Harmonic Analysis

[D!NNy]

I have a case study for the harmonic analysis some one please help me on this.

We have 1.5MVA step down t/f from 12.47kV to 0.48kV feeding the 2000A MCC with 80% non linear loads connected (VFD's). So We considered point of common coupling as MCC main bus. Based on what i have studied in the textbooks in the school harmonics are introduced by all switching devices(Power electronics). Why the harmonics are travelling in either way from the VFD to line side and load side as well? Because the switching is done at the VFD i thought the harmonics will be travelling downstream to the motor. in fact they even travel towards utility effecting transformer and power grid?


Thanks

 

[petersonra]

The harmonics are generated because the load only exists during part of the cycle when the DC bus is being recharged from the line so the current demand from the feed is choppy.

Downstream of the VFD it is also not a sine wave because a VFD actually uses PWM to generate a feed to the motor that simulates a sine wave but is actually a series of DC pulses.

 

[Besoeker]

I have a case study for the harmonic analysis some one please help me on this.

We have 1.5MVA step down t/f from 12.47kV to 0.48kV feeding the 2000A MCC with 80% non linear loads connected (VFD's). So We considered point of common coupling as MCC main bus. Based on what i have studied in the textbooks in the school harmonics are introduced by all switching devices(Power electronics). Why the harmonics are travelling in either way from the VFD to line side and load side as well? Because the switching is done at the VFD i thought the harmonics will be travelling downstream to the motor. in fact they even travel towards utility effecting transformer and power grid?


Thanks

I'll try.
The VFD power circuit comprises three elements.

• An input rectifier that converts the AC supply to DC.
• The DC link.
• An IGBT output that converts DC to a variable frequency and voltage AC usually using pulse width modulation (PWM).

For sure, the voltage being sent to the motor isn't sinusoidal - it's a series of square wave pulses. But the inductance of the motor smooths the current to make it a fairly good sinewave.
This switching is effectively decoupled from the line side by the DC link which has a fairly big capacitor bank. So, forget that in terms of the supply side.

So what about the line side connected to your PCC?
That's where the harmonics issue lies. The input rectifier. That takes non-linear current. For most variable frequency three phase input drives, this is a plain six pulse diode bridge.
There are two current peaks for each half cycle of the supply so it presents a distinctly non-sinusoidal load on the supply. That's where the harmonics come from.

 

[D!NNy]

The harmonics are generated because the load only exists during part of the cycle when the DC bus is being recharged from the line so the current demand from the feed is choppy.

I understand that current and voltage of the input wave forms are partly used but how would that effect on the harmonics?

Downstream of the VFD it is also not a sine wave because a VFD actually uses PWM to generate a feed to the motor that simulates a sine wave but is actually a series of DC pulses.

Can you explain this in little more detail?

 

[D!NNy]

I'll try.
The VFD power circuit comprises three elements.

• An input rectifier that converts the AC supply to DC.
• The DC link.
• An IGBT output that converts DC to a variable frequency and voltage AC usually using pulse width modulation (PWM).

For sure, the voltage being sent to the motor isn't sinusoidal - it's a series of square wave pulses. But the inductance of the motor smooths the current to make it a fairly good sinewave.
This switching is effectively decoupled from the line side by the DC link which has a fairly big capacitor bank. So, forget that in terms of the supply side.

Do this Capacitor bank filter the harmonics going towards line side?

So what about the line side connected to your PCC?
That's where the harmonics issue lies. The input rectifier. That takes non-linear current. For most variable frequency three phase input drives, this is a plain six pulse diode bridge.
There are two current peaks for each half cycle of the supply so it presents a distinctly non-sinusoidal load on the supply. That's where the harmonics come from. Here the two current peaks are after rectified? If yes there must be only positive current drawn from the supply side?

This VFD at the point of starting the motor, voltage will be gradually increased until the motor reached to full load mean while current from the supply is charged partly to full. correct me if i am wrong here

Thank you for spending your time on this.

 

[Besoeker]

Thank you for spending your time on this.

Do this Capacitor bank filter the harmonics going towards line side?

First, let's look at what a harmonic is in the electrical context.
For a repetitive non-sinusoidal waveform, you have a fundamental frequency component, often 50Hz or 60Hz input side of the VSD and variable on the output side.
Harmonics are multiples of this fundamental frequency.
The pulse width modulation (PWM) on the output side is a switching frequency, typically in the several kHz range and not strictly related to the fundamental output frequency so not a harmonic of it.
But yes, the capacitor on the DC link provides a low impedance source for the load side switching frequency components so they don't appear on the supply side.

Back to supply side harmonics and what they are.
Take a simple square wave like this:

It is just that. A simple square wave. Now, sometimes you may want calculate how it will interact with other parts of say, a network.
One way of doing this is to mathematically describe the waveform as a combination of sine wave - sine waves are easier to work with than waveforms with discontinuities like the square wave.
The square wave can be represented mathematically as an infinite series of sine waves at multiples of the fundamental frequency. In this case, all the odd multiples - three times, five times etc.
But note that this is a mathematical tool to aid analysis. A useful tool it is. But a tool nevertheless. The harmonics don't create the waveform.

For the three phase variable frequency drive, the input circuit is usually like this:

The components that follow that have a significant bearing on the shape of the current drawn from each leg of the three phase supply.
This is typical:

As you can see, the current is discontinuous. There is no simple equation that can define it and that's where the harmonic analysis comes in handy. The waveform can be analysed as a series of sine waves.
Cue Fourier.

 

[Besoeker]

Small correction.
In post #6 the SCRs should have been diodes but the arrangement shows what was intended.

 

[D!NNy]

First, let's look at what a harmonic is in the electrical context.
For a repetitive non-sinusoidal waveform, you have a fundamental frequency component, often 50Hz or 60Hz input side of the VSD and variable on the output side.
Harmonics are multiples of this fundamental frequency.
The pulse width modulation (PWM) on the output side is a switching frequency, typically in the several kHz range and not strictly related to the fundamental output frequency so not a harmonic of it.
But yes, the capacitor on the DC link provides a low impedance source for the load side switching frequency components so they don't appear on the supply side.

Back to supply side harmonics and what they are.
Take a simple square wave like this:

It is just that. A simple square wave. Now, sometimes you may want calculate how it will interact with other parts of say, a network.
One way of doing this is to mathematically describe the waveform as a combination of sine wave - sine waves are easier to work with than waveforms with discontinuities like the square wave.
The square wave can be represented mathematically as an infinite series of sine waves at multiples of the fundamental frequency. In this case, all the odd multiples - three times, five times etc.
But note that this is a mathematical tool to aid analysis. A useful tool it is. But a tool nevertheless. The harmonics don't create the waveform.

For the three phase variable frequency drive, the input circuit is usually like this:

The components that follow that have a significant bearing on the shape of the current drawn from each leg of the three phase supply.
This is typical:

Is this wave form a output of the square wave applied to the rectifier unit? can you explain me how the current curve is being restricted to 15 Deg to about 50Deg?

As you can see, the current is discontinuous. There is no simple equation that can define it and that's where the harmonic analysis comes in handy. The waveform can be analysed as a series of sine waves.
Cue Fourier.

VFD is meant to use variable speed application but in this case VFD is to provide the smooth start they run at the constant speed once after they have been started. So do we still have harmonics even after reaching to full speed. Yes that is because we are always switching inside VFD even for 60Hz am i correct? Will it be good idea to bypass VFD after reaching to FULL SPEED that way there wont be any harmonics in the systems?

 

[D!NNy]

Small correction.
In post #6 the SCRs should have been diodes but the arrangement shows what was intended.

SCR instead of diode will that make any difference to harmonics?

 

[kingpb]

If you want to ramp up and then go full speed, you should consider reduced voltage starter. You must however, take into consideration the amount of starting torque you will need to overcome. VFD works on Volts/Hz wheras RVS works on simply reduced voltage. There are starting torque limits, i.e. won't work in every situation.

I noticed you are using the MCC as your PCC, why not use the input terminals at each drive as the PCC? that way you won't have system issues. Many manufactureres like to use the MCC becasue they will use other loads on the system to try and cheapen up their equipment, but that does not help the end user.

Jsut something to consider.

 

[Besoeker]

VFD is meant to use variable speed application but in this case VFD is to provide the smooth start they run at the constant speed once after they have been started. So do we still have harmonics even after reaching to full speed. Yes that is because we are always switching inside VFD even for 60Hz am i correct?

You will still have supply side harmonics. But not due to switching which is what I have been trying to explain. The line side harmonics are a result of the input rectifier section taking non-linear current.

Will it be good idea to bypass VFD after reaching to FULL SPEED that way there wont be any harmonics in the systems?

You can change over to fixed voltage at full speed. Needs a clever bit of electronics to do so without a bump. I've yet to see one that does so entirely seamlessly. And yes, that would remove the non-linear input rectifier.
Whether you still have harmonics on the system would depend on what else is connected and contributing. But your now fixed speed motor wouldn't be one of the contributors.

 

[Besoeker]

SCR instead of diode will that make any difference to harmonics?

Depends on how they are operated.
I've seen arrangements where SCRs were used to charge up the DC link capacitor in a controlled fashion and eventually phased fully on.
But most that I've seen these days have a plain rectifier and a pre-charge resistor that gets shorted by a contactor when the capacitors are charged.

If the SCRs are use to vary the DC link voltage, then that will make a difference to the harmonics. That used to be done before the introduction of IGBT technology.

 

[D!NNy]

If you want to ramp up and then go full speed, you should consider reduced voltage starter. You must however, take into consideration the amount of starting torque you will need to overcome. VFD works on Volts/Hz wheras RVS works on simply reduced voltage. There are starting torque limits, i.e. won't work in every situation.
why not VFD can be used to meet the starting torque of the motor all the time? can you tell me what is limitation?

I noticed you are using the MCC as your PCC, why not use the input terminals at each drive as the PCC? that way you won't have system issues. Many manufactureres like to use the MCC becasue they will use other loads on the system to try and cheapen up their equipment, but that does not help the end user. Also can you explain me how this PCC at the MCC is not helping the END user?
If i use PCC at the VFD input that will lead me to use the harmonic filters at 6 different locations which will be a space issue.

Jsut something to consider.

Thanks

 

[D!NNy]

Depends on how they are operated.
I've seen arrangements where SCRs were used to charge up the DC link capacitor in a controlled fashion and eventually phased fully on.
But most that I've seen these days have a plain rectifier and a pre-charge resistor that gets shorted by a contactor when the capacitors are charged.

If the SCRs are use to vary the DC link voltage, then that will make a difference to the harmonics. That used to be done before the introduction of IGBT technology.

Is there any way that i can see detailed schematic of the VFD from any manufacturer? would you recommend me to whom to go with ?
Thanks

 

[D!NNy]

You will still have supply side harmonics. But not due to switching which is what I have been trying to explain. The line side harmonics are a result of the input rectifier section taking non-linear current.


You can change over to fixed voltage at full speed. Needs a clever bit of electronics to do so without a bump. I've yet to see one that does so entirely seamlessly. And yes, that would remove the non-linear input rectifier.
Whether you still have harmonics on the system would depend on what else is connected and contributing. But your now fixed speed motor wouldn't be one of the contributors.

Perfect thanks

 

[kingpb]

Torque explanation for RVS

http://www.powerqualityanddrives.com/reduced_voltage_starting_torque/

 

[D!NNy]

Thanks every one. Information provided by all of you guys helped me a lot.