Is it possible to "island" Enphase inverters with generator

[Naptown]

No the impedence is too high on the generator and either the inverters or the generator will produce expensive smoke and flame.
In addition a generator will go out of phase every time the load changes and the engine has to catch up or slow down causing the Enphase to trip off line due to out of parameter attempt to sync.

[bonaire]

I'm curious, If the enphase inverters can go in sync. with the utility , Why cant they do the same with a good whole house generater? Wouldnt the generater behave as if no load?

Let's say the solar system + generator are at peak output and that may be say 7000 watts and the grid is gone and you only have your refrigerator running drawing 1200W. Where will the extra power produced go? Electric production needs a consumer at the other end.

That's why hybrid systems and off-grid systems have a load source so that once batteres are charged, there is a load put on the system to draw out the power - such as a heater or other heavy-load component to absorb the extra power.

You can always do some type of hybrid system but there is not much available other than the forthcoming SMA TL models (3,4,5 KW) which provide for an up-to 120V 12A power plug during the mid-day when the grid is down. But that's no good for running appliances at night.

[johansen]

i can't speak for myself but i have read of folks on another forum doing this with Xantrex and other inverters.

I can't see why the Enphase inverters would not work with a generator, provided you put a big capacitor (like say a few amps of reactive current) across the generator to make it appear a bit lower impedance than it is.

The enphase inverters are typically 150, 200 watts at 240 volts right? so they are pushing an amp into the grid. that's nothing, not into a generator either.
The question is do they turn off when handling the fluctuating Frequency of the generator under variable load, do they see that as a grid "failure" or just ride through it.

I would like to know because i'm contemplating buying one of these enphase inverters to play with it a bit, see if i can't turn the 200 watt rating into 2000 by changing out the dc-dc converter side, by "upgrading it" for the direct purpose of islanding an ac power system disconnected from the grid. (i'm not intending to void my fire insurance.)

as far as i know, having spoken with someone who is familiar with the UL1741 standard, grid tie inverters check to see if the grid is present on a Cycle by Cycle basis.
meaning, at the zero crossing they check to see if the grid is there or not.
this will be a problem if you managed to get say, 10KW of grid tie inverters running on a 5Kw generator-- there's no reason this couldn't be done provided the generator is acting as an ideal synchronous motor... they aren't. they are a bit loose producing harmonics themselves...

someone upstream said it would not work because the inverter could push your local grid to, too high a voltage.
beyond 130 volts those expensive inverters are either going to shut down or they will back off on the power.
pushing 130 volts into a generator is going to cause the AVR to reduce the field strength which is going to cause the generator to consume reactive power.
hmm... i can't recall which direction that is going to go.. consuming capacitive reactance or inductive? --this could be the biggest problem

[johansen]

No the impedence is too high on the generator and either the inverters or the generator will produce expensive smoke and flame.
In addition a generator will go out of phase every time the load changes and the engine has to catch up or slow down causing the Enphase to trip off line due to out of parameter attempt to sync.

is this your experience or opinion?
if it actually happened, then i refuse to buy one of those inverters, and thank you for the information.

(there's no reason an inverter should not be able to turn off within 5 microseconds of the output being reverse connected to 240 vac of the wrong polarity!)

[Naptown]

is this your experience or opinion?
if it actually happened, then i refuse to buy one of those inverters, and thank you for the information.

(there's no reason an inverter should not be able to turn off within 5 microseconds of the output being reverse connected to 240 vac of the wrong polarity!)

This is true of all grid tie inverters not just the enphase. In best case the inverter will sense frequency out of tolerance and will shut down probably never to restart with the constant frequency shifts of small generators with changing loads. They will however sync to another inverter in what is known as an AC coupled system. However Enphase voids the warranty on their inverters in this type of system.

[russ]

i can't speak for myself but i have read of folks on another forum doing this with Xantrex and other inverters.

Be very careful about what you read on the net or see on youtube - the great majority is just people chattering to make themselves feel important.

[Naptown]

The xantrex and Magnum inverters are a bit different as they are designed to Island and have separate generator inputs for battery charging and carrying the load.

[fuzzelogic]

Most automatic transfer switches will wait a minute or so when the grid comes back up to make sure it is stable before disconnecting and returning the power to the grid. The generator will run for about a minute of so to cool down before shutting down.

SO, if this was a system at your house, would you be willing to have the enphase inverter tied into the main breaker panel while the power is transferred back from the grid.

You are correct, it's a automatic transfer switch that is in the main breaker panel. Above the transfer switch will not be powered, but the breakers below will get power when the power is out.

I forgot about the startup period regarding the inverter.

Scenario
Backup generator is running and power is out.
Power from the grid comes back online and the transfer switch starts to transfer power from the generator and get it from the grid. 30-45 seconds transfer period
In between this phase, there is a limbo of power transferring from generator and the grid.
At this point, the enphase inverter will start the boot up process. Is it possible that inverter "could" get fried during this transfer period of 30-45 seconds?

Note: Would prefer to use Enpase inverters and not use a central inverter such as SunnyBoy with Sunny Isalnd due to the system getting complicated and expensive VERY quickly

[Naptown]

SO, if this was a system at your house, would you be willing to have the enphase inverter tied into the main breaker panel while the power is transferred back from the grid.

You are correct, it's a automatic transfer switch that is in the main breaker panel. Above the transfer switch will not be powered, but the breakers below will get power when the power is out.

I forgot about the startup period regarding the inverter.

Scenario
Backup generator is running and power is out.
Power from the grid comes back online and the transfer switch starts to transfer power from the generator and get it from the grid. 30-45 seconds transfer period
In between this phase, there is a limbo of power transferring from generator and the grid.
At this point, the enphase inverter will start the boot up process. Is it possible that inverter "could" get fried during this transfer period of 30-45 seconds?

Note: Would prefer to use Enpase inverters and not use a central inverter such as SunnyBoy with Sunny Isalnd due to the system getting complicated and expensive VERY quickly

There is never a limbo period it is either grid or generator period. Perhaps a momentary flicker when the transfer back to the grid occurs.
Again how this works
Power down about a minute later generator starts. Runs till grid comes back up.
power comes back up generator runs for a set period of time (still disconnected from grid) to establish that the grid is back up and stable.
Once a stable grid is established the transfer switch will switch the generator circuits back to grid power. Generator continues to run for a preset time to cool down and shuts off.
At no time is the generator connected to the grid when it is up.
Be sure you tie the inverters(s) into the side that is in your case above the transfer switch and will connect to the grid only.
As far as inverter choice goes any will work it doesn't matter as long as the inverter(s) are connected to the grid side of the transfer switch.
If you are not trying to create a hybrid system which you did not mention in this post a single inverter may be less expensive depending on the size (up to about 10 modules the Enphase are comparable in cost above that level they get more expensive)
Add batteries to the mix and the complexity and cost grow exponentially.

[fuzzelogic]

Naptown, thanks for repeating what I typed so we are both on the same page.

I do want to avoid the hybrid system and you are correct that the generator would disconnect, BUT still run for a set cooldown period as you mentioned in the post above.

The system would have some panels on a west facing roof and some on a south facing roof, with neither set probably having enough power to turn on a single inverter which is why I was considering enphase. I had not mentioned this before and probably would have helped if I had clarified that in the 1st post.

[inetdog]

... with neither set probably having enough power to turn on a single inverter which is why I was considering enphase.

Most GTI inverters have a minimum voltage for startup, which is a matter of proper string sizing for Voc and Vmp, but not a particularly high minimum power requirement. They might not operate very efficiently at low power, but they will operate. My 3kW SunnyBoy will chug along merrily showing a production of 100 watts or less. Can you clarify what you are referring to?

Remember that both Voc and Vmp are pretty much independent of the amount of light falling on the panels. What changes is Imp.

[Naptown]

A single inverter will work fine as long as the different facing strings are coplanar. meaning if there are two strings one entire string is facing one direction and the other entire string is facing the other. You cannot have a one string facing in 2 directions.
If say you have 20 panels and you can get 9 facing south and 11 west then a Power1 inverter with the dual mppt inputs will work with each direection on an input. As long as the voltage is within specs for each string it will be fine.

[inetdog]

A single inverter will work fine as long as the different facing strings are coplanar. meaning if there are two strings one entire string is facing one direction and the other entire string is facing the other. You cannot have a one string facing in 2 directions.
If say you have 20 panels and you can get 9 facing south and 11 west then a Power1 inverter with the dual mppt inputs will work with each direection on an input. As long as the voltage is within specs for each string it will be fine.

Actually, as discussed in another thread on this and other forums at great length and substantiated by research (from either SMA or Schneider/Xantrex?) you can put non-coplanar arrays on one MPPT input as long as the following three conditions are met:

1. The panel types are identical (this one can be relaxed a bit with careful attention)
2. The number of panels in each string is identical (actually that the full sun Voc and Vmp of the two arrays be equal, see above)
3. And this is the most important after the obvious stuff above: Neither array has partial shading of panels at a time when both are producing useful power.

The reason for the last condition is that if one or more panels (or even panel segments) are being cut out of the series string because their bypass diodes are being activated, it will have the effect of changing Vmp for the shaded string, causing a violation of condition 2. The result will be that the MPPT input cannot match the differing Vmp of the two arrays well enough to get good power from both of them. The exact results will depend on the details of the MPPT algorithm used, and this will vary from brand to brand and even model to model.

If both arrays consist of unshaded panels, with the same overall Voc, then the Vmp of the two arrays will be so close to identical that it does not matter, even when the Imp of the two arrays is vastly different (for example East and West facing arrays in morning and evening).