Cable substitutions to save money

Alphonse53

Member
Location
Louisiana US
Occupation
Retired I&E
I've seen threads posted on this forum discussing the use of smaller higher temperature rated cable in lieu of larger lower temperature rated cable to save installation cost. Does the NEC offer an engineering calculation basis that allows such practice, including any heat loss cost calculations?
 
I've seen threads posted on this forum discussing the use of smaller higher temperature rated cable in lieu of larger lower temperature rated cable to save installation cost. Does the NEC offer an engineering calculation basis that allows such practice, including any heat loss cost calculations?
Remember that the client will be paying for the additional I²R losses for the life of the installation. If the load is high, those costs will outweigh the savings over the life of the installation, especially for industrial installations were the loads often run 24/7.
 
Remember that the client will be paying for the additional I²R losses for the life of the installation. If the load is high, those costs will outweigh the savings over the life of the installation, especially for industrial installations were the loads often run 24/7.
That's exactly what I'm referring to. I guess the question is whether there's a program available that can account for the plethora of variables to find which route is more cost effective, or whether we gonna have to wait for AI?
 
The only people who are allowed to make cable substitutions is the utility who will gladly land your new 200A service on the existing 4AWG aluminum triplex running overhead.

Honestly though, real world terminals are rated at 75c. I have yet to find a counterexample even though I am always searching for one. There are various code sections that demand the 75c table is used in this and that circumstance. Voltage drop becomes a problem quickly if you size things to the 90c table, especially at less than 277v to ground. In the end using high temp rated wire doesn't actually save money.

To your question about the code providing calculations that allow it, no the code does not. If what you mean is does the code allow it under engineering supervision; no, not to my knowledge. There are specific sections that allow wire to be smaller than 310.16 in limited use cases, notably 240.5(B)(2), and some motor/AC unit rules will override 240.4(D), but all in all it is what it is for a reason.
 
I've seen threads posted on this forum discussing the use of smaller higher temperature rated cable in lieu of larger lower temperature rated cable to save installation cost. Does the NEC offer an engineering calculation basis that allows such practice, including any heat loss cost calculations?
310.15; used the Neher-McGrath equation to allow use of 18AWG 260C nickel plated wire to replace 10 AWG in a military application that had NEC as part of the system specification. Used only a few cm of the 18 AWG as a righ reliability fuse for a NWE application. Cost of < 1W loss not important in that application.
 
You don't need to wait for AI, you just need to craft a spreadsheet.

Start with run length and required ampacity.

Next determine the expected current and required voltage drop. (Expected as opposed to NEC calculated, use your understanding of what the real load will look like. Required voltage drop per energy codes or equipment requirements.)

This give you the parameters for the minimally acceptable cables. You could use Al or Cu, parallel runs, 75 or 90C...just list a bunch of the possibilities. Keep in mind that some will add additional termination costs, some will add conduit costs, some will be cheaper to pull...

Shop these different configurations from your supplier to get a wire cost.

Estimate the installation cost for each approach.

Finally for each configuration calculate the power lost in the wire (just current squared times resistance) at the expected current level, multiply by the number of hours in a year and the cost of electricity to get the annual running cost of the wire.

Not an easy spreadsheet to create, but doable in a few hours for a simple feeder.

I expect that the additional cost of terminating 90C wire negates any savings in the wire. Al wire will be cheaper for the same ampacity and have very similar losses for the same ampacity.

IMHO the place where you will likely see installation vs running cost trades is in number of parallel runs. Parallel installations give you more amps per circular mil, which means more loss per amp and higher running costs.
 
Thanks to all you guys. When looking at repurposing a building, redoing the wiring can make or break the deal. You guys had some great advise, and I think what Winnie had to offer fit the question best. Although time consuming to set up, Excel should help a lot in keeping things organized and doing the calculations. I hope all of us can keep and use everything this thread has to offer.
 
Note there are other cost saving strategies (some have been mentioned, just listing all for completeness).

1. Use AL conductors.

2. Winnie mentioned an oft forgotten one, which is using smaller parallel sets. For example take the ampacity of something large like 750 CU, which has a very low amps/cmil. Compare to the size needed for the same ampacity with two - three parallel sets.

3. Use a 100% rated breaker.

4. If allowed change wiring method to something like AL MC feeder cable.

5. Use wire from priority cable that is mismarked and smaller than what the printing says - oh wait, they won't credit you the cost difference once you discover it (😡) so you won't save any money except maybe a little from less conduit fill.
 
Remember that the client will be paying for the additional I²R losses for the life of the installation. If the load is high, those costs will outweigh the savings over the life of the installation, especially for industrial installations were the loads often run 24/7.
Exactly.
But conversely, the load that's heating the wires may be intermittent or seasonal.
Or, the waste heat may be useful.
Or, the actual load may be far less than the calculated load, and the smaller wire was fine.
Or, not.

There are too many ORs.
 
Exactly.
But conversely, the load that's heating the wires may be intermittent or seasonal.
Or, the waste heat may be useful.
Or, the actual load may be far less than the calculated load, and the smaller wire was fine.
Or, not.

There are too many ORs.
Seems to be a lot more trouble than it's worth, unless there are some big quantities involved.
 
Seems to be a lot more trouble than it's worth, unless there are some big quantities involved.
I often upsized the conductors for industrial motors one size, where that did not require an increase on the conduit size with a payback of less than two years. However these are known loads that run pretty much 24/7.
 
Not a temperature thing but I have always wondered why 600Y/347 is not used more in the lower 48. Seems like the bump from 480 to 600 would be a instant savings on wire size and lots of gear is rated for 600V.
Now since 2017 or whenever they bumped 600 up to 1000V you could even design to 1000Y/577, the PV people got the memo but I have not seen it catch on yet.
Data centers seemed to go the other way with 416Y/240.
 
Not a temperature thing but I have always wondered why 600Y/347 is not used more in the lower 48. Seems like the bump from 480 to 600 would be a instant savings on wire size and lots of gear is rated for 600V.
Now since 2017 or whenever they bumped 600 up to 1000V you could even design to 1000Y/577, the PV people got the memo but I have not seen it catch on yet.
Data centers seemed to go the other way with 416Y/240.
Because the computers use 240.
 
I often upsized the conductors for industrial motors one size, where that did not require an increase on the conduit size with a payback of less than two years. However these are known loads that run pretty much 24/7.
Interesting. Did you calculate the savings beforehand or just knew it would eventually pay off?
 
Honestly though, real world terminals are rated at 75c. I have yet to find a counterexample even though I am always searching for one. There are various code sections that demand the 75c table is used in this and that circumstance.
Equipment terminals, overcurrent devices and such are usually 75C on newer items anyway. But there are terminal blocks, splicing devices, etc that are rated 90C. You would need 75C conductor for some length at each end then use 90C splicing method and go with conductor selected at 90C ampacity to do this. Thing is after going through all that for a short length you maybe didn't save anything, if you do it for a longer run, well you may still have voltage drop to take into consideration as well as the I2R losses
 
Top