Summary:

The unit was shut down due to a Gas Valve Leak (GGV leak), and recently, we reinstalled our gas valve in the turbine and realigned it due to high vibration. Upon starting the unit, it was observed that the speed of the Gas Turbine (GT) did not hold at 100% during the Full Speed No Load (FSNL) state; instead, it continuously increased up to 106%. Consequently, we issued an emergency stop for safety concerns.

Unit Specification:

- Unit: Frame 6b (PG6541P)
- Control System: Mark V
- No DLN system installed

Sequence of Events:

1. The unit stopped 7 months ago due to high vibration and heavy gas leaks from the Gas Control Valve (GCV) valve.
2. The gas valve was serviced and inspected for gas leaks and operation at the site, and it was found to be okay.
3. Realignment was performed.
4. A serviced gas valve was installed.
5. GCV and SRV were calibrated and found to be okay, Vibration sensitivity changed (R core - IO.conf downloaded and rebooted).
6. GT was started, but the speed did not hold at FSNL.
7. SRV and GCV servo were replaced.
8. The P2 pressure transmitter was calibrated and found to be okay.
9. The hydraulic oil supply is confirmed to be okay.

Despite these efforts, the overspeed issue remains unresolved.

 

@vscontrol,

So, you really want to do this dance? You haven't provided a single alarm--Process or Diagnostic.

You didn't say whether this machine has a TMR Mark* V or a SIMPLEX Mark* V. (Which might help explain why you only downloaded to <R> if the Mark* V is a SIMPLEX unit. But not if the Mark* V is a TMR panel.)

Why was it felt the vibration sensitivity needed to be changed?

What was the vibration sensitivity changed from and to?

Did the vibration sensitivity change have a positive or negative effect on the vibration?

You say there were heavy leaks from the GCV. Meaning the GCV couldn't full stop gas fuel flow? GCV1 or GCV2? Or that it was leaking around valve stem seals or the casing?

What kind of SRV/GCVs does this machine use? Does it have the combined SRV/GCV assembly (two valves in one cast steel assembly)? Can you post a picture of the gas valves (all THREE of them)?

You said the "original" gas valve was serviced at the site, inspected and found to be working properly--using what procedure? Which gas valve(s)? The SRV, GCV1 and GCV2?

Are you certain the P2 pressure transmitter(s) are all calibrated correctly?

Are you certain all the manual valves on the gas valve assembly are in the proper positions for running? (There are usually three-way valves which have to be in one of two positions for running/controlling. It's unfortunately quite common for those valves NOT to be in the proper positions which can hamper proper operation. And if the gas valves were removed, serviced and re-installed (we'll get to that in the next question) were ALL of the valves returned to the proper positions?

You say a serviced gas valve (SRV, GCV1 and/or GCV2???) was installed. Was that the original valves which were serviced and tested on site, or valve(s) obtained from off-site?

Were servo-valve polarity checks performed on each gas valve which was removed and replaced? (How about any gas valve(s) that weren't removed and replaced--have the servo-valve polarity checks been done on those to verify or modify to correct operation?)

Why are there two GCVs (Gas Control Valves)? You said there was no DLN combustion system but there's something about Secondary on the fuel nozzles.?.?.?

We can't see what FSR was.

Also, we can't see what the IGVs are doing--are they stable at FSNL?

You do know that P2 pressure reference increases as speed increases, right? So, as speed increases the P2 pressure will increase--which will make for increasing fuel flow through the GCV? But why is the GCV not trying to control speed--or is at some kind of minimum limit (which I don't think I recall seeing, but I haven't looked at a Mark* V CSP in a couple of years...)? I think there was some kind of Minimum FSR control which was ostensibly in place to prevent flameout on generator trip events (when the turbine wasn't also tripped). Is the Mark* V running on that minimum FSR at FSNL and the speed is still increasing? That would mean there is still a leak of GCV1--and/or GCV2, wouldn't it?

FPG4 looks like "primary" gas fuel nozzle pressure (since it's only slightly above CPD). Please confirm.

The generator frequency was 50.24 Hz--which translates into 1.0048 PU (Per Unit) of frequency.

Take 5135 RPM and divide it by 1.0048 (PU frequency--which is also PU speed (since speed and frequency of a synchronous generator are directly related) and that means the machine's 100% rated speed is about 5109/5110 RPM. Which means that 5135 RPM for a 5110 RPM machine means the machine was running at 100.48% of rated speed--which ALSO doesn't appear on the Main Display. (And don't get me started on the Exhaust Temperature display--disgraceful. Just DISGRACEFUL. Full stop. Period.)

I'm CERTAIN someone at the site believes the machine should have been tripped on exhaust spread (combustion trouble) because several of the bargraphs are red and some are yellow. And there is some adjacency there, too, but I'm not calculating that because it wasn't in the original post. (BIG HINT: You can check adjacency by looking at the JXDn arrays....) [Those bargraphs with the red/yellow switching are so friggin' misleading it's pathetic--and, yet, GE still mis(uses) them. And what's with not identifying the bargraphs with numbers or the locations with alpha characters? THAT'S also disgraceful.]

So, it would appear there is some slight problem with some fuel nozzles--and there was zero mention of that in the original post. Which CAN'T be caused by the Mark* V, by the way.

Anyway, there's no way we can help you without a chance to look at the alarms. It looks like the HMI uses the WorkstationST Alarm Viewer--so it's possible to post ONLY the alarms for this machine (if it's a multi-unit site). AND why do people leave useless columns in the display? Is the alarm 'Description' different from the 'Primary Language Alarm Description'? If not, why not delete or hide the 'Primary Language Alarm Description instead of taking up valuable alarm message information with duplicate rows??? Again, disgraceful.

Nope; no help here with the information provided. None. Zilch. Nada. Zip. Niente. Nunca. It does look like the machine is in Droop Speed Control, and it's at 14HS, in AUTO and on GAS (whatever kind of gas GAS is--something else we weren't told). Why is there a second GCV??? ARE YOU 110% CERTAIN THAT SECOND GCV ISN'T LEAKING AND CAUSING THE MACHINE SPEED TO INCREASE WITHOUT CONTROL?

If you really want help, you gotta post alarms--ALL of them for this machine. Full stop. Period. AND if the HMI has WorkstationST then it should also have some kind of trending information and that would be very helpful, also. (Actually, there's not much we can do without it.)

So you think you can dance? (That's the actual name of a popular dance contest show on American television, but appropriate for this thread, too.)

Even without information you have some checks/verifications which will help eliminate causes and may help find the actual cause in the process. So, you have more than nothing, but n

 

• Mark v TMR
• Vibration sensitivity changed by Control TA, as he wants to use a 200mv vibration sensor in place of 150mv
• Not sure why they downloaded R core only ( later we performed a download for all (R, S, T)
• Initially, there was a leak at the GCV -1 stem gland pack and it was rectified by replacing the gland
• The unit has combined young Franklin SRV/GCV
• Servo polarity checked and found ok
• GCV used for CO2 emission control
• IGV was stable
• The unit was running with minimum FSR at FSNL and speed still increased
• Inspected and found GCV-2 had some minor leaks and it was rectified but still the speed increase

 

latest alarm display

 

Hi All

After reading the alarm messages it seems that LVDT showing "Out of limits" and "Voltage unbalanced" do you know for with device these LVDTs are for ?

As @WTF? asked are you confident with SRV GCVs 1 &2 LVDT calibrations ?

We will try to support remotely !

Also who is packager of the unit GEEPE Belfort , Nuovo pignone or John Brown Massood..??

James

 

Hi All

After reading the alarm messages it seems that LVDT showing "Out of limits" and "Voltage unbalanced" do you know for with device these LVDTs are for ?
- "S" core CSGV (kindly find the attached photo)
- S core - FSGR abnormal
- R core 125VDC power supply abnormal (S&T is equal +64 , - 63VDC)

As @WTF? asked are you confident with SRV GCVs 1 &2 LVDT calibrations ?

- calibrations are done by the control TA from the maintenance team and I plan to redo calibration and start all investigations from tomorrow onwards as the team left the site with this problem due to management issues.

We will try to support remotely !

- thanks for your kind words sir

Also who is packager of the unit GEEPE Belfort , Nuovo pignone or John Brown Massood..??

- European gas turbine (EGT)

James

 

Thanks for the information. It's going to take me a while to go through the .csv file.

From the alarm photos, I really don't like the 125 VDC Battery Ground alarm.... If it's intermittent and dithering (changing frequently) it's not a good indication.

Also, there are several TCQA alarms regarding position feedback. REG1 is probably the SRV; if so, REG2 is most likely GCV1, and I think REG8 (from foggy memory--it's been a few years) is probably GCV2, but that might not be depending on the type and manufacturer of valve.

If as you say the machine is running at MIN FSR at FSNL and the speed is increasing--that suggests MIN FSR is too high (and why would it change??? has fuel supplier changed--fuel (does this machine burn LNG?)) and/or the LVDT calibration IS NOT correct, and the TCQA alarms seem to point to that as a possible problem.

When the machine is at zero speed and not running, and the Aux. Hyd. Pump is not running can you measure the gap between the SRV and GCV1 valve stems and the actuator rods (the valve stem is above the actuator rod, and has the bar with the LVDTs attached to it. Take a large pry bar and pry DOWN on the LVDT bar from above CLOSE TO THE VALVE STEM AND GENTLY (don't bend the LVDT bar). This is to ensure the actuator rod is fully down. Then use the pry bar to pry UP on the LVDT bar--again CLOSE TO THE VALVE STEM AND GENTLY so as not to bend the LVCDT bar. (Often it is necessary to use a block of wood under the end of the pry bar to make it easy to move the pry bar without needing a lot of vertical space for the end of the pry bar you are using. If you start with the pry bar close to horizontal just before prying while in contact with the LVDT bar with the bar by using a block at the far end it will be easier and simpler to pry down or up.)

After prying down on the LVDT bar, and then prying up on the LVDT bar it should be possible to see a small gap between the bottom of the valve stem and the top of the actuator rod. Use an Imperial set of feeler gauges (or convert the dimensions to metric) and you should find the gap to be somewhere in the range of 0.030 inches, plus-or-minus 0.010 inches, or between 0.030 inches and 0.050 inches. If that gap is not present or is too great that can have a large impact on the LVDT calibration.

Speaking of LVDT calibration, prior to starting the LVDT valibration procedure it is important to use the pry bar to create ("open") the gap and then insert feeler gauges between the top of the actuator rod and the bottom of the valve stem AND LEAVE THE FEELER GAUGES IN PLACE DURING THE CALIBRATION AND THE VERIFICATION OF THE CALIBRATION! This position is the true 0% stroke position of the valve(s) (the actuator rod pushes the valve stem UP to OPEN the valve). It's also at this position (0% stroke) that one should adjust the LVDT zero stroke voltages to approximately 0.700 VAC RMS, plus-or-minus 0.020 VAC RMS. It is VERY important to use a TRUE AC RMS voltmeter that is capable of frequencies in the 3 kHz range!!! (A lot of newer Fluke multimeters say they are TRUE AC RMS--but ONLY up to 1 kHz!!!) Using an improper meter can result in very erratic and intermittent readings. (I should note, that while looking at the AutoCalibrate display for a particular valve the LVDT voltage is displayed, though it usually has a negative sign in front of it the magnitude is the value you're interested in, not the "polarity".)

I would also recommend checking the ACALIB.DAT file in the unit-specific directly to make sure the MIN_POS_SAT and MAX_POS_SAT values (or some similar text) FOR THE LVDTS BEING CALIBRATED is 0.00 and 100.00. (I personally like to have the MIN_POS_SAT value at about 0.010 just because that ensures the valve is being allowed to fully close when the reference is 0.00%, and it's a very small error that is negligible. Not critical, but it's worked well for me on some sites with some unusual valve issues.)

Then with the LVDT 0% stroke voltage properly set for BOTH LVDTs, AND the feeler gauges STILL IN PLACE, start the AutoCalibrate procedure. PAY ATTENTION to any warning or error messages which may appear in the AutoCalibrate display. (There is usually a couple/three Diagnostic Alarms that are annunciated when performing an AutoCalibration, but they're just because GE factory engineers didn't think it was important to "bypass" those alarms when someone was performing an AutoCalibration, because, and I quote, "Everyone knows those alarms don't apply when performing an AutoCalibration." TRUE story. Sad; but true.)

It's also very important PRIOR to performing LVDT calibrations that the servo current polarities for all three processors are correct.... (I know, I should have--and could have--put this at the top of the procedure, but it should be obvious to anyone performing LVDT calibrations that the servo currents have to be correct PRIOR to performing LVDT calibrations (or even verifications).

That should be the process, and it should clear up those TCQA Diagnostic Alarms.

 

I want to add that when the feeler gauges are removed from the gap between the valve stem and actuator rod it's common for the LVDTs to indicate a negative position (because the valve stem has dropped below (less than) the true 0% stroke position). That's normal and it shouldn't be more than -1% or -2% and won't affect anything at all--again, because it's normal and to be expected. Sometimes the valve stem and LVDT bar won't drop immediately after removing the feeler gauges; sometimes it won't drop until the next time the machine is started and trips or shuts down. But, again a very slight negative position feedback from the LVDTs is normal and expected and doesn't create any problems (unless the feedback is less than -5% (the typical alarm value)).

That gap helps to ensure there is nothing preventing the valve(s) from closing when the reference is zero or less. It's important and needs to be there for safety purposes.

The other important thing to ensure is the roll pin that is used to "pin" (secure) the LVDT bar to the valve stem is installed. It's very common for that roll pin to not be re-installed or too small a roll pin is used (and it vibrated out)--which allows the LVDT bar to slide down on the valve stem which messes up the LVDT calibration.

So, don't worry and don't let others try to tell you a very slight negative position indication is bad--it's not. It's normal and to be expected. (The design of those Y&F gas valves is not typical of most other fuel control valves, but they have worked for several decades without serious problems. Yes; the valve stem packing does wear, and some contaminants in gas fuel (sand; silica; small rocks) can even damage the valve plug and valve seat over time--but that's not a valve problem, it's a fuel cleanliness problem. MANY natural gas suppliers no longer do a very good job of cleaning contaminats out of the gas fuel they pump into a pipeline. Pipeline companies don't get paid for supplying clean gas so they don't filter or remove contaminants--and they often are the source of some bad contaminants (compressor lubricating oil; petrol and/or diesel used for cleaning pipes and valves; etc.). They also don't get paid for maintaining a particular dewpoint of the gas, which can change depending on where the gas comes from and ambient conditions and the number of pressure drops the gas goes through before reaching the power plant. And, there are still pressure drops as the gas fuel passes through control valves (SRV and GCV) and fuel nozzle orifices.

But, I digress. A slight negative gas valve position indication is NOT unexpected, and doesn't affect normal operation (unless it gets too large--say from someone forgetting to put in a roll pin or using too small a roll pin--and there have even been cases of roll pins shearing (too many turbine trips!!!).

 

Thank you for these informations

Ok so its GEEPE called before EGT (European Gas Turbine)

Did you contact them for this issue ? Is that TA worked for or with GEEPE (EGT) if yes what they said for this issue?

 

Thanks for the information. It's going to take me a while to go through the .csv file.

From the alarm photos, I really don't like the 125 VDC Battery Ground alarm.... If it's intermittent and dithering (changing frequently) it's not a good indication.

- 125 VDC Ground alarm cleared (found ground from 11KV Generator breaker control circuit)

Also, there are several TCQA alarms regarding position feedback. REG1 is probably the SRV; if so, REG2 is most likely GCV1, and I think REG8 (from foggy memory--it's been a few years) is probably GCV2, but that might not be depending on the type and manufacturer of valve.

- all diagnostic alarms cleared

• The S-core TCQA card was replaced (all FPRG, GSCV, and LVDT diagnostic alarms were gone)
  
• The R-core TCPS card was replaced and TCQA 125VDC fail diagnostic alarms cleared

If as you say the machine is running at MIN FSR at FSNL and the speed is increasing--that suggests MIN FSR is too high (and why would it change??? has fuel supplier changed--fuel (does this machine burn LNG?)) and/or the LVDT calibration IS NOT correct, and the TCQA alarms seem to point to that as a possible problem.

- Plan to replace the GCV - LVDTS (will update calibration and polarity checks results later here)

When the machine is at zero speed and not running, and the Aux. Hyd. Pump is not running can you measure the gap between the SRV and GCV1 valve stems and the actuator rods (the valve stem is above the actuator rod, and has the bar with the LVDTs attached to it. Take a large pry bar and pry DOWN on the LVDT bar from above CLOSE TO THE VALVE STEM AND GENTLY (don't bend the LVDT bar). This is to ensure the actuator rod is fully down. Then use the pry bar to pry UP on the LVDT bar--again CLOSE TO THE VALVE STEM AND GENTLY so as not to bend the LVCDT bar. (Often it is necessary to use a block of wood under the end of the pry bar to make it easy to move the pry bar without needing a lot of vertical space for the end of the pry bar you are using. If you start with the pry bar close to horizontal just before prying while in contact with the LVDT bar with the bar by using a block at the far end it will be easier and simpler to pry down or up.)
After prying down on the LVDT bar, and then prying up on the LVDT bar it should be possible to see a small gap between the bottom of the valve stem and the top of the actuator rod. Use an Imperial set of feeler gauges (or convert the dimensions to metric) and you should find the gap to be somewhere in the range of 0.030 inches, plus-or-minus 0.010 inches, or between 0.030 inches and 0.050 inches. If that gap is not present or is too great that can have a large impact on the LVDT calibration.

• As found "zero" clearness between the valve stem and the actuator rod and not able to pull down the actuator rod by following your method (all oil circuits are switched off)
• this unit does not have an auxiliary hydraulic pump and requires crank speed every time for the calibration
• Plan to inspect the GCV actuator (will update the clearness later here)

Speaking of LVDT calibration, prior to starting the LVDT valibration procedure it is important to use the pry bar to create ("open") the gap and then insert feeler gauges between the top of the actuator rod and the bottom of the valve stem AND LEAVE THE FEELER GAUGES IN PLACE DURING THE CALIBRATION AND THE VERIFICATION OF THE CALIBRATION! This position is the true 0% stroke position of the valve(s) (the actuator rod pushes the valve stem UP to OPEN the valve). It's also at this position (0% stroke) that one should adjust the LVDT zero stroke voltages to approximately 0.700 VAC RMS, plus-or-minus 0.020 VAC RMS. It is VERY important to use a TRUE AC RMS voltmeter that is capable of frequencies in the 3 kHz range!!! (A lot of newer Fluke multimeters say they are TRUE AC RMS--but ONLY up to 1 kHz!!!) Using an improper meter can result in very erratic and intermittent readings. (I should note, that while looking at the AutoCalibrate display for a particular valve the LVDT voltage is displayed, though it usually has a negative sign in front of it the magnitude is the value you're interested in, not the "polarity".)

I would also recommend checking the ACALIB.DAT file in the unit-specific directly to make sure the MIN_POS_SAT and MAX_POS_SAT values (or some similar text) FOR THE LVDTS BEING CALIBRATED is 0.00 and 100.00. (I personally like to have the MIN_POS_SAT value at about 0.010 just because that ensures the valve is being allowed to fully close when the reference is 0.00%, and it's a very small error that is negligible. Not critical, but it's worked well for me on some sites with some unusual valve issues.)

• Please find attached Photos for your reference

Then with the LVDT 0% stroke voltage properly set for BOTH LVDTs, AND the feeler gauges STILL IN PLACE, start the AutoCalibrate procedure. PAY ATTENTION to any warning or error messages which may appear in the AutoCalibrate display. (There is usually a couple/three Diagnostic Alarms that are annunciated when performing an AutoCalibration, but they're just because GE factory engineers didn't think it was important to "bypass" those alarms when someone was performing an AutoCalibration, because, and I quote, "Everyone knows those alarms don't apply when performing an AutoCalibration." TRUE story. Sad; but true.)

It's also very important PRIOR to performing LVDT calibrations that the servo current polarities for all three processors are correct.... (I know, I should have--and could have--put this at the top of the procedure, but it should be obvious to anyone performing LVDT calibrations that the servo currents have to be correct PRIOR to performing LVDT calibrations (or even verifications)

That should be the process, and it should clear up those TCQA Diagnostic Alarms.

- all alarms cleared (kindly guide me If i miss anything here) and thank you for your support

 

Briefly, the procedure for "opening" the gap between the valve stem(s) and the actuator rod(s) is to push the LVDT bar down to push the actuator rod fully down into the actuator. Then push the LVDT bar up to push the valve stem up to contact the valve plug. THIS should "open" any existing gap between the actuator rod and valve stem. If there is no Aux. Hyd. Pump, then as long as the machine is at rest (zero speed) there should be no hydraulic pressure to push the actuator rod up, pushing the valve stem up. It usually takes several minutes, sometimes longer, for the hydraulic pressure to bleed to zero--especially if there is a Hydraulic Accumulator (that is properly charged and properly valved in (block valve open; bleed valve closed)). Sometimes, if there is a Hydraulic Accumulator, all that's necessary to quickly bleed off any hydraulic pressure is to open the Hydraulic Accumulator bleed valve for 30 seconds to one minute, and then close it again; hydraulic pressure should be at 0.

So, it's basically a two-step process: force the LVDT bar down (to force the actuator rod down), then force the LVDT bar up (to force the valve stem up to contact the valve plug). Excessive force should not be required for either step. If you do both of these steps and there is no gap, then whoever disassembled and reassembled the valve(s) did not read the NOTES section of the Gas Valve Drawing (the drawing number is stamped on the nameplate, but it seems to have been damaged). The gap is clearly shown on the drawing, as well. Either there is something amiss with the actuator (leaky seal of the actuator piston), or it was not assembled and measured properly, or there is something amiss with the valve stem/valve plug interface. The purpose of the gap is to ensure that nothing is holding the valve plug from being fully seated--to prevent leaking.

If the SRV has no gap then that means it is probably leaking gas fuel at least a little bit which can be a problem during starting. That would be evident from gas exiting from the goose-neck vent above the Gas Valve Compartment--when the unit is at rest and there is gas fuel pressure upstream of the SRV. That gas will leak into the intervalve area between the SRV and GCV1, and it will flow out of the Fuel Gas Vent Valve Solenoid, 20VG-1, and into the goose-neck vent above the Gas Valve Compartment. If GCV1 has a gap then one could force 20VG-1 solenoid to close--thereby trapping the gas in the intervalve area. The pressure will increase, probably slowly if it's a small leak, or quickly if it's a large leak, as seen on the P2 pressure gauge on the combined Gas Valve Assembly.

If GCV1 is also leaking (has no gap) this would also cause some gas fuel to flow into the combustors. This would not be a good thing. And, it would also prevent the intervalve pressure to increase slower than it otherwise might (if GCV1 was not leaking) if performing the test with 20VG-1.

But, both the SRV and GCV1 should have some gap between the actuator rod and valve stem--0.030 inches to 0.050 inches is the factory specification, but if the gap is slightly less than 0.030 inches, that's not all bad. If it's between 0.005 inches and 0.000 inches, that's probably not very good. And, if the gap is very small or nonexistent for GCV1 then that could be why there's too much fuel flowing into the machine at FSNL and the speed can't be controlled.

I'm presuming the servo current null bias is correct for both gas valves (SRV and GCV1) and GCV2 is NOT leaking by. If there's some kind of manual isolation valve upstream of GCV2 I suggest closing it for testing at FSNL.

I'm also presuming the servo-valve polarity for each processor/servo coil was correctly checked and corrected if necessary.

If one replaces a TCQA in a TMR Mark* V turbine control panel then the LVDT calibrations for all devices need to be done again.

I also note that the POSITION_NEG_SAT and POSITION_POS_SAT values for the IGVs are incorrect. 34 and 84 are the values of CSKGVMIN and CSKGVMAX, respectively--NOT the minimum and maximum mechanical positions of the actuator ring stops. This have to be properly measured and put into the appropriate fields. Typically, the factory settings for minimum and maximum mechanical stops are approximately 32 and 86 or 87 DGA, but that's just the desired settings; sometimes it's a little more or less at one end or the other. But, 34 and 84 are just default values put into ACALIB.DAT because some value is needed for factory testing; the commissioning person is supposed to measure and input the actual values found during commissioning.

 

@WTF? , @ControlsGuy25 ,

Problem solved

• As per your advice, I made a clearness between the GCV valve stem and the GCV actuator rod
• GCV- LVDTs replaced and calibration and polarity checks were carried out
• The unit started and it was able to hold speed at FSNL(100.15%) and all other parameters are within the limits

Thank you so much for your help and support.

 

@vscontrol

Happy to hear good news from your site

Thank you for this update and feedback

Regards,
James

 

Summary:

The unit was shut down due to a Gas Valve Leak (GGV leak), and recently, we reinstalled our gas valve in the turbine and realigned it due to high vibration. Upon starting the unit, it was observed that the speed of the Gas Turbine (GT) did not hold at 100% during the Full Speed No Load (FSNL) state; instead, it continuously increased up to 106%. Consequently, we issued an emergency stop for safety concerns.

Unit Specification:

- Unit: Frame 6b (PG6541P)
- Control System: Mark V
- No DLN system installed

Sequence of Events:

1. The unit stopped 7 months ago due to high vibration and heavy gas leaks from the Gas Control Valve (GCV) valve.
2. The gas valve was serviced and inspected for gas leaks and operation at the site, and it was found to be okay.
3. Realignment was performed.
4. A serviced gas valve was installed.
5. GCV and SRV were calibrated and found to be okay, Vibration sensitivity changed (R core - IO.conf downloaded and rebooted).
6. GT was started, but the speed did not hold at FSNL.
7. SRV and GCV servo were replaced.
8. The P2 pressure transmitter was calibrated and found to be okay.
9. The hydraulic oil supply is confirmed to be okay.

Despite these efforts, the overspeed issue remains unresolved.

Have you check and calibrate the SRV and GCV LVDT feedback? Insure if you don’t do that.
Check FSR (GAG at 100%)

 

@Inst control,

Please read response #12; it states the GCV1 LVDTs were replaced and the LVDT feedback was calibrated and the problem was solved.

The FSR GAG alarm was not present in any of the photos of the WorkstationST Alarm Viewer the original poster attached to this thread. AND, gagging FSR only LIMITS the amount of fuel that is flowing into the combustors; it can't be used to increase the fuel flow--and the problem was about too much fuel flow causing FSNL speed to increase