About to finish my first tank of 89 octane mid-grade. No performance differences or sounds but I got 20 miles for this tank and low fuel light hasn't come on yet. I'm sold for now...

 

I believe that isn't exactly accurate. The computer doesnt know what octane fuel is being used; it knows to recognize 'knocking' or 'pre-ignition' of sorts. Lower octane fuel burns faster and easier than higher octane, thus is more prone to pre-ignition. If somebody uses lower octane fuel, the computer may sense more 'knocking' and then retard timing, etc, to keep the engine safe. But to say the computer retards timing because of the octane rating of the fuel is innacurate. If an engine is being fed 87 octane, and there is no pinging, the car will run perfectly, at full timing advance allowed by the ECU.

 

DMP: Absolutely true. There is no "octane sensor" that looks at the gas. Only the O2 and knock sensors working together to detect premature ignition (aka "knock").

 

I burn 87 octane ethanol reformalated gas all winter. No problems flooding starting/stopping cold etc. Seems to run fine. I tried 93 octane and the power seems the same. I'm hopping ethanol keeps the engine cleaner. Anyone know?

 

Sorry for the "non-technical" explanation. But you are correct that it is based on detonation. It's not like a car can sense octane #'s. However the ECU in many of today's high performance cars will detect pinging & detonation long before the person driving the car would & adjust faster (retarding the engine) than we could. You might not feel the retardation of the timing but the loss of hp is there. If I took away 10-15 hp from this car, I would hazard a guess that not many people would be able to know that with a "butt dyno." In normal around town driving it will pretty much be unnoticeable, but would come into play in high performance/track driving. Similar to what happened to the M3 in the article.