• General

  • 'Don't use battery during off peak' setting not working

inferior-opposite

I've found the same. The activity log says a hold is in place but the battery is still active. As it happens, the sun has unexpectedly come out so I'm happy that the hold didn't work as my battery is charging

Thinking out loud… the battery should be paused for the duration of the off-peak period when the charge is complete, but the charge is never completed because the target value is never met (in my case it's always 1% short). Could that be what's going on?

So there's several different things going on.

Hold charge is not being used on the regular charging schedules. I would expect the exact same behaviour you are seeing if you applied the charging parameters via the Giv app or portal, including adjusting charge rate.

For the Agile charge slots, where we apply the hold, we effectively set discharge power to 0.

For some Giv Hybrid inverters this does actually not work as you'd expect, as it can allow a few hundred watts through. More on that in this thread: https://community.wonderwatt.com/d/110-release-363-octopus-agile-window-hold-charge-functionality/51

I get the same result when charging at max rate. But it worked fine back in November.

I don't think the inverter is the problem, because we can see there was no SOC wobble issue up until mid December 2023. The last update to the BMS was on 2023-11-30, but I can see the battery was holding charge in off-peak just fine on 2023-12-08.

In Mid December the WW variable charge rate feature was introduced. The battery was no longer able to meet the target, typically 3–6% short. I worked around this by overstating the battery capacity by up to 15%. I can see the SOC wobble is evident on December 26, when it completed a charge early.

I will try and set a timed charge via the GE portal and see what happens.

@admin As you correctly predicted, I got the same SOC wobble when setting via the Giv portal 'Timed Charge' settings. The wobble is actually pretty large, draining the battery at the inverter's full rate of 2.6kw – so I think we can discount the possibility that it's the inverter 'leaking' a few hundred watts.

Looking back in the logs, back in November/December WW would send a command to 'Enable Eco Mode', which is no longer sent. Could it be that this command prevented the issue I am seeing?

    Avenir

    We don't touch Eco-mode for charge schedules, so your default should be to have Eco-mode on, have you not got Eco-mode enabled currently?

    We only set Eco mode back to true after a Timed Discharge.

        admin Hi, thanks for the response. Eco mode is currently on.

        The charging logs up to mid December look like this, with the Enable Eco mode command, which I assume is being sent by WW.

            Avenir Thanks for digging that out.

            So back then we were using the presets, but they were causing too many Inverter Timeouts.

            I don't think it's related to Eco-mode. It's more likely the Enable DC Discharge being set to False.

            We may need to do this during charge windows, and possibly during on hold.

                7 days later

                admin in the meantime, a workaround would be to allow the charge rate buffer value to be user configured. For example a +4% value might allow me to better cover the full period, and avoid the SoC bounces.

                    Avenir Yeah, though I think at lower charge rates for Giv, the end SOC target ends up being generally less accurate anyway. So I'm doubtful a simple percentage based offset would work, of course unless you specify a high enough buffer.

                    We can have a look.