GE Frame7AE Exciter Control


Thread Starter


I've got a quick question about generator voltage control. First some background:

I operate a GE frame 7AE GTG coupled to a HRSG and associated STG in a COGEN plant. We generate power for use on-site but also for merchant power sales. Recently, we had to start a very large compressor motor. Our procedure is to have the utility tap up our transformer, raising the plant's 35kV bus voltage from 34.8kV to 35.6kV. We are also asked to move our GTG's MVARs from its usual amount (+11 MVAR) to -10MVAR. I didn't have to adjust this, as the MVARS dropped on their own as the bus voltage climbed.

When the compressor motor is started there is a momentary drop in bus voltage (-1500 volts) and the GTG MVARs according shoot up from -10 MVAR to +33 MVARs.

My question is, why does the GTG MVAR output drop when the bus voltage is raised by the utility? Shouldn't the exciter increase the current flow and try to maintain the original MVAR output? Our Alstom STG has a physical setpoint for generator terminal voltage (13.8kV) but I don't see anywhere on the Mark VIe to input a generator terminal voltage for the GTG... all I've got are raise and lower buttons that seem to bump the VARS up or down by about 2.5MVAR per click.

If anyone can shed some light on the excitation control for the GTG it would be greatly appreciated.


Thank you for the background; it's most helpful in helping describe what's happening and why.

You (and most power generator operators) don't know it, but when you are clicking on RAISE- or LOWER VOLTAGE (and the exciter is in Automatic (AC) Regulator--which it is most of the time!) you are changing the generator terminal voltage setpoint. You are correct; the value of the setpoint is not displayed anywhere (that I've ever seen), but that's because when the operator is clicking on RAISE- or LOWER VOLTAGE he (or she) is almost always looking at the VAr (or MVAr) meter or the Power Factor meter.

The difference between the generator terminal voltage and the grid voltage determines the VAr flow and the Power Factor. When the amount of excitation being supplied to the generator is exactly equal to that required to make the generator terminal voltage equal to the grid voltage (at the generator terminals) then the VAr (or MVAr) meter will read 0, and the Power Factor meter will read 1.0 (Unity).

When something happens that causes the generator terminal voltage to differ from the grid voltage (at the generator terminals), then the reactive current will increase (VArs; MVArs) and the Power Factor will decrease (from 1.0). So, if the operator increases the generator terminal voltage setpoint (that unseen setpoint...!) from the point at which the VAr (or MVAr) meter reads 0, and the Power Factor meter reads 1.0, what will happen is that the generator terminal voltage will increase (but often the amount of increase is very difficult to see on any meter or display (we're talking tens or hundreds out of about 13,800 volts in some cases--but tens or hundreds of volts out of tens of thousands of volts, anyway)), and VArs (or MVArs) will increase from 0 in the Lagging direction, and the Power Factor meter will decrease from 1.0 in the Lagging direction.

On the other hand, if, the VAr (or MVAr) meter is reading 0 and the Power Factor meter is reading 1.0, and the grid voltage decreases with respect to the generator terminal voltage, then the VAr (or MVAr) meter will increase in the Lagging direction, and the Power Factor meter will decrease in the Lagging direction. So, during the day, as grid voltage changes (and it usually does, especially when the grid load is high) the VAr (or MVAr) meter and the Power Factor meter will also change--and the operator has to manually adjust excitation (by changing that unseen generator terminal voltage setpoint) to maintain whatever VAr (or MVAr) or Power Factor the plant operations supervisor wants to maintain.

When you tell the utility to change the transformer tap--the value of the grid at the generator terminals changes, BUT the generator terminal voltage setpoint doesn't change. If the value of the grid voltage increases with respect to the generator terminal voltage (which is the same thing as decreasing the generator terminal voltage with respect to the grid voltage) then VArs (or MVArs) will flow into the generator in the Leading direction.

I am presuming, by the way, that the generators are NOT being operated with VAr- or Power Factor Control active--that they are essentially in what's being called "voltage control mode." If VAr- or Power Factor Control mode was active (and only one of the two modes can be active at any time), then the control system would adjust the (unseen) generator terminal voltage setpoint to maintain either the desired VAr- (MVAr) or Power Factor setpoint (depending on which was selected and active).

Many people might say, "But, when the generator is synchronized to the grid the generator terminal voltage is always equal to the grid voltage." And, they would be technically correct. When the amount of excitation being applied to the generator rotor is exactly equal to the amount required to make the generator terminal voltage equal to the grid voltage then there is no reactive current flowing between the generator and the grid. BUT, if the operator increases that unseen generator terminal voltage setpoint by clicking on RAISE VOLTAGE the amount of excitation being applied to the generator rotor increases to make the generator terminal voltage--and the grid voltage--increase. If the grid is "strong" where the generator is located, you might not see a very large generator terminal voltage/grid voltage increase; and if the grid is "weaker" where the generator is located, you might see a larger increase in generator terminal voltage/grid voltage. (A high resolution voltmeter would also help--again, we're talking about hundreds out of tens of thousands of volts). This is technically called "boosting" the grid--you are trying to (and in some cases more than others) increase the grid voltage with your generator(s). The difference between how much the generator can actually increase the grid voltage and how much additional excitation is applied causes Lagging reactive current to flow in the generator stator windings--and this is usually considered to be positive VArs (MVArs) out of the generator and on to the grid.

The opposite happens when the operator decreases the excitation to try to "buck" the grid voltage--reduce it. Only in this case, Leading reactive current flows in the generator stator windings--and that is usually considered to be negative VArs (MVArs) into the generator from the grid.

But, really, any time the relationship between grid voltage and generator terminal voltage changes--be it because the operator or the control system changes the excitation, OR because something on the grid causes a difference (such as in your case, changing the transformer tap(s)) there will be a difference in reactive current flow (and Power Factor). But, when it's an "external" change (from the grid) and not a change of the generator's unseen terminal voltage setpoint, there will still be a change in VArs (MVArs) and Power Factor.

It's all relative, you see. (I hope it's clearer.)

By the way, did you know that when you are clicking on RAISE- or LOWER SPEED/LOAD you are actually changing the turbine speed reference--and trying to make the turbine speed increase or decrease? BUT, when it is synchronized to a grid the speed can't change--and the difference in fuel flow--that would cause the turbine speed to change if it were not synchronized to the grid--becomes amperes and causes the power out of the generator to increase or decrease? For the GT the Turbine Speed Reference is called TNR, and you can observe it changing when you click on RAISE- or LOWER SPEED/LOAD. But, to my knowledge, you don't normally see a generator terminal voltage setpoint, and the generator terminal voltage you do see is a reflection of the grid voltage when the unit is synchronized to the grid.

Isn't this fun?

By the way, you did not mention what happened to the ST generator voltage when the tap was changed (and I'm presuming the ST and the GT are connected through the same tap changer--but that's a SWAG, and if I'm wrong, please correct me). If the ST is connected to the grid through its own transformer and NOT through the same tap changer as the GT--or the ST transformer has its own tap changer and the tap wasn't changed on the ST transformer tap changer when it was changed for the GT generator output--then it's conceivable the ST generator voltage (and VAr (MVAr) value) didn't change.

Thanks for the clarification!


Wow, first of all thank you very much for such a detailed reply CSA, it is much appreciated!

I've experimented since with the RAISE/LOWER MVAR buttons and, as you say, I can see my adjustments changing both the MVARs flowing out of the machine as well as changing the 35kV bus voltage (On click equaled 2.5VARs which equated to ~120V change on the bus in my case).

As for the ST, it is connected to separate Bus with its own utility breaker. The two buses can be joined by a tie-breaker, but we do not do this when both generators and utility transformers are in use. (something about reactive current flowing in loop). As such, the ST terminal voltage and associate bus voltage were stable during this operation.

A further question, I can see from my GT's MkVIe where the operating curves for my machine sit. It appears to be happy between the regions of -40MVARs (leading) and +80MVARs (lagging). We are always expected to operate with a slightly lagging power factor, supplying VARs to the grid. As I mentioned, during the voltage swing when we started the large motor my VARs jumped wildly (from -11 to +33) momentarily. Is there a protection relay that will trip my machine if it were to be operated outside the generator curves provided? I understand these are related to machine heating limits. I have found alarm codes for under/over excitation but there is no literature to specify at which point (if any) an excitation exceedance may trip a generator protection relay?

Also, you assumed correctly that our machine is being operated in VOLTAGE CONTROL. I see that there is an option for PF control, however I've been told that we do not operate this way (although no one can tell me why that might be...)

Thanks again!

The excitation system, sometimes called the exciter regulator, sometimes called the AVR, has upper and lower limits of excitation)usually the amount of current that can be applied to the generator rotor field windings). Too much current, called over excitation, causes excessive heating of the generator rotor field windings, and too little excitation current can result in excessive heating of the stator winding end-turns or worse, a condition referred to as "slipping a pole" which can cause great physical damage to the generator, the coupling between the turbine and generator and even the turbine rotor.

If the excitation exceeds the limits programmed into the exciter (probably an EX2100e if the turbine control system is a Mark VIe) then there will be Process Alarms and the current will be limited to protect the machine. These settings can usually be found using ToolboxST with the exciter device (as it's called in ToolboxST parlance).

Yes, there are also generator protection relay(s) to sense if conditions are potentially dangerous that can also alarm and trip the turbine. The settings for these protections should be found in the Generator Protection Panel elementary (schematic) drawing.

There should be a display on the HMI which lists most everything that can trip the turbine (the Trip Display). That's a good place to start understanding what can Trip the turbine. But, some Trip conditions, like 86G, can be caused by multiple different conditions (some of which may be individually alarmed by the Mark VIe and most of which are not). So, it's best to create a description for each trip conditions that accurately describes the setpoint for each condition that can trip the turbine. Not the easiest thing to do, but the sites which have done so have excellent reliability records and good operators who know how to respond to ("manage") alarms and minimize trips.