In the new millenium, my ultramarathon cycling career will continue partly on a recumbent! My interest in these speed-machines dates back to 1997, when during my last quarter at Stanford University, I began designing a ‘bent of my very own. Unfortunately, I never finished that project. However, three years later, I have just building a recumbent based on George Reynold’s semi-low racer frame, the Reynolds Wishbone SWB (short wheelbase).
The advantages of a recumbent over a traditional upright bicycle include a more aerodynamic and comfortable riding position, arguably better safety, and typically higher attainable cruising speeds on flats and downhills. All of the human-powered land speed records have been achieved on recumbents. Their main Achilles Heel is climbing, although some riders, including famed Race Across America rider Pete Penseyres, claim that they are at least as fast on their recumbents as their upright bicycles up to grades of 10%.
Alas, after years of doing double centuries on my regular bike and recumbent, I cannot climb nearly as well on my ‘bent as on my “upright” road bike. Actually, on short and non-steep climbs, I climb about the same. In fact, on some rolling courses, I ascend faster on the recumbent just because the entry speed is higher from the downhill. For long or steep climbs, however, there is no doubt in my mind the recumbent is at a big disadvantage. See analysis below.
Comparison of Recumbent vs. Upright Bicycle
Below are excerpts from an email I wrote in March 2003 on this topic:
Thanks for your message. I have always meant to post something on my website about this topic, but it has been way down there on the to-do list. Anyhow, here’s my honest opinion, with some time comparisons for the same rides.
Let’s start out with the time comparisons.
- 1998 Solvang Double (Cannondale): 14:10
- 1999 Solvang Double (Cannondale): 14:26
- 2000 Solvang Double (Reynolds): 12:12 (my quickest double ever)
The Solvang is a flat/rolling course, with notoriously strong headwinds when riding northbound (which is like 33% of the time). There is only one semi-long hill of maybe 15 minutes at the end. I definitely felt that the Reynolds was quicker on this course, though probably not as much as the time difference suggests (because in 2000, I was so comfortable, I stopped much less!)
The next time comparison was the climb up Mt. Diablo in Northern California (goes up ~3800 feet in 11 miles, or a 6.5% average grade):
- in the 1999 Devil Mountain Double (Cannondale): 1:25
- in the 2000 Devil Mountain Double (Reynolds): 1:35
So here the recumbent wasn’t TOO much slower, but it is definitely at a disadvantage on the climbs.
To illustrate, here is another example: the Knoxville Double, with >12,000 feet of climbing over 200 miles (320 km), and hills seemingly everywhere:
- 2000 Knoxville Double (Reynolds): 19:05 (stopped a lot due to riding with someone else)
- 2001 Knoxville Double (Reynolds): 17:35 (did not stop much)
- 2002 Knoxville Double (Cannondale): 14:12! (stopped about the same amount as in 2001 or maybe even more due to getting 4 flat tires)
I was in better shape in 2002 than in 2001, but any way you look at it, this ride pretty much conclusively made me concede that I prefer to do any ride with significant hills on my Cannondale.
Last example: the Death Valley Double (200 miles, 9000 feet of climbing, with most climbing concentrated in a 25-mile stretch):
- 1997 Death Valley Double (Cannondale): 15:30, after riding a 5:30 first 100 miles (fastest century ever)
- 2002 Death Valley Double (Reynolds): 15:25
This course features Salsberry Pass, which is 16 miles long, and took me ~2 hours to climb the front side on the recumbent. and then ~1 hour to climb back up the back side. The rest is flat with nothing exceeding, say, a 3% grade. I thought it would be a good recumbent course, and indeed, my time was fair (better than my 1997 time in fact, but in ’97 I had multiple flat tires plus an injured ankle, etc., in the 2nd half), but after finishing the ride I really felt I would have been faster on my Cannondale. This is because I was so wasted after going up Salsberry Pass twice that my legs had nothing in them to take advantage of the flat course afterwards!
So in conclusion: for a flat/rolling course, the recumbent is faster (unless drafting while riding the Cannondale… which I’d tend to do more of [ride in pacelines] on the ‘dale than the Reynolds). However, despite the Wishbone having a reputation for being a good climber (for a recumbent), in my opinion, for a course with either long or steep hills, it’s definitely slower than a diamond-frame bike mainly because of 1) you are always using the same set of muscles, whereas on a regular bike you can alternate between standing and sitting and give your muscles a break, and 2) at 26 lbs (fairly light for a recumbent), it’s about 6 lbs. (2.7 kg) heavier.