Random Column Packing Escaping Column 2

[riddance]

Hey all -

Need some wisdom troubleshooting one of my batch stripping columns. Its 3' in diameter, and has a ~25' tall bed of 3/4" 316SS random packing rings.

Lately we've been finding intact, undamaged rings in the strainer at the column bottoms pump. It would make sense that rings are slipping by the bed support plate, but we've opened the bottom manhole, and the plate is perfectly centered with a 3/8" gap between the plate and the wall -- too small for rings to be slipping by. The centering screws are all fully extended properly and haven't vibrated out of place.

Last time we opened the top body flange of the column to PM the liquid distributor, we realized the top of the shell was significantly out-of-round. The bottom is still round. We strip at ~360F -- could thermal cycling due to batch operation be widening the gap between the support plate and wall enough to allow packing to drop through?

The bigger puzzler is that we are finding intact column rings in the reactor pump strainer as well. If rings were coming through the top of the column, they'd have to bypass the hold down plate, distributor, mist eliminator, a 50' vapor line, condenser, separator, and pump to make it back to the reactor. Through the bottom they'd have to bypass the support plate, bottoms strainer, and bottoms pump undamaged. How in the world??

If hydraulic shock was to blame, I'd expect to see some broken/crumpled rings, but all the ones we've found are intact.

Have any of you folks experienced wandering packing before? What other potential failure modes am I missing here?

Many thanks for your assistance,

-JAR

 

[racookpe1978]

Without getting into your proprietary lineups, what is the P&ID for those pumps, tower, and locations where you are finding the packing?

I'm suspicious (as you are) of the out-of-round tower walls, but also of over-filling that "might" be pumping the packing back over and underneath into the pump suction.

 

[georgeverghese]

Think we need to see the PIDs for this section of the plant to be able to help. Does sound like there is no way these packing rings could escape the stripper bottom holdown grating. Maybe you've got reverse flow through some line that is leading back to the stripper - a pump min flow recycle line going back to the column packing section that has no check valve in it, or maybe a poorly installed check valve ?

 

[riddance]

Following the adage that "one good measurement is worth 1000 theories," I had Maintenance open the column again. The packing hold-down plate had been pushed upward and torn away from the supports. This allowed it to "float" in the 6" tall area below the distributor, and likely gave the packing support plate enough room to tip up and allow packing into the bottoms.

This immediately screamed "high bottoms liquid level" to me, so we rigged up a new hold-down plate, topped off the packing, buttoned up, and ran a batch. Found out that the high-level alarm for the column level was reading/tripping improperly, so we were getting liquid impinging on the steam inlet. This is almost certainly the source of our internals damage, and this explains packing ending up in the column pump strainer.

Still unclear as to how packing is making it back to the reactor. You folks seem to be pointing to the P&IDs - I will be tracing lines to ensure the as-built drawings are correct, and holding an FMEA with Operations/Production to see if there are any flow paths I'm unaware of.

I appreciate the help.

-JAR

 

[racookpe1978]

Thank you for continued information!

My/Our questions about the P&ID is give us a chance to think about "other failure modes" flow paths that might be covered up by the busted plate and supports. I'm suspicious of what I might find, but don't know what that might be.