Total Harmonic Distortion

Status
Not open for further replies.

Chamuit

Grumpy Old Man
Location
Texas
Occupation
Electrician
I have a new toy that gives/monitors THD. How can that help me? I have a very light-weight understanding of harmonic distortion.
 

Chamuit

Grumpy Old Man
Location
Texas
Occupation
Electrician
Thanks Rob!

That was an interesting paper. I knew quite a bit of what was presented in there from my CEU classes. In fact, the PE that taught one of my classes went on about needing to make sure your neutrals were sized correctly when dealing with computer centers. But, he didn't lay it out as well as that paper did.

So, in a typical residential setting, this particular feature will more than likely not be of any help or significance for me. Right?

I don't have any customers that have VFDs that I can think of off the top of my head.

What would be a "bad" THD. It measures it in %'s.
 

Chamuit

Grumpy Old Man
Location
Texas
Occupation
Electrician
THD what?

Voltage or current?

I thought you could figure that out.

It would be voltage. The meter measures and logs voltage, sags, swells and implulses. It also measures % THD and logs periods of 3% (line to neutral.)
 

Electric-Light

Senior Member

One thing about that is the recommendation about computer loads is going to become obsolete fairly soon.

All 80Plus certified power supplies are power factor corrected and offer efficiency in excess of 80%. They're becoming more common in OEM units due to globalization as non-power factor corrected PCs are prohibited in the EU and having to have US only and export power supplies creates demand-supply forecast and inventory management difficulty.

This is also why you see RoHS certified stuff in US these days.
 

wirenut1980

Senior Member
Location
Plainfield, IN
Thanks Rob!

That was an interesting paper. I knew quite a bit of what was presented in there from my CEU classes. In fact, the PE that taught one of my classes went on about needing to make sure your neutrals were sized correctly when dealing with computer centers. But, he didn't lay it out as well as that paper did.

So, in a typical residential setting, this particular feature will more than likely not be of any help or significance for me. Right?

I don't have any customers that have VFDs that I can think of off the top of my head.

What would be a "bad" THD. It measures it in %'s.

I have only seen harmonics cause issues in industrial settings, and most of these issues were related to capacitor banks failing due to resonance. I have not had any experience in data centers, though.

In a residential setting, harmonics have not caused any issues in my experience. In fact, IEEE is going to raise the harmonic limits in IEEE 519 for the next update of the document, because there are very few problems associated with harmonics.

I can't give you a bad THD percentage, because harmonics are everywhere, but there really isn't much equipment that is sensitive to them. It really is situation dependent.
 

Electric-Light

Senior Member
I have only seen harmonics cause issues in industrial settings, and most of these issues were related to capacitor banks failing due to resonance. I have not had any experience in data centers, though.

In a residential setting, harmonics have not caused any issues in my experience. In fact, IEEE is going to raise the harmonic limits in IEEE 519 for the next update of the document, because there are very few problems associated with harmonics.

I can't give you a bad THD percentage, because harmonics are everywhere, but there really isn't much equipment that is sensitive to them. It really is situation dependent.
3 phase motors don't like 5th harmonic.

CFLs and switching power supplies cause odd order harmonics. 5th harmonic causes three phase motor to pedal "backward" which cause reduction in efficiency and sometimes failures. Computers are usually the worst offenders since the VA draw is significant. Externally ballasted fluorescent lights are not really an issue, because they're required to have front end that limits harmonics.

Power supplies used in data center equipment or the double conversion UPS' front end portion these days have active PFC like the type used in fluorescent ballasts, so they're generally not a bad offender when it comes to causing harmonics.

Here, harmonics from an office building with a lot of computers served on the same PoCo transformer as the machine shop caused damage to motors.

Also, switching power supplies pull power from the wave peak and everything downstream essentially operates from 170v DC. Put enough of these loads and the waveform will become trapezoid and peak voltage will drop enough to cause similar loads to not operate properly. This is called "flat topping"

http://www.fluke.com/Fluke/usen/solutions/pq/Case-Study-Machine-Shop-Motor-Failures.htm
 
Last edited:

robbietan

Senior Member
Location
Antipolo City
Thanks Rob!

That was an interesting paper. I knew quite a bit of what was presented in there from my CEU classes. In fact, the PE that taught one of my classes went on about needing to make sure your neutrals were sized correctly when dealing with computer centers. But, he didn't lay it out as well as that paper did.

So, in a typical residential setting, this particular feature will more than likely not be of any help or significance for me. Right?

I don't have any customers that have VFDs that I can think of off the top of my head.

What would be a "bad" THD. It measures it in %'s.

a "bad" THD would be a 5% Voltage THD. that level of harmonics may disrupt computer operations, heat up transformers, etc. In a typical residential setting, the amount of harmonics entering the utility transformer would be very small and would not be a problem.
 

wirenut1980

Senior Member
Location
Plainfield, IN
3 phase motors don't like 5th harmonic.

CFLs and switching power supplies cause odd order harmonics. 5th harmonic causes three phase motor to pedal "backward" which cause reduction in efficiency and sometimes failures. Computers are usually the worst offenders since the VA draw is significant. Externally ballasted fluorescent lights are not really an issue, because they're required to have front end that limits harmonics.

Power supplies used in data center equipment or the double conversion UPS' front end portion these days have active PFC like the type used in fluorescent ballasts, so they're generally not a bad offender when it comes to causing harmonics.

Here, harmonics from an office building with a lot of computers served on the same PoCo transformer as the machine shop caused damage to motors.

Also, switching power supplies pull power from the wave peak and everything downstream essentially operates from 170v DC. Put enough of these loads and the waveform will become trapezoid and peak voltage will drop enough to cause similar loads to not operate properly. This is called "flat topping"

http://www.fluke.com/Fluke/usen/solutions/pq/Case-Study-Machine-Shop-Motor-Failures.htm

Interesting study, thanks. I was not aware that around 5% voltage THD was considered a problem for fully loaded motors. I wonder if there was some kind of study behind that. I also wonder what the voltage unbalance was at the motor. A fully loaded motor can only take 1% of voltage unbalance before derating is needed.

I am curious what loads in an office building produce 5th harmonic, because computers and flourescent lighting are mostly 3rd and very little 5th. Maybe 3 phase heating and cooling with a VFD? I have noticed some significant 5th harmonic voltage distortion on utility feeders that has nothing to do with the load, plus the loads on the feeder were primarily residential.

I still stand by my statement that the vast majority of locations do not have any problems due to voltage or current harmonics.
 

Electric-Light

Senior Member
A fully loaded motor can only take 1% of voltage unbalance before derating is needed.
5th harmonic is like an additional load, because its more or less like applying brakes.

I am curious what loads in an office building produce 5th harmonic, because computers and flourescent lighting are mostly 3rd and very little 5th.
Take a look at the graph on the left under HARMONIC ANALYSIS & WAVEFORM CAPTURE header. About half way down the page:
http://www.mymeterstore.com/p7789/elog2009.php

Not endorsing them, but that's what I found using Google. Note the quite significant amount of 5th harmonic.
 

wirenut1980

Senior Member
Location
Plainfield, IN
5th harmonic is like an additional load, because its more or less like applying brakes.


Take a look at the graph on the left under HARMONIC ANALYSIS & WAVEFORM CAPTURE header. About half way down the page:
http://www.mymeterstore.com/p7789/elog2009.php

Not endorsing them, but that's what I found using Google. Note the quite significant amount of 5th harmonic.

I stand corrected on the 5th harmonic for the computers (and probably lighting as well). The attached picture shows the waveform from my laptop. The 5th is almost as much as the 3rd in that waveform. I guess I am used to monitoring close to the main where voltage harmonics typically are not all that high due to the harmonic loads being a small part of the total load.

I understand how the negative sequence harmonics act like additional load. My point was that the voltage unbalance also could have caused heating of the motor windings, but that was not mentioned in the article.
 

Attachments

  • laptop computer power supply voltage and current.jpg
    laptop computer power supply voltage and current.jpg
    52.7 KB · Views: 0

Electric-Light

Senior Member
Where did you take that measurement? You can see the flat topping in the voltage waveform. Was this in a suite in an office building?

I stand corrected on the 5th harmonic for the computers (and probably lighting as well). The attached picture shows the waveform from my laptop. The 5th is almost as much as the 3rd in that waveform. I guess I am used to monitoring close to the main where voltage harmonics typically are not all that high due to the harmonic loads being a small part of the total load.
THD current is usually 100-150% (relative to fundamental) on uncorrected rectifier bulk storage loads.

Electronic loads which use DC rail internally that represent a significant portion of kVA demand (such as VFDs and electronic ballasts) have a harmonics correction which also corrects power factor. Ballasts for commercial use must have correction. It is also required for computers in some countries, so in an effort to reduce SKUs, active PFC computer PSUs are becoming more common in the US too.

Lighting isn't a big factor. Older electronic ballasts were rated <20% THD. Newer ones with active PFC are rated <10% and power factor is usually over 0.98. Screw-in integral ballast CFLs however are not corrected, but they usually don't represent a significant amount of kVA.

BIG battery chargers are an issue too as batteries behave similarly to capacitors in that it only starts taking in charge only on higher ends of the cycle. When this golf course had a tournament and charged up a ton of golf carts all at once, it burnt out the PoCo's tranny.
http://www.fluke.com/fluke/nlnl/opl...teit/case-study-rural-transformer-failure.htm

I understand how the negative sequence harmonics act like additional load. My point was that the voltage unbalance also could have caused heating of the motor windings, but that was not mentioned in the article.
Doesn't look like it...
 
Last edited:

wirenut1980

Senior Member
Location
Plainfield, IN
The measurement was taken at my desk at work (office building). You can see some flat topping in the voltage waveform, but the voltage THD is still only 1.4%.
 

Attachments

  • Harmonic spectrum.jpg
    Harmonic spectrum.jpg
    61 KB · Views: 0
Status
Not open for further replies.
Top