Inaccurate mpg reading?

[rdin]

Hi all,

First post on this forum. Today I believe I might've found something odd with the mileage computer.

When I got my I2, I filled up the gas tank and reset trip A. Today was the first time since then that I've gotten gas, and I filled up the tank again.

The display for trip A showed these stats:
46.9 average MPG
382.9 total trip length

However, the gas pump receipt showed that I pumped in 8.642 gallons, which is weird, because 382.9/8.642 = 44.3 mpg.

Does anybody else get discrepancies?

Thanks!

[buglermcd]

Yep, just get used to it. The computer over-estimates things by 5-10%. Apparetly most hybrids do this, except for the Honda Civic.

[jpleong]

Some over-state, some under-state. A few other Insight 2 owners have reported a 1-3mpg discrepancy while others have reported it is spot on.

There are other factors that can affect your calculations such as using the same pump every time, filling only to the first click, whether the odometer is actually correct, etc...

JP

[waterdog]

I wonder if some of the consistent downward discrepancy is due to idling time. If I were designing a mileage calculator for a car, I would not include the gas used during idling, but rather only the mpg while moving. I don't know how they actually work, but this approach would make sense to me - and would be one factor that would tend to make the MID reading consistently lower than the calculated mileage from total fuel used (the other factors noted earlier would create discrepancies, but not necessarily consistently high or low).

[jpleong]

Originally Posted by waterdog

If I were designing a mileage calculator for a car, I would not include the gas used during idling

Maybe I'm reading your post wrong but I think excluding fuel usage at idle would give too rosy a picture of mpg (ie, true mpg would be far, far less than MID reported mpg). Idle is also a huge fuel-wasting activity (that's why most cars' city mpg is so horrible) and the reason auto-stop exists.

JP

[euro sport]

IMO, I feel that unless you top off the tank to the brim every time, then calculating mileage manually can be fairly inaccurate. I don't care if you visit the same station and use the same pump, it still can produce different results depending on when you decide to stop pumping gas.

As an example, I ran my Insight until the MID said I had 9 miles remaining on my tank of gas. As I filled up at an Exxon station, the pump's auto-stopped feature cut out at 9.6 gallons. I have a tendency to fill my tanks close to the brim and in this situation, I was able to add another full gallon to the tank before I decided to officially stop (required a lot of clicks, but whatever). 10.6 total gallons went into my tank.

So here is the potential range in MPG:

550 miles / 9.6 gals = 57.29mpg (after 1st click)
550 miles / 10.6 gals = 51.89mpg (after the billionth click)

So in my situation, whenever I decide to stop pumping, could affect my overall perceived MPG. My MID said I got 53.0mpg as I pulled up to the gas station. If I called it quits when the gas pumped registered 10.4 gallons, then the MID would have been virtually spot on. I think having such a small tank can produced crazy differences from one driver to the other.

I have no problems just relying on the MID for a fairly accurate measurement.

[waterdog]

Originally Posted by jpleong

Maybe I'm reading your post wrong but I think excluding fuel usage at idle would give too rosy a picture of mpg (ie, true mpg would be far, far less than MID reported mpg). Idle is also a huge fuel-wasting activity (that's why most cars' city mpg is so horrible) and the reason auto-stop exists.

JP

My point is that I would want the MID to show the mpg excluding idle time, because I can easily calculate the aggregate total mpg including idle time from the total gas used and the odometer reading (subject to all the caveats noted above), but I have no way of calculating my mpg while actually driving unless the car does it for me. Therefore, I would rather have the car calculate the number than I can't calculate, rather than the one I can. I have no idea if that is the way the car designers at Honda think, but if they do, then the calculation approach could explain a consistently slightly higher mpg from the MID than from a hand calculation.

With regard to your comment that the mpg excluding idle time would be far less than the total - I think that might be true with a non-hybrid in city driving, but for a hybrid, I would not expect the difference to be large, and the 1-3 mpg difference that many people see between the MID mpg and their hand calculated mpg in my view could easily be caused by the MID mpg excluding the idle time in the calculation.

We would likely need a Honda design engineer to chime in to know whether my theory holds any water or not.

[jpleong]

Originally Posted by euro sport

IMO, I feel that unless you top off the tank to the brim every time, then calculating mileage manually can be fairly inaccurate. I don't care if you visit the same station and use the same pump, it still can produce different results depending on when you decide to stop pumping gas.

As an example, I ran my Insight until the MID said I had 9 miles remaining on my tank of gas. As I filled up at an Exxon station, the pump's auto-stopped feature cut out at 9.6 gallons. I have a tendency to fill my tanks close to the brim and in this situation, I was able to add another full gallon to the tank before I decided to officially stop (required a lot of clicks, but whatever). 10.6 total gallons went into my tank.

So here is the potential range in MPG:

550 miles / 9.6 gals = 57.29mpg (after 1st click)
550 miles / 10.6 gals = 51.89mpg (after the billionth click)

So in my situation, whenever I decide to stop pumping, could affect my overall perceived MPG. My MID said I got 53.0mpg as I pulled up to the gas station. If I called it quits when the gas pumped registered 10.4 gallons, then the MID would have been virtually spot on. I think having such a small tank can produced crazy differences from one driver to the other.

I have no problems just relying on the MID for a fairly accurate measurement.

I totally agree with this. The only "perfect" system I can imagine for real-world MPG measurement is to run the car completely empty and then only use gallon (or 5 gallon) gas/petrol cans until you, again, run the car completely empty. That way you'd know for sure how much fuel the car consumed over the variable distance.

Without knowing with absolute certainty I guess we'll have to say, "In MID we trust."

Originally Posted by waterdog

My point is that I would want the MID to show the mpg excluding idle time, because I can easily calculate the aggregate total mpg including idle time from the total gas used and the odometer reading (subject to all the caveats noted above), but I have no way of calculating my mpg while actually driving unless the car does it for me. Therefore, I would rather have the car calculate the number than I can't calculate, rather than the one I can. I have no idea if that is the way the car designers at Honda think, but if they do, then the calculation approach could explain a consistently slightly higher mpg from the MID than from a hand calculation.

Can't you do this already? Simply reset A (or B) whenever you start moving from idle. In stop-and-go driving you'd have to remember some numbers and calculate the average but it should still enable you to do as you wish.

With regard to your comment that the mpg excluding idle time would be far less than the total - I think that might be true with a non-hybrid in city driving, but for a hybrid, I would not expect the difference to be large, and the 1-3 mpg difference that many people see between the MID mpg and their hand calculated mpg in my view could easily be caused by the MID mpg excluding the idle time in the calculation.

We would likely need a Honda design engineer to chime in to know whether my theory holds any water or not.

In regards to "idle:" I am differentiating a vehicle at idle from a vehicle that has engaged autostop. An IMA hybrid at idle is just like any other car and continues to consume fuel. The Insight 2 is even more prone to finding itself at idle rather than autostop because of its cost-saving compromises. Suburban driving in morning rush-hour traffic bore this out for me last weekend as I watched my numbers tick away as I waited for the car to warm up while trying to get into the city.

I think the calculation system is fine, as is, since a vehicle at idle is expending X amount of fuel to go 0 miles and with autostop is expending 0 fuel to go 0 miles.

I'm sorry if I seem to be missing the point of your previous posts. You're just trying to offer up some explanations as to why the MID is off, correct?

JP

[waterdog]

I don't think we are really arguing about anything useful - I was primarily offering a possible explanation for the typical discrepancies between MID mpg and hand calculated mpg - not trying to convince anyone of anything.

As a secondary comment, I was noting that that my personal preference (does not need to be shared by anyone else, and I don't need to convince anyone else) would be to have the MID do the calculation in the manner that I speculated about (i.e., don't count the idling time). If my speculation is generally correct, it would be interesting to know whether the system distinguishes between idling while in D or S (unintentional idling) and idling in P or N (intentional idling).

[jpleong]

Originally Posted by waterdog

I don't think we are really arguing about anything useful - I was primarily offering a possible explanation for the typical discrepancies between MID mpg and hand calculated mpg - not trying to convince anyone of anything.

That's what I figured

As a secondary comment, I was noting that that my personal preference (does not need to be shared by anyone else, and I don't need to convince anyone else) would be to have the MID do the calculation in the manner that I speculated about (i.e., don't count the idling time). If my speculation is generally correct, it would be interesting to know whether the system distinguishes between idling while in D or S (unintentional idling) and idling in P or N (intentional idling).

I know from my own idling that it does, indeed, count idling in Park. I did that once to keep ice cream from melting after a trip to Aldi and my trip fuel economy was an abysmal 26.7mpg.

Overall, I would like to see a more thorough and advanced computer for the MID. It would incorporate your suggestion of moving fuel economy, total fuel economy, etc... and other things like speed to mpg curves, etc... if this were linked to the Navi it would actually make it worth the extra $$$.

JP