Hybrid Mileage Challenge: Part 5


In’s latest Mileage Challenge, we relied mostly on each car’s onboard mileage readout, a feature integrated into the trip computer on many cars these days. Ultimately we chose to use the readouts over our own at-the-pump calculations. In our final installment, we’ve decided to look at just how those readouts work — and more specifically, how accurate they are compared to calculations made at the pump.

Experts were split: Some defended the onboard displays as significantly more accurate than at-the-pump calculations. Another said that the differences tend to even out over time, while another warned that mileage readouts have their own set of potential inaccuracies.

So what makes these things tick?

Generally speaking, they’re able to gauge instantaneous and cumulative mileage by factoring vehicle speed against how much gas your fuel injectors squirt into the engine. Most run on standardized calibrations, which produce “pretty repeatable and accurate” results, GM’s Terry Connolly said. Connolly is director of the automaker’s North American Energy Center.

The systems aren’t always airtight. Ford technology spokeswoman Jennifer LaForce noted that changes in fuel density, which can vary depending on where you fill up, can affect accuracy, as can tire wear or wheel size. As vehicles age, fuel injectors can get dirty, but LaForce and Connolly said even if that happens, internal sensors will keep accurate tabs on the amount of fuel passing through. It’s all part of a “closed-loop” system, Connolly said.

How do these systems compare to more traditional mileage calculations — specifically, pumping a dozen-plus gallons and doing some quick division before you reset the odometer? Honda spokesman Jon Fitzsimmons said such methods can yield varying results because of fuel pump inaccuracies. Connolly echoed the sentiment: “Pumps have widely varying flow rates, which affects nozzle shutoff substantially,” he said. “The difference in fill levels by a half-gallon is not unusual at all.”

At-the-pump calculations can be especially dicey if you’ve only filled up with a few gallons, University of West Virginia’s James Smith said. Smith teaches mechanical and aerospace engineering at the school, and he’s also the Society of Automotive Engineers’ 2009 president-elect. At the end of a fill-up, gas pumps click off because an internal sensor tells them there’s enough backpressure in your tank that it’s full, Smith said. As such, their accuracy can vary: “They weren’t made to be accurate, they were made to be safe,” he said.

That doesn’t mean you should top off your tank at every stop, but it does suggest a potential for variance. “When you’re looking at the pump, say it says 14 gallons, you could be off a half of a 10th either way,” Smith said. But if you’re only putting in a few gallons, that difference could amount to 5% or 10%.

A litany of other variables, from the design of the gas tank and filler tube to the outside temperature and density of the fuel, can also interfere with fill-up accuracy. You can even affect the point at which the pump clicks off by how you insert the nozzle into your tank, Smith said.

All told, this gives plenty of reason to trust in-car mpg readouts over calculations at the pump in the short run, but it doesn’t mean at-the-pump calculations are worthless: Over several weeks and multiple tanks of gas, the variations should even out, and you should get pretty comparable numbers, Smith said. He regularly calculates mileage in his Ford Explorer, both at the pump and through the trip computer’s mileage readout.

Ford’s LaForce agreed.

“Generally, the value should be pretty close, but there are variables,” she said. “The best method is to check gallons consumed and distance traveled over two to three fill-ups to estimate your mpg, and to crosscheck against the [onboard mileage] meter.”



This is good, interesting and useful information. And the reason I like this website. The blog about how comfortable the LX570 is for an office? Not so much........


Based on 3 cars with trip computers and years of calculating mileage at the pump, I am convinced some cars' computers are accurate, some are not. My 2000 BMW 3 series was within 1-2% either way, so yes, it evened out. However, my 2005 Saab 9-3 and 2008 Nissan Sentra's mpg trip computer consistently overstate mileage by 5-7%. The pump calculations are certainly more accurate if the discrepancy is only going one way. And think about it, wouldn't manufacturers want people to think they are getting 5-7% better fuel economy than they actually are?


Some new cars use flexible "bladder" style gas tanks, which can vary in size depending on temperature and other conditions. A "fill-up" will vary by amount from time to time, so use the car's onboard mileage readout to get consistent readings on fuel use.

Shin K.
page 5 Table 1 Phase II SAE J1634 fuel efficiency compared to the onboard computer reported fuel efficiency.

Onboard computer fuel efficiency percentage above the SAE
J1634 fuel efficiency results

Civic 1 AC off +21.7%
Civic 1 AC on +21.0%
Insight 1 AC off +11.0%
Insight 1 AC on +11.7%
Prius 1 AC off +15.7%
Prius 1 AC on +14.7%

I trust the experts from SAE committee and national lab more than the experts from auto manufacturer sales department.


I thought the article was very weak. I was expecting you guys to do accurate measures (put in exact amounts of fuel and drive till empty) and show if the computer was close or not. My guess was that the computer readings are usually not that accurate. And sure enough the post by Shin K. shows this to be the case (great post by Shin - more useful that your article), especially around high mileage vehicles which are really trying to make the driver feel like they are getting great mileage. Notice that none of the computer mileages were lower than actual.

This is an interesting article. I've always wondered how readouts work and how accurate they are.

Post a Comment 

Please remember a few rules before posting comments:

  • If you don't want people to see your email address, simply type in the URL of your favorite website or leave the field empty.
  • Do not mention specific car dealers by name. Feel free to mention your city, state and brand.
  • Try to be civil to your fellow blog readers. This blog is not a fan or enthusiast forum, it is meant to help people during the car-buying process and during the time between purchases, so shoppers can keep a pulse on the market.
  • Stay on topic. We want to hear your opinions and thoughts, but please only comment about the specified topic in the blog post.
view posting rules

If you have a TypeKey or TypePad account, please Sign In

Search Results

KickingTires Search Results for

Search Kicking Tires

KickingTires iPhone App