I ran across an accident report posted by an English climber who's partner fell to his death while rappelling in Zion. It sounds as though he's sharing his account of the accident in an effort to insure that it is not duplicated elsewhere. While he does not know for certain what caused his partner's death, it appears as though the double figure-eight knot that they had been using to join the ropes failed. Apparently this knot tends to invert (roll over itself and move closer to the end of the rope)quite readily when loaded.
I don't use this knot myself for a number of reasons, but thought I'd share the account just in case there's anyone out there who does. It's rather long, but you can easily skip to the portions directly concerned with the accident if you wish to.
I've posted the account below if you're interested in reading further:
"On May 21st I was descending with my friend Ross from Spaceshot on the Leaning Wall. During the last abseil Ross fell to his death. Ross and I are from the UK and were on a trip visiting various crags in the US.
There is a lot of stuff spinning around in my head as I write this, but my main thought is to let people know what (it seems) was the cause of this accident. The main factor in this has surprised a good number of the climbers I have talked to. I know there has been some discussion of this on the web already. Hopefully by telling the whole story - however irrelevant some of it might be - all of the various questions might be answered. I will try to reply to any questions where I can tell you something vaguely useful.
%==== The long story [skip ahead for the facts] ====%% On Monday we climbed the first four pitches and returned to the ground, leaving ropes in place to jug the next day. All the anchors we used were fixed, except maybe for the one at the top of the first pitch. Pitch 1 is slightly grotty 5.6 climbing. Pitch 2 is a pretty nice 5.7 flake and ends at the left end of a large sandy ledge. We fixed a 60m rope ("the blue" 60mx10.5mm) to this anchor, having got beta saying this would just reach the ground. Pitch 3 is a mixed bag of sandy 5.5 and ends at the base of a huge smooth clean red wall, the stuff we came to do. We fixed "the green" (55mx10.5mm) to this anchor and chucked it back down to the sandy ledge (top of pitch 2). Pitch 4 is where it gets fun. I lead the pitch (C2 aid) and Ross followed, cleaning the gear. We fixed our 60m lead rope ("the yellow" 60mx10.5mm) to this anchor and abseiled down. Then down the green to the sandy ledge. Then down the blue (carefully checking it reached) back to the ground. It didn't quite reach the dirt, but left us with maybe 20ft of trivial down-shuffling to get back to our bags. We left the 3 ropes in place and headed off for a beer.
Tuesday morning we jugged the ropes. Amongst all the other crap you take aid climbing, we had a 9mm rope. We planned to lead on the yellow (the top fixed rope) and take the 9mm to deal with the double-rope abseils on the descent. We would chuck the green down to the big sandy ledge as we went past it, and then could retrieve the green and the blue by jugging just the blue on Wednesday and abseiling down.
I set off first, Ross followed. I got to the top of pitch 4 as Ross arrived at the top of pitch 3. Ross had got some two-way radios earlier on the trip and we chatted on the radio: the weather forecast had been slowly deteriorating for the last 3 days, today was 50% chance of afternoon rain, there were a lot of gloomy clouds brewing above us, the sandstone is all bad in the wet, we were not super fast aid climbers...there were a lot of reasons for continuing, mostly that I didn't want to have to lead that C2 pitch again!! A brief spot of rain actually hit us and we decided to bail. I pulled up the 9mm rope, tied it to the yellow, stripped the anchor and descended to the top of pitch 3. Meantime Ross had been untying the green from this anchor and getting ready to set up a double-rope abseil. I got down to him, chucked him the end of the yellow to tie to the green and started pulling the ropes down from above.
Ross headed off down to the big sandy ledge as I coiled the 9mm and put it on my back. He radioed me to say "rope free" and I headed down. I arrived on the big sandy ledge about 10-15ft away from the anchor - Ross was off to my left, already clipped into the anchor and sorting out the blue rope, ready to set up the last abseil. I chucked the loose end of the yellow to Ross and started pulling the ropes from above. I was unclipped at this point - being a very bad boy, even though it was a huge ledge. This was actually the only thing that struck me as unsafe about our whole day. As the knot came down, I stopped and untied it to free the yellow, which was now all tangled up in plants and rocks on the ledge. Ross fed it over the edge as I untangled it from everything on the ledge. I started pulling the green down as Ross sorted himself out over at the anchor. I was coiling the green rope as Ross called over to say "see you at the bottom in a few minutes", he saw me coiling the green and offered to carry it, since I had the 9mm already on my back, but he already had our daysack on so I said I was fine taking it down. I turned to just finish up coiling the green and at that moment he fell.
I rushed over and there was nothing there - our ropes had gone, Ross had gone, the anchor was fine, untouched. Everything floated for a moment, slipped sideways and turned unreal - then I started shouting...I knew I had to get down in case by some impossible chance there was something I could do to help him. I was yelling down to the road and got someone's attention, they flagged down one of the shuttle buses and shouted that help was coming. I had the 55m green and the 50mx9mm ropes with me. I couldn't get to the ground in one go but I knew there was another anchor (top of the Alpine Start for those that know it) that I would be able to reach. I set up the double rope abseil and set off down. The ropes tangled around everything - it was a complete shambles. I saw the rangers and the ambulance arrive; the rangers were racing up the hill to Ross. I set up the second abseil, it was all taking so long...as I reached the ground one of the rangers came over to tell me what I already knew.
%%==== Some stuff that I do know ====%% Ross was found with the two ropes correctly through his belay device.The ropes extended about 10feet "above" him (the other 190feet being "below" his belay device) and the ends were not tied together. Throughout this trip we had always been tying ropes together using a fig-eight knot (more below). The only other abseil Ross set up that same day (from top of pitch 3 down to the big ledge) he had used the fig-eight knot with no back up knot on the tails. The knot was neat, I don't remember exactly how long the tails were but they didn't cause me a second glance. I could not see exactly what Ross was setting up on that last abseil - he was 10ft or so to my left and was sitting (while clipped in) so that he obscured my view of the anchor.
The fig-eight I refer to is tied as follows: The two ends you want to join are held parallel with the ends "pointing" in the same direction. You grab both ropes together and then tie a regular single fig-eight knot in both ropes at once.
What we did NOT use: The only other way that might be confused is when you have the ends pointing in opposite directions. Tie a single fig-eight in one rope then follow this through with the other rope - we did NOT do this.
%%==== The important bit ====%% Some guys that were helping me out played around in their yard with this fig-eight method, tying it and trying to pull the knot apart. They found some worrying things.
-The way the ropes pull on this knot on a double-rope abseil deforms the knot badly.
-If the knot is not perfectly "dressed", in particular if there is a single slack loop anywhere on the fig-eight, they could pull the knot through even with 6 INCHES of tails, just pulling the ropes apart as happens naturally on an abseil. 6 inches of tails is NOT ENOUGH. If you use this knot, tie a back up knot and leave LONG tails. It scares me to think that I could have innocently/ignorantly made this same catastrophic mistake.
%%==== My thoughts (not facts) ====%% The only plausible explanation of this accident I have come up with is that the knot slipped off the ends. I won't go through all the alternate scenarios and my objections to them here. I hope it doesn't sound contradictory to say that Ross was a safe climber. I never saw him rig a belay that I thought was unsafe, never saw him do anything that made me think "does he realise that's pretty dodgy". We were not in a big rush getting down. We were moving quickly and efficiently but with no sense of panic or anything like that. Ross knew that the last abseil was a long one and we would be a bit tight on rope. I can imagine that would make him want to keep the knot pretty near the ends, but I do not believe he would only leave something ridiculous like one inch of tails. I think he must have tied the knot with something like 6inches of tails, thinking this was plenty (go tie the knot - it looks good with this much rope sticking out of it) and maybe he didn't make it all neat and snug. I think when he set off he was happy with his set-up, not thinking at all that the tails were dangerously short. The first 30feet of this abseil are a little slabby - and with two 60m ropes you do have to feed armfuls through your belay device at the top - the first few feet of such an abseil are always a bit jerky. I guess he fed through a couple of armfuls of rope and hence bounced the knot just a couple of times, which caused it to fail.
While I will never know for sure what happened, I do know what any of you can prove to yourselves - that you can get this knot to fail even with 6 inches of tails. I did not know that the necessary margin for safety was so wide for this knot, I am sure Ross did not realise this either.
The ropes involved (the blue and yellow) have been sent to one of the testing guys at Black Diamond who is going to run some relevant tests involving this fig-eight knot. I will post anything they find that might be of interest.
%%==== Last words ====%% Thoughts of Ross are vividly etched in the minds of almost everyone he met. We miss him terribly. The only other thing I want to say here is that the Rangers at Zion were incredible; the way they dealt with the incident, the diligence of their investigation and the compassion that they showed me...I have only praise for everything they did. I was overwhelmed by the generosity of so many other people in Springdale - it's a small town of wonderful people. Despite everything, I have some very fond memories of Zion and the people I met. It is a beautiful place - you should go there and climb those amazing walls.
I'm interested in this story since I use this knot regularly. I usually only use an overhand rather than a figure eight (overhand with an extra twist). I also use at lest 12 to 18 inches of tail. It seems to me that the single overhand creates a smaller knot that is less prone to turning inside out. Does anyone know if this is true? My reason is based on simple observation - tie both and pull them apart, the overhand snugs nicely while the fig eight twists around a bit. Fern, are you picturing a figure eight follow through (each end protruding from opposite sides of the knot) rather than the EDK variation with both ends out the same side? Otherwise, I'm not sure what you mean about the ropes not being tied entirely together.
I mean I think it is possible the ropes were not tied together. No knot at all, nothing, just two ropes completely separate. The other scenario (fig8 rolling over,eating up all the tail until the ropes part) that has been implicated is also possible. We will never know.
Kevin, the system you use sounds 'not totally dangerous' to me. Long tails, snug overhand knot. If you die on rappel one day it's not my fault though. There are pro-cons to every rappel system but it's always sketchy.
Wow. What a tragic incident. Thanks for sharing the report. It's hard to know what happened, but the knot inverting and untying itself could have been the problem.
I feel compelled to also comment on another variation of the figure 8 knot that I use (and many of my partners as well) to tie into the rope. If you pass the tail back through the knot and pull it all the way through the knot can intert if there is a lot of force placed on it in a fall. The knot comes untied if it inverts twice.
If you like this tie in method better than using a fishermans or other backup knot, the way to make sure it doesn't invert is to leave a few inches of rope out the top of the turn and to completely dress the knot. If the knot inverts it will catch itself on the tail and only invert once, not twice.
Ehmmic, I'm having a hard time putting your words into a mental picture. Are you referring to the "figure nine?" (at that's what I've heard it called) where you tie a regular 8 and then the tail passes back through the knot in such a way that it get's the tail out of the way and makes it easier to untie after being fallen upon?
Some additional commentary/testing info on the overhand figure-8 knot...
I learned this weekend at the International Technical Rescue Symposium that the figure eight version of the Euro Death Knot is being actively taught to climbers in Canada. I had thought that only the overhand version was in widespread use and that everyone pretty much recognized the figure eight to have the potential to be a really quick trip to the bottom of the crag.
Failure of the figure eight version of this knot has already caused a fatality in1994 at Seneca Rocks and an accident in 1995 in Salt Lake.
Both of these knots would politely be called "mis-loaded" and impolitely would be called, well - the "Euro Death Knot". The failure mode for the figure eight version is to flip/invert/capsize, which then becomes the identical twin of the first knot, just with shorter tails. After enough of these events there are no tails left and the knot fails. Leaving at least a foot of tail is recommended.
I spent a few hours yesterday in my front yard with a come-a-long and a load cell to try to get some decent information on this. The results still leave room for plenty of argument. People who don't like the figure eight will say, "See, it slips at really low loads!" People who do like it will say, "See, if you dress it right, pretension it well, and leave long tails, it doesn't fail. Besides, I've been using it for years and I'm still alive!"
If you're too impatient to wade through the results below, the short answer is that if you do all those things, you should be ok. My question is - why would you take the chance? If you're in a situation where a stuck rope would be catastrophic, use the overhand. It has all the same advantages and not nearly as much risk.
Be safe - the body we have to scape off the rock may be yours.
- Tom Moyer Salt Lake County Search and Rescue
Rope A: Mammut 11 mm static - used Rope B: Unknown manufacturer red 11 mm dynamic - used Rope C: ABC/Sterling 11 mm static - new Rope D: Blue Water II+ 11 mm static - new Rope E: ABC 8mm static - new Rope F: 1" Tubular Webbing
Test #1: RopeA/RopeA - figure 8 - well dressed and pretensioned Capsized at 750 lb, Rope broke at 2520 lb
Test #2: RopeB/RopeB - figure 8 - well dressed and pretensioned Capsized at 590 lb, Capsized at 2280 lb, Rope broke at 2560 lb
Test #3: RopeB/RopeB - figure 8 - well dressed, pretensioned loosely Capsized at 290 lb, Stopped Test at 2800 lb
Test #4: RopeB/RopeB - figure 8 - sloppy, crossing strands and loose Capsized at 110 lb, Capsized at 140 lb, Capsized at 340 lb, Capsized at 420 lb, Capsized at 530 lb, Stopped Test at 2500 lb
Test #5: RopeB/RopeB - overhand - well dressed and pretensioned Capsized at 1400 lb, Capsized at 1940 lb, Capsized at 1990 lb, Rope Broke at 2070 lb
Test #6: RopeA/RopeA - overhand - well dressed and pretensioned Stopped Test at 2540 lb
Test #7: RopeC/RopeC - figure 8 - well dressed and pretensioned Stopped Test at 2500 lb
Test #8: RopeD/RopeD - figure 8 - well dressed and pretensioned Capsized at 2170 lb, Stopped Test at 2550 lb
Test #9: RopeB(11mm)/RopeE(8mm) - figure 8 - well dressed and pretensioned Capsized at 1330 lb, Capsized at 1550 lb, 8mm broke at 2700 lb
Test #10: RopeB/RopeB - figure 8 - well dressed and pretensioned - WET Capsized at 470 lb, Rope broke at 2790 lb
Test #11: RopeB/RopeB - figure 8 - well dressed, pretensioned loosely - WET Capsized at 290 lb, Rope broke at 2470 lb
Test 12: RopeF/RopeF (webbing) - well dressed and pretensioned Webbing broke at 2070 lb
quote:Originally posted by Off White: Is there some reason why you wouldn't use a double fishermans knot for tieing your ropes together for a rappel? I don't think I've ever used anything else.
If I'm rapping down a blank face and the odds of the rope getting caught in a crack are low, I go with the double-fisherman's. If it's windy, and/or there's a reasonable chance that the rope will get hung up on the way down I go with the EDK (backed up with an overhand knot immediately behind the EDK itself and a long tail).
Loc: trying to climb into my navel
Thank you very much JayB.
I looked at the capsize issue a long time ago. I strongly feel that a figure eight loaded cross ways is an inappropriate use of the knot. This is particularly true of larger diameter ropes; as your tests seeem to indicate. Small diameter ropes will often hold better on this knot. I also think that kern-mantel ropes are the most suseptible to capsize>
The only advantage I see in this knot is that it is fast to tie. Useing ehmmics technique may make the knot capsize proof but if tied wrong ( the loose tail is in the wrong place or there is no loop) i think that it fails upon the first capsize.
It seems that if you re-weave a figure eight so that each tail comes out of the opposite end; the knot is fail proof. Is this not the proper way to tie (and load) a figure-8 for connecting 2 ropes?
The person you feel you are inside is inconsequential. It's what you do, that counts.
Uncle Tricky - the knot may be called a figure 9. I'm not sure. It is a standard figure 8 follow through knot with the tail passed back through the knot loop on the bottom as a backup knot. I think we're talking about the same thing based on your description.
On fisherman's knots for rappeling, if you tie a square knot before tying the fisherman's you will always be able to untie the knot.
The figure-9 is a figure 8 on a bight with an extra half-twist, making it easier to untie after loading (see "On-Rope" for an example). I've never heard of it tied as a follow-through knot. I use it fairly regularly for clipping in at the harness when I know I will be hanging out.
As for the EDK, I have also heard the warnings about the fig-8 var. from several sources. The overhand w/ good tails is the way to go as far as I know.
If you do use a double-fisherman, just throw in a double sheet bend or square between the barrels to untie easily. Even under a rescue load, this knot unties easily.
JayB, thanks for the objective data and analysis. That is a good response to an accident rather than the usual second-guessing from monday-morning qb's.
Just a note concerning the data/commentary in my post above. Except for the fist sentence, it's all from Tom Moyer, of Salt Lake City Search and rescue, in a post of his that I found during a search of rec.climbing. Anyhow - hopefully it's useful information.
Loc: trying to climb into my navel
JayBee...dude; I thought you did all that testing. I was really, very impressed and envious of your motivation. I guess I have to read more carefully. Oh well, great info, thanks for sharing...you rock any way.
The person you feel you are inside is inconsequential. It's what you do, that counts.