modeling new fuel combustion...combustion combustion creck modeling group kaust future fuels...

Modeling new fuel combustion

Tiziano FaravelliCRECK modeling Group

http://creckmodeling.chem.polimi.it/

KAUST Future Fuels Workshop

March 7-9th 2016

http://creckmodeling.chem.polimi.it/


Outlines

Hydrocarbon combustion

Biofuels and their impact on combustion

oxygen atom

changing bond energies

introducing new reaction classes

double bonds

changing bond energies

introducing new reaction classes

losing molecule symmetry: increasing mechanism complexity

lumping

reduction

numerical methods

Next challenges

new biofuels

soot

Conclusion

CRECK modeling Group


Combustion

Pyrolysis:C-C bond cleavage

Oxidation:from C-H bondsto C-O and H-O bonds

nC7H16

CO2

H2O

O2

HEAT

COMBUSTION

COMBUSTION



Combustion complexity

Low temperature High temperature

+ O2 Oxidation

Pyrolysis

Pyrolysis

+ O2 Oxidation



High TemperatureMechanism(Eapp 30000 cal/mol)

Intermediate Temperature Mechanism(Eapp 19000 cal/mol)

+ O2

OH• + Cyclic Ethers

OH• + •RCHO + CnH2n

HO2• + R=

b-Decomposition Products

RH

R

+ O2

ROO

QOOH

+ O2

OOQOOH

OQOOH + OH•

Oxidation of alkanes

DegenerateBranching Path


NTC

con

vers

ion

Reactor Temperature


New Biofuels

Oxygenated molecules

Syngas

Biooil

Alcohols (methanol, ethanol, propanol, butanol)

Ketones (acetone, EMK, DEK)

Ethers (DME, DEE, EME, MTBE, ETBE)

Esters (methyl and ethyl esters)

Furans (methyl furan, di-methyl furan)



Oxygen atom effect on the closest bonds

99.7(0.0)

97.2(-2.5)

86.587.5

92.0(-7.7)

97.2(-2.5)

98.6 86.291.9

102.7(+3.0)

93.0(-6.7)

98.2(-1.5)

92.083.986.8

87.5(-12.2)

89.2(-10.5)

82.280.8

94.5(-5.2)89.6

(-10.1)

99.9(+0.2)

82.481.8

82.4

81.8

84.284.8

93.1(-6.6)

98.2(-1.5)

94.1(-5.6)

n-butane n-butanol n-butanal

MEK methyl-butyl ether methyl-butanoate



Reactivity of oxygenated fuels

0.01

0.10

1.00

10.00

100.00

1000.00

0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7

Ign

itio

n D

ela

y Ti

me

[m

s]

1000/T [K]

fuel/air, F=1.0, 10 atm

n-butane n-butanol

n-butanal MB

MEK propane



Low T alcohol reactivity

OH

+O2

O

HO2+

OH OH OH OH

OH

OO

OHOO OH

OO

+O2 +O2 +O2

branching

99.7(0.0)

97.2(-2.5)

86.587.5

102.7(+3.0)

93.0(-6.7)

98.2(-1.5)

92.083.986.8



Low T ketone reactivity

O

O O O

O

O

O

O

O

O

O

O

O

O

O

O

methyl vinyl ketone tetrahydrofuran-3-one

HO2+ +OH

+O2+O2+O2

99.7(0.0)

97.2(-2.5)

86.587.5

94.5(-5.2)89.6

(-10.1)

99.9(+0.2)

82.481.8

82.4


Hoppe et al., Fuel 2016Scheer et al., PCCP, 2014


Intermediate T ester reactivity

O

O

O

O

O

O

O

O

O

O

OO

O

O

OOH

O

O

OOH

O

O

O

O

+O2

HO2+

99.7(0.0)

97.2(-2.5)

86.587.5

92.0(-7.7)

97.2(-2.5)

98.6 86.291.9


methyl-crotonate


Intermediate T aldehyde reactivity

O

O O

+CO O+

80%

80%

20%

20%

60% 40%

Temperature [K]

Rat

e co

nst

ant

HO2 + C3H7CHO → H2O + C3H7CO

HO2+C3H8 → H2O + nC3H7

104

1010

109

108

107

106

105

500 600 700 800 900 1000

T = 850 K



Oxygen atom: new four center molecular reaction classes

OHH

+ H2O

Alcohol molecular elimination (dehydration)

OCH3

H

Ether molecular elimination

+ CH3OH

O

O

Ester molecular elimination

+ CO2

H3C

O

O

+ CH2O

H2C

H

O



Importance of dehydration molecular reactions

1-butanol 2-butanol

i-butanol tert-butanol

Frassoldati et al., Comb. & Flame 2011



Double bonds reduce reactivity at low temperatures

Resonantly allylic

stabilized radicals

Double bonds in the

unsaturated methyl esters

Lower reactivity of the

unsaturated methyl esters at

low temperature

O

OCH3CH3

C.K. Westbrook, W.J. Pitz, S.M. Sarathy, M. Mehl, Proc. Combust. Inst. (2012)

Methyl linolenate

C19H32O2

H

HH

H

H

HH

H

1.0E+02

1.0E+03

1.0E+04

1.0E+05

1.0E+06

0.8 1 1.2 1.4 1.6

Ign

itio

n D

ela

y Ti

me

[µ

s]

1000/T [1/K]

MSTEAMEOLEMLINOMLIN1

P = 13.5 atm; φ = 1

1.0E+02

1.0E+03

1.0E+04

1.0E+05

1.0E+06

0.8 1 1.2 1.4 1.6

Ign

itio

n D

ela

y Ti

me

[µ

s]

1000/T [1/K]

stearateoleatelinoleatelinolenate

P = 13.5 atm; φ = 1

1.0E+01

1.0E+02

1.0E+03

1.0E+04

1.0E+05

0.7 0.9 1.1 1.3 1.5

Ign

itio

n D

ela

y Ti

me

[µ

s]

1000/T [1/K]

MD expMDMPA MSTEA

P = 16 atm; φ = 1

1.0E+02

1.0E+03

1.0E+04

1.0E+05

1.0E+06

0.8 1 1.2 1.4 1.6

Ign

itio

n D

ela

y Ti

me

[µ

s]

1000/T [1/K]

MSTEAMEOLEMLINOMLIN1

P = 13.5 atm; φ = 1

1.0E+02

1.0E+03

1.0E+04

1.0E+05

1.0E+06

0.8 1 1.2 1.4 1.6

Ign

itio

n D

ela

y Ti

me

[µ

s]

1000/T [1/K]

stearateoleatelinoleatelinolenate

P = 13.5 atm; φ = 1

1.0E+01

1.0E+02

1.0E+03

1.0E+04

1.0E+05

0.7 0.9 1.1 1.3 1.5

Ign

itio

n D

ela

y Ti

me

[µ

s]

1000/T [1/K]

MD expMDMPA MSTEA

P = 16 atm; φ = 1



Double bonds new reaction classes: Waddington mechanism

methyl oleate - CH3-C18H33O2

O

OCH3CH3

O

OCH3CH3

OH

O

OCH3CH3

OH

OO

O

OCH3CH3

O

OOH

O

OCH3CH3

O

O

+OH

+O2

-OH

Nonanal + 9-oxo methyl-nanoate



Effect of Waddington mechanism

0.0E+00

1.0E-04

2.0E-04

3.0E-04

4.0E-04

5.0E-04

500 700 900 1100

Mo

le f

ract

ion

Temperature [K]

Methyl oleate

0.0E+00

1.0E-04

2.0E-04

3.0E-04

4.0E-04

5.0E-04

500 700 900 1100

Mo

le f

ract

ion

Temperature [K]

Methyl linoleate

Methyl esters/benzene oxidation in JSR (P = 106.7 kPa, t = 2s)

Rodriguez, A. et al., Comb & Flame, 2016



Double bonds new reaction classes: molecular decomposition

methyl linolenate O

O

O

O

O

O

+ butadiene + methyl-tetradeca-8,11-dienonate



Effect of molecular decomposition

with without

Rodriguez, A. et al., Comb & Flame, 2016

Methyl esters/benzene oxidation in JSR (P = 106.7 kPa, t = 2s)

0.0E+00

2.0E-05

4.0E-05

6.0E-05

8.0E-05

1.0E-04

500 700 900 1100

Mo

le f

ract

ion

Temperature [K]

C4H6

Methyl linoleate

Methyl stearate

Methyl oleate

Methyl stearate

Methyl oleate

Methyl linoleate

b)

0.0E+00

2.0E-05

4.0E-05

6.0E-05

8.0E-05

1.0E-04

500 700 900 1100

Mo

le f

ract

ion

Temperature [K]

C4H6

Methyl linoleate

Methyl stearate

Methyl oleate

Methyl stearate

Methyl oleate

Methyl linoleate

a)



Double bonds new reaction classes: internal cyclo-addition

O

OCH3CH3

O

OCH3CH3

methyl linolenate

CH3

O

OCH3

CH3

O

OCH3

CH3

O

OCH3

H abstraction reactions on the

favored allylic position

Formation of cyclic unsaturated

molecules (aromatic precursors)



Effect of internal cyclo-addition reactions

Exp data: P. Dagaut, personal communication

Stoichiometric oxidation of rapeseed methyl ester in JSR (P = 1 atm, t = 0.07 s)

0.0E+00

2.0E-05

4.0E-05

800 1000 1200 1400

Mo

le F

ract

ion

Temperature [K]

C6H6 with

without



Increase of number of species

OH

Number of primary radicals

2

5

butane C4H10

butanol C4H10O

Loss of simmetry

Carbon number Alkanes Alkenes

5 3 5

10 75 377

15 4374 36,564

20 366,319 4,224,993

25 36,797,588 536,113,477

Number of isomers

From: Galtier P., in Advances in Chemical Engineering, ed. G. Marin, 32, 2007



Biodiesel characterization of model compounds

Biodiesel is composed by saturated and unsaturated heavy methyl esters.

R

O

CH3O

Methyl esters

O

OCH3

CH3

O

OCH3CH3

O

OCH3CH3

O

OCH3CH3

The five major components are:

O

OCH3CH3

C.K. Westbrook, C.V. Naik, O. Herbinet, et al., Combust. Flame (2011)

“Detailed” kinetic scheme

methyl palmitate (MPA) – CH3-C16H31O2

methyl stearate (MSTEA) - CH3-C18H35O2

methyl oleate (MEOLE) - CH3-C18H33O2

methyl linoleate (MLINO) - CH3-C18H31O2

methyl linolenate (MLIN1) - CH3-C18H29O2



Mechanism dimensions

Adapted from: T.F. Lu, C.K. Law, Prog.

Energy Comb. Sci., 35 (2009)

biodiesel (POLIMI)

biodiesel (LLNL)

Biodiesel + NOx + soot (POLIMI)

computational cost associated with such mechanisms is usually very high

need of reduction methods, numerical techniques and computational tools to make:

-use of large kinetic schemes computationally efficient

-easy their integration in new and/or existing numerical codes

Lumping and reduction methods can result in effective approaches to face the problem



Handling mechanisms

RANS LES DNS

Accuracy

Size of kinetic mechanisms

Computational cost

Detailed mechanisms: not directly applicable in large-scale computations

3 objectives:

Set up a robust and efficient framework for ad hoc mechanism reduction.

Address skeletal reduction to customtargets, beyond reactivity and ignition delay

Obtain the optimal trade-off between sizeand accuracy

Lumping and Skeletal Reduction: more compact mechanisms with the same accuracy

Ranzi E. et al. (2014) International Journal of Chemical KineticsSeveral time scales involved

Fuel%

CO%

C16H10%

NO%

0.00#

0.20#

0.40#

0.60#

0.80#

1.00#

1.0E*09# 1.0E*07# 1.0E*05# 1.0E*03# 1.0E*01# 1.0E+01#

no

rmal

ize

d%m

ass%

frac

8o

n%[

:]%

t%[s]%

C3H8#

NO#

C16H10#

CO#

Isothermal PFR C2H4/air @ 1800 K



What next?

New fuels:

Heavier Alcohols (hexanols +) and aldehydes

Oxymethylenether CH3O-(CH2O)n-CH3

Bio-oils (characterization, surrogate definition and kinetics are open challenging problems)

New molecules from genetic modifications of microorganism

…

enough work for the coming years, despite the employment crisis (in Italy)

Funds?



Soot problem 1

Combustion of oxygenated biofuels can lead to a reduction of CO, UHC and particulate mass (PM) emissions compared to fossil fuels

(US EPA, Technical Report, 2002) (Lemaire & Therssen & Desgroux, 2010)

Engine-out emissions for

varying biofuel contentLocal soot volume fraction in a

burner for varying biofuel content



Soot formation from a simple biofuel: Ethanol

Exp data: A. Eveleigh et al. / Combustion and Flame 161 (2014) 2966–2974

0.5(54)+17

= 62%

54+17

Reduction in soot formation: C bond to oxygen does not contribute to soot formation.Assuming the same sooting yield of the C and C atoms forming C2 species, it is possible to calculate the different sooting tendency:

Expected: C = 50 % - C = 50 %Measured: C = 62 % - C = 38 %

Relevant role of the dehydration reaction, which allows C to be included in the soot.

CH3CH2OH

100 C atoms

4654

CH2=CH2 CH3CHO

CO

23

CH3CHCH

23

Soot18

17

CH4

6



Soot problem 2

Oxygenated fuels form less soot, but:

Number of emitted particles might not be reduced

Shift of particle size distribution towards smaller particles

Increase of ultrafine particles(<100 nm)

First indications that PM produced by biofuel combustion show an increased cytotoxicity (e.g. Gerlofs-Nijland 2013, Yanamala 2013)

(Fukagawa et al., 2013)

Shift towards ultrafine particles

with increasing biofuel content



Conclusion

Biofuels require new kinetic attention, because they:

include oxygen atom

present one or more double bonds

are complex molecules whose oxidation involves many species

have a different environmental impact not fully understood



Acknowledgment

This presentation is the result of the work of many colleagues and especially

students.

I have to acknowledge:

Eliseo Ranzi

Alberto Cuoci

Alessio Frassoldati

Matteo Pelucchi

Chiara Saggese

Alessandro Stagni



THANK YOU FOR YOUR ATTENTION

modeling new fuel combustion...combustion combustion creck modeling group kaust future fuels...

Documents