Sorry to reply to my own post, but after re-reading one of Bill Winney's posts I realize my assumption about the invisible mileage register keeping a rolling average of just one tankful is probably incorrect. His observations of the towing mileage dropping over a 2000 mile period indicates that the invisible register MPG is probably averaged over several tankfuls.

I have a question for all of you: how accurate do you think the fuel gauge is as it drops near zero? I have assumed it is more accurate near zero than near full but have no evidence. I ask because I want to know how much trust I can put in the MTE reading near zero. Some replies to my post say that you can't trust it, which means that my method of trying to compare fill-up volumes from station to station is a waste of time. Is this the case?

RayP

 

This still leaves my last question: how accurate is the gas gauge near zero? If it is pretty accurate, I agree with you. If you can't trust the gauge, and can't trust the rolling average MPG, then you may have a problem.

RayP

 

Very accurate.

Bottom Line: When you hit MTE 0 you have 1.3 gallons of fuel left in a 2005 to 2008 MY, and 0.8 gallons left in a 2009 MY ( based upon Gary G's 2009 FEH ).

How far you can go past zero is relative to your MPG during that last gallon or so. Since the "rolling ave" or long-term whatever average is not trustworthy over short distances, it is up to the driver to know how far that last gallon will last.

HTH,
-John

 

P.S. Go out and experiment and report back, if you can. Less data is available for the 2009's. If you do this in the city, you can creep along in EV 1 mile min, 2 miles max. to a gas station if you run out, or you could carry a gas can with you.

I base my post(s) on running out of fuel 4 times. Once, while towing, just last week. I was getting 15 MPG and ran out of fuel at 18.9 miles past zero, which indicates to me, the 1.3 gallons of liquid left at zero widely reported is accurate.

( I was 0.25 mile from a gas station and towed just fine in EV mode at 22-25 MPH. )

 

This evokes a memory from 2005. I was fortunate to attend the one and only "Ford MoCo 'Open House'" for Escape Hybrid Owners in the Fall of 2005 in Dearborn, Mich. Dozens of engineers who worked on the Hybrid project were present to meet & greet the public, and answer questions and gather feedback.

While there, I asked of the engineers: "Why does the MTE go down while I'm driving in EV mode?" One of them looked at me like he'd seen a ghost, turned to another engineer, and said "I thought we'd fixed that." I don't think they had a good answer for me at the time, 'cause I sure don't remember anything. I think the answer was "It's not supposed to. We'll get back to you on that."

Here is what I THINK it does based upon 4.5 years of driving one ( 85,000 miles ). It recalcs the MTE at approx. each 10% tank level drop. This is when I see big moves, jumps, "corrections", or if you don't like that term, big adjustments to the MTE.

Hypothesis:
At 100% tank level, based on the long term average, you get a MTE value. For 10% of your gas tank, it drops the MTE 1 per 1 mile on the odometer.
At 90% tank level, it checks, or recalcs to see if it's still on track, and compares tank level to rolling ave MPG and recomputes. This would explain the "jumps" I often see. ( sometimes the MTE jumps up as well )

In the intervals between "adjustments", based on a shorter term MPG average, the MTE sometimes drops faster or slower than 1 per 1 on the odometer.

It is not simply gallons x ave. MPG at all times. Only at the start, and at 10% intervals, perhaps. At each key turn also, perhaps, as my MTE is not always where I left it when I come back after a key off. How about you guys?

I think we are spending way too much time and effort on this, but hey, it's kind of fun to reverse engineer things!

 

I just got back from a short drive that started with surface streets and then went to freeway. My typical mileage on surface streets, sub 40MPH and with minimal interruptions, is around 45MPG. This is a combination of pure EV and ICE/Charging modes. Now, getting on the freeway is another story. Accelerating to 65MPH and holding it there for a mile or two my average is probably only 25MPG or less. So I did a quick experiment. I watched the MTE for the last mile or so before entering the freeway, and then for the mile that included acceleration and 65MPH cruise. The MTE ticked over 1 mile per mile of the odometer in both cases. So, at least on my '09 and under the conditions described, instantaneous mileage is not considered in the MTE calculation.

RayP

 

Good observations.
I think the sample distance/time was too short.

Watch for any "adjustments" at appox. 10% tank level drops.
Thanks!

It would be wastefull, but would be fun to watch: What occurs when you run the engine for an extended time with the car in PARK! aHA!

Does anyone doubt the MTE will go down, even though the odometer isn't changing???

 

Well, the change was EXACTLY mile for mile, and based on approx 45MPG vs 25MPG it would have been almost a 2:1 change, so I figured even 1 mile is enough. Longer may be better but try to go 5 miles in one driving mode and then 5 miles in another. Hard to do.

Your point about MTE dropping in park is good, and I can verify that it does indeed drop, but very slowly.

RayP

 

That is what I would predict. See!? Instantaneous fuel consumption does play a role. Obviously, in park, fuel consumption is very low. Also think about all the stopsigns and stoplights people are at with the engine running.

Here's what I think: The car makes predictions and uses certain assumptions based on averages and then occasionally takes measurements and recalcs, for real life conditions. The MTE calculation is not continuous, but periodic.

 

That's not the conclusion I would make. If instantaneous consumption made a difference, then my experiment at different FE levels would have shown it. What happens when the car is running in park is the average FE goes down gradually, very gradually, and the fuel level also goes down gradually, thus pushing down the MTE.

I absolutely agree that the car makes predictions based on your driving habits, historical FE, and even the max RPM you use for acceleration (my observation). My proposed simple algorithm is probably not all there is to the MTE calculation, or the AVG MPG calculation for that matter. I like your idea about the recalcs at the 10% intervals and will watch for that behavior.

RayP