November 19, 2003Mark F. Ruth and John L. Jechura

National Renewable Energy Laboratory

Incorporating Monte Carlo AnalysisInto Techno-Economic Assessment

of Corn Stover to Ethanol

Presented at the 2003 AIChE Annual Meeting, November 20, 2003, San Francisco, CaliforniaNREL/PR-510-35313

Corn Stover to Ethanol Process

Feedstock Handling

CO2

EthanolEnzyme

Corn Stover

LigninResidue Steam

Electricity

Steam & Acid Liquor

Pretreatment S/L Separation

ConditioningSaccharification&

FermentationDistillation &Ethanol Purification

WastewaterTreatment

Burner/BoilerTurbogenerator

Lime

Steam

Gypsum

Process Analysis Technique

ConceptualProcess Design

Material & Energy Balance

Capital & Project Cost

Estimates

EconomicAnalysis

EnvironmentalAnalysis

R&D

ASPENPlus™ Process Simulation Engine (Input Language)

63 Components

164 Unit Operation Blocks

82 Control Blocks

Microsoft Excel™

Process Analysis Technique

ConceptualProcess Design

Material & Energy Balance

Capital & Project Cost

Estimates

EconomicAnalysis

EnvironmentalAnalysis

R&D

ASPENPlus™ Process Simulation Engine (Input Language)

63 Components

164 Unit Operation Blocks

82 Control Blocks

Microsoft Excel™

Batch ASPEN file:

Runs faster than GUI

Easier to search and modify for a small group of engineers

Technoeconomic Model

In 2002, NREL published an updated target-case design report– Greenfield corn stover to ethanol process– NREL/TP-510-32438– www.nrel.gov/docs/fy02osti/32438.pdf

Minimum Ethanol Selling Price ($ per gallon ethanol) is the primary result

Design Case Economic ResultsPlant Size: 2200 tons (2000 MT) Dry Corn Stover/Day (Greenfield Site)

Corn Stover Cost: $30/dry ton

Economic Parameter (Units, $2001)

Value

Minimum Ethanol Selling Price ($/gal)

$1.07

Ethanol Production (MM gal/yr) 69

Ethanol Yield (gal/dry ton stover) 90

Total Project Investment ($ MM) $197

TPI per Annual Gallon ($/gal) $2.86

Production Cost ($/gal) $0.58

Monte Carlo Analysis

Uses random numbers within defined probability functions to predict the uncertainty of modeled systems– Packaged software (e.g., Crystal Ball)

makes it easier with Excel

Used in the environmental, safety, business and other fields

Use of ASPENPlus™

EXCEL

&

VBA .sum File

ASPEN

Engineer Enters

Parameters

Engineer starts

transfer routine

Initial Monte Carlo Technique

Crystal Ball

EXCEL

&

VBA

.sum File

ASPEN

Engineer Selects

Parameter Function

Engineer Transfers

Parameters

Engineer starts

transfer routine

Monte Carlo Technique

Crystal Ball

EXCEL

&

VBA

Text File

.sum File

ASPEN

Engineer Selects

Parameter Function

VBA routinewrites text file

ASPEN Fortran blockreads text file

Monte Carlo Technique

Crystal Ball

EXCEL

&

VBA

Text File

.sum File

DOS BatchFile

ASPEN

Engineer Selects

Parameter Function

VBA Routinelaunches batch file

DOS batch filelaunches ASPEN

Monte Carlo Technique

Crystal Ball

EXCEL

&

VBA

Text File

.sum File

DOS BatchFile

ASPEN

Engineer Selects

Parameter Function

VBA routine “senses”

completion of ASPEN and reads

“.sum” file

Monte Carlo Technique

Crystal Ball

EXCEL

&

VBA

Text File

.sum File

DOS BatchFile

ASPEN

Engineer Selects

Parameter Function

Probability Distributions

Feedstock variability– Measured data

Yields– Estimated probability functions

Corn Stover Collection Locations

xx

xx

xx

x

xx

xxx

xxxx x

x

x

x

x

x

xxx x

xx xx

xx

xxxx

x

xxx xxx

xxx

xx

x

xx

xx

x

x

738 Samples

Stover Composition Ranges

Structural Glucan Xylan Lignin Protein Acetyl Uronic

Acids

Structural Inorganics

(Silica, Ash)Soil

SolubleSolids

Total

Minimum(% dry wt.) 27.9 14.5 11.5 1.3 0.9 1.4 -1.2 0.9 2.0 90.0

Maximum(% dry wt.) 39.6 25.5 20.4 7.0 3.9 3.9 10.2 1.7 19.6 101.9

Span (% dry wt.) 11.7 11.0 8.9 5.7 3.0 2.5 11.3 0.8 17.5 11.9

Mean(% dry wt.) 33.8 20.0 15.8 3.6 2.7 2.9 4.2 1.3 8.2 97.4

Standard Deviation

(% dry wt.)2.0 1.6 1.4 0.7 0.5 0.3 1.6 0.1 2.2 1.7

738 Samples

Feedstock Probability Functions

Structural glucanOther structural sugarsXylan fraction of other structural sugarsLignin

Constant ratios– Galactan/mannan/arabinan– Ash/acetate/protein/soluble solids

General Feedstock Composition

35%

26%

20%

15%4%

Structural GlucanOther Structural SugarsLigninRemainingSoluble Solids

Component Breakdown

21%2%3%1%

20%

5%2% 3% 5% 4%

35%

Structural GlucanXylanGalactanArabinanMannanLigninAshAcetateProteinSolsolidsExtractives

Feedstock Probability Distributions

Yield Probability Distributions

Monte Carlo Analysis Results

Histogram of MESPs for 1195 Monte Carlo Simulation Runs

0

10

20

30

40

50

60

70

80

90

100

$0.96

$0.98

$1.00

$1.02

$1.04

$1.06

$1.08

$1.10

$1.12

$1.14

$1.16

$1.18

$1.20

$1.22

$1.24

$1.26

$1.28

$1.30

$1.32

$1.34

$1.36

$1.38

MESP ($/gal)

Num

ber o

f Occ

uran

ces Mean: $1.15/gal

Monte Carlo Analysis Results

Cumulative MESPs for 1195 Monte Carlo Simulation Runs

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

$0.96

$0.98

$1.00

$1.02

$1.04

$1.06

$1.08

$1.10

$1.12

$1.14

$1.16

$1.18

$1.20

$1.22

$1.24

$1.26

$1.28

$1.30

$1.32

$1.34

$1.36

$1.38

MESP ($/gal)

Perc

entil

e

Mean: $1.15/gal

90% Confidence Interval:$1.06-$1.25/gal

Conclusions

Monte Carlo is useful for confidence interval estimates– A technique has been developed to use

ASPEN batch techniques and Excel– Reams of data improve function definition

but estimates can be useful – This analysis gave a 90% confidence

interval of $1.06-$1.25/galMonte Carlo can also be used to estimate design constraints

Acknowledgement

This work supported by the Office of the Biomass Program of the

U.S. Department of Energy

Office of Energy Efficiency and Renewable Energy