The other day, one of my friends commented in our bike chat WhatsApp group about his flat pedals and how they were "quite stiff". He was wondering how much power this would be costing him.
I told him that if he was curious enough to spend a few minutes looking into it, he could measure the resistive torque and calculate the associated power loss from that.
I did a back-of-the-envelope calculation of the power cost, based on 80 rpm cadence (see below).
The resistive torque he can feel when turns the pedals is either coming from either poor bearings, or more likely (as the pedals were new), from stiff seals.
I told him he could measure the resistive torque by hanging a weight off the pedal and measure the distance from the pedal axle. Progressively adding weight until the pedal turned would then give a resistive torque (weight in Newtons multiplied by moment arm). Multiplying by cadence in rad/s then provided the power lost for one pedal. Multiply by two to give total bike power loss. Strictly speaking, what's measured with this method is stiction, or rather the resistive torque due to stiction, whereas we really want the friction losses when the pedal is turning. That's more difficult to measure simply though. I think the stiction torque would give a conservative (slightly high) estimate of the power cost, which is good enough I think.
The results? He found he needed 161g of weight at a 5cm moment arm to turn the pedal. He calculated this to be 1.35 Watts of power loss using my equation below. So not significant, but not nothing either.