modeling new fuel combustion...combustion combustion creck modeling group kaust future fuels...
TRANSCRIPT
Modeling new fuel combustion
Tiziano FaravelliCRECK modeling Group
http://creckmodeling.chem.polimi.it/
KAUST Future Fuels Workshop
March 7-9th 2016
KAUST Future Fuels Workshop
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
KAUST Future Fuels Workshop
Combustion
Pyrolysis:C-C bond cleavage
Oxidation:from C-H bondsto C-O and H-O bonds
nC7H16
CO2
H2O
O2
HEAT
COMBUSTION
COMBUSTION
CRECK modeling Group
KAUST Future Fuels Workshop
Combustion complexity
Low temperature High temperature
+ O2 Oxidation
Pyrolysis
Pyrolysis
+ O2 Oxidation
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
NTC
con
vers
ion
Reactor Temperature
KAUST Future Fuels Workshop
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)
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
Hoppe et al., Fuel 2016Scheer et al., PCCP, 2014
KAUST Future Fuels Workshop
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
CRECK modeling Group
methyl-crotonate
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
Importance of dehydration molecular reactions
1-butanol 2-butanol
i-butanol tert-butanol
Frassoldati et al., Comb. & Flame 2011
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
Double bonds new reaction classes: molecular decomposition
methyl linolenate O
O
O
O
O
O
+ butadiene + methyl-tetradeca-8,11-dienonate
CRECK modeling Group
KAUST Future Fuels Workshop
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)
CRECK modeling Group
KAUST Future Fuels Workshop
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)
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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?
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group
KAUST Future Fuels Workshop
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
CRECK modeling Group