Posted in

Bike Chain Efficiency Testing: Wax vs Oils, Old vs New Chains, Singlespeed vs Gears, OSPW

Bike Chain Efficiency Testing: Wax vs Oils, Old vs New Chains, Singlespeed vs Gears, OSPW

Bicycle drivetrains come in many different configurations, and as a result, bike chain efficiency – the percentage of power that actually makes it from your pedals to the rear wheel – can vary quite a bit.

Fortunately, the cycling community has a strong appetite for testing the technical side of things, always looking for ways to squeeze out a little more performance. And then there are people like me, who enjoy digging into the data and sharing what it all means.

In this article, we’ll take a closer look at 18 different drivetrain setups and how they compare in terms of efficiency. Let’s dive in!

Russell Bridge’s Recumbent Test Rig

Russell Bridge designed and built a custom test rig to evaluate a wide range of recumbent drivetrain configurations. This allowed him to compare setups such as rear-wheel drive (RWD) vs front-wheel drive (FWD), old vs new chains, waxed vs wet-lubed chains, small vs large idlers, different idler positions, chain tubes, oversized pulley wheels, and more.

For each configuration, Russell pedalled on the rig for two minutes at a steady 200 watts. To calculate drivetrain efficiency, he used two power meters – one at the pedals and another at the hub. By comparing the power input at the pedals with the power output at the hub, he could determine how much energy was lost through the drivetrain.

For example: if the pedals show 200 watts and the hub measures 194 watts, the drivetrain efficiency is about 97% ((194 ÷ 200) × 100).

The results obviously depend on the accuracy of the power meters, namely the Garmin Rally RS200 pedals and the Powertap G3 hub, both of which are typically accurate to within ±2%. At 200 watts, that margin equates to several watts of potential discrepancy.

To improve reliability, Russell repeated some tests multiple times and averaged the results. Even so, it’s worth noting that this isn’t a controlled lab test with highly precise instrumentation, so small differences should be interpreted with caution.

You can see Russell’s full test HERE.

Singlespeed Drivetrain Efficiency

Singlespeed with Tensioner (97.0%)
It’s no surprise that a singlespeed setup proved highly efficient. With only a single jockey wheel in the tensioner, drivetrain efficiency measured 97%.

Singlespeed without Tensioner (97.5%)
Removing the tensioner delivered a small efficiency gain, with Russell extracting an extra 0.5% from the drivetrain. This setup provides a useful benchmark for the derailleur drivetrains that follow.

Derailleur Drivetrain Efficiency

Derailleur with Old, Worn Chain (90.6% Efficiency)
The least efficient derailleur setup came from using an old, worn chain, dropping drivetrain efficiency to around 90%. As the chain wears, it no longer meshes precisely with the chainring and cassette teeth, increasing losses.

Derailleur with Smaller 48-tooth Chainring (95.2% Efficiency)
Switching to a new chain brought efficiency back up above 95%. With a 48-tooth chainring, efficiency sat around 95%, but the tighter curvature increases chain articulation, which leads to slightly higher energy losses.

Derailleur with Standard 61-tooth Chainring (95.0 to 95.8% Efficiency)
With the 61-tooth chainring back on for the rest of the tests, measurements across two cassette sprockets showed efficiency ranging between 95% and 96%. A larger chainring essentially improves efficiency by reducing how much the chain has to bend.

Derailleur with Slack Chain (95.5% Efficiency)
Running a slightly longer (slacker) chain appeared to offer a small efficiency gain, though the difference falls within the margin of error, so it’s not a definitive result.

Derailleur with Waxed Chain (96.1% Efficiency)
Waxing the chain delivered a modest improvement, pushing efficiency just above 96%. While the watt savings are small, the bigger advantage is reduced wear on chains, cassettes, and chainrings. I have an article on the best chain waxes HERE.

Derailleur with Upper Idler (96.5% Efficiency)
Adding an upper idler appeared to increase drivetrain efficiency compared to a standard setup. However, this is likely an anomaly. Idlers typically introduce extra friction through both the bearing and added chain articulation. As Russell only ran a single test in this configuration, the result should be taken with caution.

Derailleur with An Oversized Rear Pulley Wheel (97.5% Efficiency)
The most efficient derailleur setup used an oversized rear pulley wheel with a ceramic bearing, reaching 97.5%. This makes sense mechanically, as the larger pulley reduces chain articulation while the low-friction bearing minimises losses. The trade-offs are typically reduced shifting performance (or more finicky setup) and shorter pulley bearing lifespan.

Recumbent FWD Drivetrain Efficiency

Derailleur With Small Idler (88.1%)
The least efficient FWD recumbent setup used a small idler between the derailleur and chainring, resulting in roughly 88% efficiency. The tight radius forces the chain to bend more sharply, increasing articulation losses.

Derailleur With Two Chains (91.5%)
Efficiency improved notably when a dual-chain system was used. Here, a standard derailleur drivetrain was linked to a second singlespeed chain driving the front wheel. Despite the added complexity, overall efficiency remained relatively high.

Derailleur With Large Idler (94.0%)
Using a large-diameter idler between the chainring and cassette brought efficiency up to around 94%. The 6% gain over the small idler highlights how critical idler size is in reducing chain articulation losses.

Derailleur With Large and Small Idlers (94.0%)
This configuration paired a large idler with a smaller one on the return side. Interestingly, adding the second idler had little impact on efficiency, as the smaller pulley handled the lower-tension return chain. This result is still higher than you’d be expect, as the small idler undoubtedly increases friction compared to not running it.

Recumbent RWD Drivetrain Efficiency

Chain Over Dual Idlers, Old Worn Chain (88.0%)
The least efficient RWD recumbent setup combined a worn chain with two idlers – a large lower idler and a small upper idler on the return side – resulting in about 88% efficiency. As expected, the worn chain significantly increased drivetrain losses.

Chain Over Large-Diameter Upper Idler, Return Chain In Tube (93.0%)
Efficiency improved to around 93% by using a large upper idler while routing the return chain through a tube (not shown in picture). This setup reduces contamination and keeps the chainline tidy.

Chain Over Large-Diameter Lower Idler, Return Chain Over Small-Diameter Idler (94.3%)
Using a new chain with two idlers – a large one for the power side and a small one for the return – pushed efficiency up to 94.3%. This is over 6% higher than the same configuration with an old chain! This brings the system close to the efficiency of a typical derailleur drivetrain.

Chain Over Large-Diameter Idler, Return Chain In Tube (94.7%)
The most efficient recumbent configuration used a large lower idler for chain tension, with the return chain running inside a tube. At 94.7%, this setup is likely within 1 to 2% of a standard derailleur drivetrain.

Summary

There we have it. A single test rig, a wide range of drivetrain configurations, and some insightful efficiency data!

The results highlight just how important it is to minimise chain wear. Run a chain for too long, and you could be giving up around 6% of your pedalling power!

Oversized pulley wheels (OSPW) and chain waxing do offer measurable gains, but they’re clearly marginal. Nice to have, but far less impactful than keeping your drivetrain clean, well-maintained, and within its wear limits.

Finally, and only if applicable to your bike, the data also reinforces the value of using large-diameter idlers on the loaded (power) side of the chain to minimise chain articulation, while idler size is less critical on the return side.

Leave a Reply

Your email address will not be published. Required fields are marked *