Hello From California

Can you be a bit more specific: do you know exactly what kind of Enphase inverters you have?  (I assume they're microinverters; regardless, could you provide a part number?)

This feature is a work-in-progress right now with the Genetry Solar inverter (actually on today's programming agenda!)  Basically, if the Enphase grid-tie inverters will throttle down with an AC input frequency shift, it should work very well with a GS inverter connected to a battery bank. 

Basically, UL regulations require that a grid-tie inverter shut down if the AC input frequency exceeds 62Hz.  Some grid-tie inverters will linearly throttle down their output between 60-62Hz AC input--and this is what the GS inverter will use to "throttle back" the grid-tie array as needed.  Depending on what model of Enphase inverter you have, this may be a standard feature, or maybe you can configure it for this desired operation.

Battery charge: if a grid-tie system is connected to the AC output of a regular low-frequency inverter (i.e. GS inverter), it will actually "backfeed" power through the inverter and charge the batteries.  (Obviously, the low-frequency inverter must be designed to handle and control this, otherwise the batteries will get overcharged and/or damaged.)  GS inverters will detect this reversed AC power direction, and automatically switch to "grid-tie master mode", where it regulates the input power from the grid-tie array as needed for either battery charge and/or AC output loads.

• If there is no need for power (i.e. no load, no charge), the GS inverter will shut down the grid-tie inverters.
• If there is no sunshine (i.e. nighttime), the GS inverter will provide power from the battery to run any necessary loads.
• If there's full sun, a small 1500W load, and no battery charge required, the GS inverter will throttle the grid-tie system down as needed to match the load
• If there's full sun, and 5100W of load, the GS inverter will provide the needed power beyond the output of the grid-tie inverters (grid-tie inverters full throttle).
• If the batteries need charged, any excess power produced beyond the loads, would charge the batteries (with a configurable current limit).

In short, to convert your grid-tie system to an off-grid system, you would only need to add a GS inverter and a battery bank for a completely functional off-grid power system.  (This assumes the grid-tie inverters have the required frequency-shift throttling function.)

Yes it's a microinverter system. The SKU is M215-60-2LL-S22-IG. It was installed in 2015 so I don't believe it's their IQ system which I think are the ones that'll throttle down like you're saying. I believe these will go full throttle untill they're outside of their Hz range but I could be wrong. But even if I'm right and they're either full on or off could that possibly be a feature the gs inverter could one day incorporate and basically chance the frequency between 60 and 62+ Hz depending on whether the bank has enough buffer for the full power or not?

Looked it up...appears you're right on that: the M and S-series microinverters don't support frequency shift, but the IQ system does.

Bit of a bummer there.

Obviously, if there's a way to programmatically handle an off-grid system, you bet we'll be able to do it.  As long as there is no hardware change required, we can tweak firmware all day long.  Yes, it would be possible to have a setup for "grid-tie system size", and the inverter wait until the combined need for charge/AC loads exceeds that, and then bump it on.  The only problem is that battery charge (amperage) could not be limited (remember, it's on or off!)...and if you have those nice puffy clouds (sunny / cloudy / sunny / cloudy every few minutes), that throws another whole monkey wrench into the issue.  Because if tripped off due to "AC Input Error" (which a frequency shift would be), most microinverters take 5 minutes before they will start back up again.

I'm afraid that trying to battery-back a fixed on/off grid-tie setup would be an extremely crude proposition at best.

Seems to me that Enphase could provide a microinverter firmware update for their M-series inverters--throttle shift by frequency is a super easy function to implement!  (On the GS inverters, it's literally a single line of code.  The Enphase microinverters are already regulating the input with the MPPT algorithm--and it's a piece of cake to tweak that "set" value based on the input frequency offset from the "root" frequency.)  But big corp usually doesn't care about the little guy...and even less about someone trying to be thrifty.

You know what I'd suggest?  If you truly want to go off-the-grid...ditch the microinverters.  (Yeah, that's a bridge-burning operation right there!)  If you only have 3kw of solar, that should be very easy to rewire and connect to an MPPT-based battery charge controller.  Assuming you settle for a 48v battery bank, you would only need a single 60A MPPT to fully harness 3kw of solar.  (Average "float" voltage say 50v * 60A = 3000W.)

• The Morningstar Tristar TS-MPPT-60 is a very efficient (no cooling fans), solid-designed controller--my only complaint with it is that it seems a bit easy to blow up (in my experience...though Morningstar did have stellar warranty response).  And it is a smidge pricey at >$600.
  • Keep in mind that the maximum input solar voltage is 150vDC; planning for cold weather (even though you're in Cali) is a good idea, lest a once-in-a-century cold snap come through and blow out the charge controllers with overvoltage.  Yes, solar panels get THAT much more efficient as the temperature drops.
• Another option would be the Epever Tracer AN-series units; I have two of the Epever Tracer 8420ANs (both connected to a 4kw array each)...and apart from a somewhat slow voltage regulation response, they've thus far been solid units (again, no cooling fans).  They are also much cheaper than the Morningstar.
  • The xx20AN units are rated for 200vDC solar input, which is great for the newer solar panels with higher voltages.

Most battery-backed low-frequency inverters will allow you to connect an AC input (which could be the grid) to them.  In that case, if the battery is low, they can switch back over to grid and/or charge the battery from the grid. 

I'm not really looking to go 100% off grid, I was wanting to just have the ability to go off grid incase of a grid outage so ditching the microinverters isn't really an option right now. Even if i managed to go 100% off grid I'm pretty sure the law here in California requires me to still have electricity service connected to my house so even if I had the main breaker off id still be paying their minimum fees. One can hope that enphase would release a firmware update that would allow throttling but like you're saying they probably care more about the bottom line than they do about the little guy, why release a free update when they can sell you their newest system. If I'm lucky they'll discontinue the M series and my microinverters will soon go out under warranty and be replaced with the IQ system haha.

What about some kind of variable dump load for the extra electricity instead of trying to turn them off and on? I assume that would require an extra device ontop of the gs inverter but just a thought?

That would definitely solve the issue.  Yes, it would require an external device; I can't just burn up the extra power in the GS inverter 😉

Easiest way would be to find a battery-side dump load: if the battery voltage gets too high, the dump load automatically shunts it down.  I actually am aware of a company that makes products that do exactly that...so that actually might be quite a viable option.  So we would have an LF inverter + batteries + precision battery shunt setup.

Would probably also work with any other "dumb" low-frequency inverter on the market too.

You know, if you want to pursue it, that should be completely feasible.  Dump load would have to be able to handle the full output of the grid-tie system though (assuming no charge, no load).

Ok so if my production was greater than my usage the gs inverter would see this and put the excess power back into my bank and then if the voltage in the bank became to great than the dump load system would kick in and keep it under a preset value? I'm assuming I can set the max charge on the gs inverter a hair higher than the dump is set at as a fail-safe at which point the Hz would raise above the microinverters limit and shut them down for the 5 minutes or so? Am I understanding that correctly? Is the 6k inverter capable of charging at that rate assuming the bank is sized to accept that kind of current?

That is the basic premise, yes.  The reverse power flow charging the batteries is actually an uncontrollable "feature" of the total inverter design--the only way the GS inverter can regulate or control it would be to throttle down the grid-tie inverters (or shut down).

If you're going with a 48v battery bank, 3kw works out to about 62A, which should be quite feasible.  24v would be 125A, which may be possible...but might also overheat the FETs (inverter would shut down).  Depends on the ambient temperature and total load.  The GS inverter will automatically throttle down maximum charge current based on the internal temperature--which with a fixed input, would cause the inverter to shut the grid-ties down if it got too hot.

Yes, charge settings are completely configurable.  You'd basically have to set the inverter up for "100% power load" (i.e. max charge current, etc.) so the power could be safely consumed--otherwise it'll trip the grid-tie inverters off.

If for whatever reason the dump load couldn't keep up, battery charge current went over the limit, overheat, etc.--the GS inverter would automatically start a frequency shift (to throttle down a compatible unit).  This would immediately fault the Enphase microinverters out...the GS inverter would revert back to the normal frequency (because no power coming in), and the microinverters would run their default 5-minute timeout, and the cycle would repeat.

Crude?  Yeah.  Will it work?  Pretty good chance of it.  Sure beats being completely without power, though 😉

Ya might be something worth exploring for the peace of mind even if outages aren't to common in my area. What company makes the products that can be used for a dump load system? Maybe some of your balancers on steroids? 😄 

Umm...the balancers are absolute maximum 6vDC, and max amperage of 5A, so only 30W.  Little bit shy of the 3kw you'll need if there's no load/charge, etc.

So the company that I'm aware of is actually on this forum, it's EPC Corp @paul-z...it sounds like they make 1500W dump loads or possibly even higher.  We might be able to get a pretty nice setup for you here 😉

We may have been wrong about enphase not releasing the firmware for the M series.

https://enphase.com/en-us/storage-m-series

I assume this would hopefully make my microinverters fully compatible with the gs inverter future throttle feature?

That would be extremely nice (was actually testing the frequency-shift throttle feature yesterday and today)...

Looking at the link you provided, I do see a potential red flag:

"PLC Over AC Line"...this could be something different, sort of like older X-10 "home automation" systems, or AC-wall "intercoms", etc.  Unfortunately, this will likely be a proprietary communication protocol--and while it could be sleuthed, the GS inverters do not have capability (yet!) for overlaying communications on the generated output power.

It may well now support frequency shift throttling...but without an official word from Enphase (or a grid-tie inverter to test), I can't say for sure.  I did notice that the "datasheet" for the Envoy S indicates that it can communicate with "up to 600 microinverters."  Not sure whether this refers to the limits of a unique indexing system, or more the length of wire, etc.?

So close.  Maybe it does have the functionality?  Maybe someone else has already tested?  Did a few minutes' web searching, and really didn't find anything clear.  TBH I'm not that familar with grid-tie systems...but supporting 'em is certainly something we at Genetry Solar will do to the best of our ability.

Well darn. The crude method might still be worth doing. Keep me posted on your tests and whether or not you think it could work in my situation. 

I would have to test an actual Enphase M250 to find out.  But it sounds like you have to have the "Envoy S" in order to update the firmware in the M250 units--which means they likely have a pretty sophisticated communication going on over the power line.

The crude setup likely will work, to some extent or another.

Another suggestion I saw online was to configure the M250 units to different root frequencies (if this is possible--you would have to find out).  Let's say that 8 of them are set to 60.00Hz, 8 of them to 59.5, 8 of them to 59.0, 8 of them to 58.5...or something like that.  This way, when the GS inverter starts to throttle upwards (to reduce output power), the M250 microinverters start to trip off in batches.  The further up the frequency goes, the more microinverters trip off.  They will have a 5-minute "restart" time...still somewhat crude, but a lot better than dumping all the power.

The basic premise there being that if they are configured to trip off at (frequency + 2Hz), if you can change the [frequency] component, you can in theory change the trip-off.  So if it's set to 60Hz, that's 60Hz + 2 = 62Hz shutoff.  If set to 59.5, that's 59.5Hz + 2 = 61.5Hz shutoff.  59.0 + 2 = 61.0Hz shutoff, etc.

Gotta love that Envoy S...looks like it rolls in at approximately $600.  A 4kw Epever Tracer 8420AN MPPT costs less than that 😉

Another idea:

If you already have an Envoy S with your setup of M250s, can you set it to "zero export" mode?  If so, we could certainly hack something around that to "mis-use" that provided throttle for the intended purpose...actually quite easily I think 😉

Zero export mode requires a current sensor on the house breaker panel, and the purpose of "zero export" is to reduce the electric consumption to zero...but don't put any power back into the grid.  Which requires precision throttling.  And that precision throttling can be completely leveraged by the GS inverter.  I'd basically need to throw together a quick little project to "squelch" the output of the "house breaker panel current sensor" based on the frequency shift of the GS inverter.  Compared to the GS inverter project, that's a piece of cake 😉