G100 devices

[openspaceman]

Things get trickier with some other kit that doesn't have such a well coordinated ecosystem.

Yes as I am finding with my growatt ac coupled inverter, the supplier has told me the warranty will be negated should I charge the battery from a standalone array and MPPT solar charger, whereas this is an advertised method with the victron multiplus ESS.

I shall probably risk it because it will cost me five grand to replace the battery and inverter with victron kit anyway.

[Oldgreybeard]

Things get trickier with some other kit that doesn't have such a well coordinated ecosystem.

Yes as I am finding with my growatt ac coupled inverter, the supplier has told me the warranty will be negated should I charge the battery from a standalone array and MPPT solar charger, whereas this is an advertised method with the victron multiplus ESS.

I shall probably risk it because it will cost me five grand to replace the battery and inverter with victron kit anyway.

This is similar to the position I'm in with the Sofar AC inverter. I'm part way through adding another 2kWp of PV and did think at first of just using it with an MPPT charger to charge the batteries, in parallel with the Sofar. I think this could work OK as long as the second DC charger was set to cut off at a slightly lower voltage than the Sofar.

It is a bit of a compromise though, as it would mean that the reported data from the Sofar would be wrong, there would be more charge than it would know about. There are probably ways around that, by adding charge measurement from the new DC charging side, but it all seemed a bit messy.

What I've opted to do instead is add another inverter, so the second PV system is only AC coupled to the house. I can still use the Sofar PV monitoring, as that uses a CT around the live from the existing inverter and it's easy to run the live from the second inverter through the same CT. That will automatically sum the two currents that the Sofar measures.

The only other problem I've had to solve is to find a way to limit the total PV export to no more than the 6kW that I have G59 approval for (now G99). Because the new array faces more or less east, and the existing one more or less south, they most probably aren't ever both going to exceed 6kWp. The new array will mostly just provide early morning power, useful at this time of year when the south-facing array doesn't really wake up until mid-morning.

The solution I've come up with to make a G100 limiting device is to use another current sensor, that has an adjustable current switch, around the live wires from both PV systems and use switch signal to operate a timer relay. This timer relay will disconnect the second inverter for an adjustable time if the combined output of both PV systems exceeds 6kW. I very much doubt this will ever happen, but it will keep the DNO happy as nothing has changed as far as the grid is concerned. The timer function is needed just to try and prevent the tiny risk of the thing switching off and then back on again too quickly. I've bench tested this and it seems to work just fine. Not perfect, as it's restricting generation, rather than export, but fairly simple (and cheap) to do.

[Krill]

This is similar to the position I'm in with the Sofar AC inverter. I'm part way through adding another 2kWp of PV and did think at first of just using it with an MPPT charger to charge the batteries, in parallel with the Sofar. I think this could work OK as long as the second DC charger was set to cut off at a slightly lower voltage than the Sofar.

Forgive me for what may be a stupid question but would it not make more sense to have the MPPT set to cut off at the higher voltage and have the Sofar set at a lower voltage? Would there be a way to get the Sofar to pull the extra charge out to force an export from an earlier part of the day, so the export limit isn't breached at peak generation?

[Oldgreybeard]

Forgive me for what may be a stupid question but would it not make more sense to have the MPPT set to cut off at the higher voltage and have the Sofar set at a lower voltage? Would there be a way to get the Sofar to pull the extra charge out to force an export from an earlier part of the day, so the export limit isn't breached at peak generation?

The snag with that is that the Sofar is controlled by the Pylontech data connection, so will start to reduce the charge current and eventually shut off charging when the batteries command it to. It would have no control over the MPPT charger, wouldn't know it was there, and so it would falsely believe that less energy had been put into the batteries than was the case.

It would be possible to disconnect the data cable and set the Sofar up as if the batteries were lead acid, but that would mean losing the data that the batteries send through that data connection and which the Sofar makes available via the RS485 port.

[openspaceman]

This is similar to the position I'm in with the Sofar AC inverter. I'm part way through adding another 2kWp of PV and did think at first of just using it with an MPPT charger to charge the batteries, in parallel with the Sofar. I think this could work OK as long as the second DC charger was set to cut off at a slightly lower voltage than the Sofar.

This is my original plan as propounded by joeboy and seemed straightforward until Nowty raised a concern in a castaway comment.

It is a bit of a compromise though, as it would mean that the reported data from the Sofar would be wrong, there would be more charge than it would know about.

This did not worry me as I expected to be able to deduce the state of charge from voltage, but this only when the battery under little charge or load, but for the fact the battery is constantly monitored by the inverter. The inverter also reports back to China and has had firmware upgrade remotely.

So the fear is the inverter could sense the battery parameters had changed and decide a fault condition has arisen.

[Oldgreybeard]

This did not worry me as I expected to be able to deduce the state of charge from voltage, but this only when the battery under little charge or load, but for the fact the battery is constantly monitored by the inverter. The inverter also reports back to China and has had firmware upgrade remotely.

So the fear is the inverter could sense the battery parameters had changed and decide a fault condition has arisen.

The correlation between terminal voltage and SoC isn't always that tight, I think. This is data just pulled out of the Sofar that illustrates this, the terminal voltage tends to go up and down a fair bit without much change in SoC, as the system switches from charge to discharge, I believe:

SoC vs voltage.jpg (32.72 KiB) Viewed 1322 times

[openspaceman]

Yes I had commented on this previously which is why I qualified the comment by saying only checking under conditions of low charge or discharge.

I take it that the problem occurs because of the battery internal resistance??

[Oldgreybeard]

I take it that the problem occurs because of the battery internal resistance??

I don't really understand the cause, to be honest. Internal resistance will contribute to it, but is so low that it can't be the sole, or even the primary, cause of the variations. I suspect the culprit may be the BMS, specifically the on resistance of the FETs in the charge/discharge path, but can't be sure. Another possibility may be down to the way the battery and inverter communicate, perhaps. It's possible that there's a bit of latency, and that this contributes to the small variation.

All really speculation on my part, though.

[spread-tee]

Oh dear,

my plan is the same as you mention above, MPPT charger in parallel with the inverter. The problem you mention regarding the doubt about the SOC, is that just an issue of not having accurate data for your records or could it be potentially damaging for the battery or something else? I too was thinking of setting the MPPT to cut of say 0.1V below the inverter setting.

Desp

[openspaceman]

my plan is the same as you mention above, MPPT charger in parallel with the inverter. The problem you mention regarding the doubt about the SOC, is that just an issue of not having accurate data for your records or could it be potentially damaging for the battery or something else? I too was thinking of setting the MPPT to cut of say 0.1V below the inverter setting.

Desp

Joeboy didn't report any problems with his pylontech battery charged by victron charger off additional panels and that now runs in parallel with his home assembled battery of prism cells. I cannot remember his inverter details offhand.

It may well be a non problem even with the growatt system and it does get around the need to make another application to the DNO even though the MPPT charger is probably more expensive than a cheap stand alone solar inverter.

Personally I don't need to know how much my extra panels are producing but knowing how much battery charge is left is a bit important to me. I just want to extend my electricity self sufficiency a bit further into the "shoulder" months.

As a post script this is how my supplier put me off:

"I wouldn't advise connecting the off grid panels to the battery from so little information as I am unsure and unwilling to say if it would be safe to do so. If any damage is caused in the process that would of course void the warranty on the hardware. "

I have sent full details of my additional panels and mppt charger in order for them to make a more informed comment.

Evenso I remain unhappy about my uninformed choice to go with the growatt system as I wish I had bought a more open system of Victron ESS and pylontech battery.

As it stands the growatt does work well.