Major inspection maintenance was carried out on one of our Gas turbine. It is Frame VI (PG6581B) using MKVI controller system. But it was discovered later that the performance of the Gas turbine system in terms of Megawatt output after the MI is slightly lower than it's performance before the MI. Please I welcome very good suggestion about what might be likely cause of the drop in Megawatt, since the essence of MI is to have better output performance and more Megawatt.
Thanks

 

Sirdee
Welcome to Control.Com!
How much is "Slightly"?
Regards, Phil Corso

 

Please consider also ambient temperature before and after MI.
secondly have you made any changes in the system /parameters except inspections ?
How frequently washing online/offline is carried out
look at fuel system, fuel quality , air quality, air intake filters conditions and finally calibration of load transducer.

 

Sirdee
Welcome to Control.Com!
How much is "Slightly"?
Regards, Phil Corso

Thanks.
Slightly is about 3 to Megawatt.
The MI maintenance job was contracted out to company. We have eight GTs on site with the same capacity and ratings and one of the machine was given out for MI.

 

Any diagnostic /process alarms displayed ?

What are the major components which have been replaced /changed...

Fuel type used ?

 

Sirdee...
What is (are) the rated MVA or MW of the Generator(s)?
Regards, Phil Corso

 

Please consider also ambient temperature before and after MI.
secondly have you made any changes in the system /parameters except inspections ?
How frequently washing online/offline is carried out
look at fuel system, fuel quality , air quality, air intake filters conditions and finally calibration of load transducer.

Good day. The ambient air temperature is the same, there is no difference. Then the system parameters was not changed during the maintenance period.The air inlet filter has been changed after the incidence. Although, the load transducer has not been calibrated because we think that the two transducer cannot be bad at the same time. Thanks.

 

Any diagnostic /process alarms displayed ?

What are the major components which have been replaced /changed...

Fuel type used ?

Good day There is no process or diagnostic alarms displayed..The following are major components changed during the maintenance period.
Turbine buckets(1st Stage, 2nd Stage and 3rd Stage).
Turbine Nozzle (1st Stage, 2nd Stage and 3rd Stage).
Transition Pieces.
The Compressor was brought out and all the blades were cleaned and inspected.
IGV was opened up, the blades were taking out , the IGV cylinder too and so many things were inspected,and replaced the suspected bad ones

 

You--and the Mechanical Department--don't want to hear this, but ... The problems is MOST LIKELY NOT the turbine control system. If you didn't change any parameters in the Mark* VI AND if you're sure the following things are correct:

a) IGVs were properly installed and measured and the IGV LVDTs are properly calibrated after the upper casing was removed and the IGV hydraulic actuator was removed and replaced (possibly with a new one)

b) the exhaust thermocouples were properly installed during reassembly--INCLUDING the wiring terminations of the thermocouple leads (thermocouple wires are very particular; this topic has been covered MANY times before on Control.com; you can use the Search feature to find previous threads about thermocouple terminations)

c) you are ONLY comparing the power output of the machine before and after the MI while operating at Base Load, and both measurements should have been taken with the machine internals at stable temperatures (typically defined as NO wheelspace temperature has changed more than 5 deg F (approximately 2-3 deg C) in 15 minutes

d) the fuel make-up (LHV and HHV) hasn't changed since before the MI

e) you're 100% certain the CPD pressure transmitters are all calibrated correctly and the manual valves in the lines to the transmitters are all in the correct positions.

Item 'a' is a two-fer--meaning it has two points. If the IGVs were removed--even if they weren't ALL individually removed from the bellmouth halves to replace one or more IGV blades and/or the IGV bushings--and the IGV hydraulic actuator was removed (meaning the IGV LVDTs were removed since they are attached to the actuator assembly) it's CRITICAL that the mechanical stops be checked and set one the upper half of the bellmouth is back on and the actuator assembly is in place BEFORE the IGV LVDTs are calibrated! This is often not done correctly and can affect IGV position. This includes making any adjustments to the IGV position indicator ("pointer") on the side of the axial compressor casing if necessary. The air flow through the axial compressor and turbine is JUST AS IMPORTANT as the fuel flow, so the IGVs (and the IGV LVDTs MUST be calibrated correctly to compare the outputs.) AND, it is probably being ASSUMED that the IGV LVDTs were properly calibrated BEFORE the MI--which unless the calibration was checked and verified (and changed if necessary) then it's not really possible to compare the output before and after the MI, even if all the IGV system reassembly and LVDT calibration was done correctly.

The second part of item 'a' is IGV LVDT calibration--it needs to be done properly AND it needs to be verified. This, too, has been covered many times before on Control.com. To use the "auto-calibration" feature of the Mark* VI it's necessary to know PRECISELY what angles the IGVs are at when they at the full-closed mechanical stop AND the full open mechanical stop for the auto-calibration to work properly.

AND all of this has to be done by knowledgeable people who know how to read the device being used to measure the IGV angles (usually a machinist's protractor)--and MOST machinist's protractors have to have had the long blade machined down to about half of its normal width to fit in between adjacent IGV blades when the IGVs are fully closed against the mechanical stop.

Item 'b' is also very important--because it's not enough to just "connect the yellow lead to the yellow lead and the red lead to the red lead", it has be done in a very specific manner OR inaccuracies will be introduced into the readings which can lead to problems with performance readings. This, too, has been covered many times before on Control.com.

Item 'c' is VERY important--because the OEM makes NO performance guarantees at any load other than Base Load--which by definition means the IGVs are at the maximum operating angle. AND, that the machine internal temperatures (the wheelspace temperatures) have been steady and stable for a period of time, and usually after a start from a "cold" condition that takes about 4 hours of operation at Base Load to achieve stable and steady wheelspace temperatures (machine internal temperatures).

Item 'd' is also very important. People don't realize it but fuel make-up (fuel chemistry) is very important and can change over time, and if there is a different supplier (for example, if the machine burns liquid fuel and there is a new liquid fuel supplier it should be tested to make sure it meets the requirements the OEM set for the machine when it was new). Even natural gas can change over time, and if a new field is brought on to the pipeline or one is taken off-line. Fuel is not fuel--it has to be tested and known/proven to be as per specification.

Item 'e' is also VERY IMPORTANT to proper operation at Base Load. AND, if the machine has CPR biasing of the maximum allowable exhaust temperature then the calibration of those transmitters ARE ALSO VERY CRITICAL. And, again--unless the calibration of these transmitters wasn't verified BEFORE the MI and they have been adjusted during calibration verification during the MI then its very difficult to say if performance is high or low.

That's all the controls-related stuff--for Base Load Operation. If you're comparing operating at any other load, all bets are off. Full stop. Period.

Now, if that's all good and verified and acceptable, then the problem is mechanical. Either the axial compressor bleed valves are passing (leaking) or the fuel isn't what is was or the clearances of the side seals between the transition pieces are correct or the combustion hardware wasn't installed correctly. Or the exhaust duct back-pressure isn't what it was before the MI, or isn't per specification. Or the inlet air filters aren't the same specification as the OEM or even the previous filters (MANY times the Sourcing Department finds a less expensive filter for a lower price and buys that without considering what effect it will have on machine operation and performance; it's just a filter, right? WRONG). If the openings in the combustion liners and transition pieces weren't checked and verified that can cause issues. If the turbine nozzle segments were refurbished or even not supplied by the OEM it's possible they don't meet the OEM spec. What about the seal strips on the turbine blade ends?

There are actually a LOT of things which can cause problems--and it's even very likely that more than one thing is wrong and together they are contributing to the issue of perceived low output/performance. But, again, if you're not comparing output/performance at Base Load before and after the MI, and if the critical machine instrumentation (IGV LVDTs; CPD transmitters; barometric pressure transmitters (if used)) wasn't verified before the outage it's really hard to say what the problem is. If proper attention wasn't paid by site personnel during reassembly, if the parts used weren't OEM or known good quality third party equivalents and if the IGVs and IGV LVDT calibratin wasn't done correctly and the CPD transmitters are calilbration correctly (and the barometric pressure transmitters calibrated and manual valves KNOWN to be in the proper positions) then all manner of things can be happening. BUT, comparing machine output at Base Load is critical. And, if you're confident the controls stuff is all good--then it's something mechanically incorrect.

It's easiest to go through all of the controls stuff first and prove it's all correct (because EVERYONE is going to blame the Mark* VI first--best to prove it, and its inputs and outputs are correct so that can be eliminated as the cause of the problem). After that, if the problem persists--it's mechanical and someone has to decide what the next course of action will be to find and resolve the problem.

As a former contributor to Control.com used to say, "Hope this helps!" It's not a good problem to have. Ever. That's why supervision is SO important during any maintenance outage, but especially during a MI.

 

The

Sirdee...
What is (are) the rated MVA or MW of the Generator(s)?
Regards, Phil Corso

Good day.
The rating if the generator is 42MW and 50MVA

You--and the Mechanical Department--don't want to hear this, but ... The problems is MOST LIKELY NOT the turbine control system. If you didn't change any parameters in the Mark* VI AND if you're sure the following things are correct:

a) IGVs were properly installed and measured and the IGV LVDTs are properly calibrated after the upper casing was removed and the IGV hydraulic actuator was removed and replaced (possibly with a new one)

b) the exhaust thermocouples were properly installed during reassembly--INCLUDING the wiring terminations of the thermocouple leads (thermocouple wires are very particular; this topic has been covered MANY times before on Control.com; you can use the Search feature to find previous threads about thermocouple terminations)

c) you are ONLY comparing the power output of the machine before and after the MI while operating at Base Load, and both measurements should have been taken with the machine internals at stable temperatures (typically defined as NO wheelspace temperature has changed more than 5 deg F (approximately 2-3 deg C) in 15 minutes

d) the fuel make-up (LHV and HHV) hasn't changed since before the MI

e) you're 100% certain the CPD pressure transmitters are all calibrated correctly and the manual valves in the lines to the transmitters are all in the correct positions.

Item 'a' is a two-fer--meaning it has two points. If the IGVs were removed--even if they weren't ALL individually removed from the bellmouth halves to replace one or more IGV blades and/or the IGV bushings--and the IGV hydraulic actuator was removed (meaning the IGV LVDTs were removed since they are attached to the actuator assembly) it's CRITICAL that the mechanical stops be checked and set one the upper half of the bellmouth is back on and the actuator assembly is in place BEFORE the IGV LVDTs are calibrated! This is often not done correctly and can affect IGV position. This includes making any adjustments to the IGV position indicator ("pointer") on the side of the axial compressor casing if necessary. The air flow through the axial compressor and turbine is JUST AS IMPORTANT as the fuel flow, so the IGVs (and the IGV LVDTs MUST be calibrated correctly to compare the outputs.) AND, it is probably being ASSUMED that the IGV LVDTs were properly calibrated BEFORE the MI--which unless the calibration was checked and verified (and changed if necessary) then it's not really possible to compare the output before and after the MI, even if all the IGV system reassembly and LVDT calibration was done correctly.

The second part of item 'a' is IGV LVDT calibration--it needs to be done properly AND it needs to be verified. This, too, has been covered many times before on Control.com. To use the "auto-calibration" feature of the Mark* VI it's necessary to know PRECISELY what angles the IGVs are at when they at the full-closed mechanical stop AND the full open mechanical stop for the auto-calibration to work properly.

AND all of this has to be done by knowledgeable people who know how to read the device being used to measure the IGV angles (usually a machinist's protractor)--and MOST machinist's protractors have to have had the long blade machined down to about half of its normal width to fit in between adjacent IGV blades when the IGVs are fully closed against the mechanical stop.

Item 'b' is also very important--because it's not enough to just "connect the yellow lead to the yellow lead and the red lead to the red lead", it has be done in a very specific manner OR inaccuracies will be introduced into the readings which can lead to problems with performance readings. This, too, has been covered many times before on Control.com.

Item 'c' is VERY important--because the OEM makes NO performance guarantees at any load other than Base Load--which by definition means the IGVs are at the maximum operating angle. AND, that the machine internal temperatures (the wheelspace temperatures) have been steady and stable for a period of time, and usually after a start from a "cold" condition that takes about 4 hours of operation at Base Load to achieve stable and steady wheelspace temperatures (machine internal temperatures).

Item 'd' is also very important. People don't realize it but fuel make-up (fuel chemistry) is very important and can change over time, and if there is a different supplier (for example, if the machine burns liquid fuel and there is a new liquid fuel supplier it should be tested to make sure it meets the requirements the OEM set for the machine when it was new). Even natural gas can change over time, and if a new field is brought on to the pipeline or one is taken off-line. Fuel is not fuel--it has to be tested and known/proven to be as per specification.

Item 'e' is also VERY IMPORTANT to proper operation at Base Load. AND, if the machine has CPR biasing of the maximum allowable exhaust temperature then the calibration of those transmitters ARE ALSO VERY CRITICAL. And, again--unless the calibration of these transmitters wasn't verified BEFORE the MI and they have been adjusted during calibration verification during the MI then its very difficult to say if performance is high or low.

That's all the controls-related stuff--for Base Load Operation. If you're comparing operating at any other load, all bets are off. Full stop. Period.

Now, if that's all good and verified and acceptable, then the problem is mechanical. Either the axial compressor bleed valves are passing (leaking) or the fuel isn't what is was or the clearances of the side seals between the transition pieces are correct or the combustion hardware wasn't installed correctly. Or the exhaust duct back-pressure isn't what it was before the MI, or isn't per specification. Or the inlet air filters aren't the same specification as the OEM or even the previous filters (MANY times the Sourcing Department finds a less expensive filter for a lower price and buys that without considering what effect it will have on machine operation and performance; it's just a filter, right? WRONG). If the openings in the combustion liners and transition pieces weren't checked and verified that can cause issues. If the turbine nozzle segments were refurbished or even not supplied by the OEM it's possible they don't meet the OEM spec. What about the seal strips on the turbine blade ends?

There are actually a LOT of things which can cause problems--and it's even very likely that more than one thing is wrong and together they are contributing to the issue of perceived low output/performance. But, again, if you're not comparing output/performance at Base Load before and after the MI, and if the critical machine instrumentation (IGV LVDTs; CPD transmitters; barometric pressure transmitters (if used)) wasn't verified before the outage it's really hard to say what the problem is. If proper attention wasn't paid by site personnel during reassembly, if the parts used weren't OEM or known good quality third party equivalents and if the IGVs and IGV LVDT calibratin wasn't done correctly and the CPD transmitters are calilbration correctly (and the barometric pressure transmitters calibrated and manual valves KNOWN to be in the proper positions) then all manner of things can be happening. BUT, comparing machine output at Base Load is critical. And, if you're confident the controls stuff is all good--then it's something mechanically incorrect.

It's easiest to go through all of the controls stuff first and prove it's all correct (because EVERYONE is going to blame the Mark* VI first--best to prove it, and its inputs and outputs are correct so that can be eliminated as the cause of the problem). After that, if the problem persists--it's mechanical and someone has to decide what the next course of action will be to find and resolve the problem.

As a former contributor to Control.com used to say, "Hope this helps!" It's not a good problem to have. Ever. That's why supervision is SO important during any maintenance outage, but especially during a MI.

Thank you very much. The control aspect has been checked, the Value of CPD, the value obtained from 96CS was also look into, the ambient and exhaust temperature was also carefully look into. The quality of the fuel is the same as that of the one going into other unit and other GTs are performing better. From the HMI, the IGV angle us 85 degree, FSR is okay and all the parameters in FSR control page also looks good.
Then from observation, whenever we did HGPI, we always get improved performance, but now we did Major Inspection we get low performance.
All the point you raised will be discussed with all the parties concerned.
If there any other thing to add ,please I will be expecting from you.
Thanks and God bless

 

The barometric/atmospheric pressure transmitters I was referring to are 96AP devices; they can be very difficult to calibrate and insects usually find some way to build nests in the pressure port(s) because the sintered metal covers usually get removed or damaged and not replaced with proper replacements.

I'm a little leery of an indicated IGV angle of 85 DGA.... The usual Control Constant name for the maximum operating angle is CSKGVMAX, and for ,most GE-design Frame 6B heavy duty gas turbines the value is typically 84 DGA. So, if the indicated angle is 85 DGA and CSKGVMAX is 84 DGA, something is amiss (the servo current polarity is not correct; the LVDT calibration was not properly verified (using proper measuring technique), and/or the IGV regulator null bias value is incorrect. A LOT of sites use 34 and 84 as the values for minimum and maximum mechanical stroke in the auto-calibration routine--and that is INCORRECT. The minimum and maximum mechanical stop values MUST BE PHYSICALLY MEASURED to determine what exactly (or reasonable close) the mechanical stops were set at, and those values must be entered in the appropriate fields of the auto-calibrate routine. AND, it usually takes a machinist's protractor with a ruler blade cut down to about half its normal width (by a knowledgeable machinist using properly milling tools to prevent warping the ruler) to fit between IGV blades when they are closed and at minimum operating angle (usually 34 DGA--Control Constant value CSKGVMIN). The mechanical stops (closed and open) should be OUTSIDE the minimum and maximum operating angles, something like 32 DGA and 86 DGA are pretty typical values, but I've seen some sites with 30.5 DGA and 87.75 DGA--and, again, these measurements must be taken using proper technique (to negate the normal hysteresis from the rack mechanism while holding the IGV in the position it will normally be in when the unit is running). This has to be done for EVERY measurement, both before the IGV auto-calibration and after the IGV auto-calibration when verifying the accuracy of the calibration. AND, every IGV LVDT calibration should be done with the normal, typical value of regulator null bias current AND AFTER the polarity of the servo currents (all three of them) have been properly checked and corrected if necessary.

All of this has been covered many times before on Control.com. Auto-calibrate ONLY calibrates LVDT feedback; nothing else (meaning nothing to do with the IGVs or the IGV actuator or the servo-valve)--only the IGV LVDT feedback which must be verified to confirm it is as close as possible to the actual IGV angle as measured with a machinist's protractor. And this must be done with properly verified servo current polarities AND using a physical measuring device (such as a machinist's protractor with a proper blade/ruler) and proper measuring techniques. Auto-calibrate has to be TOLD what the minimum and maximum mechanical stop positions are to work properly, and those have to be measured to be certain of their positions (they vary from every machine to every other machine, even from the factory, and especially after a MI has been done).

It would be helpful to a LOT of people if you could update this thread with the progress and the findings that resolved the perceived loss of performance (because we still don't know if this is happening at Base Load or Part Load).

 

Also, I'm presuming the unit(s) have conventional combustors, NOT DLN-I combustors. (You haven't told us a lot about the machines....)

 

Also, I'm presuming the unit(s) have conventional combustors, NOT DLN-I combustors. (You haven't told us a lot about the machines....)

The unit has conventional combustion system and not DLN machine