PTQ Q4 2023 Issue

Cyclone design parameters

Primary inlet velocity Secondary inlet velocity

55-65 ft/sec 65-75 ft/sec

Primary outlet Secondary outlet

90 ft/sec

130-150 ft/sec

Fragmentation dominated

Flux rates

Cone section

90 lb/ft 2 sec (max) 110-135 lb/ft2 sec 150 max

Diplegs

Abrasion dominated

Table 2

100

300 Threshold velocity 200

400 500 600 Jet velocity, ft/sec (%)

700

most likely to occur during start-ups when moisture may be present. Uncovering the dipleg and applying a pressure bump may clear the problem. Catalyst losses can soar if the bed level falls below the height to seal the cyclones. This is more likely to happen if the losses exceed catalyst additions. Each unit has an opti- mum range of regenerator bed levels and should operate accordingly. Increasing the bed height will increase the level of cat- alyst in the dipleg. If the level is too high, the losses will go up, and the fines (0-40 microns) are preferentially lost. The stack can ‘puff’ periodically as the diplegs repeatably fill with catalyst and then empty. Damaged valves Damaged trickle or flapper valves used to seal the diplegs can also cause large losses. Coke can form on the reactor valves and keep them from moving and sealing properly. Damage from impinging catalyst or aeration jets might also compromise collection efficiency. Gas flowing up the dipleg can increase the losses. If the valve stays open, the gas can flow up the dipleg. The vertical flapper plates used to seal the secondary diplegs have elliptical openings, which need to be partially closed with a plate welded to the top of the opening to prevent reverse gas flow (see Figure 5 ). A shroud is used to keep the flapper from flipping up and staying open. The flapper should be hung at a three-degree angle to ensure it properly seals. Sudden increases in catalyst losses are frequently associated with outages caused by a power failure. If the regenerator slumps, there is a chance the air distributor will become partially plugged. A damaged or partially plugged air distributor will increase catalyst losses. The air goes through only a portion of the holes, leading to high catalyst loading to the cyclones above the active portion of the grid. This increases the pressure drop through those cyclones and can cause flooding and increased erosion. Losses can more than double in this situation. The bed level in the regenerator will not be uniform and may affect the dipleg seals. More catalyst attrition may occur due to the higher velocities through the open nozzles. The bed density readings may provide a clue to the problem. If they are not close to one another, the higher reading will indicate a lack of fluidisation in that area. Cyclone temperatures may also show the problem. Plate air distributors have a catalyst seal around their

Figure 4 Attrition regimes divided into two domains

perimeter that can fail and allow air to go up the regenerator walls. Hot spots occur around the regenerator, and catalyst losses may increase when the blown seal grows. Bed densi- ties could also significantly change. Measuring attrition The attrition resistance of cracking catalysts has changed dramatically over the years. Earlier catalysts had Davison Attrition Indexes (DIs) that measured 25-35. The ecats were 10-15 numbers lower. Soft catalysts with DIs of 40-50 could double the catalyst losses from the unit. One unit went from 2 to 5 t/d in catalyst losses due to a catalyst change. There was also concern that harder catalysts might ‘ball-mill’ softer catalysts and notably increase losses. When the environmental limits on catalyst losses were imposed in the 1970s, all the catalyst suppliers started making cata - lysts with lower DIs and higher densities to reduce losses. Typical DIs are now between 4 and 8, and densities (ABDs) have increased from 0.60-0.70 to 0.75-0.90. Hardness is

Gas probably leaks through top of valve

Catalyst likely to de-aerate

Flow restrictor (Half moon plate)

1/3 d

Trickle-valve outlet area is 1/3 blanked o

Figure 5 The lines should be designed for minimum pressure drop, with few bends and be as short as possible

PTQ Q4 2023