,

 

I want the fans to idle at 10% duty cycle at room temp, and max out at 18v @ ~40C

That 555 circuit will only give a small change in PWM duty-cycle for the expected change in thermistor resistance (10k to 5kΩ), not give your desired range of duty-cycle.
Would you consider a circuit using a LM339/393 comparator circuit instead, which can go from 0% to 100% duty-cycle?

Then we can talk about voltage.

 

I took a different approach with an almost identical circuit.
I put the thermistor where the 0.1uF capacitor is, and then two fixed resistors above it. That gives you a standard hysteresis control, so the fan is either on or off.
Brushless DC fans don't take kindly to a low supply voltage, they don't rotate, but the internal circuit takes current and generates heat in the motor, which is effectively stalled.
If you choose the "fan on" temperature and the "fan off" temperature, you can calculate the two resistor values.
If you have 24V fans, you need a 24V supply. You aren't going to get 18V from a 12V supply (easily).
Is that a Victron MPPT? The colour is unmistakable!

 

I'll build the suggested circuit next then, thanks for the suggestion.
Hands on projects like these are the best way to learn IMO
Didn't know that the NE555 was limited in such a way - I do now!

I wouldn't recommend it - for a fan with a brushless motor, it won't start until the PWM gets to about 50% and below that it will sit the with the motor effectively stalled. Use a standard hysteresis controller, unless you can figure out a way that the PWM never goes below 50%.

 

Given that you have 6 fans, if the LM3914 still existed, I'd use its outputs to switch on one fan at a time.
In the absence the LM3914, I'd use a LM339 to switch fans on at different temperatures, so you can have from zero to six fans running depending on the temperature. I suspect would be rather more efficient than attempting to speed-control the fans.
By the way, you should blow cold air into the enclosure, rather than blowing hot air out. The fans run cooler that way and last longer.
No point at all in running the fans at night!
Another thought - if you switched the fans according to the current not the temperature, you would preempt the heating - you would start cooling before the heat from the buck regulator made its way through the heatsink.

 

Will consider multiple stage operation, that's a good idea I had not considered.
Efficiency is no problem here, primarily want to extend longevity of hardware.
The 6 fans maxed out @24v/8ma. could only ever pull 11.52w
If I shut the solar off and ran them independently (theoretically) I would get somewhere in the neighbourhood of 130 hours / 5 days of runtime off a charged battery - suffice to say probably double that If I ran them at 18v.

the 3 fans in the image are intakes, the other 3 are exhausts.
They hit the SCC and the 3 exhaust fans take the hot air from scc and the inverter, it's goes left to right. takes about 3 seconds for e cig vapour to pass through (did a test)
Victron recommend 100mm space around the inverter however, this active cooling system should circumvent the need for it. But it is definitely necessary unless I prop the door open, if more than a few 00 watts is being used.

I'm surprised such a circuit doesn't exist prebuilt, have looked but to no avail.
Perhaps I am not looking hard enough?
I'm determined to get this working as I want it to but didn't think it'd be this tricky.

I would suggest putting all the fans on the intakes, and just some vents on the exhausts.
You would get much longer life for a fan that is not powered up!