Getting the most out of multiphase

[benjamindees]

For most, it seems the reason to use multiphase is to smooth out the DC going to the batteries once rectified.  For this reason, many designs wire their coils for 3 (and some up to 5) phases.  This smooth output helps batteries last longer.

However, transmitting multiple phases ineveitably involves more wires, more losses, or less flexibility.  You must either put your batteries next to your wind generator, transmit DC and put up with losses, or run enough wires for each of the phases you generate.

What is the reason others, like power companies, prefer multiphase?  They are not charging batteries.  They don't want to run more wires than they have to.  Power companies typically generate and transmit no more than 3 phases.

I submit that the real benefits of multiphase are in alternator design and efficiency.  Furthermore, I submit that most designs fail to take advantage of this efficiency.

Here is my understanding of the benefit of 3-phase design.  With a single phase, you have a coil of wire passing a magnet.  One leg of the coil produces a positive voltage.  The other, negative.  This produces a full sine wave, or one phase.  However, with the physical constraints of this design, there must be space within the coil of wire that is basically wasted.  No current is produced as this empty space passes the magnet.

Here's where other phases come in.  When we instead take two coils, placing the leg of one within the empty space of the other (and vice-versa), we have effectively increased the efficiency of our design.  We now generate twice as much power in only 33% more space, with only an investment in more copper (not magnets).  This is a 2 phase design.

Is there reason to go beyond 2 phases?  Perhaps to help satisfy the original reason for using multiphase in axial flux alternators, to smooth the output to the batteries.  There may also be geometric reasons for adding more phases to the arrangement.

Thoughts?  Additions?  Criticisms?

[scoraigwind]

I use multiple phases in alternators mainly because they run smoothly without vibration.  They also make more efficient use of the stator by allowing more copper in there as you have noted.  With single phase there are only certain positions the copper can go but with multiphase it can go anywhere you can fit it and you are only limited by the ability to turn around at the ends.

I don't worry about the battery ripple current particularly but I do like to keep loss in the cables down.  The most efficient way to do this is to use DC (assuming you are not using transformers and higher voltage).  Using AC transmission is wasteful because the current varies and the high currents lose a lot more than the low ones (square of current) so overall you lose more with AC than with DC.  The rms is going to be higher.

On the grid they use 3 phase for similar reasons.  More efficient alternators and also better motor starting.  Single phase can be taken off at a lower voltage.  So the 3 phase is more efficient than single phase.  However we do not see this benefit on DC systems, because the DC will win out.  3 phase feeding a rectifier is always less efficient (cable loss wise) than DC.

5-phase is really just a gimmick, but it seems to me that it is often simpler to have a lot of coils in parallel in a low voltage stator and I prefer to use separate rectifiers to prevent parasytic currents so they might as well be a lot of different phases as only 3.  You could have 11 phases no problem.  But keep the rectifier near the stator.

[Flux]

Yes I basically agree with most of this except the bit about the batteries liking the dc as smooth as possible.

With the exception of gel batteries which object to ripple when on float voltage, I don't think the battery is bothered, in fact there is some evidence that a pulsating current may be helpful.

You are right that the advantage lies in alternator efficiency. For normal ac distribution a single phase alternator needs to be 50% bigger than a polyphase.

When you charge batteries the single phase comes out even worse as current only flows when the emf is above the battery voltage and there are periods where it is doing nothing. This lowers efficiency and also results in a seriously pulsating torque that shakes the whole machine.

Two phases or more produces a reasonably smooth dc and conduction never stops and the thing runs smoothly.

For power distribution 3 phase is the logical choice as the vector sum of the line currents is zero, a balanced load only needs 3 wires that carry equal currents and a neutral if used is only carrying any un balance in the load.

If you rectify, I don't think 2 phase would be significantly worse than 3, I am hoping that Windstuff Ed and some others will come up with some results.

2 phase does not make best use of the rectifier but I doubt that it will be an issue.

Different phase numbers commutate differently in the rectifier, the peak dc voltage is 1.4 x the ac voltage feeding the conducting pair of diodes.

This ac voltage is 1.73 for 3 phase 1.9 for 5 phase and 2 for 6 phase as the two phase voltages are directly opposite.

As you increase load into a battery these all seem to tend to the same current at high values, so the 6 phase has the lowest cut in volts followed

 by 5 and then 3 phase. The 6 phase does give a better match to the propeller than 3 phase. The 3 phase case gives a slightly better overall efficiency but this could easily be more than compensated for by better prop matching.

[Flux]

Hugh's post came up when I was posting mine.

I must stress that what I said about 3 phase and cables was refering to ac distribution. If you use dc then it is more efficient to use cables to transmit the dc than the 3 phase ac.

Flux

[bluGill]

The utilities go 3 phase because motors like it.   A 3 phase motor doesn't need complex start up circuits, and can be easily reversed by just changing the order of then phase.    Homes don't need this advantage enough to make it worth running 3 phase, but big industry does.

3 phase is also great when you don't want to increase the voltage, or wire sizes, yet you still need to keep the power down.   run 3 phase to a neighborhood, and tap each house off a different phase - good load balancing, and should something happen to one phase you just cross that phase to a different one (with much greater line losses) until you can get it fixed.    (Course this is bad for industrial customers on the same line)  This only comes into play because industry makes them generate 3 phase anyway.

The utilities do not transmit power in 3 phases.   Long distance transmission (state to state) uses DC at very high voltages.   There is a maximum voltage that can placed on a wire (I'm not sure why, something about air ionizing), and DC lets them get closer.    Then near the city they are sending power to they have a big inverter that turns the power from one wire in to 7 different high tension AC lines to substations, each with 6 wires on them!  (I'm thinking of a particular inverter not far from my house that has the above numbers)

The short answer is because motors work best on poly-phase, and 3 is the smallest number of phases that gives the advantage of poly-phase.

[laskey]

I'm afraid you're not entirely correct.  Power companies do transmit via three phases... most places in fact, BUT they only do it when there is a reasonable distance to be covered knowing that there will be tonnes of transformer substations along the way to feed cities and towns and the like.  This is because the transmission losses are very low at 250,000 volts AC, and they don't need as much copper as DC because at high ac levels the current tends to flow at the surface of the cable (hysteresis effect) so they can make the cable core steel cable rope, which gives the cable very high breaking strenghts.  

They transmit DC over "very" large distances where there isn't anybody at all who needs a feed in between.  For example between islands, or across the state of north dakota.  They only need one wire, because they use ground return, but they need a very, expensive inverter installation at the reception end.

I've never read anything about low voltage DC ground return... I wonder it it would work?  Only feed the positive voltage through the wires, and bury a fairly large copper plate in the ground at each end of the wire, and connecting the negative side to that.  I wonder if it would work as well as in a high voltage DC system?

[ghurd]

The Hoover dam transmitted through a hollow copper cable.

They may have changed over to something else now.

I bought a piece of the original cable, interesting,

kind of made of 10 pieces of 'tonge and groove'

or 'ball and socket' copper,

formed into a 'pipe'.

Fit is good but not too tight,

they can be seperated and reassembled.

'The guy' said they ran 3-phase,

but I'm not sure he understood the question.

G-

[nickelbender]

Some large industrial power cables are hollow this is to make better use of the copper do to the effects of skin effect.

DC power transmission is nice because it allows faster power control and lower Ir losses, but they require complicated HV electronics to manage.

Long AC transmission lines are compensated with series capacitors and inductors. There is a good "Shaum outline" from McGraw-Hill that explains this along with the calculations required to compensate a power line.

6 phase is also commonly used industry in places like copper refineries to reduce riple in the DC for the tanks but this is done using a transformer connnection to go 3 to 6.