Yes, but sometimes no. When you start getting over 600 amps of 48 volt batteries, the battery cabling, bussing, and breakers get pretty big.
Advice on two systems to be tied together
- Thread starter GuyG
- Start date
zanydroid
Solar Wizard
I was thinking of the management complexity not an increase in direct costs. Also, you may have a increase in battery costs if you have big appliances in both buildings. With two battery banks both may need to individually be big enough to handle this. It depends on if hybrid can be configured to export from battery in both directions, and in a load demand-driven basis.
Keep in mind that exporting to grid is NOT really load demand driven. The grid has an infinite demand, it’s up to the export policy and AC coupling to contain it.
We really need to have a naming convention for the buildings. A = utility connection building, B = other building. B is AC coupled follower of A. If a large load clicks on in A, that it’s local battery bank can’t handle, it seems quite unlikely that A’s hybrid knows how to frequency shift up B’s export in response. And highly debatable whether frequency shift can click on quickly enough anyway. So A needs to be sized big enough, inverter and battery wise, to handle its local loads.
zanydroid
Solar Wizard
Hmm that makes sense. Though another view is that At some point the batteries need to be AC coupled (to achieve higher voltage interconnect / smaller wires), or HVDC rather than 48v. Though I appreciate that 48v may be what is practical today.
Hybrid inverters parallel with the grid. And yes they have circuits that allows that to be done as well as disconnect if the grid goes down. 2 together must be in communication to parallel with each other properly and not back feed if zero export is selected. Now introduce a 3rd hybrid inverter some distance off that parallels to the the output of the combined 1st grid. But it is not set for zero export, indeed the OP wants it to feed back to the 1st grid depending on loading scenarios. How exactly does it know what the other grid needs?
zanydroid
Solar Wizard
I don’t think you can expect that third inverter to do any load sensing. As I said a few posts up, hybrids don’t sense what the utility grid needs, that’s not even part of the operating model. I think AC coupling to facilitate shifting a constant amount of power, with slow time scale of adjustment, is the most likely one that can be achieved
Could you explain what you mean by AC coupling?
BTW. The utility grid is considered a near infinite load for solar grid tie purposes. Up until recently all grid tie inverters sought to send as much as they could to the grid. Limits are now being sought by utility companies since all the solar power setups can be a problem if it starts taking all the loads away from the operating generators.
Last edited:
zanydroid
Solar Wizard
Combine the system by 120/240 power connection only, and use 1741SA curtailment, to address the limits issue you mention. With Frequency Watts being the only one I think implemented by grid forming inverters.
Utility generators (spinning, thermal, etc) typically respond much slower than PV, so it makes engineering sense to put the onus on PV inverters to hit the brakes, or inject more (if they were previously limited) to help with grid stability
Exactly, and as I mentioned in the OP, it would be the bigger of the two. It has the bigger loads, but only when it’s in use. And I don’t want to build a big system that has capacity for peak loads, without being able to share it with the smaller but more regularly used structure B.
And I’m not as worried about the ability of system B contributing upstream to system A, which will still be grid-tied and able to meet its temporary peak demands in excess of its capacity. It’s more important that system A’s large capacity not be wasted on the reg, and to do that it needs to share itself with B.
zanydroid
Solar Wizard
If B is connected as a load of A, A should be able to sense the power demand there and dispatch in real time as needed. Pretty normal hybrid capability.
zanydroid
Solar Wizard
B is still grid tied, right? It’s going to have a path to the grid via A’s hybrid. Since hybrids generally directly connect the grid AC to inverter output. So B’s inverter A’s inverter and grid are paralleled together.
Those were exactly my thoughts. The harder question is whether B can help out A sometimes through this admittedly unconventional form of AC coupling.
Which goes back my concern that UL1741SA inverter manufacturer (not to many of them) likely not approving this type of setup. Unapproved setups often void Warranty.
It would be interesting what the engineers at SolArk would say about it.
The plan is to cut the cord on B after everything is set up and running. B’s “grid connection” would come from the inverters at A.
zanydroid
Solar Wizard
I think when on PV, this is exactly the standard use case of FW controlled AC coupling. So it should work?
Battery export is required by SA to follow FW curtailment function. And supposedly tested by UL.
Not sure how many people here use FW to AC couple a battery to a grid forming inverter. Usually the battery is grid forming or paralleled with another inverter in a closed loop way.
zanydroid
Solar Wizard
You mean because the hybrid generally does not follow a grid former? (Microinverters are used this way all the time, I don’t know if it’s explicitly allowed)
And/or because the hybrid generally does not follow on battery power to a grid former?
I know the grid forming side often has a lot of conservative constraints on what is allowed to follow. In terms of kW output of the following inverter. And those are explicitly listed so more likely to be a cause for warranty issues.
It’s not hard to exceed those limits.
SilverbackMP
Solar Addict
- Joined
- Apr 4, 2022
- Messages
- 928
This sounds overly complicated. Why not pick one of the buildings as the main? Or better yet, build a solar shed.
I’m going to be powering my mom’s house, my shop/office/temp house, and other outbuildings from one system located in a 12x12 shed (which has been rebuilt and is now structurally strong AF).
It might even power my future house but I suspect I’ll need more amps by then.
I might tie it together but it’s gonna be two independent systems with transfer switches in case one goes down.
With SMA Sunny Boys, I can add AC power at any point.
I’m going to be powering my mom’s house, my shop/office/temp house, and other outbuildings from one system located in a 12x12 shed (which has been rebuilt and is now structurally strong AF).
It might even power my future house but I suspect I’ll need more amps by then.
I might tie it together but it’s gonna be two independent systems with transfer switches in case one goes down.
With SMA Sunny Boys, I can add AC power at any point.
zanydroid
Solar Wizard
sure. Another way to logically decompose this project is that you delete B’s grid connection, feed it from A’s grid connection. And then added Solar into it.
zanydroid
Solar Wizard
One reason I can think of to do an inverter in both buildings is to shorten the length/number of DC circuits. Even if the shed with all inverters is exactly in the middle.
OTOH I think the extra back breaking of pulling more cables needs to be traded off against simpler system design and ongoing operations. You don’t need that many brain cells to pull more conductors, and it only needs to be done once.
Solark touts its ability to accept AC coupling through its gen port or a load side sub panel. This would be through a panel. The only difference is that it would be coming from the downstream inverter’s “grid export” instead of its load output.
zanydroid
Solar Wizard
Yeah I think Sunny Boys are even explicitly listed for/have official system schematics showing how to AC couple to storage? I think putting those in the remote buildings to handle the PV on roof and ground mounts next to it, and then AC coupling back to one place with batteries and grid connection, is a pretty standard approach.