The great e-bike experiment: the data

(Previously: A bicycle snob takes on an e-bike)

Last time around, we met the e-bike, and discussed how it came to pass. I was starting to feel comfortable with how it rode, and I was curious to see how it handled my commute.

Speed, or lack thereof

The first lesson in e-biking is that it isn't really about speed (that is, on a legal in Canada, pedal assist bike). Yes, there are circumstances where you're moving dramatically faster than you would on a non-e-bike. But overall, if you're a somewhat fit cyclist, your speed on flats will not be much greater than what it would be on a regular bike. At times, I found myself getting dropped by fit guys on road bikes. They most definitely would have dropped me if I was on a regular bike, and perhaps a bit more quickly, but it is far more embarrassing to not keep up when a motor is assisting you. I would feel an urge to rush off after them to explain the challenges of a 32 km/hour speed limit, but that would have required an ability to push through that limit (which is more or less impossible due to all of the additional weight and resistance of an e-bike).

Almost immediately, one of the main premises of my experiment was shot. No matter how I sliced it, the e-bike was going to be a fair bit slower than a car and not much faster than a regular bike. Still, the scientist in me required rigorous documentation to prove this out. I spent the next few months with a Garmin GPS jammed in my pocket.

What I found is that I was about 13-15% faster over flat ground. On my regular bike, my average speed was in the 25-26 km/hour range, while on the e-bike it was in the 29-30 km/hour range. It's difficult to maintain a speed at the limit of the bicycle, so your cruising speed ends up sitting just below. My one wish for the Bosch system that powers this bike is for them to put some more work into how the power tapers off near the speed limit.

Things get interesting on hills, as the bike will astonish the first time you motor up a moderate grade at 30 km/hour. There are more variables when climbing, but when I was able to isolate a straight stretch of road with a consistent climb, I found that my average speed was nearly 60% faster, jumping from an average of 16 km/hour all the way up to 25 km/hour. It's no free ride, as it does take a fair amount of effort to keep that speed, but you are getting a lot more bang for your buck.

Putting it all together, my commute times did drop, but not by a large margin (see graph below for details, and thanks to Hypothesis Media for the help on this). There were not enough hills (only 119m of vertical climbing over the entire route) to take advantage of the massive hill climbing delta, and the speed difference was not great enough to matter much on the flat bits. Overall, I dropped less than 10 minutes from my roughly hour long bike ride, and I still took up to twice as long as in my car (depending on traffic and time of day).

The largest thing working against me was that my commute was against the flow of traffic, and it was too easy to get in my car and drive. Plotting a hypothetical commute in the opposite, with traffic direction, my commute-by-car times ballooned, to the point where driving vs. e-biking would very nearly be a wash. Clearly, these things could work quite well under certain circumstances.

The New New Job. Same as the old Job

Halfway through the test, I decided that the new job wasn't for me. So I went back to a new job in my old organization. Once again, I was looking at a reasonable 12 km commute by bike each morning. This is where the dilemma of the e-bike took hold again. A 12 km/30 min (one way) bike ride is plausible each and every day, and it would be lazy to ride an e-bike that distance. But I am too lazy to ride my bike for an hour each day, so I don't. But riding an e-bike is lazy compared to riding the non-e-bike ride that I knew I should be riding, but wasn't. It's like a really nerdy chicken vs. egg conundrum of laziness. Still, in the name of science and data, I jammed the Garmin back in my pocket, hopped on the e-bike (and a few other modes of transport) and timed my routes. Once again the data surprised me.

Relative to my longer commute, my new/old commute has what feels like a fair number of hills. It still only has 100m of climbing, but over a much shorter distance. For much of the ride, you're either going up or down something, and this is the thing that keeps me in bed most mornings. I thought the e-bike would be my secret weapon, and I actually expected to gain enough time to challenge the commute time in my car.

But it was not to be. I only dropped 4 minutes or so from my times vs. my non e-bike, and I was still 5+ minutes off what I could do in my car. The optimists will note that this is a much closer result than seen compared to the driving times on the longer commute. I did feel like I was getting much closer to an optimum use case for an e-bike. But at the same time, it almost seemed like it was hardly worth it when the times were compared to those on the regular bike. My conclusion was that if this e-bike was going to change my life, it wasn't going to be due to time saved. And that will be what we explore next time.

Next: Conclusions