Possible rewiring of my ESS

[wwhitney]

Going back to the OP, if your final load calc comes out to 200A or less, and you are OK with limiting your total inverter capacity to 200A (via the GW CT-based feature), then I believe you can do what you propose without any new panels.

Namely, you can set up a 200A feeder with the following connections in order starting at the service disconnect:

200A service disconnect
Gateway
200A Spur to load panel 1, which needs a 200A main breaker
200A Spur to load panel 2, which needs a 200A main breaker
CTs for monitoring inverter current, limiting it to 200A
200A Spur to generation panel 1
Generation panel 2

That is, you could set up a really big junction box with (5) circuits coming into it: grid/GW, load 1, load 2, generation 1, and generation 2. The neutral and ground conductors just all get connected with 5 port Polaris-type connectors. For each hot you could use a 3 port Polaris and a 4 port Polaris, along with an internal jumper (short wire segment that doesn't leave the box). The 4 port connector gets grid, load 1, load 2, and jumper. The CTs go on the jumper. And the 3 port connector gets jumper, generation 1, and generation 2.

[Actually if the CTs are physically large enough to put (2) 2/0 Cu wires through, you could use 5 port Polaris connectors on the two hots as well, and just put each CT on both generation 1 and generation 2 conductors of a given leg. But if only (1) 2/0 Cu wire fits through the CT, you'd need to use a jumper like I described.]

Pretty sure that complies with the feeder tap rule in 705.12(B)(1). And the stupid thing about 705.12(B)(3) is that if you replace the Polaris connectors with busbars, and the connections with 200A breakers, it doesn't comply. Even though electrically it's equivalent.

Cheers, Wayne

 

[h2ofun]

Going back to the OP, if your final load calc comes out to 200A or less, and you are OK with limiting your total inverter capacity to 200A (via the GW CT-based feature), then I believe you can do what you propose without any new panels.

Namely, you can set up a 200A feeder with the following connections in order starting at the service disconnect:

200A service disconnect
Gateway
200A Spur to load panel 1, which needs a 200A main breaker
200A Spur to load panel 2, which needs a 200A main breaker
CTs for monitoring inverter current, limiting it to 200A
200A Spur to generation panel 1
Generation panel 2

That is, you could set up a really big junction box with (5) circuits coming into it: grid/GW, load 1, load 2, generation 1, and generation 2. The neutral and ground conductors just all get connected with 5 port Polaris-type connectors. For each hot you could use a 3 port Polaris and a 4 port Polaris, along with an internal jumper (short wire segment that doesn't leave the box). The 4 port connector gets grid, load 1, load 2, and jumper. The CTs go on the jumper. And the 3 port connector gets jumper, generation 1, and generation 2.

[Actually if the CTs are physically large enough to put (2) 2/0 Cu wires through, you could use 5 port Polaris connectors on the two hots as well, and just put each CT on both generation 1 and generation 2 conductors of a given leg. But if only (1) 2/0 Cu wire fits through the CT, you'd need to use a jumper like I described.]

Pretty sure that complies with the feeder tap rule in 705.12(B)(1). And the stupid thing about 705.12(B)(3) is that if you replace the Polaris connectors with busbars, and the connections with 200A breakers, it doesn't comply. Even though electrically it's equivalent.

Cheers, Wayne

Confused on the comment limit total inverter capacity to 200A? Is the the solaredge inverters. 2 11400's it like 95amps output max?

 

[wwhitney]

Confused on the comment limit total inverter capacity to 200A? Is the the solaredge inverters. 2 11400's it like 95amps output max?

By inverter I mean both PV inverters and the inverters built into each PW. Going by breaker size (which includes the 125% continuous use factor), you have 330A of inverters. 2 x 60A (PV) and 7 x 30A (PWs).

Cheers, Wayne

 

[Vines]

Going back to the OP, if your final load calc comes out to 200A or less, and you are OK with limiting your total inverter capacity to 200A (via the GW CT-based feature), then I believe you can do what you propose without any new panels.

Namely, you can set up a 200A feeder with the following connections in order starting at the service disconnect:

200A service disconnect
Gateway
200A Spur to load panel 1, which needs a 200A main breaker
200A Spur to load panel 2, which needs a 200A main breaker
CTs for monitoring inverter current, limiting it to 200A
200A Spur to generation panel 1
Generation panel 2

That is, you could set up a really big junction box with (5) circuits coming into it: grid/GW, load 1, load 2, generation 1, and generation 2. The neutral and ground conductors just all get connected with 5 port Polaris-type connectors. For each hot you could use a 3 port Polaris and a 4 port Polaris, along with an internal jumper (short wire segment that doesn't leave the box). The 4 port connector gets grid, load 1, load 2, and jumper. The CTs go on the jumper. And the 3 port connector gets jumper, generation 1, and generation 2.

[Actually if the CTs are physically large enough to put (2) 2/0 Cu wires through, you could use 5 port Polaris connectors on the two hots as well, and just put each CT on both generation 1 and generation 2 conductors of a given leg. But if only (1) 2/0 Cu wire fits through the CT, you'd need to use a jumper like I described.]

Pretty sure that complies with the feeder tap rule in 705.12(B)(1). And the stupid thing about 705.12(B)(3) is that if you replace the Polaris connectors with busbars, and the connections with 200A breakers, it doesn't comply. Even though electrically it's equivalent.

Cheers, Wayne

Love it and that is how industrial electrical distribution has been done for many years. Electrical equivalent as you described even though it would be arguably safer with the breaker stabs onto a busbar though a violation without a truly large busbar.

The CT meters can also be Y-split with kits designed for that. Not sure about a pair of 2/0 cables through a single CT.

Since the load calculations are less than 200A, by definition he doesn't need more than 200A fed into them from inverters.

 

[CrazyRabbit]

@h2ofun use 3/0 cable from grid to you 400/600 amp panel. 2/0 or whatever that feeds your two loads panels is fine.

 

[wwhitney]

@h2ofun use 3/0 cable from grid to you 400/600 amp panel. 2/0 or whatever that feeds your two loads panels is fine.

Whether a 200A feeder in the OP's situation would need to be 2/0 or 3/0 if Cu is an interesting question. [On other than a 200A residential service, or running between 90C connectors (like two Polaris connectors), a 200A feeder would always need to be 3/0 Cu.] But there's no reason that the answer would be different for different parts of the OP's dwelling unit service.

Also, for large feeders, Al is still a more economical choice. Less $ and less weight per amp.

Cheers, Wayne