1 | Bioenergy Technologies Office

1 | Bioenergy Technologies Office eere.energy.gov

2013 DOE Bioenergy Technologies Office

(BETO) Project Peer Review

2.6.1.2 Analysis for Production

- Technical & Market

May 20, 2013

Technology Area Review: Biochemical Conversion

Sue Jones, Aye Meyer

Organization: Pacific Northwest National Laboratory PNNL-SA-95160

This presentation does not contain any proprietary, confidential, or otherwise restricted information


Goal Statement

• Support the DOE Bioenergy Technologies Office’s goal to reduce the estimated mature technology processing cost for converting cellulosic feedstocks to hydrocarbon fuels via biochemical and catalytic means by developing analysis tools and techno-economics to identify pathways with the potential to meet the $3/gallon gasoline equivalent target

• Perform analysis as requested by the Biochemical and Catalytic Conversion Technology Area


Quad Chart Overview

• Project start date: FY03

• Project end date: FY15

• FY11-FY12 100% complete, FY13 50% complete

• At-A. Lack of Comparable, Transparent, and Reproducible Analysis

• Bt-J. Catalyst Development

• Bt-K. Biochemical Conversion Process Integration

• Funding for FY11: $250k



• 11 years funding at an average

of $328k/year

Timeline

Budget

Barriers

• INL for modeled feedstock costs

• NREL for consistent modeling assumptions and work on Tech Memos

• Project management – communication as needed with other labs, and DOE

Partners


Project Overview

History

• FY03-07: Products focus -“Top 10” products from biomass report, Products from lignin report, Co-products analysis in a biorefinery

• FY08-11: Economics of ethanol production via fungal organisms

• FY10-12: Economics of co-products, economics of microbial production of hydrocarbon compatible fuels

• FY13: Preliminary metabolic modeling of oleaginous organisms to support understanding of hydrocarbon production

Context

• BETO’s portfolio has expanded to include hydrocarbon fuels

• Need for preliminary analysis of candidate technology pathways

Objective

Focusing on fungal processing, the objective is to provide economics and gap analysis in support of the BETO goals to enable the production of renewable transportation fuels, particularly, gasoline, diesel and jet blendstocks


Approach

TECHNICAL APPROACH • Quick screening, then detailed

analysis

• Iterative process

• Sensitivity analysis to understand variance & significant impact areas

• Results used to measure progress towards research goals

MANAGEMENT APPROACH Develop project management plans

• Statement of work and how it relates to DOE goals

• Quarterly milestones to minimize project slippage

• Go/No Go decision point

Frequent project communications

• Telecons with platform leads and labs – share information, use consistent assumptions

• Quarterly formal reporting to HQ

Input from

researchers,

clients, literature

Metabolic & Process Modeling

Economics & sustainability

Excel, Chemcad,

Aspen

quick analysis for Go/No Go

GOAL

Guide

Research

NEW PROCESS,

FUEL, PRODUCT


This project has three separate analysis tasks:

• Task 1: Screening economics for production of co-products in an ethanol biorefinery (FY11&12)

• Potential chemical candidates

• Avoidance of chemical market saturation

• Task 2: Screening economics for hydrocarbon production through an oleaginous organism and determination of maximum practical yield (FY12 &13)

• Preliminary economic model

• Metabolic modeling coupled with experimental work (from a separate project) to inform the economic model

• Task 3: This project also partially funded the analysis and document preparation for the BETO pathways analysis for sugars conversion (FY13)

Technical Accomplishments / Progress / Results


TASK 1: Co-Products in an Ethanol Refinery

Enzymatic hydrolyzate slipstream (10-20%) used for co-product production. Four non-food, non-pharmaceutical organic acids produced.

Acids produced in blocked continuous fashion. Equipment added for chemical production and storage.

Basis: 2011 Biochemical Ethanol Report (Humbird et al) and organic acid conversion data from literature. Analysis in 2007$.

On-site Enzyme

Production

Enzyme Hydrolysis

Ethanol Fermentation

Distillation+ Mol. Sieves

+ Solid Separation

Steam and Electricity

Generation

Wastewater Pretreatment

Feed HandlingDilute Acid

Pretreatment and Conditioning

Biogas

Crude Enzyme

Process Steam

Electricity

Corn Stover2000 mtpd

Seed cultivation

Inoculum

Process water

Steam

Sulfuric acid

Ammonia

Glucose

Nutrients

Pretreated corn stover

Organic Acid Fermentation

and Purification

Slipstream

Organic acid co-products

Ethanol

Solid slurry (cell mass+lignin)

Nutrients

Nutrients

Nutrients

Wastewater Wastewater

Solid slurry (lignin+cell mass+syrup)

A700Product and Feed Chemical Storage

A900

Utilities

A100A200

A400

A300

A1000

A800

A600

A500

Technical Accomplishments Task 1


Acids chosen as examples of good match between market size and productivities

Technical Accomplishments Task 1 - Results

Ethanol production 52 mm gal/yr

Ethanol Yield per feedstock 67 gal/dry ton

Total co-product production 77 mm lb/year

Citric acid: 20.1 106 lb/yr Itaconic acid: 13.5 x 106 lb/yr

Lactic acid: 18.8 106 lb/yr Succinic acid: 24.6 x 106 lb/yr

Key Outcomes:

10-20% hydrolysate slipstream diverted to chemicals production has potential to reduce ethanol costs by $0.15-0.45/gal

Additional capital offset by co-product value

Multiple co-products prevent market saturation


Basis: 2011 NREL Biochem Ethanol Report. 2011$

Yields and product recovery from literature:

Base case lipid yield = 0.17 g oil/g sugar produces 25 gallons of hydrocarbon/dry short ton of biomass


Task 2: Hydrocarbons from Oleaginous Yeast

Relevance:

1st look at oleaginous yeast pathway

Model will be updated in FY13 with experimental results


Improvements needed to reach $5/gal: • Improve lipid yield

• Increase tolerance to inhibitors (such as substrates, intermediates and products)

• Reduce aeration & nutrients

• Reduce fermentation time

• Auto-release triglyceride

• Produce products in middle distillate range, e.g. C14

Lignin usage for fuel or products needed to reach $3/gal

Literature Based Screening Economics

Technical Accomplishments Task 2 - Results

HT = hydrotreating, WHSV=HT weight hourly space velocity

Conclusions: yield and carbon usage dominate costs – FY13 modeling and experimental work will provide directions for improvement


101

840

AREA 100

Feed Handling

M104 - M105

201

104

2201

2232

232

P-201

T-201

5516Vent

2203

2217

203

T-20 3

M-203

215

105

M-206

202

QM-204

M-207 207

204

T-204

2209

T-209

273

A-2 0 1

2011

2012

224

213

222

P2 03 - P2 08

223

209

228

275

254 227

217

225

2013

12

252H-244

255

CWS

CWR

1244

1245

2252

253

AREA 200 Pretreatment and Conditioning

2215

710

Acid P r etr eatm ent

Conditioning

St i rred tankSt i rred tank

T-310

310

P-306

306

T-306

CSL

DAP

301

P301-P302

312

311

3304

AREA 300 Fermentation and Oil Separation

S acchar ification

S eed cultivation

Fer m entation

H-301

211212

203

214

307

23Air450 454

NH3

SO2

CSL

401

F-401

P-401

4401

407

408

453

F-400400

T-405

405

401

Glucose

Process H2O

27

4402

402

409

406

405

404

435 VENT

Cellulase P r oduction

Cellulase S eed Pr od.

AREA 400 Cellulase Product Fermentation

P-40 5

4405

443

440

CSL

CNUTR

T-406

400

P-406

406

4406

4164416

410

442

441

SO2

NH3

302

P-31 0

3310

H-310

3306

306

H-300

1304

1305

3000

13011300

420

Filt ration

AREA 500 Hydrotreating and Product Separation

403

4421

6602

903

Wel l water

45

260

To WWT

606

T-606

CIP

6606

603

604 605

6602

Anaerobic

S611-Fi l ter

611

B-60 6

6061

T-608 608

Biomass prod

44609

611

6608

VENT

630

632

Caustic

631

621

617

Aerobic

13

16

622

Air

Biomass prod

630

609

625

6609

P-609

Fi l ter R-609620

6616

P-616

6620

624

6610

613

602

AREA 600 Wastewater Treatment2260

8821

6624

S-610 6100

B-14

6282

6281

B-1 3

6629

RO629

B-106010B-12

640

614

B-4A

6007

628

6008

B-821

606

6005

B-3A

6003

M -64 0C

640

6012

6015

627Brine6007

6607

EVAP

HC from SCO fermentation

Front End as in Biochem 2011 Design Case

2,000 metric ton/day corn stover

Sludge

Area 800 Combustor and Steam Turbine

T 114 C

P 62.2 atm

W 340000 kg/h

T 277

P 13.0

T 27

P 1.0

T 28

P 1.0

T 165

P 60.5

T 165

P 60.5

To S 216To S 215

T 246 C

P 9.5 atm

W 104 kg/ h

804

803

815

820

830

831

835

806

811

863

861851

852

853

868

828

65 67

68

69

70

8817

8823

8812 815

8816

872

849

39

25

615

B iogas

8022

623To M -803

WWT Sludge

Lignin

600

T 117

P 13.0

H-801

H-811

8818 T 277

P 60.5 31

821 T 2 7 7

P 6 1

806

8049

30

T 5 5 7

P 1

805 T 870

P 1.0

814888

778216

8215

8814

S team user8811

P-811

8152

8151

863

T804

P-804

8161

H-812

8812

T-826

P-826

826

8826 8130

812

BFW

T 8 2 4

P 1

T-828

8211

8212

From S8211

62

8813

8150

T 113

P 62

W 340000

T 33

P 5

W 34600

822

8811

8805

8806

8807

216

218

83

250

2218

2812 H-812

862

251

W 7735 kg/ h

W 772 050 kg/ h

3100

3060

1310

1312

CWS

CWR

W 6 0 2 3 9 2 k g / h

W 220 603 kg/ h

M-401

451 452

1401 CWS

1402

CWR W 134 579 kg/ h

HP steam

HP steam

5501

M-811 cond

865

8153

816

1811

1812

CWS

CWR

304

F-300 coi ls

8803

8043

8044

6623

T 192

P 1.0

92

To S904

626

904

9944

Process H2O

Process H2O

NH3

CWS

CWR

301

F-301 coi ls

3011

1320 1321

1322

4420

Chi l led H2O

310195

Chi l led H2O

1330

1331

1332

T-310 Coi ls

3310

974400

44014400 4403

F400,F401 coi ls

98

VENT

423

4423

Chilled H2 O

W 155 8484 1 kg / h

LP STEAM USED FOR EVAP.

6640

To 6640

From 640

To S8821

BFW MU

811

4000

274

To part icul ate removal

AIR

309

CSL

303

DAP 3301310

60

F-301

F-300

550

H 2581 582

583

5800 5801 5802 5803

551

584

5804

T 40 8

P 7

T 15 0

P 7

T 31 9

P 19

T 15 0

P 19

Guard Bed

51 09

51 02

H ydrotreat ing

11 1

63AIR

30 0

30 0

33 00

33 17

31 5

Fe rm enting glucose

Fe rm enting xy los e

33 0111 3

33 0233 0333 04

Fe rm enting minor s ugars and suga r poly m ers

VENT

31 51

31 5231 5331 54

31 55

31 5633 05

33 6

33 06 33 0833 0731 57 31 59

Ga la ctose

M annoseArabinose

13 03

Glu Olig .

Xy lo Ol ig. Ga la c Olig . M ann Olig . Arabi Olig .

LP STM

560

TO W W T

5530535

572

P-530

To Combusto

121TO B U R N ER

H-201

H-201

H 2SO4 Sol.

80

AIR

78

79

70

571

Lignin+Bioma

81

82

86

501

Solvent Extraction

Solvent Evapolat ion

83

89

LP STEAM

585 T 36 2

P 59

5100

340

343

345

348 349

823

3350

350

351

3351

351

352

Hexane Recyc led

M U HEXANE

560

17

3501

3502

3505

35573558

3507

3559

5800

599

Air

Biogas

8021

804

5560

6615

596

597

5599

D iesel

99% of theoretical maximum

FROM HT

5571

813

318

319320321322

325

33093311331233133314324

Ce llu loseXy la nGa la ctoseM annoseArabinose

W 385000

BFW

20601

P-602

C-502C-501

P-209+239

31

8040

T 436 C

P 1.0 atm

W 497484 kg/h

M -804 FAN

T 139

P 3

305

570

549551

553

550

T 323

P 59

T 380

P 59

T 523

P 59

T 140

P 58

T 294

P 59

W 8823.11

552

541

14

540

593

TO WWT

552

From Str 94

T 464

P 59

580

32 36

92 7

Air for

Fermenta tion and Seed Culv ita tion

5100

P-610

33

544 15

Technical Accomplishments Task 2 (FY13)

Task 2 – Oleaginous Yeast Metabolic Model Lipomyces starkeyi

• Goal: improve modeled economics – what are the practical yield limits?

• Approach: combine process modeling & economics, metabolic engineering with experimental work

• Leverages published yeast models to reconstruct L. starkeyi metabolism

• 30-Liter bioreactor data from 2.4.1.2 informs both metabolics and economics

Experimental Data Metabolic Model

Technoeconomic

Model


Network Reactions and Genes Stoichiometric matrix


Metabolic Network Modeling Methods

• Constraint-based reconstruction and analysis (COBRA) approach is being used.

• COBRA integrates metabolic network stoichiometry into a system of linear equations.

• The COBRA Toolbox provides a MATLAB-based simulation environment for predicting

optimal distributions of mass fluxes within the metabolic network that maximize an

objective function (typically, cell growth and/or specific metabolite production).

• The stoichiometric matrix and constraints on nutrients and excreted metabolites (from

experimental data) determine the flux predictions (such as for lipid accumulation).

• Becker SA, Feist AM, Mo ML, Hannum G, Palsson BØ, Herrgard MJ. Nature Protocols 2007;2(3):727-38.

• Schellenberger J, Que R, et al., Nature Protocols. 2011 Aug 4;6(9):1290-307.

Reactions

Me

tab

oli

tes

atp[c] + D-glucose[c] -> h[c] + g6p[c] + adp[c]

(YALI0B22308g or YALI0E15488g or

YALI0E20207g)

g6p[c] <=> f6p[c]

YALI0F07711g

atp[c] + f6p[c] -> h[c] + fdp[c] + adp[c]

YALI0D16357g

fdp[c] <=> g3p[c] + dhap[c]

YALI0E26004g



Pathway Tools DB for metabolic network curation

L. starkeyi Metabolic Modeling Progress to Date

• Developed Pathway Tools database:

organizes L. starkeyi genes, enzymes, reactions, and metabolites for metabolic network curation

• Leveraged published models: Orthologous Markov Clustering matched L. starkeyi enzymes with enzymatic reactions within previously published COBRA models for the yeasts: Y. lipolytica, S. cerevisiae, S. stipitis, and P. pastoris

Next Steps: Finish integration of preliminary metabolic network and new experimental into full COBRA model for L. starkeyi to

• Predict optimal carbon yield

• Guide metabolic engineering

Key Project Outcome: Enable timely progress towards solving the complex problem of maximizing carbon yield to fuels


Task 3 Pathways Analysis

• FY12 NREL and PNNL

collaboration to perform analysis

for hydrocarbon biofuels pathways

• Integrated efforts with INL for

feedstock interface and costs

• Vetted assumptions and basis with

researchers and scientist for

pathways

• DOE BETO selected 2 sugars

based pathways for further

development; published as Tech

Memos



Milestones/Metrics and Progress

Title/Description Due Date Completed

Complete Screening for co-product routes that meet the 2012 Reference Case

target of $2/gal of Ethanol

31 Dec 2011

Completed model for fungal base co-products production with goals to meet the

2012 target of $2/gal Ethanol

31 Mar 2012

Complete model for hydrocarbon production via biochemical route to assess

potential for meeting the 2017 target of $3/gal gasoline

30 Jun 2012

Complete sensitivity analysis for both cases to identify research needs which

meet reference case targets

30 Sep 2012

Complete the generation of a pathway-genome database for L. starkeyi,

comparative analysis with other members of the Saccharomycotina fungi, and

draft metabolic network reconstruction

31 Dec 2012

Complete refinement of L. starkeyi central metabolism pathways 31 Mar 2013

Complete refinement of L. starkeyi fatty acid synthesis pathways 30 Jun 2013 On schedule

Complete preliminary TEA for L. starkeyi based process and improvements

leading to $3/gal gasoline equivalent

30 Sep 2013 On schedule

Complete analysis reports and models, create draft manuscript 30 Sep 2013 On schedule


Relevance

• Directly supports the Bioenergy Technologies Office development of new pathways and technical targets for the 2017 to 2022 timeframe

• Helps DOE and industry understand critical issues related to hydrocarbon production

• Enables the Biochemical Conversion Area’s strategic goal “to develop commercially viable technologies for converting biomass feedstocks via biochemical routes into energy dense, fungible liquid transportation fuels, as well as bioproducts or chemical intermediates, and bioenergy.”


Critical Success Factors

Hydrocarbon fuel production from biological processes requires a combination of tools to successfully direct and accelerate future research and ultimately reduce costs

To be successful, our metabolic model for the

oleaginous yeast L. starkeyi must be able to:

– Integrate the best available data to predict the

maximum metabolically feasible lipid yield from the

wild-type strain

– Identify targets for modifying the metabolism of the

wild type strain that raise the lipid yield nearer to the

theoretical maximum of 33%, thus reducing the cost

of cellulosic diesel towards the BETO goal of

$3/gallon gasoline equivalent


Future Work

Overall:

• Continue collaboration with experimentalists to refine metabolic and economic

models and model outcomes

• Develop understanding of yield limits on glucose and eventually, complex

substrates

• Develop strategies to control unwanted side reactions (e.g. secreted

exopolysaccharides that increase broth viscosity)

ML, DL or

Go/No Go

Description FY13

Q3

FY13

Q4

FY14

Q1

FY14

Q2

FY14

Q3

FY14

Q4

M.1.ML.3

Theoretical Yield Estimates & Economics: Refine fatty acid

synthesis pathways

M.1.ML.4

Theoretical Yield Estimates & Economics: Complete updated

TEA for L starkeyi

M.1.DL.1 Theoretical Yield Estimates & Economics: Final Report

N.1.ML.1

Theoretical Yield Estimates & Economics: exercise models to

direct next research steps

N.1.ML.2

Theoretical Yield Estimates & Economics: assess experimental

results for model use

N.1.ML.3 Theoretical Yield Estimates & Economics: refine models

N.1.DL.4 Theoretical Yield Estimates & Economics: Final Report

N.1.GN.1 Go-no go decision


Summary

Approach: Combine multiple strengths: economics, metabolics, experimental

Technical accomplishments:

• Completed economic analysis of a co-product scheme

• Completed literature based process model for hydrocarbon production

• Developing metabolic model for L. starkeyi

• Published (with NREL) 2 Tech Memos for sugars conversion

Relevance: Supports BETO goal of economic hydrocarbon fuels from sugar

Critical Success factors and challenges:

• Obtaining sufficient experimental data for metabolic & economic

models

• Deployment of metabolic & economic model to influence research

directions

Future Work:

• Leverage L. starkeyi experimental work to improve models

• Exercise metabolic and process models to suggest research directions

Technology transfer: Publication of results


Acknowledgements

Bioenergy Technologies Office – Leslie Pezzullo, Joyce Yang, Bryna

Berendzen, Katy Christiansen

PNNL Team – Aye Meyer, Iva Tews, Jim Collett

NREL – Mary Biddy


Additional Slides


Responses to Previous Reviewers’ Comments

2011 Reviewer Comments • The analysis provides a useful decision making and analysis tool to assist researchers. For the long term planning [2012-

2017] the assumptions the models are based upon a most likely highly speculative since little experimental information exists

on many of these. Rigid adherence to these could lead to the adoption of wrong research paths to pursue as well as

elimination of potentially promising avenues of investigation.

• This type of project reminds me of the chicken or egg controversy - of which came first. The project seems necessary in

order to determine which avenues are feasible and worth further consideration, i.e. funding to generate more data. On the

other hand, it is difficult to make realistic determinations of "feasibility" when the data available on a given topic is

sparse. Hence, I see the project as important but the focus must be on systems/processes for which there is considerable

data upon which to make judgments

• The researcher may want to avail them self of work done in other institutions such as Stanford Research Institute which may

be very beneficial. With this I believe the bioinformatics would be more firmly based.

Response to review comments The reviewers generally recognize the need to establish some criteria for determining research goals and measuring progress

towards goals, but regretted the lack of experimental data and reliance upon sparse literature information. To remedy this, the

economic analysis for hydrocarbon production is beginning to be supplemented with the development of metabolic modeling and

experimental results aimed at informing both the economic and the metabolic models.

As per the reviewers’ suggestion, we have employed the Pathway Tools bioinformatics software package from SRI (Stanford

Research Institute) to integrate genomic sequence data and biochemical literature information into highly curated metabolic

pathways. We are assembling these pathways into a process model of the stoichiometric reaction network within the oleaginous

yeast Lipomyces starkeyi to inform techno-economic analysis and to identify metabolic engineering strategies that will reduce the

cost of cellulosic hydrocarbon fuels toward the BETO goal of $3/gallon.


Publications, Presentations, and Commercialization

Publications: FY13: PA Meyer, IJ Tews, JK Magnuson, SA Karagiosis and SB Jones, “Techno-economic analysis of corn stover

fungal fermentation to ethanol”. Accepted for publication by Applied Energy

FY13: Biological Conversion of Sugars to Hydrocarbons Technology Pathway

http://www.nrel.gov/docs/fy13osti/58054.pdf

http://www.pnl.gov/main/publications/external/technical_reports/PNNL-22318.pdf

FY13: Catalytic Upgrading of Sugars to Hydrocarbons Technology Pathway



Presentations: FY11: A poster entitled “Novel Microbial Conversion of Biomass to Hydrocarbon-Compatible Fuels: A Survey” was

presented at the 33rd Symposium on Biotechnology for Fuels and Chemicals in early May.

FY12: Iva Tews presented a summary of the PNNL analysis work on Biochemical Pathways to Fuels at the Recent

Advances in Fermentation Technology (RAFT IX) Conference entitled “A Techno-economic Assessment of Cellulosic

Bio-Conversion Routes: Ethanol and Beyond.” November 2011

FY12: Pimphan (Aye) Meyer ACS National Spring Meeting March 2012, San Diego CA. "A Techno-economic Analysis

for Ethanol Production from Corn Stover by Fungal Fermentation".









http://sim.confex.com/sim/raft9/webprogram/Paper20297.html






1 | Bioenergy Technologies Office

Documents

Transcript of 1 | Bioenergy Technologies Office