TECHNOLOGY Stepwise method determines source of FCC catalyst losses

Ray Fletcher
Akzo-Nobel Chemicals Inc.
Houston

A set of guidelines for fluid catalytic cracking unit (FCCU) monitoring and a logical, stepwise approach to troubleshooting FCC catalyst losses will help process or operations engineers find the causes of such losses.

Periods of high catalyst losses from an FCC unit typically occur directly after a unit turnaround or toward the end of a long operating cycle as cyclone efficiencies begin to deteriorate. As a result, FCC engineers may never have the opportunity to troubleshoot high losses during their entire rotation on the FCC unit.

For this reason, when losses occur, there generally is little experience on which to draw, which may result in less-efficient troubleshooting methods and greater uncertainty regarding the root cause of the losses.

Unit monitoring

A thorough understanding of the entire catalyst system during normal operations establishes the base line data necessary for troubleshooting. A comprehensive, ongoing analysis of catalyst losses includes the following tests:

Catalyst balance -- A periodic catalyst balance will establish normal reactor and regenerator-side losses.
Catalyst balance is the primary method of calculating losses from the regenerator. Losses are calculated as: Fresh shipments 2 fresh hopper inventory changes 2 regenerator inventory changes 2 spent catalyst hopper inventory changes 2 spent catalyst shipments.
Losses from the reactor are calculated using the slurry production and catalyst concentration present in the slurry. Total losses less reactor-side losses yield regenerator losses.
It is difficult to distinguish abnormal operations from normal operations without such a base line. These data should be plotted weekly.
Fresh catalyst physical properties -- Careful attention to the apparent bulk density (ABD), particle size distribution, and attrition index of the fresh catalyst and any purchased equilibrium catalyst (E-cat) will help prevent unexpected loss problems. The use of additives (such as SOx reduction and bottoms conversion additives, ZSM-5, and metals traps) also should be monitored carefully.
Equilibrium catalyst properties -- Careful attention to the ABD and particle size distribution (PSD) of the E-cat also is required. Gradual shifts in the physical properties of the circulating inventory, while the fresh catalyst physical properties remain constant, often indicate deteriorating cyclone performance.
Fines particle size distribution -- Periodic PSD analysis of regenerator and slurry catalyst fines will provide a direct indication regarding cyclone performance and integrity, or a possible attrition source. These data should be plotted monthly. (Interpretation of fines PSD will be discussed later.)
Pressure surveys -- Occasional pressure surveys of the regenerator air and reactor stripper steam manifolds during normal operation are recommended. These surveys will provide base line data if air grid or stripping steam grid damage is suspected and a pressure survey is required.
Line and restriction orifice record -- Keeping complete list of all steam purges, aeration lines, and instrument taps into the reactor and regenerator is recommended. Line size, media (steam, gas, or air), design flow rate, restriction orifice size, and orifice inspection history should be maintained.

If drilled gate valves are used in place of restriction orifices, these valves should be color-coded, with the normal position clearly indicated. This information will simplify troubleshooting efforts if an attrition source develops in the unit.

Troubleshooting losses

A logical, step-by-step approach to troubleshooting high catalyst losses reduces the time spent finding the problem and results in a much quicker return to normal operations. This method uses a four-step approach to quickly identify the root cause of losses:

Review operations
Review catalyst physical properties
Review cyclone performance
Search for attrition source.

Operations review

The process engineer can use a logic diagram as a troubleshooting aid after establishing that losses have increased using the catalyst balance (Fig. 1) (116965 bytes). Calculated high losses should be supported by an increased stack opacity reading, an increased fines collection rate in the electrostatic precipitator (ESP), or a higher slurry fines content.

Step 1 requires a verification that unit operations are normal. Some items to check are:

Increased charge rate -- At charge rates in excess of the design rate, catalyst loading in the cyclones may increase substantially. This increase is the result of cyclone velocities and loadings exceeding design rating because of increased combustion air and catalyst addition rates. Increased velocity generally will lead to higher cyclone efficiency. Catalyst losses, however, will increase with higher velocity if: attrition occurs, a hole develops in the cyclone because of increased erosion, or the dipleg backs up into the dust bowl because of the cyclone pressure balance.
Increased addition rate of catalyst or additive -- Catalyst losses will increase if the catalyst addition rate is increased. The addition rate typically is increased to improve conversion, to control metals, or because of an increased charge rate.
Increased bed level -- Any increases in regenerator bed level over the minimum transport disengaging height will result in increased losses. The pressure gauges used to measure the bed level should be blown down before concluding that the bed level is normal.
Corrections to the bed height calculations should be made if the E-cat density changes appreciably. It is recommended that the bed density be checked occasionally using pressure taps that are sealed continuously with catalyst.
Increased air rate -- Increases in air rate will lead to higher catalyst losses. This is because of increased catalyst entrainment in the cyclones, which may result in cyclone overloading because of both mass flow and increased pressure drop. Attrition may increase if both the catalyst and air pass through a plate or dome distributor. Higher air rates will result from an increased charge rate, higher conversion, poor stripper efficiency, or declining feed quality.
Decreased pressure -- A decrease in the regenerator pressure at a constant air rate will result in increased cyclone velocities because of the volume expansion of the combustion air. Losses will be higher if the increased velocity leads to: catalyst attrition, a hole caused by erosion, or a high level in the dipleg because of increased pressure drop.

The regenerator pressure typically is reduced to help unload the combustion air blower.

A return to normal operations is recommended if any of these variables are contributing to high catalyst losses. Losses should be rechecked after the unit has achieved steady-state conditions. If the source of losses has not been identified, proceed to Step 2 in the logic flow sheet (Fig. 1) (116965 bytes).

Physical properties

Step 2 is to review the physical properties of all purchased catalyst, both fresh and equilibrium. Collecting a representative sample from the catalyst truck or rail car is recommended.

Additive physical properties also should be reviewed. Specific items to check are:

Attrition index -- Watch for any large increases in attrition index over that of the catalyst typically delivered to the unit. An increase in attrition index implies a softer catalyst.
Particle size distribution -- Watch for large increases in the 0-20m range. This will be especially critical for units that do not retain fines efficiently.
Apparent bulk density -- A large decrease in the ABD of the E-cat will result in an increase in the minimum transport disengaging height. With a constant air rate, this will result in increased cyclone loading and may increase losses.

Regenerator and slurry fines should be collected and tested for metals content and physical properties to determine whether the unit is losing primarily fresh catalyst or E-cat. Problems with off-specification catalyst deliveries should be corrected before moving on to Step 3.

Cyclone performance

Regenerator cyclone efficiency and mechanical integrity can be monitored easily for units with third-stage cyclones, ESPs, or flue gas scrubbers. The reactor cyclones can be monitored by analysis of fines filtered or centrifuged from the slurry oil.

This method requires a PSD analysis of the fines using an analytical method with high resolution in the range of 5mand smaller. The output is then used to plot the percentage captured at each particle size.

Possible indications are:

Normal operations -- A set of cyclones in good condition will show a normal bell curve for the fines PSD. A typical regenerator fines PSD for cyclones in normal operating conditions is shown in Fig. 2 (38570 bytes). The peak of the bell curve coincides with cyclone efficiency and, usually, will be in the 30-40mrange.
Damaged cyclone or plenum -- A bimodal fines distribution will be observed if a hole develops in one of the cyclones or the plenum. The second peak will be shifted to the right of the normal peak, in the direction of the full-range E-cat average particle size (APS). The abnormal peak usually appears at 40mor more.
Fig. 3 (38194 bytes) shows the fines PSD for an FCCU with a hole in one of the reactor cyclones. The second peak is the result of full-range catalyst being drawn into the cyclone or plenum.
Consider a situation in which the normal peak of the fines PSD is at 35m, with the full-range E-cat APS at 75m. If a hole develops in one of the cyclones, full-range catalyst will be drawn into the cyclone because of the lower pressure in the cyclone system. The fines PSD then will show the normal peak at 35m, with a second peak appearing at about 60-70m.
The actual position and magnitude of the second peak will vary according to the size of the hole, and whether the hole is in the plenum or a cyclone. The number of sets of cyclones, and whether the cyclone is in the first, second, or third stage, also will influence the position of this peak.
Attrition source -- If an attrition source develops in the unit, a bimodal distribution will be observed, with the second peak shifted to the left of the normal peak. This second peak generally will appear at about 2-3m. The generation of "microfines" is the result of catalyst fracturing because of the presence of a high-velocity stream striking and damaging the catalyst. The regenerator-fines PSD for an FCCU experiencing attrition is shown in Fig. 4 (34737 bytes).
Dipleg flooding -- Flooded diplegs also can be identified using the PSD approach, assuming a good base line exists for normal fines PSD. The 0-40mfraction of the fines PSD will show a significant decrease, while the 40+mrange will show a significant increase.
Differences in cyclone outlet temperatures will be observed when catalyst and air are flowing preferentially to one set of cyclones. "Stack puffing" also may be observed. Flooded diplegs will result from: excessive solids loading caused by poor air and catalyst distribution in the regenerator, a restriction in the dipleg, a stuck trickle valve, or a partially slumped bed. Additionally, excessive cyclone velocities can cause dipleg backup into the dust bowl because of the pressure balance of the cyclone.
Loss of cyclone efficiency -- Cyclone efficiency will decline over the length of the operating cycle. This efficiency loss is due to wear in the cyclone abrasion lining, weld cracks, and small holes. This loss will be observed in the fines PSD plot as a decrease in the magnitude and a shift of the normal peak to a larger particle size. An example of this efficiency loss is shown in Fig. 5 (39636 bytes). Periodic fines PSD analyses are recommended to enable the engineer to monitor efficiency over the FCC operating cycle.

Each of the analyses just described should be substantiated by analysis of the equilibrium PSD. Fig. 6 (57716 bytes) provides an indication of the qualitative shifts that should be observed after the unit establishes a new PSD equilibrium.

In general, the E-cat APS will increase as the 40mcatalyst fraction is lost when a hole develops in one of the cyclones. With an attrition source, the E-cat APS will decline as the concentration of smaller catalyst fragments increases in the inventory.

The shifts in the E-cat physical properties will be more difficult to observe with dipleg flooding. This is a result of a wide range of catalyst particle sizes being lost as the level in the dipleg enters the dust bowl.

Analysis of the regenerator and reactor fines should allow quick determination of whether a hole has developed in a cyclone or plenum, or whether an attrition source has developed. A hole in a cyclone dipleg may be sealed temporarily by raising the regenerator bed level, if the hole is low on the dipleg. Periodic analysis provides the base line information for monitoring and correctly interpreting these data.

Step 3 analysis for refiners without an ESP or flue gas scrubber will be more difficult. One potential method for catching a representative sample of regenerator fines is to use isokinetic sampling.

This technique withdraws flue gas and fines across the entire radius of the flue-gas stack through a stack fitting, in a manner similar to a "pitot tube traverse." Environmental contractors commonly use this method to measure compliance with particulate regulations.

Attrition source

Step 4 should be initiated only after having eliminated operations, catalyst qualities, and cyclone performance as possible causes for catalyst losses. Attrition is caused by a high-velocity stream impacting the catalyst at velocities greater than 200 fps.

The most common sources of attrition in the regenerator are:

Hole in air grid (verifiable through a pressure survey of the air manifold)
Torch oil steam purge rate too high
Excessive emergency fluidization steam
Steam condensate contacting the catalyst
Tube leak in catalyst cooler
Tube leak in bed coils
Excessive catalyst cooler fluidization media
Excessive bed or cyclone velocities (High bed velocities are attributable to high charge rates, high air rates, and low operating pressure.)
Missing or eroded restriction orifices
Instrument-tap steam purge rate too high.
If a loss of equilibrium catalyst microactivity test (MAT) activity and surface area coincides with increased losses, the source is likely high-velocity steam damaging catalyst in the regenerator. The loss of activity and surface area suggests steam deactivation.
Potential attrition sources in the reactor vessel are:
Hole in stripper grid (evaluate using a pressure survey)
Hole in stripper purge ring (evaluate using a pressure survey)
Hole in dome steam line (evaluate using a pressure survey)
Excessive dispersion steam media (steam or lift gas)
Plugged feed nozzle
Poor feed nozzle design or high charge rate leading to excessive velocities
Missing or eroded restriction orifices
Excessive emergency fluidization steam
Excessive instrument-tap purge rate
Steam condensate contacting the catalyst
Excessive cyclone velocity (High bed velocities are attributable to high charge rates, high steam or lift gas rates, and low operating pressure.)
Common sources of attrition in standpipess are:
Excessive lift or fluidizing media velocities
Excessive instrument-tap purge rates
Missing or eroded restriction orifices
Excessive emergency fluidization steam
Steam condensate contacting the catalyst.
Potential attrition sources in the load lines are:
Excessive velocities
Long or convoluted load lines.

The Author

Ray Fletcher is a technical service engineer in the FCC catalyst division of Akzo-Nobel Chemicals Inc., Houston. Before joining Akzo-Nobel a year ago, he worked as a process engineer for Shell Oil Co., and for Texaco Refining & Marketing Inc. He worked on the FCCU and most other process units for those companies. He has a BS in chemical engineering from the University of Washington.