Dear Colleagues,

After completion of GE frm6's Major overhaul we experienced unexplainable vibrations during startup which, reach their maximum at FSNL:

-Both seismic sensors of turbine brg indicated: 13.4 mm/s @ FSNL

-Both seismic sensors of compressor brg indicated: b 7 mm/s @ FSNL

By measuring the vibration using 2 other different external instruments, the values became 3.5 mm/s in turbine brg side and almost similar values in compressor brg which were considered acceptable.

So we proceed as follow to trouble shoot the problem:

-Replaced the 4 seismic sensors by other good ones

-Secured firmly the 4 probes

-Installed a new temporary shielded cable directly from field to control panel(MKV)

-Checked the configuration of TCQA in IO Config.

-Checked the revision of this card and compared with the IO configuration

-Replaced the TBQB card in MKV panel

-Checked all signals of TCQA in DIAG

-Checked for grounding in MKV

All these checks which were found normal led us to a slight improvement, the values changed form 13.4 to 9 mm/s in turbine bearings and 6mm/s in compressor bearing.

I share with all of you this weird problem in order to achieve to a convincing solution to this problem.

Thanks for all.

 

Ridha,

Were all these things done in a "shotgun" approach--meaning, where they all done at the same time, or was each one done and then any change recorded before doing the next step?

> So we proceed as follow to trouble shoot the problem:

> -Replaced the 4 seismic sensors by other good ones

Good idea, but why, if both sensors on each bearing indicated the same vibration?

> -Secured firmly the 4 probes

Could have been done to the original probes, which were likely removed before the MI and replaced after the MI.

> -Installed a new temporary shielded cable directly from field to control panel(MKV)

Are you still using the temporary cable? If not, did the signals change when the temporary cable was installed and then removed? If so, how did they change?

> -Checked the configuration of TCQA in IO Config.

Why? Was one of the TCQA cards replaced during the outage?

> -Checked the revision of this card and compared with the IO configuration

Why? Was one of the TCQA cards replaced during the outage?

>-Replaced the TBQB card in MKV panel

What were the results of changing the TBQB?

>-Checked all signals of TCQA in DIAG

What were the results of this check?

>-Checked for grounding in MKV

Low-level signals aren't affected by 125 VDC grounds.

> All these checks which were found normal led us to a slight improvement,
> the values changed form 13.4 to 9 mm/s in turbine bearings and 6mm/s in
> compressor bearing.

Still sounds like the vibration is elevated after the MI, and you didn't tell us what the vibration levels on each bearing were before the MI.

If you did each step above, one by one, which one resulted in the biggest improvement in the readings? It's extremely difficult to say what the problem is or might have been when several actions were taken at once. But, it still sounds like the vibrations are elevated--which means it's not really a controls problem. (Sorry, Mechanical Department!)

Presuming there were no configuration changes to the vibration sensors in the Mark V during the MI and that the interconnecting cabling was the same as before the MI, and the sensors were the same ones removed before the MI, and that they were properly re-installed after the MI, there's no reason to believe the elevated vibration levels are the result of a problem with the control system.

It's most likely a problem with the balance of the rotor, since most likely some rotating components (axial compressor blades or turbine buckets) were replaced and that would affect the rotor balance to some degree. Also, some sites will perform some hand-cleaning of axial compressor blading during a MI, and sometimes it's not done with the best attention to detail.

We don't know all of what was done to the turbine and axial compressor rotor, but you certainly seem to have eliminated the Mark V and the interconnecting cabling as causes of the elevated vibration readings.

 

Dear Sir,

Thank you for your reply
Please take note of these points:

-The tests were not done in the same time.

-The replacement of TBQB did not lead to different results
We always reads high values 13.4 mm /s.

-The readings in TCQA DIAG were all normal.

-The problem is that the vibrations were also measured using an external vibration tester along with its sensors and cables.
The results were normal:

3.5 mm/s in turbine and 2,8 mm/s in compressor brg

That is why I did the different checks in Control (TCQA config, revision…) although the card was not replaced during the outage,to eliminate any doubt from the control side.

-Now I normalized the situation by putting back the original sensors and cables, then back to the high values (13,4 mm/s ....).

Thank you for the help and I look forward for any other suggestion.

 

Ridha,

> -The problem is that the vibrations were also measured using an external
> vibration tester along with its sensors and cables.
>The results were normal:

>3.5 mm/s in turbine and 2,8 mm/s in compressor brg

How were the external measurements performed?

I'm not clear about how the vibration measurements changed when you put the new sensors in place of the old ones. Sensors do go bad--but it's very difficult to believe that two sensors on the same bearing will both go "equally" bad at the same time.

It's very common for the Speedtronic to be blamed for high vibration readings after a maintenance outage. Very common. But, after a lot of time is spent proving that the problem is not in or being caused by the Speedtronic it's usually found that alignment issues or balance issues were the real cause.

If you had more believable readings with the new sensors, that would seem to suggest the original sensors may have been or are degrading. This presumes the new sensors have the same sensitivity (volts/inch-per-second, or, converted, volts/mm-per-second). It's pretty common, also, to find the sensors in the spares warehouse do not have the same sensitivity as those originally mounted on the machine.

A Mark V will "fan out" the seismic vibration inputs terminated on the TBQB to all three TCQA cards in each of <R>, <S> and <T>. And there would be Diagnostic Alarms if the values in any of the processors were different from the others.

Again, the fact that you have said the readings of both sensors on each bearing were similar indicates that both sensors on each bearing are sensing the same movement. And if both new sensors on each bearing also indicated the same vibration (even if was a slightly lower vibration) then it would certainly seem there is increased vibration (regardless of what the Mechanical Department says or believes).

That's all I can suggest based on the information provided. Please write back to tell us how the troubleshooting progresses, and, ultimately, how the issue is resolved.

 

Dear CSA,

Thank you for the attention you give to this subject.

From my side I Strongly believes that MKV has nothing to do with this problem for the simple reason that my work has been limited in the control system to the routine tests and general cleaning. On the other side the vibration during the preoutage and before starting the MI have not exceeded 2.8 mm/s in all bearings. What follows after this, is the MI, that is to say the mechanical work in the field side and not on the MKV side, so why do we blame MKV.

Anyway our company decided to use a third company specialized in vibration measurements and analysis, to diagnose this problem.
I will keep you informed about the evolution of the situation.
Thank you

 

Hmmmm..before the outage/shutdown, was this a controlled shut down?

How much run time has been added since this maintenance outage? FSNL...part load etc..

CSA, if nothing changed during maintenance, and this vibration is stress induced. Run out possible?

 

Dear CSA,

As you previously predict the cause of the vibration was a misalignment.

Lot of wasted time to finally reach this conclusion well known. People always tend to reject responsibility to the others. At present, the mechanical staff is trying to fix the problem.

Thank you for your support

 

Ridha,

Thanks for the feedback! I'm currently on a re-start of a GE-design heavy duty gas turbine that is also experiencing some problems that can only be attributed to mechanical causes and there is pressure--by the mechanical department--to blame the problems on the control system.

It's, unfortunately, a very common waste of time and money that always results in lost time and production. It's never ceased to amaze me how management is so quick to want to blame the control system--at the insistence of the Mechanical Department--and spend so much money only to come to the inevitable conclusion that the problem is mechanical.

Worse, this has happened repeatedly at the same sites over a period of many years. The control system is the Evil Empire of many plants around the world, the Axis of Evil, on which all problems--real and perceived--are blamed. I promise you this: this is, unfortunately, not the first time you will experience this in your career. I've often wished I'd had been given 10% of the money I've seen wasted on these very foolish exercises in frustration. I could retire early and handsomely on that money.

And the worst part of the whole thing is that you are never given the acknowledgment for having been proven correct. In fact, sometimes you are reprimanded for not having been more forceful in your early protestations! And, I have been at sites where the same people do the same thing again and again because that "thing with all the wires and LEDs and that infuriating alarm horn" simply must be the root of all evil and problems after any mechanical maintenance outage.

Welcome to the world of turbine controls. It's not all it's cracked up to be.