Nov-2023

Identifying high catalyst losses

Multiple approaches to assessing FCC catalyst losses require an assessment of the unit’s mechanical system’s impact on fluidisation efficiency.

Warren Letzsch
Warren Letzsch Consulting PC

Article Summary

High catalyst losses cost the refiner in many ways. Economically, the cost of extra catalyst can be substantial, with fresh catalyst selling for $3,000 to $4,000/short ton. Yields may also deteriorate, which might cost more than the extra catalyst. When catalyst losses go up, the fines (0-40 microns) content usually goes down.

The viscosity of the fluid beds increases while the quality of fluidisation declines. Bubbles are larger, and mass transfer declines, affecting catalytic reactions, stripping, and regeneration. Catalyst circulation may become erratic; more hydrocarbons could enter the regenerator, and afterburn might be higher.

Higher catalyst losses cause additional wear in the flue gas system. More catalyst goes to the secondary cyclones, plenum chamber, flue gas slide valve, orifice chamber, and waste heat steam generators. The tertiary recovery system also sees higher costs, and increased catalyst going to tankage can be expensive.

Data is required to answer some fundamental questions about the problem. This includes the catalyst densities and bed levels in the unit, the catalyst circulation rate, the catalyst entrainment rates, and the level of catalyst in each cyclone dipleg. Superficial velocities and the transport disengaging heights need to be determined. Cyclone efficiencies and the losses from each side of the unit are needed, and these numbers represent the loss history of the unit.

Assessing catalyst losses
The catalyst inventory of the unit can be calculated from the pressure balance around the reactor, regenerator, and connecting piping equipment. Pressure differentials around the gas distributors and grids should be a part of the balances since changes in the delta Ps could indicate a plugged or eroded piece of equipment. The reactor side must include the gas plant since the wet gas compressor suction pressure is the starting point for the reactor pressure. The catalyst levels in the regenerator and stripper are determined from the taps in the side of the vessels. The equation for the bed level contains the bed density and is:

Bed height = 62.4 (meter reading)/bed density

If the meter is a manometer containing water (62.4 lb/ft³ density), and the level taps are placed 62.4 in apart, the reading will give a direct measure of the bed height above the lowest tap. Many refiners do not measure the bed density but assume one. It should be measured because the value changes with catalyst type, gas velocity and density, and particle size distribution. A foot difference can affect the dipleg seals and the catalyst level in the cyclone diplegs. The catalyst dipleg height is given by:

Catalyst height in dipleg = (cyclone delta P + bed density x length of submerged dipleg)/dipleg density

It has been recommended that losses be plotted vs the dipleg level to find the point where excessive losses occur to ensure cyclone flooding is avoided. The levels in the diplegs are not constant and fluctuate with the delta P across each cyclone.

The total losses are equal to the total catalyst additions minus the catalyst withdrawals and minus the gain in unit inventory. The inventory will be constant if the catalyst densities stay constant and the regenerator bed level does not change. Changes in the fresh and equilibrium catalyst silo inventories are a part of the calculations. When catalyst is added to the silo, it packs with a density equal to the ABD of the catalyst. If it sits for days, the density approaches the compacted density. If some aeration of the silo is possible, a more consistent density can be used for silo calculations.

Catalyst and additive analyses are needed for the fresh materials as well as the equilibrium catalyst and fines from the reactor and regenerator. This will help track the losses of each catalyst and additive added and the possible cause of increased losses. The particle size distributions give clues to the loss mechanism. If the APS of the ecat increases significantly, the losses are probably mechanical or operational. An increase in the 0-20 micron content of the fines is due to attrition. If the losses increase and the 0-40 micron content of the ecat does not go down or even increases, then attrition is occurring.

Measuring accuracy
The losses from the reactor are the bottoms yield times (multiplied by...) the catalyst content. Since most units have a close-connected riser termination system, the loading to the cyclones is the catalyst circulation rate. Therefore, the efficiency of the separation system is easy to obtain. It is worthwhile to look at the efficiency vs the catalyst circulation rate to see if lower loss rates occur at lower catalyst circulation or if losses go up when operated past the design feed rate.

Regenerator losses are the total losses minus the reactor losses. The recovered catalyst from the downstream collection equipment subtracted from the regenerator losses will give the amount leaving with the flue gas. This could correlate with the stack opacity. If a representative sample of flue gas is taken at a point just downstream of the regenerator, the particle size distribution and the number of solids would also give an independent measure of the regenerator losses.

Measuring regenerator cyclone efficiencies requires an estimate of the catalyst entrainment rate. The entrainment rate is shown in Figure 1 as a function of the gas density, the particle density and the effective gas velocity. The effective velocity equals the superficial gas velocity when the bottom of the primary cyclone is located above the transport disengaging height (TDH), and the air is well dispersed through the bed.

In Figure 2, the entrainment rates for a 60,000 b/d catalytic cracker are shown for various superficial velocities. A velocity of 2.5 ft/sec gives an entrainment rate about half the catalyst circulation rate, while at 3.3-3.5 ft/sec the two are about equal. Above this rate, the entrainment rate becomes very large. The TDH also increases with velocity. Even a 0.1-0.2 ft/sec increase may overload the cyclones when operating at 3.5 ft/sec. 

If increased losses are occurring, there are fundamental questions that need to be answered. These are listed in Table 1. The most serious of these is item 4, when the losses are high and have suddenly occurred.

Causes of failure
The causes of a sudden increase in catalyst losses can be the mechanical failure of a cyclone due to corrosion/oxidation, high-temperature stress/creep, and/or the high differential expansion of the metal components. Weld or metal failure can be caused by cooling sprays impinging on the cyclone system. Cyclones have been known to fall to the bottom of the regenerator, resulting in losses of up to 50 t/d. Water sprays are not usually included in modern designs due to the damage they can cause.