gas mileage: actual vs reported...

[FRSowner]

Some posters say that the fuel mileages readings on our cars are optimistic by 1.5 - 2 MPG.
I don't understand how the onboard fuel mileage calculator can be so incorrect.
Onboard readout: The computer can measure EXACTLY how much fuel the engine is consuming through injector pulse times(s) while at the same time, monitoring the distance traveled. The amount of fuel in the tank, the octane, the weather is all moot - what matters is the amount of fuel that has passed through the injectors compared to the distance traveled. It's basic math but is COMPUTER MONITORED. Should be dead on accurate.
Driver calculations: It seems to me that recording the distance traveled and how much fuel used would have more variables and chances for errors such as the accuracy of the gas pump and what volume you are actually getting compared to what it tells you it is dispensing as well as the accuracy of your odometer which can be in error by wrong sized tires, rear end gear changes, etc. and also, math errors.

I was going to call BULLSHIT on the people that say that the MPG readout in our cars was incorrect by 1.5 - 2 MPG, then I read this...

http://www.edmunds.com/fuel-economy/...s-fibbing.html

It would appear (to me at least) that the car manufacturers are scamming us. There is NO reason that the readout should be incorrect. All the excuses given in the article are nonsense. Thoughts?

[W2]

From article:
"I can't imagine any reason that any automaker would want to make drivers think they can get better fuel economy than they were getting," Williamsen adds.

I don't know to sell cars maybe?

[LeftFootBrake]

I don't doubt the inaccuracies across the board but can report that my dash readout and fuelly currently match identically at 7.9L/100km (US 29.9mpg) after 2730km.

http://www.fuelly.com/driver/86gtsat/86

[Thunderchicken]

Originally Posted by FRSowner

Some posters say that the fuel mileages readings on our cars are optimistic by 1.5 - 2 MPG.
I don't understand how the onboard fuel mileage calculator can be so incorrect.
Onboard readout: The computer can measure EXACTLY how much fuel the engine is consuming through injector pulse times(s) while at the same time, monitoring the distance traveled. The amount of fuel in the tank, the octane, the weather is all moot - what matters is the amount of fuel that has passed through the injectors compared to the distance traveled. It's basic math but is COMPUTER MONITORED. Should be dead on accurate.
Driver calculations: It seems to me that recording the distance traveled and how much fuel used would have more variables and chances for errors such as the accuracy of the gas pump and what volume you are actually getting compared to what it tells you it is dispensing as well as the accuracy of your odometer which can be in error by wrong sized tires, rear end gear changes, etc. and also, math errors.

I was going to call BULLSHIT on the people that say that the MPG readout in our cars was incorrect by 1.5 - 2 MPG, then I read this...

http://www.edmunds.com/fuel-economy/...s-fibbing.html

It would appear (to me at least) that the car manufacturers are scamming us. There is NO reason that the readout should be incorrect. All the excuses given in the article are nonsense. Thoughts?

Here is the skinny... The best way to calculate MPG is distance/amount of fuel pumped. If you say that it's BS its not because all fuel pumps in the USA are required to be calibrated by the US Department of Agriculture every year. If they aren't accurate they get fined and possibly loose there license to sell.

Your on board computer doesn't calculate by how much fuel it uses voltage readings from the air sensor, throttle body, and O2 sensor and it spits out a number no math involved it's all programmed on a table, however it does give you a ball park idea of what you are getting.

My 08 STi was always 2-4 MPG higher then what I calculated.

[Kwaziekeller]

I've always found that onboard mpg calculators are kind of on the high side... if you don't believe it... go driving @ 1am on a desolate stretch of interstate with your low fuel light on... bet you'll worry if you're really getting 31.5mpg then! lol. I do the math based on miles driven/fuel added. I always try to top off my tank as well. No doubt in my mind I'm over 30mpg because I go a lot further than my SRT4 or SRT8 did on the same amount of fuel. I can get 350+mi of out a tank on the interstate without much trouble or dramatics and it'll only be like 11gal of gas.

[Leonardo]

I just got back from a 2700 mile road trip. I averaged 34 -40 mpg. I would fill up, reset trip A, and then drive till i was about at half a tank. Then fill up again and do the math. My dash says my avg is 34.5mpg. I believe it. It said 29mpg before the trip. I would avg 55-80mph in traffic. (4-5 over what is posted) I do have an auto, which is geared different. I have almost 5000 miles now.

[smbrm]

Some food for thought

You can also reset the average fuel consumption readout by press and hold. This will allow you to get an average for whatever interval you want, like a single tank. If you don't reset for each tank, then you are looking at a lifetime average since the last reset. This is seperate from and odometer reset.

Thinking about the math of averaging and rounding might also help. When using the computer I look at the average over each tank by reseting the average indicator after each fill. If you do the average over say 10 tanks, individual variations have less impact on the average even if it is computer calculated.

Also think about the instaneous fuel consumption readout. I have seen readouts in 1rst gear that approach 50L/100km. but the average is effected by not only that number but how long I am generating that fuel consumption. So while high if you upshift earlier to 2nd, the impact on the average is less.

Yes the computer should capture this, but there is still math and it still reports in this case of L/100 Km. to one decimal place. So on a recent tank I averaged 5.7L/100km based on the fuel pump fill/odometer. The meter showed 5.6 until just before I pulled into the station when it went back to 5.7. so even when you consider that you are recording to one decimal (0.1L/100km) being on one side or the other of 0.1 L/100km represents an error of 1.75% in this case. If in miles/gallon, rounding represents only about 0.3% for 0.1mpg variation on 33mpg. Does that mean the computer is more accurate in measuring mpg vs. L/100km. No, but the readout is still only accurate to its reporting capability and rounding criteria.

Now it is also pretty much common practice for fuel to be priced based on a standard density at 15 degrees C. so as to minimize the impact of fuel temperature. At least in some places! So that will also influence dispensing pump accuracy.

As someone else eluded to fuel pump accuracy is regulated and subject to inspection, so they are usually going be as accurate as the regulations require.

Fuel economy competitions and Formula One as examples, measures fuel consumption by mass to minimize volume temperature issues, but that is a bit more complex. Plus as someone else suggested there is no guarantee regarding how these instantaneous and average fuel consumptions are actually generated.

I agree there are many other variables that contribute to differences as well.

The important part should be to use a consistent method for measurement/calculation since comparison relative to how you drive and use your vehicle is what I think you are trying to capture rather than absolute values comparing one car to another.

The fill/odometer method still appears to have the greater opportunity for accuracy as long as you ensure you use the same procedure to decide when the tank is full!

[Memphis]

I just got back from a trip to Nashville from Memphis and got 40.8 MPG. Auto FTW.

[sierra]

Originally Posted by Memphis

I just got back from a trip to Nashville from Memphis and got 40.8 MPG. Auto FTW.

Very impressive!!!!

[ToxicSneakers]

With 1800 miles I decided to drive my AT in snow mode replicating the Prius economy methods. What I got was 250 miles at 43.3 MPG, almost all city streets. During this time on the roads I maintained normal traffic flow and speeds, but did modulate throttle at minimum for conditions. Amazing for the car I will be taking to autocross this weekend, and I can't think of any other choice as versatile: Can you?

[SPEEDGOD]

Originally Posted by ToxicSneakers

With 1800 miles I decided to drive my AT in snow mode replicating the Prius economy methods. What I got was 250 miles at 43.3 MPG, almost all city streets. During this time on the roads I maintained normal traffic flow and speeds, but did modulate throttle at minimum for conditions. Amazing for the car I will be taking to autocross this weekend, and I can't think of any other choice as versatile: Can you?

Yes that's impressive and I think you have it bang on, it's the versatility that sealed the deal for me to get AT

[Memphis]

Originally Posted by ToxicSneakers

With 1800 miles I decided to drive my AT in snow mode replicating the Prius economy methods. What I got was 250 miles at 43.3 MPG, almost all city streets. During this time on the roads I maintained normal traffic flow and speeds, but did modulate throttle at minimum for conditions. Amazing for the car I will be taking to autocross this weekend, and I can't think of any other choice as versatile: Can you?

I am going to have to try that lol. That is insane.

[Boosted2.0]

Originally Posted by ToxicSneakers

With 1800 miles I decided to drive my AT in snow mode replicating the Prius economy methods. What I got was 250 miles at 43.3 MPG, almost all city streets. During this time on the roads I maintained normal traffic flow and speeds, but did modulate throttle at minimum for conditions. Amazing for the car I will be taking to autocross this weekend, and I can't think of any other choice as versatile: Can you?

I have done the same thing. I was able to average 44 MPG on my manual trans at speeds of 0 - 60 MPH on city roads and country backroads using the same hyper-miler egg under the foot driving we use to confirm proper fuel economy for Prius.

[Boosted2.0]

Originally Posted by smbrm

Some food for thought

You can also reset the average fuel consumption readout by press and hold. This will allow you to get an average for whatever interval you want, like a single tank. If you don't reset for each tank, then you are looking at a lifetime average since the last reset. This is seperate from and odometer reset.

Thinking about the math of averaging and rounding might also help. When using the computer I look at the average over each tank by reseting the average indicator after each fill. If you do the average over say 10 tanks, individual variations have less impact on the average even if it is computer calculated.

Also think about the instaneous fuel consumption readout. I have seen readouts in 1rst gear that approach 50L/100km. but the average is effected by not only that number but how long I am generating that fuel consumption. So while high if you upshift earlier to 2nd, the impact on the average is less.

Yes the computer should capture this, but there is still math and it still reports in this case of L/100 Km. to one decimal place. So on a recent tank I averaged 5.7L/100km based on the fuel pump fill/odometer. The meter showed 5.6 until just before I pulled into the station when it went back to 5.7. so even when you consider that you are recording to one decimal (0.1L/100km) being on one side or the other of 0.1 L/100km represents an error of 1.75% in this case. If in miles/gallon, rounding represents only about 0.3% for 0.1mpg variation on 33mpg. Does that mean the computer is more accurate in measuring mpg vs. L/100km. No, but the readout is still only accurate to its reporting capability and rounding criteria.

Now it is also pretty much common practice for fuel to be priced based on a standard density at 15 degrees C. so as to minimize the impact of fuel temperature. At least in some places! So that will also influence dispensing pump accuracy.

As someone else eluded to fuel pump accuracy is regulated and subject to inspection, so they are usually going be as accurate as the regulations require.

Fuel economy competitions and Formula One as examples, measures fuel consumption by mass to minimize volume temperature issues, but that is a bit more complex. Plus as someone else suggested there is no guarantee regarding how these instantaneous and average fuel consumptions are actually generated.

I agree there are many other variables that contribute to differences as well.

The important part should be to use a consistent method for measurement/calculation since comparison relative to how you drive and use your vehicle is what I think you are trying to capture rather than absolute values comparing one car to another.

The fill/odometer method still appears to have the greater opportunity for accuracy as long as you ensure you use the same procedure to decide when the tank is full!

There are huge variations in pump speed and flow rate from pump to pump and gas station to gas station. Fuel purchased divided by distance traveled only becomes accurate over a large number of fill ups. To accurately measure a single tank you actually need to drain the tank, fill it with a measured quantity of fuel, and then drain and measure the remainder when done.