Wire Sizing

[actionjoe]

We ran 3 sets of 500MCM Copper for a 1200 Amp Main service underground in PVC. The conduits come up into a 1200 Amp Main CT Cabinet with vents at bottom and top,lugs are rated for 75 degrees. Is this OK or do we need to change the wire to 600MCM or the breaker to 1000 amp?

Thanks Joe Orsini

 

[infinity]

Re: Wire Sizing

Three sets of 500MCM are only good for 1140 amps(380x3). Once you hit 801 amps or higher the conductors must be rated to their full ampacity. 800 amps and below allow you to go up to the next standard size. Take a look at 240.4(C) and 240.6.

[ July 09, 2005, 12:35 PM: Message edited by: infinity ]

 

[bob]

Re: Wire Sizing

Is this OK or do we need to change the wire to 600MCM or the breaker to 1000 amp?

Joe
You need to caculate the load first and then determine the conductor size and breaker size. If the caculated load requires a 1200 amp breaker then you need to install 600 kcm. You don't adjust the breaker to match the wire size.
The designer may have wanted to add spare capacity for future growth.

 

[charlie b]

Re: Wire Sizing

Originally posted by infinity: Once you hit 801 amps or higher the conductors must be rated to their full ampacity.

The question is about conductor ampacity, not breaker settings. The conductors must always have an ampacity equal to or greater than the calculated load, as Bob has said. What I think infinity means here is that above 800 amps, you can?t use a breaker with a trip setting that is higher than the full ampacity of the conductors.

 

[tx2step]

Re: Wire Sizing

The service entrance conductors must always be sized equal to or exceeding the calculated or actual load, whichever is higher. That being said, another approach would be:

If this was a 1200A MLO main panel instead of a 1200A MCB main panel, then the 3X500's might be OK if they equal or exceed the calculated or actual load.

If you had a 1200A MLO main panel, you could have up to six (6) circuit breakers in it to serve various distribution or branch panels or other loads.

The trip size of the 6 CB's could add up to more than 1200A total, as long as each of the feeder circuit conductors & CB's were properly sized together.

In other words, you could have a 1200A MLO main panel with (for example) 4 each 200A feeders and 2 each 400A feeders. This arrangement is OK as long as each of the feeders are adequately sized and the total load on the main panel doesn't exceed the ampacity of the service entrance conductors. In essence, some of the 6 feeders are sized heavier than their load actually requires.

While this arrangement is code compliant, it places more burden on the EC's that work on the installation later. They must be certain that they don't inadvertently overload the service entrance. The safer (more idiot proof) way is to have a MCB that is sized to match the service entrance conductors -- that would prevent an inadvertent overload in the future. It's pretty easy for stuff to be added later such that the actual load exceeds the original calculated load.

But the MLO main panel with the 6-circuit rule is pretty common.

Edited to correct my lousy wording and spelling.

[ July 11, 2005, 01:23 PM: Message edited by: tx2step ]

 

[actionjoe]

Re: Wire Sizing

Thanks guys, I believe infinity is right and I either have to change 1250 feet of 500 mcm to 600 mcm or change the main breaker to 1000 amp.

Thanks again

 

[tx2step]

Re: Wire Sizing

Originally posted by actionjoe:
I believe infinity is right and I either have to change 1250 feet of 500 mcm to 600 mcm or change the main breaker to 1000 amp.

I agree -- if you are going to stay with a Main Circuit Breaker configuration, then you will have to either upsize to #600, or add another parallel set of #500 (did you install a spare conduit?), or reduce the main CB trip to 1000 amps.

You didn't mention the voltage involved -- if it is 480Y/277V then the 1000A Main CB will have to have ground fault protection (230.95 & 215.10).

That's one of the reasons why MLO Main Panel configurations are common -- it usually keeps the CBs below the 1000A threshold.

 

[dvmont]

Re: Wire Sizing

Joe,

Be careful here. Since the conductors are underground, you have to derate them. The amount of the derating depends on the thermal charactaristics of the dirt and the configuration of the conductors in the duct bank.

I have seen 500MCM conductors burn up at about 200 amps per phase when installed underground.

 

[iwire]

Re: Wire Sizing

Originally posted by dvmont:
Joe,

Be careful here. Since the conductors are underground, you have to derate them.

Sure would like to see an NEC section to back that up for conductors under 600 volts.

[ July 16, 2005, 07:41 PM: Message edited by: iwire ]

 

[hey_poolboy]

Re: Wire Sizing

Ok, bear with me here,(I'm just trying to learn a little bit), but is it possible that dvmont is thinking of derating as it would apply to UG ducts.

Just for kicks, how screwed up of an installation must one have to burn up 500 w/ only 200 amps?

 

[tx2step]

Re: Wire Sizing

Probably had 4 conduits with all of A phase in one, all of B phase in 2nd, all of C phase in 3rd and the neutrals in the 4th?

In metal pipe that can cause some real interesting heating problems, whether underground or overhead.

 

[dvmont]

Re: Wire Sizing

tx2step

Actually, the installation was even more screwed up than that. It was supposedly a 4000A circuit. There were 6 conductors per conduit (2 A phase, 2 B phase, 2 C phase). I believe there were 6 conduits - three on top and three below.

If you look at the derating in the appendices that aren't a part of the code but are included for reference, you arrive at an extremely low ampacity.

The inspector decided the circuit was fine and signed it off. When I asked him, he said the cables should be good for 380A each.

As load was added at the plant, the circuits burned up at about 200A. That was only a little bit lower than the ampacity you arrive at when you look at the appendix that isn't a part of the code.

 

[iwire]

Re: Wire Sizing

dvmont

That is a very common installation that usually works fine, I am having a hard time believing that 500 Kcmil copper in separate conduits 'burned up' with 200 amps of load.

Are you sure there were not other contributing factors?

Water, poor connections, etc.

I have seen 500MCM conductors burn up at about 200 amps per phase when installed underground.

It was supposedly a 4000A circuit. There were 6 conductors per conduit (2 A phase, 2 B phase, 2 C phase). I believe there were 6 conduits - three on top and three below.

I come up with a capacity of 4128 amps.

 

[iwire]

Re: Wire Sizing

I say common but actually we would normally run 10 sets of 600 Kcmil copper for a 4000 amp service.

We do not normally put more than 3 current carrying conductors per service or feeder raceway.

[ July 17, 2005, 09:06 AM: Message edited by: iwire ]

 

[james wuebker]

Re: Wire Sizing

Don't forget about the breaker that should be a GFCI.
Jim

 

[tx2step]

Re: Wire Sizing

Originally posted by dvmont:
Actually, the installation was even more screwed up than that. It was supposedly a 4000A circuit. There were 6 conductors per conduit (2 A phase, 2 B phase, 2 C phase). I believe there were 6 conduits - three on top and three below.

6 current carrying conductors in one conduit would have to be derated to 80% per table 310.15(B)(2)(a)

Originally posted by dvmont:
If you look at the derating in the appendices that aren't a part of the code but are included for reference, you arrive at an extremely low ampacity.

I presume you mean "Annex B" -- I've never had an occasion to use that Annex.

Originally posted by dvmont:
The inspector decided the circuit was fine and signed it off. When I asked him, he said the cables should be good for 380A each.

They should have been derated to 380 X .80 = 304 A capacity per wire.

Originally posted by dvmont:
As load was added at the plant, the circuits burned up at about 200A. That was only a little bit lower than the ampacity you arrive at when you look at the appendix that isn't a part of the code.

I still can't imagine that they burned up with a load of only 200A per conductor.

If there really was much of a problem with needing to derate conductors installed in underground duct banks, I can't imagine that it wouldn't be directly addressed in section 310. Underground duct banks (direct buried or concrete encased) are the most common method of installing large service entrances. There must be a million of them installed that way, operating without problems.

There must be some other factor involved that made this installation unique. I still lean toward it being likely that they hooked up 12 conductors in 2 pipes to one phase (repeated for each phase). This would be easy for an inspector (or anyone else) to miss -- he might look directly at it and it still not dawn on him that the wires were hooked up wrong.

As I see it, the installation that you describe would have been rated:
380 X .80 X 12 = 3,648 A total capacity. (not 4000 A)

What am I missing here, or what other factors made such a difference in this particular instance?

 

[don_resqcapt19]

Re: Wire Sizing

The only reason that we don't burn up underground duct banks more often is the result of Article 220 load calculations. They as so excessive that the actual feeder or service conductor is often 50% or less that the calculated load. This results in a huge safety factor that prevents the burn up of underground duct banks, however in many cases Article 220 is not used for industrial load calculations. The engineers doing these calculations use much more realistic numbers resulting in a calculation that is very close to matching the actual load. When conductors are installed in underground duct banks using the Tables in Article 310, as opposed to those in the Annex, the result is often a burnt up set of conductors.
If you go back a few code cycles and look at the ROP and ROC you will find this was a very hotly debated subject. The ampacity tables found in the annex were actually placed in the main code body for one cycle with an effective date three years later. In the next code cycle the debate continued and these tables were placed in the annex. I think that this was in 90 or 93 but am not sure.
Don

 

[dvmont]

Re: Wire Sizing

There are several ways to look at it.

1. If you do it by the code, you're legal. You might not be right, but you are legal. Electricity doesn't care much about what the lawyers say.

2. There is software available that computes the ampacity of duct banks. The software we used calculated an ampacity of less than 2000A for the 4000A circuit.

3. I worked out a Neher-McGrath calculation based on some assumptions about the heat transfer charactaristics of the installation and came up with an ampacity of about 220A per conductor - 220 x 12 is about 2500A.

4. I looked in the code annexes that aren't a part of the code and came up with an ampacity of about 220A per conductor - 220 x 12 is about 2500A.

5. No one knows exactly what the current was when the circuit burned up. But we do know it was a whole lot less than the code allowed ampacity. The best guess I heard was that it was about 1900A.

The fact is, the circuit burned up. The code tables can drop dead for all I care. I don't want to put in circuits that burn up.

There is a whole body of work about estimating duct bank ampacities. The general conclusion of the work seems to be "don't believe you're sizing stuff for free air."

I spoke with the electrical inspector about why the stuff in the appendix was taken out of the code. His answer was that "no one understands it."

Maybe they don't understand it. But underground circuits still burn up a whole lot faster than circuits in air.

 

[dvmont]

Re: Wire Sizing

tx2step

I looked at the installation after new conductors were put in. They were hooked up 2Aphase,2Bphase,and 2Cphase per conduit.

We dug down and measured the temperature at the outside of the conduits. The temperatures were about 195 deg Farenheit (90.5 deg Centigrade) on the outside of the conduits. The current was about 1950A in the circuit.

The conduits were improperly installed. No spacers were used. The conduits were laying against each other.

I'm fairly sure that the circuit would have been good for about 2400-2500A if the the conduits were properly spaced.

 

[tx2step]

Re: Wire Sizing

Hmmm... Sounds like I'll have to study Annex B.

So what do you recommend when we are asked by a commercial or industrial customer to install a specific capacity new service entrance or large feeder using an UG duct bank?

It's not that uncommon for someone to be building a light industrial building and to tell us that they want it to have, say, a 1200 A main service.

They can't give us all of the load data on their machinery -- they haven't settled on exactly what they are going to buy & install. The machinery will be selected and installed after the building is built -- we're just to install all of the provisions in the initial stage. But they do know that they want the service to be 1200 A.

There is no PE involved.

We did one like this not long ago, and installed 4 parallel sets of #350 Cu in 3" PVC.

I think it would have been hard to sell them on the idea of doubling the conductor size to maybe 6 sets of #600. Or to convince them that 4 parallel sets of #350 would not really give them over 1200 A capacity.

In situations like this, what do you suggest that we do?