load balancing in 100A/3P/4W commercial panel

[malachi constant]

Engineer here. Trying to figure out how far to enforce the load balancing specifications. Spec requires loads to be tested and balanced on all equipment - transformers, switchboards, panelboards. Spec says if any two phases within a panelboard are not balanced within 20% then rebalance/retest.

After initial testing the worst one is 22.3A/18.4A/18.0A. 22.3/18.0 = 124% -> a 24% difference. This is a 480V lighting panelboard. Contractor is arguing that within smaller panelboards a couple phases being off by 20% is not a big deal.

So what are the ramifications of an unbalanced panel? What about an unbalanced transformer? Service? At what point is balancing a concern, and why?

 

[jghrist]

If you have a spec, then follow it. You need to define unbalance, however, to apply the spec. If the definition is (Max -Average)/Average or (Average - Min)/Average, then you only have 14% unbalance.

On re-reading the original post, I see that it is defined.

 

[charlie b]

With a lighting panel, the loads will vary greatly from moment to moment, as you turn lights on and off in various rooms. So there is no meaning to having an imbalance at any point in time, for you need only wait until someone enters or leaves a room, and the degree of balance will change.

I should like to know what the loading condition was, when you did the test. Were all lights on throughout the entire building? Was that defined in the spec requirements?

 

[charlie b]

At what point is balancing a concern, and why?

It is not a concern until the loading approaches 100%. That is because an imbalance could mean that two phases are running at 90% of the conductor ampacity and the third phase at 110% of the conductor ampacity . Unless and until there is a risk of one phase being overloaded (i.e., beyond its ampacity), there is no risk of damage to any conductor. I think that it would be worth the owner's time to consider removing that spec requirement.

 

[malachi constant]

It is not a concern until the loading approaches 100%.

Is that true? If a 200A panel is normally loaded at 50A/100A/150A for 10+ years there is no risk of physical damage to the equipment or feeders?

I recall as a newbie right out of school, being on a field visit with a engineer w/ 30 years experience. He pointed to a pole-mounted utility transformer and said he could tell from the visible damage to (...unfortunately I don't remember what he was pointing at...) that the service was unbalanced. Something to do with magnetic fields or something.

But if the only real long-term danger is that a highly-loaded panel might get its loads unevenly distributed...well, yeah, it seems like a dumb thing to put in the spec - esp since most panels calculated as "highly loaded" are in actuality loaded much less.

So could anyone else weigh in on the above?

Does the same hold true for transformers? Do they "transform" worse or are otherwise damaged if they are highly unbalanced? Thanks!

 

[markstg]

Is that true? If a 200A panel is normally loaded at 50A/100A/150A for 10+ years there is no risk of physical damage to the equipment or feeders?

I recall as a newbie right out of school, being on a field visit with a engineer w/ 30 years experience. He pointed to a pole-mounted utility transformer and said he could tell from the visible damage to (...unfortunately I don't remember what he was pointing at...) that the service was unbalanced. Something to do with magnetic fields or something.

But if the only real long-term danger is that a highly-loaded panel might get its loads unevenly distributed...well, yeah, it seems like a dumb thing to put in the spec - esp since most panels calculated as "highly loaded" are in actuality loaded much less.

So could anyone else weigh in on the above?

Does the same hold true for transformers? Do they "transform" worse or are otherwise damaged if they are highly unbalanced? Thanks!

Yes what Charlie says is true. A 50% unbalance as you describe is not a risk of damage to cables or electrical equipment.

I would keep the balancing requirement in the specification. A highly unbalanced system, can lead to inadvertant overloads quicker than a balanced system. Attempting and being fairly succesful at balancing the connected loads is better than not.

Transformers do not transform better balanced than unbalanced within their ratings.

 

[dbuckley]

After initial testing the worst one is 22.3A/18.4A/18.0A.

Requring loads to be balanced at this level of current (assuming the installation has a 100A service or more!) is lunacy.

You need to get the definitions changed on the spec.

 

[malachi constant]

Requring loads to be balanced at this level of current (assuming the installation has a 100A service or more!) is lunacy.

You need to get the definitions changed on the spec.

After reading comments here I agree. Any suggestions as to when load balancing should be considered? In practice should it ever?

 

[DataCenterGuy]

White Paper on Phase Balance

White Paper on Phase Balance

A recent white paper claims to prove that energy savings can be gained from phase balance. It was written with the Data Center environment in mind, but should apply to anything if it's accurate. (I didn't write it, nor am I endorsing it. Just providing it for what may be worth.) It's too large to attach, but the URL to Download is:
http://www.servertech.com/whitepape...-Few-Inches-Of-a-High-Efficiency-Power-System

 

[markstg]

After reading comments here I agree. Any suggestions as to when load balancing should be considered? In practice should it ever?

Yes, load balancing, IMHO, should always be considered. The designers usally do a pretty good job, and the Panel Schedule reflects that, both in design and in practice.

By balance checking, this verifies if rebalancing is needed. There can be no hard and fast numbers that will work for all cases, which is why if I put out a number in the specification, I might add "or accepted by the engineer or inspection authority", which allows acceptance by the authority to override an otherwise acceptable installation.

By not having anything in your spec, 1. Balancing will never be checked, 2. and if it is, you have nothing to stand on to change it.

 

[jim dungar]

He pointed to a pole-mounted utility transformer and said he could tell from the visible damage to (...unfortunately I don't remember what he was pointing at...) that the service was unbalanced. Something to do with magnetic fields or something.

It is possible that a multi-core transformer which is fully, or even over, loaded and severely unbalanced could have a large magnetic flux imbalance which could cause an excess amount of flux to 'flow' through a portion of the tank, in turn causing a portion of the exterior paint to become discolored.