Microgrid with generator?

[LATTC Student]

I have a newly installed grid-tied solar PV system with microinverters. We have a gas powered (Honda EU7000is) inverter generator that also back-feeds the main panel to the house on a manual transfer switch.

During a utility outage - if the generator is wired upstream from the microinverters and powered up they should 'see' a new voltage/frequency source and turn on correct ? If yes, and microinverters are set to "grid no-export", what happens when the PV panels start making more power? Same microinverter limiting function as with utility power? Would this damage the generator?

I assume the generator hardly needs to be running at that point if the PV is producing, just enough power to provide VAC/freq reference to the microinverters?

 

[jaggedben]

No, you can't have a generator on a micro-grid with solar inverters that is just handled by a transfer switch between grid and generator. And yes, damage to generator or PV inverters is a possibility. So is just shut downs that leave you without power. There have been a number of threads on this here already. The same goes for batteries and generators, generally.

Possible work-arounds are:
- Connecting the PV on the line side of the the transfer switch
- having a manual interlock that prevents you from connecting the generator without disconnecting the solar
- installing an integrated, listed solution (such as Enphase Enpower) that is capable of starting, stopping, connecting, and disconnecting the generator appropriately

 

[LATTC Student]

No, you can't have a generator on a micro-grid with solar inverters that is just handled by a transfer switch between grid and generator. And yes, damage to generator or PV inverters is a possibility. So is just shut downs that leave you without power. There have been a number of threads on this here already. The same goes for batteries and generators, generally.

Possible work-arounds are:
- Connecting the PV on the line side of the the transfer switch
- having a manual interlock that prevents you from connecting the generator without disconnecting the solar
- installing an integrated, listed solution (such as Enphase Enpower) that is capable of starting, stopping, connecting, and disconnecting the generator appropriately

Sorry I didn't search through the forum enough to find the appropriate posts. I do thank you JaggenBen, once again, for replying timely and informatively! I feel the go to solution would be the Smart Switch but that requires a MPU which is out of budget.

For now I will probably just remove the MTS and interlock from my main panel to keep it compliant to ditch the back-fed generator deployment altogether (I have had the MTS installed for a few years and never had to use it as SCE is really stable here at our location).

 

[analog8484]

You can also use a voltage monitoring relay/contactor with manual reset on the PV output to automatically disconnect on grid outage.

 

[winnie]

My understanding of the failure mode is that when the PV array is producing more power than used locally, it will try to 'export ' power to the generator...which cannot absorb it. The system voltage climbs until something shuts down or fails.

But if the inverter is capable of guaranteeing 0 export, doesn't this solve the problem?

AFAIK there is no physics reason that PV and generators cannot be used together, but rather a failure in the design of standard PV systems to function correctly with standard generators.

Jon

 

[Hv&Lv]

I know Tesla makes a switch to switch a generator on automatically when the battery is depleted, but it won’t work with generator, battery and solar??

 

[electrofelon]

My understanding of the failure mode is that when the PV array is producing more power than used locally, it will try to 'export ' power to the generator...which cannot absorb it. The system voltage climbs until something shuts down or fails.

But if the inverter is capable of guaranteeing 0 export, doesn't this solve the problem?

AFAIK there is no physics reason that PV and generators cannot be used together, but rather a failure in the design of standard PV systems to function correctly with standard generators.

Jon

My thoughts too. It really wouldn't be that hard. Seems like it would just be a few lines of code, of course coupled with a transfer switch.

Actually there is a potential problem that I could see happening: it is my understanding that PV inverters are somewhat limited on how far they can drive the VI curve in either direction to lower the power harvest. So if you had very small power needs and full sun, I'm not sure the inverter could handle that. I imagine a workaround may be that the inverter couldit could just pwm the panels to shed as much power as needed, but after a point that would probably get into making "dirty" AC.

Does anyone know what "grid no export" means from the OP? Sounds like it has the capability do not send power back to the grid, which would of course require CTs around the service conductors. If it does have that capability, it's a cool feature.

 

[Hv&Lv]

They do have that feature.
We require them to be in the No Export mode before they are anti island tested.

edit.
Let me rephrase that.
The ones that do have that feature are required to be in no export mode before we test if they are turned on.
the ones without that feature are required to be off until we test.

 

[LATTC Student]

My thoughts too. It really wouldn't be that hard. Seems like it would just be a few lines of code, of course coupled with a transfer switch.

Actually there is a potential problem that I could see happening: it is my understanding that PV inverters are somewhat limited on how far they can drive the VI curve in either direction to lower the power harvest. So if you had very small power needs and full sun, I'm not sure the inverter could handle that. I imagine a workaround may be that the inverter couldit could just pwm the panels to shed as much power as needed, but after a point that would probably get into making "dirty" AC.

Does anyone know what "grid no export" means from the OP? Sounds like it has the capability do not send power back to the grid, which would of course require CTs around the service conductors. If it does have that capability, it's a cool feature.

Yes, in CA we have grid profiles that have to be "CA Rule 21" compliant. We use the CA rule 21 "Grid No-Export" profile at our house. The microinverters rate-limit themselves to "match" consumption. Actually it is slightly just under the consumed amperage, by anywhere from 20W to 60W so we import a little. Our consumption CT's are in between the meter and the main panel as you mentioned. Fortunately the system we have supports feeds of over 500' 18GA CL3R from the controller (Envoy) to the CT's. Pretty sweet.

SocalEdison (SCE) is our provider and they pay very little for customer generated power. I would rather export nothing than send it to them when they make 70% margin's off of my PV generation. Our main purpose for PV deployment here is Energy Independence.

 

[electrofelon]

Yes, in CA we have grid profiles that have to be "CA Rule 21" compliant. We use the CA rule 21 "Grid No-Export" profile at our house. The microinverters rate-limit themselves to "match" consumption. Actually it is slightly just under the consumed amperage, by anywhere from 20W to 60W so we import a little. Our consumption CT's are in between the meter and the main panel as you mentioned. Fortunately the system we have supports feeds of over 500' 18GA CL3R from the controller (Envoy) to the CT's. Pretty sweet.

SocalEdison (SCE) is our provider and they pay very little for customer generated power. I would rather export nothing than send it to them when they make 70% margin's off of my PV generation. Our main purpose for PV deployment here is Energy Independence.

Neat. I am in a specific niche here in upstate New York where we only use string inverters, and we still have basic net metering so I haven't had to deal with any of this.

So as far as you know the system can throttle down as far as it needs to without issue? I would be curious how exactly that is done, I assume they must be shutting off entire micro inverters as needed.

 

[GoldDigger]

My thoughts too. It really wouldn't be that hard. Seems like it would just be a few lines of code, of course coupled with a transfer switch.

Actually there is a potential problem that I could see happening: it is my understanding that PV inverters are somewhat limited on how far they can drive the VI curve in either direction to lower the power harvest. So if you had very small power needs and full sun, I'm not sure the inverter could handle that. I imagine a workaround may be that the inverter couldit could just pwm the panels to shed as much power as needed, but after a point that would probably get into making "dirty" AC.

Does anyone know what "grid no export" means from the OP? Sounds like it has the capability do not send power back to the grid, which would of course require CTs around the service conductors. If it does have that capability, it's a cool feature.

The capability to throttle output can be added to a grid-interactive inverter, and driving the panel array away from its Mpp is simply a matter of limiting the current. There is already, in effect, a PWM capability present in the voltage-transforming Mppt front end. So it may be mostly a matter of software. But at a minimum there would, as mentioned, have to be CTs on the utility/generator feed. And it would inevitably cost more.
A limited number of traditional hybrid (grid-interactive and standalone) inverter systems are available with generator support. An important part of this operating mode is a relaxed specification for input voltage and frequency and the disabling of any anti-islanding tests that rely on the low source impedance of the grid.
There are also command-throttleable as well as "no export" grid interactive inverters to meet POCO requirements for grid stability, initially in Hawaii. Some of them can also be commanded to deliver current at a non-unity displacement power factor. I think that the latter capability first appeared in commercial scale inverters.

 

[GoldDigger]

...
...
SocalEdison (SCE) is our provider and they pay very little for customer generated power. I would rather export nothing than send it to them when they make 70% margin's off of my PV generation. Our main purpose for PV deployment here is Energy Independence.

As long as you are not actually losing money by it, net export is better for our enjoyment of our planet. (The planet itself will be fine regardless of what we do to destroy the environment.)

 

[electrofelon]

As long as you are not actually losing money by it, net export is better for our enjoyment of our planet. (The planet itself will be fine regardless of what we do to destroy the environment.)

I had 4600 KWH surplus at the end of this year's annual zero out period which just ended a few months ago. (we have net metering within that year, but at the year any surplus is paid at the avoided cost + reset to zero). So yeah $90 isn't that much, but I like to see it get used and the POCO make some money on it, and I already leave pretty much lots of lights on and I'm not stingy with the thermostat.

 

[electrofelon]

No, you can't have a generator on a micro-grid with solar inverters that is just handled by a transfer switch between grid and generator. And yes, damage to generator or PV inverters is a possibility. So is just shut downs that leave you without power. There have been a number of threads on this here already. The same goes for batteries and generators, generally.

Possible work-arounds are:
- Connecting the PV on the line side of the the transfer switch
- having a manual interlock that prevents you from connecting the generator without disconnecting the solar
- installing an integrated, listed solution (such as Enphase Enpower) that is capable of starting, stopping, connecting, and disconnecting the generator appropriately

So considering the system the OP is talking about, seems like this would work fine. Why wouldn't it? This is of course assuming the arrangement of the generator connection and those CTs is appropriate.

 

[jaggedben]

...
But if the inverter is capable of guaranteeing 0 export, doesn't this solve the problem?
...

Possibly, but problematic in many cases, and getting into the weeds more than the OP initially warranted.

In my experience all zero export systems rely on CTs at the point of export. Placement of these CTs needs to be downstream of the generator and not just downstream of the grid. Also in my experience the zero export function is dependent on a software layer to limit export on a kWh scale and may not actually stop short term (i.e seconds long) exports that would be a problem for the generator.

 

[jaggedben]

So considering the system the OP is talking about, seems like this would work fine. Why wouldn't it? This is of course assuming the arrangement of the generator connection and those CTs is appropriate.

The OPs equipment and setup (Enphase, he didn't say so but easily deduced) is not listed to work with a generator without batteries and an MID. See last post. I'm far from convinced the no export function is fast enough to provide stability off -grid.

 

[Hv&Lv]

The OPs equipment and setup (Enphase, he didn't say so but easily deduced) is not listed to work with a generator without batteries and an MID. See last post. I'm far from convinced the no export function is fast enough to provide stability off -grid.

Generac is the only one I know how to modify settings and turn off the non export function.

Is the enphase as easy as the generac to do? I didn’t see a screen to change settings on an enphase.
Or is it all changed through software?

 

[jaggedben]

Generac is the only one I know how to modify settings and turn off the non export function.

Is the enphase as easy as the generac to do? I didn’t see a screen to change settings on an enphase.
Or is it all changed through software?

Enphase is microinverters and the grid profiles need to be updated one by one for each microinverter by the Envoy. It's definitely slower to change than systems with one inverter. Also as an AC system it presents very different issues than the DC coupled Generac.

 

[synchro]

In my experience all zero export systems rely on CTs at the point of export. Placement of these CTs needs to be downstream of the generator and not just downstream of the grid. Also in my experience the zero export function is dependent on a software layer to limit export on a kWh scale and may not actually stop short term (i.e seconds long) exports that would be a problem for the generator.

An inverter generator will have no way to "sink" current coming from the PV inverter other than to charge up its DC bus capacitors through the bypass diodes that are across its IGBT or FET output devices. This could easily cause a bus overvoltage in a manner similar to a VFD that's ramping the frequency down too quickly so that it's being driven by current from the motor. The inverter generator bus overvoltage would only be limited by the PV inverter shutting down under this condition. Even if damage is avoided, I can envision some sort of "limit cycle" occurring involving the PV inverter shutting down and turning back on, the inverter generator engine cycling between idle and higher speed, etc.
Perhaps a traditional synchronous alternator in a home generator could absorb a significant reverse current for several seconds, but I don't know.

Any home generator is going to have a significantly higher source impedance than the POCO service. And so the noise and distortion voltage created by the PWM current of the PV inverter will also likely to be noticebly higher. This might cause problems with AFCI's, GFCI's, and/or home electronics.

With some exceptions (as GoldDigger mentioned), the output current generated by PV inverters will typically be in-phase with the AC voltage. By supplying in-phase current that would otherwise be supplied by the generator, a PV inverter will decrease the power factor seen by the generator. The power factor of residential loads typically doesn't depart too far from unity, and so this might not be a problem in a typical case.

 

[LATTC Student]

The OPs equipment and setup (Enphase, he didn't say so but easily deduced) is not listed to work with a generator without batteries and an MID. See last post. I'm far from convinced the no export function is fast enough to provide stability off -grid.

Yes my setup is Enphase. They recently started offering generator integration/support with the (~$1300) Enpower Smart Switch including grid-sensing generators that are triggered "on" by the Switch when power goes out. It's a pretty slick MID with it's own Neutral Forming Transformer with a ton of advanced features (like 4 circuit PCS, load shedding, PV shedding, "Power Start" for dynamic surge management, and the generator integration mentioned above). Unfortunately I would need an MPU to get the switch installed - out of budget

I was looking for a workaround without this Switch but looks like there isn't one.