I'm not familiar with TT, but N/E bond to be provided when disconnected from grid (running from battery and/or PV), not when still connected to grid and adding power from battery in parallel with grid power (don't know if your Deye does, my brand does.)
First drawing posted above by Bop shows L1/L2/L3 in various configurations.
The others only show L.
OP has L1/L2.
OP also has N, but not G. So this is a case where lost N leaves single phase loads on L1 or L2 in series with N, same lost neutral issue we have in the US, can be over-voltage and brownout.
I'm glad to hear that N is supposed to bond to G when inverter gets isolated from the grid, same as US so I understand it.
Input 3-pole 60A breaker isolates N from grid as well as L1 and L2, like a mobile application in the US.
Following this 3-pole breaker is RCD, meeting the requirement of breaker protection for RCD that was mentioned by Bop.
I think OP should add another RCD followed by inverter, so:
60A breaker --> RCD --> inverter --> RCD --> breaker panel busbar
(because based on the photo it doesn't appear feasible to connect inverter after existing RCD)
OP needs access to a ground wire for the inverter to do N/G bond.
If OP's system matches TT drawing given by Bop, equipment chassis does NOT connect to N. That is something which can be checked with DMM, turn off main breaker and use ohm meter. Can also check that chassis of multiple devices are connected together. But don't know if that is because they happen to share a water pipe, or because there is a ground rod. Would be nice to know there is a dedicated ground.
OP also mentioned to me he saw a dangling green wire on a pole outside, suspicion is grid has lost its N/G bond.