Mileage on full and empty tank

July 14, 2004 9:44 AM

I was on a road trip recently, and we fell to wondering whether the car uses gas faster when full or nearly empty. My friend said one thing, I had heard another, and harbor a suspicion that it's all urban myth anyway. Any ideas?
posted by kaibutsu to Travel & Transportation (21 answers total)

The car is significantly lighter when nearly empty and therefore uses less fuel per unit distance.
posted by Capn at 9:58 AM on July 14, 2004

Ah, but if you're wondering what your fuel needle does, that's a different kettle of fish totally.

A car's gastank isn't the same shape top to bottom, it's an odd shape wedged in to the available space, but the fuel gage does not know this. So if your tank is more like a pyramid, the needle will go down faster when then tank is full and slower near the end. If your tank is more like an inverted pyramid, it will go down faster near the end.
posted by Capn at 10:01 AM on July 14, 2004

Significant? I don't know about that. Most cars have 40-60L tanks. That's 30-45kg of gas, perhaps 70 to 100 lbs of fuel. For a typical car weighing 1 to 2 tons, that's only 2 to 5% of the vehicle curb weight. The calculators I know about don't even bother to consider fuel weight when calculating fuel efficientcy.

Btw, if I can give an unworthy sorta-self-link: one of our sister divisions produces calculators for these sorts of things based on real vehicle measurements. The emission calculator is the source for those "you emit (bignumer) kg of CO₂ on a typical commute" comments you see in Greenpeace adverts and such.
posted by bonehead at 10:21 AM on July 14, 2004

As the tank empties, you're more likely to be using "dirty" gas, filled with dust, sludge, whatever. My car runs significantly better with a full tank than a low tank.
posted by five fresh fish at 11:05 AM on July 14, 2004

All other things being equal, less fuel == less mass == less energy needed to motivate. There are possible external effects from such things as dirty fuel, as fff noted, but I would imagine that they would be even smaller than the already small factors we're dealing with.

So, let's say, as long as your car is fairly new and you keep fresh gas in it regularly, you'll get marginally better mileage as the tank goes down, but don't bother trying to measure it.
posted by jammer at 11:40 AM on July 14, 2004

Gas gauges are approximate, at best, even if you factor out tank shape. I've had a couple out and I was amazed at how simple they are.

Anyway, the fuel weight is pretty much the same as a young teenager, to put it in human terms. While the mileage would be theoretically less with that teenager in the car, you'd be hard pressed to be able to measure it well enough to say for sure it was a factor.
posted by tommasz at 12:36 PM on July 14, 2004

Hmmm....

So the gas is cleaner when full, but you get to throw a kid out the window as you go. So it's pretty indeterminate.

Is it true that there's some leeway 'under' the E to make sure you can get to a gas station?

(It's kinda academic as I don't drive; I just ride along on a lot of road trips...)
posted by kaibutsu at 1:26 PM on July 14, 2004

Is it true that there's some leeway 'under' the E to make sure you can get to a gas station?

Absolutely. I have a good 30 miles of highway driving after the warning light goes on in my car. Saved me more than once.

Of course, knowing this is like intentionally setting your watch 5 minutes fast in an attempt to fool yourself into getting everywhere a little early; I just push my luck against 9:05 instead of against 9:00, or against E+30 miles instead of against E. In order to get any benefit from that leeway or buffer zone, you need to not know about it. So forget I said anything.
posted by ChasFile at 1:39 PM on July 14, 2004

Is it true that there's some leeway 'under' the E to make sure you can get to a gas station?

Fuel tank capacity of my car: 14.5 gal. Number of miles that my trip computer says I still have until empty when the hand points to E and the fuel warning light comes on: 50.
posted by kindall at 2:08 PM on July 14, 2004

Is it true that there's some leeway 'under' the E to make sure you can get to a gas station?

I drive an old Toyota Corolla to work and back every day... on a particularly terrible week, I was on empty and facing a very long drive through the country. Since it was just me, I figured what the heck, let's see just how far this car will go before it sputters. I stopped to fill up a gas can for the finale, and took off down the road.
After the light came on, I traveled 87 miles before it finally died.
posted by bradth27 at 2:09 PM on July 14, 2004

Pretend this is at the end of my last post:

... most I've ever put in the tank at one fill-up: 13 gal.
posted by kindall at 2:10 PM on July 14, 2004

while i doubt that it's necessary to keep the tank full to get "clean" petrol, you really don't want to run it to zero. there is some gook in there - i ended up having to take the carburettor apart and clean out the needle valve because it became blocked with dirt after running out of fuel once.
posted by andrew cooke at 2:26 PM on July 14, 2004

Parents car is 2 years old, and has run clean out of gas at least 3 times. It's fuel injected, is this bad?

Also, I heard running your car very low on gas every so often is good for your gas tank, as the deposits at the bottom are rather minute early on. Getting a small amount into your engine is better than a lot at once. Is this true?
posted by Keyser Soze at 2:32 PM on July 14, 2004

Based on an article I read years ago in Automobile magazine: The needle or digital display for the gas tank does not move in a linear fashion. They are designed to remain at "full" for a while, suddenly drop to half a tank, and then meander down to empty. This is a feature, not a bug, as it is intended to give the impression of greater fuel economy for the vehicle.

In the article, a Cadillac with a digital gas readout in gallons was described. When the car initially shipped, the gauge went from "Full" to 17 gallons as the first gallon of gas was consumed. The company was deluged with complaints about the poor fuel economy of the car. Their fix was to change the gauge from moving from "Full" to 17 to "Full" to 16, thus displaying "Full" for twice as long. No other change was made to the vehicle. The complaints ended.

Caveat: I did read this a long time ago and am too lazy to dig up a citation. I believe this to be correct but, you know, if it's important verify it elsewhere.
posted by stet at 2:40 PM on July 14, 2004

Tangential, but another reason to avoid running out of gas is that the gas pump is cooled by the gas itself, and can overheat if not cooled.

That being said, I can get almost 80 after my light goes on. On a 15 year old accord.
posted by sauril at 3:45 PM on July 14, 2004

Is it true that there's some leeway 'under' the E to make sure you can get to a gas station?
This varies between car models and between individual cars to some extent. The only way to really know would be to run the car until it stops, taking note of how far you had travelled after the needle read E.

The only variable between full and empty would be weight, so there would no doubt be some difference in economy between the two, simply because of the weight difference, but it would be minor at best. Fuel gauges are approximate at best and are in no way a guide to fuel consumption.

Keyser Soze, running a fuel-injected car out of fuel is much worse, because the fuel injectors themselves contain minute holes (that cannot easily be seen by the naked eye) and they are sensitive. Just before the fuel runs out, it starts to contain lots of air bubbles and these can damage the injector nozzles. The tiniest particle of foreign matter can block the nozzles, too. Having said that, modern cars have filters in the fuel line, so the risk of getting foreign matter into your engine is very slight anyway.
posted by dg at 4:46 PM on July 14, 2004

Will the air bubbles end up "removing themselves" after a prolonged period of time?
posted by Keyser Soze at 5:58 PM on July 14, 2004

I had* a 1990 BMW 525 with a gas needle that consistently jumped down from 1/3 to 1/10 with almost no stopping in between. So, you're thinking to yourself, no worries 'cuz I have a slick Euro-car that gives a digital readout of projected range left in the tank. Too bad that at that gas level the range tends to drop 3 or 4 miles for every real mile travelled. It was not very comforting.

*Actually, I still have it, and am looking for a worthy cause to donate it to that will show their appreciation by coming to get it and by giving a generous assessment of said vehicle's value. Any suggestions?
posted by NortonDC at 7:47 PM on July 14, 2004

Driving up and down hills can also sometimes affect the fuel gauge. In my Subaru, driving down even a slight incline can push the petrol needle lower, and in some cases, switch on the Empty warning light...then when I level out, the fuel gauge rises and the light goes off. Conversely, driving up a hill can make it look like I have more fuel left than I really do.
posted by Jimbob at 8:11 PM on July 14, 2004

I've always liked my friend's notion of if the car has enough gas: When going up and down hills, if the gas needle fluctuates, you're good to go!
posted by jmd82 at 9:42 PM on July 14, 2004

The only way to really know would be to run the car until it stops, taking note of how far you had travelled after the needle read E.

Hey, I've done that!

Unfortunately, I hadn't planned in advance, and didn't note the mileage when it had first touched E.

I also didn't have a jerry can.

What a pain in the arse.
posted by five fresh fish at 11:58 PM on July 14, 2004