Ah! I didn't realise the board was feeding sub mains.
I thought it was an overkill seeing as you are using Compacts hence no fly leads!
For others here thinking wtf, consider a metal cu on a TT system. The earthing arrangement is likely nowhere near as good as with TN systems. The incoming meter tails are stranded copper cores that are sheathed and insulated, in effect double insulated but it's actually wrong to call them that even though the sheath is insulating! They will come in through a plastic gland. Quick CAD sketch (in lieu of my earlier fag packet sketch - felt the thread deserved better):
You have in the cu for want of a better word "fly leads" from the main switch to the individual RCDs in the case of a split load board. These fly leads are single insulated. They're also pretty flexible by construction. Add to that the modern move away from two screw terminations in main switches to one bfo screw (bad move imo). That's a big bone of contention on the basis it's easy for it to loosen. NEVER, EVER think that these screws are factory tightened properly, always check.
So with potentially dodgy single screws and springy, single insulated, flexible conductors, if one comes adrift it could easily touch the metal enclosure...and that's BAD! Wasn't a huge issue before when a TT enclosure had to be made from "plastic".
So the way round this, in a TT is to fit an upfront TD RCD in place of the normal DP main switch. Wire comes adrift, touches the case, TD RCD trips. The 100mA rating is there to give fault protection for the single insulated cables that feed the 30mA RCDs. If you put a normal, no time delay 100mA RCD in front of the line of 30mA RCDs it doesn't guarantee it won't trip first hence the time delay element. Only a TD 100mA provides proper discrimination.
There's a lot of confusion over time delay but think of it like this: A fault occurs on an outgoing circuit in excess of 100mA you would think that it should trip the 30mA RCD first if that's before a 100mA RCD. But normal BS EN RCDs (& RCBOs) only have to trip within 300ms. The issue is that the 100mA RCD if non time delay may react quicker than the 30mA RCD. Making the 100mA RCD time delay gives the 30mA a chance to trip first. Hope that makes sense.
If you look at the SBS Compact assembly the bus bars take the place of the single insulated fly leads. They're affixed at multiple points and can't really come undone and spring against the enclosure.