Rappel Test for Sudden Accidental Disconnect

[Scott McCrea]

Thud. "Uggh." I heard it this weekend—the sound of coverall-ed flesh smacking limestone. Luckily, very luckily, the caver fell backwards wedging across the top of the canyon. He had a death grip on his rack, which was rigged to the rope, but no longer attached to him.

He was a victim of Sudden Accidental Disconnect (SAD). Under the right circumstances, it is easy to disconnect a rappel rack from a carabiner. If the carabiner, connecting the rack to the harness, is unlocked and the rack is momentarily unweighted and things line up just right, the rack will pop out of the carabiner. Here's a simplified example of how fast this can happen.



SAD has claimed the lives of at least two cavers and almost got my buddy. He was crossing a simple, standing rebelay. He rigged his rack, fiddled with the bars and removed the cows tail. When he sat down on the rack, he quickly realized it was no longer attached. Somehow, in the fiddling and reaching, the rack eye had levered open the carabiner gate and slid out. A simple mistake and a some luck is all it takes.

Don't worry, it is very easy to prevent with the Rappel Test. Since everyone knows a QAS should always be above an untested descender, simply slide the QAS down and rappel a couple inches proving the descender works and is attached. Then, and only then, can the QAS be removed and properly stowed. The Rappel Test can catch other mistakes, not just SAD.

Rappel Test
1. Engage QAS.
2. Rig descender.
3. Rappel an inch or two.
4. Circle one: pass/fail. Fail=repeat test.
5. Stow QAS.

It is also possible to lessen the threat of SAD by replacing the locking carabiner with something more secure or a screw link. However, nothing will lessen importance of always doing the Rappel Test before each descent. Note that SAD is not an equipment problem. It is fully due to operator error.

Please, take the Rappel Test. Everytime.

[Carl Amundson]

That is why I always use a quick-link to attach my rack to my D-link.

Here is photo that illustrates the same thing:

"The above photos illustrate how the gate of a carabiner, which has unscrewed itself back into the open position, can become forced open due to the motions of the bottom of a rack pushing against it. Such a situation can, and has, occurred when a caver stands up on a ledge in order to disconnect the small cowtail (i.e. safety) from the hanger supporting the rope onto which he has just threaded the rack. Needless to say, this can be a FATAL mistake.
Of course, if one treats every transfer from one rope to another as a full rebelay (i.e. placing the top ascender of the frog system on the rope just below the hanger) and stands up on the footloops, instead of a conveniently located ledge, in order to disengage the cowtail – you would be connected to a safety tether and LIVE!"

[jharman2]

Very scary.

It is also possible to lessen the threat of SAD by replacing the locking carabiner with something more secure or a screw link.

It is my personal opinion that a carabiner should NEVER be used as the attachment between the D-Link and a rack.

It is very easy to load a biner across the gate when crossing a rebelay. It is impossible to cross load a properly sized screw link.

Still like Scott said nothing replaces keeping that 2nd point of attachment.

[NZcaver]

It is also possible to lessen the threat of SAD by replacing the locking carabiner with something more secure or a screw link. However, nothing will lessen importance of always doing the Rappel Test before each descent. Note that SAD is not an equipment problem. It is fully due to operator error.

Please, take the Rappel Test. Everytime.

Excellent post, Scott.

[ek]

EDIT: Clarification: I am not arguing that SAD is less of an issue that Scott is saying it is. It's a huge, major issue, like Scott says, and you avoid it by engaging in proper practices. Rather, in this post I argue that carabiners, used correctly, are safe to use to attach racks. Used correctly.

It is very easy to load a biner across the gate when crossing a rebelay.

Suppose the attachment carabiner for a descender becomes loaded along the short axis. Then you would lock off, put on an ascender-with-footloop, and fix it. What's the chance that a carabiner, loaded along its short axis, will actually break? For a carabiner to be UIAA/EN certified, its short axis strength must be at least 7 kN. Any descender (even auto-braking) will slip before that. An ascender will cut the rope before that point. (Prusiks tend to slip around there, but may not.)

I acknowledge that it's possible to break a carabiner loaded along the short axis, if there is a major shockload (though in descent, your descender would have to be locked off). I agree that you shouldn't go around rigging things with your carabiner loaded along its short axis. But I fail to see how loading a locked carabiner along its short axis while rappelling carries a substantial risk of breaking the carabiner.

When it comes to accidental disconnection, this can happen with a carabiner that is unlocked, and it can also happen with a maillon that is not screwed shut. It is perhaps worth mentioning that a carabiner of remotely modern construction loses no strength when the gate is left unlocked, whereas a maillon loses substantial strength. Of course, for a maillon attaching a descender to a harness maillon, it's probably more obvious when it's not screwed shut. For the harness maillon itself, it's not obvious at all.

Finally, it's important to mention that while a small maillon is safer than a carabiner for attaching a rack, it is dangerous to use one to attach a standard bobbin since its small size makes it possible for it to twist and operate the spring clip, which will open the bobbin and disconnect it from the rope.

[jharman2]

EDIT: Clarification: I am not arguing that SAD is less of an issue that Scott is saying it is. It's a huge, major issue, like Scott says, and you avoid it by engaging in proper practices. Rather, in this post I argue that carabiners, used correctly, are safe to use to attach racks. Used correctly.

I am not argueing anything rather stating some opinions. Please note that as in the original post I am only talking about the RACK.

1. The safest way to attach a rack to a D-link is directly. No screw link, no biner. You will never cross load anything and you only have 1 screw gate to tend.

2. The second safest way to attach a rack to a D-link is using a properly sized screw link. The screw link will not load across the gate. The screw link engages more threads with more friction than a biner. It is obvious when a screw link is not locked. However, it is good form to check each time before the screw link is loaded and of course keep your QAS on until everything checks out.

3. The third safest and last acceptable way to attach a rack to a D-link is using a locking biner. You MUST ensure the orientation of the biner when weighting the rack. You MUST ensure the biner is screwed shut.

I know of no "trail of bodies" of cavers that use screw links. Sadly as Scott said there are two documented cases with biners.

But I fail to see how loading a locked carabiner along its short axis while rappelling carries a substantial risk of breaking the carabiner

Sure the biner is strong (>7kN) across the gate. However, loaded across the gate is using it outside of its intended purpose. Do you want to cross a rebelay at +160' only to find that your biner is cross loaded? What if you don't catch the blunder and rappel 5 feet only to have the biner orient itself correctly? Your hand gets pinched between the biner and the rack. You were using a micro on fast 9mm and your brake hand lets go of the rope to try to free your stuck control hand. You have a very quick 160' ride and the inside of your cave suit becomes as dirty as the outside.

[Scott McCrea]

1. The safest way to attach a rack to a D-link is directly. No screw link, no biner. You will never cross load anything and you only have 1 screw gate to tend.

Unless you are using a ropewalker. When changing over you need the slop provided by an extra link between the rack and harness. Try it. But, do it on a rope rigged thru a pulley, so you can be lowered down if need be. Some day, I'll write about this fun little hazard, too.

Screw links are safe. Carabiners are safe. Humans are not. The Rappel Test saves humans. And not just from SAD, also suicide rigged racks, racks rigged to a lanyard instead of the rope, fig 8's that didn't get clipped back into the biner, bobbins that didn't get closed all the way, etc, etc.

[ek]

Sure the biner is strong (>7kN) across the gate. However, loaded across the gate is using it outside of its intended purpose. Do you want to cross a rebelay at +160' only to find that your biner is cross loaded? What if you don't catch the blunder and rappel 5 feet only to have the biner orient itself correctly? Your hand gets pinched between the biner and the rack. You were using a micro on fast 9mm and your brake hand lets go of the rope to try to free your stuck control hand. You have a very quick 160' ride and the inside of your cave suit becomes as dirty as the outside.

Cross-loading a carabiner is not a use outside of its intended purpose. Carabiners are designed and tested to withstand short-axis loading in all but the most severe loading situations.

When you say "control hand" do you mean the hand that holds the rope below the device (which, for fixed friction devices and "hybrid" devices like the micro rack, is called the "brake hand"), or do you mean the hand that controls spacing of the bars? I believe the former is the standard meaning of the term "control hand" among rack-users.

The situation you're describing where my hand is pinched between the carabiner and the rack sounds unlikely, since the hand would have to actually be in the space at the top of the carabiner. If my hand is touching the improperly oriented carabiner, how would I not know it's improperly oriented? I accept that this is conceivably possible, for instance if I had no feeling in my hand due to cold. But unlikely. I would be much more worried about being unable to attach my rack because I lack the necessary dexterity to screw a maillon shut. (Of course, if the rack is attached directly to the harness maillon, then that's not a problem either.) Furthermore, if my hand that I was using to space the bars were to become incapacitated in any way, I would not let go with my other hand. I would use the hyperbar.

I'm right-handed. If my left hand did get stuck between the carabiner and the rack then I would not let go of the brake end of the rope with my right. I would use the hyperbar. I can actually lock off a micro-rack with a Bassett twist with one hand only, but if I lacked the dexterity to do so then I suppose that after running the rope over the hyperbar I would, without letting go of the rope, open up my Croll and attach it to the brake end of the rope. This would allow me to free my right hand and attach my upper ascender with its footloop, and I would then stand up in the foot loop and pull out my left hand. I accept that this is not an ideal procedure for me to have to engage in. But it would most likely succeed, and the situation requiring it is unlikely to occur.

I accept your hierarchy of safety in rack attachment, at least in terms of detachment risk. On the other hand if your rack is stuck--say, jammed up against a knot, it's good to have the option of disconnecting it from yourself. This is easy to do with a carabiner, possible to do with a separate maillon, and impossible to do if the rack is connected directly to the harness maillon. Many rack users, especially male users and myself included, find it downright painful to climb with a rack swinging around between our legs, so we use a detachment connector (carabiner or maillon) to enable us to remove it from the harness maillon when we're not using it. The pain can be reduced but, for some, not entirely eliminated by clipping the top of the rack to a gear loop. The additional time and effort that it takes to operate a maillon as compared to a carabiner complicates any already-complicated procedure and could worsen any already-bad situation. Both carabiners and maillons can get stuck screwed shut, but this is a much more significant problem for maillons, and when it happens with carabiners it can usually be fixed by putting the carabiner back under load and unlocking it. Finally, if you keep your rack directly attached to your harness maillon on descent but clip it to your gear loop on ascent, then if you have to change over, you'll end up needing to attach it to your maillon with a carabiner or maillon. Then it helps if you have experience rappelling in this way, since most circumstances requiring a changeover are somewhat trying.

Actually, the reason that compels me to use a carabiner rather than a maillon to attach my rack to my harness maillon is that, when transferring the rack to or from the harness maillon while suspended from the rope, there is a risk of dropping the rack off the open maillon or the open maillon off the rack. With a carabiner, you can trap the rack between the open gate the lower-end (i.e. hinge-end) of the carabiner.

[Ralph E. Powers]

In over 30K feet of rappelling with a rack I've used a triangle shaped steel mallion directly connecting my harness and my rack together. Occasionally I will use a steel locking biner with the gate turned so that it's pressing against my crotch and it's a pain to turn it around to open it.
I was taught this over 15 years ago by one of my mentors about the dangers of not ensuring that a biner is locked or has a chance to be clipped open by the rack. Thankfully it's never happened to me nor anyone else that I've known or trained.
One of the things I've learned is that there is no ONE right way to do things safely... but there are dozens upon hundreds of ways to do them wrong and usually only once.
It is sad that folks have learned this the hard way and that their companions and friends learn the lesson at the expense of those who have had this happen... but it is good that we have this broad network of communication between vertical lovers to ensure that it won't happen often. To say it won't happen again is futile. Accidents DO happen irregardless. It is up to US to learn our gear so that we can do it in our sleep. It is the responsibility of everyone on the trip to account for everyone else's safety. Egos need to be parked and left in the car or better at home and safety checks requested by others in the party. It is far too easy to overlook a simple safety step and I've been guilty of that myself a time before... very scary and it drives home why there are things that we must do before getting over that lip. Asking someone to safety check your gear/harness is just simple common sense.

Sorry... off soapbox.

[fuzzy-hair-man]

It might be worth mentioning:

If you don't use a QAS the long cowstail can fill the same role, you can clip it into the rebelay anchor or the rope above the rebelay (the rebelay loop), the second option will mostly give you a little more slack so it's easier to avoid the situation where you test descend too far and weight your long cowstail. You will be prone to this if your short and long cowstails aren't all that different lengths.

And even if you do use the QAS the above is still an option and a long cowstail is going to be better at catching a fall than an ascender, as the long cowstail is less likely to cause damage to the rope after a fall.

Scott: I know what you mean about the ropewalker and descender attachments, I have a friend who has been finding that out the hard way because there's no one around here to learn the finer points of a ropewalker from, I think she's got it sorted now though.

[Bob Thrun]

But I fail to see how loading a locked carabiner along its short axis while rappelling carries a substantial risk of breaking the carabiner.

See http://www.caves.org/section/vertical/nh/47/uiaa2.pdf

I had this kind of loading happen to me when I put my weight on a ledge at the start of a rappel. Fortunately I was using a maillon rapide.

[ek]

But I fail to see how loading a locked carabiner along its short axis while rappelling carries a substantial risk of breaking the carabiner.

See http://www.caves.org/section/vertical/nh/47/uiaa2.pdf

I had this kind of loading happen to me when I put my weight on a ledge at the start of a rappel. Fortunately I was using a maillon rapide.

Good point.

This is not really short-axis loading, but it is a fatal way to load a carabiner against the gate.

While this typically happens to people using figure 8's, I suppose it would be possible with a rack. Using with a U-rack, like a micro-rack, I think would eliminate this risk.

I've thought of using a little piece of thin accessory cord, tied to the bottom of my micro-rack and barrel-knotted to a small maillon, to keep the rack attached to the maillon. If I did this, I'd feel good about using a maillon to attach my rack. But I'm too afraid that in a moment of delirium/distraction/hypothermia/exhaustion/idiocy, I'd attach the assembly to my harness maillon with the rack outside the maillon and only attached by that bit of cord.

I suppose I could mostly eliminate that risk by using 5mm Spectra/Dyneema cord (or 5.5mm Spectra/Dyneema with a nylon sheath), with the triple-fisherman version of the barrel knot. Then if I make that mistake, the cord should be able to hold me up...

What do you folks think?

[Scott McCrea]

What do you folks think?

KISS

[Scott McCrea]

This is not really short-axis loading, but it is a fatal way to load a carabiner against the gate.

While this typically happens to people using figure 8's, I suppose it would be possible with a rack. Using with a U-rack, like a micro-rack, I think would eliminate this risk.

I tried this a few years ago. I was fully prepared to shatter a biner and crash to the padding on the floor. But, no matter what I tried (including bouncing and bouncing a two person load), I could not break one. Anyone else care to try?

[Carl Amundson]

What do you folks think?

KISS

Amen brother.