High fuel consumption without reason until restart (ECM bug?)

[vwengineer]

Hey guys!

It's been a bit more than one year that I own the car now and I encountered a strange situation. Yesterday I was driving on a small road and after 20 min, engine was at temperature and I did a strong pull after an intersection up to 50 mph in 2nd speed, then up-shifted directly to 6th speed (constant speed). Something felt strange, not sure if it's simply that I rarely to a big up-shift like this but the engine seemed to be noisier than usual in 6th speed (like if I gave more throttle than I actually was). I continued to drive, eventually had to slow down and change speeds several times.

At this point, I noticed that the live MPG seemed to be really off of what they should be. On a slight downhill at 30mph, either the engine was using more fuel than on the flat or none with sufficient gradient. Then went back to 50 mph, 6th speed, engine response was good but fuel consumption was very high (at least according to the MFA). On a flat road at 50mph in 6th I got something like 23 mpg continuously while normally I would be at 40 or more.

I decided to stop as it seemed something was really wrong, went out of the car and the engine did not seem to be noisier than usual at idle, coolant temperature was normal, no oil pressure light, no engine light. I stopped the car, waited 30 seconds and restarted. I continued on the same road, same conditions, flat road, but now I got more than 40 mpg and during the rest of the journey the fuel consumption was normal (which made the journey MPG improve a lot).

So not sure if there is a situation that would trigger such a behavior, as far as I know the fuel trims are limited to 20% so not sure that it could be a bad sensor reading (my LTFT was -0.9% last time I checked). Not sure either if it's related to this pull or not, I thought to an overreaction to knock but I am running premium fuel and the air temperature was low. As anyone noticed something like this?

I am thinking about a one time problem or bug caused by a conjunction of events but I wanted to know if there was a legitimate reason for the ECM to deliberately run a rich mix (like to raise exhaust gases temperature when cold)?

Thanks


PS: I am running the stock software

 

[Fred930]

Interesting - can't say I've ever seen that situation with my 2009 MkV GTI (stock TSI - CBFA), but I seldom have the MFD set to instantaneous MPG, rather preferring to leave it on average (since last fuel fill / reset).

Obviously it would be interesting to connect VagCom and watch several measuring blocks while attempting to re-create the situation. Short term fuel trim ; (and I think from memory) injector pulse width ; maybe throttle position ; coolant temperature ; MAF - things that could make the computer think a rich condition was required. But I would think even short malfunctions of any would set the CEL, or at very least give an ECU fault code.

Hum - even if your CEL was not on, you have checked for ECU fault codes - right?

That said - now that I think about it, am unsure how the instantaneous MPG is actually calculated. Assume it's simply integrating injector pulse width over some short time period, and converting that to fuel usage per unit time. Then that very short value could again be integrated over longer time to accumulate the "average" MPG. But can't say I've ever read any discussion of how it's actually done.

Assuming you can get it to occur again, the first question is whether the car is actually running rich, or maybe the problem is simply in the algorithm used to calculate the MPG display. A rich running condition will often cause a strong fuel smell, but that would require someone closely following you to detect.

Will be interested to know what you discover.

 

[vwengineer]

Thanks for your answer. Yes I think those items would be interesting to log, should it ever happen again. I did not have to scan the car yet but I don't think it will show any error.

For me the average mpg is pretty spot on and it also decreased with the decrease in instantaneous after the restart so I am pretty sure that the car was indeed burning that fuel.

That being said, I have an oil leak at the vacuum pump level. I was suspecting a boost leak in the vacuum hose through a defective check valve (from the intake mainfold to the vacuum line). Today I found a more likely explanation that a defective PCV can leak boost in the valve cover and so in the vacuum pump causing it to leak oil past its seals.

Now if it's a defective PCV (partially defective or intermittently defective) I think that could be the explanation, I still have to check that. Idle is getting rough when removing the oil dip stick so the PCV is probably not completely gone.

But it seems that a defective PCV can indeed lead to increased fuel consumption.

 

[ROH ECHT]

Can you log group 032 and post LTFT in block 2 just at idle is good as the long term data will be stored? (multiplicative/partial load)
Then reset it (clear codes) and report what it's additive is in about one week?

 

[ROH ECHT]

It is likely to just be a momentary improper signal from a sensor or combination that caused it all.

 

[Fred930]

Thinking about this thread this morning, I just couldn't help myself. Set the MFD to read instantaneous MPG, did a second gear pull to red-line, then shifted into 6th and coasted for a bit. The MFD showed exactly what I expected it would - nothing unusual.

So I did it again! Still all systems normal.

Guess it will be hard to suggest much more, unless / until you are able to consistently re-create the problem.

 

[vwengineer]

Thanks ROH ECHT for your advice, I will try to do it and report back, I have to log that block while driving, right? I just checked, there are no DTC, engine stopped, the idle part (first item of block 032) is at 0.5%. But I am pretty sure it was just a glitch somewhere or a intermittent failure of the PCV if that's possible (like stuck open creating a vacuum leak?).

Thanks also for the experiment Fred930 but I am not sure the issue is reproducible, it just happened once in a year. My main concern was to understand if it was something "normal" under certain conditions and if by switching off the car I interrupted a legitimate action of the ECM (for example diesel cars sometimes enrich the mix to clean the DPF) or if it was a glitch somewhere.

 

[vwengineer]

I did a log tonight, about 20 miles mostly at 45 mph, some parts at 75 mph. Fuel consumption was fine with an average of 47 mpg on this trip (very gentle driving).

On group 032, I have an idle fuel trim at 0.4% and partial which I assume is short term was for most of the trip at -0.4%. Always between 0% and -0.4%. So MAF seems fine.

Leaning toward a vacuum leak due to the PCV last time.

 

[ROH ECHT]

Wrong..."partial" is long term multiplicative-or the off idle fuel adjustment. This is the value of adjustment when driving normally when off boost and off idle. It isn't a reflection of how the MAF is...it's a reflection of air/fuel reports to the ECU by the O2. Your value looks very good at -0.4%. My partial/off idle is usually between +6% and +7%. If mine has a small vac leak, or small post MAF leak in the intake to throttle body, it will go over +10%.

Maf reporting is in group 3/Air Mass. But you need to know what it was when you had a new MAF or know what your HP is. The max Mass Air is only available at near the end of revs in a certain gear...3rd or 4th near redline is preferable.

Here...watch the vid linked at the end; http://wiki.ross-tech.com/wiki/index.php/Fuel_Trim_Info

 

[vwengineer]

Thanks for the link, very informative. I know what to look for should I have the problem again. I understand that the trim is a reflection of the reading of the O2 sensor and that MAF is not involved.

My understanding is that if the MAF readings are incorrect to begin with, the ECU will "plan" an incorrect amount of fuel, then the feedback from the O2 sensor will tell him that it's not what is expected and it will then "trim" the mixture accordingly (or struggle to trim it because I don't think the MAF error will be linear), on the other hand, if the MAF reading is correct and there is no vacuum leak, the amount of fuel will match perfectly the amount of air so that no further trims will be needed.

Your trim seems high in comparison (even though not that high) but keep in mind that your car is tuned and also run a custom map, which as good as it is, maybe doesn't match 100% the mods you have on your car, so that your trim might be just a shift that makes it match.

I think there was someone on the forum having huge issues with fuel trims, CEL and running lean, it turned out to be its tune if I remember correctly.