2
\$\begingroup\$

I am getting close to investing in a micro-inverter (Enphase M250) based grid-tie solar array for my home, but am concerned I won't be able to use the solar power during grid outages. I found a similar post here that hints towards spoofing the solar array to power on.

First, is this actually possible? I understand I would need to disconnect from the grid before enabling the solar array and would turn all but the necessary circuits off at the panel.

In my mind, I'm thinking a pure sine wave split phase inverter, like this, hooked to my truck battery and fed in through my 220v, 50A welding outlet would enable the micro-inverters to start producing power again. I would then charge the battery with a float charger. Is it that simple?

Second, if that would work in theory, how much load would be on the inverter? Let's say I am producing 10kW from the solar system and have a load of 7kW. I'd like to think the 4kW inverter wouldn't have any load on it, but I have no idea. I'm concerned it would try to pull from both (or possibly only the inverter?) and fry the split phase inverter or destroy my truck battery.

Third, if I am producing 10kW from the solar array and only using 7kw, is that going to be an issue? Normally the grid would consume the remaining power.

Lastly, I do realize there are plenty of options, such as a battery backup system, or a hybrid inverter, or a string-array with a SunnyBoy inverter, which has a usable outlet during grid outtage. I'd like to enable the solar array with very little extra cost then what I'm already prepared for and without modifying the system, since it's all under warranty unless I modify it (as in tapping into the DC lines from the panels). Also, a very similar question was asked here, but I'm hoping to answer the technical aspects of 'is it possible', not as much of the 'is it advisable' question mostly addressed in other post.

EDIT: I also found a document from enphase about AC coupling their micro-inverters here, but it doesn't spell out specifics. I'm convinced it's possible...

EDIT: I've contacted my installer, who agrees that it should be possible, and is contacting the manufacturer for more details. I still have my fingers crossed.

FINAL EDIT: For anyone else looking for a similar setup, I received a response from my installer. The Enphase microinverters absolutely will work with any "clean" power source. As the accepted answer points out, the microinverters are already designed for this very purpose. They don't care what the source is, they'll match it as long as it's a stable power source. My installer did say if I don't use a clean power source and fry the microinverters, they won't be covered under warranty, so be careful of the selected power source.

\$\endgroup\$
  • \$\begingroup\$ Just a note on units: your inverter output power is measured in kW. The amount of energy given by the system is power x time = kW x h = kWh. You might want to correct this in your question. \$\endgroup\$ – Transistor May 6 '16 at 18:34
  • \$\begingroup\$ This is, in practice, pretty non-feasible with an inverter intended for grid-tie (unless it is one that is hybrid and grasps the concept of not being grid-tied.) \$\endgroup\$ – Ecnerwal May 6 '16 at 19:26
  • \$\begingroup\$ @Ecnerwal, any chance you could help me understand why? According to Enphase, the "grid" sine wave has to be within +/- .02hz from 60hz. Granted, the inverter I linked to claims to be within +/- .03hz, but for the sake of gaining knowledge, let's say I found an inverter that was within +/- .02hz, is there something else that would cause it not to function? \$\endgroup\$ – Cam May 6 '16 at 19:34
2
\$\begingroup\$

I don't have experience with the specific models you are referencing but I do have experience with phase locking AC power sources. The bicycle analogy above is exactly how mechanical generators work and with solid state power inverters to the point of the inverter "pushing" on the grid however you don't want the grid to ever "push" on your inverter. The inverter watches the 60hz sine wave from the grid connection to trigger the driver to keep the output in phase with grid power. This is going to be a high impedance detection circuit so any source should do the job such as your pure wave inverter off a car battery. The listed tolerance on the sine wave is well outside the normal tolerance for mains. How precise is the frequency of the AC electricity network?

While connected to the grid it is critical the inverter is synced to the grid frequency and phase as tight as possible to prevent the inverter from seeing the grid as a load. If the grid were to have just gone down and your system started automatically the phase could be 90deg out of sync when the grid power is restored which would be worse than shorting the output of your inverter. This is why they inverter relies on an external clock source and shuts down if one is not present. Disconnected from the grid you have a self contained power system and the phase and frequency relationship to the grid is meaningless but the frequency does need to be close to 60hz for any devices you maybe running on that power.

\$\endgroup\$
2
\$\begingroup\$

The main issue is that most grid tie inverters are not designed for standalone operation, they are designed to always pull the maximum amount of power possible from the panel. That way it always harvests the maximum possible amount of energy each day giving you the maximum possible savings (assuming you get paid for feeding power back to the grid). Standalone operation is different as the inverter will need to regulate its output so it pumps out only as much energy as you need, otherwise it'd overload something. There are grid tie inverters that work in standalone operation as well but they cost more (smaller market share, economies of scale etc.) Unkess it explicitly states both standalone and grid tie operation, assume it only does one or the other (there are plenty of standalone only inverters too so be careful)

\$\endgroup\$
1
\$\begingroup\$

This answer only addresses the grid-tie aspect of your question.

I suspect the problem is that your inverter is relying on the "strength" of the grid to keep it in synch.

When a generator is connected up to the grid it's a bit like an additional rider getting onto a 10-man "tandem" bicycle. If he tries to accelerate the bike he will end up taking more and more load but making very little difference to the overall speed because he's only 10% of the "system".

enter image description here

Photo 1. A bicycle built for 10 in 1896. Source: Leicestershire and Rutland CTC.

If the rider tries to slow down the bike (on a tandem there's no freewheel unless all freewheel) the others will drive the slow rider, much like the grid will "motor" a slow generator.

... spoofing the solar array to power on.

My understanding of the problem is that your inverter is somehow relying on being pulled into synch by a very strong grid. Whatever you use to "spoof" the grid would need to be a strong, stable AC power source itself.

The only option I can think of to overcome this would be to somehow convert the unit to run on its own reference frequency when the mains is isolated. I suspect you don't wish to attempt that for warranty reasons.

\$\endgroup\$
  • \$\begingroup\$ So it sounds like my theory of a truck battery powered inverter would most likely be unable to turn the micro-inverters on? I'm more willing to give it a try if it's not just going to blow things up. \$\endgroup\$ – Cam May 6 '16 at 21:49
  • \$\begingroup\$ I'm afraid I haven't enough experience with them to comment further. \$\endgroup\$ – Transistor May 6 '16 at 21:51
  • 4
    \$\begingroup\$ It's a regulatory requirement for US and CE that grid-tie inverters do not output power without first syncing to the grid for some period of time. That's more than likely why they wouldn't turn on... \$\endgroup\$ – Some Hardware Guy May 6 '16 at 22:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.