# Car average mpg display correctness

I have been wondering about the correctness of the average MPG displays on modern cars. Mine in particular is a Renault Modus. What I did is fill the tank, drive around and then refill the tank to see how much fuel I used.

Based on my calculations I made 557 miles (roughly) on 50 litres of diesel... which I agree is very good and comes down to roughly 50.6 mpg. (using the UK gallon here).

The display, at the time of refuelling, was showing 53.4 mpg average.

My question here is, how do these displays work and how much off can they be? Is it expected to be a good 3 mpg off over a 550 miles trip?

Thanks everyone!

-

I think that the computer's fuel consumption statistics don't actually come from the tank float. The consumption estimate comes from the injector duty cycle and fuel pressure.

The computer probably has a map or graph between duty cycle and fuel pressure which tells how much fuel the injector will spray at any given cycle and pressure. Using this it continually estimates how much fuel it has sprayed since the mpg calculator was last reset. It then takes this amount and divides it into the miles traveled since reset to give you the miles per gallon.

There are definitely going to be inaccuracies in the graph between the expected fuel spray and the actual spray that occurred so the MPG estimate will never be 100% accurate but I wouldn't expect it to be anymore than 5-8% off.

Also as previously mentioned when doing your own calculations you shouldn't use your entire tank capacity as the fuel consumption. It would be better to use the amount that you actually refilled. As long as you refill to the same point in the tank every time this should give you an accurate reading of mpg. Understand that you might not get the exact same fuel level each time though which means there will be slight error in this calculation as well but for all intents and purposes is accurate enough.

-
A couple of points to remember: 1. the inaccuracies in distance measurements will apply to all organics calculations (could try GPS for true distance over ground) 2. filling to the same point on the tank is nearly impossible to do accurately (you'd need to drain the tank and then add a measured gallon of fuel). – Bob Cross Jun 28 '13 at 16:55
Neither of those points add anything to the problem at hand. – Mike Saull Jun 28 '13 at 17:39
The inaccuracies in the distance will not factor into discrepancy between the calculations the odometer and computer ecu would both run off of the output shaft rev counter of the transmission. I mentioned point #2 in my answer and as I stated before it would be accurate enough for any regular purposes. Emptying the tank would be overkill. – Mike Saull Jun 28 '13 at 18:22
Your assumptions as to how the consumption is calculated is correct. – Juann Strauss Jun 29 '13 at 15:56

Tallmaris' means of determining fuel consumed per distance is correct, but could be made more precise by repeating the test several times until at least one thousand, and preferably two thousand, miles have been driven, keeping close track of the fuel consumed traveling those miles... starting with a full tank and ending with a full tank.

Mike, too, is correct in that the computer doesn't use the fuel tank float to calculate average mpg - it uses what it THINKS the current fuel consumption rate is, and constantly recalculates a running average, based on what it thinks the flow rate through the injectors is.

Based on the original test, the original comparison between calculated average and real average shows very nearly a 10% error. That's not actually very bad. A speedometer is assumed to be accurate only to within about 10%, and a fuel gauge that reads within 10% is a VERY GOOD fuel gauge.

-