Optimal generator design

[eddiematos]

Hi all,

I am starting a thesis in my university on using pico hydro generation solutions (under 2kW) to power mobile phone masts. In developing countries such as Nigeria, where no other form of remote communications exist, it is currently too expensive to make grid connections to these very rural transceiver masts. Along with other off-grid renewable solutions, hydro power could make it feasible for large multi-national telecom companies such as Motorola to install these much needed mobile phone networks. (For more background info on the progress in the solar and wind parts of this project see: http://direct.motorola.com/hellomoto/windandsolartrial/)

In order to make this commercially viable (which is really the only way these kind of things will ever get done!), one of the things I need to do is ensure the generator design works as efficiently as possible with the turbine in terms of torque and speed. In other words I need to choose a generator from a range of suppliers and manufacturers to fit my requirements. I've been looking around the forum for a while but there seems to be very little in terms of any design advice. Due to the admirable homemade nature of these designs, decisions on whether to use AC/DC, one/three phase, number of poles, etc, all seem to be made on what available parts one can scrap together or, if given the option, "what you feel best with"!

I am looking for any reading (i.e. website links, or book/journal references) anyone might be able to recommend on the design of generators of less than 2kw power. If you think you might be able to help me, and could spare a few minutes of your time, I would also be more than happy to contact you personally if this is what you prefer! (Any advice given would then be posted for the benefit of the other readers in this forum).

Thank you very much in advance for any help you can offer,

Ed

P.S. In case these are useful, roundabout requirements for the systems I will be designing are:

-> Low head, high flow stream so probably looking at more rpms and less torque than your average hydro generator.

-> Looking to charge a 48V battery bank which will power the mobile phone mast.

-> The mobile phone masts need high tech yearly inspections anyway, so access to high tech parts and maintenance would mean that the generator design would not necessarily be limited to low-tech appropriate technologies.

Resectioned.

[Flux]

There must be thousands of alternator manufacturers out there. For a commercial venture quantity and reduction in price will no doubt be the main influence.

Hydro is generally easier than wind, the conditions are much less variable and absolute efficiency at low output is less important. You will probably end up with an iron cored pma with neo magnets. The air  gap machine may compete but you are not so worried about low load losses. The iron cored machine will probably have a good enough efficiency over the range you want if you choose it right.

The only advice I can offer is forget single phase, any multiphase machine will do, but again it is unlikely that you need to consider other than 3 phase.

You will be rectifying to dc so the number of phases has little effect. It is unlikely that you will see much benefit from other than simple diode rectifiers unless you choose an odd machine design with high leakage reactance.

I would strongly suggest that you aim for a voltage that will get you charging 48v in the worst water conditions. You are not afraid of technology and you will simplify the turbine requirements immensely if you use some form of buck converter to match the turbine to the battery load. You could incorporate mppt if you want it to adapt to water conditions automatically. Basically I am suggesting you choose your alternator to cover the lowest output case and use something similar to a MX60 to match your load. You may need to look at a higher voltage device if you want to cover all cases with a single alternator or you can have several alternators to cover your range.

Don't be tempted to cover a voltage range by using star/ delta, the delta option is not good with rectifier loads. Series parallel winding connection would be more practical.

If you go for small quantity production you will likely be forced to use standard motor punchings. These are not ideal for low speed PMAs but will work.

The number of poles doesn't matter a lot, but generally the lower the speed the larger the diameter and the higher the pole number for best economy. Motor cores are intended for high speed and small pole numbers, you can increase pole numbers but you end up with deep slots that may not be to your advantage.

Stacks of big vehicle alternator punchings may be a good starting point as cost may be more of a factor than the fact that they use the cheapest core materials.

Flux

[ebby1234]

Have a look at my website It will give you some info on water turbines. Email me if you want more info

[asheets]

Part of what you'll need to look at is what equipment the cell mast is going to have.  I'm currently working as a contractor for a cell equipment producer, and from what I can tell even the most basic tower will need more power than the <2KWh turbine you've mentioned will produce (unless I've read this wrong, or unless you intend to offsite things like the local SS7 and signaling controllers, or unless your mast is simply a repeater and not a full-blown GSM site).

[asheets]

I specifically refer you to http://www.motorola.com/mot/doc/6/6682_MotDoc.pdf , page 8.  Motorola refers in the chart to a base GSM 4/4/4 to need 2300W.  I would also guess that since your proposal is a rural solution, you'll need more power on the transponders, plus more power to support the longer station-to-station interconnections and backhauls.

[spinningmagnets]

One of the benefits of Hydro is that with a fairly large feed pond, you may be able to specify a fairly constant RPM.

There is a lot of info here on Permanent Magnet Alternators (PMA's) because overall they are very useful in a variety of RPM's, which is the power-harvesting profile of wind-generators.

When using Hydro, steam, or Diesel/WVO, it is possible to control and meter the RPM's, so you have the option of using an "induction-motor-conversion". Bearing in mind that PMA's are still a useful option.

With low-head hydro, it may be possible to make up for low pressure with a large diameter wheel, but in order to get an effective magnet speed, it may need to run a large diameter PMA (Google "Cameroon water wheel"), or use gearing (causing high inefficiency) to speed up a small PMA or induction motor conversion. These are not the most desirable options.

For low-head/high-flow in a remote location, it may easier and more "doable" to make a turbulent flat paddle wheel that is big enough to do the job. "Efficient design" can mean getting the job done with a smaller wheel, and a Poncelet is a style with closed sides and curved buckets, so the flow makes a smooth 90 degree turn from the rim, up towards the central axle (capturing most of the available energy), then stops and flows back out (capturing a little more energy).

Such a wheel can be manufactured in a remote location using primitive supplies. Automobile axles can be used for their cheap and available bearings which are robust.

Best of luck with your research and final design...

[eddiematos]

Hello!

Just thought I would express my thanks for your comments. Though they didn't feature so much in my final dissertation (which has now just been handed in!), they were essential to get me off the ground with hydro power!

I ended up focussing on maximising generator efficiency over various operating speeds (=different stream flow rates). I make steps towards designing a control system that would moderate the current in a field wound rotor.

Anyway, I've uploaded the final report here: <http://www.otherpower.com/images/scimages/9387/An_Investigation_to_Increase_the_Efficiency_of_Pico_Hydro_Electricity_Generation.doc>

Enjoy, and feel free to challenge it, or ask questions! It would probably help for when my viva comes along!

Cheers,

Eddie