Control.com - Nerds in Control

Need help on the following. Suddenly the intervalve pressure in Frame V Mark V controlled GT has begun to fluctuate. Usually the values recorded were between 12.9 to 13.0 bar. Suddenly they have begun to change from 12.88 to 13.5 bar. The effect is quite visible in the analog gauge in the field. The visible movement of the pointer has made operators uneasy. The turbine was run in preselect and float modes. Same fluctuations were noticed. The frequency is constant. No maintenance was carried out in last few months. Lube oil and hydraulic oil pressure is constant. Oil level is normal. Second turbine at site shows no fluctuation in p2 pressure. Speed is constant too. Fluctuates between 5115 to 5118 RPM. Inlet gas pressure is steady at 250 PSI.

Please help.

As you are mentioning that the inlet gas pressure is steady, speed is not fluctuating and the hydraulic pressures are Ok, I feel, it is worthwhile, you must look at your Moog valve and the Hydraulic cylinder of Gas Ratio valve, operation of this may not be smooth and may be sticky, causing the fluctuations in P2 pressure.

As you are aware, Speed Ratio valve is controlled by a Cascade loop that consists of P2 pressure controller as the Primary (or Master) and the Position controller as the secondary. Type 77 servo controller does just that.

Assuming that no settings have been changed, the following factors can affect the presssure:

1. TNH ( in this case it is steady, right?)
2. The pressure reported by the three P2 pressure transmitters
3. LVDT feed backs

There are some utilities available in Mark V that allows you to collect data at the frame rate of 32 Hz. Some of the more useful ones are:

* VIEW2
* VIEW2T
* VIEWPV

For instance, if one of the three P2 pressure transmitter output fluctuates, the median value also will fluctuate. This in turn will cause P2 pressure to fluctuate.
How quickly you can locate the problem depends on the signals selected for viewing using the utilities mentioned above.

Last, but not the least, make sure that the following oil pressures are stable and at the rated values:

* Lube oil
* Control (Trip) oil
* Hydraulic oil

I had experience of undersized restriction orifice in the Control (Trip) oil header causing the system to "misbehave" during the peak summer months.

In essence, to locate a problem of this sort, you need to look into all possible sources of troubles - elctrical, hydraulic and mechanical.

Is this a SIMPLEX or a TMR Mark V?

Is there one P2 pressure transducer, or three?

Have you looked at the Prevote Data Display to see what the prevote values of P2 pressure and TNH are?

Is there a single LVDT measuring the Stop/Ratio Valve position, or are there two?

For units which operate for long periods of time at Base Load with a steady gas fuel supply pressure, there have been cases of worn Stop/Ratio Valve actuators and valve stems. The actuator sealing ring(s) wear the cylinder walls in one area, which can either cause excessive leakage of hydraulic fluid past the ring(s) or sticky valve actuation. The valve stems usually develop problems where they pass through the packing "gland", causing sticky operation.

Also, as has been noted recently in another case, the core of the LVDT can rub on the inside of the LVDT stator and cause intermittent problems over time, necessitating LVDT replacement.

As one respondent has noted, if the pressure fluctuations are not causing load fluctuations, then the problem is not too severe, but does warrant some kind of action. As another respondent has noted, you need to use one of the VIEW tools to gather some date (usually VIEW2) and analyze it.

No one has mentioned that the P2 pressure transducer (if there's a single transducer) or one of the P2 pressure transducers (if there are three P2 pressure transducers) may be operating erratically and causing a problem. I have seen, on units which run on gas fuels with high liquid "content" (compressor oil carry-over, entrained liquids, etc.) which develop a blockage of the sensing port and/or tubing lines to the P2 pressure transducer(s).

It's most likely the problem is NOT with the Mark V, but with the instrumentation (P2 pressure transducer(s), LVDT(s), hydraulic actuator, valve stem, sensing tube blockage, etc.) or, it may even be as simple as tightening the screws of the instrumentaion (P2 Pressure transducer(s), LVDT(s), servo-valve outputs) in the entire circuit (from the Mark V terminal boards, through intermediate junction boxes, all the way to the device(s)).

when we opened the front plate below the srv & gcv compartments in the running turbine, we saw several jets of oil from the hydraulic block connected to the power cylinder's bottom (the block on which the moog is strapped). the jets were coming out from plugs (1/2") of the hydraulic block of both gcv and srv. we replaced o-rings, which were cut, and the jets stopped. but when we restarted the turbine we found that the srv fluctuations have increased. infact the turbine started in three attempts. the first two attempts resulted in flame out as the P2 was varying from 0.5 to 4 bar. we allowed the turbine to reach fsnl. synchronised the generator. at lower loads the variation was severe. the turbine was loaded to base load. the mw variations increased to 1.8 mw. (before arresting the leakage it was about a mw). we re-examined the hydraulic assembly below the gcv & srv compts. it is now seen that oil is also jerkily coming out (not as jets, though) from the top vent hole (a 1/2"
opening) of gcv's power cylinder. we could not observe this leakage earlier bcos of other oil leaks.

we noted some parameters which i reproduce below :

a)fsg : value varies from 53.82 to 56.56%
b)fag : value varies from -2.78 to -4.28%
c)fsgr : value varies from 41.77 to 46.30%
d)fagr : value varies from 3.37 to -8.52%
e) fsrout : value varies from 54.11 to 56.67%
f) fpgrout : value is steady at 12.90 bar

with the above info i wish to ask following questions:

i) mvar fluctuations are high. between 9 to 14. is it bcos of 1.8 mw of load fluctuations?

ii)is it safe to run the turbine at base load in these conditions?

iii) will leakage from power cylinder of gcv have such an effect on P2 and cause srv to hunt?

iv) do we have to calibrate the gcv/srv if we replace the power cylinder?

v) any other advice you feel would help us solve this problem

thanks in advance

In general, during operation you shouldn't really see any oil streams during steady-state operation. The spurting you are describing can probably be attributed to the fluctuations you have noted in the values you can see on the operator interface. When the valve changes position (specifically when it closes) oil must be "relieved" from the actuator by the electro-hydraulic servo-valve; when the valve is commanded to open oil is supplied to the actuator by the servo-valve.

If the gas control valve is unstable, the intervalve pressure (P2) will be unstable and the stop-ratio valve will change position to attempt to maintain a stable P2 pressure. It can be difficult to determine if the gas control valve is causing the stop-ratio valve to fluctuate or vice versa. One of the easiest ways to do this is to "gag" FSR using the MANUAL FSR function *while operating at part load.* Reduce the Manual FSR value until it is less than FSRN (Speed Control FSR value) and the gas control valve *should* stabilize--if it doesn't, then something else is wrong. If the GCV does stabilize and the SRV stops fluctuating then the source of the problem is the GCV actuator, the servo-valve, or its reference/servo current, or it's position feedback. If the SRV doesn't stabilize with a steady GCV, then the problem is the SRV actuator, the servo-valve, or its reference/servo current, or it's position feedback or the P2 pressure feedback (since the regulator for the SRV is a pressure-control loop with position feedback).

VArs are a function of excitation--not fuel. Typically, the VArs may fluctuate if the unit is very lightly loaded (just one, two or three MW) and the power is also fluctuating. I would say you have an exciter regulator stability problem is you are experiencing VAr fluctuation. Power is directly proportional to fuel (torque); reactive current (VArs) is directly proportional to excitation.

FPRG is the P2 pressure reference; FPRGOUT is the signal name assignment that "connects" FPRG to the servo-valve output and it's regulator. If FPRG and FPRGOUT is steady, that's a good thing because FPRG is a function of speed feedback and if speed feedback were unstable then FPRG, and FPRGOUT, would be unstable.

Safe? It's not good for the turbine to be thermally cycling the hot gas path components by these fuel flow fluctuations. Also, the exhaust temperature would probably be close to the alarm/trip setpoints due to the fuel fluctuations; if they were to suddenly start fluctuating even more erratically, the unit will be tripped. But the question of "safety" is relative; are you talking about machine safety or personnel safety? It's not good for the unit, especially the turbine.

If there is excessive leakage from the GCV actuator and the servo-valve can't keep up with the required flow to maintain stable GCV position, then, yes, it will cause instability.

One has to wonder if some settings in the Speedtronic turbine control panel were "adjusted" during initial troubleshooting which need to be reset to their original values. This thread is an excellent example of how problems outside the turbine control panel can be attributed to the turbine control panel and because most people (supervisors and operators and plant managers, especially) always believe the source of every control problem is the control panel ("Just look at all the wiring and all those micro-chips--the problem *must* be in the panel!" is one refrain I've heard time and time and time again; people always suspect the thing they least understand). In this case, the turbine control system is probably trying very hard to overcome the problems with the actuators, and *appears* to be the source of the problem. The (un-)natural reaction is always to try changing this or that in the control system (doesn't involve any wrenches or getting dirty!) and when that doesn't work, then the wrenches come out (reluctantly!).

The are *two* times LVDT feedback has to be recalibrated: When the LVDTs are replaced, or when something is done to physically affect the stroke of the device (such as disassembling the valve for internal maintenance, or if the bar connecting the LVDTs to the valve stem is removed/replaced, or the IGVs have been disassembled for maintenance/repair, etc.). Replacing the servo-valve does nothing to change the physical stroke of the valve or IGVs; replacing the actuator of a Gas Control Valve (of a Young & Franklin-manufacturer combined SRV/GCV assembly) does nothing to change the physical stroke of the GCV. Only when something is done to affect the stroke of the valve or to affect the LVDTs which provide the indication of the position of the valve (or IGVs) does the LVDT feedback need to be re-calibrated.

Another thing which could cause GCV instability would be a problem with the LVDTs. Many times, the movable core of the LVDT will chafe and rub the inside of the armature of the LVDT over time and this will cause the LVDT feedback to be unstable at a particular point in the range. This usually happens on older LVDTs of units which operate at Base Load for long periods of time, but not always. The Mark V has two diagnostic functions, VERIFY POSITION and VERIFY CURRENT, which can be used to check LVDT feedback and servo-valve outputs to determine if there is a problem with the feedback or the servo. To make this work, one has to run AutoCalibrate on the device first, then run the diagnostic functions (they produce graphs which can be printed for analysis). Check the Mark V Maintenance Manual for info on these two functions, available in the AutoCalibrate function. The unit must be shut down to perform these tests.

Our problem is solved. thanks for all your support and help.

v did the following:
1) arranged for a power cylinder retrieved 20 yrs ago from an old discarded gcv assy. it wasnt a parker cylinder, though, but specs matched to the T.

2) stopped the turbine. the turbine while stopping tripped after reaching 3300 rpm. "tripped on flame out" signal.

3) v replaced the 'miller' make power cylinder. the leaky power cylinder had lots of dark black viscous oil layer on it's rod. on opening it we found the 'o' rings had folded inwards and some part worn. the black oil it seems was a result of 'o' ring wear. but v rnt sure.

4) v also replaced a 20fg sov. cos v found some oil leaks from 20fg too before taking the shutdown.

5) v started the turbine. the gcv opened at firing speed but srv did not.

6) v didnt know how to use autocalibrate function and manual fsr functions. so v had no option but to test while running.

7) after lots of troubleshooting v replaced the srv's moog valve. this time the valve opened.

8) checked the 'faulty' moog. the pencil filter was coated with the same black oil seen in the leaky power cylinder.

now the system is running. though v r not sure if it was the moog that was a problem or leaky power cylinder.

parameters after this exercise seem to have yeilded better results:
fsg : 57.39 to 58.77
fag : -2.56 to -4.79
fsgr : 46.24 to 45.20
fagr : -1.61 to -3.55
fsrout : 58.12 to 59.09
fagrout : 13.08 to 13.09
P2 : 12.96 to 13.17
mw : 19.1 to 19.7 @ base load
cdp : 7.70 to 7.73 bar

thanks for all ur help and support

There are many possible cause of light fluctuation occured in your system. But the important thing is , could the fluctuation in inter-valve pressure affect the generating load ? I mean were there any sudden change in load ? If not, you don't have to worry much about it. The inter-valve pressure always fluctuates to a little extent due to servo valve's action to maintain inter-valve pressure to a calculated value continuously changing with frequency change.

AhsanMM

We have also had interstage fuel pressure cycling, from day one, on all 9 MKV controlled 7FA peakers at this site. The swings run 6-12 PSI w/ ~ 2 second peak to peak freq. In all cases FPRGOUT is rock solid. Fuel gas supply is reduced from 600 PSI to 475 PSI at 5 metering and regulating stations depending on which units are lined up and running.

Typical troubleshooting techniques of locking the upstream pressure controllers in "Manual" and checking for AC noise on the SERVO drive signals have been performed. The servos have also been checked for smooth operation. The swings do not appear to affect MWs, MVARs, NOX or Combustion Dynamics. We have talked to several MK6 controlled 7FA users and they have indicated that P2 pressure operates very steady.

If this were a control valve controlled by a PID controller this issue would appear to be a tuning condition. This brings me to the question
of the MKV IO Configurator Servo Regulator Definition settings. I am very familiar with the function and purpose of "Current Bias", but lack the same knowledge when it comes to "Current Gain", "Integrator Convergence Gain" & "Position Reference Gain" and have not been able to get or find a good explaination. What affect do these parameters have on valve response? Feel
free to relate to PID controller parameters if there is a relationship at all. We do not want to make an experimental parameter change without
knowing what to expect.

Thanks
Greg Ponto
Lead Combustion Turbine Specialist
Elwood Operations
(815) 423-9883 x2232
Email: greg.ponto@dom.com