Dear All,
Sorry for the late reply.

We had the opportunity to change the servos; since we suspect the Servo of SRV and Splitter we had changed both the servos with the old servos which was used earlier (working fine). Before replacing the servo, the polarity check done on all the servos & stroke check performed on all the valves and found good. After replacing with the old servos, the polarity check and stroke check performed and found good. (No other changes were made on unit software or IOCFG screen etc,)

The unit was started successfully, synchronized with grid and started to load. We could observe that there was no MW hunting until we select Premix Steady State Mode (PMSS), the unit loaded very smoothly as like earlier startups and the unit started to swing soon after the selection of PMSS. The same was verified again by taking the unit back to Piloted Premix mode and observed that the unit was steady with no hunting, again selected PMSS and its started hunting. As this unit is equipped withDLN2.0 combustion system, only Quaternary valve will come in to line and primary valve starts purging. The healthiness of Quaternary valve is good in trends during PMSS, i.e. there is no fluctuation in its position as comparing with other valves (its having ~0.2%/sec jump but the frequency of jump is slightly more comparing with earlier trends on base load). Before selecting PMSS, none of the valves have fluctuation, after the selection of PMSS, except Quat all the valves started to fluctuate.

Again we suspected the DLN Constants (FXKPTSP_1, FXKPTSP_2 & FXKPTF_4-->PMSS Split constants for splitter valve) and looked in to it and found that all these constants will play its role only when the unit is selected to PMSS and all these constants are from single interpolator block. The constants FXKPTSP_1, FXKPTSP_2 will be in control only during the time of loading and will not be in line at Base load and other constant FXKPTF_4 will give the input in terms of TTRF1 for the calculation of split for Splitter valve, i.e., when TTRF1 goes above 2330.6 deg F the split will be 83% , (earlier 2350 deg F). We suspected this constant and decided to put back the original constant (2350 deg F). To test this we had taken the unit out of PMSS and changed the constant and again put back the unit to PMSS and found still hunting.

The above observation had taken to OEM and they suspected the servos of SRV, Splitter and Quat and suggested to recalibrate all the valves. Not sure how far it is going to help, since the servos which are in service are the old servos and worked fine earlier with same null bias and gain.

If any one have any suggestion and that can be tested\experimented at online, we can execute the same to avoid this hunting.
Please let me know if you need more data\inputs.

Thanks for helping,
A2A.

 

A2A,

This problem started after a major inspection. You've returned at least a couple of the original servo-valves to service, verifying that they work correctly.

I don't have access to a CSP or .m6b file for a DLN 2.0 system, so I can't review the block.

However, the system did work before the MI, right?

So, one has to concentrate on what changed during the MI that is causing the issue?

Was DLN tuning done after the MI? If so, precisely which Control Constants were changed <b>during</b> the tuning and which Control Constants were changed <b>as a result</b> of the tuning? In other words, sometimes come Control Constants are used during tuning to affect some changes/adjustments and those Control Constants <i>are supposed</i> to be returned to normal values but don't always (shocking, I know, but it's true!), and which Control Constants were actually changed from the pre-tuning values that were left in the post-tuning values to achieve the desired emissions.

Again, I go back to the fact that everything worked before the outage (it did, right?) and after the outage and subsequent DLN tuning (the unit was tuned (DLN tuning) after the outage, right?) the problem started. The servos have been replaced, and re-replaced (though not ALL of the servos have been re-replaced, right?).

I would be looking at trending using VIEW2 all of the inputs to the various servo-valve reference blocks and the outputs from the servo-valve reference blocks to try to understand what starts the "swinging".

Are you sure the IGVs are stable? And the IBH control valve?

What is grid frequency doing when this happens.?.?.? (Meaning, what is turbine speed doing when this problem occurs???)

 

Dear All,

Yes, it was working good before the outage.

I have compared the constants before and after the outage (through constant check window) and only the above said DLN Constants were changed. No other constants were changed.

Yes, we have replaced only SRV and Splitter servos, rest of the servos are new one.

IGV and IBH are stable, no issues.

No grid fluctuations, it is good.

The problem starting only when we select Premix steady state, if we come out of Premix steady state, the unit is stable and no hunting observed.

Thanks,
A2A.

 

A2A,

Again, I'd be looking at using VIEW2 to try to ascertain what is initiating the fluctuations. One of the problems with the Mark V is that servo current only gets "reported" or "recorded" at a 4- or 8-hz rate, even though it's really changing at a 128 Hz rate. So, it's difficult to see exactly when the servo current starts changing or even the min and max values of fluctuation. But, it should be possible to see if any of the valve reference outputs from the block in the CSP to the TCQA card start "oscillating" first.

You also mentioned an auto unload sequencing change implemented by GE. I wonder if some oscillation in gas fuel supply pressure when going to PMSS. GE are FAMOUS for not putting hysteresis and deadband in comparator functions (just look at the Wheelspace Temp alarm comparators, or any GE Steam Turbine Mark V or Mark VI sequencing for metal/bowl temperature alarms--NO hysteresis/deadband whatsoever, and the nuisance dithering/toggling of alarms is a real PITA (Pain In The Arse)!).

I'm re-reading all of the posts to this thread and trying to think of all the changes effected during the maintenance outage. You need to do the same, and use VIEW2 to try to capture the data.

One thing you probably can't capture is hydraulic pressure fluctuations. Are you sure the Accumulator(s) is(are) charged and the valves are properly set? This is something that affects a LOT of units because people don't properly connect the gauge when checking or charging, and the valves are not usually labeled at all (another GE shortcoming on their heavy duty gas turbines--lack of valve identification).

Please keep us informed about the progress of this troubleshooting.