You are quite right - an 800 watt ideal  load at the end of 30m of 1.5mm2 or even 1mm2  cable will be absolutely fine from a voltage drop perspective - indeed you could drop the MCB to 5A as well.

But..

1) Certain types of flood light have surprisingly high inrush (ten times running perhaps) and struggle to 'strike' if the voltage droops too far during that phase.

2) Mechanical strength - 2.5 mm is not too fat to work with easily, but is slighly  less likely to be damaged by pulling in, and has a copper size that fits the sort of terminals that folk  use in outdoor gear rather better.

And no one wants to dig it up to do it twice when asked to add an one extra light or socket.

Mike

Thank you for your asnwer:

1) Certain types of flood light have surprisingly high inrush (ten times running perhaps) and struggle to 'strike' if the voltage droops too far during that phase.

- for this design, I'm using C type breaker rather than B, to solve the inrush current. But I'm not sure whether this is a practical solution. Need to check it again.

2) Mechanical strength - 2.5 mm is not too fat to work with easily, but is slighly  less likely to be damaged by pulling in, and has a copper size that fits the sort of terminals that folk  use in outdoor gear rather better.

-in my humble opinion, the damaged done by cable pulling work is due to poor workmanship. Even if I use 1.5sqmm Cu/PVC cable for 2 nos x T8 LED tube, if the electrician / worker have wrong method / poor workmanship and etc. the cable will still get damaged. 

And no one wants to dig it up to do it twice when asked to add an one extra light or socket.

- this justification is acceptable, I do agree with you, in the event where there is a necessity to add one or more lighting, then my existing 1.5sqmm wont be sufficient anymore.  

Thank you once again for sharing your input. 

the breaker firing  is not usually the biggest inrush problem - or at least if it is, then other things are also probably wrong. But during the inrush, the volt drop is potentially 10 times higher - this does not matter too much if the device can still strike at the reduced voltage of 210V or whatever, and  I suspect a lot of this advice is generally for traditional pulse starter discharge lights and maybe less of a problem with electronic drivers and LEDs.

Mike.

It seems like I get your point (or maybe not...) please let me know if my understanding now is align with your explanation.

In normal operation, both my breaker and cable capacity won't have any problem.

during the starting (inrush current), the C-type breaker which will trip between 5-10 times rated current also will not operate , however the cable (in this case 1.5sqmm, current carrying capacity of 17.5A) won't be able to carry the inrush current and resulting in cable melting/damage etc.

so the voltage drop that I calculated is only meant for normal operation, in practical I have to consider the inrush current as well. is this what you are trying to explain?

Thank you.

however the cable (in this case 1.5sqmm, current carrying capacity of 17.5A) won't be able to carry the inrush current and resulting in cable melting/damage etc.

Nooooo . How long can 1.5mm take  30-40A ? Well probably getting on for an hour, but certainly far longer than the any inrush - cable does not spontaneously combust just because you have passed two or three times the current through  it than its steady state rating. The steady state assumes a temperature rise then needs minutes, perhaps tens of minutes to occur. Keep the  overload short relative to that timescale and let it cool afterwards, and all will be well.

As an example I have jump started a car in the past - drawing probably quarter of a kA of peak cranking current for a few seconds, - using the cores of a length of 2.5mm2 , & yes it was a bit hot afterwards, but it was fine for the few seconds needed and apart from the fact I had no use for a 6 foot length with the ends where contact had been made slightly cooked, the middle of it would have been perfectly serviceable.

(Rather amusingly much to the horror of the alarm installer whose car battery was serving as host- he clearly knew enough about current ratings to be very worried, but not enough about thermal inertia to be reassured.)

The problem is that the volt drop during inrush can be so much that a discharge type lamp flickers and pops but never actually strikes..

Mike

however the cable (in this case 1.5sqmm, current carrying capacity of 17.5A) won't be able to carry the inrush current and resulting in cable melting/damage etc.

Nooooo . How long can 1.5mm take  30-40A ? Well probably getting on for an hour, but certainly far longer than the any inrush - cable does not spontaneously combust just because you have passed two or three times the current through  it than its steady state rating. The steady state assumes a temperature rise then needs minutes, perhaps tens of minutes to occur. Keep the  overload short relative to that timescale and let it cool afterwards, and all will be well.

As an example I have jump started a car in the past - drawing probably quarter of a kA of peak cranking current for a few seconds, - using the cores of a length of 2.5mm2 , & yes it was a bit hot afterwards, but it was fine for the few seconds needed and apart from the fact I had no use for a 6 foot length with the ends where contact had been made slightly cooked, the middle of it would have been perfectly serviceable.

(Rather amusingly much to the horror of the alarm installer whose car battery was serving as host- he clearly knew enough about current ratings to be very worried, but not enough about thermal inertia to be reassured.)

The problem is that the volt drop during inrush can be so much that a discharge type lamp flickers and pops but never actually strikes..

Mike

cable does not spontaneously combust just because you have passed two or three times the current through  it than its steady state rating

The problem is that the volt drop during inrush can be so much that a discharge type lamp flickers and pops but never actually strikes..

okay, i got it.