I'm not familiar with Acura. But I suppose there's a knocksensor on it?
About differences in FE: the energy content of a fuel can be different, independant of the octane rate. Very hard to compare. Theoretically you should investigate in a lab what's the energy content of the different fuels you're using on the runs.

Yep knock sensor. Knock events start happening like crazy under heavy load and ignition timing gets pulled way back.

Under normal driving or cruise it didn't knock at all.

Usually when manufacturers use the same engine across cars, the only difference in mileage comes in with the overall weight of the vehicle.

That 1mpg difference is likely due to the weight and maybe some aerodynamic differences.

The good news is that, with newer (at least 2004) vehicles the new O2 sensors and ethanol requirements will allow you to drive lower octane. The trick is to drive conservatively for the first few tanks. The onboard computer is going to adjust the fuel injector timing during this time as "knocks" and emissions change due to the lower octane.

Of course, you should not go back to higher octane without first repeating the same process.

Good luck and post some mileage logs ok?

Ethanol has a higher octane than gasoline.
Thats not true with cars built in the last 10 years. Modern engine management systems adapt in real time. Long term fuel trim is no longer an issue.

A knock sensor requires knock to work. That kind of knock might not blow your engine up on the first day, but inadequate octane (causing knock) over a hundred thousand miles or more will cause head gasket problems, smushed crank bearings, etc. My point is that every engine should be designed to avoid knock in the first place, not to use knock as a tuning parameter.

If there are alternate ways to avoid knock (not just react to it), that's fine too. That BMW system sounds pretty slick.

-BC

Hi Bob,
I can't agree with your vision.
It's different: an engine should be designed to not have knock with the lowest required octane rate.
A knock sensor helps you getting the benefit of higher octane rate (e.g. my Civic min. requirement is 95RON, but I can put 98 also in it). This benefit can be lower fuel consumption or higher power.
On top of it: the knock sensor protects your engine for unexpected problems (carbon, very bad fuel, a bit diesel residue, too high temperature...etc) that can cause knock.
The knock captured by the knock sensor is not dangerous at all as it's already captured from the very first signs of knock and doesn't continue for long seconds.

I'm not saying there should not be knock sensors - I agree that there could be problems (fuel issues, carbon buildup, etc.) but for this the knock sensor functions as a safety device. I don't think that the OEMs should program too much timing into their motors and just hope for the knock sensor to find the best timing, like is how I believe things are now.

I just saw a Civic in the shop last weekend that had a head gasket problem with less than 100k miles on the motor. In fact I see lots of head gasket problems (i.e. the yellowish goo under the oil cap) on 100k mile cars, on cars that are otherwise in good condition. In my opinion there are way too many head gasket failures, and I attribute this to incorrect timing from the OEMs. Not to mention that I also hear knock when I drive cars, even though the "correct" octane fuel is being used.

And a few seconds of light knock won't hurt anything, but years of knock will. I say this because I see it too often.

I'm not exactly sure what you don't agree with - do you think it's a bad idea to have an economy (low octane) / performance (high octane) switch for the driver?

- BC

Well Bob, I don't agree when you say knock shouldn't be used as a tuning parameter. It's a wonderfull parameter to adjust your timing to the limit.
It's not fair to relate headgasketproblems to existance of a knocksensor. The very limited 'allowed' knock (before ignition is retired) by the engine ECU can never cause such problems.
The way people drive is far more destructive for the headgasket (high load with cold engine), just as a broken cooling system.
An economy/performance switch for the driver seems very inconvenient to me for 2 reasons:
1)Independant of the octane rate I may need power in some situations. Those situations can be overtaking a car, crossing a street... Dangerous and I don't want to depent on a switch.
2)Economy: I want to have my engine always as efficient as possible, no matter what octane rate, when I don't need power.
All this is done automatically by the knocksensor.

Bob, which brands do you observe more of that with, and which brands do you see less?

I'm going to guess that brands with 100,000 mile warranties have safer timing...they have something to lose by making their vehicles die early.

It's fine for an OEM to use knock (although a 5-gas analyzer is much more indicative of what's going on in the cylinder) to develop an initial tune. But once they've established an optimum timing table (which should not knock except under unusual conditions), then they can use that table in the production vehicles. They put thousands of man-hours into tuning these engines, and by the time the engines reach the consumer I consider any knock to be unacceptable (provided the consumer is using the required octane level).

By the same token, I also would like to see OEMs add some retard during transitional events (i.e. throttle movement, shifting, etc.). Maybe some or all of them do this now, but the programs I've been in don't consider those events for timing.

As far as what vehicles- recently I recall seeing a Northstar, a civic (the guy just bought it for a "bargain" unfortunately), a Volvo, and a Trailblazer with foamy oil caps. But since I only work at the shop one day each week, I probably only see about 7 - 10 cars each week, and I don't really keep a log of what vehicles have problems and which ones don't.

Headgasket technology and machining practices are so far advanced nowadays that cars nowadays should all last 200K+ miles without a major failure.

-BC