Sol-Ark Grid connection options

[solarken]

I could use some feedback on three options I put together for a Sol-Ark system design. 2023 NEC applies here. Attached is simple 1-line of each. I don't think option A is feasible because our utilities generally require an external accessible and lockable PV System disconnect. It is not clear to me that Option A does not meet NEC 690.13 requirement for PV System disconnecting means, as this option does provide a means to disconnect the PV system by turning off the Service breaker and moving the Bypass switch to the bottom position, albeit that also cuts power to the home as well.
Option C provide the most extensive capabilities and avoids a supply side connection, but it requires two breaker enclosures and the bypass switch, so a lot of wall space outside the house and extra expense.
Option B seems to meet requirements to me with only one breaker enclosure, but is a supply side connection
The other detail I need to work out is where to run a neutral with the L1 and L2 between all this equipment. It seems I will need to leave a neutral out of one run in the triangle formed by the Inverter having connections to the breaker enclosure and the bypass switch, or there will be a loop in neutral path.
Any thoughts or suggestions are welcomed

 

[scrubbin]

I basically did my Sol-ark 15k like your example C except I used a fusible disconnect for the PV system disconnect. Mind you that this is/was my first and only solar install and I'm just a industrial maintenance technician. I just ran my neutral through the PV disconnect enclosure (not connected) to the Sol-ark 15K and from there to the bypass switch. My system is only 100 amp but because I used a Sol-ark 15k my utility wants me to upgrade their transformer at the tidy sum of $3400.00.. The install did pass inspection last month so I'm just waiting for a battery I ordered so I can get the install finished. The Sol-ark and the bypass switch are mounted opposite of the outside disconnects.

 

[jaggedben]

The basic problem here is that the PV inverter is in the same piece of equipment as the MID. So you can never give the utility their PV disconnect without letting them also disconnect the load.

I don't think B has a compliant service disconnect; I don't see the overcurrent protection. Are there manual transfer switches that are service rated? If so wouldn't they need fuses?

 

[tortuga]

I could use some feedback on three options I put together for a Sol-Ark system design. 2023 NEC applies here. .../...
Any thoughts or suggestions are welcomed

1) It appears to be a 15kw (63A) optional standby feeding a 200A loadcenter,
If the the sol-ark is going to function as an ATS optional standby system does it not need to meet 702.4(A)(2) in the 2023? The 2023 NEC seems to require a Listed Energy Management System (EMS), and no longer allows the old load shed modules.

2) Regarding the neutral loop it would seem the 'bypass switch' would need be a 3 pole.

3) Why do you need the bypass switch? Is that a code issue or customer request?

I don't think B has a compliant service disconnect; I don't see the overcurrent protection. Are there manual transfer switches that are service rated?

There are manual transfer switches that are service rated, you can also modify a ATS to be manual only by adding a toggle switch.

 

[wwhitney]

2) Regarding the neutral loop it would seem the 'bypass switch' would need be a 3 pole.

The other way to deal with neutral loops is to avoid them by collapsing any schematic loops into lines as far as the actual conduit runs, and then running only a single neutral conductor within each conduit run. So while the ungrounded conductors may form loops, the neutral conductors would form a tree.

E.g. in diagram A in the OP, the feeder from the Service Disconnect to the Sol-Ark could be routed through the Bypass Switch. The conduit from the Service Disconnect to the Bypass Switch would have two pairs of ungrounded conductors but just one neutral. Obviously this arrangement would have ampacity adjustment implications.

Cheers, Wayne

 

[wwhitney]

The other detail I need to work out is where to run a neutral with the L1 and L2 between all this equipment. It seems I will need to leave a neutral out of one run in the triangle formed by the Inverter having connections to the breaker enclosure and the bypass switch, or there will be a loop in neutral path.

Agreed that a neutral loop is not allowed, but just leaving the neutral out of one side of the triangle is also not allowed. It would violate 300.3(B).

Cheers, Wayne

 

[tortuga]

The other way to deal with neutral loops is to avoid them by collapsing any schematic loops into lines as far as the actual conduit runs, and then running only a single neutral conductor within each conduit run. So while the ungrounded conductors may form loops, the neutral conductors would form a tree.

Good idea that would work if both sets of ungrounded line conductors were either service or feeder conductors, he can't mix service and feeders in the same raceway though.

 

[wwhitney]

Good idea that would work if both sets of ungrounded line conductors were either service or feeder conductors, he can't mix service and feeders in the same raceway though.

Seems to me the service disconnect has to be the first enclosure after the meter, and the segment from the meter to that enclosure is never part of the loop. So the loop is always all feeders.

Cheers, Wayne

 

[tortuga]

Seems to me the service disconnect has to be the first enclosure after the meter, and the segment from the meter to that enclosure is never part of the loop. So the loop is always all feeders.

Cheers, Wayne

I think thats the case in diagram A and C but not B

 

[wwhitney]

I think thats the case in diagram A and C but not B

In B the "bypass switch" is labeled "service disconnect", so the triangle is all feeders.

Cheers, Wayne

 

[tortuga]

In B the "bypass switch" is labeled "service disconnect", so the triangle is all feeders.

Cheers, Wayne

Take a closer look the service conductors from the meter go to the red highlighted dot, when red is selected the bypass switch is the service disconnect and presumably the loadcenter is the OCPD for the service.
When the blue highlighted position is selected the service OCPD and disconnect is the PV system disconnect enclosure.
The red highlighted conductors and the blue could not share a raceway.

 

[wwhitney]

When the blue highlighted position is selected the service OCPD and disconnect is the PV system disconnect enclosure.
The red highlighted conductors and the blue could not share a raceway

I see from 230.91 that the service overcurrent device does not have to be integral with the service disconnect, but if it is not, it must be "immediately adjacent thereto." I guess that would permit the PV system disconnect, service disconnect/bypass switch, and loadcenter to all be in a row in adjacent enclosures. Kind of odd that the location of the service overcurrent device depends on the setting of the bypass switch, but I guess there's nothing prohibiting that?

In any event, with immediately adjacent enclosures, there would be no need for a raceway between enclosures, you could just use a chase nipple.

Cheers, Wayne

 

[tortuga]

I see from 230.91 that the service overcurrent device does not have to be integral with the service disconnect, but if it is not, it must be "immediately adjacent thereto." I guess that would permit the PV system disconnect, service disconnect/bypass switch, and loadcenter to all be in a row in adjacent enclosures. Kind of odd that the location of the service overcurrent device depends on the setting of the bypass switch, but I guess there's nothing prohibiting that?

In any event, with immediately adjacent enclosures, there would be no need for a raceway between enclosures, you could just use a chase nipple.

Cheers, Wayne

In terms of 230.7 I am pretty sure I have herd of inspectors turning down people for having other than service conductors in a chase nipple with service conductors, a common case is when people mount a ATS next to an existing meter.

 

[wwhitney]

In terms of 230.7 I am pretty sure I have herd of inspectors turning down people for having other than service conductors in a chase nipple with service conductors, a common case is when people mount a ATS next to an existing meter.

I don't think a chase nipple counts as a raceway, it's a fitting. A close nipple, that's less clear, the only thing distinguishing it from a full stick is the length.

But if you have to deal with that interpretation you could do put OCPD on the line side of the bypass switch. Or maybe use a short piece of non metallic innerduct to treat the chase nipple as two separate raceways for the purpose of 230.7, but treat it as one raceway for the purpose of 300.3(B).

Or, for a lot of trouble, you could cut aligned rectangles out of the sides of the two enclosures and replace them with non ferrous panels. And then use two non-ferrous chase nipples or the like. Then there's no 300.3(B) issue.

Cheers, Wayne

 

[tortuga]

Good points, though I am not a fan of a15kw (63A) optional standby generator automatically transferring to a 200A loadcenter, I'd go for a 70A critical loads panel and leave the other loads in a main panel.

 

[jaggedben]

Good points, though I am not a fan of a15kw (63A) optional standby generator automatically transferring to a 200A loadcenter, I'd go for a 70A critical loads panel and leave the other loads in a main panel.

A 200A load center means nothing without a load calc though. Of the over a hundred battery backup jobs I've done, I could probably go back to all but about three of them and pull data showing their consumption never exceeds 15kW. And yeah, most of those have 200A service that could provide five times as much power as they ever use.

 

[tortuga]

A 200A load center means nothing without a load calc though. Of the over a hundred battery backup jobs I've done, I could probably go back to all but about three of them and pull data showing their consumption never exceeds 15kW. And yeah, most of those have 200A service that could provide five times as much power as they ever use.

That would be interesting to see if the systems can record peak demand.
Here its pretty typical to have electric resistance water heater 4.5kw , electric range 11-12kw, 5kw dryer, some electric resistance heat and or heat pumps, plus all your standard fancy kitchen appliances not many people on propane anymore and methane gas utilities are just in bigger towns and just in downtown areas.
About 10 years ago when propane whole house generators were all the rage we had several people overload and stall 20kw generators on a 200A service, not that bad just necessitated replacing caps on the generator, but made me look more closely at the loads being automatically transferred.

 

[jaggedben]

Your milage may vary. Nameplates do not always mean than much especially on ranges. 20kw isn't anywhere near 200A. All I'm saying is a 200A service or loadcenter means basically nothing.

 

[PWDickerson]

I have done quite a few systems like this. I have generally followed your option C. With regard to the neutral loop, your choices are to use a 3-pole bypass switch so you can switch the neutral, or arrange the circuits so that both paths from the splice point to the bypass switch are run together like Wayne suggested (arranging all the equipment above a large gutter helps with this). I have done it both ways. Switching the neutral is the superior way to go in my opinion, but it can be tough to source that switch.

 

[solarken]

So I am still having trouble with a final approach to this problem. I switched to having the Sol-Ark Load output powering a critical loads panel instead of whole house backup and attached a diagram. I have tried to see how I can get the neutrals in a tree shape but I don't see a way to do that. The supply side conductors between MDP and PV disconnect can't occupy the same conduit as the conductors between the 100A bypass breakers in MDP and the Tfr switch, and I have to run N with the Hots in both paths. So with the Tfr switch in the solar position, power flow is from the supply side taps thru PV disconnect thru Sol-Ark (assume no sun) microgrid interconnect thru Tfr switch, to critical load panel (CLP), and return on Neutral from CLP thru Tfr switch, at which point it can split on way back to Grid thru the upper path and the lower bypass path, even though the Bypass hots carry no current. The 100A Tfr switch is readily avail in 2-pole and inexpensive at around $180. If I need to switch the neutral in the Tfr switch, it will be harder to find and likely around $1200 to 1400 bucks. I don't see how a gutter will help with this, even if I were to put a neutral bar in the gutter.

Any more ideas to avoid the 3-pole xfr switch? I know just omitting the bypass would get rid of the problem, but I am trying to provide a solution that eliminates problems the homeowner would have to deal with if the Sol-Ark inverter has major issues and does not pass thru power properly. In that case all their critical loads would be dead until I or another electrician could get onsite and route around the inverter.