Drivetrain efficiency is an interesting and often overlooked opportunity for going faster. A new Shimano Ultegra groupset is said to cost approximately 20W when pedalled at 250W (90rpm)* — in other words, your rear tyre receives 92% of what you put into the pedals. And scarily, that number only gets worse as your power output increases.
Fast chain lube and a clean chain can make a measurable difference here, while other upgrades, such as different bearing seals and lubricant can have a more marginal impact, too. 100% efficiency is impossible, but narrowing the gap is not.
Retailing from US$500/AU$650 for what’s effectively a carbon fibre derailleur cage and ceramic bearing-equipped aluminium pulley wheels, CeramicSpeed’s OSPW (OverSized Pulley Wheel) system is immediately one of the most polarising speed pursuit upgrades. Many assume it’s the well-marketed ceramic pulley wheel bearings (fitted with light contact seals and fast lubricant) that provide the claimed 2-4W savings, and while they play a role, the real benefits are in the oversized pulley wheels.
The data suggest the OSPW will indeed save you watts, although that speed is anything but free. But does it also impact shifting and drivetrain reliability? Is it noisier? And if it’s so good, why don’t we see more pros racing with it?
CeramicSpeed claims the OSPW can save you between 2-4W over a stock Shimano Dura-Ace derailleur cage system (at 250W, 90rpm), and that figure progressively improves the worse your chain maintenance and selection of chain lubricant is (or as it wears off and gets contaminated toward the end of a race). In reality, with an optimised or even well-maintained chain in use, that figure is likely closer to just 1-2W – but still, watts are watts.
Larger cogs are more efficient because of reduced chain articulation. As compared to the SRAM Red22 short cage on the left, each chain link articulates far less when it’s running through the CeramicSpeed OSPW.
Just as we know using larger cogs and chainrings are more efficient as a result of reduced chain link articulation, the same applies to derailleur pulley wheels. With regular 11-tooth pulley wheels, each link of a chain must articulate about 33 degrees as it enters and exits each pulley wheel. But with a 19-tooth pulley wheel (the largest available in any OSPW), each chain link only needs to articulate 19 degrees.
While it’s the most commonly seen, the most expensive, and often considered the best, the CeramicSpeed OSPW is hardly the only oversized pulley wheel system. German company Berner was the first to offer an aftermarket derailleur cage upgrade focussed on increasing the pulley wheel tooth count. Other brands, including CeramicSpeed and a number of lower-cost imitations, have since followed.
Formerly the world’s independent voice for all matters of drivetrain efficiency, Jason Smith is now CeramicSpeed’s technical officer. Still, much of his earlier test findings from his Friction Facts days remain. Back in 2013, FrictionFacts measured a 0.49W efficiency difference between a 10T-10T pulley combination and a 15T-15T pulley combination, with all other variables held constant. Reducing derailleur cage tension then offers similar benefits, and CeramicSpeed’s OSPW makes it easy to select between three spring tensions.
In addition, a larger pulley wheel spins slower for a given cadence, and in this case, each pulley also runs on premium hybrid ceramic bearings with light-contact seals and a low-friction oil. It’s easy to see how this product can improve drivetrain efficiency.
For CeramicSpeed, each OSPW is unique to the derailleur model it’s designed to be fitted to, and there’s a great variety in the exact pulley wheel tooth counts used. For example, the SRAM mechanical OSPW kit I tested here features dual 17T pulley wheels, whereas the original setup used 11T ones. Some other cages, such as for Campagnolo or certain Shimano models, feature a smaller 13T top pulley combined with a larger 19T pulley.
The exact configuration depends on derailleur geometry and just how much CeramicSpeed believes the limits can be pushed before shift performance suffers.
The CeramicSpeed OSPW for Shimano R9100/R8000-series derailleurs features a 13T upper pulley, and an enormous 19T lower.
So exactly how many watts will an OSPW save you? Tooth counts aside, not all stock derailleur pulleys are created equal — different derailleurs run different tensions, and even tooth profile matters, too. There are simply too many variables at play to give an exact figure, but the data safely point toward it being at least a watt.
The process for installing an OSPW varies based on the derailleur model, and each kit includes clear instructions. Following these instructions makes it a medium-difficulty install, with elements such as fitting a slightly longer chain and ensuring the derailleur hanger is 100% straight adding to the complexity.
Installing the OSPW involves removing the stock cage assembly. By using a removable tension stop, CeramicSpeed has made it an easier process than installing the stock cage.
With a longer cage and bigger wheels, an OSPW-equipped derailleur is certainly more sensitive to adjustment and alignment. But once set up, the shifting is surprisingly good, and in the case of my Red22 drivetrain, remained suitably crisp.
I was surprised to feel play between the cage and the derailleur’s B-knuckle. This is relatively common according to other mechanics, but the consensus is that it doesn’t seem to impact shifting or performance. It’s commonly agreed that the play is worse with the simple pin retention on SRAM derailleurs, and the latest OSPW models for Shimano are far more solid – something I can attest to having recently received a test bike fitted with a Shimano R9100/R8000 version of the OSPW.
The OSPW offers three tension settings. Once installed, you’ll need to remove the cage in order to change it.
The higher the spring tension, the snappier the shifting, and the better the chain retention. CeramicSpeed recommends running the lowest of three spring settings you can to get the most efficiency, but general road riding/racing will likely benefit from a higher tension. I ended up settling on the highest (and highest-friction) spring tension, simply because the chain slap was otherwise too great on my poorly surfaced local roads. The high setting is roughly equal to what the stock cage was set on.
Weight-wise, the OSPW sample for SRAM mechanical added 22g to my derailleur: the original Red22 short cage is 49g, and the CeramicSpeed OSPW is 71g. Interestingly, the OSPW I fitted effectively turned my derailleur into a long cage version, and CeramicSpeed suggests that up to a 32T can be fitted, much like SRAM’s own medium cage Red22 rear derailleur which is also 20g heavier than the regular short cage version (28T capacity).
The Dura-Ace R9150 Di2 derailleur fitted with its OSPW weighed 230g, approximately 30g heavier than a stock derailleur, but like the SRAM version, the fitting of the OSPW grows the lowest gear allowance to 32T (30T stock).
Out on the road, I quickly forgot that my bike was carrying an extra $500 on one of its more easily damaged components. And despite the play from where the cage connects and the increased flex from the extended cage, my shifting remained reliable and crisp. The same can be said for the Dura-Ace R9150 Di2 version I’ve recently begun using, with the robotic shifting kept quick and precise.
I may have forgotten it was installed, but others looking closely at my bike certainly hadn’t. There is no missing this modification, and the aesthetics are questionable at best.
The large aluminium pulley wheels are noisier than small plastic pulleys, but not so noisy to be a nuisance over the whirr of your tyres or the hum of nearby traffic. The only time I noticed the additional noise was on the indoor trainer.
I’d be lying if I said I could feel the difference the OSPW provided. To be fair, my Red22 derailleur was already fitted with SRAM’s ceramic bearing pulley wheels, and they were worn in and well-kept. My bottom bracket was already set up with high-end bearings and obsessively preloaded for minimal resistance. Likewise, my chain is treated with Molten Speed Wax – one of the fastest options.
As a result my bike is one where an OSPW may only save a watt, if not less. Still, there’s something extremely satisfying about the smoothness provided through a sum of efficient parts – and the OSPW was certainly a contributor toward that.
The OSPW is pricey, but with the right maintenance, it could outlast the derailleur it’s fitted to. The metal pulley wheels should wear more slowly than resin versions, more because their larger size spreads the load further, rather than because of the materials used. And CeramicSpeed’s ceramic bearings are famously durable, too.
“In general, with the [regular] CeramicSpeed bearings, you can bring them back to life as the races and bearings don’t get surface rust on them,” said Zach Edwards of the Boulder Gruppetto workshop, in Boulder, Colorado. “I travel with a guy during the cyclocross season, and last year we had CeramicSpeed bottom bracket and hub bearings, and they lasted a whole season of back-to-back-to-back muddy weekends with lots of power washing. You can just clean and flush out the bearings and re-grease to keep them going even after they have seized.”
In case you were still concerned about the durability – and if the OSPW wasn’t expensive enough – then for an additional $100, CeramicSpeed has a Coated version, which features a harder and more corrosion-resistant coating on the bearing races, and is backed by a six-year guarantee from CeramicSpeed.
A small bottle of CeramicSpeed oil is provided with the system. It’s simply applied through the ports on the backside of the cage. How often? Well, that depends on the conditions you ride in.
Out of the box, the OSPW bearings run on light oil, and CeramicSpeed provides a small bottle of lubricant to keep the bearings running smooth. The cage has small lube ports built-in, although removing the pulleys to access the bearings and seals is better. For extended durability or more care-free use in poor conditions, Edwards recommends doing a half fill with CeramicSpeed grease instead of the supplied oil.
One argument I’ve seen is that the watts an OSPW would save over an already optimised drivetrain are negated by increased aero drag, and not surprisingly, that argument is disputed by CeramicSpeed. But nevertheless, it’s a question I posed to Silca’s Josh Poertner, a former technical director at Zipp and current technical consultant to a number of top WorldTour teams. Poertner said that although he hasn’t specifically tested the OSPW in a wind tunnel, it’s also never come up as a concern, and that whatever drag it may create, it’s likely within the tolerance of the tunnel anyhow.
Poertner agrees that CeramicSpeed’s claims for an OSPW saving 2-4W are overstated and that the figure is indeed closer to 1-2W. However, in the pursuit of marginal gains, Poertner did say it’s a product he’s recommended to teams, at least for use in time trials.
Due to its obvious presence, pro teams with drivetrain sponsorship agreements aren’t allowed to ride a product like the OSPW.
Speaking off-record, a mechanic from one of the top WorldTour teams said they had also tested the CeramicSpeed OSPW and found it offered real benefits. I was told the team would choose to use the system if it didn’t directly conflict with a far more valuable drivetrain sponsorship agreement.
And right there is likely the reason for why we don’t see more drivetrain modifications in professional cycling: sponsorship conflict. Step into other disciplines where lucrative drivetrain contracts are rarer, such as Ironman, and OSPWs and similar modifications are a common sight.
Jason Smith, CeramicSpeed’s technical officer, suggested that a few of the top teams are likely using hidden drivetrain drag tricks, at least for important races and stages. And while he suggests that an OSPW is the second best upgrade you can do to reduce drivetrain drag (after chain preparation), it’s also one of the few mods that’s impossible to hide.
All of this begs the question: why do the likes of Shimano, SRAM, and Campagnolo stick with small pulley wheels when the data suggest they’re not the best way?
“It’s hard to say with certainty,” said Smith. “It could be a corporate belief/philosophy that drivetrain efficiency is not a priority. It could be because drivetrain efficiency is not considered by the ‘Big Three’ as a mass-market requirement yet. Or perhaps, it’s simply because the Big Three, while industry leaders in general, are not industry leaders when it comes to the finer aspects of efficiency.
“Arguably, if efficiency was a priority, they’d be manufacturing oversized pulley wheel systems. We can say with certainty and based on data, that it’s not because smaller pulley wheels are more efficient.
“We must also not rule out the depth [to] which standard-sized pulley wheels hold cross-compatibility for many derailleur models. Manufacturing equipment, design, and compatibility should be considered, too.”
I reached out to Shimano’s R&D department for insight into the matter and simply got a “no comment”.
Given that you can find an entire Dura-Ace Di2 rear derailleur (or an entire Shimano Tiagra groupset!) for the cost of this pulley cage assembly, one does have to question its value for money. I’m not going to beat around the bush: this product is overpriced for what you get.
Yes, CeramicSpeed has indisputably invested more into research and competitive testing than any other company in the space. Yes, quality and tolerances are great. Yes, the OSPW features top-class ceramic bearings and a quality fibre-reinforced composite cage. But you can’t ignore that price.
The price becomes a little more palatable when comparing the benefits it offers over more common (and also expensive) ceramic bearing pulley wheel upgrades, but even the price of those is something most will (and should) choke on. The decision becomes tougher again when you consider the option for larger pulley wheels that fit stock derailleur cages. For example, Kogel now offers 12/14T pulleys to fit directly into the latest 11-speed Shimano road rear derailleurs, and the company even has 14/14T pulleys to fit into 12-speed mountain bike derailleurs. Such an upgrade will cost you US$150, or less than a third of an OSPW.
Still, if you’re in the pursuit of marginal gains and spending $500 to save a couple of watts seems like the next step, then the OSPW is exactly for you. It should provide better efficiency gains than a bottom bracket or hub bearing upgrade, and those benefits only amplify the harder you pedal and the dirtier your chain becomes.
A bike fit, coaching, faster tyres (and/or latex tubes), aero wheels, an aero helmet, and even an optimised chain will all provide a vastly superior return on watts-saved-per-dollar when compared to a product like the OSPW. But after you’ve done all that, and you’re at a point where you can obsess over the one-percenters, then go for it. An OSPW will certainly make more of a difference than the titanium bolt kit and carbon bidon cages you’re eyeing, and the benefits will transfer to your Zwift efforts, too.
Another view of the CeramicSpeed OSPW for newer Shimano 11-speed road rear derailleurs. Here it’s fitted to a R9150 Di2 rear derailleur, and shifting is kept buttery smooth.
Pictured is a version specific to SRAM Mechanical. The way it attaches, the geometry, and even the tooth count vary depending on the derailleur brand and model it’s designed for.
I’m not denying the actual benefits CeramicSpeed’s OSPW can offer, but those small gains certainly come at a significant price.
15mm Door Window Rollers Wheels
Our mission is to bring you the best content in the world that showcases the beauty of cycling in all its glory. We want to inspire people to love the sport as much as we do, and ultimately pull you away from the computer to get out for a ride.
From exclusive content to premium benefits, join us and experience all we have to offer. Or feel free to continue reading the article.
Kactus Sport, Carbon Fiber, Bike Guide Wheel System, Quick Release - Kactus,https://www.kactussport.com/