Protection Considerations during starting of Gas Turbine Generator with LCI

[rockman7892]

I was looking for some good references for protection considerations during the startup of a Gas Turbine generator when the generator is being driven as a motor by an LCI drive.

Since the generator will in essence be a synchronous motor during the starting phase I'm thinking some of the principles of synchronous motor protection apply during the starting process.

From some of the information I have seen, some consideration must be given to the differential protection since the LCI may fall within the differential zone and this may cause mis-operation of the differential element. Is it recommended to have some protection elements that have two settings groups for both starting and normal conditions?

It also appears that some of the most critical settings during startup are V/Hz, Overexcitation, and Overcurrent, along with phase and ground fault protection.

I'm not really familiar with LCI's and their controllers but I'm assuming that the units have a controller or protective device that provides certain protection elements of the LCI and generator during the starting process. I would also think that the excitation controller would have similar protection functions.

Thanks for the help!

 

[davidbeach]

The LCI drive should provide the bulk of the protection. The generator differential should not be impacted. As far as I know it starts as an induction motor on the damper bars and the field isn't energized until near synchronous speed.

 

[ScottyUK]

I'm fairly sure GE's LCI 2100 has the field energised from startup. Rotation is a bit 'lumpy' until it picks up speed.

 

[rockman7892]

davidbeach

You bring up an interesting point with the damper bars. I was not familiar with this starting method but investigating further I see it is a common practice to start a synchronous motor as an induction motor until it reaches speed. So with that being the case do we simply treat the motor as an induction motor in terms of protection with the LCI drive protection/controller? It appears that most drive controllers have the necessary protection elements for starting the generator as a motor.

Are there any settings in the excitation controller that need to be considered for starting, or is the excitation typically out of the picture until the motor/gen is up to speed.

If the LCI is in the middle of the differential zone, would this not impact the differential?

 

[pwrengrds]

The differential is usually turned off, even if it's in the zone the differential relays don't do well with harmonics. What I have seen is overcurrent protection phase and ground protection during LCI operation.

 

[ScottyUK]

From

http://www.ge-mcs.com/download/control-solutions/GEI-100792A.pdf.

My highlights.

"During startup, the turbine control supplies the run command and speed setpoint signals to
the LS2100e control. The static starter operates the power converter in closed-loop mode
to deliver variable frequency current to the generator stator. [highlight #FCE94F]By controlling generator field
voltage and stator current[/highlight], the static starter is able to adjust the torque produced by the
generator and control the acceleration and speed of the turbine generator set.
Note
[highlight #FCE94F]During startup, the field voltage setpoint is provided by the LS2100e control[/highlight] to
regulate the generator stator terminal voltage."

You need to understand how the specific type of drive in question operates before you can protect it. Regardless of that, most drives of this power level have fairly sophisticated internal protection of their own.

 

[MatthewDB]

LCI stands for load commutated inverter. It uses a 6 SCR bridge rectifier back to back with a 6 SCR inverter. The back EMF of the motor commutates the inverter SCRs. An LCI draws VARs from the motor.

Given that an induction motor does not generate VARs, it is impossible for a LCI to drive an induction motor.

For startup before the motor generates enough back EMF to commutate the inverter, the output SCRs are commutated by forcing the DC link current to zero with the rectifier SCRs. When this happens, the torque is pulsing to zero 6 times per electrical rotation during commutation so the motor runs real rough.