Less Power, More Speed on the Bike: calculating the benefits of aero testing

Rewind to February 2022. The British time trial cycling season is a still a long way off, training is still very much in the “base” phase and I find myself driving down the M4 towards Newport. The destination: Newport Velodrome (or to give it its full name: the Geraint Thomas National Velodrome of Wales!). The aim: to get faster on the TT bike.



I’ve had my share of TT bike fits over the years (from very relaxed long-distance triathlon positions to the more aggressive short-course position I have these days), but always relying on the bike fitter’s eye, or a system like Retul. I’ve never done any real aerodynamic testing.


For someone with too many skinsuits, helmets and wheelsets it felt like it was time to bite the bullet and see what I could find out in a more controlled environment. With a wind tunnel session being more expensive than I could justify, a trip to an indoor velodrome seemed a good option (still expensive, mind you!).


Aero testing at the velodrome

The plan was simple. I’d take my bike and a few kit options to Newport; we’d fit the bike with an aerodynamic sensor paired to the power meter and speed sensor. We’d then do runs of about 9-10 laps of the track at 10-mile power and look for ways to improve my aerodynamic efficiency for the same power.


The first run was a baseline to get the bike in the right gear (the rear mech was then disconnected to help improve consistency across the runs) and to get the power about right (in actual fact, I later learned I’d gone a bit too hard, making the ninth and tenth runs bloody hard work!).


After the first two benchmark runs, Jacob (my coach and also ‘aero consultant’ for the day) showed me some video and immediately suggested my saddle was wrong. He felt he could improve the profile of my back which would in turn improve airflow (and no, I’m not telling you exactly what we did, sorry!).


And he wasn’t wrong. According to the aero sensor, the next run showed a saving of 9.4 watts (equivalent of up to 45 seconds in a 25-mile TT).


Next Jacob wanted to look at my arms, specifically the position of the elbow rests. Switching out the Drag2Zero rests for some WattShop rests and repositioning them yielded another 2.3 watts saving on the next run (another potential 11 seconds off the clock). It’s worth pointing out here that the ‘tolerance’ for results using the aero sensor and velodrome is about 1-2 watts. So, anything under that can’t be relied on for accuracy.


TT skinsuits for the win

Next up, we tested the clothing. First, we tested with and without gloves. As a rule, no gloves generally tests faster but my run without gloves showed a 4.7 watt penalty. I think everyone was surprised, so we ran the test again with the gloves back on and got exactly the same 4.7 watt saving. Unfortunately, not a 2022 saving as I’ve been running these gloves for over a year (mostly to stop my hands getting cold!).


Then it was time to test the two skinsuits I had brought with me. Despite one suit costing more than twice the other, there was only 1.2 watts in it (in the favour of the more expensive suit). But as above, 1.2 watts is within the margin of error, so it could have been even closer.

Next came the helmets. I had two options with me and WattShop had another two for me to try. If you read back to my blog on helmet testing (on the road) from last year, you’ll know that I was able to determine quite marked differences between three helmets. In particular, comparing the POC Tempur against the HJC Adwatt, I found them to perform comparably with a tailwind, but quite differently (in the favor of the POC) into a headwind.



Obviously, an indoor velodrome doesn’t have headwinds and tailwinds, so it was interesting to see that there was a 2.3 watt difference between the two helmets on the track. Testing the two helmets supplied by Wattshop, one was demonstrably slower (if only by 1.2 watts) which was again surprising as it’s the helmet that they’ve perhaps had the most success with for other riders, and the other was marginally (again, 1.2 watts, but in the other direction) faster than the POC.


Ultimately, the best of the helmets (under these test conditions) would have improved my 25-mile TT time by an estimated 21 seconds compared to the worst performing.


We then did a couple of control runs to re-test some earlier kit choices (which confirmed the results shown above) as well as some different overshoe options (top secret, sorry).


Overall, between the first control runs and the best runs of the day, I saw a total combined saving of 14.1 watts. Or a calculated improvement over 25 miles of something like 70 seconds. It hasn’t escaped my attention, however, that’s there’s a certain irony in the biggest saving of the day (the saddle position change) being achieved through the good old mk1 human eyeball (although confirmed by the speed sensor) – the skill of the bike fitter is very much alive and well!



Of course, road time trials are never held in such controlled conditions, so I was going to have to wait until the early season races to find out if those savings would transfer from track to road.


For that, stay tuned for the next update!