In the past week, Remco Evenepoel won his fourth consecutive World Championship in the time trial. He rode the 39km course at an average speed of 52km/hour – an entire kph faster than his closest competitor. It’s looking really hard to see how someone could actually beat him in the medium term. This got us thinking, how hard actually is it? How many watts are needed at different drag coefficients to beat Evenepoel?
It’s clear that Remco is on another level to everyone else. They’ll all be scratching their heads wondering how to beat him…
How good was his performance?
Fiddling with the various values that go into myWindsock, we were able to achieve a time on the exact GPX file that he uploaded to Strava that’s just two seconds off what he actually uploaded.
Evenepoel has a reported FTP of 425W and we’ll assume a system mass (rider + kit + bike) of 72kg, not that is makes a huge difference on such a flat course. We will set CRR values and drivetrain resistance values that assume he’s on the fastest possible set up. Presumably they weren’t using slow tyres…
This is the modelled average cda of Remco, including some sitting up for cornering and more. The minimum cda is the value recorded in the wind tunnel and reported.
Using the file that Remco uploaded to Strava, we can play around with the values until we’ve matched his time. This gives us an average cda of 0.194 during the course of the TT (however, this includes all sitting up, cornering etc) so during the long flat periods of the TT his cda will be significantly lower. Evenepoel’s reported cda is 0.17 in the tunnel so this chimes with our estimate including him sitting up and cornering.
What happens to the power requirement as we change the cda from 0.194 up to 0.21? This represents a decrease in aerodynamics of just 3.5%.
cda
Power needed for 45 minutes (W)
0.194
391
0.200
401
0.204
409
0.210
419
This table of values assumes that every single other aspect of your set up is exactly the same as Evenepoel including system weight, rolling resistance and such. In reality, it’s very unlikely that these numbers would exactly match.
The performance by Evenepoel required him to do an average power of 391W which is 6.2W/kg. If Evenepoel was just 3.5% less aerodynamic the required relative power would suddenly be 6.65W/kg. While W/kg is not particularly relevant for time trials, it does showcase the vast difference that a small change in aerodynamics can make in power requirement.
What about an 83kg rider with a reported cda of 0.19? Asking for an Italian mate…
Ganna rode at an average speed of 51.3kph during the race. Adjusting all of the other values to meet this speed comes out at a required estimated average power for Ganna of 440W for his roughly 46 minute effort. This is in line with Ganna’s reported threshold values. In order to achieve the time that Evenepoel did, we can adjust the power values as Ganna is extremely aerodynamically optimised. It turns out, he’s 15-20W short.
Those of you reading this in the UK will, undoubtedly, have had multiple occasions in your life where you’ve wished it was possible for you to change the weather. Well, we have good news for you. For the low price of your annual myWindsock subscription you will now be granted this power. Here’s how…
Open up myWindsock’s Surface platform
Scroll down the panel on the left hand side of the screen
Click on the “Weather Override” drop down menu option
Input values to change the weather
Unfortunately, for legal reasons, we are only able to change the weather within the course simulation on myWindsock, we are not able to adjust the weather outside. That said, it’s still interesting…
You go faster when it’s hot
The human body doesn’t love hot temperatures, especially a British one. That said, there’s an offset – as the air resistance of a bicycle decreases when it warms up so you need to produce less power to go the same speed. Bare with us while we do a little bit of maths.
This expression tells us about the power needed (P_aero) to ride at a given velocity (v). You’ll notice there’s a little symbol that looks like the letter p, that’s a Greek letter called rho and in this equation it represents air density. What this tells us is the power required to ride at a specific speed, v, is directly proportional to the density of the air.
Now, air is made up of tiny little particles, mostly:
Nitrogen molecules ((N_2)) — 78%
Oxygen molecules ((O_2)) — 21%
Argon atoms ((Ar)) — 0.9%
Carbon dioxide molecules ((CO_2)) — 0.04%
Water vapour molecules ((H_2O)) — variable
When it gets really hot, all of these particles whizz around quite a bit faster which means in every cubic meter of air, there’s actually less particles than when they’re moving slowly. When you’re cycling, you have to push each of these particles out of the way – that pushing is the force we feel from the wind and we call it “air resistance”.
Air density is proportional to atmospheric pressure (P – not power this time) and inversely proportional to temperature. When temperature goes up, air density goes down (if all other things are fixed).
This is why you go faster when it’s hot. In real life, the temperature does impact your ability to produce power. But on myWindsock – we can just pretend that isn’t the case…
How hot does it need to be for me to break the Poggio KOM?
The Poggio di San Remo segment on Strava is held by Tadej Pogacar and Filippo Ganna who some how managed to ride up it at 39kph. Neither of them share their power data on Strava but presumably they did quite a lot of watts.
As we’ve already discussed, air resistance decreases when the temperature goes up. That means, you ride faster for the same watts. Imagine a 76kg rider on a 6.8kg bike doing 500W on this climb in a relatively aerodynamic road bike position, with a cda of 0.23 on a windless day that’s 30 degrees. This is 6.5W/kg for around 6 minutes, so pretty good but not alien level world tour numbers.
This gives us a prediction of 6 minutes and 16 seconds. A pretty good time on this segment. But what happens if we change the weather?
Getting the perfect day
We are now going to calculate the conditions required for our imaginary rider to get the KOM on this climb. We will start with the wind…
The Poggio di San Remo segment is twisty enough that if we turn the wind up too much, the headwind section slows us down more than the tailwind speeds us up…
Wind Speed (m/s)
Segment Time (min:sec)
0
6:16
5
6:09
10
6:07
15
6:12
As it turns out, 10m/s is the optimal wind speed (at 30 degrees with this cda) on this climb. Any faster, and the headwind section slows us down more than the tailwind speeds us up.
We need to find 37 seconds to get that time down to 5:30. We will use our optimised wind direction and wind speed to increase the temperature until we find that extra 37 seconds.
Temp (degrees C)
Segment Time (min:sec)
30
6:07
40
6:05
100
6:00
Well, looks like we aren’t getting that KOM then…
It turns out, a regular human has absolutely no chance competing on the Poggio with Ganna or Pogacar. Ah well, we tried.
Every grand tour used to contain a long flat time trial but these days they’re becoming more and more rare. Truthfully, we miss them and the organisers of the Vuelta must have been reading our minds because they’ve bought it back, though we’d prefer for it to be 50km like the good old days.
After making their way out of El Puerto de Santa María, riders will sette into their pace for a prolonged high-speed effort. There are a few twists and turns in Lomopardo and Estella del Marqués, where there’s also a time check. The route is tailor-made for pure time trial specialists so look out for big performances from Küng and Hayter.
The course map
The course is pretty simple, there are around 10 corners that will cause the riders to have to take some kind of action, be that get off their aero bars or slam the anchors on, but in general it’s a flat TT that goes from one town to another – the kind we’ve seen a million times.
The wind map and surface friction
Anyone who’s ridden their bike in coastal European towns will be familiar with the wind. We’ve all left a town into the open areas and found ourselves suddenly being thrown around by the wind between hedgerows and buildings. That’s why we developed the surface friction mapping at myWindsock.
The myWindsock surface friction plot displays the area’s topology. Shaded areas are where the wind will move slower than forecasted due to buildings, trees or other physical features of the landscape. As it happens, for this particular time trial the wind is relative mild but we can still see areas of the course where buildings slow the wind down around the start and finish towns. The wind is forecasted to be around 5m/s (18kph) for the time around which the GC riders set off.
The race is forecasted to be majority tailwind which means that the windiest part of the day is actually the fastest.
As you can see, the wind adjusted elevation plot shows the course to be effectively net down hill, despite finishing at a higher altitude (slightly). Essentially, riding the course will feel “down hill” as riders are pushed along by the wind. Especially in the middle section between the two towns.
Riders will be setting off between roughly 2pm and 5pm for this time trial. Broadly speaking, conditions get slower throughout the afternoon giving the time trial specialists who are a long way down on GC even more of an advantage against the GC riders. The only GC rider who has a track record of beating the TT purists is Roglic, who will be setting off during the slowest part of the window.
There’s a ten second difference between setting off at 15:00 compared to 17:00, so there’s not a lot in it but we’ve seen world tour time trials of this distance won by smaller margins.
How hard is it to gain two minutes?
To have any chance of winning this Vuelta, Roglic will have to ride this TT about 2 minutes faster than Mas. Given this will be the target, how good a performance is actually needed to pick up two minutes on a “bad” time trialist?
A bad time trialist in the Vuelta is a pretty good time trialist:
Enric Mas will probably not mind us saying that, relative to other riders at his level, he’s never been the best time trialist. Reportedly, he weighs about 61kg and has a sustained 40 minute power of around 370W. Let’s say he’s about 4% lower power on the TT bike than the road bike (this is the upper end of “average”) then we can guess he will ride around 355W for this TT. We’ll assume his cda is 0.2, it’ll likely be a little lower but as he’s not a TT specialist he will also sit up a bit, so that’ll drag his average cda up slightly. We will assume his mechanics have him on fast tyres and a fast drive-train.
Plugging all of these assumptions into myWindsock gives Mas a predicted (but large error bars here as this estimate really was a bit of a guess for things like power and cda) of 40:29.
In order to challenge on GC, a top GC rider will need to produce a time of 38 minutes and 30 seconds. What kind of performance does this require? Let’s go back over some of our assumptions and put them into “peak TT Roglic” mode.
Back in the day, Roglic wasn’t just a ‘good for a GC guy’ time trialist, he was great. He’s reported to be at a weight of 65kg, which is a couple of kg heavier than Mas. During this Vuelta, he’s been climbing at a similar, but slightly slower, rate as Mas. Based on his track record though, we can estimate him to be slightly more aerodynamic as well as have a slightly lower power loss in the TT position.
We will assume his sustained 40 minute relative power is 1% lower than Mas at the moment, based on their climbing performances so far during this race. This gives him an estimated 390W to play with, but we’ll assume he loses only 2% of this relative to Mas in the TT bars instead of Enric’s 4% drop off. That means our estimated TT power for Roglic during this TT is 383W. We will also assume a cda of 0.19, we could be more aggressive with this estimate but he’s currently struggling from the impacts of a crash in training before the Vuelta started.
All of these assumptions lead to a time of 38:56, about a minute and a half faster than Mas.
Now, in order to get that two minute gap back, he will have to take some risks on pacing. Let’s now assume he’s magically slightly more aero at 0.18 and loses no power in the TT bars. This is essentially the best possible day that Roglic could conceivably have…
This would give Roglic just over 2 minutes back. He needs a generational day, but it’s within the bounds of possibility.
This blog scratches the surface of how World Tour Cycling Teams use myWindsock to prepare for time trials. If you want a detailed look at Stage 18 of the Vuelta – you can check out the forecast here.
myWindsock’s new version, surface, allows you to customise the information panel on the right hand side of the user interface. While, of course, your account comes pre-installed with a large array of pre built panels focussing on many aspects of cycling including weather, aerodynamics, terrain and more – each rider is different and this blog will focus on how to go about setting up your own panels.
Panel Basics
This is a screengrab of the right hand side of the screen on desktop with the Ironman 70.3 World Championships course pre loaded into myWindsock. By clicking the three dots on the right hand side you’re able to see a drop down menu which contains the panel you’re in, the ability to edit it and an option for panel library. Next to the three dots, there’s a a star, which allows you to input a text based description of what you want and myWindsock will automatically adjust the panel based on what you want.
The pre loaded Panel Library comes with six options…
Default – A broad overview of the course.
Performance Deep Dive – Focussing on effort and power delivery.
Wind Analysis – This is self explanatory.
Course and Terrain – Focusing on the elevation and physical aspects of the course.
Race Plan – A panel designed for race preparation
Aerodynamics and CdA – Focussed on aerodynamic analysis
Users are also able to update the pre loaded panels, as well as design their own from scratch. Today, we’ll focus on editing the Race Plan panel to go from the default panel to one more focussed on triathlon.
Editing a Panel
There are two ways to edit a panel, the option from the drop down menu or the text input option.
The data panel can be customised with a prompt.
In order to decide what to prompt the panel we need to decide what we want to know in order to prepare for the race. Preparing for a triathlon is a unique challenge as we need to carry a lot, it’s not actually a maximal effort and we need to get the gearing correct. It’s also, in the case of professional racing, a “semi drafting” race where some advantage is to be gained by being among a group of riders even if they must stay 20m apart. This means that energy investment is less linear than during a pure time trial as energy applied at seemingly non optimal points in the course might actually make sense in the context of non-draft triathlon.
We can use the Panel Index to see what’s currently contained in the Race Plan. With this, we can see what’s currently contained in the plan and decide what we want to delete or add. The first thing we’ll do is prompt the text box to delete the “W’ Balance Risk” plot as I don’t have up to date critical power information for this race – just a rough idea of the power I could hold, so we’ll delete this as it provides a false precision for the pacing plan in my specific case.
We can ask myWindsock to delete the panel for us in natural language like this. The new panel index now contains four charts.
How many charts is too many?
For a course that’s 90km with mountains, it’s possible to summarise all the pacing and race preparation information between 5 and 10 charts. The exact charts you feel you need prepared is left for the user.
The first thing we will do is add two charts, one which has an elevation profile and a second which shows us the distribution of gradients during the ride.
Our new Panel Index looks like this. We can now go onto looking at the Pacing Strategy Section.
The interval sessions that appear in your training plan are sometimes in significant disagreement with the terrain around you. This can make route planning extremely difficult. The demands of races dictate your training sessions, but the demands of your training sessions can dictate your route. This blog will take you through how a training session can be planned using myWindsock.
The session we will plan today is a roughly 2.5 hour ride in Andorra with 12 x 3 minutes low cadence intervals with 1 minute between them. This means the climb needs to be long enough to do the whole block, steep enough to tolerate low cadence training but shallow enough that the rest in between is easy enough at a normal cadence.
Before the 12 x 3 minutes we have a warm up including 4 x 12s sprints with 48s rest and after the 12 x 3 minutes we have a 1 hour block of “zone 2” – so we’ll need a pretty much continuous up hill for that too.
This is the session structure. But the tricky thing is the start point for the ride is at 1800m, so we need to get down low enough to start while still doing some warm up sprints (not downhill). As such, we’ll break the ride up into 3 points, warm up (20 minutes), 3 minute efforts (48 minutes) and zone 2 (60 minutes).
The warm up
The goal is to get down as low as possible between 20 and 30 minutes while finding a small climb to do the warm up sprints on. The warm up starts with 15 minutes easy pedalling, and the road down hill is quite easy to pedal the whole way down as it’s pretty straight and pretty wide, so we’ll pop that into myWindsock and see how far we can get in 15 minutes…
15 minutes of easy pedalling gets us here, almost to the entrance of the tunnel which is flat and a safe enough place to do the warm up sprints as there’s a nice wide bike path through it and, luckily, I don’t sprint very fast. One advantage of making the route with myWindsock is that it allows me to plan to go through the tunnel, whereas Strava’s route planning tool seems to think bikes aren’t allowed through it.
myWindsock is predicting that we’ll get out the tunnel and have our 5 minutes post sprints recovery in just under 30 minutes.
By clicking the timeline at the bottom of the screen, we can add intervals. Adding in the sprints and recovery reduces our time estimate to the bottom of the climb to around 28 minutes.
3 minute efforts
The bottom of the longer climbs the other side of the tunnel start at around 1200m above sea level. The intervals total 48 minutes and during this our rider will climb at a rate of 1100m/hour so we need to gain around 900m for the duration of the intervals.
myWindsock is now estimating that we’ll get to the turnaround point at 1:13 into the ride. Technically, the recovery after the intervals into the zone 2 is meant to be 5 minutes, but we’ll hit the descent and see how long it’ll take to get down.
The Zone 2
The pleasant way back is via the Col d’Ordino but this doesn’t take me an hour.
The myWindsock prediction has us getting to the bottom of Ordino at around 1:30 into the ride and we’ve got 60 minutes of zone 2 left to do. The zone 2 up hill will take up another 900m of elevation or so.
From Ordino town to the very top will take about 35 minutes and bring the total ride duration to just under 2 hours and 10. That means we need to descend down the other side and ride back up toward the starting point. We’ll be at the bottom of the descent two hours and 15 minutes into the ride with 25 minutes of up hill riding to go which conveniently will take me right to the front door of my favourite cafe.
The total elevation profile of the ride looks like this. Including descents the estimate is a total of 2:50 minutes which is a little longer than the planned session, but it’s a price worth paying to avoid going out and back.
After planning the route in myWindsock, I’m able to send it directly to my bike computer as the two are connected.
This coming weekend the RTTC National 10 Mile Championships are taking place on the O10/1 course near Doncaster in South Yorkshire. The women’s and junior races takes place on Saturday and the men’s race is on the Sunday. Both days the first rider is off at 10am and the riders will be on course until around 11:20 on the Saturday and 12:10 on the Sunday. The course is identical both days though the weather forecast varies slightly.
The course itself is a pretty simple out and back so the only real variation in pacing will come from wind and blow ups.
Both days are currently slated to be dry, temperate and relatively windless so our forecasts will focus mostly on which day is the fastest. We will set up the riders each day as the same, with a cda of 0.21, a system mass of 75kg and an average power of 300W then see how the time evolves across the start time of each day.
The forecast
This graph shows how the conditions evolve throughout the day. This graph shows the conditions relative to neutral weather. When the points are at -1, this means that the course is 1 second faster, 0 means the conditions are completely neutral and above zero is that many seconds slower.
There is a brief period on the Saturday where the wind makes the course marginally slower. On the Sunday, the conditions get marginally faster throughout the morning. That, alongside traffic picking up, means that conditions for the final rider should be slightly faster but barely noticeable.
The margins are tight, and both days are pretty fast. Due to the shortness of a 10 mile TT it’s not an unrealistic scenario that the podium could be split with just a handful of seconds. We should expect to see some pretty fast times though, with good conditions for racing forecast.
On Sunday the 16th of August the RTTC National 12 hour time trial championships will be held in Monmouthshire on the R12/16 course. The first riders roll out at 6:01 and will be on the gas for the following 12 hours, completing as many laps of the course as they can in that time.
12 hour TT courses can be a little harder to model performance on seeing as we don’t know exactly what laps the rider will do. Typically, this decision is made on the ground by the race director as to which part of the course riders use due to conditions on the road and other logistical considerations. Obviously, the part of the course with the big hill in it is most likely slower, so if riders spend more time on there than the flat bit, their modelled speed will be slower too.
The Weather
A 12 hour time trial is quite a bleak endeavour, even for the most enthusiastic riders there’s always at least one portion of the race that leaves you questioning why you entered. Nothing would accelerate that moment quite like the freezing cold and the rain, however conditions appear to be on the side of the participants this year.
myWindsock’s weather panel has precipitation at 0 throughout the event, with a patch of rain moving in at 8pm. If that comes early it could make the last hour quite unpleasant, but by that point in a 12 hour time trial, it’s all unpleasant anyway. On top of the precipitation, the wind forecast is low throughout most of the day but it also picks up a little in the last hour, though 5.8m/s is not a particularly high wind speed.
The Elevation Profile
One of the toughest things about a long time trial is the back breaking work of spending 12 hours in the TT bars. Every time you sit up, you lose a couple of metres from your result. Steeper hills and low wind speed sections can give riders the opportunity to sit up without paying a penalty.
Unfortunately for these riders, there’s limited opportunities to get out the saddle. Even though the elevation profile is rolling, it would appear that none of these climbs are particularly steep providing limited opportunities to sit up without losing time. On the other hand, this can help during a 12 hour TT as high torque efforts come with muscular damage that feels absolutely dreadful in the second half of the race. A practical take away for riders is they should sit up during the parts of the course they’re travelling their slowest on.
As you may have noticed from the graphs, myWindsock has a new look thanks to our surface platform, which is currently under development. One new feature of this is we can write extensions, which are small snippets of code in a “high level” language called Lua. A high level language is one that’s reasonably close to written english, making the syntax easily decipherable such that learning to write these extensions can be done by anyone curious enough to try.
An example of extensions
We can use these extensions to take a look at the average speed changes as a rider sits up more or less frequently. Setting up a time interval, we can see what happens to the average speed if our rider sits up for 60s at the following intervals (for an 80kg system mass with an “in position” cda of 0.23 riding at 240W)…
Every 3 hours: 40.1kph
Once an hour: 40.1kph
Every 30 minutes: 40.0kph
Every 10 minutes: 39.6kph
This is the cda plot of our extension, obviously the “real life” value will look absolutely nothing like this – the real world is not this neat. It does help us quantify the impact of a small thing like sitting up for 60s. The flexibility of myWindsock extensions allows riders to test the difference in speed of sitting up every 15 minutes for 30s against sitting up every half an hour for a minute – or some other such variations. Obviously, there’s an infinite number of possible variations you can try, many of which are specific to individual riders. This is why we built the ability for you to write your own extensions.
All of the CTT courses are available in myWindsock. You can check it out here.
It’s well understood in cycling circles how good you need to be to win World Tour men’s races with a summit finish, but the data as to how fast you need to ride up a hill in the women’s field is a little more sparse and not reported on as widely. In 2026, the Tour de France Femmes goes up Mont Ventoux and the QOM on Strava is currently held by Illi Gardener with a time of 57:35. While Illi isn’t a World Tour rider, she’s an extremely accomplished climber having won the British Hill Climb Championships three times. As such, we can expect a similar-ish time from the fastest women in the Tour de France Femmes this year.
Mont Ventoux – The Segment
At 1,910 m (6,270 ft), it is the highest mountain in the region and has been nicknamed the “Beast of Provence”, the “Giant of Provence”, or “The Bald Mountain”. It has gained fame through its inclusion in the Tour de France and over the years we’ve seen some epic battles on this climb.
The QOM has a VAM (a measure of climbing speed) of almost exactly 1400m/hr. We can see the peak climbing performances from the women’s world tour mapped below.
Peak Vollering has shown herself capable of 5.1W/kg for an hour. She seems in pretty good shape so far, though there hasn’t been a pure W/kg test so far.
A power to weight ratio of 5W/kg on 8.6% will lead to a VAM of (wind dependent, of course) around 1600m/hr, which will lead to a climbing time in the 50-52 minute range though this assumes no wind. The stage leading into the climb is pretty straightforward, which will lend itself to fast climbing.
The run in to the base of the climb is mostly flat with a nice warm up ramp on the Col de Suzette. It’s forecasted to top out at 28 degrees on Friday on the climb which is mercifully cool for this time of year.
How fast do you need to be?
The following table shows you the power to weight ratio, power at 60kg and the time you might expect to ride the climb on Stage 7 in order to achieve the listed results. We are assuming that, as is the case each year, the level continues to rise and it’ll take an all time great climbing performance to win the stage. For a climb of this length, that’ll be around 5.1W/kg to win the stage.
The Mont Ventoux segment on Strava can be uploaded directly into myWindsock. Simply “favourite” the segment within Strava and it’ll appear in the menu within myWindsock for deeper analysis. Notice the headwind on the final section of the climb. This is something that teams have been complaining about for a long time as many of the best performances on Ventoux have been achieved with this headwind. This edition of the Tour de France Femmes will see the riders take on the segment with a total of 58% of headwind.
To win
5.1W/kg
306W
56:24
Top 20 on the stage
4.7 W/kg
282W
1:00:01
Top half of the field
4.5 W/kg
270W
1:02:06
We wanted to add another row of what it’ll take to make the time cut, but it’s hard to know what that’ll be. It can be very dependent on the race dynamics. The grupetto will have to ride the climb much harder if they’re dropped early. The projected times here also display how outrageously good Illi Gardener’s effort on this climb was as it’s likely that the majority of the field will go slower than her.
All in, if you’re riding a climb of around an hour in length above 4.5W/kg, you’ve got what it takes to get round in a mountain stage of the Tour de France Femmes, presuming your bike handling and race sense is up to scratch. In order to win a mountain top finish you’re likely having to ride an hour long climb at or in excess of 5W/kg.
Top teams and riders will be using myWindsock to prepare for the stage. You can too by clicking here.
Tour de France Femmes 2026 Route stage 4 preview: Gevrey-Chambertin – Dijon. The race’s only individual time trial and the only course preview that takes account of the weather!
The 4th stage of the Tour de France Femmes is a 21-kilometre individual time trial. Just before the halfway mark, the riders face short climb of 1.8 kilometres at an average gradient of 6.9%.
The first major GC shakeup is expected in the first stage of the Tour de France Femmes in the TT that takes place in the land of mustard, finishing in Dijon.
For the first 7 kilometres, the riders head straight towards Dijon. A left-hand turn in Marsannay-la-Côte takes the riders onto a false-flat drag uphill. Gradually, the gradient increases until they reach the climb proper. This is the Lacets de Marsannay, officially a 1.8-kilometre climb averaging 6.9%, although it is perhaps more accurately described as a 3-kilometre drag averaging 5.7%. This TT’s profile is not dissimilar from the men’s ITT stage a couple of weeks back, where we saw Remco Evenepoel deliver a clinic in how to time trial on a 5% gradient.
The route continues on rolling terrain for another 2 kilometres or so, before a 6-kilometre descent takes the riders to Dijon. There are no bends to speak of in this section, allowing the riders to descend at breakneck speeds and make full use of gravity. The descent will take place in the aero bars – important for assumptions for our prediction. The race finishes in the city centre of Dijon.
What will it take to win it?
Last year’s Tour de France Femmes was without a time trial (which should be illegal), however three years ago there was a similarly profiled lumpy TT in which Reusser prevailed. This year, our opinion is that Vollering will take victory, however we’re going to forecast what it’ll take to win.
Demi Vollering has produced two historic climbing performances in recent years with her 10 minute effort at an estimated 6.4W/kg being arguably the greatest climbing performance by any woman on a bike in history. Credit for this graph goes to @CyclingGraphs who are well worth a follow on your social media platform of choice.
In order to win this TT, we think a rider will need to hit the climb in the TT bars at around 6.2W/kg, which for a 60kg rider is a staggering power of 372W. This, combined with a world class cda of 0.175 on the flat and assumptions that the mechanics are setting the riders up perfectly, leads us to a predicted winning time of 26:37.
The forecast for this TT prediction that we’re about to talk through can be found on myWindsock here. Given that, unfortunately, none of these teams invite us to the wind tunnel with them, our estimates for their aerodynamic data are best guesses based on what we’ve seen in the past from top female cyclists. Most of the uncertainty in our estimates for courses like this can be traced to not being sure on values related to equipment.
The starting point for this prediction was that, in order to be competitive with the performance of Vollering (assuming she’s here in peak shape), a rider will need to hit 6.2W/kg on the climb without breaking their aero position. They’ll go over their Critical Power and start using up their W’ allowance – essentially emptying the tank over the crest of the climb and holding on toward the finish.
The above time split used a Critical Power of 350W and a W’ of 16kJ for a 60kg rider. This particular pacing strategy leaves that rider with a minimum W’ balance of roughly 2kJ – which tells us there’s scope for a rider to potentially go harder than this on the climb. If this estimate is wrong, we’re forecasting that it’s more likely to be too conservative and it’s possible a rider will go toward the 26 minute mark, though breaking this seems unlikely.
The fact that the finish is tailwind descent also reduces the consequences for imploding on the time, as there’s a a pretty hard limit on the amount of time you can lose on that descent. In fact, if you stop pedalling after the top of the climb, you’d lose less than a minute on the run in to town with this pacing strategy. This means that riders may be more inclined to take a risk in their pacing on the climb, which further increases the chance our prediction will be beaten.
There’s only one time check on course, at 9.7km just before the proper crest of the hill. We should not expect massive swings from the intermediate check to the finish.
Any significant time loss amongst GC contenders will be on the climb but it will be lost due to air resistance, not gravity. Going up this drag sat up will lead to heavy time losses!
Riders hoping to be in the overall GC battle will need to ride the course under 28 minutes. This will limit their losses to around 90 seconds, anything more than that and you’re pretty much fully out the game. Race day conditions currently point to net tailwind for most of the TT which will help limit the losses of the weaker time trialists.
Runners are getting faster, whether it’s shoe tech, altitude or something in the water there’s no denying that the times are getting fast. As we know, aerodynamic resistance increases with the cube of speed which means the impact of aerodynamics becomes important very quickly. Recently, we’ve seen an example of this.
When Josh Kerr broke the mile world record he added his name to a list of record holders that had not been changed for 27 years. Clocking 3:42.66, the British 28-year-old took 0.47 off the world record running at a speed of 25.93 km/hour.
Kerr’s attempt was planned months in advance. Rather than quietly targeting the record, he publicly committed to it, building an entire training cycle around running one mile faster than anyone in history. Every decision was made with a single objective in mind.
Training, of course, was the foundation, but the margins came from everything around it. Kerr and coach Danny Mackey leaned heavily on physiological data, monitoring sleep, recovery and heart rate variability to balance workload and reduce the risk of overtraining. Away from the track, he installed an altitude chamber in his bedroom and an oxygen-enriched training room at home, allowing him to maximise adaptation between sessions.
World records aren’t solo efforts, and Kerr’s was no exception. Two pacemakers guided him through the opening laps at precisely the speed required for a 3:42 mile split, allowing him to settle into rhythm saving energy while drafting behind them. Kerr went through the first 400 metres in 55.3 seconds and then through 800 metres in 1:51.1.
By the final lap, with the pacemakers gone and a roaring London crowd behind him, the race became a battle against history. Kerr accelerated down the back straight and pulled clear of rival Yared Nuguse. Forty-five seconds later, he crossed the line in 3:42.66, becoming the fastest miler the sport has ever seen.
Josh Kerr didn’t visit a wind tunnel once during the preparation for this attempt, as far as I can tell. Should he have? He ran just shy of 26kph, speeds more commensurate with cycling than running. Aerodynamic drag in running is worth between 10 and 15 percent of the total work done against resistive forces. Perfect drafting is worth around a quarter of a second per lap, though this was “banked” and he used the first three laps worth of energy saving to empty his anaerobic energy stores in the final lap. This is about 0.75s worth of energy saved, slightly more time than the amount he broke the record by – it’s clear drafting helped him here.
During the final lap, without drafting, Josh Kerr ran 54.9 for the final 400m, this is a speed of 26.2kph and at this pace a 5% reduction in cda would save him a 10th of a second. To put that into “time trialese”, that would mean finding a saving of around 3W, this begs the question as to whether or not Josh Kerr should be wearing aero socks and aero sleeves.
The key takeaway for normal runners here is probably headwind related. It’s pretty hard for us mortals to get up to an air speed of 26kph unless we’re running into a headwind, but the energy savings of drafting if we’re running into a headwind can be significant, in the realms of 5 to 10 seconds over the course of a 5k Parkrun.
UK Time Trial Events
