Guidance on Long-Run DC Cable Design for Lighting

Any special considerations when selecting cables or performing cable calculations for DC circuits

Not really. AC calculations are done on the basis of "r.m.s." values - i.e. the d.c. equivalent - so Ohms Law and related resistance based calculations should be just the same. (For larger cables inductive/reactive effects will be rather different of course, but for smaller cables typically only the resistive component is considered anyway).

Relevant standards or documents I should refer to.

If it's in the UK or simialr, BS 7671 would be a good starting point I would have thought.

Use of overload and RCD protection

This bit is tricker. Fuses and MCBs should properly be d.c. rated, as breaking d.c. arcs is much more difficult than for the equivalent a.c. situation. RCDs I suspect will be very problematic - I don't know of any available ones that work on a d.c. supply. B-type ones for instance are usually good for spotting d.c. residual currents, but sill need an a.c. supply as far as I know.

   - Andy.

Thank you very much for your support.
Yes, this project is based in the UK.

Regarding RCD protection : do you think it would be advisable to install an RCD upstream of the AC/DC converter (i.e., the power supply unit)? I don’t want to create any issues by doing so, but I’d like to confirm whether such an RCD would actually be effective in detecting and protecting against leakage currents on the DC side of the circuit.

Again, I greatly appreciate your support. 

 do you think it would be advisable to install an RCD upstream of the AC/DC converter (i.e., the power supply unit)?

Depends on how the unit works internally and how the d.c. side is referenced to Earth (or not), but if it's the usual design where the DC side is separated from the AC side then the RCD will likely provide no benefit at all, as faults on the d.c. side won't result in a L-N imbalance on the a.c. side.

You might want to take a step back here - 500V d.c. is a pretty unusual system (and there might be good reasons for that). Have a think about how to provide shock protection - ADS is only one option of several and might be fairly awkward to achieve (e.g. for 500V you'd need max 0.1s disconnection time and without RCDs you'd conventionally have to rely on overcurrent protective devices - and so to achieve 0.1s you'll require very substantial prospective fault currents - which in turn may be difficult to guarantee from a PSU arrangement unless it's massively oversized and you'll likely need pretty thick cables too to keep the loop impedances down). 

I guess there's more to this project than meets the eye - presumably there's some reason for preferring d.c. over conventional 230/400V a.c. systems. SELV/PELV is more conventional for d.c. systems - but extra low voltages tend to have much more onerous voltage drops so are likely to be impractical for circuit lengths of many tens let alone hundreds of metres. Double/reinforced insulation might be a better bet, but does need some extra care and on-going supervision (which may or may not be suitable for the situation you have in mind). If you are sticking with ADS, lowing the voltage a little might help - 400V or below and you can have a 0.4s disconnection time, and 1s for 230V or below - which should make things rather easier. If you can arrange the supply to be centre-tap earthed - e.g. so it's 115-0-115 rather than 0-230 you can have the same voltage available for the loads, but half the voltage available for shocks (to Earth) (below 120V you don't have to disconnect at all).

   - Andy.

 do you think it would be advisable to install an RCD upstream of the AC/DC converter (i.e., the power supply unit)?

Depends on how the unit works internally and how the d.c. side is referenced to Earth (or not), but if it's the usual design where the DC side is separated from the AC side then the RCD will likely provide no benefit at all, as faults on the d.c. side won't result in a L-N imbalance on the a.c. side.

You might want to take a step back here - 500V d.c. is a pretty unusual system (and there might be good reasons for that). Have a think about how to provide shock protection - ADS is only one option of several and might be fairly awkward to achieve (e.g. for 500V you'd need max 0.1s disconnection time and without RCDs you'd conventionally have to rely on overcurrent protective devices - and so to achieve 0.1s you'll require very substantial prospective fault currents - which in turn may be difficult to guarantee from a PSU arrangement unless it's massively oversized and you'll likely need pretty thick cables too to keep the loop impedances down). 

I guess there's more to this project than meets the eye - presumably there's some reason for preferring d.c. over conventional 230/400V a.c. systems. SELV/PELV is more conventional for d.c. systems - but extra low voltages tend to have much more onerous voltage drops so are likely to be impractical for circuit lengths of many tens let alone hundreds of metres. Double/reinforced insulation might be a better bet, but does need some extra care and on-going supervision (which may or may not be suitable for the situation you have in mind). If you are sticking with ADS, lowing the voltage a little might help - 400V or below and you can have a 0.4s disconnection time, and 1s for 230V or below - which should make things rather easier. If you can arrange the supply to be centre-tap earthed - e.g. so it's 115-0-115 rather than 0-230 you can have the same voltage available for the loads, but half the voltage available for shocks (to Earth) (below 120V you don't have to disconnect at all).

   - Andy.