Timed import and export no-longer working on AC3 since yesterday

As of yesterday, it seems that the timed import/export option on my AC3 inverter is no-longer working, and neither does it work via the "App - Quick Settings - Discharge at Full Power" and "App - Quick Settings - Charge at Full Power" options, which ultimately are changing the same set of registers on the AC3.

On poking around, it seems I always get inverter timeouts on two of the registers in the Charge/Discharge schedule settings:

I've tried resetting the inverter, and it didn't help. Is anyone else experiencing this, or does anyone have any suggestions? Everything else seems to be working normally, i.e. the ECO mode is fine, and the battery is charging/discharging normally according the solar input and energy requirements of the house.

Fraz I think the per-discharge slot limit is relatively new, so your inverter might not have it. But the AC charge limit has been there for a long time.

When you say they don't work, do they completely ignore the request, or does the battery go idle (as if it had already reached some limit) ?

On the timeouts, I'd be trying things like snooping the network to see if the requests are getting through, but that sort of thing does require a certain amount of familiarity with such tools.

DD I don't have a lot of time to play around with it right now, but it appears that it just ignores it and stays in ECO mode, pretty much. I see the requests and responses in the logs, which look fine, e.g.:

and these settings match the settings that appear in the logs when it's been working in the past.

Re. the network snooping, I've used wireshark and ngrep in the past, although not for several years now. I'll maybe give them a whirl this evening.

In fact, on looking closer, things seem to be getting worse. It's not been charging since around 8:20 AM this morning, which is about an hour earlier than I started tinkering with the settings to see if I could figure out what was happening:

And here are my logs for the period before and after the time the battery stopped charging despite there being excess solar:

Fraz battery level seems to drift quite a lot despite there being no apparent power drawn from it.

DD Not in my experience, especially since one of the battery firmware updates last year. But anyway, that's not what is being shown here. It's the battery not charging despite there being power exported to the grid.

For the last hour or so I've been exporting 3+ kW, and the battery (which is in ECO mode), is no-longer charging:

Fraz seemed to drift down from almost 70% to around 55% ? But anyway...

Quite high consumption between 08:30 and noon - roughly following solar? EV charger or eddi or something ?

I avoid the quick-settings on the app. I've never been quite sure how/where it remembers the previous settings when you pause, for example, to be able to put them back as they were afterwards.

You mentioned 'resetting the inverter' earlier - is that reset as in restart, reset as in restore-settings-to-default ?

DD

seemed to drift down from almost 70% to around 55% ? But anyway...

It's genuinely not drifting downwards. I just turned off some of the overlayed graphs in the first picture I posted. This shows the battery load correlates with the charge dropping:

Quite high consumption between 08:30 and noon - roughly following solar? EV charger or eddi or something

A solar iBoost for the hot water tank (same idea as an Eddi, but older), along with various machines being on in the house.

You mentioned 'resetting the inverter' earlier - is that reset as in restart, reset as in restore-settings-to-default ?

I've done both - a reset to factory defaults and also a restart of the inverter. Neither helped 😞

Fraz Sorry, can't really think of much else that might be helpful. Have you tried the bbc basic app? Not sure how well it works for AC3. Available for various platforms - see https://community.givenergy.cloud/d/438-how-accurate-are-the-import-from-grid-metrics/16
(Bit clunky on linux, but works well enough.)
Or for android, https://givenergymonitor.wixsite.com/tutorials

Both are using modbus to get details directly from the inverter. If you know python, you can see everything using a library (though again, AC3 support may be a bit behind)

DD Thanks and no worries. I don't really doubt the metrics (and I have a Zappi/myenergi setup that is showing the same things), but those are interesting links. Also I'm very familiar with Python, as it happens. But your prompting took me to https://pypi.org/project/givenergy-modbus/, which looks interesting also.

I've sent an email to support@givenergy.co.uk, so we'll see what they say about things.

Fraz That version is probably very out of date, and wom't support latest firmware which requires crcs.
I'm currently running my own fork: https://github.com/divenal/givenergy-modbus-async with some random scripts in https://github.com/divenal/modbus-scripts but that probably only works with hybrid devices, and maybe first-generation single-phase AC.

There's also a copy in https://github.com/britkat1980/giv_tcp/tree/dev3/GivTCP/givenergy_modbus_async which supports a wider array of products. That's bundled inside the GivTCP repository - author seems to have much closer access to GivEnergy than the rest of us. We're slowly working towards getting them merged together.

Incidentally, if you're on IOG, I have scripts in https://github.com/divenal/givenergy-scripts which can get hold of charging slots, etc.

DD
Thanks. Took me a while to work out what IOG meant, but I'm on Octopus Flux right now. I generally switch to (unintelligent) Octopus Go in September time and then back to Octopus Flux in early April. Although this winter I'll probably switch to their Cosy Octopus tariff, given I'll have a heatpump by then.

Fraz I generally switch to (unintelligent) Octopus Go in September time

With a zappi you should qualify for intelligent go. (That's my setup.) I decided to stick with IOG for now, rather than trying Flux, but probably quite marginal.

Wow, so Givenergy support got back to me very quickly (kudos to them!), and it turns out that my installer changed a setting remotely, presumably by accident/mistake. They had no reason to do this:

Snap...

... however in my case the DNO has mandated the change upon the installer.

I wonder how the update firmware / recalibrate action is meant to work when export limit is zero?
I'm holding off the D0.612-A0.612 update for this reason - I don't want it to fail.

Fraz It's interesting that the 'grid export limit' should have an effect even when 'enable export limit' is false (whereas for Icarus it got set to true)
And it's unintuitive that a grid export limit should prevent battery charging (ie effectively force export).

Icarus Whaaat? Why would the DNO not want you to be able to export?! That's bonkers!

DD

The issue you are having is related directly to the settings your engineer has set on the 31st of July around 13.27PM. The Grid Export Limit was set from 3613W to 0W. Setting this to 0W will stop the battery from taking charge off solar even if it's on ECO Mode.

Yes, I was very surprised by this also. (Pasting the text here in full rather than it being part of a screen-grab, to help with search). Although it's also not-quite true, as my battery was definitely charging for a while this morning, when this setting was still active.

Icarus I wonder how the update firmware / recalibrate action is meant to work when export limit is zero?

The battery can still be discharged to home load.

Fraz Why would the DNO not want you to be able to export?! That's bonkers!

Setting the grid export limit to zero on a AC coupled system doesn't restrict export. It only prevents discharging the battery whilst the panels are exporting.

Rubikcube Setting the grid export limit to zero on a AC coupled system doesn't restrict export. It only prevents discharging the battery whilst the panels are exporting.

Thanks @Rubikcube I didn't realise that.
I will verify with a battery export tonight when dark just to satisfy myself.
Then I really need to wait for a decent plunge period some night to do the update firmware / recalibrate malarkey.

Icarus I will verify with a battery export tonight when dark just to satisfy myself.

Just FYI Rubikcube I found that on my particular AIO, forcing an export after dark ( when solar PV is zero) would not allow more than about 100 watts towards the grid.
As far as I can tell my zero export limit does indeed seem to be constraining what I can give back to the grid after dark.
So for now I'm not going to risk a firmware upgrade (unless GE can clarify that the limit temorarily gets overriden during calibration). Maybe this is just a scenaro that has never been considered (because they assume at least 3680 watts export is posssible) ?