Increase in computer vs. hand calculation MPG error

I don't really understand the point of your thread.

You state that your tires were changed, your trips are shorter, etc. Both of these factors could have affected your observed mpg difference.

Just take the displayed mpg with a huge grain of salt, it's an estimate based on a complicated algorithm after all, and you'll never be disappointed.

SCH-I535

 

Also, you'll get a better sense of the error if you calculate the percent of the error, not just the numerical difference. You say the original error was typically 2.4 mpg. As an example,

There's a world of difference between:

30 vs 32.4 mpg (8.0% error)

And

50 vs 52.4 mpg (4.8% error)

 

Grain of salt, definitely. But why is it a "complicated algorithm"? The odometer is reasonably accurate and fuel level is known as indicated by info available at the OBD port. Simple calculation.

 

The fuel level is estimated by the ECU though. All the variables like load, rpm, air density/temp, ignition timing, throttle position, etc. all factor in. But in the end, it's still an estimate, an educated guess.

SCH-I535

 

Indeed all those factors effect the actual mpg value. However, the calculated mpg is simply miles divided by gallons. Any inaccuracies are due to the measurement of those two variables. Miles are very accurate, gallons no so much. Bias by the Toyota software unknown

 

I find the displayed mpg, and my calculation are fairly consistant. The average error I'm finding is 8%, and it's a very consistant error.

Per CaliforniaBear I'm not seeing this "educated guess": you travelled a distance, it took so much to fill up, what's not to like, LOL.

 

Have you considered that there might be a slight diameter difference between the NEW tires vs. the OLD tires? For instance, the OEM Yokohama Avid S33D tire turns 831 revolutions-per-mile, while the OEM Bridgestone Ecopia EP20 tire turns 833 revolutions-per-mile.

Thus, the ECU would be measuring & calculating fuel consumption via the same algorthm but the milages (both yours' and the computers') would be slightly OFF because the true number of tire revolutions-per-mile will have changed from what the ECU was originally programmed with.

 

Consumer Reports did a article comparing the computer MPG vs doing by hand. They found the computer MPG was correct. ( don't yell at me..just reporting.)

 

remember the calculated MPG is just a computer estimating. Id venture to say that the milage traveled is the "known" value and its best guess on fuel it has used.

The computer should have an idea on how many "ounces" of fuel it squirts each time....but the question is how accurate is the measurement of each squirt at each particular RPM and load.

yes the program knows how much is it supposed to use.., the question..is the injector really that accurate....

So the computer is using a known value (milage) and its best guess on fuel being consumed. That's all it really is... the better the data the better the results....

yes I know that Ive gone 500 miles on a tank. the computer is telling me my MPG is 50. but when I fill up i put in 10.3. and calculate 48.4 mpg...

the dilemma is none of us have any way of knowing that when we filled it up if we have returned the fuel to the exact level it was at when we started the tank. not every pump will click off at the same fill level, heck I'd venture to say the the same pump will click off at different levels based on temperature, age of the pump handle, how fast the fuel is flowing, angle of the handle....ect.. All variables that we really have no control over.
one could argue that the computer may be more accurate at calculating the MPG than we are by doing it long hand.

Both the car and us have the same mileage number. the only other variable is the amount of fuel used. the computer is estimating how much fuel was used, and essentially...so are we...


and if you really think about it. even the mileage is a suspect number. too high or too low tire pressure, worn tires, different tire makers...again. all variables.

remember GIGO errors.....garbage in = garbage out

 

112,000 miles in our 2010 Prius. In Summer computer is about 2.5 mpg or so off optimistic than hand calculated.
During Winter months 3.5 mpg or so off optimistic than hand calculated....

My ex had a 2006 Civic Hybrid. That car's computer always under. Computer read 48 mpg, hand calculated would be 50 mpg or so..

 

FWIW, our old '11 Cruze computer was constantly 5-7% optimistic...but seemed to be become more accurate the longer & continuous the trips.

My "guess" is that city-driving tended to have TOO much time spent in 'open-loop' (ECM) operation, while long, continuous, freeway-driving had minimum 'open-loop' operation and thus better knowledge of actual fuel consumptions.

 

My error fluctuates a lot. I assumed it was because I fuel at different gas station and thus pumping errors.

 

Our previous '06 Civic Hybrid was either spot-on, or slightly low. If for example it said 5.0 liters per 100 km, calculated was either 5.0 or 4.9 (lower number being better).

I suspect accurate mpg display is not that common. I've even read that the new Accord hybrid is prone to rosy reports. OTOH, it's also being reported that Prius c and v (and maybe the plug-in?) are at least closer. Maybe due to customer feedback?