Electrical Duct Bank Calcs

1. The Rho values in Table B.310-. are for the soil.

2. Both the Table 9 resistance and the Table B-310-7 ampacities are based on 75°C conductor temperature, so the heat of the duct bank is taken into account. If you were going to use ampacity tables (like B.310.1) that consider higher conductor temperatures, and were going to load the conductors to the maximum limit, you would have to adjust the resistance when calculating voltage drop.

I would use rho = 90 as a starting point. The code book says that is good for 90% of USA. You did not indicate the system voltage, ampacity, and number of circuits, but assuming you are at 480 VAC, then table B.310-7 includes different colums for derateing based on the number of circuits. Use ambient derating factors at the bottom of the table, also.

If medium voltage, go to B.310-77 or somewhere close by. If medium voltage, you also need to derate for induced shield currents. Don't forget that the conductor jacket temperature rating cannot exceed the temperature rating of the conduit (usually 90 deg C, so MV-105 is out for PVC).

You don't exactly "derate" for voltage drop. To allow for Vd:
1. Figure a minimum wire size for a maximum Vd (I use 3% for feeders) at full load amps;
2. Divide the ampacity of the wire determined in step 1 by each of the derating factors in turn to get a final ampacity requirement;
3. Find the required wire size from the value in the proper column in the proper table.

To figure Vd for a given wire size, use a worst-case power factor (I use 0.7) and the corresponding X and R values from table 9.

William

The effect of PVC in the thermal model is small and most cases are neglected.

The ampacity calculation is based primarily in the concrete and soil thermal resistivity.

[sub]Enclose see a few typical values of Thermal Resistivity of various materials:
- Plastic Ducts 480 oC cm/watt
- Concrete 60-to 85-oC cm/watt
- Average common US soil 90 oC cm/what
- Sand 120 oC cm/watt
- Thermal fill 60 oC cm/watt
- Insulation + extruded shield 350 oC cm/watt [/sub]

See a typical ampacity calc in a UG ductbank:

Thanks to all for the helpful replies.

There are two types of voltages being used; 480V and 11.5kV.
Weh3, good point about PVC conduit being rated for 90C but the local utility comany has requested to use PVC conduit (type EB) with thier 11.5kV duct bank.

For the 480V duct bank, there will be (1) row of (5) 4"C and (2) rows of (5) 4'C all with (4) 500MCM copper conductors. None of the details in the NEC list these setups. How do you determine which column to use in table B-310-7 if the set up of the duct bank does not match any of the NEC details?

If you are using 90C wire, is there a table or ampacity factor to apply to table B-310-7?

Question about ductbank depth: In a portion of the run, the duct bank will be installed 35'-0" below grade level so that the duct bank can be installed under a building. When you take 6% of 35' you get a deratinfg factor of 2.1. Doesn't sound right.

MV-105 has a conductor temperature limit of 105°C. The temperature of the outside of the jacket, in contact with the conduit, will be considerly lower. PVC is not necessarily out.

In Annex B of National Electrical Code there is a 15 duct bank that uses 2,000 KCM conductors. See figure B.310.5 . Since for 6 or more ducts the main thermal bottleneck you can figure 500 KCM at 168.75 amps per cable for natural cooling. For 15 ducts you MUST use concrete encasement because that is the only kind of backfill that will not break the ducts and you need the thermal conductivity of the concrete. This would give you 2,531.25 amps total. This ONLY applies to the part that has no more than 4 feet of dirt on top.

Because of the deep portion and because of the puny capacity of each cable, you need artificial cooling. Your ducts are theoretically big enough to contain four 0.625 inch O.D. polybutylene tubes that form 2 cooling circuits in each conduit. You will need chillers at each end and you should use a total of 6 cooling circuits, 2 for each row with a total 12 chillers. Your chillers should be hooked up with electrical plugs and hydraulic quick connects so that you can swap in a spare chiller.

Polybutylene tubing is very reliable as long as the connector uses an elestomeric gland such a a nylon or rubber gland. DO NOT USE HOSE CLAMPS!!!!!!!!!!!!!! Hose clamps will break the tubing.

B310.60.C.(2)b states that for deeper portions an ampacity derating factor of 6% per foot shall be permitted. Based on this, cvirgil calculates a derating factor of 2.1 which indeed doesn't sound right. Put in another way, if a cable is installed in a duct 16.6-ft deep it will have no rating whatsoever therefore not even a 1 Amp load would be permitted using this concept; not realistic either.

I'm certainly not sure what the NEC means but it could be a "compound interest" type of calculation where the figure (1-6%)or 0.94 is raised to the power of the depth. For example, the 35' case would be 0.94^35 = 11.5% deratng; at least something realistic.

I might mention that certain European cable companies supply tables for derating due to depth and typically the factor becomes larger as depth increases up to 10 meters at which point the factor becomes constant, about 10-15%. This too seems realistic

Finally, what is the meaning of the NEC wording "shall be permitted"? why isn't it stated "shall be required"? Is it due to the fact that an earlier paragraph allows a ampacity determination based upon "engineering supervision"? In this case the rather simple formula stated in the NEC just before Para. 310.60 could be used which has no parameter for depth - which doesn't seem right either.

I would appreciate any comments whatsoever on this issue and a way to present a calculation to an inspector who could understand and accept a reasonable derating factor.

CharlieD

The rather simple formula in 310.15(C) is a little deceiving. The trick is in determining RCA, the effective thermal resistance between the conductor and surrounding ambient. This effective resistance is the result of several complicated formulas that take depth into account.

A good book on the subject is Rating of Electric Power Cables, George J. Anders, McGraw-Hill/IEEE Press, 1997.

The NEC allows the Neher-McGrath calculation, but gives little guidance that would make it generally useful, beyond a couple of examples.

If you were required to get the ampacity just right, then it ought to be cited as a requirement, and data for every kind of soil, insulation type, and cable configuration would be provided. As it is, I have never had sufficient information, nor time to get it, to use N-M, so engineering judgment has had to suffice.

In a recent case, however, it might have proved helpful. There were so many constraints that, had a single one been the least bit more restrictive, I doubt I would have been able to find a solution. I wonder, whether the N-M formula would have been more restrictive, or less?

E.g., I was restricted to MV-90 because of PVC conduit, to a maximum cable diameter because of the 40% fill, and to a minimum conductor size because of voltage drop and pulling tension, etc., etc.

Let me put it this way: if economics had not been a consideration, then anything would have been possible.

William

I need to edit my post a little. For 15 ducts the surrounding soil and concrete is the thermal bottleneck. At this capacity forced cooling is a viable option. At 35 feet deep and any capacity forced cooling is mandantory. You cannot fool the laws of thermodynamics.

Forced cooling would also allow you to use fewer conductors which might actually cost less overall. You did not state how long the duct bank is.

Shall be permitted is both a design option and also a means to prevent inspectors from requiring overkill. Shall be required is a stronger wording for what is mandantory rather than optional.

Another option at this capacity is to use a packaged indoor substation that uses a dry transformer with an 11.5 KV primary. Dry transformer technology goes all the way up to 46,000 volts primary. This is how Coca-Cola did their third 480 volt service out at their Twinsburg, Ohio plant. For this option you would need a metal enclosed 15.5 KV primary switch with fuses and for conversational purposes it is a just a jigh voltage version of a 600 volts switch - more to it than that but that is the idea. You can get these substation with a compartment for utility secondary metering as well as a draw out main circuit breaker.

The way that you want to wire your 480 volt service is the hard way. Move the 11.5 KV primary transformer indoors and you will be much happier with the results.

Still another twist; what happens if the ductbank is below the water table as it may very well be at 35' or even 10' in certain areas?

I believe that a certain natural flow of water would normally occur plus a flow that would be brought about by convection currents. Any flow at all would carry heat away and the presence of water certainly would preclude the soil from drying out, so rho would remain constant. None of these factors would be taken into account by the NEC tabulations nor the abbreviated formula given.

Another off-the-wall thought; if 105-deg cable were used and temperatures approach that point, the water would boil and absorb heat, thus maintaining the temperature at slightly below the cables rating - so no derating at all!

But in general, what would the real ampacity be with water present in sufficient quantity and how would it be calculated?

As weh3 has so well stated; "I have never had sufficient information, nor time to get it, to use N-M, so engineering judgment has had to suffice". But how to convince an inspector is the question.

CharlieD