REFORMULATED GASOLINE WILL CHANGE FCC OPERATIONS AND CATALYSTS

July 2, 1990
Gary M. Stokes, Charles C. Wear, Wilson Suarez, George W. Young Davison Chemical Division W.R. Grace & Co. Baltimore Operation of fluid catalytic cracking units (FCCUS) will be significantly affected by new regulations that will in all probability require gasoline to be produced with lower aromatics and olefins contents, lower vapor pressure, and a minimum oxygen content.

Gary M. Stokes, Charles C. Wear, Wilson Suarez, George W. Young
Davison Chemical Division
W.R. Grace & Co.
Baltimore

Operation of fluid catalytic cracking units (FCCUS) will be significantly affected by new regulations that will in all probability require gasoline to be produced with lower aromatics and olefins contents, lower vapor pressure, and a minimum oxygen content.

A study was conducted to better define the basic relationship between operating variables, including catalyst and naphtha quality, in the context of reformulated gasoline. The study helped to define specific operating strategies, potential problem areas, and opportunities for improved FCC unit and catalyst technologies.

FCC feedstock quality can have a significant influence on the composition of FCC naphtha. However, even extremely paraffinic or aromatic feeds can yield substantial levels of both olefins and aromatics in FCC naphtha, particularly when compared to the levels proposed in a reformulated gasoline pool.

Catalyst type, reactor temperature, and catalyst/oil (C/ 0) ratio have only minor influence on the aromatics content of full-range FCC naphtha.

Increasing conversion, regardless of the mechanism used, increases gasoline aromatics to a slight extent.

Unlike aromatics, olefins content of FCC naphtha is a strong function of catalyst type, reactor temperature, and C/O ratio. Traditional means of increasing octane via catalyst and reactor temperature does so by increasing gasoline olefinicity. On the other hand, increasing C/0 severity decreases olefinicity.

Olefins in FCC naphtha are concentrated in the C5-C7 range, especially in the C5'S Aromatics are concentrated in the C8-Cl2 range.

Rejection of aromatics by separation techniques (extraction, etc.) and possibly by mandated gasoline endpoint reductions will dramatically decrease the volume of FCC naphtha available to the gasoline pool.

The amount of aromatics rejected to heavy naphtha cuts is not strongly influenced by catalyst type or operating variables.

Yield Of C5 olefins, especially isoamylenes, are enhanced by the use of ultrastable Y (USY) octane catalysts and ZSM additives. Using FCC isoamylenes for tertiary amyl methyl ether (TAME) feedstock and alkylating the normal C5 olefins would, in addition to providing excellent reformulated gasoline components, be an effective control over FCC gasoline olefinicity.

Operating strategies that enhance yield and selectivity of total olefins, C5 olefins, and isoamylene are also effective in the production of butylenes for C4 alkylation and methyl tertiary butyl ether (MTBE) production.

Although environmental legislation will probably include provisions to reduce gasoline olefins, the most effective FCC strategy may well be one that maximizes olefins to provide feedstocks for other processes.

FCC'S ROLE IN REFORMULATION

Most proposals defining reformulated gasoline have many elements in common: A minimum oxygenate level and reductions in aromatics, olefins, sulfur, and 90%-off boiling point.

Table 1 compares today's estimated U.S. unleaded gasoline pool with one possible set of reformulated gasoline specifications.

Also included in Table 1 are typical properties of today's FCC full-range naphtha. FCC naphtha is the largest volume component of the gasoline pool, at 35%. It is also the major source of olefins and, along with reforming, a major contributor of aromatics.

Reforming accounts for 32 vol % of the pool, alkylation 12%, isomerization 5%, butanes 5%, MTBE 2%, and other 9%.

While this is problematic to reformulation specifications, the FCCU is also the major source of light olefins for alkylate and ether production, both of which should be key elements in a future reformulated pool.

The evolution of the FCC process and catalyst technology has focused on high-quality gasoline production. Opportunities now exist for the FCCU to have some other mission.

For example, the FCC may find its future role as primarily a light olefin generator. W.R. Grace & Co. researchers are exploring this.

FEEDSTOCK EFFECTS

The quality of the feed has always been one of the most influential variables in determining the yields and product qualities obtained from an FCC unit. Changing feed type affects the paraffins-olefins-naphthenes-aromatics (PONA) analysis of FCC naphtha.'

Besides reducing the naphtha yield at constant operating conditions, a change from a very paraffinic (K factor of 12.5) to a very aromatic feed (K factor of 11.5) has the expected result of sharply increasing the concentration of aromatics.

The same change in feed also decreases the concentration of olefins.

Analogous shifts in naphtha properties occur with FCC feed hydrotreating, whereby a low-K stock is pumped with hydrogen to reduce feed sulfur and aromatics.

This approach increases gasoline yield and reduces sulfur in liquid products. This strategy is sound to meet not only reformulated gasoline requirements, but also high-quality diesel (0.05 wt % sulfur) as well.

In the short term, most refiners have limited flexibility to select feedstocks based on FCC yields alone. However, variables, such as catalyst type, reactor temperature, and C/O ratio, can usually be changed quite readily to adjust product yields and qualities.

The Davison study focused on the extent these traditional operating variables have upon FCC naphtha properties and yields of light olefins for downstream uses. All of the study work was done in Davison's circulating riser pilot plant, using a normal Gulf Coast feedstock.

TEST EQUIPMENT

The circulation riser is a small-scale FCCU pilot unit riser reactor, available under license from Davison .2 3 it features full computer process control with both adiabatic and isothermal modes of operation.

In its adiabatic mode, the pilot plant has demonstrated its ability to closely simulate the performance of several commercial FCCUS. The adiabatic mode was used for the study.

Feed was transferred from one of the twin storage tanks into the corresponding feed weigh cells. A metering pump precisely controlled the feed rate as feed was pumped from the load cell through a preheater to the feed nozzle. Steam was used as a feed dispersant.

Catalyst and dispersed feed passed into the vertical riser reactor equipped with adiabatic heaters. Internal thermocouples independently monitored the actual catalyst/vapor mix along the riser.

Catalyst residence time in the reactor was approximately 6 sec, and vapor residence time was about 3 sec. After passing through the reactor, the hydrocarbon products, catalyst, and dispersant passed into a stripper disengager.

Product gases exited through the top of the disengager to the product collection system.

The gaseous hydrocarbon products from the stripper passed through a primary stabilizer column that provides a product cut between C4 and C5.

The C4 and lighter gases were metered and analyzed.

The liquid products (C.5 and heavier) were condensed and batch collected for subsequent distillation.

Spent catalyst, dropping from the disengager into the vertical stripper standpipe, forms a dense-phase fluidized bed above the stipper slide valve. This slide valve is used to control catalyst flow from the stripper to the regenerator.

Part of the stripper-regenerator spent catalyst transfer line consists of a jacketed heat exchanger that provides a precise and reliable method for the direct, continuous display of catalyst circulation rate.

After passing from the stripper, spent catalyst enters the regenerator disengager. The spent catalyst drops to the regenerator dense-phase fluidized bed where it is burned clean with air.

The excess air and combustion products pass up through the disengager and then to the flue-gas control system where the gases are cooled, volumetrically metered, and analyzed for O2, CO, and CO2. The regenerator is equipped with multiplezone heating for complete, independent control over regenerator temperature.

For this study, the riser top temperature was used as the primary set point to control the regenerator slide valve, Feed temperature was used to change C/O ratio at constant riser and regenerator temperature.

Detailed gasoline composition and gas chromatograph (GC) octane numbers (research and motor) were determined using a proprietary Grace model. CFR engine octanes were also run on most samples.

CATALYSTS USED

In order to bracket a wide range of performance, three Davison catalyst systems were chosen as models: XP750, Astra-378, and Additive-OHS.

Properties of the catalysts are shown in Table 2, along with those of the standard feedstock.

XP-750 is non-rare-earth-treated, ultrastable zeolite catalyst with a relatively high activity matrix. Expected attributes of this catalyst are maximum octane and maximum olefin yield due to its low hydrogen-transfer characteristics.

Astra-378 is a dual zeolite catalyst (USY and Y) with high rare-earth content and moderate matrix activity. Performance of this catalyst is characteristic of octane-barrel grades which have intermediate octane potential and high gasoline yields relative to pure octane grades like XP-750.

Additive-OHS is Davison's ZSM-5 octane additive. ZSM-5 is known to enhance octane of FCC naphtha at the expense of yield, and to increase C,3-C4 olefin yields.' For this study, Additive-OHS was blended with Astra at the 4 wt % additive level.

All catalysts were steamed at 1,500 F. for 4 hr with 1 atm of steam pressure prior to riser testing.

Catalyst/oil severity curves were generated for XP, Astra, and Astra/OHS at 960 F. reactor temperature. In addition, 925 F. and 1,000 F. runs were made for Astra378 only.

Feed preheat temperature was varied from 300 to 700 F. which allowed C/O to be varied from 3 to 9. Regenerator temperature was maintained at 1,3000 F. and system pressure at 25 psig.

In order to determine selectivity for yields and gasoline quality, data were interpolated from the C/O severity curves, typically at constant conversion. For the study results, conversion is defined as the disappearance of all material boiling above 450 F. true boiling point (TBP). Because of the lower conversion levels for Astra at 925 F., this condition was omitted from the selectivity analysis, but it is included in selected figures.

FULL-RANGE FCC NAPHTHA

The constant conversion comparison for the set of variables is shown in Table 3. Note that the different cases illustrate traditional ways octane can be enhanced in the FCC: Reactor temperature, octane catalyst (USY), and octane additive (ZSM-5).

It is also worthy to note the relative differences in octane achieved in each case relative to the base operation of Astra at 960 F. Additive-OHS had the same octane effect as a 40 F. increase in reactor temperature, whereas XP gave an even larger octane boost.

The inverse relationship between naphtha selectivity and octane is well known.6 Generally, changing a variable to increase octane causes a drop in naphtha selectivity and vice versa.

The exception is the effect of higher C/O, which increases conversion and octane, but decreases naphtha olefin level. The effect of C/O ratio on octane is less pronounced than the effects of catalyst and reactor temperature.

The olefin content of the FCC naphtha can be significantly altered by these variables.

In fact, this has been common in the industry for the past several years as octane demand has shifted.

Test data also point out, however, that a substantial FCC naphtha octane penalty is in store for those who seek lower gasoline range olefins by these conventional routes.

The aromatics content of the FCC naphtha is not sensitive to any of these variables.

C/O ratio and conversion have some influence.

A 10 liquid-volume (LV %) increase in conversion increases aromatics content in the FCC naphtha, but only by about 1-2 LV % of the naphtha.

Catalyst type and reactor temperature appear to have virtually no effect.

PONA COMPOSITION

Fig. 1 shows the paraffins, olefins, naphthenes, and aromatics breakdown by carbon number of the naphtha produced with Astra-378 at base conditions.

There is a high concentration Of C5-C7 olefins and paraffins (mostly isoparaffins).

These materials amount to about half of the naphtha by volume. C5 olefins comprise almost 40% of the total olefins in the naphtha, and the C5 portion of the naphtha is about half olefins.

Aromatics, on the other hand, are concentrated in the back end of the naphtha. More than 54% of the CB and higher material is aromatic.

The implication of these data is clear. Reducing aromatics involves dealing with the heavy end of the FCC naphtha. On the other hand, controlling olefins in the naphtha in more of a front-end problem.

ENDPOINT REDUCTION

Preventing the formation of aromatics in FCC naphtha is not possible with today's technology, so minimizing pool aromatics will probably require aromatic rejection via distillation or extraction.

In the study, we chose to examine the effect of reducing FCC naphtha endpoint on the yield and properties of the remaining, undercut, light naphtha. Table 4 gives these data for the base Astra conditions.

Decreasing endpoint from 450 F. to 300 F. (TBP) reduces naphtha volume by about one fifth, which has a concentrating effect on both olefins and residual aromatics. The 350 F. cutpoint rejects about one quarter of the aromatics from the corresponding full range cut; the 300 F. cutpoint rejects about half.

As expected, Rvp increases slightly, and the octane of the FCC naphtha is enhanced by the decrease in endpoint.

Fig. 2 shows that the amount of aromatics rejected to heavy naphtha as a result of endpoint reduction is remarkably similar for the different sets of catalysts and reactor temperatures used in the study.

It is disappointing, therefore, that none of these operating approaches have any particular advantage when an endpoint strategy is used to control aromatics.

While the aromatics level is significantly reduced as a result of lowering the endpoint, the olefins content is increased by concentration because the olefin content of the heavy tail is negligible. Fig. 3 shows the level of olefins in 300 F. endpoint naphtha for a full-range material.

Like the full-range naphtha, the light-naphtha olefin content can be significantly affected by catalyst type, reactor temperature, and C/O ratio.

The octane of the light-naphtha cut is still influenced by the olefin content as shown in Fig. 4, so that a tradeoff must be made between octane needs and olefin requirements.

C5 OLEFINS

As previously discussed, traditional approaches to enhance FCC octanes increase gasoline-range olefins. Rejection of aromatics via distillation or extraction will also concentrate olefins by reducing FCC naphtha volume.

For all of the conditions tested, C,5 olefins accounted for 30-40% of the total naphtha olefins present. Although some selectivity differences exist between the different operating strategies, high C/0 conversion appears to be the major driving force. That is, high C/O forces total olefins lower and the selectivity to C5 olefins higher.

Of all of the gasoline-range olefins, C,5's are potentially the most valuable, and indeed may actually become a desirable product in an era of reformulated gasolines. Alkylating these materials has recently gained some following as a technique to reduce pool Rvp.7

Alkylating C5's has the double advantage of replacing volatile olefins in the pool with alkylate.

Tests showed XP-750 boosts C5 olefin yield to about 9 LV % of feed over Astra-378; therefore, the catalytic approach (USY and/or ZSM-5) is an efficient way to enhance C5, olefin yield.

C5 olefins also have potential as ether feedstocks .8 The production of TAME is accomplished by reacting isoamylenes (2-methyl butane-1 and 2-methyl butane-2) with methanol.

This high-octane oxygenate could be a valuable contributor to the reformulated gasoline pool.

Fig. 5 shows how isoamylene yield is affected by catalyst type, reactor temperature, and C/O ratio. High reactor temperature, ZSM-5, and high-octane catalysts increase naphtha olefinicity, and these same factors also increase isoamylene yield. The strongest influence is that of octane catalyst, followed by ZSM-5 and reactor temperature.

Isoamylene selectivity is also affected by these variables. XP-750 gives the highest isoamylene/total C5= ratio, and in fact approaches the thermodynamic equilibrium ratio of about 0.5. ZSM-5 also increases isoamylene selectivity, but reactor temperature has very little effect.

C4 OLEFINS

Alkylate derived almost exclusively from FCC olefins and isobutane accounts for about 12% of the gasoline pool and a significant percentage of pool octane barrels.

And it is well-known that C4 olefins, in particular, yield the highest-quality alkylate.9 10

Table 5 shows the breakdown of the C4 stream as a function of catalyst and operating conditions in the study. This comparison was made at constant total C4 production because this can be a common bottleneck. Total butylene yield is illustrated in Fig. 6 and isobutane-to-butylene ratio in Fig. 7. The same trends in olefin production that were established earlier are repeated here.

Both catalytic approaches (USY and ZSM-5) enhance butylene selectivity compared to reactor temperature. As expected, the conditions that produce high olefin yield result in a decrease in isobutane/butylene ratio. The correct operating strategy is, thus, dictated by balancing alkylation capacity vs. isobutane availability.

The isobutylene component of the C4 olefins is of obvious interest as a feedstock for MTBE and ethyl tertiary butyl ether (ETBE). Currently, about 25% of FCC isobutylenes are finding their way into ether production in the U.S.

In Figs. 8 and 9, isobutylene yield and selectivity, as a function of total C4 yields, are shown to be influenced in the same way as the other olefinic products.

XP-750 gives the highest ratio of isobutylene to total butylenes.

The addition of ZSM-5 also increases isobutylene selectivity. In this case, the thermodynamic equilibrium ratio is about 0.45.

XP-750 is the only catalyst that comes close to this level, while Astra-378, which has the advantage in isobutane yield for alkylation, is far below XP-750. In general, variables that increase isobutylene decrease isobutane.

All of these data point out that a refiner that adopts a strategy of light olefin maximization for alkylation and/or ether requirements has several variables from which to choose.

The octane catalyst approach, perhaps in conjunction with ZSM-5, would be an effective option for reaching this goal.

ACKNOWLEDGMENT

The authors express their thanks to T.J. Dougan and J.A. Goytisolo Jr. for their valuable assistance in this work.

REFERENCES

  1. Young, G.W., et al., Davison Catalagram, Davison Chemical Division, W.R. Grace & Co., Baltimore, No. 76, 1987.

  2. Young, G.W., and Weatherbee, G.D., paper presented at the AlChE annual meeting, San Francisco, November 1989.

  3. Young, G.W., Weatherbee, G.D., amd Davey, S.W., Paper No. AM88-52, National Petroleum Refiners Association annual meeting, San Antonio, 1988.

  4. Cotterman, R.L., and Plumlee, K.W., paper presented at the 198th ACS national meeting, Miami Beach, Fla., 1989.

  5. Pappal, D.A. and Schipper, P.H., "ZSM-5 in Catalytic Cracking: Gasoline Composition Analysis,' 198th ACS national meeting, Miami Beach, Fla., September 1989.

  6. Wear, C.C., Davison Catalagram, Davison Chemical Division, W.R. Grace & Co., Baltimore, No. 80, 1990.

  7. Transcripts, Section IV, B-12, of the National Petroleum Refiners Association annual Question & Answer Session on Refining and Petrochemical Technology, 1989.

  8. Chase, J.D., and Woods, H.J., "MTBE and TAME-A good octane boosting combo," OGJ, Apr. 9, 1979, p. 149.

  9. Hammershaimb, H.U., and Shah, B.R., "Trends in HF Alkylation," Hydrocarbon Processing, June 1985.

  10. Masters, K.R., and Prohaska, E.A., "Add MTBE Unit Ahead of Alkylation," Hydrocarbon Processing, August 1988.

  11. Thomas, R.X., "Worldwide Production Trends for Fuel Ethers," AlChE national meeting, 1989.

Copyright 1990 Oil & Gas Journal. All Rights Reserved.