Chain Power Loss According to Muc-Off

Just reading about Muc Off’s nano-tube lube chain that costs several thousand dollars. That’s interesting but absurd…even more interesting is some of the data they let slip about chain loss.

This is data taken from Dura Ace 11sp chains. Just looking at some old data I have & that’s a pretty good chain. Definitely a top performer in terms of loss. One got nano-tube lube…the other got UFO drip…so two top lubes.

The thing I note is that chain loss can be a LOT. 14 watts after a little under 4 hours of use. That’s for a good chain, a good chainline, a good lube, good weather.

So next time your pedal based power meter is reading 10 watts higher than your hub-based trainer…think about this data. This is just the chain, too! No derailleur pulley.

By the way…can we say that UFO drip is a crit lube? :wink: I think so…

Not saying you aren’t raising some valid points but you need to consider the source of your data here

The company that sells fully optimized chains telling you how great their product is - this isn’t a surprise

Take these things with a grain of salt unless they are coming from an impartial source


Which part of the data do you find suspicious? It all looks pretty consistent with other data sets I have for chain wear…given that these are Dura Ace chains. If they were Campy chains…yikes…we’d have to turn our heads away from the screen.

I’m not going to get sucked into advocating for Muc Off chains…I think I’ve already said they are absurd…but I would certainly entertain a discussion about what a reasonable envelope for chain loss might be.

The thing is, you kind of are advocating for them, or at least their methodology.

This shows you think their research of their product and their competitors product is valid and legitimate.

Take a look at some impartial testing of chain lubes and you don’t see this dramatic of changes in this amount of time, nor any reasonable amount of time

Which part of the data do you find suspicious? It all looks pretty consistent with other data sets I have for chain wear…given that these are Dura Ace chains.

I like how you refute the data he provided but don’t provide any of your own. Could you share some of the “impartial testing of chain lubes” you seem to have lying around?

I agree that the data that manufacturers provide is curated to make their arguments look stronger. That said, it’s hard to find unbiased studies done on drive train loss. It seems the only folks who really have the incentive to pay money for these tests are the components manufacturers themselves.

I think the rate of degradation is suspicious particularly since they compare directly to a single competitor. I also think proprietary blends such as this that aren’t available for general consumer testing are always a bit nebulous - think power meters that haven’t been made available for testing by DCRainmaker and GPLama. They aren’t necessarily bad, it’s just hard to trust the data without an impartial review


You may have fallen prey to the clickbait headline. While the “product development process” may have cost thousands, the chain is $145 or you can get the lube yourself for $65.
Additionally, the watts of loss on their charts of the 2 treated chains also is presented only in relative terms to another product, not in absolute terms. It’s possible that their “14 watts of loss” at some 3+ hour point in time could be measured at a driven power of 2,000 watts. Sure, I’d love to gain 12-14 watts for $145, but from this data you don’t know if they’re presenting a (14/2000=) 0.7% increase in power to the cassette or a (14/100=) 14% increase in power to the cassette.


Going back to '98, I’ve used SRM, Quarq, Powertap – sticking with the Powertaps since 2011. I’ve never seen a significant power gain from a new or clean chain.

I have both a powertap GS and powertap P1 on my Domane and I see about a 10% difference in power between the pedals and hub. I’m not falling on either side of this argument. Just saying there is some power loss, which I believe we all agree. Question is if this loss through chain can be improved upon enough to matter. I do believe at least half of the 10% loss does come from some frame flex on my domane.

Yeah. That’s the crux of it, really. I’ll offer three more points…one that supports this data set but is anecdotal and two that maybe cast a little shade on the muc off data…

First, when I brought up UFO drip on another thread the anecdotal experiences that folks shared seem to (predominately) support what this data show: UFO drip doesn’t have much staying power. So I guess I was a little pre-disposed to accept what Muc Off is trying to conclude about UFO drip. Further supporting the notion that UFO Drip doesn’t have much staying power is Jason’s own advice that you should re-apply every 120 miles. (Note that Muc Off simulated a 112 mile ride!)

Second, remember Jason’s data for chain wear? He just took a campy, dura ace, and sram chain out of the bag, slapped it on the tester, and let it run for several hours. Here is what that data look like:


After Jason went to work for CeramicSpeed he did a similar experiment with stock chains of various manufacture vs same chains UFO prepped (UFO prep is more than just UFO drip…looks like the chains are dusted with PTFE as well). Here is what that data look like:

My main point here is, that’s a lot of data from a lot of chains treated a lot of ways & none of those curves look anything like what Muc Off showed for the UFO drip chain. So, what’s up, Muc Off? Jason seems to have repeatable data. Is your data repeatable, Muc Off?

Final point…notice the conditions under which Muc Off claim to have tested: “a 112 mile ironman stage” simulation completed in 4 hours. ! Ummm…that’s like, record time, right? Why pick such an extreme simulation point? I suspect they picked it because, although it is at the extreme, they can still say it’s possible…but at the same time it sort of maximized chain tension which maximized chain wear for a thinner lube like UFO drip. So the power difference at the end of four hours is well overstated compared to a more mundane power output over that same time period.

But that’s neither here nor there to the main point I took away from this data set…which is that there can be substantial power loss in a drivetrain JUST FROM THE CHAIN. What Andy Starykowicz can perpetrate on a chain in 4 hours I can also do…it just takes me several more hours.

Negative. That loss is the absolute loss, stated in watts, of the chain. At T0 that loss is between 4 and 6 watts for both chains. That’s pretty consistent among all the data sets I have.

By that logic, if I pedal at 5 watts, the hub sees between 1 watt and -1 watt?
It makes sense that at certain conditions of speed and torque the loss varies. The choice of conditions may be giving this a greater significance than it holds for actual cycling.

It’s an interesting graph Muc Off uses in the video. I would need to see more info on the number samples and the exact details of the test before taking the bait though.
I’m not really familiar with pre treated chains, but it would seem like the extra cost is fairly insignificant compared to buying a standard chain and preparing it yourself. I know from personal experience that getting rid of all the original lube is most of the battle.
The Muc Off chain is around $120 AUD delivered locally.

Personally, I find that sudden increase in friction very strange.

Which may also be the normal variance between units. I’ve had different hubs on the same bike that have 10-15w differences, which is within the normal operating range.

So it may be measuring at the pedal vs the hub (also – the P1s tend to read high compared to other PMs), or it may just be that +/- 1.5% doesn’t always mean what users think it does.

As a side point, I have never bothered with watt-matching adjustments in software. Just ride both PMs equally, manually calibrate the FTP to split the difference. All this stuff – WKO, Training Peaks, Xert, Strava Premium, TrainerRoad – are attempts to use data to proxy what’s going in the cells. Use the data but don’t be a slave to it, and remember that feel is often the best indicator. The body knows more than the software, so listen to it, and if the PM gadget is a little high or a little low, so what…

I agree with @hammerkill. You have to include the load on the chain when presenting these measurements because frictional losses increase with chain tension.

1 Like

:laughing: I like the way you’re thinking! But, yeah, if the drive train overall offers a constant resistance at or more than 5 watts and you only pedal at or less than 5 watts…well…the pedals won’t turn! But it’s not quite that simple because the drive train is more efficient at lower cadence BUT less efficient at lower rider output.

But I certainly agree that Muc Off cherry picked conditions of the exp’t to highlight their product. You’re right on, there.

Do they specify the power, chain speed and chainline in their test data? You can’t compare drivetrain power loss between tests without knowing the test conditions. 10W of power loss is enourmous if testing at 150W with a straight chainline, but very small at 2000W with a significant chainline offset, and insignificant if you are doing the same with a very slow chain speed.

This is an open response from Ceramicspeed camparing what they believe to be Muc-Off’'s procedure with their own:



The Friction Facts data shows that this in fact is not the case. The power losses shown by MUC are unrealistic, as they are generated with a chain that’s constantly in tension rather than cycling through a tension cycle every chain turn; this simple difference has a major impact on chain lubrication. Also not that none of the tests (MUC or Friction Facts) talk about chainline, which is the biggest contributor to drivetrain losses variability. In other words - there is more drivetrain performance to be gained from careful shifting than from brand X vs brand Y of lubricant.

1 Like