The voltage drop Wilko and I estimated for the connection between your inverter and the main intake was between 2(me) to 3(Wilko) Volts at 21 Amps when you export 5kW. Wilko cleverly included the drops across isolators, mcb and at connections which explains his higher figure. As you are so close to the transformer I reckon any volt drop along the cabling between your home and it is of the order 0.5-1V (20 A x 0.05 Ohms = 1V).
In sum between your inverter ac output installation and the transformer a worse case volt rise of circa (nb 3 + 1 =4V) of 235 + 4 = 239V. Not troublesome if your supply voltage is circa 235V since the disconnect figures are typically 255V rolling 10 minutes (IIRC) time averaged, 260V peak for 1 sec and 261 peak immediate - and 239V < 255V.
What we don't yet know is the effect the other solar pv systems at peak sun have on elevating your incoming mains voltage so be be very interesting to find out.
I turned up this article which you might find interesting which helps to explain our line of thinking on your peak sun disconnect problem. For example you will read about the cable between the inverter and the mains intake and the suggestion to double up on it if the run is long - hence our volt drop estimates - to see if this might make a difference. It also explains inverter cycling at high voltage rise. Wilko and I don't understand why there is no fault code reported by your inverter which tends to indicate this is not happening in you case unless the internal logic is that this is not a
fault per se but rather an
event - I recognise this is semantic thinking. Let us hope for some sunny days next week.
Solar Voltage Rise - why you should care | MC Electrical - https://www.mcelectrical.com.au/solar-voltage-rise-explained/