Sounds like something for section 5.5 of appendix 4 of BS 7671. You probably need a few more details (like how well loads are balanced across phases) to get a complete answer, but as I read it an ordinary  full size N (i.e. 100% of L) will cope with up to 15% distortion.

  - Andy.

Without further information on the % of each harmonic, you could only assume that most of the distortion will be triple-n harmonics.

Table 4Aa of BS 7671:2018+A2:2022 indicates that the neutral can be sized simply for (line current x 1.0) with 3rd harmonic content of 5 %. 

The above advise is current guidance for UK installations.

For triple-n harmonics exceeding 33 %, the factor is applied to a calculated neutral current, not the line current.

Example calculations are given in 5.5.2 in Appendix 4 to BS 7671:2018+A2:2022.

thank you for your response Sir, so from BS 7671 I can see it uses "assumptions" of wether the load MAY or MAY NOT produce harmonics. Does this mean I just have to "predict" (for lack of a better word) if a load would produce harmonic distortions?

The manufacturer should provide information on harmonic distortion of equipment one way or another, especially in the UK or the EU. This might be:

• For relevant equipment, conform to the relevant conducted harmonic emission standard (you will then be able to assume the harmonic distortion does not exceed the limits of the standard); or
• For equipment where there is no standard, or the use of a standard is optional, the actual harmonic emissions from type tests.

In the UK, and in a number of EU countries, there are limits for total combined harmonic distortion of installations - in the UK the installation designer must demonstrate conformity to ENA Engineering Recommendation G5 (current version I believe is G5/5)

I did wonder why this has not been moved to 'Wiring and the Regulations' as it's BS 7671-related ?

Interestingly, the design software I use automatically provides a Total Harmonic Distortion percentage based on the type of load, which can be quite helpful.

is your load 3-phase 3-wire, like most motor drives, or a collection of single phase connections? if the disturbing load isn't connected to the neutral, then it can't draw any current through the neutral 

if the three phase 3-wire load has one of it`s lines broken, I think it would create an unbalanced load in the neutral even though ithe neutral isn`t connected to it

Collins Ezenyinwa said:

if the three phase 3-wire load has one of it`s lines broken, I think it would create an unbalanced load in the neutral even though ithe neutral isn`t connected to it

Think again - there can't be current in the N if there's no N conductor. Loads can either be connected directly between lines, in which case there's just current flowing in two lines if one is broken - potentially with two of the loads connected in series between the two remaining lines. If loads are connected in star without a N connection, then the voltage on the star point will move away from 0V until things equalize (similarly some loads connected in series across lines).

  - Andy.

AJJewsbury said:

Think again - there can't be current in the N if there's no N conductor.

That's an interesting perspective, but I'm not sure that's always the case either .... if there are single-phase 400 V loads (or broken three-phase 400 V delta loads) connected alongside fully balanced single-phase 230 V loads, and 400 V TP&N loads, in a three-phase 4-wire system, harmonics can still travel in the 4-wire neutral upstream of a 3-wire circuit, and unbalance even in only delta-connected or line-line connected loads could cause neutral current based on volt-drop causing unbalance in the star-connected and single-phase loads.