Warpverter

[Warpspeed]

Is the Growatt AC coupled to Warpverter?

Ordinarily, Warpverter is just an inverter. Should also work as battery charger. I've thought separate monitoring of battery voltage could be used to alter frequency. Also enable dump loads.

Yes direct 235v ac connection between Growatt and Warpverter.

Warpverter is just a dumb crystal locked inverter, no real advantage in moving its frequency away from 50.00Hz.
The Growatt would just chase it around in frequency anyway, with no net effect.

Excess power from the Growatt goes into the battery as you would expect. And this has its dangers as well as advantages.
It really needs a control system to prevent possible overcharging.
We are coming into winter down here, so overcharging is not likely to be a significant problem right now, and becoming less so.
A dump load would fix it, and so would shutting down strings of panels feeding the Growatt.

The way this whole project is going, its going to pay for itself pretty quickly I would think

 

[Hedges]

Dump load (or DC shunt regulator) would be one way. Available for wind, so an off-the-shelf solution.

If Growatt does frequency-watts, then Warpverter increasing frequency would tell Growatt to reduce wattage output.

That is a newer grid-tie standard here in the US, part of California's Rule-21.

SMA has done that for a couple decades with Sunny Boy in Island mode, AC coupled to Sunny Island.

There seem to be some other ways to command reduced power from GT PV inverters. One is by Ethernet, Modbus parameters. Another I think is digital pulses.

SI + SB work smoothly, hovering at a frequency where power produced matches load & charging needs.

 

[Warpspeed]

I am completely unfamiliar with commercial equipment or the official grid tie rules and regulations.
I always design and build my own equipment from scratch. Much more interesting, and I get whatever features I want to have, and its for off grid anyway.
But you are quite right.
A detailed read of the Growatt book of words might suggest a cunning scheme.

 

[rogerdw]

Is the Growatt AC coupled to Warpverter?
Were loads in excess of what Growatt put out, so Warpverter simply reduced its own output?
Or have you implemented frequency shift to regulate Growatt output?
Or, although Growatt exceeds loads, it simply backfed through Warpverter charging batteries (which didn't reach full)?

Ordinarily, Warpverter is just an inverter. Should also work as battery charger. I've thought separate monitoring of battery voltage could be used to alter frequency. Also enable dump loads.

Warpspeed's already answered some of this and can explain better than I can ... but yes, directly into the AC-out of the Warpverter.


Yes, the loads were often up to 6 or 7kW over and above the 5kW the Growatt is contributing ... although first thing in the morning there's not much load on the system ... say <1,000watts.

Here's an example with actual numbers ... 9.25am 28/04/24

All 4 charge controllers each putting out ~1,500w ... the Growatt showing 4,440w

Battery volts showing 56.25v (so clearly being charged ... normally only 48 or 49v at that time of morning unless the sun has started really early)

AC volts at 244.5v ... normally at or below 241v

43 amps coming OUT of the Warpverter (took me a while to wake up to the fact it was coming OUT and not going IN)

152 amps feeding INTO the battery.

At the same time the power meter reading the power out of the Warpverter was showing 2,154 watts ... and this really confused me because I knew the current house load was <1,000 watts.

So the power meter is possibly showing backfeeding as if the Warpverter is supplying a bigger load rather than the actual house load. And I suppose the Warpverter is supplying more power ... the extra is going into the battery.


I keep an eye on it manually at present as I work in the next section of the shed anyway, but I will need to automate it somehow for safety.

The highest I've seen it push the battery voltage was to 60 volts ... which is okay as the battery equalises at 61 and could go to 62 ... but as soon as I add a load (6kW floor heating) it all goes back to pretty much normal.

So a simple solution may be to monitor the battery and add more loads via relay or contactor if it gets too high (disabled during any equalisation charging). In summer time, I may need to shut down some arrays perhaps to keep it under control.

I discovered in the settings that the Growatt will start to limit if the AC-out gets to 253V and I could change that if I can get into the setup system.

 

[Hedges]

Frequency shift may emulate what a rotating generator does under light vs. heavy load. It spins up higher with less load. Sunny Island battery inverter emulates that (but only above nominal frequency.)

Also, an induction motor used as generator, if driven at a fixed RPM, would deliver more power if grid feeding it dropped to lower frequency, deliver less power as grid frequency increased, ultimately start drawing power if grid frequency exceeded its rotational speed. That seems to be what Sunny Boy GT PV inverter emulates (again only at elevated frequencies.)

Such a frequency shift scheme might help your inverter couple with rotating generator fed by hydro. If you install GT PV like the Growatt at a remote location, assuming is supports frequency-watts, that would control it without additional wires.

Of course if your generation sources are near the battery, or have separate wires back to the battery, they can DC couple.

SMA has a scheme the works with multiple inverters distributed over an area, has used this for a couple decades, so a good example to learn from. They say is supports GT PV up to 2x the wattage of battery inverter.

 

[rogerdw]

If Growatt does frequency-watts, then Warpverter increasing frequency would tell Growatt to reduce wattage output.

Thinking about it ... the only time I've seen the AC-out over 241.5 has been when the Growatt is pushing heaps of current back through the Warpverter ... and as I mentioned above, presently the Growatt is set to limit the output if the AC gets to 253V.

I saw it get to about 246v on the first morning and panicked a bit ... so shut off a couple of arrays. Once I thought about it, I realised I could use that power and so switched on the heating ... and turned the arrays back on ... and it all went back to normal looking levels.

Maybe I could bring the Growatt max voltage number down a bit which should throttle it back earlier. Then I can simply rely on the existing charge controllers to look after the battery level. Might be as simple as that ... though if I have loads that could use the power, I'd rather be using the power.

 

[Hedges]

I think you're going to get AC voltage drop with load and rise if backfeeding to charge battery.
Feed-forward regulation means it doesn't correct for that, but within some tolerances (IR drop) it will be OK.

Does your Growatt throttle, or just drop offline?
Some settings might knock it offline for 5 minutes, others just switch off output, restart immediately when grid back within spec. or maybe that would oscillate.
I prefer continuous adjustment.

besides frequency-watts, there is also volts-watts. Don't know what if any are in the firmware for your market. But it might have firmware for world market including US.

 

[Hedges]

I discovered in the settings that the Growatt will start to limit if the AC-out gets to 253V and I could change that if I can get into the setup system.

If that is Volts-Watts, it keeps AC within regulation.

But you still want to avoid over-charging battery. Unlike my Sunny island, your Warpverter will take everything Growatt delivers and stuff it in the battery.

How about a battery high-voltage signal to a relay to disconnect Growatt? (leaving charge controllers to carry the rest of absorption.) That is, if enabling a dump load isn't sufficient.

 

[rogerdw]

I think you're going to get AC voltage drop with load and rise if backfeeding to charge battery.

Yes.

Feed-forward regulation means it doesn't correct for that, but within some tolerances (IR drop) it will be OK.

Agreed

Does your Growatt throttle, or just drop offline?

Haha, very good question.

I prefer continuous adjustment.

Definitely, me too.

besides frequency-watts, there is also volts-watts. Don't know what if any are in the firmware for your market. But it might have firmware for world market including US.

Yeah ... so how do I find out? Ask someone who knows? Where do I find such a person? Read all the setup data and somehow interpret that?

What if I simply wind the Warpverter output up to 253 and see what happens? Am I game?

 

[Hedges]

Yes, monitoring Growatt amps vs. Warpverter Volts and Warpverter Hz is a good test.

1) RTFM
2) Go through the configuration choices of Growatt. (i.e. read the setup data.)

People sometimes set the wrong country code and it has different limits from what's needed. Like Japan 100V vs. US 120V.

Sunny TriPower was supposed to support Rule 21, but that wasn't in its drop-down menu choices. I asked SMA what update would have that, installed the update they sent, still didn't see it. But, the frequency-watts and other parameters were already there, and by simply widening its UL-1741 frequency range to UL-1741-SA limits, I already had it working.

I've seen "Island Mode 60" in a screen shot for Sunny Boy Storage (HV battery inverter), but SMA support denies the product has that mode. It may be in a different firmware update (not one the battery I've got supports.)

"Firmware" &!@#$!

That's something you mostly avoid with your Warpverter implementation.

 

[Warpspeed]

Official supply voltage spec in oZ is 210v to 265v.
The power utilities guarantee to stay within that, and any approved household appliance to Australian Standards will work within that range too.
So voltage is not that critical, the grid goes up, down, and sideways all the time, and everything still works just fine.

That is why I chose feed forward to control the Warpverter. It offers much better dynamic response to rapidly changing loads, to reduce light flicker issues. The long term static steady state voltage regulation is not perfect, but then it doesn't really need to be.
It will sag under load, and rise when back feeding.
Fitting current compensation to the feed forward will fix that. And it works in both directions. Only three extra components are required to do it.

"Firmware" &!@#$!

Warpverter is absolutely guaranteed to be totally software free.

It should be possible to buy a directional AC power meter ? Don't know never really looked into that.

 

[rogerdw]

1) RTFM
2) Go through the configuration choices of Growatt. (i.e. read the setup data.)

I have read all that I can find ... but none of it mentions the parameters that I saw online while connected to the history of the actual inverter I bought. That's where I saw the 253V maximum figure. Wish I'd at least screenshot all the settings.

Frustratingly I can no longer even find the option on the site to view that again!!!

"Firmware" &!@#$!

That's something you mostly avoid with your Warpverter implementation.

Haha, yep. I don't disagree with Warpspeed's decision to avoid that altogether.


Anyway, I did wind up the voltage to 255V four seperate times and I learned a few things.

1) When it reaches 253 - 255V the Growatt does not shut down

2) It does appear to throttle back

eg. It was outputting 4,400 watts and I wound Warpverter volts up to 248V ... and the Growatt dropped back to 3,900w

At 253V it dropped back to 3,400w

At 255V dropped to 2,900w

So wound back to 240V and Growatt up to 3,700 ... and over a couple minutes 4,240

(watching the other charge controller outputs it seemed consistant all the way through during my testing)


I've mentioned before about the extra noise the Warpverter makes if an electric blanket is running ... or any time the Growatt is running while connected ...

... well on my fourth test, I switched off all the other DC charge controllers so the Warpverter was running off the battery ... and the Growatt was running at full power ... 5,000w. The AC was at 244V which happens if the Warpverter is back feeding a fair bit.

Adjusting the voltage seemed more responsive ... didn't take as long to change and settle ... and after 30s I had it at 250V and the Growatt was down to 4,500w.

As I wound it around the 252-253 mark it suddenly went a lot quieter which was a surprise and the Growatt was down to 3,600w

I did take it to 255 but the phone ran out of memory so I don't remember what the final output was. Not much lower I don't think.


So I did learn something ... just need to process it all.

I do know I'll be in the city tomorrow, so I might just shut off the Growatt to prevent potential overcharging ... and leave the Warpverter ticking along like it used to. The charge controllers seem to run fine and prevent any overcharging.

Warpverter is absolutely guaranteed to be totally software free.

Haha, thought you may say that.

It should be possible to buy a directional AC power meter ? Don't know never really looked into that.

Had never even thought of that. I do know it would provide a lot more help if I could see which way the power was going ... and the actual amount too. There's a few options on aliexpress ... I shall research some more.

 

[Hedges]

It should be possible to buy a directional AC power meter ? Don't know never really looked into that.

Wattnode has some power monitors, can be used for single, split, or 3-phase.
These have communications rather than display. Often used by inverters to measure power at grid, shave imports and exports.

https://www.powerwisesystems.com/application/files/2215/1230/2952/Data_Sheet_WNC_MODBUS.pdf