METAL BELLOW SEAL FAILURE.

[shyamshins]

Dear members,
I am facing problem of seal failure repeatedly.This is SINGLE INSIDE CARTRIDGE seal. Seal is found leaking from QI mentioned in drawing.

Fluid : DOWTHERM - A
Suction press:2kg/cm2
Discharge press:3.90 bar
Box Press: 2.83 bar
Vapour Press:2.16Bar.
Tempreture :310 Deg.celcius
Specific Gravity: 0.81Cp - 1.06 cP
Viscosity: 0.22 - 4.29 cP
RPM : 1475

This seal leakes after 20 to 25 days in operation.Bellow doesnt gets cracked.Pls. help me to resolve this problem.

 

[flexibox]

Few questions for you
1 do you have a quench turned on?
2 is quench steam or N2?
3 if seal leaks after 20 days sounds like a face leak.
4 what is the condition of the face and gaskets?
5 does the pump cavitate?
6 I assume it is continuous operation not stop start?
7 is the process fluid clean? Any solids in the system?
8 do you have any pictures of the seal faces?

Await to hear back from you.

 

[shyamshins]

1) Yes quench is turned on.
2) Quench is nitrogen.
4) Some deposits are observed between faces & gaskets are intact.Black coloured deposits are observed at I.D. of bellow.
5) No
6) Yes its continuous operation.
7) Process fluid is clean. can there be solids in dow therm A ?
8) No.

 

[flexibox]

What is meant by some deposits are observed between the faces? This infers solids enter the faces and faces come out scored? If not I suspect face distortion. I don't particularly like the rotating face arrangement! Have the seals been operating ok then started to fail?
Sounds like you are coking, do you have sufficient N2 as quench?
Yes there can be solids in any system, is the system clean? I suggest if it fails again check seal face flatness. Check N2 supply is on permanently. I saw a case last week operators used quench for start up only then turned it off! Send some photos of seal faces, and is seal repaired by original seal manufacturer? If not that could be another problem!

 

[Airsporter1st]

Unfortunately, I am unable to open this drawing.

Dowtherm is notorious for production of cokes and if the bellows has a carbon face, these tend to build up at the I.D. to the point that the face geometry/balance actually changes - at which point the opening force starts to exceed the closing forces and significant leakage ensues.

As Flexibox says, you do need to check the N2 quench, but also a stationary bellows with a flow guiding baffle for the quench N2 which ensures that it arrives at the critical juncture of the faces, is the best solution for this application.