Transformer feeder sizing, again

[wallypiper]

This has been discussed here in a variety of ways but I'd like some clarification/confirmation on a particular situation.

A european machine manufacturer commonly provides a transformer with their machine which runs on 400/3. They provide all secondary wiring and overcurrent protection (at the machine, based on tap rules) and even the wire trough for the secondary wiring.

The transformers are typically significantly oversized for the secondary load. They admit this and offer no technical explanation. It is easy to see when looking at their data. Presumably it has to do with availability, stocking decisions etc.

As part of our role in projects involving this equipment we have to list all of the electrical loads being installed. It's easy enough to get the KVA of the transformer and list the load based on that and the primary voltage (nearly always 480/3), but that results in oversized primary feeders based on the secondary load that will be served.

The rules for primary overcurrent protection seem to specify only the MAXIMUM allowed. As is often the case in the NEC, exact reading of the words is important. Can the primary feeder be sized based on the secondary load as long as the overcurrent protection for that feeder is also sized that way? The difference can be significant. I've seen it as much as 200 KVA. The transformer is not general purpose. It is dedicated to this one machine although it is installed separately. That is, it is not physically part of the machine but is installed next to it.

You don't have to buy the transformer from them, but if you provide it yourself you have to get a move on because availability of 480-400 transformers is not good in the US so they are frequently ordered from the machine supplier to avoid issues with delivery. If you bought it here, you would definitely try to get something closer in size to the actual secondary load. For what it's worth, we're talking about 400 to as much as 1200 KVA.

 

[WPiper]

57 views, no replies.

Are we "over our head"? Hope not. We list this type of info for dozens of different pieces of equipment on every job. This particular question, though, has us scratching our heads. We know what the secondary load is but since the transformer is rated much higher than that, we're not sure how it should be handled. We want to know not only that we are providing the correct information but also that the EC or consultant that specifies the wiring gets it right.

 

[texie]

I think I understand your delemma. This is sort of like having a VFD that is larger than required but by code you have to size the circuit to the drive. This has never been explained to me, though I have tried find the rational behind this. But in your case I can't think of code that would require this for a transformer circuit. As a practical matter though, I would think you would have to be careful not to go to low in the percentage of circuit to xformer capacity so you don't risk tripping the OCPD on xformer energizing, especially if it is unloaded.

Hopefully some of the more experienced guys can comment. Just thought my comments might help get something going for this.

 

[WPiper]

Sort of similar to the VFD issue for sure. Although the NEC wording on VFDs is pretty unequivocal, I have run into that exact situation with a large air compressor. The manufacturer stated that the circuit breaker feeding the compressor should be 600 amp but the input rating on the VFD was 614 amps. Although I could never get an explanation from the compressor manufacturer, I assume that they were not allowing the VFD to load beyond some point. I believe the motor was actually nameplated at 400v, 50Hz so it is certainly possible that they were not going to spin it at 60Hz and that would reduce the actual load it carried. But I don't see any wording in 430.122 that allow for that. Of course, you could put in (2) 600 kcmil and maybe still use a 600 amp breaker. Might have trouble with terminating wire that big on that breaker, not sure. Air compressors seem to be odd cases a lot of the time. Many are routinely run in the service factor of their motor which, you would think, would require special treatment. But I see most get wired based purely on the nameplate HP of the motor.

 

[wallypiper]

I think I have found a thread that answers my question.

http://forums.mikeholt.com/showthread.php?t=119324

The conclusion there seems to be that is is both acceptable and somewhat common to size the primary OCPD and wire for less than the rating of the transformer. The OCPD protects both the wire and the transformer.

It just needs to be big enough to handle the load created by the secondary and no bigger than the maximum allowed by the code.

Is that a correct statement?

 

[augie47]

Correct, but keep in mind texie's comment referencing the transformer inrush current. If you "downsize" the primary OCP too much you will have nuisance tripping from inrush current.

 

[wallypiper]

Thanks, Augie.

I'll wander over to the Calculations forum to see what I can find out about calculating transformer inrush. I'm aware of it but don't really know how large it is or how long it lasts. Some more research.

 

[kwired]

JMO, but a transformer is not a load, and neither is a VFD. I don't know of any rules that you have to size a circuit supplying a transformer to max current of the transformer, VFD's - I personally thing the rule is pointless, and have violated that rule with no regrets. The load on the feeder or service does not know you have a 100kVA load connected to a 200kVA transformer, it only sees the 100 kVA load.

 

[wallypiper]

By the same token, a motor is not a load. The device it is connected to is the load. If I connect a 50 HP motor to a centrifugal pump that can only generate 20 brake HP of load, then that's all the load that circuit is going to see. But the NEC says I have to size the branch circuit serving that pump based on the very coservative FLA tables in the code and the nameplate HP on the motor. It would be much easier for me to tweak a VFD to get more load out of the driven machine that it would be to tweak a pump connected to a fixed speed motor. Refrigeration equipment, on the other hand, gets a "free pass" in the code with branch circuits based on RLA. RLA is a value that the equipment manufacturer calculates based on the maximum load he is going to let a machine take on. I might install a chiller with an RLA of 250 amps but I could actually force it to draw much more by manipulating the limits in the built in controls. The only thing that stops that chiller from pulling more than RLA in many cases is a programmed value for maximum amps and/or maximum head pressure. With the right password, those limits can be changed. York sells a chiller with multiple small scroll compressors on it. You can order it with or without external compressor motor overload protection. With it, the nameplate MCA is much lower. The compressor motors don't change, just the current limiting controls in the compressor starter circuits. For the largest unit in the line, the MCA with overloads is 208 and without them it's 308.

 

[kwired]

By the same token, a motor is not a load. The device it is connected to is the load. If I connect a 50 HP motor to a centrifugal pump that can only generate 20 brake HP of load, then that's all the load that circuit is going to see. But the NEC says I have to size the branch circuit serving that pump based on the very coservative FLA tables in the code and the nameplate HP on the motor. It would be much easier for me to tweak a VFD to get more load out of the driven machine that it would be to tweak a pump connected to a fixed speed motor. Refrigeration equipment, on the other hand, gets a "free pass" in the code with branch circuits based on RLA. RLA is a value that the equipment manufacturer calculates based on the maximum load he is going to let a machine take on. I might install a chiller with an RLA of 250 amps but I could actually force it to draw much more by manipulating the limits in the built in controls. The only thing that stops that chiller from pulling more than RLA in many cases is a programmed value for maximum amps and/or maximum head pressure. With the right password, those limits can be changed. York sells a chiller with multiple small scroll compressors on it. You can order it with or without external compressor motor overload protection. With it, the nameplate MCA is much lower. The compressor motors don't change, just the current limiting controls in the compressor starter circuits. For the largest unit in the line, the MCA with overloads is 208 and without them it's 308.

I guess you could look at it that way. There also is no minimum overcurrent protection for motors. Most of the time I have lower rated circuit breakers for a motor circuit than what the maximum allowed is.

Take a 10HP 480 volt 3 phase motor. According to NEC I can go 2.5 times current which would allow me to use a 35 amp breaker. I have put many 10HP motors on 20 amp breakers and even seen some on 15 amp that do not trip when starting. Square D tables usually recommend 25 amp for this motor. You still have to size conductors for current listed in tables at end of section 430, which I guess is a little like the situation with the transformer in the OP. So I don't know what the right anwer is. What about motors that are basically a 10 HP motor, but they relabel them as open - air over and with a 12-15 HP rating? I understand an increase in ampacity for supply conductors - but starting current is still going to be same if it is physically same windings as the 10 HP and you are trying to accelerate the same load.

 

[wallypiper]

There is minimum circuit ampacity though based on the FLA tables in the code. So while "undersized" OCP may be technically allowed you have to use a breaker or disconnect that can handle the minimum allowed wire size.

The real point here I guess is that the NEC is not perfect and some of the details seem arbitrary and oversimplified when examined very closely. It is a life safety and fire prevention code. One could always plead special cases to the AHJ if the cost penalty for strict compliance is significant and there is a solid engineering basis for an exception.

 

[kwired]

One could always plead special cases to the AHJ if the cost penalty for strict compliance is significant and there is a solid engineering basis for an exception.

Good answer. You might want to discuss it with AHJ before installing it though and not after he finds it.

In the absence of a (legally) required permit or inspection the installer and/or designer becomes the AHJ. That doesn't give them a free ticket to just do what they please, it means even more liability rests on them if there is ever a problem.