C-max MPG not calculated correctly

[DaveofDurham]

From the OBD manual:

 

"Fuel consumed" is continuously calculated based on PCM fuel pulsewidth summation as a percent of fuel tank capacity.

 

I don't see any sign of a flow meter in any of the discussions so it looks like they calculate it based on how long the injectors are on. Probably not the most accurate way to do it.

 

Thanks.

 

What does  >>>"Fuel consumed" is continuously calculated based on PCM fuel pulsewidth summation as a percent of fuel tank capacity.>>> mean and how is "pulsewidth" converted to gallons?

[DaveofDurham]

I do not know or understand how all this is calculated - I tend to do very simple math.....however, I recently started using my remote start to warm up the car when the morning temps were down in the 40s and below.  Took me a few days to figure out what was happening, but it appears that when my car is running in remote start, using gas and warming up, the trip meter doesn't register the time or the fuel used.  So all of the sudden it was cold and my FE was better than in the past, which sent me looking for an explanation.  My normal mpg heading to work is about 47.  During the cold snap when I let the car warm up in remote start mode I was getting 52.

 

Don't know that this helps.  I hope it didn't confuse the issue.

 

Fascinating.  By trip meter do you mean the individual drive (trip meter) that comes up on the small left screen when you turn the car off when you get to your destination OR do you mean the trip odometers?

[spyburn]

I mean trip 1 and trip 2.  I reset trip 1 with each fill up.  I reset trip 2 for short trips - to see how I did on my way to and from work.  I'm in the habit of resetting after I get where I'm going, so when I start out the next time, trip 2 will have just a few seconds on it and no distance.  When I started my car in remote mode, then got in it later (about 10 minutes), I confirmed that it didn't count that time because I looked at the trip 2 display and it had less than 5 seconds on it.

 

And for the one that shows up after you turn off the car, I never thought about comparing that to my trip 2.  I will try to remember to do that next time.

Edited November 17, 2013 by spyburn

[mbedit]

From the OBD manual:

 

"Fuel consumed" is continuously calculated based on PCM fuel pulsewidth summation as a percent of fuel tank capacity.

 

I don't see any sign of a flow meter in any of the discussions so it looks like they calculate it based on how long the injectors are on. Probably not the most accurate way to do it.

 

Ok, this is probably the issue right here. Most gas tanks will typically hold more gas than the rated capacity. Thus, if the rated capacity is 11.5 Gal and I'm sticking 12.5 Gal in it the computer will mess up.  Sounds like I need to run an experiment where I pump it up to 10 Gal, and then when it says 10 Gal is consumed, pump ten more and do that  several times to see if its accurate that way. in otherwords, I'm wondering if over filling could cause the issue?

 

spyburn... thats just odd... don't know what to say about that.

Edited November 17, 2013 by mbedit

[jdbob]

Thanks.

 

What does  >>>"Fuel consumed" is continuously calculated based on PCM fuel pulsewidth summation as a percent of fuel tank capacity.>>> mean and how is "pulsewidth" converted to gallons?

 

As an analogy, think of a garden hose with a spray handle. Let's pretend that if you hold the spray handle all the way open for 1 second that it will dispense 0.1 gallon. What happens if you do that 100 times? It's just 100 * 0.1 = 10 gallons.

 

A fuel injector is a small, electrically operated valve that passes gasoline instead of water. How long it is open can be precisely controlled by the PCM (Powertrain Computer). At some point someone figured out how the amount of fuel dispensed varies with how long the valve is opened. The PCM just adds up all those very small amounts of gas each time a fuel injector is opened and converts that fuel usage into gallons or liters. Of course if the fuel pressure varies, so will the calculation, hopefully the PCM compensates for that.

[ptjones]

Also I think the more Ethanol in gas the more of a difference there will be too. As an experiment fill your tank up with Pure Gas and see how close the difference is. :)

 

Paul

[DaveofDurham]

Also I think the more Ethanol in gas the more of a difference there will be too. As an experiment fill your tank up with Pure Gas and see how close the difference is. :)

 

Paul

How many fill ups have you done w pure gas and have they all shown better agreement btw the gallons used per the trip odometer and gallons added per the pump?

 

What difference btw pure gas and ethanol/gas blend do you think leads to better agreement with pure gas than with ethanol/gas blend?

[ptjones]

How many fill ups have you done w pure gas and have they all shown better agreement btw the gallons used per the trip odometer and gallons added per the pump?

 

What difference btw pure gas and ethanol/gas blend do you think leads to better agreement with pure gas than with ethanol/gas blend?

Pure Gas seems to get the best match up to Smart Gauge and EPA testing is done used it.  I have 3 fill ups with Pure Gas and all have had better MPG. I believe most everyone else has had similar results. But Pure Gas $4/gal here and not worth using regularly.

 

Paul

[DaveofDurham]

Pure Gas seems to get the best match up to Smart Gauge and EPA testing is done used it.  I have 3 fill ups with Pure Gas and all have had better MPG. I believe most everyone else has had similar results. But Pure Gas $4/gal here and not worth using regularly.

 

Paul

 

I get the better MPG with pure gas than with a gas/ethanol blend.  Gasoline has higher energy content than ethanol, so pure gas gives better MPG than a gas/ethanol blend.

 

What I have been trying to understand is [1] how much data (how many fill ups with pure gas) you have to lead you to conclude that there is better agreement between the gallons used per the trip odometer and gallons added per the pump when you use pure gas than when you use gas/ethanol blend AND [2] why there would be better agreement between the gallons used per the trip odometer and gallons added per the pump when you use pure gas than when you use gas/ethanol blend?

 

If you have only filled up with pure gas 3 times then it seems to me that you don't have much info to go on to conclude that there is better agreement between the gallons used per the trip odometer and gallons added per the pump when you use pure gas than when you use gas/ethanol blend.  

 

I have only used gas/ethanol blend.  Sometimes I get very good agreement btw the gallons used per the trip odometer and gallons added per the pump, frequently I pump more than the car says I have used and sometimes I pump less - over 43 fill-ups the car says I have used 4.7% less gallons than I have pumped.

[mbedit]

I've used pure gas and my results are the same.

 

The amount of energy in the gas shouldn't make a difference. Its a simple calculation of Miles divided by gallons. If you use gas with lower energy you should simply go less miles  for the same number of gallons.

 

I guess the issue is actually the gauges according to this article http://www.edmunds.com/fuel-economy/your-fuel-economy-gauge-is-fibbing.html

 

If this article is to be believed, the gauge counts the number and duration of fuel injections and makes an assumption of the volume for each injection, and it seems they are all off because 92 percent of cars use the same gauge.

 

But since mine went from about 4%  to 8%  after the reprogramming, I think they certainly can get it closer to accurate.

[obob]

http://www.tirerack.com/tires/tiretech/techpage.jsp?techid=177

 

"New, Full-Treaded Tires Travel Farther per Tire Revolution Than Shallow-Treaded, Worn Tires"

"While the 100-mile test distance didn't change, essentially the vehicle's odometer overstated the distance traveled by about 1.5% when equipped with the worn out tires."

 

"New, Full-Treaded Tires Generate More Rolling Resistance Than Shallow-Treaded, Worn Tires"

"Tire rolling resistance gradually drops by about 20% during the life of a tire as the tread wears from its original molded depth to worn out.  

 

.... drivers should expect to experience a potential 2% to 4% decrease in mpg."  (going to new tires)

 

Though it might not be noticeable as measured by the car because the tires are bigger. 

 

Maybe someday miles in the car will be measured by GPS - or perhaps a combination, to use traditional measurement until the GPS becomes reliably online.

Edited November 19, 2013 by obob

[HPRifleman]

To say that the C-Max is not calculating mileage correctly assumes the pump at your local gas station is accurate. We could just as easily say that the pump is over reporting the amount of fuel it is putting into the tank.

 

Let's look at some actual data. Below are the car and pump volumes for my last 10 fill ups with the percentage difference.

 

Tank---Gal (car)---Gal (pump)---Diff

...1........11.50..........11.62..........-1%

...2........12.47..........12.52...........0%

...3........12.37..........12.81..........-3%

...4........11.03..........11.50..........-4%

...5..........8.12...........8.04...........1%

...6........10.80..........11.13..........-3%

...7........10.08..........10.00...........1%

...8........11.90..........12.56..........-5%

...9........11.51..........12.54..........-9%

..10.......12.63..........12.14...........4%

 

A negative difference means the pump displayed more fuel than the trip meter did. While there is a difference (about -2% overall) between the two there are several instances where the pump reported volume is lower than the trip meter.

 

I just want to illustrate that we should be cautious before making statements that the car is "incorrect". It may or may not be but we don't know.

[obob]

To say that the C-Max is not calculating mileage correctly assumes the pump at your local gas station is accurate. We could just as easily say that the pump is over reporting the amount of fuel it is putting into the tank.

 

Let's look at some actual data. Below are the car and pump volumes for my last 10 fill ups with the percentage difference.

 

Tank---Gal (car)---Gal (pump)---Diff

...1........11.50..........11.62..........-1%

...2........12.47..........12.52...........0%

...3........12.37..........12.81..........-3%

...4........11.03..........11.50..........-4%

...5..........8.12...........8.04...........1%

...6........10.80..........11.13..........-3%

...7........10.08..........10.00...........1%

...8........11.90..........12.56..........-5%

...9........11.51..........12.54..........-9%

..10.......12.63..........12.14...........4%

 

A negative difference means the pump displayed more fuel than the trip meter did. While there is a difference (about -2% overall) between the two there are several instances where the pump reported volume is lower than the trip meter.

 

I just want to illustrate that we should be cautious before making statements that the car is "incorrect". It may or may not be but we don't know.

 

So while a gas station pump adjusts fuel volume for the temperature of the fuel, the car computer probably doesn't or perhaps approximates it with ambient temp. 

 

http://answers.yahoo.com/question/index?qid=20070608101130AA79rH6

 

This kind of says for every 15 degree change in the temperature of the fuel, there is a 1.425 % change in volume.

 

I do know it is substantial for I have experienced getting gas and topping it off just before parking my car in the sun and then coming out later and there is spilled gasoline.

Edited November 19, 2013 by obob

[DaveofDurham]

So while a gas station pump adjusts fuel volume for the temperature of the fuel, the car computer probably doesn't or perhaps approximates it with ambient temp. 

 

http://answers.yahoo.com/question/index?qid=20070608101130AA79rH6

 

This kind of says for every 15 degree change in the temperature of the fuel, there is a 1.425 % change in volume.

 

I do know it is substantial for I have experienced getting gas and topping it off just before parking my car in the sun and then coming out later and there is spilled gasoline.

 

Does your state require that gas pumps adjust fuel volume for temperature of the fuel?  While searching for some info about how accurate gas pumps are supposed to be, I found several places that said that in the US the pumps do not adjust fuel volume for temperature of the fuel and that in Canada they do.  In defense of not adjusting, there was a discussion of how/why the fuel temp does not vary that much because of underground tanks.  

 

I am going to check with someone in Raleigh on what the NC regulations are for fuel pump accuracy.  I know from their website that the check is done with a calibrated 5 gallon container and that the container is filled twice (one rapid fill and one slow fill) to check that the pump is within the required accuracy.

Edited November 20, 2013 by DaveofDurham

[Adair]

I do not know or understand how all this is calculated - I tend to do very simple math.....however, I recently started using my remote start to warm up the car when the morning temps were down in the 40s and below.  Took me a few days to figure out what was happening, but it appears that when my car is running in remote start, using gas and warming up, the trip meter doesn't register the time or the fuel used.  So all of the sudden it was cold and my FE was better than in the past, which sent me looking for an explanation.  My normal mpg heading to work is about 47.  During the cold snap when I let the car warm up in remote start mode I was getting 52.

 

Don't know that this helps.  I hope it didn't confuse the issue.

HUH??? I was told it work exactly the opposite, so I've been hesitant to use the remote start, even tho morning temps are in the 20s some days.

[mbedit]

To say that the C-Max is not calculating mileage correctly assumes the pump at your local gas station is accurate. We could just as easily say that the pump is over reporting the amount of fuel it is putting into the tank.

 

Let's look at some actual data. Below are the car and pump volumes for my last 10 fill ups with the percentage difference.

 

Tank---Gal (car)---Gal (pump)---Diff

...1........11.50..........11.62..........-1%

...2........12.47..........12.52...........0%

...3........12.37..........12.81..........-3%

...4........11.03..........11.50..........-4%

...5..........8.12...........8.04...........1%

...6........10.80..........11.13..........-3%

...7........10.08..........10.00...........1%

...8........11.90..........12.56..........-5%

...9........11.51..........12.54..........-9%

..10.......12.63..........12.14...........4%

 

A negative difference means the pump displayed more fuel than the trip meter did. While there is a difference (about -2% overall) between the two there are several instances where the pump reported volume is lower than the trip meter.

 

I just want to illustrate that we should be cautious before making statements that the car is "incorrect". It may or may not be but we don't know.

 

HPRifelman, that's nice data. but that's not at all indicative of the data I'm seeing. I'm never seeing a positive difference. I'm always seeing between 3/4 of a gal and 1 Gal negative amount. Also... I've had the luxury of having two c-Max's (I jest because one was lemoned) and both behaved this way... never once reporting a positive difference, only negative. Also... gas pumps are actually calibrated and certified to pump the correct amount of fuel. Saying that all pumps are that inaccurate is a stretch. It doesn't add up to me.

 

Obob, I think you have an interesting point. There should be an easy way to test that idea. I  think your suggesting that filling up in say the peak of the day when its warmest, should result in a positive difference since the fuel will expand so it takes less to fill up the same volume.  The thing I would counter is that fuel tanks are buried and the temp of fuel that is pumped doesn't actually change dramatically, thus the argument of fuel expanding or contracting seems to be a dead end to me.

 

Since pumps are strictly calibrated and tested to meet those standards and the fuel doesn't actually expand or contract as much as people may think, I'm still highly skeptical of the way the car is actually calculating the value. More so, I find it odd that after the firmware upgrade it went from being 1/2 a gal off to 3/4-1 gal off.

[fbov]

Two things..

 

First, I agree that in the contiguous US, temperatures 10 feet below ground level are the same just about everywhere, all the time. Pumps are tested for accuracy... I'll take that data over what the car thinks any day.

 

Second, is there a consistent difference in fueling procedures that could explain the discrepency? Do results change if you top-off compared with less-than-maximum fill levels? I see a consistent discrepancy in my fueling data, but I'm a topper...

 

Have fun,

Frank

[ptjones]

Two things..

 

First, I agree that in the contiguous US, temperatures 10 feet below ground level are the same just about everywhere, all the time. Pumps are tested for accuracy... I'll take that data over what the car thinks any day.

 

Second, is there a consistent difference in fueling procedures that could explain the discrepency? Do results change if you top-off compared with less-than-maximum fill levels? I see a consistent discrepancy in my fueling data, but I'm a topper...

 

Have fun,

Frank

It doesn't matter what you do , because in the long run it all averages out. I've have over 70 fillups and almost all of them were more gal. at the pump. I think the Smart gauge is calibrated for Pure Gas, because that is the way the car is tested. IMO :)

 

Paul

[DaveofDurham]

It doesn't matter what you do , because in the long run it all averages out. I've have over 70 fillups and almost all of them were more gal. at the pump. I think the Smart gauge is calibrated for Pure Gas, because that is the way the car is tested. IMO :)

 

Paul

 

What characteristic of pure gas (compared to a gas/ethanol blend) or of the smart gauge system do you think could/would result in the smart gauge being more accurate for pure gas than for a gas/ethanol blend?

[DaveofDurham]

To say that the C-Max is not calculating mileage correctly assumes the pump at your local gas station is accurate. We could just as easily say that the pump is over reporting the amount of fuel it is putting into the tank.

 

Let's look at some actual data. Below are the car and pump volumes for my last 10 fill ups with the percentage difference.

 

Tank---Gal (car)---Gal (pump)---Diff

...1........11.50..........11.62..........-1%

...2........12.47..........12.52...........0%

...3........12.37..........12.81..........-3%

...4........11.03..........11.50..........-4%

...5..........8.12...........8.04...........1%

...6........10.80..........11.13..........-3%

...7........10.08..........10.00...........1%

...8........11.90..........12.56..........-5%

...9........11.51..........12.54..........-9%

..10.......12.63..........12.14...........4%

 

A negative difference means the pump displayed more fuel than the trip meter did. While there is a difference (about -2% overall) between the two there are several instances where the pump reported volume is lower than the trip meter.

 

I just want to illustrate that we should be cautious before making statements that the car is "incorrect". It may or may not be but we don't know.

 

I spoke to someone at the NC Dept of Agriculture, Standards Division, Measurement Section this afternoon.  I asked him how much variation is allowed when gas pumps are tested.  He told me that the same standards apply throughout the US.  These standards are presented in NIST Handbook 44 which is adopted by the Nat'l Conference on Weights and Measures.  This handbook is available online at

 

In NC the testing is done using a calibrated 5 gallon container.  The permissible variation is described in table T.2 on page 3-20 of the handbook - see the statement after the * at the bottom of the table (the statement after the * is for "test drafts" less than 10 gallons).   For a new pump or a pump refurbished within 30 days of testing the permissible variation is +/- 3 cubic inches per 5 gallons (meter reading on the pump).  One gallon is 231 cubic inches, so 5 gallons are 1155 cubic inches and +/- 3 cubic inches is +/- 0.26 percent.  For a pump in service for more than 30 days the permissible variation is +/- 6 cubic inches per gallon which is +/- 0.52 percent.

 

So, the permissible variation at the pump is much less than the variation we are seeing between the pumped gallons and the car's report of gallons used. 

 

Here is a video from a company in Canada about a more modern and sophisticated way to check pump accuracy.

 

[ptjones]

What characteristic of pure gas (compared to a gas/ethanol blend) or of the smart gauge system do you think could/would result in the smart gauge being more accurate for pure gas than for a gas/ethanol blend?

From  what I understand all EPA testing is done with Pure Gas. It is interesting that with Pure Gas the Smart Gauge works more accurately for gal. and MPG.

 

Paul

Edited November 21, 2013 by ptjones

[HPRifleman]

I spoke to someone at the NC Dept of Agriculture, Standards Division, Measurement Section this afternoon.  I asked him how much variation is allowed when gas pumps are tested.  He told me that the same standards apply throughout the US.  These standards are presented in NIST Handbook 44 which is adopted by the Nat'l Conference on Weights and Measures.  This handbook is available online at

 

In NC the testing is done using a calibrated 5 gallon container.  The permissible variation is described in table T.2 on page 3-20 of the handbook - see the statement after the * at the bottom of the table (the statement after the * is for "test drafts" less than 10 gallons).   For a new pump or a pump refurbished within 30 days of testing the permissible variation is +/- 3 cubic inches per 5 gallons (meter reading on the pump).  One gallon is 231 cubic inches, so 5 gallons are 1155 cubic inches and +/- 3 cubic inches is +/- 0.26 percent.  For a pump in service for more than 30 days the permissible variation is +/- 6 cubic inches per gallon which is +/- 0.52 percent.

 

Thanks for doing this research.

[mbedit]

Thanks for all the data folks. PtJones, I guess I too am interested in finding out more about how the "Smart" Gauge works, as I would assume (and it sounds like in correctly) that the a gauge would want to accurately report volume. I guess I'm not understanding what or why knowing the "energy content" of the gas is needed to make a calculation for the MPG or any of the trip information?

 

I've concocted my own experiment... and I guess some members can tell me if they think this is flawed... Upon the car notifying me that the fuel is low, I promptly (with-in a few miles) fueled up exactly 10 gallons. I'm interested to see how close to 10 gallons it will be when I refuel again.

 

My assumption here is that the refuel light/gauge comes on about the same time each time. I'm also assuming that is a different sensor than the smart gauge used for fuel calculation. I don't think it'll be dead nuts accurate, but I would expect to be able to be with-in a tenth of a gallon accuracy. That would be it working.

 

What I'm hoping to prove or disprove here is that when I fill up and top off, perhaps (as suggested) as the fuel warms up and expands, i'm actually loosing some to leakage? I would have assumed that the tank was supposed to be a closed and sealed system as some states (like California) have laws that are supposed to prevent fuel spillage, but maybe something is wrong and the capless system isn't sealing right?

 

Anyhow, Since I'm not topping off, that should remove that variable, and it'll be interesting to see how close to 10 gallons the trip gauge says I am when the refuel light comes on.

 

Of course, I'll probably need to do this 3 or 4 times to get some trending data.

 

And Happy Thanksgiving everyone!

[ptjones]

Thanks for all the data folks. PtJones, I guess I too am interested in finding out more about how the "Smart" Gauge works, as I would assume (and it sounds like in correctly) that the a gauge would want to accurately report volume. I guess I'm not understanding what or why knowing the "energy content" of the gas is needed to make a calculation for the MPG or any of the trip information?

 

I've concocted my own experiment... and I guess some members can tell me if they think this is flawed... Upon the car notifying me that the fuel is low, I promptly (with-in a few miles) fueled up exactly 10 gallons. I'm interested to see how close to 10 gallons it will be when I refuel again.

 

My assumption here is that the refuel light/gauge comes on about the same time each time. I'm also assuming that is a different sensor than the smart gauge used for fuel calculation. I don't think it'll be dead nuts accurate, but I would expect to be able to be with-in a tenth of a gallon accuracy. That would be it working.

 

What I'm hoping to prove or disprove here is that when I fill up and top off, perhaps (as suggested) as the fuel warms up and expands, i'm actually loosing some to leakage? I would have assumed that the tank was supposed to be a closed and sealed system as some states (like California) have laws that are supposed to prevent fuel spillage, but maybe something is wrong and the capless system isn't sealing right?

 

Anyhow, Since I'm not topping off, that should remove that variable, and it'll be interesting to see how close to 10 gallons the trip gauge says I am when the refuel light comes on.

 

Of course, I'll probably need to do this 3 or 4 times to get some trending data.

 

And Happy Thanksgiving everyone!

I don't think you are loosing any gas. Thinking about the problem I have come up a couple of ideas that it could make a difference. Speed, Hwy VS City and style of driving (P&G vs not). On my record runs I used a lot of P&G and my gal. were very close vs HWY, I'm usually off by as much as a gal.  I take a IPhone pic of every fill up so I could compare to Fuelly #'s, you could do the same comparing the two,  City vs HWY. I don't know if there is anything we can do about this problem. I just know by experience that I can go 12.8gal with no worries and also when fuel light comes on I've got 2gals left. :)

 

Paul

Edited October 12, 2014 by ptjones

[userw014]

Does anyone know about how the gasoline volume/flow is measured by the C-Max?  Does it need temperature compensation?  (I'm sure that the temperature fluctuations in any car are a lot greater than in the pump - the fuel coming from underground tanks is going to govern the temperature of the fuel gauges in pumps at the stations, whereas the temperature of the gasoline in the automobile fuel tank is going to be ambient temperature - -5F to 105F is what I experience in Ann Arbor.

 

If the fuel gauge reports different values for the same volume of gas, then there might be an issue there.  (I.e.: imagine the conduit through which the gas flows increasing or decreasing in size as temperature increases or decreases.)

 

But I Am Not A Mechanical Engineer - certainly not a fuel-systems engineer.