Author Topic: High voltage (Read 4242 times)

jaketheironhead · « **on:** October 01, 2009, 08:11:36 PM »

I own a sizable property and was considering a windmill to help with my pv panels however the best location I have for building is approx 1 mile in distance up on a high hill. My thought was to make a windmill for at least 240 volts (perhaps more)to reduce line loss and run it to my batt bank to a transformer step it down to 48 volts then a rectifier and charge controller. Now with that in mind will I still need a dump load because I believe the transformer primary will act as a continuous load and then the charge controller could just cut off the secondary when the batt are fully charge. Also whats anyones thoughts on using a MPPT with a setup like this

Flux · « **Reply #1 on:** October 02, 2009, 12:44:58 AM »

For a 1 mile transmission you are going to have a hefty cable cost even at high voltage. Single phase presents lots of problems and 3 phase is probably justified even though it adds more costs.

Yes you will need a dump load or some sort of controller. The transformer primary shouldn't act as any form of load if it is designed properly. It will take magnetising current but that is reactive and will provide no load. there is some inevitable iron and copper loss from a transformer running light but if well designed it should only be a few % of full load rating.

Without exception I would expect a mppt machine to perform better than one without, but I am not sure what sort of choice of mppt controller you have at the voltage needed for this project. The grid tie 2windy boy2 works up to 600v but I don't know of any battery charging scheme that works at this voltage.

You could use a mppt converter on the rectified dc from your transformers if you chose the transformers correctly but you would have to give serious consideration to voltage spikes especially if this long cable run is above ground.

Flux

joestue · « **Reply #2 on:** October 02, 2009, 10:23:37 AM »

Build your own MPPT. (stepwise tap changer, maybe 4-6 taps, PWM the DC side). Wind your own transformer, should get 1.5-3x the power from a traditional 'dumb' system, depending on the wind at your location.

What power level/what size turbine are you looking at building?

I would go for as high a voltage as possible but this is limited to practicality.

120hz at 400 volts into a 240v 60hz transformer will present almost no load to the turbine.

I recommend building a proper 12 pulse rectifier, it adds the complexity of a second set of secondary coils, but this adds redundancy (one fails you can operate at 60%) and will remove the sometimes strange things that can happen when you have a 5th harmonic of 600HZ @ 12%, and you have a mile of 14 gauge wire feeding this...

scottsAI · « **Reply #3 on:** October 02, 2009, 12:43:10 PM »

Wire loss calculator. (bottom of page) http://www.powerstream.com/Wire_Size.htm

480vac 3 ph. 1 mile, Current 4 amps.

Loss is 15.24% using 10awg Al wire 3 ph.

No idea what this will cost.

Have fun,

Scott.

Ungrounded Lightning Rod · « **Reply #4 on:** October 02, 2009, 03:30:30 PM »

For a mile I'd get some "pole pig" transformers and step it up to about 8 kV (scaled down by <your actual voltage> / 240V). Run your cable on poles with standard power line hardware and be sure to include the surge arresters on both ends (though you can skip the fuse/cutout switch).

Figure out the voltage/Hz ratio for your genny. Anything under 240/60 is fine (or 240/50 for European transformers.) Transformers saturate if the frequency is too low UNLESS the voltage is reduced in proportion. Fortunately with a fixed field excitation (i.e. permanent magnet or saturation-regulated self-excited alternators) the generated voltage is proportional to generated frequency, so one transformer fits all wind speeds.

It's tempting to wind the genny for 8 Kv @ 60 Hz and skip the transformers on the genny end. But getting 8 KV out of the stator and down the tower without arcing - and untwisting the drop wire with the wind up if you DID figure out how to wire it - would get problematic. B-)

= = =

If you have enough power to charge your batteries at a non-trivial rate you also have enough power to OVERcharge them. So either it's a play-toy system suitable for a couple dashboard lamps or you need the dump load controller.

jaketheironhead · « **Reply #5 on:** October 03, 2009, 06:57:38 AM »

Thanks to everyone that replied your thoughts have been most helpful

I'm still very interest in doing this I can see that this is going to be a challenge to accomplish but I am set on doing this even though it would be easier to relocate the genny closer this is where i believe the best wind would be this hill is already the highest thing for miles couple that with a tall tower of say 120 ft and i should get a lot of wind so i am thinking big around a 20ft dia genny this should produce a considerable amount of power As far as getting it to the homestead I would like to bury the line inside polyethylene pipe so I'm not sure just how high of a voltage I can get away with but I would say 8kv is too much but I like the idea of using a step up transformer as apposed to winding the genny to high voltage however some of you suggested that a transformer would present no load on the windmill I'm no electrician so I have a little difficulty understanding that but with that said how would I put a load on the genny so it doesn't over speed and explode since with the transformers in use the genny is effectively on its own circuit with the primary on the the transformer and from what I've seen is that a dump load to control this is tied in the back side of the batt bank

ghurd · « **Reply #6 on:** October 03, 2009, 02:37:08 PM »

I am wondering how great the idea to run 1 mile lines, as opposed to finding a less perfect site a lot closer with a little taller tower and a bit more swept area.

The wind goes around that hill somewhere.

G-

oztules · « **Reply #7 on:** October 04, 2009, 02:14:56 AM »

For a nominal 240v mill, 2km is about the upper sensible limit... so your 240v and transformer scenario is fine.

We use the 240v AWP turbine into a three phase transformer and then to 48v batts. A dump load is necessary....

It is also important that your 240v and up is at a reasonable frequency. The AWP is a 30 pole machine, and starts to switch in at around 45hz and about 100hz+ at full power. It is currently only about 500m away from the batts (seems a lot further when you have to climb up there though).... but losses are not material...in truth it probably helps match the load.

Remember, if cut in is around 200v, then full power for us is about 450v-500v.... (500 rpm) plenty of volts to sacrifice... bit of line loss is insignificant I feel... the load pulls it down to 250v anyway.

I think best to site the thing where the wind is, and cop the losses. If you have a good wind resource, line loss is the least of your problems with a HV machine.

Some idea can be seen here:

http://www.otherpower.com/images/scimages/5171/awp_1.html

including the transformer and controller.

..........oztules

oztules · « **Reply #8 on:** October 04, 2009, 02:30:19 AM »

It is interesting to note, that recently I have rewound this stator... all 90 coils of it. I decided to try a few less turns (about 20% less).

The result of this was to increase the output from about 1.2-1.3kw to around 1.5-1.7kw. So we were losing more volts in the stator from armature reactance and resistance loss than was obviously going down the drain in the line.

If we were to weigh the tail down a bit, I think we may get a touch more.. I have seen 30A@57volts when it has been getting tossed about trying to furl itself out of the wind.

.........oztules

scoraigwind · « **Reply #9 on:** October 04, 2009, 02:21:38 PM »

All good comments. I agree that it is definitely worth going a way up the hill and using high voltage. The wire is surprisingly small sectional area, even over a mile.

I usually include a relay that disconnects the transformers at low volts (to help with startup) and a second trip circuit that detects high voltage above the normal and applies a big dump or a short circuit for safety.

The MPPT can be done with transformer taps if you are clever.

Chose a transformer based on volts / Hz. If Hz is low then volts must be low in proportion, to avoid saturating the transformer.

SparWeb · « **Reply #10 on:** October 05, 2009, 08:49:15 AM »

Oz,

That write up is excellent. Where can I find part 2?

I especially liked the parts about design-versus-manufacture. At work I'm called upon to design a lot but usually can't get involved in much of the manufacture, (except when there's a complaint!)

...and I really get off on pictures of skewed, tightly-wound stators...

Ungrounded Lightning Rod · « **Reply #11 on:** October 05, 2009, 02:50:27 PM »

Remember that rewinding the stator with the same total cross-section of copper in the coils (just divided into many thin turns or few thick ones) doesn't materially change the power out at a given RPM or the max power at a given RPM and stator heating. It just adjusts the voltage/current tradeoff. The current density and heating from it, when driving an appropriate load and pulling a given number of watts, is the same. (Unless you go SO thick you're winding the coils with bar stock and the eddy current losses climb into significance.)

When hooked to batteries of a fixed voltage, adjusting the voltage/current tradeoff will adjust the matching of the mill to the batteries and thus make some difference in the power you can extract as a result. But if you're winding for feeding a transformer that will then step it to another voltage and you get to chose the transformer ratio, too, there's not much to be gained by one turns ratio versus another. (The only thing that gets modulated is losses in the wire from the mill to the transformer, and even then only if you don't change the thickness of that wire in proportion when you change the thickness and number of turns in the coils.)

oztules · « **Reply #12 on:** October 05, 2009, 05:42:44 PM »

Thanks Stephen, the second part is here:

http://www.fieldlines.com/story/2007/6/17/141131/265

"..and I really get off on pictures of skewed, tightly-wound stators"........ you'd be surprised how they lose their wow factor when you find yourself rewinding the 90 coils. I wound them 15 at a time... just to keep the terminations to a minimum.

........oztules

oztules · « **Reply #13 on:** October 05, 2009, 06:07:00 PM »

Yes, it does change the load matching.... but

In this case, it is the armature reactance that limits the performance. It current limits at the max output. I changed the turns from 48 to 38 per coil, and the wire remained at .8mm in each case

The R went down, which is nice, but importantly, the ampere turns changed, and we were able to extract more current from these lesser turns. Obviously the rpm changed, and we chainsawed some new blades out of a macrocarpa log, which were a few inches shorter.

The rpm didn't change as much as you would think, as with the armature reactance being lower, it didn't change the max rpm very much at all. With the higher turns, it was running away, with the lesser turns, it was still developing more power. Tricky beasts these mills.

Heating of the stator is of no importance, as you can run it shorted, and it still survives perfectly well ie no heat damage at all..... in fact shorting does not stop the mill unless there are only mild winds.

Shorting in higher power winds (900W and up) only serves to speed up the blades, as the amp/turns limits the power out, and so power in. This obviously does not result in stall. Only by loading it with a 1.5kw dump load can you stop it in higher winds.... and then only in a slight lull. Then the load gets the chance to overpower the blades and cause stalling to occur... the blades then quickly come to a very low rpm. Then you can short it to a virtual stop.

..........oztules

lucasdalebultema7321 · « **Reply #14 on:** October 06, 2009, 08:03:33 PM »

oztules...what on earth do you do for a job? are you a mechanical engineer, electrical engineer? the amount of knowledge in your head is amazing.

I'm trying to tackle medium size wind turbine for my senior design project (mechanical engineer) and finding more and more stuff I should probably add to the turbine system for longevity. Some of the stuff you were talking about I have now idea about but will be researching it.

I really appreciate your ability to fabricate stuff out of nothing. I pride myself on doing the same.

Great writeup on the AWP, I really enjoyed reading all that and the pictures...gives me inspiration to keep going and design something that nice and then get to fabricate it myself up to level of quality I would expect. Thanks!

Ungrounded Lightning Rod · « **Reply #15 on:** October 07, 2009, 08:33:45 PM »

Sorry, oztules. I wasn't thinking about reactance limiting.

Guess my head is still stuck in the coreless, low frequency, potted-stator designs rather than many-pole iron-cored stuff. Should have read your post more closely before sounding off. B-)

TroubledGus · « **Reply #16 on:** October 15, 2009, 05:54:49 PM »

Some things in this thread trouble me.

I see no discussion of grounding and bonding or lightning protection.

Next issue I have is with 8Kv transmission. Thats scary and just for starters you need to consider all insulation as live unless you are using shielded cable.

If it were me I would look at some 1000 V teck cable burried. Step up to that level is fairly easy if you can manage 120 volts @ 60hz then any large 600 to 120 volt power transformer could be connected like an auto trasnformer to get to the 1000 volt mark. A 3 Kva unit is easy to find from companies like Hammond or a savy fellow could rewind what ever is handy provided the core can be stacked ( not welded together )

Armature reaction brings up an interesting point. Alternators by their nature have high output impeadence. You can't get a decent power transfer if your load has a lower impeadence than your alternator. I don't think the way people are currently winding coils and building alternators is not as deaply thought out and well planned so to speak as they could be for best performance.

Trouble is take the iron out of the equation, add the rare earth magnets, and axial design and 100 years of conventional easy machine theory/design flies out the window.

Flux · « **Reply #17 on:** October 16, 2009, 02:25:52 AM »

"Armature reaction brings up an interesting point. Alternators by their nature have high output impeadence. You can't get a decent power transfer if your load has a lower impeadence than your alternator. I don't think the way people are currently winding coils and building alternators is not as deaply thought out and well planned so to speak as they could be for best performance."

Interesting comments. Wind machine alternators need very strange requirements if you want to charge batteries without any form of electronic intervention.

Nearly all the early design was done on iron cored alternators and with the early ferrite magnets they were very reactive. Matching the blades was a compromise that gradually evolved.

"Trouble is take the iron out of the equation, add the rare earth magnets, and axial design and 100 years of conventional easy machine theory/design flies out the window."

You seem to be one of the very few to recognise this. It changes the approach you need to blade matching dramatically. The problem with reactive alternators was to maintain enough load in higher wind speed and blades ran away.

The air gap machines are only resistance limited and load too quickly in high winds and drag the blades to stall and without a mppt converter blade match is a serious compromise but it is predictable.

Although conventional alternator design goes out the window for the air gap machines, they are easier to predict the performance and as long as you forget all you ever learned about machine design and just stick to the basic emf equation they are easier to deal with without reactance to consider. There are tricks that you need to learn about eddies in the conductors once you get away from slots but in general the things are very predictable. The internal resistance doesn't seem to behave as expected when driving rectifiers but the factor seems fairly constant once you get an empirical value.

Unless you use an electronic mppt converter then you must consider the overall power out and not be guided by best alternator design criteria. Blade matching determines output much more than alternator efficiency and you will have to trade alternator efficiency if you want the best power out. The non optimal loading of the alternator is in fact essential to get a good overall compromise if you must clamp the alternator to a rectifier and battery. If you are looking at this from the point of view of a machine designer used to conventional loading the ideas here may seem strange indeed.

The alternator needs to work at constant voltage, variable speed and have an input power related to speed cubed. Without field control this can only be approximated to and the presence or not of reactance changes the compromises you have to make.

Flux

TroubledGus · « **Reply #18 on:** October 16, 2009, 12:08:16 PM »

I don't fully understand all your terminology....

Not exactly what do you mean by an "electronic MPPT converter".

From my limited understanding of these things. I could see some advantage to running the output to rectifier, DC link and then chopper. This would let you float your DC bus voltage instead of trying to pin it to the load for changes in the wind and load ( and it would also be a good place to add a dynamic breaking resistor independent of load ). So is this the sort of thing you are talking about?

And a question for Hugh Piggott I see in your original alternator designs you used a conventional distributed lap winding over the stators you were using for back iron. But this was abandoned in the Axial flux design for single salient pole windings. Has anyone tried a conventional distributed winding?

I've given this some thought and I don't think it would be that hard to add some measure of field cotnrol by change the design a little and removing one magnet disk and making a wound field on one side of the stator. Where my thoery fall apart is can you make this work and not saturate iron....

A realy smart guy ( not me ) would have to sit down and create chart of the forces at play within the syncronous reactence of the machine. This is easy to do with a regular iron core machine at constant speed. This make a bench test easy to predict what the machine will do under load....

Flux · « **Reply #19 on:** October 16, 2009, 03:25:18 PM »

The mppt converter is an electronic buck converter that effectively acts as a dc transformer allowing the rectified voltage to rise with wind speed and increasing the current to the constant voltage battery. The mppt bit just means maximum power point tracker and lets the alternator load the propeller at its maximum power point.

I can't answer the question you ask of Hugh, I will leave that to him but conventional machines tend to have a fairly standard size of slot and the more slots per pole there are the more sections each pole group is divided into. The distributed slots are one means of obtaining a sinusoidal waveform from a conventional alternator. For rectifying to dc waveform is not important but a non distributed winding would require a very strange core and wouldn't make good use of the copper or iron.

Once we abandon the iron core things are somewhat different. In a slotted iron cored machine the flux flips from tooth to tooth and links the coils in the slot instantaneously. This doesn't happen without the iron core and to get the same waveform you would need to wind the coils with almost single wire sized sides. The coils as used here don't link all the turns instantaneously the flux slides across the coil gradually linking more turns so the thing is inherently a distributed winding. There is nothing to be gained from physically splitting it into series connected sections.

The conventional lap winding is not so easy to do with an axial machine and it has been found by experiment that it is more effective to leave out coils and arrange as a single layer winding that is not entirely wound than to wind the whole circumference with overlapping coils. the end windings need to be longer to allow the overlap and the resistance becomes higher. When reactance is the dominating factor the copper length has little effect but with resistance only determining the output the lowest resistance winding works out best.

Combining field control with a permanent magnet is not easy. Permanent magnets behave as a saturated air gap and field coils will not increase the flux significantly . They can buck it if configured correctly but the mmf needed to have much effect on neo is large and you will use a lot of field current doing it. The " Polar " alternator sort of does this but it is in effect two field systems one pm and one wound field and both act on the same coils. The purpose in this case is to allow regulation of voltage at fixed speed. It is a complex and costly machine and the control winding only handles a small change in flux. It would probably not be very efficient for the 3: 1 or more change in flux required for wind power.

There are mechanical methods of altering air gap or moving one magnet disc in relation to the other but they are not simple. I believe it can be done but I prefer the buck converter approach.

Flux

TroubledGus · « **Reply #20 on:** October 16, 2009, 08:36:13 PM »

Hmm....

Interesting this MPPT. I'm not a solid state guy so I am realy not up on this. This thing does same as I was thinking only completely different in function . Where could I do some reading up on this " MPPT " thing?

From some scope screan shots I have seen the salient pole would coils do seem to produce a good sin wave. I understand what you are saying on why it better than a distributed winding. Its not that I don't believe you I just have sit down spin some coils I guess and see for myself and try and get a better feel for it. Again I think its because I am still thinking in terms of iron.

What kind of flux density in the air gap do you think you guys are getting?

Sounds like the field control ideas have been kicked a round from your reply....

Flux · « **Reply #21 on:** October 17, 2009, 12:40:08 AM »

With the ironless machines it seems best to work at a flux density about half Br of the magnet. Typically something between 600 and 700 mT. Small air gaps and high flux density leave you little winding space and the necessary mechanical clearance is a large portion of the total air gap. Large gaps and low flux seem to cause excessive flux loss to fringing to adjacent poles.

Also from the magnetic circuit point of view it also makes you come to the half Br figure to position the magnet half way up the second quadrant of the magnetising curve at BH max.

This is low compared with the tooth saturation figures of iron cored machines but when you consider the teeth are only about half the air gap area the figures don't look so bad. The things do need more magnet but the absence of reactance does help them to keep the load up in high winds. If you design an iron cored machine not to inductive current limit at the top end you have to make it bigger than a conventional design.

Even if you master a way of flux control to achieve the cube law input power you still have a problem with the winding resistance being a big factor when bogged down to a rectifier and battery. It would be a great improvement in terms of prop matching and may in fact perform well but without tap changing, transformers on the ac side or a dc converter you will have trouble keeping the electrical efficiency high.

One of the problems of the air cored machine is disposing of the heat with no direct thermal transfer from coils to the iron. It makes life infinitely easier if you can let the voltage rise with speed and keep the current down. This seems to be the final limitation of direct connection.

Flux

TroubledGus · « **Reply #22 on:** October 17, 2009, 07:41:53 AM »

Very interesting read.....

Especialy the part about the MPPTs.

I have been wathcing and reading this site on and off over the years. I've never built anything just reading.

Seems to me some of the things I have been thinking about are coming to pass like floating the alternator voltage and running at much higher voltage in general.

Thank you for taking the time to answere my questions.

Maybe I will take more of an interest in this and build something in the future.

News:

Author Topic: High voltage (Read 4242 times)