Qest T. Silverclaw said:
So it's a pain in the butt, destroys my engine (almost certainly with a voided warranty for doing this) and can kill the paint if I'm not careful? I'm gonna pass.
I agree.
p220sigman said:
I wouldn't try it. I always fall back to if it was that simple, the car manufacturers would already have an acetone injection system.
... or the oil companies would put it in the gas as a premium feature. Like all easy answers, this is one is not valid.
From the linked site:
Acetone said to improve the fuel's ability to vaporize completely by eliminating the surface tension that causes an increase in particulate vaporization temperature. (PESN; March 18, 2005)
A growing number of people are reporting their results, as tabulated here. Most have noted modest increased mileage (e.g. 2-15%)...
If 2% to 15% of the fuel were not vapourizing and thus not burning the engine's emissions would be horrendous. If the stuff works, a different mechanism is at work.
Again from the linked site:
Acetone is a solvent that can be used as an additive for boosting octane levels in gasoline.
There you go; a plausible explanation. If a vehicle with a modern engine management system is fed fuel which is inadequate in octane level, it will adjust (such as by retarding ignition timing) to prevent preignition and suffer in both power and economy. Correcting the octane deficit thus improves power and fuel economy. Octane requirements - and thus the potential for improvement - obviously will vary by make and model, and even by individual vehicle. Is acetone the best way to boost octane, or even a rational option?
I did not read all (or really any) of the anecdotal evidence (since I have no intention of adding acetone), but if I were to look for useful experimental results I would want a comparison in rigourously controlled conditions of regular gasoline, premium (higher octane) gasoline, regular with a commercially available octane-boosting additive, and regular with acetone at multiple concentrations. Anything less is playing, not research. Two consecutive runs of 100 miles on real roads (for example) are unlikely to produce consistent results within 2% even with identical fuel; they cannot distinguish the effect of a fuel change to that precision.
This is all, of course, just my opinion. In case it is not evident here, I'm not a fan of snake oil.
