Bicycle Tire Watt Requirements

[Jim Rogers]

@Nessism1

Here are some data to back up my statement that, depending on the tires compared, the difference in absolute watts needed could be up to 50%.

They are from Jan Heine's nicely-generated, well-controlled, long-term, highly respected data as published in Bicycle Quarterly over two decades:
______________________
René Herse Extralight 35–42 mm
Total rolling losses (pair): ~18–22 W


23 mm race tire (older generation, stiff casing)
Total rolling losses (pair): ~28–35 W


21 mm tire (1990s–2000s style)
Total rolling losses (pair): ~35–40+ W
________________________

Note that these data represent real‑road total energy losses, not drum cRR, which is a critical distinction.

In fact, it was Jan's insight that drum cRR doesn't tell us much and that the key is vibrational loss to the water molecules in the riders' bodies that enabled him to disprove the myth and lore of skinny tires that was so prevalent for so long.

So, pick a RH tire from the low end of its range (~18W) and a skinny tire from the high end of its range (~35W for a 23mm tire and ~40+W for a 21mm tire), and you can indeed see a 50% increase in the watts needed to maintain speed on a skinny vs fat tire.

I don't blame you for being incredulous on this as I didn't believe it at first either. So, how do we know these data are actually correct?

I would say that the fact that the pros are now commonly running 28mm tires, and some are even using 30mm tires (unheard of back in the 1990s when I was racing!) is pretty strong evidence to support Jan's data.

Many credit Jan with revolutionizing bicycle tire theory. And how did he do that? with definitions and data, and not just accepting the pro's conventional wisdom of "I won the TdF of skinny tires, so they are the fastest!"

It took a lot of time and a lot of data to convince me that tires could make such a big difference, but I now believe that all arrows are pointing to this as being a fact.

 

[rijndael]

I went down this rabbit hole too. It started when Latex tubes became more common. I was shocked at the feel difference they offered.

There's also a good bit of data offered on https://www.bicyclerollingresistance.com/

 

[Nessism1]

50% extra rolling resistance doesn't mean the rider needs to exert 50% more effort (watts output for the same speed). Wind resistance is the big issue, not rolling resistance.

 

[Jim Rogers]

50% extra rolling resistance doesn't mean the rider needs to exert 50% more effort (watts output for the same speed). Wind resistance is the big issue, not rolling resistance.

Yes, wind resistance is huge as it goes up by the cube of the speed. So, I should have been more precise with my language, so let me re-frame:

"If rolling‑resistance power goes up by 50%, the rider does need to supply 50% more power for that component."

 

[IndyIan]

50% extra rolling resistance doesn't mean the rider needs to exert 50% more effort (watts output for the same speed). Wind resistance is the big issue, not rolling resistance.

+1 We've done some gravel group rides with a range of bikes and tires, and for coasting down a moderate ~5% hill, tires or bike type didn't seem to matter as much as weight.
With all of us tucking to our best ability The 250lb guy on a fat bike with well worn mid fat tires, coasts a bit faster than the 150lb guy on a road bike with ~20mm tires. My "gravel bike", old XC bike with bald-ish 2.1 XC tires and 190lb me on it, coasts about the same as the 150lb road biker.

Now my trail bike with 2.8 full tread depth knobby sticky tires, is dead slow compared my gravel bike, so tall squirmy treads do eat watts!

 

[MRC01]

An important factor often disregarded is the surface being ridden - specifically, how smooth it is. There's a reason track racers on billiard smooth surfaces still today use skinny high pressure tires. And at the opposite end of the spectrum, why tubeless tires on rugged MTB trails are great at 20 PSI.

This factor cuts both ways. It's one of the primary reasons why for decades, road cyclists rode on 23mm tires at 140 PSI, which is counterproductive. Yet it's also one of the reasons why so many people now believe the opposite, that "wider and low pressure is always faster", which is also false. The truth is in between: the ideal tire size & pressure depends on the surface being ridden. Roads in the real world are not as smooth as velodrome tracks, but they aren't gravel trails either (though in some cities having poor street maintenance, they might seem that rough).

 

[MRC01]

It started when Latex tubes became more common. I was shocked at the feel difference they offered.

Back in the 1980s when I first tried latex tubes I thought it was a myth. But the first time I rode on them I could feel a difference. Just a bit more comfortable at the same pressure, and it felt faster. So I stuck with them and still use them today. Only decades later did I learn that difference was real, they really are faster. Just a smidge, but that difference can be squashed if the tire is not compliant enough.

 

[tcp71]

What was the weight of those test tires? If the "extralight" wide tires are significantly lighter, wouldn't that effect the outcome?

 

[jadmt]

I subscribe to bicyclerolling resistance and have tried so many tires....I have not had great luck with rene hersey tires as too many flats...I do currently have two sets mounted now on 26" wheels the naches pass and rat trap passes on my all around bike ...I will give up some rolling resistance to my Conti GP5000's in 700x32 with aerothan tubes and not worry about flats.

 

[Jim Rogers]

+1 We've done some gravel group rides with a range of bikes and tires, and for coasting down a moderate ~5% hill, tires or bike type didn't seem to matter as much as weight.
With all of us tucking to our best ability The 250lb guy on a fat bike with well worn mid fat tires, coasts a bit faster than the 150lb guy on a road bike with ~20mm tires. My "gravel bike", old XC bike with bald-ish 2.1 XC tires and 190lb me on it, coasts about the same as the 150lb road biker.

Now my trail bike with 2.8 full tread depth knobby sticky tires, is dead slow compared my gravel bike, so tall squirmy treads do eat watts!

Yes, the extra gravitational force on the heavy guy will tend to allow him to overcome air drag in the way a baseball compares to a whiffle ball.

But just to be clear (in case you were making this point), that extra weight is mostly due to the rider-- differences in bicycle weights don't make a huge difference. Some difference, but not a lot.

The two biggest factors by far, I believe, are wind resistance and tire resistance. I would estimate that on the average road ride over varying terrain, the breakdown as far as watts needed would be (just my estimates from reading-- no link):

-- ~60% of your watts go to pushing air
-- ~20% to tire and vibration losses
-- ~10–15% to lifting your total mass on the uphills
-- ~1–2% to the just bike’s weight itself.

An important factor often disregarded is the surface being ridden - specifically, how smooth it is. There's a reason track racers on billiard smooth surfaces still today use skinny high pressure tires. And at the opposite end of the spectrum, why tubeless tires on rugged MTB trails are great at 20 PSI.

True, and that's why smooth drum cRR measurements have turned out to be useless. I see that https://www.bicyclerollingresistance.com/ (which I've not visited in some time) is using a textured surface.

That's an improvement, but without some liquid (maybe a bag of gel?) added to the test rig to reproduce the vibrational losses from that texture, it's kind of meaningless.

This factor cuts both ways. It's one of the primary reasons why for decades, road cyclists rode on 23mm tires at 140 PSI, which is counterproductive.

I'm pretty sure it came from smooth drum testing that showed a lower cRR fo the tires. This can produce drag that is real, but it is minor.

Yet it's also one of the reasons why so many people now believe the opposite, that "wider and low pressure is always faster", which is also false.

The data I've seen seem pretty clear that it's not false, even at surprisingly wide widths. Do you have a link to the contrary I could look at?

The truth is in between: the ideal tire size & pressure depends on the surface being ridden. Roads in the real world are not as smooth as velodrome tracks, but they aren't gravel trails either (though in some cities having poor street maintenance, they might seem that rough).

This is correct-- and it's surprising the degree of "rough surface" a typical well-maintained, paved street is. All the micro texturing of concrete can create a surprising amount of micro flex in a tire, meaning that hysteresis losses are far greater than one might think (and explains why smooth drum tests are not helpful).

Back in the 1980s when I first tried latex tubes I thought it was a myth. But the first time I rode on them I could feel a difference. Just a bit more comfortable at the same pressure, and it felt faster. So I stuck with them and still use them today. Only decades later did I learn that difference was real, they really are faster. Just a smidge, but that difference can be squashed if the tire is not compliant enough.

I've never tried latex tubes, but the data I've seen does confirm that they are noticeably faster.

What was the weight of those test tires? If the "extralight" wide tires are significantly lighter, wouldn't that effect the outcome?

The Extralights are heavier-- here are some weights I found:


RH Extralight 35–38 mm = ~260–300 g
42–44 mm = ~330–360 g
Classic 23 mm race tires (Michelin Pro Race, Conti GP3000, Vittoria Open Corsa CX, etc.) 23 mm = ~200–230 g

I subscribe to bicyclerolling resistance and have tried so many tires....I have not had great luck with rene hersey tires as too many flats...I do currently have two sets mounted now on 26" wheels the naches pass and rat trap passes on my all around bike ...I will give up some rolling resistance to my Conti GP5000's in 700x32 with aerothan tubes and not worry about flats.

I've actually never tried RH tires specifically because of my aversion to flats. Jan claims up, down, and sideways that the lower pressure means far less stiffness in the face of sharp objects (i.e., the tires tend to deform around sharp objects rather than be punctured by them).

But it seems to me that a lot of people report a lot of flats on RH tires-- I don't know the truth on that, but it is a barrier (along with cost) to me using them.

I do plan on giving them a test someday, though.

 

[rijndael]

-- ~60% of your watts go to pushing air
-- ~20% to tire and vibration losses
-- ~10–15% to lifting your total mass on the uphills
-- ~1–2% to the just bike’s weight itself.

Drivetrain loss is real too .... especially if you're a cross chainer.

 

[jadmt]

Yes, the extra gravitational force on the heavy guy will tend to allow him to overcome air drag in the way a baseball compares to a whiffle ball.

But just to be clear (in case you were making this point), that extra weight is mostly due to the rider-- differences in bicycle weights don't make a huge difference. Some difference, but not a lot.

The two biggest factors by far, I believe, are wind resistance and tire resistance. I would estimate that on the average road ride over varying terrain, the breakdown as far as watts needed would be (just my estimates from reading-- no link):

-- ~60% of your watts go to pushing air
-- ~20% to tire and vibration losses
-- ~10–15% to lifting your total mass on the uphills
-- ~1–2% to the just bike’s weight itself.



True, and that's why smooth drum cRR measurements have turned out to be useless. I see that https://www.bicyclerollingresistance.com/ (which I've not visited in some time) is using a textured surface.

That's an improvement, but without some liquid (maybe a bag of gel?) added to the test rig to reproduce the vibrational losses from that texture, it's kind of meaningless.



I'm pretty sure it came from smooth drum testing that showed a lower cRR fo the tires. This can produce drag that is real, but it is minor.



The data I've seen seem pretty clear that it's not false, even at surprisingly wide widths. Do you have a link to the contrary I could look at?



This is correct-- and it's surprising the degree of "rough surface" a typical well-maintained, paved street is. All the micro texturing of concrete can create a surprising amount of micro flex in a tire, meaning that hysteresis losses are far greater than one might think (and explains why smooth drum tests are not helpful).



I've never tried latex tubes, but the data I've seen does confirm that they are noticeably faster.


The Extralights are heavier-- here are some weights I found:


RH Extralight 35–38 mm = ~260–300 g
42–44 mm = ~330–360 g
Classic 23 mm race tires (Michelin Pro Race, Conti GP3000, Vittoria Open Corsa CX, etc.) 23 mm = ~200–230 g


I've actually never tried RH tires specifically because of my aversion to flats. Jan claims up, down, and sideways that the lower pressure means far less stiffness in the face of sharp objects (i.e., the tires tend to deform around sharp objects rather than be punctured by them).

But it seems to me that a lot of people report a lot of flats on RH tires-- I don't know the truth on that, but it is a barrier (along with cost) to me using them.

I do plan on giving them a test someday, though.

I use the rene hersy psi calculator to set psi and still more flats than either contis or schwalbe tires
https://www.renehersecycles.com/tire-pressure-calculator/

 

[Jim Rogers]

Drivetrain loss is real too .... especially if you're a cross chainer.

Yes-- true! That's why my numbers don't add up to 100%. Drivetrain + Misc is probably in the 3-5% range.

I'm a classic '70s Suntour Super-6 freewheel guy, where I get 6 gears in the same space as the standard 5-gear freewheel and the resulting 120mm dropout spacing.

With modern offset spoke bed rims, I always get true dish-less wheels and no crossover issues. And, oddly enough, I'm the rare user of half-step gearing (done with two smaller-sized chainrings), so all gears are full available to me.

My drivetrains are custom and (I think) great!

I use the rene hersy psi calculator to set psi and still more flats than either contis or schwalbe tires
https://www.renehersecycles.com/tire-pressure-calculator/

You are certainly not alone in reporting that. I would really love to use RH tires, even at the high cost, in order to experience that great ride and fast speeds.

But I ain't payin' a ton to get tires that are going to slow me down with flats.

 

[Arc]

It's always been a known fact that different tires produce different rolling resistance. This isn't anything new and it's always been a tradeoff between performance and traction depending on the surface the tire has to ride on.

 

[Jim Rogers]

It's always been a known fact that different tires produce different rolling resistance. This isn't anything new and it's always been a tradeoff between performance and traction depending on the surface the tire has to ride on.

That was always the conventional wisdom, but it turned out to not be true.

Rolling resistance (as indicated by the dimensionless cRR) turns out to play a very minor role in tire drag and, surprisingly, there turns out to be no tradeoff to using wider tires when desiring higher speeds.

Turns out that wider is faster-- that realization was very new!

 

[MRC01]

... so many people now believe the opposite, that "wider and low pressure is always faster", which is also false. ...

... The data I've seen seem pretty clear that it's not false, even at surprisingly wide widths. Do you have a link to the contrary I could look at?

I was careful to use the key word: always. Just because wider lower pressure tires are sometimes faster, does not mean they are always faster. The evidence to the contrary is obvious: smooth drum rolling resistance testing, which resembles conditions of ultra smooth surfaces like an indoor velodrome track. This testing shows, for example, that the Conti GP 5000 in 30mm consumes 10.0 watts of rolling resistance while the Conti GP 5000 in 28mm consumes 8.3 watts. The 30mm is likely faster on real-world pavement, but the 28mm is faster on a smooth surface like a testing drum or indoor velodrome.

and, surprisingly, there turns out to be no tradeoff to using wider tires when desiring higher speeds.
Turns out that wider is faster-- that realization was very new!

Here's what I think is a more accurate way to describe new (the past 5-10 years) information about wider tires. We've always known that wider tires with lower pressures provide better traction and comfort, but we pay a penalty in speed/efficiency/weight for this. It turns out that penalty is smaller than people expected, and the impact of real world roads not resembling the smooth surfaces of rolling resistance measuring drums, is bigger than people expected. The combined effect of both is that the fastest most efficient tires are wider and lower pressure than previously believed.

This doesn't imply that "wider is faster" and "there is no tradeoff". It remains true that all else equal:

• On a smooth surface, narrow high pressure tires have lower rolling resistance
• Compliant (wide, low pressure) tires absorb surface imperfections more efficiently than stiff (narrow, high pressure) tires which transfer them to the bike & rider

On a typical road bike on typical paved roads, 30mm tires at 75 PSI are probably faster than 25mm tires at 100 PSI. But that doesn't imply that a 40mm tire at 50 PSI is even faster (even if would fit, which it probably wouldn't). On smooth pavement the 40 is probably slower. Once the tire is compliant enough to sufficiently absorb surface imperfections, going wider and lower pressure only increases rolling resistance and mass. That's the tradeoff.

 

[MRC01]

It's always been a known fact that different tires produce different rolling resistance. This isn't anything new and it's always been a tradeoff between performance and traction depending on the surface the tire has to ride on.

I mostly agree, though what is new is the discovery that the optimum tradeoff on real-world pavement is wider and lower pressure than previously believed.

 

[Jim Rogers]

I was careful to use the key word: always. Just because wider lower pressure tires are sometimes faster, does not mean they are always faster. The evidence to the contrary is obvious: smooth drum rolling resistance testing, which resembles conditions of ultra smooth surfaces like an indoor velodrome track. This testing shows, for example, that the Conti GP 5000 in 30mm consumes 10.0 watts of rolling resistance while the Conti GP 5000 in 28mm consumes 8.3 watts. The 30mm is likely faster on real-world pavement, but the 28mm is faster on a smooth surface like a testing drum or indoor velodrome.

Ok, yes, in the velodrome where the surface is glass-smooth, the cRR becomes the major contributing factor to tire drag.

But are you then going to say that because of that very narrow surface-case, that virtually no one rides on, that all the data are wrong that show wider tires (of equal construction quality) are pretty much always faster on any real-world road?

Because in fact, what your example shows sort of proves is the point I was making-- that cRR of a tire (measured on a smooth drum) is so minor compared to the vibrational losses, the only time it's relevant is on the velodrome track.

So, that basically confirms Heine's work rather than rebutting it.

Here's what I think is a more accurate way to describe new (the past 5-10 years) information about wider tires.

I believe the data shows that the only way your statements (that will be following) are accurate is if you posit no cap on tire sizes and allow crazy-wide tires.

While it's true that I didn't explicitly state a cap, I didn't mean to imply the relation would go to infinity. Perhaps I should have been more clear by stating a cap of "up to 60mm."

We've always known that wider tires with lower pressures provide better traction and comfort, but we pay a penalty in speed/efficiency/weight for this. It turns out that penalty is smaller than people expected, and the impact of real world roads not resembling the smooth surfaces of rolling resistance measuring drums, is bigger than people expected. The combined effect of both is that the fastest most efficient tires are wider and lower pressure than previously believed.

Not true, at least by the data I've seen. It seems pretty clear that, assuming a quality tire, up to at least 42mm (and possibly up to 60mm, depending on tire construction) there is no penalty. Here is link with plenty of articles related to Heine's data:

https://www.renehersecycles.com/category/testing-and-tech/

This doesn't imply that "wider is faster" and "there is no tradeoff".

Per the data I cited, it doesn't imply it because it doesn't have to-- there is actual data to show it (linked above).

However, happy to look at alternative data if you provide it.

It remains true that all else equal:

• On a smooth surface, narrow high pressure tires have lower rolling resistance
• Compliant (wide, low pressure) tires absorb surface imperfections more efficiently than stiff (narrow, high pressure) tires which transfer them to the bike & rider

Agree.

On a typical road bike on typical paved roads, 30mm tires at 75 PSI are probably faster than 25mm tires at 100 PSI.

Agree, and note that the acceptance of that fact took Jan Heine more than a decade of research, data, and debating to convince the world of. You're stating it as accepted fact (which is true), but it's only very recently accepted!

But that doesn't imply that a 40mm tire at 50 PSI is even faster (even if would fit, which it probably wouldn't). On smooth pavement the 40 is probably slower. Once the tire is compliant enough to sufficiently absorb surface imperfections, going wider and lower pressure only increases rolling resistance and mass. That's the tradeoff.

Disagree, and I would like to see data supporting that. 40s are not slower, and often faster (at least by the data I've seen and summarized at the link).

 

[MRC01]

But are you then going to say that because of that very narrow surface-case, that virtually no one rides on, that all the data are wrong that show wider tires (of equal construction quality) are pretty much always faster on any real-world road?

Your words, not mine. I didn't say that.

Because in fact, what your example shows sort of proves is the point I was making-- that cRR of a tire (measured on a smooth drum) is so minor compared to the vibrational losses, the only time it's relevant is on the velodrome track.

So, that basically confirms Heine's work rather than rebutting it.

I'm not rebutting Heine's work. My point is simple: the degree of relevance of cRR, and the optimum tire width & pressure, depends on the road surface.

... While it's true that I didn't explicitly state a cap, I didn't mean to imply the relation would go to infinity. Perhaps I should have been more clear by stating a cap of "up to 60mm."

That's better. Though the cap isn't a fixed number, but depends on the road surface, of course.

 

[Jim Rogers]

Your words, not mine. I didn't say that.

Yes, it was a question. Because the rest of your post clearly implies that, so I was asking for clarification.

I'm not rebutting Heine's work. My point is simple: the degree of relevance of cRR, and the optimum tire width & pressure, depends on the road surface.

I understand, but I'm saying unless you get super-extreme (glass-smooth velodrome or cooled magma flows), that's actually not correct and I provided a link with evidence of that.

Do you have any links I could look at that summarize or explain your position?

That's better. Though the cap isn't a fixed number, but depends on the road surface, of course.

Yes, sorry for the omission-- didn't think it would be necessary in the context of this kind of technical thread.

For the record, the bottom cap is 21mm.