Why *average* efficiency is what counts in derailleur v HGB comparisons

[Andre Jute]

[An article I posted to another forum where the slightly aggressive tone is necessary to break through the prejudices of the resident luddites.]

A figure currently thrown about on with abandon on [another bicycle forum] is that bicycle chains and sprockets are 98% efficient. That isn't untrue in certain very limited circumstance, but the minute you ride that bike to which those limited circumstance apply beyond the bottom of your drive, the efficiency falls and undermines the 98% assumption.

The 98% assumption *requires* a single-gear or fixie bike with a dead straight chainline, sprockets and chain rubbed in to each other (not new, off the shelf) but not worn beyond the point of lowest friction either, and all parts to be dead clean and lubricated to perfection.

Any cyclist with his brains in gear knows this case doesn't perpetuate itself once you're out of the garage. The lube attracts filth to itself and the efficiency heads south.

***

So you have to cover that chain, not a perfect answer but one that with modern materials and design is possible without a huge further downgrading of efficiency by either additional weight or friction. I have in mind here the Hebie Chainglider which just about totally encloses your chain. I have about 30K miles of experience with various chain covers in a deliberate longterm experiment, and the only one I will recommend is the Chainglider. Anything else lets in dirt, and there goes the efficiency.

But you can't fit a good enclosure about a derailleur transmission. Oops.

***

If you're going to be forced to operate a single-gear bike anyway to preserve average efficiency, you may as well use a hub gearbox, which to the gestalt of the bike appears like a single-gear system, and works with the virtually no-cost (in efficiency terms) Chainglider.

***

A little, a very little thought* will persuade you that the efficiency of an open-chain single-speed bike, or a derailleur bike, inevitably goes downhill from the moment it is first used, while the hub gearbox efficiency increases as it is run-in and then remains largely stable for most of its lifespan (in a Rohloff unknown, but assumed by reference to the experience of world circumnavigators to be north of 200,000km, in a Shimano 8sp Nexus Premium said to be 50,000km).

On average for any ride or for its total life, a hub gearbox, especially if the chain is suitably covered, is likely to be more efficient than a derailleur system or a fixie.

I will bet that my Rohloff hub gearbox equipped bike (perfect chainline, chain running for its entire life on the factory lube, stainless steel sprockets, chain totally enclosed in Chainglider) in your garage is as efficient as your brand-new derailleur bike, and five miles from your house is more efficient than your brand-new single gear/fixie as well.

It is the gradual recognition of these facts that drive more and more road-cyclists to fit hub gearboxes to their winter training bikes.

Copyright © 2019 Andre Jute
*A little, a very little though will suffice -- John Maynard Keynes

[martinf]

Agree to some extent, but my own experience tells me that a well-maintained derailleur bike is about as efficient when riding as a similar bike with a decent internal hub gear, not less, also not significantly more.

I consider the Rohloff, the Nexus Premium version of the Shimano 8-speed, and the old oil-lubed Sturmey Archer S5/2 to be decent internal hub gears.

Mechanical efficiency of a Shimano Nexus 7 (plain bearings in the planet gears rather than the needle bearings in the Nexus Premium and Alfine versions) seemed less than other systems, so some hub gears are best avoided. I found that the mechanically very efficient Sturmey Archer and SRAM 3-speeds, which I have used a lot, were in most cases less efficient when riding than other gear systems due either to the large jumps between gears not allowing an optimised pedalling cadence or the need to walk up very steep hills. The same goes for single-speed or fixed gear bikes, which aren't really suitable where I live.

For me, the big difference between derailleurs and decent hub gears is the time saved on routine maintenance. Even without a chaincase, a hub gear system has fewer exposed parts that need to be cleaned, so the job is quicker and easier, and with no derailleur dangling down near the road it also picks up slightly less muck. Before getting Chaingliders for most of my bikes I reckoned on cleaning the transmission on my derailleur bikes about every 300 to 500 km, more like every 800 km on the hub gear bikes. With all the gubbins on a derailleur bike, the job also took me about twice the time. So overall the derailleur system was less efficient for me when the total time of riding plus maintenance was counted.

With a Chainglider chaincase, routine maintenance is reduced still further, but is still necessary for a bike used for off-road riding and commuting in extreme weather conditions. The Chainglider doesn't stop water from getting on the transmission during heavy rain and, although it is pretty good at keeping muck off the chain it isn't completely sealed to dust/fine sand/mud.

For a bike used only on dry tarmac roads, I don't think there is any real advantage in going for a hub gear over a derailleur system. This changes if the bike is used in all weathers and/or on tracks and paths, especially the sandy coastal paths I often use.

[PH]

Reading with interest, though it’s hard to take my experience and separate the fact from the perception.  There will be rides where I feel the Rohloff bike is dragging like an anchor and others where I’m sure whichever bike I’m on has grown wings, then I download the data and the differences are minimal or nonexistent.
My thoughts mainly are that I’m riding a bike not measuring a drivetrain.
There are times on hilly rides where I know I’d be faster on the derailleur bike, but it’s hard to work out the gains from always being able to change gear.  I do ride different depending which it is, as an example when I approach a junction I may need to stop at, I’ll change down to a pulling away gear on the der bike but not on a Rohloff.  Or on a short decent I might not change up a chainring knowing I’ll soon have to change down again, yet will change up the hub. You can’t measure that efficiency.

[Andre Jute]

...then I download the data and the differences are minimal or nonexistent.
My thoughts mainly are that I’m riding a bike not measuring a drivetrain.

The difference between a well-maintained, clean, derailleur and the better hub gears is anyway too marginal, in my almost exclusively tarmac experience anyway (see Martin's post above), to be easily separated from one's prejudices.

I did once have a revelatory experience. When I lived in Millbrook House, on the Bridewell River, I rode every afternoon along exactly the same route. The bike was a Gazelle Toulouse, described by the Dutch as a "stadssportief" (a sporting commuter) or a "vakansiefiets" (a holiday bike, because only the naive would leave a bike like that chained up in front of the station all day). It had a Nexus Premium box. Then I bought a Trek Smover, developed by the Benelux division of Trek for exactly the same "vakansiefiets" market. Same Nexus box, only difference was that it had fully automatic gears and electronic adaptive suspension. Same ride every afternoon. I also had a heart rate monitor cum bike computer with download facilities to my Mac and a pattern soon emerged: the automatic bike was noticeably faster because it changed gears more optimally in response to the terrain and my input.

But what I had in mind in the original post is not that large and obvious a difference, more the sort of difference that is measured in a laboratory under ideal conditions that have nothing to do with real life. There a derailleur wins by a small margin over the hub gearbox. But the moment you take them on the road into real life the hub gearbox, especially with an enclosed chain, comes close to maintaining the laboratory finding while the derailleur bike drifts further and further away the longer it is ridden from its pristine starting condition.