PDA

View Full Version : Got called out to replace a shaft seal







Tycho
11-05-2015, 07:06 PM
Got called out to replace a leaking shaft seal on a Howden XRV 204, got told the machinists had been struggling with intermittent leaks on it, where during operation it would suddenly spew out a cloud of gas for a few seconds and then work fine for a day or two making them think it was OK, then as soon as they turned their back to it it would poof out a cloud of NH3 again...

Gathered my tools and a replacement shaft seal and sped of towards the ship in distress, with the leaking smelly shaft seal...

Ran on board to close the valves on the compressor unit, only to get into the machinery room only to see that the machinists had already taken care of the problem....

13757



I grabbed my tools, got back in my car and angrily called my boss to tell him to stop sending me out to fake shaft seal leaks

RANGER1
11-05-2015, 10:12 PM
Tyco, did you receive my email?

This would be pretty standard kit on our jobs :D

Machine in your picture looks like it just came out of the box.

RANGER1
11-05-2015, 10:37 PM
Seriously though it could be carbon face drive lugs worn, causing seal to get "hung up".
This can allow faces to open up, sometimes occasionally, sometimes all the time.

In one case pressure testing an overhauled compressor, seal faces would suddenly open up & close or flutter.

Was pressurizing into suction of compressor with dry nitrogen, believe balance piston behind shaft seal with it's large surface area, was pushing from from behind trying to equalize the pressure into seal cavity.
Balance piston was moving on shaft back & forth momentarily opening up seal faces.

Tycho
14-05-2015, 04:34 PM
Tyco, did you receive my email?

This would be pretty standard kit on our jobs :D

Machine in your picture looks like it just came out of the box.

Got your mail :)

did you get my answer?

The compressor is on an MMC SX unit (RSW) we installed on a ship two months ago

Tycho
14-05-2015, 06:00 PM
Seriously though it could be carbon face drive lugs worn, causing seal to get "hung up".
This can allow faces to open up, sometimes occasionally, sometimes all the time.

In one case pressure testing an overhauled compressor, seal faces would suddenly open up & close or flutter.

Was pressurizing into suction of compressor with dry nitrogen, believe balance piston behind shaft seal with it's large surface area, was pushing from from behind trying to equalize the pressure into seal cavity.
Balance piston was moving on shaft back & forth momentarily opening up seal faces.

When I was wiping of the seal to inspect the surfaces I noticed that the carbon ring in the rotating half was quite loose, or at least loose enough for my brain to say "This isn't normal" :) when I measured the total play it was around 5mm. (hold the steel housing in one hand and the carbon ring in the other and try to "unscrew" it)

On the new seal there was 2 mm play.

Another thing I noticed was pitting on the face of the carbon seal, usually we see this on systems where there is a problem with the oil heating, or with a lot of condensation in the oil separator, it looks like white spots on the seal face. According howden this happens when the liquid droplets in the oil turn to gas and collapse at super sonic speed, creating a mini explosion between the seal faces... Both these systems are in a closed of room and the stand still oil heating keeps the oil at 45C, so with the pre-lube pump I find it hard to believe that there was any refrigerant in the oil.


I found it a bit strange that there should be wear on the seal that might indicate some kind of torque, the motor is on a VFD and has a ramp up time of 10 seconds, plus the units have a pre-lube pump.
The pre-lube pump is also set to start once every 24 hours to prime the compressor during standstill.

The prime function is something we have implemented after way too many shaft seal failures on XRV compressors, especially on those with VFD, which we attributed to the longer startup time making it take longer for the differential oil pressure to build up.
Another factor was the running hours the compressors got on a purse seiner during a year, which could be between 3-4000 hours, so the prime function ensured that the shaft seal and bearings gets lubricated every day even when the ship is at port for 3 months.


I have also experienced what you mention during pressure testing in the work shop (and a few times out on systems), I figured the same thing as you, that the balance piston was pushing on the shaft seal, but I wasn't sure until now when we are two with the same opinion :)

The second time it happened I learned to take the pressure up slowly instead of doing the "yeah yeah, git'r done" :)

The few times it has happened on new systems, it's been when the compressor units were delivered and installed almost a year before the pressure test... in those instances we usually just replace the shaft seals.

RANGER1
14-05-2015, 09:50 PM
Tyco,
Thanks, the email I send had all the contact details, have not seen anything since then, will check tonight.
We are not a big fan of Howden seals, even with their latest design.
To many "0" rings to fail for a start.
A seal company we work with developed a single welded metal bellows with cast iron stationary seat.
It is secured on shaft with a method that does not damage shaft.
Also good for long idle periods, as cast iron holds oil.

I I think your neck of the woods has alternatives as well.

Tycho
29-05-2015, 04:16 PM
That sounds very interesting Ranger1.

I have not heard of anything like that. the seals howden delivers now have a carbon rotary ring and carbon stationary ring, the stationary ring is a lighter color tho, so maybe a different composite.

I know the howden seals used to have a cast iron stationary ring a few years back, but we had even more problems with that type than the current one.

The issue, I think, is not the seals, but the amount of time they are actually rotating. We service 80% of the purse seiner fleet in Norway, and the ships have evolved to a cargo capacity that enables them to fill their quotas in a month or two and then spend the next 3-5 months in port before fishing is opened on the next kind of fish.

on smaller vessels, with XRV 127 and 163 compressors, that are fishing most of the year we don't experience any shaft seal failures.

Most of the failures are on XRV 204 compressors of varying rotor lengths on ships that spend 7 months of the year in port.

the instruction manual says that the shaft should be rotated by hand every few days, and once every two weeks (i think) the system should be started and the compressor should be run for an hour until it has operating temperatures.
This is to rotate the bearings on the compressor and motor, to lube the shaft seal and to lube the compressor o-rings and gaskets.

Very few machinists do this while the ship is in port, and they don't start the system until the day before they are heading out, and then when something fails when they cold start it, without giving the shafts a few turns before before pushing start they complain that the compressor/system/design is crap.

the pre-lube pump used to be an option, but nobody wanted it because it cost a few bucks more, now it is no longer an option and is part of all the systems we deliver... it makes it a bit more costly, but that is the price we pay for lazy machinists :)

In our latest PLC upgrade we are also adding a "run log" it logs the 10 last starts of the compressor, showing
-start date
-start time
-total run time for each start


*gripe gripe gripe* :D

RANGER1
29-05-2015, 10:38 PM
That sounds very interesting Ranger1.

I have not heard of anything like that. the seals howden delivers now have a carbon rotary ring and carbon stationary ring, the stationary ring is a lighter color tho, so maybe a different composite.

I know the howden seals used to have a cast iron stationary ring a few years back, but we had even more problems with that type than the current one.

The issue, I think, is not the seals, but the amount of time they are actually rotating. We service 80% of the purse seiner fleet in Norway, and the ships have evolved to a cargo capacity that enables them to fill their quotas in a month or two and then spend the next 3-5 months in port before fishing is opened on the next kind of fish.

on smaller vessels, with XRV 127 and 163 compressors, that are fishing most of the year we don't experience any shaft seal failures.

Most of the failures are on XRV 204 compressors of varying rotor lengths on ships that spend 7 months of the year in port.

the instruction manual says that the shaft should be rotated by hand every few days, and once every two weeks (i think) the system should be started and the compressor should be run for an hour until it has operating temperatures.
This is to rotate the bearings on the compressor and motor, to lube the shaft seal and to lube the compressor o-rings and gaskets.

Very few machinists do this while the ship is in port, and they don't start the system until the day before they are heading out, and then when something fails when they cold start it, without giving the shafts a few turns before before pushing start they complain that the compressor/system/design is crap.

the pre-lube pump used to be an option, but nobody wanted it because it cost a few bucks more, now it is no longer an option and is part of all the systems we deliver... it makes it a bit more costly, but that is the price we pay for lazy machinists :)

In our latest PLC upgrade we are also adding a "run log" it logs the 10 last starts of the compressor, showing
-start date
-start time
-total run time for each start


*gripe gripe gripe* :D

Our seal was developed for the same reason, long idle periods.
The cast iron face was used as it is porous & can hold oil.
Most other machines have a oil seal in back of shaft seal chamber with a oil vent hole at top for circulation, this mean oil always around seal no matter what so it does not dry out.
Howden also changed "0" ring supplier a number of years ago, as they were failing prematurely.

Sometimes improvements are are made by manufacturers, but you revert to what you know is proven
in the field, as it costs you, not supplier.