Feeder size for multiple EVSEs

[tls]

Local electrician I drink with sometimes made a fairly convincing case to me yesterday that if I am putting in a second charger and thus swapping my 2-space panel used as a disconnect for a 4-space panel used as a disconnect/j-box, I am not required to follow the 125% continuous-load rule for the wiring from subpanel to main panel. That wiring now becomes a feeder serving two continuous loads which I am not required to assume will be simultaneously at full current, rather than a branch serving one continuous load.

OK so far. Wondering what @FlasherZ in particular thinks of this. But here's the galaxy-brain part.

My present setup is a single EVSE connected by about 30' of #4 XHHW in conduit. This is under a deck and though re-pulling would be possible, I don't care to do so just to get another 8A of charging ampacity by replacing with #3. Everything involved is rated for 75C so I get 85A ampacity, which I am allowed to round up the next standard breaker size, 90A. This lets me run my charger at its 90A breaker/72A load setting. The charger is a Tesla Wall Connector Gen2, which does load sharing, so I can add a second charger and they will split the 72A amongst themselves, though strict code compliance might be understood to require me to put each on its own branch circuit (the only reason it might not is that the Tesla charger manual, which is within the scope of its UL approval, says that the power wiring for two HPWCs in a load-sharing arrangement must be connected in a junction box, and in theory, if the approved device's manual says you can do it, you should be allowed to do it).

With me so far? Now hear this.

My buddy Sparky suggests that I can legally run just a couple of feet of #3 *from the new 4-space subpanel to each charger* and then place each charger on a 100A breaker, rather than a 90A breaker, even though the panel itself is fed by what is now a 90A feeder. Each (2' long...) branch circuit from subpanel to charger is compliant because the cable and OCPD are rated for 125% of the 80A continuous load. The 30' feeder back to my pain panel is also compliant because the 125% rule does not apply to feeders.

This makes a sort of perverse sense to me, and would appear to be a way to legally wire two 80A EVSEs with a single run of #4 copper from a main panel. But it has to be prohibited somewhere in the NEC, right? We have both looked for a reference forbidding a subpanel from containing a branch circuit OCPD larger than that protecting the feeder circuit for that subpanel, and we are coming up dry... taken to an extreme, this would seem to permit any number of things that really ought not be allowed so long as that small-panel-used-as-disconnect right before the EVSE has a couple of extra spaces - like an 80A EVSE and a 20A convenience outlet, together on that subpanel fed by a #4 feeder.

Speaking of "how the poco does it"...

Sponsored

 

[PV2EV]

How much had he drank before the suggestion?

Not derating sounds correct (and a cool tip) to the panel. The thing I question is going from a #4 to #3, just does not sound kosher.

I had #2 AL SER run to a NEMA box, and am replacing the box with a sub-panel to a. serve as a disconnect, b. add a 120v quad box, and c. downsize the cable to land in the 50A EVSE. So 100A/#2AL to 70A/#6 CU, with an option for an easy EVSE upgrade in future.

 

[Maquis]

Your feeder must be sized for 125% of the largest load plus 100% of the other load(s). So you would need a minimum feeder size of 180A to run 2 Charge Station Pros. There is an exception to allow a smaller feeder if the EVSEs are “smart” meaning they communicate to limit the charge current.

 

[FlasherZ]

[...]if I am putting in a second charger and thus swapping my 2-space panel used as a disconnect for a 4-space panel used as a disconnect/j-box, I am not required to follow the 125% continuous-load rule for the wiring from subpanel to main panel.

That wiring now becomes a feeder serving two continuous loads which I am not required to assume will be simultaneously at full current, rather than a branch serving one continuous load.

When it becomes a feeder, you have to follow the feeder rules instead of branch circuit rules. But what will trip you up is that feeders are also subject to load calculations and the code still requires you follow 125% of continuous load (2017 NEC 215.2(A)(1)(a), 220.40). There is an exception for load demand factor of 75% if you have 4 or more appliances installed in a dwelling unit (220.53), but you only have 2.

If you can reasonably guarantee (and not just through human process "don't do that" means) that both EVSE's won't be in use at the same time, 220.60 allows you to use the largest EVSE to size the feeder. For example, if you had a Tesla HPWC and a Ford CSP serving the same garage bay and the cords couldn't reach anywhere else, you could rely upon this.

My present setup is a single EVSE connected by about 30' of #4 XHHW in conduit. This is under a deck and though re-pulling would be possible, I don't care to do so just to get another 8A of charging ampacity by replacing with #3. Everything involved is rated for 75C so I get 85A ampacity, which I am allowed to round up the next standard breaker size, 90A. This lets me run my charger at its 90A breaker/72A load setting. The charger is a Tesla Wall Connector Gen2, which does load sharing, so I can add a second charger and they will split the 72A amongst themselves, though strict code compliance might be understood to require me to put each on its own branch circuit (the only reason it might not is that the Tesla charger manual, which is within the scope of its UL approval, says that the power wiring for two HPWCs in a load-sharing arrangement must be connected in a junction box, and in theory, if the approved device's manual says you can do it, you should be allowed to do it).

Unfortunately, it doesn't allow you to use your charger at the 90A breaker/72A load setting (legally). You determine conductor ampacity based on load, and then breaker size to protect the conductors. You can't "back in" to a smaller conductor because of a next-up breaker size. This is also what invalidates 2/3 NM cable for 100A installs (60 degC column @ 95A).

Take 72A charging load times 1.25 = 90A. Conductor ampacity must be minimum 90A, and 85A doesn't cut it (legally).

2 gen-2 wall connectors can be viewed as a single appliance for load calculation reasons if they reasonably guarantee that, combined, they will manage their load to never exceed a configured load. Tesla gen 2 connectors do that.

My buddy Sparky suggests that I can legally run just a couple of feet of #3 *from the new 4-space subpanel to each charger* and then place each charger on a 100A breaker, rather than a 90A breaker, even though the panel itself is fed by what is now a 90A feeder. Each (2' long...) branch circuit from subpanel to charger is compliant because the cable and OCPD are rated for 125% of the 80A continuous load. The 30' feeder back to my pain panel is also compliant because the 125% rule does not apply to feeders.

The 125% rule does apply to feeders:

215.2(A)(1)(a):
" Where a feeder supplies continuous loads or any combination of continuous and noncontinuous loads, the minimum feeder conductor size shall have an allowable ampacity not less than the noncontinuous load plus 125 percent of the continuous load. "

Speaking of "how the poco does it"...

We shouldn't get started there, PoCo's are permitted to do things that NEC shrieks at.

 

[FlasherZ]

Your feeder must be sized for 125% of the largest load plus 100% of the other load(s). So you would need a minimum feeder size of 180A to run 2 Charge Station Pros. There is an exception to allow a smaller feeder if the EVSEs are “smart” meaning they communicate to limit the charge current.

Even 180A will not work. All EV charging loads are continuous, and 215.2(A)(1)(a) requires 125% of that, or 200A feeder required for 2.

 

[Maquis]

Even 180A will not work. All EV charging loads are continuous, and 215.2(A)(1)(a) requires 125% of that, or 200A feeder required for 2.

You’re correct….unless exception 1 is met, but we shouldn’t go there!

 

[tls]

When it becomes a feeder, you have to follow the feeder rules instead of branch circuit rules. But what will trip you up is that feeders are also subject to load calculations and the code still requires you follow 125% of continuous load (2017 NEC 215.2(A)(1)(a), 220.40). There is an exception for load demand factor of 75% if you have 4 or more appliances installed in a dwelling unit (220.53), but you only have 2.

I mean, it is obvious how to arrange to have 4 - this just requires a 6-space panel - but adding a quad convenience outlet as a code-lawyering exercise seems pretty dubious.

Unfortunately, it doesn't allow you to use your charger at the 90A breaker/72A load setting (legally). You determine conductor ampacity based on load, and then breaker size to protect the conductors. You can't "back in" to a smaller conductor because of a next-up breaker size. This is also what invalidates 2/3 NM cable for 100A installs (60 degC column @ 95A).

Frustrating to me on this one is that the existing #4 on 90A setup was literally suggested by the local inspector after he saw we'd protected it at 80A. Words were something along the lines of "it's like how you can wire a 60A spa with 6/3" which, let's just say, seems to be the standard local interpretation in out in the sticks where I live, but one I wouldn't try to get away with in the big smoky city where I work...

I suppose I ought to crawl under the deck (more of a porch really), find those pull elbows, and pull the 2 #3 and bare ground that'll just meet the fill rule for the conduit that's already there. I was putting that off until I decided what to do about the HIS (which would require more / larger pipe than I have now) but SunRun kind of decided that one for me.

 

[FlasherZ]

Frustrating to me on this one is that the existing #4 on 90A setup was literally suggested by the local inspector after he saw we'd protected it at 80A. Words were something along the lines of "it's like how you can wire a 60A spa with 6/3" which, let's just say, seems to be the standard local interpretation in out in the sticks where I live, but one I wouldn't try to get away with in the big smoky city where I work...

The AHJ (inspector) is the only person you really have to please here and it shifts liability to him. I'm a bit concerned that you have an inspector not knowing the code, but hey...

 

[jefro]

There are a few evse's that communicate to allow for sharing of power. That isn't what you have. You have a load calculation where some end point(s) could be a continuous load. No amount of beer can change that.

 

[tls]

There are a few evse's that communicate to allow for sharing of power. That isn't what you have.

Actually it's exactly what I have. I have a Tesla HPWC gen 2, and can swap the Ford charger (if it ever arrives) for another which is at a relative's house. The gen 2 Tesla chargers communicate to limit total current.

 

[adoublee]

This is covered clearly in the EVSE section. All continuous unless auto load management system.

Sponsored