Accuracy of Honda Gas Mileage Gauge

My 04 Civic Hybrid has a digital gauge that displays the current miles per
gallon consumption as well as mpg for distance on the trip meter. Can
anyone tell me how these guages work and what are the spec tolerances?

I always set the trip meter at each fill-up to automatically get mpg, but I
also calculate the mpg based on the gas pump fillup quantity. Those results
are always different, sometimes by almost 1 mpg over 500 miles. Differences
vary so neither is always higher or lower. I believe gas pumps are quite
accurate for amount dispensed, but the difference probably is caused by
inconsistency in gas pump cut-off nozzles so that the gas tank is filled to
same level. I usually try to use the same pump but my guess is that the
nozzle trip point is not accurate and is affected by the outdoor
temperature.

Anyway, I didn't know which was likely to be more accurate and wondered if
someone knew. Thanks

 

VW wrote:

<snip>

I have found that the pump shutoff point varies quite a bit. I Measured
over 10 fill ups where I picked pumps more or less at random (but always
at the same station). Driving conditions were consistent during the
entire time. The average gas mileage during the test was 32.3 MPG. Using
this number I calculated the expected number off gallons needed to fill
the tank and compared with the actual. I found that the actual ranged
from 0.98 gallons below expected to 0.82 gallons above expected for a
range of 1.8 gallons. Standard deviation was 0.64.

I did a control experiment using the same pump every time 9 times in a
row. During this test the average MPG was 32.9. During this test, actual
was from 0.35 gallons below expected to 0.25 gallons above expected for
a range of 0.6 gallons. Standard deviation was 0.20.

 

Fuel delivery at a gas station is set to automatically adjust the volume
to account for variations in temperature. In this part of the world
it's adjusted to 15*C. If it's warmer than that then you will actually
receive more fuel than the pump will read. This could possibly explain
part of the variation.

Just out of curiosity what's your mileage been like in that vehicle?

Peter H

 

Randolph,
Thanks for the data. It confirms that the fill-up quantity is quite
variable. What size tank do you have? That 0.6 gal diff on 11 gal fill-up
would really throw off the mileage calculation. 0.2 gal is much better but
hard to always get same pump. Would like to know specs on the digital gauge
which I hope is more accurate. --V

 

Peter,
Thanks for your input. In the DC metro area commuting, I can get 50+ mpg
with careful driving in the right lane at 55-65 mph. If I go in faster
lanes over 60, the mpg goes down to about 43. --V

 

If worrying about 1 mpg over a distance of 500 miles is your only worry in
life then you are doing pretty good. Additionally you have too much damn
time on your hands to micro manage your mpg.

 

Look, I hate to point out the obvious, but this whole thread is really
fucked up.

You measure your mileage over an extended period of time, with respect
to miles driven. The possible variance delivered by the how much you
'top off' becomes insignificant after a few tanks. Let me provide an
example:

At mile x, you fill the car. When the low fuel light comes on, you fill
it again, recording how much gasoline you put in. In order to get a
reasonable estimate from tank to tank, you try to fill it pretty much
the same way each time (i.e. til the first click off at the pump, as
full as you can possibly get it, three squeezes for luck, or whatever)
but this doesn't matter over the long run. Let's say you put in 13.44
gallons. If you wish to know an approximate mpg for that tank, you also
write down how many miles you've driven between fillups and divide that
number by 13.44.

Repeat this over 10 tanksful, just as an example. As long as at the end
of the tenth tank you have the tank about as full as it was back at mile
x when you first filled it up, you will have a very accurate picture of
your mileage over that period. Irrespective of exactly how full you got
the tank each individual time, your gas mileage is the total number of
miles you've driven since mile x, divided by the total number of gallons
you've put in. This is true even if several times out of the 10
fillups, you only put in $5 'cause that's all you had with you. The
fact that you didn't fill the tank all the way, or even half way, makes
no difference at all. You *do* want to fill the car pretty much to the
same level as it started on the fillup where you're going to measure -
in this example, the 10th fillup.

If you put in 13.44 gals, 13.76 gals, 2.55 gals, 12.54 gals, 13.12 gals,
14.65 gals, 12.88 gals, 3.14 gals, 13.23 gals and 12.89 gals, your total
gallons put in is 112.2. Let's say your miles driven (current mileage -
x) is 2964; your mileage over that period of travel (2964 miles) is
26.42 miles per gallon.

This does not depend on micromeasurement, you don't have to do any
calibration, you don't have to buy anything but a pencil (and the gas,
of course), you don't have to go to the same pump (or station), you
don't have to buy all your gas during the same season - in fact, the
more tanksful you get, the less possible effect any variable can have on
your mileage calculation, and the accuracy of the gas gauge doesn't
enter into the picture in *any* way. All it requires is a very basic
knowledge of 5th grade math.


RFT!!!
Dave Kelsen

 

You can't make it any simpler than that---People want to make it Rocket
Science. The only basic variable(s) are the accuracy of the gas pump
metering system and the writing down of the numbers. To view the
variability of the calculations all one has to do is to compare individual
gas fillups against the long term, high volume/long miles. It becomes very
obvious the longer between calculations the more the small stuff washes out
and accuracy increases.
MLD

 

Sorry if this didn't come across clearly in my original posting, so maybe
this will clarify: I was trying to get data to compare the accuracy of the
computer/electronic gauge compared to the traditional fillup/divide method
which everyone knows has variability. I agree that averages over more
fillups will give a good average long term mpg, however, at least I
personally sometimes want to know mpg for specific shorter trips.

What I really hoped is that someone had the specs on the digital gauge or
had other experience/knowledge with them as they're getting more common in
cars. If the computer is reasonably accurate I won't bother to do the
fillup recording and calculations (which I've done for years to monitor my
car performance). However, my experience with electronics is that these new
gauges might not be very accurate, say 5% over total range, to keep their
cost down (as with fuel tank gauges).

And quite often there are web postings from people unhappy with "low" gas
mileage which can be difficult to assess if it is not measured accurately or
the different operating conditions are invalid for comparison. Even the EPA
ratings on new cars are subject to debate.

Thanks for the the responses anyway.

 

I would say that your best bet with respect to measuring specific
shorter trips would be to do the same as I outlined before, but
additionally to reduce the variables as much as possible. Go to the
same station/pump when possible, and fill up the same way if possible,
for example stop when the pump first clicks off automatically.

My (2001) Odyssey's gauge and low fuel light are famously inaccurate;
Odyssey owners have found that in general, when the gauge reads empty,
you have *at least* three gallons left. Some owners report nearer to
five gallons! I don't know of any existing generalized data with
respect to the accuracy of modern gauges.

RFT!!!
Dave Kelsen

 

Dave,
Thanks for your response. I should explain that the new digital computer
gauge in the Civic Hybrid provides both a instantaneous reading of mpg and
it also provides the mpg for trips set by the 2 trip odometers. By reseting
the trip odometer at fill up, it will automatically displays mpg for that
trip until I fill up again, or anyother distance or trip. This is easier
than recording fillup quantity and doing the calculation (even if not hard).
My interest (as a techie type and being gas conscious) is how accurate is
the gauge? I assume it uses a flow metering device and sensor to measure
gas consumption but I don't know if it is more accurate than the gas pump
method. Hopefully it is, but thats why I was asking. I am also going to
post on Cadillac newsgroup since they've had these type of computer gauges
for quite some time. Cadillac computer/f"uel data center" even provides an
estimate of the miles you can drive before running out of gas (which I heard
is not very accurate). Probably bored everyone so sorry. -V

 

Gotcha; I had one of these in my 96 Chrysler LHS; they have been in many
cars for many years now - my brother-in-law has one in his Trailblazer,
my mother-in-law had one in her Lincoln.

My experience was that it was a useful guide, but nothing delivers the
accuracy of my own measurements. They may have gotten to the point
where they *do* though, now. It's been a few years.

Good luck.


RFT!!!
Dave Kelsen