Fan voltage is a funny thing.
I use 24V for my printer and have several cases where I use 12V fans.
In most cases, I wire 2 identical fans in series. While this is perfect in theory, the reality is that one of the fans burns out before its time, since they aren’'t identical enough.
For a single fan, 50% duty cycle is 200% power dissipation in the motor coils. To have about the same power, a 24V fan has approximately 4X the DC resistance of a 12V fan. Because power = volts X amps, if you double the voltage, given the same resistance, you also double the amperage. This results in 4X the disipated power.
If you double the resistance, you keep the same current, but at double voltage, you’re still double power. Power dissipation is what burns out those motors.
So, for a single fan, it’s a cap of 25% duty cycle that will give the intended fan life at double voltage. It also gives the intended motor power.
This gets more complicated if you use an RC network to buffer the motor, since a 50% duty cycle can then be used to produce a constant half voltage. The fact that the motor is an inductive load also changes this a little. In short… It’s complicated.
In reality, you’ll have to figure out where the actual power setting is for any given model of fan. It’s tricky without using a scope to measure actual power in the circuit. Expect that you’ll go through fans much quicker than normal. I would suggest that you try it with a max duty cycle of 25% (safe) but expect that by about 35% you’re going to start losing fans.