Question about 120 percent NEC rule with Tesla Gateway 2

[holeydonut]

I'm ready to lose it. Going green sucks. Getting solar sucks. Getting ESS sucks. Bust most of all, PG&E sucks.

The engineer I'm trying to argue with is claiming the following setup will not work for a 8 kWp DC system with 3x Tesla Powerwall 2's using a Tesla Energy Gateway 2.

Main Service Panel rated at 200 A (solar ready Square D SC2040M200PF with 225 A bus bar)
Main Service Panel Houses a 200 A main breaker
All this feeds Tesla Energy Gateway 2
Left side of Gateway 2 houses 3x 30 A breakers for three AC Coupled Powerewall 2s
Left side of Gateway 2 houses 1x 35 A breaker for 240v 6.68 kWp (AC) system (via Enphase Envoy)
L1 Backup Lug uses 6 AWG to feed a internal sub-panel that has operated for years with a 90 A breaker
L2 Backup Lug uses 6 AWG to feed a internal sub-panel that has operated for years with a 70 A breaker

PG&E's engineer's beef is that since the Tesla Energy Gateway 2 is rated for 200A, then my main service panel needs to be sized 120% (240 Amps). So PG&E says I will only be allowed to use a 125 A main breaker and force the contents of the Tesla Energy Gateway 2 down to 100 Amps.

This means I cannot use the L2 backup lugs and need to move those off to a separate line coming out of my Main Panel.

Ugh.

 

[nwdiver]

PG&E's engineer's beef is that since the Tesla Energy Gateway 2 is rated for 200A, then my main service panel needs to be sized 120% (240 Amps). So PG&E says I will only be allowed to use a 125 A main breaker and force the contents of the Tesla Energy Gateway 2 down to 100 Amps.

.... that's 24kW. Are you running a grow house? I've been running my entire house through a 30A breaker for ~5 years.

 

[wwhitney]

The PG&E Engineer is mistaken. The "120% rule", or NEC 705.12(B)(2)(3)(b) is not the only rule available for protecting a panel's busbar. For the main panel, you are using the "sum of all breakers" rule, or NEC 705.12(B)(2)(3)(c). And it sounds like all your other panels have only a single source of supply, so they are not subject to 705.12(B).

A reasonable question for the engineer to ask is whether your 200A feeder complies with NEC 705.12(B)(2)(1), and a one-line diagram showing the connection details and conductor ampacities should be enough for that. But don't bring it up unless he asks.

BTW, it seems to me you have listed 3 different things to connect to the 2 lugs on the load side of the gateway: the internal subpanel (used for a generation panel) and the two existing load panels you mentioned. Also, all the feeder segments for which the only OCPD between that segment and the grid is the 200A main breaker will have to have an ampacity of 200A (* 0.83 for the residential dwelling rule), and the panels they feed will need to have main breakers.

Cheers, Wayne

 

[holeydonut]

The PG&E Engineer is mistaken. The "120% rule", or NEC 705.12(B)(2)(3)(b) is not the only rule available for protecting a panel's busbar. For the main panel, you are using the "sum of all breakers" rule, or NEC 705.12(B)(2)(3)(c). And it sounds like all your other panels have only a single source of supply, so they are not subject to 705.12(B).

A reasonable question for the engineer to ask is whether your 200A feeder complies with NEC 705.12(B)(2)(1), and a one-line diagram showing the connection details and conductor ampacities should be enough for that. But don't bring it up unless he asks.

BTW, it seems to me you have listed 3 different things to connect to the 2 lugs on the load side of the gateway: the internal subpanel (used for a generation panel) and the two existing load panels you mentioned. Also, all the feeder segments for which the only OCPD between that segment and the grid is the 200A main breaker will have to have an ampacity of 200A (* 0.83 for the residential dwelling rule), and the panels they feed will need to have main breakers.

Cheers, Wayne

Thanks Wayne - your knowledge is endless. I wish you worked for PG&E haha.

I'll reference the NEC 2014 705.12(B) stuff. Although now I think they're just blocking me for the sake of blocking me. After all they did say me going solar is taking money from PG&E's pocket. I wonder if I actually need to file a complaint with the California Office of Planning and Research, CPUC, or some other governing body.

On that last point you raised that the things connecting to the Gateway 2 need their own breakers... Based on what Sunrun spec'ed are you saying I need to move the 35 A solar and 3x 30 A Powerwall breakers to a separate generation panel and put a breaker on that panel? The Gateway 2 makes it sound like you can directly feed generation/Powerwall stuff directly to its integrated busbar.

 

[holeydonut]

Ok so PG&E says that they cannot accept the internal subpanel of the Gateway 2 does not have its own OCPD. So they are assessing that the continuous energy output of the Solar + 3x Powerwalls is 100 Amps. They say the first available OCPD that goes on top of this needs to be 125% of 100 = 125 amps. They say that means the Tesla Energy Gateway 2 itself cannot be connected to more than a 125 amps breaker in my main service panel.

I'm trying to explain to them that the generation/clean-energy side (non-backup) of the GW 2 has an OCPD on the internal panelboard. Why does the homeowner need to do all this extra work? This is stupid!!!!

 

[miimura]

AFAIK, the left side of the GW2 is Non-Backup only. It makes no sense to connect Powerwalls there. Maybe I misunderstood your posts.

 

[miimura]

If you want the internal panelboard to be used for Powerwalls, you have to use one of the sets of backup lugs at the bottom left. If you still have two sub-panels that need to be connected to the backup side, you would have to use Polaris connectors or something else external, if that's even code compliant.

 

[Kodemonkey]

The PG&E Engineer is mistaken. The "120% rule", or NEC 705.12(B)(2)(3)(b) is not the only rule available for protecting a panel's busbar. For the main panel, you are using the "sum of all breakers" rule, or NEC 705.12(B)(2)(3)(c). And it sounds like all your other panels have only a single source of supply, so they are not subject to 705.12(B).

A reasonable question for the engineer to ask is whether your 200A feeder complies with NEC 705.12(B)(2)(1), and a one-line diagram showing the connection details and conductor ampacities should be enough for that. But don't bring it up unless he asks.

BTW, it seems to me you have listed 3 different things to connect to the 2 lugs on the load side of the gateway: the internal subpanel (used for a generation panel) and the two existing load panels you mentioned. Also, all the feeder segments for which the only OCPD between that segment and the grid is the 200A main breaker will have to have an ampacity of 200A (* 0.83 for the residential dwelling rule), and the panels they feed will need to have main breakers.

Cheers, Wayne

Have you seen the NEC load calculation worksheet that Tesla Engineering submitted to PG&E? Mine had a lot of errors in it. Like my house was sized 50% more in sq feet and a 50A calculation for a Spa that doesn't exist was left in it.

Also, isn't City/County planning that approves your electrical layout? In LA it is done by the county planning or city (depending if you are in an incorporated city or not)

 

[miimura]

Have you seen the NEC load calculation worksheet that Tesla Engineering submitted to PG&E? Mine had a lot of errors in it. Like my house was sized 50% more in sq feet and a 50A calculation for a Spa that doesn't exist was left in it.

Also, isn't City/County planning that approves your electrical layout? In LA it is done by the county planning or city (depending if you are in an incorporated city or not)

I agree. I don't understand why PG&E engineer is involved with this at all. The AHJ (city / county) has jurisdiction over these issues.

 

[holeydonut]

AFAIK, the left side of the GW2 is Non-Backup only. It makes no sense to connect Powerwalls there. Maybe I misunderstood your posts.

Got it, so if this schematic is used I only have one set of lugs on the bottom to do the backup loads. I misunderstood Sunrun's line diagram. They meant:

Main Service Panel rated at 200 A (solar ready Square D SC2040M200PF with 225 A bus bar)
Main Service Panel Houses a 200 A main breaker
All this feeds Tesla Energy Gateway 2
Left side of Gateway 2 houses 3x 30 A breakers for three AC Coupled Powerewall 2s
Left side of Gateway 2 houses 1x 35 A breaker for 240v 6.68 kWp (AC) system (via Enphase Envoy)
Backup Lug uses 6 AWG to feed a new sub-panel. Within this sub panel will be a 90 A breaker and 70 A each feeding existing home loads

 

[holeydonut]

I agree. I don't understand why PG&E engineer is involved with this at all. The AHJ (city / county) has jurisdiction over these issues.

[quote="kodemonkey]Have you seen the NEC load calculation worksheet that Tesla Engineering submitted to PG&E? Mine had a lot of errors in it. Like my house was sized 50% more in sq feet and a 50A calculation for a Spa that doesn't exist was left in it.

Also, isn't City/County planning that approves your electrical layout? In LA it is done by the county planning or city (depending if you are in an incorporated city or not)[/quote]



I'm going through Sunrun, not Tesla. But I'm getting Tesla's hardware installed for the ESS and Gateway. I'm in unincorporated Contra Costa County. The county is fine with what Sunrun submitted. PG&E is the blocker and says they need full approval alongside Contra Costa.

PG&E insists that they are worried about the scenario where I tell all of my Powerwalls and divert my max solar array production to discharge back into their grid. And they're worried what will happen. This is ironic since the last thing I want to do is to push any of my energy onto their grid.

 

[miimura]

Got it, so if this schematic is used I only have one set of lugs on the bottom to do the backup loads. I misunderstood Sunrun's line diagram. They meant:

Main Service Panel rated at 200 A (solar ready Square D SC2040M200PF with 225 A bus bar)
Main Service Panel Houses a 200 A main breaker
All this feeds Tesla Energy Gateway 2
Left side of Gateway 2 houses 3x 30 A breakers for three AC Coupled Powerewall 2s
Left side of Gateway 2 houses 1x 35 A breaker for 240v 6.68 kWp (AC) system (via Enphase Envoy)
Backup Lug uses 6 AWG to feed a new sub-panel. Within this sub panel will be a 90 A breaker and 70 A each feeding existing home loads

If you replace both instances of "Left side of Gateway 2" with "Internal Panelboard wired to Backup", I would say it all looks good.

 

[miimura]

I'm going through Sunrun, not Tesla. But I'm getting Tesla's hardware installed for the ESS and Gateway. I'm in unincorporated Contra Costa County. The county is fine with what Sunrun submitted. PG&E is the blocker and says they need full approval alongside Contra Costa.

PG&E insists that they are worried about the scenario where I tell all of my Powerwalls and divert my max solar array production to discharge back into their grid. And they're worried what will happen. This is ironic since the last thing I want to do is to push any of my energy onto their grid.

The thing they're worried about doesn't happen. In a Powerwall system, only solar production feeds to the grid. I think there are different versions of the Interconnection Agreement with different limitations on storage operations. I believe one of them is called something like "Non-Export Storage Interconnection". Or not.

 

[holeydonut]

Ugh, so I don't know if PG&E is intentionally being obtuse or if Sunrun screwed up. But it seems the solution to this mess is to get the official line diagram approved by a licensed electrician and by the county permitting office to include the language:


NOTE: PRODUCERS STORAGE DEVICE(S) WILL
NOT CAUSE THE HOST LOAD TO EXCEED ITS
NORMAL PEAK DEMAND. NORMAL PEAK
DEMAND IS DEFINED AS THE HIGHEST
AMOUNT OF POWER REQUIRED FROM THE
DISTRIBUTION SYSTEM BY PRODUCERS
COMPLETE FACILITIES WITHOUT THE
INFLUENCE OR USE OF THE ENERGY
STORAGE DEVICE(S).

THE INSTALLED BATTERY IS NOT CHARGING
FROM THE GRID.

THE BATTERY IS EXPORTING TO THE GRID
UNDER NEM AND TOU.


Yay I guess I have a path forward?

But whyyyyyyyyyyyyyyyy is this even a thing? I cannot believe I am the first customer to get this experience?????

Why would this language even matter for getting a 200 A main service panel approved?

 

[wwhitney]

On that last point you raised that the things connecting to the Gateway 2 need their own breakers... Based on what Sunrun spec'ed are you saying I need to move the 35 A solar and 3x 30 A Powerwall breakers to a separate generation panel and put a breaker on that panel?

First, you could put a main breaker in the internal Gateway 2 panelboard by backfeeding a breaker, like they show when that is used for non-backed up loads (they do that because it is required). You could that when using it as a generation panel, although the jumper layout wouldn't be as convenient. And the panel is limited to 125A breakers, so you'd be limited to 125A of generation breakers.

Second, the rule on the main breaker in a panel is when a panel with loads is supplied by a feeder that has separated connections from the utility and from local generation (inverters for solar or in the Powerwalls). When using the internal panel for generation connection only, you don't need a main breaker.

Ok so PG&E says that they cannot accept the internal subpanel of the Gateway 2 does not have its own OCPD. So they are assessing that the continuous energy output of the Solar + 3x Powerwalls is 100 Amps. They say the first available OCPD that goes on top of this needs to be 125% of 100 = 125 amps. They say that means the Tesla Energy Gateway 2 itself cannot be connected to more than a 125 amps breaker in my main service panel.

That rule doesn't exist. They made that up. There's no need for a single master generation breaker. You can have multiple generation breakers. You would have (4) separate ones, (3) 30s and a 35, rather than one 125A generation breaker.

For example, 2017 NEC 705.12(B) starts off:

705.12(B) Load Side. The output of an interconnected electric power source shall be permitted to be connected to the load side of the service disconnecting means of the other source(s) at any distribution equipment on the premises. Where distribution equipment, including switchgear, switchboards, or panelboards, is fed simultaneously by a primary source(s) of electricity and one or more other power source(s), and where this distribution equipment is capable of supplying multiple branch circuits or feeders, or both, . . .

That clearly contemplates multiple local power sources, not all combined into one master generation breaker.

But as I mentioned above, you could put a 125A breaker in there. The only problem is that the internal panelboard only has 6 spaces. A 125A breaker would use 2. You could get a quad 30A double/30A double that allows 2 Powerwalls to be installed in 2 spaces instead of 4. A quad 30A double/35A double would then let you attach the the 3rd Powerall and the solar inverter, if such a thing exists, I'm not sure it does.

BTW, even without a main breaker, you'll need one quad breaker to connection (4) double pole sources in just 6 spaces.

Cheers, Wayne

 

[wwhitney]

Main Service Panel rated at 200 A (solar ready Square D SC2040M200PF with 225 A bus bar)
Main Service Panel Houses a 200 A main breaker
All this feeds Tesla Energy Gateway 2
Left side of Gateway 2 houses 3x 30 A breakers for three AC Coupled Powerewall 2s
Left side of Gateway 2 houses 1x 35 A breaker for 240v 6.68 kWp (AC) system (via Enphase Envoy)
Backup Lug uses 6 AWG to feed a new sub-panel. Within this sub panel will be a 90 A breaker and 70 A each feeding existing home loads

Right, so this the "new sub-panel" would need a main breaker, as it's supplied by a feeder that is connected to multiple sources of power. That #6 AWG has to be an error, though, it's only good for 65A. The obvious thing to do would be to use #2/0 Cu or #4/0 Al and use a 200A main breaker in the "new sub-panel", which should also be the specs for the feeder from the new service panel to the Backup Gateway.

As to the note requested, it's stupid and incorrect, but if it will let you install it the way you want and compliantly with the NEC, it's probably the least annoying path forward.

Cheers, Wayne

 

[holeydonut]

That rule doesn't exist. They made that up. There's no need for a single master generation breaker. You can have multiple generation breakers. You would have (4) separate ones, (3) 30s and a 35, rather than one 125A generation breaker.

Thanks Wayne - I'll see if Sunrun is ok trying to put the 3x 30A and 1x 35A on the internal panelboard (sounds like these will fit since they won't need to also squeeze a 125A on there.

 

[holeydonut]

Right, so this the "new sub-panel" would need a main breaker, as it's supplied by a feeder that is connected to multiple sources of power. That #6 AWG has to be an error, though, it's only good for 65A. The obvious thing to do would be to use #2/0 Cu or #4/0 Al and use a 200A main breaker in the "new sub-panel", which should also be the specs for the feeder from the new service panel to the Backup Gateway.

As to the note requested, it's stupid and incorrect, but if it will let you install it the way you want and compliantly with the NEC, it's probably the least annoying path forward.

Cheers, Wayne

Yeah, thanks for the catch on the #6 AWG thing. Sunrun also has #1 AWG going from my 200 A main service panel to the Gateway through 1 1/4" conduit. This was approved by two engineers and had a county approval as well. So yeah, I wonder who's fault it would be if they tried to push lots of amps through tiny ass wires.

Nit-picky thing though, but isn't 2/0 copper only rated for 195 amps? Don't I need 3/0 Cu or 4/0 Al?

BTW, who's job is it supposed to be to make sure the engineered plan is "ok" on residential projects? When I was going through the sales/design process, Sunrun simply wouldn't send me the actual permit, line diagrams, site plan, etc. They'd send me a gussied up "laymans" info sheet, but not the level of detail necessary to audit/review that I'm accustomed to on commercial projects with an architect.

Sunrun said their "experts" were the "best in the business" and it was against protocol for a homeowner to get that involved. Clearly, I should have pounded the table harder with them back in August.

Actually what I should have done was go through @Vines. I am an idiot.

 

[jjrandorin]

We are extremely lucky to have obvious industry veterans like @wwhitney and @Vines not only as members here, but also so forthcoming with their time to help our members.

Thank you both very much for all of the information you provide here, we all appreciate your time, energy and effort VERY much.

 

[wwhitney]

Nit-picky thing though, but isn't 2/0 copper only rated for 195 amps? Don't I need 3/0 Cu or 4/0 Al?

2/0 Cu has a 90C ampacity of 195A, which means you can run 195A through it in a 30C ambient environment when you have at most 3 current carrying wires together in a cable or conduit, and the conductor temperature shouldn't exceed 90C. So if you have wire insulation rated for 90C, you won't damage it.

However, you can almost never actually install it in practice for a 195A circuit, because the terminations at the ends are typically only rated for 75C, and so the most you can get is the 75C ampacity of 175A. The 90C ampacity is only useful as a starting point for ampacity adjustment and correction for the case when you have more than 3 current carrying conductors or the ambient temperature is above 30C. Those factors don't apply at the terminations.

So why is 175A ampacity sufficient for a 200A feeder? For a residential dwelling unit, the NEC recognizes that it's own load calculation process is quite conservative, and it lets you apply a 0.83 factor to service conductors or a feeder that carries the entire load of a dwelling unit. That means a 200A feeder only has to have a ampacity of 166A.

BTW, who's job is it supposed to be to make sure the engineered plan is "ok" on residential projects? When I was going through the sales/design process, Sunrun simply wouldn't send me the actual permit, line diagrams, site plan, etc. They'd send me a gussied up "laymans" info sheet, but not the level of detail necessary to audit/review that I'm accustomed to on commercial projects with an architect.

It's the engineer's job to produce correct plans, the permit department's job to find the errors, and the contractor's job to only install compliant work. And by job I mean "legal requirement". So when everybody misses something and the wrong thing gets install, if push comes to shove, you sue both the engineer and the contractor. I think the permit department has immunity typically, their passing something is not to be construed as a positive sign that it is a code compliant, only as a sign that they've failed to find anything that isn't code compliant.

Cheers, Wayne