[News] Electricity from Water (run-of-the-stream)

THE REPOSITORY IS THE MOST UP TO DATE SOURCE OF INFORMATION!
WARNING: ALL BELOW MAY BE DISCUSSION OR OUTDATED!

Update:
This project won’t use a turbine anymore due to very low head and very few kinetic output (110…300 W max power, see image below). The nearby stream is just waay too small (and not steep enough) to do miracles with a turbine. We’ll have to use a waterwheel! There’s no other way! Design is almost there, but for integrating the math into the engineering toolkit, we’d need a lot of more time. It’d take us 2 full time weeks at least. And time is rarer than the rarest material on this planet.


This might show some of the complexities that we face when trying to not just randomly construct things but to figure a working system a priori. I’m not yet convinced it works as expected. Once I’ve constructed my system we’ll see if the maths was correct.

In the above image cells highlighted in red pose potential danger, e.g. melting of the generator windings.

A run of the stream system - despite water’s desity = 1000x air density - is no miracle power system. As can be seen clearly in the spreadsheet the water flow speeds are so small, that not more power is possible than with a common wind system.

An advantage is that the power output is steady. This allows either for island-operation without inverter, batteries and all these efficiency-destroying and expensive complications.

Or - if an asynchronous IM for phase lock is available - even for non-island (grid-tie) or (with some circuitry like grid-disconnect switch) even 1-island (off-grid) operation.

Correction: synchronous IM is wrong, Asynchronous IM is what is much easier to find (it’s wide-spread) and what I meant. Synchronous motors are those that operate at their synchronous speed (and have torque there), while asynchronous motors never have any torque at their rated synchronous speed!

They develop torque only if driven a bit (approx. 3%) less rpm than their rated synchronous speed.


Update:
This project won’t use a turbine anymore due to very low head and very few kinetic output (110…300 W max power, see image in first post). The nearby stream is just waay too small (and not steep enough) to do miracles with a turbine. We’ll have to use a waterwheel! There’s no other way! Design is almost there, but for integrating the math into the engineering toolkit, we’d need a lot of more time. It’d take us 2 full time weeks at least. And time is rarer than the rarest material on this planet.

initial design. Critics and feedback welcome. Let’s finally turn the tide. The world is spinning out of control if we are passive. We need to open affordable living standard quick-fix technology now - or we’ll never stop the war-waves of this world.




  • The materials are chosen arbitrarily. (whatever you have available)
  • The design is following the building blocks principle.
  • The axle with bearings (ball or slide) and mount are missing.
  • The generator must be attached to the axle depending on the generator type. If it has low Kv/RPM, then the transmission ratio can be decreased drastically, potentially decreasing losses here. For common induction motors as generator the transmission ratio will be quite big.
  • The surrounding will be modeled when the prototype is built to validate the design (as soon as I get the metal, probably steel, either zinc-coated, rust-protection-painted of stainless steel if it can be afforded. I think the prototype isn’t worth it.).

Avoid using it as underflow waterwheel. Power efficiency c_p will rise significantly if you surrect an aquaduct/penstock and divert some water of the stream. The water height energy will be captured by the buckets! It’ll improve the situation a lot if compared against the purely kinetic underflow waterwheel. Use overflow (pitch-back) whenever possible.

More on how pitch-back is designed here for the nearby stream, later.

Generator 3phase → 1phase: If the way from the stream to the electric loads is long, consider using the 3 phase machine as 1 phase (which will double line-to-line voltage, info is in the web). This way one transformer only at the consumer side is enough. The loss formulas will be integrated in the open ecological engineering document when we have time. Then it’s easier to see when it’s worth it and when not. Best is to not at all use transformers (if losses are low enough due to short connection or multiple wires soldered together – in AC, thick wires won’t help due to skin effect!).

Off-grid systems may prefer using permanent magnet generators. Otherwise self-excitation capacitors need installation across the generator windings. A controller will be necessary in off-grid systems if the power source (flow of water) or the load power demand vary a lot. Otherwise voltage and rotation of the motor will change, thus the induction generator may lose more and more torque. Excitation current frequency must be increased, or a load controller be used to stabilize voltage and generator rpm.

Electronic Load Controller
There is an ELC design in the works. The situation currently just not allows to further develop it. It’s not a lost case, yet, but as here we have grid-tied connection, it’s a lot easier, as reactive power can be drawn from the giant (~infinite) electrical grid. Probably a basic ELC will be installed nevertheless. Though it’ll not be the PWM one, that is in development. The chokes alone cost hundreds, it’s not worth it for low power (pico) systems like this project (about 100…1000W continous per hour)!

Steel, plexiglas variant. Note the sheet depth (thickness) is as reference only. If the sheets are made of steel, the thickness can be decreased.
Will calculate stress analysis for common materials once I got to it. Wood, steel, aluminium, plexiglas, Bamboo?

I’d like to try omitting one of the cross links of the base rectangles. It’s in the way of the water. I don’t like this. The water must have free way.

The bucket/blade’s outer rim might need to have another sheet attached to keep the water in longer. Not really sure if it’s worth it, because the most efficient point has already been surpassed … and if water is high, then the water might stay in too long - or perhaps that’s even an advantage (less turbulence)?

Plexiglas zu teuer, Bamboo zwar nett, aber kein heimischer Rohstoff, Alu, vermutlioch auch teurer im Vergleich mit Stahl und ausserdem nicht so belastbar. Ergo: Stahl.

Oder Holz, wenns billig und schnell gehen soll, man kann das ja immer mit Proof-of-Conzept rechtfertigen … wobei, Holz hat sich durchaus schon etliche Jahrhunderte in der Menschheitsgeschichte für sowas bewährt. Aber als Tischler bin ich da vielleicht etwas parteiisch :wink:


Was ich ansonsten noch sagen wollte:

nette Grafiken, Du musst mir gelegentlich mal erklären, wie das mit dem Raytracing in Blender funktioniert. Aber bitte, nimm für die Blades ne andere Farbe, sonst denkt jeder die wären aus Kupfer. Das hat mich schon bei Deinem Entwurf für die CNC-Mill neulich irritiert :wink:

Was den oberen Entwurf angeht, da sollten die Schaufeln meines erachtens aus Stabilitätsgründen seitlich eine ganze Scheibe am Stück haben, Baukastensystem in allen Ehren aber das dürfte einfach stabiler sein.

Und was den unteren Entwurf angeht: Ich glaub nicht das das Rahmengestänge sonderlich stört, aber ev. wäre noch ne Alternative seitlich einfach große Platten (aus Holz ? :wink:) als Rahmen zu verwenden, vielleicht sogar mit einer Art Einmünmdungs-trichter.

Aber, wenn wir schon soweit sind, irgendwie taucht aus den Tiefen meines Gehirns das Wort „Pelton-Turbine“ auf, wäre das nicht eigentlich das Mittel der Wahl für diese Anwendung ?

Noch eine andere Sache, die ich nicht verstehe:

Dein Spreadsheet. Die Details sind mir zwar unklar, aber grob und abstrakt hab ich dem folgendes entnommen:

Du wendest da ein Spreadsheet für Windströmung auf Wasserströmung an, was ok sein mag, da beides fluide strömende Medien sind und Du den Unterschied durch einen Dichte-Faktor von rund 1000 kompensierst. Und kommst dann zu dem Ergebnis, das da unterm Strich bei Deinem Bächleien wohl nicht allzuviel an Strom unten rauskommen wird. Was ja durchaus schonmal eine handfeste bzw. konkrete und damit nutzbringende Aussage wäre.

grins ich vermute mal, damit ist wohl auch die Idee, die 10000L Milch talwärts strömen zu lassen vom Tisch ? :wink: Sorry, die Bemerkung konnt ich mir grad nicht verkneifen, gleichwohl finde ich es aber grundsätzlich richtig erstmal sämtliche Möglichkeiten in Betracht zu ziehen.

Wieauchimmer, an einer Stelle wird gesagt, das eine bestimmte Menge an energie in der Strömung steckt und es darum geht diese auf eine Fläche zu übertragen und man aber nicht mehr rausbekommen kann, als drinne steckt.

Wiso denkst Du nun, das ein Wasserrad anstatt einer Turbine da mehr rausholen könnte ? Eigentlich gilt doch eine Turbine als effizienteres Konstrukt bei der Stromgewinnung ?

Gruss, Oliver

PS: Was bedeutet in deinem Spreadsheet der negative Torque an einer Stelle ? Meint das, man muss erstmal soundsoviel Energie dafür aufbringen, ehe sich das Ding überhaupt erst anfängt zu drehen ?

Plexiglas zu teuer, Bamboo zwar nett, aber kein heimischer Rohstoff, Alu, vermutlioch auch teurer im Vergleich mit Stahl und ausserdem nicht so belastbar. Ergo: Stahl.

Agreed. Steel and wood is enough. Though I’d like to test bamboo for developing countries. My focus is not on Germany. I forget about countries, the world is my concern.

Und was den unteren Entwurf angeht: Ich glaub nicht das das Rahmengestänge sonderlich stört, aber ev. wäre noch ne Alternative seitlich einfach große Platten (aus Holz ? > :wink:> ) als Rahmen zu verwenden

That’s a nice variant for a cheap solution. Though it’s planned to anchor the base rectangle to the stream’s ground. Of course that’d work with a side wood/metal sheet as mount too. It might even improve stability if the anchor/post is driven into the ground at an angle. Have to think about it.

vielleicht sogar mit einer Art Einmünmdungs-trichter.

[/quote]
Is planned if the waterwheel is placed in-stream. Though I’ve received high criticism for the placement. It’s been said the (next) flood will crush the wheel. Therefore I’m forced to place the waterwheel out-of-stream, which is really suboptimal from an energy standpoint. Let’s see, perhaps two will be built.

The simplest way to funnel the water towards the waterwheel is to use two tree trunks (or chunks thereof).

Of course, as the design is used as a (redirection) overflow pitch-back waterwheel, there’s a need to surrect a penstock. Thus funnelling will happen there. Though for the in-stream placement the extra funnelling by the two wood chunks is desirable.

nette Grafiken, Du musst mir gelegentlich mal erklären, wie das mit dem Raytracing in Blender funktioniert. Aber bitte, nimm für die Blades ne andere Farbe, sonst denkt jeder die wären aus Kupfer. Das hat mich schon bei Deinem Entwurf für die CNC-Mill neulich irritiert > :wink:

True, and what I feared even more is that people think the waterwheel blades need to made from gold. :wink: Of course they don’t! Don’t do it! But make it happen!

Was den oberen Entwurf angeht, da sollten die Schaufeln meines erachtens aus Stabilitätsgründen seitlich eine ganze Scheibe am Stück haben, Baukastensystem in allen Ehren aber das dürfte einfach stabiler sein.

I’d really like to get the stress analysis going. Calculation alone probably will take weeks. But I need to get the maths done first.

Du wendest da ein Spreadsheet für Windströmung auf Wasserströmung an, was ok sein mag, da beides fluide strömende Medien sind und Du den Unterschied durch einen Dichte-Faktor von rund 1000 kompensierst. Und kommst dann zu dem Ergebnis, das da unterm Strich bei Deinem Bächleien wohl nicht allzuviel an Strom unten rauskommen wird. Was ja durchaus schonmal eine handfeste bzw. konkrete und damit nutzbringende Aussage wäre.

That’s it. Power of 110W is … nothing … not even 1/30 of the average consumption. => The problem is the low flow speed! => Kinetic-only turbine (Savonius, Helix, Crossflow, …) out of the question. Impulse turbine like Pelton or Turgo don’t work well with lower head than 30m. Here it’s just 2m or at max 3m.

At first one thinks as alternative one could convert the height into kinetic energy (speed up the water). And while that’s true and an experiment and subsequent comparison would be nice in general - here the problem is the crossflow calculations were for an effective area of 1m (height) * 2m (width along axis) = 2 m^2. => There’ll not be enough water to cover such a giant area! It’s impossible. And the crossflow turbine needs to be 2m high then too. Just as high as the waterwheel.

And there we go: the Romans and Ancient Greek were smart, it seems the waterwheel is still a good choice if the circumstances are not right for any other system. Because contrary to the kinetic only turbines, it also extracts height energy in an easy and efficient way (okay the Romans maybe didn’t use the overflow or pitch-back design too often, but who really knows).


The goal is combining turbine and waterwheel. Though that’s nothing for a quickfix … and quickfix the current design is. Easy to get parts. Just a bunch of different types of parts. And easy to assemble. That’s all I need … until the metal is here. Then maybe the flaws will become imminent. :smiley:


Was bedeutet in deinem Spreadsheet der negative Torque an einer Stelle ? Meint das, man muss erstmal soundsoviel Energie dafür aufbringen, ehe sich das Ding überhaupt erst anfängt zu drehen ?

The torque is negative only if you turn on voltage control. It’s not a too useful option but it shows that to make the crossflow turbine spin up to nominal speed, the current needs in theory needed to be traded for voltage to make up for the difference to the nominal voltage (which is 380V for most 3phase motors here).
If the resulting current is falling < 0, then the torque will become negative too as torque relates to current. Of course this means it then just won’t spin at all. But that’s just theoretical. In practice, the motor current can’t simply be reduced and the voltage increased on-the-fly (it’s possible in a limited way by changing terminal connection/winding series vs. parallel, but that’s just not really useful and may lead to issues)! It’s possible by modifying/exchanging the motor too. So if you don’t want to exchange motor/generator - or have different windings configurations for switching inbetween on the fly, then keep voltage regulation off!

For the motor torque the spreadsheet uses Kt = 1 / Kv. The motor torque coefficient (Kt) may seem a bit low. The motor torque = I * Kt too.

But the newest version calculates turbine torque too.
For flow speed of 1m/s, the turbine torque is 1.6Nm after the 70:1 transmission (in the ideal case of no losses there). This is 1.3Nm higher than the torque calculated via current * torque constant.


Note this all is obsolete now with the waterwheel. The waterwheel calculations lead to different values.

ich vermute mal, damit ist wohl auch die Idee, die 10000L Milch talwärts strömen zu lassen vom Tisch ?

Not entirely. Letting the milk flow down the 300m height difference is still a huge (and unused) energy source. And think of the petrol saved because the milk truck doesn’t need to climb the hilltop. The energy crisis could be solved. But people fear technology. And they fear ecologists. And demonstrants. And thus they don’t find a solution for extracting water energy an environmental friendly way. Archimedian screw is one. Waterwheel the other. Underwater turbines a third. Mountain stream the next. Airflow upwards in the alps follows, … how about deccelerating space scrap and extracting the energy, then beaming it to Earth using the famous Ionosphere?

When looking at the quantities bill of the waterwheel (the prime-mover component of the electricity from waterflow system), the 160 nuts and 80 M12 bolts are shocking.

Thus a new overhaul: a minimalistic version (Git branch).

  • Less bolts,
  • easier drilling of less holes,
  • enhanced buckets,
  • overall easier placement,
  • hopefully easier to manufacture sizes, e.g. 20cm instead of 16.666…cm).

Disadvantages:

  • A bit risky from a stability perspective. No risk, no fun.

While redesigning the waterwheel, a potentially disastrous mistake was found - and fixed. The buckets are plug and play. Each is shifted sidewards by the side plate’s thickness. But they didn’t fit into each other due to a badly welded sheet metal! A new revision for the non-minimalistic waterwheel (master branch) is underway.

Hi,

welchen Durchmesser hat eigentlich das Rad ?

Das Dreieck vom Ständer würde ich ev. noch durch einen Mittelholm verstärken, so ähnlich wie ich das auch bei dem Ständer von meinem Warmluftkollektor gemacht habe:

Zur Seite hin könnte der Ständer auch etwas schlabberig sein, ich weiss allerdings nicht, inwieweit hier seitlich starke Scherkräfte auftreten können oder nicht.

Gruss, Oliver

Thanks for the input. Pretty good idea. Will add something like your side-to-side stabilizers at the bottom of the flipped-V holders, too. Else the single weld will probably break.

Another update:
Because the 40 bolts and 80 nuts of the minimalistic version - while better than the 80 bolts and 160 nuts of the original version - were still quite shocking. The waterwheel just underwent another redesign.

Now there’s a fully pluggable waterwheel version. Just 16 bolts and 32 nuts now (for the cross beams).

We must not give up. For a greener and better world. Once that beast is constructed I’ll either flip to a wind power plant and micro grid or to the electric vehicle (oh, the ELC is first of course, and let’s hope we can show RODOS there). Also thinking about that furnace, not happy how opensourceecology is handling all that, such a furnace is of huge importance for recycling old metals. It should be a priority.

Somewhere in a .gif there are some dimensions included, but while in general the design is scalable it is shown as 2m diameter (center bucket to center bucket).

The metal supply is settled, 4.90€ / 1kg Aluminium (cheaper than Stainless Steel) … furthermore a wood-compatible construction is reaching the top of my priority queue.
Several blender extensions have been driven forward over the course of the last weeks … (They have been holding up this project much to my regret - I wished freecad and librecad finally merged and blender integrated tightly with both.)