mechanical throttle for crossflow turbine

[robl]

I've been running a 600 watt/240VAC/60Hz home-built Michell-Banki for the last 18 years (12 inch diameter by 7 inches wide turbine)and would now like to add a practical, proven valve mechanism to decrease water usage during low flow times. It can be hudraulic, electronic or mechanical. Right now I am opening the unit up and inserting a steel flap. There's got to be a better way.

Thanks in advance for any suggestions.

Rob

[robl]

PS. I have posted a photo of the unit before the generator was attached in the photo upload area.

[DanB]

Hi Rob - welcome to the board!  Sounds like you have an excellent setup..  I can't offer much advice, but your setup sounds super.

You can insert your pictures into your postings here pretty easily if you want.  Ill do it for you this time!

It'd be fun to hear more about the system.

[Flux]

I am not sure how you have made your nozzle or what head you have, but the most efficient way seems to have the outside edge of the nozzle fixed and to alter the angle of the inner edge to change the jet area. The jet follows the outer edge and enters the wheel at the same angle.

You may be able to arrange a hinged vane with an adjusting screw to move it backwards and forwards. Maybe something like conveyor belting would be flexible and provide the seal.

you also have to seal the edges, but a good fit and some leather or more belting may work, depends on the head.

Flux

[robl]

Thank you for the assist Dan,

I'll scan in more pictures that show the completed system in a few days. I think I do have a shot showing the interior construction kicking around.

Gross head is around 25 feet, the pipe is 6"PVC and the system moves around 60-100GPM. Basically, the turbine was constructed almost completely from scrap steel, stainless and aluminum. The draft tube is experimental so it's constructed out of plywood. Only the the angled steel frame and the bearings are new. The generator is a 1Hp 3-phase Hyundai induction motor rewired as a 1-phase 240VAC/60Hz generator.

The drive is via a 0.75 inch wide timing-type belt connecting a 14' main pulley to a 3.5" generator pulley.

Regulation is a simple Triac diversion system that starts to kick in at 250VAC and is fully engaged at 280VAC. The load dump itself is a 1KW/240VAC baseboard heater. Last winter there were three houses connected to the system, two through battery chargers and one directly (basically just using lights and a radio). We are in the process of relocating the intake so I'm not sure what the output capabilities will be this winter.

Cheers

Rob

[Flux]

Try www.ossberger.de and find the OSSB turbine.

They have a cunning version of what I suggested using a pivoted vane. This seems to use a double nozzle, both of which are reduced together. The pivoted vane may be easier to adapt than the one movable one as the forces are balanced.

hope this helps.

Flux

[ghurd]

How about a gate or ball valve in the water pipe? To a stepper motor. To a (an 'inverse'?) charge controller.

As the v increases, the gate closes and vice versa.

112/224 VAC begins to open the gate, 118/236V begins to cose the gate.

Higher demand = higher flow

The ball or gate valve for that size wouldn't be cheap, but the IC circuit guys here could probably draw a circuit faster than the scanner could read it!

Came to mind from an old (HP) magazine article. The man had a piece of plywood at the top, connected to a pulley clothes line type setup. If the lights got too dim or bright, he pulled the line accordingly. Out a window if I recall.

G-

[Nando]

ROB:

We have done a variable nozzle that in a sense squeezes the nozzle horizontally to insure and avoid vibration.

We have used a motor equivalent to a car motor window glass movement with a belt reduction to close the mouth of the nozzle, some were just one side closing, but the best has been when both sides were closed, that keeps the overall velocity of the water but reduce the volume.

We have done another that is a wedge driving by a screw that goes in to the edge of the nozzle ( from the inside ) to perform the nozzle area reduction, using the same type of motor.

Also, the standard Banki control that has a TONGUE in the lower part of the nozzle that is brought up to reduce the flow -- we did not like it, it caused water velocity reduction with eddy currents ( some times).

The DC motor may be driven by an electronic H bridge for directional motor rotation.

Plates used to reduce the volume does generate a lot of eddy currents.

Another system that we were testing was the nozzle divided in three vertical sections isolated each by common wall between sections, so just closing one valve will close a section (1/3). the valves could be manipulated by electric means for variable regulated output -- the initial testing was great  -- I did not know the final results -- my cousin was killed and everything folded -- .

Regards

Nando

[Flux]

Splitting the nozzle into sections is not a bad idea for a 3 or 4 step control.

Throttling before the nozzle would not give good part load efficiency so each nozzle should be all or nothing but a bit of prportional control on one valve could take care of intermediate demands.

Flux

[robl]

Nando, Ghurd and Flux,

Hmm, just hit "post" and lost the whole message. Hopefully it won't reappear or I will have posted two very similar replies...

Thanks for the suggestions.  I've seen a video of a successful sectioned system running somewhere in the Fraser valley (British Columbia, Canada). It was about four feet long, sectioned in at least four places and running on less than 5 feet of head.

Appreciate the suggestion of window motors. I am also partial to the linear drive motors from old analog satellite dishes. I plan to draw up some of the suggestions and check them for feasibility/ease of construction and maintenance.

I do have one design drawn up. It consists of a small single-acting hydraulic ram that pushes the nozzle vane into the water stream (no linkage this time). Theoretically, the water-pressure will push the vane (and the ram back) down when the hydraulic pressure is released. The ram would be powered by a four-ton hand operated pump (used in car-body repair) via a four-foot flex line. I've actually bought all the parts but am daunted by the machining required to insert the ram into the nozzle. Bit of an irrevocable step. The reference to eddy currents is not very reassuring either.

Below is another photo of the unit after a winter of continuous running, except for a monthly lube stop. The black flap covers the access to the main bearing lube connector. The belt shroud was for visitor protection, falling leaves and icicle build-up.

Regards

Rob

[BurksFallsMan]

Robl: I had in my files. Will it help?. Could I ask you a little more about your system?

I am planning to build one for my creek approx. the same parameters than yours.

Thanks, Wilson      WFGonzalez2@cs.com

[BurksFallsMan]

Robl: Old age ! I forgot to include the picture:





Wilson