diy solar

diy solar

Parallel Inverter Load Balancing

webbbn

Solar Enthusiast
Joined
Aug 9, 2023
Messages
345
Location
Arizona
My garage is on the opposite corner of the house, almost all of the high-current loads are located there, so I am considering putting in a subpanel with battery backup. I may add solar, but that would likely be AC-coupled, unless there's some advantage to running it into the inverters directly. What I would like to do is utilize a pair of existing 50A 240V lines that run to the exact spot where I would put the inverters and panel.

Can I connect each of the 50A lines to an 8-10kW (6kW might work as well) parallel-capable inverter, connect the outputs in parallel, and draw more than 50A from the output in passthrough mode and/or in battery mode? Will they load balance between the inverters to split the load? Can I charge a battery bank from the pair of inverters as well with it balancing the load?

If the answer is yes, then any suggestions on inverters to consider or avoid would also be apricated if someone has experience running in such a configuration. The inverters would need to be split-phase 240V output, but most of the load is 240V, so balancing the 120V legs should not be too much of an issue.

FWIW, it may or may not be grid-tied, depending on what I end up doing with the solar. I don't think the battery has to be grid-tied, though.

TL;DR: Do parallel inverters balance load across their AC inputs so that the combined total output can exceed the load capacity of the inputs individually? Is that code compliant, or is it equivalent to connecting two circuits together to try to get more capacity? My assumption is that each inverter could be configured to not draw more than 50A (or 40A if it has to be derated), so there wouldn't be an issue with overloading one line if the other disconnects.
 
Without meaning to be rude, the post is kind of rambling all over the place, is difficult to ascertain the objective. You will get better responses if you can summarize your goals.

AC coupling is primarily only done when there are existing grid-tie inverters and a battery back up inverter is added. Secondarily, in special cases where someone with NET metering installs a combination AC and DC coupled back up inverter. Not sure where the idea of AC coupling fits within the context of your situation.

Later the post mentions the system may not be grid-tied at all. Does that mean you don't intend to sell back or don't intend to have the grid connected at all? Based on the question of utility and inverter power blending together, the answer is you would need to use a higher end, more expensive inverter that is grid-interactive. UL1741SB. There is a 120% rule that states the combination of utility power and PV power cannot be more than 120% of the bus bar rating in the subpanel.

Typically PV and battery back-up inverters are installed to reduce monthly utility costs and keep things like refrigerators on in an outage not multiply the available power due to an appliance(s) that is oversized compared to the panel and you need more power.

Several of the popular inverters discussed on this forum have 60A (or more) pass-thru transfer relays so your existing 50A wiring and breaker could be used for the AC Input to a hybrid inverter.
 
Ok.. so on an inverter that has an internal transfer switch which can connect its AC input to its AC output, when it does so, the inverter itself has no control over what passes through that circuit other than to open the circuit actively, or to passively pop a fuse or breaker.

So if for example you install 2 inverters in parallel which both do 'ac passthrough', the only thing actually dictating how many amps flow through one or the other is the relative resistance of the two parallel paths. If they are similar, they will flow similar amounts of current. But the inverters do not actively 'control' that.
 
Regardless of the inverter capabilities, code forbids conductors small than 1/0 to be paralleled.
 
I am in europe, and i run 230v, but i have voltronic max axpert inverters, and when put in parallel configuration they are designed to share the load. They have a special cable to connect to the two inverters that will enable the load balancing between the two.

You are in us, and i dont know what inverters are you looking into, but i am sure that you will find in their manuals the specifics of load sharing or current sharing between them when you put them in parallel if they have this capability.

What i can tell you is that voltronic max axpert 8kw inverters when put in parallel will share the load and share the batteries. You can charge the batteries from the grid in usb mode too. They are economic and well tested.
 
That makes some sense. The Victron Quattro 10k has specs like this:

4 sets of three units can be parallel connected to provide 96W / 120kVA inverter power and more than 1600A charging

The input is designed for 6AWG wire, which, of course, will not support 120kVA. It does say "inverter power", but I don't see how you could connect 96kW (the specs must have a typo) of output and support passthrough without some sort of load balancing.

I'll have to take a look at the Voltronic and see if they would have a configuration that fits my needs.
 
Can I connect each of the 50A lines to an 8-10kW (6kW might work as well) parallel-capable inverter, connect the outputs in parallel,
No
As stated above, you can not parallel these existing circuits. Because they're too small to be paralleled.
You can use one or the other to feed the stacked AIO's. Or feed one to each, with separate loads.
What I would do is run the stacked AIO's off grid completely. And feed one or both existing circuits to chargeverters for grid assist.
 
Without meaning to be rude, the post is kind of rambling all over the place, is difficult to ascertain the objective. You will get better responses if you can summarize your goals.

AC coupling is primarily only done when there are existing grid-tie inverters and a battery back up inverter is added. Secondarily, in special cases where someone with NET metering installs a combination AC and DC coupled back up inverter. Not sure where the idea of AC coupling fits within the context of your situation.

Later the post mentions the system may not be grid-tied at all. Does that mean you don't intend to sell back or don't intend to have the grid connected at all? Based on the question of utility and inverter power blending together, the answer is you would need to use a higher end, more expensive inverter that is grid-interactive. UL1741SB. There is a 120% rule that states the combination of utility power and PV power cannot be more than 120% of the bus bar rating in the subpanel.

Typically PV and battery back-up inverters are installed to reduce monthly utility costs and keep things like refrigerators on in an outage not multiply the available power due to an appliance(s) that is oversized compared to the panel and you need more power.

Several of the popular inverters discussed on this forum have 60A (or more) pass-thru transfer relays so your existing 50A wiring and breaker could be used for the AC Input to a hybrid inverter.

Thank you. I probably have more information than was necessary, most of it not relevant to what I was wanting to know, other than for suggestions on inverters, which could be different if it was intended for grid-tie. The AC-coupling comment was in relation to the "special case" that you mention, which doesn't (in my limited experience) seem to be a special case, but the norm. All of the solar installers that I've talked to seem to prefer AC-coupled microinverters and AC-coupled storage.

I don't know what the rest of your response has to do with my question. I'm not trying to overload a panel. My question has to do with wiring/breaker size on the input of parallel connected inverters.
 
No
As stated above, you can not parallel these existing circuits. Because they're too small to be paralleled.
You can use one or the other to feed the stacked AIO's. Or feed one to each, with separate loads.
What I would do is run the stacked AIO's off grid completely. And feed one or both existing circuits to chargeverters for grid assist.
I just noticed the Victron 3kW has 6AWG terminals, the others have bolts, so my assumption that it has to be load sharing is not necessarily valid. I thought I'd seen other examples where the input wire size seemed undersized compared to the maximum parallel output, but I don't know for sure.
 
All of the solar installers that I've talked to seem to prefer AC-coupled microinverters and AC-coupled storage.
Of course they do, a fully integrated professionally installed, Enphase system for example will cost 2-3 times as much as a DIY or Partial DIY system. If your budget is sufficient, there is nothing at all wrong with going that route and having the manufacturer warranty. Since this is a DIY forum, I may have wrongly assumed you were interested in cost considerations as well as answers to some technical questions.
 
I just noticed the Victron 3kW has 6AWG terminals, the others have bolts, so my assumption that it has to be load sharing is not necessarily valid. I thought I'd seen other examples where the input wire size seemed undersized compared to the maximum parallel output, but I don't know for sure.
Stacked AIO's do in fact share the loads if paralleled.
Not exactly 50/50 but very close to it.
You can definitely undersize the input. But when it switches to bypass. The source breaker could trip due to overload.
It really depends on what you want to accomplish. In order to determine what will work best.

Battery backup.
More power at the remote location.
Shifting power usage to a cheaper time of day.
Moving loads from grid to solar.
 
Of course they do, a fully integrated professionally installed, Enphase system for example will cost 2-3 times as much as a DIY or Partial DIY system. If your budget is sufficient, there is nothing at all wrong with going that route and having the manufacturer warranty. Since this is a DIY forum, I may have wrongly assumed you were interested in cost considerations as well as answers to some technical questions.
In my case, installing panels involves installing on and ripping up a second story tile roof, which is beyond what I want to DIY, and I want to make sure it's done right and the roof is warrantied, etc. My current impression is that microinverters don't add to much to the entire panel installation, and it seems like it would give me more freedom on the storage side, which I believe I can tackle and do much less costly.

Another solution that I've considered is just to have an AOI (Sol Ark 15k preferred) installed by the panel installer and go from there, but, as I said, the electrical panel is not in a good location for such an install.
 
Stacked AIO's do in fact share the loads if paralleled.
Not exactly 50/50 but very close to it.
You can definitely undersize the input. But when it switches to bypass. The source breaker could trip due to overload.
It really depends on what you want to accomplish. In order to determine what will work best.

Battery backup.
More power at the remote location.
Shifting power usage to a cheaper time of day.
Moving loads from grid to solar.
Battery backup and time of use.

Maximum usage for the entire house runs around 9kW, with my Vue showing a peak usage of 14kW (likely when the EV charger was running). The plan is to move around 8kW of that to the battery backed circuit, so normal total usage should not exceed a single 50A circuit. The parallel configuration is intended more as a buffer and to pick up surges.
 
Battery backup and time of use.

Maximum usage for the entire house runs around 9kW, with my Vue showing a peak usage of 14kW (likely when the EV charger was running). The plan is to move around 8kW of that to the battery backed circuit, so normal total usage should not exceed a single 50A circuit. The parallel configuration is intended more as a buffer and to pick up surges.
There are many different ways to accomplish your goals. But connecting the two existing circuits in parallel can not be one of them.
As I stated before. If you want to utilize the second circuit. I would use a chargeverter to do so.

If you plan to use a professional installer. Your choices are limited to what they offer. Which is almost always micro inverters and a grid-tied system.
Because it's the easiest to install and is the most profitable for them. If they offer a battery backup solution, it will be outrageously expensive. To be worth it to them. Due to complexity and customer misuse. Which translates to call backs. (Profits lost)
 
If you plan to use a professional installer. Your choices are limited to what they offer. Which is almost always micro inverters and a grid-tied system.
Because it's the easiest to install and is the most profitable for them. If they offer a battery backup solution, it will be outrageously expensive. To be worth it to them. Due to complexity and customer misuse. Which translates to call backs. (Profits lost)
It is amazing that in Italy we have the exact same situation with professional installer.
 
Stacked AIO's do in fact share the loads if paralleled.
Not exactly 50/50 but very close to it.
You can definitely undersize the input. But when it switches to bypass. The source breaker could trip due to overload.
It really depends on what you want to accomplish. In order to determine what will work best.
There are many different ways to accomplish your goals. But connecting the two existing circuits in parallel can not be one of them.
As I stated before. If you want to utilize the second circuit. I would use a chargeverter to do so.
I'm trying to understand this better, and I found some actual documentation on how to wire the XW Pro in a parallel configuration that might explain the problem. The problem in this case is that the bypass switches might not all switch at the exact same time, so one could get the full load for a fraction of a second and trip the breaker. They sell a special switch for dealing with this situation, which I think would fit my use case, but it's pretty expensive.

That brings me to the question of, is this a code issue, or a potential issue in actual use, or have I missed the point entirely? I can pop a breaker on any line in my house by plugging in too much stuff. The documentation on the XW Pro indicates that you only need the extra switch if you drive more than 60A total across multiple units. Would an inspector add up all the loads on the panel to ensure that the load cannot exceed 60A in a parallel XP Pro configuration?

If it's a problem with wiring code, then can someone explain how one would, say wire 6 EG4-6500x in parallel, which could output a maximum of 35kW (not including charging) when it only supports (AFICT) 4 gauge wire? That's a serious question, since I'm pretty sure I've seen suggestions of using a larger wire to the panel, and branching off to smaller wires to the inverters, which seems worse than individual smaller wires with breakers.
 
I'm trying to understand this better, and I found some actual documentation on how to wire the XW Pro in a parallel configuration that might explain the problem. The problem in this case is that the bypass switches might not all switch at the exact same time, so one could get the full load for a fraction of a second and trip the breaker. They sell a special switch for dealing with this situation, which I think would fit my use case, but it's pretty expensive.

That brings me to the question of, is this a code issue, or a potential issue in actual use, or have I missed the point entirely? I can pop a breaker on any line in my house by plugging in too much stuff. The documentation on the XW Pro indicates that you only need the extra switch if you drive more than 60A total across multiple units. Would an inspector add up all the loads on the panel to ensure that the load cannot exceed 60A in a parallel XP Pro configuration?

If it's a problem with wiring code, then can someone explain how one would, say wire 6 EG4-6500x in parallel, which could output a maximum of 35kW (not including charging) when it only supports (AFICT) 4 gauge wire? That's a serious question, since I'm pretty sure I've seen suggestions of using a larger wire to the panel, and branching off to smaller wires to the inverters, which seems worse than individual smaller wires with breakers.
It is a code issue.
And I agree that the same thing is being done with a distribution panel ahead of the stacked AIO's. And the only difference is that the distribution panel is in the same proximity as the AIO's. And dedicated for that purpose.
You have to understand that when power in a single circuit is fed from two different breakers. If you turn one off. The wires connected to it are still live. (Being fed by the other breaker)
Personally, I don't know if this dedicated distribution panel actually meets code. But, it seems to have been slipping through. Maybe because you can disconnect both ends (Inverter input and output) and isolate it.
On the load side (Inverter combiner box) it makes more sense. Because the multiple sources are working in conjunction.
 
It is a code issue.
And I agree that the same thing is being done with a distribution panel ahead of the stacked AIO's. And the only difference is that the distribution panel is in the same proximity as the AIO's. And dedicated for that purpose.
You have to understand that when power in a single circuit is fed from two different breakers. If you turn one off. The wires connected to it are still live. (Being fed by the other breaker)
I thought that was supposed to be prevented by anti-islanding. Are you saying that if I turn off power to one inverter, the other inverter will back feed through to the other line? I wouldn't expect the wires to be directly connected. One set should power one inverter, one the other. Obviously it's against code to directly wire two circuits together.
 
I thought that was supposed to be prevented by anti-islanding.
Anti-islanding applies to grid-tied and exporting.
You asked about EG4 6500 off grid AIO's.
Are you saying that if I turn off power to one inverter, the other inverter will back feed through to the other line?
In the case of stacked AIO's, yes. If you only turn off one unit. It's output wires are still live. (Being powered from the paralleled output connection)
I wouldn't expect the wires to be directly connected. One set should power one inverter, one the other.
You are correct in regard to the input circuits.
The two circuits aren't paralleled, until after they pass through the AIO.
Now, I'm not sure if it would pass code or not. It would be up to the interpretation of your local AHJ. It seems to be acceptable for most. When the distribution panel is in close proximity.
And electrically, it's no different at a further distance.
I take back my earlier stance.
 
Back
Top