Load end RCBO's, upstream BS88-2 in combination with RCD (30mA)

Two tiers of 30mA RCDs sounds distinctly odd (if I've understood what you're proposing correctly). Normally the upstream RCD would be at least 100mA or 300mA time-delayed (S-type) (and design to avoid the need for additional protection by 30mA RCD for the distribution circuits - e.g. choose wiring systems with a concentric c.p.c. rather than soft sheathed cables for routes embedded in walls etc.) For decent earth fault currents (e.g. >5x IΔn) you should get disconnection within 0.15s which is usually fast enough to protect the cables for currents too low for the fuse.

   - Andy,

Load end RCBO's, upstream BS88-2 in combination with RCD (30mA)

Two tiers of 30mA RCDs sounds distinctly odd (if I've understood what you're proposing correctly). Normally the upstream RCD would be at least 100mA or 300mA time-delayed (S-type) (and design to avoid the need for additional protection by 30mA RCD for the distribution circuits - e.g. choose wiring systems with a concentric c.p.c. rather than soft sheathed cables for routes embedded in walls etc.) For decent earth fault currents (e.g. >5x IΔn) you should get disconnection within 0.15s which is usually fast enough to protect the cables for currents too low for the fuse.

   - Andy,

Thanks Andy, This problem was the BS88's don't trigger with a Zs =20 ohms. Therefore the upstream solution incorporates 300mA time delayed (S type)  RCD's. By removing the BS88's this has corrected the design. Many thanks for you help