Visbreaker Fractionator Fractionator reflux seal

[relentlessratt]

can someone provide advice about this service?
i am having frequent seal failures with large amounts of coke buildup in all areas of a double seal. the plans are plans 41/52. coke was found in the primary and the backup seal bellows,faces, pumping ring, gland ports etc.
the product is cooled from 620 F in the heat exchanger to the outlet temperature of about 300 F. is this flush temperature too high? maybe the product is coking with increased temperature generated by seal faces. the seal components are designed for high temperature service with solids - graphalloy packings, hard faces, inconel 718 bellows assembly.
the product has coke chips - the reason for the cyclone separator. other than the coke, faces seem to be in okay condition when disassembled.
i am thinking to go to a plan 32/52. The flush material is compatible but very expensive.

 

[flexibox]

Sounds like you have a dual unpressurized high temperature metal bellows seal. If you are cooling from 620F (327C) to 300F (150C) then what may be coking is the barrier fluid not the Product itself. What is the barrier fluid? Oil or Diesel? Try changing the barrier fluid to Royal Purple GT34 before you change to Plan 32/52.
What does the Plan 52 vessel operate at? Do you have a cooling coil in the veesel? What size is it?
send some duty conditions and we may have a better look at it for you.
trust this helps

 

[relentlessratt]

flexibox,
Thank you for you input. The buffer fluid is shell cassida HF 32. At some point in the past we used to use royal purple GT34 but was changed for some reason (most likely capital cost).
The buffer tank operates very close to atmospheric pressure (flare system pressure of 2 psig max).
The cooling coil in the buffer tank is 1/2" diameter tubing.
The seal sleeve diameter is 2.5 inches.
From inspection of the failed seal components, I found alot of coke on the seal flush side and in the stuffing box. This is why I thought the flush fluid was coking.

The pump product and flush fluid is visbroken gas oil.
Hope i clarified. Thanks.

 

[flexibox]

If you have COKE in the stuffing box then this must be coming from the product not the seal. When a seal has Coking it is normally inside the ID convelutions not the OD as you mention. Check the Cyclone is not blocked, and dirty product is contaminating the seal. It sounds to me that you need to go back to Plan 32 or convert to a dual pressurized seal. Without seeing any pictures you may HAVE to use Plan 32 in either design.
trust this helps

 

[relentlessratt]

flexibox,
you perfectly highlighted my train of thought when i was troubleshooting the failure. I know the product has coke and the cyclone may not always work properly, but i was just wondering if the seal itself was contributing to the coke.
The product is quite dirty which is why i considered plan 32.

thanks for your advice.