mechanical seal failure of pump with API plan 52. 2

[I2P]

Friends,
We havew a centrigugal pump handling condensate, and the pump has plan 11 and plan 52 for sealing fliud. Recently we have had a seal failure. I have a few basiic questions.
1.0 On the sight glass provided on the seal return line to the seal pot,we could never observe the circulation of fluid.When the pump vendor was contacted during commissioning he said that circulation is happening but is not too distinctly visible.Also there was no significant difference between the seal fluid supply and return line temperature. Again when this was pointed out to the vendor he said that since the fluid is at ambient temperature there wont be any heat buildup. Also we have not provided any cooling water supply to the seal pot as per his recommendation. Upon inspection after failure we found the seal had cracked, and root cause investigation is under progress.I am fairly convinced thatif the seal has been rightly installed then the flow would be clearly visible in the sight glass, and if flow is ensured, then irrespective of process fluid temperature being ambient( around 30 deg C), there would be a heat build up and the return seal fluid would be at a higher temperature.Am I right in thinking that it is a case of seal fluid not circulating at all which has led to seal failure.
Presently we have replaced the seal with a new one from another vendor and trial was taken and we observed circulation from sight glass and also seal return line getting appreciably hotter than the supply.
2.0 Now our problem is that having not connected cooling water to the seal pot we have no means of cooling the seal fluid. Since we do not have cooling water utility we propose to provide external fins on the seal return line.We want to make our own inhouse calculation and arrive at the fin area required. Could anybody suggest how to make these calculations to determine the no, dia,of fin?
Since it has become a warranty issue I would appreciate if someone could clearly tell me that our thinking is right that seal failure has occurred due to improper installation?
Secondly I want to know what can be the cause of fluid not circulating if externally the connections seem ok. Within the seal what is the motive force that forces the seal fluid back to the seal pot?
Thanks to all in advance.





Inappropriate post?

 

[pmover]

looks like u really want an answer.

see posting in heat xsfer forum.

 

[kiwi2]

I2P,
A dual seal installed in a Plan 52 arrangement should have a pumping ring as part of the seal internals. This is usually a scroll or a paddle wheel that literally pumps the seal fluid. You should be able to see fluid movement in the return line sight glass. It will be obvious when the old seal is dismantled if the pumping ring has been installed correctly.
The seal return line should always be hotter than the supply line.
Your seal supplier should be able to assist with heat transfer calcs. Appendix B of API 682(second edition) has calculation for heat generated.
Burgmann Seals supply finned tubing for seal oil systems - product code WEL1000. They should be able to assist with sizing.

Regards

 

[longeron]

Skip the pump vendor, if you still have the seal find your local seal vendor and have him take a look.

You're right, the friction of the seal faces will heat the buffer fluid, it can become quite hot. That being said, you still might not need to have cooling water or fins added to the reservoir or tubing.

Installation of the seal reservoir and tubing are of the utmost importance. Sight glasses, valves, narrow tubing, sharp bends in the tubing, and flowmeters should all be avoided. The tubing should be sloped at about 3 cm per meter back towards the seal, no high spots in the tubing, NONE. The fill line of the reservoir should be about a meter higher than the center line of the pump, and should be no further than about a meter from the seal. Your seal vendor should be able to provide literature with a nice diagram of an ideal installation.

The rotation of the seal will provide some windage to promote circulation of the buffer fluid, not much we're talking a centemeter or so of head. Pumping rings provide a few cm of head and can be added to the seal, if not already installed. The temperature difference is what really drives the circulation of the buffer fluid. Your seal vendor might call this Thermosyphoning. The heat generated by the seal faces and the cooling effect of the tubing and the mass of the reservoir promote this. This is why minimal friction losses are important to the tubing installation. (generally pipefitters like to make thier work as neat as possible. do not allow them to tie the tubing to and from the seal together, keep the tubing lines apart so that they do not marginallize the temperature difference between them and reduce flow.)

Additionally, the way that the buffer fluid is added to the system can determine if the system will work. You must fill the reservoir and seal so that no air or vapor remains in the seal or tubing.

With an OEM pump package for an ANSI or DIN installation you're likely to have a seal with connections on either side of the shaft. This works pretty well if you're carefull about the way you add the buffer fluid. The connections are usually angled and called "tangential" ports. This is done again to reduce frictional losses in the system... it makes it easier for the buffer fluid to enter and leave the chamber in the seal. These connections do not make it easy to ensure that this chamber is liquid full. There may also be a connection at the top of the seal. Check the drawing provided with the seal to see if this port connects to the chamber between the two seals. If so, crack this port open when filling the seal letting the buffer fluid flow out of the seal along with the bubbles of air it's displacing. When no more bubbles come out of the seal, it is full.

Some seals have connections at the top and bottom of the seal and do not angle the ports. This works well too, the seal "automatically" vents the air or vapor through the tubing. This can also be done with the standard ANSI or DIN seals by installing them 90 degrees off, as long as your bearing housing accomodates the tubing connections.

Be aware that the seal crack could have occured when a bubble of vapor finally budged allow flow to the seal... the cooler fluid suddenly hitting the hot seal faces could have caused thermal shock of a seal face.

I hope this helps.

 

[abcmex]

If you do not have Cooling water you could use an air cooler.
RGS

http://www.flowserve.com/seals/literature/lit_FSD197_Airfin_data.pdf

 

[abcmex]

I2P
sorry I was in hurry. Lets try to get some facts:
1. Your pump is moving condensate (cold around 30 degC)
2. The original installation was with an Plan 11 and Plan 52 (mechanical seal in tandem?, Was really 52 or 51?).

I suppose your pump is installed after the hotwell of a condenser in a power plant. Usually the recommendation is to use a buffer fluid only to avoid the possibility of a leak (from outside to inside).

If you process temperature is that low it is very likely that the only heat produced by the mechanical seal would disipated (cooled down by the condensat flow from the API 11) and not heat the buffer fluid fluid.
If this happen you probably do not need a rotating ring or cooling water and due to the missing temperature difference, no flow may occur (thermosyphon).

Which kind of plan was installed by the second supplier? did it had a circulation ring?

The piping is very important as Longeron describe to guarantee a vented system.
RGS

 

[ChickenSoop]

ensure there is no air in your line - put an air relief valve on it and bleed it fefore startup, during startup and after startup, then bleed it off some more until you are satisfied no air is in the line. Your seal might have a pump ring, but the flow will be mostly be convection. The tank must be located close to the pump, and be about 18 inches above the seal.