Actual MPG (KmPL)

[freeztar]

A few months ago I started an experiment. Every time I filled my 2004 Subaru Outback with 87 gasoline, I recorded the gallons at the tank and reset my odometer. The next time I refueled, I repeated the process. My theory is that EPA-rated mpg for cars is not indicative of real-world driving and as a consequence is much higher than actual mpg. I put that theory to the test and here are my results:
(note: I wrote the above paragraph before doing these calculations to prevent a perceived bias in my theory)

Gallons of 87 gasoline     Miles travelled        Mpg
     
13.937                          348.3                  25.0
13.902                          279.4                  20.1
14.542                          338.9                  23.3
10.718                          292.5                  27.3
14.650                          372.7                  25.4
13.908                          296.8                  21.3
13.543                          280.1                  20.7

Average MPG = 23.3 (9.9 KmPL)

According to this site, my Subaru Outback (second one down on that list) is rated by the US EPA to have mpg's of 19 (city) and 25 (highway) and 21 (combined). Since my driving consists of both highway and city miles I took the average of city and highway miles just as another benchmark, 23. Also notice that one user has posted results for this car as 22.7.

So from my experiment it is clear to see that my theory was incorrect. My average of 23.3 mpg is much higher than the city and combined rated mpg's. Furthermore, my average is only 1.7 miles below the EPA listing of 25 mpg highway. Also note that the EPA average for city and highway miles is only 0.3 from my results.

I attribute these differences to the fact that my sampling period was during the warmer months of the year, which gives a slightly better fuel economy I believe. My 5-speed habits may come into play as well.

All in all, a fun and informative experiment!

Anyone care to share their results? Or even just comment on mine?

[Zythryn]

The mileage you get can vary widely on how your drive. For example, after you fill up, accelerate hard every time you start from a stop or need to get up to speed (as long as you do so safely of course). After measuring your mileage for that tank of gas, accelerate as slow as you safely can for the next tank of gas. There will be a significant difference.
Or, underinflate your tires by 5psi, then inflate them fully and there will be a distinct improvement.
This year the EPA changed the way they measure mpg to more closely reflect real life conditions. Previously almost no one got the mpg listed by the EPA.
We typically beat the EPA listing by a couple of mpg. I have heard of people termed 'hypermilers' who tend to get quite a bit over the listed mpg, however some of what they do to get there is not recommended by safety eperts

[Cedars]

Almost two years ago I had to replace a part on my truck (4 cyl S-10). My mechanic couldnt find an exact and ended up putting a heavy duty part on and told me my mileage would drop some. I lost 3mpg, dropping it down to 22 instead of 25.

So this year I decided to reduce my speed to 50-55 hwy, 60-65 freeway (which I dont travel much now that I dont work in the twin cities). By reducing to 50-55 I have gained back what I lost to 25 mpg highway.

[LaurieAG]

Hi Freeztar,

Originally Posted by freeztar

Anyone care to share their results? Or even just comment on mine?

My rough figures are here Ethanol/ULP Mixes and Mileage Equivalence

I was trying to work out how much mileage I was losing as a result of the amount of ethanol added to the ULP (Un Leaded Petrol).

It seems that the amount of ethanol in Australian ULP varys between 5 and 30-40 percent of the mix depending on the supplier, even though our law says 10% max.

p.s. I get 13.5 Km/L from my 1992 Daihatsu Charade (993cc).

[freeztar]

Originally Posted by Zythryn

The mileage you get can vary widely on how your drive. For example, after you fill up, accelerate hard every time you start from a stop or need to get up to speed (as long as you do so safely of course). After measuring your mileage for that tank of gas, accelerate as slow as you safely can for the next tank of gas. There will be a significant difference.
Or, underinflate your tires by 5psi, then inflate them fully and there will be a distinct improvement.

I remember CraigD mentioning in some thread that he used to over-inflate his tires which, coupled with driving at an optimal highway speed, saved him MPG.
I keep my tires inflated to the maximum PSI and keep it at that. Sometimes I accelerate quickly, other times not.

This year the EPA changed the way they measure mpg to more closely reflect real life conditions. Previously almost no one got the mpg listed by the EPA.

Interesting. When I looked up the MPG for this thread, I seemed to remember the numbers being much higher when I purchased my car. So I'm not going bonkers!

We typically beat the EPA listing by a couple of mpg. I have heard of people termed 'hypermilers' who tend to get quite a bit over the listed mpg, however some of what they do to get there is not recommended by safety eperts

I would like to increase my fuel economy, but not at the sacrifice of safety.

[freeztar]

Originally Posted by Cedars

Almost two years ago I had to replace a part on my truck (4 cyl S-10). My mechanic couldnt find an exact and ended up putting a heavy duty part on and told me my mileage would drop some. I lost 3mpg, dropping it down to 22 instead of 25.

So this year I decided to reduce my speed to 50-55 hwy, 60-65 freeway (which I dont travel much now that I dont work in the twin cities). By reducing to 50-55 I have gained back what I lost to 25 mpg highway.

Good deal. That sounds like a good trade-off all around.
I wonder where the diminishing returns occur when reducing velocity?

[freeztar]

Originally Posted by LaurieAG

Ethanol/ULP Mixes and Mileage Equivalence

I remember the thread and realizing that you beat me to the punch. It would be nice to do some ethanol experimentation, but I don't know where to get any around where I live.

It seems that the amount of ethanol in Australian ULP varys between 5 and 30-40 percent of the mix depending on the supplier, even though our law says 10% max.

That must be frustrating.

p.s. I get 13.5 Km/L from my 1992 Daihatsu Charade (993cc).

Which equates to 31.7 MPG! Not bad for a 15 year old car.

[LaurieAG]

Hi Freeztar,

Originally Posted by freeztar

It would be nice to do some ethanol experimentation, but I don't know where to get any around where I live.
Which equates to 31.7 MPG! Not bad for a 15 year old car.

You are doing the right thing already, the only extra thing you need to do is make separate records for the different fuel suppliers (or just add another column).

If you go through my posts you will see that CraigD provided a figure for Ethanol mileage equivalence of around 67% average. This means that for every 3 miles/Kms you get from ULP, you will get only 2 miles/Kms from Ethanol. By working backwards from any mileage difference you can work out how much Ethanol is in the mix. You should hopefully be able to get 100% ULP with no Ethanol to use as the base for your calculations if your local suppliers are telling the truth.

1. Record Mileage from a full tank of 100% ULP (0 % Ethanol)
2. Record Mileage/Supplier from a full tank with unknown Ethanol %
3. The Ethanol Mileage Equivalent Loss (EMEL) = 1.-2.
(i.e. 1/3 of ULP mileage is the loss)
4. The total miles under Ethanol in mix 2. = EMEL x 3
(i.e. 3 times the loss, EMEL x 1/(1-0.67))
5. The Ethanol % in mix 2. = 4./1. x 100 percent

If you get a negative figure in 3. your base probably isn't 100% ULP.

Like me you will probably get different figures from different companies, and like me you will only go to the companies that don't rip you off.

p.s. All of my calcs were based on 92 octane fuel, it's good that you keep that consistent too.

[Cedars]

Originally Posted by Zythryn

The mileage you get can vary widely on how your drive. For example, after you fill up, accelerate hard every time you start from a stop or need to get up to speed (as long as you do so safely of course). After measuring your mileage for that tank of gas, accelerate as slow as you safely can for the next tank of gas. There will be a significant difference.
Or, underinflate your tires by 5psi, then inflate them fully and there will be a distinct improvement.
This year the EPA changed the way they measure mpg to more closely reflect real life conditions. Previously almost no one got the mpg listed by the EPA.
We typically beat the EPA listing by a couple of mpg. I have heard of people termed 'hypermilers' who tend to get quite a bit over the listed mpg, however some of what they do to get there is not recommended by safety eperts

Those are good examples of variables that impact mpg. I would add in environment also. Driving with A/C on lowers mpg as does driving with the windows down at highway/freeway speeds (messes with the aerodynamics).

One trip to Duluth the variation in MPG was impacted greatly because the trip up we were fighting a very strong north wind. In my mind I have loss of 5 mpg on the trip north and normal mileage on the trip south (it was a LONG time ago). But the winds had died down alot on the trip south. We had headed north to watch a storm blow in.

Originally Posted by Freeztar

Gallons of 87 gasoline     Miles travelled        Mpg
     
13.937                          348.3                  25.0
13.902                          279.4                  20.1
14.542                          338.9                  23.3
10.718                          292.5                  27.3
14.650                          372.7                  25.4
13.908                          296.8                  21.3
13.543                          280.1                  20.7

Average MPG = 23.3 (9.9 KmPL)

It would be interesting to see if some of the variation in your mileage was impacted by weather conditions.

[freeztar]

Originally Posted by Cedars

It would be interesting to see if some of the variation in your mileage was impacted by weather conditions.

I'm sure it is, but there are so many variables that it would be hard to determine the extent. Rainy days always equal more traffic. You waste more gas idling, but the trade off is that you generally drive slower.

I suppose a lot more could be discerned if I kept a mileage journal day-by-day. My normal work commute could be the best test factor. Unfortunately, I would have to fill up every day to see the trends. I still might do it...