Underground duct bank feeding MCC

[cmoore7]

We are feeding two new 1600A MCC's via an underground bus from an existing substation (with its switchgear fed by a Delta-Delta transformer).

I have three questions relating to this, as I haven't been able to find the answers in the CSA code book. I would really appreciate your input on any of them.

1) Is there any documentation (CSA or NEC) stating the separation between duct banks such that further derating is not required?. We were planning on having two - six
conduit banks side by side, where five of them contained 3c750 mcm Teck90 (XLPE).

x x x x x x
x x 0 x x 0

2) The short question is whether Canadian or US inspectors have required the ground wire on a 3c750 mcm cable to be terminated on one end only (in this situation), as the MCC's will have a connection to the ground grid?

I have been told by an electrical inspector that if single conductor cables were used, I wouldn't have to add any ground wires. I suppose that the ground wire on a 3c750 mcm cable could be terminated on one end only, since I understand that the sheath of these cables should be terminated on the supply side only.

Originally I would have thought that the safer option was to terminate both ends of the ground wire, regardless of the possibility of different potentials (between the two ground grid connections), to deal with the rare case of one end having its ground grid connection severed.

3) There is a 400 ft building separation between the substation switchgear and MCC room, so I wanted to confirm that I am correct that a separate fusible disconnect is NOT required for the MCC's? If the customer asks for it anyway, I don't believe I need it to have ground fault protection, as the switchgear has it.

Thanks,

Chris

 

[dpc]

I know nothing about CSA, but the NEC does NOT require derating for duct bank heating for circuit below 1000V. It is required above 1000V. In the 1980s, the NEC attempted to apply the same derating requirement for low voltage circuits, but there was a revolt among electricians and the requirement was removed. It doesn't make engineering sense, but on the other hand, there are few documented cases where ignoring this has ever caused a problem, probably because the NEC ampacities are conservative to start with.

As far as grounding, I think grounding and bonding is required at both ends per NEC - see Article 250.32. You also must have a grounding electrode at each separate building. There is an exception that allows a grounded circuit conductor to be used for grounding, but your metal-clad cable has ground wires so you have to use them, and you don't have a grounded circuit conductor anyway.

Regarding disconnect at the separate building, the NEC basically requires this, see Article 225.31, but there are exceptions that might apply. Personally, I would put in some type of disconnect. When the disconnect is 400 feet away and it's raining or snowing, it becomes very tempting to work on energized equipment to avoid that long walk to the disconnect and back. If the owner wants to take it out, then it's on him.

 

[davidbeach]

The IEEE Brown Book, chapter 13, has some information on doing the derating. If your installation used all 12 ducts, the cables would be good for .522 of the single cable rating.

 

[cmoore7]

Thank you for your input, dpc and DavidBeach.

I have a few questions/comments on your posts.

DPC
1) Do you happen to know the NEC section that states derating is not required for any distance between <1000V duct banks? I only found the NEC Table (B-310-2) that relates to the individual bank arrangements and the conductor ratings for them.

2) The dual point connection for the ground wire was something I couldn't prove to the customer, so I will at least refer to the NEC article you gave me if I can't find it in the CSA code book. Even though there are separate grounding electrode connections for each building, I am surprised that if they pulled 3 individual conductors (instead of a 3-phase cable), the inspector implied that they did not need a ground wire.

I know that CSA mentions grounding the sheath on the supply side for single conductor armoured cable, so I extended that to include three conductor cable too.

-------
David Beach - Unfortunately, I don't have the IEEE Brown Book, but you mentioned 12 ducts instead of 2 x 6 ducts. I figured it was worthwhile separating the ducts to reduce the derating factor and wondered what that minimum spacing was. CSA Table D14B has 0.62 of the single cable rating when all six ducts are in use.

 

[dpc]

Actually, the range goes up to 2000V, not 1000V as I had in my first post.

There's nothing in the NEC that says you DON'T have to de-rate - the code doesn't work that way. There is just nothing for conductors up to 2000V that says you DO have to de-rate (for duct bank heating - there are other conditions that do require de-rating). The basic requirements are given in Article 310.15 for conductors 0-2000V. There is no requirement for derating given. If you look at Article 310.60 for conductors rated 2001 -35,000 V, you will find this requirement, along with other requirements that don't apply to lower voltages.

If your engineering judgement is that you should de-rate the cables, then that's what you should require.

 

[davidbeach]

The Brown Book number for 2 rows of three ducts is .641 for 750kcmil. I don't know how far apart the two duct banks would have to be to get that number. If all of the calculations are done, you will get a slightly higher number for using only 5 of the 6 positions, and slightly higher yet if you leave the spare as the middle of the bottom row rather than in the corner. Lots of calculations if you want a real answer. BTW, the annex B tables in the NEC are information only, not part of the code requirements.

 

[waross]

Re grounding;
Three conductor cables. Use the internal ground wire and connect both ends. The internal ground in teck cable is usually sufficient for equipment grounding.
Single conductor cable with a concentric ground, or a sheath or both, see Canadian Electrical Code appendix "B" notes on rules. Section 4-
Basically ground the supply end and insulate the load end. A seperate ground wire of the correct size is run for equipment grounding. The concentric ground wires will behave the same as a sheath in regards to sheath voltages and currents.
respectfully

 

[cmoore7]

dpc - Unfortunately, the CSA tables for underground duct banks refer to 0-5 kV. I noticed that there was software called "AmpCalc" for calculating for individual bank arrangements, but I'm not sure if it deals with multiple sets.

As for my poorly worded question, the code book would be a wee bit larger if it mentioned all things you DON'T have to do.

 

[dpc]

AmpCalc can be used - it doesn't care if the conductors are part of the same circuit or not - just gives a derating factor based on the geometry and heat transfer coefficients. It's based on the Neher-McGrath equations, just like the NEC tables.

If you might be using Okonite C-L-X, your local sales engineer can probably run the calc for you. AmpCalc is pretty expensive for what it does, IMHO, although it would be nice to have sometimes.

For multiple runs of 3/C armored cable/phase, you have to be a little careful about ground wire sizing, at least in the NEC world. In general, the NEC requires a full-sized ground conductor be run in each conduit or cable when you have multiple conductors per phase. This has been a problem in the past when inspectors would require a large ground conductor in the MC cable than was provided. (e.g. if you run 2 -250 kcmil/phase, then each ground wire should be sized for a 500 kcmil feeder - but the MC cable grounds would both be sized for 250 kcmil feeder). I seem to recall hearing from Okonite that this impasse had been somehow resolved, but I don't see anything in the 2005 NEC. Since the C-L-X sheath also qualifies as a grounding conductor, maybe they figure this is adequate. But you might want to check with local authority before pulling it in.

 

[waross]

I have been told by an electrical inspector that if single conductor cables were used, I wouldn't have to add any ground wires. I suppose that the ground wire on a 3c750 mcm cable could be terminated on one end only, since I understand that the sheath of these cables should be terminated on the supply side only.

This may be a misunderstanding, or the inspector may have accepted the existing or alternate connection to the ground grid as an adequate grounding measure for the equipmenmt.
If it was a general statment about single phase teck cable with a concentric ground then the statment is in errror.
In my experience, inspectors don't make this kind of error, so I must assume that he deemed the connection to the grid as an adequate ground.
Using 3c TECK cable I would ground the ground wire at both ends and use normal grounding connectors to ground the sheath at both ends. If the installation is such that connectors are not needed at the cable ends, I would just insulate the sheath/armour but connect the ground conductor at both ends.

I know that CSA mentions grounding the sheath on the supply side for single conductor armoured cable, so I extended that to include three conductor cable too.

The problems with sheath induction and voltages and currents are mainly the result of unbalanced magnetic and electrical fields surrounded by a conductor (Ie; the sheath or a spiral armour or the concentric ground conductors or a metalic conduit).
In a three conductor cable these forces tend to cancel each other and do not induce sheath voltages, and no derating or insulation from ground at one end, or other special measures are required for three conductor cables in regards to sheath voltages or currents.
Single conductor, ground the sheath ay one end.
Three conductor cable. No special treatment. Connect the grounds at both ends.
respectfully

 

[cmoore7]

Thanks for your comments, waross. As I just joined Eng-Tips yesterday, I've been impressed by the number of times you, dpc, DavidBeach, and others have spent answering questions in these forums.

The inspectors comment about grounding wires not being required if we used single conductor cable came after mentioning that 'engineers often included them anyway'. I presume he believed the ground grid connection at the MCC was sufficient. It was the customer that mentioned terminating only one end of the ground wires on a three conductor cable, and being skeptical, I thought that I would see if someone else here believed the potential difference between two ground connections was indeed a concern.

As for the sheath termination, I did know that it was compulsory to reduce the sheath currents generated by single conductor armoured cable, but guess there isn't too much value in treating the 3c sheath as a noise shield in a power cable.

 

[waross]

dpc If this is an installation under Canadian codes, One large ground cable is not required, as long as the sum of the ampacities of the grounding conductors in the individual cables is adequate. The ground cables in parallel runs of 3c Teck are sufficient. If there are any exceptions to this, or particular configurations that I'm not aware of I'm open to correction.
respectfully

 

[dpc]

As I mentioned before, I know nothing about Canadian codes. However, if there is a ground fault in one of the cables, nearly all of the fault current will return in the ground wire in that cable. The other ground wires will not carry any appreciable current. (This is especially true for steel conduit, which is not an issue here.) That is why the NEC requires that each ground wire be sized as if it were the only one. The ground wiring sizing in the NEC is based on having enough cross-section to allow the upstream protective device to clear the ground fault without damaging the ground wire.

With MC cable, you also have the sheath in parallel with the internal ground wires, so I suspect the standard MC cable ground wire size is, in fact, probably big enough if the sheath is properly bonded at each end. But this NEC requirement has been a problem for MC cable in the past.

 

[elecsun]

Querry for the experts: How much short citcuit current that delta-delta transformer can supply during the fault to decide the ground wires sizing?