Feeder Sizing (80% rule)

[RogerRoger]

I'm reviewing the electrical code to teach other engineers the rules on feeder sizing and want to verify I'm still on the right track. It seems like there's still a lot of confusion among my colleagues with the "80% breaker standard". I'm using NEC 2023.

Example 1:
The below table shows a feeder calculation. Per my understanding of 215.2, the feeder size would be determined by taking the noncontinuous loads plus 125% of the continuous loads. I first determine the noncontinuous load demand factors per article 220 and add those loads to 125% of the continuous load. The final demand value comes out to 284.5A. Other engineers say that breakers are 80% rated and have required that the minimum feeder size be determined by taking an additional 125% of the 284.5A, resulting in a feeder size of 400A (next size up from 355.63). Since I've already applied the 125% to the continuous loads in the calculations, that seems redundant. In this case, I would have the minimum feeder size be the next size up from the demand - 300A. Please note I'm just wanting to determine the minimum feeder size and not taking into account the good design practice of leaving 20%-25% spare capacity for future additions.

Non-Continuous & Continuous Feeder (Non-Dwelling)
LOAD CLASSIFICATION	CONNECTED LOAD	DEMAND FACTOR	ESTIMATED DEMAND
R - RECEPTACLE LOAD	20000 VA	0.75 (220.47)	15000 VA
X - LARGEST MOTOR	20000 VA	1.25	25000 VA
MN - MISC NONCONTINUOUS	50000 VA	1	50000 VA
MC - MISC CONTINUOUS	10000 VA	1.25	12500 VA
TOTAL (VA)	100000 VA		102500 VA
TOTAL (A @ 208V/3-ph)	277.6 A		284.5 A

Example 2:
Let's say we don't have continuous loads and the feeder only serves non-continuous loads. Per the below table, I would have the minimum feeder size to be 225A (I know this is tight... again, this is only the minimum size and I would typically go bigger in real life to allow for future wiggle-room or owner changes) while other engineers would bring up the 80% breaker rule and say the minimum feeder size would be 300A (222.1x1.25=277.6).

Non-Continuous Feeder (Non-Dwelling)
LOAD CLASSIFICATION	CONNECTED LOAD	DEMAND FACTOR	ESTIMATED DEMAND
R - RECEPTACLE LOAD	50000 VA	0.60	30000 VA
MN - MISC NONCONTINUOUS	50000 VA	1	50000 VA
TOTAL (VA)	100000 VA		80000 VA
TOTAL (A @ 208V/3-ph)	277.6 A		222.1 A

Let me know what you think!

Additional references:
80% Rule Thread (Jraef's post was particularly helpful)
Schneider Article on 80% Rated Breakers
UL 489

 

[wwhitney]

I first determine the noncontinuous load demand factors per article 220 and add those loads to 125% of the continuous load. The final demand value comes out to 284.5A. Other engineers say that breakers are 80% rated and have required that the minimum feeder size be determined by taking an additional 125% of the 284.5A

No, "other engineers" are wrong, full stop. You have followed the computation specified in 215.2.

Let's say we don't have continuous loads and the feeder only serves non-continuous loads. Per the below table, I would have the minimum feeder size to be 225A (I know this is tight... again, this is only the minimum size and I would typically go bigger in real life to allow for future wiggle-room or owner changes) while other engineers would bring up the 80% breaker rule and say the minimum feeder size would be 300A (222.1x1.25=277.6).

Again, "other engineers" are wrong, 225A is the minimum size specified in 215.2.

Cheers, Wayne

 

[d0nut]

The short version is you are correct.

Maybe if you click your heels together and repeat "There is no such thing as an 80% breaker" you can escape the land of compounding 125%s?

There are some assumptions about breakers being used in standard ambient conditions, properly sized conductors, etc., that go into the following statements.

1. Breakers don't trip below their rated value. A 100A breaker won't trip at 80A.
2. Breakers are tested under certain test parameters. A continuous load at the breaker's rating won't allow the breaker to properly dissipate the heat generated from the load or from adjacent breaker's heating the subject breaker.
3. To account for this phenomenon, when doing a load calculation for a standard breaker you take 100% of the non-continuous loads plus 125% of the continuous loads, ignoring other demand factors. The number you get from this calculation is the load. The conductors need to be sized for this value, and the breaker needs to be sized to protect the conductors. There are no other adjustments to perform on your load or breaker sizing.
4. 100% rated breakers are special. They are typically mounted in individual enclosures and are better at dissipating heat because there aren't adjacent breakers heating them. Because of this improved cooling, you are allowed to calculate the load at 100% of the non-continuous loads plus 100% of the continuous loads. Again, when you get this value, there are no other adjustment to perform on the load or breaker sizing. Size the conductors for this value and size the breaker to protect the conductors.

 

[ramsy]

It takes an engineer to read the OP dissertation question, which effectively filters free opinions from most contractors & inspectors on this forum, who favor shorter questions.

 

[infinity]

It takes an engineer to read the OP dissertation question, which effectively filters free opinions from most contractors & inspectors on this forum, who favor shorter questions.

This is in the Electrical Calculations/Engineering Forum so the length of the post is required to get the point across. If you don't like the game, dont play.

 

[ramsy]

Again, when you get this value, there are no other adjustment to perform on the load or breaker sizing.

dOnut & Wayne have educated this OP, with responses more worthy of compensation than forum opinion.

Will ad the most overlooked case where thermal imaging reports show overheated equipment occurs, with inductive ballast & motor loads, where current rises with voltage drop.

Tossing the 125% baby for NEC 220.18 inductive loads, with the 80% breaker bathwater is a typical knucklehead maneuver.

 

[infinity]

This is an excellent question that RogerRoger has asked because so many fail to understand the 80%/125% rules when applied to OCPD's. Wayne said it best.

No, "other engineers" are wrong, full stop

 

[ramsy]

If you don't like the game, dont play.

All observation should be welcome, if accurate, and free of charge.

 

[infinity]

All observation should be welcome, if accurate, and free of charge.

Agreed. No one here is getting paid, including the mods that run this thing.

 

[RogerRoger]

Thank you for the prompt responses! d0nut, assuming I'm understanding you correct, I'm not sure I agree with your item #3 because 220.40 would be irrelevant and could potentially conflict with 215.2. there are instances where the demand factor you would apply to the non-continuous load could be greater than 1, such as for the largest motor. Unless the largest motor would be considered a continuous load, in which case, I'd be wrong

In the example below, it appears you are saying we would take 100% of the noncontinuous load without applying demand factors (25kVA+50kVA=75kVA) plus 125% of the continuous load (12.5kVA) which would result in a minimum feeder size of 250A. Per the load calculation from 220.40, and without accounting for the 125% of the continuous load, the feeder load would be 31.25kVA+50kVA+10kVA=91.25kVA or 253A (@208V,3-phase). That could mean a 50A feeder size difference. I know I'm probably in the weeds at this point though.

Example 3
LOAD CLASSIFICATION	CONNECTED LOAD	DEMAND FACTOR	ESTIMATED DEMAND
X - LARGEST MOTOR	25000 VA	1.25	31250 VA
MN - MISC NONCONTINUOUS	50000 VA	1	50000 VA
MC - MISC CONTINUOUS	10000 VA	1.25	12500 VA
TOTAL (VA)	85000 VA		93750 VA
TOTAL (A @ 208V/3-ph)	235.9 A		260.2 A

 

[PD1972]

A helpful reference to understand how to do a load calculation would be to run through NEC Annex D examples.

There is no need to create your own material as those example calculations (albeit simple) perform a good breakdown of how to categorize your loads (noncontinuous, continuous, Table 220.3 loads). It also comes straight from the NEC so there should not be an argument regarding whether it is a good source of information.

 

[eric stromberg]

There is no such thing as an 80% rated breaker. Breakers can handle up to their full rating. If, however, you have a load that will have sustained maximum current for three hours or more, you have to build a circuit that is 25% larger than what you need. Circuits are always designed from the load up to the source. Not from the source down to the load. Kind of interesting that the articles in the Code go from Load to source. 210 is branch circuits, 215 is feeders, 220 is calculations, 230 is services. (220 will become 120 in 2026)

 

[wwhitney]

There is no such thing as an 80% rated breaker. Breakers can handle up to their full rating. If, however, you have a load that will have sustained maximum current for three hours or more, you have to build a circuit that is 25% larger than what you need.

But on that last sentence, only if the breaker is not 100% rated. I.e. the circuit only needs upsizing because of the limitations of the breaker. Saying a non-100% rated breaker is "80% rated for continuous loads" strikes me as a convenient and accurate short hand for that limitation. It is surprising to me that it seems to be the source of so much confusion.

Cheers, Wayne

 

[eric stromberg]

True. I didn't put the 100% rating in my response. 80% is, indeed, shorthand. I have also seen where it causes confusion. I installed a 500 amp circuit and was asked how many amps it was good for. I said "500 amps." The load was a fairly short-time load. I think, part of the issue, stems from electrical courses and electrical exams. "Continuous" is a favorite topic for these. This carries over into the real world where there are very few continuous loads to deal with. Lighting has been historically continuous, but now with lighting controls and energy conservation codes, a case could be made that a percentage of lighting is no longer continuous. Data centers, on the other hand are a totally different situation. They typically run full load 24/7.

 

[d0nut]

Thank you for the prompt responses! d0nut, assuming I'm understanding you correct, I'm not sure I agree with your item #3 because 220.40 would be irrelevant and could potentially conflict with 215.2. there are instances where the demand factor you would apply to the non-continuous load could be greater than 1, such as for the largest motor. Unless the largest motor would be considered a continuous load, in which case, I'd be wrong

I was trying to keep my example simple and focused on the continuous versus non-continuous question rather than getting lost in all the possible demand factors. When I mentioned ignoring the other demand factors, I was trying to convey that my example was ignoring the demand factors outside of continuous and non-continuous. In the actual calculation, all demand factors would come into play, including receptacles, motors, elevators, x-rays, kitchen equipment, etc. But the general guidance is still valid. When you have the load calculated including all demand factors, you don't need to adjust that load number anymore.

 

[RogerRoger]

Thank you all for your time and comments!