AIChE Annual Meeting, 11/11/2009

First principles based simulations of steam cracking of ethane/propane and ethane/butane mixtures M.K. Sabbe, K.M. Van Geem, M.-F. Reyniers, G.B. Marin

Laboratory for Chemical Technology http://www.lct.ugent.be

1

AIChE Annual Meeting, 11/11/2009

Steam cracking • • • •

Ethane, naphtha, …  ethene, propene, … Ethene production 113 106 tpa Cracking (850–1100K) in tubular reactor Energy consumption: 8% of total chemical industry primary energy use 2

C - C scission

+

+ +

H abstraction scission

+ 2

AIChE Annual Meeting, 11/11/2009

Objective Reactor simulation for hydrocarbon radical chemistry Reactor model

Reaction network

Solver

Kinetic and thermodynamic data

Develop a consistent data set based on ab initio calculations for the simulation of a steam cracker

?

> 1000 reactions using group additive models based on ab initio data Larger species Evaluate performance for: Co-cracking of Ethane/propane/butane mixtures

3

AIChE Annual Meeting, 11/11/2009

Outline Introduction Group additive method Thermodynamic and kinetic parameters Reactor simulation Conclusions

4

AIChE Annual Meeting, 11/11/2009

Group additivity: Benson Benson group

X1 C

X4

H C Cd Ct Cb C–(X1)(X2)(X3)(X4) X= C° Cd° Ct° Cb°

X2

X3

Group additivity for thermochemistry f

GAV f (group i ) i

Group additive values (GAV)

NNI j

f

o o o H , S , C S S f int p int

R ln(

j

nopt

)

Corrections for Non-nearest Neighbor Interactions (NNI)

… 5

AIChE Annual Meeting, 11/11/2009

Group additivity for kinetics ‡

Ea

Transition state for addition

H

1

fH

‡

n RT

fH

TS

reactants

GAVTS C , X ,Y , Z

‡

1

GAVreactants

n RT 1

‡

n RT

C , X ,Y , Z

GAV (Ci )

1

‡

n RT

C , X ,Y , Z

Ea ,ref

GAV (Ci ) C

Arrhenius equation

k

Ae

Ea RT

Group additivity for Arrhenius parameters 3

Ea (T )

GAVEoa (Ci )

Ea,ref (T ) i 1

log A(T )

Number of single events

~ log Aref (T )

3

GAV (Ci ) log ne i 1

ne

nopt,‡ nopt, j

j j ‡

j

Proposed by Saeys et al. for activation energies (AIChE J. 2004, 50 (2), 426-444. )

6

AIChE Annual Meeting, 11/11/2009

Group additivity for kinetics Example: radical addition (tert-butyl + iso-butene), 1000 K CH3

H3C

•

C3

CH3

H

H

H

CH3

C3

C2

C1 •

CH3

H

CH3

CH3 C2

+

CH3 H3C

C1 CH3

Ea= Ea,

Ref

CH3 + C2H4 43.1 kJmol-1

H3C

CH3 C3

H3C

CH3 C2

C1• CH3

+ ΔGAVo(C1) + ΔGAVo(C2) + ΔGAVo(C3) C1-(C)2

C2-(H)2

C3-(C)3

-3.1 kJmol-1

0 kJmol-1

-12.1 kJmol-1

Ea,GA = 27.9 kJmol-1 CH3 + C2H4 8.968

‡

C1-(C)2 +0.091

log(AGA /m3mol-1s-1)= 5.209

Ea, ab initio= 29.0 kJmol-1 C1-(H)2 +0

C3-(C)3 ne -1.151 +log2

logAab initio= 5.175 7

AIChE Annual Meeting, 11/11/2009

Outline Introduction Group additive method Thermodynamic and kinetic parameters

Reactor simulation Conclusions

8

AIChE Annual Meeting, 11/11/2009

Thermodynamics: initio based GAV Thermodynamics: group ab additivity CBS-QB3 + 1D-HR Number of GAVs determined Family Determined New NNI Alkanes 4 0 2 Alkenes 17 2 3 Alkynes 28 9 1 Aromatics 18 5 0 Radicals 34 25 2 Total 95 41 10

2-methylnonane

= 1

+ 6

+ 3

MAD between 14 GA predicted and CBS-QB3 calculated values (kJmol-1; J mol-1 K-1)

Hfo(298K) MAD

1.7

So(298K)

Cpo(300K)

Cpo(1000K)

2.9

2.2

0.5

Sabbe et al., J.Phys.Chem.A, 109:7466 (2005) Sabbe et al., J.Phys.Chem.A, 112:12235 (2008)

9

AIChE Annual Meeting, 11/11/2009

Kinetics: group additivity CBS-QB3 +1D HR forming/breaking bond + Eckart tunneling Family Number of GAV° (1000 K) C° addition/ -scission 46 1.6 H° addition/ -scission 31 1.6 H abstraction 26 + 4 resonan. correc. 1.8 Total 103 + 4 resonan. correc.

=

Sabbe et al. J.Phys.Chem.A 2007, 111 (34), 8416-8428 Vandeputte et al. J.Phys.Chem.A, 2007, 111 (46), 11771-11786. Sabbe et al. ChemPhysChem 2008, 9 (1), 124-140. Sabbe et al. ChemPhysChem, 2009, in press. Sabbe et al. PhysChemChemPhys, 2010, DOI: 10.1039/b919479g

Group additivity performance 10

1000 K 8

logkGA

Factor of deviation with ab initio rate coefficient

kcalc for kcalc kAI kAI kAI for kAI kcalc kcalc

6 H additions

4

H β scission C addition

2

C β scission H-abstraction

0 0

2

4

6

8

log(kAI/m³mol-1s-1;s-1)

10 10

AIChE Annual Meeting, 11/11/2009

Outline Introduction Group additive method Thermodynamic and kinetic parameters Reactor simulation Reaction network Reactor model Pilot plant simulations Industrial simulation

Conclusions 11

AIChE Annual Meeting, 11/11/2009

Reaction network Reaction network Automatically generated Only elementary reactions 130 species 37 molecules 93 radicals

1514 reactions reversible reactions

thermodynamic consistency incorporated range: H2, CH4, …

benzene, toluene

K eq

k forward k reverse

12

AIChE Annual Meeting, 11/11/2009

Reactor model: pilot experiments preheating

P

P

P

P

P

T

C5+ - analysis

Effect of varying process conditions in 12 experiments?

reactor zone

mixing

water

hydrocarbons

TLE

Pilot plant cell 2

cell 3

cell 4

cell 5

cell 6

cell 7

- process gas temperature measurement Furnace characteristics and Process variables

Furnace Furnace length Furnace height Furnace width

Process variables 4.0 m 2.6 m 0.7 m

Reactor coil

Reactor length Reactor diameter Material

C4- - analysis

cell 1

23.14 m 10-2 m Incoloy 800HT

Feed flow rate ethane propane butane methane Steam dilution COT COP

3.1 – 4 kg/h 10 - 68 wt% 19 - 77 wt% 0 - 24 wt% < 3 wt% 0 – 0.3 kg/kg 1004-1121 K 0.152-0.296 MPa

Small reactor diameter →1D reactor model is sufficient

13

AIChE Annual Meeting, 11/11/2009

Simulation results: pilot: ethane cracking Product yields for 7 pilot experiments 100

Varying process parameters

Simulated Product Yield (wt%)

90

ethene ethane

80

Steam dilution

0.25-0.6 kg/kg

COT

1063-1163 K

70 7

60

Hydrogen

6

50

5

40

4

Methane Ethyne Propene

3

30

1,3-Butadiene

2

20

1

10

0

Benzene

0

1

2

3

4

5

6

7

0 0

20

40

60

80

Experimental Product Yield (wt%)

Sabbe et al., submitted to AIChE J.,

100

Perfect conversion and major product yields, without a single adjusted parameter 14

AIChE Annual Meeting, 11/11/2009

Simulation results: pilot: C2/C3 Main process variables

Feed

50 wt% C2 50 wt%C3

COT

1111 K

COP

0.298 MPa

Steam dilution

0.3 kg/kg

45 40

Experimental

35

Simulated

30 25 20 15 10 5 0 CH4

CH4

C2H4

C2H6

C3H6

C3H8

1,3-butadiene

15

AIChE Annual Meeting, 11/11/2009

Simulation results: pilot: C2/C3/C4 Varying process variables

Feed

10 wt% C2 66-77 wt% C3 12-23 wt% C4

COT

1091-1123 K

COP

0.158-0.164 MPa

Steam dilution

0 kg/kg

Ab initio predicted yield (wt%)

35

methane 30

ethene 25

ethane 20

propene 15

propane 10

1,3-butadiene 5

n-butane 0

0

5

10

15

20

25

30

Experimental yield (wt%)

35 16

AIChE Annual Meeting, 11/11/2009

Simulation results: pilot: C1/C2/C3/C4 Varying process variables

Feed

3 wt% C1 67 wt% C2 22 wt%C3 8 wt% C4

COT

1005-1119 K

COP

0.152-0.157 MPa

Steam dilution

0 kg/kg

Excellent conversion and major product yields, without a single adjusted parameter

7

methane 60

Ab initio predicted yield (wt%)

Ab initio predicted yield (wt%)

70

ethene ethane

50

propene 40

propane 30 20 10

1,3-butadiene 6

1-butene

5

n-butane

benzene 4

H2 3 2 1

0

0 0

10

20

30

40

50

60

Experimental yield (wt%)

70

0

1

2

3

4

5

6

Experimental yield (wt%)

7

17

AIChE Annual Meeting, 11/11/2009

Simulation results: pilot experiments z=21.56m ; T=1151.5 K. F0 = 0.92 g s-1 ; δ = 0.583 kg/kg COT = 1101 K ; COP = 1.91 105 Pa

Rates relative to C2H6 decomposition

18

AIChE Annual Meeting, 11/11/2009

Simulation results: pilot experiments Main benzene formation pathway

Under steam cracking conditions (T