Power Distribution Harmonic Content

[Ground_Loop]

I'm involved with an industrial application where the controls vendor has concern about the power supply harmonic content in the sense that it could cause damage to his electronics.  When I look at IEEE 519, the suggested harmonic content should not exceed about 8%.  IEEE goes on to use words like 'may, might, could' to describe possible risk of damage due to harmonics.  And even then only due to elevated heating effects.  In my searching I have not been able to find ANY documented case of equipment failure due to excessive harmonic content.  Nonetheless, the vendor is asking that we install harmonic filtering equipment before they will warrant their product against failure.  It is expected through modeling that the resultant power distribution system THD will be between 10% and 12% (in excess of IEEE 519 limits) once all the equipment is on line and running at expected power levels.

So my question:  Have any of you seen or have knowledge of confirmed equipment failure due to excess harmonic content and if so, what the THD percentage was?

[jonpaul]

Bonjour GroundLoop!

Have toiled  in power electroncis since 1970s, 

1/ PFC is a benefit to the power distribution system as MW are billable and MVAR are not, the apparent power causes added heating of the conductors but is not useful work.

2/ Old systems were most resistive or inductive loads with aPFC capacitor able to correct the PF of a lagging inductive.

Since 1970s, more and more load percent is electronic, with peak charging of rectifiers causing a different type of low PF problem, with lots of harmonic currents.

The various regulatory bodies established levels of PF that vary by the type of load (eg lighting, computers) and by the wattage range.
USA, Canada, EU, CE, IEC, IEEE all have up to date specifications, which can be required for import of a device into those regions.

3/ Compliance is usually with a PFC boost preregulator.

4/ Non-compliant equipment works OK but will fail a compliance test for harmonics and PF.

5/ In large industrial connections the power distribution company will penalize the customer for excessive VARS, with a higher rate to make up for their copper losses.

6/ Finally in a 3 phase WYE system, the neutral will carry the harmonic currents and excess VARS due to heavy electronic loads can overlod and overheat the WYE distribution neutral conductor.
That can be a safety issue!

So the answers depend on the level of power, distribution system,  import or export of equipment, etc.

I hope this clarifies this complicated issue.

With Kind Regards,

Bon Chance!

Jon





 

[f4eru]

Very good summary.
Fom equipments that can fail: small electronic devices with dropper cap PSUs will experience a lot of stress because harmonic will strain them a lot. I saw LED lights fail due to faulty switches arcing (making a lot of harmonics)

[David Hess]

The only equipment that I would expect to be affected are inverters if they are supplying the power.  The low impedance of the power line considerably ameliorates the effects of harmonic current draw.

If equipment was affected, then I would expect it to be so sensitive to power lines problems that it would fail anyway.

[f4eru]

Any equipment can fail unexpectedly due to unusually very high harmonics
- input filter caps can fail due to the HF current strain
- Input fuses can trigger and rectifiers can fail due to the added current on filter caps, especially on non-PFC stuff
- Coils and other isolation can fail due to the strain on insulation
- capdropper circuits can overpower (see previous)

Many failure modes possible....

[coppice]

I don't have experience of equipment being damaged by high harmonic content, but I do have experience measuring THD in a variety of places, and I do have experience of equipment crashing due to harmonic content.

Its common for the THD on the mains to exceed 20% in industrial areas. Its not uncommon to find 20% at the top of a large commercial tower during parts of the day when the load is high. At the bottom of the same tower you might only see a few percent THD at the same time.

Some equipment is quite prone to crashing when certain other high load equipment is turned on, and increases the THD. This is especially true when the additional load is very spikey, like welding equipment, which can produce harmonic content which may not be huge, but is biased towards very high harmonics. I think generally equipment has become more robust over the years. It used to be that we could crash our million dollar DEC10 computer by letting a cleaner plug a vacuum cleaner into certain nearby sockets.

One area of damage you might see is inlet filters overheating when the harmonics are very high frequency ones. For example, some things really don't like running from the square wave power produced by some UPSes. Anything that might have a cap drop supply somewhere inside, running some small measurement or sensing function, might also have heating issues with high frequency harmonics.

[dmills]

Not control gear exactly but I have had a delta-star transformer overheat due to neutral overload from triplen harmonics.

The load was a mix of theatrical dimming and some old and rather shonky discharge lighting (MAC 500s with magnetic ballasts, I think the PFC caps had mostly crapped out), when I investigated I saw roughly 220A RMS on each phase, and well over 400A on the neutral (Neutral busbar was hot enough to discolour the lugs).

Last time I dealt with dimmer loads I had the installer double up the neutral conductors, put paid to that nonsense quick like.
 
The other one I have seen is a passive PFC cap bank that did NOT appreciate high order voltage harmonics, we only learned the thing was there when the fire alarm went off and we found the switch room to be full of exploded cap and smoke from same.

[WattsThat]

I work in the variable frequency drive industry (motor inverters) and harmonics are big concern in several vertical industry sectors that utilize the products. Where you see this most often is where people are specifying for third parties and spending other people’s money to meet the requirements of IEEE 519.

To answer your question directly, no, I’ve never seen devices fail due to high harmonics. I’ve seen some rather crappy power supplies in industrial environments and devices survive as expected. Purely magnetic devices like induction motors and transformers can certainly run hotter but outright failure, no, not common at all. Things tend to fail in industrial environments due more to the all too common transients from things like PFC caps being switched out of circuit on the plant side of the transformer. MOV’s can only absorb so much energy before something gives up its smoke.

Your estimate of 10-12% THD isn’t uncommon in industrial environments. It’s within spec for IEEE 519 where the current short circuit to load ratio is 50-100.

If they’re just claiming the need for 5%, I’d really question the supplier of the equipment as to the basis for their requirements and are they based on real world data or is it some arbitrary desire for a CYA issue to run away from possible warranty claims or other non-performance issues. It just doesn’t pass the smell test for me.

BTW, the 1992 version of 519 has a great deal more information about harmonics in general than the current dumbed down 2014 version. It can easily found online.

[Ground_Loop]

I work in the variable frequency drive industry (motor inverters) and harmonics are big concern in several vertical industry sectors that utilize the products. Where you see this most often is where people are specifying for third parties and spending other people’s money to meet the requirements of IEEE 519.

To answer your question directly, no, I’ve never seen devices fail due to high harmonics. I’ve seen some rather crappy power supplies in industrial environments and devices survive as expected. Purely magnetic devices like induction motors and transformers can certainly run hotter but outright failure, no, not common at all. Things tend to fail in industrial environments due more to the all too common transients from things like PFC caps being switched out of circuit on the plant side of the transformer. MOV’s can only absorb so much energy before something gives up its smoke.

Your estimate of 10-12% THD isn’t uncommon in industrial environments. It’s within spec for IEEE 519 where the current short circuit to load ratio is 50-100.

If they’re just claiming the need for 5%, I’d really question the supplier of the equipment as to the basis for their requirements and are they based on real world data or is it some arbitrary desire for a CYA issue to run away from possible warranty claims or other non-performance issues. It just doesn’t pass the smell test for me.

BTW, the 1992 version of 519 has a great deal more information about harmonics in general than the current dumbed down 2014 version. It can easily found online.

Thanks for the info. I'll have to look up the older version of 519. So far I have not been able to find the technical basis for the recommended limit of 8% THD. It seems that someone threw out a number out of a sense of obligation.  And yes this is a case of a third party spending someone else's money.

[mansaxel]

Not control gear exactly but I have had a delta-star transformer overheat due to neutral overload from triplen harmonics.

The load was a mix of theatrical dimming and some old and rather shonky discharge lighting (MAC 500s with magnetic ballasts, I think the PFC caps had mostly crapped out), when I investigated I saw roughly 220A RMS on each phase, and well over 400A on the neutral (Neutral busbar was hot enough to discolour the lugs).

Last time I dealt with dimmer loads I had the installer double up the neutral conductors, put paid to that nonsense quick like.
 
The other one I have seen is a passive PFC cap bank that did NOT appreciate high order voltage harmonics, we only learned the thing was there when the fire alarm went off and we found the switch room to be full of exploded cap and smoke from same.

At a large outdoor festival in Sweden some 20 years ago, a mains 10KV/400V transformer caught fire from too many VariLites run on the same feeder, without other consumers to even out their quirks. 

This being Europe, with a decent mains supply voltage, the option of running the VariLite PDU in Delta mode did not exist, and Wye mode had to be selected, making the neutral conductor carry all harmonics.