Today is...
Monday, July 24, 2017
Welcome to Control.com, the global online
community of automation professionals.
Featured Video...
Featured Video
A demonstration of EtherCAT control of linear motors using the CTC EtherCAT master.
Our Advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive
Turbine Low Speed Shaft (LP) Could Not Cut In
GE frame-3 turbine trip on startup due to LP SHAFT LOCKED trip alarm

We have Two shaft frame3 Gas Turbine installed at our plant commissioned in 1985 whose control system was upgraded in 2014 to MARKVIe by GE. Turbine is loaded with Seimens two Compressors through a gear box. Recently GT overhauled inhouse and successfully maintenance carried away. After maintenance completion a week before, unit attempted to start, however unit tripped on LP SHAFT LOCKED alarm at 55% HP SPEED.

Everything startup sequence seems to normal, e.g., Fire at 24%, Accelerating normally, fuel gas pressure, fuel flow, PCD, Servo valves, exhaust temp are normal.

This is happened first time, we rechecked LP TURBINE speed pickups found normal. Is there any other cause plz advise?

Plz. Note that a year before we have upgraded compressors and gearbox too. After that machine started successfully many times.

1 out of 2 members thought this post was helpful...

Haroon,

What were ALL the Process- and Diagnostic Alarms which were present during the start and when the LP shaft did not break away?

Did anyone actually observe if the LP shaft rotated during the start attempt?

Have you checked to see if there is something that is causing the LP shaft to be unable to rotate? The speed pick-ups might be good, but if the load on the LP shaft is too great, or the seals/bearings are too tight, there may not be enough torque from the HP to break the shaft away from zero speed.

Also, you didn't say if the unit has variable second-stage nozzles, and if so, what the angle was during the start? That is very important, and if the nozzles were not in the proper position they wouldn't be transferring as much power to the LP shaft as they should be.

0 out of 1 members thought this post was helpful...

Thanks for your response.

First, there is not any process or diagnostic alarm prior to LP shaft Locked alarm and unit trips.

2nd, how can we observe whether LP shaft actually rotating during start?, because we are just relying on LP speed pickups, which is showing zero speed till 55% of HP speed.

Yes, we have also attempted to decouple LP turbine from load i.e. from Gearbox, LP and HP compressors and rotate manually (by hands) the load, which rotate smoothly without any significant effort.

As for as variable second stage-nozzle is concern; during startup, variable nozzle angle was found +15 degrees same as unit startup before. We also manually checked nozzle operation which is quite good.
An important point may be, first time we changed variable Nozzle hydraulic Cylinder and LVDTs (02 nos) with new one during this recent GT overhaul. We properly checked nozzle cylinder position before replacing and install new one carefully with as found position.

Can you suggest us what actual and proper nozzle position should be during startup, so that we may catch the problem.

2 out of 2 members thought this post was helpful...

Haroon,

You stated that maintenance had been performed on the gas turbine immediately prior to this problem. Was anything done to lock the LP shaft during this maintenance? (Doing so might permit fired testing of the gas turbine while maintenance on the load compressor was still being performed.) If so, has anyone verified that the "lock" has been removed?

0 out of 1 members thought this post was helpful...

It's hard for me to believe the below was over looked on a startup after a maintenance of this sort....but stranger things has happened!

it's possible the maintenance done has stiffened the LP shaft, {new bearings, seals, etc..} and the speed dictation time elapsed, therefore triggering alarm/shutdown.

>Was anything done to lock the LP shaft during this maintenance? Doing >so, might permit fired testing of the gas turbine while maintenance on
>the load compressor was still being performed.) If so, has anyone
>verified that the "lock" has been removed?

0 out of 1 members thought this post was helpful...

Haroon,

Double check LVDT's [TMR] voted.

0 out of 1 members thought this post was helpful...

Thanks CSA;

As I stated earlier, this problem appeared just after GT maintenance, and before maintenance there was not such problem. One thing, I would like to mention that during machine startup, LP shaft normally cut-in (starts to rotate) between 45% to 55% of HP speed depending upon machine behavior. The problem seems that LP is "still" after HP speed reaches to 55% and hence safety triggers to trip the machine.
Could you suggest us to increase this setpoint to 2 ~ 3 % and would not be there any problem to machine specially to LP turbine to raise this setpoint.

Would you like to tell us more about nozzle position with respect to startup so that we may progress ahead.

Dear CSA,

We could not get out of the problem, your expert opinion is still awaited please.

0 out of 1 members thought this post was helpful...

Haroon,

It's impossible for me to understand what's occurring without actionable data (trends recorded during the failed LP breakaway).

As such, I'm extremely hesitant to recommend changing any parameter.

If the speed pick-ups checked out okay (and we don't know how they were checked nor what the results were--a simple continuity check doesn't mean they're properly gapped or working correctly), and the Speedtronic doesn't report any LP speed it's reasonable to presume someone has been watching the shaft to see if it spins or not during a start attempt. I believe it was reported that the compressor shaft was rotated and it was okay, but did anyone try to rotate the LP shaft (with L.O. running) with the unit at rest?

Also, it's not clear how the second stage nozzle LVDTs were set/calibrated. The mechanical department should have set the stops on the nozzle actuator ring to the proper values per the drawing and the Control Specification setting. Then, the LVDTs should have had their zero-stroke voltage set/adjusted and a calibration performed. There is usually an indicator/pointer on one of the nozzle rods which gives a visual indication of nozzle position, and this should have been checked/adjusted during re-assembly of the nozzle/casing. Then, this pointer is used when calibrating/verifying calibration of the LVDTs.

The minimum- and the maximum second stage nozzle angle should be listed in the Control Specification. Some newer Mark VIe's are NOT being provided with Control Specifications (which is completely unacceptable), but unless the unit were mechanically upgraded at the time the Mark VIe was installed/commissioned the values from the original Mark* Control Specification are still valid. AND, if the unit was mechanically upgraded then the supplier of the upgrade should have provided any new settings/parameters in some kind of document that updates the Control Specification.

If I recall correctly, the maximum and minimum operating angles for a Frame 3 two-shaft machine were typically +15 DGA and =5 DGA. The minimum stops for the mechanical actuator/ring were usually set a degree or two beyond each of those settings.

Before I could recommend changing any setting, it would be only after a review of trend data AND some confirmation that the LP shaft is free to rotate. A LOT of times, when new journal bearings are installed during a maintenance outage they can be at the minimum end of the clearance range, and that can also cause difficulty with initial operation (including vibration and higher than normal bearing metal temperature during initial operation, sometimes a period of weeks). Also, I have seen new bearings installed that weren't properly aligned, and seals that were too tight. And, for compressor-drive units sometimes the compressor is loaded during breakaway and acceleration. I have been to a couple of sites where even though the Operations Supervisor was aware of an initial change in standard operating procedures after the maintenance outage (a change in driven compressor valve line-up resulting in a need for more torque during starting), it was CERTAIN the problem was being caused by the Speedtronic turbine control (which it wasn't).

The alarm is telling you: The LP shaft isn't rotating. It either is or it isn't, and sometimes one has to visually verify the condition and not rely solely on instrumentation. If the shaft isn't breaking away, and you know the unit (turbine, gear box, and driven compressor) were worked on (disassembled; reassembled) during the outage then it's entirely possible there is something mechanically preventing the shaft from turning (maybe a wrench or bolt left inside the casing--stranger things have happened!). It's entirely reasonable to use a strap or some other mechanical means to try to rotate the LP shaft while L.O. is running with the unit at rest to see if it's free to rotate or not. And, it should be possible to tell if force being required to rotate the shaft is within an acceptable range--especially if the LP shaft is uncoupled from the load. (I know--the mechanical department is POSITIVE there is nothing wrong with the machine, and the problem is that darn Mark VIe--but it's most likely NOT the Mark VIe, but some mechanical issue. Especially if there aren't any other Process- or Diagnostic Alarms as has been reported.)

If the same settings worked before the outage, then it's safe to assume they should also work after the outage--based on the information provided. Again--it might just be that internal bearing or seal clearances are very tight, but they shouldn't be so tight the shaft can't be turned. And, we don't know if anyone has tried to turn the LP shaft.

To another poster's point, it's common practice to de-couple the LP shaft and put a "block" on the LP coupling to keep the shaft from turning and perform an HP overspeed trip test after a maintenance outage. And, then, the block is removed and the LP left free to spin and an LP shaft overspeed trip test performed (sometimes the load can't be spun up to the overspeed trip setpoint for some reason, so the LP is tested uncoupled). Once the two overspeed trip tests are completed satisfactorily the LP shaft is coupled to the load and initial operation of the entire unit (turbine and driven device) are begun. Was any of this done after your maintenance outage before this problem began?

I look forward to your reply.

1 out of 1 members thought this post was helpful...

Hi,

You May have to check the end float with charge coupling installed. If it's very tight you have to implement a spacer between LP rotor and coupling accordingly set value.

We had fixed same problem while upgrade of our turbine MS3002.

Good luck.

Dear CSA,

Further to above discussion I would like to share step by step status and action performed to resolve the issue.

As you are aware that, LPshaft Locked trip is setpoint in Mark-VIe logic which cause HP turbine & machine to Trip if LP turbine does not rotate at specific HP speed. Actual set point was 50% of HP has already increased to 59% of HP in logic. That is HP will continue to rotate till 59% NHP , so that Low speed shaft could rotate.

We have checked the speed pickups, SRV & GCV calibration, Nozzle calibration. The details of step by step action is appended below:

1- LP turbine was decoupled from gearbox and manually rotated, it rotated freely and no abnormality was observed.

2- LP & HP compressors were also rotated manually with gearbox, hardness was observed in rotation.

3- Unit was started with LP turbine decoupled from BHS gearbox. LP turbine cut-in at 25% NHP just after flame approved.

4- Calibration of variable nozzle was checked and found ok, Variable nozzle position was adjusted at 12.5 degrees from 14.5 degree, just to create additional jet . Unit was started, but no improvement observed.

5- Boroscopic inspection of LP/HP buckets, diaphragm casing, variable nozzle and first stage nozzle was carried out and found O.K.

6- Variable nozzle position was readjusted at 15.5 degrees from 12.5 degree ( back to normal). Unit was started, but no improvement observed.

7- Coupling of LP compressor and HP compressor was removed. Unit was started (LP turbine only coupled with LP compressor through gearbox), no improvement was found.

8- Coupling of LP compressor to BHS gearbox was removed. Unit was started with LP turbine coupled with only gearbox, the LP cut-in at 26% NHP.

9- Checked the LP compressor bearings, seals and replaced floating ring seals. Mechanical torque is reduced from 110 Newton meter to 40 Newton meter. Unit was started again but LP shaft unable to rotate again.

10- No activity performed at LP or HP rotor of turbine during fault rectification.

11- Turbine parameters Fuel, exhaust temp, CPD, Seismic Vibration are same at 50% , 55% and 59% NHP as noted during above tests, and are similar to what recorded earlier before Major overhaul.

Looking forward for your expert opinion.

0 out of 1 members thought this post was helpful...

In my experience, it doesn't take much to break the LP shaft away from zero speed and accelerate it--as long as the torque/load on the LP shaft isn't excessive. It would seem you have proved the same with your testing.

I maintain--if the machine has been working for decades with similar parameters (prior to the Mark VIe upgrade, and after the Mark VIe upgrade) and now, after the maintenance outage it's not operating similarly then it's NOT the control system or the control system settings. Continuing to increase the locked rotor trip speed with no acceptable results is causing undue thermal stress on the LP turbine section (nozzles and buckets). At low speeds the air flow (mass flow) through the machine is not very high and internal temperatures are high--this is the reason for a locked rotor trip. If the LP shaft isn't breaking away from zero speed it's not good for the machine to keep firing it, and so the fuel is shut off.

During normal operation (at rated speeds, above 14HS/14LS), the second stage nozzles are used to control the HP speed and fuel is used to control the LP speed. But, during low speed operation such as during starting the second stage nozzles are allowed to send a lot of the hot gases to the LP shaft to help break the LP shaft away from zero speed. And when the LP is accelerating the second stage nozzles begin to modulate to control HP speed and fuel is used to control LP speed. (That's GE's two-shaft control philosophy.)

It would seem part of the problem (as is usually the case) is that the Mark VIe is tripping the turbine. I would be extremely surprised if the previous control system (Mark II or Mark IV) DIDN'T ALSO have a locked rotor trip during starting. And, that the setpoint wasn't the same as it is in the Mark VIe. I don't believe the problem is the Mark VIe; it's either something is amiss with the second stage nozzle set-up/calibration, or the torque required to accelerate one of the compressors (as has been noted) is excessive.

It's entirely too-common refrain: "It worked FINE before that Mark VIe was installed!" And, still no one is answering the question: If it worked with the Mark VIe before the outage with the same Mark VIe parameters as now, why doesn't it work now? The mechanical department is certainly saying everything is perfect, but if nothing was changed in the Mark VIe and only mechanical work was done during the outage, how can it be the Mark VIe that's causing the problem? (Answer: It can't. And it almost never is.)

I believe one respondent has also talked about clearances/alignment of the LP shaft.

So, based on the information provided--the problem is not the control system, it's mechanical. It spins up very easily when uncoupled from one of the compressors. What would change in the Mark VIe to make the LP shaft not spin when coupled to both of them?

As has also been said, if the LP turbine bearings were replaced, it's sometimes common for them to be very tight--and if a bearing alignment wasn't also done at the same time, that can make the tight fit even worse. But, it doesn't really sound like it's a LP bearing thing, or a Mark VIe thing.

Were the HP and LP overspeed tests done?

I wish I were able to help with this problem, but based on the information provided my best estimate is that it's not a control system problem. And that continuing to raise the locked rotor trip setpoint in an effort to get the LP shaft to break away when the LP shaft is coupled to both compressors is potentially harming the LP nozzles/buckets. That's all I can say, based on the information provided.

Please write back to let us know how you fare in resolving the problem!

Esteemed Sirs,

Following this thread A few thoughts came to mind, that may be worthwhile exploring.

[96NC] through [65NC]

There was no mention of a trend or observation of movement upon startup. It's possible to have calibration done, and still does not function upon sequencing and acceleration [ACC_CTRL].

What are the voltage levels for 77NC's? And V/%

I don't believe you have a mechanical problem. From what has been said so far. If you're under the gun & have not retired unlike CSA.
This is my handle c^2 and address (vcegce at gmail). It would be my pleasure.

PS: Not contradicting CSA in no way... but I sense something else.
I had a look at an original 2 shaft control scheme, and see a possible over looked control glitch in its carry over to a MKVIe.

Keep us posted!

I'm all for finding a way to help the original poster with his problem. I don't have access to the Mark VIe .tcw file from the site, and we only have the information he and Adeel have provided.

I am more than a little concerned about the number of thumbs-down responses to the posts here; it seems a little over the top based on similar past threads and attempts to provide help.

It should be noted that this site is not a replacement for having a knowledgeable technician come to site to assist with a problem if the issue is of an urgent nature.

And everyone is entitled to their opinion and estimate of what the problem may be. Mine is based on the information provided, and that it seems the unit operated properly after the Mark VIe upgrade up until the time the maintenance outage was performed (still unconfirmed by site personnel), and the information provided to date. It would seem the unit can break the LP shaft away when not coupled to the HP- and LP compressors, and that when the HP and LP compressors were manually rotated with the gearbox "...hardness was observed in rotation."

There is some confusion between Steps 7 & -8. Step 7 says the HP- and LP compressor couplings were removed (presumably from between the gearbox and the compressors), but that the LP turbine shaft was still coupled to the LP compressor. I presume the LP turbine shaft drives each compressor (HP and LP) through a gearbox (the BHS gearbox) with two output shafts, one connected to each compressor shaft (HP and LP). Something seems amiss with these details.

Step 8 says the LP turbine shaft was then tested only coupled to the gearbox (no compressors coupled to the gearbox), and the LP turbine shaft was able to break away.

It seems only the LP compressor shaft bearings and seals were examined, and the floating ring seals were replaced with a noticeable decrease in torque required to manually rotate the shaft (it's not clear if that was just the LP compressor shaft, or both compressor shafts (HP and LP)).

Anyway, any help is appreciated, and everyone doesn't need to agree on what the problem might be. My opinion is that there's nothing wrong with the turbine (HP and LP shafts are working fine), and without being able to see the application code running in the Mark VIe to know if it somehow knows when the LP shaft is coupled to a load (the gearbox/compressors) and might be affecting the control system/unit operation it's virtually impossible to provide much more comment. To my way of thinking, if the control system application code and parameters haven't changed from before the outage, and the unit (turbine and driven compressors) worked fine before the outage, and only mechanical work was done during the outage--the problem is likely mechanical. But, again, that's based on the information provided--and without being able to see the prior turbine control system's configuration/programming and the Mark VIe's configuration/programming. I do know that GE is now incorporating driven (centrifugal) compressor control into the Mark VIe, so that could be a possible source of the problem--but we can't know because we haven't been told and we can't see the application code in the Mark VIe.

Hoping to hear back from Haroon and/or Adeel! And, thanks for the help, Paul_c^2.