Bridges in modelling and simulation of steam cracking ... · steam cracking: from fossil to renewable feedstock and from ... Steam cracking Consumer goods from chemical industry Naphtha

Laboratory for Chemical Technology, Ghent University

http://www.lct.UGent.be

Bridges in modelling and simulation of steam cracking: from fossil to renewable feedstock and from

molecule to furnace

Kevin M. Van Geem and Guy B. Marin

1

66th Canadian Chemical Engineering Conference “Sustainability & Prosperity”October 16-19 (2016) Québec City (Canada)

From molecule to industrial plant

2

66th Canadian Chemical Engineering Conference “Sustainability & Prosperity” October 16-19 (2016) Québec City (Canada)

Molecular level

Process level

Steam cracking: from fossil to renewables

3


atozforex.com; pnnl.org; districtenergy.org; scade.fr; schmidt-clemens.de; Linde Group

Crude oil

Natural gas

Bio-based feeds

Steam cracking Consumer goods fromchemical industry

Naphtha

Light gasoil

crude oil

Natural gas liquids

Gas-condensates

Hydrode-oxygenated

FAME

Ethene

Propene

Butadiene

Aromatics

4

Steam cracking: Ghent University history


Steam cracking: hot section

5

D

A

C

A: Radiation sectionB: TLEC: Steam drumD: Convection section

Endothermic process 1050–1150 K

Preheat feed and other utility streams

Rapidly quenching of reactor effluent

B

45%

50% 800K

700K

1050–1150 K

5%

1450 K

FPH

Saturated steam

BFWECO

Feed

Dilution steam

SSH

HTC

Steam


Detailedfeedstock

composition

Simulation model

Goal Reactor geometryOperating conditions

Product specs

ModelingMicrokinetic

modelReactor model

Fundamental reactor model

Detailedproduct

composition

Feedstock properties

Reaction and reactor model

6

GC×GC

Analytical techniques


Toraman, H.E. et al., Journal of Chromatography A, 1460, 135-146, 2016Ristic, N.D. et al.,Journal of Visualized Experiments, Issue 114

7

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor

• Process

• Conclusions


8

Outline

• Introduction

• Feedstock: characterization

• Kinetics

• Reactor

• Process

• Conclusions


918-10-2016

State-Key Laboratory of Chemical Engineering@ECUST September 22, 2009

GC�GC chromatogram: 2 partsConventional 1D part � C4-Comprehensive 2D part � C5+

3D view

9

On-line GCxGC


Pyl, S.P. et al.,Journal of Chromatography A, 1218, 3217-3223, 2011

SIMCO: ANN or Shannon entropy

10

Simulation model

Reactor geometryOperating conditions


Product specs

Microkinetic model

Reactor model


Detailedfeedstock

composition

Detailedproduct

compositionGC×GC

Standard methods

GoalAnalytical techniques

Feedstock reconstructionSIMCO


• Average molecular weight

• Elemental composition• Specific density • Global PINA analysis• Boiling point data

(e.g. D2887 simdist)• Aromatic Sulfur

Maximization of Shannon Entropy

11

Feedstock properties Detailed composition• Species identity

Feedstock reconstruction

± 20 properties More than 100 unknown mole fractions

Shannon entropy

maximization

( ) ( ) ∑∑==

=−=MM N

ii

N

iiii yyyyS

11

1 with ln MAX

Constraints from mixing rules (example):

∑∑=

=MN

i iii

ii

y

y

1exp Mwd

Mw

d

1∑

=

=MN

iiiy

1exp MwMw

• Mole or mass fractions


S.P. Pyl et al., AIChE Journal, 56, 12, 3174-3188, 2010

SIMCO results: Hydrocarbons

12

n-paraffins i-paraffins

Naphthenes Aromatics


13

Outline

• Introduction

• Feedstock: renewables

• Kinetics

• Reactor

• Process

• Conclusions


Thermochemical conversion of biomass

14


LIGNOCELLULOSIC BIOMASS

BIO-OIL CHAR

CO

CO2

H2OCH4

H2

GAS

NH3

HCNC2H4

5wt% Torrefaction35wt% Slow pyrolysis13wt% Fast pyrolysis85wt% Gasfication

20wt%30wt%75wt%5wt%

75wt%35wt%12wt%10wt%

Toraman, H.E.et al.,Bioresource Technology, 207, 229-236, 2016

Model components for biomass pyrolysis

15


CO CO2H2O CH4 H2 C2H4

Study molecules withstructural moieties found in biomass

For example:

T

Gamma-ValeroLactone (GVL) pyrolysisReaction families1) Scission2) Hydrogen abstraction3) β-scission/addition4) Concerted ring opening

16


De Bruycker, R. et al.,Proceedings of the Combustion Institute, 35, 515-523, 2015De Bruycker, R. et al., Combustion and Flame, 164, 183-200, 2016

17

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor

• Process

• Conclusions


CRACKSIM: steam cracking kinetics

18

Simulation model

Goal Reactor geometryOperating conditions


Product specs

Modeling Microkineticmodel:CRACKSIM

Reactor model


Detailedfeedstock

composition

Detailedproduct

composition


19

Outline

• Introduction

• Feedstock

• Kinetics: reaction network

• Reactor

• Process

• Conclusions


Families of elementary reactions

Bond dissociation and radical recombination

20

R1 R2 R1 R2• •

+

Hydrogen abstraction (inter- and intramolecular)

R1 H R1R2••

R2 H++

Radical addition and β-scission (inter- and intramolecular)

R1 R2 R3R1•

R2 R3•+


Decomposition scheme: n-hexane

21

RH

β species

PSSA to µ radicals


µ and β networks: CRACKSIM

22

β network µ network

CRACKSIM

Bi- and monomolecularreactions for β radicals

Monomolecular reactionsfor µ radicals

R1 R2 R1 R2• •

+

R1 H R1R2••

R2 H++

R1 R2 R3R1•

R2 R3•+

R1 R2 R1 R2• •

+

R2-R1 H•

R2-R1H•

R1 R2 R3R1•

R2 R3•+

R1

R2R3

•R1-R2 R3 •

1324 reversible reactions51 molecules43 radicals

13584 schemes676 molecules


Validation

23


Experimental yields during steam cracking of bio-derived hydrocarbonsFeedstock composition: MWaverage = 230g/mol 51wt% normal alkanes – 49wt% branched alkanesFHC,0 = 0.04 g/s, FH2O,0 = 0.02 g/s, P = 0.17 MPa

calculated

Network generators

24

●●

●●

●●●

●

●

●

●

●

●

●

●

●

●

RINGDaoutidisMinneapolis

GenesysPRIM(-O)ReNGePLCT, Ghent, Belgium

RMGGreenCambridge

NetGenBroadbeltDelaware

EXGASBattin-LeclercNancy, France

RDLMavrovouniotisEvanston

RDL++VankatasubramanianWest Lafayette

CASBPorolloJosjkar-Ola, Russia

MAMOX(+)(++)RanziMilan, Italy

REACTIONBlurockLund, Sweden

COMGENTruongSalt Lake City

MECHGENHébergerBudapest, Hungary

KINGDi MaioCosenza, Italy

RAINUgiMunich, Germany

GRACE, YonedaKUCRS, MiyoshiTokyo, Japan

RNGLedererPrague, Czech Republic

MECHEMValdes-PerezPittsburgh


GENESYS -GENEration [of reacting ] SYStems

25

A new program for kinetic model construction


Van de Vijver, R. et al. International Journal of Chemical Kinetics, 47 (4), 199-231, 2015

Genesys: Kinetic model construction using chemo-informaticsVandewiele, N.M.; Van Geem, K.M.; Reyniers, M.-F.; Marin, G.B.Chemical Engineering Journal, 207-208, 526-538, 2012

26

Graph theory yields powerful algorithms

PatternRecognition

MoleculeIdentifiers

1

2

3 4

56

8

7

12

9

11

10

1

2

3

4

5

6

8

7

12

9

11

10

Reactant -> product


Vandewiele, N.M. et al., Journal of Computational Chemistry, 36 (3), 181-192, 2015

“Kinetics of Chemical Reactions : Decoding Complexity”G.B. Marin and G.S. Yablonsky,Wiley-VCH Verlag, 446 pages, 2011

27

Outline

• Introduction

• Feedstock

• Kinetics: ab initio

• Reactor

• Process

• Conclusions


28

Objective: data base

Reactor model

Solver

Reaction network

Kinetic and thermodynamic data

Develop a consistent data set based on ab initio calculations

Reactor simulation for hydrocarbon radical chemistry

? Larger speciesusing group additivity


Van de Vijver, R. et al., Chemical Engineering Journal, 278, 385–393, 2015

29

Benson’s group additive methodBenson group

X: Central atomValence ≥ 2{ C, Cd, O, CO, CCO, C•, Cd

• }

A, B, C, D: LigandH, C, Cd , O, CO, CCO, C•, Cd

•, O•,CO•,CCO•

Group additivity for thermochemistry

)ln( ,, opt

oint

ooint n

RSSCSHf pf

σ+°=°∆=

Corrections for non-nearest-neighbor interactions (NNI)� hydrogen bonds� gauche interactions� other interactions

Group additive values(GAV)

)group( ∑∑ +=j

ji

if NNIGAVf

Notation: X-(A)(B)(C)(D)

BXD

C

A


30

From small to large species with group additivity

O

OOH

2-methoxy-2-methylbut-3-enoic acid additive groups

Cd-(H)2

C-(C)( Cd)(O)(CO)

Cd-(C)(H)

O-(C)2

C-(O)(H)3

C-(C)(H)3

CO-(C)(O)

O-(CO)(H)

O

OOH

Additivity

= + + + +

+ + + +

O

OH

Atoms in large molecules

Atoms in small moleculeswith similar surroundings

Group additivity

O


31

Outline

• Introduction

• Feedstock

• Kinetics: thermo first

• Reactor

• Processen

• Conclusions


Ince, A.et al. AIChE Journal, 61 (11), 3858–3870, 2015

32

Thermo e.g. for oxygenates: GAV data base

Database of thermodynamic data, ∆fH°, S° and Cp° (300 K-1500 K)• 450 oxygenate compounds• CBS-QB3 methodology• 1D-HR approximation for all internal rotors

• 157 GAVs using Benson’s GA method

• 26 NNI corrections (mainly hydrogen bonds)

• 77 HBIs for the thermochemistry of radicals

NNI8 Hbr_H-O-C-CO

2-hydroxy-2-methyl-propanal


33

Outline

• Introduction

• Feedstock

• Kinetics: rate coefficients

• Reactor

• Process

• Conclusions


34

Kinetics: computational approachConventional Transition State Theory (high pressure limit)

The CBS-QB3 ab initio method is used.

Electronic barrier ∆E0

• Ideal gas approximation• Hindered Rotor (1D-HR)

Partition functions q

• Eckart

Tunneling coefficient κ

RT

E

mBA

B eVqq

q

h

Tk)T()T(k

0‡

∆−∞ = κ

‡E

A + B [AB] ‡ C

Transition state

∆‡Eelectronic

Reaction coordinate

‡

Reactants

Products


35

Group additivity for kinetics:data base of ∆GAVo

Proposed by Saeys et al. for activation energies (AIChE J. 2004, 50 (2), 426-444.)Extended by Sabbe et al. for pre-exponential factors (Phys. Chem. Chem. Phys. 2010, 12, 1278-1298)

Transition state for hydrogen abstraction

‡opt,

‡opt,

σ

σ∏∏

= jj

jj

e n

nn

Group additivity for Arrhenius parameters

Number ofsingle events

RT

Ea

Aek−

=

Arrhenius equation

ores

iii

ii

iref AYXCTATA

~log)(GAV)(GAV)(GAV)(

~log)(

~log

3

1

ο

A~

log

3

1

ο

A~

log

2

1

oA~

log∆+∆+∆+∆+= ∑∑∑

===

oE

3

1

ο

E

3

1

ο

E

2

1

oE ref,aa resa,aaa

E)(GAV)(GAV)(GAV)()( ∆+∆+∆+∆+= ∑∑∑=== i

iii

ii

YXCTETE

RT

E

Eckart eAnknka~

eGAe

−== κκ

Primary Secondary Tertiary

C2

K3K2

K1

C1 HX2 Y2

Y1

Y3X3

X1

K6

K4

K7 K8 L7 L8

L2L3

L4

L6

L5

L1

L9K9

K5


Kinetics: data base of ∆GAVo

Hydrogen abstraction ( ethyl + ethenyl methylether)

CH=CH2

resEEEE EYGAVXGAVCGAVCGAVTETE ,a1o

1o

2o

1o

refa,a )()()()()()(aaaa

∆+∆+∆+∆+∆+=

C1 H C2

H

H

H

H

‡

• OCH3+CH2 CH3 CH3

O CH2CH3 CH3 + CH2

OCH2

36

+4.7 kJmol-1 +0 kJmol-169.7 kJmol-1 -18.2 kJmol-1

Ea, ab initio= 62.0 kJmol-1Ea,GA = 62.6 kJmol-1

C1-(C)(H)2 C2-(O)(H)2CH3 + CH4

-0.866 +0.4805.268 +0

logAab initio= 5.203log(AGA /m3 mol-1s-1)= 5.095

-0.053

O-(C2)(Cd)C-(C1)(H)3 H-

+6.8 kJmol-1 -0.4 kJmol-1

C1-(C)(H)2 C2-(O)(H)2CH3 + CH4 O-(C2)(Cd)C-(C1)(H)3 H-

-0.512


Paraskevas, P. et al.,Journal of Physical Chemistry A, 119 (27), 6961-6980, 2015

37

Kinetics: validation of group additivity

Sabbe et al. ChemPhysChem 2008, 9 (1), 124-140Sabbe et al. ChemPhysChem, 2010, 11(1), 195-210Sabbe et al. PhysChemChemPhys, 2010, 12 (6), 1278-1298

0

2

4

6

8

10

0 2 4 6 8 10

logk

GA

log(k AI/m³mol -1s-1;s -1)

H additionsH β scissionC additionC β scissionH-abstraction

1000 K


38

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor

• Process

• Conclusions


39

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor : pilot scale

• Process

• Conclusions


40

Steam Cracking Pilot Plant

Gas-Fired Furnace + Reactor

Online AnalysisSection

Control “Room”

High temperaturesampling system

HCFeed

H2O


41

Pilot results: C1/C2/C3/C4Process conditions

Feed 3 wt% C167 wt% C222 wt%C38 wt% C4

COT 1005-1119 K

COP 0.152-0.157 MPa

Steam dilution 0 kg/kg

0

10

20

30

40

50

60

70

0 10 20 30 40 50 60 70

Ab

initi

o pr

edic

ted

yiel

d (w

t%)

Experimental yield (wt%)

methane

ethene

ethane

propene

propane

0

1

2

3

4

5

6

7

0 1 2 3 4 5 6 7

Ab

initi

o pr

edic

ted

yiel

d (w

t%)

Experimental yield (wt%)

1,3-butadiene

1-butene

n-butane

benzene

H2

without a singleadjusted parameter


42

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor: 3D alternatives

• Process

• Conclusions


Coke formation


Mitigation by- Feed additives- Metallurgy & surface technology- 3D reactor technology

Deposition of a carbon layer on the reactor surface

Thermal efficiency

Product selectivity

Decoking procedures

Estimated annual cost to industry: $ 2 billion

2L. Benum, "Achieving Longer Furnace Runs at NOVA Chemicals," in AIChE Spring National Meeting, 14th Annual Ethylene Producers’ Conference, New Orleans, Louisiana, 2002.

43

Coke formation: 3D reactor technologies

44


~ 15 °C

~100 °C

Reduce convective heat resistance

Better mixingIncrease surface area

Cokes formed hereT ↑ � coking rate ↑↑

~60 °C

*Borealis.com, kubota.com, Technip.com

Computer Models

Quantify

coking &

selectivity effect

“There ain’t no such thing as a free lunch”

45


* Milton Friedman - TANSTAAFL

3D reactor

technology

Economics Selectivity ↓

Increasedheat transfer

Reducedcoking rate

Longer run lengthMore capacity

Increasedfriction

Higheraverage reactor

pressure

Flow Chemistry

3D CRACKSIM

46

Helicoidal finned tubes

α Pitch

RadiusRadiant coils in Borealis Furnace, KBR

D

e

Computational domain: 1 fin with periodic boundaries

66th Canadian Chemical Engineering Conference “Sustainability & Prosperity” October 16-19 (2016) Québec City (Canai

47

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor : CFD models

• Process

• Conclusions


Reynolds-Averaged Navier-Stokes (RANS)Single model for all scales, additional equations to provide closures

Large Eddy Simulation (LES)Resolve relevant energy containing scales, model the smaller energy dissipating eddies

Direct Numerical Solution (DNS)Fully resolve all time and length scales 48

Resolving turbulence


49


Radial temperature profiles

More uniform gas temperatureLower metal temperature

Bare Straight Helix SmallFins

Optimized

Axial wall temperature and coking profile

50


Cracking reaction model26 components

13 radical species212 elementary reactions

Coking model

Cokemodel

- 50°C

- 49%

S. Wauters and G.B. Marin; Chemical Engineering Journal, 82, 267-279, 2001S. Wauters and G.B. Marin, IEC Research, 41, 2379-2391, 2002

Effect on start of run yields

51


- 0.7 wt% + 0.3 wt%

Radical reaction model26 components

13 radical species212 elementary reactions

52

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor: “run length”

• Process

• Conclusions


Run length simulation


SOR 96h48h

Increasing run length

Millisecond propane cracker

54


• Feedstock 118.5 kg/h propane• Propane conversion 80.15 % (± 0.05%)• Steam dilution 0.326 kg/kg• CIT 903.7 °C• COP 170 kPa

Different geometries simulated• Same reactor volume• Same axial length• Same minimal wall thickness

Bare c-RibFin

Non-uniform coke layer growth

55


SOR (0 hrs)

48 hrs

Fin c-Rib

SOR

10 days

Tube Metal Temperature

56


Max. TMT increasesThermal resistance coke layer

Bare

c-Rib

Fin

Pressure drop

57


Pressure drop increasesCross-sectional flow area

decreases due to coke

Bare

c-Rib

Fin

58

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor

• Process

• Conclusions


59

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor

• Process: fire box

• Conclusions


60

Coupled reactor-furnace simulation

External coil temperature

Heat flux to reactors

Reactor

(COILSIM1D)Furnace simulationconvergence

FURNACE/COILSIM1D


Hu, G. et al.,Industrial & Engineering Chemistry Research, 54 (9), 2453-2465, 2015

61

Ultra Selective Conversion (USC)

� 100% floor burner

� Fuel composition mol%: CH4(89%)-

H2(11%)

� U coil

� Feedstock: Naphtha

Coupled modeling� 3D CFD furnace model

� 1D reactor model (COILSIM1D)

� Detailed cracking kinetics

(CRACKSIM)

Full furnace simulation


Zhang, Y. et al.,AIChE Journal 61 (3), 936-954, 2015

62

detailed simplified

Detailed : long flame burners


63

� Detailed case

• Stronger turbulence

• Faster reaction

detailed simplified

Methane mole fraction

Flue gas: velocity and concentration fields


64

This

information is key

forcontrol

Tube wall temperature field: local hot spots


65

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor

• Process: convection section

• Conclusions


66

Steam cracker convection section

EvaporatorEvaporator

Steam

Feed

Nozzle

Steam

Feed

Nozzle Feed-steam mixture overheater-1

Feed-steam mixture overheater-1

Feed Evaporator

Steam super heater

Mixture over-heater-1

Mixture over-heater-2

Flue gas out ~ 700K

Flue gas in~ 1400 K

Mixing nozzle

To radiation section

Schematic of convection section

Heavy feed


Mahulkar, A.V. et al., Chemical Engineering Science, 110, 31-43, 2014

67

Gas condensate: multicomponent mixture

• Wider stick regime as compared to that in single componentdroplet regime map

• For splash and limited splash, the no. of daughter droplets formedis greater than predicted by correlations available in literature

Rebound

StickSplash

Limited Splash

0

200

400

600

800

1000

1200

480 530 580 630 680 730 780

No

rma

l W

eb

er

nu

mb

er

Wall temperature (K)

Stick

Rebound

Splash

Limited Splash

820TLF

critNormWe


Mahulkar, A.V. et al., Chemical Engineering Science, 130, 275-289, 2015

68

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor

• Process: hot section

• Conclusions


69

The COILSIM1D package

Reactor Coil

Transfer Line ExchangerConvection section

Firebox

Long Flame Burners

Wall Burners

70

The COILSIM1D package

Downstream

simulators

...

Upstream

simulators

…


71

Outline

• Introduction

• Feedstock

• Kinetics

• Reactor

• Process

• Conclusions


72

Conclusions

integration of computational chemistry methods with engineering tools at larger time and length scales and experimental validation provides a powerful tool for the optimization and/or design of industrial units

• Shannon Entropy maximization to reconstruct feedstocks in terms required for a microkinetic model

• Consistent data set for thermochemistry and kinetics of hydrocarbon and oxygenates radical chemistry

• Ab initio simulation of steam cracking of C2/C3/C4 and oxygenates

• Compatibility of renewable feedstocks with existing plants

• Emergence of 3D reactor technologies based on CFD

• Integration of reactor/convection section/furnace


Acknowledgments


The Long Term Structural Methusalem Funding

73

74

Acknowledgments

• Profs. Marie-Françoise Reyniers, Geraldine HeyndericksMaarten Sabbe, Tony Arts, Feng Qian

• Drs. Steven Pyl, Amit Mahulkar, Nick Vandewiele, Ruben Debruycker, Carl Schietekat, Thomas Dijkmans, PaschalisParaskevas, Marco Djokic, Andres Munoz

• PhD students David Van Cauwenberghe, Pieter Reyniers, Brigitte Devocht, Ruben Van de Vijver, Laurien Vandewalle, Nenad Ristic, Alper Ince, Ezgi Toraman, Yu Zhang, Marco Virgilio


People

76

Glossary

• SFT: Swirl Flow Tube, a tube with a helicoidalcenterline. The helix amplitude is smaller than orequal to the tube radius.

• Swirl flow: a whirling or eddying flow of fluid.• Swirl number: ratio of tangential over axial

momentum transfer• Wall shear stress: component of stress parallel

with the wall. It is the product of the viscosity andthe derivative of axial velocity with respect to theradial coordinate.


Documents

Bridges in modelling and simulation of steam cracking ... · steam cracking: from fossil to renewable feedstock and from ... Steam cracking Consumer goods from chemical industry Naphtha