“Be kind, for everyone you meet is fighting a hard battle” - Often attributed to Plato but likely from Ian McLaren (pseudonym of Reverend John Watson)

Sunday, October 26, 2008

The purpose of hypermiling

As mentioned repeatedly, I'm a frequenter of a web site devoted to maximizing fuel efficiency through all available techniques. These include the operational techniques I've implemented in my driving as well as minor and major modifications to vehicles. It's a wonderful site, occupied by people with a variety of philosophies.

Mine is to minimize both my cost per mile, and my overall fuel expenditures (given the fuel hog that I drive). But there are others whose goal is to maximize the miles per gallon irrespective of other considerations. Doesn't their goal assure my goal? It doesn't. Many of these hypermilers will choose a longer route if they can achieve higher miles per gallon, even if that route entails sufficient extra mileage to cause an overall increase in fuel consumed. In other words, these hypermilers treat maximizing the miles per gallon realized as something of a sport.

Is there anything wrong with this? Of course not. As the saying goes, "ya pays your money and ya takes your choice." Certainly, these men and women (mostly men) are not using huge amounts of gasoline to make these choices. I suspect that most, if not all, of them use less fuel than I do over the course of a year. And their efforts are communicated to the group, thus giving those of us who seek to minimize total costs additional data.

So what, in my efforts, controls the overall expenditures on gasoline? Two things are key: miles driven and gasoline price per gallon. Note that miles per gallon achieved are conspicuously absent. It's much easier to save on gasoline costs by driving less and by purchasing cheaper gasoline than by utilizing economy maximizing driving techniques.

Lest people conclude that driving technique matters little, I need to clarify. After purchasing my Land Rover LR3 HSE, I attempted to use the techniques that were effective in my Jeep Grand Cherokee Limited. I found that it was difficult to exceed the E.P.A. estimates and that I was hard pressed to make much difference. This led me to drive the LR3 "normally," that is, as most would drive it. As gasoline ran through $3.00, then $4.00 per gallon I redoubled my efforts. It did make a difference, and if one considers the graph of Cost per Mile as a function of Gasoline Price, it literally separates into two distinct data sets. And the average mileages during each of these phases stand at 16.3 and 20.9 respectively.

And actually, that underestimates what can be done, since the "before" data includes my earliest efforts at trying to save fuel in the LR3 and thus is higher than "normal," and the "after" data is significantly higher in the later fill ups, as I refine technique.

But for the "after" data plotted alone with Cost per Mile as a Function of Cost per Gallon, the so-called "coefficient of determination" is greater than 0.81. In other words, more than 80% of my cost per mile is determined by what I pay for fuel, my nibbling around the edges with driving technique accounts for some of the remainder, and the nature of the driving during the tank full (stuck in traffic, city driving, pure freeway driving, etc.), and other random factors account for the rest.

Thus, regardless of what else I do, I'll leave more money in my pocket if I drive fewer miles and buy cheaper fuel. It's a good thing I have a strong mathematics background, it serves me well in deep analyses such as this.


Anonymous said...

I don't get the usefulness of the graph on Oct 18.

Drag increases exponentially, as should its percentage. This graph shows it tapering off.

Do you get the same speed crossing if you graph Vdrag and Vroll with their actual values?

King of the Road said...

The graphs actually represent the fraction of total drag each component represents at a given speed. Unfortunately, colors aren't well rendered in that graph, but at 0 m.p.h. (or a tiny amount above) rolling drag is all of the external resistance, aero drag has no effect yet. As speed increases, the portion of total resistance represented by rolling resistance decreases and that of aero drag increases. For my vehicle, they split the load 50-50 at about 22.5 meters per second (where the curves intersect).

It's true that total resistance goes up with speed, and proportionally to the square of speed, since aero drag dominates at high speed.

The graph doesn't show total drag as a function of speed, only the fraction of the total represented by each component.

Thanks for asking!