Have anybody encountered a failed motor mechanical seal (John Crane Type 1A) in a vertical pumpset application due to the following reasons?

The seal is running dry because the stuffing box was not vented in a vertical application.

The stuffing box is running in a vacuum because the supply tank is not vented properly.

A stuffing box vacuum can blow open unbalanced seals.

This message has been edited. Last edited by: Josh,

Josh,
What is a Motor mech seal? I am not familiar with this term.

Mechanical Seals definitions:
http://en.wikipedia.org/wiki/Seal_%28mechanical%29

Vee,

The mech seal (John Crane type 1A) is installed at the close couple (muff type) which connects a vertical submersible motor (wet winding type) to a pump on top of the motor. So I call it the motor mech seal because its function is to prevent the potable water leakage from the motor.

The potable water supplied from an overhead tank located on the platform deck provides cooling to the motor windings and lubrication for the thrust and journal bearings. It enters the motor at the bottom and returns to its tank at the top.

This message has been edited. Last edited by: Josh,

Josh,
Every mechanical seal suffers badly if it runs dry, because the hydrodynamic opening force is absent. Horizontal shaft seals can pick up some fluid from the stuffing box, even when it is vented poorly, so they are better off than seals on vertical shafts.
On the latter, seal faces can run completely dry unless the stuffing box is flooded. This makes the venting of stuffing boxes a vital activity. The vent hole in the stuffing box body muzt be above the plane of the seal faces, of sufficient size and we have to ensure it is working.
There can be a lot of heat at the seal faces, partly from the motor, and partly by seal viscous friction. Steam can be formed quite easily, depending on the pressure in the stuffing box. The lower this pressure, the greater the generation of steam. If this is not vented continuously, the seal face will be running in steam instead of water. One measure of this is the rate of rise in water temperature in the tank, assuming some circulation is established.
A simple calculation will show that:
- if seal face dia. is say,100 mm, in one rev. the contact faces travel 300+ mm
- if shaft speed is 1500 rpm, the faces rub along for 500 m every minute
- if seal runs dry for, say 5 minutes, the seal faces will have rubbed over a 2.5 km distance!

THAT is why the seal fails, not because it is John Crane or Borg Warner, Burgmann or Durametallic. In conclusion, it is VITAL to ensure that seals never run dry. It is easy for dry running to happen in vertical shaft seals. Venting efficiency is important, but dont forget to check vent hole location wrt seal face plane.
I have experienced vertical ethane pump seal failures for the same reason, and once a way was found to keep the seals wetted, the problems just went away.

This message has been edited. Last edited by: Vee,

Vee, thanks for good explanation. I have to study it but one thing still puzzles me:
What mechanism causes the water to return to its tank against gravity (30m up) because I don't see any pump to do so. Is it by convection due to heating by the motor windings?

Do you still have the the write up for your ethane pump? I've searched internet for similar problems but couldn't find one.

I agree it's not because of the brand which I mentioned so that you can understand the term quickly.

This message has been edited. Last edited by: Josh,

The motor is submrsible. I was looking for the vent line but couldn't find it. Reading the pump installation, it only mentions to fill the motor completely by circulating the potable water and pressure test hold for 30 minutes to ensure tightness inclucing the mech seal. Did I miss something?

Josh,
What is obvious to you is not always clear to people '000s of miles away; perhaps you can give us some drawings to show relevant details.

As attached

Vertical_submersible_motor.bmp (945 Kb, 47 downloads)

Josh,
Thank you for the drawing. It appears from this that
- the outlet tapping is well above the plane of the mech seal; that eliminates one source of trouble.
- you said earlier that you use potable water as the coolant; I have not seen potable water used for this service, usually it is distilled water with very low conductivity. You may want to check the manual if potable water is OK.

The water moves up by convection; the column of warm water 30' high is lighter than the column of cold water. heat imnputs are from the motor and bearings.

You mentioned a vacuum condition earlier; please clarify what you mean. As far as I can see, the seal can never see a pressure less than about 1 barg., but I may be mistaken, so please check what the operating conditions at the seal are supposed to be. Not venting the stuffing box makes the seal run in air at 1 barg, because there is a column of cold water feeding the motor at the bottom. Running the seal in air is just as bad, but I am not able to understand the vacuum bit.

Possible problem areas to check are whether any of the water passages are blocked; you can find out if this is a real possibility by checking whether there is a steady flow in and out of the tank. Also the return pipe temperature will be quite high.

Do you have photos of bthe seal faces? Are there radial crazing cracks on the hard faces? Or blue marks?

As always remote RCA is always difficult.

Reply as follows:

Use of potable water is specified in the OEM manual. Alternative is gycol.

Water returns to its tank by convection - Nice to confirm this is possible.

Vacuum - Well, I try to understand why the seal failed. I read in the McNally Institute the following statements:

"The seal is running dry because the stuffing box was not vented in a vertical application.

The stuffing box is running in a vacuum because the supply tank is not vented properly."

In our case, the potable water is supplied from an overhead tank so I'm looking into whether not venting the seal or the tank is causing the problem. Do we really need to vent the water tank which needs to be topped up at 70% min level? In the manual, it is said to completely fill the motor & seal prior to start up to purge out all air.

Blockage of water passages - Actually the problem faced by the operator was that the water level cannot be maintained at 70%min level ie keep losing water because no return, suspected excessive leaking at the seal because the piping pressure tested to be tight after removal of the pumpset from its caisson.

The hard face is ceramic and soft face is carbon. In the failure, the ceramic was broken which I have posted in the thread below. Are blue marks supposed to be visible on ceramic?

http://maintenanceforums.com/eve/forums/a/tpc/f/119103451/m/9001053162

Thanks for your educating comments.

This message has been edited. Last edited by: Josh,

Josh,
So far so good. But we still dont have any idea of why the seal face shattered.

quote:

The hard face is ceramic and soft face is carbon. In the failure, the ceramic was broken which I have posted in the thread below. Are blue marks supposed to be visible on ceramic?

I did not know the face was ceramic (once again what is obvious to somebody at site is not to guys like me who are guessing). And there will not be any blue marks on ceramic, only on stellite or otehr metallic face. But if the ceramic face is held in a metal ring, see if there are any marks there.

Is there a spare pump? How is it doing? Can you measure return water temperature?

I would pursue the following hypothesis and see if there is any supporting evidence.
- seal running in air instead of water
- ceramic gets very hot
- it shatters due to uneven heat distribution

Spare pump failed after 3 months of the first pump failure but this time the ceramic face was not broken. Instead, the thrust pad bearing worn out & excessive end play (4.5 mm vs design axial clearance of 3mm). Another phenomenon to be explained.

Measure return water temp - will look into this.

I'm wondering how can the air get inside the seal, together with water? Do you mean seal running in air instead of water is the same as seal running dry?

Josh,
Maybe the ceramic face shattered for the reason that the thrust bearing was overloaded in the first pump as well! Without evidence it is hard to say, but surely a viable hypothesis?

If the water temp rises a lot in the motor, some steam can form, and get trapped in the top end of the stuffing box. Water starts boiling at about 60 deg C at atmos pressure. Any dissolved air, and there is always that in potable water will also stay there. Without good continuous venting of steam/air, there is always a potential for seal overheating. But this is only ONE possibility.

If the return flow is somehow restricted, and you have look for evidence, this air/steam pocket is in effect the wrong medium for the seal (I call it dry running, but you know what I mean).

I would look for thrust bearing overload signs first.

Well, wear of thrust bearing is very evident in the failure of the standby pump B but not in the first pump failure.

I understand thrust bearing overload is maximum during start up. Are there any other instances where thrust bearing overload would be significant? Will alternate operating policy of the pump A & B such as swapping weekly contribute to excessive thrust bearing wear?

Josh,

Thrust bearing loading depends on the pressure balance within the pump. Are pump impellors balance-opposed or is there a balancing piston and line etc.?
I think we are clutching at straws with no hard data, and such remote guesswork is unlikely to help you. Suggest you get a structured RCA going. You will understand if I do not respond further, as I dont like to keep guessing what might have been happening.