I have run into a situation where a utility had decided some time ago, for whatever reason, to replace there main feeder cables with shielded EPR when they need to be upgraded.  Their old standard was concentric neutral cable.  For example, a feeder getaway would be upgraded from 750 Al 15kv XLPE CN to 1000 Al 15kv EPR w/8 mil cu shield.

My problem with this is that they are not pulling in a separate neutral with the shielded cables.  My contention is that shielded cable is not meant to be a direct replacement for concentric neutral and that they are effectively eliminating the system neutral.  The shields main intention is to evenly distribute the electric field about the conductor, a concentric neutral on the other hand has a dual purpose as being a neutral and to distribute the field.

The shield is 8 mil x 5.25" /phase.  My calculations says that this equates to about a #7 cu neutral.  In contrast to a 1000 mcm Al CN cable which comes with 20-#10 cu /phase for a neutral which would equate to about 622 mcm of copper.

I also realize the implications of circulating current in the shield and the safety hazards that can be created.

Has anyone else run across this question/problem?  I can find nothing in writing to state that the shield is not a neutral replacement.  My assumption is that it is not in writing because everyone knows it to be.

Your feedback would be appreciated.  Thanks.

Concentric neutral provides a definite purpose beyond shielding as you have stated.  But it is not always required.  If you are on a 3 phase/3 wire system (no phase to neutral loads) you don't need CN, you only need to ensure your shield can carry ground faults without damage to the cable.

Thanks gordonl,
I apologize.  I should have stated the the system that I am referring to is a grounded Y 12470/7200.  It would be a waste of money to buy CN cable for a delta feeder.

NESC 354D3.

If you added unbalanced load current to the shield current (in addition to induced circulating current) in calculating the cable ampacity, you could probably prove the need for a separate neutral.  A quick modification of calcs for 500 mm² Al 15 kV tape shielded cable indicates a reduction in ampacity from 400A to 300A with a 10% neutral current.

354D3 may not apply since it deals with random separation and direct buried cables. Try D4 if in plastic duct and random separation does apply.

For a more general rule, try 096C.

Stevenal,
 my system is in plastic duct.  So, if appears to me that all of section 35 will not apply as it is called "Direct-Buried Cable".  Correct?  
Section 96 is very general and does have some bearing, I'll agree.  I am continuing to search the NESC for more.

jghrist,
 you are on the right track with your thinking.  One of my arguments is that they are spending alot of money on these cables and then installing them in a way as to have to derate them significantly.  By adding neutral and circulating currents to a piece of foil heats the cables significantly.  Thus leading to the decreases in ampacity that you have demonstrated.  In my mind, that is not money very well spent.

Thanks

Working against the shield loss argument is the fact that if there is a phase imbalance, there will be less current in at least one of the phase conductors, and less heat generated in that phase conductor.  This is why the NEC does not require the neutral to be counted when using ampacity tables.  My quick calculation did not take any reduction in phase current into account, but I suspect that because the resistance of the shield is so much higher that that of the conductor, there will be a significant reduction in ampacity.

Let me correct my first post as I had done some quick calculating and then got interrupted.  The shield is equivalent to somewhere between a 2/0 and 3/0 Cu Oops.  Bit of a difference, but still way below the rating of a CN.
This is adequate under the right conditions to functions as a neutral but at the cost of ampacity of the phase conductors.

2/0 to 3/0 Cu equivalent seems too high for an 8 mil tape shield.  I think you were closer at first.  You need to consider the increase in effective length of the shield because of the helical winding.  If the shield is LC type, the resistance would be lower, but still about the equivalent of #4 Cu not 2/0 or 3/0.  What type of shield do you have?

354D4 would seem to be mis-filed (section 35), but looks like it would apply to cables in duct with random separation.

Correct me if my calculation is wrong for the amount off neutral.
The cable is 1000 mcm Al 15kv 100% LC shield.
This would give me a diameter of about 5.12" at the shield.
5.12 = 5120 mils.  5120 x 8 mils x 3 phases = 122,899 sq mils.  122,899 sq mils/0.7854 = 156,480 cir mils.
2/0 = 133,100 cm    3/0 = 167,800 cm
This is how I came to the equivalent copper size.

354D, It seems to me a little strange that the neutral and its size is really only discussed in the Direct Buried section and not so much in the Cables in Underground Structures.  I couldn't find much else in any other chapter.

I forgot to multiply by 3.

I also get between 2/0 and 3/0 equivalent for all 3 cables.