Articles in English

Cowstail as a fall arrester

Tuomas Pöysti 2021

Let’s make one thing clear right away: Cowstail is not a fall arrest device. It must be used for suspension or positioning only, never as a belay system. Simply put, any slack in a connected cowstail should be carefully avoided.

This article is purely ‘academic’. It should not be used as a practical guideline of any sort. If something practical, this text is meant as part of discussion on how cowstails basics should be taught.

Good theory, wrong application

Cowstails are strictly banned as a belay method, and rightly so, no one should disagree. However, this is often accompanied by horror stories about bolts breaking or cowstails snapping due to enormous impact forces. This decade old article from DMM is widely and wildly cited. The article is good and not in any way obsolete. Like any semi-scientific text, though, it can be read in a thousand ways, most of which are wrong. This Black Diamond user instruction has an actual image of a bolt failing in this context.

Another aspect is the story of fall factor (FF). As the linked Wikipedia article and for example this Petzl article imply, the concept relates to climbing with dynamic ropes. Unfortunately, it is commonly applied to cowstails and other short tethers as well.

Do ends justify means?

It might just be me being a nitpicker, but I don’t like to justify false claims by good intentions. It not even a wise strategy, if one does not care about dishonesty.

Let’s imagine an instructor who says on a basic rock climbing course that a fall factor 1 fall on a dyneema cowstail with a knot connected to a steel anchor will break the bolt, the lanyard or the climber’s back. Now, the credibility of the whole thing relies on the hopes that no one on the course will not end up actually taking such a fall. Of course, it should never happen for safety reasons alone, but on the other hand, a 60 cm drop on a 60 cm cowstail probably just slams the climber against the rock in a nasty little pendulum.

Not good, might twist an ankle, shows poor supervision and guidance. But no snapped cowstails, ripped bolts or broken backs, either. How credible should the participants now find the instructor?

The test setup

I don’t have a nice lab – I don’t even have a good, solid anchor handy that I’d like to use in winter. What I do have is a wall bars unit in my daughter’s room. They are nice, solid wood, quite sturdy at the end of each bar. But surely not as rigid as a steel bolt in a massive concrete beam.

I used myself as the test mass (around 83 kg with gear), wearing an EN 12277 C harness (Ocun Webee Quattro), a CT Multi Chain Evo PAS, which is claimed to be made of dyneema, and a Rock Exotica enForcer loadcell.

In the drop tests I have done before, I have suspended the test mass with a piece of retired rope and used a knife to “release”. This is a nice and cheap way, but on the other hand makes it hard to achieve accurate and repeatable drop heights. That’s why I bought a Tylaska T5 snap shackle. It is for sailing; that is, it looks like a toy and comes with a ridiculous price tag. That little key fob actually cost 180€ in Finland. But it does seem to work fine so far!

The test setup at 60 cm length and about 25 cm fall distance. For each drop I tightened the rope, measured the distance to lift, pulled myself up using the Microtraxion and released the snap shackle.

The Multi Chain can be adjusted to varying lengths by clipping different links. I selected 60 cm as a typical length, 35 cm as a short one and 105 cm as very long – the chain is even longer, but I did not have enough vertical space for it. The length is of course measured from end to end, although the tether is tied to tie-in points of the harness (using the method shown in the instructions) and thus replaces the belay loop.

I tried to maintain the drops as close to the vertical line of the anchor as possible. In case of the 60 and 105 cm cowstails, this was no problem, since I was able to use the same anchor, suspending the snap shackle on either end of the load cell. The 35 cm cowstail did not leave room for this and required a different anchor, and it inevitably was a couple of centimetres off the vertical axis.

Body posture during the fall really makes a difference. I tried to keep the same posture through the tests, and it seem to have paid, since the results are quite coherent. On the other hand, someone else would probably get different results, although coherent as well.

The results

The results are the maximum impact forces recorded by the load cell (500 Hz sampling rate). Unlike in some of my other articles with different contexts, these values are directly from load cell, and they include my weight. Here they are as a plot of force against fall factor:

Hardly promising, unless we differentiate cowstail lengths:

I have fitted trendlines for each series, which can of course be very misleading. Let’s try to keep in mind we have no reason to assume any linearity in broader scope.

The forces seem to be roughly linearly proportional to fall factor, but also depend on cowstail lenght. However, fall factor is also a function of cowstail length. If we plot against drop distance alone:

We find way more promising correlation. Again, the trend lines are not to be taken as final truth, but they happen to bring out three lines with almost the same slope. The maximum force is roughly

Fmax = 1 kN + x * 9.75 kN/m + k,

where x is fall distance in meters and k is -0.2 kN for 105 cm lanyard, 0 for 60 cm lanyard and 0.2 kN for 35 cm lanyard. This equation has no relevance beyond showing how small the deviation was within the scope of the study. This is clearly visible in the plot, too.


Fall factor is not a very useful concept when it comes to short falls (and static materials). As I concluded here, the main energy absorbing element in this case is not the cowstail but the human body. Lanyard length has an effect to the impact force, but it is small compared to the effect of fall distance.

Imagine falling on a concrete floor, bottom first. A 10 cm fall might be unpleasant, but probably wouldn’t do any harm. The floor can be seen infinitely rigid in this context. That is, it does not allow any room for deceleration. How is it possible, then, for any object to be stopped by the floor without an infinite deceleration and thus infinite force?

This is an easy one, of course. Like a rubber ball bouncing, the human body deforms absorbing the kinetic energy. The mass center of a ball or human body does not decelerate in zero distance. In case of human body, the limbs alone, making a swinging motion, will significantly shift the center of gravity. We are wobbly piles of flesh.

By the way, dropping a 80 kg steel weight on a concrete floor might actually be quite scary – the idea of the weight crushing the floor and falling into the basement could be taken way more seriously than with a 80 kg human. Got the subtle reference to stories about breaking bolts?

We intuitively understand it is quite ok to take small falls on a floor, however rigid it might be. Actually, this is something everyone who has learned to walk has tried many, many times. If someone has difficulties believing a similar drop on a cowstail could do any more harm, they are probably right – the cowstail is not even as “rigid” as a concrete floor.

I think the key is to have all participants on basic courses understand that falling on a cowstail can be as bad as falling on a concrete wall bottom first. Small fall – it hurts a bit. One meter fall – of course you may get injured. It does not take concepts like fall factor to comprehend this, and the story does not need to contain concrete floors breaking.

And then the discussion about nasty pendulums, snagging and wrenching carabiners and other real problems of falling on a cowstail can begin.