33

Why does power transmission use three lines with three different phases? Why not three lines all in the same phase? Does it have to do with the alternators used for generating the power, or is there less loss when the phases of the three lines are all different?

My question is somewhat the reverse of "Why three-phase power? Why not a higher number of phases?" (cf. "Why is three-phase offset by 120 degrees?").

Geremia
  • 471
  • 1
  • 5
  • 12
  • 6
    Great question! I tried explaining this to two kids on a long car journey (and without taking my hands off the steering wheel). I think I made myself understood. One of them became an electrical engineer, anyway. – Mick Sep 01 '16 at 23:35

3 Answers3

30

Why not three lines all in the same phase?

  1. Because then there is no return path.
  2. Because single phase has no "rotation". Three phase makes it very simple to make a rotating motor with phase sequence determining the direction of rotation. Swap two phases and the direction is reversed.

Is there less loss when the phases of the three lines are all different?

  1. Three phase power distribution requires less copper or aluminium for transferring the same amount of power as compared to single phase power.
  2. The size of a three phase motor is smaller than that of a single phase motor of the same rating.
  3. Three phase motors are self starting as they can produce a rotating magnetic field. The single phase motor requires a special starting winding as it produces only a pulsating magnetic field.
  4. In single phase motors, the power transferred in motors is a function of the instantaneous power which is constantly varying. In three-phase the instantaneous power is constant.
  5. Single phase motors are more prone to vibrations. In three phase motors, however, the power transferred is uniform through out the cycle and hence vibrations are greatly reduced.
  6. Three phase motors have better power factor regulation.
  7. Three phase enables efficient DC rectification with low ripple.

enter image description here

Figure 1. Resultant DC from three-phase rectifier.

  1. Generators also benefit by presenting a constant mechanical load through the full revolution, thus maximising power and also minimising vibration.
Transistor
  • 168,990
  • 12
  • 186
  • 385
  • "_Three phase power distribution requires less copper or aluminium for transferring the same amount of power as compared to single phase power._" Wouldn't this also be true if 1 phase were split into 3 lines? – Geremia Sep 01 '16 at 17:55
  • Where is your return line in that scenario? – Transistor Sep 01 '16 at 18:11
  • @Transister: What is the "return line" with 3 phase? – Geremia Sep 01 '16 at 19:39
  • 3
    Look at my Figure 1. Current from each phase returns to the generator through the other two. This is basic 3-phase theory. You've missed something on your journey to this point. – Transistor Sep 01 '16 at 19:49
  • 2
    Excellent explanation. May I add one point. Three phase motors need no brushes and are therefore more reliable and durable. – Uwe Sep 02 '16 at 07:48
  • @Transistor I come from a physics background and have no formal training in EE. What does the corresponding _I_ vs. _t_ diagram look like for each of the three phases of your Figure 1? – Geremia Sep 02 '16 at 19:34
  • @Geremia: That's a different question and it would take an effort for me to cover it well. I can't find a good diagram on an image search other than [this](http://www.motioncontroltips.com/how-to-handle-imbalances-from-light-loads-on-vfds/). The waveform is right but is complicated (as they explain) because the voltages they're showing are phase-phase rather than phase-neutral. Sketch it out yourself and see if you can figure out what's happening. – Transistor Sep 02 '16 at 19:49
  • @Transistor I've open [a new question](https://electronics.stackexchange.com/q/255768/72280). Let me know if it's stated properly. – Geremia Sep 02 '16 at 20:11
  • @ZekeC. your edit into my point 1 (neutral return) is not correct. Most three phase systems don't use a neutral. The return path is, depending on the phase balance, through the other two phases. Others have rejected the edit for me. Thanks. – Transistor Dec 18 '20 at 14:10
10

Good answer from @Transistor. To add a bit more: -

Three phase is inherently current and voltage balanced from a standpoint of it generating interference. At any one point in time (and a reasonably balanced load) the magnetic emission is low because all the magnetic fields cancel out due to the currents being balanced.

There is net voltage balance in the close far-field - important for reducing EMI. This is not true of a single phase and return wire because the net AC voltage field seen in the close far-field is half the AC field at the live terminal. This can generate EMI.

Clearly you can make an argument to say that under imbalanced conditions there will be a net magnetic field but, to counter this, on a large high-power transmission line, the imbalance will normally be only a few percent maximum: -

enter image description here

So, for a balanced 30 A load (per phase), due to the 120 degrees balance the net sum of the individual three current phasors is zero.

Another benefit is that when converting to DC, 3 phase produces a much lower ripple voltage due to the fact that there are always two diodes conducting: -

enter image description here

Andy aka
  • 434,556
  • 28
  • 351
  • 777
5

I'll focus my answer on transmission alone, without explaining why 3 phase is useful in general because other answers did that.

Transmission of power is a compromise. A compromise between transmission efficiency and ease of conversion. The most efficient way of transmitting electric power is DC. This is why most superlong lines are HVDC (high voltage direct current). However, DC is the worst for converting it to HV when you want to send it from power station, and back to LV when you want to feed it to consumers.

AC on the other hand is very convenient to convert - just put a transformer. However the transmission sucks. Eg. AC radiates some of the energy away, but that's not the main concern. If you look at sinusoidal graph, you'll realize that AC wire doesn't actually work 100% of the time. While DC cable carries useful current all the time (one can think of DC as 100% duty cycle PWM), AC cable carries current only part of the time. This means that for the same peak voltage (which dictates cost of insulating the line) and for the same peak current (which dictates size and cost of conductors), AC can transmit only part of the power.

Here comes the idea of multi-phase. Of course multi-phase alone doesn't mean a thing, you can have 3 phases on 6 conductors (3 pairs completely independent of each other). The key here is sharing of the wires between phases. It's like a hot bunk on a warship - 2 seamen share 1 bunk, when one guy awakes and starts his shift, the other ends his shift and goes to sleep. The point is to not have an empty bunk just wasting space, and 3-phase AC works on the same concept: in the time when one phase "rests", another phase is re-using one of it's wires to transmit own current. It's not clear at first sight because it's very fluid, one falling towards 0 while the others rise, and there never is a time when one phase as a wire all to itself. But the point is to re-use the idle time of the wires.

Why 3? Because 2 is too small, you can't have 2 phases on 2 wires. 3 is the minimum number of phases that can share all the wires. Why offset? Because one phase on X conductors is same thing as 1 conductor X times thicker.

When you compare the 3 phase system to a 1 phase system, you can clearly see that with adding just 50% more wires you get 3 times more current.

3-phase transmission uses the wires TWICE as effectively as 1-phase. So you can use half as much copper when building the line.

Agent_L
  • 1,725
  • 10
  • 14
  • I upvoted for the concept but this answer could be more useful if you was more concise. – Antonio Sep 02 '16 at 09:03
  • @Antonio Well, I kept it to minimum, but the background leading to the reasoning is the point here. Answers just saying it's more efficient without explaining why and how has already been posted here. – Agent_L Sep 02 '16 at 09:23
  • Sorry but your bottom line statement is just wrong. Single phase systems don't have to be unipolar. – Peter Green Sep 02 '16 at 16:06
  • @PeterGreen I don't understand. As far as I know, single phase transmission lines are always bipolar. Only some HVDC are unipolar, using ground as return. – Agent_L Sep 05 '16 at 13:11