ccrause

To some extent yes. One can interface with e.g. a smart LG product via the LG ThinQ integration, unfortunately this still runs via LG's cloud "service" rather than a local network. Similarly smart Sasmsung products can be controlled via SmartThing which also requires a Samsung account. It seems as if many commercial "smart" products are just honeypot traps designed to lock the user in a proprietary ecosystem. Note that I do not have experience automating these proprietary smart devices, I use smart switches running opensource ESPHome or Tasmota firmware. Maybe one day something like the Matter standard will put the power of home automation back in the hands of users.

My app does change the arrow direction to indicate charging/discharging, so your Power Flow view is unexpected. To double check go to the Battery view, it should show a negative value for power when charging.

Wanted to check firmware version this morning and noticed screen was off. Left the screen on and it switched off again. So problem appears to be gone. Firmware seems relatively old, so screen issue apparently not due to updated version: M 6.0.2.2 S 1.7.2.4 CE 4.2.A

Also noticed a similar issue with my 8 kW Sunsynk. The screen seems to end up on the date/time settings page and stay on. Need to check whether firmware version changed from last requested update. System installed 2 years ago, only noticed this a week or two ago.

Getting advance warning of a potential strike requires electrostatic monitoring. These units seem to be much more expensive than the lightning detector that abd7 suggested (hundreds vs thousands of bucks). So the cost appears to rise steeply for the increased statistical protection. Unless you try to DIY something.

As mentioned already, shielded cable is your best bet to reduce interference when running ethernet cable close to other cables. Note however that there are several different types of shielding: https://www.truecable.com/blogs/cable-academy/ethernet-cable-shielding-types So do not just search for "shielded" cable, also check the shielded cable type. Also note that shielded cable should be properly grounded, use shielded (metal) connectors to terminate the cables.

I'm based near Sasolburg. I've checked the data over the last two weeks and the largest deviation I could spot was 0.5 Hz away from 50. Below the grid frequency trend for yesterday. The daily plot over the last two weeks never looks noticeably worse than this.

The grid frequency seems unbelievable, especially the 0 Hz portion when the grid voltage seemed OK. My suspicion is that there is a frequency measurement problem on the inverter side.

The short answer is to discharge your battery at night so that excess PV during the day can be stored (as pointed out by Scorp). To achieve this, set you early morning slot (which ends around the time your PV start picking up) to something like 20% (or whatever low value you think will treat your batteries with respect). But also consider how much backup energy you need in case of loadshedding. To estimate this, monitor your night/early morning usage, then decide whether you need 2 or 4 hour reserve.