , Stephen Polasky, Peter Hawthorne. Land Clearing and the
    Biofuel Carbon Debt. Science. 29 February 2008:Vol. 319. no. 5867, pp. 1235 - 1238
•   Kim, S. D., Dale, B. E., 2006. Ethanol Fuels: E10 or E85 – Life Cycle Perspectives, The International Journal
    of Life Cycle Assessment, 2006, Vol 11, No 2, March 2006. (5 pp)
•   Kim, S. D., Dale, B. E., Cumulative Energy and Global Warming Impact from the Production of Biomass for
    Biobased Products, Journal of Industrial Ecology, 2003 (Vol. 7) (No. 3/4) 147-162




                                                                                                             158
CHEMINFO




•   Kim, S., Dale, B., 2002. Allocation Procedure in Ethanol Production System from Corn Grain. International
    Journal of Life Cycle Assessment. Volume 7. Pages 237-243.
•   Larson, E. D. A review of life cycle analysis studies on liquid biofuel systems for the transport sector. Energy
    for Sustainable Development. Vol. X No. 2. June 2006. p. 109-126.
•   LBST. 2002. GM Well-to-Wheel Analysis of Energy Use and Greenhouse Gas Emissions of Advanced
    Fuel/Vehicle Systems – A European Study. September 2002.
•   Leng, R., et al, 2003. Life cycle inventory and energy analysis of cassava-based Fuel ethanol in China.
    Environmental Science & Technology, Journal of Cleaner Production.
•   MacLean, H.L. and L.B. Lave. Evaluating Automobile Fuel/Propulsion Technologies. Progress in Energy
    and Combustion Science. 29. 2003. 1-69.
•   Matthews, H.S., Lave, L.B. and H.L. MacLean. Life Cycle Impact Assessment: A Challenge for Risk Analysts.
    Risk Analysis. 22(5). 2002. 853-860.
•   Moreira, J., R. 2004. Global Biomass Energy Potential. Paper prepared for the Expert Workshop on
    Greenhouse Gas Emissions and Abrupt Climate Change: Positive Options and Robust Policy, Paris 30
    September – 1 October 2004. http://www.accstrategy.org/draftpapers/Moreira.doc
•   Mortimer, N.D., Cormack, P., Elsayed, M.A., Horne, R.E. 2003. Evaluation of the Comparative Energy,
    Global Warming and Socio-Economic Costs and Benefits of Biodiesel.
•   Natural Resources Canada, Analysis and Modelling Division, Canada's Energy Outlook: The Reference Case
    2006, October 2006.
•   Patyk, G. A. Reinhardt IFEU – Institut für Energie- und Umweltforschung Heidelberg GmbH (Institute for
    Energy and Environmental Research Heidelberg). Bioenergy for Europe: Which Ones Fit Best?
    http://www.oeko.de/service/bio/dateien/en/BLT%20Biofuels%20II.pdf
•   Piringer,G.,. Steinberg, L.J.,. 2006. Reevaluation of Energy Use in Wheat Production in the United States.
    Journal of Industrial Ecology, 10:1-2, 149–167
•   Reinhardt, G., Jungk, N. 2001. Pros and Cons of RME Compared to Conventional Diesel Fuel. IFEU.
    Proceedings of the 3rd International Colloquium "Fuels 2001", 17 - 18 January 2001.
•   Rochette, P., Angers, D.A., Belanger, G., Chantigny, M.H., Prevost, D., Levesque, G. 2004. Emissions of
    N2O from Alfalfa and Soybean Crops in Eastern Canada. Soil Science Society of America Journal. 68:493-
    506.
•   SAIC, Cheminfo Services Inc., 2007, Operationalizing MAPLE-C With NOx and SO2 Emissions Capabilities
    for Canada’s Electric Power Generation Sector, For Natural Resources Canada.
•   Seungdo Kim, Bruce E. Dale. Cumulative Energy and Global Warming Impact from the Production of
    Biomass for Biobased Products. Journal of Industrial Ecology. Volume 7, Number 3–4.
•   Sheehan, J.; Camobreco, V.; Duffield, J.; Graboski, M.; Shapouri. H. Life Cycle Inventory of Biodiesel and
    Petroleum Diesel for Use in an Urban Bus: Final Report.
    http://www.nrel.gov/vehiclesandfuels/npbf/pdfs/24089.pdf
•   Sinum Eco-Pro: http://www.sinum.com/htdocs/e_software_ecopro.shtml)
•   State of California Office of the Governor. Executive Order S-1-07, The Low Carbon Fuel Standard. January
    18, 2007.
•   Statistics       Canada.          2007.          Snapshot        of             Canadian           Agriculture.
    http://www.statcan.ca/english/agcensus2006/articles/snapshot.htm



                                                                                                               159
CHEMINFO




•   Swanston, J. S., Newton, A.C., Mixtures of UK Wheat as an Efficient and Environmentally Friendly Source
    for Bioethanol, Journal of Industrial Ecology, 2005, Vol. 9, No. 3, Pages 109-126 2005
•   Timothy Searchinger, et al, Use of U.S. Croplands for Biofuels Increases Greenhouse Gases Through
    Emissions from Land Use Change. Science. 29 February 2008: Vol. 319. no. 5867, pp. 1238 – 1240.
•   United Kingdom Department of Transport 2006. Renewable Transport Fuel Obligation Programme (RTFO).
    www.dft.gov.uk/roads/RTFO
•   United Kingdom Department of Transport 2006. Renewable Transport Fuel Obligation Programme (RTFO).
    www.dft.gov.uk/roads/RTFO
•   University of Nebraska Lincoln (2007), BESS Biofuel Energy Systems Simulator Users Guide.
•   US EPA, Draft Technical Report, A Comprehensive Analysis of Biodiesel Impacts on Exhaust Emissions,
    October 2002. EPA420-P-02-001. http://www.epa.gov/OMS/models/analysis/biodsl/p02001.pdf
•   US EPA, MOBILE6.2. http://www.epa.gov/otaq/m6.htm
•   Wang, M., et al, 2004, Allocation of Energy Use in Petroleum Refineries to Petroleum Products. International
    Journal       of        Life        Cycle        Assessment        9        (1)       34        –        44.
    http://www.transportation.anl.gov/software/GREET/pdfs/IJLCA-2004.pdf
•   Wang, M., et al, February 2007, Operating                      Manual     for    GREET:      Version    1.7.
    http://www.transportation.anl.gov/pdfs/TA/353.pdf
•   Wang, M., et al, June 2001, Development and Use of GREET 1.6 Fuel-Cycle Model for Transportation Fuels
    and Vehicle Technologies. http://www.transportation.anl.gov/pdfs/TA/153.pdf
•   Wang., M., et al. August 1999. GREET 1.5 -- Transportation Fuel-Cycle Model.
    http://www.transportation.anl.gov/software/GREET/pdfs/esd_39v1.pdf
•   Weidema, B.P. 1999. System Expansions to Handle Co-products of Renewable Materials. 7th LCA Case
    Studies Symposium SETAC-Europe.
•   Weidema, B.P. 2001. Addendum to the article “Avoiding co-product allocation in life-cycle assessment”
    (Journal of Industrial Ecology 4(3):11-33, 2001).
•   Weise, A.M. January 1998. Impacts of Market-Based Greenhouse Gas Emission Reduction Policies on US
    Manufacturing Competitiveness. API Research Study #090.




                                                                                                           160
CHEMINFO




10.1 Websites Used to Access LCA Models, and Others

All websites were access between October 2007 and March 2008.

 •   Boustead Consulting http://www.boustead-consulting.co.uk/products.htm
 •   CIT Ekologik LCAiT (http://www.lcait.com: website could not be accessed).
 •   http://buildlca.rmit.edu.au/downloads/BACKGROUNDREPORTFINAL.PDF
 •   http://deville.tep.chalmers.se/commdb/DbStart.htm
 •   http://ew.eea.europa.eu/Industry/Cleaner/Theme_2/b/Tools_for_analysis_and_evaluation/URL999072874/
 •   http://faculty.washington.edu/cooperjs/Research/database%20projects.htm
 •   http://faostat.fao.org/site/567/default.aspx
 •   http://glwww.mst.dk/indu/03040000.htm
 •   http://nrel.colostate.edu/projects/century5/reference/html/Century/overview.htm
 •   http://www.aspentech.com/products/aspen-plus.cfm
 •   http://www.athenasmi.ca/tools/docs/Athena_Institute-Software.pdf
 •   http://www.bess.unl.edu/download/
 •   http://www.bess.unl.edu/download/
 •   http://www.bfrl.nist.gov/oae/software/bees/
 •   http://www.bfrl.nist.gov/oae/software/bees/please/USDA/bees_please.html.
 •   http://www.bfrl.nist.gov/oae/software/bees/registration.html
 •   http://www.bfrl.nist.gov/oae/software/bees/registration.html
 •   http://www.boustead-consulting.co.uk/download/Modelcontents.pdf
 •   http://www.earthshift.com/simapro7.htm
 •   http://www.ecobalance.com/uk_team.php
 •   http://www.ecobalance.com/uk_team02.php
 •   http://www.ecobalance.com/uk_wisard.php
 •   http://www.ecoinvent.org/
 •   http://www.ecoinvent.org/fileadmin/documents/en/ecoinventData-v2.0_Contents.pdf
 •   http://www.eco-shop.org/Resources/lcasoftwarereREVIEW.pdf
 •   http://www.eiolca.net/use.html
 •   http://www.epa.gov/NRMRL/lcaccess/pdfs/appendices_lca101.pdf
 •   http://www.epa.gov/NRMRL/lcaccess/pdfs/appendices_lca101.pdf
 •   http://www.gabi-software.com/



                                                                                                      161
CHEMINFO




•   http://www.gabi-software.com/gabi/databases1/extensiondatabases0
•   http://www.gabi-software.com/gabi/gabi-4/aboutgabi41
•   http://www.ghgenius.ca/
•   http://www.icmm.com/uploads/35Eco-Indices.pdf
•   http://www.its.ucdavis.edu/publications/2002/UCD-ITS-RR-02-02.pdf)
•   http://www.its.ucdavis.edu/publications/2002/UCD-ITS-RR-02-02.pdf
•   http://www.ivl.se/rapporter/pdf/B1390.pdf
•   http://www.kmlmtd.com/demodld/index.html
•   http://www.kmlmtd.com/demodld/lcapix20_manual.pdf
•   http://www.lca-center.dk/cms/site.asp?p=4633
•   http://www.leidenuniv.nl/cml/ssp/software/cmlca/faq.html
•   http://www.leidenuniv.nl/cml/ssp/software/cmlca/faq.html
•   http://www.leidenuniv.nl/cml/ssp/software/miet/index.html
•   http://www.nrel.colostate.edu/projects/century/
•   http://www.oeko.de/service/gemis/en/index.htm
•   http://www.pre.nl/eco-indicator99/default.htm
•   http://www.pre.nl/eco-it/download_eco-it.htm
•   http://www.pre.nl/eco-it/eco-it.htm
•   http://www.pre.nl/simapro/default.htm
•   http://www.sematech.org/docubase/document/4238atr.pdf
•   http://www.sinum.com/htdocs/e_software.shtml
•   http://www.tme.nu/english/index_uk.htm
•   http://www.tme.nu/pdf/PIA-demo-instruction.pdf
•   http://www.tno.nl/index.cfm
•   http://www.transportation.anl.gov/software/GREET/greet_1-8a_download_form.html
•   http://www.transportation.anl.gov/software/GREET/greet_2-8a_download_form.html
•   http://www.transportation.anl.gov/software/GREET/index.html
•   http://www.umberto.de/en/product/prices/index.htm
•   www.pre.nl/download/manuals/SimaPro7IntroductionToLCA.pdf




                                                                                     162
CHEMINFO




                                   11. Appendix

11.1 Selected Bioethanol Papers

11.1.1 Journal of Industrial Ecology


11.1.1.1 Mixtures of UK Wheat as an Efficient and Environmentally Friendly Source for
   Bioethanol

J. Stuart Swanston, Senior research scientist and cereal quality specialist in the Genome
Dynamics program at the Scottish Crop Research Institute, Dundee, UK.

Adrian C. Newton, Cereal pathologist and currently Head of the Host Parasite Co-evolution
program at the Scottish Crop Research Institute, Dundee, UK.

Concerns about access to oil supplies have encouraged the exploration of renewable fuel and
energy sources. Industrial ecology offers tools to compare the energy implications and benefits
of differing strategies, but using botanical sources of raw materials to replace nonrenewable ones
also requires appreciation of plant science, especially the variation in genetic potential within
species. Whereas cultivation methods determine whether genetic potential is realized, different
methods impact the environment to varying degrees. Experience with barley variety mixtures,
aimed at reducing chemical input, has shown them to improve yield and reduce disease, while
maintaining or even enhancing quality. Yield improvements still occurred in the absence of
disease and increased in proportion to the number of component varieties. Because other
research showed mixtures to be similarly effective in wheat, a protocol to grow and exploit a
complex mixture of soft wheat is proposed, offering a cost-effective and energy efficient
feedstock for a possible bioethanol industry in the United Kingdom. Ethanol would be produced
initially from grain, with the straw used for heating or electricity generation. Fertilizer
production and use and vehicle fuels have been shown as the main forms of energy consumption
in growing a crop, and targets for enhancing the energy balance, by growing mixtures under an
integrated farming system, are postulated. A close but negative association between grain protein
and alcohol yield is demonstrated and a mixture giving comparable grain yield, but superior
alcohol yield, to its best component is identified. Mixing varieties differing in plant morphology
may also increase total biomass yield and, therefore, the energy generated from the crop.
Pesticide reduction has another positive, though small, effect on the energy balance, from using
mixtures. Eliminating prophylactic spraying also reduces vehicle fuel consumption, and may
provide the low-toxicity benefits of organic agriculture without the yield penalty. A range of
alternative uses for straw and other by-products is also discussed.


                                                                                              163
CHEMINFO




11.1.1.2 Cumulative Energy and Global Warming Impact from the Production of Biomass for
   Biobased Products

Seungdo Kim, Research associate

Bruce E. Dale, Professor of chemical engineering and materials science at Michigan State
University in East, Lansing, Michigan, USA.

The cumulative energy and global warming impacts associated with producing corn, soybeans,
alfalfa, and switchgrass and transporting these crops to a central crop processing facility (called a
“biorefinery”) are estimated. The agricultural inputs for each crop are collected from seven states
in the United States: Illinois, Indiana, Iowa, Michigan, Minnesota, Ohio, and Wisconsin. The
cumulative energy requirement for producing and transporting these crops is 1.99 to 2.66
megajoules/kilogram (MJ/kg) for corn, 1.98 to 2.04 MJ/kg for soybeans, 1.24 MJ/kg for alfalfa,
and 0.97 to 1.34 MJ/kg for switchgrass. The global warming impact associated with producing
biomass is 246 to 286 grams (g) CO2 equivalent/kg for corn, 159 to 163gCO2 equivalent/kg for
soybeans, 89 g CO2 equivalent/ kg for alfalfa, and 124 to 147 g CO2 equivalent/kg for switch-
grass. The detailed agricultural data are used to assess previous controversies over the energy
balance of bioethanol and, in light of the ongoing debates on this topic, provide a needed
foundation for future life-cycle assessments.


11.1.1.3 Comparing the Land Requirements, Energy Savings, and Greenhouse Gas
   Emissions Reduction of Biobased Polymers and Bioenergy: An Analysis and System
   Extension of Life-Cycle Assessment Studies

Veronika Dornburg, Ph.D. student

Iris Lewandowski, Senior researcher

Martin Patel, Assistant professor in the Department of Science, Technology, and Society of the
Copernicus Institute for Sustainable Development and Innovation at Utrecht University in the
Netherlands.

This study compares energy savings and greenhouse gas (GHG) emission reductions of biobased
polymers with those of bioenergy on a per unit of agricultural land-use basis by extending
existing life-cycle assessment (LCA) studies. In view of policy goals to increase the energy
supply from biomass and current efforts to produce biobased polymers in bulk, the amount of
available land for the production of nonfood crops could become a limitation. Hence, given the
prominence of energy and greenhouse issues in current environmental policy, it is desirable to
include land demand in the comparison of different biomass options. Over the past few years,


                                                                                                 164
CHEMINFO




numerous LCA studies have been prepared for different types of bio-based polymers, but only a
few of these studies address the aspect of land use. This comparison shows that referring energy
savings and GHG emission reduction of biobased polymers to a unit of agricultural land, instead
of to a unit of polymer produced, leads to a different ranking of options. If land use is chosen as
the basis of comparison, natural fiber composites and thermoplastic starch score better than
bioenergy production from energy crops, whereas polylactides score comparably well and
polyhydroxyalkaonates score worse. Additionally, including the use of agricultural residues for
energy purposes improves the environmental performance of bio-based polymers significantly.
Moreover, it is very likely that higher production efficiencies will be achieved for biobased
polymers in the medium term. Biobased polymers thus offer interesting opportunities to reduce
the utilization of nonrenewable energy and to contribute to GHG mitigation in view of
potentially scarce land resources.



11.1.1.4 Reevaluation of Energy Use in Wheat Production in the United States

Gerhard Piringer, A Visiting Professor Department of Civil and Environmental Engineering at
Tulane University in New Orleans, Louisiana.

Laura J. Steinberg, An Associate Professor At the Department of Civil and Environmental
Engineering at Tulane University in New Orleans, Louisiana.
Department of Civil and Environmental Engineering Tulane University, New Orleans, LA
70118.

Energy budgets for agricultural production can be used as building blocks for life-cycle
assessments that include agricultural products, and can also serve as a first step toward
identifying crop production processes that benefit most from increased efficiency. A general
trend toward increased energy efficiency in U.S. agriculture has been reported. For wheat
cultivation, in particular, this study updates cradle-to-gate process analyses produced in the
seventies and eighties. Input quantities were obtained from official U.S. statistics and other
sources and multiplied by calculated or recently published energy coefficients. The total energy
input into the production of a kilogram of average U.S. wheat grain is estimated to range from
3.1 to 4.9 MJ/kg, with a best estimate at 3.9 MJ/kg. The dominant contribution is energy
embodied in nitrogen fertilizer at 47% of the total energy input, followed by diesel fuel (25%),
and smaller contributions such as energy embodied in seed grain, gasoline, electricity, and
phosphorus fertilizer. This distribution is reflected in the energy carrier mix, with natural gas
dominating (57%), followed by diesel fuel (30%). High variability in energy coefficients masks
potential gains in total energy efficiency as compared to earlier, similar U.S. studies. Estimates
from an input-output model for several input processes agree well with process analysis results,
but the model's application can be limited by aggregation issues: Total energy inputs for generic
food grain production were lower than wheat fertilizer inputs alone, possibly due to aggregation
of diverse products into the food grain sector.


                                                                                               165
CHEMINFO




11.1.2 The International Journal of Life Cycle Assessment


11.1.2.1 Ethanol Fuels: E10 or E85 – Life Cycle Perspectives
Seungdo Kim and Bruce Dale
Seungdo Kim, PhD Department of Chemical Engineering Room 2527 Engineering Building
Michigan State University East Lansing, MI 48824-1226 USA
Bruce E. Dale Department of Chemical Engineering & Materials Science Room 2527,
Engineering Building Michigan State University East Lansing, MI 48824–1226 USA

Goal and Scope

The environmental performance of two ethanol fuel applications (E10 and E85) is compared
(E10 fuel: a mixture of 10% ethanol and 90% gasoline by volume, and E85 fuel: a mixture of
85% ethanol and 15% gasoline by volume).

Methods

Two types of functional units are considered here: An ethanol production-oriented perspective
and a traveling distance-oriented perspective. The ethanol production-oriented functional unit
perspective reflects the fact that the ethanol fuel supply (arable land or quantity of biomass used
in ethanol fuel) is constrained, while the traveling distance-oriented functional unit implies that
the ethanol fuel supply is unlimited.

Results and Discussion

In the ethanol production-oriented functional unit perspective, the E10 fuel application offers
better environmental performance than the E85 fuel application in terms of natural resources
used, nonrenewable energy and global warming. However, in the calculations based on the
traveling distance perspective, the E85 fuel application provides less environmental impacts in
crude oil consumption, nonrenewable energy and global warming than the E10 fuel application.

Conclusions and Outlook

The choice of functional units significantly affects the final results. Thus the functional unit in a
descriptive LCA should reflect as nearly as possible the actual situation associated with a
product system. Considering the current situation of constrained ethanol fuel supply, the E10
fuel application offers better environmental performance in natural resources used,
nonrenewable energy and global warming unless the fuel economy of an E85 fueled vehicle is
close to that of an E10 fueled vehicle.


                                                                                                 166
CHEMINFO




11.1.3 Environmental Science & Technology, Journal of Cleaner Production

11.1.3.1 Life cycle inventory and energy analysis of cassava-based Fuel ethanol in China
Rubo Leng , Chengtao Wang, Cheng Zhang, Du Dai and Gengqiang Pu

RM330, Mechanical Engineering BLD, Institute of Life Quality via Mechanical Engineering,
School of Mechanical and Power Engineering, Shanghai Jiao Tong University, No. 800,
DongChuan Road, Shanghai City 200240, PR China.

The Chinese government is developing biomass ethanol as one of its automobile fuels for energy
security and environmental improvement reasons. The cassava is an alternative feedstock to
produce this ethanol fuel. Its performance of environmental impacts and energy efficiency is the
critical issue. Life cycle assessment has been used to identify and quantify the environment
emissions, energy consumption and energy efficiency of the system throughout the life cycle.
This study investigates the entire life cycle from cassava plantation, ethanol conversion,
transport, Fuel ethanol blending and distribution to its end-use. Product system of cassava-based
ethanol fuel is described and it is divided into six unit processes. The environmental impacts and
energy consumption of each unit process are quantified and some of the potential effects are
assessed.

11.1.3.2 A review of assessments conducted on bio-ethanol as a transportation fuel from a net
   energy, greenhouse gas, and environmental life cycle perspective
Harro von Blottnitz and Mary Ann Curran

Chemical Engineering Department, University of Cape Town, Private Bag, 7701 Rondebosch,
South Africa
Environmental Protection Agency, National Risk Management Research Laboratory, Cincinnati,
OH 45268, USA
Received 20 September 2005; accepted 3 March 2006. Available online 12 May 2006.


Interest in producing ethanol from biomass in an attempt to make transportation ecologically
sustainable continues to grow. In recent years, a large number of assessments have been
conducted to assess the environmental merit of biofuels. Two detailed reviews present
contrasting results: one is generally unfavourable, whilst the other is more favourable towards
fuel bio-ethanol. However, most work that has been done so far, to assess the conversion of
specific feedstocks to biofuels, specifically bio-ethanol, has not gone beyond energy and carbon
assessments. This study draws on 47 published assessments that compare bio-ethanol systems to
conventional fuel on a life cycle basis, or using life cycle assessment (LCA). A majority of these
assessments focused on net energy and greenhouse gases, and despite differing assumptions and
system boundaries, the following general lessons emerge: (i) make ethanol from sugar crops, in


                                                                                              167
CHEMINFO




tropical countries, but approach expansion of agricultural land usage with extreme caution; (ii)
consider hydrolysing and fermenting lignocellulosic residues to ethanol; and (iii) the LCA
results on grasses as feedstock are insufficient to draw conclusions. It appears that technology
choices in process residue handling and in fuel combustion are key, whilst site-specific
environmental management tools should best handle biodiversity issues. Seven of the reviewed
studies evaluated a wider range of environmental impacts, including resource depletion, global
warming, ozone depletion, acidification, eutrophication, human and ecological health, smog
formation, etc., but came up with divergent conclusions, possibly due to different approaches in
scoping. These LCAs typically report that bio-ethanol results in reductions in resource use and
global warming; however, impacts on acidification, human toxicity and ecological toxicity,
occurring mainly during the growing and processing of biomass, were more often unfavourable
than favourable. It is in this area that further work is needed.

11.1.3.3 Life-cycle analysis of emissions from fuel ethanol and blends in Australian heavy
   and light vehicles
Tom Beer and Tim Grant

CSIRO Environmental Risk Network, CSIRO Atmospheric Research, Private Bag 1, Aspendale,
Victoria 3195, Australia
Centre for Design, RMIT, Melbourne, Victoria 3001, Australia
Accepted 21 June 2006. Available online 20 September 2006.

Because carbon dioxide emissions from the combustion of a renewable fuel are not
anthropogenic greenhouse gases, there are significant greenhouse gas benefits in using ethanol
that is derived from sugar or wheat, especially from waste feedstock. However, if the ethanol is
used as an additive (as in diesohol or petrohol) then some of these greenhouse gas benefits are
lost because ethanol is less efficient as a fuel.

The vapour pressure of petrohol is higher than that of either petrol or ethanol, so that it is unclear
whether there are, or are not, air quality benefits associated with the use of ethanol.

A measurement program that surveys a significant proportion of E10 alternative fuel vehicles
should be undertaken, along with a parallel program to test the emission variations that result
from the changes in the petrol. The performance of overseas models in relation to the Australian
situation is unknown, and a combined modelling and measurement program is needed to
determine its validity.




                                                                                                  168

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:264
posted:3/9/2011
language:English
pages:173