Gordon W. Brown
Refining Consulting Services
Englewood, Colo.
Direct measurement of fluid catalytic cracking catalyst fluidizability provides refiners a tool to routinely check for circulation problems in an FCCU.
If a circulation problem develops, comparative test data may show quickly if the problem is catalyst, operating condition, or hardware related.
Refiners prefer direct measurement of any parameter over a calculated or an inferred method. The fluidizability of FCC catalyst is no exception.
Catalyst properties, operating conditions, and unit design may affect how a catalyst flows through a unit. When there is a catalyst circulation problem, one rarely knows whether it results from changes in catalyst, operating conditions, or hardware.
Geldart and Radtke state, "The direct test of an aerated bulk property is clearly preferable. The only disadvantage is that it requires a relatively large amount of powder......" 1
Fluidization properties of FCC catalysts have been directly measured by a smaller scale test apparatus developed from a larger scale apparatus described in the literature. The smaller unit is more convenient to use because of the smaller sample size and apparatus employed. The apparatus has been validated by test data compared to published literature data.
The apparatus was used to evaluate fluidizability by measuring the standardized collapse time (SCT) and the minimum fluidization and bubbling velocities (Umf and Umb)
The smaller apparatus operates with less than a gallon of sample. This eliminates Geldart and Radtke's reservation regarding the fluidizability tests.
Nineteen equilibrium catalyst samples from 10 FCC units were tested in this new apparatus. Fresh FCC catalyst samples from five of these units were also tested. One unit had circulation problems, and samples were tested both before and after changes were made to improve circulation.
PAST METHODS
In the past, most refiners did not use any method to determine the fluidizability of FCC catalyst. When circulation problems were encountered on a fluid unit, they were attributed to insufficient catalyst fines (more catalyst fines would improve fluidizability). For some enlightened engineers who wished to determine fluidizability, it was done only on the basis of correlations.
One of these correlations in use is the velocity ratio: the ratio of minimum bubbling to minimum fluidization velocity, Umb/Umf. Both velocities are described later. Often engineers calculated both of these velocities due to the lack of apparatus and the tedious nature of the tests.
Another calculated fluidizability method is Raterman's F-Prop Method, which appeared in 1985 .3 This method required several catalyst parameters to calculate fluidizability of the catalyst for a particular unit.
The value determined by F-PROP must be calibrated for any given commercial unit. It cannot be used in the absolute sense to compare catalysts taken from different units.
Both of these correlative methods fall short of the results gained when one uses direct testing.
STANDARDIZED COLLAPSE TIME
In their paper, Stenge, et al. 2 describe the use of the bed collapse test to measure the fluidizability of a catalyst. They cite a reference 6 that goes back to 1967 for this technique.
Their results indicate the test should be used on a routine basis to compare the fluidizability of equilibrium catalyst at different times.
Geldart and Radtke took it a step further and devised the standardized collapse time. Interestingly, they cite an earlier 1953 reference 7 for the bed collapse test.
The SCT uses similar apparatus to the bed collapse test. Air passes through a bed of catalyst from the bottom to fluidize it.
Then the time required for the bed to collapse upon termination of aeration is measured. Additionally, the SCT employs measurements of the bed expansion, so the distance the bed falls is included.
The SCT thus is time per distance rather than just time. Geldart and Radtke state, "The usefulness of the collapse rate test as a diagnostic tool lies in combining the bed expansion with the collapse rate to give a standardized collapse time ... Ranking of performance in a commercial unit by F-PROP, Umb/Umf, or maximum stable bubble size is not as good as by standardized collapse time."
Geldart and Radtke correlated the SCT and the measured velocity ratio. The resulting equation (Table 1, listed as Equation 13) contains a power term (1.5) on the velocity ratio, Umb/Umf. This means that the SCT is more sensitive to changes in powder properties than Umb/U,f.
MINIMUM VELOCITIES
The pressure drop across a fluidized bed is plotted against the superficial velocity as shown in Fig. 1. On this diagram the minimum fluidization velocity, U,f, is usually defined as the intersection of the horizontal fluidized bed line EDF and the sloping packed bed line, OAB.
From a practical point of view it usually coincides with the velocity when the bed first expands, as indicated in Fig. 2.
The minimum bubbling velocity, Urb, usually is at the initial peak of bed expansion (Fig. 2), but should be confirmed by observation of relatively large bubbles bursting at the surface of the bed.
Umf is used in many correlations in the literature. Investigators have tried to derive good correlations for its calculation, using fines content, density, viscosity, particle size, etc.
Abrahamsen in his master's dissertation states, "Of the 50 or more correlations available to predict Umf, the large amount of scatter goes some way to explaining why there are so many." 5
The best of these correlations was selected and is presented in Geldart's book, as Equation 2.28. (See the footnotes of Table 1.) In the equations shown in Table 1, pp is the particle density, pg is the vapor density, dp is the particle diameter, u is the gas viscosity, F is the fraction of fines less than 45 u, hd is the height of the dense bed, hs is the height of the aerated bed, Uc is the catalyst velocity, and Uw is the velocity of the vapor.
Abrahamsen reports only four correlations to calculate the minimum bubbling velocitY, Umb. The best is equation 2.29 or 2.30 in Geldart's book.
Equation 2.30 is a simplification when a powder is fluidized with air at ambient conditions and F (fraction of fines less than 45 u) is about 0.1. Table 1 shows calculated values of U,f (and Equation 2.28) from vendor supplied densities and particle size analyses. Table 1 also shows measured test results for Umf.
All of the calculated values are significantly below the measured values, indicating that perhaps a 51st correlation is needed. Abrahamsen also states in his dissertation, "Therefore, it is unlikely in the foreseeable future that an ideal Umf correlation will exist."
Table 1 likewise shows calculated values of Umb (Equations 2.29 and 2.30) with the corresponding measured values. The calculated values of Umb correlate much better with the measured test results than those for Umf.
VALIDATION OF APPARATUS
To quantitatively measure the validity of the smaller apparatus, a mathematical relationship correlating measured values was selected that has a theoretical basis.
In the paper, "Behavior of Gas Fluidized Beds of Fine Powders,"8 Abrahamsen and Geldart show a relationship between the maximum nonbubbling bed expansion ratio, Hmb/Hmf, (height, non-bubbling and height, expanded bed) and the velocity ratio, Umb/Umf. Their equation (Equation 26) equates the bed expansion to the velocity ratio raised to the 0.22 power as shown below.
[SEE FORMULA]
They also show that theoretically the exponent should be 0.25. They state their experimental work agrees well with the theoretical as shown by the 0.22 exponent.
The exponent by regression for the 19 catalysts tested in this article is 0.24. In fact, it is closer to the theoretical than in the above reference primarily because FCC catalyst particles are more rounded (more ideal) than the numerous powders included in the data in the reference.
When the calculated values of the velocity ratio of Table 1 are plotted on the same graph as the above equation, the points do not fall on the line. This indicates the poorer dependence of the expansion ratio on the calculated velocity ratio.
Densities measured in the test apparatus compare reasonably with catalyst vendor supplied densities.
APPLICATION OF APPARATUS
When an FCC unit has trouble circulating catalyst, it often is a time consuming process to pinpoint the reason and to take corrective action. Numerous articles on the subject have been prepared, including the recent, "Troubleshooting FCC Circulation Problems-Practical Considerations."9
This article outlines the indicators of circulation problems, causes, and a procedure for troubleshooting such problems. To evaluate if changes in catalyst properties are the cause, a calculated velocity ratio is used.
This has been shown to be less indicative than either the measured velocity ratio or, more particularly, the SCT.
The SCT is the most useful of the tests made in the apparatus. Periodically testing the equilibrium catalyst for the SCT quantifies the fluidizability of the catalyst.
When problems of circulation develop in a unit, the trend of the SCT will aid in determining if the problem is catalyst related. More frequent tests can be made to track the results of efforts made to resolve the problem.
Whenever a new catalyst is added to the operating unit, more frequent tests should be made to quickly spot a trend, either good or bad.
It is desirable to operate a standpipe close to, but above, the minimum fluidization velocity (U,f) for maximum pressure buildup. This is not practical, due to compression of the gas in the standpipe.
At the minimum bubbling velocity (Umb) bubbles begin to form in the catalyst bed, so it is desirable to be below this gas velocity, The larger the differential between Umb and U,f, the wider the range of fluidized flow without bubble formation.
This is why the velocity ratio is useful as a measure of fluidizability. Therefore, knowledge of both Umf and Umb can be helpful in designing or operating a standpipe.
Additionally, both velocities are employed in many correlations. The apparatus can measure both Umf and Umb with very good repeatability.
Each test uses a weighed catalyst sample. From the volume measured during testing, the apparent bulk density (ABD) is calculated at any bed velocity.
If the particle density is known, then the voidage may be calculated. Both ABD and voidage are used in a number of correlations, and for adjusting standpipe aeration.
TEST RESULTS
Table 2 presents the results from testing 19 catalyst samples from 10 FCC units from four refining companies. All values in the table were measured in the apparatus, excepting the vendor densities and particle size analyses, which were not available for all samples.
Refining Company A with four refineries ranked its FCC units in order of relative ease of circulating catalyst. The SCT results gave the same ranking of fluidizability as the ease of circulating catalyst.
None of the units had poor catalyst circulation. The SCT ranged from a low of 23.3 to a high of 33.5.
Company D on a single unit, took samples both when circulation was good and poor on two separate occasions. The SCT showed very clearly the markedly reduced fluidizability of the catalyst when the circulation was poor.
On those occasions, the SCT was 14.2 and 20.3. Corresponding high values of the SCT were 24.2 and 29.2 when circulation was satisfactory.
Three changes were made to the unit in a period of 26 days which caused the SCT to go from 20.3 to 29.2. These changes included:
- Smaller average particle size of fresh catalyst
- Recycle of recovered catalyst fines
- Lower regenerator superficial velocity.
Changes in particle size were not sufficient to indicate improved fluidizability. It appears that a unit probably does not have circulation problems until the SCT is in the lower 20's.
Companies C and D supplied fresh catalyst samples as well as the equilibrium samples. The SCT for two of the five fresh samples was significantly lower than the corresponding equilibrium samples. Two were higher, but one was essentially the same (Company C, Unit 2).
The fresh sample for Company C, Unit 3 was calcined at 1,300 F, for 16 hr and the tests performed again. The SCT of the calcined sample came up from 29.0 to 36.2, reaching the same value as the equilibrium sample (36-3). Density fell as expected.
This result says that the fluidizability of that fresh catalyst sample improved by calcining. There were no other changes, such as particle size, to reach the same fluidizability as the equilibrium sample.
These results imply that there are parameters involved in fluidization correlations not yet determined. However, fresh and equilibrium samples were both included in the validation equation (bed expansion ratio vs. the measured velocity ratio), so the results appear valid for fresh catalyst in the state as measured.
REFERENCES
- Geldart, D., and Radke, A. L., " The Effect of Particle Properties on the Behaviour of Equilibrium Cracking Catalysts in Standpipe Flow," Powder Technology, Vol. 47, 1986, pp. 157-165.
- Steenge, W.D.E., Dane, F., and Parker, W.A., "Fluidization Behaviour of FCC Catalyst: Effect of Catalyst Properties and Gas Distribution," Katalistiks' 8th annual Fluid Cat Cracking Symposium, Budapest, Hungary, June 1-4, 1987.
- Raterman, M.F., "FCC catalyst flow problem predictions," OGJ, Jan 7, 1985, p. 87.
- Geldart, D., editor, Gas Fluidization Technology, John Wiley & Sons, New York. 1986.
- Abrahamsen, A.R., Dissertation, University of Bradford, 1980.
- Rietema, K., Processing International Symposium on Fluidization, Eindhoven, 1967, p. 154.
- Diekman, R., and Forsythe, W.L., Industrial Engineering Chemistry, Vol.45,1953, p. 1174.
- Abrahansen, A.R., and Geldart, D., Powder Technology, Vol. 26, 1980, pp. 35-46.
- Englehard, The Catalyst Report, TI-844, February 1989.
Copyright 1990 Oil & Gas Journal. All Rights Reserved.
