pump errors on small fill-ups
 

We've all seen it here before. People fill up 1 gallon, or even 2 gallons and they suddenly have 10 more MPG then they've ever had before. Some peope have filled up less than a gallon and have experienced great fuel economy...

Over time the law of averages evens out people's gaslogs, so it isn't much to worry about.

I would like to discuss however how and why this happens. How many gallons must you fill up in order to eliminate this bizarre occurance?

At the suggestion of ThisIsntJared I'm going to make a regression plot to see if there is in fact a correlation between number of gallons filled and gas mileage. I suspect that there will not be one, but it's worth a shot.

Filling up less than a gallon is inaccurate. I've tested this diligently.

I think the minimum fill up should be 150 miles.

Yeah I got 50mpg on a 2 gallon fill up. I don't have a scangauge, but i belive it was true or near true. It was straight highway with some drafting. and I had a warm engine to start out.

I have had a few other small fills with accurate (well, not out of the norm)results.
36/39.5/36.7/39.5/33.7

gas pump calibration checks are done with a 5 gallon tank so the stations ripping people off have maps on the pumps that pump 5 gallons accurately but more and less gallons short according to a 20/20 show years ago. That's another reason I try to pump 10 gallons.

The only gaslog entries I post are complete fillups. The smallest fillup I have posted is slightly more than 7 US Gallons. Most of mine are 10 to 11 gallons.

It would be interesting to see where the cut off is. I had several 3-4 gallon fills when I was testing the affects of weight and vapors on FE with minimun fuel and the MPG's were stable compared to those with 10-12 gallon fills. I think that the biggest problem is there is a lot of short fills after a highway or test run which will give a big increase in the FE over normal anticipated FE.

What I do.
 

I don't usually fuel-up unless theres at least a half-tank gone (6-8 gallons).

With respect to experiments, the last tank was killed by experiments as it was done mid-tank to preserve the "1/2-tank" rule. All in the name of Science! (I guess :o )

RH77

Full tank fill ups for me and every mile as well, no skips.

I put in $20 at a time when my low fuel light comes on. $20 gives me about 8-9 gallons of gas. i use that fuel number and the miles i've driven till the next time the low fuel light comes on. i don't do alot of driving and if i did fillup would be a month before i'd need gas again! :)

no data from me then... might as well not even be here


guess i'll just lurk here now....

I think it all averages over the long haul. I think the important thing is recording every fill with out skips and not just the ones with more than a set amount of gallons. The goal is to keep increasing and listing all tanks will give you a better trend. When looking at it you can always throw out the abnormal fills.

LxMike. Everyone is welcome gas log or not.:)

Agreed. Tiny, inaccurate fills don't matter in the long run.

They're not much good for bragging rights, though, since you have to take them with such a large grain of salt.

Keep on keeping on, people.

I have a big discrepency between a couple of fills because I didn't totally fill once so it looks like my mileage spiked but the average of the next fillup to the top and it brought it right in line with prior and subsequent tanks so put them all in and the graph shows the reality of them all.

i think only the 10 cars with the most gas logs should be used, each with different colors and plotted with gallons at the x coordinates and mpg at the y coordinates.

i predict we will see some interesting trends. obviously the further to the right you go the more gallons and thus the standard deviation of mpg will decrease making the readings narrow for each car. the standard deviation will probably increase exponentially as the fillup decreases.

however, i also anticipate, with the increase in standard deviation, an increase in the average mpg simply due to evaporation within the tank and probably a few other explanations to which i am not sure.

i expect the changes in these statistics to be fairly minor. also there is no need to do any data gathering. its all sitting in the usergarages ready to be plotted.

when i originally mentioned this to matt i was reffering to the difference in mpg between filling at a half tank, to 5/8, 3/4, 7/8, and limping to the pump with a tablespoon of gas.

I understand your point there. i just overreacted a tad. is 8 gallons a dinky fillup? that what i usualy put in. it's nice only driving 10 miles a day to work! :D

my tank capacity is 12.8 gallons. when i had my mustang i'd run it down as low as i could stand and could only get 10 gals into a 14 gal tank.

I know I can get an 1.9 gallons extra in my Jetta by fooling gas tank vent. I always vent my tank when I do mpg calculation.

could be!