Three phase neutral

[studentspark]

Three phase and neutral supply

Makes my mind boggle a bit.

any guidance appreciated

In a three phase system, the return path is the phase difference?

In a three phase and neutral the return path is the phase difference, and the neutral is there if you need a single phase load?

Any current in the neutral on a three phase and neutral is an imbalance between the phases?

So if I put a clamp on the neutral on a 3P and N. Any current on the neutral would show an imbalanced load?

But what is an imbalance load, each phase must have different demands on it depending on what is being used.

few scenarios..

If i had three kitchens, each on a phase, say 100A supply each

1. Two kitchens in use, and drawing 60 A
2. Three kitchen in use and drawing 60A each
3. One kitchen drawing 105A
4. kitchen 1, 50A kitchen 2, 90A Kitchen 3, 10A

What current would be expected on the neutral with these situations

Many thanks

 

[AdrianUK]

Hows your vector maths??

In a 3 phase (no neutral system) the current delivered on one phase must leave on the other phases. That doesn't mean the phase currents must equal. One can still have unbalanced loads but they are powered line-line (ie. at 400V in the UK) not line-neutral.

For a 3 phase & neutral system, the vector sum of the phase currents & neutral must equal zero.

Assuming a completely resistive load, a balanced three phase system would pull exactly the same current on each phase and the current in the neutral would be zero. Indeed, if the system were balanced, the neutral could be disconnected & nothing would happen (DO NOT TRY THIS ON A REAL SYSTEM! - it will NOT be purely resistive)

An unbalanced system has different currents flowing in each phase & the "unbalance" current flows in the neutral.

Most real installations are unbalanced. The most common balanced device is a three phase motor (lets ignore delta loads for the moment!)

If the phase currents are L1, L2 & L3 respectively, then the neutral current is give by N = sqrt(L1^2 + L2^2 + L3^3 - L1*L2 - L2*L3 - L3*L1).

For your scenarios ....

(Assuming these kitchens are purely resistive loads... which they won't be!)....

1). L1 = 60A, L2=60A, L3= 0A, N = 60A
2). L1=60A, L2=60A, L3=60A, N=0A
3). L1=105A, L2=0A, L3=0A, N=105A ...... Although one could argue, on a 100A supply... L1 = blown fuse!
4). L1=50A, L2=90A, L3=10A, N=70A

 

[studentspark]

Thank you for taking the time to answer my questions, really appreciated.

So, in real life situations, current on the neutral is to be expected.
I will play around with the formulas you gave , so I can work out your answers

 

[AdrianUK]

Yes, most, if not all real installations will have a neutral current.

As a "rule of thumb" the neutral current won't exceed the highest phase current. That was true in days gone by, but today, because of none linear loads and harmonics it is possible for the neutral current to be bigger

 

[studentspark]

Thanks again

With just one neutral , I was wondering about overloading the neutral.
But the neutral is not the (only) return path, as such ,the phase difference enables the return.
And the neutral current will not exceed the phase current, as you say ( with caveat)
And if the phase current is so high, the protective device will operate, so dangerous current can't arrive on the neutral?

(sorry for simplistic reasoning)

 

[AdrianUK]

With a purely resistive load, the neutral current cannot exceed the highest phase current, so overloading the neutral is not possible. The worst case would be full load on one phase & no current on the other two, hence the neutral current = phase current. Additional current on the other phases will only lower the neutral current. In previous editions of the wiring regs it was permitted to use this fact to reduce the size of the neutral conductor, often by one standard size, but this is no longer allowed. Don't forget, the neutral does not normally incorporate any form of overload protection - the only exception being 4 pole linked MCBs, but these are not that common.

Modern loads tend to use switch mode power supplies (SMPS) & other phase control techniques. These are known as 'none linear' loads since they do not pull a current over the whole of the mains cycle. This has the effect of generating harmonics currents on the neutral. The problem here is that certain harmonics, particularly the 3rd order, do not cancel out in the neutral - they add. If the installation contains a lot of equipment which uses SMPS (eg computers, LCD monitors, inverter drives etc) then the addition on the neutral can be significant and it is possible for the neutral current to exceed the phase current. This can lead to cooked neutrals in older installations. Some schools of thought are now suggesting that the neutral size should be increased to allow for this, perhaps by one standard size.

 

[studentspark]

Thanks for the very good and clear response

 

[mattylad]

The current on the neutral got me puzzled a few months ago, I was testing a 3ph appliance (all resistive ISTR) and decided to just have a look at what the neutral current was, it puzzled me that the more I turned on the lower the current.
So I did as you have just done, I asked about it, found out and learnt a little more.
Which also explained the one of the reasons we try to balance the loads when wiring a 3ph appliance.

 

[ban-all-sheds]

I will play around with the formulas you gave , so I can work out your answers

The other thing you could do, to visualise it, is to play around in Excel (other spreadsheet programs are reportedly available) with different currents for each phase, just add N=L1+L2+L3 at each point in the cycle, and plot a chart.

e.g. your example #1:

#4:

Note, as I was plotting the curves for each phase I figured I'd use the peak values, but it's all relative, so doesn't really matter for visualising the magnitude of N wrt L1/L2/L3.

 

[plugwash]

Modern loads tend to use switch mode power supplies (SMPS)

Modern high power SMPSUs often extremely have power factor damn near unity. The trick is to add an additional boost converter between the rectifier and the main primary capacitors and then modulate the boost converter so that on short timescales current is proportional to voltage. Many SMPSUs are subject to EU mandates requiring such correction.

Of course older stuff, cheap chinese tat and PSUs that are too small to be subject to the EU mandates are another matter.

 

[SUNRAY]

The other thing you could do, to visualise it, is to play around in Excel (other spreadsheet programs are reportedly available) with different currents for each phase, just add N=L1+L2+L3 at each point in the cycle, and plot a chart.

e.g. your example #1:

View attachment 156923

#4:

View attachment 156924

Note, as I was plotting the curves for each phase I figured I'd use the peak values, but it's all relative, so doesn't really matter for visualising the magnitude of N wrt L1/L2/L3.

Got it wrong BAS your phases are 60° apart, they shoud be 120°