Strange battery discharge

[CliveRichardson]

My battery is discharging overnight, although I have the system set to not discharge, as per the recent change in software settings, between 00:30 and 07:30. The SOC reduces from about 65% t0 25% over a period of an hour and a half, from about 02:00 but looking at the overall stats, the rate of discharge is constant during the 00:30 to dawn of about 0.12kW (not sure why it's not zero as it is set at). The discharge does not change when usage peaks (dishwasher comes on) but remains constant. The battery is 4 x HV2600, 2.6kWh units so this 40% discharge represents about 4kWh, not the 0.6kWh represented by the 5 hours at 0.12kWh shown discharge.
Could one of my battery modules be faulty? How do I check?
screenshot of the general graph and the SOC graph

.

[canton7]

I'm seeing the same thing - a constant discharge of ~120W when it shouldn't be discharging at all.

Interestingly 120W is my base load (with the fridge compressor off).

Here, the no-discharge period is from 2am to 5am.

[calum]

Curious. My system (which usually seems to be a magnet for all possible weirdness) is showing no discharge other than when expected:

@Clive, the fact you're losing more SOC than the 120W for 5 hours would suggest makes me wonder if you're running downlevel firmware on the BMS and battery? I had a lot of problems with my batteries jumping from 80% to charged, and then dropping 20-30% overnight when in theory there was no or minimal load. Updating to the latest BMS and battery firmware has made a world of difference there.

That said I'm assuming @canton7 you're already on updated f/w - and your SOC is only moving as much as you'd expect?

What I'm getting at is that there's maybe two different issues here - an SOC tracking problem, and a 120W battery discharge problem, which appear to be able to co-incide but can also occur separately?

[canton7]

I'm still on old versions - 1.48/1.02/1.45.

Looking at soc and kwh remaining, I'm losing almost exactly as much charge as I'd expect.

I suspect you might be right on the two issues point.

[CliveRichardson]

Thanks for your thoughts Calum. I am on Master 1.54, Slave 1.02, Manager 1.57. Do you know if they are the latest and are any of these the BMS software? Also where do you download updates from. The manual talks about getting them from customer support.

[calum]

The versions you mention are the inverter firmware, not the batteries or BMS. Those versions are visible in the same menu as the ones you've mentioned.

Firmware files not generally available for download, the kit is typically upgraded remotely via the network (this is one of the jobs of the WiFi dongle). In theory your installer would have an "Agent" account on Fox's systems to enable them to do this, in practice they don't usually have it set up properly and either they end up asking Fox to do it, or you do.

It is apparently possible to do the upgrade by downloading a file and putting it on a USB stick that plugs in to the inverter, which I assume is for the genuinely off grid installs, but for the "normal" residential stuff I've not heard of anyone getting hold of firmware files from Fox.

Contact emails and phone numbers for Fox are here: https://www.fox-ess.com/get-in-touch/

[Bobinho]

I noticed something very similar and am happy to find this thread when searching for the solution.

Some time ago I called my installer qith question:
"why, when battery reaches Min SoC [15%], does inverter display -130W instead of -15W [expected standby power]?"
He justified that this display in some situations displays strange things and probably it was my household background power (not the power generated/drawn on "On Grid" inlet).
It was discrepancy with what Chint meter was showing but anyway, reasonable as it were last sunny days, no charge periods set, so flat battery was seen only during the night.
SoC was dropping slowly, to about 12-13% before PV starts charging (BTW is it normal rate of self-discharge - 6-12%/24h?).

Recently I switched to peak/off-peak tariff and enabled Force Charge period overnight without charging from grid.
When battery is flat it still works as above, despite charging period being set.
But if batteries have some juice, inverter displays -15W and monitoring (both FoxEss Cloud and Home Assistant) shows 110W of Battery Discharge. It is followed by significant SoC drop which is also visible on inv display (haven't checked powers on display yet though) so I would push back on metering problem.

Logically thinking it means that battery pack for some reason needs about 110W. If it can be drawn from batteries, it is and is reported/measured. If batteries are flat, this power is drawn from grid.
I would call it significant power lost, as, assuming it is actually drawn all the time, 24h x 110W > 2,45kWh - nominal HV25 capacity. I would expect BMS + batteries to draw in standby amount similar to inverter, e.g. 15W. Of course the less is better

Inv Ver:
Master 1.853
Slave 1.02
Comm 1.66

Bat Ver:
M 00.01.004
Sx (x=1,2,3) 0x.00.01.01

Charge period 22:00 - 06:00 (no charging from grid during this)

[calum]

See this discussion https://github.com/nathanmarlor/foxess_ ... nt-7036872 on Github - the long and short of it is that the hybrid inverters are pretty power hungry