We have 02 nos. of Frame-6 machines controlled by Mark-IV systems, installed in 1988

1st incident: we observed that the machine tripped (during start up & after detecting 02 nos. of flames) on "SRV Servo Trouble" because of servo current goes below -6.0 m amp after Firing command at firing speed.

We carried out SRV calibration (Done OK) and GCV Stroke trial (Found O.K).

2nd incident:
During 2nd start up we have observed that for a Gas turbine, the FSR value and FSG value are not matching. And probably due to this, during recent start up it took more then 45 min. from zero speed to FSNL. We found that the FSR value reached to ACCLE_FSR (@27.3%) but the FSG was still at 17.8%.

I doubt the 1st incident also happened due to this mismatch of FSR & FSG.

I would like to discuss the matter to know that how this happened & solution for the same.

 

You say you have two Frame 6s, but you didn't say which unit(s) these problems were occurring on. Did one problem occur on one unit, and the other problem occur on the other unit? Or did both problems occur on the same unit?

When did these problems start? After maintenance outage(s)? After servo replacements? After servo calibrations?

How long has it been since the L.O. was tested and/or replaced? Has it ever been "polished" (cleaned by some method such as a centrifugal separator or other highly effective "filtration" method)? If so, how recently?

What were the L.O. temperatures (tank and header) during each of these events?

Have you verified that the polarity of the servo current being applied to each coil of the servo-valve is correct?

When the unit is running at steady state conditions, producing power, what are the three values of servo current for both the servos on the unit(s) which experienced the problems?

I don't understand why the SRV LVDT calibration was necessary; did you find the LVDT feedback to be out of calibration. In other words, when you measured the physical position of the valve was the indicated position significantly different?

The SRV is a pressure loop, not a position loop. The SRV will be moved to whatever position is necessary to make the P2 pressure feedback equal to the P2 pressure reference, <b>even if the feedback is grossly out of calibration.</b>

In other words, if the SRV LVDT feedback indicates 20% stroke but it's really at 30% position (physically measured position), BUT the P2 pressure equals the P2 pressure reference when the LVDTs say the valve is at 20% stroke when it's really at 30% stroke, the Speedtronic doesn't care. It is going to position the SRV (regardless of what the indicated- vs. actual position is, or even the indicated- vs. reference positions are) to make the P2 pressure equal to the P2 pressure reference.

If, on the other hand, the SRV is not moving to the make the P2 pressure equal to the P2 pressure reference, that's either a servo problem or an actuator problem or a valve problem. But it's not an LVDT feedback problem.

You never told us if the actual P2 pressure was equal to the P2 pressure reference.

The GCV should move to the same indicated position (from the LVDT feedback) as FSR when running on gas fuel. If it's not, then something is not allowing the GCV to be moved, and the GCV servo current should be reflective of that (a large negative value would be trying to open the valve, since for all GE servos a negative current <b>*at the servo*</b> will increase the flow of fuel.

<b>NOTE: The Mark IV kind of plays games with the servo current; it indicates positive when it's actually putting out negative current, and vice versa. This can be seen in the Mark IV Speedtronic elementary on the servo output circuits; there is a "roll" or "twist" on the output traces. So, be careful when stating servo current polarities.</b>

What was the exhaust temperature (TTXM) during this slow acceleration event? What was the gas fuel supply pressure? What was the P2 pressure during this slow acceleration event? Was AUTO or REMOTE mode selected during this slow acceleration event, or was FIRE mode selected?

What Diagnostic Alarms were annunciated during the second incident?

But, I would say if the Speedtronic were trying to make the unit accelerate faster (FSR at 27.3%) and the GCV wasn't moving to the required position, then something is preventing the GCV from moving.

You didn't say how the unit responded after reaching FSNL? Did it operate correctly? Was the unit synchronized and loaded, and if so, to what load? Were they any problems during synchronization and loading?

Are there any IS barriers installed between the Mark IV and the servos and/or LVDTs? Have they been checked to make sure they are working correctly?

What were the servo currents, for each of the three processors, during the second incident?

Was MANUAL FSR active during the second incident?

There's a lot of information that's missing to be able to make a good analysis. And, unless you have great records and manually recorded lots of information (or made a lot of screen prints), we'll probably never know all of the required information to be able to make an educated guess.

As for calibrating LVDT feedback, did you physically measure the actual SRV position prior to, during, or after calibrating the SRV LVDT feedback to see if the indicated position was close to the actual position? What is the difference between indicated position and actual position and reference position?

When you stroked the GCV did you physically measure valve position to compare it to indicated position and reference position?

If you want to continue to discuss this thread, you will need to provide all of the data you have in response to all of the questions above, and provide answers to <b>all</b> of the questions, even if it means saying, "We don't have the data," or, "We don't understand the question." No picking and choosing which questions you think are relevant to provide answers to; it's all or none.

 

Dear CSA

Clarifications to your queries

Both Problems occurred in same machine & during the same start up activity. We faced the problem, when we were re-starting the machine after a tripping of machine on exhaust Over Temperature (We replaced the servo for IGV, along with Pall filter for IGV Hyd Skid & successful Calibration of IGV, we also replaced the Pall filter for SRV/GCV).

For the cleaning of Oil, we are using ELC on regular bases. During the startup the lube oil temperature was normal i.e. 44 deg C. We did not replace the servo valve for the SRV in near past.

STATUS___C-KING__F-NG___W-U____ACC
PN_PR_____1973___1938___1934___1942_____RPM
TNH______21.03__22.56__22.97__25.15____% SPD
FSR______00.00__18.7___14.4___19.7_____%
FSG______0.31___11.89___7.72__12.28____% STR
FSGR_____ NA ___00.93___00.68__1.10____% STR
FPRG1____-0.18___02.56___02.65__3.08____BAR
FPG1_____-0.16___02.59___02.71__3.09____BAR
TTXM________67______89___117____218_____DEG C

Regarding the behavior of the machine after synchronization, the difference between FSR & FSG exists around 9-12% during part/base load. The governing system responds to the FSRT/FSRN properly.

With regards