AM-95-64

FCC CATALYST TRENDS:

RESPONDING TO THE CHALLENGES OF THE 1990'S

By

Joseph B. McLean, Executive Technology Specialist
Edward B. Bovo, Technical Service Engineer
BASF Corporation - Petroleum Catalysts Group
Houston, Texas

Presented at the

1995 NPRA
ANNUAL MEETING
March 19-21, 1995
The San Francisco Hilton
San Francisco, California

 

Introduction

The FCC unit has traditionally been the major gasoline producer in refinery operations.In recent years the operating objectives for FCCU's have become increasingly complex as refiners have adapted to the changing demands of the marketplace. These factors include octane demand due to lead phaseout, increased use of "opportunity feeds" such as purchased gas oils and resid, implementation of reformulated gasoline production and integration with petrochemical operations as a feedstock producer. FCC process technology advances, such as improved feed distribution and installation of advanced riser termination designs have been incorporated into a growing number of units. These changes have had a marked impact on FCC unit operations. One of the key variables in addressing these continually changing needs is the catalyst. Customization of fresh catalyst properties to address specific unit needs has become standard practice. The interaction of the fresh catalyst properties, catalyst management policy, feedstock and operating variables ultimately dictate the properties of the equilibrium catalyst inventory. BASF maintains an extensive data base of equilibrium FCC catalyst properties for operating units worldwide compiled from our ongoing equilibrium catalyst analysis service. An analysis of some of the trends in activity and selectivity related parameters through the first half of the 1990's offers insight into how refiners have adapted this key part of their FCCU operation to meet their ever changing demands.

BASF's Equilibrium Catalyst Analysis Service

The FCC unit is somewhat unique in that the composition of the circulating catalyst inventory can be continually changed through daily catalyst additions and withdrawals. Thus it is important to monitor the properties of the equilibrium catalyst as a tool to aid in optimizing the unit's operation. BASF routinely analyzes samples of equilibrium catalysts as a service to refiners.

Samples from refineries in North America, South America, and the Asia / Pacific region are analyzed in our laboratory in Iselin, New Jersey while samples from Europe, Africa, and the Middle East are analyzed in our laboratory in Terneuzen, the Netherlands. Selected analyses are also performed in our Japanese laboratory. Together these locations monitor approximately 150 operating units and process over 600 samples per month. About 60% of these samples are from units using BASF catalyst and 40% from competitors. For the U S. alone, about 400 samples per month are analyzed from approximately 100 units. About 40% of these units, accounting for 60% of the samples, are BASF catalyst users. This data base for U.S. refineries was used to generate the comparisons reported here.

The equilibrium catalyst properties monitored include physical chemical, and catalytic properties A detailed description of the tests used and interpretation of their results can be found in Reference 1. For this study, those parameters which reflect important activity and selectivity characteristics were compared and related to developments in the refining industry. These parameters included:

Microactivity (MAT) - Reflects in-unit activity of equilibrium catalyst
Coke Factor (CF) - Reflects catalyst's relative coke selectivity
Total and Matrix Surface Area (TSA/MSA) - impact both activity and product selectivities
Rare Earth Oxides (REO) - Control hydrogen transfer potential, impact activity and product selectivities
Nickel and Vanadium (Ni/V) - Contaminant metals, reflect feedstock being processed and impact activity and product selectivities

The data for each refinery monitored were averaged on a yearly basis for each year from 1991-1994, and the yearly averages were then compared for both overall average trends, geographic trends, and several specific distributions reflecting known differences in specific types of unit and refinery configurations.

Historical Trends

Figures 1 through 10 show the average trends and ranges for the parameters monitored from 1991 through 1994. The average values reported are not weighted to reflect differences in unit capacities or catalyst usage, and the number of samples from individual refineries vary widely as well but this is not reflected in the averages. Thus, all refineries monitored contributed equally to the calculated averages reported.

The MAT activity results (Figure 1) show an upward trend overall, as well as a broadening of the range. One reason for the identified upward trend is the growing number of units which have undergone revamps to "short contact time" cracking, which in turn demands a higher catalyst activity for optimum performance(Ref. 2). This will be discussed in some detail later. The overall average is up approximately one MAT number, from 67 to 68, over the four year time frame. Similar increases were noted for both the first (lower) and fourth (upper) quartile indicating that the upward trend holds across the broad market. Interestingly, Figure 2 shows that for units monitored by BASF, the noted increase in the overall average is entirely due to increases by BASF catalyst users (up by 2.5 MAT numbers on average), while the average activity for competitors' catalyst users remained constant.

Figures 3 and 4 show the trends for total and matrix surface areas respectively. The TSA shows an upward trend, in line with the upward trend in MAT activity, while the MSA shows a slight downward trend Thus there has been an increase in the overall zeolite to matrix ratio in the market. To some extent this is due to the growing use of catalysts such as BASF's Reduxion(Ref. 3) which have been specifically developed to provide coke and gas selectivities commonly associated with lower matrix catalysts while achieving bottoms upgrading performance comparable to higher matrix catalysts. The upper end of the total surface area range has moved progressively upward, with a number of units now operating in the 190 m2/gm range. While the range on matrix surface area has narrowed somewhat, it is still quite broad with nearly a factor of four difference from low to high. From Figures 5 and 6 it is apparent that BASF catalysts on average hold higher surface areas, both total and matrix, than competitors, in line with the higher activities noted in Figure 2. The trends toward higher TSA and lower MSA noted in the overall averages are noted here for both BASF and competitors.

The Coke Factor (Figure 7) shows a slight downward trend, indicative of catalyst technology advances toward more coke selective products. This is especially notable in combination with metals trends discussed later. The broad range in coke factors reflects primarily the wide range in contaminant metals levels.

Rare Earth Oxide levels, which declined through the 1980's in response to octane demand as a result of lead phaseout from gasoline, have remained constant during the early 1990's (Figure 8). Some change might have been expected here due to the implementation of reformulated gasoline production, but to this point no significant move has occurred. This is discussed in more detail later.

Metals levels (Figures 9 and 10) show a wide variation from very low (hydrotreated VGO feeds) to quite high (resid feeds), and on average have increased over the time period shown. The average Nickel level is up about 15% while the average Vanadium increased over 30% in 1994 relative to 1991. This is probably due primarily to heavier feeds being processed, but in some cases a contributing factor is also the continuing use of improved metals tolerant catalysts. This is especially notable when comparing the coke factor trend (decreasing) to the metals trend (increasing), since ordinarily higher metals would lead to higher coke selectivities Figure 11 shows the average relationship between coke factor and metals for 1991 and 1994. While there is a wide variation in this relationship for individual units due to the many specific differences in unit feed, catalyst, and operating variables, on average a 10% reduction in relative coke selectivity at constant E-cat metals is observed.

These average trends verify that technology advances in the FCC catalyst industry have provided the refining industry with more active, more selective more metals tolerant catalysts through the first half of the 1990's. In order to relate the information in our data base to more specific refining industry trends, several more detailed comparisons were generated relating geographical and unit specific trends.

1994 Geographical Trends

For the geographical comparison the U. S. was divided into five regions as shown in Figure 12. The numbers of FCC units included in each region which were monitored in 1994 were:

East Coast - 9
Mid West - 19
Gulf Coast - 34
Mid Continent - 25
West Coast - 7

Figures 13 through 18 show the average regional variations for 1994. West Coast units clearly run higher MAT activities and REO levels than the rest of the country, due to the generally more aromatic crudes processed as well as the need to meet California RFG specifications. High octanes can be achieved from those feeds even at high REO levels, and the need to meet gasoline olefin or Bromine number restrictions also dictates a high activity, high REO operation. Metals levels are highest for the East Coast and Mid West, while the West Coast is significantly lower on average, due to a higher percentage of units having feed hydrotreaters (see below). The MAT activity, REO level, TSA and MSA are all lower than average for the East Coast, in part a reflection of processing higher resid levels as indicated by the higher average metals. This is also the only region with higher Nickel than Vanadium, reflecting different crude sources. The Gulf Coast represents the largest fraction of units in the survey, so it is not surprising that all values are close to the overall average for this region. The Mid Continent region features the highest average MSA, reflecting the high premium placed on bottoms upgrading in this region.

Advanced Riser Termination Design Units

All FCC licensers currently offer advanced riser termination designs for achieving quick and efficient separation of catalyst and oil to minimize non-selective post-riser cracking. These designs are available for both new and revamped unit applications. Proper catalyst selection is a key variable in optimizing unit operation for such designs, with higher activity catalysts preferred for most applications(Ref. 2). Approximately 15% of the units in the survey are known to have undergone revamps to incorporate some version of an advanced termination design through 1994. These include a variety of both original unit designs as well as termination system designs from different licensers. Table 1 shows a comparison of the average 1994 catalyst properties for these units vs. those for the rest of the survey. The average MAT activity, TSA, MSA, and REO are all significantly higher for units with advanced riser termination. Also shown are the average values for the same units in 1991, which in most cases represent pre-revamp designs. There are no significant differences between these values and the overall survey values for 1991. This verifies that the catalyst differences noted for 1994 are in fact in response to needs dictated by the process and hardlware changes from the riser termination revamps (along with any other changes in specific units) and not to any pre-existing features for this set of units.

Feed Hydrotreating

FCC feed hydrotreating offers a number of potential yield and product quality benefits, and clearly impacts the optimization of FCCU operating conditions, including catalyst. A recently published survey(Ref. 4) was used to define which units in the survey had at least some portion of their feed hydrotreated. This represented approximately 30%   of the FCC units monitored in 1994. This classification is not clear cut since percentage of feed hydrotreated, degree of hydrotreating severity, and fresh hydrotreater feed quality all vary widely within this group. Nevertheless, the average catalyst properties shown in Table 2 for units with feed hydrotreating show clear differences from those without hydrotreating. As expected, metals levels are significantly lower. MAT activity, TSA, and REO are all higher for these units. Coke factor is lower (as expected with the lower metals), and MSA is about the same. The use of hydrotreated feed in many cases allows operation at higher severity (conversion and/or throughput) which can be accomplished through the use of higher activity catalysts.

Reformulated Gasoline Production

Refiners faced with the challenge of producing reformulated gasoline have a myriad of options on how to get there. One clear distinction is whether or not they choose to build ether production units {principally MTBE, but also TAME and ETBE) which in most cases will rely on the FCCU to provide iso-olefin feedstocks. FCC catalysts providing low hydrogen transfer characteristics can be a key tool in maximizing iso-olefin production(Ref. 5&6). Properties from this survey which would characterize such catalysts would include low REO, along with high TSA/MSA (to counteract the low REO to some extent) and low MAT activity (to maximize cat/oil ratio). A published list of ether units(Ref. 7) was used to cross reference those FCC units which might have maximum iso-olefin production as an objective. These units comprise about 40% of the units in the survey. Somewhat surprisingly, Table 3 indicates that the average catalyst properties for these units do not follow these trends. In fact, the average REO level is somewhat higher for this group as a whole. Since it is also possible to improve olefin yields by incorporating advanced riser termination designs, those units which fell into this category and also had MTBE production were factored out separately. These units feature higher REO and MAT activity in general, as discussed previously. The properties for the remainder of the oxygenate producing refineries fall in line with the overall survey averages as indicated. Also a higher percentage of the West Coast units have ether production than for the overall survey, and as previously noted the average catalyst properties for the West Coast differ significantly from the rest of the country. However, as is evident from Table 3, even with the West Coast units factored out, the properties for the remainder of the oxygenate producing refineries are not significantly different from the overall average for 1994. Of course, it is not possible from this survey to determine how many refiners have chosen to increase olefin production through other means such as higher riser temperature or use of ZSM-5 additives. But it is clear that on balance the driving force to maximize olefins (at the expense of gasoline) has not been strong enough to force significant catalyst changes. It will be interesting to see if this trend continues through 1995 and beyond.

Resid Feed Processing

It was noted earlier that the average metals levels have increased steadily through the 1990's, which is indicative of higher levels of resid processing. The appropriate catalyst properties for resid processing involve tradeoffs and compromises. Low coke and gas selectivities are important, which would tend to favor low MAT activity, low REO, TSA, and MSA. On the other hand, activity stability under severe regeneration conditions is also desirable, favoring higher MAT, REO, and TSA. Also the need to crack heavy oil molecules to obtain high conversions and low bottoms yields favors higher MSA. So the optimum properties for a "resid catalyst" will be specific for each unit operation and will involve tradeoffs. BASF does not have sufficient information to characterize all the FCC units in this survey according to feed quality differences, but the E-cat total metals levels provide a good relative ranking of the degree of resid processing. Figures 19 through 22 show scatter plots of relevant catalyst properties vs. total metals. There is only a slight downward trend for MAT and TSA, and virtually no trend for REO and MSA with metals. Also the degree of variation in each parameter is just as high for the high metals (resid feed) units as for the low metals (VGO feed) units. Clearly the choice of optimum properties for a resid catalyst is not universal and will depend on specific unit objectives and constraints.

Summary and Conclusions

FCC equilibrium catalyst properties have shown increases in average activity and metals levels and improvements in coke selectivity through the first half of the 1990's. Geographical differences are observed which reflect regional differences in crude types processed. Strong correlations have been noted between catalyst properties and refinery process advances such as the installation of advanced riser termination designs and FCC feed hydrotreaters. No significant changes have yet been noted for the purpose of maximizing iso-olefin production for reformulated gasoline. As feedstocks get heavier, catalysts for resid processing become increasingly important, although the exact properties desired will be unit specific. Overall the increasingly complex needs of individual refiners will demand continued flexibility from their FCC catalyst supplier to customize catalyst formulations for their specific needs.

References

1. "Interpretation of Equilibrium Catalyst Data Sheets", The Catalyst Report, BASF Publication EC-6726, 1994.

2. "Improved FCC Yields Through a Combination of Catalyst and Process Technology Advances", J. B. McLean and E. B. Bovo, NPRA 1994 Annual Meeting Paper AM-94-44

3. "Advanced FCC Catalyst Matrix Technology for Reduced Coke and Slurry Yields'' A. Maglio, C. F. Keweshan, R. J. Madon, and J.B. McLean, NPRA 1994 Annual Meeting Paper AM-94-59

4. Worldwide Refining Survey, Oil & Gas Journal, December 19, 1994

5. "Reformulated Gasoline Catalyst's Impact on FCCU", J. B. McLean, G.S. Koermer, R.J. Madon, and W.S. Winkler, NPRA 1992 Annual Meeting Paper AM-92-45

6. "Iso-Olefin Production for Oxygenates Using IsoPlus", J. B. McLean and A. Witoshkin, NPRA 1993 Annual Meeting Paper AM-93-17

7. "Worldwide Oxygenated Fuel Survey", Fuel Reformulation, March/April 1994