Originally Posted by JCL

When an engine manufacturer specifies a certain AKI for its engines, it has tested the engine to perform adequately a large percentage of the time, under a variety of operating conditions (temperature, altitude, etc) using fuels that are represented to be 89, 91, etc at point of sale. In reality, fuels range widely across North America in terms of actual delivered properties, as do operating conditions.

The best way to think of it is as a normal distribution. Population on the y axis, required AKI on the x axis. The points making up the shaded blue area represent a variety of owner actual requirements for AKI ,in order to minimize the DME having to retard the timing too often, because of all those variables. BMW needs to make sure that they cover the full range of variables, as much as they can. They don't like getting "low power" complaints. And they don't really care that the 91 fuel costs more. So let's say they specify the requirement (91 AKI) where the red line is, at one standard deviation. That means that for 84% of owners, the DME won't need to intervene, and the owners will get full performance out of their engine. But that also means that most of the owners (by definition.....) are buying a higher AKI fuel than they actually require, to cover all the eventualities.

Your engine doesn't know what AKI it is burning. It has no way of telling. I don't know that you can even measure AKI directly; I recall running test fuels in a single cylinder laboratory test engine at very low speeds and then running a standard fuel, and if they knocked at the same point, then that was the AKI of the fuel (or RON, or MON, or whatever).

All your engine knows is if it needs to retard the timing. That is why 89 will work for most people, because most of them are on the left of the red line. It is also why you buy 85 octane at high altitudes, and it works fine. Thinking that 91 is some sort of magic requirement is wrong. It is a safe operating practice, just as changing the oil at 15,000 miles when the lights say to do so is a safe operating practice. But some would argue it is less than optimal.

Do what M5James did. If it runs crappy on a certain fuel, get a better one. He ran his 740 on 87 and it fell on its face. Good reason not to run 87 in that engine under those conditions. But if it ran fine on 89, then there would be no reason to put 91 in it except to satisfy the original risk management strategy of the BMW marketing department, who spec'd it to account for the widest range of variables that their engines would see in the real world.