Frame 9E-Mark vie Servo Valve

Y

Thread Starter

Yasiin

Hello,

We have Frame 9E-Mark Vie system GT with Natural gas & Syngas fired. At present we don't have syngas, so GT running on NG. When turbine is running, servo currents for R, S, T coils found to be -2 to -3 % ma (ideally should be -2.67%) which is understood. but when GT was in turning gear (110 RPM), servo current across VSR-1(NG) found to be 80% on EGD page & 8.8 ma in marshalling. In VSRVSR-11 (Syngas SRV) it was 100% in EGD & 10.98 ma in marshaling. Please help me to understand this, as i can't understand why the servo current value, when valves are closed, are 100% & 80%.

Best Regards,
Yasin.
 
Yasin,

The answer to your question is the servo-valve's reference signal. In general, in the past GE heavy duty gas turbine control systems have had a negative reference for fuel control valves when shut down. For example, many machines have a negative reference of from -25% to -40%, which drives the servo-valve output to high positive value to try to achieve a negative feedback. This was done to try to ensure the control valve would not be slightly open at a zero reference if the calibration if the control valve feedback was incorrect. Positive servo-valve current acts to shut off the flow of fuel; negative servo-valve current acts to increase the flow of fuel. And "zero current" tries to maintain the present fuel flow (with the null bias current "added" to it). When the actual feedback equals the reference the error between the two is zero, but to keep the control valve maintaining the actual flow equal to the reference the null bias current is required to overcome the servo-valve's failsafe spring tension, so the regulator output at zero error is slightly negative (which is the polarity required to increase flow).

When the reference is-25% or -40% and the feedback is zero, the regulator will put out more positive current trying to get the feedback to match the reference.

Now, it may be that GE is doing something different with the reference when the unit is at zero speed versus when it's on Cooldown than they've done in the past. If GE-Belfort was involved in the programming/configuration of the Mark VIe you can be pretty much assured they have a "better idea" than the American designers.

Hope this helps! Please write back to let us know what you discover.
 
Thanks CSA for reply.

Regarding 80% & 100% current during servo valves closed, as per my understanding, when reference is given to close servo valves, the mechanism inside servos(armature, flapper, springs, spool) would allow more hydraulic to pass,that will move the actuator to close & as soon as feedback equals reference it would shut off the hydraulic to move actuator & would send minimum(null bias ) current.

So is it not true when valve is closed, as as soon as reference equals feedback, servo current should again come to -2 to -3%, then why it is 80%/100% in this case as closed feedback already received so why it sends continuous 80%/100% current & allows hydraulic to pass.

Pls correct me if I might have misunderstood.

Thanks & regards,
Yasin.
 
Yasin,

Did you look at the references in the application code for VSR-1 & VSR-11?

I'd bet large gazoonies (whatever currency you would prefer) the references for both these valves when the unit is not running is either -25% or -40%. And the feedback (from the LVDTs) is approximately 0%, +/- approximately 3%.

That would mean the error between the reference and the feedback is very high, and the Mark VIe would therefore be putting out a high (magnitude) value of servo current to try to reduce the error to zero. Which can't happen because the valves can't physically close below approximately 0% (presuming the LVDT feedback was calibrated properly).

As was written before, the GE heavy duty gas turbine control philosophy is to put a negative reference to the servo-operated devices (most of them, anyway) when the unit is not running to try to ensure that even if there is hydraulic pressure the device will be trying to shut off or reduce the flow of fuel or air to zero or a minimum (in the case of the IGVs).

By working backwards from the servo-valve outputs of the Mark VIe you can get to the block (or macro, as the case may be) where the valve references are derived (written to, in ToolboxST parlance). You will likely find a negative reference when the unit is not running, as per GE heavy duty gas turbine control philosophy.

It sounds as though there was an assumption that the reference would be zero, and that because the feedback should be zero the output should be close to null bias current. That's a reasonable assumption, but it's incorrect--and easily proven (or disproven) by simply reviewing the application code in ToolboxST.

Hope this helps!

 
Top