multiple pennstocks

[camillitech]

Hi guys,

been running a 230vac 200w turbine for over a year now at a head of 130' and a flow of around 10gpm As my burn (Scottish for creek :-)) can flow as little as 5gpm in May/June but 100gpm in the winter I've just bought a Stream Engine with 4 nozzles and plan to feed it with at least 3 different pipes at different heights in the header tank so as the flow increases it spills into the next pipe.

Has anybody ever done this? are there any pitfalls? I'm using a 40mm ID pipe for the first pipe a 50 for the second a 90 for the third and my pipe run is 900' The project is on going and there's a description on here http://www.navitron.org.uk/forum/index.php/board,11.0.html

Any input would be appreciated.

Cheers, Paul

[dbcollen]

When you have one partially filled penstock you will see a small drop in power due to more windage in the turbine from the one slow jet. Other than that and the air bubbles making noise and causing surging in the penstock it should work fine. In my oppinion you would be better off closing off or installing smaller nozzles when the water volume drops, to eliminate the windage in the turbine and maximize power production.

Dustin

[camillitech]

Hi Dustin,

thanks for you quick reply, I thought my system may not be as efficient as it could be but I'm trying to make it simple and automatic. The problem being that my turbine is 470m away from my battery bank, through a wood and down a cliff :-( The header tank however is quite accessible so do you think I'd be better off with valves near the header tank that I could open only if there was sufficient flow to fill them?

Thanks, Paul

[Ungrounded Lightning Rod]

You don't want a half-full penstock.  The water wastes a bunch of power flowing down to the level where the pipe is full, then comes out of the jet at a lowered velocity that doesn't work well with the spinning wheel.  You use water but waste most of the power in it - essentially all of it if the penstock is no more than half-full   .  You want the jets to go from full-off to full-on and back.

I had an idea for auto-regulating this with a standpipe and intermittent-siphon arrangement:

 - A standpipe acts as a "filter capacitor" after the "resistance" and "inductance" of the initial plumbing run to near the turbine.

 - Taps at various levels serve as "comparators and switches" to turn on multiple jet loads (much like switching on multiple dump loads when you're trying to maintain a voltage level in a battery rather than a water level in a reservoir).

 - The intermittent siphon arrangement of each tap gives the "dump load controllers" some hysteresis, so they switch from full-off to full-on rather than turning on partially.  If the hysteresis is sufficient they shouldn't oscillate (except a very slow relaxation-oscillation if they end up regulating the reservoir level).

See my comments in this article for a description.

Haven't actually built one of these puppies so this is all design speculation.  My potential R.E. site is on a high desert in a rain shadow.  (Gosh I'd love a water power source.  But I've got lots of wind and sun...)

[dbcollen]

The problem with the valves at the top is:

1.  need to vent the top of the penstock so it empties and does not create a vacuum (plastic pipe can be sucked flat)

   
   

2.  filling the penstock quickly can burst the pipe at the  bottom, because when you open the valve all the water will flow down the pipeline pushing out the air, untill the air is gone. then the water will slam into the nozzle and you will have a huge water hammer.

   
   
   

I have a water heater Pressure Relief valve installed near the nozzle, to limit the pressure excursion to 150 PSI if I were to plug a nozzle.

I have 140 PSI static in my penstock (325 ft head, 925 ft long) My 2 inch poly penstock is only rated to 100 PSI so I try to keep my dynamic pressure near 100 PSI.

Dustin

[Ungrounded Lightning Rod]

Water hammer can be mitigated by a capped, upward-pointed, piece of pipe teed off the delivery end.  The incoming water compresses air ahead of it, trapping some of the air in the pipe.  When the water gets to the jet and can't continue moving as rapidly, some of it goes into the stub and compresses the air further, rather than having an arbitrarily high pressure spike when the last of the air leaves.

The compressing air in the stub acts as a nonlinear spring to bring the water to a smooth stop and then push it back toward the outlet.  Excess air also leaves via the outlet.  You achieve equilibrium with the bubble extending to near the bottom of the stub.

If left running a long time the air might eventually dissolve in the water and lose the spring action.  But in a switched penstock application (with a relief vent at the inlet end) every time you turn it off the bubble is refreshed.

[camillitech]

Thanks for the input guys,

certainly more to think about than I first envisaged, looks like I've quite a bit of experimenting ahead of me.

I will keep you informed of progress.

Many thanks, Paul

[camillitech]

Morning all,

well it is here in Scotland  many thanks to all who posted reply's way back in February and for those who may be interested here' how I got on.

I went ahead as planned with the three pennstocks at different heights in the header tank as planned http://www.navitron.org.uk/forum/index.php/board,11.0.html The first at 40mm internal the second at 50mm and the third at 90mm each one feeding a different nozzle on a high voltage 'Stream Engine'. I commissioned the system about a month ago after a dry spell and with flow down to 6.3gpm .4lt/sec which after a little playing about with nozzles and adjusting the field gave me about 100w from 130' 40m of head.

When the second nozzle comes into play it gives around 300w and the third 800w as pointed out there is a little loss of efficiency when one pipe is not fully filled but it's not a great deal and the autonomy of the system more than.

It's early days yet but we've had a good variety of flows down the stream and so far I'm well pleased with it, I'm sure that I could get more output by adjusting the field at the higher speeds but like I said I'm striving for an automatic system.

I've been taking power output readings twice daily and as soon as I can figure out how to do a graph I'll post results.

Cheers, Paul

[Ungrounded Lightning Rod]

did you try the "intermittent siphon" trick to avoid partly-full penstocks?