2005 OBP Biennial Peer Review

Processing Integration

Dan Schell

Biochemical Platform

November 15, 2005

Project Goals and Objectives

• Investigate integrated processing to reduce risk of industry-led efforts to commercialize biomass refining technology • Improve understanding of process chemistry and

integrated performance

• Identify problems and showstopper issues

• Develop integrated testing methods, tools, and capabilities

Overview

Budget• USDA/Universities• Subcontracts

• Baylor University

• Hauser Laboratories

• Hazen Research

Partners

• Barriers• Process integration

• Pathways• Agricultural Residues

• Energy Crops

Work Objective

$0.0

$0.5

$1.0

$1.5

$2.0

$2.5

Funding ($M)

FY04 FY05 FY06(planned)

Platform Fit with Pathways

Program Outputs

FeedstockR&D

SugarsR&D

ThermochemicalR&D

ProductsR&D (from)

•Residual Starch Conversion•Fiber Conversion

Systems-level demonstration and validation by 2009

Corn Wet Mill Improvements

(Corn)

Element Strategic Goals

Corn Dry Mill Improvements(Corn, Grain)

Agricultural Residue Processing(Corn Stover, Wheat Straw, Rice Straw)

•Biomass Fractionation•Sugars Production

Sustainably supply biomass to

biorefineries

Low-cost sugars from lignocellulosic

biomass

Biomass Program Strategic Goal

Cost-competitive biorefinery technologies

for the nation’s transportation, chemical

and power industries

IntegratedBiorefineries

•Residual Starch Conversion•Fiber Conversion•Milled Grain Fractionation

Energy Crops(Perennial Grasses,

Woody Crops)

Pulp and Paper Mill Improvements

(Mill Wastes, Wood)•New Fractionation Processfor hemiicellulose removal

Chemical building blocks from

lignocellulosic biomass

Fuels, chemicals and power from bio-based sugars and chemical

building blocks

Systems-level demonstration and validation by 2012

Systems-level demonstration and validation by TBD

Systems-level demonstration and validation by TBD

Systems-level demonstration and validation by 2010

•Biomass Fractionation•Sugars Production

Approach

• Focus on integrated performance testing using a model feedstock (corn stover) and a baseline process based on thermochemical dilute acid pretreatment followed by enzymatic cellulose hydrolysis

• Directly addresses process integration barrier

• Working to understand current performance and demonstrate progress towards the sugar platform cost target, while improving integrated testing capabilities and identifying showstopper issues

• Measure performance relative to technical targets established by economic analysis

2.0Biochemical Platform

2.1Pretreatment and

Enzymatic Hydrolysis

2.2Feedstock-Biochemical

Interface

2.3Process Integration

2.4Targeted Conversion

2.5Biochemical Platform

Analysis

2.1.1Pretreatment and

Enzymatic Hydrolysis

2.1.1.1CAFI 2 Support

2.1.1.2Feedstock Qualification

2.1.1.2.1Extended Fiber

Pretreatment

2.1.1.3Forest Biorefinery

2.1.1.4Enzymatic Hydrolysis

2.1.1.5Exploratory Pretreatment

2.1.3Integration of Leading Biomass PretreatmentTechnologies (CAFI 2)

2.2.4 Preprocessing and Storage Systems

Development/Qualification

2.2.5Preprocessing Feedstock

Supply

2.3.1 Processing Integration

2.4.1Targeted Conversion

Research

2.4.1.1 Chemical Conversion

Fundamentals

2.4.2Biological Processing

Fundamentals

2.4.3Plant Cell Wall Deconstruction

2.4.4BSCL and Genomics

2.4.4Industrial MembraneFiltration & Short Bed

Fractal Separation

2.3.1Feedstock Variability

2.3.2Integrated Processing

2.3.3Analytical Methods

7.04.2.GO41221Rheology and CFD Modeling

Work Breakdown Structure

NRELAcademiaIndustryEarmark

Project Structure

Processing Integration

Feedstock Variability

Analytical Methods

Integrated Processing

Test integrated performance High solids operation Assess advanced enzymes

Develop and improve methods

Distribute methods

Understand breath and impacts of feedstock variability

Barriers

Commercial Success Barriers

Price of Sugars from “Cellulosic” Biomass

Major General BarriersFeedstock Cost

Sugars CompositionSugars Yield

Conversion RateSugars Quality

Capital Investment

R&D Technical BarriersFeedstock-Sugars Interface

Biomass PretreatmentEnzymatic Hydrolysis

Sugars ProcessingProcess Integration

Feedstock Variability Stover Compositional Database

• Comparison of commercial (FY02) and non-commercial (FY05) corn hybrids

• The addition of non-commercial hybrids in the sample set has expanded the range of cellulose and xylan compositions

Feedstock Variability Stover Compositional Database

• Lignin content varies only modestly between commercial and non-commercial hybrids

• Further data mining is necessary to answer questions such as, “How does carbohydrate content correlate with lignin content and/or other components?”

• We believe the corn stover database now captures the extent of feedstock compositional variability

• No additional survey work is planned for corn stover

Understanding Risks

Frequency Chart

.000

.004

.008

.012

.016

0

8

16

24

32

$0.9987 $1.0888 $1.1789 $1.2689 $1.3590

2,000 Trials 8 Outliers

Forecast: MESPComposition used in 2002

Design Report

Analytical Methods Identifying the Problems

Cellulose Xylan Lignin

Extractives

Other Hemi.

Uronic Acid

Acetyl Ash

Protein Sucrose

Corn Stover

6.6%

60.3% 30.7%

3.6%

1.9%

2.4%

Pretreated Corn Stover Solids Liquor Furfural

Other XyloseGlucose

Pretreatment

Improving HPLC-Based Sugar Analysis Implementing Solutions

• Improved baseline resolution produces more accurate and reproducible compositional data

• Still need reliable method to measure fructose in hydrolysates

BioRad HPX-87P, RI Detector

Shodex SP-0810, RI Detector

New

Old

BioRad HPX-87P, RI Detector

Improving Lignin Analysis

• Current wet chemical methods for lignin determination

• Behavior-based definition: Lignin = Acid Insoluble Residue • Valid assumption for wood

• Invalid for agricultural residues and herbaceous materials

• Interferences from protein, carbohydrate degradation products, extractives, and silica

• Unacceptably high error, ± ~25%

• Application of method gives inaccurate mass closures• Poor correlation with spectra limits ability to develop

reliable rapid analysis method

Strategies for Improving Lignin Analysis

• Investigate alternative analytical methods for measuring lignin such as those now used in the food sciences

• Improve understanding of the fate of protein and extractives and their effects on lignin measurements

• Develop functional group-based lignin determination

• Use this new information to develop more accurate spectroscopic-based rapid analysis methods for lignin

Understanding Extractives

• Subcontract issued to Baylor University (work began in May 2005)• Goal

• Identify 90% or more of the extractives and develop analytical methods for measuring the concentrations of extractive components

Progress• Have identified several

major constituents that comprise most of the extractives

• Currently developing analytical methods for these materials

Kramer Feedstock

4.15.6

6.5 6.7 6.7 6.8 6.8

1.0

1.41.5 1.5 1.5 1.6 1.7

1.0

1.31.4 1.5 1.5 1.5 1.5

3.5

5.9

5.9 6.07.0 7.0

7.9

0

2

4

6

8

10

12

14

16

18

0.00 0.04 0.05 0.10 2 4 6

Hours in water extraction

Per

cen

t d

ry w

eig

ht

Hour in Water Extraction

Dry

Wei

gh

t (%

)Unknown SucroseAshProtein

Improving Rapid Analysis Methods

• Real-time methods are needed to support commercial biorefinery operation

• Process control and optimization

• Improved efficiency for in-house and CRADA research projects

• Reduces cost

• Increases number of samples/experiments that can be run

On-line Testing and Validation

Optics over conveyor weigh belt

SpectrometerControls

Direct light spectrometer installed in pilot plant

Moving Toward Deployment

• Methods have been transferred to Vision® Software

• Provides a more flexible and robust platform for using these methods on other instruments

R2 = 0.84

R2 = 0.87

R2 = 0.87

R2 = 0.79

0

5

10

15

20

25

30

35

40

45

50

0 5 10 15 20 25 30 35 40 45 50

Wet Chemistry Values (%)

NIR

Model

Pre

dic

ted V

alues

(%

)

GlucanXylanLigninProtein

Drive Towards High Solids Operation Pretreatment

• High solids operation significantly increases sugar concentrations

• Pretreatment is possible at 30% solids loading without significant yield loss

• Still need to improve hemicellulosic sugar yields

• Significant reduction in minimum ethanol selling price (MESP) can be realized through reduced operating and capital costs

$1.00

$1.10

$1.20

$1.30

$1.40

15 20 25 30 35 40

Reactor Solids Loading (wt.%)

ME

SP

($/

ga

l)

20

40

60

80

100

120

140

160

180

15 20 25 30 35 40Solids Loading (wt.%)

Su

gar

Co

nce

ntr

atio

n (

g/L

) Monomeric xylose

Total xylose

Total sugars*

Drive Towards High Solids Operation Enzymatic Cellulose Hydrolysis

• Inhibition by sugars and mass transfer limitations become important issues at high solid concentrations

• Further cost reductions are possible with high solids enzymatic cellulose saccharification

$0.96

$0.98

$1.00

$1.02

$1.04

$1.06

$1.08

$1.10

$1.12

18% 20% 22% 24% 26% 28% 30% 32%

Solids to Saccharification (wt. % )

MESP (

$/g

al)

30%

40%

50%

60%

70%

80%

90%

100%

5% 10% 15% 20% 25% 30% 35%

Insoluble Solids (w/w)

7-D

ay C

ellu

lose

Co

nve

rsio

n

0

20

40

60

80

100

120

140

Glu

cose

Co

nce

ntr

atio

n (

g/L

)

Washed Solids

Whole Slurry

Total glucose

Glucose generated by enzymatic hydrolysis

Genencor Spezyme (40 mg/g)

Understanding Process Relevant Performance Recycle Water Studies

• Performance is significantly affected at modest solids concentrations and recycle water ratios

• Inhibitors besides acetic acid are responsible for poor performance

0

10

20

30

40

50

60

70

80

90

100

1 2 3 4

Yie

ld o

r C

onve

rsio

n (%

)

Glucose Conversion

Xylose Conversion

Ethanol Yield

20% Solids25% Recycle Ratio

20% Solids10% Recycle Ratio

15% Solids10% Recycle Ratio

15% Solids25% Recycle Ratio

Hydrolysate liquor conditioned by overliming and then fermented using xylose-utilizing Z. mobilis

Understanding Process Relevant Performance Integrated Performance Testing

S/L Separation

Mixing

Lime MixingS/L Separation

Acidification

Fermentation

Saccharification

Water

Liquid

Solids

Lime

Water

Acid

Gypsum

Enzyme

Nutrients

Ethanol

Pretreated Corn Stover

Cells

0

10

20

30

40

50

60

70

0 1 2 3 4

Time (d)

Co

nce

ntr

atio

n (

g/L

)

CellobioseGlucose

XyloseEthanol

0

10

20

30

40

50

60

70

80

90

100

CelluloseConversion

Xylose Consumed Ethanol Yield

Yie

ld o

r C

on

ver

sio

n (

%)

Partially washed solids recombined with liquor conditioned by overliming

Ethanol yields are low due to unutilized glucose (25% left as mono- and oligo-glucose) and incomplete xylose utilization

Integrated processing illustrates limitations of current ethanologens

Cellulose hydrolysis with Genencor Speyzme (40 mg/g, 45C, pH 4.8) and fermentation with Z. mobilis (35C, pH 5.0)

Performance Summary

2002 Estimate

Latest Test Results

2012 Target(FY06

Budget)

2020 Market Target

Minimum Ethanol Selling Price $4.13 $1.73 $1.07Feedstock Feedstock Cost ($/dry ton) $53 $45 $30Pretreatment Solids Loading (wt%) 19% 30% 30% 30% Xylan to Xylose 68% 70-75% 90% 90% Xylan to Degradation Products 16% 8% 5% 5%Conditioning Xylose Sugar Loss 13% 13% 13% 0% Glucose Sugar Loss 12% 12% 12% 0%Enzymes

Enzyme Contribution ($/gal EtOH) $1.61 $0.22 $0.10Saccharification & Fermentation Total Solids Loading (wt%) 13% 20% 20% 20% Combined Saccharification & Fermentation Time (d) 10 7 7 3 Overall Cellulose to Ethanol 86% ~75-80% 86% 86% Xylose to Ethanol 76% ~25-50% 85% 85% Minor Sugars to Ethanol 0% 0% 0% 85%

Interim Stage Gate Overview

• Processing Integration Task Interim Stage Gate Review

• Stage B project

• Meeting held September 15, 2004

• Reviewers:

• Rob Anex (Iowa State University)

• Susan Hennessey (Dupont)

• Dale Monceaux (Katzen International)

• Quang Nguyen (Abengoa Bioenergy)

• Amy Miranda (DOE OBP)

• Jim Spaeth (DOE GO)

• Stan Bower (NREL)

Review Meeting Feedback

• Feedback received in four categories• Large view issues

• Feedstock variability

• Analytical methods

• Integrated processing

Large View Issues

• The answer to many questions raised at the meeting was that the activity resides in a different part of the Biomass Program. However, systems solutions are required and fragmentation of the overall effort into manageable sized projects should not be allowed to silo the Program.

• There is no capital cost reduction target. NREL has capital cost modeled, but inclusion of depreciation in the MESP is not an adequate reflection of the barrier of raising large capital. Capital reduction should be targeted and tracked.

Feedstock Variability

• Efforts to piggyback on work being performed by the USDA and others are good and should continue. You not only obtain well-characterized samples at very low cost, but add substantial value to the studies being performed by the researchers that provide the samples.

• We have continued this effort and over 200 new samples were acquired this year.

• Need to extend the variability studies to determine the impact of corn stover variability on pretreatability (sugar yields), enzymatic cellulose hydrolysis, and fermentability. There is a theoretical impact based on carbohydrate content; how does it play out in final yields?

• We acquired several new large lots of corn stover that enable this work to proceed.

• Need to expand interface to other areas to allow studies of impact of storage on feedstock composition, pretreatability, enzymatic cellulose hydrolysis and fermentability.

• This work should be accomplished in the Feedstock-Biochemical Interface Area with NREL providing bioconversion processing and analytical support.

Analytical Methods

• Develop on-line monitoring capability, especially for monitoring the enzymatic saccharification reactor. Enzymatic saccharification is the least understood of all the unit operations. The second area for application of on-line monitoring would be in the fermentation reactor.

• Moved forward last year with on-line feedstock monitoring• Other on-line monitoring development activities planned in future years

• There is little point in trying to develop process control strategies based on on-line monitoring. There is no clear target for what is being controlled and the control parameters will be process specific.

• Will not be done

• Functional group based lignin determinations is an area that should be pursued, as well as work to characterize chemical changes to lignin during and after pretreatment.

• Work planned in future years, probably via subcontract

Integrated Processing

• Perform thin studies that indicate problems and generate representative results, that is, determine the problem, skip the solution. In-depth studies that provide solutions are not justified because they will not be generally applicable across a range of processes.

• Integrated process testing this year identified problems with high recycle water use and with achieving good C5 sugar conversion.

• Enzymatic saccharification time is too long and needs to be characterized with unwashed materials, that is, with background components (non-sugars) present during enzymatic saccharification. Determine components that are inhibitory to the cellulases (e.g., Maillard reaction products). Perform spiking studies to determine what chemicals inhibit cellulases.

• Initial performance results in the presence of background sugars were presented. New work going forward in the Pretreatment & Enzymatic Hydrolysis Task.

• Move forward efforts to characterize waste streams; three years away is too late for those studies to be useful. Also generate real data on thin stillage evaporate. What is in it besides water? Are any of the streams or residues appropriate for putting back on the fields?

• Initial effort began this year to understand effect of recycle water on process performance.

Integrated Processing

• Examining the gypsum question is low priority and should not be undertaken. The fact that gypsum is an issue was an important recognition, but the solution will be unique to each process and approaches for handling gypsum are well understood from existing industries.

• Work in this area was eliminated

• Advance efforts to understand new feedstocks and new pretreatments. • Work began this year in the Pretreatment & Enzymatic Hydrolysis Task

• Interface Question: What is the root cause of biomass recalcitrance? Generate residue that can be characterized, both compositionally and structurally.

• This work resides in Targeted Conversion Research Task

Future Work Feedstock Variability

• Survey of corn stover composition finished• Maintain collaborations with USDA and

academic institutions performing field studies to advance understanding of the effect of environmental and genetic factors on stover composition, primarily through our expertise with rapid biomass analysis techniques

• Transition feedstock procurement activities to Idaho National Laboratory (INL)

• Develop collaborations with INL and other laboratories to explore variability issues for other promising feedstocks (e.g., switchgrass)

SWITCHGRASS

Future Work Analytical Methods

• Improve rapid analysis methods hand-in-hand with improved wet chemical methods• Direct Light spectrometer for feedstocks

• Demonstrate ability to measure stover composition on-line (FY06)

• Fourier Transform Infrared (FTIR) in-line probe for pretreated slurries and solids (FY07)

• Near Infrared (NIR) opti-probe for fermentation broths (FY08)

• Improve wet chemical methods for agricultural and herbaceous materials (FY07 and beyond)• Lignin, extractives

• Automation

• Distribute and publish new methods as developed (ongoing)• 2500 hits on EERE/Biomass web site accessing laboratory analytical methods in last

quarter of FY05

Future Work Integrated Processing

• Determine effect of corn stover compositional and structural variability on pretreatment hemicellulose hydrolysis yields and enzymatic cellulose digestibility (FY06)

• Investigate integrated performance of new advanced enzyme preparations from Genencor and Novozyme

• One new preparation will be tested this year (FY06)

• Other preparations will be tested in outyears (FY07 and beyond)

• In collaboration with the thermochemical platform, produce representative lignin-rich process residues for thermochemical conversion testing (FY06)

• In collaboration with the Feedstock-Biochemical Interface, determine effect of wet storage on process performance (FY07 and beyond)

Future Work Integrated Processing

• Improve hemicellulose conversion yields in dilute acid pretreatment (FY07 and beyond)

• Collaborative effort across Biochemical Platform

• Continue to supply process materials to stakeholders, industry, and universities (ongoing)• In 2004 and 2005, we supplied over 100kg of raw stover and over 1500 kg

(wet) of pretreated stover to 5 industry stakeholders, 12 academic institutions, and 2 government laboratories

End