Being a mountainboard setup with pneumatics, it’s pretty much gonna be worse compared to PU tires, always.
If it’s a really warm day and I don’t gun it hard, I can get 14Wh/km (~30km/h, 18mph), which I would say is decent for pneumatics and terrible in comparison to thanes. During the winter, with thicker air due to the cold and much more of a washboard road surface I getting around 22-23Wh/km with the 8" numies. The thing here is that it’s a long distance journey and bad road surface is very likely to be found at some point and I don’t feel like destroying my feet.
Temperature has a big effect on the overall efficiency, when it’s warmer the belts are able to deform and make the bend around the pulley easier, the tire’s rubber more easily deforms against the road surface and doesn’t loose as much kinetic energy there, air is thinner, battery pack is warmer, so it has lower internal resistance.
This is a great deep dive into electric motor efficiency. 41:10 he goes in-depth with the motor efficiency calculations and factors.
But essentially the reason why the motors are more efficient at lower rpm can be explained with an example like so:
Motor X, needs 1 Amp to be able to turn itself, this would be called the no-load current, it’s the torque needed to defeat the bearing friction and stator cogging torque.
So say, you output from your ESC a 1V RMS voltage to the motor phases and it spins up to whatever speed it does and uses 1 Amp at 1 Volt, so it doesn’t do any work, but it uses 1 Watt of energy just to spin the motor itself.
But say you output 50 V, as that’s your battery voltage, so the motor is now spinning much faster and the applied voltage is 50 V, but it still needs that 1 Amp of current to be able to defeat the internal losses, so now it’s using 50 Watts just to run itself and it’s only because it’s spinning faster.
Now let’s apply this logic to a imaginary esk8 example. You skate at a steady speed of 20mph and your drivetrain is optimized for high torque, so say the motor is spinning at max duty cycle and you have a 50 V pack and you know that the motor needs that 1 Amp just to spin itself, so the total power needed is whatever power you need for your board to defeat the rolling resistance, drivetrain losses and air resistance + 50 W, just to keep the motors also running. Let’s say you need 400 Watts to go that steady 20mph, so your motors are operating at 400W/450W = ~89% efficiency
You then optimize the gearing ratio to have the motor spin slower at the wanted steady speed and sacrifice some torque. Your motor now turns half the speed it was before, so it now only has 25V on it and it therefore now only needs 25 Watts to spin itself, so now the efficiency rises as you still need the 400 Watts to keep your speed at 20mph, but the motor itself doesn’t need so much power to spin itself the efficiency rises to 400W/425W = ~94%
Hopefully you understood the idea of lowering the motor rpm and why that can be used to raise the total motor efficiency. There are also other factors, when the motors run slower, but you need the same power as before with lower current, this means your motor current will be higher and more power will be lost in the copper of the phase windings due to resistive losses. I would also expect that the belt drive might be a little more efficient as the belt speed lower, so the belt needs a little bit less energy as it deforms less often, which takes a little bit of energy. Also as the motor spins faster you will start to get more losses in the rotor itself, called hysteresis losses. The stator also exhibits eddie losses as the winding’s cause the electromagnetic field to flip around and the internal structure of the silicon steel laminations need to flip it’s internal magnetic path.