AM-89-50

CATALYTIC MEANS TO MAXIMIZE FCC OCTANE BARRELS

By

E. L. Leuenberger, Executive Technical Service Engineer
R.A. Bradway, Senior Research Engineer
M.A. Leskowicz, Staff Enigneer
J.J. Stegar, Manager of Research and Development, Petroleum Catalysts
BASF Corporation
Edison, New Jersey

Presented at the

1989 NPRA
ANNUAL MEETING
March 19-21, 1989
San Francisco Hilton
San Francisco, California

Abstract:

FCC gasoline selectivity and octane will both be maximized if the equilibrium catalyst satisfies two conditions:

1. The zeolite must be accessible to the unconverted gas oil molecules.
2. The zeolite acidity must be kept below the threshold at which gasoline volume is lost with little beneficial octane gain.

FCC unit operating constraints will determine which other catalyst features are desirable. Optimal catalyst designs for units with regenerator metallurgical limits, dry gas handling limits and C3/C4 handling limits are discussed.

Introduction: The Value of an Octane Barrel

The gasoline produced in fluid catalyst cracking units (FCCUs) often has sufficient octane to be sold as regular unleaded gasoline without upgrading its octane. The value of the FCC gasoline produced can thus be approximated by the refinery selling price, or "rack price", of unleaded regular gasoline. However,there are costs associated with failure to produce gasoline with the 87 road octane specification of regular gasoline and there are benefits associated with exceeding the octane spec. For example, expensive high octane components must be added to the FCC gasoline if it falls short of the unleaded regular gasoline blending target. On the other hand, if the FCC gasoline exceeds the regular grade octane specifications, relatively cheap low octane gasoline can be blended into the FCC gasoline to make regular unleaded.

Refinery planners use the concept of octane barrels to represent the benefits associated with increasing gasoline octane. Octane barrels are calculated by multiplying the barrels of gasoline times the difference between FCC gasoline octane and unleaded regular octane. The octane barrel adjustment to the FCC product value is determined by the refiner's cost of upgrading gasoline octane times the number of octane barrels. Using this definition, there is no adjustment to the FCC gasoline's value if it has exactly the octane required to meet the regular blend specification. There is a cost associated with failing to produce FCC gasoline equal to the regular gasoline specification, so the octane barrel value is negative if the specification is not met. Exceeding the octane specification adds to the FCC product value, so in this case the octane barrel value is positive.

Although the value of an octane barrel of gasoline depends on refinery economics, it can be estimated from the selling prices of unleaded regular and unleaded premium when the demand for premium gasoline exceeds supply. Under these conditions, the value of an octane barrel is the difference between premium and regular gasoline "rack prices" divided by the octane difference between leaded and premium. At the end of 1988, this calculation indicated the value of a road octane barrel was 59 cents. Refiners who own gasoline stations may also consider "pump prices" in calculating the octane barrel price. The twelve cents/gallon premium versus unleaded price differential at the end of 1988 justifies an octane barrel value of $1.00. These high values of an octane barrel and the strong demand for premium justify the use of FCC catalysts designed to maximize octane barrels.

Catalyst Fundamentals that Affect Octane Barrels

The definition of octane barrels implies that an FCC catalyst must maximize both gasoline yield and octane to be an octane barrel catalyst. Increases in FCC gasoline octane often are obtained at the expense of gasoline selectivity, so maximizing both yield and octane is no easy task. Both the zeolite crystals, where the strongest acid sites that promote cracking are located, and the matrix support for the zeolite must be optimized to maximize octane barrels.

The zeolite properties that affect the yield and octane of an FCC catalyst are:

1. The physical size and pore structure of the zeolite crystal
2. The acid strength of the cracking sites

Matrix surface area and the strength of matrix acid cracking sites compliment zeolitic properties in maximizing octane barrels. The role of each of these catalyst properties in the design of an octane barrel catalyst is discussed in this paper.

Reducing Diffusional Effects Improves Gasoline Selectivity

Mass transfer within the cracking catalyst zeolite has a significant effect on gasoline selectivity. Chemical engineers who study the fundamentals of catalysis have known since the 1950's(Ref. 1) that diffusional effects can reduce the selectivity of chemical reactions for intermediate products. In catalytic cracking, C4 and lighter hydrocarbons are the end products of the reactions. Gasoline and diesel fuel are the valuable intermediate products whose yields can be reduced by diffusional limitations. It follows that gasoline yields will be improved if a cracking catalyst is designed to minimize diffusional resistance.

Since the early development of zeolitic cracking catalysts, researchers have also known that it was difficult for gas oil molecules to diffuse into the center of zeolitic crystals(Ref. 2). This diffusional limitation is caused by the 8 angstrom size of the zeolite pore openings. Since many gas oil molecules are too large to easily fit in the pores, the cracking sites in the crystal interior serve primarily to crack smaller molecules like gasoline. Eliminating diffusional resistance by increasing the accessibility of the zeolite to gas oil molecules should improve gasoline selectivity because it will increase the rate of gasoline formation and reduce gasoline "recracking".

Catalyst Features that Improve Zeolite Accessibility

The most straightforward way to improve zeolite accessibility is to increase the exterior surface of the crystals by reducing the crystal size. For equal weights of zeolite, decreasing the diameter by a factor of three will triple the outer surface area.

A recent study(Ref. 3) demonstrated that the reduction in zeolite crystal size described above is very effective in improving gasoline yields. Samples of zeolite crystals were prepared with average sizes that measured 0.3 and 0.9 microns in diameter. These crystals were incorporated into experimental catalysts that were otherwise identical. Results of this study are reproduced in Figure 1. The smaller crystals yield from 1 to 10 weight percent more gasoline.

An alternative to reducing the crystal size of the zeolite is to develop a network of larger pores within the crystal. Pores that are greater than 8 angstroms in diameter will allow more gas oil molecules to penetrate to the center of the zeolite. This reduces the amount of relatively inaccessible surface area available for gasoline recracking, promotes gas oil cracking, and improves gasoline selectivity.

Thermal Aging of Zeolite Creates a Network of Pores

It is possible to develop a secondary pore network with pores at least 40 angstroms in diameter by exposing zeolite to elevated temperatures(Ref. 4,5). This thermal aging process results in the collapse of a portion of the zeolite crystal, creating a network of holes and cracks. The collapse of the alumina silicate framework of the zeolite crystals is the result of steam extraction of the aluminum in the framework.

The longer a catalyst is aged, the more extensive the secondary pore structure becomes. Figure 2 shows large140 angstroms diameter channels develop and 40 to 60 angstroms pores double in number as aging is increased(Ref. 4). In this figure, unit cell size is used as a measure of the amount of aging the zeolite has experienced. The fresh zeolite has a unit cell size of 24.72 angstroms and the completely aged zeolite has a cell size of 24.30 angstroms or less. Unit cell size can also be used as a measure of zeolite acidity and framework alumina content, as will be discussed later.

Zeolite can be aged or dealuminated either in a commercial unit or during FCC catalyst manufacture. If dealumination conditions are properly controlled during manufacture, silicon atoms from the collapsed zeolite framework will diffuse through the crystal to fill the zeolite defects left by the extracted alumina. This "healing" process strengthens the remaining part of the crystal. The uncontrolled dealumination in a commercial FCC regenerator results in crystals with higher zeolite defect density and less stability.

The defects left by the extracted aluminum disturb the X-ray diffraction pattern by creating irregularities in the zeolite crystal lattice. This results in a smearing of the sharp X-ray diffraction pattern that is characteristic of defect free zeolite. The more defects in the crystal, the broader the width of the normally sharp bands in the pattern. BASF uses the line broadening of the XRD pattern to calculate a number which we call the zeolite defect index. The lower the defect index, the sharper the X-ray pattern and the fewer the number of defects that remain in the crystal. We have found a low defect index results in better stability. Figure 3 shows how one measure of stability, zeolite surface area retention, is improved for a catalyst with a lower defect index.

In addition to greater stability, zeolite dealuminated during its manufacture can provide improved accessibility for gas oil reactants. When zeolite is aged in a commercial FCC unit, the alumina extracted from the crystal framework accumulates in the zeolite pores(Ref. 4,5). This non-framework alumina obstructs the access of gas oil molecules to the zeolite interior. When zeolite is dealuminated during its manufacture, acid solutions can solubilize and wash away this alumina(Ref. 6,7).

Higher Acid Strength Increases Octane but Lowers Gasoline

Although zeolitic diffusional resistances have long been known to affect FCC yields, recent catalyst development work has concentrated on the effects of zeolite acid strength. This property of zeolitic cracking catalysts results in a trade-off between gasoline yield and octane. Pine and coworkers(Ref. 8) demonstrated both octane and gasoline selectivity can be correlated with a single parameter that is related to the zeolite acid strength: the unit cell size.

The unit cell size is a measure of the physical size of the alumina silicate units that repeat to form the framework of the zeolite crystal. It is also a measure of the concentration of aluminum atoms in the zeolite framework because aluminum atoms are slightly larger than the silicon atoms they can replace. The larger size of the aluminum atoms increases unit cell size at higher aluminum levels.

The acid sites in cracking catalysts are associated with the aluminum atoms.The highest acidity aluminums are surrounded by silicon and oxygen atoms. Low unit cell sizes provide the high zeolite silicon levels needed to isolate the aluminum atoms to form strong acid sites.

The strength of the acid sites is responsible for increasing gas yields(Ref. 8,9,10), and high gas yields adversely affect gasoline selectivity. However the isolation of the acid sites prevents bimolecular hydrogen transfer reactions that saturate olefins produced by catalytic cracking. Since olefins in the gasoline improve octane, reducing unit cell size increases FCC gasoline octane at the expense of gasoline selectivity.

Unit Cell Size Should Equilibrate Above a Minimum Level

Calculations based on a random distribution of aluminum atoms in a zeolite crystalline framework show that the minimum distance between sites increases rapidly as the unit cell size is reduced below 24.45 angstroms. In Figure 4, we show that the distance between sites is expected to be 16 angstroms if the unit cell size is reduced to 24.30. A distance of 16 angstroms is larger than the size of most gas oil molecules, so it is likely that a smaller cell size is not needed to prevent hydrogen transfer.

Coke formation has been described as a condensation reaction of aromatics accompanied by hydrogen transfer(Ref. 11), so we would expect that coke selectivity would be reduced at low unit cell sizes until the acid sites were far enough apart to prevent hydrogen transfer. A recent article(Ref. 12) states that a unit cell size of 24.33 angstroms is small enough to minimize coke selectivity. Coke selectivity was reduced as equilibrium unit cell size was lowered from 24.6 to 24.33 angstroms, but further unit cell size reduction did not improve the coke selectivity.

Hydrogen transfer reactions are also responsible for the saturation of gasoline olefins, a reaction that reduces research octane. As Figure 5 indicates, we observed only a marginal decrease in hydrogen transfer and we expect little octane improvement below 24.30 unit cell size. However, the stronger acid sites associated with smaller cell size will promote gasoline recracking. It is, therefore, best to design an octane barrel catalyst so its equilibrium unit cell size is as close to 24.30 as possible.

Commercial Operations Have a Unit Cell Size Distribution

The zeolite unit cell size does not remain constant as FCC catalyst ages. The high temperature, steam rich environment in a fluid cat cracker's regenerator promotes the loss of aluminum from the zeolite framework. Laboratory experiments(Ref. 13)   where catalyst aging is accelerated by raising both the steam concentration and the steam temperature show the loss of aluminum is relatively rapid. Under the 1450 degree F steaming in Figure 6, most of the dealumination occurs within the first hour. However, the equivalent time of commercial deactivation represented by that hour of accelerated aging is approximately five days.

During its aging time, the high unit cell size catalyst contributes a disproportionately high fraction of the total activity in the unit. Suppose the average catalyst age in the unit is 50 days and that the average catalyst properties are represented by ten hours of accelerated aging. Then Figure 7 shows the activity of the freshest catalyst is more than three times the average activity.

The activity versus time relationship in Figure 7 can be combined with the first order catalyst age distribution typical of a commercial cat cracker. Integrating activity over the age distribution allows the calculation of the fraction of the total activity contributed by catalyst in each range of ages. Figure 8, the results of the calculation, shows catalyst with one half hour of accelerated aging contributes 15% of the total activity in the unit.

Significance of the High Unit Cell Size Catalyst in a FCCU

Since the high unit cell size fraction of the zeolite contributes 15% of the activity in a commercial unit, its hydrogen transfer properties affect at least 15% of the gasoline. The hydrogen transfer reactions promoted by the fresh catalyst saturate high octane olefins, so high unit cell size fresh catalyst is expected to depress octanes in at least 15% of the gasoline product. In addition to the negative effects on octane, the high unit cell size portion of the inventory also catalyzes gasoline recracking.

In the set of experiments(Ref. 13) presented in Figure 9 mildly aged catalyst with a 24.47 unit cell size was compared to more severely aged catalyst. The higher unit cell size catalyst depressed gasoline selectivity by two weight percent. Part of the loss in gasoline selectivity can be attributed to bimolecular condensation reactions that form coke.These reactions are favored by the decreased distance between acid sites and the high hydrogen transfer ability that are associated with higher unit cell size. Figure 10 shows that half of the two weight percent loss in gasoline selectivity can be explained by coke formation.

The remaining drop in gasoline selectivity for fresh catalyst has been attributed(Ref. 13 ,14) to mass transfer effects within the zeolitic crystals. Fresh catalyst has so many active cracking sites that a gas oil molecule cannot penetrate far from the crystal surface before it is cracked. The interior of the crystal is then available to recrack gasoline because it is free of gas oil molecules. Although the individual acid sites in a high unit cell size fresh catalyst do not have sufficient acidity to crack gasoline to ethane, ethylene, or methane, they have sufficient acidity to crack gasoline to C3's and C4's. The relatively large number of sites available for gasoline recracking in fresh catalyst further reduces the gasoline yield by producing LPG saturates, as shown in Figure 11.

The effect of high unit cell size catalyst on research octane and product distribution can be further demonstrated by the fresh catalyst addition study summarized in Table I. In this experiment, a base operation was first established in a circulating catalytic cracking pilot unit, and then the unit was monitored for 4 hours after 5 weight percent fresh catalyst was added to the circulating inventory. The fresh catalyst addition depressed gasoline selectivity by 2.8 weight percent of conversion. Increased hydrogen transfer reduced research octane by half an octane number, increased coke yield, and increased the concentration of saturates in the LPG.

The poor gasoline yield and low octane observed above can be explained by the combination of the mass transfer and hydrogen transfer properties of the fresh size portion of the catalyst inventory, which promotes the formation of coke and LPG saturates from gasoline. A simplified equation to represent these effects is:

Aromatic Gasoline + Gasoline Olefins --------> Coke + LPG Saturates

Hydrogen is transferred from the aromatics in the gasoline to olefins in both the LPG and gasoline fractions. The hydrogen transfer causes gasoline aromatics to condense to form coke. Gasoline olefins also are recracked to LPG. The consumption of both gasoline olefins and aromatics depress octane as well as gasoline yield.

Optimal Zeolite Acidity for Octane Barrel FCC Catalysts

The unit cell size distribution of cracking catalyst in the inventory of a commercial unit will vary from that of fresh catalyst to an equilibrium value that is influenced by rare earth content, sodium level and unit operating conditions such as regenerator temperature. Test results demonstrate that the high unit cell size of fresh catalyst will depress gasoline selectivity by making coke and LPG. The same reactions that make coke and LPG simultaneously depress octane. It is thus desirable to use a fresh catalyst with a unit cell size below the 24.5angstroms level to maximize octane barrel production.

Data on catalyst hydrogen transfer potential indicates that a unit cell size below 24.30 does not significantly improve octane, but lower unit cell size zeolite can have high acidity that promotes gas formation. The acidity that maximizes gasoline octane without unnecessarily increasing gas make, therefore, occurs at approximately 24.30 unit cell size.

Investigation of zeolite acidity effects on octane barrel production thus demonstrates that a narrow band of equilibrium catalyst acidity is best. Fresh catalyst unit cell size should be as low as possible, preferentially less than 24.5 angstroms. Catalyst properties should stabilize the unit cell size at 24.30 angstroms. This narrow range of equilibrium cell sizes is best for gasoline selectivity while only marginal improvements in octane can be obtained by allowing equilibrium unit cell size to stabilize at unit cell sizes less than 24.30.

The Role of an Active Matrix in an Octane Barrel Catalyst

The matrix of a catalyst is the support for the zeolite crystals. If matrix is devoid of acid sites, it is unable to affect the product selectivities. An active matrix contains acid sites associated with aluminum atoms, but most of these sites are too weak to crack light gas oil. Only aluminum sites that by chance are isolated from adjacent aluminum atoms have strong enough acidity to crack most hydrocarbons, and due to the random arrangement of silicon and aluminum atoms in the matrix, these sites are relatively rare. Since the number of acid sites that promote cracking in the matrix will be much fewer than the number of aluminum atoms, they will be widely separated. This will prevent hydrogen transfer and maximize the octane of any gasoline formed on the matrix. In addition, the pore structure of a catalyst matrix is designed to be accessible to large molecules. This allows gas oil molecules to successfully compete with gasoline for matrix cracking sites, and usually prevents gasoline recracking on the matrix.

The matrix cracking sites are also accessible to large diameter gas oil molecules that cannot penetrate into the zeolite pores. FCC catalysts typically have 30 to 100 square meters per gram of matrix surface area at equilibrium conditions. Although the same catalyst may have up to 150 square meters per gram of zeolitic surface area, the external zeolitic surface area will be less than 2 square meters for catalyst with a crystal size of 0.2 microns(Ref. 15). Matrix surface area is needed in addition to zeolitic surface area to assure that large molecules have sufficient surface area on which to crack. When these larger molecules are aromatic, matrix cracking can produce aromatic gasoline which will improve FCC octane barrels.

An active matrix with accessible sites for bottoms cracking also may be necessary to make full use of the zeolite mesopore structure. A recent study(Ref. 16) used dealuminated zeolite incorporated in an inactive matrix to minimize the influence of matrix on the results. It was found that the mesopores of the zeolite could "act as traps for large hydrocarbon molecules." When the acidity on the pore walls was not sufficient to crack these molecules, they condensed to form coke that plugged the mesopores, hindered diffusion of smaller molecules into the zeolite interior, and reduced gasoline selectivity.

Our data suggests an active matrix prevents heavy hydrocarbon molecules from plugging the zeolite pore structure. When a bed of active matrix material was used to filter a zeolite bed in a MAT cracking experiment, conversion was increased by 18 weight percent over the level when the beds were reversed. Selectivity to gasoline was also substantially improved. This experiment, which is outlined in Figure 12, demonstrates that an active matrix can convert heavy hydrocarbons to lighter molecules that will not coke up the zeolite pore structure. The resulting increase in zeolite accessibility improves both activity and gasoline selectivity.

PyroChem Processing

BASF has developed a new proprietary manufacturing process called PyroChem processing that reduces the zeolite defect index of FCC catalyst, develops a secondary pore structure, and removes a substantial fraction of the alumina that blocks the pores. When this process is used to reduce the zeolite fresh unit cell size of an FCC catalyst, and when the zeolite is combined with an active matrix, gasoline selectivity improves due to improved zeolite accessibility. Table II shows a one weight percent improvement in gasoline selectivity can be obtained by using the PyroChem process to reduce the fresh unit cell size from 24.65 to 24.41.

PyroChem processing is combined with BASF's patented in-situ zeolite crystallization process in our new PRECISION™ FCC catalyst family. Together, these two BASF processes produce small zeolite crystals with a secondary pore network that improves zeolite accessibility for maximum gasoline yields. The in-situ FCC catalyst manufacturing process used to make PRECISION™ catalysts also produces enough active matrix surface area to crack heavy molecules that could potentially plug the zeolite pores. The active matrix thus contributes to the high gasoline selectivity by protecting the zeolite pore structure from coke formation. An additional benefit of cracking these heavy molecules on the matrix is the conversion of heavy fuel oil to diesel and other more valuable products.

A low defect index is also a characteristic property of catalysts manufactured using PyroChem processing. Fewer defects result in an increase in zeolite surface area retention at constant aging conditions. Figure 13 demonstrates that PRECISION™ catalysts manufactured using PyroChem processing retain more of their fresh zeolite surface area than catalysts dealuminated by other processes.

PyroChem Processing Improves Octane and Reduces Delta Coke

The zeolite mesopore structure developed by PyroChem processing improves zeolite accessibility, and the improved accessiblity minimizes recracking reactions promoted by the interior of the zeolite crystal. Recracking of gasoline has the empirical formula: Gasoline olefins plus gasoline aromatics yield coke plus LPG saturates. Minimizing this recracking maximizes the yield of high octane gasoline, reduces the isobutane to olefin ratio in the light hydrocarbons produced, and lowers coke yield.

Laboratory data in Table II shows the magnitude of these benefits of PyroChem processing. Delta coke is lowered by 20 relative percent, isobutane to butylene ratios are reduced 15 relative percent, and both research and motor octanes are improved by more than half a number.

Unit Constraints Influence Optimal Catalyst Design

We have found that gasoline selectivity and octane will both be maximized if the FCC catalyst zeolite accessibility to uncracked gas oil molecules is increased. Maximum zeolite accessibility is achieved in a catalyst with small diameter zeolitic crystals that have been dealuminated to a unit cell size below 24 .5 angstroms by a process that creates a "highway access system" of mesopores for the reactant molecules. Low fresh unit cell sizes also minimize coke formation on the unaged fraction on the catalyst inventory.

BASF's PyroChem process produces this type of low fresh unit cell size, accessible zeolite.

Other catalyst features may be useful to optimize octane barrels in some units, but less useful in other units. We have conducted laboratory and computer process simulation studies to provide guidelines for optimizing catalyst rare earth content. The results of these studies are summarized in Table III. We found rare earth should be adjusted so the catalyst equilibrates to a unit cell size of about 24.30 angstroms to maximize octane barrels in FCC units that are at their gas handling limit.

Units that have plenty of gas compressor room may find it beneficial to use a catalyst that equilibrates to a minimum unit cell size because this produces a small octane gain. The regenerator temperature limited case and the LPG limited case in Table III show maximum product value when a zero rare earth catalyst was used to minimize equilibrium unit cell size. Both the regenerator temperature and the LPG yield depend on the extent of the hydrogen transfer reactions. Since these reactions are practically eliminated when the cell size is reduced to 24.30 angstroms, a further cell size reduction will not greatly change LPG yield or the coke make that determines regenerator temperature. Conversion is not depressed and octane barrels increase as long as the gas produced by a zero rare earth catalyst does not exceed unit limits.

BASF offers both a zero rare earth PRECISION™ and a PRECISION™ catalyst that contains 0.6 weight percent rare earth. The zero rare earth catalyst will maximize octane barrels in regenerator temperature limited units and in units with LPG handling limits, while the rare earth version is optimal for units that have gas handling limits.

References

1. Wheeler, A., Adv. Catal. 3, 250 (1951)

2. Thomas, C.L. and Barmby, D.S., Journal of Catalysis 12, 341 (1968)

3. Rajagopalan, K., Peters, A.W, and Edwards, G.C., Applied Catalysis 23, 69 (1986)

4. Patzelova, V., and Jaeger, N.I., Zeolites 7, 240 (1987)

5. Lynch, J., Raatz, F., and Dufresne, P., Zeolites 7, 333 (1987)

6. Scherzer, J., and Bass, J.L., Journal of Catalysis 46, 100 (1977)

7. Engelhardt, G., Lohse, U., Patzelova, V., Magi, M., and Lippmmaa, Zeolites 3, 239 (1983)

8. Pine, L.A., Maher, P.J., and Wachter, W.A., Journal of Catalysis 85, 466 (1984)

9. Corma, A., Planelles, J., and Tomas, F., Journal of Catalysis 94, 445 (1984)

10. Planelles, J., Sanchez-Marin, J., Tomas, F., and Corma, A., Journal of Molecular Catalysis 32, 365 (1985)

11. Gates, B.C., Katzer, J.R., and Schuit, G.C.A., "Chemistry of Catalytic Processes", McGraw-Hill, New York (1979)

12. Raj agopalan, K. and Peters, A.W., Journal of Catalysls 106, 410 (1987)

13. Leuenberger, E.L., Moorehead, E.L., and Newell, D.F., paper AM-88-51, N.P.R.A. Annual Meeting, San Antonio, TX, Mar. 20-22, 1988

14. Corma, A., Herrero, E., Martinez, A., and Prieto, A.C.S. Symp. on Advances in FCC, New Orleans, Aug. 30 - Sep. 4, 1987

15. Farcasiu, M.X. and Degnan, T.F., Ind. Eng. Chem. Res. 27, 45 (1988)

16. Addison, S.W., Cartlidge, S., Harding, D.A., and McElhiney, G., Applied Catalysis 45, 307 (1988)