Battery problem, no output: 09/03/24 – 10/03/24

I have an AC3 gen 1 inverter and 9.5kWh battery, installed in December 2022. Apart from teething problems with SoC it has been working reasonably well. Although the max. output is stated as 3Kw it regularly outputs 3.3kW when the load is there. I have a separate PV system and the AC3 has a current clamp on the PV mains cable to provide solar output data.

Problem is that it decided to stop outputting any power for about 24 hours.

At 11:31:28 on 09/03/24 the AC3 output peaked at 3.424kW. After that for 10 minutes or so it discharged at about 100 to 200 watts and then at 11:46 it ceased output altogether. The load ran at 50 to 400watts and had peaks of over 2Kw during the afternoon. The inverter just ignored these demands completely, sitting at 0watts. The data showing this comes from the GivEnergy portal. Throughout this incident the SoC was 70 to 80%, so plenty of charge available. The inverter was set on ECO mode and no scheduled commands were in place.

At about 07:30am on the 10th the sun came out, and the inverter imported the generated power as it should. It was still not responding to house loads although it was allowing them to be serviced by the solar generation.
I did not notice this problem until 11am on the 10th as I was busy. At that time we put the oven on creating a total house demand of about 2.6kW and that was when I noticed that it was being delivered mainly from the grid, which was wrong of course.

I carried out several interventions including rebooting the inverter and powering it down completely (including battery off), and restarting it, all of which it ignored, until 11:45am when it started working correctly, and has been since. I think the inverter decided to work when I set it to export at 3kW.

The normal frequency of data recorded on the GivEnergy portal seems to be one row every 5 minutes. However on the morning of the 10th there were cases of more than one reading per minute. During the minute at 11:01 there are 11 data points listed for the 60 second period. For the hour, 11am – 12 noon, there are 420 data points.

Any ideas what was going on here?

666tim666 The normal frequency of data recorded on the GivEnergy portal seems to be one row every 5 minutes. However on the morning of the 10th there were cases of more than one reading per minute. During the minute at 11:01 there are 11 data points listed for the 60 second period. For the hour, 11am – 12 noon, there are 420 data points.

I think it does this if the app is running: the app tends to update every 10 seconds or so, and when it gets it data, that also seems to go to the portal.

DD Thanks for your reply, and yes I use the app, so that would explain the high frequency of data records on the portal. I will carry out some experiments.

666tim666 If you haven’t already, do a ‘reset to defaults’ in the portal. This can cure most ‘confused inverter’ issues

geoffreycoan Thanks for your reply. I will try 'reset to defaults' if it goes wrong again although its working alright now.

666tim666 max. output is stated as 3Kw

In the datasheet max output is stated as 13A.

To get the power you need to multiply the current by the (grid) voltage. 230V x 13A =2990W.

Your peak output would suggest the grid voltage may have been 3424W / 13A = 263V. The inverter is supposed to shut down when the voltage exceeds 262V and it looks like that's might have been what happened.

Check the VAC in the portal readings to verify if this is the cause. Judging by the way the battery was charging, looks like it was sunny at 11am on the 9th. Maybe some other properties in your neighbourhood have had solar installed over the winter.

The inverter should have restarted automatically when the grid voltage reduced, but that's another question.

If you use my android app it shouldn't upload to the portal when monitoring an AC3.

In the history graph you can also see the grid voltage:

Note the gaps in the export (just after 16:00) when the grid voltage exceeded the limit.

Rubikcube Thanks for your suggestions. The grid voltage here is always a bit high at 240V - 245V or so but at the time of the incident it was about 243V. This is as reported in the portal stats and there are recordings every few seconds, and so I don't think the voltage went exceptionally high.

243V x 13A is 3159W and not the 3.4kW output actually recorded.

Now my inverter regularly reports an output of 3.3kW, and given that the voltage is rarely over 250, it must be supplying over 13 amps. It doesn't matter to me, in fact its convenient to have a more powerful device, but I don't want the electronics to cook.

But I still don't know what caused the thing to cease output for 24 hours. Maybe there was some internal scheduling malfunction with a control register, which expired after 24 hours, so the inverter started working again. Or maybe my command for it to export woke it up.

And no, there aren't many adjacent properties here with pv even though this is west Cornwall with its longer dawn to dusk timings.

I thought I might be wrong ... and I was right, I was wrong! 🙂

No other ideas. Hope it doesn't happen again. If it does, try not to panic and rush to change things randomly. An extra hour to do things methodically won't make a noticeable difference. Start by toggling eco off and back on. Then try restarting by command. Try reset to defaults, and restart again. Finally, when nothing else has worked turn it off and back on.