induction motor generator puzzle

[mab]

Hi all,

my hydro system has been running for about 3 weeks now and I've been getting about 240w from it. it uses a 2Hp 3phase induction motor with 3 100uf caps in the c-2c configuration to give single phase output.

I've been using a variac to adjust the (resistive) load, and by comparing the ammeter and voltmeter to get max power.

Trouble is I'm still learning how it behaves, and in trying to maximise the output I've noticed something odd:

initially I was getting about 160w (1.6a 100v) at about 27Hz, but suddenly it changed  - the freq dropped to about 24Hz and the volts went up, but when I adjusted the variac for max output I ended up with about 2.4a at 100v. I didn't know why it had changed but obviously I was happy to be getting 240w rather than 160w.

anyway, it's been quite dry lately and low water together with fallen leaves blocking the strainer resulted in air in the penstock (and a small drop in output - 23hz, 1.8a, 100v).

So today, although the water level is still marginal I stopped the genny to allow the air to rise out of the penstock, then restared.

trouble is it's now running 27hz, 1.6a 100v again - i.e. worse than with air in the system and I can't think why it should run at higher frequency but lower power  .

I was hoping one of the (relative) experts on here might tell me what I'm doing wrong. In the meantime I'm going to make a smaller nozzle in case the weather stays dry...

[Flux]

This could be a bit complicated.

What is the motor rated at, it sems as though you are running it about half nominal speed so I assume it should be a 230v ( ish) motor on its rating plate.

For the turbine you need it to run at half no load speed. If you have a tachometer you could run it unloaded ( disconnect capacitors)  and check that the turbine speed virtually doubles.

I am not sure whether you are matching the working speed to half no load speed or whether your variac is optimising the motor for minimum iron loss. The two conditions may be quite significantly different.

You may need to change capacitors to get the best excitation while dragging the turbine down to half speed.

If it was dc or a synchronous alternator it would be easy as you only have one variable but you have 2 variables to conternd with.

You may in fact be able to choose capacitors roughly and then fine tune things by using the variac to alter the effective reactance of one of the capacitors. I think you can depart a bit from the C  2C relationship to get a better match. You also need the capacitors on the correct phases to suit the direction of rotation but as it is working I suspect you have this right.

Flux

[mab]

Hi Flux,

Yes it's 240v/50hz nominal.

I don't have a tacho' so I can only guess by sound that I'm running well below no-load speed - though I guess that as I got more power at lower Hz that may be where it should be running.

I do have two DPDT switches between the caps and the motor though - one switches the phases so that the 2C is either leading or following the C (output) phase. I can switch between them and stick with the one that gives a better output - although the higher output setting once found doesn't change.

The other switches the 3rd cap between parallel with one cap or series with the other (100/200uF or 50/100uF). I consistently get more power (and lower Hz) with the 100/200 setting.

All the above variation was with the 100/200 setting and the same lead/follow setting - though I did try changing them just to see if I could get it back to it's 24Hz mode, but no luck.

I can't figure how to get back to the 24hz higher power mode - mostly because I don't know what changed to make it go into that mode before. 

[hydrosun]

if you have a pressure guage you can use that to make sure the pipe is always full. How do you have the nozzle aimed? If that isn't secure it can change output. After checking that the input power is the same check wiring connections. Sounds like you have switches and other varables that might have intermittant changes.
I haven't played with induction motors for generators so I can't advise you on that. I don't know if capacitors can change ( try different caps?) Or if your motor has some interior short happening.  But i know that connections can change even when tight. I have one now that heated up and I took it apart and recrimped probably had corrosion inside the crimp.
That's all the varibles that I can think of to check.
Chris

[oztules]

Hmm interesting
I can only suspect the capacitors are changing their value, or your switch is suspect.

If we have a lower rpm, and higher emf and current, we know that the field is stronger.
One thing with electric motors is that torque is proportional to current which is proportional to amp turns in the windings.

The windings are fixed in the stator, and the secondary winding in the rotor is 1 turn.. so hard to see a change.

The amp turns in the stator induce a current in the rotor, that in turn gives us the magnetising field to generate the emf in the stator..and so proportional to the torque.... bit chicken and egg really....

If we have more current in the stator, I can only see that it comes from more reactance from the caps to supply more magnetizing current if it is at lower rpm.

So I am looking for some way to get more magnetizing current... without increasing rpm... in fact dropping it..... weird caps?..... ...... what else to blame to get more magnetizing current....

Maybe another way would be to get more slip..... but your controlling the output current and so the slip as well... ie if you loaded it up more, the slip would increase. Normally this would be done by trying to drive the rotor faster, and getting more current from the stator. It is an asynchronous motor, and so an asynchronous generator too. The frequency of the rotor does not necessarily reflect the frequency of the output... which is why we can use it for grid tie with varying speeds of the rotor, but fixed frequency in the stator... as a free running unit without the grid for magnetizing current, this may not apply... never tested it like that.... so I am left floundering at the moment.

Possible loose wire in the rotor???.... has happened but very infrequently...generally well embedded in the ring........a  once in a lifetime for me....may see some sign of burning/heatspot on the aluminum ring....


Just my thoughts on it... maybe it will trigger some thing somewhere with you or someone else's thoughts.


Yeah.... I got nuthin I guess


...................oztules

[Flux]

Try to find some method of measuring the no load speed of the turbine. It may well be that the correct operating frequency is under the figure you are getting but you haven't enough capacitance to excite the generator at the lower speed.

At less than half nominal frequency the capacitors will be rather on the big side anyway. Depends whether the motor is star or delta but 100/ 200uf may not be enough.

Having established the correct capacitor sequence for your direction of rotation you can then remove that switch to eliminate one possible variable.

Things can be faulty, capacitors can change but they almost always loose capacity and don't recover. Unless it is an old and abused motor I would not expect trouble there, motors can develop cage faults but it is not a common problem. Any other winding fault will liberate smoke sooner or later.

I think it is instability in the excitation or actual changes in the turbine. Chris mentioned the need for secure fixing of the nozzle and if the position is not dead on changes in flow may affect the way the jet impacts on the buckets, if you have air and cavitation this may be uncertain.

Try to measure that no load speed then you will know the frequency you ought to be loading to. Get the turbine on the peak of its power curve then you may be able to fine tune the excitation to reduce iron loss as far as possible but the system basically relies on saturation for stability, the less you make the iron loss the more unstable it becomes.

It's a cheap simple and reliable generator but not a particularly efficient arrangement and works best when there is plenty of water and you can afford to run the core saturated. For marginal water supplies there are better forms of generator.

Flux

[mab]

Thanks for the replies folks!

As there was a common theme of bad connections re- the excitation caps I had another jiggle of the switches (always suspect in a humid environment) but that didn't cure it.

Decided to look a bit deeper and found that I'd managed to screw a terminal down onto insulation on one of the capacitor connections (Doh!) and that the copper end could have been intermittently making contact. It was where two caps were connected (series) and the bad connection went to the motor (via the switches).

The good new is I'm back up in the 240w ball-park, though now at about 28Hz - I'm not too worried about Hz as long as I'm getting the watts (and in fact 28Hz gives me more wiggle room on the transformers) so I'm happy again  though I'm still wondring if I should be looking for another dodgy connection.

I did try wiggling the nozzle to see if the alignment was causing trouble, but it seemed OK.

I don't think I'm going to get more power without more rain though - the penstock's gurgling again so I've obviously got air in it again  . I haven't fitted the smaller nozzle yet as I decided to make some repairs to the washed-out ford and If I reduce the water-flow through the penstock, my temporary chute may overflow and wash away the new concrete.

...It's a cheap simple and reliable generator but not a particularly efficient arrangement and works best when there is plenty of water and you can afford to run the core saturated. For marginal water supplies there are better forms of generator.

I would like to try a PMG - possibly a small, high-rpm axial-flux - at some point but it's a question of finding the time, and it was a choice of either this genny for this winter or nothing until next winter - and 240w 24/7 is cracking-good compared to what I'm getting from the PV right now (had about a 3 weeks of high-pressure grey overcast here recently).

cheers for the help folks.

[Flux]

Glad you got it sorted. 240W continuous is indeed extremely useful.

Flux

[mab]

puzzle solved - I did have a 2nd bad capacitor connection (it's a wonder it worked at all) - now running 24Hz again and pushing 320w*  - I've started leaving lights on willy-nilly just because I can.

Actually all my wattage estimates were a bit high - I'm using variacs to adjust the loads ATM and I think they introduce a <1 powerfactor, so my 320va on the meters is actually nearer the aforementioned 240w in real power.

[Flux]

At 24 Hz you need a 230v variac to handle 110v. They are closed ring core or strip wound and go into saturation rather sharply. If you are using a 110v variac the peak magnetising current will be high. It will cause a power factor issue from the distorted waveform as well as the lagging magnetising current.

Your capacitors will correct the magnetising current but can do nothing about the severe 3rd harmonic waveform.

Flux

[mab]

Don't worry - they're rated 250v 47-400Hz  (they're oldies); also, my max-power-point seems now to be at about 80-90v.

Actually, I was just wondering if I could get away with putting the motor back to star (currently wired delta) to reduce transmission losses:-  1.73*80= 138v (would need to switch to 50/100uF) and in theory that would still be 24Hz I think, so:

138/24= 5.76v/hz
250/47= 5.32v/hz

hmmm.. probably better stick with what I've got.

The long term plan it to use a modified ATX psu for battery charging, and an electronic PWM controller for dumping excess into heating elements, and then I can change the motor to 3ph, star C-C-C excitation, and rectify the output and send back as d.c, which will hopefully be more efficient, but I suspect it'll take me 'til spring to get the electronics ready.

[mab]

Your capacitors will correct the magnetising current but can do nothing about the severe 3rd harmonic waveform.

Sorry Flux, I didn't really register that sentence before - I suppose that's from the ohter phases of the  3phase motor? I never thought of that - I just assumed that the C-2C method of getting 1phase out would give me a simple sinewave.

I must fire up the scope and have a look...