Frederick L. Dryer

Department of Mechanical and Aerospace Engineering Princeton University

First Annual Conference of Combustion Energy Frontier Research Center (CEFRC)September 23–24, 2010, Princeton, New Jersey

1

Fuels Combustion Research Laboratory’s contribution to CEFRC effortsGeneration of motivating fundamental experimental observations that elucidate the interactions of bio-derived molecular components and structures with components found in petroleum derived fuels.Provision of experimental validation data for development and refinement of detailed kinetic models for esters and alcohols.Experimental determination of elementary kinetic rates important to the decomposition and oxidation of hydrocarbons and hydrocarbon oxygenates.

AccomplishmentsChemical Kinetics of iso-Propanol and t-Butanol Pyrolysis and Oxidation

Auto Ignition of n-heptane/Butanol Blends in Ignition Quality Tester

Methyl Formate Decomposition, and Oxidation in Low Pressure Flames

An Updated Kinetic/Transport Model for H2/O2

2

} Bio-derived fuels

} Sub-model improvements

3

Z. Serinyel1, J. S. Heyne2, S. Jahangirian2, F.L. Dryer2

1Department of Chemistry, National University of Ireland, Galway:Collaboration with H. J. Curran (NUIG)

2Department of Mechanical and Aerospace Engineering, Princeton University

Flow reactor study on iso-propanol and tert-butanol

MotivationLower soot / polyaromatic hydrocarbon emissions.Few studies on next generation, higher molecular weight oxygenated fuels Experimental data needed for extensive model validation.Experimental determination of the dehydration rate constants forn/i-propanol and tert-butanol.Detailed speciation data for alcohol oxidation obtained at T<950 K.Detailed mechanistic study is needed to understand the combustion of these alcohols and interactions with two stage kinetics of large carbon number hydrocarbons.

4

5

Experiment T (K) P (atm) Molar conc’n(ppm)

Pyrolysis History 800 –1000 12.5 5000

Pyrolysis Reactivity (τ = 1.8 s) 550 –1025 12.5 5000

Oxidation Reactivity (τ = 1.8 s, φ=1) 520 – 1010 12.5 2500Pyrolysis with radical trapper (1,3,5-TMB) 900 - 1000 12.5 1200 equimolar

Experiment T (K) P (atm) Molar conc’n(ppm)

Pyrolysis History 720–1000 12.5 2500

Pyrolysis Reactivity (τ = 1.8 s) 500 – 975 12.5 2500

Oxidation Reactivity (τ = 1.8 s, φ=1) 676– 918 12.5 2500Pyrolysis with radical trapper (1,3,5-TMB) 1020, 1080 3.0 1540 equimolar

Iso-propanol

Tert-butanol

Level to which water and olefin yield in equal quantities gives evidence of radical trapper methodology. Rate of dehydration reaction extracted experimentallyiso-propanol: i-C3H7OH ↔ C3H6 + H2Otert-butanol: t-C4H9OH ↔ C4H8 + H2ORadical trapper addition to extract water elimination rate constant

6

0 1 2 30

400

800

1200

M

olar

con

cent

ratio

n (p

pm)

1,3,5-TMB i-PrOH H2O C3H6

t (s)0 1 2 3

0

400

800

1200

1,3,5-TMB i-PrOH H2O C3H6

Mol

ar c

once

ntra

tion

(ppm

)

t (s)

Iso-propanol pyrolysis with radical trapper:

1200 ppm i-C3H7OH1200 ppm 1,3,5-TMBP = 12.5 atm

Oxidation reactivity profiles (no reactivity at T < 775K)

500 600 700 800 900 10000

1000

2000

3000

4000

5000

Mol

ar c

once

ntra

tion

(ppm

)

T (K)

iso-PrOH O2*0.2 CO CO2

H2O*0.2

700 800 900 10000

800

1600

2400

3200

4000

4800 tert butanol H

2O/4

iC4H8

CO/5 CO2/2 O

2/10

Mol

ar C

onc.

(ppm

)T (K)

Oxidative conditions:P = 12.5 atm

iso-propanol t-butanol

F. M. Haas, A. Ramcharan, F.L. Dryer

Department of Mechanical and Aerospace EngineeringPrinceton University

7

Motivation Isomers of butanol have been identified as promising candidates of 2nd generation bio fuels.Little fundamental work available on how these alcohols interactwith hydrocarbons found in conventional petroleum derived fuels-likely end use scenario for these bio generated materials.

8

ObjectivesScan ignition behavior trends in isomeric structures.

Engine-relevant pressure, temperature and fuel loading.Identify “interesting” behaviors to focus more fundamental studies.

Examine kinetic coupling of components in fuel blends.

Experiments: A study of comparative ignition qualityTrends in reactivity extendable to conventional HCSI, DICI engine applications as well as high-efficiency PCCI future designs.Comparisons among butanols at similar volumetric blending in n-heptane and a real low-cetane diesel fuel.

9

Pure Component Avg. Measured DCN MON

t-butanol <7.18 94

i-butanol 8.51 94

2-butanol 8.54 91

n-butanol 12.02 78

n-heptane 53.8† 0†

IncreasingReactivity

1-butanol (+/-)2-butanol i-butanol t-butanol(TBA)

CH3 OH CH3CH3

OHCH3

OH CH3

CH3

CH3

OHα-hydrogen C-H bond sites indicated

10

Increasing alcohol content Increasing alcohol contentDCN sensitivity of -0.33 DCN/vol%

t-butanol shows significantly different behavior Blended with diesel

fuel (solid lines)Blended with n-heptane

11

Radical pool generation relatively slow in pure TBAUnimolecular processes slow at IQT conditions, mostly abstraction reactions governing ignition.Low temperature O2 addition reactions evidently less important (Cullis & Warwicker) than for large carbon number hydrocarbons.β hydrogen atoms less prone to abstraction than α atoms

BDEs of ~101 vs. 93-95 kcal/mol.Subsequent radical decomposition leads to straight-chain reactions and stable products.

X + TBA → XH + •CH2(C)(CH3)2OH (abstraction)•CH2(C)(CH3)2OH → i-butene + OH (β-scission)

for X=OH, product is H2O

Key difference between TBA & other BuOHs is selectivity of OH generation process: other BuOHs will mostly produce less reactive HO2radical.

S. Dooley1, T.I. Farouk1, F.L. Dryer1, B. Yang2, T. Kasper2, N. Hansen2

, T.A. Cool3

1Department of Mechanical and Aerospace Engineering, Princeton University2Combustion Research Facility, Sandia National Laboratories

3School of Applied and Engineering Physics, Cornell University

12

Motivation – (1) Diesel CombustionMethyl formate (MF) is major intermediate in Dimethoxy methane (DMM) oxidation and dimethyl ether (DME) oxidation in presence of NOx.Understanding chemical mechanism of MF oxidation => refine understanding of DMM/DME oxidation.

Motivation – (2) Biodiesel CombustionBiodiesel consists primarily of long alkyl chain methyl esters.Role of ester functionality on biodiesel oxidation remains to be fully understood quantitatively.Study of smaller carbon number esters can lead to better understanding of functional behavior in large carbon number species.

Methyl formate is simplest methyl ester.

13

Dooley et al. Int. J. Chem. Kin. 2010.Validation - High temperature shock tube ignition delays.

- Laminar burning velocities.

14

MF shock tube ignition delay times. Experiment (symbols) and simulation (lines).

0.6 0.8 1.0 1.2 1.4 1.616

20

24

28

32

36

40

Lam

inar

bur

ning

vel

ocity

, S0 u /

cm s

-1

Equivalence ratio in synthetic air

Laminar burning velocities of MF/O2/N2 freely propagating flames at 1 atm. Experiment (symbols) and simulation (line).

0.5 0.6 0.7 0.8 0.9

10

100

1000

0.5% MF 1.0% O2 2.7 atm 0.5% MF 0.5% O2 2.7 atm 0.5% MF 2.0% O2 5.4 atm 2.5% MF 5.0% O2 9.4 atm

Igni

tion

dela

y tim

e, τ

/ μs

103 K / T

2000 1800 1600 1400 1200 Temperature / K

Flow reactor speciation data measure methanol, formaldehyde and methane from MF oxidation.

Pyrolysis measurements showlarge amounts of methanol andcarbon monoxide, indicative of molecular decomposition of methyl formate.

Kinetics of MF decompositionhighly contentious, reported activation energies span48‐77 kcal mol‐1(!)

0.0 0.1 0.2 0.3 0.4 0.5 0.6

0.0000

0.0002

0.0004

0.0006

0.0008 CH3OH

CH2O

CH4

Spe

cies

mol

e fra

ctio

n

Time / seconds

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

0.000

0.002

0.004

0.006

0.008

0.010

0.012 MF O2 CO CO2 H2O

Spec

ies

mol

e fra

ctio

n

Time / seconds

Speciation data from flow reactor oxidation of 0.005/0.01/0.985 MF/O2/Ar, φ=1.0, at 3.0 atm and 975K (top) and 900 K (bottom). Experiment (symbols) and simulation

(lines). Rate constant data for MF decomposition.

15

0.0 0.1 0.2 0.3 0.4 0.5 0.610

100

10003600

3800

4000

4200

4400

46005000 ppm MF at 903K & 3 atm

Methyl formate Carbon monoxide Methanol

Spe

cies

con

cent

ratio

n / p

pm

Time / seconds

Diffuser section

0.6 0.8 1.0 1.2 1.410-6

10-5

10-4

10-3

10-2

10-1

100

101

102

103

104

105

106

107

108

k / s

-1

1000K/T

Dooley et al.1

Metcalfe, Simmie & Curran4

Westbrook et al.5

kMethyl formate

kMethanol

kCarbon monoxide

1800 1550 1300 1050 800

Temperature / K

16

Calculated concentration of CH3OH* in the reactor environment at 975 K, 3 atm, 14.344 g/s flow rate of 5000 ppm of MF in N2.

Axial profile of MF and gas phase (methanol) vs potential heterogeneous catalytic (methanol*) products for 5000 ppm MF in N2 at 3 atm at the center line (lines and symbols represent calculations and experiments, CO omitted for clarity)

900 K 975 K

Potential effect of wall catalytic reactions evaluated by conducting LES of the entire VPFR geometry.

Products of wall catalytic processes confined in the viscous sub-layer region.

Heterogeneous surface perturbations cannot significantly affect the measurement of the gas phase reaction (of negligible quantity at the center line less than 4%)

CH3OCHO(surface) ↔ CH3OH*(surface) + CO*

(surface)

17

0 5 10 15 20 25 300.0

0.1

0.2

0.3

0.4

0.5 Ar CO2 H2O O2 H2 CO MF

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

0 1 2 3 4 5 6 7

0.000

0.005

0.010

0.015

0.020

0.025 C2H4 CH3OH CH4 CH2O CH3 C2H2

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

0 1 2 3 4 5 6 70.000

0.005

0.010

0.015

0.020

0.025 C2H4 CH3OH CH4 CH2O CH3 C2H2

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

0 5 10 15 20 25 300.0

0.1

0.2

0.3

0.4

0.5 Ar CO2 H2O O2 H2 CO MF

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

0 5 10 15 20 25 300.0

0.1

0.2

0.3

0.4

0.5 Ar CO2 H2O O2 H2 CO MF

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

0 1 2 3 4 5 6 70.000

0.005

0.010

0.015

0.020

0.025 C2H4 CH3OH CH4 CH2O CH3 C2H2

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

0 5 10 15 20 25 300.0

0.1

0.2

0.3

0.4

0.5 Ar CO2 H2O O2 H2 CO MF

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

0 1 2 3 4 5 6 7

0.000

0.005

0.010

0.015

0.020

0.025 C2H4 CH3OH CH4 CH2O CH3 C2H2

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

Φ = 1.0 Φ = 1.2 Φ = 1.4 Φ = 1.6 Φ = 1.8

0 5 10 15 20 25 300.0

0.1

0.2

0.3

0.4

0.5 Ar CO2 H2O O2 H2 CO MF

Spe

cies

mol

e fra

ctio

n

Distance from burner surface / mm

0 1 2 3 4 5 6 70.000

0.005

0.010

0.015

0.020

0.025 C2H4 CH3OH CH4 CH2O CH3 C2H2

Spec

ies

mol

e fra

ctio

n

Distance from burner surface / mm

Model captures experimental measurements well. Manuscript in review with further details and analyses.

Experimental data (symbols) and simulation (lines) of MF oxidation in MF/O2/Ar , Φ = 1.0-1.8, 22-30 Torr flames

M. P. Burke1, M. Chaos2, Y. Ju1, F. L. Dryer1, S. Klippenstein3

1Department of Mechanical and Aerospace Engineering, Princeton University2 F. M. Global

3Argonne National Laboratory

18

19

H2/O2 kinetic sub-model - Fundamental topic w/ applications:Hierarchical development of all HC kinetic models.Accurate kinetic/transport models needed for syngas and high hydrogen content fuels to be used in IGCC applications.

High pressure, low flame temperature, fuel-lean and/or diluted combustion conditions used in most applications (to control NOx emissions).

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0 5 10 15 20 25 30Pressure (atm)

Mas

s bu

rnin

g ra

te (g

cm

-2 s

-1)

Present experimentsTse et al. (2000)Konnov (2008)Li et al. (2007)Saxena & Williams (2006)Sun et al. (2007)O'Connaire et al. (2004)Davis et al. (2005)GRI-MECH 3.0 (1999)

H2/O2/He, φ=0.85Tf ~1600K

0.00

0.20

0.40

0.60

0.80

1.00

1.20

0 5 10 15 20 25 30Pressure (atm)

Mas

s bu

rnin

g ra

te (g

cm

-2 s

-1)

Present experimentsLi et al. (2007)O'Connaire et al. (2004)Saxena & Williams (2006)Davis et al. (2005)Konnov (2008)Sun et al. (2007)GRI-MECH 3.0 (1999)

H2/O2/Ar, φ=2.5Tf ~1600K

0.000

0.020

0.040

0.060

0.080

0.100

0 5 10 15 20Pressure (atm)

Mas

s bu

rnin

g ra

te (g

cm

-2 s

-1)

USC-MECH IIUSC-MECH II with Davis H2 SubsetUSC-MECH II with GRI 3.0 H2 Subset

C2H2/air, φ=0.43Tf ~ 1600K

Substantial H2/O2 modeling difficulties in high-pressure dilute flames

H2/O2 modeling difficulties affect high-pressure HC

flame predictions

M.P. Burke, M. Chaos, F.L. Dryer, Y. Ju, Combustion and Flame 157 (2010) 618-631.

M.P. Burke, F.L. Dryer, Y. Ju, Proceedings of the Combustion Institute (2010) in press.

20

‣ Updated model uses the 19-reaction mechanism of Li et al.To incorporate recent studies on elementary reactions.To improve model performance at high-pressure flame conditions.

‣ Treatment of the following processes/reactions were revised:Temperature and pressure dependence of HO2 formation / consumption reactions:• H+O2(+M)=HO2(+M) • HO2 + H/O/OH/HO2

Other reactions updated:• e.g. H2O2(+M)=OH+OH(+M), OH+H2O2=H2O+HO2, H+OH+M=H2O+M.Transport updated:• e.g. H, H2.

‣ Major sources of remaining uncertainties:Temperature (and pressure) dependence of HO2+H/OH/O/HO2.Fall-off behavior and mixture rules for H+O2(+M)=HO2(+M).

M.P. Burke, Y. Ju, F.L. Dryer, S.J. Klippenstein, “An updated model and discussion of challenges for modeling the H2/O2 reaction mechanism in high-pressure flames,” in preparation for IJCK (2010).

21

Practical combustion=>multi-species bath gas.At best, models assume linear mixture rule:

At worst, common modeling approaches yield n times high-pressure limit =>butDifferent components of a multi-component bath gas “interact” in unimolecular reactionsNonlinearities stronger when:

<ΔE> difference is largerEach component has nearly equal contributions to rate constantReaction is in intermediate fall-off

Preliminary calculations and those of Troe (1980)show ~10% nonlinearities in low-p limit~10% mixing effects yield 15% differences in burning rate predictionsBehavior highly dependent on <ΔE> values

ii

i XPTkk ⋅= ∑ ),(

• M.P. Burke, Y. Ju, F.L. Dryer, S.J. Klippenstein, “An updated model and discussion of challenges for modeling the H2/O2 reaction mechanism in high-pressure flames,” in preparation for IJCK (2010).

1. J. Troe, Ber. Bunsenges. Phys. Chem. 84 (1980) 829-834.

0% 20% 40% 60% 80% 100%H2O mole fraction in H2O/Ar bath

Dev

iatio

n fro

mLi

near

Mix

ing

Rul

e increasing pressure for P < P1/2

0.00

0.01

0.02

0.03

0.04

0 5 10 15 20Pressure (atm)

Mas

s bu

rnin

g ra

te (g

cm

-2 s

-1)

w/o mixing effectsw/ mixing effects

H2/O2/He, φ=0.3Tf ~1400K

22

0 2 4 60

100

200

300

400

500

Equivalence ratio

Flam

e sp

eed

(cm

s-1

)

ExperimentsUpdated ModelLi et al.

0 0.05 0.1 0.15 0.2 0.25 0.30

0.2

0.4

0.6

0.8

1

Time (ms)

Mol

e fr

actio

n (%

)

H2O2H2OUpdated ModelLi et al.

1 1.5 20

0.1

0.2

0.3

0.4

Equivalence ratio

Mas

s bu

rnin

g ra

te (g

cm

-2 s

-1)

ExperimentsUpdated ModelLi et al.

0 10 20 300

0.5

1

1.5

2

Pressure (atm)M

ass

burn

ing

rate

(g c

m-2

s-1

)

ExperimentsUpdated ModelLi et al.

0.5 0.6 0.7 0.8 0.9 1

10-2

100

102

1000/T (K-1)

[O2] τ

ig (m

ol li

ter-1

μs)

ExperimentsUpdated ModelLi et al.

0 10 20 300.02

0.04

0.06

0.08

0.1

0.12

0.14

Pressure (atm)

Mas

s bu

rnin

g ra

te (g

cm

-2 s

-1)

ExperimentsUpdated ModelLi et al.

0 0.2 0.4 0.6 0.8 10.01

0.02

0.03

0.04

0.05

0.06

Equivalence ratio

Mas

s bu

rnin

g ra

te (g

cm

-2 s

-1)

ExperimentsUpdated ModelLi et al.

0.00

0.01

0.02

0.03

0.04

0.05

0 5 10 15 20Pressure (atm)

Mas

s bu

rnin

g ra

te (g

cm

-2 s

-1) Konnov

Li et al.Sun et al.USC-MECH IISaxena & WilliamsO'Connaire et al.Davis et al.Updated modelGRI-MECH 3.0

Figure 10.

Figure 11.

Figure 12.

Figure 13.

Figure 14.

Figure 15.

Figure 16.

Figure 17.

H2/O2/diluent1 atm

H2/O2/HeHe

Ar

N2

H2/O2/N23.44 atm, 975K

20 atm

10 atm

15 atm

H2/O2/Ar1 – 87 atm

H2/O2/Hef =0.85

Tf~1600K

H2/O2/Hef =2.5

Tf ~ 1700K

1600K

1500K

H2/O2/HeTf ~ 1400K

H2/O2/Hef =0.3, Tf ~ 1400K

1 atm

10 atm

5 atm

Reproduces previous validation targets of Mueller et al. and Li et al. including flow reactor speciation, ignition delays, and flame speeds.

Shows significant improvements against high-pressure, low-flame-temperature data.Predicts wide range of flame speed targets within 20%.

M.P. Burke, Y. Ju, F.L. Dryer, S.J. Klippenstein, “An updated model and discussion of challenges for modeling the H2/O2 reaction mechanism in high-pressure flames,” in preparation for IJCK (2010).

23

Continuing efforts to contribute elementary rate constant measurements and kinetic system validation data using flow reactor techniques.

Contributions to the small molecule chemistry that is important to high pressure oxidation kinetics of these materials.

Development of chemical kinetics model for bio-derived fuels.

24

Research Staff: Drs. Stephen Dooley, Saeed Jahangirian, and Tanvir Farouk

Graduate Students: Francis M. Haas, Joshua S. Heyne, Michael P. Burke, Zeynep Serinyel

Undergraduates: Amanda RamcharanTechnical Staff: Timothy Bennett, John Grieb, Joseph Sivo

CollaborationsDr. Steven Klippenstein (Argonne National Laboratory)Drs. Nils Hansen and Bin Yang (Sandia National

Laboratories)Dr. Terrance Cool (Cornell University)Dr. Henry Curran (National University of Ireland, Galway)