MPG v GPM

[Hodag'stiva]

Forget MPG - let's move to GPM

but the Duke profs think we should be moving to a gallons per mile model here in the U.S. That way, people might begin to see that trading a 14mpg SUV for a 21mpg hybrid version, for example, saves more fuel than trading in a 35mpg sedan for a 50mpg Prius. Right now, very few consumers realize this when they're out debating which car to buy.

I guess this is what some folks would call a "distinction without difference." Frankly, I'm not sure where saying that we're getting 2 gallons per 100 miles instead of 50 miles per gallon is going to make a difference in folks purchasing new cars.

C'est la vie, I guess. But I'm not going to start rethinking how I compute mileage (and the unit converter on my palm pilot won't do GPM anyways)

B

[theholycow]

I can see the point. The savings aren't linearly different. It's hard to tell and not immediately intuitive to most people that when you compare 14 to 21 vs 35 to 50, that those 7 mpg are worth a lot more money than the 15mpg.

In ounces per mile (I hate talking about unit per 100 units, gallons per 100 miles is totally lame to me), here's the difference:
14mpg vs 21: 9.14 oz/mi vs 6.1
35mpg vs 50: 3.66 oz/mi vs 2.56

So you can easily see that 7mpg difference is worth WAY more than the 15mpg difference...

Still, I don't think it's a good idea to change the standard. It's not broken enough to fix, considering the HUGE cost in money, confusion, obsolete equipment, reprogramming software and sites like this, etc...

[GasSavers_rGS]

Originally Posted by Hodag'stiva

I guess this is what some folks would call a "distinction without difference." Frankly, I'm not sure where saying that we're getting 2 gallons per 100 miles instead of 50 miles per gallon is going to make a difference in folks purchasing new cars.

C'est la vie, I guess. But I'm not going to start rethinking how I compute mileage (and the unit converter on my palm pilot won't do GPM anyways)

B

Mathematically, it's called the "inverse function." Basically, it's purpose is to amplify the small differences into bigger numbers. Bigger numbers are supposedly better than smaller numbers, e.g. 20 mpg < 30 mpg. Another factor is one of psychologically sensitive values. Cases in point...

MPG | GPM
----+----
0   | 0
5   | 0.2
10  | 0.1
15  | 0.067
20  | 0.05
25  | 0.04
30  | 0.033
35  | 0.029
40  | 0.025
.   | .

See, the marketing and advertising people know that people may not understand how a 0.2 gpm value is worse than a 0.025 gpm value due to the possible assumption that most consumers are mathematically challenged therefore aren't sensitive to the "big deal/difference" between those two figures. That's why some people try to show the "difference" and "significance" by using "....per 50 miles" or "per 100 miles" to show a mathematically detectible difference like as follows...

MPG | GPM    | GPM/50 miles | GPM/100 miles
----+--------+--------------+---------------
0   | 0      | 0            | 0
5   | 0.2    | 10           | 20
10  | 0.1    | 5            | 10
15  | 0.067  | 3.333        | 6
20  | 0.05   | 2.5          | 5.0
25  | 0.04   | 2.0          | 4.0
30  | 0.033  | 1.667        | 3.333
35  | 0.029  | 1.429        | 2.857
40  | 0.025  | 1.25         | 2.5
.   | .

Then, now "Salesman math" comes into play. Let's say you've got a car that gets 25 mpg and the salesman is trying to sell you a 40 mpg model. So they could say something like "If you buy this 40 mpg model, you will save up to 150% on your fuel costs!" Let's assume your current car is paid off. Soo, in order for you to "save up to 150% on your fuel costs" you have to increase your other costs such as car payments (since your current car is paid off), increased car insurance premiums (since it's a new car as well as financing requires full coverage which is relatively more expensive than liability only), and other hidden costs that the salesperson isn't telling you.

However, GPM is a more practical number to use. That's because you can use it to calculate estimated: 1) fuel used for an errand and 2) approximate cost to drive that errand.

Let's say you know for your car that you get approximately 0.04 gpm based on last fill up calculation. Let's say your friend, who lives 5 miles away, asks for you to drive them somewhere else 10 miles away. Then let's say it's 5 miles from your friend's destination back to your home. So that's a total of 20 miles. So to calculate the quantify of fuel used for that errand round trip would be...

0.04 gallons    20 miles
------------- x -------- = 0.8 gallons of fuel.
   1 mile            1

Since there are 128 ounces in a gallon, 0.8 gallons of fuel works out to be 102.4 oz or over 8 bottles/cans of soda. So, now that you know the quantity of fuel estimated, you can now calculate an estimated cost for the errand.

0.8 gallons   $4.50
----------- x -------- = $3.60
              1 gallon

So assuming current gas prices are $4.50/gallon, your estimated errand cost would be about $3.60. So now you've got a calculated figure that you could reasonably ask your friend to pay you for the gas to accomplish the task so YOU might be able to break even.

Now, mathematically, it would be more difficult if you use mpg instead of gpm because you'd have to remember to divide miles by mpg instead of multiply miles by gpm to calculate estimated fuel burn. So it's relatively easier math if you use gpm because here's the equations:

gallons per mile x miles = estimated fuel used (gallons)

estimated fuel used x fuel cost = estimated cost of errand

So in conclusion, MPG is a sales/advertising/marketing number, GPM is a practical and useful number.

[friz]

It shows the diminishing returns of improvements to MPG at the high end. It may take quite a while to see a return on a mod that yeilds .001 GPM improvement. This can be several MPG on the high end around 60 and it only represents 1 gallon saved every thousand miles. These numbers really have made me rethink some mods.

[R.I.D.E.]

6.3 cents per mile works for me.

regards
gary

[DracoFelis]

Interesting thread, that makes you think about the math involved in FE. FWIW I noticed that:

Fuel cost per mile = GPM * Cost of a gallon of fuel
or
Fuel cost per mile = (1 / MPG) * Cost of a gallon of fuel

And for you people outside the USA who buy gas per liter, and measure your distances in kilometers, you get essentially the same formula with the units just being converted to what you use. i.e.

Fuel cost per kilometer = LPK (Liters per kilometer) * Cost of a liter of fuel

For example (and no these aren't real numbers from my car), if you get 45 miles per gallon, and gas cost 4.20 a gallon, than your cost per mile (just for fuel) is:

Cost = GPM * cost_per_gallon
or
Cost = (1 / MPG) * cost_per_gallon
or
Cost = (1 / 45 ) * 4.20
or
Cost = 0.093333333333333333333333333333333
i.e.
In that (hypothetical) case, the cost of driving is a mile is around 9 1/3 cents per mile just for the fuel. And these numbers don't factor in any of the other costs, such as: tires, insurance, oil changes, car payments, maintenance, car FE mods, etc. This number is just the fuel costs of going a mile.

Personally I think that cost per mile (or cost per kilometer) is a number that is personal/meaningful enough that most people could wrap their mind around it easily. And that cost per mile (for fuel) is easily computed by just multiplying the current fuel price by your GPM (or your liters per kilometer) values.

Of course, to really get a "real" cost per mile (or cost per kilometer) you have to take your other car expenses and amortize them over the miles (or kilometers) you expect to benefit from them, and then add in those costs/mile (or costs/kilometer) to your fuel costs.

But once you do have a "real" cost/mile (or cost/kilometer) number, you can easily use that number to give decent estimates as to how much any given trip will cost you to make. All you really need to do in that case, is just multiply your cost/mile (or cost/kilometer) number by the distance of your trip, and you have a good estimate (it's only an estimate/approximation, as we all know that costs fluctuates all the time due to such factors as driving conditions) of the cost of making the trip. And once you have a decent estimated cost of making any given trip (you are thinking about), its much easier to make an (informed) decision as to if the cost of that trip is "worth it" to you personally. Because how can you make an informed decision about the trip being "worth it", if you don't have a clue what the real monitory cost of that trip is?.

NOTE: The same reasoning can be used to evaluate the cost/benefit of FE mods. If you can estimate how much savings/mile (or savings/kilometer) you expect from a FE mod, and you know about how many miles (or kilometers) you put on the car in a year, you can use those numbers to get a good estimate of how long it would take for the "break even point" where the mod pays for itself (before the break even point you are essentially behind, after the break even point the gains are all "profits" for as long as the mod and your car lasts). And again, this savings/mile (or savings/kilometer) numbers are pretty easy to calculate, when you look at (expected from the mod) differences in GPM (or LPK for you non-USA folks) and multiply those numbers by an estimated cost for fuel, and then divide that savings/year number into the cost to do the mod...

i.e. The "break even point" (in years) of a FE mod is:

break even years = cost of the mod / savings per year
or
break even years = cost of the mod / (savings per mile * miles per year)
or
break even years = cost of the mod / (savings per kilometer * kilometers per year)

[GasSavers_rGS]

Dracofelis,

Unfortunately, the actual cost of fuel depends if you're using either the cash accounting method or the accrual accounting method. Cash accounting method states that income is recognized when you received funds or funds equivalents and expense is recognized when you spend funds or funds equivalents. Accrual accounting states that income is recognized when earned and expenses are recognized when used up.

There are three ways to calculate costs under accrual accounting:

1. First In First Out (FIFO)
2. Last In First Out (LIFO)
3. Average

Let's use FIFO. So let's say you have a car with a 15 gallon capacity. Let's say it's completely empty. Let's say the current cost of gas is $4.00 / gallon. So you fill up the tank. So the value of gas in the tank is $4.00 x 15 gallons = $60.00. Then you drive around and then go to refuel your tank, which for the sake of this example, took 10 gallons to refill the tank. Also, the second fillup cost you $4.50/gallon. So, using FIFO, you've got 5 gallons of gas @ $4.00 / gallon and 10 gallons of gas @ $4.50/gallon which works out to be $20.00 + $45.00 = $65.00 worth of total gas in the tank. Then let's say you drive around some more and refill your tank, which magically is another 10 gallons to fill up. Well, that 10 gallons you filled up represents the 10 gallons burned. So to calculate the value of the 10 gallons you recently burned, you have to take the "leftover" 5 gallons of gas @ $4.00/gallon, then use 5 gallons of gas from when it was $4.50/gallon, which works out to be (5 gallons x $4.00/gal) + (5 gallons x $4.50/gal) = $20.00 + $22.50 = $42.50. Let's say that you travelled 200 miles between fillups. So that works out to be $42.50/200 miles = $0.215/mile.

Yes, you also need to take into consideration the other costs like oil changes, tire wear and tear, and if you wanna get really nitpicky, the daily cost for car insurance divided by the number of miles driven on that day.

Soo, in conclusion, it's easy to calculate GPM. It's a bit more mathematical work to calculate $PM.

[DracoFelis]

Originally Posted by rGS

Soo, in conclusion, it's easy to calculate GPM. It's a bit more mathematical work to calculate $PM.

Wow, that takes me back. I haven't thought about the details of "accrual accounting" much for years (I took a couple of accounting classes in college years ago).

And I agree with you from a strict accounting standpoint. You are correct that the cost of a gallon (or cost of a liter) of fuel can be complicated, as you have to factor in the fact that the fuel was purchased at different prices over time. And ditto for such matters as "depreciation" on the car, how the insurance is to be amortized over the miles, etc.

However, you were looking at this $PM figure from a strict accounting principles standpoint, whereas I was looking at it more as an "estimated" number for decision making purposes. And for decision making purposes, you don't necessarily need to follow strict accounting practices to get an "estimated" number that is "good enough" to be very useful.

And you can easily estimate what the cost of a gallon of fuel in our car is. If nothing else, we could just use the current gas price at the pump (which will admittedly overvalue gas previously bought at less than that price, and visa versa, but over time will tend to average out). And once you get a decent number for the cost of fuel per gallon (or cost of fuel per liter), you can go ahead and use that (admittedly "approximate") fuel costs, to calculate a usable (if not technically 100% spot on) number to help us make decisions about the cost of taking this or that action.

So, if you are saying that the calculations are more complex (although would still be easy to program into a FE computer, to automatically handle for you) if/when you need/want to follow strict accounting standards, I would have to agree. But that shouldn't stop us from estimating some of these numbers (such as the cost of fuel in our tank), and going ahead and arriving at an (admittedly estimated) $PM figure. Because unless our estimated input numbers are way off, that "estimated" $PM figure will likely still be close enough to give us some real useful insight into what is the financial cost to make this or that decision.

For example, a decent estimated $PM figure can still help you decide about (approximately) how much a planned trip will cost you, or if a proposed FE mod is "worth it" financially. For such real world questions, a good "approximation" is usually more than "good enough" to give the needed insights into the decision making process.

[GasSavers_rGS]

DracoFelis,

Originally Posted by DracoFelis

Wow, that takes me back. I haven't thought about the details of "accrual accounting" much for years (I took a couple of accounting classes in college years ago).

And I agree with you from a strict accounting standpoint. You are correct that the cost of a gallon (or cost of a liter) of fuel can be complicated, as you have to factor in the fact that the fuel was purchased at different prices over time. And ditto for such matters as "depreciation" on the car, how the insurance is to be amortized over the miles, etc.

However, you were looking at this $PM figure from a strict accounting principles standpoint, whereas I was looking at it more as an "estimated" number for decision making purposes. And for decision making purposes, you don't necessarily need to follow strict accounting practices to get an "estimated" number that is "good enough" to be very useful.

Back in my post #3, I addressed the issue of "estimates" by the block starting with "Let's say you know for your car that you get approximately 0.04 gpm based on last fill up calculation...So now you've got a calculated figure that you could reasonably ask your friend to pay you for the gas to accomplish the task so YOU might be able to break even." In other words, making a "should I go?" decision. The more detailed "accrual accounting" analysis is needed to calculate the ACTUAL [historical] accrual accounting cost of "how much did it actually cost to drive?" decision. The reason why in my example I used the figure "0.04 gpm" is because I calculate both the mpg and the gpm at every fillup. As the saying goes, your mileage may vary, so your gpm value will fluctuate as will your mpg value. So for "estimating" purposes, why not simply use the last gpm figure you achieved?

Originally Posted by DracoFelis

And you can easily estimate what the cost of a gallon of fuel in our car is. If nothing else, we could just use the current gas price at the pump (which will admittedly overvalue gas previously bought at less than that price, and visa versa, but over time will tend to average out). And once you get a decent number for the cost of fuel per gallon (or cost of fuel per liter), you can go ahead and use that (admittedly "approximate") fuel costs, to calculate a usable (if not technically 100% spot on) number to help us make decisions about the cost of taking this or that action.

So, if you are saying that the calculations are more complex (although would still be easy to program into a FE computer, to automatically handle for you) if/when you need/want to follow strict accounting standards, I would have to agree. But that shouldn't stop us from estimating some of these numbers (such as the cost of fuel in our tank), and going ahead and arriving at an (admittedly estimated) $PM figure. Because unless our estimated input numbers are way off, that "estimated" $PM figure will likely still be close enough to give us some real useful insight into what is the financial cost to make this or that decision.

For example, a decent estimated $PM figure can still help you decide about (approximately) how much a planned trip will cost you, or if a proposed FE mod is "worth it" financially. For such real world questions, a good "approximation" is usually more than "good enough" to give the needed insights into the decision making process.

I agree that you need to be careful in your calculations for decision making purposes to determine if something is "worth it." Just like anything else in life, you can plan things, then Murphy's Law gets involved to add the "reality factor" to the execution of the plan.

[thornburg]

Originally Posted by rGS

Let's say you know for your car that you get approximately 0.04 gpm based on last fill up calculation.

Originally Posted by rGS

So in conclusion, MPG is a sales/advertising/marketing number, GPM is a practical and useful number.

Your thinking still hasn't swung all the way around to GPM (or L/KM). You don't get .04 gpm, you use .04 gpm.

This is way we ought to be thinking about fuel usage. It isn't "I get 50 miles for each gallon of fuel", it's "I use .02 gallons of fuel for each mile I drive" (or, "I use one gallon of fuel for each 50 miles I drive").