A Reflection on the Evolution of Estimating Future U.S. Ligno-cellulosic Biomass Feedstock

Managed by UT-Battelle for the Department of Energy

1 Managed by UT-Battelle for the U.S. Department of Energy

Robin L Graham Oak Ridge National Laboratory

University of Wisconsin April 14, 2011

Potential ? Economic? Sustainable? A reflection on the evolution of

estimating future U.S. ligno-cellulosic biomass feedstock supplies

Mark Downing (ORNL) Bob Perlack (ORNL)

Anthony Turhollow (ORNL) Laurence Eaton (ORNL) Matt Langholtz (ORNL) Richard Nelson (U of KS)

Marie Walsh (U TN) Burt English (U TN)

Daniel de la Torre Ugarte (U TN) Chad Hellwinkle (U TN)

Chuck Noon (U TN)

2 Managed by UT-Battelle for the Department of Energy


Material Science is important - Example materials issue with gasifier: Feed auger degradation

Top of auger used to raise wood chips into the biomass gasifier

Surface of auger flight showing cracking and spalling of the wear resistant coating

Area near top of auger where wear resulted in a complete

penetration of the central tube

ORNL is currently working to identify more suitable auger materials


Pyrolysis Oil Corrosion Studies

•  Four test systems are being used for atmospheric pressure studies of corrosion by pyrolysis oil

•  Several tests are being conducted in which alloys with chromium contents ranging from 0% to 18% are exposed for a total of 500 hours at 50°C

•  Sample configurations include general corrosion coupons, samples for study of localized attack, and u-bend samples for study of stress corrosion cracking

•  All tests are being conducted in pyrolysis oil provided by NREL

Two test systems used for pyrolysis oil

corrosion studies Expect to receive oil samples from several other sources for additional corrosion tests


Nanoporous Carbon Fibers

Vehicle Uses (e.g., capacitors, ELCD)


The BioEnergy Science Center •  Funded by the US Department of Energy

Office of Science at $135M over five years •  World-class cross-disciplinary science and

proven ability to rapidly impact biomass to biofuel conversion

•  Positioned to attack the most important current barrier to the emergence of a cellulosic biofuels industry – biomass recalcitrance

•  Unique anchor facilities at the core partners •  Home base at the UT/ORNL Joint Institute for

Biological Sciences •  http://bioenergycenter.org

• Oak Ridge National Laboratory • University of Georgia • University of Tennessee • National Renewable Energy Laboratory • Georgia Tech • Samuel Roberts Noble Foundation • Dartmouth • ArborGen • Verenium • Mascoma • Individuals from U California-Riverside,

Cornell, Washington State, U Minnesota, NCSU, Brookhaven National Laboratory, Virginia Tech


A Two-pronged Approach to Increase the Accessibility of Biomass Sugars

Both utilize rapid screening for relevant traits followed by detailed analysis of selected samples

Modify the plant cell wall structure to increase accessibility

Improve combined microbial approaches that release sugars and ferment into fuels



9

Vehicle aging and catalyst durability testing

•  Vehicle emissions testing and aging –  Matched vehicle sets –  Aging on mileage accumulation

dynamometers OR –  Aging on track

On basis of results, EPA approved the use of E15 in 2007 or newer cars and lightweight trucks


ORNL partnered with Underwriters Laboratories to evaluate the compatibility of fuel dispensers with gasoline-ethanol blends

•  Two prototype dispensers containing CE25a and CE85a

•  Dispensers operated 15 weeks •  ~100 organic compounds leached out

from the hose material into the fuel

CE25a

Initial Condition 6 weeks 15 weeks

CE85a CE25a CE85a CE25a CE85a

•  CE25a fuel was found to be more degrading to the hose material than the CE85a fuel


https://www.bioenergykdf.net/ BioenergyKDF


Biomass Energy Data Book – getting data to the public

•  An online document covering bioenergy-relevant information, first published in October 2006.

•  Accessible through the Bioenergy KDF or at http://cta.ornl.gov/bedb/


Current Sustainability projects •  Sustainability metrics (Dale) •  Optimizing landuse for Biomass production (Dale)

http://blosm.ornl.gov

•  Regional Water quality & Aquatic biodiversity (Jager)

•  Global feedstock supply, modeling, and analysis (Oladosu)

•  Developing a new Landuse model ( Bhaduri/Kline) •  Great Lakes Bioenergy Center- Modeling bioenergy

feedstock production (Post & Izauralde) •  Savannah River site- field experiment studying

water quality and hydrology in high yield pine energy plantations (P. Mulholland, R. Jackson, J. McDonnell, J. Blake)

•  Establishment in 2009 of the Center for Bioenergy Sustainability – Virginia Dale Director


•  Monthly Forums on topics relating to bioenergy sustainability

•  International collaborations for analysis of cellulosic supplies, landuse change, and sustainable performance measures

•  Website with information, publications, and presentations

•  Sponsor of workshops –  Land-Use Change and Bioenergy

Workshop (6/2009) –  Sustainability of Bioenergy Systems:

Cradle to Grave ( 9/2009) –  A Watershed Perspective on Bioenergy

Sustainability ( 2/2010)

•  Organize Symposium at National Meetings

http://www.ornl.gov/sci/besd/cbes


Wood Storage Gasifiers Fuel Gas Oxidizer Boiler Economizer ESP Stack Ash Handling

ORNL Biomass Steam plant


Two decades of evaluating biomass supplies

•  1990’s focus on perennial energy crops –  Technical potential –  Regional and local economic potential –  Recognizing the issue of scale –  Recognizing environmental outcomes

•  2000-2007 Agricultural and forest residues become important –  Agricultural Sector model (POLYSYS) adapted for

energy crops –  Billion ton study –  Corn stover removal analysis

•  2008-2011 incorporating environmental constraints –  Enhancing spatial resolution –  Billion ton update


Approach • Major Land Resource Areas • 1982 & 1987 NRI database for land base • Herbaceous and woody energy crops • Assign potential yields by soil capability class • Use expert opinion for maximum potential yields • Estimate possible production on all lands

The beginning – what’s the land base suitable for energy crops and how much could it produce?

Graham, R.L. 1994. “An analysis of the potential land base for energy crops in the conterminous United State”. Biomass and Bioenergy 6(3):175-189.


Where is the land? Hectares suitable for woody energy crop production ( total = 91.1M ha)


How much biomass could be produced on that land? Tons of woody energy crop biomass that could be produced (total = 1,294M Mg)


Downing, M. and R.L. Graham 1996. “The potential supply and cost of biomass from energy crops in the Tennessee Valley Authority Region” Biomass and Bioenergy 11(4):283-303.

How will economics affect regional feedstock supplies?

Relying on breakeven- prices to determine feedstock price


Modeling breakeven prices by subregion and landuse class

Modeled land rents


Arriving at a regional supply curve


Graham, R.L., Downing, M, M.E. Walsh. 1996. “A framework to assess regional environmental impacts of dedicated energy crop production.” Environmental Management 20(4):475-486.

Next - How can we quantify the potential environmental impact of energy crop production?


Testing out the framework in the TVA region – Comparing current crops to switchgrass using EPIC


Replacing conventional crop production with switchgrass production had environmental benefits in Tennessee


Graham, R.L. and M.E. Walsh 1996. “Evaluating the economic costs, benefits and tradeoffs of dedicated biomass energy systems: the importance of scale”. Proceedings of the Second Biomass Conference of the Americas. Pp 1428-1436.

How does choice of scale make a difference?

Economic scale

Environmental scale

Enterprise Farm Biorefinery

Fields Plant Site

Community Watershed

Industry/ nation

Biomes/ globe


http://www1.eere.energy.gov/biomass/pdfs/final_billionton_vision_report2.pdf

THE 2005 BILLION-TON RESOURCE ASSESSMENT

Key Attributes National estimates – no spatial information

No cost analyses – just quantities

Long-term, inexact time horizon (2005; ~2025 & 2040)

No explicit land use change modeling

2005 USDA agricultural baseline and 2000 forestry RPA/TPO

Crop residue removal sustainability addressed from national perspective; erosion only

Erosion constraints to forest residue collection


•  Forest resources

–  Logging residues

–  Forest thinnings (fuel treatments)

–  Conventional wood

–  Fuelwood

–  Primary mill residues

–  Secondary mill residues

–  Pulping liquors

–  Urban wood residues

•  Agricultural resources

–  Crop residues

–  Grains to biofuels

–  Perennial grasses

–  Perennial woody crops

–  Animal manures

–  Food/feed processing residues

–  MSW and landfill gases

•  About one-half of the land in the contiguous U.S.

–  Forestland resources: 504 million acres of timberland, 91 million acres of other forestland

–  Agricultural resources: 342 million acres cropland, 39 million acres idle cropland, 68 million acres cropland pasture

2005 Billion ton resource base


•  Total cropland resource approached 1 billion dry tons/year including perennial energy crops

–  Continuation in yield growth trend for corn and small grains

–  Shift to conservation tillage and no-till

–  Improvements in residue collection equipment

–  Perennial energy crops (40 - 60 million acres)

2005 Billion ton future cropland potential


•  Forestland residue potential was about 370 million dry tons

– Most currently used biomass comes from forestlands

–  Unused primary sources easily exceed 100 million dry tons (logging, other removals, & fuel treatment thinnings)

–  Conservative assumptions based on accessibility, recoverability, and merchantability

2005 Billion ton future forestland potential


http://www1.eere.energy.gov/biomass/pdfs/final_billionton_vision_report2.pdf

•  Objectives of update –  Address biomass resource

availability (quantities and prices/costs) collectively and spatially •  Timeframe – 2010 - 2030 •  Spatial resolution – US County

(~3100) –  Impose sustainability constraints to

feedstock production/collection –  Explore impact of possible

technology –  Make the data and analysis

transparent and available to others •  Public report (late 2010; 200 pages) •  All output will be available on the

Web

Current Activity – the Billion ton update


Oak Ridge National Laboratory Robert D. Perlack Craig C. Brandt Anthony F. Turhollow Lynn L. Wright Laurence M. Eaton Anna M. Shamey Jacob M. Kavkewitz Matt H. Langholtz Mark E. Downing Robin L. Graham

Idaho National Laboratory David J. Muth J. Richard Hess Jared M. Abodeely

Kansas State University Richard G. Nelson

State University of New York Timothy A. Volk Thomas S. Buchholz Lawrence P. Abrahamson

Iowa State University Robert P. Anex (now U of WI)

CNJV LLC Bryce J. Stokes

University of Tennessee Chad Hellwinckel Daniel De La Torre Ugarte Daniel C. Yoder James P. Lyon Timothy G. Rials

USDA Agricultural Research Service

Douglas L. Karlen Jane M. F. Johnson Robert B. Mitchell Kenneth P. Vogel Edward P. Richard John Tatarko Larry E. Wagner

University of Minnesota William Berguson Don E. Riemenschneider

Texas A&M University William L. Rooney

USDA Forest Service Kenneth E. Skog, Patricia K. Lebow Dennis P. Dykstra Marilyn A. Buford Patrick D. Miles D. Andrew Scott James H. Perdue Robert B. Rummer Jamie Barbour John A. Stanturf David B. McKeever Ronald S. Zalesny Edmund A. Gee

USDA National Institute of Food and Agriculture

P. Daniel Cassidy

USDA Natural Resources Conservation Service

David Lightle

University of Illinois Thomas B. Voigt

Contributors


GENERAL APPROACH •  County-level feedstock supply curves for major primary

cropland and forestland resources (bales at the farm or chips at the forest landing)

•  Agricultural policy model (POLYSYS) used to estimate supply curves (price –quantities) and land use change for cropland resources (crop residues and energy crops) http://www.agpolicy.org/

•  Resource cost analysis used to estimate supply curves (cost-quantities) for forestland resources

•  Several Technology scenarios developed through workshops


Scenarios Baseline scenario

–  USDA Baseline forecast for crop yields, acres, etc., extended to 2030

–  National corn yield of 160 bu/ac in 2010, increases to 201 bu/ac in 2030

–  Assumes a mix of conventional till, reduced till, and no-till

–  Stover to grain ratio of 1:1

–  No residue collected from conventionally tilled acres

–  Energy crop yields increase at 1% annually attributable to experience in planting energy crops and limited R&D

High-yield scenario(s) –  Same as Baseline Scenario except for the following

–  Corn yields increase to a national average of 265 bu/acre in 2030

–  Higher amounts of cropland in no-till to allow greater residue removal

–  Energy crop yields increase at 2%, 3%, and 4% annually (attributable to more aggressive R&D)


GENERAL APPROACH - Agricultural lands •  POLYSYS - dynamic model of the U.S. agricultural

sector •  8 major crops (maize, soybeans, wheat, sorghum, oats,

barley, rice, cotton) and hay, livestock, food/feed markets

•  Model requires meeting projected demands for food, feed, fuel, and exports

•  Biomass feedstocks = Stover, straw, energy crops (perennial grass, coppice and non-coppice woody, annual)

•  3,110 counties

•  Assess land use change (with and without a biomass market)

•  Data and input assumptions important •  Land base includes cropland (250 million acres), cropland

pasture (22 million acres), hay (61 million acres), permanent pasture (118 million acres) •  Pasture can convert to energy crops if forage made up

through intensification •  Restraints limiting land use change

•  Carbon and energy flows – can be linked to model predictions

•  Forest resources exogenous to the model

For model background see: De la Torre Ugarte, Daniel G., and Darrell E. Ray. 2000. “Biomass and Bioenergy Applications of the POLYSYS Modeling Framework,” Biomass and Bioenergy 4(3):1-18, May.

University of Tennessee - Agricultural Policy Analysis Center (APAC) (http://www.agpolicy.org/)


AGRICULTURAL CROP RESIDUES •  Grower payments – based on nutrient value plus organic matter ($1/dry

ton) plus $10/dry ton

–  Stover ($24.60 - $27.50) and straws ($23.90 - $26.20)

–  Did not value potential affects on tillage, yields, compaction, etc.

•  Crop residue collection costs = function of amount removed per hectare

–  Moderate Removal: Combine residue spreader disengaged, bale windrow left behind

–  Moderately High Removal: Bar rake run over standing stubble, bale windrow

–  High Residue Harvest: Flail shredder cutting standing stubble and collecting residue, bale windrow

•  Total farm gate costs = grower payment + collection cost


CROP RESIDUE SUSTAINABILITY •  Sustainability - residue retention coefficients estimated using RUSLE2, WEPS, and SCI for

erosion and soil carbon (developed with residue removal tool – Wilhelm et al 2010. Industrial Biotechnology 6(5):271-287)

–  Separate coefficients for reduced till and no-till; no residue removal under conventional till

–  County specific based on ag soils in county and cropping practices

•  Collection efficiency (equipment constraint)


DEDICATED ENERGY CROPS •  Perennial grasses (e.g.,

switchgrass), non-coppice woody crop (poplar, pine, eucalyptus), coppice woody crop (willow), and a annual energy crop (energy sorghum)

•  Production costs –  Site preparation –  Planting material –  Weed control

(establishment) –  Fertilization in years

•  Harvesting (mow, rake, bale, move/stack roadside)

•  Productivity based on field trials and assumed to increase by 1-3%/yr


Energy Crop Sustainability & Restrictions

•  Assumed BMPs for establishment, cultivation, maintenance, and harvesting of energy crops

•  Energy crops not allowed on irrigated cropland & pasture •  Conversion of permanent pasture and cropland used as pasture

constrained to counties east of the 100th meridian except for Pacific Northwest

•  Energy crops returns must be greater than pasture rent plus additional establishment and maintenance costs

•  A set of restraints used to limit the amount of cropland, cropland used as pasture, and permanent pasture switching to energy crops in a given year and in total (e.g., 10% of cropland per year and 25% in total)

•  Annual energy crops (i.e., energy sorghum) limited to non-erosive cropland and part of multi-crop rotation


•  Fuel treatment thinnings (reduce fire risk)

–  Costs = fn (stumpage, harvest, skid, chip)

•  FIA data (~37,000 permanent field plots)

–  Exclude roadless areas and reserved, steep, and wet lands

–  All fire regime condition classes

–  Thin over 30-year period

•  Logging residues (based on Forest service projections for conventional logging)

–  Costs = fn (stumpage and chip)

•  Transition to integrated harvest system

–  Conventional harvesting and logging residue plus thinnings for fire hazard reduction and health treatments

FIA plot

Harvest cost (FRCS) = fn (30% max SDI, slope, …) Average

skid distance Chip costs

Stumpage

USDA/FS – Ken Skog, Jamie Barbour, Dennis Dykstra, Patricia Lebow, Pat Miles, Marilyn Buford

GENERAL APPROACH - Forest lands


•  Only non-merchantable biomass for residues and thinnings –  Tops and limbs; small trees (West – 1 to 7 in. dbh;

North/South – 1 to 5 in. dbh) •  Regional stumpage prices

–  Assume stumpage price of $4/dry ton for tops/branches, increases to 90% of pulpwood stumpage

•  Costs of collection/harvest at landing based on operability, removal volume

•  Logging projections based on Forest Service projections

FOREST RESIDUES (cont.)


•  Land base – 504 million acres of timberland & 91 million acres of “other forestland”

•  Evaluated the “state-of-the-science” for biomass removal and implications for erosion, soil nutrients, biodiversity, soil-organic carbon, and long-term soil productivity

•  Developed “conservative” woody retention levels by slope classes within the context of the science review

–  Logging residues •  30% left on-site

–  Fuel treatment thinnings •  Slope is <40% – 30% of residue is left on-site •  Slope is >40% to <80% – 40% of the residue left on site •  Slope is >80% – no residue is removed (no limbs or tops yarded)

–  Removed steep and wet sites –  Excluded sites requiring road building

•  Made cost assumptions based on the use of integrated logging systems and the use of Best Management Practices; Used USFS Fuel Reduction Costs Simulator model

FOREST RESIDUE SUSTAINABILITY

A few preliminary results for illustration

Crop Residue Simulated Supply Curves

COUNTY-LEVEL CORN STOVER SUPPLY

Supply of stover at farmgate price of $50/dry ton Supply of stover at farmgate price of $60/dry ton

Energy Crop Simulated Supply Curves – Baseline Scenario

•  Supplies increase over time due to yield growth and woody crop production •  Energy crops displace mostly commodity crops at low supply curve prices and

move onto pasture at higher prices

Baseline 2030 High-‐yield (4%)

Energy Crop Simulated Land Use Change •  Land use change at highest simulated prices by 2030

–  22 to 30 million acres cropland –  40 to 50 million acres pasture

State Potential to Supply Crop Residues and Energy Crops

•  Potential supplies are generally widely distributed –  Considerable

perennial grass potential in Southern Plains

–  Residue in Midwest and Northern Plains

–  Woody crops in the North and South

Baseline scenario - $60/dry ton; year 2030

2030 county estimates

Forest Residues Results •  Forest residues are widespread in the Southeast, North, and

Northwest

Summary Findings •  Forest residue biomass potential is somewhat less – removal of unused

resources, decline in pulpwood and sawlog markets •  Crop residue potential is less – consideration of soil carbon, no residue from

conventionally tilled acres •  Energy crop potential is greater – permanent pastureland, POLYSYS modeling

Key RD&D needs (our experience) •  Large field trials of energy crops where

–  Commercial yields are achieved –  Commercial production costs can be measured –  Environmental sustainability criteria can be measured and tested

•  Soil carbon/erosion •  GHG emissions (NOx, CO2, CH4) •  Water quality/consumption •  Wildlife/biodiversity

•  Better conventional and energy crop yield projections •  Economic models that explicitly capture geographic variation and

crop management & land use drivers at a relevant scale •  Environmental models that can be easily implemented and capture

geographic and crop management features •  Linking economic and environmental models at the appropriate scale •  Sustainability indicators that are relevant and easily measured •  Model transparency especially underlying assumptions

What’s next? •  Improving input data and gaining more spatial specificity

•  Factor in climate change effects on potential yield •  Greater geographic specificity in management inputs

•  Introducing uncertainty in the modeling •  Yield •  Prices

•  Link environmental outcomes to projected landuse change •  Soil carbon •  Fossil fuel use ( done) •  Fertilizer use (done) •  GHG emissions from soil processes ( CH4, Nox) •  Water quality & quantity

•  Optimization with multiple criteria – economic and environmental

Thank you!

Questions?

Documents

A Reflection on the Evolution of Estimating Future U.S. Ligno-cellulosic Biomass Feedstock