On my 01 E320 I have always run 91 and I track every tank of fuel. Several months ago I decided to try running 87 in it. I ran it on 87 for probably 2-3 months, then ran it for about a month or 2 on 91 again. I can tell no different in power and MPG is the same. I am getting anywhere from 25-28mpg with both types of fuel. I have talked to some folks who run these only on premium and others say it is perfectly fine to run them on 87 but you might have lower power and less mileage. But, as I tracked for myself, power seems the same and mileage is identical. Anybody else tracked this and had similar experiences?

if you're easy on it you will probably be ok.

I would run premium only but then again I have a heavy foot sometimes

Here's a real time knock test on different fuels on a m113 (not mine):

https://www.youtube.com/watch?v=iWl4h-HmExI

I own 2 M112s (E320 wagon C320 wagon) and run both on 87 octane. Here's the fuel consumtion logs.
Big Bertha (Mercedes-Benz E320) | Fuelly

The ugly one!!1 (Mercedes-Benz C320) | Fuelly

I tried 91 octane on the c320 initially and switched back and forth a few times between 87 and 91. I saw no difference in FE or acceleration. I also monitored real time ignition timing with my icarsoft i980. On 87 octane it only pulls timing in hot weather (90f+) and full throttle acceleration. In all other situations ignition timing is exactly the same. So I run both cars on 87 octane. The only time I put 91 octane in either is if I'm driving in extremely hot weather through the mountains. Think Tejon pass in August. Or if I'm pulling a trailer with the e320 in hot weather. The other 99% of the time it's 87 octane. The E320 has 181k miles and the previous owner also used 87 octane. The C320 has 170k miles, no idea what the previous owner used. But it was a $1100 car. I doubt he bothered with 91.

EDIT: I also did some 0-60 mph tests with the C320 on 87 and 91 octane in cool weather (50-60f). It ran the same 6.7 sec with both fuels.
Yeah I saw that video and tried the same test with my Icarsoft on the C320. It didn't pull timing like that on 87. I think he let the engine heat soak at idle for a long time before accelerating in order to get that result. Thats the only way I was able to replicate his result. In fact I can get it to pull timing like that on 91 octane as well if it idles for a couple of minutes. You can predict if it's going to pull timing based on the IAT sensor live data. It can get up into the 150f plus range after a bit of idling. In normal driving IAT hovers around 5-15f above ambient temperature.

__________________
CENSORED due to not family friendly words

I did not think about it making a difference between summer and winter temps. I will have to look at my records but I am pretty sure I was running 87 during winter months.

I have a friend of mine who used to work for IBM. He was a guy that made databases for everything in life. He has a 1992 300SL M104, and of course he made a data base of all his trips. His conclusion was miles per dollar would remain the same weather he used premium or regular. I'm sure that would not hold true today at lease here in the states with the current super wide difference in the cost of premium and regular. I bet hes using regular right now.

__________________
To see my 129 parts for sale visit:
http://stores.ebay.com/The-Mercedes-SL-Store
John Roncallo

Here is more food for thought. Not advocating one way or another for using 87 over 91. The compression ratio on the 112/113 is 10:1 I believe. Just because I am curious I looked up the engines that GM puts in the Sonic and Cruze (we used to have several of these as company cars) and they call for 87. Those engines are 10.1:1. I just wonder how it is that GM engines use 87 with the same compression ratio as the MB engines. Maybe they leave their timing retarded all the time.

Detonation depends on both CR and combustion chamber geometry (and other issues such as inlet air temperature/pressure and coolant temperature), plus different manufacturers have different philosophies, and they're always concerned with warranty claims.

Modern cars have detonation sensors with multiple spark advance maps. If an engine sees a lot of detonation due to low octane fuel it will default to a less aggressive map. This will usually result in less low end torque/torque power that most users probably won't notice and maybe high fuel consumption due a less than optimum spark advance map.

Some owner's manuals might say "91 recommended" rather than "91 minimum", but I doubt if most recognize the difference in these two statements.

Bottom line is that most modern engines will operate fine on 87 PON despite what the manufacturers say, but it's up to individual owners to test and determine at least SOTP any difference in performance or fuel consumption and listen for detonation.

Duke