259603

Research ArticleAnalysis of Process Variables via CFD toEvaluate the Performance of a FCC Riser

H C Alvarez-Castro1 E M Matos1 M Mori1 W Martignoni2 and R Ocone3

1School of Chemical Engineering University of Campinas 500 Albert Einstein Avenida 13083-970 Campinas SP Brazil2PETROBRASAB-RETROT 65 Republica do Chile Avenida 20031-912 Rio de Janeiro RJ Brazil3Chemical Engineering Heriot-Watt University Edinburgh EH144AS UK

Correspondence should be addressed to M Mori morifequnicampbr

Received 25 August 2014 Revised 2 January 2015 Accepted 15 January 2015

Academic Editor Deepak Kunzru

Copyright copy 2015 H C Alvarez-Castro et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Feedstock conversion and yield products are studied through a 3Dmodel simulating the main reactor of the fluid catalytic cracking(FCC) process Computational fluid dynamic (CFD) is used with Eulerian-Eulerian approach to predict the fluid catalytic crackingbehavior The model considers 12 lumps with catalyst deactivation by coke and poisoning by alkaline nitrides and polycyclicaromatic adsorption to estimate the kinetic behavior which starting from a given feedstock produces several cracking productsDifferent feedstock compositions are considered The model is compared with sampling data at industrial operation conditionsThe simulation model is able to represent accurately the products behavior for the different operating conditions considered Allthe conditions considered were solved using a solver ANSYS CFX 140The different operation process variables and hydrodynamiceffects of the industrial riser of a fluid catalytic cracking (FCC) are evaluated Predictions from themodel are shown and comparisonwith experimental conversion and yields products are presented recommendations are drawn to establish the conditions to obtainhigher product yields in the industrial process

1 Introduction

Since the first FCC commercial riser many improvementshave been achieved which have helped the process reliabilityand its capacity to transform heavier feedstock at relativelylow costs currently the FCC process remains the primaryconversion process in the petrochemical industry For anumber of refineries the fluid catalytic cracking remains themain source of profitability and the accomplishment of itsoperation decides the market competiveness of the crackingunit Approximately 350 FCC units are in operation world-wide with over 127 million barrels per day as total capacityMost of the existing cracker units have been designed ormodified by six major technology licensers [1]

The design of each FCC unit can be different but theircommon target is to upgrade low-cost hydrocarbons to morevaluable products FCC and ancillary units such as thealkylation unit are responsible for about 45 of the gasolineproduced worldwide Papers have flourished in recent years

in the attempt to describe and simulate numerically thephenomena observed in such process To predict the solidand gas phase behavior the Eulerian-Eulerian approach hasbeen used due to low computational effort required [2] Inthis study the Eulerian-Eulerian approach is used where thesolid phase is treated as a continuum [3ndash5] Computationalfluid dynamic (CFD) was implemented to solve discretizedequations a hybrid mesh (tetrahedral mesh with refiningprisms at the wall) was used as the calculation grid The 12-lump kinetic model proposed by Wu et al [6] with catalystdeactivation was coupled with the hydrodynamic model toevaluate the full problem The lumping approach has beenstudied to describe the kinetic behavior of catalytic crackingwith a large number of components where each lump isconstituted by hundreds of kinds of molecules in a specificrange of molecular weights The methodology is shown tobe very powerful when a large number of components areinvolved [7ndash11] The simulation model uses a 12-lump effectthat the variation of different process variables has on the

Hindawi Publishing CorporationInternational Journal of Chemical EngineeringVolume 2015 Article ID 259603 13 pageshttpdxdoiorg1011552015259603

2 International Journal of Chemical Engineering

conversion network which has the advantage of representingwith good reliability the products and presents the option ofrepresenting the feedstock through three different lumpsThepurpose of this study is to predict the yield and conversionbehaviors at different operating conditions in the industrialriser of a FCC unit with a 12-lump kinetic model Differentoperational conditions have been studied in order to estimateproduct yields

2 Riser Process

The riser is the main equipment of the FCC unit Inside theriser the feedstock is fed through nozzles and mixture withthe catalyst and the accelerant steam in the injection zoneThe performance of the nozzles to guarantee fast vaporizationof the feedstock and a good contact of the gasoil dropletswith the catalyst is key to improve the FCC riser efficiencythe feedstock nozzles are positioned about 5ndash12 meters abovethe bottom of the reactor In accordance with kind of FCCdesign the number of feedstock injections can be from 1 to15 Practically all of the riser reactions take place between 1and 3 s Reactions start as soon as the feed enters in contactwith the hot catalyst

The increasing velocity due to the vapor production actsas the means to carry the catalyst up in the riser The hotsolid supplies the necessary heat to vaporize the feedstock andbring its temperature to the temperature needed for crackingcompensating also for the reducing in temperature dueto endothermic behavior of riser reactions Standard risersare designed for an outlet velocity of 12ndash18ms During theoperation coke deposits on the catalyst declining the catalystactivity and thus representing a concern for the efficiency ofthe cracking reactions [12]

3 Mathematical Model

The fluid dynamic equations and kinetic model are summa-rized in Section 31 and taken and adapted from Alvarez-Castro [13] the catalytic cracking kinetic models are takenfromWu et al [6] and Chang et al [14] In order to study theheterogeneous kinetics and the particle phase deactivation(15)ndash(20) were implemented in the CFX code

31 Governing Equations for Transient Two Fluid Models

Governing Equations

(1) Gas-solid fluid model (Eulerian-Eulerian) [15]

120597

120597119905(120576119892120588119892) + nabla sdot (120576119892120588119892u119892) = 0

120597

120597119905(120576119904120588119904) + nabla sdot (120576119904120588119904u119904) = 0

(1)

where 120576 is the volume fraction 120588 is density and u isthe velocity for each phase

(2) Momentum equations

120597

120597119905(120576119892120588119892u119892) + nabla sdot (120576119892120588119892u119892u119892)

= nabla sdot [120576119892120583119892(nablau119892 + (nablau119892)

119879

)] + 120576119892120588119892g minus 120576119892nabla119901 +119872

120597

120597119905(120576119892120588119892u119904) + nabla sdot (120576119904120588119904u119904u119904)

= nabla sdot [120576119904120583119904 (nablau119904 + (nablau119904)119879)] + 120576119904120588119904g minus 120576119904119866nabla120576119904 minus119872

(2)

where119901 is the pressure 120583 the viscosity119866 themodulusof elasticity 119892 the acceleration of gravity and 119872 theinterphase momentum transfer

119872 = (1501205762

119904120583119892

1205761198921198892119904

+7

4

10038161003816100381610038161003816u119904 minus u119892

10038161003816100381610038161003816120576119904120588119892

119889119904

)(u119904 minus u119892)

For dense zones where 120576119904 gt 02

119872 = (3

4119862119889

10038161003816100381610038161003816u119904 minus u119892

10038161003816100381610038161003816120576119904120576119892120588119892

119889119904

)(u119904 minus u119892)

For dilute zones where 120576119904 lt 02

(3)

where 119889119904 is the solid diameter and 119862119889 is the dragcoefficient

119862119889 =044

120576265119892

(Re gt 1000

adequate for inertial effects to govern viscous effects)

119862119889 =1

120576265119892

24

Re(1 + 015Re0687)

(Re lt 1000 viscous and inertial effects are significant)

119866 = exp [119862119866 (120576119904 minus 120576119904max)]

(4)

(see [16]) where 120576119904max is maximum volume fractionand the packing limited about 065

(3) Turbulence equations

120583119892 = 120583lam119892 + 120583tur119892 (effective viscosity) (5)

(a) The 119896-epsilon mixture model [17]

120583tur119892 = 1198621205831205881198921198962

120598 (6)

International Journal of Chemical Engineering 3

where 119896 is the turbulence kinetic energy 120598 is theturbulence eddy dissipation and 119862120583 is constant

120597

120597119905(120588119892119896) + nabla sdot (120588119892u119892119896) = nabla sdot [120583lam119892 +

120583tur119892

120590119896

+ nabla119896]

+ 119875119896minus 120588119892120598

120597

120597119905(120588119892120598) + nabla sdot (120588119892u119892120598) = nabla sdot [120583lam119892 +

120583tur119892

120590120598

+ nabla120598]

+120598

119896(1198621205981119875

119896minus 1198621205982120588119892120598)

(7)

where 120590119896 120590120598 1198621205981 and 1198621205982 are constants 119875119896 is

the turbulence production

119875119896= 120583tur119892 (nabla119906119892 + (nabla119906119892)

119879

) (8)

(4) Heat transfer model

120597

120597119905(120576119892120588119892119867119892) + nabla sdot (120576119892120588119892119906119892119867119892)

= nabla sdot (120576119892120582119892nabla119879119892) + 120574 (119879119904 minus 119879119892)

+ 120576119892120588119892sum

119903

nabla119867119903

120575119862119903

120575119905minus 119876119877 minus 119876119881

(9)

120597

120597119905(120576119904120588119904119867119904) + nabla sdot (120576119904120588119904119906119904119867119904) = nabla sdot (120576119904120582119904nabla119879119904) + 120574 (119879119892 minus 119879119904)

(10)

where 119867 is enthalpy 119879 temperature 120582 thermalconductivity 119876119877 heat of cracking reactions and 119876119881energy lost in gasoil vaporization

120574 =Nu120582119889119904

(11)

where 120574 is the interphase heat transfer coefficient 119889119904is the diameter of the catalyst and Nu is the Nusseltnumber

Nu = 2 + 06radicRePr03 (12)

(see [18])(5) Energy lost in gasoil vaporization transfer by hot

solid

119876119881 = 400 kJkg of gasoil (13)

(see [5])(6) Kinetic model [6]

Variation of the chemical species

120597

120597119905(120576119892120588119892119862119892119868) + nabla sdot (120576119892120588119892119906119892119862119892119868) = nabla sdot (120576119892Γ119894nabla119862119892119868) + 119868

(14)

where Γ is diffusivity 119862119892119868 is the concentration ofspecies 119868 in the gas phase and 119868 is consumption orformation of each species

(61) The rate equation for the generic reaction

119868119903 = minus119896119903 sdot 120588119901 sdot (120588120572119894) sdot 120601 (119905) sdot 119865 (119873) sdot 119865 (119860) (15)

where we have the following 120601(119905) catalystpoisoning due to coke content 119865(119873) alkalinenitrides 119865(119860) polycyclic aromatic adsorption119896119903 kinetic constant 120588119901 particle density and(120588120572119894) the mass content of species 119894 in gaseousphase

(a) Decay model based on coke content

Φ (119905) = e(minus120572119905) (16)

where we have (119905) time and 120572 constant(b) Alkaline nitrides

119865 (119873) =1

1 + 119896119873119862119873119905119862119865119888119900

(17)

where 119896119873 is the adsorption factors ofnitrides119862119873 themass content of nitrides 119905119862the relative detention time of catalyst 119865119888119900the catalyst-to-oil ratio in the feedstock

(c) Polycyclic aromatic adsorption

119865 (119860) =1

1 + 119896119860 (119862119860 + 119862119877) (18)

where 119896119860 is the adsorption factor of aro-matics 119862119860 the mass content of aromaticsand 119862119877 the mass content of resins in thefeedstock

(d) Arrheniusrsquo equation

119896119903 = 1198960

119903exp(

119864119903

119877119879) (19)

(e) Arrhenius equation for any temperaturedependent on the holdup of solids

119896119888 (119879 120576119904) = 119896119903550∘C (120588 120576119904) exp [minus119864119903

119877(1

119879minus

1

550∘C

)] (20)

4 Simulation

The system of governing equations twelve-lump catalyticcracking kinetic model solid influence and catalyst deacti-vation functions was solved by employing the finite volumemethod technique using the commercial software ANSYSCFX 140 The relevant results and the calculations steps areanalyzed and discussed in detail in the following sections

4 International Journal of Chemical Engineering

Riser46m

Catalystwater vapor

Side inlet

Water vaporfluidization

Inlet

Feednozzles

Gasoil

Productsand slurry

(unconverted)

Outlet tofraccionator

90∘

45∘

2m

2m

5m

05m

07m

1m

06m

Figure 1 Riser geometry

41 Geometry andGridsGeneration Steamor fuel gas is oftenused to lift the catalyst to the feed injection In most designsthat incorporate a ldquoWyerdquo section for delivering the catalystto the feed nozzles a lift gas distributor is used providingsufficient gas for delivery of dense catalyst to the feed nozzlesIn other designs the lift gas rate is several magnitudes greaterwith the intent of contacting the gasoil feed into a moredilute catalyst stream In this work the geometry of the riseris considered according to industrial reactor specificationstaken from Alvarez-Castro [13] as shown in Figure 1 whichreports a typical riser with Wye section

The geometries considered are meshed according to theprocedure described above previous works [4 19] showedthat the CFD utilized there and adopted in this work is meshindependent and meshes of 700 to 900 thousand controlelements are recommended for a good representation ofindustrial risers A hybrid mesh with 800 thousand controlelements was built and applied in this work Details of outletand inlet mesh can be seen in Figure 2

42 Model Setting-Up To implement the numerical simula-tion the hydrodynamic configuration of themodel was set upfirst and then the 12-lump kinetic model was linked with thehydrodynamic equations Appropriate specific subroutinesthat is user defined function (UDF) were implemented inthe model and solved in the CFX code in order to considerthe heterogeneous endothermic kinetics and catalyst deacti-vation

421 Hydrodynamic Setup The setup considered in thiswork considered steam as the fluidization agent which wasfed into the bottom of the riser a side inlet was used forfeeding in the particle phase A small amount of the steam(3 to 7wt of the total steam) was fed together with thecatalyst and 12 nozzles 5 meters above the riser base wereused to feed gasoil the zone where the nozzles are located is a

Inlet Outlet

0 2000 40001000 3000(m)

Figure 2 Mesh details

Table 1 Operating conditions

Item ValueReaction temperature (K) 79315Reaction time (s) 322Flux of fresh feedstock (th) 12446Inlet temperature of fresh feedstock (K) 54315Catalyst temperature at riser inlet (K) 91315Ratio of catalyst to oil 81

very significant one since it is responsible for guaranteeingfast vaporization of the liquid gasoil recent technologieshave led to development of high-efficient nozzles [20ndash22]which implies a time for complete vaporization of about 3(around 005 to 02 seconds) of the total reactant residencetime in the reactor in typical operation conditions In thepresent simulation it was assumed that the feedstock is totallyvaporizedThe nonslip and free slip condition at the walls wasused for the phases

Gasoil properties and operating conditions used in thepresent work were taken from Wu et al [6] and Chang et al[14] and are summarized in Tables 1 and 2 respectively

According to Nayak et al [5] 400 kJkg is the heat to beadopted in the simulation needed for the evaporation of theliquid droplets

422 Kinetic Model Setup A 12-lump model was used torepresent the products and feedstock behavior [23] Suchmodel can undergo a large number of reactions (56 reactions)

International Journal of Chemical Engineering 5

Table 2 The property of feedstock

Item ValorDensity (kgm3 at 29315 K) 924Hydrogen content (wt) 121Group analysis (wt)

Saturates 6605Aromatics 2525Resins + asphaltenes 87

Distillation (K)HK lt5781510 6641530 7091550 7341570 77115Alkaline nitrides content (mgg) 1750Conradson carbon residue (wt) 233

SS[1]

GS[5]

Lp[8]

LO3[9]

GO[6]

SA[2]

SR[3]

GA[7]

LO4[10]

DI[4]

CK[12]

DR[11]

Figure 3 Twelve-lump kinetic model network [6]

leading to a large number of products depending on thedifferent types of feedstock The kinetic paths are shownin Figure 3 and Table 3 summarizes the different rangesof products and the feedstock characterization The valuesof the kinetic constants activation energies and catalystdeactivation constant are listed in Table 4 In heat transfermodel (9) 119876119877 is estimated by the amount of coke producedin cracking reactions this factor 119876119877 is equal to 9127103 kJmultiplied by the mass of coke which is corresponding toendothermic reactions in riser of FCC [6 23]

43 Convergence Transient expressions were estimated viathe second-order backward Euler method The convectiveterms were interpolated through a second-order upwindscheme ldquohigh-resolution methodrdquo

In the simulation was used a time step of 10minus3 secondsto provide a lower Courant number in order to ensure

Table 3 Lumps of the 12-lump kinetic model [6]

Lump symbol Lump Boiling range119878119878 Saturates in feedstock 61315 K+

119878119860

Aromatics infeedstock

119878119877

Resin and asphaltenein feedstock

119863119868

Diesel withoutpretreating LCO 47715ndash61315 K

119866119878 Saturates in gasoline C5 - 47715 K

119866119874 Olefins in gasoline

119866119860 Aromatics in gasoline

119871119901 low carbon alkanes C3 + C4

1198711198743 Propylene

1198711198744 Butene

119863119877 Dry gas C1 + C2 + H2

119862119870 Coke

simulation results were not dependent on the time stepselected and monitoring the simulation with Courant num-ber less than one The convergence for progressing in timeimplied a residual squaremean less than 10minus4The simulationswere solved using computers provided with Xeon 3GHzdual core processors About twelve days of calculation wasnecessary to predict a period of time (15 [s]) long enough toshow that the variables had a cyclic behavior

The following section reports the numerical results aimedat evaluating how the variation of the different operationvariables affects the heat transfer the chemical reaction andthe hydrodynamic behavior of the riser

5 Results and Discussion

Comparing model predictions for industrial reactors withplant data is not an easy task because the computationalmodel requires detailed information about the feedstock aswell as the design and operating conditions of the industrialsetups and petroleum companies normally do not releasethese data on industrial risers

51 Validation of the Simulation Results The catalyst distri-bution profile for the riser is shown in Figure 4 Figure 4(a)shows a rendering of volume for catalyst volume fractionsalong an axial extension for the first six meters of the riserheight where it can be seen that just after the expansion zoneand the nozzles the feedstock has reacted and consequentlyis produced and the gas velocity increases due to less productsdensity so the catalyst moves at the high velocities imposedby the gasoil injection and its distribution becomes increas-ingly uniform with increasing height Catalyst distribution isshown on eight radial contour planes in an axial directionin Figure 4(b) On the first radial planes it can be observedthat just after expansion the solid phase (dense region)tends to agglomerate at the center of riser This is due tothe high velocity of the injected gasoil which prevents thephenomenon of catalyst agglomeration on thewalls known as

6 International Journal of Chemical Engineering

Table 4 Kinetics constants and activation energies of reaction [6 14]

Reaction path Activationenergy (kJmol)

Exponentialfactor (m3kgs) Reaction path Activation

energy (kJmol)Exponential

factor (m3kgs)1 rarr 4 27251 05496 4 rarr 6 130832 000785

1 rarr 5 09432 01478 4 rarr 7 122401 004245

1 rarr 6 12912 0728 4 rarr 8 82513 000853

1 rarr 7 72956 001707 4 rarr 9 37156 000294

1 rarr 8 130682 000221 4 rarr 10 34688 000479

1 rarr 9 89495 000824 4 rarr 11 163068 000765

1 rarr 10 7787 000289 4 rarr 12 102884 006204

1 rarr 11 88766 002903 5 rarr 6 14641 001914

1 rarr 12 95764 002268 5 rarr 7 157166 000595

2 rarr 4 47964 05068 5 rarr 8 153998 106119864 minus 05

2 rarr 5 40451 009092 5 rarr 9 13124 000982

2 rarr 6 141004 00178 5 rarr 10 128934 004039

2 rarr 7 135735 002794 5 rarr 11 182895 000547

2 rarr 8 07088 003926 5 rarr 12 19805 000055

2 rarr 9 34203 00579 6 rarr 5 126572 006655

2 rarr 10 37921 002698 6 rarr 7 89658 01029

2 rarr 11 47483 002206 6 rarr 8 135233 125119864 minus 13

2 rarr 12 33867 004335 6 rarr 9 121083 00297

3 rarr 4 101081 004164 6 rarr 10 121945 00246

3 rarr 5 143479 002781 6 rarr 11 146554 001485

3 rarr 6 158237 01043 6 rarr 12 113696 000878

3 rarr 7 1601057 001088 7 rarr 8 140169 130119864 minus 06

3 rarr 8 09537 03375 7 rarr 9 119348 001566

3 rarr 9 19214 01208 7 rarr 10 104221 008629

3 rarr 10 135212 008769 7 rarr 11 102512 009008

3 rarr 11 40009 005663 7 rarr 12 93636 005506

3 rarr 12 39143 006459 8 rarr 11 303051 0002563

4 rarr 5 144455 0006942 10 rarr 11 385004 0000683

119896119860= 0003854 119896

119873= 0002009 120572 = 0002543

coral annulus and guarantees keeping a much more uniformdistribution (better homogenization) throughout the riser

The fluidization velocity of the steam at the bottom of theequipment has a major effect on catalyst residence time inthe reaction system as presented in previouswork byAlvarez-Castro [13]

The model developed for the riser simulation was usedto simulate the plant data reported by Chang et al [14] inorder to validate the model and compare the product yieldsand conversions behavior Products distributions that is theaverage yields along the height of the riser are shown inFigure 5 Red and green curves represent the main yieldproducts (gasoline and diesel resp) it can be seen that after25meters an asymptotic behavior is achieved at the end of theequipment with less conversion due to overcracking Blueyellow and brown curves represent LPG dry gas and cokerespectively The black curve shows the total unconvertedslurry Results show good agreement between simulation andexperimental data

Conversions and final products yields simulations modeland the industrial data are shown in Figures 6 and 7 respec-tively Measurements were taken at the riser outlet in orderto compare the accuracy of the model simulation with thepredicted results

52 Operational Variables Data obtained from Petrobrason the multipurpose pilot unit U-144 (height of 17m anddiameter of 052m) in which different tests were carriedout by changing the feedstock temperature the catalysttemperature and the catalyst-to-oil ratio are reported inTable 5

The sensitivity of the conversions and products yields toprocess variables based on the validated simulation modelwas studied The conversions and yields were found to bevery sensitive to variations in feedstock temperature catalysttemperature and catalyst-to-oil ratio the differences in theconversion and product yields were in the range of 1 to 5

International Journal of Chemical Engineering 7

Table 5 General behavior of the multipurpose pilot unit U-144 studied

Item Catalyst to oil ratio Catalyst temperature Temperature of freshfeedstock Residence time

78 to 86 680 to 720 (K) 530 to 550 (K) 1 to 22 [s]Slurry (unconverted) Decrease Decrease Decrease DecreaseDI diesel Decrease Decrease Decrease DecreaseGasoline Decrease Decrease Decrease DecreaseLPG Increase Increase Increase IncreaseDR dry gas Increase Increase Increase IncreaseCK coke Increase Increase Increase Increase

(a)

25m

15m

7m

6m

53m

43m

35m

010

009

008

007

006

004

003

002

001

000

(b)

Figure 4Volume rendering and contour profiles for axial and radialplanes of catalyst volume fractions

Industrial data

0 5 10 15 20 25 30 35 40 45 500

10

20

30

40

50

60

70

80

90

100

(wt

)

Z (m)

Feedstock (unconverted)GasolineDieselLPG

Dry gasesCoke

Figure 5 Model simulation results and industrial data

05

1015202530354045

DI diesel Gasoline LPG

Model simulationIndustrial data

()

Slurry

feedstock)(unconverted

CK coke

DR drygases

344

279

5

383

0

158

7

516

924

492

274

5

392

9

153

3

475

826

Figure 6 Comparison between product yields industrial data andthe simulation model

3830 3929

2795 2745

1587 1533

924 826516 475344 492

0102030405060708090

100

Model simulation Industrial data

GasolineDI dieselLPG

CK cokeDR dry gasesSlurry (unconverted feedstock)

()

Figure 7 Comparison between industrial data and the simulationmodel for each case

Followed results obtained for the three variables studied inthis order

(a) feedstock temperature

(b) catalyst temperature

(c) catalyst-to-oil ratio

8 International Journal of Chemical Engineering

Table 6 Operating conditions with variations in feedstock temperature

Item Case A Case B Case C Case D Case EValue Value Value Value Value

Reaction temperature (K) 79315 79315 79315 79315 79315Fluidization steam () 3 3 3 3 3Flux of fresh feedstock (th) 12446 12446 12446 12446 12446Inlet temperature of fresh feedstock (K) 44315 49315 54315 59315 64315Catalyst temperature at riser inlet (K) 91315 91315 91315 91315 91315Ratio of catalyst to oil 81 81 81 81 81

Case

A

Case

B

Case

C

Case

D

Case

E

9133

7957

6781

5605

4429

(K)

Figure 8 Temperature profiles for the axial plane with variationfeedstock temperature

521 Feedstock Temperature Different case studies for tem-peratures ranging between 44315 K and 64315 K were testedwhile holding the other operating conditions constant asshown in Table 6

(1) Comparison of the Hydrodynamics Profiles for DifferentFeedstock Temperatures The global temperature (two phases)was calculated as arithmetic average contour planes for allthe case studies as shown in Figure 8 the profile for case Ahas the lowest inlet feedstock temperature and profile for caseE has the highest It can be observed that the temperaturedistributions are similar in all cases with an approximatevariation of 50 [K] between the first and last cases A and E

Figure 9 contains the profiles for average temperature(two phases) along the center line of riser height whichwas also calculated as arithmetic average the temperaturedecreases significantly after the feeding area due to theendothermic nature of the reaction

(2) Dependence of Product Yield on Feedstock TemperatureThe percentage of yield and conversion products for eachcase presented in the previous section is shown in Figure 10The yields were broken down into the followingmain groupsgasoline diesel LPG dry gas and coke Feedstock cracking is

700

750

800

850

900

950

0 5 10 15 20 25 30 35 40 45 50

Tem

pera

ture

(K)

Case A = 44315KCase B = 49315KCase C = 54315K

Case D = 59315KCase E = 64315K

Z (m)

Figure 9 Temperature profiles through riser with variation infeedstock temperature

3935 3872 3830 3781 3762

2897 2857 2795 2768 2742

1527 1588 1587 1590 1610

825 855 924 929 1030442 457 516 545 580419 368 344 342 288

0102030405060708090

100

()

GasolineDI dieselLPG

CK cokeDR dry gasesSlurry (unconverted feedstock)

Case

A=443

15

K

Case

B=49315

K

Case

C=543

15

K

Case

D=59315

K

Case

E=643

15

K

Figure 10 Products yields for each feedstock temperature

represented by complex series-parallel reactions where gaso-line and diesel are intermediate products fromwhich the finalproducts (LPG dry gas and coke) are produced If feedstockrate of conversion is too high because of high temperaturethe secondary reactions of the intermediate products causethe rate of yield to decrease due to overcracking or generationof more final products

International Journal of Chemical Engineering 9

Table 7 Operating conditions with variations in catalyst temperature

Item Case A Case B Case C Case D Case EValue Value Value Value Value

Reaction temperature (K) 79315 79315 79315 79315 79315Fluidization steam () 3 3 3 3 3Flux of fresh feedstock (th) 12446 12446 12446 12446 12446Inlet temperature of fresh feedstock (K) 54315 54315 54315 54315 54315Catalyst temperature at riser inlet (K) 81315 86315 91315 96315 101315Ratio of catalyst to oil 81 81 81 81 81

419

289

7

393

5

152

7

442

825

368

285

7

387

2

158

8

457

855

344

279

5

383

0

158

7

516

924

342

276

8

378

1

159

0

545

929

288

274

2

376

2

161

0

580

103

0

05

1015202530354045

Slurry(unconverted

feedstock)

DI diesel Gasoline LPG CK coke

Case A = 44315KCase B = 49315KCase C = 54315K

Case D = 59315KCase E = 64315K

()

DR dry gases

Figure 11 Comparison of the product yields for different feedstocktemperatures

Feedstock temperature has an important role in theprocess A comparison of the product yields and conversionfor all cases studied is reported in Figure 11 where it canbe seen that cases A B and C have higher gasoline anddiesel yields but a lower feedstock conversions while casesD and E have lower gasoline and diesel yields but a higherdiesel conversion The temperature is lower at the highergasoline and diesel yields the importance of which shouldbe evaluated by a cost analysis of feedstock reprocessing orproduction of dry gases and coke in order to improve theplant targets

522 Catalyst Temperature Different cases with catalysttemperatures ranging between 81315 [K] and 101315 [K] weretested while holding constant the other operating conditionsas shown in Table 7

(1)TheEffect of Catalyst Temperature onRiserHydrodynamicsThe global temperature (gas and solid) was calculated asarithmetic average contour planes for the different casestudies are shown in Figure 12 Case A is characterized bylower average overall temperature in the riser while cases BC D and E show a drastic increase in the average overalltemperature in the riserwith higher temperature in the profilefor case E It may be noted that small changes in the catalystfeed temperature cause a significant increase in the overalltemperature

Temperature profiles plotted along the riser height areshown in Figure 13 for all cases studied andwere calculated as

Case

A

Case

B

Case

C

Case

D

Case

E

10156

8724

7292

5861

4429

(K)

Figure 12 Global temperature profiles for the axial plane withvariations in catalyst temperature

arithmetic average (gas and solid phases) It can be observedthat catalyst temperature has a strong effect on the overalltemperature in the riser showing that the temperature pro-files with a variation of 50 [K] similar to the inlet temperatureof the catalyst have a much greater effect

(2)Dependence of Product Yields onCatalyst TemperatureThepercentages of conversions and product yields for each casestudied are shown in Figure 14 The percentages of convertedgasoil and product yields are reported

The product yields for each case studied are shown inFigure 15 Case A has higher gasoline and diesel yields but alower conversion of diesel while case E has lower gasolineand diesel yields and a higher percentage of final productssuch as light gases coke and LPG In the latter case thefeedstock conversion is higher due to the higher temperaturewhich causes the intermediates to undergo overcrackinggenerating lighter products of lower commercial value

523 Catalyst-to-Oil Ratio Study Catalyst-to-oil ratios from61 to 101 with step increases ratio of 1 for all cases werestudied while holding all other variables constant as shownin Table 8

10 International Journal of Chemical Engineering

Table 8 Operating conditions with variations in catalyst-to-oil ratio

Item Case A Case B Case C Case D Case EValue Value Value Value Value

Reaction temperature (K) 79315 79315 79315 79315 79315Fluidization steam () 3 3 3 3 3Flux of fresh feedstock (th) 12446 12446 12446 12446 12446Inlet temperature of fresh feedstock (K) 54315 54315 54315 54315 54315Catalyst temperature at riser inlet (K) 91315 91315 91315 91315 91315Ratio of catalyst to oil 61 71 81 91 101

700

750

800

850

900

950

1000

1050

0 5 10 15 20 25 30 35 40 45 50

Case A = 81315KCase B = 86315KCase C = 91315K

Case D = 96315KCase E = 101315K

Z (m)

Tem

pera

ture

(K)

Figure 13 Temperature profiles through riser altering catalyst tem-perature profiles for the riser with variations in catalyst temperature

3935 3872 3830 3781 3762

2857 2818 2795 2761 2729

1477 1574 1587 1574 1636

809 891 924 1018 1082433 491 516 529 576439 350 344 345 287

0102030405060708090

100

()

GasolineDI dieselLPG

CK cokeDR dry gasesSlurry (unconverted feedstock)

Case

A=813

15

K

Case

B=863

15

K

Case

C=913

15

K

Case

D=963

15

K

Case

E=1013

15

K

Figure 14 Products yields for each catalyst temperature

(1) Dependence Riser Hydrodynamics on Catalyst-to-Oil RatioThe catalyst-to-oil ratio is an important variable since it hasa direct effect on the conversion and selectivity of gasolineand diesel Figure 16 shows the profile of the catalyst volume

439

285

7

398

6

147

7

433

809

350

281

8

392

1

157

4

491

891

344

279

5

383

0

158

7

516

924

345

276

1

378

2

157

4

529

101

8

287

272

9

365

9

163

6

576

108

2

05

1015202530354045

Slurry(unconverted

feedstock)

DI diesel Gasoline LPG CK coke

()

Case A = 81315KCase B = 86315KCase C = 91315K

Case D = 96315KCase E = 101315K

DR dry gases

Figure 15 Comparison of the product yields for different catalysttemperatures

Case

A

Case

B

Case

C

Case

D

Case

E010

008

005

003

000

Figure 16 Catalyst volume fraction profiles for catalyst-to-oil ratio

fraction for the different case studies with case A having alower catalyst-to-oil ratio and case E having a higher one incomparison to all cases studied In both cases A and B it canbe noted that the fraction of catalyst is lower along the riser

International Journal of Chemical Engineering 11

Case

A

Case

B

Case

C

Case

D

Case

E

91817

82411

73006

63601

54195(K

)

Figure 17 Temperature profiles for the axial plane with variationsin catalyst-to-oil ratio

700

750

800

850

900

950

0 5 10 15 20 25 30 35 40 45

Tem

pera

ture

(K)

Z (m)

Case A = 610 catoilCase B = 710 catoilCase C = 810 catoil

Case D = 910 catoilCase E = 1010 catoil

Figure 18 Temperature profiles for the riser with variation incatalyst-to-oil ratio

height and higher at the side where the catalyst is fed in Incases D and E catalyst fraction is higher and more uniformalong the riser height

At the bottom of the riser where the feedstock is injectedthe temperature profile is very complex and chaotic due to thecontact between hot catalyst reagents and steam

Figure 17 shows the temperature profiles for a contourplane Case A is characterized by a lower catalyst-to-oil ratiowhile case E represents a higher catalyst-to-oil ratio Thetemperature profiles increase from case A to case E

Figure 18 contains the temperature global profiles (gasand solid) along the center line of the riser which can beobservedwith variations in catalyst-to-oil ratioWhen the gasencounters the barrier formed by the catalyst particles whichbegins the reaction the temperature decreases slowly alongthe riser due to the endothermic nature of the reaction

4070 3974 3830 3669 3508

2969 294 2795 2645 2486

1289 1388 15871646 1732

750 86 924 1157 1319424 48 516 606 758545 41 344 238 170

0102030405060708090

100

()

GasolineDI dieselLPG

CK cokeDR dry gasesSlurry (unconverted feedstock)

Case

A=6

10

cato

il

Case

B=7

10

cato

il

Case

C=8

10

cato

il

Case

D=9

10

cato

il

Case

E=1010

cato

il

Figure 19 Product yields for each catalyst-to-oil ratio

545

296

9

407

0

128

9

424

750

41

294

397

4

138

8

48

86

344

279

5

383

0

158

7

516

924

238

264

5

366

9

164

6

606

115

7

170

248

6

350

8

173

2

758

131

9

05

1015202530354045

Slurry(unconverted

feedstock)

DI diesel Gasoline LPG CK coke

()

DR dry gases

Case A = 610 catoilCase B = 710 catoilCase C = 810 catoil

Case D = 910 catoilCase E = 1010 catoil

Figure 20 Comparison of the product yields for catalyst-to-oilratio

(2) Dependence of Product Yields on Catalyst-to-Oil RatioTheconversions and yields for each case study are reported inFigure 19 with case A characterized by a lower catalyst-to-oil ratio and case E by a higher catalyst-to-oil ratio It can benoted that this variable has a large impact on product yieldespecially for gasoline and diesel

A comparison of the product yields in the case studiedis presented in Figure 20 Case A has higher gasoline anddiesel yields but a lower conversion of feedstock on thecontrary case E with a higher catalyst-to-oil ratio has lowergasoline and diesel yields but higher percentages of lightgases coke and LPG Case A has a higher kinetic gas buton the other hand diesel yield is low because of a lowercatalyst-to-oil ratio Higher catalyst-to-oil ratio undergoesan overcracking which generates lighter and lower valueproducts

12 International Journal of Chemical Engineering

6 Conclusions

The kinetic and hydrodynamic behavior of the riser in a FCCprocess has been simulated employing the 12-lump kineticsmodel in conjunction with ANSYS CFX 140 software Themodel has been validated against industrial data showingthe ability to capture the relevant features characterizing theindustrial FCC riser behavior Systematic investigations havebeen carried out to study the influence of the catalyst temper-ature the feedstock temperature and the catalyst-to-oil ratioon the riser performance Specifically it has been shown thatwhen the inlet conditions (of the feedstock and catalyst) arefixed the yield of the wanted product can be increased bycontrolling the temperature of the riser and the catalyst-to-oil ratio Conditionswhich lead to a better homogenization ofthe flow avoiding unwanted hydrodynamic features such asthe core-annulus flow which could lead to poor conversionhave also been identified By comparing the simulated resultswith the experimental data it can be concluded that themodel mimics well the process therefore the model can beemployed as a tool helping the design operation and controlof industrial FCC risers

Nomenclature

119862119894 Molar concentration of component 119868 [kmolmminus3]

119862119889 Drag coefficient [-]119862119866 Constant of elasticity modulus function [Pa]119862120583 Constant 0091198621205981 Constant 1441198621205982 Constant 192119889 Particle diameter [m]119864 Activation energy [Jmolminus1]119892 Gravitational acceleration [m2sminus1]119866 Elasticity modulus [Pa]119867 Static enthalpy [Jmolminus1]119896 Kinetic constant of reaction [m3kmolminus1sminus1]

or turbulent kinetic energy [m2sminus2]1198960 Preexponential factor [m3kmolminus1sminus1]119896119888 Deactivation constant [kgcatkmolminus1]Nu Nusselt number [-]119901 Static pressure [Pa]119901119896 Shear production of turbulence [Pasminus1]

Pr Prandtl number [-]1199021 Specific coke concentration [kmolkg

minus1

cat]119877 Reaction rate [kmolm

minus3sminus1] or universal gasconstant [Jmolminus1Kminus1]

Re Reynolds number [-]119879 Static temperature [K]u Velocity vector [msminus1]119876119877 Heat of cracking reactions [JKgminus1]119876119881 Energy lost in gasoil vaporization [JKgminus1]

Greek Letters

119872 Interphase momentum transfer [kgmminus3sminus1]120576 Volume fraction [-]

120598 Turbulence dissipation rate [m2sminus3]0 Catalyst decay function [-]120574 Interphase heat transfer coefficient

[Wmminus2Kminus1]Γ Diffusivity [kgmminus1sminus1]120582 Thermal conductivity [Wmminus1Kminus1]120583 Molecular viscosity [Pas]120588 Density [kgmminus3]120590119896 Constant 100120590120598 Constant 3001198621205981 Constant 3001198621205982 Constant 300120593 Tracer concentration [kgm3]

Subscripts

119892 Gas phase119904 Solid phase119877 Reactionlam Laminarturb Turbulent

Conflict of Interests

The authors declare that their research is only for academicpurposes there is not any financial gain or other kind ofbenefits influenced by secondary interest

Acknowledgment

The authors are grateful for the financial support of Petrobrasfor this research

References

[1] R Sadeghbeigi ldquoProcess descriptionrdquo in Fluid Catalytic Crack-ing Handbook R Sadeghbeigi Ed chapter 1 pp 1ndash42Butterworth-Heinemann Oxford UK 3rd edition 2012

[2] F Durst D Milojevic and B Schonung ldquoEulerian and Lagran-gian predictions of particulate two-phase flows a numericalstudyrdquoAppliedMathematicalModelling vol 8 no 2 pp 101ndash1151984

[3] X Lan C Xu G Wang L Wu and J Gao ldquoCFD modelingof gasndashsolid flow and cracking reaction in two-stage riser FCCreactorsrdquoChemical Engineering Science vol 64 no 17 pp 3847ndash3858 2009

[4] G C Lopes L M Rosa M Mori J R Nunhez and WP Martignoni ldquoThree-dimensional modeling of fluid catalyticcracking industrial riser flow and reactionsrdquo Computers andChemical Engineering vol 35 no 11 pp 2159ndash2168 2011

[5] S V Nayak S L Joshi and V V Ranade ldquoModeling of vapor-ization and cracking of liquid oil injected in a gas-solid riserrdquoChemical Engineering Science vol 60 no 22 pp 6049ndash60662005

[6] F Y Wu H Weng and S Luo ldquoStudy on lumped kineticmodel for FDFCC I Establishment of modelrdquo China PetroleumProcessing and Petrochemical Technology no 2 pp 45ndash52 2008

[7] J Ancheyta-Juarez F Lopez-Isunza E Aguilar-Rodrıguezand J C Moreno-Mayorga ldquoA strategy for kinetic parameter

International Journal of Chemical Engineering 13

estimation in the fluid catalytic cracking processrdquo Industrial ampEngineering Chemistry Research vol 36 no 12 pp 5170ndash51741997

[8] H Farag A Blasetti and H de Lasa ldquoCatalytic cracking withFCCT loaded with tin metal traps Adsorption constants forgas oil gasoline and light gasesrdquo Industrial amp EngineeringChemistry Research vol 33 no 12 pp 3131ndash3140 1994

[9] I Pitault D Nevicato M Forissier and J-R Bernard ldquoKineticmodel based on a molecular description for catalytic crackingof vacuum gas oilrdquoChemical Engineering Science vol 49 no 24pp 4249ndash4262 1994

[10] V W Weekman Jr ldquoModel of catalytic cracking conversion infixed moving and fluid-bed reactorsrdquo Industrial amp EngineeringChemistry Process Design and Development vol 7 no 1 pp 90ndash95 1968

[11] L C Yen R EWrench and A S Ong ldquoReaction kinetic corre-lation equation predicts fluid catalytic cracking coke yieldsrdquoOiland Gas Journal vol 86 no 2 pp 67ndash70 1988

[12] R Sadeghbeigi ldquoProcess and mechanical design guidelines forFCC equipmentrdquo in Fluid Catalytic CrackingHandbook chapter11 pp 223ndash240 Butterworth-Heinemann Oxford UK 3rdedition 2012

[13] H C Alvarez-Castro Analysis of process variables via CFD toevaluate the performance of a FCC riser [PhD thesis] ChemicalEngineering Departament University of Campinas 2014

[14] J Chang K Zhang F Meng L Wang and X Wei ldquoComputa-tional investigation of hydrodynamics and cracking reaction ina heavy oil riser reactorrdquo Particuology vol 10 no 2 pp 184ndash1952012

[15] T B Anderson andR Jackson ldquoFluidmechanical description offluidized beds Equations of motionrdquo Industrial amp EngineeringChemistry Fundamentals vol 6 no 4 pp 527ndash539 1967

[16] D Gidaspow Multiphase Flow and Fluidization Continuumand KineticTheory Descriptions Academic Press Boston MassUSA 1994

[17] F R Menter ldquoTwo-equation eddy-viscosity turbulence modelsfor engineering applicationsrdquo AIAA journal vol 32 no 8 pp1598ndash1605 1994

[18] W E Ranz andWRMarshall Jr ldquoEvaporation fromdrops partIrdquo Chemical Engineering Progress vol 48 pp 141ndash146 1952

[19] H C Alvarez-Castro E M Matos M Mori and W P Marti-gnoni ldquo3D CFD mesh configurations and turbulence modelsstudies and their influence on the industrial risers of fluidcatalytic crackingrdquo in Proceedings of the AIChE Spring AnnualMeeting Pittsburgh Pa USA 2012

[20] KBR-Technology ATOMAX-2 Feed Nozzles 2009[21] J Li Z-H Luo X-Y Lan C-M Xu and J-S Gao ldquoNumerical

simulation of the turbulent gas-solid flow and reaction in apolydisperse FCC riser reactorrdquo Powder Technology vol 237 pp569ndash580 2013

[22] L M Wolschlag and K A Couch ldquoNew ceramic feed distribu-tor offers ultimate erosion protectionrdquo Hydrocarbon Processingpp 1ndash25 2010

[23] J Chang F Meng L Wang K Zhang H Chen and YYang ldquoCFD investigation of hydrodynamics heat transfer andcracking reaction in a heavy oil riser with bottom airlift loopmixerrdquo Chemical Engineering Science vol 78 pp 128ndash143 2012

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Mechanical Engineering

Advances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Distributed Sensor Networks

International Journal of

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Antennas andPropagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014