[romad]

On Saturday when I filled up the tank v e r y s l o w l y, the indicated mileage was ca. 46.5, yet when I divided the 433 miles by the 7.73 gallons, I had an mpg of 56!

On Monday, I again filled up with an indicated average 49.5 mpg. Again I put in 7.73 gallons but this time I'd only driven 276 miles so my actual mpg was 36.7.

There is an obvious problem with the way the MFD calculates the "average" mpg. Is there a way to calibrate it or should I take it in to be fixed?

 

[mrv]

The MFD uses "ticks" from the #1 cylinder and then extrapolates the MPG from that.

The by-hand calculations on a per-tank basis doesn't really work too well. With the US Prius' bladder tank, its size changes depending on the outside air temperature. So, "full" in the summer is a larger tank than "full" in the winter. Besides that, how do you know that you filled it to the same point each time?

 

[DanMan32]

After several tanks, like say 5, you should have a better idea of your true mileage, and then you can compare that to your MFD average mileage over those same tanks.

Some have tried putting in the number of gallons based on the MFD MPG and the MFD distance. That may work for a bit, but either the MFD calculation will have you come up short in as in my case, in which you will have less and less in your tank, or have you come up over, in which case you will eventually overflow. Also those little roundoffs make a big difference.

Bottom line, make comparisons over at least 5 tanks.

 

[SkipHuffman]

On Saturday when I filled up the tank v e r y s l o w l y, the indicated mileage was ca. 46.5, yet when I divided the 433 miles by the 7.73 gallons, I had an mpg of 56!

On Monday, I again filled up with an indicated average 49.5 mpg. Again I put in 7.73 gallons but this time I'd only driven 276 miles so my actual mpg was 36.7.

There is an obvious problem with the way the MFD calculates the "average" mpg. Is there a way to calibrate it or should I take it in to be fixed?

So, over the two tanks you got an average of 45.9 mpg, by your calculations. And the system calculated 47.9. (both normalized by miles driven). Does not sound that inaccurate to me. Check it over then next eight tanks or so.

 

[Sanny]

Verrrry interesting...

My Prius is not broken in yet (only 3100 miles so far), but I can already see that the gas mileage is improving. That's probably due, in part, both to breaking in, and also to my changing driving habits. Over a moving four tank average (calculated from gallons per fill and miles driven), I'm getting:

39.13
40.36
41.84
41.60
42.73
43.54

Sloooowly creeping upwards. Wheee!

 

[Thomme]

Sanny -

Don't underetimate the effect of warming weather too. I got my Prius last June, and got good mileage right off the bat. When the weather got "cold" here last winter, and I was regularly using heat, defog, etc, my mileage fell off rather alarmingly. With warmer weather, I'm back to my original mileage level.

 

[Sanny]

Sanny -

Don't underetimate the effect of warming weather too.

Warm? Where is it warm? Ok, we did have 40 today, so that's something. But shoot, we had snow three times in the past 2 weeks. Don't shake your "warm" at me! :? I've had to defrost the bunnies' (yes, furry bunnies) water bottle almost every morning.

That said, I do anticipate even better mpg when Spring eventually arrives.

 

[KTPhil]

I just use the MFD values regardless. Other than curiosity (and bragging), what use is it? The absolute value isn't imortant unless you are trying for that last mile before you run oit of gas.

On all cars, I've used it as a measure of health--a drop-off not explained by driving or weather means it's time to check things out. Harder to do with a Prius that is so weather-dependent, but of some value regardless.

 

[Thomme]

Don't shake your "warm" at me! :?

Wouldn't dream of it. I learned my lesson after the restraining order.

 

[romad]

Average over 7 months (18 fill-ups) and 6,859 miles: 44.684 mpg

Temperature average during the week of the last two fill-ups: 82 degrees Fahrenheit, or 27.7777... Celsius

As I said I fill it very, very slowly, stopping when it starts clicking off after only 1 cent of gas goes in. I also remove the nozzle periodically to let any air bubbles escape.

 

[ehurtley]

Sanny -

Don't underetimate the effect of warming weather too.

Warm? Where is it warm? Ok, we did have 40 today, so that's something. But shoot, we had snow three times in the past 2 weeks. Don't shake your "warm" at me! :? I've had to defrost the bunnies' (yes, furry bunnies) water bottle almost every morning.

That said, I do anticipate even better mpg when Spring eventually arrives.

hehe.. Oregon, where Feb/March is usually in the 30's and 40's, has been having a VERY warm Winter, with highs in the 70s last week, and in the 60s for the few weeks before that.

And I have seen an accompanying milage increase, as well. (5mpg above my average for Dec-Jan.)

 

[jshore]

On Saturday when I filled up the tank v e r y s l o w l y, the indicated mileage was ca. 46.5, yet when I divided the 433 miles by the 7.73 gallons, I had an mpg of 56!

On Monday, I again filled up with an indicated average 49.5 mpg. Again I put in 7.73 gallons but this time I'd only driven 276 miles so my actual mpg was 36.7.

Doesn't it seem a tad bit unlikely to you that you really got 56 mpg on one tank and 36.7 on the next?!?! You should trust the MFD numbers more...They are taken from measurements of how long the fuel injectors are on and are very precise.

It may be true that over the very long term, the numbers from many fillups will be more accurate but they won't be on a fillup-per-fillup basis. And, having now had my Prius for close to 7000 miles, I can tell you that the lifetime gas mileage determined from the MFD and from the number of gallons of gas I've filled it up with only differ by ~1.0mpg (or a little over 2%). This finding that the MFD numbers for MPG run a few percent (say, 2-3%) high seems to be consistent with what others see.

 

[Ken S]

Verrrry interesting...

My Prius is not broken in yet (only 3100 miles so far), but I can already see that the gas mileage is improving. That's probably due, in part, both to breaking in, and also to my changing driving habits. Over a moving four tank average (calculated from gallons per fill and miles driven), I'm getting:

39.13
40.36
41.84
41.60
42.73
43.54

Sloooowly creeping upwards. Wheee!

I'm seeing roughly the same thing with mine. I've only got abotu 1500 miles on mine. It has been warming up a bit here too (it's Spring Break and it can get downright hot when driving along the beach).

 

[dontcheff]

Sanny -

Don't underetimate the effect of warming weather too. I got my Prius last June, and got good mileage right off the bat. When the weather got "cold" here last winter, and I was regularly using heat, defog, etc, my mileage fell off rather alarmingly. With warmer weather, I'm back to my original mileage level.

It is like -10C here in Scandinavia and I am getting less tham 40. Usuallly 38. But I have the auto A/C on...

 

[romad]

On Saturday when I filled up the tank v e r y s l o w l y, the indicated mileage was ca. 46.5, yet when I divided the 433 miles by the 7.73 gallons, I had an mpg of 56!

On Monday, I again filled up with an indicated average 49.5 mpg. Again I put in 7.73 gallons but this time I'd only driven 276 miles so my actual mpg was 36.7.

Doesn't it seem a tad bit unlikely to you that you really got 56 mpg on one tank and 36.7 on the next?!?! You should trust the MFD numbers more...They are taken from measurements of how long the fuel injectors are on and are very precise.

Miles PER Gallon (MPG) means the number of miles driven by the number of gallons used(replaced at fill-up). The MFD display is SUPPOSED to be showing the AVERAGE MPG since the last fill-up. It really is only guessing at it and is NOT a true indicator. The fault is in the way the fuel system is designed. The bladder is either not pliable enough in winter or too pliable in summer. Also, there should be a time/distance based flow meter in the fuel line that would send the flow rate to the computer which would then use it and the the distance traveled to compute an M/G or L/100 KM. If the MFD display was a LIFETIME average, I wouldn't be concerned, since, as I said in my previous post, the lifetime is close to the average shown on the MFD.

I still intend on looking into a way to replace the bladder with a regular tank. Regarding the PZEV rating, I didn't buy the Prius because of that, so the normal tank vapor recovery system is fine for me.

 

[DanMan32]

Maybe your car has the MFD show accurate MPG, but mine is consistenly about 2MPG below calculated. I have yet to figure out if the error is multiplicitive or additive, but since it is probably an error in flow rate of the injectors, probably multiplicitive, somewhere around 3%.

If I were to fill up the amount the MFD suggests, adding .1 more for rounding errors, I will gradually have less and less in the tank per fillup, until eventually will have little fuel left to get to the next station.

 

[mdacmeis]

On Saturday when I filled up the tank v e r y s l o w l y, the indicated mileage was ca. 46.5, yet when I divided the 433 miles by the 7.73 gallons, I had an mpg of 56!

On Monday, I again filled up with an indicated average 49.5 mpg. Again I put in 7.73 gallons but this time I'd only driven 276 miles so my actual mpg was 36.7.

Doesn't it seem a tad bit unlikely to you that you really got 56 mpg on one tank and 36.7 on the next?!?! You should trust the MFD numbers more...They are taken from measurements of how long the fuel injectors are on and are very precise.

Miles PER Gallon (MPG) means the number of miles driven by the number of gallons used(replaced at fill-up). The MFD display is SUPPOSED to be showing the AVERAGE MPG since the last fill-up. It really is only guessing at it and is NOT a true indicator. The fault is in the way the fuel system is designed. The bladder is either not pliable enough in winter or too pliable in summer. Also, there should be a time/distance based flow meter in the fuel line that would send the flow rate to the computer which would then use it and the the distance traveled to compute an M/G or L/100 KM. If the MFD display was a LIFETIME average, I wouldn't be concerned, since, as I said in my previous post, the lifetime is close to the average shown on the MFD.

I still intend on looking into a way to replace the bladder with a regular tank. Regarding the PZEV rating, I didn't buy the Prius because of that, so the normal tank vapor recovery system is fine for me.

Why do we always find someone who suggests the MFD is "guessing?"

Let's look at the facts. The Prius uses the injector on time multiplied by the rated fuel flow through the injector over a given time to arrive at the fuel used. The flow is calculated as a result of injector on times that are milliseconds in duration. This is your high resolution flow meter. The injectors are tested before assembly to the engine to verify they have no more than 1.2% error at their rated pressure and volume. The nominal error is less than 0.5%. The fuel used over time is divided by the mileage traveled. The mileage calculation is based on the rolling radius used, the speed sensor conversion (revolutions per mile), and conversion errors. The resolution on this circuit is 1/100th MPH or KPH depending on how the calculation is done. Most now use KPH. The total error in this conversion is about 2%. Now we have distance and the amount of fuel used. The resulting calculation does not result in additive error. This rarely occurs as normally the errors are of opposite sign. But let's think worst case. So I have 2.5% error. At 8 gallons used, that is 0.2 gallons. At 350 miles distance (typical distance for 8 gallons of fuel), 8 gallons used is 43.75 MPG, 8.2 gallons used is 42.68 MPG, and 7.8 gallons used is 44.87 MPG. The effect on actual versus displayed fuel economy is only 1.1 MPG!

As for my personal experience, having logged every fill-up over 10,500 miles, I find the difference between the calculated MPG (miles on ODO/gallons pumped) versus the MFD (back calculating ODO/MPG displayed = gallons I should have pumped) is within 0.7% of each other. Given the bladder variation, this error may further decrease with time as more fill-ups further minimize this effect, but my result is close to the 1% error one should expect from this type of calculation, and further supports Toyota's claim that the displayed MPG is "highly accurate."

The MFD displayed mileage is not a "guess." It is a 16 bit calculation with more inherent accuracy than the fuel pump you fill your car with.

 

[DanMan32]

I agree it doesn't guess, but I beg to differ on the the percent error you state. After 18K, I have a total of 3% loss of fuel. For the miles that I have included the MFD MPG, which is 16K, I filled the car with 307.30 gallons, but the MFD says I used 297.58 gallons, for an difference of 9.72 gallons, about a whole fillup or 3.1% error relative to gallons filled.

I think you are off by the injector tolerance. According to service manual, volume for 15 seconds can deviate from 36-46cm^3 with a 10cm^3 between injectors. That's a +/-12.5% deviation tolerance from average, and a 25% tolerance between injectors! The closed loop operation of the ECM would deal with this deviation, though it would not know why it has to adjust (it would think it was for environmental reasons) and therefore could not inform the MFD about the flow rate deviation.
Now if there were a macro flow rate meter, then it could fine tune the assumed injector flow rate.
The MFD measurement of the MPG is precise, but not accurate.

Leakage tolerance is 1 drop per 12 minutes, probably not relevant to the error.

 

[mdacmeis]

I agree it doesn't guess, but I beg to differ on the the percent error you state. After 18K, I have a total of 3% loss of fuel. For the miles that I have included the MFD MPG, which is 16K, I filled the car with 307.30 gallons, but the MFD says I used 297.58 gallons, for an difference of 9.72 gallons, about a whole fillup or 3.1% error relative to gallons filled.

I think you are off by the injector tolerance. According to service manual, volume for 15 seconds can deviate from 36-46cm^3 with a 10cm^3 between injectors. That's a +/-12.5% deviation tolerance from average, and a 25% tolerance between injectors! The closed loop operation of the ECM would deal with this deviation, though it would not know why it has to adjust (it would think it was for environmental reasons) and therefore could not inform the MFD about the flow rate deviation.
Now if there were a macro flow rate meter, then it could fine tune the assumed injector flow rate.
The MFD measurement of the MPG is precise, but not accurate.

Leakage tolerance is 1 drop per 12 minutes, probably not relevant to the error.

Two elements here. First, the error you see has to be rationalized carefully. What I see happen here time and time again is people assume the pumps at the stations are accurate. Most are checked very infrequently and the tolerance allowed is several percent. Until someone takes a precision measuring device to the station and measures each gallon added, the only way to account for error in the fuel economy calculation is to utilize what is known, and that for now (for me, anyway) is the tolerances of the various players that comprise the calculation. Even my experience can be attributed to error on both sides - the vehicle calculation and the pump, and each are occuring simultaneously. A station that pumps 9.9 gallons when the pump reads 10.0 is likely not going to get reported or caught. In fact, I believe this is within allowable error. That is 1%!

Second, the injectors are indeed 1.2%, last I knew, or very close to it. You cannot maintain emissions or good combustion control with 10%+ tolerance injectors. Even aftermarket injectors are built to these tolerances. Check http://www.mrgasket.com for example. Their aftermarket injectors are +/- 1.5%. The service manual is a different animal. There is always a large tolerance placed on parts before they are advised for repair. Otherwise the defective component is not able to be located. The measurements in the manual are generally shown for definatively "broken" versus problematic, and account for the range and tolerance of acceptable fuel pressure from the fuel pump. With only one primary O2 sensor for the 4 cylinders, an error this large would result in leaning or richening of the other cylinders and could cause problems. One injector at +10% and another at -10% would be a real problem. However this does demonstrate the real issue here - if the injector flow rates are out of tolerance, the fuel used calculation is off. But until you remove the error from the pump at the stations, which no one I know is currently measuring, what you calculate versus what you used cannot be resolved.

 

[Sanny]

This discussion makes my head hurt. So does the $2700 computer one. Ouch.

But carry on...I'll just sit in the peanut gallery with Hep and Melgish in the peanut gallery, and sip amaretto on the rocks. Maybe we'll even find a high spirited F for Mr. Fred!