1.     Introduction

For years always wanted to build something that harvests energy but I could never quite manage to actually start the projects off. I like the idea of recycling waste heat in air conditioning or waste water to extract old heat energy but have never had the confidence to re-plumb the house. I did once make a very large and heavy aluminium plate heat exchanger that could pull the energy from waste heat exhausted from a house (e.g. the bathroom) and would warm the incoming air.

 

2.     My Thoughts

While we all think that rainfall is a great resource for generating energy and indeed companies harvest that power by creating hydro-electricity. Sometimes when its raining heavy I wish I too had a turbine in my rainwater downpipe to catch that wasted energy. However the cloud-burst is often short lived and I know the turbine would stop and the 'lights' would go out. Nevertheless that energy is just wasted and I thought that it needed collecting/concentrating prior to use.

Black and White photo of Hoover Dam

This Project14 challenge gave me the idea opportunity to try something out, perhaps not quite on the scale of Hoover Dam in the above photo [source: www.en.wikipedia.org]

 

3.     The Physics and Numbers

Water has a great amount of energy when stored at a height, this is known as Potential Energy (PE). The amount of energy held within a mass of water (m) in kilograms at a height (h) in metres is:

 

Energy = mass, m(Kg) x gravity, g(9.8m/s/s) x height of storage, h(m)  [Eq. 1]

"What amount of rain water would I expect to gather?"

The answer varies between nothing for days or 'lots' for a few minutes. Scientifically I can get an approximate rainfall per month value form the internet and currently this is just over 90mm per month for October in my area (based on previous years' weather). I'm going to use a figure of 3mm per day.

 

That doesn't sound much but it is falling over the entire roof surface. But the volume captured also depends on the roof area. Assuming a 100m2 roof with equal divides front and back (or side to side) I'd see 3mm x 50m2 of water collected per day. This is 0.15 m3. Water is about 1000Kg/m3 so this equates to 150Kg of water. That's falling from an average two-storey building, which is ~5m from guttering to ground level.

 

If I could capture all that energy (ignoring losses due to friction and conversion) I'd have:

     Energy = 150Kg x 9.8 m/s/s x 5

                 = 7350 Joules

 

"Is that very much juice?"

Assuming we managed to capture all of that energy then we could use it up slowly over a long period of time by running a low power device, for example I guarantee we could run a standard low-current LED for a long time from it. Or we could use that energy quickly with a much more powerful device for example a 1kW bar heater...it wouldn't be powered for very long. Recall that power is the rate of conversion of energy, and 1 Watt = 1 Joule / second

 

That theoretically makes my predicted collection system capable to power 7kW (for 1 second) or 700W for 10 seconds, 70W for 100 seconds or 7W for over 15 minutes. It is now apparent that even saving this up will allow me to run a small/low power gadget for only 0.25 hour at the best. And that is what the Trickle Charger will aim to do: store up the energy over the day until the user wants to charge their phone, run an LED outside light etc.

 

"I Want a Cup of Tea"

It takes 1200J to heat 1Kg by 1 degree Kelvin. I could raise a 200g mug of water by:

     (7350/1200) x (1000grams/200g) = 30.6 degrees (above ambient)

 

Maybe just enough for a lukewarm cup of instant coffee but tea needs boiling water, no I think I won't even manage a cuppa each day . This is also assumping there are no losses whereas taking time to warm the water will result in a constant heat loss, wasting much of that energy.

 

How About Some More Rain, aka Cloudburst

[Source: www.Sunshinegutterspro.com ]

It only occasionally rains hard enough that the downpipes are bursting. Perhaps I could store the water up high and when required open a valve to drop it and generate hydro-electricity, but that would require a fairly large water store on the roof...most UK houses have concrete tiles pitched at an angle. Therefore the slight trickle during a light rain shower is lost and worse still most of it trickles down the inside surface of the downpipe due to surface tension. For now my thoughts moved on to using a supercapacitor.

 

4.     Welcome to the "Trickle Charger"

Based on charging via the rain water trickle, I think this project will be a true trickle charger and hence the name and logo !

Time to order some bits. I'd need a DC motor (as they use Permanent Magnet and can therefore be spun up to generate a voltage) and a supercapacitor at the very least....over to CPC Farnell where I ordered the following essentials:

  • CA06062          1.5F, 5.5v Supercapacitor
  • MC0274069     3v DC Motor
  • MC0194769     1.5-4.5v DC Motor

 

5.     Motor Spin Test

I cannot imagine the motor spinning at 1,000s rpm in the same way it will with 5v driving it and this was confirmed when it arrived. Spinning quickly between the fingers produced a meager 600mV at most (photo below shows my Tektronix DMM TX3 on Minimum Hold as that was the polarity I was generating in the direction of spin).

The supercapacitor can take up to 5v and that would then allow it to be used for powering something useful. This millivolt signal is also so sporadic that it would be wasted if I attempted to have a power IC upconverting this using buck-boost type topography. Instead I need something much more basic. I decided to use an old fashioned voltage multiplier made from capacitors and diodes to bring this voltage up to a level where it could pump charge into the supercapacitor. As a rough aim I though about 10v would be great and then I can use a Zener diode to limit the supercapacitor at 5.1v

 

Onwards with the building....

 

6.     The Turbine and Housing

My fingers would soon get worn out testing this little motor so I decided to go straight into a build. I would need a turbine and housing for the motor. There were several ideas I had for orientating this, placing the motor outside, down stream, using square downpipe or round cross-section. Once mounted I also thought I might be able to blow or suck air through the pipework to spin my turbine and use that for test purposes.

 

In the end I opted for a round design where the motor and turbine sit directly in the flow. The motor is mounted upstream and protected by a 3D housing. Down-stream needed a mounting to help hold the turbine steady e.g. a lower bearing.

6.1     Turbine

Design:

I actually had no idea how to make a turbine or fan nor do I understand the fluid mechanics necessary to make one that is most efficient at extracting the energy from the water. However to make it spin with smaller flows I wanted to reduce the mass and therefore sought after a fairly thin turbine blade. I looked around ThingyVerse and found a great OpenSCAD script by a user caIled Denise Lee that details fan and turbine manufacture, I modified the parameters until I got a turbine fan that seemed like it would work! Below is my OpenSCAD drawing. Thank you Denise.

Printing:

My first attempt at printing wasn't good and after 4-5 layers I abandoned it as I could see the filament sagging at the end of each pass. The remedy was easy: in Cura Lulzbot I just clicked the 'support' box and ran the print again. What came out appeared to be a solid mass with some turbine fan design inside. With pliers, cutters and knife I was able to easily free the turbine fan from the support structure...I'm quite impressed.

 

 

6.2     Lower Bearing Mount

Design:

The lower mounting will have a small brass pin that the turbine can rest on. This will prevent wobble and also guard against the turbine becoming detached from the motor with heavy through put.

Printing:

This printed well first time. Although it took me a while to realise there was a small amount of support material still on one side !

 

6.3     Upper Bearing Mount and Motor Shield

Design:

This top bearing serves several purposes. It will fit into the inside of the pipe and form a waterproof housing for the motor to sit inside whilst protected against the main ingress of downward water fall. The nacelle helps reduce the loading on this component by guiding the falling water into the next section. Additionally the chamfered edge wall will help guide minimal rainfall from clinging to the pipe's sidewalls and into the turbine. I purposely haven't added a wire hole as I'm not sure yet where I wish to bring them out.

Printing:

I printed this upside-down to save on support structures, however it didn't print properly and I had to abandon it early on. When I investigated what was made I did find the completed nacelle inside the mess. On closer inspection of my OpenSCAD images I think there was a discontinuity between the nacelle and the motor shield. So I adjusted the code and printed again.

{gallery} Upper Mounting

7.     Assembly of Turbine

I made the motor housing oversize by a few millimetres as I wasn't sure how I wanted to retain it and also didn't want the fit too tight and end up splitting the plastic. I used PVC electrical tape to build up the body of the motor until it became a good friction fit. I drilled a hole for the motor wiring to leave the housing. If the wires hang down afterwards there will be less chance of water egress via tracking and I will also seal the hole with glue. I drilled out the turbine pilot hole to 1.5mm diameter on both sides. I used a pillar drill to ensure the hole was perpendicular and have the best chance of the turbine spinning without wobble. The turbine was a snug fit onto the motor shaft.

 

The lower bearing was planned to have a small brass pin but when I drilled the cross beams I realised I should have made the middle larger, with a squat cylinder. After drilling and inserting a short pin I added some glue to reinforce the area....then I let the glue dry.

8.    Test Time

8.1     A Quick Test

Obviously I was eager to see it in operation so before I mounted it in a small section of 68mm down-pipe I connected up my Tektronix TX3 and held the turbine under the cold water tap. Things were not as expected. I only managed 70mV-80mV. I guess the water and turbine was spinning much slower than those peak spins I could get from my fingers. Also as the water stream broke up and landed on other parts of the turbine it actually slowed it down !

 

I also realised I had fitted the turbine the wrong way, but the results were no better after reversing it. Perhaps it just starts spinning easier with the curved edge leading? Below is a photo from later after I reversed it.

8.2     A Better Test

Time for a wet outside test. The botton bearing was loose in the 68mm drain pipe so I wrapped PVC tape (blue of course) around to build it up. This made it fit well but then came my next problem...my hand doesn't fit inside that pipe to position it. Using some wire to suspend the bearing, as edge-on it falls straight through, I was able to use a wooden stick to level and bed it about 100mm down the pipe. Below is a view looking down the pipe.

I then placed the motor and turbine into the pipe. Placing the pipe over an outside drain I sprayed the garden hose into the pipe whilst looking at my DMM: it showed 0v.

After realising the turbine had jammed on the bottom bearing, and freeing it up, I went to a 'jet' test. On full jetspray I was only able to get approx 150mV DC.

What can I say .

 

My bearing design needs improving as there is too much friction. A hardened pin onto a slight centre-punched hole on a metal plate would be much better. Also the overall design needs changing so I can assemble the unit and then slide it into the pipework as a single module.

 

9.     Voltage Multiplier

I experimented building a capacitor-diode ladder to make a voltage multiplier but initial tests when spinning the motor by hand gave almost zero output. I spent a while probing the breadboard circuit I had made wondering why this didn't work. The capacitors were quite old and maybe the leads were slightly corroded? Was the values incorrect? I parked this section for a few days and then realised what a huge mistake I had made....my input signal was DC, varying wildly, but nevertheless still DC. These capacitor-diode ladders were to multiply up AC voltages.

 

I would need a Switch-Mode or Buck-Boost type circuit to get my measly ~0.1v up to over 5v. As the voltage could be sporadic I didn't really want to waste energy by driving a switching supply and also at 0.1v almost nothing could be powered up anyway.

 

I need to do some more experimenting here but have got to the end of the Project 14 monthly challenge time. I'm thinking about the Joule Thief types of circuits or transformer but still need that switching component: with less than 0.6v I don't think there is much to be extracted. Perhaps I could have used the turbine and a hall-effect switch to generate it? Another thought would be to use a DC motor that can generate more voltage i.e. a 24v design?

 

10.     Charging the Super Capacitor

I never got to play with this component, the first Super-Capacitor I have bought. At least I have one to experiment with now.

11.     Utility

In reality there are much better sources of wasted energy to get tapped into rather than the infrequent rainwater flow in a downpipe - light wind/breezes and solar will likely generate more energy over a longer period. But for extra fun....then I hope you enjoyed reading about the Trickle Charger. Such a turbine could be used where there is a fall in height of a stream but much simpler mechanisms like a water wheel would likely be a better choice.

 

12.     Summary and Thoughts

Sadly I believe I am almost out of time to the deadline and more so to my actual free time . I am disappointed that I haven't finished this project to the point it is fully installed in my house downpipe, whilst also very pleased with the success of OpenSCAD and the Lulzbot 3D printing quality. Actually I learnt a lot from the printing and really was amazed at the turbine script as well as the way the support structures can be broken out.

 

Maybe I need to add another 3D structure prior to the turbine, to guide water off the pipe's inner surface and funnel it into a single spout before it lands on the turbine.

 

I therefore have a few things I could still do:

  • Buy a 68mm square-to-round and a round-to-square adapter
  • Add a pre-filter to stop any debris clogging up the turbine
  • Upstream water routing/conditioning
  • Work out a suitable voltage multiplier that gets the DC to above 5v
  • Store that harvested energy !

 

I hope you have enjoyed reading my journey and ideas. Please comment if you have suggestions or need something explaining in more detail

 

13.     Example OpenSCAD Code

13.1     Top Mount

// Top Mount for
// Trickle Charger v1
// 14rhb
// An Element14 Project
// v2
// Nacel wasn't part of main print and detached !

echo(version=version());
bottom_mount();

//Variables for user modification
inner_radius=32;        //of the pipe it fits into
top_height=5;
mount_thickness=4;
cross_mount_width=3;
bearing_radius=0.75;
motor_shield_inner_radius=12;
motor_shield_wall=2;
motor_shield_length=40;
nacel_length=10;

module bottom_mount()
{
    difference(){
       {
       union(){ 

               
               //cross mounting
                union(){
                    {
                        cube([inner_radius*2, cross_mount_width, cross_mount_width], center = true);
                    }
                    {
                        cube([cross_mount_width, inner_radius*2, cross_mount_width], center = true);
                    }

                    
                } //union of cross supports
            
                //motor shield
                
                    //the outer ring
                    translate([0,0,(-motor_shield_length/2)])
                    
                    union(){
                        difference(){
                        {
                            cylinder(h = motor_shield_length, r1 = motor_shield_inner_radius+motor_shield_wall,r2=motor_shield_inner_radius+motor_shield_wall, center = true);
                        }
                        {
                            translate([0, 0, 0])
                            cylinder(h = motor_shield_length*2, r1 = (motor_shield_inner_radius),r2=(motor_shield_inner_radius), center = true);
                        }
                        } //diff motor mount
                        
                        translate([0,0,motor_shield_length/2])
                        cylinder(h = motor_shield_wall, r1 = (motor_shield_inner_radius+motor_shield_wall),r2=(motor_shield_inner_radius+motor_shield_wall), center = true);
                        
                    }//end union to cap off cylinder
                
            //the outer ring
            difference(){
                {
                    cylinder(h = top_height, r1 = inner_radius,r2=inner_radius, center = true);
                }
                {
                    translate([0, 0, 0])
                    cylinder(h = top_height*2, r1 = (inner_radius-mount_thickness),r2=(inner_radius-0), center = true);
                }
            }//diff for outer ring
            
            
            //And add in a nacel cone
            translate([0, 0, (nacel_length/2)])
            cylinder(h = nacel_length, r1 = (motor_shield_inner_radius+motor_shield_wall),r2=(0), center = true);
            
            
        }//union cross support and ring edge
        
        
        
        
        
    }//end of the bits I want !
    
    
    
    
    {  //Now define a cookie cutter to take the protruding edges off 
        //Outer cookie cutter
    difference(){
    {
        cylinder(h = top_height, r1 = inner_radius*2,r2=inner_radius*2, center = true);
    }
    {
        translate([0, 0, 0])
        cylinder(h = top_height*2, r1 = (inner_radius),r2=(inner_radius), center = true);
    }
    }//cookie diff    
    }
    
}//diff
    
    
}

13.2     Bottom Mount

// Bottom Mount for
// Trickle Charger v1
// 14rhb
// An Element14 Project

echo(version=version());
bottom_mount();

//Variables for user modification
inner_radius=32;        //of the pipe it fits into
top_height=5;
mount_thickness=4;
cross_mount_width=3;
bearing_radius=0.75;

module bottom_mount()
{
    difference(){
       {
       union(){ 
            
           //Cut bearing hole in cross support
           difference(){
               //cross mounting
                union(){
                    {
                        cube([inner_radius*2, cross_mount_width, cross_mount_width], center = true);
                    }
                    {
                        cube([cross_mount_width, inner_radius*2, cross_mount_width], center = true);
                    }

                    
                } //union of cross supports
                
                //bearing hole
                cylinder(h = top_height*2, r1 = bearing_radius,r2=bearing_radius, center = true);
            } //diff, cut bearing hole
                
                
            //the outer ring
            difference(){
                {
                    cylinder(h = top_height, r1 = inner_radius,r2=inner_radius, center = true);
                }
                {
                    translate([0, 0, 0])
                    cylinder(h = top_height*2, r1 = (inner_radius-mount_thickness),r2=(inner_radius-0), center = true);
                }
            }//diff for outer ring
        }//union cross support and ring edge
    }//end of the bits I want !
    
    {  //Now define a cookie cutter to take the protruding edges off 
        //Outer cookie cutter
    difference(){
    {
        cylinder(h = top_height, r1 = inner_radius*2,r2=inner_radius*2, center = true);
    }
    {
        translate([0, 0, 0])
        cylinder(h = top_height*2, r1 = (inner_radius),r2=(inner_radius), center = true);
    }
    }//cookie diff    
    }
    
}//diff
    
    
}

 

13.3     Turbine

I'm not including the code for this as it was made available on ThingyVerse by a user called Denise Lee. Her project is OpenSCAD Turbine Propeller Generator