KMC Chains, specs, speeds and wear-resistance

[wafter]

Rightly or wrongly I've fallen into the habit of exclusively buying KMC chains. I think this was initially because they offered quick links when Shimano didn't; but they seem well-regarded, are used by many OEMs and my experiences have been good.

In the 8, 9 and 11 speed fitments I require there are a lot of variants and choosing can be difficult. Having gone down this rabbithole again I have a better understanding so thought I'd share as an aid-memoir that hopefully might be of use to others

KMC do two main lines for 8-12sp bikes; "X" and "E"; the number of speeds denoted by an appropriate suffix - for example X8, X9 etc for 8 and 9 speed X-series chains repectively.

The X-series are the core range, the E-series are aimed at E-bikes; I think the big difference being a different pin configuration in the latter to handle more load. They're all offered in various different (aesthetic) finishes which won't be covered here, as well as different models which are broadly summarised below.

Of key importance to those of us looking for longevity is the chain's resistance to wear. KMC communicates this using the term "SP" for "stretch proof" and assigns a rating (single, double, triple or 1, 2, 3) depending on wear-resistance (which I guess is a reflection of component hardness rather than coatings). Higher numbers are more resistant to wear. I'm not sure if KMC's ratings are intended to be taken as a direct reflection of longevity (i.e a 2-rated chain lasts twice as long as a 1) or whether these are just to signify relative superiority of one over another.

Anyway, we have (in ascending order of typical cost for each series):

X8 - X12: Basic model - X8 & X9 are SP1 while X10, X11 & X12 are SP2.
X8 - X12 EPT: Corrosion-resistant coating - X8 & X9 are SP1 while X10, X11, X12 are SP2.
X10 - X11 EL: "Extra light" with cutouts in sideplate - X10 & X11 are SP3 (X8, X9 and X12 have no EL variant).
X9 - X11 SL: "Super light" with cutouts in sideplates and hollow pins - X9 is SP2 while X10 & X11 are SP3 (X8 and X12 have no EL variant).

E8 - E11 Sport: Basic model - E8, E9, E10 & E11 are all SP2 (E12 has no E-Sport variant)
E8 - E11 Sport EPT: Basic model with corrosion resistant coating - E8, E9, E10 & E11 are all SP2 (E12 has no E-Sport EPT variant)
E8 - E12 Turbo: Posh model - E8, E9, E10, E11 & E12 are all SP3 - Seems these aren't available in the UK
E8 - E12 Turbo EPT: Posh model with corrosion resistant coating - E8, E9, E10, E11 & E12 are all SP3 - Seems these aren't available in the UK


Takeaways from the above are that the 8 and 9 speed variants of the X-series chains are all one SP-rating lower than the 10-12 speed variants across the range, while the E-series are all apparently made to the same standard regardless of speed / fitment. 10 and 11 speed chains have the greatest product range, with chains of higher wear resistance offered in the X-series compared to those available for 8, 9 or 12sp.

All other factors notwithstanding if you want the most wear-resistant KMC chain in the UK your options are:

8 speed: E8 Sport (badged simply as "E8" in the UK) @ SP2 (probably an E8 EPT as this is all most suppliers tend to stock)
9 speed: E9 Sport (badged simply as "E9" in the UK) @ SP2 (again, probably the EPT flavour) or X9 SL @ SP2 (more expensive but lighter with different finishes)
10 speed: X10 EL or S10 SL (lighter, more expensive) both SP3
11 speed: X11 EL or S11 SL (lighter, more expensive) both SP3
12 speed: X12 or X12 EPT; both at SP2


Some advocate fitting chains from systems with more speeds (i.e. a 9sp chain on an 8sp drivetrain) for improved quality / wear resistance. In the case of what's currently available from KMC, to get a meaningful improvement in quality on 8 or 9sp drivetrains would require a 10 or 11sp chain which will physically fit but may or may not work correctly due to the chain's narrower outer width. There's apparently nothing to be gained in terms of wear resistance by fitting either a 9sp chain to an 8sp drivetrain, or 11sp chain to a 10sp setup.

It seems that currently the best value chain within the whole range in terms of wear-resistance is the vanilla X11, which can typically be had in the basic grey finish for £17-20 depending on where you're looking.. perhaps offering an argument for trying on drivetrains of lesser speeds as a cost-effective alternative to more expensive "proper" options.

It's also worth noting that the market's evidently flooded with fake KMC chains - so it pays to buy from a legit, established bike shop and not one of the shifty online market places.

I hope this might help some get their heads around what's available / best for their application - as it took me long enough

 

[Ajax Bay]

Great post. I used to favour KMC chains but after three separate link side plate fails (all caught by reasonably frequent chain @mickle ing - fractured plates catch on the paper towel I use) I moved to SRAM with no further failure issues.
Maybe these KMC ones were fakes but all bought through large retailers on and off line.

 

[presta]

My Horizon came with a KMC Z51, but all the ones I've bought have been SRAM for no other reason than an LBS where I bought the first one had SRAM, and I wanted to stick with them to compare like with like. They've all been PC58 on 8 speed, and PC971 after I switched to 9 speed, each were the middle one of three models ascending in price.

This is the wear data I've collected over about 45,000 miles, plotted alongside data from wear tests published by Wippermann, and some figures quoted on here by 'Reiver':

When comparing chain wear rate with small samples though, it's worth bearing in mind that:

"there is an extremely large natural variation in roller chain wear rates. Many engineers are surprised to see so much variation"
John L Wright, Standard Handbook of Chains.

 

[Ming the Merciless]

I bought a 50m reel of KMC X9 grey chain. The price worked out at approx £5 per normal length chain.

My last chain off the reel, lasted 9,543km of riding in all weathers. That’s on my recumbent where the chain is equivalent to 2 1/4 standard length chains. Thus assuming wear scales with chain length, we get 4,200km per normal length KMC X9 chain. I use ISO 64 chainsaw oil mixed with a lighter mineral oil around ISO 25.

I tend to replace the chain in winter and before my main Audax of each year. Given the relative price it works out at, I don’t try and squeeze out every last km before replacing.

 

[Bristolian]

Rightly or wrongly I've fallen into the habit of exclusively buying KMC chains. I think this was initially because they offered quick links when Shimano didn't; but they seem well-regarded, are used by many OEMs and my experiences have been good.

In the 8, 9 and 11 speed fitments I require there are a lot of variants and choosing can be difficult. Having gone down this rabbithole again I have a better understanding so thought I'd share as an aid-memoir that hopefully might be of use to others

KMC do two main lines for 8-12sp bikes; "X" and "E"; the number of speeds denoted by an appropriate suffix - for example X8, X9 etc for 8 and 9 speed X-series chains repectively.

The X-series are the core range, the E-series are aimed at E-bikes; I think the big difference being a different pin configuration in the latter to handle more load. They're all offered in various different (aesthetic) finishes which won't be covered here, as well as different models which are broadly summarised below.

Of key importance to those of us looking for longevity is the chain's resistance to wear. KMC communicates this using the term "SP" for "stretch proof" and assigns a rating (single, double, triple or 1, 2, 3) depending on wear-resistance (which I guess is a reflection of component hardness rather than coatings). Higher numbers are more resistant to wear. I'm not sure if KMC's ratings are intended to be taken as a direct reflection of longevity (i.e a 2-rated chain lasts twice as long as a 1) or whether these are just to signify relative superiority of one over another.

Anyway, we have (in ascending order of typical cost for each series):

X8 - X12: Basic model - X8 & X9 are SP1 while X10, X11 & X12 are SP2.
X8 - X12 EPT: Corrosion-resistant coating - X8 & X9 are SP1 while X10, X11, X12 are SP2.
X10 - X11 EL: "Extra light" with cutouts in sideplate - X10 & X11 are SP3 (X8, X9 and X12 have no EL variant).
X9 - X11 SL: "Super light" with cutouts in sideplates and hollow pins - X9 is SP2 while X10 & X11 are SP3 (X8 and X12 have no EL variant).

E8 - E11 Sport: Basic model - E8, E9, E10 & E11 are all SP2 (X12 has no E-Sport variant)
E8 - E11 Sport EPT: Basic model with corrosion resistant coating - E8, E9, E10 & E11 are all SP2 (X12 has no E-Sport EPT variant)
E8 - E12 Turbo: Posh model - E8, E9, E10, E11 & E12 are all SP3 - Seems these aren't available in the UK
E8 - E12 Turbo EPT: Posh model with corrosion resistant coating - E8, E9, E10, E11 & E12 are all SP3 - Seems these aren't available in the UK


Takeaways from the above are that the 8 and 9 speed variants of the X-series chains are all one SP-rating lower than the 10-12 speed variants across the range, while the E-series are all apparently made to the same standard regardless of speed / fitment. 10 and 11 speed chains have the greatest product range, with chains of higher wear resistance offered in the X-series compared to those available for 8, 9 or 12sp.

All other factors notwithstanding if you want the most wear-resistant KMC chain in the UK your options are:

8 speed: E8 Sport (badged simply as "E8" in the UK) @ SP2 (probably an E8 EPT as this is all most suppliers tend to stock)
9 speed: E9 Sport (badged simply as "E9" in the UK) @ SP2 (again, probably the EPT flavour) or X9 SL @ SP2 (more expensive but lighter with different finishes)
10 speed: X10 EL or S10 SL (lighter, more expensive) both SP3
11 speed: X11 EL or S11 SL (lighter, more expensive) both SP3
12 speed: X12 or X12 EPT; both at SP2


Some advocate fitting chains from systems with more speeds (i.e. a 9sp chain on an 8sp drivetrain) for improved quality / wear resistance. In the case of what's currently available from KMC, to get a meaningful improvement in quality on 8 or 9sp drivetrains would require a 10 or 11sp chain which will physically fit but may or may not work correctly due to the chain's narrower outer width. There's apparently nothing to be gained in terms of wear resistance by fitting either a 9sp chain to an 8sp drivetrain, or 11sp chain to a 10sp setup.

It seems that currently the best value chain within the whole range in terms of wear-resistance is the vanilla X11, which can typically be had in the basic grey finish for £17-20 depending on where you're looking.. perhaps offering an argument for trying on drivetrains of lesser speeds as a cost-effective alternative to more expensive "proper" options.

It's also worth noting that the market's evidently flooded with fake KMC chains - so it pays to buy from a legit, established bike shop and not one of the shifty online market places.

I hope this might help some get their heads around what's available / best for their application - as it took me long enough

Thank you for taking the time to research and post this information

 

[Pat "5mph"]

Useful info @wafter, thank you.
I'll pin this thread for easy reference.

 

[Big John]

Very useful, thank you. I've used KMC whenever they've been available and found them to be reliable chains. I've used SRAM, Shimano and other odd ones over the years but only because I was desperate and the shop didn't have KMC. I now tend to buy online and keep a stock of them in my shed.

 

[Gunk]

I moved to SRAM

Me too.

 

[HLaB]

Great post. I used to favour KMC chains but after three separate link side plate fails (all caught by reasonably frequent chain @mickle ing - fractured plates catch on the paper towel I use) I moved to SRAM with no further failure issues.
Maybe these KMC ones were fakes but all bought through large retailers on and off line.

I've had one KMC chain fail prematurely (warped at a few hundred miles) but all my others have been great. I put it down to a fake even though it was through a large on line retailer.

 

[Fastpedaller]

Another 'thank you wafter' very useful and informative. That X10 I recently bought for the Brompton is identified as a good purchase
The E chains may well be the best value for money - I know when I had Sedisport chains years ago they never seemed to wear, even though I did mega-miles compared to todays.

 

[wafter]

Thanks folks - tbh I wasn't sure this' post would get any attention at all given how potentially niche and OCD it is. I'm very happy that some have apparently found some value from my appetite for knowledge and humbled that it's considered sufficiently important to be pinned - thanks @Pat "5mph"

Thanks @presta for the detailed info; I envy your many data points! Currently I can only dream of having enough information to create a graph While there's a fair bit of variation your points show a pretty solid linear trend, and your examples have evidently worn a lot less than the other, published data (which in itself seems to show a pretty severe wear rate!).


While I've had good experience with KMC stuff that's not to say there aren't potentially better alternatives. Unfortunately they're pretty under-represented in Zero Friction Cycling's chain testing; others seem to do better but I think some of the KMC chains tested might be older variants.

While I reckon the humble X11 is probably hard to beat in terms of value, if you're throwing bigger money at higher-end chains the ZFC data suggest that offerings from SRAM, Shimano and Wippermann might offer better value if prices are comparable.

From ZFC's testing high-end SRAM stuff acquits itself very well, although is often eye-wateringly expensive and personally I've never really paid much attention as I find their website uninformative, model naming convention confusing, compatability uncertain... while I've never much liked them as a company (probably because I associate them with 1x, electronic shifting and other horrors of the modern world). The DB8 brakes on my MTB are quite nice, however

Shimano XTR chains also get a good rep for wear, although I believe contains PTFE so I'll not be buying those on principal.

I do think the Wippermann stuff is worth a look - seems to wear well, has an innovative quick-link (which relies on shape rather than an interferance fit to secure, so presumably would present less of a security issue when worn compared to more conventional designs). The 10sp item I have left over from my road bike seems nicely made.. That said not many UK companies seem to carry them, while they remain relatively pricey; perhaps one to keep an eye on incase someone decides to discount them heavily

 

[PpPete]

I've generally had very good experience with the "EL" range...... and very bad with the "EPT" range: An X9 EPT failed with less than 1000km and and X11 EPT failed around 300km
https://www.cyclechat.net/threads/kmc-ept-chain-avoid.294904/

 

[wafter]

I've generally had very good experience with the "EL" range...... and very bad with the "EPT" range: An X9 EPT failed with less than 1000km and and X11 EPT failed around 300km
https://www.cyclechat.net/threads/kmc-ept-chain-avoid.294904/

Thanks - I've just responded to that thread as it happens. Seems those two examples had something seriously wrong with them!

 

[presta]

Thanks @presta for the detailed info; I envy your many data points! Currently I can only dream of having enough information to create a graph While there's a fair bit of variation your points show a pretty solid linear trend, and your examples have evidently worn a lot less than the other, published data (which in itself seems to show a pretty severe wear rate!).

The linear appearance suggests to me that there isn't anywhere near enough data to be reliable, chain wear actually follows a cube root law if you plot enough points.

Again, from Wright:

"Two important concepts can be derived from Figure 3-10. One is that wear per pitch in roller chains progresses in rough proportion to the number of cycles to the 1/3 power. The other is that there is an extremely large natural variation in roller chain wear rates. Many engineers are surprised to see so much variation in wear rates. But the results of these tests agree well with those of wear tests done on many different types of bearings using many different materials. Testing shows that there is a very large amount of variation in all types of wear."

The 'extremely large' variation in wear may well account for the difference between Wipperman, Riever & Presta above.

I don't think most cyclists realise that chains wear fastest when they're new, and that the wear rate reduces as they get older. It's evident if you think about the way the link pins bear on the clearance holes they run in: you have a small circle resting inside a slightly larger circle, and when the chain is new they are near-perfect circles, so the contact area will be very small. As the wear progresses, the contact area gets larger and larger (rapidly so at first), so the thickness of the layer required to remove equal volumes of material gets smaller, hence the corresponding increase in chain length gets smaller.

It's this slowing of the wear rate that causes the tops of the teeth to round off and the chain to skip when you stand on the pedals, if you keep the chain too long. If the wear rate were linear, the time the roller spends at each point up the height of the tooth face would remain constant, and the amount worn from the tooth at each point would also be constant, maintaining the profile of the tooth.

If the chain wears much slower than the tooth even when new you have the converse problem: the base of the tooth becomes undermined before the chain has worn enough to rise further up, and you get 'hooked' teeth. In that case, you're looking for measures that will slow the tooth wear relative to the chain. One way is to buy gears made from harder alloy, another is to share the wear over more teeth by not using the same gear all the time, and the other is by not rotating a set of three chains. A set of three chains just simulates one chain that wears three times slower, and will therefore undermine the base of the tooth even faster.

The teeth on my middle chainring get hooked because my riding style leaves the chain on the middle ring most of the time, with all the wear is being taken by just 32 teeth. I've never had to replace a small or large ring, but the middle ones lasted 8000 & 9000 miles (6000 out of a lighter alloy). By comparison, the cassettes aren't hooked even after 19,000 miles because the wear's distributed over a total of 189 teeth. I've never rotated multiple chains. The data I plotted above weren't measured until long after I quit cycling, but in retrospect, had I been making more considered decisions at the time I don't think I would have changed the chains as early as some of those were.

 

[silva]

I became a "know what you buy" adept and with KMC chains, I don't know.
"corrosion resistant coating" is not a specification.
Also this:
"ECOPROTEQ COATING: The latest milestone of KMC in rust protection. The non-toxic EcoProTeQ treatment doubles the rust resistance compared to KMC‘s RB anti-rust treatment."
"EcoProteq": says nothing.
"RB" stands for "Rust Buster": says nothing.
"Rust resistance": is property nor unit, and a "doubling" of it doesn't tell more.
KMC decided to hide the properties of metal and kinds of coating, as to obstruct making an objective decision.
For industrial customers, steel grades, steels treatments, coating materials and thicknesses are data sheet items.
To illustrate the point: I once bought "Rust Buster" singlespeed chain. What was the rust protection: lilely some kinda zinc bath or spray alike they do for static applications. After a couple rides all the zinc on places of mating surfaces ended on the streets and leaving the zinc where it doesn't have any chain life benefit, aka completely useless, at most cosmetical when looking at the side.
So ordinary "over"zinc, but KMC labeled it "Rust Buster" alike it's some dedicated technology of their own.
Also, "over"zinc increases dimensions of parts, meaning it voids designed engagement gaps, increasing wear on the zinc, plus zinc is soft, so the overzincing job adds cost for near zero gain.
The only "gain" is that the first days your bicycle chain runs more silent than normal. Because the soft zinc dampens the engagement noises by diverting the energy towards deformation, that results in it coming loose and thus temporary.

Another example: the ceased Z1X INOX.
An as stainless steel marked chain.
Stainless steel has mechanical properties halved relative to steel, which was compensated for by plates twice as thick (and thus also longer pins), effectively doubling the mating surface.
Thumbsup, BUT, and what KMC decided to NOT list, the pins are NOT stainless.
I held a spare bike, typically stalled 6 months, thought no need to put oil to avoid rust, surprise, pin ends brown and some even upto the point that the riveting surface didn't suffice anymore to prevent the plate to detach.
Something you notice after the first ride in wet conditions, the rust-expanded steel then loses the remaining mechanical strength in no time, dissolves, and all of sudden pedaling without chain.
Why bothering for stainless steel rollers and plates to then ruin the gain anyway by leaving a (rust)weak part?

https://web.archive.org/web/2013013...nts-for-Bosch-e-Bike-Drive-System-BIK005583W/

To cope with these high torsion forces the world’s biggest in bicycle chains, KMC, developed a special chain. This type is fitted with reinforced pins that can hold the side plates of the chain when heavy forces are applied. The shapes of the side plates are specially formed to allow for a better shifting performance which brings down the forces applied to the chain.

"Specially formed side plates"
"reinforced" steel pins
= the same nothingsayers.
A chain breaks when the plate - pen joint is too weak to transmit the force.
The real reason is of course that for non motor-chain drive bicycles, they had cut down costs along material & quality to a least acceptable degree, and that an E-bike motor that puts more force than most people do, reverted what was accepted to not anymore accepted.
So no special formings and reinforcements, but just a return to a past point of bang for the buck, back to a previous normal, that is. And hence now more bucks than back then.