Strange load graph?

[andrewv]

Hello,

I had a Fox system installed last month (7 panels, 2xEP11 batteries and a 5kW H1 hybrid inverter). I've also got a Zappi EV charger.

I was monitoring the app graphs from when we were away for a week and noticed something strange. This is from the first day we were away:

The three load spikes in the early morning are the dishwasher. Apart from that, the only load is the fridge compressor going on and off, and a background level of appliances in standby etc. This is what I would expect, and have seen before.

This is from the next day, when the battery got to 100% for the first time, exporting to the grid from about 13:30:

When the solar falls away at the end of the day about 18:50, the battery takes over, but the load takes on a strange shape for a few hours - much more irregular, and sometimes even dropping to 0W. Then after a few hours the regular pattern takes over again at about 22:30.

This happens again for each of the next days till we return. The times vary a bit, but the load seems to start going strange when the full battery takes over from the solar, then gets back to normal when the battery is at 97%. Here are some more graphs:

(this one above has someone in the house in the afternoon, using some electrical items)

When we returned, all our appliances seemed to be working OK, and there was no sign of a power cut (e.g. the cooker clock was still correct). Now we're back, the battery hasn't got near 100% again due to greyer weather and us using more electricity, and our use in the evening is perhaps masking anything unusual. But I can't imagine what would make the entire household load go to 0W several times over a few hours each day. It definitely seems linked to running off the battery, with the battery getting to 100%.

The battery won't get to 100% again till I get the export MPAN and start force charging overnight, and I'll monitor what happens then. Is this just a measuring glitch?

[Dave Foster]

It does look a bit different but if you weren't at home there wouldn't have been any of the normal base load, lights, phone battery chargers etc.. - it's only small stuff but without that your house load would be very low (apart from the fridge compressor).

A couple of thoughts - that graph is very small, you might be better going to the foxesscloud.com website, log on and at the bottom of the screen on the graph - click on the labels 'output power', 'feed-in power' and 'gridconsumption power' and it will remove them from the graph so you are only looking at house load - it may be easier to see whether the load really was absolute 0 or was really a small amount above that which would seem more realistic.

The inverter draws power to keep itself running when there is no solar from the battery - the statistics tend to include it in the house load so you do sometimes see strange effects with 50-70 watts being shown in the load, sometimes in the battery discharge.

I guess finally to say the data samples are only taken every 5 minutes so it's not got the best resolution for this kind of precision, things can turn on and off in between the data samples and the load would never show them. If you are considering going to home assistant it connects directly to the inverter and the data samples will be very precise and approx every 10 seconds which makes it much easier to really see what is going on using that.

[andrewv]

Thanks for your reply.

I do think it looks strange - even though we weren't in, I'd expect to see the night-time load pattern just continuing for all 24 hours - nothing special happens in the evening. Normally, with the fridge compressor going on and off, it's going from 0.12kW to about 0.25kW and back again repeatedly, all day and all night.

Here's the website graph from 24/8/25 - scrubbing the cursor along shows the load goes to exactly 0kW four times (the label shows this when I do it, but blocks the graph, so I didn't screenshot that).

I get that samples can miss readings (I've boiled the kettle a few times, and it's totally missed it), but something strange definitely happened each evening when it switched from solar to battery at 100%, then it took till it went down to 97% before the load went back to looking normal again.

If this isn't a known issue then I guess home assistant might be the best way to look in more detail.

[Dave Foster]

It’s not an issue i’m familiar with but you’ve said a couple of things (100% - 97% which is where the BMS won’t allow charge) that make me wonder whether your had extremely low house usage and because the battery was at 100% the standby usage was not being subtracted from the load.

[andrewv]

Yes, I think it's a measuring issue to do with a full battery and a low load, as nothing in the house seemed affected. I'll check again when I next expect to see those conditions (probably early morning after a forced charge when I start doing that). It was a bit disconcerting to see my brand new, never been fully charged before system do it when I was 150 miles away though!

[andrewv]

One more quick question about the graphs: I sometimes see a spike where it imports from the grid, even though there is plenty in the battery and the load is not over the inverter size (5kW), like at 07:41 below. I would expect it to import at 18:31 as the load was > 5kW then, but what causes the occasional grid import at other times? Is it just an artefact of sampling and it just happened for a second? I guess I could try home assistant and see what's going on at a 10 second level, but just wondered if this is normal.

[Dave Foster]

Yes it's just a sampling artefact - when a load changes the battery has to stop supplying power instantaneously but with the capacitive nature of the design it takes a short time to adjust during which the power has to go somewhere i.e. the grid is used as a sink (or a source if the load switches the other way) - you still see it with HA but as its so fast it really is just a transient.