Sticker on a bike rack reading “53 miles per burrito”

If I could somehow metabolize gasoline…

Published 2022-06-10

I did this math the last time gas crossed $4/gallon. If I could magically consume gasoline I would get about 718mpg.

Of course, I have to eat boring old food, so I more reasonably get something like 48mpg (of 2% milk) or 82mpg (of cooked white rice).

On the other hand I can burn much more exciting fuel than milk or rice. Like butter (652mpg), or beer (40mpg), for example. My car can only burn gasoline but it’s also fun to see how much this would all cost if it burned butter or rice.

Butter almost has the same caloric density as gasoline, by the way, but is significantly more expensive!

My actual cost per mile to travel by bike (at 15mph) is generally less than 5¢/mi. Compare to driving (at 60mph) at 23¢/mi. I do get there four times faster though, so we could discount that to about 6¢/mi. This, of course, sets aside the weird thing where riding a bike expands time but driving dilates it, which is like trying to comparison shop rainbows.

Our 2018 Subaru Ascent, on the other hand, gets about 24mpg on a steady highway cruise at 60mph. Obviously this is about four times faster than I can ride a road bike, without drafting, on flat terrain with no wind. But on the other hand it does considerably worse in the usual stop-and-go situations when it’s actually moving 15mph, and we pretty much never drive 60mph on an open freeway. So this whole exercise is heavily tilted toward the car.


If we assume the following caloric expenses…

mi/hr kcal/hr kcal/mi
Paul on a bike 15 600 40
Subaru 60 71,865 1198

…then we can obtain the following efficiencies.

Bike Subaru
1 gallon of… USD$/gal kcal/gal $/kcal mi/gal $/mi mi/gal $/mi
regular unleaded $5.49 28,746 0.0002 718.7 $0.01 24.0 $0.23
2% milk $3.69 1,952 0.0019 48.8 $0.08 1.6 $2.26
cooked rice $6.58 3,280 0.0019 82.0 $0.08 2.7 $2.40
butter $29.52 26,112 0.0011 652.8 $0.05 21.8 $1.35
beer $16.87 1,581 0.0107 39.5 $0.43 1.3 $12.78

I just went shopping (Trader Joe’s) and filled the car up (Space Age) so these prices are all very current for my neighborhood.

My bike riding efficiency estimate is highly contingent. For starters, I am not usually on my lightest road bike, the weather is seldom windless, and I like to climb hills. Drafting on another rider offers significant efficiencies as well. I based my estimate here (600kcal/mi) first on this ancient source which seems roughly accurate. I believe I am more efficient than this, however. My own power estimations — using heart rate data, therefore a second-order calculation — suggest that I burn less than the 704kcal on that chart. For example, on this solo ride to Mt. Angel in 2019, where I averaged 14.5mph, my HRM suggested I burned 3,658kcal in 5.5hrs or ~665kcal/hr, although most of this effort occurred on two climbs. A year later I rode the same route with Thom at almost the same speed and my HRM suggests I burned 3,118kcal, or ~567kcal/hr.


This would be a fun game to turn into a web app. Enter your weight, your bike speed, and a gallon of some random energy source (apple juice, honey, chocolate ice cream, mushed up bananas, oatmeal, burritos…) and see how far that gallon will take you.

The whole exercise is really different with electric cars. You could compare mechanical efficiency bikes vs. cars pretty easily (watts/pound to travel distance d)…but reckoning the cost of the energy source becomes a lot more abstract. Bike riders “burn” (and ICE’s literally burn) calories directly to produce watts, electric cars don’t burn anything, they just move electrons from one place to another. How many calories are in a battery, anyway? Is that even a rational question?

For our internally-combusting car, at almost every step I have put my finger on the scale for the car. I reckoned the engine to be more efficient than it actually is. I chose the least expensive gas for sale near our house. I picked pretty expensive food — although not the fancy organic stuff — bulk rice in particular would be a lot cheaper than the packaged basmati rice at Trader Joe’s priced above.

But “cars” is one of those topics about which we ceased long ago to be swayed by facts or numbers or reality. As a means for navigating a city (or suburb), bikes so badly outclass cars that clearly something else keeps Americans so invested in cars. I don’t imagine anyone will read this, who doesn’t already agree with me, and think “gee, it would be great if our lives were structured so we didn’t have to spend so much time/money/effort on cars.” Even in hikeybikey Portland only about 5% of most trips are made by bike; I sincerely doubt 90+% of Portlanders have health conditions that prevent them riding their bikes a few miles to work or school or wherever. The task of rebuilding our individual lives seems so unthinkable that we have conceptually walled off the possibility, such that mere facts cannot breach. Riding your bike instead of driving your car, for any given trip, is an eminently imaginable possibility, if we allow ourselves to imagine it.

Addendum 2022-06-13

I’ve had one of those “53 mpb” (miles per burrito) stickers for years. But in actuality I only get about 27mpb: