2005 GT.

Bought the car with a current emissions test. Drive it about 200 miles from purchase, got down to a 1/4 tank, refilled it with cheap, 87 octane gas.

20 miles later, emissions instrument light went on. Haven’t pulled codes yet.

Advice and comment?

Don't use 87 octane. 91 octane or higher is required not recommended.

For now you're stuck with a tank of low octane. Two choices risk driving with it or have someone drain it. The drain option isn't so simple.

To ensure no damage while driving with low octane: easy acceleration no drag stripping.

We in the UK have 95 or 98.5 premium, in the US I believe its 93 premium.
Depends on how much fuel you stuck in your tank, id fill it up with 91 or higher as 1eapplebaum suggests at least then you may not distory the engine.Then keep filling up with the right fuel as the tank empties.

Maybe some octane boost??

Engine is running fine, is it possible the emissions light is unrelated? I generally don’t believe in co-incidences, but maybe?

i don’t see why the octane of the gas would affect oxygen sensors and whatnot.

I admittedly am guessing. Performing a scan will reveal something more substantive.

My premise is based on the knock sensor detecting x amount of knocks as a result of low octane. Indeed the vehicle should run adequately. The knock sensor will detune the engine to protect itself. The knock sensor however is not a tool for justifying low octane gas.

There should be no relationship with the tank of low octane fuel and the CEL. The ECU(s) are designed to adjust to that condition. Power will be reduced (might not even be noticeable at the seat of the pants driving level). I hope the codes are published here.

The ECU(s) use two timing tables. The primary table is maximized for the recommended octane fuel. In the US that should be 91 (the Euro standard is different so 95-98 RON equals 91 to 93 US octane). The second (secondary) timing table is created to maximize timing with lower octane fuels such as 85/87 US octane.

As Mr Applebaum stated, when the knock sensors sense knock (from a lower octane fuel, if used), the ECU immediately reduces timing by leaving the high octane table and re-selecting a timing for the same given cylinder air mass and RPM from the low octane table, but for a value somewhere in between the two tables, enough to eliminate the knock. As the low octane fuel continues to cause knock, the ECUs continue to adjust until they end up on the low octane table. This will continue until the low octane fuel is gone and is replaced with a higher octane. HOWEVER, the return to the higher octane table takes a far longer time than it took to get to the lower one. It could take as much as 50 to 100 miles of driving with the refreshed higher octane fuel. The engine will protect itself from knock, but reduced timing is not as immediately harmful, just power robbing, so the ECUs start this LONG journey increasing timing ever so slightly, while not seeing new knock from the sensors until they find themselves back using maximized timing for the now correct 91 or 93 octane US fuel. One could immediately intervene by deleting LTFTs, which would have the same effect as the ECUs starting over with STFTs to reestablish LTFTs, but just drive it and refill with the correct octane, maybe a couple of short fill times to blend up octane.

Here are a couple of pics from a different engine (and one that I personally tuned). The top table is the high octane table, the bottom low octane. I had to clip part of the lower air mass rows to get the pic in, but you can match a given spark air mass with a given RPM to see the difference in timing the ECU will use to optimize what you've put in the tank. Bottom line, although it will not harm anything, it is not optimal to have the ECUs constantly responding to knock and constantly reducing timing, living on the lower table.







I would like to have the invasive SW for Bentley that I've used for tuning other cars, but a little lost finding it in the marketplace. I know there is at least 50 hp left on the table by Bentley (common with car manufacturers for engine preservation). I know the SW exists, just cannot find anyone willing to sell it.

Thought I'd add a little more about the parameters the ECUs use to adjust for knock (retard). It gets a little too detailed, but the engine management design allows the tuner to adjust the "learning" curve the ECUs use when knock occurs. The tuner can adjust how many occurrences and how big the spark knock is to either accumulate a certain number of knocks in a given time, then erase memory, start counting over again, so needless permanent adjustments are not needed when the ECU sees a single stray knock. There are many reasons a tuner might make these adjustments based on expected performance outcomes.

Here is a typical EMS parameters for another engine, not Bentley.

You must use 99 octane.. these are highly engineered cars.. don’t spoil the ship for a ha’peth of tar! 🙂

As BWings already stated, the ECU can self-adjust. In Germany I've used 102-octane (RON) as it's widely available, that would be around 97.6 octane in US (RON+MON/2) and the car feels even quicker than usually. At one point I had to settle for 95 (90.7) octane and everything was fine as long as I feathered the throttle until I could get a refill with 98 or higher.

There are differences between chains in what the actual octane rating is. I noticed this back in the 90's when we tried to squeeze the last bit of performance out of an engine, refueled between dyno runs and found out that it can take a bit more timing advance and boost than we had dialed in earlier. With the "same" 98-octane pump unleaded, except from a different chain station. We used this information to our advantage when we were racing in a pump fuel only class and even had the fuels lab tested. One particular "98" was actually around 100.1-100.4 octane so we could use an even thinner head gasket to up the compression and a bit more advance, to make more power. This was a higher reading than with the hyped Shell V-Power "99-octane".

One thing I've noticed as I keep track of fuel economy at every filling is that with 98 you get up to 10% better mileage than with 95, as an average from half a dozen cars and a few 100k's of miles. This is with modern engines that are capable of adjusting fuel and timing according to knock and other sensor feedback. Then there's the fun factor too, I have a 3-gallon can of pure toluene (121/114 octane!) in the trunk of my GT-R:s at all times, to up the knock threshold. I can use race fuel maps when it's mixed with regular pump gas, which makes quite a difference in power.

So my car failed its MOT this morning. It looks like my emissions is very high compared to what it should be (in red) so is that likely a lambda sensor problem? Or could 95 Fuel be causing it? It's had a major service late last year so all plugs etc are new. Thoughts welcomed?

Note, car runs well and no issues on dash. I have done a scane a couple of weeks ago and nothing but electrcial things showed that I kinda new about and will get to them. Car temp runs at 90 every day so no over heating either.

Hi @CUE99T

I am assuming, by what you've described, the emissions test is a device connected to the tailpipe(s) while the engine is running and not an OBD test only?

If that is the case, and the EMS is not showing a CEL and no P-Codes, then more likely than anything else, your Cats are no longer effective. That would be #1 for me, with no other info. There could be other more outliers, if you are starting to burn (consume) oil yet no P-Codes showing.

The EMS (any OBD II EMS) is designed to maintain stoichiometry for gasoline (@ 14.7:1) which produces the most complete combustion for gasoline. This exhaust waste is very little challenge for a CAT to do a final cleaning. However if, for some reason, fuel trims were going south (would have to be rich, not lean) this could eventually damage a CAT, but the system is designed to "complain" very loudly with faults that would produce CELs far ahead of damaging CATS. Then again, CATS don't last forever.

Don't know if any of this helps,

Thanks BWings, yes its a tail pipe test I imagine as part of the ITV for road worthiness in Europe. I don't think its burning oil, and temp is normal at 90c.

I am not that mechanically minded but I get your point. My Andorra mechanic thought maybe lambda sensors. It's good to know some opinions so when I speak to my mechanic this week I can run things past him. Hes pretty good so hopefully will get to the bottom of it.

As usual the Bentley isn't cheap to run which is fine, and I expect it, however things like Cats scare me for the cost a bit as not sure if it can come as a part or needs a full exhaust type of thing, and in general I'd rather try and get used parts like this maybe from Flying Spares so need to know what I'd be ordering. I have a parts usb from them so will maybe take a look.

Guess maybe the thing that surprised me is that 7 months ago in September there was no issues and had a good bill of health mechanically. Does a cat just eventually fail like that?

I'll let you know what is said this week and we can see if you hit it on the head!!