Hey forum.

I am designing a small energy harvester/measuring unit. A propeller is placed in a stream of water and will act as a energy harvester. It should be able to generate power enough from the flow to send data twice a minute through an Xbee

The propeller will be rotating at about 50RPM.

Hardware:

• I will be using an arduino Fio - it seems to have a charging circuit and Xbee is easily mounted. I will add a 1000mAh battery to have some power in reserve, should the water stop flowing.

• Xbee pro to send data

The general approach i thought about taking was to connect a rectifier circuit to the alternator to get a DC voltage. Then to a DC/DC converter that regulates the voltage to > 3.7V (compatible with Arduino Fio) and then directly into the Arduino Fio since it will take care of charging the battery.

Question 1) How does that plan sound???

Question 2) HOW DO I CHOOSE THE ALTERNATOR????
I am aware that i can use a BLDC motor, but how do i determine how large it has to be? It would be great if i could use a fan from a PC - knowing that the flow is almost constantly such that i will have 50RPM on my propeller i dont need to charge in bursts but rather a lower constant charge.

Question 3) I obviously need some kind of gearing since the low RPM wont give me any voltage from the alternator. But how many RPM will i be aiming for - would i want to aim for half the no load speed of my motor/alternator as the speed/power characteristic plot of a motor indicates?

Any help is greatly appreciated.

Bear

 

How do you know that the rpm will be ~ 50 rpm?
What is the torque at 50 rpm?
What is the diameter of the propeller?

 

How do you know that the rpm will be ~ 50 rpm?
What is the torque at 50 rpm?
What is the diameter of the propeller?

180mm prop
I know the RPM will be 50 ish, because i have calculated it. Modeled like a screw but with some slip due to loss.
I am not sure about the torque. I think the torque is calculated by: T=C_p*P_0/omega Cp is Betz limit, P_0 is than available power (P_0=0,5*A*density_of_water*velocity^3) and omega is the angular velocity (about 5,2rad/sec at 50RPM).

But i am not 100% on the calculations in the sense that it is actually not a creek but flow in a pipe - the available energy here should be higher since Betz limit should (in my head) be higher since the water has no ware else to flow then through the pipe, hence you can build up a larger pressure difference over the prop.

Thx

 

Question 1) How does that plan sound???

Question 2) HOW DO I CHOOSE THE ALTERNATOR????

Those two questions are related. At low voltage, consider that your rectifier will lose two diode drops in voltage. That could be as high as 1.4V, so if you're generating only 3V, the rectifier loss is almost half your power. You can reduce the loss using Schottky diodes but the problem is still there.

One solution is to generate a higher voltage at lower current. The rectifier ∆V loss remains the same, but the percentage of the power is then lower. Another possibility is to use a transformer on the AC coming from the alternator to boost the voltage before rectification. That loses power also and is less efficient than using the ideal alternator, but might be better than losing so much in the rectifier.

I think you're going to want a motor or alternator wound for more voltage at lower rpm. I believe that will be finer wire in the windings, and lot of turns. A car alternator is wound with heavier gauge wire and fewer turns, lots of amps, and runs at several thousand rpm. Your gear ratio should be as low as possible, I think, for maximum efficiency. Not sure about that, but a high gear ratio seems like a bad idea.

 

The plan is ok...but the steps are backwards. First you need to build the water box, to see how much power you have to work with. After you know that.....then you can scrounge for a dc or ac motor of appropriate size. Then, depending on what motor you can find......you can work on circuitry.

 

Not to expensive to find out.

http://www.ebay.com/itm/DC-motor-DC...607?pt=LH_DefaultDomain_0&hash=item2592830b57

 

I'm not sure if this is a better approach or not but I am struck by the fact that gearing is pretty inefficient and feel that getting a higher speed from the prop directly would be better. Can you introduce a venturi to reduce the diameter and increase the speed? although this would mean a smaller diameter prop which is less efficient.

How much energy do you need? Is it sensible to turn the prop/pipe into an alternator directly, e.g. magnets on the edge of the prop, stators on the inside or outside of the pipe?

 

The plan is ok...but the steps are backwards.

Actually, optimizing for end-to-end efficiency is an iterative process. Each step needs to be optimized in consideration of the other steps. It's one of the problems with alternative energy installations. Windmill prop design depends on the end load and control strategies. Choosing the wire gauge for the alternator windings requires an estimate of the current and voltage needs. If you want to squeeze out the most performance, you can't design the steps independently.

In this case, the choice of generator depends on available pressure drop and flow, dimensions of the tube, required power output profile and so forth.

Or, you forget efficiency and just over-design it to get the job done.

 

Nice guys.......

About the voltage drop - thought about MOSFETs but think i have ended up with a motor designed for fewer RPM.............. like this guy maybe:

http://www.ebay.com/itm/1pcs-120V-D...415?pt=LH_DefaultDomain_0&hash=item2a3bd060af

(?)
Maybe i don't have to gear up too much to get enough power out of it. The arduino FIO can charge with 5V and 340mA (1,7W) max.
I agree with the fact that large gear reductions are unwanted. As said, the water is moving quite slow so not much kinetic energy in it.

I Would love to avoid the venturi if i could. would prefer the simplest and easiest scalable design. There is also a matter of pipe maintenance. I am not saying it cant be done, just that is an extra "thing" on the system. I could make the waterbox but we are talking very large quantities of water so i would have to make a small scale model - of cause, that is also an option.


thx