MPG -- and Insight into the Female Mind

True or false: you get better MPG when your car's tank is more than half full.

That was Jason's question on this week's show (You can here it right here.). And there's more riding on it than just enlightening our caller in Casper, Wyoming -- because Jason's wife thinks the answer is a resounding, "yes!"-- and she's the brainy one in the family, too.

Tom and Ray's take? The MPG is all the same, whether your tank is topped off or running on fumes. So, why does it always seem that the fuel gauge slides toward "E" more quickly on the second half of the tank? Because of the way the fuel is measured in the tank, that's why.

What do you think? Do you notice his funky fuel phenomenon? Do you think it's a conspiracy on the part of manufacturers, to have us believe that we're all getting fabulous gas mileage? Let us know what you think!

Yeah, no. The reason for Jason’s wife’s observation is simple and obvious: gauge system error. One can fill a tank to the Full mark on the fuel level gauge, and then continue to put fuel in the tank. This is because there is error in both the sender unit and the display unit. It is also why when the gauge hits the empty mark, the car doesn’t immediately shut off for lack of fuel. Also, depending on how much fuel one puts into the tank, one can cause fuel to back up into the filler tube.

Come on you guys, the lady’s just plain wrong. If my car has a full tank, it has to lug around the weight of all that gas, so it gets WORSE gas mileage when it’s full. When the tank is near empty, there’s a lot less weight in the car, so it gets better gas mileage when it’s near empty. You’d have to have pretty fine measuring tools, but we’re not talking about “perceptions” or gas needles and gauges, we’re talking about Jason’s honor, and his wife is just plain wrong.


Gas weighs about 6 pounds a gallon. If the tank holds 14 gallons, full it weighs about 84 pounds more than empty. Simple physics tells you that it takes more power from your car’s engine to accelerate a heavier car, not to mention more brake pad to stop it, and more tire loss for any directional change.

If this were true - F1 cars (which now start the race with a full tank for the entire race) would continue to do about the same lap times the entire race. This of course is not true, as they start out slower and then make the fastest lap times toward the end of the race (tire wear notwithstanding).

I’ve observed the same thing in my 2001 Saturn. When I fill up, I set the trip odometer and have tracked my gas mileage for months. Consistently, from full to 3/4, I can go 150 miles, from 3/4 to 1/4 I can go another 200. That last 1/4 tank has about 50 miles in it that I use as a buffer to payday.

Jason’s wife might also want to notice that when she shuts off the car that once the gauge is under half, that when she restarts the car that the gauge doesn’t come back up as far as it was when she shut it off. That’s due to a kind of “mechanical delay” between the float and gauge, and, best as I can figure, gravity that keeps the needle from coming back up once under half. LOL

Okay. First, understand the geometry of Fuel senders.
They are a float on the end of an arm, attached to a wiper that sends a resistance value to the Fuel gauge. They pivot from basically horizontal at full, down to about 30 degrees at empty. (By the way, you can apply this week’s Puzzler concept about the cylindrical Diesel tank to this same situation!)
If manufacturers bothered to make the wiper resistance match the geometry of the tank, we could get exact measurements, but they don’t…it’s a linear measurement (i.e. 5 degrees movement at the top of the tank is the same resistance-change as 5 degrees at the bottom.)
So, what Tom & Ray said was basically correct. The upper section of the fuel tank lowers the float slower per inch of fuel level drop than the bottom section.
Now…addressing the “Fuel Weight” idea that someone posted… Fuel weighs about 6.5 pounds per gallon (If I recall my Air Force technician days correctly), So a 15 gallon tank weighs about 100 lbs. That is so negligible an amount (since a driver is 150-200…okay, Ray is 260) as to make no difference in fuel economy.
End result…does the gauge move faster? Yes. Does the MPG change? Not really.

Three assumptions are being made here.
(1) that the sender and gage are both linear
(2) that the sender and gage are both full scale, top to bottom.
(3) that 1/2 of the volume of the gas equals the 1/2 point on the gae

None of these three assumptions are true.
(1) The gas level, and thus the float, drop at different rates at different places throughout the range due to the irregular shape of gas tanks.
(2) The float is not full range. It never hits the actual tank bottom, and it never hits the top of the tank. It runs out of range before doing either.
(3) Typically the range of the float is set to where it’ll read empty before actually being so, in order to protect the pump.

In reality, the added weight of gas reduces gas mileage, because it takes more energy to move 3,065 pounds than it does to move 3,000 pounds. But you’d generally be hard-pressed to measure the difference.

I don’t understand why Tom and Ray have not articulated ( is that too big a word?) the reason that fuel gages fall faster when less than half full. They have certainly been under enough cars to have noticed that fuel tanks (usually behind the rear axle) have bottoms that are sloped upward at the back (to provide ground clearance). A vehicle is designed fore and aft with increased ground clearance from the axles to the bumpers. This deacreases the volume of the fuel tank in it’s lower half. Manufacturers provide only a basic linear sensor on their fuel float sensors, resulting in a more rapid decline in the fuel level reading for the last half of the fuel tank. Perhaps this will be a future improvement that auto manufacturers will provide to help distinguish (and sell) their product.

Erich Dalton

The reason the guage seems to stay on full longer is simple. When you fill the tank it arrives at the full sensor before it is full to the top. The tank continues to fill beyond the sensor. It will not move off of full until the gas above the sensor is used. When it empties past that it will begin to register down ward.
Picture any thing that works with a float. Say the ball on a stick in your toilet. When the water level pushes the float to the level of the top the water stops filling. If water continued to fill the tank the float would submerge. It would not function as intended - to add water in this case until the water level went below the full setting.
In the gas guage parallel there is gasoline above the full setting.

My response would be: Why does it matter?
Why spend so much time and effort to figuring out how much one should put into the tank to get the best fuel economy when it’s just easier to fill the tank and forget about it.

Mostly I think that it doesn’t matter. Its actually a perfectly fine practice to keep the gas tank above half full.

But if Jason & his wife really want to settle it they should just do it with empirical evidence. Choose a number - maybe 3000 miles. Spend 3000 miles calculating actual gas mileage (miles/gallons) while running the tank down below 1/4 tank on the gauge. Then spend 3000 miles (or whatever) doing the same never letting it go below a 1/2 tank. Compare.

Regardless of the fuel tank shape or the potentiometer linearity, they could calibrate the fuel gauge meter scale to accurately show the MEASURED fuel tank contents. In other words, just put a 1/4 tank mark where the needle settles after adding 1/4 of the tank capacity of fuel to an empty tank, then add 1/2 and 3/4 marks, etc.

Digital fuel gauges could be calibrated in software to match the actual tank geometry by measuring it.

This is how Thomas Edison did things. He once asked a new hire to determine the volume of a lightbulb glass shell. After the engineer spent signficant time measuring the bulb dimensions and performing exhaustive calculations, he showed his result to Edison, who then proceeded to fill the bulb with water, which he poured into a graduated cylinder, and told the new engineer that his measurements were close, but his choice of measuring tools needs improvement.

Use Edison’s approach to calibrate the fuel guage!

Of course, if a man values his love life, he must choose between being right or being lonely.

I’d sure hate for her to drive…and logic…my wife’s 79 Chevy stepside.

Those fuel guage floats are so far off that you can drive 80 miles before it even begins to drop from full.
Since it had to be one or the other I had them both adjusted ( bent ) to accurately guage empty rather than full.
But MY wife knows this and does not attempt to twist the logic that MPG is better when full.

Excellent point. I’ll bet that even Thomas Edison would agree!

Several of those responding so far have it partly figured out. As the former Automotive Sales Manager for US Gauge Division of Ametek, which makes the dashboard instruments for several lines of cars and trucks produced in the USA, I had this question arise on many occasions. The sensor in the tank is indeed a float which reaches its top position before the tank is completely filled. When you continue to put fuel in the tank, that additional fuel must be consumed before the float can go down, so the gauge reads “full” during that period. So you truly do use more fuel for the gauge to go from “full” to “half” than you do for the second half of the tank. The other reason this phenomenon occurs is that the gauge and sender are calibrated to show “empty” when in fact some amount of fuel remains. This is done intentionally so the driver will not run out fuel while the gauge still indicates a partial tank remains. Can you imagine how many angry drivers there would be if that happened? Thus, when the gauge reads “empty” there is still some fuel in the tank, further contributing to the “uneven” movement of the needle between the first and second halves of the tank.

Tony Siegel, Fort Collins, CO

It takes fuel to carry fuel.

I was in my car and noticed that my gauge showed that my tank was exactly half-full. As a test I pulled into my favorite gas station and asked the poor attendant to fill my tank while the temperature was 3 degrees and the wind was howling. My 11.9 gallon tank (2000 Toyota Echo) took only 5.54 gallons. That means I still had 6.36 gallons left. WHAT’S WRONG WITH MY TANK?

Nothing, in fact your fuel gauge is better than most. I have a 93 Caprice that will go 100 miles before the needle drops to the full mark. Then another 100 to go to 3/4 and a 100 more to half. The next 100 miles the needle drops from half to empty. It’s not something I lose any sleep over.

Ed B.