Carabiners for Litter Connections?

[ron_miller]

but you still have not answered my question.

What question would that be, Andy? If it was

Okay then, if you do not think it unsafe, then why are you questioning it?

then I think I pretty thoroughly answered that. I never said I do not think it unsafe, so the premise of your question is false. In fact, I questioned it because I DO think it is unsafe.

I fundamentally disagree that in a rescue or rescue training scenario, "unsafe" = "someone is about to die or be injured". In my mind, "unsafe", especially in an introductory-level teaching scenario, can include "a practice that can foreseeably result in failure of a primary safety system component." From my perspective, needing to rely on the primary belay connector's backup - the harness connection to the presumably already injured patient - to take the entire force of a falling rescue load, does NOT make the system SAFE. On the contrary, if an injured patient must hold the force of a falling rescue load onto his/her harness, there is significant potential for further injury to the patient - in my mind, that makes this scenario UNSAFE.

If the belay is keeping pace (or even close to it) I don't see how the peak dynamic forces you describe could actually be generated in this scenario.

I concur. That said, the whole point of this discussion is not to get overly caught up in the specifics of this particular scenario, but to use this photo as a starting point for a broader discussion of the use of carabiners as litter connections, especially at the introductory level of cave-rescue training. How many times have you seen significant slack develop in the belay line at some point in a haul or lower operation? I've seen it happen numerous times.

The British Columbia Council of Technical Rescue (BCCTR) Belay Competence Drop Test Method (BCDTM) was developed to test foreseeable loading scenarios on belays, with a 200 kg (or 280 kg, depending on the desired rating) mass dropping one meter on three meters of rope (presumably, with the idea that the most likely scenario for a main line failure is at the lip). Over and over again in BCCTR BCDTM testing, forces well above 7 kN are generated in the belay system. If you don't think the BCCTR test is reasonable for defining "safe", then I would be interested to hear what you propose as the alternative.

[Anonymous_Coward]

That said, the whole point of this discussion is not to get overly caught up in the specifics of this particular scenario, but to use this photo as a starting point for a broader discussion of the use of carabiners as litter connections

Sorry Ron, I didn't realize that the focus was this narrow and that discussing the specifics of the photo was off-topic. I also didn't realize that you were planning on dropping the patient one meter. That is what sounds unsafe, and could easily translate into "about to die". I know this is the BC standard, but it should never happen. For comparison, letting go of a munter belay, even for one second, is also unsafe. But it should never happen. Yes, I know that sometimes things do happen that shouldn't. This is why we over-engineer.

Let me offer these observations and then I will take my answer off the air. The biner in the photo only slipped sideways because the belay line was not continually tensioned. The reason it lost tension is because it is a belay line operated by a human being. A belay line by definition is a backup. In the photo, the backup was backed-up by another backup. The biner on the main line did not rotate, because it was under constant tension. In many real-life applications, there is no belay. I would offer that in these situations, there would be very little chance for the biner to rotate and cause this problem. Even if it did, aren't people watching all this? If you have a vigilant rescue team, someone will notice this problem, yell STOP! and fix the problem just like Jansen did. If this kind of vigilance is impossible, then add belays.

If you take this discussion out of reality, and into the realm of the theoretic, then I am done with it. I don't have time for that kind of abstract physics geekery. If we do keep it in the real, based on the photo we have two contingency plans for a main line failure. That seems pretty safe to me. Also, in reality, we're gonna make damn sure there is no way we ever drop this guy anything close to one meter. Once again, THAT would be unsafe. How about we teach level 1's that. Don't do things during a rescue that cause peak dynamic forces! You may think I am baiting you, but I am not. It is very similar to "don't let go of the munter belay" or "don't let go of the steering wheel when you are driving." In the photo, it does not appear that the patient would fall more than an inch or two, and I'll betcha that the cross-loaded carabiner would hold in that situation.

For those of you with more rescue experience than me: How common is main line failure? How often have you seen it?

[SDcaver]

Ron- you stated- "a practice that can foreseeably result in failure of a primary safety system component." As I asked in my last post, please provide me with data, test results, something that proves your claim! As an example I will do so- the following use biners for connecting belays and mainlines to a rescue liter:

Dave Merchant, Life on a Line 2
Ken Laidlaw, Considerations in Rope Rescue 2009
Speleo Secours Francais, Cave Rescue Manual
NSS, On Call
CMC, Rope Rescue Manual
NCRC, Manual of US Cave Rescue Techniques
Rick Lipke, Technical Rescue Riggers Guide

These are just a few from a quick glance of my library. Granted everyone could be wrong and you right, but I need something to convince me of that other than your opinion.

When we are looking at biners please keep in mind that they are different and have different ratings. Stating that aluminum biners have a cross-loaded rating of 7kN is a generalization. I have some that are rated at 9 kN, which would hold up to the BC drop test for a rescue load.

Ron, you also stated “if an injured patient must hold the force of a falling rescue load onto his/her harness, there is significant potential for further injury to the patient - in my mind, that makes this scenario UNSAFE.” Then what do you do when you are hauling/lowering someone not in a liter? Keep in mind that the scenario in the picture is NOT a rescue load, but a single person load. A rescue load is defined as two people plus gear.

[PseudoFission]

Friends,

Great discussion and interesting points. It's crazy how quickly people trend to argue about the nuances of diction and ignore the actual technical setup at hand. Did anyone look closely at the picture besides Marc? Because it would answer most your questions (and I'll reiterate some of your points, Marc):

* The patient IS tied in with the tail of the knot (at least to a biner at his harness, which is hopefully oriented correctly under the packaging). The cross-loaded biner will take the load of the ferno at the top bridal (though it looks like the patient might take load first since the line is already under tension to his harness). The failure of the questioned biner would cause the patient to be held by his harness, still in the ferno and holding the weight of the setup (probably upside down).

* The BCCTR tests are for a TWO person (440+ lbs) load, ie patient and litter bearer. The belay in question in the photo is for a one person load. This is a big difference, as the BCCTR tests (which I wouldn't take too seriously) produce mainline failure peak 'shock loads' in the ranges of aluminum cross-loaded biners. Considering the weight difference, it is likely that the belay would hold even during a cross-loaded shock load on an typically-rated aluminum biner.

* 'Safety Factors' and estimating 'shock loads' are extremely difficult to apply to dynamic belays in the first place because of the number of variables for different setups and the lack of testing. These factors include rope type, length, bridal materials and connections, the complexity and rope friction of the vertical environment, the type of belay and technique of running it, the amount of slack in the system, to name just a few. There has been very good advancements in computer models and mannequin drop-testing over the last few years, but it's still very difficult (impossible) to obtain live volunteers for specific mainline failure tests. As Marc stated, I think it rather silly to start claiming 'dangerous and life threatening' practices without having any documentation or test results to prove it so. Let's talk science and mechanics, not speculative drama or accusations.

* Cross-loaded equipment doesn't always stay cross loaded under sudden load. With any slack in the belay, there will be a moment of free-fall in which the movement and final orientation of connections will be uncertain. Some newer biner designs incorporate smoother locking-mechanism profiles to prevent snaggings/cross loads.

* All connections have safety limitations. Knots can be tied wrong and lack versatility in some situations. There is no 'right' answer for everything, especially for the complex and varying circumstances of cave rescue. Usually it's a compromise between versatility and minimal gear. Lots of high angle rescue teams with complex bridal/spider connections to the mainline use giant steel or aluminum rings to avoid cross-loading of components. These teams all use 2-person+ loads and have different system considerations.

* At some point the human variable must take responsibility for the success of the system operation. Knots, friction hitches, lowerings/raisings all require an attuned perception and skill of the user to run correctly. If the litter is left alone to be pulled into a slot, or the rope is loaded on a dangerous pendulum, do we re-design the litter and rope system? How do we design for rock fall, flooding, and lightning? Despite our best designing of systems, there are some variables that must be managed strictly by human diligence. Again I would say there is usually a compromise between this diligence and system redundancy, with the former being more important in my opinion.

* The key is not to let the main line fail in the first place! I've been affiliated with a high-angle rescue team that has been around for 60 years. They average ~4 big high angle missions a year, and practice it at least a half dozen times a year, often with brand new people. In those 60 years, there has never been a single mainline failure.

-Andrew

[Clem Akins]

Gentlemen,

Thank you for bringing the tone of this discussion to a rational, reasoned one. Now we're getting somewhere.

Marc, it's interesting to find your interpretation of the same reference books I have is different from mine. I'm glad to hear it! That's the reason such details are worth discussing. I approached those same books with the assumption that "loose" carabiners (those whose orientation I couldn't be sure of) were not used if there was any doubt, where you have assumed they were used. My copy of "On Call", my 2002 copy of Laidlaw, and my Revised 1st Edition of Lipke all show an unclear picture, or simply use terms such as "attendant is connected" that I interpreted to mean a knot. LOAL (2) clearly shows carabiners, and I don't have the other works you mentioned at hand. My notes from a Rigging for Rescue class show a knot used to tie the attendant and the patient, but I didn't write down a definitive rule about not using a 'biner.

Lipke's book is the most pertinent, I think (though I don't have his latest edition here, dangit.) He stresses (as does the Rigging for Rescue literature) that the litter is attached with a single carabiner, and two interlocking long-tail bowlines. Had the litter in the photo above been rigged this way the cross-loaded biner would have been eliminated. This is a much better way to rig than by simply clipping two 'biners to the litter bridle and hoping that the belay one stays safe. Yes, this is an example of using a 'biner on a litter, something I ranted against in my original post. I'll eat those words if you'll accept Lipke's illustration as a better way than the photo shows. The interlocking of the two long-tailed bowlines is important, as it ensures that the belay's attachment loop is very close to the haul's attachment loop--they are tied together.

Andrew, your post is also fascinating, thank you sir. I accept your notion that sometimes it's better not to have a belay at all, and that the risks of the main line breaking are really very small, in practice. That said, I do take issue with your defense of the cross-loaded 'biner. I think that it was ready to fail if the main had broken, and that the patient's only connection then would have been at his waist (a poor place to take a fall) and that it would be of the type that had already failed once--a "floppy" carabiner that could have just as easily been a knot that would be less likely to fail. Why not use a tool that you know will work in an already disastrous scene? I'd much rather not have a belay at all and thus be aware of the care I need to take of it than to have a belay that gave a false sense of security.

I'm trying to say that carabiners whose orientation can't be predicted are giving that false sense of security. Arguments such as "use a bigger 'biner" aren't good enough. Saying "just don't break the main line" isn't good enough, either, I'm sorry.

There really is a pile of bodies associated with broken and unclipped carabiners. I personally know two TAG cavers who were injured in a fall from a broken carabiner. These days I use a Maillion Rapide to hold my rack to my harness as a result--I don't trust a 'biner to stay loaded correctly. Ironically, a well-known caver died as the result of his bobbin coming unclipped because he used a Maillion instead of a 'biner, because he didn't trust his carabiner. (The MR was the wrong tool for the job, as it wasn't wide enough to keep his rappel device closed. That's my understanding of how he died, anyway. I helped to carry his body from the scene.) Rock climbing has a well-documented history of deaths as a result of carabiner failures. The original poster denied the validity of the comparison, and insisted that only "dead bodies" from the failure of a belay carabiner on a cave rescue was a suitable analog. I disagree with that, and maintain that the rock climber's experience is directly applicable and that he knows more than anyone about how carabiners versus knots behave in a fall--it happens in his lap all the time.

Finally, I don't teach people to drop a load on a cross-loaded 'biner, but I certainly do teach people to build belay systems that can carry the full force of the load. A system that contains a cross-loaded 'biner cannot do that. If a properly tied knot will eliminate that problem, I say use it! It's a really bad idea to rely too much on just "clipping in." I would be really happy if all this ranting and posting just had the effect of causing people to think twice about trusting a carabiner that's not under load. Will a knot work better, there?

Thank you,

Clem

[PseudoFission]

Clem:

"that the litter is attached with a single carabiner, and two interlocking long-tail bowlines. Had the litter in the photo above been rigged this way the cross-loaded biner would have been eliminated."

Not necessarily. If you use a carabiner in any application, at all, under load, it has potential to be cross-loaded. Do you have a picture of what you are describing with the long tail bowline setup - specifically where the tails go? If the bridal is connected to said carabiner and attached to interlaced bowlines, you would still need some kind of backup connection to the bridal (I've seen one tail clove hitched to the rail before). Otherwise if the biner blows you'll load the patient. Granted the likelihood of the biner becoming cross loaded in the first place is less, because it would theoretically be under tension the entire time, but that doesn't count for much to me. There are plenty of difficult scenarios I could imagine in which the litter could be temporarily unweighted and allowing the biner to 'flop. I will say I agree with you on it being a better method, on the agreement that every rigging style has it's place somewhere.


"That said, I do take issue with your defense of the cross-loaded 'biner. I think that it was ready to fail if the main had broken"

Fair enough. But why? Let's assume an aluminum NCRC carabiner was being used (SMC Locking D) with a cross loaded minimum breaking strength of 7kn, or 1,575 lbs. How can you assert to me a shock load greater than 1575 lbs would occur to the belay line if the main line were cut? The belay line is tensioned in the photo - not much of a fall. Looks like there is a good amount of rope out to help take the fall. And it is the weight of a single person. I will say in my experience that would be incredibly hard to do, considering most 2 person (400+ lb loads) belays with considerable slack in the system (such as the 1m fall on 3m of rope BCCTR tests) generate loads in the 7-10 kn range, and that is absolute worst case (very little rope out, huge slack, and lack of live load testing). I am happy to share some resources with you if you have interest in such tests.

I completely agree with you that connections (carabiners and rapides) generally kill more people than anything else. But you left out a very important part of that - and that is nearly all of those deaths resulted from MISUSE of the carabiner/mallion, and not the gear itself failing. Sure there are the cases of racks and eights leveraging off carabiner gates, but that's an incredibly specific loading requiring a big lever arm. Certainly no clear mechanism like that for some rope knots and webbing knots clipped into a biner.

"I certainly do teach people to build belay systems that can carry the full force of the load."

Again - what exactly is 'the full force of the load'? Until you give me a thoughtful understanding of the forces involved, such a statement won't mean much to me.

Good thoughts.

-Andrew

[NZcaver]

Good to see this has evolved into a thoughtful discussion. Now that I'm back above ground I can finally make time to address a few points.

Clem has brought up a number of valid concerns. However in practical use I still don't see an overwhelming need for an outright ban on carabiners for use as litter connections. Even for level 1 NCRC students. Andy, Marc and Andrew have each brought up some excellent points, and there's no need for me to repeat them. I do have a few extra points to make, but first I'd like to give a personal account of the situation shown in the photo.

Andy made a comment earlier about vigilance. He was correct in that I was closely observing the situation, as I simultaneously shot photos of it. As soon as I spotted the carabiner, I quickly assessed the entire situation known to me and determined - despite the disoriented carabiner - that the patient was in no immediate danger. Rather than stop the haul, I immediately alerted the rigging task force (who were a couple of body lengths above the patient in the photo) and had them keep tension (but not haul) on the belay line. As I recall, I instructed the person closest to the edge to correct the carabiner orientation as soon as possible. Once safely at the top of the drop, the patient (Jeff Burns) was briefly checked and then transferred to a horizontal guide line before being assisted along a rather slick and narrow muddy path back towards the entrance.

If people really feel strongly about the whole carabiner/litter issue (as Clem and Ron obviously do), I suggest spelling out your concerns as clearly and concisely as possible and making a submission directly to the NCRC. If you ask, I'd be happy to provide photos. The National Coordinator is Anmar Mirza, and the National Training Coordinator is Steve Hudson. Contact details can be found at http://www.ncrc.info Ultimately curriculum decisions come from a majority vote of the Board of Regional Coordinators, a couple of whom have already posted replies in this topic.

Now Clem, please clarify this for me. Are you recommending interlacing the main and belay lines with the litter bridle (3 interlaced bights) for vertical litter orientation, and with the spider (effectively 4 or 6 interlaced bights, depending on rigging configuration) for horizontal litter orientation? And when using a SKED, I assume this means interlacing with the bight by the head end for vertical orientation, and with the loops on the two horizontal lift straps for horizontal orientation? Even though the SKED manual shows that big locking steel carabiner as the key connection point. Or perhaps that's just Skedco dogma? Which reminds me...

I'll be happy to discuss this further, as I'm not one who merely stands behind a dogmatic position. Further, I'll do it here under my own name, and not a clever alias.

With your lengthy contributions to this discussion, I can now see that you don't merely stand behind a dogmatic position. This is good! OK then, my "absolutist dogma" terminology may have been a little misdirected. But given your initial statement about carabiners (outside of this forum), I don't think my comments were an entirely inaccurate assessment. Or... perhaps it is you who is really the pragmatist, and the carabiner-devotees who have the dogma problem??

As for using "your own name and not a clever alias" - I assume that veiled insinuation was directed at me, but I'm not sure why. Many of our forum members have a user name which is different to their legal name, including people like Andy and Marc. It's OK - it doesn't mean we have evil alter-egos or anything. If you look under my avatar, you'll see my name, NSS number, location etc clearly listed. It's hidden about as well as the sideways carabiner in that photo. I realize you're not on the forum too often, but most of the regulars around here know who I am. Plus Ron made sure he spelled out lots of identifying information about me in his initial post, which seems a little redundant and frankly rather childish. Perhaps that was what you meant?

Also, regarding copyright and use of photos (previously mentioned by Ron and Andy), another forum member has started a different topic about this. When time permits, I will reply to that topic rather than confusing this one. And now back to our regular topic.

According to at least one manufacturer's usage literature, a carabiner in a cross-loaded orientation is not safe. http://www.petzl.com/files/all/technical-notice/Pro/M73SL_VULCAN_SL_M735100B.pdf The little "skull and bones" icon means that life is in danger.

This is Petzl stating that this particular carabiner is not safe in a cross-loaded configuration. Whereas this is Petzl stating another model of aluminum carabiner is usable (to 15kN) when loaded directly onto the gate. Yes, I realize the Omni is a whole different design than what we generally use with litters. My point is while most of us realize a carabiner loaded on the gate is a bad/undesirable thing, not all manufacturers declare all their carabiners as unacceptable failure risks in that configuration. Many state the major and minor axis MBS specs, and leave the application decisions to the user.

A recurring theme here is the old adage - "is it safe, does it work?" I would argue that although the system works (both in general and specifically the case shown in the photo), technically it was and NEVER WILL BE "safe." But guess what? Tying in directly without the use of carabiners is not safe either. Getting on rope is not safe. Even with a belay. Neither is going caving, or driving around town, or walking down the street. Safe is a relative term, not a yes/no option. It's about acceptable and unacceptable risks, real and imagined, seen and unseen. If "safe" in this context means an acceptable degree of risk (and it generally does), then can carabiner litter connections be considered safe? I think they can, but obviously that's a matter for debate.

Good stuff. Carry on.

[shibumi]

Anmar Mirza here, speaking on behalf of the NCRC (just so I'm not hiding behind a clever alias ;)

While Clem and Ron bring up some valid considerations, Marc, Andy, Andrew, and Jansen have all addressed them pretty well IMO.

I am going to address a few things:

First off, NCRC has deviated away from some of the "accepted standards" for some very good reasons. The first and foremost is that we are teaching CAVE rescue. Yes, I know many of the techniques are similar across disciplines, but what we have decided to focus on is the techniques that we as an organization, not just as one or two individuals, feel are the most suitable for cave rescue. To this end the reasoning may be subtle, and it may still lack any real data, but it's a conglomeration of what some of the most highly experienced folks in cave rescue, many of whom train cross-discipline, have come to accept. Even then we are not in total agreement at all times. We may from time to time accept a slightly higher risk in one area to address a more pressing concern in another area. Is it perfect? Of course not. Will we change over time as better data and better understanding comes about? Sure.

NCRC teaches several different acceptable methods for litter connections. While we do not forbid it, we've come to recommend against interlocking knot connections at the head of the litter. One of the tenets we try to espouse is to not sacrifice efficiency for unnecessary redundancy. Far too many times we've had students get the litter hung up at the most dangerous part of the whole operation, the edge, where simply being able to quickly and easily disconnect one line or another would solve the problem. Disconnecting it then immediately reconnecting it is both faster than untying a knot, and safer for the patient in the long run, and allows for the reconnection that whipping out the knife and cutting the rope does not. We've evaluated the relative risks (and the key word here is relative) and this is the conclusion that we as an organization have come to.

Yeah, NCRC is involved in some controversial stuff with regard to techniques we are teaching. There's a degree of comfort in just going along with what the crowd is doing. What we are trying to avoid is both having to adapt less than optimal techniques to the specific enviroments we deal with, and simply being different for the sake of being different. Just because we do or don't do one specific thing does not mean that that thing has not been carefully considered by both the education committee and the entire Board of Regional Coordinators. We are also trying to avoid the trap of doing things just because that's the way we've been doing them for years. As much as possible we are trying to actually gather data and use that data to evaluate our curriculum, but that cannot always address the myriad of considerations that must go into rigging a system.

[Clem Akins]

Thanks, Shibumi (Anmar). I'm happy to concede that wiser heads than mine have considered the relative risks of the various litter rigging options, especially at the head of the litter, and chosen a path that makes more sense than what I know, especially for multi-pitch cave-specific rigging.

Maybe this a point where we have to "agree to disagree" but I'm still not convinced that a cross-loaded carabiner should be considered safe. Neither should a carabiner be considered safe if it's connected in a "floppy" manner, such as on the tail of a litter line that's not under tension. Jansen, the Petzl Omni is a specialized item for harnesses only, and the way I read the literature it's not rated for gate-loading anyway. The diagram shows it as being safe for 3-way loading, which is a whole lot more than conventional carabiners allow, but the only time a gate is pictured under load it's with the little skull and bones.

I might be able to understand a position like "Though we know that a cross-loaded 'biner is unsafe, we choose to take the risk of cross-loading because we feel other risks outweigh that one." I wouldn't like it, and I'd lobby hard for keeping on hand better ways of dealing with the situation, but I might understand it.

When I'm using carabiners for connections where the orientation isn't certain (connecting rappel device to harness, belay device to belay rope, etc.) I am super-vigilant about cross-loading. If I see a 'biner cross-loaded on a student or colleague's equipment I warn him of it immediately, and lecture about the dangers of cross-loading. I *never* say "No worries, you're only exerting a kN or two, and the gate is solid for at least seven!" You're scaring me by asserting that cross-loaded 'biners are safe--I'm just not buying it.

The math is harder to quantify, but I'll give it a shot. As I understand it, when Larson developed the BCCTR he did it as a way to envision the worst-case scenario. That kind of thinking makes sense to me. I'll agree that the scenario in the photograph would probably not break the cross-loaded 'biner (going along with your math) but not that it should be considered safe. Even Jansen saw it as a problem to be corrected. Isn't that, by definition, unsafe? Back to the BCCTR, isn't a force of 7-10kN on a cross-loaded 'biner unsafe? It might not break under that force, but I just can't call it safe. I like a 10:1 margin on my rigging between the actual force and the manufacturer's minimum breaking strength. Below that ratio, I get really nervous, and as it approaches 1:1 I've already changed to something stronger or safer. Using rigging equipment with a 1:1 safety factor is no safety at all, in my book. Especially not when human lives are at stake, and better alternatives are at hand.

But wait--there's more! Cross-loading is far from the only thing that can happen to a carabiner if it's rigged in a "floppy" way. There is also gate loading, where the force comes onto the gate and is directed inward. This can easily happen with ropes, and more easily (and more dangerously) with lever arms such as rappel racks. I have watched in horror as a rope unlocked a carabiner and unclipped it, as it came under load. More often, I've seen the result of such an action but not the act itself, as it happens in a fraction of a second. (It was all kinds of funny when, as I watched from the bank, my canoe unclipped itself and headed downstream, for a lighthearted example.) It's not all about the manufacturer's rated MBS in a dynamic event.

Have we beaten this topic to death, yet? If not, perhaps I'll let someone else fight the good fight for a while. I love this kind of talk, but it's exhausting!

Clem

[NZcaver]

Jansen, the Petzl Omni is a specialized item for harnesses only, and the way I read the literature it's not rated for gate-loading anyway. The diagram shows it as being safe for 3-way loading, which is a whole lot more than conventional carabiners allow, but the only time a gate is pictured under load it's with the little skull and bones.

Yes, the Omni is designed for harness connections (I've been using one myself for about 5 years). Check out the little diagrams at the top of the first page in the PDF. It shows the Omni rated to 20kN for 3-way and end-to-end 2-way loading, and indicates 15kN for 2-way loading directly outward onto the reinforced gate (from the inside). Although in all fairness, I see the written instructions below now discourage users from loading directly onto the gate. You are correct in that a force applied inward on the gate would be very bad, and that's the example Petzl identifies with the skull and crossbones.

You were talking about the possibility of a rappel rack in a certain configuration leveraging inward force on a carabiner gate. Coincidentally we were recently discussing this in another topic on the forum. Personally I do what you do. I use a Maillon Rapide as the attachment between my rack (usually my BMS Micro) and my Petzl Omni harness link. If I switch back to my Stop at any time, I also switch my attachment back to a locking carabiner. Is using an MR with a rack more safe than a carabiner? Many people think so, but other experienced cavers are perfectly happy with carabiners there. I don't feel a need to correct them, because I don't feel the practice is unacceptably unsafe. Especially considering these people are generally super vigilant about how their gear is rigged and oriented before committing to it.

So why do I use an MR and not a carabiner with my rack? Yes, the gate loading factor is a consideration - but it's not the only reason. My 10mm Zicral oval MR is as light as a locking carabiner, and more compact. Therefore it's more convenient for me, especially since I rarely remove it from my harness. OK, time to move away from racks and back to the subject at hand. As Andrew said, most link failures (carabiners and screw links) are caused by MISUSE (or human error), not equipment failure. I can already hear you saying "all the more reason to tie in directly!" But as you know it's not always that simple, and many attachments would be downright inconvenient - and possibly unsafe - if we were to eliminate carabiners and screw links.

You're scaring me by asserting that cross-loaded 'biners are safe--I'm just not buying it.

I didn't mean they were "safe." I wouldn't buy that either! To the contrary, I prefer to be the pessimist who thinks "nothing is ever safe."

I might be able to understand a position like "Though we know that a cross-loaded 'biner is unsafe, we choose to take the risk of cross-loading because we feel other risks outweigh that one." I wouldn't like it, and I'd lobby hard for keeping on hand better ways of dealing with the situation, but I might understand it.

It seems we have common ground. To me, the "risk" hinges on 3 factors. 1 - How likely is it for the carabiner to accidentally orient itself in such as way that it would experience gate loading? 2 - If it does migrate into that position, how likely/unlikely is it that sufficient force will be applied resulting in carabiner failure? And 3 - What is the foreseeable result of a carabiner failure, and will it lead to total system failure? And central to this discussion, to what degree will we lose some efficiency in the overall function of our system by removing the carabiners from our litter? Does the KISS (Keep It Simple and Safe) mantra apply more to the use of carabiners, or to hard tie-ins?

Have we beaten this topic to death, yet? If not, perhaps I'll let someone else fight the good fight for a while. I love this kind of talk, but it's exhausting!

My friend, we have not yet begun to scratch the surface.
Welcome to the world of technical debate ad nauseum. More commonly known by its scientific symbol,

[shibumi]

My friend, we have not yet begun to scratch the surface.
Welcome to the world of technical debate ad nauseum. More commonly known by its scientific symbol,

Or, for those of us who have been around the net for a while: "It was a typical day on USENET: an angry mob beating a greasy spot on the pavement where used to lie the body of a dead horse."

Some people enjoy discussing technical details to the nth percentile. That's great here where we can take the time to do so. Just don't let me catch the students doing it deep in the cave while the patient just sits there not moving towards the entrance! On a somewhat humorous note: During the Sides Cave rescue Jim Johnson who I'd put in charge of evac came to the sudden realization that it had been 20 minutes since the patient had moved, and "Anmar was going to be back soon and was going to be pissed!" and that's when he got the litter team moving in the right direction

[NZcaver]

On a somewhat humorous note: During the Sides Cave rescue Jim Johnson who I'd put in charge of evac came to the sudden realization that it had been 20 minutes since the patient had moved, and "Anmar was going to be back soon and was going to be pissed!" and that's when he got the litter team moving in the right direction

Incidentally, the preparation for the haul shown in the carabiner photo took an inordinate amount of time due to delays at the bottom. Everything was ready in place at the top, but positioning the patient below seemed to take forever. It was like waiting for speleogenesis. I'm sure there was a good reason. It must have been extremely difficult to carefully balance that cross-loaded carabiner perfectly for the benefit of the photos.

But from what I saw overall that week, thankfully delays like that were the exception rather than the rule.

[Steve Hudson]

In case some of you haven't heard or seen any of the several new carabiner models now available;

Especially for those who are wanting a faster way to connect than a knot or screw link but are worried about possible dangerous short axis and gate loading issues, ANSI has the answer.
Although OSHA regulations have not yet been updated to reflect this change, the 2007 ANSI Z359.1 standard now requires a connector's gate to withstand 3600 pounds force in all
directions. Longitudinal strength requirement remains 5000 pounds. They have BIG gates!

Cheers,

Steve Hudson
Who is not promoting anything in particular or speaking for any organization, just pointing out what is now available to possibly solve a problem some may have.

[Scott McCrea]

In case some of you haven't heard or seen any of the several new carabiner models now available

Oh, yea. I forgot about that. This is the new Black Diamond Gridlock.

I'll let y'all decide if it is rescue worthy, but it would lessen the chance of cross-loaded biner.

[Bill Putnam]

Steve Hudson
Who is not promoting anything in particular or speaking for any organization, just pointing out what is now available to possibly solve a problem some may have.

Nice disclaimer. ;-)