Scooterbine: The Next Chapter

Renewable energy from unwanted, unloved motors, and a mean scooter.

Similar projects worth following
Living in Wellington, there is a lot of wind to be harnessed. There are also a lot of unwanted electric scooter motors from the numerous electric scooter hire businesses. Surely these motors can be used for something a little more useful than making an impressive mountain of ewaste?

This is where the Scooterbine comes in. I got my hands on a number of unused scooter motors, wanting to use them for energy generation in some way. With my house being up on a hill exposed to a lot of strong wind, wind power seemed like the ideal way to do this.

Currently, I have worked out a process for analysing the avaiable wind resource and developing an effective rotor. However, energy output is minimal (would you like to power a single LED? I've got you covered).

Huge thanks to Daniel (One of - if not the - best lecturers), and to my dad for supporting me to get this project to where it is. You guys are awesome!


Hey, thanks for checking out my project! As I said in the description, this project aims to develop a low cost method of recycling electric scooter motors into a small scale home energy generation system.  This section will provide a brief outline of where the project is at, and where it is headed. Project logs provide technical details on individual elements.

Before I get into the project details - I just want to elaborate on the thanks in the description. Thank you so much to the people that helped me get this project to where it is. Thanks to Daniel (One of the best - if not the best - lecturers) for not only giving me the push I needed to get it made, but also for talking through a lot of the technical details, especially when I was deep down the fluid dynamics rabbit hole. Thanks to my dad for being there with some really solid practical help and advice - I think that the tower would have fallen over the first time I tried a live test if it hadn't been for some of the rock solid mounting methods you came up with. Another big thanks to the engineering techs at my university, for some more solid advice.

As it stands, the project breaks down into three major elements:

  • Development of blades and rotor suitable for harvesting local wind
  • Development of generator capable of maximising energy output for local wind conditions
  • Development of an electronic control system for braking, and for handling energy output into a battery bank or local load

Currenty I have put a lot of time into the first bullet point, and developed a methodology for creating a rotor suitable for specific wind conditions. This process has also highlighted the fact that the scooter motor may not be suitable for operation by a wind generator, as in order to ensure a low cut-in speed the operational RPM of the motor must be reduced to a point where output power is minimal.

Blade Development

In the final trimester of my engineering degree I worked on this project as a self directed study course, giving me a lot of time and resources to develop elements of the system. During this trimester I worked on developing a methodology for analysing local wind conditions, and developing a rotor to those conditions. 

This process involved a lot of iterative design, and avoided in depth fluid dynamics (as I was studying electronic engineering, mechanical was a bit beyond the scope of my course). In order to do this, I used a software package called QBlade to handle fluid simulation. This was paired with research into airfoils developed for small wind turbine applications, and resulted in blade profiles targeting specific tip speed ratios (TSR). 

Further detail is provided in my blade development project log (and in the report I wrote for this section, found here), but if you just want to see the rotor operating at maximum speed (With no generator gearing attached), check out this video:

Generator Development

Part of the self directed study work looked into details of the scooter motor as a generator. The brief amount of research into this topic is detailed in a log, however in summary the research found that wind generators (and generators in general) tend to have what is effectively a maximum power point on their power vs RPM curve.

Additionally, it was found that the scooter motor needs to run at a reasonably high RPM to begin approaching this point. Further analysis of the motor power rpm curve could provide insights on exactly where on this curve the scooter motor should be operated for maximum efficiency.

However, this produced another issue. The turbine can drive the generator at it's maximum speed, however it cannot self-start. In order for the turbine to self start, a low gear ratio needs to be introduced. This lowers the starting torque of the generator, reaching cut in speeds  as low as 2.1m/s. 

So a low cut in speed was achieved, but another issue reared its head. With a low gear ratio, the resulting...

Read more »


Spreadsheet for calculating blade shape. Incorporates equations from Wind Energy Explained to provide parameters for modelling each airfoil into a blade for fabrication.

spreadsheet - 48.95 kB - 04/30/2022 at 04:19



This is the first report I wrote for this project as part of my self directed study. It covers the details of my wind analysis process, and the methodology I have used to develop an effective rotor.

Adobe Portable Document Format - 8.30 MB - 04/30/2022 at 01:21



This report is the second report, and covers the process of integrating the rotor with the scooter motor as a generator. It covers the nacelle construction, and the method of selecting gear ratio in the drive train.

Adobe Portable Document Format - 9.67 MB - 04/30/2022 at 01:22



The final report written for my self directed study. This report covers the testing process I went through, detailing how well the rotor behaved and the issues encountered with the generator start up torque.

Adobe Portable Document Format - 734.88 kB - 04/30/2022 at 01:22


  • Integration & Testing

    Sam Griffen04/30/2022 at 22:50 0 comments

    This project log summarises the work in the Integration Report and the Testing Report. For more detail on any of the stuff in this log, check those reports out.

    With a rotor design complete, it was time to work out how that rotor was going to be mounted to the fenceline, and how the drive shaft was to be linked to the generator. Figure 1 shows the CAD model of the nacelle and tower. This OnShape model is linked to on the main project page. 

    Figure 1: Nacelle model

    The rotor discussed in the previous log clamps onto the drive shaft. There are then two steel braces holding the driveshaft in place with two bearing blocks. These braces are bolted to a baseboard, which is connected to the main nacelle platform. This platform is a recycled scooter, with the main shaft of the scooter sliding into a steel tower pole that is firmly mounted to the fenceline. With this bearing sitting at the top of the pole, the turbine can be designed to have any desired range of motion, to allow for wind tracking. 

    In the video above, this setup can be seen. The rotor is connected to the driveshaft, and the nacelle is mounted in the tower post. As this was an initial test of the rotor there is no link between the driveshaft and the generator.

    Rotor Velocity Testing

    Calculations in the Blade Design Report indicated that for a tip speed ratio (TSR) of 7, in an average windspeed the rotor would operate at around 700RPM. Figure 2 shows an analysis of the audio spectrum of a rotor test. Highlighted in red is a major frequency band that was only present while the rotor was operational, and died off when the rotor was stopped.

    Figure 2: Spectrum analysis of operational turbine

    Figure 3: Rotor speed calculation

    Given three blades on the rotor, it is reasonable to imagine that this frequency will be three times the operational frequency of the rotor. With the band at approximately 35Hz, calculations shown in Figure 3 found the rotor RPM to be approximately 700RPM. Having this number align with the theoretical model was pretty exciting, it indicates that the blade development methodology has some weight behind it.

    Generator Integration

    Once the rotor was shown to effectively self start, and spin up to expected speeds, it was time to integrate the actual generator and look at whether power could be effectively extracted. A set of sprockets and used chains were obtained from a bike shop, and my plan was to use these to modify the startup torque required from the rotor, and the operational speed of the generator. 

    Figure 1 roughly shows how these were to be set up in relation to the rest of the setup. 

    Figure 4: Drive shaft to generator drivetrain

    Figure 4 gives a close up image of the developed gear train. A number of chains were developed to allow for testing of a range of sprocket configurations. 

    Testing & Results

    Initial testing indicated that for a TSR of 7, the startup torque required was too high and could not be generated by any of the fabricated rotor configurations. There is detailed discussion in my Testing Report on this topic, but the result was a shift to a TSR of 5, and a wider root section. A lower TSR results in a lower operational speed (Approximately 500RPM) allowing for a lower gear ratio, and a wider root section will increase the generated starting torque. 

    Testing then pivoted to analysing the impact of gear ratio on the cut in speed for this TSR 5 rotor configuration. Figure 5 shows the results of these tests. With a ratio of 0.44 or below, a cut in speed below 3m/s was achieved. This was important as the NIWA data obtained indicated that with a cut-in speed between 2 and 3m/s, 85% of the available wind could be effectively harnessed.

    Figure 5: Gear ratio tests

    In terms of power output, there is detailed discussion on the small amount of power that was produced in the report. In summary, it wasn't much. At most the rectifier output 5.8V, producing between 0.2W and 0.6W. This is obviously far from ideal, given that the maximum...

    Read more »

  • Blade Development

    Sam Griffen04/30/2022 at 03:32 0 comments

    As this project was accellerated by working on it as a self directed study during the final year of my degree, I have written reports detailing the process. This log effectively just summarises my Blade Design report, found here.

    One of the most important aspects of a wind turbine is the efficiency with which energy can be extracted from the wind. A large part of this efficiency depends on the aerodynamic properties and configuration of the blades. It is therefore critical that any turbine design considers the impact of all relevant blade parameters, and implements some process to optimise aerodynamic efficiency.

    For the design of this turbine design there was one location suitable for installation. The goal of this setup is to provide power for a small greenhouse, and the fenceline adjacent to the greenhouse is exposed to a large amount of wind. There is no other really suitable location for a turbine on the property, and therefore there wasn't much opportunity to analyse various positions.

    The images below show the location for installation:

    Figure 1: Turbine install location, looking out over the valley

    Figure 2: Turbine install location

    Analysis of the wind resource in this location couldn't be completed in detail, so NIWA data has been used for the local area to estimate an average windspeed. This analysis showed that the average windspeed is approximately 7m/s, 

    Figure 3: Local windspeed distribution (From NIWA)
    Figure 4: Calculation of maximum power available (Based on the Betz limi

    With this average windspeed analysis completed, the theoretical absolute maximum power output was calculated as 145W (Figure 4). This is significantly more than the greenhouse would need to operate, which is good as the turbine is unlikely to achieve a higher output than 50%-60% of this figure due to losses in the system.

    Once I had a model for rough local wind behaviour, I selected a range of airfoils designed for small wind turbines, and airfoils that have been recommended for small wind turbines. For further detail on airfoil selection, refer to the full report, pages 5-6. 

    Figure 5: Catalog of airfoils to be analysed. Screencap from my "Blade Design Report"

    Each of these airfoils has been recommended as a turbine airfoil, however it was not feasible to fabricate a blade with each to test. Therefore a simulation package called QBlade was utilised to produce drag polars and glide coefficient vs angle of attack plots for each airfoil. These results are shown in Figure 6. 

    Figure 6: Drag polars for each selected airfoil

    The first plot is referred to as a drag polar. This plot shows a relationship between coefficient of lift (Cl) and coefficient of drag (Cd). In the literature there is discussion of a wide "drag bucket" being a desirable trait for turbine blades, which leads to an minimal drag coefficient over a wide range of conditions. For more on this see page 6-7 of my report.

    Figure 7: Relevant blade terminology

    In the second plot, there is a more useful representation of airfoil behaviour. This plot shows how the glide ratio (Cl/Cd) changes with angle of attack (Alpha). Figure 7 shows how angle of attack is defined (for more detail see page 1 of my report). With these values in mind, the second plot shows for what range of relative wind vectors the blade will have maximum lift, and therefore produce maximum torque.

    Airfoils with a high peak are good as they represent a larger lift output for the given windspeed. However, it is important to acknowledge the fact that a wide peak is almost as important. A wide peak means that the airfoil will have an optimal glide ratio (and therefore optimal lift) for a wider range of relative wind vectors. This is desirable, as the relative wind vector will change based on rotor velocity, which means that for varying windspeeds the turbine will have good energy output.

    Figure 7: Airfoil models developed for testing

    Of the airfoils...

    Read more »

View all 2 project logs

Enjoy this project?



Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates