Hi all,
I installed a 1200A Square D I-line breaker today and am getting some very odd amp readings on the load side of the breaker.

The facility is a large screen printing factory with 208v, 2500A, three phase service. Approximately 60,000 sq. ft., with 1/3 production and offices and 2/3 warehouse. Most of the machinery, Hvac, etc. is three phase. The lighting is all 120v.

Customer called and said they were having problems with an 800A breaker tripping. This breaker is the main for an I-line sub-panel which in turn feeds several subs with a wide variety of loads. We checked the load and the breaker was carrying 760A. Today we changed out the breaker to a 1200A.

When we were doing the load check we noticed that the two 500mcm parallel conductors were not sharing the load equally. We wrote that off as a poor installation with unequal length parallel conductors. Wrong! Our new installation shows the same symptoms.

Today we pulled out all the old conductors and installed a new 1200A breaker with three parallel runs of 500mcm CU on the load side. All the phase wires were cut to exactly 88" . The three neutral conductors were all cut to 110" .

The problem is that the parallel runs show an imbalance of 30% on any given phase. The lugs are torqued. What's up with that???

How much imbalance is normal and acceptable with parallel conductors??

GJ

Time to look at load downstream. You may have loads that are not balanced from the sub panels. Look for harmonics and surge type loads. eg a shear or press that cycles its motor(s) on and off perhaps quickly.

What kind of amperage numbers are we talking? How many conduits?

nesparky,
I don't see how that would cause unequal division of current on a parallel feeder.


[This message has been edited by Redsy (edited 09-23-2001).]

That is exactly my question!

Parallel feeders are resistors in parallel. Ohms law says that if the currents are unbalanced, we know the voltage is the same, the resistance has to be out of balance by the same amount as the current.

What kind of amperage numbers are we talking? How many conduits?

The feeders are run through the switchgear, no raceways, except a nipple between the two cabinets. Downstream the load is very diverse. Sub-panels for lighting and HVAC. Three phase loads for equipment, every other type of load you would expect in the typical commercial/industrial setting.

We were measuring with two different cheap Amprobe clamp on meters. Both meters showed the same results, I don't know if these read true RMS or not.

The load was very light. We did this on Saturday, with the factory shut down. No equipment was operating.....the only loads were lighting and Hvac. Normal load with the plant operating is 760A on this feeder.....on Saturday we were seeing about 200A.....the three parallel feeders were showing numbers like 55A/63A/78A....all three phases were similar.

GJ

FWIW there is considerable phase imbalance. At full load of 760A the neutral is carrying about 150A of current. This seems to be a separate issue. Some of the load needs to be shifted from A & B to C phase to balance everything.



[This message has been edited by golf junkie (edited 09-23-2001).]

GJ,
you're defying the physic's of electricity here

Sparky,
I'm not that clever!
Either the parallel resistance is out of balance or the meter is lying.....I suspect the later.

GJ

The other possibility is that something is happening that I just don't understand, which is why I'm asking for your expertise.

GJ,
I had a similar situation a while back. While I don't have an answer for you, I can tell you this. My situation was a temp generators. We ran 12 4/0 type W per phase from each generator to a termination cabinet to parallel the two generators. All cables were exactly 40ft in length. While load testing the generators with a load bank (a purely resistive load) I took some amp readings on the cables out of curiosity. To my surprise there was a variation of about 50A in while pulling a total of 1000A. Some cables were very close to each other some were quite different. I never could explain why especially in a controlled situation.

GJ,
i guess you need to get a few more meters in there and do a finer assessment. is it possible to get a probe over all the #500's as well?
maybe check it against the line side???

[This message has been edited by sparky (edited 09-24-2001).]

A couple of thoughts-
1) Is the unequal division of current consistent as a percentage or as an absolute value? If the imbalance remains 20-25 amp at higher load, I wouldn't be as concerned. If the imbalance increases proportionally and it looks like the conductors will exceed their individual ampacities, then I would be more worried.
2) Are the cables bundled in any manner? Do they pass through the nipple ABCN? Making sure that they are grouped together this way for the whole length might help.

How about swapping the highest reading pair in the breaker, and then re-assessing. There must be some sort of R value that is inconsistent, possibly in the new breaker and/or termination ( maybe you could check the heat on all terminations??)

Sorry if this is way off base, but I'm not too familiar with the layout of your 3-ph commercial panels.

Any chance that there's a large area of ferrous metal closer to one or more of the parallel feeders than the others? The inductance could cause an imbalance.

Why is it such a big deal to try and get exactly the same load on each conductor? I don't beleive there is any way to be that exact. Given two pieces of wire that are the same there will be slight differences. Those differences are in its terminations, bends, lengths, etc.. We install parrallel conductors all the time and yes there are always variations in the amount of load that each carries. But there has never been a concern about it.

glenn

PS , I don't think your meter is lying. Its just a fact of parrallel conductors.

One more thing. Sometimes in my business of being a lineman we must install mechanical type jumpers to perform liveline work so that we keep customers happy. It was brought up by someone in the past that when parralleled jumpers are used that they should be installed a certain way to equalize the load.

Hope this makes sense;

wrong way--------1-2----------2-1---------

right way--------1-2----------1-2---------

if done the wrong way then #1 jumper hogs all the load even though the jumpers are exactly the same length.


[This message has been edited by glenn35 (edited 09-25-2001).]

It's ok? happens all the time?

someone call the jaws of life an' help me get both feet outta' my mouth....

Not off base at all. This is what I was alluding to in my post above.

I believe a slight imbalance may be unavoidable, but depending on the magnitude, one of the conductors still may become overloaded.

Thanks for all the feedback guys!

I haven't been back to do any more checking. In fact that may not happen for some time.

I did so some comparitive measurements on a similar set-up. 1200A, 208v, 4 wire service. The main is three parrallel runs of 600mcm CU. None of the service laterals showed an imbalance greater than 5%.

A few things are different;
This is a church/school....the load is not as diverse.
The feeders are longer (100') so a few inches of difference in length would not make less of a percentage change in resistance.

GJ

GJ
The reason I suggested checking the down stream load is I once had a simular problem. A 4000 amp service had 14 runs of 500MCM to it. Kept burning cables. Not a good thing in an ethonal plant. We repulled the 500 MCM making sure all cables were the same length. When we turned the service back on amp readings were scary less than 200 on 2 or more cables over 600 on others with a range of readings in between. The engineers and the rest of us went nuts trying to find the problem.
While the plant was down because of this another crew did some repair of some computer controlled equipment and rebalanced the various loads at some sub panels.
After repulling the 14 runs of 500 MCM again,reterminating and checking every thing else back on the service came. Same problem different cables. Back to tests, discussion-sometime heated- meetings etc.
Other crew comletes their tasks using only enough power to test each individual subpanel.
Management want more tests full power run-- problem disappeared.!!!!!
What caused it is still an arguement something downstream probably causing a harmonic imbalance. What was never agreed to. When we left all 14 500 MCM / phase 3 phase 480 cables were within 4% of each other.
Hope you have better luck and actually find a cause.

Redsy,
your 9-24 post is interesting there...

Dare I ask...
Interesting in a good way or bad?

On Saturday, when we changed the breaker, there was no equipment operating.....only lighting and HVAC. The HID lighting could be causing harmonics.....in fact, thats the first place I would look. But I don't understand how harmonics can affect paralleled feeders.

Can anyone explain this????

GJ

I don't know if harmonics are an issue here, but if they are, doesn't the skin effect increase with increased frequency? If, for example, all of the A phases were next to each other, flat, the one in the middle might carry less current than the other two. Also, if improperly grouped, the proximity of the metal enclosure might cause a choking effect on the conductor nearest it. These problems could also occur at fundamental frequency.
I dunno!

Dare I ask...

interesting as a method of assessment