Is AC coupled even worth it?

[Ampster]

keeping things as AC coupled for solar to a house battery to me is very wasteful in that case.

I think the term AC coupling has different meanings depending on your perspective. I think the intent of the OP in framing the question in the title was fairly specific to the question of AC coupled solar and a hybrid inverter. In my case I would not even know how to quantify the waste since I think in terms of AC .

 

[realpinochet]

My original thought was to supplement a traditional style solar system by injecting solar generate AC current into the loads panel along with the battery inverters AC being injected. This thought came about after reading how much more efficient AC derived from a solar inverter was compared to converting DC to AC. Since I run a air conditioner all day during the summer I thought supplementing the load with a solar inverter for daytime usage might be worth looking at. I've learned throughout this thread what I really meant was AC/DC coupled. I'm debating if it's worth the trouble for me!

 

[JRH]

I don't plan on selling anything back, I don't plan on the local or federal gov having anything to do with my solar install at all. The PV inverter was going to inject AC right on the output bus to the panel along with the xw or into it's own sub panel. Are you saying there's zero way to not back feed into the grid with the PV inverter? As long as my meter doesn't roll backwards and I'm not backfeeding to hurt a lineman I don't see them caring.

That is one good aviator… and message .. go Bundy…!

 

[400bird]

3.) If you want MOST things actually are DC not AC(not that most of us would be you could run most things without ever converting to AC)

You'll have a difficult time powering them all a differing voltages. Sure, if you are in an RV, with limited roof space, it makes sense to run as much as possible on 12 v. But if this is a house with all the normal junk around the house, it's impractical. I'm not running 12 v to my TV, 19 volts to the media stuff, 5 volts to the USB alarm clock...
Sounds like a lot of time and effort running wires and DCDC converters all over the place if you want to run each and everything from straight DC.

4.) There are massive losses in being AC Coupled(which I think is really the main reason)

Wut? Do you mean something other than electrical/efficiency loses? My AC PV inverter is listed at 99% efficient. Hard to beat that when going straight to the load.
Sure, you lose some when the energy needs to go to battery and back out for overnight use. But, if you need to run AC all day...

The only reason we have these AC coupled systems mostly is because of the grid and selling power back to the grid.

That's a valid reason why they are so common, but not the only reason.

 

[Hedges]

My original thought was to supplement a traditional style solar system by injecting solar generate AC current into the loads panel along with the battery inverters AC being injected. This thought came about after reading how much more efficient AC derived from a solar inverter was compared to converting DC to AC. Since I run a air conditioner all day during the summer I thought supplementing the load with a solar inverter for daytime usage might be worth looking at. I've learned throughout this thread what I really meant was AC/DC coupled. I'm debating if it's worth the trouble for me!

AC coupling is ideal for that. And you get the GT PV inverter's kW capacity in addition to your battery inverter; that is typically sitting at zero, so full surge capability available.

If grid is connected and you don't want to backfeed, the two choices are disconnecting from grid and ramping up frequency, or having communication with the inverter. SMA supports SpeedWire (Ethernet) and some 3rd-party meter with CT for their later models.

Alternative, put a slightly undersized GT PV system behind the air conditioner contactor. "Guerilla grid-tie", it would never export.

 

[realpinochet]

AC coupling is ideal for that. And you get the GT PV inverter's kW capacity in addition to your battery inverter; that is typically sitting at zero, so full surge capability available.

If grid is connected and you don't want to backfeed, the two choices are disconnecting from grid and ramping up frequency, or having communication with the inverter. SMA supports SpeedWire (Ethernet) and some 3rd-party meter with CT for their later models.

Alternative, put a slightly undersized GT PV system behind the air conditioner contactor. "Guerilla grid-tie", it would never export.

I'm not sure how to frame this question so I'll do my best. Can you split the output from a single PV source to both charge controllers and a PV inverter? I plan on having, in the end, one large array south facing and one medium array west facing. I plan on combining these arrays in a combiner box and feeding it to the inverters. Can I split out the output from the combiner box to both the charge controllers and PV inverter? I was thinking once the batteries are full the charge controllers would shut down but the pv inverter would still be able to use the pv input and inject ac. What I'm trying to achieve is using the same arrays for both the charge controllers and pv inverter. It seems like the sunny boy or island might allow this. Do you have any input on sungrow pv inverters? A video I watched said that SMA was making there stuff in China now and sungrow was their competitor which seems to have high ratings.

 

[AntronX]

Can you split the output from a single PV source to both charge controllers and a PV inverter?

I tried that with 2 charge controllers fed in parallel from single PV array. One was Victron 150/60 charging 22V battery and another was Victron 75/15 charging 12V battery. During cloudy conditions the bigger 60 amp controller was being confused by competing MPPT algorithm from the smaller 15A controller. During full sun or when smaller 15A unit was maxed out the larger 60A unit did the tracking.

 

[Hedges]

In general you can't connect more than one SCC or PV inverter to a single PV array, but there might be ways to get away with it.

Newer inverters are transformerless, and PV+/PV- are driven to voltages related to AC line voltage peaks. Can't load that toward any DC voltage. Also can't load with much capacitance (only some PV panel types are compatible.)

Transformer type inverters have PV array either positive ground or negative ground. Some sort of ground-fault mechanism, such as a fuse or breaker which bonds PV- (usually) to ground. If you load toward a DC voltage, that will trip.

Some isolated/floating devices might operate in parallel. For instance if you got an isolated DC/DC converter you could connect it to one more more PV panels of a string to get desired voltage, and it would deliver isolated DC power. I kind of doubt any SCC would work like that, but possibly.

If you do have two MPPT operating in parallel they will probably confuse each other.

You could mechanically switch the array. Such DC voltage relays are going to be difficult to find. Manually, you can swap MC connectors or any switch rated AC or DC for the voltage, so long as no current flow. My idea is MC3 connectors inside disconnect switch with locking lid.

What I would do is just have the array go to GT PV inverter, let island-forming inverter AC couple to it and charge battery.
You can also have more panels DC coupled. I've tried a couple, experimenting, but that's not part of my system for now.

 

[realpinochet]

In general you can't connect more than one SCC or PV inverter to a single PV array, but there might be ways to get away with it.

Newer inverters are transformerless, and PV+/PV- are driven to voltages related to AC line voltage peaks. Can't load that toward any DC voltage. Also can't load with much capacitance (only some PV panel types are compatible.)

Transformer type inverters have PV array either positive ground or negative ground. Some sort of ground-fault mechanism, such as a fuse or breaker which bonds PV- (usually) to ground. If you load toward a DC voltage, that will trip.

Some isolated/floating devices might operate in parallel. For instance if you got an isolated DC/DC converter you could connect it to one more more PV panels of a string to get desired voltage, and it would deliver isolated DC power. I kind of doubt any SCC would work like that, but possibly.

If you do have two MPPT operating in parallel they will probably confuse each other.

You could mechanically switch the array. Such DC voltage relays are going to be difficult to find. Manually, you can swap MC connectors or any switch rated AC or DC for the voltage, so long as no current flow. My idea is MC3 connectors inside disconnect switch with locking lid.

What I would do is just have the array go to GT PV inverter, let island-forming inverter AC couple to it and charge battery.
You can also have more panels DC coupled. I've tried a couple, experimenting, but that's not part of my system for now.

If I'm understanding you correctly I'd just forgo SCC's and use the PV inverter in place? If that's the case I'd just split my arrays and have each one feeding a PV inverter installed at the array location, ground mounted, and then feed this back to the battery inverters. It seems I'd be able to supply direct AC for loads, charge batteries and use smaller wires ..saving money. Are these assumptions correct? What are the downsides not using a DC SCC?

 

[Hedges]

Downside of no DC SCC is that if battery dead, no AC, no AC coupling, no charging from GT PV.
I avoid that with a load-shed relay that disconnects all loads, but leaves GT PV connected and AC being produced with last usable 10% of of battery capacity.
Having some DC coupled could be useful.
In the case of Sunny Island, it won't even charge from AC if not operating due to dead battery (there is an emergency charge process to feed AC into the output, which I've never done.) I don't know about Schneider.

PV inverter in place of SCC is the way SMA systems work (although they did rebrand an SCC for a while.)
Typical inverter would be Sunny Boy 5000US, 5kW, 208/240/277V, max 600V input (my array is 480Voc, 380Vmp). You might buy for $500 used, or even new old stock if you're lucky. So $0.10/W with high voltage input.
I like to run PV wires the longer distance, AC shorter, because voltage drop on PV is pretty much unimportant. It is efficiency loss but otherwise doesn't affect operation (within MPPT voltage range.) I have 2500W strings connected by 150' (one way) of 12 awg, negligible loss.

Here's a diagram of AC coupling with Sunny Island. Schneider would be similar, but I have no experience with its behavior.

https://files.sma.de/downloads/SI6048US-DS-en-28.pdf

Example systems:

https://files.sma.de/downloads/SISYSGUIDE-KEN140421.pdf

Where AC coupling did have a hiccup for me was powering cheap VFD with diode/capacitor front end, poor power factor. That upset the Sunny Boy GT PV inverter (when it checked grid before connecting), but not the Sunny Island grid-forming inverter.

 

[Darren Orange]

Downside of no DC SCC is that if battery dead, no AC, no AC coupling, no charging from GT PV.
I avoid that with a load-shed relay that disconnects all loads, but leaves GT PV connected and AC being produced with last usable 10% of of battery capacity.
Having some DC coupled could be useful.

If some then a system should be designed to focus on DC if there is any advantage like this. Others had questions about losses so here is the problem.

DC to DC conversion which what a mppt charge controller is really doing has extremely low losses, maybe 2% at most. DC to DC is very efficient conversion.

Now converting DC to AC or back normally has around a 5-8% loss.

So lets start with DC coupled.
Solar Panels->MPPT(DC to DC -2%)->Battery(98%)->DC to AC(-6%)->Final output 92%

DC to AC to DC to AC
Solar Panels->MPPT Inverter (DC to AC(-8%)->AC to DC(-6%)->Battery(88%)->DC to AC(-6%)->Final output 80%

You effectively lose another 12% or more in real world use case. That being said this is only under the condition in which you push the power back to the battery so assume more than half your power will have this kind of loss or more because most systems with batteries will charge batteries obviously for the evening. Anyway my main point is it is less efficient and more points of failure.

 

[400bird]

You forget about using power during the day.
A quality grid tie inverter is 98-99% efficient. That is better than anything of the efficiency numbers you calculated above.

 

[Hedges]

An off-grid system which needs to save power on a sunny or overcast day for a future rainy day would get the most Wh into batteries using SCC, about 98% efficient.
That is assuming all MPPT algorithms are created equal, which they are not. With partial shading and any strings in parallel (or half-cut panels), poor algorithms can miss the peak operating point.

For me AC coupling was a no-brainer because I began with strictly grid-tie, and capacity exceeds consumption much of the time. When grid isn't present to store the surplus, there is greater kWh available from PV than battery can store so efficiency in PV --> battery path doesn't matter.

For systems where PV wattage exceeds desired charge rate, it is important to have a mechanism to measure battery current and varying what is harvested from PV to match loads plus desired charge.
Stand-alone SCC without communication can't achieve that because it doesn't know where the current is going.
SCC receiving communication from battery shunt (Victron with Cerbo, Midnight with Wiz Bang Jr) can do it.
AIO can do it
AC coupling can do it. (Sunny Island optionally uses battery shunt so external DC loads and charging are accounted for. It can't control them, but it can adjust the charging it performs from AC sources to increase current to target. Depending on firmware it may be able to receive data from SCC.)

 

[wheisenburg]

I have not decided on a inverter yet, I was looking at the xwpro or radian 8048. I plan on having a traditional system as in two inverters with two charge controllers and a rack of batteries with a ground mounted solar array. I will have grid connection to the inverters for grid assist and charging of batteries when there's no sun. But, in Texas we run the AC all day in the summer. The article said PV AC is more efficient in this case. So, I was trying to find out how to add a little PV AC into the mix for assisting with running the conditioner and a few other things and do it correctly. I would be AC/DC coupled I think. I do not want to send anything back to the grid. And, should the grid be down I'd still like to be able to use the PV AC and the incoming DC to the charge controllers. As stated it appears the way to achieve this is to couple it to the AC output side of the battery inverters and injecting AC right along that bus into the loads panel. I'm assuming the AC output from the xwpro keeps the PV inverter from shutting down...a false grid? It shouldn't back feed like this right?

Is this correct?

https://www.solar-electric.com/lib/wind-sun/Conext-XW-Pro-SW-AC-Coupling-Guide.pdf

View attachment 143059

When AC1 is connected to a working grid a relay is closed that connects AC1 and AC Out together. Since AC1 is 60 hz the GT inverters will run at 100% output (based on available sun of course). Unless you consume all of your PV power the system will export the excess back to the grid.

 

[wheisenburg]

This is data I collected this weekend. It shows that when running AC coupled IQ8s, they will come on briefly and then shut right down.

Right now Enphase has no solution for this.