Join or Manage Your Profile
Problem with centrifugal compressor axial displacement

This topic can be found at:
http://maintenanceforums.com/eve/forums/a/tpc/f/3751089011/m/10220212173

25 March 2013, 12:27 PM
Hungbka
Problem with centrifugal compressor axial displacement
Dear everyone!

Please help me.

We have a compressor - turbine system. They working well before but 2 month ago. The compressor had failure cause of high axial displacement. The system must shutdown emergency.

Two shaft position probes ( bently nevada xl 3300) and six thrust bearing have been damaged.

We replaced new thust bearings and probes. The compressor had operated in 3 weeks and similar failure happened.

And now the compressor had failure for total of 8 times.

what shoud I do with this compressor now?

Please view the attached file below.


25 March 2013, 01:49 PM
John from PA
Four questions come to mind initially.

1. Have you confirmed proper function/operation of the balance piston.

2. What type coupling (gear, shim pack, diaphragm, etc) and how lubricated?

3. Any damage to radial bearings?

4. Is lube system functioning properly?
25 March 2013, 03:30 PM
Shurafa
Regardless of the root causes, the protection ssytems have to take an action on the correct time; not after the failure.
You did not say if the machine tripped or was manually shut down.
For the damages in the photos to take place, I cannot trust the protection system. Do you have skilled people to work on the instrumentations? If you cannot garantee that, you should hire someone at least to verify the functionality of the system including the set points and output signals.

The radial vibration should have triggered something. So should have done the temperature points. If they did not send at least alarms to the operators, you have a very bad system.

This is only mho and i could be totally wrong.

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


Regards- Ali M. Al-Shurafa
25 March 2013, 04:17 PM
John from PA
quote:
If the did not send at least alarms to the operators, you have a very bad system.


I would not be inclined to blame the system, but the people who set it up or perhaps lack of operator knowledge as to the importance of alarms.
25 March 2013, 08:50 PM
Hungbka
quote:
Originally posted by John from PA:
Four questions come to mind initially.

1. Have you confirmed proper function/operation of the balance piston.

2. What type coupling (gear, shim pack, diaphragm, etc) and how lubricated?

3. Any damage to radial bearings?

4. Is lube system functioning properly?


Dear Mr.John.

Thanks you for your help.

1. My compressor do not use balance piston it use a blance disc. I we can can not comfirm that it function normally. And so how can I check it?

2. The type of coupling is gear box.

3. The radial bearings seem working well.

4. The lube system is good
25 March 2013, 09:19 PM
Hungbka
quote:
Originally posted by Shurafa:
Regardless of the root cause, the protection have to take action on time; not after the failure.
You did not say if the machine tripped or was manually shut down.
For the damages in the photos to take place, I cannot trust the protection system. Do you have skilled people to work on the instrumentations? If you cannot garantee that, you should hire someone at least to verify the functionality of the system including the set points and output signals.

The radial vibration should have triggered something. So should have done the temperature points. If the did not send at least alarms to the operators, you have a very bad system.

This is only mho and i could be totally wrong.


Dear Shurafa.

Thanks you for your attention to my problem.

The protection system working correctly. The first failure it trip the compessor when the axial displacement reached HH set point.

But the shaft still go toward the probes and break it.

After replace a new set of bearing pad. The temperature of thrust bearing raise quickly when compressor speed reach 7000RPM so we must shutdown it manually. We Take the uninstall, check, and reinstall the thrust bearing for many time but the temperation still high ( 125 °C).
25 March 2013, 11:25 PM
Ralph Stewart
Have you the name of the manufacturer, a drawing of the internals of the compressor, specs and dimensions' drawings of recommended settings and a full schematic of how it is supposed to work?

How long since last overhaul? Sounds like this started all of a sudden. Is that the way it happened?

quote:
The type of coupling is gear box.
Has the gearbox coupling to compressor been inspected for correct operation and setup? Has it been replaced lately?


Thanks and Have a Great Day,
Ralph
Senior Analyst and Instructor @
Alert Analytical
http://www.alertanalytical.com
26 March 2013, 01:07 AM
spherical
What about the turbine?
How come the shaft hit the probes; check your setting of axial float. the shaft shall not reach the probe. Unless the thrust bearing were abruptly eaten but that sounds strange.
Check the thrust collar if mounted true perpendicular to the thrust bearing.


regards
26 March 2013, 03:25 AM
Shurafa
quote:
The protection system working correctly. The first failure it trip the compessor when the axial displacement reached HH set point.

But the shaft still go toward the probes and break it.


From the pictures, I guestimate the alarm and trip settings to be around +/-18 and +/-24 mils. If you are using a typical axial probes, the minimum gap between the shaft/collar metal and the tip of the probe is >40 mils after installing new thrust bearing and properly installing the probe (while the compressor starts to take load).

For the shaft to break the probe then it has to cross the alarm setting, trip setting and continue to eat the bearing metal at least 0.030 in and then it just touches the outside tip cover. That is not an easy thing to happen (and to happen suddenly). Again, this is difficult if the protection system is healthy. By the way, which side was damaged (active or inactive)?

Please check your system for the above parameters and advise the actual settings (from both the equipment manuals and form physical field checks)

It would be much easier for the forum members to give you some meaningful suggestions if you can post the trends of the axial readings, active and inactive temperatures and speed.


Regards- Ali M. Al-Shurafa
26 March 2013, 08:03 AM
John from PA
quote:
The type of coupling is gear box


Can you clarify? By use of the term "gearbox" are you referring to a gear type coupling and if so how is it lubricated. Is there also a gearbox between the turbine and compressor?

Gear type couplings develope thrust under misalignment conditions especially if they lock up.

Something I would add to my previous list is what is the possibility you are operating near surge. Not necessarily all the time but for extended periods.

quote:
My compressor do not use balance piston it use a blance disc.


"Balance disk" is just different terminology for a device that is intended to offset the thrust generated as we move the process medium from a low pressure to a high pressure. Typically a balance disk takes a pressure drop across an axial face where the drum takes the drop across radial seals.

The thrust generated is a function of the differential pressure across each impeller. To a lesser degree the different geometries or "shape" of the two sides of the impeller also contribute. Depending on the number of stages in the compressor the net thrust can be high and design provisions must be used to cancel out most if not all of the thrust. Some machines use opposing impellers but that isn't likely for the machine you describe. Most likely , all the impellers (if multi-stage) are mounted in the same direction and after the last stage there is a balance disk or balance drum. The balance disk will have the discharge pressure on one side and the suction pressure on the other side. The routing back to the suction side often accomplished by external piping. By proper sizing of the disk/drum the differential pressure across the device generates a thrust opposing the thrust developed from movement of the process medium. It may not fully oppose the thrust but the net thrust is what is used to size the thrust bearing. The disk/drum has to have some extensive sealing arrangements since a pressure differential exists across the two sides. If those seals are worn then the drum/disk doesn't develop the intended thrust and the net thrust, that imposed on the thrust bearing increases. As I mentioned there is a line that routes the suction side pressure to the balance drum/disk. That line should be checked to insure it is open although it is always a good idea to verify OEM recommendations. Many installations use differential pressure gages to insure proper pressures exist on each side of the balance disk/drum. Installations also are sometimes equipped with a method of flow measurement of the process medium back to the suction side. If the flow becomes excessive that is an indicator that the drum/disk sealing arrangement is malfunctioning. Under such conditions, thrust go up and can cause thrust bearing failure. Proper balance drum/disk operation is critical.

The difference between a balance disk and a balance drum is whether the primary pressure drop is taken across a radial clearance (drum) or an axial face (disk). Some pumps use a combination drum/disk that incorporates both radial and axial clearances.

quote:
The protection system working correctly. The first failure it trip the compessor when the axial displacement reached HH set point.
But the shaft still go toward the probes and break it.


This statement still troubles me. If indeed the turbine was tripped then the machine should have immediately coasted down in speed. Was this the case and if so what would you estimate the coast down time to be? There seems to be indications that considerable damage to the thrust pads occured after the trip. What has happened to the lubrication system after the trip?

While on the subject of the thrust pads, were the replacement pads installed properly for the intended direction of rotation?
27 March 2013, 10:37 AM
William_C._Foiles
Where is your test data or description of your test? With 8 failures, you should have designed a test by now.

Are the replacement bearings oem or identical to keep having failures? Are they installed correctly. Is the balance setup correctly? Are you surging your compressor? Test!

Where are you on the operating envelope of the compressor? Do you go into surge at trip? Lubrication issues checked?

DId this use to operate correctly?

Straight through, back-to-back? Multiple sections, multiple input streams? Change in gas density - pressures?


Regards,
Bill

27 March 2013, 11:52 AM
Hungbka
thanks everybody for pay attention to my problem.

I'm not the man who operated this comprresor.

I'm incharge of measure instrument for this compressor include Vibration probes, Tem- probes...

I have some picture that i take from the operator and I attached it below.

After reinstall new shaft position probes and new thrust bearing pads the compressor still failed cause of hi-temperature althought the axial displacement value is small( the trip value is +0.7mm or - 0.7mm.


27 March 2013, 12:21 PM
Walt Strong
Hungbka,

Respectfully, you have a machine problem and not an instrument problem. As long as the instrumentation is installed and operating correctly, you have done your job. This will in no way correct the machine failure root cause. I doubt if anyone can help you or your associates with solving the machine problem without more information about the machine (design and operating parameters), as already requested.

Walt
28 March 2013, 07:48 AM
John from PA
quote:
Respectfully, you have a machine problem and not an instrument problem. As long as the instrumentation is installed and operating correctly, you have done your job.


Personally, I am of the opinion he has both a machine problem and an instrumentation problem.
28 March 2013, 08:01 AM
Xracer
Here is the .rar attachment changed over to Pdf....




thexracer@live.com
28 March 2013, 10:07 AM
rotary
Seriously need an expert help on this problem...ask your mechanical maintenance incharge to call in a consultant/company involved in such analysis in your area...

No doubt you will have another failure if you continue to replace sensors and bearings
28 March 2013, 01:22 PM
Shurafa
quote:
Originally posted by John from PA:
quote:
Respectfully, you have a machine problem and not an instrument problem. As long as the instrumentation is installed and operating correctly, you have done your job.


Personally, I am of the opinion he has both a machine problem and an instrumentation problem.


I second John.

The protection system (hardware, software, settings etc.) failed several times to protect the machine.
The machine or process has a serious problem that led to several failures, despite the condition of the protection system.

Solving one side of the problem may not prevent the future failures. It looks a very critical machine from the illustrations posted. More "trials" should not be attempted here.

In all cases, I strongly recommend lowering the set points for both axials and temperatures if the machine will be restarted at least to trip it before a damage will take place. I would do that even if I have a qualified staff to fix the machine; unless I enjoy destroying compressors.


An axial shutdown set point of 0.7 mm (~ 27 mils) is a bit high, anyway to me. The readings on the trends shows max +0.48 mm. Or maybe I'm not able to read the plots correctly.

Also, I see a clear deviation between the axial probes , reaching >30%. (points 1702 and 1703)
Also, the readings of active thrust temperatures are quite a way from each others, by about 40% in one of the plots. (points 1722 and 1723)
By the way, repairing and operating the machine by the same staff may not easily avoid the failure, if they will go through the same procedure they used. If you are a key player in the decisions there, consider a fresh look from an outsider who is qualified and unbiased to help. It is worth it.


Regards- Ali M. Al-Shurafa
28 March 2013, 02:02 PM
John from PA
Two major failures in 3 months = time to call on OEM
28 March 2013, 02:11 PM
Walt Strong
quote:
Two major failures in 3 months = time to call on OEM

I guess that would be to fix the instruments so the compressor would not fail again!

Walt
28 March 2013, 02:25 PM
William_C._Foiles
Your data shows both active and inactive sides have high temperatures. This can happen with a dynamic axial displacement, but if set properly, I don't see how a static axial force can do this.

'If set properly' - you can always set it up too tight. You can have lubrication issues, wrong oil/lubricant/ or dirt.

When you decide to do a proper test (or your consultant) test for both axial vibration, also.

If I see the plots correctly, one had both active and inactive high

JohnPa,

Looks like a back-to-back with a center seal.

This message has been edited. Last edited by: William_C._Foiles,


Regards,
Bill