Gas mileage computations are NOT accurate

[fordtruckman2003]

Even if you have a traditional appearing gauge it's still digital in the background. You can read actual fuel level behind the gauge with a scan tool if you want to get into details.

I just don't worry about it. It's a truck. I do get photo of every fuel up before I reset trip meter and the gas pump so I can plug into my spreadsheet if I feel like it. I did that for 5 years on my last truck to satisfy my curiosity but then I realized what I said above about the amount of fill having huge variance so stopped bothering.

Sponsored

 

[powerboatr]

I've had eleven vehicles in the last 30+ years. I'm not sure with which vehicle they started displaying MPGs, but not one has been accurate on the dash. They have all generally been within 1-2 mpg of what the dash shows.

my 2006 super duty with 4.30 gears and the 6.0 was dead on......12.6 over 78k miles empty or loaded
i prayed it would one day be optimistic ?

so far the 22 150 is close , as in .3 to .8 mpg in town
on highway for long periods i believe its a variable with no counter.?
i plan my drives on 17 mpg and then i map out fuel stops .
i can make it from my house in north east texas (quitman) to other side of mobile to buccees on one tank however

 

[hardlymoving]

Am sure that the accuracy of the gas mileage as shown on the Dash Display varies between vehicles but on my 2022 F150 with the 2.7 engine I used the Engineering Test Mode to lower the reading by 4% and now the dash reading closely agrees with the actual miles per gallon as hand calculated. I had to make the adjustment on the other F150 but can't remember by what amount, probably 3 or 4% less.

 

[PatchManager]

If you fill up your tank until the fuel pump clicks (the recommended approach by Ford), the amount of fuel you are putting in can easily vary by 1 gallon. The click point of a fuel pump is not exactly the same on every pump and can vary from one fill-up to the next using the same pump because the click point is affected by a number of factors. Therefore, you will know with good precision how much fuel you added to your tank, but you don't know if that is how much fuel you actually used.

I had an older car that I used to fill up to the point where I could see the fuel level. I found that where the pump would click off could vary anywhere from 0.5 gallons to 1.5 gallons or more. To get an accurate estimate of your MPG, you need to average it over several tanks, so the click-point variance is average out.

 

[spiritrider1]

As a couple others have mentioned, use Engineering Test mode to adjust the AFE bias. Factory stock is set at 1000. Both my F-150's computers were 1.2 mpg 'optimistic'. Things such as tire size, axle ratio, weight of truck due to options, etc. all factor in the computer's calculations. I reset the bias on my 2018 2.7 STX and my current 2021 PB. In both cases I'm usually around .2 mpg of the truck's calculations. Sometimes a bit more. That variance is what others explained is due to temperature, fuel density, and accuracy of filling. Nothing's going to be perfect but you can get it to be very accurate. I will attach a pdf with instructions for those who care.

Also, I use Fuelly app which pretty much does what those of you are manually doing with spreadsheets. I was reading a Motor Trend article about the Lightning costing .21 cents/mile to operate. Per Fuelly, I'm spending 14.3 cents/mile. I'm pretty pleased with that.
I'm also pleased that in some cases, I do better than expected. See IMG_0274. But that's the exception, not the rule. It's usually no more than .2 - .4 off.

 

[Zengineer]

Thanks for all the comments. With everything else being digital on the newer vehicles, it just seems the fuel gauge should be too. It computes out how many more miles one can travel on the fuel in the tank, so it must somehow meter the useage as well. When we buy it at the pump, they meter how much we get, why not a digital fuel guage? It would certainly be more accurate than an analog needle. Just a thought....I am still miffed at not being able to get the Nav package and Sirius/XM radio in my truck, as well as no CD player and no built-in programmable garage door opener.

Digital is how it is displayed. The gage still is at the mercy of the signal generated by approximations and with all the variables described in the thread. A digital readout is easier to read but that doesn’t make the reading more accurate.

 

[gtotco]

When the owner fills-up the vehicle and resets the economy calculator, drives 200+ miles, then fills-up again, why does the truck's computations say 17.4 mpg and my calculator say 16.76 mpg? I definitely trust my calculator much more than Ford's computer.

This has been case with every vehicle I’ve owned. I calculate by hand and resetting trip computer I’m usually around .5 mpg below the computer. Sometimes a bit more sometimes a bit less. I have never had a vehicle that wasn’t around this though.

 

[Iconic22F150]

My LOM is constantly 1 mpg higher than actual so at least it’s easy for me to know what it really is. I’m real happy for the 19 real mpg over the 10K mi I’ve clocked.

 

[scott011422]

Its always surprised me that the LOM isn't dead accurate. Its not like the injector orifice is variable. They know what the fuel pressure is and they know when and for how long each injector is open. These are hard limits. Obviously they use this data to some point because there is a fuel flow rate in the computer in I believe Lbs per hour? But even my boat has the same problem. I put 20Gal in an empty tank. At the end of the year it said I had 11.8 gallons remaining on the ECU. I drained the tank and only pulled out a little over 8 gallons. So even the fuel usages rates on modern fuel injected outboards have the same problem.

And I ask myself again, Why? Now like everything else I'm sure there is tolerance on the fuel flow of injectors. Stock GM injectors for instance are somewhere around 450cc per min. Is that +- 5cc? I doubt every injector is EXACTLY what they say it is. Say its 5cc on a v6. Thats 60oz an hour more fuel than its supposed to be using. But then again, That would be a constant. I never see the LOM off the same amount every time. I've seen my LOM at times dead on, and I think the worse was almost 5 MPG off. That like 6 gallons a tank. Thats a hell of an error.

 

[ReverendQ]

Its always surprised me that the LOM isn't dead accurate. Its not like the injector orifice is variable. They know what the fuel pressure is and they know when and for how long each injector is open. These are hard limits. Obviously they use this data to some point because there is a fuel flow rate in the computer in I believe Lbs per hour? But even my boat has the same problem. I put 20Gal in an empty tank. At the end of the year it said I had 11.8 gallons remaining on the ECU. I drained the tank and only pulled out a little over 8 gallons. So even the fuel usages rates on modern fuel injected outboards have the same problem.

And I ask myself again, Why? Now like everything else I'm sure there is tolerance on the fuel flow of injectors. Stock GM injectors for instance are somewhere around 450cc per min. Is that +- 5cc? I doubt every injector is EXACTLY what they say it is. Say its 5cc on a v6. Thats 60oz an hour more fuel than its supposed to be using. But then again, That would be a constant. I never see the LOM off the same amount every time. I've seen my LOM at times dead on, and I think the worse was almost 5 MPG off. That like 6 gallons a tank. Thats a hell of an error.

To possibly eliminate the tolerances and other variances I smoothed my data a bit and as the first fill-up it appears to be dead nuts on ? See below.

Date	Miles	Electric	Gallons	MPG(cal)	MPG(dash)
24/04/20	15684​	51​	14.694​	17.4​	18.2​
24/04/25	15901​	29.5​	11.763​	18.4​	18.5​
24/04/29	16221​	19.8​	18.692​	17.1​	18​
24/05/08	16626​	36.9​	22.45​	18.1​	18.8​
24/05/20	17028​	66.3​	22.283​	18.0​	19​
24/05/23	17476​	17.5​	22.966​	19.5​	19.8​
24/06/10	17907​	63.2​	23.767​	18.1​	18.6​
24/06/25	18328​	84.2​	23.857​	17.7​	18.6​
24/07/10	18640​	58.3​	17.397​	17.9​	18.9​
24/07/17	18906​	48.6​	15.357​	17.3​	17.7​
24/07/17	19290​	80.6​	21.034​	18.2​	19.4​
24/08/14	19681​	77.3​	21.138​	18.5​	18.9​
24/08/21	19913​	50.1​	12.963​	17.9​	19.1​
24/09/11	20659​	59.1​	18.424​	18.7​	19.5​
24/09/19	20855​	36.4​	10.549​	18.5​	19.8​
24/09/26	21142​	44.4​	14.995​	19.1​	19.2​
24/10/02	21301​	34.5​	8.621​	18.4​	19.7​
24/10/16	21640​	78.4​	18.705​	18.1​	19.6​
24/10/23	21873​	49.6​	13.16​	17.3​	19.3​
24/10/30	22007​	27.2​	8.187​	16.3​	17.2​
24/11/06	22252​	50.1​	14.196​	17.4​	18.4​
With calibrated speedometer for 34" wheels
Averages				18.002​	18.867​
AFE Bias on Averages (New Setting)					954​
First Fill Up Since Calibrating AFE Bias
24/11/20	22631​	79.2​	19.627​	19.2​	19.2​
	​	​	​		​

 

[scott011422]

You average much closer. Nice. Truck says average for the year is about 18.56. Total miles / gal for the year says 17.05. Here is mine this year:

DATE OF FILLUP		PRICE PER GALLON	GALLONS PUMPED	ODOMETER	TRIP	TRIP MPG	COST	HAND CALC	DIFF
1.11.24		3.799​	28.218​	16197.8​	446.5​	17.8​	107.2​	15.823​	-1.977​
1.22.24		4.299​	18.713​	16397.7​	199.8​	14.5​	80.45​	10.677​	-3.823​
2.9.24		4.699​	26.602​	16858.8​	461​	18.5​	125​
3.2.24		4.449​	27.981​	17284.2​	425.6​	18.2​	124.5​	15.21​	-2.99​
3.22.24		5.099​	27.195​	17697.2​	412.9​	18.2​	138.7​	15.182​	-3.018​
4.10.24		4.495​	27.624​	18142.5​	445.1​	17.6​	124.3​	16.1128​	-1.4872​
4.30.24		4.499​	28.571​	18660.2​	517.7​	20.3​	128.5​	18.119​	-2.181​
5.24.24		4.539​	28.927​	19203.4​	543.2​	19.3​	131.5​	18.778​	-0.522​
6.20.24		4.369​	29.533​	19710.7​	507.2​	18.1​	129​	17.174​	-0.926​
7.15.24		4.499​	27.649​	20178​	467.2​	18.3​	124.4​	16.897​	-1.402​
8.8.24		4.739​	26.606​	20632.9​	454.9​	18.1​	126.1​	17.097​	-1.003​
8.25.24		4.639​	29.442​	21214​	581​	20.6​	136.6​	19.733​	-0.867​
9.25.24		4.399​	28.564​	21739.4​	525​	19.4​	125.7​	18.379​	-1.021​
10.22.24		4.399​	29.039​	22287.8​	548.4​	20.1​	127.7​	18.885​	-1.215​
11.15.24		4.419​	27.127​	22776.5​	488.7​	19.4​	119.9​	16.54​	-2.86​

 

[ReverendQ]

You average much closer. Nice. Truck says average for the year is about 18.56. Total miles / gal for the year says 17.05. Here is mine this year:

DATE OF FILLUP		PRICE PER GALLON	GALLONS PUMPED	ODOMETER	TRIP	TRIP MPG	COST	HAND CALC	DIFF
1.11.24		3.799​	28.218​	16197.8​	446.5​	17.8​	107.2​	15.823​	-1.977​
1.22.24		4.299​	18.713​	16397.7​	199.8​	14.5​	80.45​	10.677​	-3.823​
2.9.24		4.699​	26.602​	16858.8​	461​	18.5​	125​
3.2.24		4.449​	27.981​	17284.2​	425.6​	18.2​	124.5​	15.21​	-2.99​
3.22.24		5.099​	27.195​	17697.2​	412.9​	18.2​	138.7​	15.182​	-3.018​
4.10.24		4.495​	27.624​	18142.5​	445.1​	17.6​	124.3​	16.1128​	-1.4872​
4.30.24		4.499​	28.571​	18660.2​	517.7​	20.3​	128.5​	18.119​	-2.181​
5.24.24		4.539​	28.927​	19203.4​	543.2​	19.3​	131.5​	18.778​	-0.522​
6.20.24		4.369​	29.533​	19710.7​	507.2​	18.1​	129​	17.174​	-0.926​
7.15.24		4.499​	27.649​	20178​	467.2​	18.3​	124.4​	16.897​	-1.402​
8.8.24		4.739​	26.606​	20632.9​	454.9​	18.1​	126.1​	17.097​	-1.003​
8.25.24		4.639​	29.442​	21214​	581​	20.6​	136.6​	19.733​	-0.867​
9.25.24		4.399​	28.564​	21739.4​	525​	19.4​	125.7​	18.379​	-1.021​
10.22.24		4.399​	29.039​	22287.8​	548.4​	20.1​	127.7​	18.885​	-1.215​
11.15.24		4.419​	27.127​	22776.5​	488.7​	19.4​	119.9​	16.54​	-2.86​

Hey Scott, Yes you have quite a bit more variance in your data than I have in mine. Are you going to try averaging to change your AFE Bias?
I am curious if averaging will help to calibrate out the tolerances (and variances) long term.
It would be great to see if others average their data to get a similar apparently more accurate result... like I did in my sample of one fill-up since changing the AFE Bias.

 

[Buyer2021]

It wasn't a priority, but I finally did tweak the AFE ratio on my 3.5EB Lariat.

I previously posted a table showing calculated vs displayed variance over multiple fills of 0-8%

Now with 14 fill-ups after tweaking the AFE, that variance is averaging less than 1%, with the largest outlier being -2%.

Sponsored