Gas Octane for my 2008 3.0si

[Mike Benvo]

When I said the car detects you are running 87, I didn't mean that the car knows exactly what octane you are running. It has no idea if you are running 87 or 100, but it can advance and retard from the preprogrammed map depending on input from the knock sensors and other inputs as well. If you go too far out of range, limp mode and other safety features will kick in.

Much of what you said is accurate but is beyond beyond scope. There is no reason to expand this to include the standard distribution of fuel quality. Lets stay within 87 octane and 93 octane which is a fair range for most of the cars out there. I never said the car is programmed to run a specific octane level.

The car will run on 87 octane just fine. It won't be optimal, but you could fill with 87 octane for the life of the car and it would drive fine. I merely made a statement about how the car EXPECTS fuel to have specific characteristics within a relevant range. It will only advance UP to what is programmed in the code, which is why adding 100 octane doesn't give you the boost that it could. Most BMW cars out there are pulling timing to some degree. I never said there was lag in switching maps, that is instantaneous. There is a definite time lapse in advancing timing, and a much smaller time lapse in retarding timing. This isn't due to map switching but the way that adaptations work.

You seem to be weighing on environmental factors more, and alluding that based on a certain set of environmental conditions, that 89 octane is fine. You will get more performance out of higher octane only to a certain degree. If you are driving around town and not calling for much ignition advance, it might run the preprogrammed map value just fine. It's only when you push the car harder, and run it in less unfavorable environments, that the benefits of higher octane/quality gas become realized. 89 octane is not going to cut it in all conditions, and neither is 91.

Put a car on the dyno, clear the adaptations, and watch it pick up 10+ wheel horsepower.

ard to answer your questions:
Q: Is it your contention that even with Knock Sensors you will get pinging? Actual, repeatable, pinging??? And that these knock sensors actually will not sense the knock and slam the timing so far that it will not ping?

>>Yes, you can get some slight pinging under certain circumstances. It won't be consistent as the car will pull timing and add it back as it sees fit up to the limited map value.

Q: Are you saying 87 will ping and the 'failsafe' knock sensing mode will not prevent it?

>> There is no "fail safe" knock sensing mode. It is always getting input from the knock sensors. The fail save mode, called "limp-home mode" or "emergency mode" is only activated if a sensor goes awry, or if there is a more severe condition it detects. If you clear the adaptations on a car, and run it hard with 87 octane in conditions which call for the use of higher octane gas (up to the preprogrammed timing limit), you will ping as the car removes timing. If it detects a misfire, the ECU will take it a step further. It's not going to ping ping and blow the motor, but you will get occasional pings as it adapts to the poorer fuel quality.

Originally Posted by JCL

Mike

Couple of comments. First, thanks for the reasoned response. This topic all too often degenerates into "If you can't afford 91 you shouldn't be driving a BMW" or something similar. I think the engineering and science is actually interesting here.

You say that the car "detects you are running 87" but I don't think that is the right way of saying it. The engine has no way of knowing specifically what you are running. What it can determine is that there is persistent knock being experienced, which could be caused by lower AKI, or higher temperatures, or altitude, etc, or a combination of the above. Since AKI is determined by comparing a fuel to a reference fuel on a test engine, a very artificial environment, and since different fuels of the same AKI can vary in their knock responses at low rpm vs high rpm, at full throttle vs partial throttle, etc, AKI or octane rating is not absolute. I agree that persistent knock can push the DME into a different map, but it is because of the sensors hearing the onset of knock, not because of the fuel in and of itself.

I understand your point that shifting between maps can come with a lag, probably due to time-based averaging, but I can only comment that I haven't heard pinging due to that issue. Even if there is a lag switching to an alternate map, that wouldn't cause the knock sensors to stop working. It may cause the engine to use a less than ideal map in terms of efficiency while the DME adapts.

You also say "BMW has recommended 91 as a minimum because they know that anything less will cause knock in unfavorable conditions". I don't agree with that, because AKI or octane rating is not an absolute. In the real world, assume that fuel qualities overlaid on varied engine conditions will result in the onset of knock according to a bell curve or Gaussian distribution, ie lots of engines will experience the onset of knock at a typical AKI, fewer will experience it at a lower AKI, and fewer will experience it at a higher AKI. It is no good designing the engine to run precisely on 91, because in that situation, half of the population would have knock (all the ones on the right side of the bell curve). If BMW wanted to make sure that no owner ever had knock, they would specify a fuel way over to the right side of the curve, so that all (or nearly all) were 'safe'. But that would mean that most owners were paying for fuel they didn't need to pay for, ie cost without benefit. So they pick a spot a little over to the right, good for most owners most of the time, and that turns out to be 91 AKI. The engine is actually designed to run on something less than that, ie the mean or peak of the bell curve, recognizing that there are a range of fuels sold as 91, and a range of operating conditions.

So, some may have problems on 91 despite the best intentions of all involved. Many will run on 91 forever, with no problems, and no need to think about it beyond choosing the 91 nozzle when they fill up. But some of us checked our local fuel, and found that the engine ran every bit as well on 89 (my example) or 87 (others). They aren't wrong, they are just fortunate to have a combination of local conditions, fuel quality, and engine condition that works fine on that AKI octane rating. My engine was new when I ran it on 89, and it may be that at high miles, it would have required a higher AKI.

Finally, in my experience, soccer moms can experience pinging at least as often as other drivers, due to the use of a light foot on the throttle on a long grade.

Jeff

[JCL]

I wasn't talking about fuels anywhere beyond 87 and 93. I was focusing on fuels around 89 and 91. Below 87 isn't necessarily safe, and above 93 can't provide any real world benefit to a stock engine.

Due to variations in real world fuel quality, BMW can't design an engine for 91 and then specify 91, simple as that. You said that BMW specify 91 because they know that less than that will result in knock in unfavourable conditions. I was just pointing out that that isn't necessarily true. BMW design it for less than 91, and by specifying 91 they reduce or eliminate complaints from owners. If they designed it for 91, half the owners would be complaining that their vehicles didn't operate well on that fuel.

My point is that a specific local example of something that is called 87 or 89 can work fine, and can in fact be optimum, but not for all owners. It depends on a multitude of variables, including local fuel quality, engine condition, and ambient conditions.