FCC MANUAL 5-The Role of Catalyst in FCC Troubleshooting
FCC MANUAL 5-The Role of Catalyst in FCC Troubleshooting
FCC MANUAL 5-The Role of Catalyst in FCC Troubleshooting
SUMMARY
Fluidized Catalytic Cracking is a complex process where problems are not easily
defined or isolated. This paper presents some background regarding the effects of
catalyst properties on operating problems. In commercial operations it is essential to
separate these effects from the influence of hardware malfunctions or changes in
process conditions. This requires a basic understanding of fresh catalyst management
and the implications of E-cat analyses.
CONTENTS
1. Introduction
2. Catalyst replacement
3. Metal poisoning
4. Catalyst stripping
5. Catalyst losses
6. Circulation problems
Continuous addition of fresh catalyst to the FCC is essential for at least three reasons:
The primary criterion governing fresh catalyst addition is activity maintenance. The
following factors may serve as guidance for catalyst management:
• the optimum E-cat activity level, determined by feed quality, unit limitations
and yield/product quality requirements
• the range of metals, primarily Ni, V, Na, to be tolerated on E-cat, on a
temporary or permanent basis
• the response of E-cat activity to fresh catalyst addition
The optimum catalyst activity level must be determined for any specific operation.
Generally, cleaner feeds allow higher E-cat activity to boost conversion. Heavier and
more contaminated feeds lead to more moderate catalyst activities, due primarily to
heat balance constraints (i.e., delta coke).
1
In principle, three alternative catalyst management policies exist:
A catalyst switch is usually made by adding the new type of fresh catalyst on top of
the old inventory. The rate of replacement depends on the size of the inventory and
the catalyst addition and withdrawal rates. Calculation of the replacement is not
straight forward due to the continuous losses of fresh catalyst together with the E-cat.
With the following equation, the replacement rate can be estimated (figure 1):
For properly operating cyclones the replacement efficiency is about 0.80, indicating
that some 20% of the fresh catalyst is not retained. With a typical addition rate of 2%
of inventory per day, 60% will be replaced in about 60 days. In most cases the effect
of the catalyst switch should be visible after 60% replacement, when most of the
active E-cat is then the new catalyst. Remaining prior catalyst is substantially
inactive.
2
Figure 1: Effect of catalyst addition on replacement
A different approach is to compare the equilibrium analyses with the analyses of the
fresh catalyst before and after the switch. In this approach it is assumed that the
composition of the catalyst lost is equal to the composition of the equilibrium mixture.
The parameters to calculate the replacement are usually the alumina and rare earth
contents. Other (trace) elements can also be applied, such as separate RE elements
like Ce or La, or additives such as TiO2.
In principle all metals entering the FCC reactor with the feed are deposited on the
catalyst. A simple mass balance adequately describes the effect of catalyst addition
on equilibrium metals. This is illustrated in figure 2. In some cases the buildup of
metals is different due to increased pickup of metals by the catalyst fines that leave
the system, or volatility of some metals at high regenerator temperatures.
Nickel (to a lesser extent vanadium and other metals) causes a large increase in the
production of hydrogen and consequently the volume flow to the wet gas compressor.
The addition of passivator (e.g. antimony) to reduce dehydrogenation is a proven
technology and is recommended for units operating at extremely high nickel levels.
For units at modest Ni levels, or operating periodically with Ni-rich feed, the
installation of an antimony dosing system may not be justified economically. In such
cases the nickel tolerance of the catalyst is an important factor. Albemarle has
catalyst zeolite and matrix technology that has demonstrated excellent nickel
tolerance. In this technology, proven superior at Ni levels over 4000 ppm on catalyst,
the nickel is encapsulated and deactivated in large pores with low acidity.
3
Figure 2: Effect of catalyst addition on equilibrium metals
If the production of C1 and C2 increases, the compressor power requirement also goes
up. A reduction in the amount of dry gas is obtained by minimizing thermal cracking
in the riser and reactor. A low regenerator temperature is beneficial, which makes the
delta coke performance of the catalyst important. A higher catalyst activity can also
be considered because it improves the balance between catalytic and thermal
cracking, provided that the delta coke does not increase substantially (figure 3).
Finally, a higher compressor inlet temperature or lower inlet pressure increases both
the gas volume and the power required. It is clear, therefore, that the term
“compressor limited” is too broad to properly define the FCC downstream processing
in terms of adjusting the catalyst management. A more thorough analysis is needed to
establish the actual limits. This requires, among other things, accurate monitoring of
the gas yield breakdown and the E-cat analysis.
4
The effects of metals, especially Ni, V, and Na, can be minimized by adding high
levels of fresh catalyst, or blends of fresh and E-cat or flush cat, to reduce metals
concentrations in catalyst inventory. A plot of Na on E-cat surface area, which
reflects catalyst activity, is shown in figure 4. Vanadium also reduces catalyst
activity, depending on regenerator conditions, and can be combated with metal traps
in the catalyst.
Poor spent catalyst stripping causes an increase in delta coke, and especially in
amount of hydrogen in coke, which result in a decrease of catalyst circulation rate at a
given heat balance duty. The regenerator temperature will increase, and eventually
the feed rate may have to be decreased, or feed quality improved. A stripping
problem may be apparent from an increase in hydrogen in coke calculated from the
flue gas analysis, but care is needed to ensure the flue gas is properly sampled and
analyzed.
The optimum steam rate to ensure proper stripping depends on the catalyst circulation
rate, the stripper design, the catalyst porosity and its equilibrium catalyst surface area.
Good stripping occurs when the delta coke is minimized. For a given design, the
stripping steam rate is optimized by adjusting it to a value slightly above the one
which minimizes the regenerator temperature. This is determined by ramping steam
rate upward until regenerator temperature stops dropping.
The catalyst architecture can directly affect the observed stripper performance.
Entrainment of hydrocarbons in the catalyst pores and the resulting carry-over to the
regenerator will be minimized by use of a large-pore catalyst. With such a catalyst, it
is easier to remove the hydrocarbons from the catalyst. An example of surface area
impact is shown in figure 5.
5
Figure 5: Effect of E-cat surface area on stripper operation
The E-cat fines content decreases with deteriorating cyclone performance. This
usually results in a continuous rise of the average E-cat particle size (figure 6). The
replacement efficiency drops as well. A procedure to troubleshoot this problem is
provided in FCC Manual 5.5; in this section we summarize the issues.
Catalyst is usually lost through the reactor cyclones to the slurry oil or through
regenerator cyclones to the stack (or downstream catalyst removal devices such as
third stage separators or ESP). The losses are acceptable as long as they are less than
target fresh catalyst makeup rate, slurry oil is within specification on ash and metals
content, and flue gas dust emissions are within legal limits. Total catalyst losses can
be calculated by unit and hopper inventories, taking the ABD as an indication of
catalyst density:
6
Figure 6: Increase of particle size with catalyst losses
7
Figure 8: Particle size distributions of equilibrium fines
An increased load can be caused by poor air distribution, leading to localized high
velocities and catalyst entrainment. Increasing regenerator bed level may help to
combat the maldistribution. Once cyclones are badly eroded, little can be done to
keep the unit operational for long periods with controllable losses. If there are holes
near the bottom of diplegs, it may be possible to raise bed level to seal them and
reduce losses.
Measured bed levels are actually pressure drops, and require calculations to determine
the physical height of catalyst. This requires two separate dp-cell transmitters, one to
determine bed density and the other to determine total bed pressure drop. For proper
troubleshooting, these must be calibrated and free of blockages. A change in catalyst
type can cause a change in catalyst bed level if the catalyst density differs. Cyclone
efficiency and fines retention usually improves with increase in E-cat density (figure
9), though fluidization actually is better with lower catalyst density.
8
Figure 9: Effect of catalyst density on fines retention
Proper fluidization of catalyst, both in the beds and transfer lines, is essential to the
operation of the FCC unit. A basic measure of the fluidization properties of FCC
catalyst is given by the ratio of minimum bubbling (Umb) and minimum fluidization
(Umf) velocity as described by the following equation*:
Umb / Umf = 2300 (ρg0.126 μ0.523 e0.716F45) / (dp0.8 ((ρp – ρg) g)0.934
*(D. Geldart and A. L. Radtke, “The Effect of Particle Properties on the Behaviour of
Equilibrium Cracking Catalysts in Standpipe Flow”, Powder Tech., 47 (1986) 157-
165.)
The ratio Umb / Umf , also called bed expansion, is larger than 1. The bed expansion
(figure 10) increases with this ratio and favors smooth circulation. The required ratio
is not an absolute value, but depends on unit characteristics and operating conditions.
9
The following factors appear to improve fluidization, according to the above
correlation:
The E-cat fines have a positive effect on fluidization as they influence both the F45
fines fraction and the mean particle size. The fines content depends on the balance
between cyclone design, unit attrition sources, fresh catalyst fines content, fresh
catalyst attrition index, amount of E-cat fines recycled, and other factors which
determine cyclone loading, e.g. operating pressure, temperature, solids concentration
and gas velocity.
10