AFRs: Shell V-Power vs. Exxon w/105 Ethanol

[avengerhed]

I've had my wideband (AEM UEGO) in for about 1000 miles now and I typically fill the car with Shell's V-Power (93 octane) gasoline. At idle and at a cruising pace I would typically see the AFR on the wideband "hovering" around 14.7 with a swing from maybe 14.4 to 15.0 .

Today, very low on fuel, I filled up with 93 Oct from Exxon which in MA now has 10% Ethanol. For the next few minutes of driving everything seemed fine until I stopped in traffic on a hill and I noticed the AFR hovering around 13.8. This continued at idle or at a cruising pase, sweeping from 13.3 to 14.3 but never in the range of 14.7 again.

My questions follow:

1. Is this a typical response from 90/10 ethanol mixtures?
2. Is the car compensating for the ethanol and running richer?
3. Does/is/will my dynoflash (for 93 Oct) cause problems?
4. Is there anything to worry about?

From what I know about ethanol mixtures, ethanol is more oxygenated than hydrocarbons like gasoline, so it provides less engergy when it combusts. this would lead me to think either a) you'd expect a leaner mixture not richer or b) the car is compensating for the reduced power/efficiency by richening the mixture.

I'm looking for someone to shed some light on the situation, I'll be going back to Shell this afternoon, thankfully I only put in 1/4 of a tank to get me to work.

The wideband O2 sensor is fairly new, I don't use alcohol injection and it's as far from the turbo as it should be (by the flange on the DP). It's also installed at the right angle.

Thanks for your help!

[value]

There is a difference with the ethanol mix. I never ran it until I moved to WA. Now my EVO seems to run richer also.

[Ted B]

Whatever functional difference exists between MTBE and the new EtOH formulations, the ECU should adjust the mixture to run the specified closed loop AFR as reported by the O2 sensor, and will apply the adaptation globally (including the open loop trims). It may take a short while for the ECU to update the compensatory fuel trims, but it should make whatever adjustment (if any) necessary.

[vrex]

Gas with Corn Oil

check out post #11. the gist of it is, we are used to an AFR of 14.7:1 because thats how gasoline burns, and thats all we've ever had to tune for. ethanol has a stoichiometrics ratio of 9:1, i.e. richer than gasoline, so you will see a decrease in your AFR with the addition of ethanol because the ethanol take up a volume per injection that used to be filled by gasoline (if that makes sense), and you will thus run richer. read the posts, they are very informative. there must be someone more qualified than me to answer your 4 questions, i.e. an engineer who blows stuff up for a living... anyone?

[avengerhed]

This is great information, thank you all.

[Ludikraut]

Originally Posted by avengerhed

I've had my wideband (AEM UEGO) in for about 1000 miles now and I typically fill the car with Shell's V-Power (93 octane) gasoline. My questions follow:

1. Is this a typical response from 90/10 ethanol mixtures?
2. Is the car compensating for the ethanol and running richer?
3. Does/is/will my dynoflash (for 93 Oct) cause problems?
4. Is there anything to worry about?

...

Thanks for your help!

I'm not an engineer, but I'll take a stab at it:

1. If you are measuring actual air/fuel ratios, then this sounds almost right. Stoich for a 90/10 mix is ~14.13 instead of ~14.7 for 100% gasoline

2. Your car should be compensating in closed loop mode, since the stock O2 sensor will generate its stoich reading at an AFR of ~14.13 now instead of 14.7 (wiki explanation of O2 sensors here). However, IIRC, your wideband sensor and display calculate the actual AFR from the lambda value that is reported by the sensor. I find it odd that it would report a lower AFR since your UEGO has not been adjusted to account for the new stoich ratio. It would be interesting to have someone chime in on how exactly the AEM UEGO calculates its AFR values.

3. As long as your AFRs don't lean out too much, it should be fine. If anything, the Ethanol will give you a smidgeon more knock resistance.

4. I'd worry about the accuracy of your UEGO. AFAIK they cannot be calibrated beyond the initial factory calibration. Any sensor that cannot be calibrated periodically is suspect, IMO.

As a final note, I have been running my car on a 90/10 gas/ethanol mix for as long as I've owned it (can't get anything else in IL anymore), and it has never been an issue. My wideband (Innovate) has always hovered around 14.7 at cruise and idle (assuming a good tune, hehe).

l8r)

[l2r99gst]

I don't want to get into a long explanation, but the 90/10 mixture will actually make your car run leaner, not richer.

What you experienced is probably just the fuel trims overadjusting due to the mixture. But, don't worry about it....that's what the closed-loop fuel trims are for. They adjust either negative or positve to maintain a stoich ratio of 14.7:1 (what it knows the ratio should be for pure gasoline). But with the mixture, this reading of 14.7:1 will actually be running a little leaner than the ~14.13:1 stoich for the mixture.

Eric

[Ted B]

Originally Posted by l2r99gst

I don't want to get into a long explanation, but the 90/10 mixture will actually make your car run leaner, not richer. . . . But with the mixture, this reading of 14.7:1 will actually be running a little leaner than the ~14.13:1 stoich for the mixture.

Be careful Eric, that's incorrect. The ECU will gradually enrichen the mixture in closed loop operation until it arrives at the same stoichiometric point as it did with the straight gasoline. It will then apply that same adjustment globally (including open loop).

The O2 sensor reports free oxygen, not AFR. Likewise, the ECU adjusts the fuel trims until it gets a signal from the O2 sensor that equates to a certain percentage of free oxygen.

Because of this, if the ECU is programmed to maintain a level of free oxygen that equates to an AFR of 14.7:1 for gasoline, that very same concentration of free oxygen equates to ~14.13 for the ethanolic mixture (assuming the stoichiometry you reported is correct). Likewise, if we were to run straight methanol, the ECU would attempt to richen the fuel mixture until it reached the same concentration of free oxygen, which would be something like 6.4:1. Obviously it would never be able to make that level of adjustment, but the point is the ECU only 'knows' the free oxygen content, not the AFR. It's up to humans to determine what the actual AFR is (for the sake of discussion), which is obviously based upon the particular fuel we use.

[Nad1370]

Your ECU just needed to re-learn stoich values on the gas you have.
It might take at least 10 min. or so and so miles to re-learn.

I only have 90/10 mix around here and always get stoich at idle.
.

[l2r99gst]

Originally Posted by Ted B

The ECU will gradually enrichen the mixture in closed loop operation until it arrives at the same stoichiometric point as it did with the straight gasoline. It will then apply that same adjustment globally (including open loop).

Perhaps I misworded some of my post, but I didn't want to get into a long explanation, but I appreciate that you did. What I meant to say was that if you run any mixture with a stoich ratio less than 14.7:1 (ie 14.1:1) then you car will run leaner, until the closed loop fuel trims readjust back to the stoich ratio for the mixture. I was simply suggesting that perhpas in his case, the fuel trims overcorrected a bit and he was seeing a richer than normal reading.

However, I don't agree with the second sentence in you post above. Fuel trims are only used during closed loop operation. They are not applied during open loop fueling (at least from my experiences with the DSM ECUs...I wouldn't expect the Evo ECU to be any different here), so you will indeed, run a little leaner in open-loop (WOT) if you are running a mixture with a stoich ratio less than (richer)that of pure gasoline.

Also, along with what you are saying, your wideband reading, whether closed or open loop, will show a little leaner that what you are actually running during open-loop. This is due to the fact of what you already explained. Stoich for gas is 14.7:1. Stoich for 90/10 is 14.13:1. So, for example, when the wideband says the mixture is stoich, it reads 14.7:1, but you are actually running at 14.13:1.

Eric

[Ted B]

Originally Posted by l2r99gst

However, I don't agree with the second sentence in you post above. Fuel trims are only used during closed loop operation. They are not applied during open loop fueling (at least from my experiences with the DSM ECUs...I wouldn't expect the Evo ECU to be any different here), so you will indeed, run a little leaner in open-loop (WOT) if you are running a mixture with a stoich ratio less than (richer)that of pure gasoline.

It's possible, but assuming the Mitsu ECU works like the Fords, the ECU extrapolates the adapted closed loop injector pulsewidths onto the open loop programming as well. Therefore, it has some means of tailoring the open loop mixture, albeit without correction, so it isn't a perfect system.

[dubbleugly01]

from my perspective (I'm a chemical engineer), the swap to a 10% ethanol mix will make your car run leaner at the same injector duty cycle. It will be up to the 02 sensor to catch the lean condition and increase the pulse width until enough fuel is being dumped to match the oxygen in the cylinder. The 02 sensor is measuring free oxygen, not any type of stoichiometric ratio.

So, we just swapped to a 10% ethanol blend here in Houston about a month ago. First tank in the EVO and I'm thinking I'll be running lean until the fuel trims work their magic. Great.... I'll take it easy, give the computer a chance to make the adjustments and get some higher mileage out of the first tank of gas. My mileage actually dropped by about 10%, which goes against the science behind it all. All I can think, is that the 02 sensor overcompensated, and forced a rich condition until things got sorted out. The second and third tanks of gas have been back to normal mileage.

I don't know about the EVO ECU, but the Toyota's definitely have a "short term" and "long term" fuel trim capability. If the short term corrections from the 02 sensor occur long enough, a correction will be applied to the open loop tables as well.

What really bakes my noodle, with the swap to 10% ethanol, we just lost injector capacity!!!

[Nad1370]

^^ Evo's also have both short and long term FT.
.

[thatsMR2u]

exxon and mobil just advertise the 10% ethanol at the pump but all gas stations including shell v-power have switched in MA to the 10% ethanol at the same time--

[NEW2006EVOIX]

I have a BRAIN CRAMP now after reading this post! I think what everyone is saying is that it does run leaner, but the O2 sensor picks up on this lean condition and increases the fuel. At the end of the adjustment range, you guys are getting the rich readings.

Here's something to try out...disconnect your O2 (assuming its failsafe is to maximize fuel enrichment). If your mileage stays the same, then running this gas causes the ECU to over-compensate to its max.

Need to stick my head in the freezer now...heard that makes computers run faster...I think my brain is stuck at 100% usage and I think it's about to bluescreen on me...might have to reboot my brain...