Hydroprocessing/FCC Synergy · hydrotreating catalyst systems included a NiMo catalyst (ART AT575) and a CoMo catalyst (ART AT775).Both catalysts have had extensive use in FCC feed

Hydroprocessing/FCC Synergy

N U M B E R 1 0 1 S p r i n g 2 0 0 7

A message from the editor...

Dear Refiner:

Recently, you have been demanding to know moreabout the synergy between the fluid catalytic crackingprocess and hydroprocessing. With the formation ofART in 2001 and our long-standing premier position inthe FCC business, Grace Davison and ART have leadthe industry in these critical areas.

In light of clean fuels and ultra low sulfur dieselrequirements, you have expressed particular interestin the dynamics between cat feed hydrotreating and

fluid catalytic cracking. Chuck Olsen, ART's Worldwide Manager of TechnicalService, has co-authored our lead article titled, " Balancing the Need for FCCProduct HDS and Opportunities for Improving FCC Performance," whichhighlights the RT/ART efforts. Chuck was joined by Joanne Deady, GeorgeYaluris and Natalie Petti in preparing the paper, which will be presented at theNPRA Annual Meeting in San Antonio in March 2007.

With our extensive experience in refining, Grace Davison and ART's TechnicalTeam are ready to assist you with optimizing your hydroprocessing and FCCunits to get the best yields possible. Please contact your technical team or goto our website www.e-catalysts.com for more information. We are dedicatedto meeting your challenges.

Regards,

Robert H. BullardVice-President and General Manager, Grace DavisonManaging Director, Advanced Refining Technologies

Catalagram 101 Spring 2007 1

www.e-catalysts.com2

IN THIS ISSUEBalancing the Need for FCC Product HDS and Opportunities for Improving FCC PerformanceChuck Olsen, Worldwide Technical Services Manager, Advanced Refining TechnologiesJoanne Deady, Vice President Marketing and R&D, Grace Davison, Columbia, MDGeorge Yaluris, Marketing Manager, Grace Davison, Columbia, MDNatalie Petti, ConsultantClean fuels regulatory demands have brought a focus on the importance of optimizedFCC feed pretreating in efforts to facilitate compliance. The proper selection of FCCPretreater catalyst and severity integrated with the selection of the appropriate FCCcatalyst design and operating conditions offer a refiner the opportunity for an optimizedrefining operation.

3

Catalytic FCC Gasoline Sulfur Reduction Citgo's Corpus Christi refinery reduces FCC gasoline sulfur by up to 31% using Grace Davison's GSR®-5 additive Lauren Blanchard, Strategic Business Marketing Manager,Advanced Refining Technologies, Columbia, MDMichael Ziebarth, Research Manager, Grace Davison, Columbia, MD Timothy Dougan, Marketing Manager, Grace Davison, Columbia, MDLow gasoline sulfur is key to operating successfully in today's environment. In order todefine an effective sulfur reduction strategy a good understanding of the sulfurspecies present in the gasoline streams is essential. Grace Davison Catalysts andAdditives are designed to remove specific compounds present at different boilingranges in the gasoline. A Citgo Refinery successfully used an additive and reachedgasoline sulfur reduction of 44% with undercutting (31% without undercutting).

16CATALAGRAM 101

Spring 2007

Managing Editor:Joanne Deady

Contributors:Lauren Blanchard

Joanne DeadyDavid A. HuntChuck Olsen

Craig BorchertMin Pu

Bill MinyardJeff KoebelNatalie Petti

Timothy DouganGeorge Yaluris

Michael Ziebarth

Please addressyour comments to

[email protected]

W. R. Grace & Co.-Conn.7500 Grace Drive

Columbia, MD 21044(410) 531-4000

www.e-catalysts.com

©2007W. R. Grace & Co.-Conn.

Understanding and Minimizing FCC Slurry Exchanger FoulingDavid A. Hunt, Technical Service Manager, Grace Davison, Houston, TX

Bill Minyard, National Technical Sales Manager, Grace Davison, Houston, TXJeff Koebel, Technical Service Manager, Grace Davison, Chicago, ILSlurry Exchanger fouling can impact FCC operation by causing reduced feedrate, lossconversion and increased maintenance costs. A good understanding of the foulingmechanisms involved is important. The selection of appropriate operating conditionsand FCC catalyst selection can reduce the incidents of exchanger fouling andimprove unit profitability.

30

The information presented herein is derived from our testing and experience. It is offered, free of charge, for yourconsideration, investigation and verification. Since operating conditions vary significantly, and since they are not under ourcontrol, we disclaim any and all warranties on the results which might be obtained from the use of our products. You shouldmake no assumption that all safety or environmental protection measures are indicated or that other measures may not berequired.

Hydroprocessing/FCC Synergy

N U M B E R 1 0 1 S p r i n g 2 0 0 7

Clean Fuels: An Opportunity for Profitability Using Gasoline Sulfur Reduction Technology Lauren Blanchard, Strategic Business Marketing Manager,Advanced Refining Technologies, Columbia, MDCraig Borchert, Valero Energy, Wilmington, CAMin Pu, Valero Energy, Wilmington, CA FCC Gasoline Sulfur is an important operating parameter for a refinery operation inefforts to meet strict regulations. Outages of processes upstream and downstream ofthe FCC can impact continued compliance with these regulations. Grace Davison andValero Wilmington worked closely together to develop a successful strategy involvingfeed and catalyst selection to stay within gasoline sulfur compliance during an FCC FeedPretreater Outage.

23


Introduction

he benefits of hydrotreatingFCC feed on product yieldsand sulfur content were

recognized some time ago anddescribed in earlier publications1,2,3,4.Recently, regulatory demands andthe drive towards clean fuels haverenewed interest in FCC feedhydrotreating to facilitate complianceand satisfy the need for improvedyields. To address these needs,Advanced Refining Technologies(ART) introduced the ApART

Catalyst SystemTM for FCCpretreatment in 20015. Thistechnology was developed toprovide increases in HDSconversion while at the same timeproviding significant upgrading ofFCC feeds. In essence, an ApARTCatalyst System is a staged bed ofhigh activity NiMo and CoMocatalysts where the relativequantities of each are optimized tomeet individual refiner goals andconstraints. The ApART technologyhas been described in detailpreviously6. ART has continued to

Balancing the Need for FCC Product HDS and Opportunities

for Improving FCC Performance

T

Chuck Olsen Worldwide Technical Services Manager,Advanced Refining Technologies,Chicago, IL

Joanne Deady Vice President Marketing and R&D,Grace Davison, Columbia, MD

George YalurisMarketing Manager,Grace Davison, Columbia, MD

Natalie PettiConsultant

develop a better understanding ofthe reactions and kinetics involvedin FCC pretreating, and, throughour partnering with Grace DavisonRefining Technologies, a detailedunderstanding of the effects ofhydrotreating on FCC unitperformance. The complexity ofcombinations of catalyst designand operating conditions for boththe FCC feed hydrotreater and theFCC unit presents a significantoptimization opportunity forrefiners to drive the combinedoperation to maximum productvalue.

FCC Pretreating

There are numerous reactions thatoccur during hydrotreating of heavygas oils making it a challengingprocess to model. A partial list ofreactions includes sulfur andnitrogen removal, aromaticssaturation, carbon (MCR) removal,and contaminant metals removal.The progression of several of thesereactions as a function of thehydrotreater temperature is shownin Figure 1. Each reactant in thefigure shows a unique response totemperature which indicates thatdifferent reaction kinetics areinvolved for each species. A widerange of conversion levels isachieved, with some reactants at50% conversion and others at90%+ conversion. As shown inFigure 1, the conversion of

polynuclear aromatics (PNA's) passesthrough a maximum as temperatureincreases, and begins to decrease astemperature increases beyond thatpoint. The conversion of totalaromatics follows a similar profile withincreasing temperature.

The data in Figure 2 examines theconversion of aromatics in more detail.This figure shows how the aromatics’conversion varies with temperature andLHSV when hydrotreating a coker gasoil feed blend. The existence of amaximum in conversion is readilyapparent. At higher temperatures theconversion begins to decrease, andthis is due to thermodynamicequilibrium constraints. Thetemperature where the conversionbecomes thermodynamically limited is

a strong function of feedstock,hydrogen partial pressure and LHSV.As indicated in Figure 2, thetemperature of maximum conversionincreases as LHSV increases and,not surprisingly, the overallconversion level decreases. A similartrend is observed for the PNAconversion although in that case theconversion levels are highercompared to total aromaticsconversion. It is important to notethat the aromatics conversion trendsobserved with increasing LHSV aresimilar to what is observed as thehydrotreating catalyst ages. Thetemperature for the maximum in PNAconversion (and aromatics) graduallyincreases as the catalyst ages fromstart of run (SOR) to end of run(EOR).

It is generally accepted thatremoving PNA's and nitrogen fromthe FCC feed will improve FCCperformance. As an example,Figure 3 summarizes FCC pilotplant data for several hydrotreatedFCC feeds. The left y-axisrepresents the FCC conversion andthe right axis represents the FCCfeed quality, shown as a ratio ofeither product nitrogen content tofeed nitrogen content or productPNA content to feed PNA content.Clearly, the PNA and nitrogencontent of the FCC feed have astrong effect on FCC performance.At the highest pretreater severity theFCC conversion actually decreases

Figure 1Reactant Conversion in FCC Pretreat

0.00

0.10

0.20

0.30

0.40

0.50

0.60

660 680 700 720 740 760 780

Temperature, ˚F

Pro

d C

on

c / F

eed

Co

nc

Sulfur

Nitrogen

PNA

MCR

Figure 2Reactant Conversion in FCC Pretreat

30

35

40

45

50

55

660 680 700 720 740 760

Temperature, ˚F

Aro

mat

ics

Co

nve

rsio

n,

wt.

%

Equilibriumconversionf (T, H2 partial pressure)

increasing LHSV


by about 1 wt.% relative to theprevious severity in the chart. Thedecrease in FCC conversioncoincides with an increase in PNAcontent from 0.48 to 0.55. Noticealso that the conversion decreasesdespite the fact that the FCC feednitrogen content is at the lowestlevel suggesting the PNA feedcontent has the larger impact onFCC conversion.

It has been previously shown6 thatthe type of hydrotreating catalyst orcatalyst system used in thepretreater is critical for FCC unitperformance. However, it is notenough to simply have the right FCCpretreat catalyst as there is also anoperating window for the pretreaterwhich maximizes the performanceof the FCC. ART and GraceDavison Refining Technologies haverecently collaborated on a studyexamining the synergy between theFCC pretreater and the FCC unit.The scope of the study covered theeffects of catalyst type (bothpretreater and FCC catalysts) aswell as pretreater operating severity.A typical FCC feed sample washydrotreated in the pilot plant withboth NiMo and CoMo hydrotreatingcatalysts across a range ofoperating severity levels, and thehydrotreated product samples werethen catalytically cracked in theFCC pilot plant. A variety of FCCcatalyst types were tested toevaluate the combined effects of

hydrotreating catalyst type, pretreateroperating severity, and FCC catalystdesign on FCC yields.

I. FCC Pretreating Pilot PlantResults

A VGO/coker gas oil blend was usedas the feedstock for the FCCpretreater in this case study. Theproperties for the feedstock, alongwith the pilot plant test conditions are

summarized in Table I. Thehydrotreating catalyst systemsincluded a NiMo catalyst (ARTAT575) and a CoMo catalyst (ARTAT775). Both catalysts have hadextensive use in FCC feed pretreatapplications around the world.Several hydrotreating severitieswere investigated that cover thetypical range encountered in FCCpretreating, and both catalysts weretested across the same range ofseverities. The resultinghydrotreated products were thenused as FCC feed in subsequentFCC pilot plant testing.

Figure 4 shows how the productsulfur from the pretreater changeswith temperature for both catalysts.At the lowest severity, the CoMocatalyst removes more sulfur thanthe NiMo catalyst, as expected. Atthe higher severities the NiMocatalyst slightly outperforms theCoMo catalyst with this feed andthese operating conditions. Theproduct sulfur content spans therange from 1000 ppm down toaround 50 ppm.

Figure 3Pretreater PNA and Nitrogen Conversion Effect FCC Conversion

66

68

70

72

74

76

78F

CC

Co

nve

rsio

n (

wt.

%)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Nitro

gen

or P

NA

Co

nten

t (Pro

d/F

eed)

FCCConversion

Nitrogen

Increasing Pretreater Severity

PNA

Table ITest Conditions and Feed Properties

Operating ConditionsLHSV, hr-1 1.0Temperature, °F 675-750H2 Pressure, Psi 1300

Feedstock SR/CokerAPI 22.7Sulfur, wt.% 1.40Nitrogen, ppm 1500Total Aromatics, wt.% 45.9PNA, wt.% 25.1Distillation, °F (ASTM D-1160)

IP 38210 59530 68950 75170 81690 89095 960EP 1060


Figure 5 shows how the productnitrogen varies with temperature forboth catalysts and, as expected, theNiMo catalyst significantlyoutperforms the CoMo catalystthroughout the temperature rangeinvestigated. The product nitrogencovers the range from about 600ppm down to <100 ppm. The widerange in nitrogen content suggeststhat clear differences might beexpected in the subsequent FCCpilot plant testing.

Figure 6 shows how the PNAconversion varies for both catalysts.Again the NiMo catalyst significantlyoutperforms the CoMo catalyst forsaturation of PNA's. The range ofhydrotreater severity was chosen tocover the range where PNAconversion is thermodynamicallylimited, and thus, PNA conversionfor both catalysts is seen todecrease at the highertemperatures in the study. The PNAconversion for the NiMo catalystvaried from around 77% to 86%compared to the CoMo catalystrange of only 70% to 78%. Again,these ranges suggest substantialdifferences in FCC feed qualitywhich should result in differences inFCC performance.

Figure 7 shows the HDS and PNAdata for the NiMo catalyst on onechart. It is useful to view the data inthis way when discussing FCCpretreater operating modes. Themaximum PNA conversion is about86%, and at that temperature theproduct sulfur (i.e. FCC feed sulfur)is about 200 ppm. This representsthe FCC feed quality for a pretreaterrunning in maximum PNA saturationmode. Another common way tooperate the FCC pretreater isconstant HDS mode. Assuming thetarget sulfur is 0.1 wt.%, the sulfurtarget is met at significantly lowertemperature than required for themaximum PNA saturation mode. Atthat temperature the correspondingPNA conversion is nearly 10 wt.%lower than that obtained in PNAsaturation mode.

Figure 4Comparison of HDS Activity in FCC Pretreat

10

100

1000

10000

660 680 700 720 740 760

Temperature, ˚F

Pro

du

ctS

ulf

ur,

pp

m

NiMo

CoMo

Figure 5Comparison of HDN Activity in FCC Pretreat

NiMo

0

100

200

300

400

500

600

700

660 680 700 720 740 760

Temperature, ˚F

Pro

du

ct N

itro

gen

, pp

m

CoMo

Figure 6Comparison of PNA Saturation in FCC Pretreat

65

70

75

80

85

90

660 680 700 720 740 760

Temperature, ˚F

PN

A C

on

vers

ion

, w

t.%

NiMoCoMo


Given the lower nitrogen and PNAcontents achieved from operating inPNA mode it is expected that theFCC feed will result in significantlyhigher FCC conversion and highergasoline yield compared to runningin HDS mode. This improvement,however, does come with someadditional operating costs for thepretreater. The PNA mode ofoperation will result in higherhydrogen consumption and ashorter run length compared to theHDS mode of operation.

The pretreater pilot plant data wasused to calibrate the ART FCCpretreatment model for the feedwhose properties are summarizedin Table I. The model was then usedto predict the hydrogenconsumption and FCC pretreatercycle length for both the HDS andPNA modes of operation shown inFigure 7. The results are listed inTable II.

Figure 8 shows the results ofmodeling the differences in FCCpretreater temperature profiles forthe two operating strategies. InHDS mode, the reactor temperatureis increased to maintain a constantproduct sulfur, 0.10 wt.% in thisexample. The end of run (EOR) istypically determined by a maximumoutlet temperature, which isreached in about 36 months in thiscase. In PNA mode, the reactortemperature is ramped up to theconditions resulting in maximumPNA conversion. The temperatureis then adjusted to maintainconstant PNA conversion. PNAsaturation activity deactivates at aslower rate relative to HDS activity,so the rate of temperature increasein PNA mode is much slower thanfor the HDS mode.

The EOR for the PNA mode ofoperation is determined by therequired sulfur level in the FCCfeed. Figure 9 shows how the sulfurconversion changes throughout thecycle for each mode of operation.In HDS mode the conversionremains constant throughout the

Figure 7Comparison of HDS and PNA Activities

0

200

400

600

800

1000

1200

1400

660 680 700 720 740 760

Temperature, ˚F

Pro

du

ct S

ulf

ur,

pp

m74

76

78

80

82

84

86

88

PN

AC

on

vers

ion

, w

t.%

Sulfur

PNA

PNA modeHDS mode

Mode HDS PNAFCC Feed QualitySulfur, ppm 1000 260Nitrogen, ppm 420 150PNA Conversion, wt.% 77 86

H2 Consumption, SCFB 395-415 470-490Cycle Length, mos 36 24

Table IIComparison of FCC Pretreater Operating Modes

Figure 8FCC Pretreat Temperature Profiles

660

680

700

720

740

760

780

0 3 6 9 12 15 18 21 24 27 30 33 36

Cycle Length, Months

Tem

per

atu

re,

˚F

PNA

HDS

EOR Temperature for HDS mode

Constant HDS vs.Optimum PNA


run. In PNA mode the sulfurconversion at SOR is much higherthan required due to the higher SORtemperature, and because HDSactivity deactivates at a higher ratethan PNA saturation activity, theHDS conversion decreases as thecycle progresses. The EOR isreached when HDS conversion hasdecreased to the target sulfur level,24 months in this case.

Figure 10 shows how the PNAconversion changes through thecycle for both modes of operation.In PNA mode the conversion staysnearly constant until close to EORwhen equilibrium constraints start tolimit conversion. In HDS mode, thePNA conversion starts out relativelylow, and increases as the cycleprogresses. The PNA conversionreaches a maximum and thenbegins to decrease through the lastpart of the cycle due to equilibriumlimits. In this mode of operation themaximum PNA conversion (i.e.maximum FCC feed upgrade) isonly achieved for a small fraction ofthe total cycle.

As Figure 11 shows, the nitrogenconversion decreases from SOR toEOR for either operating mode.Under these operating conditionsHDN activity deactivates at a higherrate than HDS activity or PNAsaturation activity, so HDNconversion is much lower at EORcompared to SOR. For the HDSoperating mode HDN activitydecreases from around 72% to 45%from SOR to EOR, while for the PNAmode the HDN conversion declinesfrom 87% to about 57% from SOR toEOR.

These data clearly show that thereare significant differences in FCCfeed quality as the cycle moves fromSOR to EOR in either mode ofoperation, and this indicates thatFCC performance will change asthe cycle progresses regardless ofthe operating strategy. Toinvestigate this further, the ARTmodel was used to simulate theFCC feed properties as thepretreater cycle progressed from

Figure 9HDS Conversion Changes with Operating Mode

PNA

HDS

90

91

92

93

94

95

96

97

98

99

100

0 3 6 9 12 15 18 21 24 27 30 33 36%

HD

S C

on

vers

ion

Constant PNA operation


Constant HDS operation

PNA mode EORdue to sulfur

Figure 10PNA Conversion Changes with Operating Mode

72

74

76

78

80

82

84

86

88

0 3 6 9 12 15 18 21 24 27 30 33 36

PN

A C

on

vers

ion

, w

t.%

PNA

HDS





Figure 11HDN Conversion Changes with Operating Mode

40

50

60

70

80

90

100

0 3 6 9 12 15 18 21 24 27 30 33 36


%H

DN

Co

nve

rsio

n


PNA

HDS




SOR to EOR for both modes ofpretreater operation. The FCC feedproperties were then used toestimate FCC conversion, theresults of which are shown in Figure12. Operating the pretreater in PNAmode provides significantly higherFCC conversion for most of thecycle, but again it comes at the costof shorter pretreater run length andincreased hydrogen consumption.These costs need to be weighedagainst the benefits in FCCperformance to determine the bestoperating strategy.

There is the potential to extend thecycle length of the pretreater whenoperating in PNA mode through theuse of sulfur reduction technologiesin the FCC unit. FCC gasoline sulfurcan be reduced by 20%-35%commercially7,8,9 which wouldenable the pretreater to run to ahigher sulfur level. Figure 13 showsthe impact of using a sulfurreduction product in the FCC unitthat delivers 20% reduction in FCCgasoline sulfur (such as D-PriSM® orGSR®-5 additives, or SuRCA®

catalyst) while operating thepretreater in PNA mode. Thepretreater product sulfur is allowedto increase in months 24 through 28while the FCC gasoline sulfurremains constant because thesulfur reduction product is used tocontrol FCC gasoline sulfur duringthose months.

The results of estimating the FCCconversion for the FCC sulfurreduction case are shown in Figure14. This chart demonstrates thatthe use of a sulfur reductionproduct in the FCC can extend thepretreater cycle length in PNA modewhile at the same time maintain thebenefits of increased FCCconversion.

Figures 9 through 14 show howchanges in pretreater operatingmode can influence FCC feed sulfur,nitrogen and PNA content whichdirectly effects FCC unit conversion.As the pretreater progresses fromSOR to EOR in either mode ofoperation, the FCC feed properties

Figure 12PNA vs. HDS Run Mode Comparison

66

68

70

72

74

76

78

FC

C C

on

vers

ion

(w

t.%

)

PNA Mode

HDS Mode

Run Length (months)

PNA Mode

HDS Mode

10 20 30 400

Reduced Run Length

Incremental FCC Yield

Figure 13Cycle Length with a Sulfur Reduction Product

PNA

HDS

90

91

92

93

94

95

96

97

98

99

100

0 3 6 9 12 15 18 21 24 27 30 33 36


%H

DS

Co

nve

rsio

n FCC Sulfur ReductionConstant PNA operation



EOR withFCC SulfurReduction

Figure 14Benefit for PNA mode with FCC Sulfur Reduction vs. HDS Run Mode

66

68

70

72

74

76

78

Run Length (months)

FC

C C

on

vers

ion

(w

t.%

)

PNA Mode

HDS Mode

PNA Mode

HDS Mode

Incremental FCC Yield

Increased Run Length fromFCC Sulfur Reduction

0 10 20 30 40


change, again directly influencingFCC unit conversion. The nextsection will show how the FCCcatalyst formulation can potentiallybe used to offset pretreater productproperty shifts, as well as tooptimize the joint operation of thepretreater and FCC units formaximum profitability.

II. FCC Unit Pilot Plant Results

The next phase of the studyinvolved taking the hydrotreatedproducts and using them as FCCfeed in an FCC pilot plant. The FCCfeeds were cracked over threedifferent FCC equilibrium catalysttypes. The pilot unit was run at aconstant reactor temperature of 990ºF, and each of the FCC e-catsamples was tested at threecatalyst-to-oil ratios (4, 6, and 8) inthe pilot unit.

The yields resulting from FCCequilibrium catalyst Type 1 will bediscussed first. The properties ofthe Type 1 Ecat are shown in TableIII.

Type 1 catalyst is high rare earth,high zeolite to matrix (Z/M) ratio,and high equilibrium unit cell size(UCS) designed for maximumgasoline selectivity. Ecat metalslevels are moderate, and thecatalyst has a relatively high Ecatactivity (MAT). The yields for theFCC feeds from the NiMo catalyst inHDS and PNA modes with Type 1Ecat are shown in Figure 15, andare presented at constant cokeyield (2.0 wt.%).

The trends indicate that the PNAmode of pretreater operation resultsin higher overall conversion (themaximum value on the y-axis).Selectivity of gasoline and LPG alsoincrease, while dry gas yield isessentially unchanged. The lowerPNA and nitrogen contents of theFCC feed generated by the PNAmode of operation result in higherFCC conversion, and production ofmore valuable products.

Figure 15FCC Yields for NiMo Treated Feed Reacted over Type 1 Ecat

53.1 55.1

11.7

12.4

6.0

6.31.55

1.56

40.0

45.0

50.0

55.0

60.0

65.0

70.0

75.0

80.0

Dry Gas

C3's

C4's

Gasoline

77.4

74.5

HDS Mode PNA Mode

FCC

Yie

lds

(wt.

%)

Table IIIEquilibrium Catalyst Properties - Type 1

MAT (wt.%) 76Total Surface Area (m2/g) 152Zeolite Surface Area (m2/g) 120Matrix Surface Area (m2/g) 32Nickel (ppm) 183Vanadium (ppm) 1527Unit Cell Size (Å) 24.34Rare Earth (wt.%) 3.72Alumina (wt.%) 42.4

Figure 16C3 & C4 Olefinicity of FCC Products from

NiMo Treated Feed Reacted over Type 1 Ecat

0.76

0.77

0.78

0.79

0.80

0.81

C3

Ole

fin

icit

y

0.42

0.43

0.44

0.45

0.46

C4

Ole

fin

icit

y

C3 Olefinicity - NiMo


HDS Mode PNA Mode


Figure 16 shows that olefinicity ofboth C3's and C4's is lower in PNAmode than HDS mode, whichsupports previously observedtrends for hydrotreated feeds10.Hydrotreated FCC feeds are morehydrogen rich, and the resultingproducts are therefore morehydrogen rich.

Comparing the performance of theFCC feed produced by the CoMocatalyst, shown in Figure 17, showslower overall conversion and lowerselectivities for gasoline, C3's andC4's than the NiMo produced feed.Again, operating the hydrotreater inPNA mode (i.e. higher severity)instead of HDS mode results inhigher conversion and selectivitiesfor gasoline and LPG. This is due tothe increased HDN and PNAremoval at higher temperature.Both NiMo and CoMo will removemore nitrogen and PNA's from theFCC feed at the elevatedtemperatures of the PNA mode,although the NiMo catalyst removessignificantly more compared to theCoMo catalyst. This results insignificantly higher FCC conversion(1.4 wt.% in either mode) for theNiMo than the CoMo catalyst.

Similar trends showing reduced C3

and C4 olefinicity for the hydrotreaterin PNA operating mode are observedfor CoMo produced FCC feed (Figure18). Overall the C3 and C4 olefinsproduced by FCC feed from the twohydrotreating catalyst types areessentially equal.

Applying typical Gulf Coasteconomics to the FCC yields allowsa comparison of the valueassociated with each pretreatercatalyst type and each operatingmode. In the following figures, theNiMo produced FCC feed in HDSmode is used as the base caseassuming a 50,000 BPD FCC unit.Figure 19 shows the observedtrends in FCC product value forType 1 Ecat.

The NiMo catalyst produces FCCfeed that results in $0.19 to$0.23/BBL higher total product

Figure 17FCC Yields for CoMo Pretreater Feed Reacted over Type 1 Ecat

52.3 53.6

11.311.7

5.86.0

1.541.51

40.0

45.0

50.0

55.0

60.0

65.0

70.0

75.0

80.0

Dry Gas

C3's

C4's

Gasoline

73.1

HDS Mode PNA Mode

74.9

FCC

Yie

lds

(wt.

%)

Figure 18C3 & C4 Olefinicity of FCC Products from NiMo and CoMo

Pretreater Feeds Reacted over Type 1 Ecat

0.76

0.77

0.78

0.79

0.80

0.81

C3

Ole

fin

icit

y

0.42

0.43

0.44

0.45

0.46

C4

Ole

fin

icit

y

C3 Olefinicity - CoMo


C4 Olefinicity - CoMoC4 Olefinicity - NiMo

HDS Mode PNA Mode

Figure 19Delta Economics for FCC Products from NiMo and CoMo Pretreater

Feeds Reacted over Type 1 Ecat

$(0.40)

$(0.30)

$(0.20)

$(0.10)

$-

$0.10

$0.20

$0.30

$0.40

$0.50

Del

ta P

rod

uct

Val

ue

($/B

BL

) NiMo

CoMo

HDS Mode PNA Mode


value than the CoMo catalyst. Inaddition, the PNA operating modeyields $0.37 to $0.41/BBL higherproduct value than HDS mode.Note that this excludes any valueassociated with lower sulfur FCCproducts which result from the lowersulfur content of feeds treated inPNA operation. However, theseproduct value benefits of operatingin PNA mode must be consideredagainst the higher hydrogenconsumption cost and theassociated costs of the reduction inthe hydrotreater run length resultingfrom higher severity operation, asdiscussed previously.

Even after adjusting these productvalues to account for increasedhydrogen consumption, the NiMoproduced FCC feed in PNA modestill yields higher product values.Accounting for the higher hydrogenconsumption cost for NiMo relativeto CoMo, the delta FCC productvalue for NiMo catalyst is $0.14/BBLhigher than the CoMo product valuein PNA mode (Figure 20), and $0.16to $0.39/BBL higher than eithercatalyst in HDS mode.

To determine the point at which theincremental conversion fromincreasing pretreater severitydiminishes from reduced PNAconversion and/or unfavorablehydrotreater operating costs for thefeed in Table I, additional modelingwas conducted using the pilot plantresults. Starting with an FCC feedproduced by a pretreater with NiMocatalyst, assuming a 50,000 BPDpretreater with a catalyst change outcost of $7 MM and a four-year FCCunit shutdown cycle, running inhigher severity mode would causeadditional pretreater outage andhydrogen consumption costs. Asshown in Figure 21, higher FCCproduct value can be obtained byrunning the hydrotreater at higherseverity. The HDS operating modeand Type 1 Ecat represents the baseoperation. Increasing the severity ofthe hydrotreater operation towardoptimum PNA mode results in higheroperating costs for the hydrotreater,which are more than justified by the

Figure 20Delta Economics for FCC Products from NiMo and CoMo Pretreater Feeds Reacted over Type 1 Ecat Adjusted for

Pretreater Catalyst Hydrogen Consumption

Del

ta P

rod

uct

Val

ue

($/B

BL

) NiMoCoMo

Adjusted NiMoAdjusted CoMo

$ 0.50

$ 0.40

$ 0.30

$ 0.20

$ -

$ (0.10)

$ (0.20)

$ (0.30)

$ (0.40)

$ 0.10

HDS Mode PNA Mode

Figure 21Economics for FCC Products and Hydrotreater Severity

for NiMo Pretreater Feed Reacted over Type 1 Ecat

Delta Product Value ($/BBL)

Higher Severity Cost ($/BBL)



$(1.60)

$(1.40)

$(1.20)

$(1.00)

$(0.80)

$(0.60)

$(0.40)

$(0.20)

$-

$0.20

$0.40

Del

ta F

CC

Pro

du

ct V

alu

e ($

/BB

L)

$(1.60)

$(1.40)

$(1.20)

$(1.00)

$(0.80)

$(0.60)

$(0.40)

$(0.20)

$-

$0.20

$0.40

Hig

her

Sev

erit

yC

ost

($/

BB

L)

IncreasingHydrotreater

Severity

BaseHD

OptimumPNA

$(1.60)

$(1.40)

$(1.20)

$(1.00)

$(0.80)

$(0.60)

$(0.40)

$(0.20)

$-

$0.20

$0.40

Del

ta F

CC

Pro

du

ct V

alu

e ($

/BB

L)

$(1.60)

$(1.40)

$(1.20)

$(1.00)

$(0.80)

$(0.60)

$(0.40)

$(0.20)

$-

$0.20

$0.40

Hig

her

Sev

erit

yC

ost

($/

BB

L)


Severity

S

OptimumPNA

Table IVEquilibrium Catalyst Properties - Type 2



higher FCC product value. At theoptimum PNA point, $0.10/BBLhigher profit would be realized afteraccounting for the higher cost ofrunning the hydrotreater in that mode.Beyond the optimum PNA point,however, the incremental FCCproduct value begins to level off, andeventually begins to decline due tothe higher PNA levels in the feed (aswas observed in Figure 3). Theincreasing cost of higher severityoperation of the hydrotreater is notjustified as it more than offsets theFCC product value realized.

FCC catalyst can also be designedto further enhance yield advantagesof hydroprocessing catalyst type orhydrotreater mode/severity ofoperation. As an example, theresulting yields for FCC feedproduced by NiMo catalyst in HDSand PNA operating modes overFCC equilibrium catalyst Type 2were investigated. The propertiesof the Type 2 Ecat are shown inTable IV.

Type 2 catalyst has moderate rareearth, with a lower zeolite to matrix(Z/M) ratio and lower unit cell sizethan Type 1 catalyst. Shifts in Z/Mand UCS will result in a yield slatewith more olefins and octane thanType 1. For some hydrotreated feeds,Type 2 catalyst will yield higheroverall conversion (see Figure 22), asthe matrix contribution from thecatalyst will crack more of the heavierfeed molecules.

Comparing Figure 22 to Figure 15,switching to Type 2 FCC catalystboosts conversion by 1.9 wt.% inHDS mode and 2.7 wt.% in PNAmode. Compared to Type 1 catalystin HDS mode, the higher conversionfor Type 2 results in $0.66/BBLhigher product value, and thatincreases to $0.89/BBL higherproduct value in PNA mode (Figure23).

Figure 24 shows the product valueincrease along with the costsassociated with the higher severitypretreater operation with Type 2catalyst. In this case, there

Figure 22FCC Yields for NiMo Pretreater Feed Reacted over Type 2 Ecat

53.155.5

13.2

13.2

7.1

7.01.69

1.58

40.0

45.0

50.0

55.0

60.0

65.0

70.0

75.0

80.077.2

79.2

Dry Gas

C3's

C4's

Gasoline

FCC

Yie

lds

(wt.

%)

HDS Mode PNA Mode

Figure 23Delta Economics for FCC Products from NiMo Pretreater Feed

Reacted over Type 2 vs. Type 1 Ecata

$(0.40)

$(0.20)

$-

$0.20

$0.40

$0.60

$0.80

$1.00

$1.20

Del

taP

rod

uct

Val

ue

($/B

BL

) Type 2

Type 1

HDS Mode PNA Mode

Delta Product Value overType #1 HDS Mode.

Figure 24Economics for FCC Products and Hydrotreater Severity

for NiMo Pretreater Feed Reacted over Type 2 Ecat

$(1.60)

$(1.40)

$(1.20)

$(1.00)

$(0.80)

$(0.60)

$(0.40)

$(0.20)

$-

$0.20

$0.40

Del

ta F

CC

Pro

du

ct V

alu

e($

/BB

L)

$(1.60)

$(1.40)

$(1.20)

$(1.00)

$(0.80)

$(0.60)

$(0.40)

$(0.20)

$-

$0.20

$0.40

Hig

her

Sev

erit

yC

ost

($/

BB

L)




Severity

BaseHDS




Severity

OptimumPNA


appears to be a smaller window ofpretreater operation where the gainin product value exceeds theincreasing operating costs. Thepoint at which the cost of higherseverity operation for the pretreateroverwhelms the incremental FCCproduct value is closer to the baseHDS operating mode, so the refinerrunning Type 2 FCC catalyst couldmake up to $0.12/BBL more profit atthe optimum PNA point, but theoperating window is much narrowerthan it was for Type 1 catalyst.Obviously the economics of eachcombination of FCC catalyst,pretreater catalyst, and hydrotreaterseverity are unique to each refinery.

For the last comparison case, theresulting yields for an FCC feedproduced by NiMo hydrotreatingcatalyst in HDS and PNA operatingmodes over FCC equilibriumcatalyst Type 3 are presented. Theproperties of the Type 3 Ecat areshown in Table V.

Type 3 catalyst properties look similarto Type 1, but Type 3 has a higherZ/M ratio. Type 3 also has slightlyhigher rare earth levels and a UCSthat favor gasoline selectivity. Themetals levels are appreciably lowerthan Type 1 and the MAT activity ishigher. For the feed produced by theNiMo catalyst in PNA or HDS mode,Type 3 catalyst yields significantlyhigher conversion than Type 1 (2.7wt.% on average over yieldspresented in Figure 18), and slightlyhigher conversion than Type 2 (0.4wt.% on average over yieldspresented in Figure 22). However, asFigure 26 shows, the delta productvalues for Type 2 catalyst are higherdue to the higher olefins yield for thistype of catalyst.

Modeling Type 3 catalyst yields andselectivities for increasing severityof pretreater operation shows thatType 3 will increase in delta productvalue and then level off across therange of severity analyzed (Figure26). The benefits of increasedseverity level off sooner for Type 2than Type 3 catalyst, and then Type

Figure 25FCC Yields for NiMo Pretreater Feed Reacted over Type 3 Ecat

53.8 55.5

13.113.5

6.9

7.11.67

1.61

40.0

45.0

50.0

55.0

60.0

65.0

70.0

75.0

80.077.5

79.7

Dry Gas

C3's

C4's

Gasoline

FCC

Yie

lds

(wt.

%)

HDS Mode PNA Mode

Table VEquilibrium Catalyst Properties - Type 3


Figure 26Delta Economics for FCC Products from NiMo Pretreater Feed

Reacted over Type 1, Type 2 and Type 3 Ecat

$0.40

$0.50

$0.60

$0.70

$0.80

$0.90

$1.00

Del

ta P

rod

uct

Val

ue

($/B

BL

)

Type #2

Type #3

Increasing Hydrotreater Severity

Type #2

Type #3

Type #2

Type #3

HDS Mode PNA Mode

Delta Product Value overType #1 HDS Mode.


2 catalyst product value actuallydeclines with the further increase inseverity while Type 3 catalystproduct value levels off. The highermatrix surface area of Type 2catalyst creates diminishing returnson FCC conversion as the pretreaterseverity increases (simulating thepretreating catalyst progressingtoward EOR). This is because thedeclining pretreater product qualityat end of run causes higher FCCcoke yield and Type 3 catalystdesigned with the higher Z/M ratiowill be more coke selective with thisfeed trend.

Figure 26 indicates that there aresignificant opportunities to adjustFCC catalyst formulations aspretreater cycles move from start ofrun to end of run, however, the datapresented are specific to the feed inTable I, the FCC catalyst typestested, and the range of severitiestested. Each refinery operationrequires individual analysis ofcandidate catalysts and operatingconditions of both FCC feedpretreater and FCC units todetermine the optimum combinationwhich maximizes profitability.

Conclusion

FCC feed pretreating offers thepotential to significantly improverefinery economics. Both thehydrotreating catalyst system andthe operating strategy for thepretreater are critical to providingthe highest quality feed for the FCC.In general, NiMo based catalystsproduce better FCC feed, with bothlower nitrogen and PNA content.The operating mode of thehydrotreater can be used to furtherimprove the FCC feed. Driving thehydrotreater to remove nitrogen andPNA's improves FCC product value,but this needs to be balancedagainst the increased costs ofhigher hydrogen consumption andshorter cycle length that result fromthis mode of operation.

The shorter run lengths fromoperating in PNA mode can beextended with the use of a sulfur

reduction catalyst or additive in thedownstream FCC unit. FCC gasolinesulfur reduction of 20% or higher froma product such as D-PriSM® or GSR®-5additives, or SuRCA® catalyst wouldenable the pretreater to run beyond thepoint when product sulfur levels exceedtarget levels, while FCC gasolineproduct sulfur remains within the limitsneeded for gasoline pool sulfurcompliance. Extending operation inPNA mode closes the gap in run lengthversus a hydrotreater in the HDS modeof operation, while continuing toprovide FCC feed of higher quality thanHDS mode.

FCC catalyst formulation can also beincluded as one of the variables that canbe adjusted to maximize profitability.With constant hydrotreater productproperties, FCC catalyst formulation canprovide up to $0.90/BBL higher productvalue. Pushing hydrotreater severity toan optimum with constant FCCcatalyst can yield in excess of$0.10/BBL net product value, whichincludes the cost of higher severityoperation of the hydrotreater. Finally,adjusting FCC catalyst formulation asthe hydrotreater moves from start ofrun to end of run can significantlyoffset decline in product quality, andpreserve product value.

All of the combinations presentedshow the need for refiners to follow anintegrated approach to managing thecatalysts and operation of the FCCpretreater and FCC units. FCC andhydroprocessing operation can becontinuously optimized throughout thecourse of the hydrotreater run, tosignificantly increase refiner revenue.

References

1. Wollaston, E. G., Forsythe, W. L.,and Vasalos, I. A., Oil & Gas Journal, August 2, 1971, pp. 64-69.

2. Huling, G. P., McKinney, J. D., and Readal, T. C., Oil & Gas Journal,May 19, 1975, pp. 73-79.

3. Campagna, R. J., Krishna, A. S.,and Yanik, S. J., Oil & Gas Journal,October 31, 1983, pp. 129-134.

4. Krenzke, L. D., and Baron, K., 1995NPRA Annual Meeting, Paper AM-95-67.

5. Shiflett, W. K., Krenzke, L. D., and Wear, C. C., Davison Catalagram 89,October, 2001, pp. 11-17.

6. Shiflett, W. K., Olsen, C. W, Watkins,B. R., Krishnamoorthy, S., and Wear,C. C., 2002 NPRA Annual Meeting Paper, Paper AM-02-39.

7. Blanchard, L., Borchert, C., Pu, M.,2007 NPRA Annual Meeting Paper,Paper AM-07-06.

8. Blanchard, L., Oishi, T., Teo, B.,Haley, J., Davison Catalagram 98,Fall 2005, pp. 5-10.

9. Blanchard, L., Davison Catalagram 100, Fall 2006, pp. 8-13.

10. Salazar-Sotelo, D., Maya-Yescas, R.,Mariaca-Dominguez, E., Rodriguez-Salomon, S., Aguilera-Lopez, M.,Catalysis Today 98, 2004, pp. 273-280.


ncreasingly stringent gasolinesulfur regulations such as the 30ppm sulfur limit imposed on

most U.S. refiners in 2006 present yetanother challenge to the refiningindustry. Refiners not only have tocomply with the new specifications butthey need to determine the mosteconomical method to reduce gasolinesulfur while preserving as muchoperating flexibility as possible. Thefluid catalytic cracking (FCC) unit is asignificant contributor to the gasolinepool and one of largest sources ofgasoline sulfur. Therefore, reducingFCC gasoline sulfur is a key componentof a refiner's overall compliancestrategy. While pre-treating the feed orpost-treating the FCC gasoline arepopular approaches, the use ofcatalysts and additives for sulfurreduction can provide increasedflexibility and improved profitability.

Grace Davison first commercializedgasoline sulfur reduction catalysts andadditives in 1996 and has beensuccessful in introducing newproducts to the market. Theseproducts have been used in 85commercial applications and havedemonstrated the ability to reduceFCC gasoline sulfur by up to 46%.

When combined with undercutting,gasoline sulfur reduction approaches60%. Refiners have found numerousways to incorporate these productsinto their short-term and long-termstrategies to:

• Improve profitability • Extend feed hydrotreater

run-life• Manage FCC feed

hydrotreater outages• Process higher sulfur

opportunity feeds • Increase FCC feed

hydrotreater throughput• Minimize undercutting• Comply with pipeline

specifications• Minimize octane loss from

post treatment• Reduce hydrotreater

hydrogen consumption

Grace Davison's, SuRCA® catalyst,which is designed to replaceconventional FCC catalyst, hasbeen Grace Davison's mostsuccessful gasoline sulfur reductionproduct. SuRCA catalyst has beenused in over 45 applicationsworldwide. In addition to reducing

Catalytic FCC Gasoline Sulfur Reduction Citgo's Corpus Christi refinery reduces FCC gasoline sulfur byup to 31% using Grace Davison's GSR®-5 additive


Lauren BlanchardStrategic BusinessMarketing Manager,Grace Davison,Columbia, MD

Michael ZiebarthResearch Manager,Grace Davison,Columbia, MD

Timothy DouganMarketing Manager,Grace Davison,Columbia, MD

I

gasoline sulfur, it has, in someinstances, reduced LCO sulfur by10-15%. Grace Davison also offersGSR-5® gasoline sulfur reductionadditive, which is based on similartechnology embodied in SuRCAcatalyst. This product providescomparable performance with theincreased flexibility and theconvenience of an additive.Typically, GSR-5 additive is used inplace of 25% of catalyst additions.Grace Davison has had sevencommercial applications of GSR-5.Citgo Petroleum Corporation'sCorpus Christi, TX refinerysuccessfully used the additive forover a year and was able to reducegasoline sulfur by up to 31%.Results from this application aresummarized in this paper.

The Origin of FCC GasolineSulfur

The proper application of existingproducts, as well as thedevelopment of new technologies,requires an understanding ofgasoline sulfur formation pathways,the species that are formed and theeffect of FCC feed sulfur speciationon the resulting gasoline sulfur.

Sulfur compounds in FCC feedinclude mercaptans, sulfides, alkylsubstituted thiophenes, thiophenols,benzothiophenes and multi-ringaromatic thiophenes. The amount

and relative quantity of each of thesespecies is dependent on a wide rangeof factors. These factors include feedsource, type of feed pretreatment(hydrotreating, etc.), and whether ornot other refinery streams such ascoker gas oil have been blended intothe FCC feed.

In the FCC unit, typically 35-45% ofthe feed sulfur is converted into H2S,45-55% ends up in the light cycle oil(LCO) and bottoms and 5% in thecoke. Only about 5% ends up in thegasoline. Gasoline sulfur species canbe determined by using a GasChromatograph (GC) equipped withan Atomic Emission Detector (AED).The typical sulfur compounds found inFCC gasoline are presented in Figure27 as a function of their boiling point.The gasoline range sulfur species

include mercaptans, sulfides, alkylsubstituted thiophenes, thiophenols,and, depending on the gasoline cutpoint, benzothiophene andalkylbenzothiophenes.

The importance of the feed sulfurtype on the amount of gasolinesulfur produced is shown in Figure28. Two FCC gas oil feeds withsimilar sulfur levels were cracked inan FCC pilot plant and the resultinggasoline analyzed for sulfur. Thegasoline produced from Gas Oil Bhas about 75% higher sulfur thanthe gasoline produced from Gas OilA, despite the fact that both startinggas oils had similar sulfur levels.

The two gas oils were analyzed byHigh Resolution Mass Spectrometry(HRMS) and X-ray PhotoelectronSpectroscopy (XPS) to determinethe types of sulfur species present.The HRMS determines the type andamount of aromatic thiophenespresent by measuring the quantityand mass of the molecular species.XPS measures the binding energyof the feed sulfur species andallows quantification of the amountof aromatic and aliphatic sulfurspecies. An analysis of the twofeeds by these methods shows thedifferences in the types of sulfurspecies present in each feed (TableVI). Gas Oil B has about twice thelevel of alkylthiophenes, twice thelevel of aliphatic sulfur species andabout 10% higher alkyl-benzothiophenes, compared to GasOil A. Gas Oil A has more multi-ringbenzothiophenes.

Figure 27Gasoline Sulfur Speciation Determined by Simulated Distillation

122 167 212 257 302 347 392 437 482 527

Temperature, (˚F)

Figure 28Sulfur Species From Two Cracked Gas Oils

100 150 200 250

Boiling Point ˚F

300 350 400 450 500

1400

1200

1000

800

600

400

200

0

Cu

mu

lati

ve S

ulf

ur

(pp

m)

Gas Oil A 2.8 wt.% S

Gas Oil B 2.6 wt.% S


the biggest contributors to gasolinerange alkylthiophenes, while the otherfeed sulfur molecules are more modestcontributors. The alkylbenzothiopheneand multi-ring thiophenes are the mainsource of benzothiophene which maynot appear in the FCC gasoline,depending on the end-point.

Understanding the mechanisms bywhich gasoline sulfur is formed hasallowed Grace Davison to develop arange of different products that targetthe different sulfur producingpathways.

Based on the feed sulfur analysis,experiments were designed todetermine if feed alkylthiophenes,alkylbenzothiophenes andsaturated sulfur compounds aremore likely to end up in the gasolineas compared to multi-ring alkyl-benzothiophenes. A hydrotreated,low sulfur gas oil was spiked withone of four model sulfurcompounds; a mercaptan, an alkyl-thiophene, an alkylbenzothiopheneand an alkyldibenzothiophene.These feeds were cracked in anFCC pilot plant and the resultinggasoline was analyzed for sulfur.The results are shown in Figure 29.The data shows that thealkylthiophenes in the feedcontribute significantly to gasolinerange sulfur species. The alkylchains on the thiophene arecracked and isomerized to producea range of alkylated thiophenespecies. The alkyl group can alsocyclize and dehydrogenate to formbenzothiophene. The mercaptansare cracked mainly to H2S and onlya small quantity ends up asgasoline range sulfur. The alkyl-benzothiophene and alkyl-dibenzothiophene produce somebenzothiophene but essentially nogasoline range alkylthiophenes.

In summary, this data shows that(C5+)-alkylthiophenes in the feed are

Catalytic Gasoline SulfurReduction Options

D-PriSM® sulfur reduction additive,another product from GraceDavison, is effective at removing anintermediate that converts toalkylthiophene and therefore ismore effective at reducing lightgasoline sulfur. This product hasbeen used in more than 25refineries worldwide. The SuRCA®

catalyst family, GSR-5 additive andthe recently commercializedNEPTUNETM catalyst are designedto reduce a broader range ofgasoline sulfur species (Figure 30).

When formulating a solution for acustomer, we consider the FCCgasoline speciation, the desiredlevel of gasoline sulfur reduction,and whether a catalyst or anadditive is preferred. Carefulselection of the appropriatetechnology for each application hasresulted in consistent productperformance that has met orexceeded customer expectations.As a result, these technologies tendto be used on a long-term,continuous basis. The averagelength for all 85 commercialapplications is 388 days, with oneuser at 6 years. Over one-half ofthe applications have lasted 100

Table VISulfur Characteristics of Gas Oil A and B

Figure 29 Gasoline Range Sulfur Distribution from Feedstock

Spiked with Model S Compounds

1200

1000

800

600

400

200

0

Su

lfu

r, p

pm

C2-Thiophenesand Lighter

C3-,C4-Thiophenes

Benzo-Thiophene

Mercaptan

Alkylthiophene

Alkylbenzothiophene

Alkyldibenzothiophene


Gas Oil A Gas Oil BSulfur Compounds byHigh Resolution Mass Spec (ppm)

C(N)H(2N-4)S

C(N)H(2N-10)S

C(N)H(2N-16)S

C(N)H(2N-22)S

THIOPHENES

BENZOTHIOPHENES

DIBENZOTHIOPHENES

NAPHTHOBENZOTHIOPHENES

526

13918

9439

1517

980

12702

7317

0

Total Sulfur XPS Analysis (wt.%)

ALIPHATIC SULFUR

AROMATIC SULFUR

0.23%

2.54%

0.50%

2.09%

days and over one-quarter used thetechnologies for more than a year.Grace Davison currently has 25gasoline sulfur reductionapplications worldwide.

Citgo Corpus Christi

Citgo, Corpus Christi needed toreduce FCC gasoline sulfur to keeptheir overall refinery gasoline pool incompliance. Citgo and GraceDavison worked together todetermine if catalytic sulfurreduction technologies couldachieve the necessary reduction ontheir 67,000 barrel per day UOPHigh-Efficiency unit. The unitprocesses 100% hydrotreatedvacuum gas oil and operates in fullcombustion. After carefulevaluation of their unit operationand objectives, and Citgo'spreference for an additiveapproach, GSR-5 additive wasrecommended. Citgo beganbaseloading the product to achieve25% in the circulating inventory.Maintenance dosage was also atthe recommended level of 25%additions in place of normalcatalyst additions.

Unit data, including feed quality,operating conditions, yields andyield quality was collected for eightmonths that encompassed both thebase period, prior to additiveadditions and the GSR-5 additiveperiod. Sulfur reduction on theheavy gasoline stream is shown inFigure 31. The data was normalizedfor changes in feed sulfur andendpoint. Sulfur reduction rangedfrom 21-31% depending onendpoint with an average reductionof 26%. Combining the effect ofundercutting with the GSR-5additive reduced gasoline sulfur by44%.

To supplement the data generatedby the refinery, FCC feed and heavygasoline samples were collected byCitgo and sent to Grace Davisonboth before and during the trial.The samples were analyzed by GasChromatograph (GC) to determinethe sulfur species present.

Figure 30 Grace Davison's Portfolio of Gasoline Sulfur Reduction Products

Preferred Technology

CATALYST

ADDITIVELight or Intermediate

Gasoline

Full RangeGasoline

Light & Intermediate

Solution

D-PriSM 10-35%reduction

35%reductionor more

20-35%reduction

Light, Intermediate and Full range gasoline

NEPTUNE

GSR -5SuRCA

gasoline

®

®

TM

®

Figure 31Normalized FCC Gasoline Sulfur for Citgo Corpus Christi

1800

1600

1400

1200

1000

800

600

425˚F D86EP 445˚F D86EP

21% Reduction

31% Reduction

44% Reduction observed whenGSR-5 & undercutting combined

Base

GSR-5

Gas

olin

eS

ulf

ur

No

rmal

ized

for

Fee

d S

Figure 32Gasoline Sulfur Reduction by Boiling Point

1000

900

800

700

600

500

400

300

200

100

0100 150 200 250 300 350 400 450 500 550

Sulfur Compound Boiling Point (˚F)

A

B

C

D

E

Average Distillation ˚F

Base

GSR-5

T95

414

413

FBP

456

460

GSR-5Base

A Thiophenes

B C2 Thiophenes

C C3Thiophenes

D C4Thiophenes

E Benzothiophenes


The performance of the GSR-5additive on individual sulfur speciesis shown in Figure 32. Both sets ofsamples were normalized to accountfor variation in feed sulfur. The baseline is steeper at the points A throughE than the GSR-5 additive lineillustrating the performance acrossthe entire range of sulfur speciespresent in the heavy gasoline. Totalgasoline sulfur reduction asmeasured by speciation is 30%.There were no adverse yield shiftswith the GSR-5 additive whenincorporated into Citgo's operation at25% of the unit inventory. Theproduct was baseloaded into the unitallowing for quick results. This isachievable because the GSR-5additive contains catalytic crackingfunctionality, which for Citgo resultedin a conversion increase of 0.7 lv%and slurry decrease of 0.5 lv%.

Citgo will use the GSR-5 additive forabout a year to help cost effectivelymeet pool gasoline sulfur limits untilalternative technologies can becommissioned. Then they willevaluate using GSR-5 on a long-term basis in conjunction with thehardware that is installed.

Citgo's experience with GSR-5additive technology is representativeof other commercial applications.Table VII contains a list of refinersthat have employed the product.With this additive, refiners can takeadvantage of FCC gasoline sulfurreduction on either a short or a long-term opportunity basis. Because theproduct is an additive, the applicationis independent of the base catalysttechnology used by the refiner.

Summary

Gasoline sulfur regulations arebecoming increasingly stringentworldwide. Since the FCC unit is such alarge contributor of sulfur to the gasolinepool, effective means of reducing FCCgasoline sulfur can significantly improverefinery profitability and flexibility.Catalysts and additives for FCC gasolinesulfur reduction have been commerciallyproven in 85 units worldwide.Understanding the pathways of FCCgasoline sulfur formation as well as thespecies present is crucial in selectingthe most appropriate technology. Properapplication leads to successful resultslike those achieved at Citgo's CorpusChristi refinery.

Table VIICommercial Users of GSR-5 Additive Technology


Refiner

A*

B*

C*

D*

E*

F*

G*

Region

NA

NA

NA

NA

NA

NA

NA

UnitType

UOP HIGH EFF

TEXACO DESIGN

UOP HIGH EFF

UOP SBS

UOP MOD V

EXXON FLEXI

EXXON FLEXI

% SulfurReduction

26

23

23

15

29

50

8

DaysUsed

382

398

368

96

231

196

196

Min

0.2

0.2

0.3

0.8

0.2

0.4

0.9

Max

0.5

0.4

0.5

1.6

0.4

1.2

2.3

Avg

0.3

0.3

0.3

1.4

0.3

0.8

1.3

Feed S, wt.%

Grace Davison environmental technologies

have helped refiners reduce FCCU emissions

for over a quarter of a century.

The FCCU is often the largest point source within the refinery for SOxand NOx emissions. The use of innovative catalytic technologiesreduces emissions from the FCCU regenerator, without the need forcapital intensive “end-of-pipe” hardware solutions.

Super DESOX® provides industry leading SOx removaleffectiveness. More and more refiners are finding that Super DESOXcan cost effectively control SOx emissions below 25 vppm from awide range of uncontrolled SOx baseline emissions.

Grace Davison researchers have studied the complexformation of NOx and developed two additives to reduce NOx.XNOx® is a low NOx combustion promoter designed toreplace conventional promoters, which often causeincreased NOx formation. For units not using a promoter,or requiring additional NOx reduction, DENOX® is theadditive of choice. NOx reductions in excess of 50%have been observed commercially with both products.With more than 50 commercial applications of NOxadditives, Davison has more experience than the restof the industry combined and would be happy toassist you in reducing FCCU NOx emissions.

Extensive customer-driven research efforts at Grace are providing new insights into improvingSOx and NOx removal. Want to find out more about our environmental technologies? Contact us at www.e-catalysts.com or call us at (410) 531-8226.Let Grace Davison help you meet your FCCU emission challenges.

Your proven FCCU emissions solution.Grace Davison:

Have questions about our clean fuels technology? Contact us at www.e-catalysts.com or call us at (410) 531-8226.

Let Grace Davison and Advanced Refining Technologies give you the custom catalytic solution to your clean fuels challenges.

Your Clean Fuels Solutions

For gasoline sulfur reduction, today's refiners are challenged morethan ever before to blend various refinery streams to meet stringentproduct specifications and government regulations for clean air.Proven in over 75 commercial units, Grace Davison's portfolio ofFCC gasoline sulfur reduction technologies includes the D-PriSM®

and GSR®-5 additives and the SuRCA®, SATURN® andNeptune™ catalyst families.

>Recently commercialized, NEPTUNE catalyst further expandsour sulfur reduction technologies to continuously improve bothperformance and cost effectiveness. This next generationtechnology is a step out improvement over earlier technologies,providing 35-50% full range gasoline sulfur reductioncommercially with full catalyst formulation flexibility.

>For refiners with FCC pretreaters, the ApART Catalyst System™utilizing combinations of ART AT575, AT775 and AT792 offers theopportunity to significantly increase sulfur removal in thehydrotreater while at the same time maximizing FCC feed quality. Theimproved performance of the pretreater results in higher gasolinepotential in addition to decreasing FCC gasoline sulfur.

>For ULSD processing, the SmART Catalyst System™ utilizes state-of-the-art catalyst technology which is staged in the properproportions to provide the best performance, while at the sametime meeting individual refiner requirements. The catalyst stagingis designed to selectively take advantage of the different reactionmechanisms for sulfur removal with efficient hydrogen usage.ART CDXi, our newest generation of high activity CoMo catalyst,efficiently removes the unhindered, easy sulfur via the directabstraction route, while ART NDXi, our high activity NiMocatalyst, then attacks the remaining sterically hindered, hardsulfur. The SmART system provides higher activity than either thetraditional CoMo or NiMo catalyst alone while effectively helpingthe refiner manage hydrogen utilization.

When it comes to commercially proven technology to meet increasingly stringent

Clean Fuels regulations, make Grace Davison and Advanced Refining Technologies

your one-stop shop for FCC catalysts, additives and hydroprocessing catalysts.


Introduction

n most refineries, FCC gasolineis one of the largest, if not thelargest contributor of sulfur to

the overall gasoline pool. Regulationslimiting the amount of sulfur in gasolinehave highlighted the importance ofreducing sulfur in the FCC gasolinestream. Grace Davison has beenproviding catalysts and additives to theindustry that reduce FCC gasolinesulfur by 35% or more, for over tenyears. These technologies have beenproven in 80 units worldwide. Anotherpopular alternative to reduce gasolinesulfur is the installation of hardware.Increasingly stringent regulations haveforced refiners to install FCC feedhydrotreaters and gasoline post-treaters to meet ultra-low sulfur gasolineregulations. Interestingly, the phase-inof hardware has not eliminated thebenefit of sulfur reducing catalysts andadditives. Grace Davison's gasoline

sulfur reduction technologies arecomplementary to hardware solutions.They are being used at numerousrefineries around the world inconjunction with hardware to driverefinery profitability by providingfeedstock flexibility, minimizing octaneloss and providing operationalflexibility during hydrotreater outages.

Proper management of FCC feedhydrotreater outages is becomingincreasingly important as more andmore refiners rely on hydrotreatingto meet their per-gallon gasolinesulfur limits. About half of all FCCunits now have feed hydrotreaters.Some are run at higher severity thanin the past to achieve these newultra-low gasoline sulfur targets.Running at higher severity increasesthe frequency of turnarounds.Conventional methods of ensuringthat the gasoline pool stays below

Clean Fuels:An Opportunity for Profitability Using

Gasoline Sulfur Reduction Technology


ILauren BlanchardStrategic Business Marketing Manager,Advanced Refining Technologies,Columbia, MD

Craig BorchertValero Energy,Wilmington, CA

Min PuValero Energy,Wilmington, CA

the sulfur limit during thehydrotreater turnaround arepurchasing low sulfur feed orreducing FCC throughput. Eitherapproach can significantly reducerefinery profitability. Another optionis to use one of Grace Davison'ssulfur reducing technologies duringthe outage to provide feedstockflexibility while maintaining sulfurcompliance.

This paper presents a case study inwhich Grace Davison and ValeroEnergy’s, Wilmington, California, USArefinery worked closely together tominimize the impact of a FCC feedhydrotreater outage. Significantplanning and preparation took place,including selecting the bestpurchased feeds to run during theoutage as well as the best sulfurreduction technology to use to meettheir objectives given the sulfurspecies in the FCC gasoline. The useof Grace Davison's GSR®-5 sulfurreduction additive allowed Valero toprocess feed that was higher in sulfurthan their routine feed yet remain ingasoline pool sulfur compliance.Valero estimates that the use of GSR-5 additive saved them $1.7 millionduring the hydrotreater outage. Theresults were so encouraging thatValero has elected to use GSR-5additive on an on-going basis and

estimates increased profits of over $8million annually.

Evolution of Sulfur ReductionTechnologies

In 1992, well before the industryrealized the need to reduce FCCgasoline sulfur, Grace Davisonanticipated this requirement andbegan a research and developmenteffort for catalytic reduction of FCCgasoline sulfur. The objective was todevelop a family of FCC catalysts andadditives to help refiners meet cleanfuels specifications. Established

technologies that evolved from over14 years of continuous R&D includeD-PriSM® additive, SuRCA® catalyst,GSR-5 additive, and NeptuneTM

catalyst.

A road map for product selection isshown in Figure 33. Whenformulating a solution for acustomer, Grace Davison expertsconsider the FCC gasoline streamtargeted for sulfur reduction, thedesired level of gasoline sulfurreduction, and whether a catalyst oran additive is preferred. Carefulselection of the appropriatelyengineered solution for eachapplication has resulted inconsistent product performanceand meeting or exceeding customerexpectations. Twenty-three refinersare currently using Grace Davisongasoline sulfur reduction productsworldwide, making Grace Davisonthe leading supplier of gasolinesulfur reduction catalysts andadditives. Many of these refinershave been using these products ona continuous basis as part of theiroverall sulfur reduction strategy.(Figure 34).

For refiners who desire FCCadditives for maximum operatingflexibility, Grace Davison's D-PriSMadditive is effective at reducingsulfur species in light andintermediate FCC gasoline. It has

Figure 33 Grace Davison's Portfolio of Gasoline Sulfur Reduction Products

Preferred Technology

CATALYST

ADDITIVELight or Intermediate

Gasoline

Full RangeGasoline

Light & Intermediate

Solution

D-PriSM 10-35%reduction

35%reductionor more

20-35%reduction

Light, Intermediate and Full range gasoline

Neptune

GSR -5SuRCA

gasoline

®

®

TM

®

Figure 34 Grace Davison Experience in Sulfur Reduction Applications

51>100 days

>365days

# of Applications

Length ofApplications

27

10 20 30 40 500 60

TIME

388Days

5.9Years

Longest Application

(ongoing)

Average Length forall 85 Sulfur ReductionApplications Worldwide

85 Total Applications WW to Date => 25 Current Users


been used in more than 25refineries worldwide. D-PriSMadditive has provided up to 35%sulfur reduction on light FCCgasoline with no FCC yielddeterioration.

Grace Davison's SuRCA catalystfamily is designed to completelyreplace the conventional FCCcatalyst in the circulating inventory.This product provides gasolinesulfur reduction of up to 35% on fullrange gasoline while maintaining oreven enhancing existing yields andselectivities. Additionally, reductionsof 10-15% in LCO sulfur have beenobserved in some applications.Over 45 SuRCA catalyst ap-plications have occurred worldwide,with 10 current users havingemployed the technology for anaverage of more than three years.These refiners have incorporatedSuRCA catalyst into their operatingstrategies for long-term profitabilityand operating flexibility.

Appreciating refiners’ desire for aproduct that provides theperformance of SuRCA catalyst butcan be used as an additive formaximum operating flexibility, GraceDavison commercialized the GSR-5additive in 2004. It is based on theSuRCA catalyst chemistry andprovides similar gasoline sulfurreduction with base crackingcatalyst functionality. There arecurrently five refiners benefitingfrom the use of GSR-5 additive.

To further expand the sulfur reductionportfolio for continuous improvementsin both performance and costeffectiveness, Grace Davison hasrecently commercialized a newgasoline sulfur reduction catalystfamily. This next generationtechnology, Neptune, is a step outimprovement, providing 35-50% fullrange gasoline sulfur reductioncommercially with full catalystformulation flexibility.

Pre Turnaround Preparation

The Valero, Wilmington refineryapproached Grace Davison fourmonths prior to their scheduled FCCfeed hydrotreater shutdown to helpthem get a better understanding oftheir options during the outage. Valero

and Grace Davison worked togetherto determine if the use of a gasolinesulfur reducing technology wouldenable Valero to improve theireconomics and remain within theirFCC gasoline sulfur limit during theshutdown.

During the outage, the refineryplanned to purchase severalhydrotreated feeds. These wouldbe different than the pretreated feednormally charged to the FCC unit.The refinery planned to blendpurchased feeds with their routinefeed. The candidate feeds weresent to Grace Davison for testing tocompare the potential effects of thefeeds on the refinery's FCC yieldsand gasoline sulfur.

Analysis of the candidate feeds andValero, Wilmington's pretreatedRoutine Feed is shown in Figures35-37. Candidate feeds (SamplesA, B, and C) were heavier and morearomatic than the Valero WilmingtonRoutine Feed (Figure 36) makingthem more difficult to crack (Figure3). Additionally, the three samplefeeds contained significantly moresulfur and nitrogen species, whileconcarbon levels were similar to theRoutine Feed (Figure 37).

The four feeds were tested using arepresentative Valero, Wilmingtonequilibrium catalyst (Ecat) sample inthe pilot plant. The results

Figure 35Feed Distillation Profile

300

400

500

600

700

800

900

1000

1100

1200

0 10 20 30 40 50 60 70 80 90 100

Percentage

Bo

ilin

g P

t(°

F)

Sample A

Routine Feed

Sample B

Sample C

300

400

500

600

700

800

900

1000

1100

1200

0 10 20 30 40 50 60 70 80 90 100

Percentage

Bo

ilin

g P

t(°

F)

Sample A

Routine Feed

Sample B

Sample C

Figure 36Feed Properties

10

15

20

25

30

35

40

45

50

55

60

°API Aromatic Carbons,Ca (wt.%)

Napthenic Paraffinic Carbons,

Routine Feed

Sample A

Sample B

Sample C

Carbons, Cn (wt.%) Cp (wt.%)


suggested that all of the candidatefeeds would suppress conversionby at least 4 wt.% (Figure 38). Yieldsinterpolated at constant coke areshown in Table VIII. All feedsshowed the potential for increasedLCO and bottoms. These feeds alsoyielded significantly less gasolinewith slightly lower octane.

In addition to shifting yields towardless favorable products, the feedsalso increased gasoline sulfur.Samples B and C increased sulfurby 200%, while Sample A more thantripled FCC gasoline sulfur relativeto the Routine Feed (Figure 39).

The gasoline sulfur concentrationfor the Routine Feed produced inthe pilot plant is significantly lowerthan what is sent to blending fromthe Wilmington FCC unit. A numberof “tramp” gasoline streamsgenerated at other process unitsare currently processed in the FCCgas plant. These streams elevatethe apparent “FCC gasoline sulfur”as it is received in blending.

To help align the estimated FCCgasoline sulfur that would resultfrom processing the candidatefeeds with predicted commercialperformance, the gasoline sulfurspecies produced in the pilot plantusing the Routine Feed werecompared to the commerciallyproduced gasoline samples. Thecommercial gasoline samples wereproduced in the Valero WilmingtonFCC unit while processing twodifferent feeds (Figure 8). There isgood agreement for both individualsulfur species, and the relativeamount of each species presentwhen comparing the two sets ofsamples. Therefore, we canreasonably replicate the sulfurdistribution of the commerciallyproduced gasolines in the pilotplant. Cut gasoline sulfur is the sumof the species (mercaptans throughC4 thiophenes) and is higher for thecommercially produced gasoline,suggesting the presence ofadditional sulfur from the “tramp”streams.

Figure 37Feed Properties

Figure 38Feed Study Conversion vs. Coke

0.00

0.05

0.10

0.15

0.20

0.25

Sulfur (wt.%) Basic Nitrogen(wt.%)

Total Nitrogen ConradsonCarbon (wt.%) (wt.%)

Sample A

Routine Feed

Sample B

Sample C

Sample A

Routine Feed

Sample B

Sample C

64

66

68

70

72

74

76

78

80

82

84

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

Coke (wt.% Feed)

Co

nve

rsio

n (

wt.

%)

Sample A

Routine Feed

Sample B

Sample C

Sample A

Routine Feed

Sample B

Sample C

Table VIIIInterpolation at Constant Coke (3 wt.%)


Another potential reason fordifferences between the pilot plantgenerated gasoline sulfur and thatproduced commercially is themethod used to measure the sulfurconcentration. Refiners typicallymeasure gasoline sulfur by the bulkx-ray gasoline sulfur method. Themeasurement of gasoline sulfurproduced in the pilot plant utilizes agas chromatograph to identify theindividual sulfur species and totalsulfur concentration.

Sulfur species in the gasolinesproduced by the candidate feedswere compared to species presentin gasoline generated by theRoutine Feed. The candidate feedsall produced the same gasolinespecies as the gasoline generatedfrom the Routine Feed except inhigher concentrations (Figure 41).

Gasoline samples for eachcandidate feed in Figure 41 werenormalized to account for the deltabetween FCC-produced and pilotplant-produced methods (fromFigure 40). Additionally, data wasnormalized to an x-ray basis toreflect the levels of sulfur that wouldbe observed on the FCC unit.Finally, gasoline sulfur for eachcandidate feed was adjusted tomaintain the same gasoline sulfur tofeed sulfur ratio observed on theWilmington FCC unit - resulting inthe data in Figure 42.

The results from the pilot plantstudy, along with information on theValero, Wilmington FCC unitoperation (both routine and during aprevious pretreater outage) werethen used by Grace Davison toevaluate the performance of variousgasoline sulfur reductiontechnology options. The GSR-5additive, which contains basecracking functionality, wasdetermined to be the best solution.Based on the normalized species inthe gasoline produced by SamplesA through C, Grace Davisonestimated that the GSR-5 additivewould reduce gasoline sulfur by 19-

Figure 39Total Gasoline Sulfur vs. Coke

Figure 40Comparison of Commercially Produced and Pilot Plant Gasolines

0

4

8

12

16

Mer

capta

ns

Thiophen

e

Met

hylThio

phenes

Tetra

hydro

Thiophen

e

C2-Thio

phenes

Thiophen

ol

C3-Thio

phenes

Met

hylThio

phenol

C4-Thio

phenes

Cut Gasolin

e Sulfu

r

Su

lfu

r (p

pm

)

1 - FCC Commercial Sample2 - FCC Commercial Sample1 - Pilot Plant2 - Pilot Plant

Includes Tramp Streams

Figure 41Pilot Plant Produced Gasolines

0

5

10

15

20

25

30

35

40

Mer

capta

ns

Thiophen

e

Met

hylThio

phenes

Tetra

hydro

Thiophen

e

C2-Thio

phenes

Thiophen

ol

C3-Thio

phenes

Met

hylThio

phenol

C4-Thio

phenes

Cut Gasolin

e Sulfu

r

Su

lfu

r (p

pm

)

Routine Feed

Sample A

Sample B

Sample C

0

10

20

30

40

50

60

70

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

Coke (wt.% Feed)T

ota

lG

aso

line

Su

lfu

r (p

pm

)

Sample A

Routine Feed

Sample B

Sample C

Sample A

Routine Feed

Sample B

Sample C


23%. This narrow range reflectsextensive understanding of thecustomer's operation and the up-front pilot plant work.

The Valero, Wilmington refinery poolgasoline sulfur limit is 30 ppm. Therefinery also must produce gasolinebelow California NOx emissionslimits, which are influenced heavilyby the sulfur and olefins content ofthe gasoline streams blended intothe pool. The results of the pilotplant testing confirmed that therefinery would need to store RoutineFeed to blend with candidate feedsduring the pretreater outage to keepFCC feed sulfur levels low enoughto remain below their limits. Basedon the pilot plant results, Valeroconcluded that Feed A was too riskyin both gasoline sulfur and FCCyields/selectivities. The decisionwas made to purchase the otherfeeds and blend them at variousratios with available Routine Feedduring the outage.

GSR-5 Additive Application

As is common in many refineries,the FCC gas plant at Valero'sWilmington facility processesstreams from outside the FCC unit.These streams contain sulfur that isnot affected by the GSR-5 additivesince the additive works byparticipating in the crackingreactions that take place in the FCCunit. Unfortunately, the Wilmingtonrefinery sampling configurationdoes not allow for direct sampling of

the FCC gasoline prior to the inclusionof the “tramp” streams. The presenceof these streams in the gasolinesamples used to evaluate GSR-5additive performance reduces theapparent performance by reducingthe calculated percentage sulfurreduction.

Valero began use of GSR-5 additivetwo months prior to the 45-day feedhydrotreater outage. Coordinatedefforts with Grace Davison allowedValero to receive material andbaseload their inventory in 14 days. Ablend of the candidate feeds alongwith the Routine Feed was fed to theFCC prior to the outage, whichincreased feed sulfur by 20-35%.Additive additions proceededsmoothly and the projectedperformance was exceeded in lessthan 30 days with gasoline sulfurreduction of 20-25%. Figure 43

depicts a year's worth ofnormalized gasoline data vsendpoint. The three periodsrepresented are typical operation(Base Period), GSR-5 additivebefore and during the outage, andfinally GSR-5 additive following theoutage. Throughout the outage,Valero remained within their FCCgasoline sulfur limits whileprocessing all candidate feeds.After the outage, continued additionof GSR-5 additive allowed them torun 10-15% higher feed sulfur.

Economics

The candidate feeds also increasedthe FCC gasoline olefins content,which combined with the increasein the projected gasoline sulfurwould have forced Valero tohydrotreat approximately five MBPDof FCC gasoline to comply withCalifornia NOx emissionspecifications. The loss of octanefrom hydrotreating the FCC gasolinewould have reduced the amount oflow octane streams, such as LightStraight Run (LSR) and Heavy CatNaphtha (HCN), that could beblended into the pool. The sulfurreduction provided by the GSR-5additive allowed Valero to avoidhydrotreating the five MBPD FCCgasoline stream, the value of whichwas calculated to be $0.25/BBL or$1.7 million over the three monthperiod surrounding the outage.

Figure 43Normalized FCC Gasoline Sulfur vs. T95

GSR-5 provides 20-25%reduction

200

250

300

350

400

450

500

550

310 315 320 325 330 335

D86 T95 (°F)

Gas

olin

e S

ulf

ur

/ F

eed

Su

lfu

r (p

pm

/wt%

)

Base Period (Days 1-71)

GSR-5 & Outage (Days 90-115)

GSR-5 After Outage (Days 116-375)

Figure 42Normalized Pilot Plant Produced Gasolines

0

10

20

30

40

50

60

70

Mer

capta

ns

Thiophen

e

Met

hylThio

phenes

Tetra

hydro

Thiophen

e

C2-Thio

phenes

Thiophen

ol

C3-Thio

phenes

Met

hylThio

phenol

C4-Thio

phenes

Cut Gasolin

e Sulfu

r

Su

lfu

r (p

pm

)

Routine Feed

Sample A

Sample B

Sample C


After the pretreater was back on-line, Valero evaluated theeconomics of continued GSR-5additive usage. They determinedthat by continuing to use it, theycould consistently feed high sulfurVGO to their FCC feed hydrotreaterinstead of medium sulfur VGO andremain under the refinery poolgasoline sulfur limit of 30 ppm.Valero estimates the incrementalprofit for processing high sulfurVGO is $4.4 million per year (usinga conservative 3 cents per gallondifferential between high andmedium sulfur VGO). Accountingfor the cost of the GSR-5 additiveand incremental SOx additiverequired to remain in SOx emissioncompliance (higher FCC feed sulfuryields higher SOx emissions) thenet profit is $3.8 million.

The Wilmington refinery targets atwo-year cycle on the FCC feedhydrotreater. The cycle length isdetermined by the catalyst activity,which is influenced by operatingseverity and throughput. Valerodetermined that by using the GSR-5additive to control FCC gasolinesulfur, they could reducehydrotreater severity (even withhigher sulfur VGO feed to thehydrotreater), which allowed themto process more VGO through theunit. Valero was able to increasehydrotreater throughput by 4%. Theexcess hydrotreated FCC feed isperiodically sold at a premium overregular gasoil (using a seven centsper gallon differential betweenhydrotreated and regular gasoil) foran estimated annual profit of $4.5million.

The total benefit from the GSR-5additive for the Wilmington refineryis calculated to be $8.3 million peryear - a return of over 18 times theincremental cost of the GSRadditive technology.

The refinery has also determinedthat managing the tramp gasolinestreams using a different process,rather than processing them in theFCC gas plant, will allow them toachieve significant flexibility in their

gasoline pool blending operation.They plan to revamp an existing towerinto an LSR splitter, which will removeC3's and C4's before sending gasolinematerial directly to blending. Theoperation of the FCC unit is expectedto shift in favor of more light olefinswith the new equipment in service.The impact of the change in FCCoperation on FCC gasoline sulfur andolefins will be evaluated to determineif GSR-5 economics remain favorablewith the new process configuration.

Conclusion

Proper management of hydrotreateroutages is becoming increasinglyimportant as more and more refinersrely on hydrotreating to meet gasolinesulfur limits. Outages of either FCCfeed hydrotreaters or gasoline post-hydrotreaters create opportunities forrefiners to incorporate GraceDavison's sulfur reduction catalystsand additives into their planning. Thiswould allow for significant savings inpurchased feeds or mitigation of thecost of constraints caused by non-routine operation leading up to andduring the outage. With hydrotreatingequipment in service, thesetechnologies can generate significantrevenue for refiners who want tooptimize operations to driveprofitability. Reduction of FCCgasoline sulfur allows for higher feedsulfur to the FCC unit or to anupstream FCC feed hydrotreating unitwithout risking gasoline sulfur non-compliance. Lower FCC gasolinesulfur can also allow for reducedseverity operation on either the FCCfeed hydrotreater or the gasolinetreating units, creating revenue in theform of octane recovery, higherthroughput, or extended cycle life.

The Valero, Wilmington case studypresented here was made possible byincorporating discussions betweenValero & Grace Davison into theplanning stages of the FCC feedhydrotreater outage. Based on thosediscussions, pilot plant testing wascompleted which assisted Valero inselecting the purchased feeds theywould run during the outage.Estimates provided by Grace Davison

showed that the GSR-5 additivewould allow Valero to achieve theirshutdown objective of keeping theirgasoline sulfur in compliance whilerunning the higher sulfur purchasedfeeds. The use of GSR-5 additiveduring the outage resulted in $1.7million in savings.

With the proven productperformance of the GSR-5 additive,Valero was then able to optimizetheir operation once the FCC feedhydrotreater was put back inservice. The FCC feed hydrotreaterunit severity was reduced, allowingfor higher throughput, and theincremental hydrotreated gasoilwas sold at a premium over regulargasoil. Additionally, Valero was alsoable to feed high sulfur VGO inplace of medium sulfur VGO to theFCC feed hydrotreater withoutexceeding the FCC gasoline sulfurlimits. Both operating changesresulted in a combined profit of over$8 million per year for the refinery.

While each refinery configuration isunique, and the economicspresented here are specific to theValero Wilmington refinery, thisexample demonstrates that GraceDavison's gasoline sulfur reductionproducts can provide enhancedoperating flexibility in any operationand significantly improve refineryprofitability.

Acknowledgements: The authors wish tothank Valero Petroleum Company and GraceDavison for permission to release thisinformation. Natalie Petti (consultant), isgratefully acknowledged for her contributionof data evaluation and critique.


lurry exchanger fouling is oftenconsidered the worst foulingservice in the FCC process. (1)

The primary problem that results fromslurry exchanger fouling is reducedheat exchanger duty in the slurry/FCCfeed preheat exchanger or the steamgenerators. The reduction in feedpreheat temperature that can resultfrom just mild fouling of the FCCslurry/feedstock exchangers canresult in reduced unit feed rate orconversion.* Additionally, excessivepressure drop or inability to cool theslurry to the necessary rundowntemperature can also requirereducing the feed rate. Certainlyexcessive slurry exchanger foulingcan be very costly to the refinery interms of lost feed rate, lowerconversion and higher maintenanceexpenses.

The purpose of this article is topresent potential sources of slurryexchanger fouling and suggestions onhow to prevent or minimize fouling. We

Understanding and Minimizing FCC SlurryExchanger Fouling

will draw on industry experience fromseveral sources to present a broadreview of the subject. Our readersmay consider this information andcase study experience as they work tounderstand and minimize slurryexchanger fouling at their refinery.

Figure 44 illustrates a typical FCCMain Fractionator Slurry circuit.Superheated FCC product vapor isquenched as it enters the mainfractionator using the reflux from theslurry pumparound circuit. Slurryexchangers, which recover thisenergy by heating the feedstockand generating steam, are oftensubject to fouling through a numberof mechanisms. When slurryexchangers foul, feed rate orreactor temperature must often bereduced.

*For those FCC units that do not havea fired feed heater and are air blowerlimited.

Figure 44Typical Main Fractionator Slurry Circuit


SDavid A. Hunt

Technical Service Manager,Grace Davison,

Houston, TX

Bill Minyard National Technical Sales

Manager, Grace Davison,Houston, TX

Jeff Koebel Technical Service Manager,

Grace Davison,Chicago, IL

Figure 45 shows how quickly aslurry/feed exchanger heat transfercoefficient can deteriorate. Over aten-week period, this refiner neededto clean their FCC slurry exchangerthree times. During each cleaningcycle they were forced tosignificantly reduce feed rate.

Potential causes of FCC slurryexchanger fouling are shown inFigure 46. Fouling can be generallyclassified as either “organic” or“inorganic” based. Several organicor inorganic slurry foulingmechanisms are possible. Each ofthese possible fouling sources willbe discussed and suggestions toprevent or reduce each type will bepresented.

ORGANIC BASED FOULING

Organic based fouling is the mostcommon fouling type. The potentialcauses of organic based foulingare very broad. It is helpful toclassify organic fouling into twogeneral sub-types: “hard” and “soft”coke fouling.

Hard Coke Fouling

Solid coke fragments circulating tothe slurry exchanger tubes thatrestrict the flow through theexchanger is an example of hardcoke fouling. This type of foulingresults in excessive pressure dropand a loss of heat transfer duty.These shiny coke fragmentsgenerally accumulate on theexchanger tube sheets at the inlet tothe tubes. Figure 47 shows howpieces of coke can block theexchanger tubes, increasingexchanger pressure drop.

These coke fragments can originatein the reactor overhead line or themain fractionator. The coke oftenbecomes dislodged following anFCC shutdown because of thethermal cycling of the surface thatthe coke is adhered to. If the cokefragments are small enough to passthrough the suction strainers on theslurry pumparound pumps, they

Figure 45FCC Slurry Exchanger Heat Transfer Coefficient Deterioration

15

20

25

30

35

40

45

Hea

t T

ran

sfer

Co

effi

cien

t, B

TU

/hr

- ft

2 -

F

6/1 7/21 9/9 10/29 12/18 2/6

ExchangerCleaned

ExchangerCleaned

Exchanger Cleaned

Figure 46Root Causes of Slurry Exchanger Fouling

Slurry Exchanger

Fouling

CorrosionProducts

Metals

Inorganic Organic

Coke Deposits

PrecipitatedAsphaltenes

Polymerization

Catalyst Particulates

Al & Si


Figure 47Hard Coke Deposits

Photo Courtesy of: GE Water and Process Technologies

Photo Courtesy of:GE Water and Process Technologies

can eventually foul the first slurryexchanger in the pumparound loop.Another scenario that can result isexcessive hard coke accumulationin the bottom head of the maincolumn that restricts the suction ofthe circulating slurry pumps.

Smaller coke fragments that passthrough the exchanger tubes canstill be problematic. These cokeparticles are either smaller cokeparticles from the main fractionatoror are formed by polymerizationreactions in the slurry pumparoundcircuit at high main fractionatorbottoms temperatures. Small cokefragments can settle onto the tubesurface and further polymerize,resulting in a barrier to heat transferand slurry flow.

In some cases, formation of a verythin, hard layer of deposit has beenobserved on the tube walls. Thistype of deposit has a hard, shinyappearance similar to varnish.These deposits are formed bypolymerization reactions on thetube surface. This type of foulingcan reduce the heat transfercoefficient.

Preventing Hard Coke Fouling

Poor feed/catalyst contacting canbe a significant source of hard cokeformation in the vapor line. This isparticularly true in units thatprocess resid feeds. Feed /catalystcontacting can be improved inseveral ways:

• Improve feed atomization;• Increase feed dispersion

steam within the limits of the feed distributor design;

• Repair damaged feed distributors or replace with a more modern design;

• Increase the feed temperatureto avoid high feed viscosity atthe injection distributors. Thisis particularly applicable to resid operations.

• Increase the catalyst/oil mix zone temperature. This is most critical in resid or low reactor temperature operations where incomplete vaporization of the feed is morelikely. Mix zone temperature can be increased using Mix Zone Temperature Control (MTC) (2) and/or higher reactor and regenerator temperatures;

• Optimal catalyst matrix design can improve feedstock vaporization. (3)

A hot wall reactor vapor line must beproperly insulated to reduce thelikelihood of liquid condensation atcool spots. Insulation must beproperly anchored and should bewatertight. Once oil droplets form,they eventually dehydrogenate to formcoke. The reactor vapor line blindflange and all the vapor line pipesupports should also be insulated ifthe designs allow, as coke canaccumulate at these locations. Do notinsulate the bolts of the blind flange,however, as this can result in theflange opening due to bolt creep.

Reactor vapor line vapor velocitiesless than 100 fps should be avoided tominimize coke formation. Theexception to this is the velocity right atthe inlet nozzle to the main column,where lower velocities are permitted.(4)

Higher velocities will reduce thelikelihood of un-vaporized oil

accumulating along the wall of thevapor line and eventually formingcoke. Vapor line velocities between100 to 120 fps are a goodcompromise to minimize both cokeformation and pressure drop. (4)

Self-draining reactor overhead linesare a design feature often used tominimize coke formation andsubsequent slurry exchangerfouling.

The maximum main fractionatorbottoms temperature is typicallybetween 680˚F and 700˚F to avoidcoke formation and slurryexchanger fouling. The maximumsafe bottoms temperature for anyunit is unit and feedstock specific. Itis also important to note that thebottoms temperature is often baseda single temperature indicator (TI)at the outlet line of the mainfractionator. Poor pumparounddistribution and liquid mixing mayresult in locally hotter temperaturesthan what is actually measured inthe bottoms outlet line, increasingthe likelihood of coking. This canbe particularly true in operationswhere quench is used to cool thebottom of the main column.

Long liquid residence time in theslurry circuit will also influencecoking. Reducing bottoms liquidlevel can minimize slurry residencetime. Also note that increasing mainfractionator bottoms temperature

Figure 48Slurry Quench and Coke Trap in the slurry circuit

Quench

Coke Trap


and slurry pumparound rate bothincrease the liquid residence time.(5)

FCC feedstock, particularly feedcontaining resid, can form coke attypical main fractionator bottomtemperatures. As a result, refinersshould ensure that no feedstock isleaking into the slurry pumparoundcircuit through an emergency feedby-pass valve or slurry/feedexchanger.(5)

The slurry pumparound returnshould be properly distributed tominimize hot spots. Thepumparound rate should besufficient to ensure the grid zone iswell wetted. Local areas devoid ofliquid flow will allow hard coke toform. A minimum flux rate of 6gpm/ft2 has been recommended.(6)

A slurry pumparound rate of 1.2 to1.5 times the feed rate is a rule ofthumb that has also beensuggested to ensure good liquiddistribution in the bottom of themain fractionator.(7)

Slurry quench, as shown in Figure48, can be used to sub-cool themain fractionator bottomstemperature and reduce hard cokeformulation.(6) A quench distributorshould be used to minimize hotspots. Slurry quench, however, iscommonly injected from a nozzlethat terminates near the wall of thefractionator. As a result, thereturning quench liquid is often notwell distributed. Therefore, do notrely on perfect mixing of the slurry

quench, as zones of high temperaturecould still be present. The refiner mayconsider monitoring bottomstemperature on a quench-free basis toaccount for potential high temperaturezones.(6)

Maintaining slurry exchanger tubevelocities greater than 6 fps will helpminimize any settling of small cokefragments or FCC catalyst onto thetube service.(8) Units often havespillback valves on the slurry productexchangers to help maintain minimumslurry flows during times of turndownoperation. Tube velocities should beless than 10 fps to avoid erosion.(8)

Using an elevated slurry exit nozzlecan reduce the likelihood of cokebeing entrained into the slurrypumparound loop.(8) However, thisresults in the bottom head of the maincolumn filling with accumulated coke

and catalyst. On units where theliquid draw is on the bottom head, acoke trap should be employed inthe bottoms suction nozzle on thecolumn and/or upstream of theslurry exchanger as shown in Figure48 to trap any entrained cokeparticles before they can foul theexchanger. G. Walker discussedapplication of a coke trap andresulting reduction in slurryexchanger fouling.(9)

Soft Coke Fouling

Soft coke fouling is organicallybased where commonly aninsulating barrier is depositedinside the exchanger tubes,reducing the exchanger heattransfer coefficient. The insulatingbarrier can be found throughout thetube service.(10) Figures 49 and 50show examples of soft coke fouling.Just a thin layer of material canresult in a costly reduction of theexchanger heat transfer coefficient.In some cases, soft coke fouling canalso result in increased exchangerpressure drop.(10) Generally,however, reduced heat transfer isapparent before excessive pressuredrop with this type of fouling.

Precipitated asphaltenes are acommon source of soft cokefouling. Asphaltenes are highlycondensed polyaromatics typicallyinsoluble in a saturatedhydrocarbon such as heptane.(11)

The concentration of these multi-ring aromatics in the slurry can be

Figure 49Soft Coke Precipitation Fouling

5Photo Courtesy of:

GE Water and Process Technologies


Figure 50Soft Coke Precipitation Fouling

Photo Courtesy of:GE Water and Process Technologies

increased by thermal condensationreactions in the slurry circuit.

Asphaltenes can become insolublein the slurry oil and begin toprecipitate onto the tube surface.The tar-like layer on the tube servicecan also trap coke and catalystparticles that are present in theslurry.(12) Figure 51, an abbreviatedversion of Figure 46, illustrates thesequence. Analyzing such tubedeposits and slurry for fusedaromatics can be insightful. Higheramounts of fused aromatics in thedeposit relative to the slurry canconfirm asphaltene precipitation.(13)

Slurry viscosities increase at thetube wall due to the locally coolertemperature. The higher viscositycan cause material to adhere to thetubes, resulting in exchangerfouling.(12) Slurry with a higherparaffinic content may be moreprone to fouling due to theinherently higher viscosity.

Preventing Soft Coke Fouling

Two sources suggest minimizingasphaltene content of the slurry inorder to minimize fouling.(5,14) ASTMD3279 can be used to determineasphaltene content. Theasphaltene content is defined asthose components in the samplethat are not soluble in n-heptane.This procedure gives the combined

amount of asphaltenes andparticulate matter.

Maintaining asphaltenes in solution iskey to preventing soft coke fouling.The composition of the slurry willaffect the solubility of the asphaltenes.Generally higher aromatic content ofthe slurry tends to keep asphaltenesin solution. As such, changing theslurry composition by dropping someLCO down the tower can increase thesolubility of asphaltenes in thebottoms material. This also has thebenefit of reducing the temperature atthe bottom of the main column.

It is often necessary to adjust thebottoms composition and temperaturein this manner during changes in feedcomposition. For example, slurryproduced from paraffinic feedstockstends to be more prone to fouling andrequires lower main fractionatorbottoms temperature to minimizefouling.

Many refiners will also adjust bottomstemperature with conversion shifts tominimize fouling. A drop in conversioncould result in higher slurry exchangerfouling. Slurry with higher API gravity,generally due to lower conversion,contains more saturated compounds,which can reduce asphaltenesolubility and increase soft cokefouling. Slurry with a high API gravityis also more viscous and more proneto fouling.

Feedstock leaking into the slurrycircuit can cause asphalteneprecipitation. This occurs becausethe feedstock is more paraffinicthan the slurry and reduces thesolubility of the slurry asphaltenes.The refiner should take allprecautions necessary to ensurethat the feed emergency by-passvalve and the slurry/feed preheatexchangers do not leak feedstockinto the slurry circuit.

The FCC catalyst can be formulatedwith features to minimize fouling.Catalyst matrix design can beoptimized to improve slurryexchanger fouling by increasingType III cracking as described byZhoa.(3) Minimizing naphtheno-aromatics and paraffinic content ofthe slurry by increased Type IIIcracking may improve asphaltenesolubility and reduce slurryexchanger fouling. Using a catalystwith proper tolerance tocontaminant metals will help avoidfouling as well. Increased catalystcontaminants that result in a loss ofFCC conversion can increase thelikelihood of fouling as discussedabove.

Hot Cycle Oil Flush at the inlet of theslurry exchangers can help keepasphaltenes in solution andincrease tube velocity, both ofwhich will help reduce slurryexchanger fouling.(5)

INORGANIC BASED FOULING

Inorganic fouling can includefouling prompted from corrosion oriron scale, catalyst or precipitatedmetals. Catalyst is often found intube deposits and can be identifiedby the presence of alumina, silica,and rare earth. Catalyst in thedeposits is often a result of organicbased fouling, since catalystgenerally accumulates onto viscousprecipitated asphaltenes or otherhydrocarbons already present onthe tubes.

Another inorganic foulant in FCCslurry service is antimony. Antimonypresent in the slurry has been found

Figure 51Catalyst and Coke Interaction with Precipitated Asphaltenes

Slurry Exchanger

Fouling

Inorganic Organic

Coke Deposits

PrecipitatedAsphaltenes

Polymerization

Catalyst Particulates

Al & Si


on tube deposits. In one instance,20 wt.% antimony was found in adeposit.(15) The antimony source wasfrom antimony injection into the FCCfeedstock to passivate equilibriumcatalyst nickel.

Preventing Inorganic BasedFouling

Catalyst losses to the mainfractionator should be minimized byproper reactor cyclone operationand good reactor cyclonemechanical integrity. The catalystitself can also be designed tominimize losses to the mainfractionator. To maximize catalystretention, the following catalystdesign parameters should beconsidered:

• Low Attrition Index (Low DI);• Low 0 to 40 micron content;• High Particle Density.

Antimony levels in the slurry shouldbe closely monitored to minimizethe possibility of antimonydepositing onto the tube service.FCC equilibrium catalyst antimonyto nickel ratio is generally between0.10 and 0.60 by weight. Theantimony chemical should beinjected to maintain the target levelon the equilibrium catalyst withoutsignificant overfeed.

Proper metallurgy in the mainfractionator, slurry piping and slurryexchanger should be used tominimize corrosion. Below is asummary of suggested materials:(6)

• Main fractionator internals TP 410 SS ;

• Exchanger tubes TP 405 or TP 410SS ;

• Slurry Lines 5 Cr - 1/2 Mo with 1/4” CA.

Antifoulants

Antifoulants have been successfullyused to prevent FCC slurryexchanger fouling. Antifoulants canbe generally classified as follows:

• Organic dispersants - prevent the agglomeration and deposition ofasphaltenes;

• Inorganic dispersants - prevent the deposition ofcatalyst fines or other inorganic foulants such as Fe compounds;

• Coke suppressants - inhibit condensation reactions, which lead to hard coke-like deposits in exchangers.

A combination of antifoulants can beused. However, the likely source ofthe fouling should be identified beforea specific antifoulant is applied.

Note that there can be somedownstream effects when using anantifoulant. For example, in somecases catalyst fines settling in slurrytanks can be impacted if an inorganicdispersant is used.

Additional Design and OperationConsiderations

There are many design considerationsfor exchanges in slurry service thatcan help to minimize the potential forexchanger fouling.

Slurry exchanger tube velocitiesshould be 6 to 10 fps.(8) Velocitiesbelow six fps can result in catalyst,coke or other particulates settling ontothe tube surface, resulting in fouling.The minimum slurry exchanger tubediameter should be one inch.(5)

Smaller tubes can be subject toexcessive fouling and are difficult toclean.

Spill back control can be used on netproduct exchangers in turndownconditions to keep tube velocitiesabove minimum values.

Slurry should be present only on theexchanger tube side. With slurry onthe shell side, it is impossible toprevent catalyst settling in theexchanger because of low localvelocities.

Vertical and Spiral slurrypumparound exchanger designstend to be less prone to slurryexchanger fouling.

Finally, having spare slurryexchangers should be consideredto minimize turndown duringexchanger cleaning.

Case Study

An FCC unit began observing severefouling of their Slurry SteamGenerator exchangers. The foulingbegan suddenly and continued forapproximately two weeks and thenstopped. During that time, it wasnecessary to clean the exchangersseveral times. The exchangersexhibited a reduction in heat transfercoefficient. Exchanger pressuredrop was not affected.

The FCC was a modern designemploying a modern riser terminationdevice and state-of-the-art feedinjection nozzles. The reactortemperature operated at 980˚F andmain fractionator bottomstemperature was typically 690˚F.Conversion normally was ~78 vol.%with a slurry API gravity of -2 API.

The feedstock was a vacuum gas oiland resid blend with the followingnominal feedstock properties:

• API 22˚ to 24˚;• K Factor 11.7 to 11.8;• Conradson Carbon ~1.0 wt.%;• 10% Greater than 1050˚F.

A deposit was taken from the fouledexchanger. Analysis of the depositshowed the following:

• 87% Carbon, 94% Organic Based (C, H, N);

• <1% Alumina;• 1 wt.% Antimony;• <1 wt.% Iron;• 52% of the sample was

Asphaltenes.


The refinery does use antimony topassivate nickel. However, antimonyhad been used for several yearswithout any previous issues. Norecent change in antimony injectionwas made and consequentlyantimony was likely not the cause ofthe fouling.

The deposit itself did not containcatalyst, as evident by the lowamount of alumina.

The high amount of asphaltenes inthe deposit confirmed asphalteneprecipitation as the likely foulingmechanism.

A review of feedstock propertiesshowed the feedstock had recentlybecome more paraffinic, as evidentby the higher API gravity and KFactor during the same time as theexchanger fouling. Figure 52 showshow feedstock API and K factorshifted. When the refinery changedthe feed source the feedstockproperties returned to typical valuesand the fouling stopped.

Many refineries recognize that somefeed and crude sources can resultin increased FCC slurry exchangerfouling. Those sources are eitheravoided, the main fractionatortemperature is reduced, and/orantifoulants are used to minimizefouling while those feedstocks areprocessed.

Final Remarks

Continuous monitoring of the overallheat transfer coefficients is critical tocatch a slurry exchanger foulingproblem early. Monitoring slurryproperties such as API gravity, ashcontent, asphaltene content, andviscosity can also alert the refinerwhen the FCC unit may be moresusceptible to slurry exchangerfouling. A shift of feedstockproperties or unit conversion may alsoincrease slurry exchanger fouling.

Slurry exchanger fouling often occursduring start-up or at turndownconditions when feedstock andoperating conditions may be atypical.Special precautions may be consideredduring these unusual operations.

Reducing slurry exchanger fouling bylower main fractionator bottomstemperature and higher slurry productrate can be costly in terms of lowerproduct value. Grace Davison canwork with the refiner to adjust catalystproperties and operating strategy tominimize fouling and any subsequentyield loss.

References

1. Barlow, R., Reduce FCC Fouling,Hydrocarbon Processing, July 1986pg 37-39

2. Meyers, R., Handbook of PetroleumRefining Processes 2nd Edition,McGraw Hill 1996, pg 3.95

3. Zhao, X. FCC Bottoms Cracking Mechanisms and Implications for Catalyst Design, 2002 NPRA AnnualMeeting, San Antonio, TX, AM 02-53

4. Wilson, J., Fluid Catalytic Cracking Technology and Operation, PenwellPublishing 1997, pg 225

5. Sadeghbeigi, R., Fluid Catalytic Cracking Handbook 2nd Edition,Gulf Publishing Co. 2000, pg 251 to253

6. Walker, P. , NPRA FCC Principles and Practices, 2004 NPRA Q&A,Anaheim CA

7. Wilson, J., Fluid Catalytic Cracking Technology and Operation, PenwellPublishing 1997, pg 257

8. Wilson, J., Fluid Catalytic Cracking Technology and Operation, PenwellPublishing 1997, pg 229-230

9. Walker, G, A Case Study in Slurry Fouling, 2003 NPRA Annual Meeting, San Antonio TX, AM 03-63

10. Fouling in FCC Units & Purification Trains, GE Infrastructure Water andProcess Technologies Presentation

11. Dickakian, G., Asphaltene Precipitation in Primary Crude Exchanger Fouling Mechanism, Oil and Gas Journal, Mar. 7, 1988, pg 47-50

12. Falker T. et al., Modeling of FCC unitMain Fractionator Bottoms Pumparound Fouling, 1995 NPRA Annual Meeting, San Francisco CA,AM-95-71

13. Shawney, K et al., The FCCU SlurryLoop Fouling Mitigation, 2002 AIChE Spring Meeting, New Orleans LA

14. NPRA Q&A Transcripts 1989, pg 68

15. NPRA Q&A Transcripts 1986, pg 54


26

25

24

23

22

21

20

11.95

11.90

11.85

11.80

11.75

11.70

11.65

11.60

11.551-JAN 21-JAN 10-FEB 1-MAR 21-MAR 10-APR 30-APR

Severe Slurry ExchangerFouling began suddenly

Fouling Stopped

API Gravity K Factor

Figure 52Feed Shift Property Shift Impacts Slurry Exchange Fouling

Davison Ref in ing TechnologiesW.R. Grace & Co.-Conn.7500 Grace Dr iveColumbia , MD 21044(410) 531-4000

Davison Ref in ing TechnologiesAsia Pac i f icW.R. Grace Pte. Ltd.501 Orchard Road#05-11/12 Wheelock PlaceSingapore 238880(65) 6737-5488

Davison Ref in ing Technologies Europe Grace GmbH & Co. K.G.In der Hol lerhecke 1Post fach 1445D-67545 Worms, Germany49 (6241) 40300

cata [email protected]

©2007 W.R. Grace & Co.-Conn.

Documents

Hydroprocessing/FCC Synergy · hydrotreating catalyst systems included a NiMo catalyst (ART AT575) and a CoMo catalyst (ART AT775).Both catalysts have had extensive use in FCC feed