NEW CATALYST IMPROVES HEAVY FEEDSTOCK HYDRO-CRACKING

Arend Hoek, Tom Huizinga, A.A. Esener, Ian E. Maxwell, Wim Stork
Koninkiijke/Shell-Laboratorium
Amsterdam

Frans J. van de Meerakker
Shell Internationale Petroleum
Maatschappij BV, The Hague

Oscar Sy
Shell Canada Ltd.
Oakville, Ont.

A new zeolite-Y-based, second-stage hydrocracking catalyst, designated S-703, has been developed by Shell.

Laboratory testing and commercial use show it has significantly improved performance with respect to gas make and middle-distillate selectivity in processing heavy feedstocks when compared to a Shell catalyst, S-753, developed earlier. Further, the new catalyst exhibits enhanced stability.

Extensive laboratory testing of the S-703 catalyst has been carried out under single-stage, stacked-bed, two-stage-flow, and series-flow conditions. Commercial experience with the new catalyst has now been obtained in several units.

To date, the commercial results have confirmed the laboratory results in terms of the superior, heavy-feedstock processing performance of the new catalyst in all respects. Because the trend toward heavier feedstocks is expected to continue, it is likely that catalysts such as S-703 will find increasing application in hydrocrackers in the future.

PROCESSING TRENDS

The importance of hydrocracking as an oil-conversion process is growing because of a number of structural trends. One of these trends is the shift toward an increasing middle distillate/gasoline product ratio in refineries (relative to catalytic cracking). This is particularly evident in rapidly developing countries; for example, in the Pacific Basin and the Indian subcontinent.

Another important trend is the increasing environmental legislation being introduced worldwide. Because hydroprocessing technology inherently yields products with generally low concentrations of heteroatoms (e.g., sulfur and nitrogen) and aromatic components, it is attractive from an environmental viewpoint.

In addition, the quality of the middle-distillate products from the hydrocracking process is, in general, very high. For example, when the product qualities of typical gas oils from hydrocracking, thermal cracking, and catalytic cracking processing routes are compared on the basis of density and cetane number (Fig. 1), the hydroprocessing route is clearly superior.

These differences in product quality can be directly attributed to the hydrogen content of the products, as Table 1 indicates.

Differences in hydrogen content reflect processing differences, such as hydrogen addition (as in hydrocracking) and carbon rejection (as in catalytic cracking).

The large differences in, for example, the sulfur contents of typical kerosine and gas oil products (also shown in Table 1) further emphasize the aforementioned environmental advantages of hydrocracking.

Another increasingly important application of the hydrocracking process is at the interface between oil refining and petrochemicals. In particular, as shown in Fig. 2, a single-stage hydrocracker can be used to manufacture not only typical oil products such as naphtha and gas oil, but also a heavy paraffinic fraction known as "hydrowax." This is an attractive feedstock for a steam cracker.

The quality of this hydrocracker-derived, steamcracker feed is also directly related to its hydrogen content. As indicated in Fig. 2, a relatively high proportion of the hydrogen taken up during processing is concentrated in this hydrowax.

Finally, another trend which is also structural to the oil refining industry is the increasing heaviness of feedstocks.

This is a response by the industry to the relatively low value of heavy fuel oil. This trend is expected to continue, and possibly even increase, in the future due to environmental concerns related to the poor quality of heavy product fractions.

The shift toward processing heavier feedstocks is also reflected in the hydrocracking process within the Shell Group, as shown in Fig. 3.1 This further demonstrates that this trend was quite dramatic during the 1980s.

The marked increase in 1979 is due to the start-up of a hydrocracking unit designed by Shell that operates on vacuum gas oil. Feedstock heaviness increased gradually until 1983, when the introduction of a zeolitic catalyst developed by Shell allowed a further increase in feedstock heaviness.

The start-up of a hydrocracking unit processing a mixture of vacuum gas oil and butane deasphalted oil is responsible for the further increase in feedstock heaviness in 1988. An important factor in the ability to achieve this significant increase in heavy feed processing has been the development of new catalysts, including those based on zeolites.

This article is intended to show developments within Shell related to new zeolitic catalysts for hydrocracking heavy feedstocks, and their applications in different process configurations; in particular, single-stage stacked beds, two stage, and series flow. The data presented are based on both laboratory and commercial operating experience.

ZEOLITIC CATALYSTS

The first commercial hydrocracking processes made extensive use of amorphous inorganic oxides as catalysts (in particular, aluminosilicates).2 But it was soon realized that the use of crystalline materials such as zeolites, with their well-defined pore structures, offered substantial advantages.3

One of the major advantages of zeolite catalysts, as compared to their amorphous analogues, is their markedly reduced coking tendency, which results in significantly improved catalyst stability. As mentioned, this improvement in catalyst performance is particularly important when heavy feedstocks are being processed, because high rates of catalyst deactivation can become a major process constraint.

Another important characteristic of zeolite-based catalysts is their uniform and relatively high concentration of Bronsted acid sites.4 This results in high levels of hydrocracking activity, which is reflected in lower start-of-run reactor temperatures, thus further enhancing catalyst life.

Typical performance data for amorphous aluminosilicate and zeolite catalysts are shown in Fig. 4. The benefits of the zeolite catalyst, in terms of both initial activity and stability, are apparent. The experiments from which these data originate were performed on a feedstock similar to Feedstock B (see Table 2).

The first zeolitic hydrocracking catalyst formulated by Shell that went into commercial use-HTY-was based on research at Koninklijke/Shell-Laboratorium in Amsterdam (KSLA). HTY consisted of an Ni/W hydrogenation function and a Y-sieve cracking function. It was introduced in the Shell refineries in 1983.

Compared to the older amorphous catalyst, it gave longer life, particularly with heavier feedstocks. And regenerability proved to be excellent. The HTY catalyst formulation was further optimized and is currently manufactured under the designation S-753. Some properties of this catalyst are listed in Table 3.

The zeolitic catalysts also exhibit different selectivities than amorphous systems. First, the zeolites are generally more naphtha-selective, and thus less middle-distillate selective, than their amorphous counterparts. Second, the selectivity of the current zeolitic hydrocracking catalysts has tended to deteriorate with on stream time, for heavy feedstocks.5 This can result in excessive production of low-value gaseous byproducts.

The new S-703 zeolite catalyst has been developed to overcome some of these disadvantages without sacrificing the reported advantages of the zeolitic system. Some typical properties of S-703 are listed in Table 3. In the following, one of the features differentiating S-703 from HTY and S-753 is briefly discussed.

It has been found that the performance of Y-sieve-based hydrocracking catalysts is markedly influenced by the unit cell size of the zeolite crystals. This can be varied by changing the "stabilization" conditions applied during manufacture of the zeolite raw material.

Reduction of the zeolite unit cell size reflects expulsion of aluminum from the zeolite structure. It also reduces the number of acid Sites6 (Fig. 5) and generally affects properties like mesopore volume and amount of extra-lattice alumina.

HTY-type catalysts typically contain zeolite Y with high acid site densities.

It has subsequently been found that, by using lower-unit-cell-size zeolite, higher middle-distillate and lower gas selectivities can be achieved without major detriment to activity. This is shown in Fig. 6, where the relationship between the kerosine selectivity, observed in hydrocracking heavy Feedstock B (Table 2), and the zeolite unit cell size has been plotted for S-703 and S-753.

In addition, the stability with respect to selectivity during catalyst aging is markedly improved. For Feedstock B (Table 2) this has been demonstrated in the laboratory by means of an aging test in which the S-703 and HTY catalysts were compared under the same process conditions.

As shown in Fig. 7, S-703 is much more stable with respect to gas make as a function of catalyst age.

As discussed later, this improved yield stability is also observed in commercial practice with heavy feedstocks under recycle conditions. It should be pointed out that our studies have focused on heavy feedstocks. With lighter feedstocks, the selectivity effects reported above tend to be smaller.

BED CONFIGURATION

Single-stage hydrocracking is the simplest configuration, often having a once-through flow of feed. In recent years the performance of such single-stage configurations has been substantially improved by the introduction of "stacked beds," combining hydrotreating and zeolitic catalysts within the same reactor configuration.

This can give much higher cracking activities than hydrotreating catalysts alone. The activity gains that typically can be achieved by means of such stacked beds when processing heavy vacuum gas oil of Arabian heavy origin (Table 2, Feedstock A) are shown in Table 4.

A drawback of these stacked beds is the lower selectivity to middle-distillate products caused by the introduction of the zeolite component. With the S-703 catalyst, this potential disadvantage has been substantially reduced, while the activity gain is maintained.

Although, in general, stacked-bed configurations may give somewhat lower quality products than single-bed systems, it has been shown that the combination of Criterion's C-324 catalyst and S-703 can produce high-quality products, particularly at high conversion levels.8 With regard to catalyst life, the C-324/S-703 combination exhibits significant improvements over the single beds for heavy feedstocks. This can be attributed to the low coke-forming characteristics of the zeolite component.

These results pertain to the relatively high pressures (100-200 bar) which prevail in conventional hydrocrackers. However, the stacked-bed catalyst systems can be equally well adapted under mild hydrocracking conditions (40-80 bar).9

SERIES FLOW/TWO STAGE

The new S-703 catalyst has also been evaluated for the more conventional two-stage and series-flow modes of operation. In the two-stage hydrocracker, the first-stage hydrotreated effluent is separated into light and unconverted products. The latter are then routed to the second-stage reactor containing the hydrocracking catalyst (Fig. 8).

By contrast, in series-flow the whole charge from the first-stage reactor is sent directly to the second-stage reactor containing the hydrocracking catalyst. Thus, light hydrocarbon products, H2S, and NH3 are present in the gas phase in the cracking zone (Fig. 8). It was the advent of high-activity, nitrogen-resistant zeolite catalysts that turned the lower-cost series-flow mode of operation into an attractive option.

In Fig. 9, the selectivity patterns obtained for the S-703 and HTY catalysts are compared under second-stage (once-through) conditions, with heavy Feedstock B (Table 2). Under these conditions, the new catalyst produces about half the amount of light gases and significantly more kerosine, without loss of activity.

In Fig. 10, the results obtained under simulated series-flow (once-through) conditions using pretreated Feedstock A (Table 2), are compared.

Again, the conclusions are similar: A significant reduction in gas make is achieved with an increase in distillate product for the S-703 catalyst at activity levels comparable to those obtained with the HTY catalyst.

Extended pilot plant studies under full-recycle conditions have demonstrated that this improved performance can be sustained for long run periods. In these studies, Feedstock A (Table 2) was used, with the S-703 catalyst in combination with Criterion's C-424 first-stage catalyst.

In Table 5, the product qualities of various fractions obtained from the series-flow recycle pilot-plant studies are given.

In these studies, a heavy vacuum gas oil of Arabian origin (Feedstock A) was used as feedstock.

These data indicate that the properties of the tops are good, as exhibited by a relatively high iso/normal ratio. The naphtha fraction exhibits properties which are indicative of excellent reformer feedstocks. Furthermore, both the kerosine and gas oil fractions are high-quality products based on their attractive smoke points, cetane indices, and low sulfur contents.

In view of the results of these extensive laboratory studies, it was decided to load the new S-703 catalyst into two Shell-designed, series-flow (heavy feed) hydrocrackers. The results obtained in both commercial units have fully substantiated the performance expectations based on the laboratory testing.

For example, in Case 1, vacuum gas oil of Arabian origin is used as feedstock (Fig. 11). Not only is the initial gas make significantly reduced, but even more importantly, the gas make is quite stable throughout the operation. This reduced gas make also has enabled the plant to increase throughput, with substantial economic benefits compared to alternative debottlenecking options. Moreover, S-703 has displayed a substantial gain in life.

These results for both catalysts are, in fact, in line with expectations based on the laboratory catalyst aging test using heavy feedstock (Fig. 7). In addition to the reduced gas make, increased yields of middle distillate (in particular, kerosine) were obtained in both commercial units. This had been expected on the basis of the laboratory testing.

The full product yield patterns for S-703 and S-753, which demonstrate this shift in product yield, are compared for the Case 2 commercial unit in Fig. 12. The product properties achieved under commercial conditions with the S-703 catalyst are also in line with the data obtained in the laboratory.

The highly satisfactory behavior of S-703 is also demonstrated by the fact that both refineries have decided to purchase additional batches of S-703 catalyst.

In addition, quite recently the S-703 catalyst was loaded into a two-stage hydrocracking unit.

REFERENCES

Maxwell, I.E., "zeolite Catalysis in Hydroprocessing Technology," Catalysis Today, Vol. 1, 1987, pp. 385-413.
Ward, J.W., "Preparation of Catalysts," Design and Preparation of Hydrocracking Catalysts, Part 3, Poncelet, G., Grange, P., Jacobs, P.A., Eds., Elsevier Science Publishers, Amsterdam, 1983, p. 587.
Breck, D.W., Zeolite Molecular Sieves, Wiley Interscience, New York, 1974.
Haag, W.O., Lago, R.M., and Weisz, P.B., "The Active Site of Acidic Aluminositicate Catalysts," Nature, Vol. 309, 1984, pp. 589-91.
Reno, M.E., Schaefer, B.L., Penning, R.T., and Wood, B.M., "New Hydrocracking Catalyst for High Quality Distillate Production," item no. AM-87-60, NPRA annual meeting, San Antonio, 1987.
Pine, L.A., Maher, P.J., and Wachter, W.A., "Prediction of Cracking Catalyst Behavior by a Zeolite Unit Cell Size Model," J. Catal., Vol. 85, 1984, pp. 466-76.
Scherzer, J., "The Preparation and Characterization of Aluminum-Deficient Zeolites," Catalytic Materials: Relationship Betweeen Structure and Reactivity, ACS Symposium Series 248, Whyte, T.E., Dalla Betta, R.A., Derouane, E.G., Baker, R.T.K., Eds., American Chemical Society, Washington, D.C., 1984, pp. 157-200.
Maxwell, I.E., and Esener, A.A., "Advances in Hydrotreating Catalysts," Improved Hydrocracking Performance by Combining Conventional Hydrotreating and Zeolitic Catalysts in Stacked Bed Reactors, Elsevier Science Publishers, Amsterdam, 1989, pp. 263-71.
Gosselink, J.W., Van de Paverd, A., and Stork, W.H.J., Catalysts in Petroleum Refining, "Mild Hydro- cracking: Optimization of Multiple Catalyst Systems for Increased VGO Conversion," Elsevier Science Publishers, Amsterdam, 1989, pp. 385-97.