Author Topic: 3ft head, 10cuft/sec.  (Read 13012 times)

0 Members and 1 Guest are viewing this topic.

Coles

  • Newbie
  • *
  • Posts: 16
3ft head, 10cuft/sec.
« on: July 26, 2010, 06:24:05 PM »
Hi folks,

I've been asked to construct a turbine/waterwheel for an ultra low head, high flow site.  The electricity produced will be used to provide light and heat for a cabin, and possibly to charge an electric vehicle.

The site has a weir with 3ft of head, a mean flow of 400 cuft/s, but as only 1.5-2kw of energy is required (and a maximum of 50% efficiency is expected), a design flow of 10cuft/s will be used.

In metric that's a static head of 0.9m and a design flow of 350 Ltrs/sec.

Firstly, is anyone aware of a similar project (so I don't go to the trouble of reinventing the wheel!), or an 'off-the-shelf' product that will suit my needs?

Also, does anyone have any thoughts on the type of turbine to base this design on?  The client would prefer if the installation was discrete, so an undershot or breastshot wheel with a diameter greater than 5ft (1.5m) would be unsuitable.  The budget is very tight (as usual) so a Kaplan probably wouldn't be a runner either (unless there's an 'off the shelf' plastic prop?).

Any thoughts?  I'll keep this thread updated if the project progresses.

Many thanks. 

 


damian

  • Newbie
  • *
  • Posts: 47
Re: 3ft head, 10cuft/sec.
« Reply #1 on: July 28, 2010, 09:23:08 PM »
Your total power available given the 10 cfs @ 3ft. is about 2.5 KW.  What you can generate depends on how much of that 2.5 KW you can capture.   

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #2 on: July 29, 2010, 07:01:56 PM »
Thanks for the reply Damian.

I'm in the process of ruling out some form of Poncelot waterwheel.  Even with a the absolute minimum ratio of 3:1 (diameter:head), the thing is just too big.  Ideally a ratio of 4:1 would be used, but then we have a massive 12ft wheel.

Here's a really handy calculator for designing a Poncelot waterwheel.  You may have to use Goggle Translate unless your Swedish is good.  Mine isn't.
Image from 24volt.eu.

Has anyone tried a ratio of less than 3:1?  Any idea how the efficiency falls off as the ratio is moved towards 2:1?

Any other avenues I should be considering?

Thanks.






damian

  • Newbie
  • *
  • Posts: 47
Re: 3ft head, 10cuft/sec.
« Reply #3 on: July 29, 2010, 10:04:22 PM »
Breast-shot could get you down to about 6 Ft. Dia.

High torque, low speed.  Gearing?  Low speed alternator?  Use the rim of the wheel as a large cog?

Damian

kurt

  • Hero Member
  • *****
  • Posts: 925
  • Country: us
    • website
Re: 3ft head, 10cuft/sec.
« Reply #4 on: July 29, 2010, 10:20:43 PM »
pardon me hydro is not my thing but would not a Banki Turbine be ideal for your situation with low head and high flow??

Ungrounded Lightning Rod

  • SuperHero Member
  • ******
  • Posts: 2865
Re: 3ft head, 10cuft/sec.
« Reply #5 on: July 29, 2010, 10:51:06 PM »
Image from 24volt.eu.

Has anyone tried a ratio of less than 3:1?  Any idea how the efficiency falls off as the ratio is moved towards 2:1?

Look at that image of the wheel, vs. the head to the left of the jet deflector.

A Poncelot gets its efficiency by letting the water wash up between the blades until it's at the height of the static head, converting any velocity to head, before letting it down again.  So the inner edge of the blades would have to be above the level, not just of the water to the left of the jet, but to the level of the water if the jet were cut off and the water level allowed to stabilize (i.e. any incoming momentum is ALSO converted to head).

As for crossflow, it ought to work.  The application chart in the wikipedia water turbine article prefers a Kaplan to a crossflow when the head is below about 12 feet.  But that might just mean a Kaplan is more efficient, not that a crossflow is too rotten to use.  A crossflow is 'way easier to build, too.

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #6 on: July 30, 2010, 02:59:21 PM »
Thanks for all the replies.

@Damian...  I doubt if I could make a 'normal' breastshot wheel with greater than 30% efficiency, but I haven't ruled it out yet.  I can use a larger flow to achieve the required output but then everything starts getting bigger and heavier. 

@ULR...  Thanks for explaining the detail of how the Poncelot achieves such a high efficiency.  Clearly I won't be able to keep it's characteristics (or it's efficiency) by changing the diameter:head ratio to the degree I'd like to.  I had imagined that the Poncelot was simply an 'impulse' waterwheel and that by lessening the diameter and making the buckets/blades smaller and more perpendicular to the water 'jet' I could increase the RPM without lowering the efficiency too much.  I would of course be losing the 'reaction' element of the water (as you explained), and the efficiency would suffer.

@Kurt, ULR...  A Banki/Crossflow?  Yes indeed.  I know that's precisely where this project is heading but I have some reservations.   3 ft is a very low head for a Crossflow.  I don't have a good calculator to figure out the dimensions, but I imagine it will have a diameter of about 10 or 11 inches and a lenght of about 4 feet.  If the flow is entering just above the centre line of the wheel, would this be the point to measure the head from?  An effective head of 2' 7" from a static head of 3'?   Perhaps I'm being a bit too precious about losing a couple of inches of head, but straight off that seems to me to be a loss of 10-15%. 

Are there any good online resources for designing a decent Crossflow? 


Ungrounded Lightning Rod

  • SuperHero Member
  • ******
  • Posts: 2865
Re: 3ft head, 10cuft/sec.
« Reply #7 on: July 30, 2010, 04:06:03 PM »
A Banki/Crossflow?  ... If the flow is entering just above the centre line of the wheel, would this be the point to measure the head from?  An effective head of 2' 7" from a static head of 3'?   Perhaps I'm being a bit too precious about losing a couple of inches of head, but straight off that seems to me to be a loss of 10-15%.

Losing a significant percentage of your head is a big issue.

You can use an airtight housing from the turbine housing to the tailrace.  Once the air is worked out of it, it will pull a slight vacuum in the turbine housing.  This lets you use the entire head, down to the tailrace water level, even though the turbine itself is up a bit.  (This trick is good up to something somewhat short of thirty feet at sea level, so you should be good to go for a foot or two.)

It also reduces water seepage and seal issues somewhat, because the air wants to go into the housing rather than water wanting to come out.  Good for keeping the genny and electronics dry.  (Downside is a much larger volume of air than water can get through a small leak.  So you have to be extra careful about the housing seals and integrity.  You may also have to look into the lubrication of your seals, since you will NEED seals and water pressure can't be counted on to put a film of water between the seal and the shaft.)
« Last Edit: July 30, 2010, 04:11:54 PM by Ungrounded Lightning Rod »

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #8 on: July 30, 2010, 04:31:30 PM »
Excellent!  I hadn't considered using a 'draft tube' on such a low head turbine (probably because I had approached this project from the waterwheel perspective), but it makes good sense.  The draft tube would also make the whole installation more discreet by reducing the noise from the turbine.   

One of the major benefits of the Crossflow would be that the RPM are so much higher and less gearing would be required, but more about that later.  First I have to get my head around the design of an ultra low head Crossflow turbine...


Ungrounded Lightning Rod

  • SuperHero Member
  • ******
  • Posts: 2865
Re: 3ft head, 10cuft/sec.
« Reply #9 on: July 30, 2010, 04:49:14 PM »
Excellent!  I hadn't considered using a 'draft tube' on such a low head turbine (probably because I had approached this project from the waterwheel perspective), but it makes good sense.  The draft tube would also make the whole installation more discreet by reducing the noise from the turbine.   

One of the major benefits of the Crossflow would be that the RPM are so much higher and less gearing would be required, but more about that later.  First I have to get my head around the design of an ultra low head Crossflow turbine...



RPM is proportional to head over radius.  So make it long and thin.  You might get enough RPM to couple it directly to a genny rather than stepping it up.

(If you want to speed-convert I'm partial to v-belts.  They handle weather well if you keep the sun off 'em and run 95-98% efficient if tensioned properly - dropping to 93% or so if you let them get loose enough to start slipping.  Adjust 'em once after they've spun a few days and then they're good for months to years.  You'll hear 'em if they need tightening, or you can use a spring-loaded idler or put the genny on a pivot to make it automatic.)
 
Do your rotor in two or three segments (maybe of power-of-two lengths), with a support/water blocking disk at the boundaries and separate control of input water to each segment.  You can turn segments on or off to seasonally adjust water usage, torque, and available output current while keeping head, RPM, and (unloaded) voltage at the design level.

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #10 on: July 31, 2010, 01:34:19 PM »
A quick search for a Crossflow design calculator threw up this Spreadsheet from Planetary Hydro.  Takes a while to figure out how to use it, but thankfully there is also a PDF manual available.
 
Very useful, and it certainly appears that a Crossflow turbine would be suitable for this particular application.  

@ULR...  A quick crunching of numbers suggests that shortest suitable runner would have a lenght of about 750mm a diameter of 275mm.  The 'nozzle' height is set as a ratio to the total diameter, and in order to maximise the output for such low pressure (and high flow) this ratio has to be changed from 0.095 to about 0.33.  I have a lot more reading up to do before I trust my sums, but at least it's a start.

Has anyone ever tried this calculator (http://www.ludens.cl/hydro/BANKI.EXE)?  I have a phobia about .EXE files so I'd rather not be the first!

« Last Edit: July 31, 2010, 01:41:21 PM by Coles »

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #11 on: August 02, 2010, 10:44:47 AM »
Lots of reading later and it seems that the crucial issue is the ratio of the width of the inlet 'jet' to the runner diameter.  Normally this ratio would be around 0.1(the diameter being 10 times the height of the jet), but then the runner lenght and width need to be huge to accomodate the flow.

I found a good design manual on Scribd, Crossflow Design Manual by Abhiroop Chattopadhyay.  Unfortunately it doesn't show how certain key constants are derived so it's not clear if it is suitable for designing an ultra low head turbine.

The formula used for determining the Jet size is...

s=(0.22*Q)/(L*Sqrt(H))

where Q is the flow, L is the runner lenght and H is the head (all in meters).

By using a constant 0.22 in the above equation, the ratio of the Jet to the Diameter is 0.084 so the turbine runner ends up massive. 

For a head of 0.9m and a flow of 0.35m3/s, the optimum runner appears to be have a diameter of  0.45m and a lenght of 2.15m running at 84 RPM.  That's pretty big! But at least the fabrication and balancing of it would be fairly simple.  Incidently, the radius of the blades would be 73mm so they could be cut from a 150mm pipe.

And how much power might it produce?  From a hydraulic power of 3kw and a 60% turbine efficiency, a generator efficiency of 85% it might produce 1.5kw.

More reading required to determine the effect of moving the Jet/Diameter ratio from 0.084 towards 0.15.
  .

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #12 on: August 10, 2010, 03:55:10 PM »
A quick update on this project.

I'm abandoning the standard Crossflow design in favour of some sort of Poncelot/Crossflow hydrid with a diameter:head ratio of about 1:1.  I'm reasonably confident that it will achieve an efficiency of 60% and will perform well if the head increases by 20%, and the flow by a corresponding amount.  I expect that the width:diameter ratio of the wheel will also be close to 1:1.

hydrosun

  • Sr. Member
  • ****
  • Posts: 399
Re: 3ft head, 10cuft/sec.
« Reply #13 on: September 09, 2010, 12:02:54 PM »
I just noticed this thread. For off the shelf solutions;
Google Ron Mcleod and his nautilus hydro turbines. They are high quality low head  propeller in a tube turbines.   He has been featured in Home Power magazine.
The other slightly lower cost option is multiple ESD turbines at microhydropower.com  A low quality imported powerpal turbine would also work in multiples. Whatever you make or buy is going to be higher cost or maintainance than a high head hydro because you are dealing with a much larger volume of water.
Chris

richard

  • Newbie
  • *
  • Posts: 24
Re: 3ft head, 10cuft/sec.
« Reply #14 on: September 09, 2010, 08:25:43 PM »
Look up Bonneyville Mills.  In Goshen indiana.   richard

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #15 on: September 09, 2010, 08:45:37 PM »
Hi Chris, thanks for the reply.

I've looked at all these options, but the cost just becomes prohibitive.  The head is just too low to make them worthwhile.  One interesting option was a siphonic kaplan, and I haven't fully ruled that posibility out yet.  I was worried that it might be problematical to keep primed but I've thought of a way to automate that.  One particular benefit of a Kaplan would be that the electrics would be above the floods.  Has anyone ever heard of anyone manufacturing cheap (but durable) plastic or resin Kaplan props?

I came across one example of a small diameter runner (0.25m x 1.5m?) being used with a flow of 120ltrs/sec and 1.2m head, but it was producing an underwhelming 400w.  35% efficient?  Maybe that's as good as I can expect...

But for now I'm working on the crossflow/poncelot hybrid.  When I get a bit further I'll post some scribbles, but any other thoughts would be greatly appreciated.

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #16 on: September 09, 2010, 09:01:30 PM »
Look up Bonneyville Mills.  In Goshen indiana.   richard
Wow!  Thanks Richard.

It has taken a bit of digging but I think I've found what you were referring to.  All the following information is from the Smokstak Forum, the Antique Gas Engine Bulletin Board.





Quote from: Jebaroni
After the water wheel had been around for a while, the "next generation" of water power started to look at water powered turbines as the replacement for the water wheel. The turbine is somewhat like a water wheel on it's side. The entire unit is completely submerged inside a vertical box or flume called the pennstock. The large shaft in the very center is the output shaft of the turbine and would be directly coupled to the mills main drive shaft to power equipment inside.

There is a smaller shaft just off to the side of the output shaft which was the control rod. The segment gear that the control rod pinion meshes with has many "fingers" or runners that go out to the edge and open/close several gates around the outside edge of the turbine. So by turning the control rod, you control how much water is going into the turbine and thereby how much power or speed the turbine puts out. The water enters from the pennstock through the gates on the side and exits through the opening in the bottom to the tail race. The tail race and pennstock are two separate compartments. Think of the Pennstock as a giant sink with a hole in the bottom and the turbine plugs the hole.

The turbine had the advantage of being able to run even during mild floods and mild droughts when waterwheels would be helpless (too much water after a waterwheel begins to submerge the waterwheel. This is kinda like trying to run a marathon through a lake). The downsides to turbines is that they were not quite as efficient (but almost) as a waterwheel, they consumed massive amounts of water (usually almost 1000 gpm or more), were prone to clogging with sticks, leaves, eels and other foreign matter in the water and servicing was difficult.


It works a bit like the wicket gates on a Francis Turbine?  More research needed...

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #17 on: September 10, 2010, 03:49:21 AM »
Apologies!  More of a vertical kaplan turbine.  In terms of size, it appears to have a diameter of 500mm (20") and an inlet height of 100mm (4").

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #18 on: September 10, 2010, 04:32:18 AM »
Ah yes.  I think I've found what Richard was really hinting at.  A 'Gorlav Helical Turbine'.  Very good.  Here's a news article about the installation.

Quote
"The single turbine at Bonneyville is a 10-kilowatt system"
10kW?  Assuming 65% efficient that would require a 1.5m3/s flow over a 1m head? 

http://www.youtube.com/watch?v=SEr6x-ZSM4M

More stuff to think about....

Coles

  • Newbie
  • *
  • Posts: 16
Re: 3ft head, 10cuft/sec.
« Reply #19 on: September 10, 2010, 05:53:55 AM »
A Gorlov Helical Turbine runner.

From University of Rhode Island.