Timothy J. Dougan
Grace Davison
Baltimore
Ulrich Alkemade, Balbir Lakhanpal
Grace Davison
Worms, Germany
Lori T. Boock
Grace Davison
Columbia, Md.
A vanadium trap technology called RV4+ has demonstrated in a variety of commercial fluid catalytic cracking (FCC) units its ability to reduce vanadium on equilibrium catalyst by more than 20%.
Reducing vanadium loading increases microactivity and zeolite surface area retention, confirming that RV4+ protects zeolites from vanadium deactivation.
Sulfur competition had prevented some previous traps from working commercially, but was not a factor with the new trap. The technology can save refiners millions of dollars per year in catalyst costs, or allow them to process feeds containing higher vanadium concentrations.
VANADIUM TRAPS
Vanadium, while not the only contributor to FCC catalyst deactivation, frequently dictates the rate of fresh catalyst addition to the FCC unit needed to maintain activity. For this reason, improvements have been made to both zeolites and matrices to minimize the effect of vanadium.'
Another method of protecting the catalyst from vanadium deactivation is to use "traps" that prevent the vanadium from contacting the catalyst. Vanadium traps are not new but frequently have shown more promise in laboratory testing than has been realized commercially." One explanation for this discrepancy is that sulfur, present in commercial operations, has been known to interfere with the traps' ability to capture vanadium.
This article focuses on seven commercial trials of a new, rare-earth-based, dual-particle trapping technology known as RV4+.
Vanadium reduction on equilibrium catalyst (Ecat) as high as 23.4% was observed with as little as 4.3% of the trap in inventory. The trap's affinity for vanadium was as high as six times that of FCC catalyst.
Improvements in Ecat microactivity directly related to higher zeotite surface area were observed-a sure sign that the effects of vanadium were being mitigated. One refiner was able to reduce fresh catalyst additions by 20% while maintaining activity. No sulfur interference was observed during the commercial trials.
Refiners can take advantage of this technology in several ways. The most obvious is to process cheaper feeds with higher metals concentrations. But the amount of resid fed to the unit also can be increased. Another option is to reduce fresh catalyst additions. Spent catalyst disposal costs would be decreased as well.
DEACTIVATION MECHANISM
All crude oils contain metals, the most common of which is vanadium. Vanadium is usually associated with organometallic compounds found in the higher-boiling-range fractions. Distillation concentrates the vanadium in the fractions typically sent to the FCC unit.
Vanadium quantitatively deposits on the catalyst, destroys the zeolite, and contributes to increased coke and hydrogen yields. Many other factors, such as inherent catalyst stability, regenerator conditions, and average catalyst age, also play a role in determining the activity of FCC catalyst. The dominant role of vanadium, however, is demonstrated by plotting equilibrium microactivity vs. vanadium level for the entire industry (Fig. 1).
The primary vanadium-containing compounds in FCC feed and resid are metal-centered porphyrins. 2 3 Within these molecules, vanadium is present in the + 3 or +4 oxidation state .4
In the FCC reactor, the porphyrins decompose and the reduced vanadium deposits onto the surface of the catalyst, along with coke. In the regenerator, the coke bums off the catalyst and the vanadium oxidizes to the + 4 and + 5 state. If present at high enough concentrations, the oxidized vanadium may be present as V205 on the catalyst surface.
Mobile vanadium species, such as vanadic acid, VO(OH)3, are formed by reaction of the oxidized vanadium with steam in the FCC regenerator. Vanadic acid migrates into the catalyst particle and destroys zeolite, reducing catalyst activity.5
Vanadic acid also can move from particle to particle, which accelerates the deactivation of fresh catalyst particles. Additionally, in the presence of steam, vanadium may react with rare earth elements in the zeolite, forming rare earth vanadates and further destabilizing the zeolite.6-8
A vanadium-tolerant catalyst is required to process high-metal feedstocks. The use of a vanadium trap is one way to achieve this goal. A vanadium trap should be able to capture and immobilize the vanadium in a nondestructive form.
With a successful vanadium trap, the rate of vanadium migration to the trap should be significantly faster than the migration of vanadium to the zeolite. Additionally, the capacity of the trap must be great enough to remove a considerable amount of vanadium from the catalyst.
This property is related to the pore structure of the trapping component. The trap also must irreversibly bind the vanadium so that it cannot migrate back to the catalyst. Finally, competition by other acidic species (sulfur) for reaction with the trap should be negligible.
HISTORY OF TRAPS
One common type of vanadium trap contains a basic species to react with and neutralize the acidic vanadium compounds. The vanadic acid reacts with the basic component of the trap according to this general reaction scheme:
2MeO + 2VO(OH)3 --
Me2V207 + 3H20
Compounds that have been proposed to react by this mechanism include barium titanate, calcium titanate, calcium carbonate, strontium titanate, and magnesium oxides. 39 10 All these basic compounds theoretically react with vanadic acid and bind it in the trap, and have proved effective in laboratory evaluations. Sulfur competition, however, negatively affects the performance of these traps in commercial units. 3 5
Sulfur oxides in the FCC regenerator flue gas can react with these alkaline-earth metals to form sulfates. On the basis of thermodynamic data, the formation of calcium and barium sulfates is favored over the formation of vanadates at typical regenerator conditions. 5 11
Other trap materials may or may not be affected by sulfur competition, depending on the SO, concentration and regenerator conditions. In any case, the effect of sulfur competition cannot be overlooked when designing effective vanadium traps.
APPROACH
A vanadium trap can be either integral to the catalyst particle or contained in a separate particle. Grace Davison employs both technologies. Each has advantages and disadvantages and neither has emerged as vastly superior to the other in testing to date.
Integral traps are closer to the zeolite and may provide better protection in units with low vanadium mobility, such as those in partial burn or those with low steam partial pressure. Incorporating the trap in the catalyst particle, however, can change the selectivity of the catalyst and its physical characteristics.
Dual-particle or separate traps, such as Davison's RV4+, must have attrition and fluidization properties similar to FCC catalyst.
The advantages of such traps are: They do not change the selectivity of the base catalyst and they have a higher theoretical capacity for vanadium capture.
Performance evaluation of dual-particle traps is usually simpler. They often can be isolated from equilibrium catalyst and analyzed for vanadium capture. Confirmation of preferential pickup on integral traps tends to be a bit more qualitative.
A possible disadvantage of dual-particle traps is that they are more dependent on vanadium mobility than are integral traps.
VANADIUM MOBILITY
Since the effectiveness of a separate-particle vanadium trap depends on the ability of the vanadium to migrate from the catalyst to the trap, a number of laboratory experiments and commercial evaluations were designed to measure vanadium mobility.
Vanadium mobility can be discussed in terms of intraparticle mobility, interparticle mobility from the catalyst to the trap, and interparticle mobility from the trap to the catalyst (irreversibility).
INTRAPARTICLE MOBILITY
In commercial trials, "Time of flight secondary ion mass spectrometry" (TOF SIMS) images of Ecat and electron microprobe scans of the RV4 + fraction show that, while the vanadium concentration may be higher on the surface of the particles, vanadium is found throughout the RV4+ particle.
The SIMS scan also shows that vanadium is found throughout the catalyst particle. This shows that, over time, there is intraparticle mobility of vanadium in both catalyst and RV4 + particles.
INTERPARTICLE MOBILITY
Fresh RV4 + blended with equilibrium catalyst (90 wt % catalyst/10 wt % RV4 +, 50 wt % catalyst/50 wt % trap, and 10 wt % catalyst/90 wt % trap) was steamed by the cyclic propylene steaming (CPS) procedure." During this short steaming time (20 hr), the RV4+ removed vanadium from the Ecat as shown in Table 1.
The trap also removed some of the nickel. This interparticle Ni mobility, although somewhat contrary to current beliefs, has been observed in other CPS experiments with catalyst blends and metal traps, and in certain commercial RV4 + trials. A publication exploring these results is in progress.
In summary, the test results and the literature suggest that, while significantly less mobile than vanadium, nickel is mobile in an FCC unit under certain conditions.'314 This is clear evidence that the vanadium trap not only picks up metals from the incoming feed, but also can remove "old" mobile metals directly from the Ecat by interparticle migration.
Microactivity testing (MAT) of the 90% Ecat/10% RV4 + sample compared to a 100% Ecat sample steamed by CPS was also performed. Results show a dramatic improvement in yields and activity (Table 2).
Electron microprobe scans of cyclic metal impregnated (CMI) Residcat-767Z4 + , which incorporates RV4+ technology, visually confirm this interparticle mobility. 12 Because the catalyst and the RV4+ were simultaneously exposed to the metals during the CMI procedure, the rate of deposition of vanadium on catalyst and trap surfaces should be similar.
The catalyst particles, however, contain virtually no detectable vanadium. In contrast, the RV4 + particles containing the active trap component contain much greater vanadium concentrations. This is another indication of particle-to-particle vanadium mobility.
IRREVERSIBILITY
Since the migration of vanadium from the catalyst to the trap was clearly observed, an obvious question is: Does the vanadium migrate back to the catalyst from the trap?
To answer this question, a vanadium trap was impregnated with 1.0 wt % vanadium and steamed. This trap was then blended with 90 wt % fresh catalyst and steamed by CPS for 20 hr. After steaming, the catalyst and trap were density-separated and analyzed for vanadium. Results are presented in Table 3.
As shown in the table, less than 6% of the vanadium migrates back to the catalyst. This is an insignificant amount of the total vanadium transferred. Additionally, because the vanadium on the catalyst may migrate back to the trap over time, the degree of reversibility may actually decrease with time.
MEASURING PERFORMANCE
The ultimate measurement of trap performance is whether microactivity increases at constant fresh-catalyst addition and metals levels, or whether the improved stability provides the flexibility to reduce additions or process higher-vanadium feed. It is helpful to have additional methods of determining success.
Dual particle traps frequently can be separated from equilibrium catalyst if the densities of the two materials are slightly different. The two fractions then can be analyzed for vanadium.
If the trap is preferentially picking up vanadium, this confirms that the technology is working, even if there is too little trap in the inventory to improve microactivity or if another variable is at work reducing microactivity.
Davison has found the ratio of vanadium on the two fractions to be an effective means of confirming trap performance. This ratio is referred to as the pickup factor (PUF) and is expressed as:
PUF = V on trap, ppm
V on Ecat, ppm
Another useful comparison is the amount of vanadium "removed" from the equilibrium catalyst. "Removed" is somewhat misleading because the measurement represents not only vanadium that has migrated from the equilibrium catalyst to the trap, but also vanadium that has deposited directly on the trap.
Had the trap not been there, all of the vanadium would have deposited on the equilibrium catalyst, so "% V removed" represents, in essence, the amount of vanadium removed. Mathematically it is expressed as:
% V removed = (V on trap, ppm)(wt % of trap in inventory)
(V on total blend, ppm)
COMMERCIAL RESULTS
Seven commercial trials have been conducted using RV4+ technology. A wide range of base catalysts, vanadium levels, unit designs, and unit operations (including a partial-bum operation) were studied. Table 4 summarizes the key results.
The wt % vanadium removed varied from approximately 5 to 25%, and correlated well with the amount of trap in inventory (Fig. 2). In all cases, the targeted amount of RV4 + in inventory was 5%. While much of the laboratory work was done with 10% blends, a 5% blend was chosen for the commercial trials to minimize possible dilution effects.
Several units did not attain the 5% level due to previously, scheduled turnarounds. In two of the cases where the targeted level was achieved - Trials A and G - vanadium removal exceeded 20%. Interestingly, vanadium removal in the partial-burn operation, Trial F, was not that much less than that of the full-burn operations.
The amount of RV4 + in inventory is a function of time. It stands to reason that the percentage vanadium removed also varies with time. Fig. 3 illustrates this relationship.
In general, for the same number of days on the trap, the unit with the greatest trap concentration in inventory provided the highest vanadium removal. There were significant variations, however, suggesting that additional factors influenced vanadium removal.
PUF was plotted against the amount of trap in inventory (Fig. 4). As the RV4 level increased toward 4%, PUFs of 4-6 were common. The highest value observed was 6.8.
PUF also could be plotted as a function of time, as was vanadium removal in Fig. 3, but the resultant plot does not contribute anything not already apparent in Figs. 3 and 4.
Removing vanadium or showing a special affinity for it are the first signs of a successful trap. As previously noted, the real indicator of success is whether the activity of the equilibrium catalyst increases.
Taking a closer look at Trial G, the refiner's objective was to run higher-metals feed without increasing fresh catalyst additions. Figs. 5a and 5b track vanadium and microactivity as a function of time.
Shortly after the introduction of RV4 + to the unit, the vanadium level increased by more than 1,500 ppm. Normally this increase would have reduced activity significantly. But microactivity remained relatively stable.
This stability had the effect of redefining the MAT-vs.-vanadium deactivation curve for the unit (Fig. 6). The shift to the right, or toward higher metals levels, can be attributed to increased zeolite surface-area retention (Fig. 7).
In this case, the same percentage. of zeolite surface area is being retained at 1,500 ppm greater vanadium concentration with constant catalyst additions. This clearly shows that the trap is protecting the zeolite from deactivation.
SULFUR COMPETITION
Sulfur competition has been the Achilles' heel of other vanadium trap technologies.
While RV4 + picks up some sulfur, that does not appear to hinder its performance. In fact, RV4 +'s propensity to pick up sulfur diminishes rapidly as its ability to capture vanadium increases, suggesting that the rare earth vanadates formed are more stable than rare earth sulfates. This decreasing sulfur affinity can be seen in Fig 8.
Also evident in Fig. 8 is the great amount of vanadium on RV4+, about 11,000 ppm. Fig. 9 shows the vanadium on the trap plotted versus the number of days on the trap for all of the trials. The highest vanadium concentration achieved was greater than 16,000 ppm.
The theoretical saturation point, however, is several times greater than this. Given more time in the unit and more favorable conditions for vanadium mobility, the trap should continue to pick up vanadium.
Fig. 10 confirms the trap's ability to capture vanadium long after the trial ended. There may, however, be some factors at play limiting the amount of vanadium the trap can capture.
Vanadium level on equilibrium catalyst, average catalyst and trap age, regenerator internals, steam partial pressure, and the amount of excess oxygen in the regenerator are just a few variables that may influence trap performance. More commercial data are needed to sort out their respective roles.
In Fig. 10, the fact that the decay curve appears quite normal indicates good unit retention of the trap. But does this mean that trap performance is unit specific? To a certain extent, yes.
It is well known that the MAT-vs.-vanadium deactivation curve is different for different catalysts. Where on this curve a refiner is operating will influence the response to trapping technology.
What may not be obvious, however, is that different unit designs have different deactivation curves and that the mode of regenerator operation also influences the MAT-vs.-vanadium relationship. Fig. 11 shows this relationship for four different unit designs all using the same family of catalyst.
Design B has a much larger inventory and older average catalyst age, accounting for its difference from the others. Even unit designs A, C, and D, however, which have similar inventories and catalyst age, show a difference at elevated vanadium loadings.
Fig. 12 breaks Fig. 1 into two curves, based on carbon-on-regenerated-catalyst (CRC) level.
Those units with CRC less than 0.05 wt % are likely to be in a full-burn mode, while those with CRC greater than 0.15 wt % are more likely operating in partial-burn mode. The more severe operations (CRC
ECONOMIC VALUE
There are three ways to capitalize on this technology:
- Keep feed and catalyst constant and allow conversion to increase as microactivity increases.
- Substitute a higher-metals feed or process more resid.
- Reduce fresh catalyst addition.
Options 1 and 2 probably will provide the greatest return, but these options require knowledge of a refinery's yields and economics to determine returns. In other words, Options 1 and 2 are very unit specific.
Some generalizations, however, can be made about Option 3.
In the case of RV4 +, assuming a PUF of 6 at 5 wt % RV4 +, one can reasonably expect to reduce fresh catalyst additions by 20%. Fig. 13 illustrates the annual savings for units with varying size and addition rates.
The savings range from several hundred thousand dollars per year for a 30,000 b/d unit adding a typical 0.25 lb/bbl, to about $4 million/year for a 90,000 b/d unit adding 1.0 lb/bbl. Each refiner must evaluate the economics on an individual basis, as these savings are sensitive to the PUF obtained. For instance, at a PUF of 12, the savings would be doubled.
Commercial experience indicates that the economics for using RV4- are favored when equilibrium catalyst vanadium levels exceed 2,500 ppm and unit conditions are such that a PUF of 5-6 can be obtained.
THE FUTURE
As noted above, increasing the PUF enhances the value of the trap. Much of Davison's effort, therefore, focuses on achieving this goal.
An improved RV-type technology with higher vanadium pickup has been developed by Grace Davison Research.
This technology was compared to RV4 +.
The two traps were blended with 90 wt % catalyst, impregnated with 2,000 ppm Ni and 3,000 ppm V and steamed by CPS. Results indicate that this new technology has a pick-up factor twice that of RV4+ and can remove up to 50% of the vanadium on the catalyst.
REFERENCES
- Flitter, R.E., Rheaume, L., Welsh, W.A., and MaGee, J.S., OGJ, July 6, 1981, p. 103.
- Wormsbecher, R.F., Peters A.W., and Maselli, i.M., Journal of Catalysis, Vol. 100, 1986, pp. 130-37.
- Scherzer, J., Octane-enhancing FCC Catalysts, Marcel Dekker Inc., 1990, New York.
- Occelli, M., Catal. Rev.- Sci. Eng., Vol. 33, 1991, pp. 241-80.
- Fluid Catalytic Cracking II: Concepts in Catalyst Design, Occelli, M.L., Ed., American Chemical Society, Vol. 452, 1991, Washington, 6.C.
- Mauge, F., Courcelle, J.C., Engelhard, P., Gallezot, P., and Grosmangin, J., 7th International Zeolite Conference, 1986, Elsevier, Tokyo, pp. 803-09.
- Gallezot, P., Feron, B., and Engelhard, P., Zeolites: Facts, Figures, Future, P.A. Jacobs and R.A. van Santen, Eds., Elsevier Science Publishers, Amsterdam, 1989, pp. 1281-90.
- Feron, B., Gallezot, P., and Bourgogne, M., journal of Catalysts, Vol. 134, 1992, pp. 469-78.
- Rajagopalan, K.R., Cheng, W.C., Suarez, W., and Wear, C.C., NPRA Annual Meeting, 1993.
- Rawlence. D.J., Gosling, K, Staal, L.H., Chapple, A.P., Preparation of Catalysts V, G. Poncelet, P.A. Jacobs, P. Grange, and B. Delmon, Eds., Elsevier Science Publishers, Amsterdam, 1991, pp. 407-19.
- Lange's Handbook of Chemistry, 13th ed., Dean, J.A., Ed., McGraw-Hill, New York, 1985.
- Rajagopalan, K.R., Cheng, W.C., Suarez. W., and Wear, C.C., NPRA Annual Meeting, 1993.
- Occelli, M.L., Psaras, D., Suib, S.L., Journal of Catalysts, 1985, 96, p. 363-70.
- Gerritsen, L.A., Wijngaards, H.N.J., Verwoert, J., and O'Conner., P., Akzo Catalyst Symposium, 1991, Akzo Chemie, Amsterdam, p. 109.
Copyright 1994 Oil & Gas Journal. All Rights Reserved.
