Using the GEN input for AC coupled solar?

[wheisenburg]

Sol Arc and the new EG4 18,000 both recommend connecting AC coupled solar to the Gen port.

Normally when your system is grid tied excess PV is routed to the grid. In a grid down situation that is not an option. The Hybrid inverter will then become the grid forming inverter and the PV inverters will couple to that an begin exporting power to the Hybrid inverter. The Hybrid inverter will need to store the power in the batteries. When the batteries are full the hybrid inverter can use frequency shifting to shut down or curtail the PV inverters

My understanding of why Sol Arc uses the Gen input is that managing the curtailment of the AC coupled inverters is not instantaneous and it is not perfect. It depends on having some room left in the batteries so that you can dump the excess power into them. Sol Arc, I know suggests that you setup your installation so that it cycles between 80% and 90% SOC. So when the SOC <= 80% it will close the GEN connections and keep it connected until SOC >= 90%. The gen input will then stay off until SOC <= 80%.

My assumption is that they are doing this because they feel allowing the AC coupled inverters to operate above 90% SOC could cause the system to become unstable. After 1.5 years of not having to worry about this because the PV inverters would not turn on when not connected to the grid, that problem was finally solved

My question is why does Connect WX Pro not use this same method? They have no way to disconnect the PV inverters from the micro grid. Is there a potential instability issue because of the inability to quickly turn PV generation off when operating with batteries that are fully charged?

 

[1201]

I don't know why they recommend the gen input. Sma sunny island has no gen input and will ac couple at 100% soc.

A lot of data so you may have to view in full screen.

The pv inverter and sunny island output are tied together in a subpanel. Sunny island controls the frequency at this subpanel to control the output of the pv inverter.


Also, the 18kpv and solark 15k should be able to throttle on their outputs, Even if the pv is not connected on the gen input.

 

[wheisenburg]

I don't know why they recommend the gen input. Sma sunny island has no gen input and will ac couple at 100% soc.

A lot of data so you may have to view in full screen.

The pv inverter and sunny island output are tied together in a subpanel. Sunny island controls the frequency at this subpanel to control the output of the pv inverter.


Also, the 18kpv and solark 15k should be able to throttle on their outputs, Even if the pv is not connected on the gen input.

I have have looked into their solution. It looks good, but I believe the system uses both an SMA Hybrid and PV inverter. Working together this pair of devices is actually able to manage the PV output more effectively. If you go with a full Enphase solution with both their battery and micro inverters, it is similar situation.

I am guessing the reason they are showing this video bragging about how they can AC couple at 100% SOC is that most other solutions can't do it.

 

[1201]

I have have looked into their solution. It looks good, but I believe the system uses both an SMA Hybrid and PV inverter. Working together this pair of devices is actually able to manage the PV output more effectively. If you go with a full Enphase solution with both their battery and micro inverters, it is similar situation.

I am guessing the reason they are showing this video bragging about how they can AC couple at 100% SOC is that most other solutions can't do it.

I made this video to show how ac coupling works at 100% soc.

Also the pv inverter is a growatt not an sma.

I too am curious to know why they recommend using the gen input. Maybe so they can track the production. I think I read that somewhere?

 

[wheisenburg]

I made this video to show how ac coupling works at 100% soc.

Also the pv inverter is a growatt not an sma.

I too am curious to know why they recommend using the gen input. Maybe so they can track the production. I think I read that somewhere?

I think they state two reasons.

1. It allows them to track your AC coupled production.

2. The ability to cycle the PV inverters on and off. The big issue with AC coupled solar is when you have rapidly changing loads or PV production. If the sun suddenly breaks out behind the clouds or an large A/C system turns off, you get a mismatch between production and usage. So the Hybrid inverter can throttle down the PV Inverters through frequency shifting. In the mean time it will need to send the excess production to the batteries.

Maybe low frequency inverters handle this better? In any case having more battery capacity apparently helps because you can push more power into them without the battery voltage spiking. This is supposed to make the system more stable. Sol Arc also recommends that DC coupled PV > AC couple PV.

 

[1201]

,

I think they state two reasons.

1. It allows them to track your AC coupled production.

2. The ability to cycle the PV inverters on and off. The big issue with AC coupled solar is when you have rapidly changing loads or PV production. If the sun suddenly breaks out behind the clouds or an large A/C system turns off, you get a mismatch between production and usage. So the Hybrid inverter can throttle down the PV Inverters through frequency shifting. In the mean time it will need to send the excess production to the batteries.

Maybe low frequency inverters handle this better? In any case having more battery capacity apparently helps because you can push more power into them without the battery voltage spiking. This is supposed to make the system more stable. Sol Arc also recommends that DC coupled PV > AC couple PV.

Ok I think I understand why.

They have operation modes the sunny island doesn't have.

There are modes where the inverter is always connected to the grid. For example when it's supporting the main panel loads.

If it's connected to the grid then the input and output share the grid frequency and the inverters cannot frequency shift to control the output of the pv inverters so the only way they can establish control eg zero export is to be able to disconnect the pv inverter and they wouldn't be able to do that unless you have the pv inverter output going into the solark or eg4.

That makes sense to me at least

 

[RCinFLA]

Using a HF hybrid inverter for AC coupling is dicey at best. The reason why SolArc recommends reconfiguring Gen port to inject AC coupled GT inverter is to give them the ability to shut down the AC connection to GT inverter by opening the Gen port pass-through relay if they get into trouble with overproduction with no place to dump excess power to. It is a safety net for them to prevent possible inverter damage.

LF inverters are inherently bidirectional and flip from sourcing to sinking on the fly. LF inverter will instantly dump excess PV power to batteries for the several seconds it takes for frequency shifting to reduce PV GT inverter output.

HF inverters must make a power direction change when they are supplementing AC output from the battery to HV DC converter and have to change over to charging battery mode to absorb excess PV GT power. This takes several milliseconds to accomplish. This time may not sound like much, but it is enough time to damage inverter.

Without a grid present to dump PV GT power to, the primary risk is when you have large GT PV generation dumping in AC coupling with AC output loads that consumes a significant portion of the PV GT inverter generated power. If a large load is switched off there will be instantaneous, and possibly significant overproduction of PV GT power the inverter must deal with.

Frequency shifting to reduce GT inverter power is not fast enough to handle this situation on HF inverters.

SolArk and Deye have a large bank of HV DC capacitors that can absorb a small amount of excess production but even the large capacitor bank is limited on what surge in power it can absorb. The primary purpose of the HV DC capacitor bank is to allow directional power change during load shaving where it switches from charging battery to supplementing an AC output for a sudden output AC load surge.

 

[1201]

Frequency shifting to reduce GT inverter power is not fast enough to handle this situation on HF inverters.

I believe that if the hybrid inverters are connected to the grid in sync, they use the grid frequency and therefore cannot curtail the production of the pv inverter via frequency shifting. They have to disconnect the gen input

 

[RCinFLA]

I believe that if the hybrid inverters are connected to the grid in sync, they use the grid frequency and therefore cannot curtail the production of the pv inverter via frequency shifting. They have to disconnect the gen input

You don't need (and cannot do) freq shifting when you have a grid to push excess PV power to.

However, you never know when the grid may take a dump. Inverter always needs to be prepared to handle a PV excess overpower situation.

 

[1201]

You don't need (and cannot do) freq shifting when you have a grid to push excess PV power to.

However, you never know when the grid may take a dump. Inverter always needs to be prepared to handle a PV excess overpower situation.

What if you want to do zero export but still be connected to the grid? Still can't frequency shift and the power has nowhere to go- that's why I think they use the gen port

 

[wheisenburg]

However, you never know when the grid may take a dump. Inverter always needs to be prepared to handle a PV excess overpower situation.

So the question I have about all of this has to do with the Schneider XW Pro 6848 inverters. They have a setting for the Max SOC or DC volts where the system will attempt to charge the batteries. Above that, presumably they would use frequency shifting to reduce or turn off power production. My system has two XW Pro inverters and three 100 amp batteries. On my best production days, I might see power output of say 9000 watts. That would require potentially pushing about 60 amps into each battery. I was planning to eventually add three more batteries and that would drop to 30 amps. Right now I don't have these extra batteries.

I have had the system for about 1.5 years. During most of that time, the IQ8+ inverters would never turn on in the event of a power failure. Enphase fixed that issue with new firmware. Normally, my Hybrid Inverter system runs as a backup UPS. I have full 1 to 1 net metering. When the grid is up I export all excess PV generation. The system is in Pass though mode unless the Grid fails. I let the batteries discharge to around 93% and then charge it back up to 100% (it cycles every 3-4 days). When the system is operating Off Grid in AC coupled mode it is supposed to have logic that will turn off charging before the batteries reach a high state of charge. There is however no way to actually physically disconnect the PV inverters to prevent a problem.

So if there were a situation where I had an outage on a really sunny summer day while the batteries were topped off, would I have a problem? I have run the system off grid to simulate an outage, but only with the batteries at 90-95%. Also, it was not a high power production day. Everything seemed to work fine.

Possible Enhancement:

Each inverter has a programable 12 volt relay. They can be programmed independently. So that I could have one 12 volt signal for SOC < 86% and another for SOC < 91%. I could also add a relay with a 240 Volt coil to sense the grid status with a "Grid Connected" signal. Now I realize there is a small chance that there was enough power on the grid to turn on the relay, but a problem with power quality that prevented the AC 1 from actually connecting. So I could have a false positive, but I would be no worse off than I am today. These three signals could be used to make sure my PV does run at high power when the batteries are full by adding 2 contactors and some relays. I have three strings that produce about 3000 watts each.

So it would work like this:

String 1: If (SOC < 86% OR Grid Connected) turn on Contactor 1
String 2: If (SOC < 91% OR Grid Connected) turn on Contactor 2
String 3: Always connected no contactor.

If I were to move forward with this, I would install the control wiring in an electrical cabinet above my Envoy box. The wires going to the PV arrays could then be intercepted and controlled. Now if I might have an issue, I would be willing to solve it this way. If I don't have a problem, then there is no reason to solve a problem that doesn't exist.

Materials:

1. 1 Electrical cabinet
2. 1 240 volt AC relay.
3. 2 12 volt DC relays.
4. 2 24 volt DC contactors.
5. 24 volt DC power supply.
6. EMT conduit.
7. Wires and fittings.

Any thoughts? Has anyone blown up an inverter due to a power surge while running Off Grid with AC coupled?

 

[wheisenburg]

What if you want to do zero export but still be connected to the grid?

If you bought Grid tied PV inverters in the first place, it likely due to the fact that you built the system out as a grid interactive system and got the required permission to operate. Most of the AC coupled Hybrid / PV inverter systems are the result of someone adding battery back up to an existing system. In Europe there are many people using AC coupled solar with AC coupled batteries that have no hybrid inverter. The system is there just to manage time of use. All the equipment is grid tied and doesn't work when the grid goes down. You can use it to store electricity during the day and use that electricity at night or sell it back to the power company. Because electricity costs far more in Europe than in the USA and they have more Time of Use rates there, these systems make sense for them economically. Grid down backup power for a whole house is more popular here, but is very expensive still.

There are settings on the Schneider Inverters that assist you in doing a number of things. They are rather convoluted and people have trouble understanding how they work. Mostly they are going to be geared towards managing power from DC coupled Solar and the Battery. You can control when and how fast it will charge the battery from DC PV or Grid AC. You can setup peak shaving. You can also setup "Grid Support", this can sell power back to the grid. These features are designed to shift around when you use your grid power, in order to take advantage of Time of Use Tariffs. You can also setup "Grid Support" to sell electricity from the battery into the grid during times when you get the best price.

AC coupled seems like a great idea until you actually try to make it work in a grid down situation. So when the grid is down.

DC coupled is "Plug and Play".
AC coupled is "Plug and Pray".

 

[1201]

.
Any thoughts? Has anyone blown up an inverter due to a power surge while running Off Grid with AC coupled?

That shouldn't if you have the recommended amount of battery

 

[RCinFLA]

What if you want to do zero export but still be connected to the grid? Still can't frequency shift and the power has nowhere to go- that's why I think they use the gen port

That is true.

On HF inverter, still most likely to happen when you switch off a large load. On SolArk, Deye, they can absorb small amount of excess in their HV DC capacitor bank for the short amount of time necessary to switch over to charging mode to dump excess to batteries.

Most inverters, when AC coupling enabled, will not allow battery to be fully charged to keep a safety reserve 'dump' margin. This applies even when grid selling is enabled due to unpredictable grid drop out possibility.

AC coupling on a LF inverter is less risky but still not risk free. Most common thing on LF inverter is over-voltaging DC input due to insufficient battery size or too much battery cable path voltage drop. Most common mistake is folks think they don't need large AH batteries because they have so much PV power available.

If maximum inverter DC input voltage is exceeded, the only option the inverter has is to shut down to protect itself.

Best way to make it fairly safe is with a quick reacting AC load dump.

 

[wheisenburg]

Best way to make it fairly safe is with a quick reacting AC load dump.

What about adding relays to shed PV when SOC is high? If the system shuts down due to DC overvoltage without causing any damage maybe it's not worth it?