Harmonic affects

[dbuckley]

Design a robust distribution system and you should not need to take 'special' measures to counter harmonic currents.

Very true, but most folks (make that - most electricians) dont like seeing panelboards and cabling significantly underrun. If you've got full size neutrals, you can't have 80% loading with continuous non-linear loads and expect your panelboard and wiring to survive, as awith (say) 100A 'board with 80A fairly well balanced phase current loads, the neutral probably has in excess of 110A on it, and, sooner or later, theres going to be trouble.

Run your 200A panel at 120A max load and you'll probably be fine.

And at the end of the day (or more accuratly, at the end of the wire) the transformer feeding the non-linear loads has to be able to live with the extra temperature rise. K rated transformers are the right answer, but using an oversized standard transformer is a technique with long history...

 

[don_resqcapt19]

Harmonics makes the cables coming into a building smoke as they are a bit hot.

Harmonics make vertical busbars melt.

Harmonics cause terminations to melt and fail.

All seen 1st hand.

I have a second hand story from a reliable source of a dry transformer that failed impressively due to harmonics.

Ok, lots of times harmonics cause no actual problems, but thats not to say they never cause problems. If you fool around with significant non-linear loads at somewhere near the rated capacity of the distribution system, be prepared to see harmonics be an issue.

Of all of the projects that you have worked one, what percentage have had these spectacular failures? My only point is that if you read the trade magazines and the manufacturer's literature, you are lead to believe that these types of failures would appear on a majority of the electrical installations that involve non-linear loads. That just is not happening. There is no question that these problems can and do occur, but not in a majority of the installations.

 

[Besoeker]

Actually we can, if it is working fine why would we search for a problem in need of a fix?

The assumption to which referred was:

the plant that I described should have lots of problems

You can't assume that without a detailed analysis.

 

[Besoeker]

Of all of the projects that you have worked one, what percentage have had these spectacular failures? My only point is that if you read the trade magazines and the manufacturer's literature, you are lead to believe that these types of failures would appear on a majority of the electrical installations that involve non-linear loads.

As I have said a number of times, the spectacular failures are the ones that get the immediate attention.
Any infrequent deleterious effects may well go unreported.

 

[__dan]

mmm

mmm

An MCC with a 50 Hp circulator and lots of tiny 3 ph valve actuator motors and the valve electronics.

One year in, the main at the switchboard and the main at the feeder both blow open under fault, smoked some of the valve electronics ... why ?

When I saw the drawing last year I told them to separate the feed to the 50 Hp circ from the rest of the load, separate clean feed to the valve electronics. I suspected they would not like getting hammered with switching transients from the 50 Hp circ during generator test switching. This load going down is $$$.

Proactive vs reactive, guys declined to make the change at design but love to fix the same problem three or four times after something trips a few times.

Point is, I am wondering if applying some PFC correction caps on the 50 Hp circ, if they would have mitigated the fast switching transients, made them small enough so they were harmless to the valve electronics. My recommendations included a TVSS but I am wondering if PFC caps would be a better application for absorbing inductive switching transients.

If so, that would be the lowest cost fix of the range of options.

 

[iwire]

Very true, but most folks (make that - most electricians) dont like seeing panelboards and cabling significantly underrun. If you've got full size neutrals, you can't have 80% loading with continuous non-linear loads and expect your panelboard and wiring to survive,

I do not see loading anywhere near that level in modern buildings designed under the NEC.

Brian Johns testing seems to bear that out. The problems seem to be in old buildings that have been changed to new uses without proper engineering.

 

[PowerQualityDoctor]

It tells me you cannot count very well, there are many moderators that have not joined in.

How about you explain how you came up with this statement?

In this posts there are 2 moderators, both expressing the same non believing approach. Besides, I said non scientific.

I sure would like to see you back up the above statement of yours as IMHO it is far from accurate. I asked you before for some statistics to back that up and you said you had none.

Then there is this one, lets clear up exactly what you mean.

Told you once, also in PM. 50 harmony equals 1kHz, which is already in the area of induction heaters. There is no statistics.

Are you saying reducing harmonics will reduce my utility payment by 30% regardless if the utility penalizes me for it or not?

Considering you brought up reducing my salary by 30% I have to assume you are saying you could reduce a utility bill by 30%.

Depends on the harmonic level and amount of cable and their cross section, this is definitely possible.


I know that whatever I will say, you will consider me as not reliable, but I still try.... I just came from a power quality conference which held almost 100 engineers in this field, including utilities, regulators, consultants, end users and solution providers. There was a lecture explaining how to estimate the losses due to harmonics with some examples. The potential for saving on reduction of harmonics is huge, depending where we make the filtration. I have asked for copy of this lecture (which includes quotes from international reports) and will post some data if I get it.

 

[StephenSDH]

I would hate to see you guys in a thread with Mike Holt and Stray voltage.

I will likely die never seeing an issue with stray voltage. I enjoy learning about. I have seen a couple issues with harmonics, and we design around harmonics, transients, noise in every project.

 

[dbuckley]

Of all of the projects that you have worked one, what percentage have had these spectacular failures?

These are rare unusual events, not common.

I do not see loading anywhere near that level in modern buildings designed under the NEC.

Again, these are rare events that only occur when you have all (or almost all) non-linear loads and a lot of them. Places like data centres and finance house trading floors. These are not places most electricians work most of the time.

But these issues do occur in office accomodation too. The panelboard on the floor of my office at work has a distinct 150Hz hum emanating from it. Our floor is all IT people, so all the power loads are computers, and the building was built in the 70s and so has a shortage of outlets - I have three four-way distribution blocks off my one socket, and I have less stuff than most on my floor.

When I hear a panel humming away at 150Hz (would be 180Hz in the US!) then thats a sign of harmonics. If the panel is humming away at 50Hz/60Hz then its probably OK.

 

[brian john]

As I have said a number of times, the spectacular failures are the ones that get the immediate attention.
Any infrequent deleterious effects may well go unreported.

Not if they are doing predicative maintenance AKA IR.

 

[brian john]

These are rare unusual events, not common.


Again, these are rare events that only occur when you have all (or almost all) non-linear loads and a lot of them. Places like data centres and finance house trading floors. These are not places most electricians work most of the time.

And in my experience are all a result of poor management, poor design, poor planning, poor ongoing maintenance.

 

[iwire]

Again, these are rare events that only occur when you have all (or almost all) non-linear loads and a lot of them. Places like data centres and finance house trading floors. These are not places most electricians work most of the time..

Actually I have worked in a lot of these types places and again the designs where such that it would not be possible to max out the electrical systems.

I can entirely understand why a building built in the 60s would have issues.

 

[Besoeker]

Not if they are doing predicative maintenance AKA IR.

How much of that is carried out in domestic premises?
And IR is not the only failure mode.

 

[Besoeker]

Actually I have worked in a lot of these types places and again the designs where such that it would not be possible to max out the electrical systems.

I can entirely understand why a building built in the 60s would have issues.

The hotel I mentioned has been open less than 10 years.
I have also had to come up with solutions for data centres in London's Docklands area. These were completely new buildings. In fairness, there was some awareness of the potential problems before the centres were fully operational.
So, not just old buildings.

 

[mivey]

Actually I have worked in a lot of these types places and again the designs where such that it would not be possible to max out the electrical systems.

I can entirely understand why a building built in the 60s would have issues.

And I suspect the average age of buildings in North America to be at least 40-50 years and the average electrical system maybe slightly younger (10 years?). Maybe someone has access to some insurance statistics that would have this type data. Of course, I suspect new england will have the oldest age profile.

 

[iwire]

The hotel I mentioned has been open less than 10 years.
I have also had to come up with solutions for data centres in London's Docklands area. These were completely new buildings. In fairness, there was some awareness of the potential problems before the centres were fully operational.
So, not just old buildings.

I very much doubt any of those where designed under the NECs requirements.

 

[iwire]

In this posts there are 2 moderators, both expressing the same non believing approach. Besides, I said non scientific.

Yeah

Told you once, also in PM. 50 harmony equals 1kHz, which is already in the area of induction heaters. There is no statistics.

That is not an answer to what was asked. But I did not expect you to do any better.

Depends on the harmonic level and amount of cable and their cross section, this is definitely possible.

But pretty damn unlikely, just like claims many sales people put out.

I know that whatever I will say, you will consider me as not reliable, but I still try.... I just came from a power quality conference which held almost 100 engineers in this field, including utilities, regulators, consultants, end users and solution providers.

So considering that you should be able to answer the question I asked about fires failures and losses.

 

[PowerQualityDoctor]

And if I'd show you statistics, you will say that it is biased, wrong, not in the relevant territory (over here we do all based on NEMA, so everything is excellent, of course for those with proper engineering).

I assume you don't need power quality analyzer, thermal camera or even torque meter (if it was closed properly on day one, why should it change?).

 

[brian john]

And if I'd show you statistics, you will say that it is biased, wrong, not in the relevant territory (over here we do all based on NEMA, so everything is excellent, of course for those with proper engineering).

I assume you don't need power quality analyzer, thermal camera or even torque meter (if it was closed properly on day one, why should it change?).

I won't, I promise

 

[iwire]

I assume you don't need power quality analyzer, thermal camera or even torque meter

Assuming is never a good plan.

Within sight of were I am sitting is a Dranetz PX5 PQ meter, a Flir P45 IR cam, my truck has two sizes of a torque wrenches and a torqure screwdriver in it. I am also a Flir certified level 1 thermographer.

(if it was closed properly on day one, why should it change?).

FWIW you really should not 're-torque' connections, you end up damaging the conductors.