Life-Cycle Analysis of Biofuels: Issues and Results Michael Wang Center for Transportation Research Argonne National Laboratory Presentation to the Special Committee on Domestic Biofuels State of Wisconsin Joint Legislative Council Madison, WI, October 14, 2008
Life-Cycle Analysis for Vehicle/Fuel Systems Has Been Evolved in the Past 30 Years • Historically, evaluation of vehicle/fuel systems from wells to wheels (WTW) was called fuel-cycle analysis • Pioneer transportation WTW analyses began in 1980s • Early studies were motivated primarily by battery-powered EVs • Recent studies were motivated primarily by introduction of new fuels such as hydrogen and biofuels • Pursuing reductions in transportation GHG emissions now demands for intensive and extensive WTW analyses • Early WTW studies were for evaluation of individual technologies or processes; the current focus has been expanded to general policy evaluation • Many studies conclude with the quantity of energy and emissions; some studies carry all the way to impact assessment
The GREET (Greenhouse gases, Regulated Emissions, and Energy use in Transportation) Model • The GREET model and its documents are available at Argonne’s website at http://www.transportation.anl.gov/software/GREET/ • The most recent GREET version (GREET 1.8b) was released in May 2008 • As of July 2008, there are 9,000 registered GREET users worldwide Vehicle Cycle GREET 2.7 Well to Wheels Fuel Cycle GREET 1.8
GREET Includes More Than 100 Fuel Production Pathways from Various Energy Feedstocks
Oils for Biodiesel/Renewable Diesel Soybeans Rapeseed Palm oil Jatropha Waste cooking oil Animal fat Starch Crops for EtOH Corn Wheat Cassava Sweet potato GREET Includes Some of the Potential Biofuel Production Pathways • Sugar Crops for EtOH • Sugar cane • Sugar beet • Sweet sorghum • Algaes • Oils • Hydrogen • Butanol Production • Corn • Sugar beet • Cellulosic Biomass for EtOH • Corn stover, rice straw, wheat straw • Forest wood residue • Municipal solid waste • Energy crops • Black liquor • Cellulosic Biomass via Gasification • Fitscher-Tropsch diesel • Hydrogen • Methanol The feedstocks that are underlined are already included in the GREET model.
The 2007 Energy Independence and Security Act Established Aggressive Biofuel Production Targets
The 2007 EISA Requires US EPA To Conduct Life-Cycle Analysis for Fuels • LCA is conducted to determine if given fuel types meet mandated minimum GHG reductions compared to 2005 baseline petroleum fuels • Ethanol produced from corn: 20% (only applies to fuel produced in new facilities) • Cellulosic biofuels: 60% • Biomass-based diesel (e.g., biodiesel): 50% • Other advanced biofuels (e.g., imported sugarcane ethanol, renewable diesel, CNG/LNG made from biogas): 50% • Life cycle analysis includes • All major GHGs (CO2, CH4, and N2O) • Both production and use of biofuels • Direct and indirect land use change impacts
GREET Ethanol Life-Cycle Analysis Includes Activities from Fertilizer to Ethanol at Stations
Key Issues Affecting Biofuel WTW Results • Continued technology advancements • Agricultural farming: continued crop yield increase and resultant reduction of energy and chemical inputs per unit of yield • Energy use in ethanol plants: reduction in process fuel use and switch of process fuel types • Methods of estimating emission credits of co-products of ethanol • Distillers grains and solubles (DGS) for corn ethanol: 0-50% • Electricity for cellulosic and sugarcane ethanol • Animal feed and specialty chemicals for biodiesel • Direct and indirect land use changes and resulted GHG emissions • Life-cycle analysis methodologies • Attributional LCA • Consequential LCA
Accurate Ethanol Energy Analysis Must Account for Increased Productivity in Farming Over Time U.S. Corn Output Per Pound of Fertilizer Has Risen by 55% in The Past 35 Years Based on harvested acreage. Source: USDA
Energy Use for Corn Farming Varies Considerably Among Corn-Producing States 2001 U.S. Corn Farming Energy Use: Btu/Bushel • Corn farming energy use varies by three times among nine corn-producing states. • From 1996 to 2001, U.S. corn farming energy use in Btu/bushel was reduced by 34%.
In General, Infrastructure-Related Activities Are Not A Major Contributor to WTW Results – GRRET Simulation of Farming Equipment for Ethanol WTW Analysis • Size of farm • Life time of equipment • Energy for producing equipment materials (the majority of equipment materials is steel and rubber) • Argonne has found that farming equipment may contribute to <2% of energy and ~1% GHG emissions for corn ethanol
Improved Technology and Plant Design Has Reduced Energy Use and Operating Costs in Corn Ethanol Plants ~1/3 of Energy is Spent on DDGS Drying 30% reduction 50% reduction There are indications that the ethanol industry continues to reduce plant energy use.
Co-Products with Biofuels • Types of co-products • Corn ethanol: animal feeds (distillers grains and solubles, DGS) • Sugarcane ethanol: electricity • Cellulosic ethanol: electricity • Biodiesel and renewable diesel from soybean and rapeseed: animal feeds, glycerin, and other chemicals • Ways of dealing with co-products • Displacement method (or the system boundary expansion approach) • Allocation methods • Mass based • Energy content based • Economic revenue based • Production plant process purpose based • Scale of biofuel production (and resultant scale of co-product production) can affect the choice of methods
Proper Accounting for Animal Feed Is Key to Corn Ethanol’s Lifecycle Analysis Source: RFA, 2008 Argonne uses the displacement method.
Key Issues Affecting Cellulosic Ethanol Results • Cellulosic biomass feedstock types • Fast growing trees • Soil carbon could increase • Fertilizer may be applied • Irrigation to be needed? • Switchgrass and other native grass • Soil carbon could increase • Fertilizer will be applied • Irrigation to be needed? • Crop residues • soil carbon could decrease • Additional fertilizer will be needed to supplement nutrient removal • Forest wood residues: collection effort could be extensive • Co-production of ethanol and electricity • The amount of electricity produced • The types of conventional electric generation to be displaced • Land use changes could have less effects on cellulosic ethanol’s GHG results
GHG Emissions of Corn Ethanol Vary Considerably Among Process Fuels in Plants GHG Emission Reductions By Ethanol Relative to Gasoline GHG effects of potential land use changes are not fully included in these results.
Approach to Address GHG Emissions of Potential Land Use Change by Large-Scale Biofuel Production • Potential land use changes • Direct land use change: regional or national scale • Indirect land use change: global scale • Both can be simulated with global general equilibrium models • The resolution level of global GE models could be a key factor • Carbon profiles of major land types • Models in the U.S. and Europe are available • Carbon profiles of land types in other parts of the world (South America, Asia, Africa) may be less understood • Time horizon of biofuel programs; “for-ever biofuels” can mathematically result in zero GHG emission changes from land use changes • At present, GREET includes the following soil CO2 sources/sinks for ethanol • Corn ethanol: CO2 source of 73 grams/gal. EtOH from soil C reduction • Cellulosic ethanol • Fast growing trees: CO2 sink of 1,250 g/gal. EtOH from soil C increase • Switchgrass: CO2 sink of 540 g/gal. EtOH from soil C increase
Modeling of Land Use Changes by Biofuel Production • Baseline definition • Global trend of demand for food and thus agricultural commodities • Global trend of supply of agricultural commodities • Current and future global land use patterns (including agricultural sector and other sectors) • Various worldwide biofuel programs: are they parts of a biofuel system or competing programs? • Growth of crop yields • Trend yield growth • Yield growth response to price increase • How to value animal feeds in modeling? – nutrition value vs. market price approach • Land use changes vs. land use intensification
Change in Soil Carbon Content Differ Among Land Use Changes and Over Time(The Three Profiles Here Are for Illustrative Purpose Only)
GHG Benefits and Burdens for Fuel Ethanol Cycle Occur at Different Stages (and With Different Players) CO2 in the atmosphere CO2 via Photosynthesis Energy inputs for farming Fossil energy inputs to ethanol plant CO2 emissions during fermentation CO2 emissions from ethanol combustion Carbon in kernels Carbon in ethanol Fertilizer Change in soil carbon DGS Negative effect via price signal N2O emissions from soil and water streams Positive effect via reduction in conventional animal feed demand In direct land use changes for other crops and in other regions Conventional animal feed production cycle
Facility-Level Certification for LCFS? Leaders Corn Ethanol GHG Reductions Industry baseline (default?) Laggards (leakage to non-LCFS states?) Individual Facility
Four FT Diesel Production Options Were Evaluated on the Well-to-Wheels Basis by Argonne • Natural gas to liquids (GTL) • Coal to liquids (CTL) • Biomass to liquids (BTL) • Co-firing of coal and biomass to liquids (C/BTL) • 85/15 C/B co-feeding • 38/62 C/B co-feeding: GHG breakeven with petroleum diesel • All options were evaluated with and without carbon capture and storage (CCS) in FTD plants
Key Issues and Assumptions for FT Diesel Plants • FT diesel plant designs • Standalone to produce diesel, naphtha, and other products • Co-generation of steam and/or electricity for export • This study evaluated standalone plants • GTL plant assumptions in this study • Energy conversion efficiency of 63% • Carbon conversion efficiency of 80% • CTL plant assumptions in this study • Based on studies by National Energy Technology Laboratory (2003 and 2007) • Energy conversion efficiency of 50% • A carbon capture and storage (CCS) case with a carbon capture rate of 90% at FT plants • BTL plant assumptions in this study • Based on studies completed in Europe • Energy conversion efficiency of 50% • A CCS case with a carbon capture rate of 90% • C/B TL plant assumptions • This is a case for diluting carbon from coal • Assumptions are based on CTL and BTL plants • Engineering details need to be examined • Logistic and costs of two separate feedstocks are a key issue
Trade-Offs Between Petroleum Reductions and GHG Reductions by FT Diesel from Different Feedstocks
Increased by 1000 times during 8 years Source: National Biodiesel Board Biodiesel Is AnRenewable Alternative to Petroleum Diesel • Produced from various biological sources (soybeans, rapeseeds, animal fats, sunflower seeds, palm oil…) via the transesterification process • In US, a majority of biodiesel is produced from soybeans • High cetane value of 50-65 (vs. 40 for petroleum diesel) • Can be blended with conventional diesel fuel in any proportion • Production & sales volume for biodiesel in U.S. has increased dramatically • Hydrogenation process can also be used to produce renewable diesel
Four Allocation Approaches Were Employed to Address Various Co-Products 1. Displacement 2. Energy value-based allocation 3. Market value-based allocation 4. Hybrid approach: combination of displacement and allocation
Co-Product Methodologies Significantly Influence Results for Soybean-Based Biodiesel or Renewable Diesel 1, Displacement; 2, Energy-value–based allocation; 3, Market-value–based allocation; 4, Hybrid
Outstanding LCA Issues • Purpose of LCAs and their models has been evolving over the past 30 years • Scope of LCAs: • Representation of LCA scoping is a key factor; misrepresentation can cause confusion to the least • Average vs. marginal analysis • Industry vs. facilities: significant implications on LCFS and carbon trading • National vs. regional analysis: national or regional LCFS; need to avoid double-counting and/or leakage • Transparency of methodologies and input data • Technology advancement over time need to be considered • System boundary of LCAs: has been a moving target • Consistency vs. intuition (an issue of resource availability) • Research vs. policy development • Aggregate effects vs. attribution (and responsibility): consequential vs. attributional LCAs
If We Are Going to Consider Different Farming Practices for GHG Regulations for Fuels, We Will Need to Address: • Chemical and energy inputs to individual farms, a given county, or a given state? • Different tillage practices? • Practicality, traceability, and verifiability of input data for the feedstock (e.g., corn0 into a specific biofuel facility • A balance between operationability of a regulation and incentivizing of advanced farming practices
Some Thoughts • While the current discussion/debate/research efforts are on land use changes by large-scale biofuel production and the uncertainties caused by them, can we separate • Technology effects and certainties – should the society continue to promote technology advancement? • Social, behavioral, and/or economic effects and uncertainties – are some of them mitigatable? • We need to design a biofuel policy with flexibility to address uncertainties, to help advance technologies, and to avoid potential risks