Sola transformer to turn on grid tied inverter

[jes_sacramento]

Has anyone tried using a constant voltage Sola type transformer with a small true sine wave battery operated inverter to turn on grid-tied inverters when the grid is down?
Of course, the problem with trying to fool the inverters to turn on is the rising voltage back feed into whatever you are using to synthesize a grid and get the inverters to fire up. A constant voltage transformer tied to your 60hz sine wave source should protect it from the rising voltage from the inverters and keep it from pushing current back into your small reference inverter, since is will only see 110V, no matter what the inverters are putting out [it would also be going through a 110v to 240V transformer to tie into the main panel and inverters, so that the 110V reference inverter will provide the 60 hz 240v "grid reference" to the solar inverters].
Obviously, this type of arrangement would necessitate enough of a load on the solar inverters to dump whatever load they can produce, but the goal here is to determine if a cheap and small battery backup could be used via a very small reference inverter to turn the system on without destroying anything.

 

[sdwilliams_solar-nube]

NEWBIE ALERT!!! Trying to avoid TX Snowmegedon2

I was also wondering if you could use a small inverter generator to "fool" the grid-tied inverter to turn on solar. If this is possible, would the solar take the load off the generator. Of course, the MAIN would be locked out during this setup. This would be a daytime use only.

What are the dangers and can they be mitigated?

 

[Keith C]

You would not want to use a constant voltage transformer. That would be akin to one of my favorite Simpson's where at the bowling alley all knocked down pins are swept into a dumpster outside. Meanwhile logging trucks are rolling up delivering whole trees that the automated system picks up and sticks into a lathe. Each tree is turned down to a single bowling pin painted and stood up in one of the alleys where the original pin(s) have been swept to the dumpster.

A constant voltage transformer is horrifically inefficient. Often idling they burn you instantly if you touch them. Used with a solar system they are the electric analog of the Simpsons Bowling Alley.

Grid-tie inverters look for line voltage and 60Hz (50Hz). If you make a system that can put out 60Hz at a voltage just about exactly what the inverter would 'like' to put out but a little more it should work. As an example setup a 12V or 24V or whatever your battery bank supplies, oscillator that runs at the correct frequency. back-feed that to a small control transformer that puts out the inverter voltage. You must galvanically isolate the battery bank from this oscillator and the inverter! The transformer does this.

The problem I'd expect is that a grid-tie is built to drive current into the grid. To do that the inverter tries to create a sinewave that is slightly leading the grid. This is how it exports into the grid. It tries to do this to the limit of the available solar. But there is no grid...

 

[macheung]

Are there grid tie inverters with a no export mode where it won’t try to push back into the grid? In that case, would a small sine wave inverter allow it to turn in safety and power local loads when the grid is off and disconnected?

 

[Keith C]

More specifically there are Off-Grid inverters that have grid-tie modes you can select. They will, for instance, charge your battery bank until it's full then transition to grid-tie to use the solar productively. They will also draw from the grid to charge your battery bank if there's not enough solar available.

However, these days it seems that if you just go looking for a "grid-tie inverter" mostly that's what you get grid-tie.

Some "All-in-ones" will also do what I describe above. I have a couple of MPP style AinOs and they do the usual MPP thing PLUS grid-tie. They're the "All-in-one Hybrids". Hybrid being the added function of grid-tie.

 

[Jordi]

Has anyone tried using a constant voltage Sola type transformer with a small true sine wave battery operated inverter to turn on grid-tied inverters when the grid is down?
Of course, the problem with trying to fool the inverters to turn on is the rising voltage back feed into whatever you are using to synthesize a grid and get the inverters to fire up. A constant voltage transformer tied to your 60hz sine wave source should protect it from the rising voltage from the inverters and keep it from pushing current back into your small reference inverter, since is will only see 110V, no matter what the inverters are putting out [it would also be going through a 110v to 240V transformer to tie into the main panel and inverters, so that the 110V reference inverter will provide the 60 hz 240v "grid reference" to the solar inverters].
Obviously, this type of arrangement would necessitate enough of a load on the solar inverters to dump whatever load they can produce, but the goal here is to determine if a cheap and small battery backup could be used via a very small reference inverter to turn the system on without destroying anything.

I try to provide you here an answer with some facts too.

The grid-tied inverter can be fooled by an off-grid inverter because it creates a "local grid" at 220V.

The grid-tied inverter offers an AC at 240V to overcome the grid resistance at 220V; so is current flow happening from inverter to grid.

The off-grid inverter also has the capacity to offer a higher AC voltage than 220V, specially when the load is low (compared to the battery capacity and state of charge) and/or simpler said; the battery voltage provided to the inverter is over 12V. Real data I can see with my inverter: It produces AC at 230V when battery offers voltage around 14,5V and AC 215V when battery voltage is around 10V. As you can see; in most of the cases the off-grid inverter can not match the AC voltage of the grid-tied inverter.

Conclusion, everything under 240V is seen by the grid-tied inverter as a load. Here appliances (loads) are seen as 220V loads. If you manage to sustain a consumption over the maximum output of the grid-tied inverter and the off-grid inverter always has to supply something; then you can run this setup temporary in near to safe conditions. BUT! may one of your loads be intermittent (eg. fridge compressor) and may the total consumption be for a while inferior to the grid-tied inverter, then the grid-tied inverter will supply to the off-grid inverter and fry it in an unknow but rather short amount of time.

One user of this forum did that and explained his experience. The first hours he managed well, but then his consumption went down and shortly after there was smoke around the off-grid inverter which became useless.

 

[macheung]

I am considering a different approach to get power off a grid tied PV system when the grid is out. Instead of fooling the grid tie inverter with a local grid, I am trying to hook up another off-grid inverter to the some of the panels in parallel. For this to work, the parallel connection point would need to be between the optimizers and the panel, as the optimizers would shut off when the grid is out due to rapid disconnect requirements.

This would work, but there are two issues that needs to be investigated / mitigated if needed:

What happens when 2 MPPT is connected to the same panel?
The MPPT of the off-grid inverter would be connected to the panel in parallel with the optimizers. Normally this doesn't happen as the off-grid inverter can be plugged in only when the grid is out, but when the grid comes back on, it is likely that both would be connected until the operator disconnects the off-grid inverter. So will there be damages when both MPPT are connected? Or just won't work well during that time? If this is a problem it can be mitigated by also disconnecting the main grid tie inverter when the grid is out, and only manually connecting it back once the off-grid inverter is connected. Not friendly, but possible.

Can panels be serially connected even when connecting to their own optimizers?
It would be cost ineffective to run a parallel sets of line down from each panel to the off-grid inverter. Ideally, a short string of 3 panels can be wired together before their optimizer so only 2 strings need to run down to the off-grid inverter taking power from 6 panels. However, I am unsure if anything bad would happen in normal operation when the main optimizers are on. This can be mitigated through relays on the roof that opens the serial connection when the grid is powered on, but that is complex and a lot more work.

Yes, I am aware that running additional lines down would negate the rapid shutdown capability, as those line will still be energized. However, that can be mitigated later by putting a relay to cut off the power from the panel where the secondary strings enters the roof. That is left out of the picture for now.

What do you guys think?