hydro experiment

[windspeed]

here are some photos of our work to date not very pretty ..yet

demensions are 17" long x 9" dia

so far we have had 1.5 amps into a 12 volt battery.... expecting more

the gearing is about 3 to 1 with a serpintine auto belt

maybe 300 rpm at the genny...which is rated at 400 w 12 v........it was very poor as a wind turbine with a 3 ft blade

we intend to extend the flume to full head height and revise the nozzle adjustment to close off the flow completely so that the flume fills up and we can regulate the flow at full .......any ideas...?

the dam is next.. thinking of a wooden beam accross the river to attach the turbine to.......more ideas ?

the frame we got in a scrapyard..... a few mods , the legs were very handy for working height and also for positioning in the river

trying to start a diary.... any tips ....can I move older posts to the diary

HUGE PIC


http://www.otherpower.com/images/scimages/7734/hydro_039.jpg

http://www.otherpower.com/images/scimages/7734/hydro_041.jpg

Windspeed

come on now windspeed over 1000KB and 3000 pixels wide give me a break.....

[Flux]

Not possible to get much idea from the pictures. There is no indication of your water flow or head but with that size turbine rotor you would be looking at a major stream to supply it and if you had a reasonable head you would be looking at quite a few kW out.

For 18W the thing is infinitely oversize and I bet your nozzle is not even running full let alone being fed with the static head of your supply.

For 18W think about a rotor 1" wide and perhaps 6" diameter as a more practical size.

You may be doing better than you think. I bet there are more iron and friction losses in that alternator than you are getting out. At that power level it may only be 20% efficient, most of which you could reclaim with a dual rotor axial machine wound for direct connection.

Flux

[windspeed]

Hi Flux

I posted the details previously this the first diary entry my intention is to have all the "hydro experiment" in the diary

here are the details,original posting

"our group in the west of Ireland have started to build a low head banki

we have .5 mtr of head and flow 33 ltr per sec average

we are building the prototype from ply and pipe slotted in a holding disc with a full backing disc 9" dia ..this is backed by a 6" saw blade ..........see pics

the axle is a 16 mm threaded bar with 2 ebay pillow block bearings either side

next task is the mounting frame which will be bolted onto the 3ft high x 6ft wide wooden dam"

as you can see the head is minimal hence the long runner to avail of the flow

18 w is I hope not going to be our final figure

our thinking is because of the low head and reasonable flow with low rpm and good torque it would be hard to avoid gears

we have to revise the vane to control the flow so the flume can fill up to head level

do you think the width of runner is too wide for this flow

why do you say the alternater is inefficient, apart from the gearing of course.

thanks

[Flux]

I shall have to translate your metric units to something I can understand.

1. /2M is a very low head for a Banki, you may do better with a Kaplan or propeller turbine but the Banki should do something.
   
33. l/sec sounds a lot but I suspect it's not that much really. I am inclined to suspect you don't have enough water to deal with that width of rotor but it is early days and I await more results.
    
    

You may get a cheaper result with gearing but at 18W the belt drive is not going to be very good.

The alternator seems to be some form of commercial pma and I suspect it has an iron core that robs you of at least 30W. At its full rating the efficiency may be quite high so stick with it for now until you deal with the turbine.

May be an idea to section off part of the nozzle and then increase the width until. the thing just manages to cope with your water supply, then you have maximum static head for your nozzle to work on. If you have things right the working speed will be close to half no load speed and you may need to play with gear ratios or alternator loading to get that right.

Flux

[Ungrounded Lightning Rod]

33 l/sec sounds a lot but I suspect it's not that much really.

523 gal/min.

[Ungrounded Lightning Rod]

Let's see...,

.5 meter = 1.64 ft

watts = head (ft) x flow (gal/min) / divisor

Divisors I've seen:

 10 (A DoE Microhydro formula taking into account a system efficiency of a tad over 50% and maybe being pessimistic)

  8 (Energy Systems and Design - turgo wheel on automotive alternator)

 2.3 (on this board - I take it that one is the energy in the water and not deducting system efficiency.)

Going with DoE's divisor (i.e. most pessimistic - still might be high if you've got a rotor that's not efficient at your head/flow combo):

1. 64 * 523 / 10 = 85.77 watts
   
85. 77 watts 24/7 is about 752 KWhr/year- a hair over 2 KWhr/day - or the equivalent to about 400 watts of solar panel in clear weather.  Won't run you house (by a factor of 12 or so) unless you just use a couple lights, but nothing to be sneezed at either.
    

[Flux]

Yes that's the right order, I doubt if you will get much over half that with a Banki on that low head. Perhaps 60W to aim for.

I don't know the present nozzle dimensions but if it is 17" x 1/2" I think you will need many times your water supply. Maybe trying 6" long would be a good start. Keep reducing until not all the water passes the nozzle, then you are about there.

Flux

[mattg]

I may be wrong , and it is hard to tell from the photos but is'nt your nozzle pointing the wrong way- as far as im aware the nozzle should fire towards the top half of the runner, imparting most of its kinetic energy on the blades. The water should then pass through the centre of the runner, imparting a smaller amount of energy as it passes  the blades for the second time.

Or is the picture showing the runner upside down? Or maybe im being stupid...

Im following your progress carefully as we have two mycrohydro projects on the go later this summer. Looks cool so far!

matt

[Flux]

I can't tell much from the photos, it even looks to me as though it may be hitting the blades in the wrong direction but I wouldn't like to deduce anything from those photos.

I did refer him to pictures of an Ossberger turbine so I hope he has got it right.

The nozzle should be sort of tangential to the rotor at 10.0'clock and conforming to its curvature to direct the water into the vanes with the outflow striking the other side of the drum at about 4o'clock.

Flux

[windspeed]

 " 33. l/sec sounds a lot but I suspect it's not that much really. I am inclined to suspect you don't have enough water to deal with that width of rotor but it is early days and I await more results. "

there are times I think the flow would suit this width

We hope to have more results in a week

"You may get a cheaper result with gearing but at 18W the belt drive is not going to be very good."

18 w was produced with 14" of head dam leaking and poor vane design

we hope to get near 20" of head

"The alternator seems to be some form of commercial pma and I suspect it has an iron core that robs you of at least 30W. At its full rating the efficiency may be quite high so stick with it for now until you deal with the turbine."

yes it will do for now and when go axial flux we can cash in

"May be an idea to section off part of the nozzle and then increase the width until. the thing just manages to cope with your water supply, then you have maximum static head for your nozzle to work on. If you have things right the working speed will be close to half no load speed and you may need to play with gear ratios or alternator loading to get that right."

for now it will be easier to go with nozzle depth adjustment

Thanks for your help

[windspeed]

"Yes that's the right order, I doubt if you will get much over half that with a Banki on that low head. Perhaps 60W to aim for."

"I don't know the present nozzle dimensions but if it is 17" x 1/2" I think you will need many times your water supply. Maybe trying 6" long would be a good start. Keep reducing until not all the water passes the nozzle, then you are about there."

Flux

we calculated the flow by the weir method .....1" over 5 ft devided by 17" (runner) should give us 3" at the runner

would be happy with 60 w

thanks

[windspeed]

Pictured two methods of water flow we are using the verticle flow

Windspeed

[windspeed]

correction this is the flow we are using in our turbine

Windspeed

[ghurd]

Not sure where this fits, so here it is.

I believe that type of water wheel needs the nozzle to cover at least 2 full vanes at a time.

That is why they control the width of the flow, in addition to the possibility of the flow into the nozzle.

Trying to control the nozzle like you are doing it looks like much of the water is not doing any work.

G-

[mattg]

Oh I see, I havnt seen the vertical flow before.I stand corrected, thanks for the pics.

matt

[windspeed]

yes you may be right but it is the easiest control to fabricate

and we do get alot of flood which would give max flow

the flow varies from one day to the next

have you any ideas on a design for controlling the width of the flow

[Ungrounded Lightning Rod]

Sure:

Put a disk in the blades about a third of the way from one end.  Put a similar divider in the nozzle.  This splits the turbine into two turbines, 1/3 and 2/3 power.

Add a cover (or other valving arrangement) to block the water flow from the 1/3 and from the 2/3 sides.

Turn on the 1/3, 2/3, or both to handle 1x, 2x, or 3x flow.

(Got a LOT of variation?  1/7, 2/7, 4/7 gives you seven steps.)

The genny will run at about the same speed and voltage regardless of the number of segments open.  The amount of water controls the max current you can get from it.

[ghurd]

What ULR said.

And the easiest to fabricate is usually the worst choice.

G-

[Flux]

Yes I think dividing the nozzle is infinitely easier than the variable area scheme for trials at least.

I suspect reducing your nozzle to 1/3 width will make a big difference. I assumed you were trying to hit one blade and get most of the output from the first stage. If your nozzle is wide enough to feed several blades then I think it is way too wide for your flow.

You ought to divide the rotor as ULR suggested but for trials you can just restrict the nozzle width until you get some idea where you are going. Dividing the rotor will also strengthen the blades but at such low head it may not be necessary.

Flux

[windspeed]

Thanks U.L

you said,

"The genny will run at about the same speed and voltage regardless of the number of segments open.  The amount of water controls the max current you can get from it."

do you mean regardless of the width of segments open...............

when we build for the long term "rotor division" will be included in the design

it would mean scrapping this unit to do it now and there are still lessons to be learnt from this one

do you think deviding the nozzle without deviding the rotor would be of any use

appreciate your input  

Windspeed

[windspeed]

Thanks Flux

we are hitting 2 or three blades and I am pretty sure we are getting something on the second stage also because the water is exiting between 6 and 8 o clock

but as you said before we are losing pressure at the nozzle

Windspeed

[Ungrounded Lightning Rod]

Don't know.

I'd guess that the momentum transfer would be about the same without the extra wall in the rotor to guide the water against sideways motion.  But if I were doing the design I wouldn't take a chance on it when it's so easy to put the wall there.  (Before the fact, of course.)

The wall shouldn't make it any WORSE.  So go ahead and test it using just the nozzle width.

Once you get it working you could temporarily dam up the stream to get a couple minutes of full-width flow and compare the charging rate vs. the charging rate for the same head into a nozzle masked to a suitable fraction of the width of the rotor.  The battery voltage will roughly regulate the speed so by measuring the current (and the voltage at the battery for making the measurements more exact) you can get a good idea whether the lack of the extra wall matters.

If you don't need the wall maybe adjusting the nozzle by using a sidewise-sliding baffle will do the trick.  B-)

[Ungrounded Lightning Rod]

I seem to recall that you get the bulk (2/3?  3/4?) of your power when the water enters the runner but you need the exit stage as well to get nearly-all.

The Banki is also nice because it mostly self-cleans (the water exiting the runner pushes leaves and crud back off) and has a flatter efficiency curve with varying load than most other turbine types.

[Jon Miller]

Hi there,

Your first post spurred me on to getting on with my stuff.

I have a video of my 'basic' set up:

http://www.youtube.com/watch?v=pRogfURdIHc

Have/can you make a video of yours working?

Regards

 

[Ungrounded Lightning Rod]

That's cute the way the "startup" picture shows the angle of the jet changing as the rotor speeds up.  Also that the water has left virtually all its momentum behind and is just dribbling down after passing back out through the rotor.

[Flux]

That's interesting.

I get the impression that the thing is running unloaded. When running the jet hardly seems to be deflected. Your peripheral velocity seems to be that of the jet not about half for maximum power. Even if the speed is too low to light the bulb your alternator should be able to hold it down to half speed even if it is at a very low voltage.

Again the whole thing looks to be far too big for the available water supply. At least you have restricted the jet width to something reasonable. A smaller runner would have got the speed up but with a suitable winding you should be able to get the volts up even at low speed.

From the water supply I can't see much power being available and a decent low speed axial air gap alternator is the only thing likely to get you a decent % of what you have. Geared motors will loose you nearly all of it. A larger diameter pair of discs with lots of smaller magnets should get you what you want.

Flux

[Jon Miller]

Hi Flux,

Thanks for the comments, indeed the turbine is running unloaded currently, the AF alternator was just not loading in it unless I shorted it out as I did towards the end of the video.

I agree with the size observation you have made many times, the turbine is far to big for the flow and head I have.  In my defense it was free as was the metal and pipe, I am a student and dont really have the time or money to fabricate a purpose built turbine sadly.  

I will still try the gear box and motor setup as for the reasons listed above but agree that an AF is the only way to go for DIY micro hydro instillations.

Flux, thanks for pointing me to the websites in the last post, they helped me understand it all a lot better.

In an attempt to bring it back to the original post and to limit the hijacking, is it necessary to hit more then one blade at a time or is this done to maximise torque?

Also, Windspeed I like how you have made your nozzle variable by using the two halfs of pipe.  Have you pre loaded one side with an adjustable spring as to get the most effective conversion of static pressure to velocity whilst minimising or stopping air from entering the system?

Keep up the interesting and good work.

Regards

[Flux]

Jon you raise an interesting point. This is a guess .

The Banki is normally used for low head with high flow, that is why it is normally a very wide runner. To be able to deal with the very high flow required for high power in low heads the thing normally runs with several blades in operation at once. Water availability control normally alters the nozzle area and hence the number of blades in contact. Conversion is done in 2 stages with about 70% of the power being extracted in the first stage. With the normal large flows there is still a lot of water to extract more power at it exits the runner.

With machines on very limited flow then I see no reason why you can't just hit one blade and it works largely as an impulse turbine much as  Turgo. In this case the recovery from the second stage may not be that high as with the very low head you will have lost much of the jet velocity on the first stage and the scatter from the first stage doesn't give you a good impact on to the blades for the second stage.

Whatever happens as long as you have the full head available to the input of the nozzle and you convert the pressure energy to velocity at the nozzle it should work reasonably well with just one blade. In the large machines with lots of flow I am not sure that maintaining maximum velocity at the nozzle is quite so important.

When you scale things down to micro hydro sizes things tend to behave differently.

Normally for low heads impulse turbines are not very practical and the choice would be some form of reaction machine but the small flows, clearances and problems of rubbish make it very difficult to make an effective tiny reaction turbine. Even propeller turbines need more water than you have available unless you get into very intricate mechanical design.

The Pelton and Turgo are not ideal from the specific speed point of view for low head and run inconveniently slowly but they do hold up on efficiency well and are less bothered by some of the other problems with low flow. You may be in the region where a Turgo would do better than a Banki, especially with a vertical shaft Turgo with runner just above water level to use the available head. For home construction the Banki is simpler to make and with normal heads and flows it is a good choice but there are always compromises.

Flux