Gas Splitter Valve Split and Stroke Out of Mark V?

L

Thread Starter

lesswood

Our power plant consists out of 3 Heavy duty GE frame 9E gas turbines equipped with LEC-III (PSM) combustion systems and Mark V. For my internship research I'm investigating possibilities to install an autotune system on the GT's or to make a manual tuning model with only installing an Combustion dynamics monitoring system (CDMS).

Now the gas fuel split set point is fixed on 83,25% split and is only tuned after an outage. However, the gas splitter valve command stroke is around the 78,8%. This is the actual GSV command stroke. My question is what's the difference between gas fuel split set point from 83,25% and the command stroke of 78,8%? in other words what's happening in the Mark V to come to the stroke of 78,8% (Com ref. temp??)?

Hope to get an answer,
 
lesswood,

I'm not familiar with the PSM LEC-III DLN combustion system, but in the GE system flow through the gas splitter valve is not proportional to stroke, so a method is used to convert flow percentage to the valve position that will result in the required flow percentage. Some control schemes refer to this as valve linearization, I believe. Sounds like something like this is being done with the valve for this system, also. I'm sure if you look through the CSP in the Mark V you will find some block (usually and ALIP (Analog Linear Interpolation) used to achieve this function.

Hope this helps!

 
CSA,

Very helpful, thank you.

I found the ALIP and relevant linearization block positions. In that way I calculated the gas splitter valve command stroke which is fixed and matches identically with the Mark V indication. So just to be sure if I've understood, the split position (83,5%) is the primary/secondary flow rate and the command stroke is the valve position. The Mark V is also indicating a feedback position (from the lvdt) of this command stroke. This is fluctuating around a range of 0,3% and seems to be independent from IGV position an combustion reference temperature. So my question is, why is this stroke feedback fluctuating constantly while the command stroke is fixed? Maybe this is a normal procedure, but I'm not familiar with it.

Many thanks for your support!!
 
lesswood,

You're welcome; glad to be of help.

Feedback signals on running heavy duty gas turbines are never rock-solid, as many people think they should be. There are all kinds of reasons for this, but they're just never rock-solid. (I'm speaking of analog values, like position feedback, thermocouple readings, RTD readings, pressure transmitter values, etc.)

The important thing is that the valve itself is not moving, or is only moving a VERY small amount. 0.3% isn't much, but in the DLN world if the splitter valve is moving and not stable that can cause emissions problems. Instability can be caused by improper servo regulator gains, electrical noise on the servo output signals, hydraulic oil pressure fluctuations, servo problems, etc. But. 0.3% is really a very small fluctuation and could just be due to noise and poor signal cable routing/isolation and poor shield drain wire grounding practices.

Again--you should probably verify that the valve itself is stable, and that means physically observing the valve stem and indicator (there's usually some kind of indicator/pointer on the valve stem). (Don't worry; valves don't bite. It's probably noisy, and may even be hot in the compartment where the GSV is located; but it's good to get out and get a little dirty and see things in operation.)

But, have a look at any other analog signal while the unit is running, and even when it's shut down--they're almost never rock solid, stable. Some of that has to do with "rounding" of signal values, some with cabling/wiring practices, and some with the process/parameter that's being monitored. But, the GSV itself should actually be very stable, or moving every so slightly as to be virtually imperceptible to the human eye.

My GE-design heavy duty gas turbine valve "stability" test is to take a medium-sized coin (as in monetary unit) with smooth edges and attempt to balance the coin on its edge on some flat surface attached to the valve stem while the valve is open to some position. If the coin will remain on its edge and the movement does not cause the coin to fall off its edge then in my professional opinion the valve is stable and suitable for operation. If the coin can't be balanced on its edge and remain on its edge while the valve is at some steady reference position, then something is amiss and should be investigated.

Ain't technology great?
 
Top