Grid-tied inverter powered by off-grid inverters

[solarhombre]

Hi there,

I have solar panels connected to a grid-tied inverter that feed a breaker in my main electrical panel. An additional critical loads panel with a manual interlock is connected to 2xEG4 6500 and 6x EG4LLv2 batteries.

Now, imagine there's a grid outage. My grid tied stops producing energy and the house is powered by the batteries. After a day or two batteries are running low.

What would happen if I open the main breaker (disconnect from the grid) and bypass the interlock so that the EG4 inverters power the grid tied inverter? I have the feeling something will break if the power generated by the panels on the grid tied inverter cannot be absorbed by the load on the EG4 inverters, including battery charging.

 

[Samsonite801]

Can you draw a picture of the described scenarios? I'm still trying to get my coffee to kick in this morning, and it's hard to follow for me.

Maybe someone else will get it better without a picture, but pictures are fun too..

 

[BentleyJ]

The EG4 6500 is an Off-Grid inverter. As far as I know it doesn't have AC coupling capability.

 

[solar8484]

If you are lucky the EG4's will just shutoff but they may also let out the magic smoke.

 

[Samsonite801]

The EG4 6500 is an Off-Grid inverter. As far as I know it doesn't have AC coupling capability.

Best I can tell, he may be referring to how when grid goes down, the grid-tied inverter, shuts down since it needs to see the grid voltage reference, so then solar production stops, so he wants to somehow feed an AC signal to the grid-tied system (from the EG4 inverter AC output), so the solar can produce and somehow get it to charge his batteries connected to the EG4.

If that is the case, I would think he would need a Chargeverter (to charge the batteries) to decouple the AC from the grid-tied system and not use the EG4 AC input while in this mode of operation.

I could be totally off though on my interpretation of what is the topology here.

 

[solar8484]

If that is the case, I would think he would need a Chargeverter (to charge the batteries) to decouple the AC from the grid-tied system and not use the EG4 AC input while in this mode of operation.

Assuming you are talking about powering a Chargeverter from the GTI AC output when the grid is out, the idea is a non-starter. What would get the GTI to produce power? Even if GTI is producing power, excess power will likely damage the Chargeverter.

 

[solarhombre]

Best I can tell, he may be referring to how when grid goes down, the grid-tied inverter, shuts down since it needs to see the grid voltage reference, so then solar production stops, so he wants to somehow feed an AC signal to the grid-tied system (from the EG4 inverter AC output), so the solar can produce and somehow get it to charge his batteries connected to the EG4.

If that is the case, I would think he would need a Chargeverter (to charge the batteries) to decouple the AC from the grid-tied system and not use the EG4 AC input while in this mode of operation.

I could be totally off though on my interpretation of what is the topology here.

Yes, this is exactly what I would like. This solution of adding the chargeverter and keeping the off-grid inverters not connected to the grid-tied inverter looks very interesting. What would happen if the panels produce more than what the chargeverter can take?


Adding two diagrams here. The second one has no grid and new lines for the off-grid inverters to feed the grid-tied one.

 

[solarhombre]

The EG4 6500 is an Off-Grid inverter. As far as I know it doesn't have AC coupling capability.

Right. The idea is to use the EG4 6500s as the "grid" so the grid-tied inverter would sync to what the off-grid is feeding it.

 

[zanydroid]

Right. The idea is to use the EG4 6500s as the "grid" so the grid-tied inverter would sync to what the off-grid is feeding it.

Right but that feature is often called AC coupling (unfortunately not standardized in industry spec sheets). It’s not as simple as you make it out to be

 

[Samsonite801]

Yes, this is exactly what I would like. This solution of adding the chargeverter and keeping the off-grid inverters not connected to the grid-tied inverter looks very interesting. What would happen if the panels produce more than what the chargeverter can take?

It doesn't work like that. The panels only produce what is needed based on the demand (draw). When there is no demand, they sit there idle, like zero throttle (basically open circuit), just wasting away Sun...

 

[zanydroid]

It doesn't work like that. The panels only produce what is needed based on the demand (draw). When there is no demand, they sit there idle, like zero throttle (basically open circuit), just wasting away Sun...

(Re earlier in thread): How would the chargeverter generate a grid for the GTIs?

Re: not sending more than asked for. GTIs have a control loop that seeks to go to the MPP of the solar panels while also pushing all of it to grid. They will shift the PWM to try to do so, I believe generally by increasing the AC voltage. Eventually something will get unhappy, GTIs will either disconnect or the charger will blow up.

The chargeverter also does not know that the GTI is power limited, it has no such control loop to detect if the inverters are sagging due to trying to pull too much power.

 

[Samsonite801]

(Re earlier in thread): How would the chargeverter generate a grid for the GTIs?

The Chargeverter is only to charge DC batteries, not to generate anything AC (since it is a DC charger).. The 6500EX (AC output) was supposed to generate the fake 'grid' AC here.

Re: not sending more than asked for. GTIs have a control loop that seeks to go to the MPP of the solar panels while also pushing all of it to grid. They will shift the PWM to try to do so, I believe generally by increasing the AC voltage. Eventually something will get unhappy, GTIs will either disconnect or the charger will blow up.

So are you indicating that GTI don't regulate or operate within any voltage limits, they just keep turning up MPPT (even if nothing is drawing down AC volts) until the moon is reached? So it wouldn't act like a gas generator, and just turn down the field coil so to speak, when nothing is plugged into it just to maintain a certain maximum voltage? (based on some kind of AVR) Don't they have CT's or something that they use for feedback?

I'll admit that I know very little about GTI's but that doesn't seem like it would be all that well thought out if it just keeps trying to pump out higher voltage until infinity.

The chargeverter also does not know that the GTI is power limited, it has no such control loop to detect if the inverters are sagging due to trying to pull too much power.

I'm not especially getting this part here, can you rephrase maybe? Which inverters? Inputs or outputs, I'm a bit lost. I'm really trying to understand.

I would have thought that a GTI would just turn itself down (relax MPPT), once some peak AC voltage is reached?

I'm honestly trying to learn something new here, since my hands-on experience is more with off-grid scenarios where generator is the closest thing we have to any grid on an input, my limited knowledge of grid-tie equipment is more textbook and forums.

 

[sunshine_eggo]

GTI do not regulate themselves. They expect an infinite grid in which to dump whatever the array can generate.

An off grid inverter must have the ability to AC couple or frequency shift to be able to serve as grid for a GTI. The EG4 do not have this.

An inverter capable of ac coupling can use the incoming GTI power to charge the batteries, but once the batteries hit absorption, there is nowhere for the power to go. The off-grid inverter then shifts the frequency away from 60Hz to force the GTI to disconnect.

 

[Samsonite801]

GTI do not regulate themselves. They expect an infinite grid in which to dump whatever the array can generate.

An off grid inverter must have the ability to AC couple or frequency shift to be able to serve as grid for a GTI. The EG4 do not have this.

An inverter capable of ac coupling can use the incoming GTI power to charge the batteries, but once the batteries hit absorption, there is nowhere for the power to go. The off-grid inverter then shifts the frequency away from 60Hz to force the GTI to disconnect.

Sounds reasonable, thanks for clearing that up grid-tie for dummys style for me, so I guess OP's proposal wouldn't work. I never liked the idea of installing solar that doesn't have a battery, and requires a grid AC in order to function, it kind of kills a lot of the benefits of having solar for power redundancy in the first place.

Say I bought a house that had a GTI system like that, the first thing I'd do is ditch the GTI altogether and install a hybrid inverter and batteries, and reconnect all the solar panels to the hybrid inverter only. To not use any equipment that can't function without a grid source.

Oh well, I know we're all just trying to work within money constraints too, we all wish we had the perfect setup, it comes in time I guess, work within your means.

 

[Hedges]

For many/most GTI systems that implement Rule 21, you could keep the system and add a battery or hybrid inverter that does support AC coupling.

Compare the cost of that with rewiring panels directly to a hybrid (may have to deal with RSD or optimizers), also the capability and quality of the different products, so see if it is what you want.

When adding an inverter that has battery, consider whether peak shaving and time shifting is desired; that varies among products.

 

[MajicDiver]

Hi there,

I have solar panels connected to a grid-tied inverter that feed a breaker in my main electrical panel. An additional critical loads panel with a manual interlock is connected to 2xEG4 6500 and 6x EG4LLv2 batteries.

Now, imagine there's a grid outage. My grid tied stops producing energy and the house is powered by the batteries. After a day or two batteries are running low.

What would happen if I open the main breaker (disconnect from the grid) and bypass the interlock so that the EG4 inverters power the grid tied inverter? I have the feeling something will break if the power generated by the panels on the grid tied inverter cannot be absorbed by the load on the EG4 inverters, including battery charging.

A good many people on this forum are doing it, but are not using EG4 equipment. I am doing it with Growatt inverters which ac couple/frequency shift. I have also experimented on my system with ac coupling turned off on my grid forming inverter and with carfull tuning of the power meter (load ramp rate) I could keep the system running. It’s a lot easier with ac coupling turned on. others are doing it with sunny island or solark.

Read my profile if you want any info on my system setup.

 

[zanydroid]

The Chargeverter is only to charge DC batteries, not to generate anything AC (since it is a DC charger).. The 6500EX (AC output) was supposed to generate the fake 'grid' AC here.

Fair enough, but I think there's a possibility of a circular loop here infinitely wasting energy unless carefully tuned.

Because the output of the chargeverter is connected to the input of the 6500EX

So are you indicating that GTI don't regulate or operate within any voltage limits, they just keep turning up MPPT (even if nothing is drawing down AC volts) until the moon is reached? So it wouldn't act like a gas generator, and just turn down the field coil so to speak, when nothing is plugged into it just to maintain a certain maximum voltage? (based on some kind of AVR) Don't they have CT's or something that they use for feedback?

For this discussion, GTI is the inverter that you're trying to AC->DC couple into the system via chargeverter.

Grid looks like an infinite load to GTIs (in the current operating regime). CTs are optional in most GTI implementations in North America. They are required to enforce export limits in other markets.

I'll admit that I know very little about GTI's but that doesn't seem like it would be all that well thought out if it just keeps trying to pump out higher voltage until infinity.

It's well thought out enough for the US to have reached gigawatts of installed GTI on the grid. It does not pump until infinity voltage because at current deployment levels on the grid, the grid can absorb all of it. Or GTI hits high voltage shutoff. However, since this is not guaranteed (imagine a grid that's saturated with 80% power from GTI), GTIs are required (by 1741SB) to respond to signals from the grid indicating that it is reaching the limit, and curtail appropriately.

In fact in the utility scale lingo, grid forming means something quite different from what grid forming means to us peons. It refers to next generation inverters/renewable power stations, etc. that can keep a stable grid in the above conditions, where today we need a minimum % of classic spinning generators to keep it stable.

I would have thought that a GTI would just turn itself down (relax MPPT), once some peak AC voltage is reached?

They shut themselves off because this indicates a problem with the infinite load (grid) assumption they are designed for. Note that the 1741 spec says nothing about protecting randomly connected devices from GTI's attempt at interacting with the grid. 1741 only guarantees protecting the grid and not starting a fire with the inverters. If the inverters kill themselves or kill attached equipment when used in an off-label way, not a design flaw with the NRTL testing or spec. There is also no guarantee in 1741SB on whether your microgrid is guaranteed to work with every GTI out there. There are lots of reports of IQ8s triggering some kind of anti-islanding detection and not working with non-Enphase microgrid.

 

[zanydroid]

The chargeverter also does not know that the GTI is power limited, it has no such control loop to detect if the inverters are sagging due to trying to pull too much power.

I'm not especially getting this part here, can you rephrase maybe? Which inverters? Inputs or outputs, I'm a bit lost. I'm really trying to understand.

OK, let's for shorthand say we have microinverters going into chargeverter. 3kW-AC nameplate. Chargeverter is set to 2kW with a knob so it's not adjustable.

We know from preceding discussion that if the microinverters are actually generating 3kW, then the microinverters will throw a fit because the load cannot accept 3kW, where the contract with the grid that it's supposed to be connected to is to accept unlimited export power. In a proper AC coupled system either the micros will be cut off / ramped down, or the charger will be ramped up.

What if the microinverters are generating 1kW. Well your Chargeverter is still hard-set to 2kW. The microinverters will not be able to satisfy this load and will shut off on some error or other. This is not an issue when microinverter is grid-tied because the grid will make up the 1kW deficit.

 

[zanydroid]

So in the previous messages, I said that GTIs are not able to respond to loads. This is not strictly true.

Enphase designed the IQ8 generation to be able to do this in combination with some extra hardware, and those can supply small loads as battery-less off-grid inverters.

I do not believe DIY community has figured out how to force IQ8s in isolation into this mode for DIY projects, without buying the whole Enphase solution and using it as a system.

 

[sunshine_eggo]

So in the previous messages, I said that GTIs are not able to respond to loads. This is not strictly true.

Enphase designed the IQ8 generation to be able to do this in combination with some extra hardware, and those can supply small loads as battery-less off-grid inverters.

I do not believe DIY community has figured out how to force IQ8s in isolation into this mode for DIY projects, without buying the whole Enphase solution and using it as a system.

Since it's confession time, I'll do the same...

A Victron eco-system can be configured to work with Fronius GTI and actually regulate them as opposed to frequency shifting for on/off.