How efficient is a bicycle ? Rohloff v derailleur.

I’ve been reading some stuff on the comparison of the efficiency between Rohloff and derailleurs.
http://www.hupi.org/HParchive/PDF/hp55/hp55p11-15.pdf

To summarise in one sentence, a Rohloff hub is 2% less efficient than derailleurs.
At first glance, it might look like a Rohloff bike would be 2% slower than a derailleur bike for the same power input.
It’s not that simple though.

How much of a riders effort is wasted through transmission losses ?
If you take in to account power losses through wind resistance, rolling resistance of the tyres and so on, does that 2% really make a difference ?

Say, for example, transmission losses only account for 10% of total losses, that’s only 0.2% overall difference.
Does anyone know what the real number are ?
I would imagine there’s a big difference in the wind resistance and rolling resistance between a road bike on skinny tyres at 25mph and a mountain bike on knobblies at 10mph.
Is there any research on this subject on line anywhere ?

How much of a riders effort is wasted through transmission losses ?

Not a lot, chain driven bicycles are extremely efficient, IIRC it can be anything up to 99% of the input goes into the wheels – hence things like PowerTaps working. That decreases as things get more shagged!

Wind resistance is a huge part, but with a Rohloff you have to put 2% more power in to be at that speed.

So yes, a Rohloff road bike with 100psi 23c tyres is quicker than an MTB with a conventional transmission, but if you stick a Rohloff on a given bike you have to put out 2% more power to go at the same speed. Which isn’t insignificant really. 6W for me at FTP.

For how long is a Rohloff less efficient that derailleurs though?

Derailleurs are a disposable gear system and don’t last long in dirty conditions (if they’re not swiped off on a rock first)

Chains still wear on Rohloff’s too, which will decrease their efficiency. No idea how that compares to losses through the whole drivetrain.

Not a lot, chain driven bicycles are extremely efficient, IIRC it can be anything up to 99% of the input goes into the wheels – hence things like PowerTaps working. That decreases as things get more shagged!

I recall reading a paper many moons ago where that oft-quoted 99% was quantified – yes you can get 99% when running a new chain between two large, equally sized, rings on a straight chain line with the system being constantly lightly lubed. Out in a muddy field, meanwhile…… 😉

I had a Nexus, which had a little drag – but it was enough to bug you.

mind you I am thinking of going for an Alfine 8, so it can’t have bugged that much.

Out in a muddy field, meanwhile……

the rohloff chain is still getting muddy (without looking at the type of maintenance shy people rohloffs <edit> may appeal to) but already with a 2% handicap.

I’m not slating rohloffs BTW, I would be quite interested in real world scientific comparisons.

I have just switched from Rohloff to a whole new XT drive train.

The XT is noticeably less “draggy” feeling, particularly uphill but that was when it was shiny and new. Will give it a few weeks in a scottish winter with the chain getting clogged with mud and chain jumping i may think differently.

In answer to the OP not aware of any research

Would things like PowerTaps not work though if there was a significant loss through the system as they measure power output ‘at the road’ as it were? Serious question like…

Surely you’d get different readings in different gears and what not. As it is you get near identical readings to a crank-based power meter, and no discernible deviations irrespective of chain conditions. As above as well, the Rohloff chain still gets just as dirty. Unless you start getting into chainsuck, or the cassette gets so clogged it skips, or the jockeys seize I can’t really see that the Rohloff surpasses a geared system.

Then there’s the weight.

Would things like PowerTaps not work though if there was a significant loss through the system as they measure power output ‘at the road’ as it were? Serious question like…

TBH, if the the drivetrian is maintained and changed when things obviously need replacing then the transmission losses from crank to hub are pretty small – so the effect on a PowerTap reading will be fairly constant and in the noise, but it must be there, even for things like running full chain cross-over ratios etc. Constant, straight, chainlines and the ability to run more robust rings/chains so it all works better for longer are the only advantages hub gears can offer over mechs in getting the power from the ring to the cog via a chain, I think…..

Are belt drives more or less efficient? Does high tension it is under affect it?

Constant, straight, chainlines and the ability to run more robust rings/chains so it all works better for longer are the only advantages hub gears can offer over mechs in getting the power from the ring to the cog via a chain, I think…..

That’s my point though, certain the chainline point seems to be a bit of a red herring – as there aren’t discernible losses in power reading depending on chain line – if I’m climbing in big/big on the road at a certain power/cadence, then change to an equivalent ratio on the inner ring and match my cadence, my power doesn’t noticeably change. I can see the point of more ‘robust’ parts, but surely there’s not a power advantage there, more a longevity one.

Think I’d seen that belt drives were less efficient, could well have made that up though, to my mind you think they’d be more efficient.

ability to run more robust rings/chains

Bigger/Thicker chains/rings = more friction? It’s probably minimal but I’d guess it’s in the same order of magnitude as running a not quite perfect chainline with gears.

It’s the weight Ithink I’d hate the most, I ike that when cranking up a rough climb the way the ligher SS rear wheel seems easier to bounce/hop up over obstacles than it’s geared equivalent, adding more weight than the geared hub would only make that worse.

Odly out on Sunday with gears on the SS bike I did a lot of climbing in one gear rather than shifting, and when I did shift it was to the very end of the cassette to sit down and rest.

I can’t reference it but I have read of Rohloff failures, and returns needed for bearing, oil changes etc.

Just mentioning it because some fanbois above seem to think geared hubs are the holy grail of reliability and no servicing. They may be better than derailleurs in that respect, but then those are the achilles heels of derailleur systems – fairly constant tweaking required, gubbins exposed to mud.

I know what I’d sooner have on my bike – XT.

I think it boils down to it all being the rounding really and the number of variables makes it tough to quatify.

In a similar vein, I percieve my Alfine 11 speed commuter to be very draggy compared to my similar specced geared comutter. I can’t split out if it is the Marathon plus tyres being dead/draggy and weighing a tonne, or the hub itself, but that bike is consistently a couple of minutes slower than the geared bike over the same commute…… and headwinds or steep hills are a complete PITA. It is pretty much zero maintenance though, so you pays yer money…. etc.

I can’t reference it but I have read of Rohloff failures, and returns needed for bearing, oil changes etc.

You want this

They don’t seem to like being completely submerged in water – search back on here for simonfbarnes experiences. That said, I get the impression he replaces the bearings in his Hope hubs every few months so his use isn’t typical. I’ve been running them for over 7 years now (on and off road) and have yet to need bearings replaced on any of the ones we run. Oil change is a simple 15 minute job at home.

The most common ‘failure’ is the freehub springs going missing when people/shops have taken the shifter block off and dropped them.

The biggest problem I have is removing the sprocket as they have a tendency to seize on. Other than that – no chainsuck, very rare to drop the chain (does happen sometimes on sus bike on and the hardtail if you let the chain get too slack), shifting always works reliably.

The couple I use are running better than ever after years of use. Running costs are replacements chains, sprockets and, less often, chainring. Oil is expensive in kits but cheap in bulk (and there are people on eBay splitting the bulk oil into sensible amounts).

I’m looking forward to the Pinion Gearbox coming in Spring. A further weight penalty above Rohloff but promises better weight distribution and even greater range.

I wonder if the draggy/ dead feeling some complain of is more to do with having all the weight in one place? Plus it’s rotating mass.

Depends on the gear. Of the 14, some are draggier than others, and anything in the low-range half of the box is draggier than the high-range. Ergo, gear it so that you’re normally in 8-14 and you’ll be fine, though it might mean that you lose the top-end. Given that I almost never use the big ring on my derailleur bikes, that’s no massive hardship.

If you really want to get to the bottom of this read “Bicycling Science” 3rd edition.

I did.

It can be quite heavy going.

However, unless you’re going up hill, wind resistance is what you want to worry about.

Well I use the Rohloff all through the winter on my off road commute and if I took account of cleaning and maintenance time when calculating average speed then the Rohloff is massively more efficient! 🙂

I don’t think it mentions in that comparison test whether the Rohloff was brand new or had been run in which might well have improved it’s results.

I love my speedhub equipped winter tank but have always felt it significantly harder work ( feels like 10%+)than other derailleur bikes I have . May be the extra weight but I’m sure it’s the friction from the hub and have never believed the 2% claims. My hub has also been back to Germany for some new bearings . Great for winter training and laziness.

…as there aren’t discernible losses in power reading depending on chain line…

From that article;

Derailleurs

Rohloff

Comparison

So, according to that, a Rohloff is more efficient in 11th gear.

Also, it looks like sprocket size has more effect than chain line. Wrapping a chain round a small sprocket wastes energy in friction.
I wonder if all the single speeders running 32×16 know they would be better off with 40×20 ?

I still don’t know the answer to my original question.
I get that idea that a 2% drop in efficiency means you would need to put in 2% more effort to go the same speed.
How much slower does a 2% drop in transmission efficiency make you though ? I would guess it’s less than 2%.

I’d say I more than make up for the 2% loss, on group rides waiting for my mates to fix gammy mechs, dropped chains etc.

🙂

How much slower does a 2% drop in transmission efficiency make you though ? I would guess it’s less than 2%.

Well if you assume all other variables are the same, and 98% efficiency – as that appears to show then it’ll be 1.96% slower.

Well I use the Rohloff all through the winter on my off road commute and if I took account of cleaning and maintenance time when calculating average speed then the Rohloff is massively more efficient!

Get that with a belt drive, but when I ride my MTB in winter the part that gets trashed and requires attention is the chain. The rest of the transmission is absolutely fine under a layer of grime.

I know that rohloff research will have been rigorous, but it doesnt seem to explain the 6/7 gear draggy whirring sound does it? I thought it was in those gears that the planetary gears were all engaged at the furtherest point from each other? I consciously try and avoid them if I can.

Im fully prepared to forego some efficiency against deraillieur just for the simplicity of a single chain ring/tensioner set up just as I have on my SS. I swap my frame between geared and SS in just 10-15 minutes because the form is so similar.

But nothing feels as efficient, sublime, silent, and taught as a perfectly lubed singlspeed.

ir_bandito – Member
I’d say I more than make up for the 2% loss, on group rides waiting for my mates to fix gammy mechs, dropped chains etc.

I used to ride with a guy who had a rolhoff, we were forever waiting for him to fix yet another pinch flat caused by his winter proof bike (1.8 tyres and rolhoff) coming unstuck at the first sign of anything lumpy.

So, according to that, a Rohloff is more efficient in 11th gear.

because 11th gear is 1:1 and there’s no “gearing” effect.

I wonder if all the single speeders running 32×16 know they would be better off with 40×20

was already considering 34×17 for a little more longevity, possibly 36×18 but bigger and you’ll ruin clearence, got a few local logs/steps etc that only just clear 32. Besides 32 rings are the norm and easier/cheaper to get hold of than other sizes.

Besides 32 rings are the norm and easier/cheaper to get hold of than other sizes.

I run my 29ers as 36:20 so that I can go to 36:16 rohloff without mucking about with chainwheels. I have bucket loads of 36T SS rings 😉

Well if you assume all other variables are the same, and 98% efficiency – as that appears to show then it’ll be 1.96% slower.

I’m not convinced it’s a linear scale like that.

Say, for example, it takes 200w to cycle at 20km/h.
Does it therefore take 190w to cycle at 19km/h, 50w to cycle at 5km/h and so on ?
Or, to put it another way, if you lose 2% of that 200w so you’re down to 196w, would you really cycle at 19.6km/h for the same effort ?

There’s a chart showing how many gear sets are engaged in each gear on the top right of this page.
http://www.hupi.org/HParchive/PDF/hp55/hp55p11-15.pdf
3rd and 5th are theoretically the worst.

Stoner, I run 32×18 SS and 32×13 Rohloff for the same reason.

I guess it must be gear No. 5 Im talking about. Cant say I ever look to check – I just keep twisting the knob one oway or the other until Im either in the right gear, or Ive run out of them 😉

Cant help but feel there’s more than a 95% to 99% spread of efficiency across the gears. Although I suppose if 1:1 is 99% efficient, introducing a planetary gear takes another 1.5%, then introducing 3x planetary sets for the loss of only 4% (from 99% “perfection”) seems proportional.

Yeah, that’s one of the things I like about the Rohloff, it’s a simple twist one way to change up, the other way to change down until it won’t go any further.
Maybe it’s because I’m not used to derailleurs that I find myself constantly thinking about which gear I might want next and how I’m going to get it, “Up one on the front, down three on the back”.

MTG – I’ve gone 1×9 on my derailleur bike because after 4 years of using the Rohloff exclusively the bit of my brain that could deal with making gear changes using both hands no longer works. I bet you also find it deeply frustrating when you realise you can’t change when stationary.

Cant say I ever look to check

Indeed. On road, with a normal 38/16 set up, I’m rarely below 7 so spend time in the upper gears. Off road tends to be dynamic enough that you just use the gear you need – it being the *right* gear for that moment outweighs any thoughts about which gear it is.

32 rings are the norm and easier/cheaper to get hold of than other sizes

Unramped DH rings easily available 36-40t.

And don’t forget Rohloff customer service.

Mine had developed a bit of sideways play. Enough that if it was Shimano, you’d just nip up the cones. But you can’t do that.
So I took it to my LBS, they sent it to Ison, they sent it to Rohloff. Its been stripped and rebuilt with new seals bolts etc, and just made its way back to the LBS.

Bear in mind I bought it second hand off this site a few years ago, its cost me absolutly nothing to get sorted. Nice.

A couple more thoughts on this…

Energy = mass x velocity squared.
A 2% increase in energy will only = a 1.41% increase in speed.
I still don’t think that a 2% increase in drivetrain efficiency is equal to 2% increase in power at the wheel though.

Something I just read on the latest 29er thread got me thinking;
A 29er wheel is approximately 10% bigger than a 26er wheel.
As it therefore rotates 10% fewer times for a given distance, does that mean it effectively makes the transmission 10% more efficient ?

MidlandTrailquestsGraham – Member
…A 29er wheel is approximately 10% bigger than a 26er wheel.
As it therefore rotates 10% fewer times for a given distance, does that mean it effectively makes the transmission 10% more efficient ?

Don’t know what the maths might tell us, but my feeble legs tell me that’s about right.

You’d get 10% less transmission loss, so instead of ~2% loss you’d get ~1.8% loss. But that’s less than the difference in losses between the various gearing combinations.

njee20 – Member
when I ride my MTB in winter the part that gets trashed and requires attention is the chain

But what about any scientific analysis of chain wear/friction? It boils my piss when I’m riding in muddy conditions and all I can hear is the transmission – but is it actually running any less efficiently? It’s too easy to go on irrational experience like this rather than proper data.

IIRC there’s a study where un-lubricated (and I mean thoroughly degreased) chains ran no less efficiently than clean lubed chains.

Gasman Jim – Member
If you really want to get to the bottom of this read “Bicycling Science” 3rd edition.

I did.

It can be quite heavy going.

However, unless you’re going up hill, wind resistance is what you want to worry about.

For road bikes…I agree, but off-road? Wind resistance is a much lower factor, and others may be significant.

epicyclo – Member

Don’t know what the maths might tell us, but my feeble legs tell me that’s about right.

It won’t be any feebleness in your legs…